UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Asymmetric synthesis from solid state photochemical asymmetric induction to catalytic asymmetric hydrogenation Liu, Zhaoqing 1995

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1995-983132.pdf [ 10.68MB ]
Metadata
JSON: 831-1.0061500.json
JSON-LD: 831-1.0061500-ld.json
RDF/XML (Pretty): 831-1.0061500-rdf.xml
RDF/JSON: 831-1.0061500-rdf.json
Turtle: 831-1.0061500-turtle.txt
N-Triples: 831-1.0061500-rdf-ntriples.txt
Original Record: 831-1.0061500-source.json
Full Text
831-1.0061500-fulltext.txt
Citation
831-1.0061500.ris

Full Text

ASYMMETRIC SYNTHESIS From Solid State Photochemical Asymmetric Induction to Catalytic Asymmetric Hydrogenation by Zhaoqing Liu B.Sc, Peking University, Beijing, China, 1985 M.Sc, Peking University, Beijing, China, 1988  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (DEPARTMENT OF CHEMISTRY) We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA APRIL 1995 0  Zhaoqing Liu, 1995  In presenting this thesis in partial fulfilment  of  the  requirements  for an advanced  degree at the University of British Columbia, I agree that the Library shall make  it  freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of department  or  by  his  or  her  representatives.  It  is  understood  that  copying  my or  publication of this thesis for financial gain shall not be allowed without my written permission.  (Signature)  Department of  ^ ^ ^  v  St^y  The University of British Columbia Vancouver, Canada  Date ^ *  DE-6 (2/88)  a  ^~.  °\S  ABSTRACT  A number of phosphine oxides were synthesized via different routes.  Grignard  displacement of menthyl phosphinates afforded different enantiopure monophosphine oxides. By coupling the monophosphine oxides, bisphosphine dioxides I were prepared. l,2-Ethynediylbis(diphenylphosphine oxide) was reacted with anthracene and its derivatives to give Diels-Alder adducts ethenoanthracenes, and two of them were further reduced to ethanoanthracenes by magnesium in methanol.  A new enantiopure  bisphosphine ligand, anthraphos (II), was prepared. Dibenzobarrelene derivatives HI form well defined crystalline complexes with triphenylphosphine oxide as well as with bisphosphine dioxides I. Irradiation of the complexes in the solid state led, via the di-7i-methane rearrangement, to pairs of regioisomeric dibenzosemibullvalene photoproducts. The ratio in which these products are formed was found to vary markedly with the structure of the phosphine oxide. Low to moderate photoproduct enantioselectivities were observed in the complexes with enantiopure diphosphine dioxides. Five of the complexes studied had their crystal and molecular structures determined by X-ray crystallography.  Correlations between the  photochemical behavior of the complexes in the solid state and their crystal structures are discussed. The new host IV forms crystalline inclusion complexes with a wide variety of aliphatic solvent molecules. The molecular and crystal structures of these complexes were analyzed by X-ray crystallography. The crystal structures show that the complexes with ethanol, 1-propanol and 2-propanol have the chiral space group 7*2,2,2], where the absolute configuration of IV was assigned. The inclusion complexes can undergo the di7i-methane rearrangement to result in dibenzosemibullvalene derivative product V in  ii  solution and the solid state. In the case of the complexes with ethanol, 1-propanol and 2-propanol, the solid state photoreaction gave product V in more than 90% e.e.; this reaction was used for the asymmetric synthesis of bisphosphine ligand VI. The absolute configuration of dibenzosemibullvalene V was also determined.  By correlating the  absolute configuration of reactant IV with that of its photoproduct V, the key structural features responsible for the enantiospecific solid state photorearrangement were elucidated. Other ethenoanthracene phosphine oxide derivatives were also examined for inclusion complex formation as well as for solid state photochemical asymmetric induction. 9,10-Dihydro-9,10-ethenoanthracene-l l-(diphenylphosphine  oxide)-12-carboxylic  acid and its ethyl ester were photolysed in different media including the solid state, and two regioisomers were obtained in each case. The medium-dependent regioselectivity was considered to result from the conformations in which the phosphinoyl group could be free or hydrogen bonded. Two different mechanisms were applied to rationalize the regioselectivity. The reactivity was correlated with the crystal structures of the starting material. An active asymmetric catalyst, [Rh(COD)(/?,/?-anthraphos)]BF4, was prepared from enantiopure anthraphos (II). The molecular and crystal structure of this catalyst was analyzed by X-ray crystallography and compared with the known structures of chiraphos and norphos.  Use of  [Rh(COD)(/?,/?-anthraphos)]BF4 to hydrogenate (Z)-cc-  acetamidocinnamic acid gave (5)-(+)-N-acetylphenylalanine in 90% e.e. Bisphosphine VI was also used for the same purpose, but gave lower enantioselectivity.  Different  ruthenium-anthraphos complexes were also synthesized by ligand exchange reactions in an attempt to use them as asymmetric catalysts in the hydrogenation of C=C, C = 0 and C=N bonds.  in  H  PPh2  PQ2H  €>f<««+0 Ar  Ar =jp-C6H4(CH3X /w-QH^O^) 3M4P3(P%)2,  3,5.(^3(013)2  II  R = —CH2CH3 —01(013)2 III Ph,P  IV  PPh2  TABLE OF CONTENTS  Abstract  ii  Table of Contents  v  List of Tables  x  List of Figures  xiii  Acknowledgment  xx  Dedication  xxi  INTRODUCTION I.  Asymmetric Synthesis  2  H.  The Topochemical Postulate  4  DX The Di-7t-Methane Rearrangement  9  (i) Reaction Mechanism  9  (ii) Reaction Multiplicity  13  (iii) Regioselectivity of the Di-rc-methane Rearrangement  14  (iv) Photochemistry of Dibenzobarrelenes and Derivatives  15  TV. Asymmetric Photoreactions in the Solid State  V.  18  (i) Absolute Asymmetric Photochemical Synthesis  19  (ii) Photochemical Asymmetric Induction in Eanatiopure Hosts  21  (iii) Photochemical Asymmetric Induction in Chiral Salts  23  Catalytic Asymmetric Hydrogenation  25  (i) Mechanism  25  (ii) Chiral Ligands  26  (iii) Chiral Catalysts  28  (iv) Substrate  29  v  VI. Organic Clathrate Chemistry  30  VII. Research Objective and Outline of the Thesis  32  RESULTS AND DISCUSSION Chapter One General Organic Synthesis  37  1.1  Synthesis of Optically Pure Phosphine Oxides with P-Chiral Center  37  1.2  Preparation of Substituted Anthracenes and Acetylene Derivatives  42  1.3  Synthesis of Dibenzobarrelene Derivatives  45  1.4  Magnesium/Methanol Reduction Reaction  51  Chapter Two Analysis of Optical Purity 2.1  NMR Chiral Shift Reagents  56 56  2.1.1 Optical Purity Determination of the Dibenzosemibullvalene Derivatives  57  2.1.2 Dibenzoyltartaric Acid as a Chiral Complexing Shift Reagent for Phosphine Oxides 2.2  59  Chiral HPLC Analysis of Phosphine Oxides  65  2.2.1 Optical Purity Determination on the CHIRAPAK OP(+) Column  67  2.2.2 Optical Purity Determination on the CHIRALCEL OD Column 68 2.3  Measurement of the Optical Purities of the Crude Photolysed Sample  Chapter Three Cocrystalline Complexes and Their Photochemistry.... 3.1  3.2  69 75  Cocrystalline Complexes of Phosphine Oxides  75  3.1.1 FTIR Spectra  78  3.1.2 Mass Spectra  79  3.1.3 X-ray Crystal Structure Analysis  81  3.1.4 Clathratocomplexes or Coordinatoclathrates  83  Solid State Photochemistry of the Complexes  91  vi  3.3 Photochemical Asymmetric Induction of the Complexes  93  3.4 Stereoselectivity of the Photoreactions of Dibenzobarrelenes 105 and 106 in the Solid State  95  3.4.1 Electronically Controlled Regioselectivity  98  3.4.2 Absolute Configuration of Dibenzobarrelene 105 and its Enantioselectivity  106  Chapter Four Inclusion Complexes and Their Photochemistry  108  4.1 Formation of Inclusion Complexes  108  4.2 X-ray Structure Analysis of Inclusion Complexes  115  4.3 Photochemistry of Host Ethenoanthracene 126  124  4.3.1 Characterization of Photoproduct 143  125  4.3.2 Asymmetric Induction  127  4.3.3 Absolute Asymmetric Photorearrangement of Host 126  128  4.3.4 Asymmetric Synthesis of Compound 143  133  4.4 Photochemistry of Inclusion Host 130  136  4.4.1 Characterization of Photoproduct 149  138  4.4.2 Photorearrangement Mechanism  141  4.4.3 Asymmetric Induction of Inclusion Complexes 130/wo-Propanol 143 Chapter Five Photochemistry of Phosphine Oxides 132 and 133  147  5.1 Photolysis of Ethenoanthracenes 132 and 133  149  5.2 Medium-Dependent Regioselectivity  152  5.3 Hydrogen bonding and Regioselectivity  154  5.4 Solid State Conformation and Regioselectivity  159  Chapter Six Asymmetric Catalytic Hydrogenation  162  6.1 Preparation of Possible Catalyst Precursors from Anthraphos  162  6.2 Asymmetric Hydrogenation of Z-(a)-Acetamidocinnamic Acid  166  6.3 X-ray Structure Analysis of (-)-[Rh(COD)(anthraphos)]BF4  170  Vll  EXPERIMENTAL I.  General Procedures  175  n.  Synthesis of Starting Materials  179  HI. Synthesis of Complexes  220  (i) Complexes of Triphenylphosphine Oxide (TPPO) with Dibenzobarrelene Derivatives  220  (ii) Complexes of l,2-Ethanediylbis(diphenylphosphine oxide) (DPPEO, 101) with Dibenzobarrelene Derivatives  222  (iii) Crystalline Complexes of the Optically Active Bisphosphine Dioxides with Dibenzobarrelene Derivatives  224  (iv) Inclusion Compounds of 9,lO-Dihydro-9,10-ethenoanthracene-l 1,12bis(diphenylphosphine oxide) (126); 9,lO-Dihydro-9,10ethenoanthracene-l ,5-dimethanol-l 1,12-bis(diphenylphosphine oxide) (129) and 9,10-Dihydro-10-hydroxy-9,10-ethenoanthracene-l 1,12bis(diphenylphosphine oxide) (130) (v) Inorganic Complexes  227 235  IV. Photochemical Studies  239  (i) Solid State Photolyses of Dibenzobarrelene Complexes with Phosphine Oxides  239  (ii) Photolysis of 9, lO-Dihydro-9,10-ethenoanthracene-l 1,12-bis(diphenylphosphine oxide) (126)  242  (iii) Photolysis of Compound 9,10-Dihydro-10-hydroxy-9,10-ethenoanthracene-l 1 , 12-bis(diphenylphosphine oxide) (130)  246  (iv) Photolysis of 9,lO-Dihydro-9,10-ethenoanthracene-ll-(diphenylphosphine oxide)-12-carboxylic Acid (133)  viii  248  (v) Photolysis of Ethyl 9,10-Dihydro-9,10-ethenoanthracene-l 1(diphenylphosphine oxide)- 12-carboxylate (132)  250  (vi) Quantum Yield of Photoreaction of Inclusion Complex  V.  126/ethanol (1:1)  252  Catalytic Hydrogenation  254  (i) Catalytic Hydrogenation of Traws-a-acetamidocinnamic Acid ...  254  (ii) Workup and Characterization of Hydrogenation Product  255  APPENDIX I  257  REFERENCES  266  IX  LIST OF TABLES  Table 1.1  Caption Chemical and Optical Yields of Aryl Grignard Reaction with Phosphinate 107  1.2  Page  39  Isolated Yields of Diels-Alder Adducts between Anthracene or Its Derivatives and l,2-Ethynediylbis(diphenylphosphine oxide)  47  1.3  NMR Spectroscopic Data of Ethanoanthracenes 134 and 135  54  2.1  Resolution of 31p NMR Signals (81 MHz) of Racemic Phosphine Oxides in the Presence of D-DBT  2.2  62  Chromatographic Data for Phosphine Oxides on the CHIRALPAK OP (+) Column (25 cm x 0.46 cm I.D.) at Room Temperature  67  2.3  Chromatographic Results with the CHIRALCEL OD Column  69  3.1  Properties of Triphenylphosphine Oxide (TPPO) Complexes  77  3.2  Properties of Bisphosphine Dioxide Complexes  78  3.3  Stretching Bands v(P=0) (cm~l) of Phosphine Oxides only and Those of the Complexes with Dibenzobarrelenes 105 and 106 as well as /7-Nitrophenol  79  3.4  Molecular Ions of FAB Mass Spectra of Phosphine Oxide Complexes  80  3.5  Oxygen-Oxygen Distances of P = 0 - H - 0 in the Crystalline Complexes 83  3.6  Crystal Data for the Complexes of Dibenzobarrelene 105 with Different Bisphosphine Dioxides  3.7  88  Photochemical Regioselectivities of Dibenzobarrelene 105 in its Crystalline Complexes with Phosphine Oxides  92  3.8  Regioselectivities of Dibenzobarrelene 106 in Crystalline Complexes with Phosphine Oxides  3.9  93  Optical Yields in Asymmetric Induction of Dibenzobarrelenes 105 and 106 in the Solid State of the Complexes with Chiral Bisphosphine Dioxides  3.10  94  0=C-C=C-C=0 Torsion Angles of Dibenzobarrelene 105 in the Complexes with Triphenylphosphine Oxide  3.11  Different Factors Influencing the Photochemical Regioselectivities of Dibenzobarrelene 105 in Triphenylphosphine Oxide Complexes  3.12  100  101  Analysis of Electronically Controlled Regioselectivities in the Photorearrangement of Complexes of Dibenzobarrelene 105 with Bisphosphine Dioxides  102  3.13  Torsion Angle-Enantioselectivity Correlation  107  4.1  Inclusion Compounds of Ethenoanthracene 126  110  4.2  Inclusion Complexes with Host Ethenoanthracenes 129 and 130  114  4.3  Partial Crystal Structure Data for the Inclusion Complexes  115  4.4  Intramolecular Contacts and Intermolecular Distances in Ethenoanthracene Compounds and Inclusion Complexes  4.5  Optical Yields of Photoproduct 143 from Single Crystal Photolyses (Pyrex) of 126 in its Inclusion Complexes with Different Guests  4.6  127  Absolute Configuration Coordination in the Photorearrangement of Inclusion Complexes of Ethenoanthracene 126  4.7  120  129  Optical Yields of Photoproduct 143 from Polycrystal Photolysis of 126/EtOH (1:1) Prepared by Different Seeding Methods  134  4.8  Photochemical Asymmetric Induction in 130/i-PrOH in the Solid State 144  5.1  Chemical Shifts (ppm) and Coupling Constants (multiplicity, Hz) from *H and 31 PNMR Spectroscopy  151 xi  5.2  Regioselectivities of Ethenoanthracene 132  153  5.3  Regioselectivities of Ethenoanthracene 133  153  5.4  Torsion Angles of 0=C-C=C and 0=P-C=C of 132 and 133  160  6.1  Asymmetric Hydrogenation of 155 in Methanol Catalyzed by [Rh(COD)((ll/U2/?)-Anthraphos)]BF4  6.2  168  Asymmetric Hydrogenation of 155 by Using an in situ Catalyst from Bisphosphine 4bR,8bR,8cR,8d/M44 and [Rh(COD)Cl2]2  169  6.3  Comparison of Structure of Diphosphine Complexes  171  Al  Incremental Shifts of the Aromatic Carbon Atoms of Mono-substituted Benzenes  257  A2  Calculated Chemical Shift for Phosphine Oxides  258  A3  Assignment of 13C NMR Spectrum of Monophosphine Oxides  259  A4  Observed 13C NMR Spectra and Their Assignment of Bisphosphine Dioxides  A5  260  Transition Energies and Intensities for the ABX System which Originate from X  A6  A7  261  l^C NMR Transition Energies and Intensities for PCH2CH2P, where A = B = 31p, x = 13C  262  Coupling Constants (Hz) of Bis(phosphine oxide)s 115 and 117  264  xii  LIST OF FIGURES Figure  Caption  Page  1  The Photochemical Reactions of trans-Cinnamic Acid Derivatives ...  2  The Reaction Cavity Concept  6  3  Stereoselective Norrish Type II Reactions of Macrocyclic Diketone 5  8  4  The Postulated Di-u-methane Rearrangement Mechanism  9  5  The Denitrogenation Reaction of Azoalkane 13  10  6  The Photolysis of Deuterium Labeled Barrelene 18  11  7  The 1,2-Aryl Shift Mechanism  12  8  Multiplicity Dependence of the Di-7c-methane Rearrangement of 28  14  9  Derealization Control of Regioselectivity of 39  15  10  Photorearrangement of Dimethyl Dibenzobarrelene Dicarboxylate (40)  16  11  The [2 + 2] Photocycloaddition of Dibenzobarrelenes  17  12  The Tri-7r-methane Rearrangement of 49  17  13  Solid State Asymmetric Bromination  19  14  Asymmetric Single-to-single Crystal [2 + 2] Photoaddition  20  15  Unimolecular Absolute Asymmetric Photoreactions in Chiral Crystals  21  16  Asymmetric Reaction in Inclusion Complexes  22  17  Typical Resolved Host Molecules  22  18  Photoasymmetric Induction of Salts via the Di-7C-methane rearrangement 24  19  Asymmetric Induction of Salt via the Norrish Type II Reaction  24  20  Asymmetric catalytic Hydrogenation of MAC  25  21  Mechanism of Asymmetric Catalytic Hydrogenation of Methyl Z-(cc)Acetamidocinnamate  5  26  xiii  22  B isphosphine Ligands for Asymmetric Catalyses  27  23  Ruthenium Based Chiral Catalysts  29  24  Typical Inclusion Host Molecules  31  25  Donors and Acceptors of Cocrystalline Complexes  33  26  Phosphine Oxide Derivatives of Dibenzobarrelene  34  27  Phosphine Oxide Derivatives of Dibenzobarrelenes  34  28  Bisphosphine Ligands  35  1.1  Preparation of Monophosphine Oxides  38  1.2  Oxidative Coupling of Monophosphine Oxides  41  1.3  Preparation of Substituted Anthracenes  43  1.4  Preparation of Ethynediylbis(diphenylphosphine oxide) (125)  44  1.5  Preparation of Ethyl 3-(Diphenylphosphine oxide)-2-propynoate  44  1.6  The Dibenzobarrelene Skeleton  45  1.7  Preparation of Dibenzobarrelene Ester-Acids  45  1.8  Dienophile of 1,2-Ethynediylbis(diphenylphsophine oxide)  46  1.9  Diels-Alder Reactions between Anthracene or Its Derivatives and 1,2Ethynediylbis(diphenylphosphine oxide)  47  1.10  Equilibrium between 9-Anthrone and 9-Anthrol  48  1.11  Preparation of 9-Hydroxyethenoanthracene 130  49  1.12  Preparation and Hydrolysis of Diels-Alder Adduct 132  50  1.13  Reduction of Double Bond Conjugated to Phosphinoyl Group  51  1.14  Mechanism for the Reduction of a Phosphinoyl "Conjugated" Double Bond  53  1.15  Preparation of Anthraphos  55  2.1  1H NMR Spectra of the Methyl Ester Groups of a Mixture of  2.2  Enantioenriched Regioisomers 138 and 139 in CDCI3  57  Chiral Shift Reagents and Substrates  58  xiv  2.3  31  P NMR Spectra (81 MHz) of Racemic Phosphine Oxide 149  inCDCl 3 2.4  31  60  P NMR Spectra (81 MHz) of Racemic Phosphine Oxide 134 in 3:1  CDC^/o-dichlorobenzene 2.5  31  61  P NMR Spectrum (81 MHz) of Racemic Phosphine Oxide 129  inCDCl 3  62  2.6  31  63  2.7  Resolution (AS = 8. - 8+) of 3 ^P NMR Spectra of Phosphine Oxide  P NMR Spectra (81 MHz) of Phosphine Oxide 143 in CDC13  143 as a Function of Equivalents of D-DBT  65  2.8  Packing Composition of the CHIRALCEL OD Column  66  2.9  Packing Composition of the CHIRALPAK OP(+) Column  66  2.10  Chromatogram of a Photolysed Sample of Ethenoanthracene 126 on a Chiracel OP(+) Column (25 cm x 0.46 cm I.D.)  2.11  31p NMR Spectra of a Sample from the Photolysis of Ethenoanthracene 126  2.12  70  71  31p NMR Spectra of a Photolysed Sample from the Enantiomorphously Pure Complex of 130/wo-propanol (1:1)  73  3.1  Phosphine Oxides and //-Donors  76  3.2  EI Fragmentation of Complex 105d (2:1 105/DPPEO)  80  3.3  Stereodiagram of Complex lOln (1:2 DPPEO/p-nitrophenol)  81  3.4  ORTEP Drawing of Complex R,R-117n (1:2 /?,/?-117//;-nitrophenol)  82  3.5  Packing Diagram of Complex 105t (1:1 TPPO/105)  84  3.6  Packing Diagram of Complex 105tt (2:2:1 TPPO/105/toluene)  85  3.7  Packing Diagram of Complex lOln  86  3.8  Packing Diagram of Complex 117n  87  3.9  Packing Diagram of Complex 5,5-105p (2:1 105/5,5-114)  89  3.10  Packing Diagram of Complex 5,5-105m (2:1 105/5,5-115)  90  xv  3.11  Di-7t-Methane Rearrangement of Dibenzobarrelene Derivatives 105 and 106 and the Diazomethane Workup  3.12  91  Mechanism of Di-u-Methane Rearrangement of Dibenzobarrelenes 105 and 106  96  3.13  Predisposition of Vinyl Substituents and Absolute Reaction Pathways  98  3.14  ORTEP Diagram of Complex 105t (1:1 TPPO/105)  99  3.15  ORTEP Diagram of Complex 105tt (2:2:1 TPPO/105/toluene  100  3.16  ORTEP Diagram of Molecules in Complex 105p (2:1 105/(5,5)-114)  103  3.17  ORTEP Diagram of Molecules in Complex 105m (2:1 105/(5,5)-115)  104  3.18  PLUTO Diagram of Molecules in Complex 105xo (2:1 105/(5,5)-114)  105  4.1  Inclusion Host Compounds  109  4.2  Thermogravimetric Analysis of Inclusion Complex 1:1 126/acetone  111  4.3  Differential Scanning Calorimetry of Complex 1:1 126/acetone  112  4.4  Differential Scanning Calorimetry of Complex 1:1 126/ethanol  112  4.5  Differential Scanning Calorimetry of Complex 1:1 126/isopropanol .... 113  4.6  ORTEP Stereodiagram of Complex 1:1 126/EtOH  116  4.7  ORTEP Stereodiagram of Complex 1:1 126/2-Propanol  116  4.8  ORTEP Stereodiagram of Complex 1:1 126/7-Propanol  117  4.9  ORTEP Stereodiagram of Complex 1:1 126/EtOAc  117  4.10  ORTEP Stereodiagram of Complex 1:1 130/2-Propanol  118  4.11  ORTEP Stereodiagram of Ethenoanthracene 130  118  4.12  Intramolecular Dipole Interactions  119  4.13  Packing Diagram of Complex 1:1 126/Ethanol  121  4.14  Packing Diagram of Complex 1:1 126/«-Propanol  122  4.15  Packing Diagram of Complex 1:1 126/1 sopropanol  122  4.16  Packing Diagram of Complex 1:1 126/EtOAc  123  4.17  Packing Diagram of Complex 1:1 130/Isopropanol  123  xvi  4.18  Photorearrangement of Ethenoanthracene 126  124  4.19  Portion of l^C NMR Spectrum and Assignment of Photoproduct 143  125  4.20  ORTEP Stereodiagram of Bisphosphine (+)-144  126  4.21  Absolute Configurations of Molecule 126 in Crystalline State  128  4.22  Absolute Reaction Pathways of Inclusion Complexes of 126  130  4.23  Molecular Predisposition by the P(l)-C(l 1)-C(12)-P(2) Torsion Angle  131  4.24  Absolute Asymmetric Photorearrangement of Ethenoanthracene 61 ..  132  4.25  Absolute Asymmetric Photorearrangement of Ethenoanthracene Salts 73 and 99  133  4.26  Equilibrium between the Enantiomers of 1', 1-Binaphthyl  135  4.27  Photochemistry of 9-Hydroxydibenzobarrelene Derivative 145  136  4.28  Photorearrangement of Ethenoanthracene 130  137  4.29  Partial  4.30  Vicinal Coupling Constants between Two Nuclei that are Bonded to  13  C NMR Spectrum of Photoproduct 149 its Assignment....  138  a Five-membered Ring  139  4.31  Dephosphorylation of Compound 149  140  4.32  Photorearrangement Mechanism of Ethenoanthracene 149  141  4.33  Cleavage of Cyclopropanol Intermediates  142  4.34  ORTEP Diagram of Inclusion Complex 1:1 130/isopropanol  145  4.35  Absolute Reaction Pathway of Inclusion Complex 130/zso-PrOH (1:1) in the Solid State  145  5.1  Ethenoanthracenes 132 and 133  147  5.2  Photorearrangement of Ethenoanthracenes 105 and 106  148  5.3  Different Hydrogen Bonded Species of Ethenoanthracenes 105 and 106 148  5.4  1,2 Aryl Shift Mechanism  149  5.5  Photorearrangement of Ethenoanthracenes 132 and 133  150  5.6  ORTEP Diagram of Photoproduct 152  150  xvii  5.7  Di-7t-methane Rearrangement Mechanism of Ethenoanthracenes 132 and 133  5.8  154  Thermodynamic Parameters for 3,3,5-Trimethylcyclohexanol Equilibrium  5.9  155  Steric Effect in the Vinyl-benzo Bridging of Ethenoanthracenes 132 and 133  156  5.10  Dissociation of Intermolecular Hydrogen Bond of Ethenoanthracene 133 156  5.11  Excited State Proton Transfer and 1,2-Aryl Shift for the Intramolecular Hydrogen Bonded Species of Ethenoanthracene 133  5.12  157  Excited State Hydrogen Transfer and 1,2-Aryl Shift in the Intermolecular Hydrogen Bonded Species of Ethenoanthracene 132  158  5.13  ORTEP Diagram of Ethenoanthracene 132  159  5.14  ORTEP Diagram of Ethenoanthracene 133  160  6.1  Preparation of [Rh(COD)((lIR, 12tf)-anthraphos)] +BF 4 "  163  6.2  Preparation of Ruthenium Complexes with Anthraphos  164  6.3  3  165  6.4  l  6.5  Attempted Preparation of Ru[ri3-CH3C(CH2)2](anthraphos)  166  6.6  Catalytic Hydrogenation of Z-(a)-acetamidocinnamic Acid (155) ....  167  6.7  Bisphosphine Ligands  167  6.8  Diazomethane workup of iV-acetylphenylalanine (156)  168  6.9  Preparation of the Dimeric Pre-catalyst  169  6.10  ORTEP Diagram of (-)-[Rh(COD)((l IR, 12fl)-anthraphos)] +  171  6.11  An Approximately C2 Symmetrical Edge-Face Arrangement of Phenyl  ! p NMR Spectrum of mww-RuHCl((l IS, 125)-anthraphos)2  H NMR Spectrum of Hydride in trans-RuHC\((lIS, 12S)-anthraphos)2 165  Groups in Rh(I)-Bisphosphine Complexes  xvin  172  6.12  Perspective View of [Rh(NBD)(S,S-norphos)]C104 (A), [Rh(COD)((lliU2tf)-anthraphos)]BF4 (B), and [Rh(COD)(£S-chiraphos)]C104 (C)  173  Al  Incremental Shifts (ppm) of the Aromatic Carbons  259  A2  Structure and Numbering of Compound 116  261  A3  13  262  A4  ABX Type Spectrum from  C NMR Spectrum of Compound 115 at 75 MHz 13  C NMR of Ethanoanthracene 135  xix  264  ACKNOWLEDGMENT I would like to thank my research supervisor Professor John R. Scheffer for his valuable guidance and encouragement on my research and study throughout the years, and I have learned much of great value from my experiences in his laboratory.  I  appreciate greatly his understanding and patience in helping me write this thesis. I would like to thank Professor James Trotter and members of his research group, Dr. Ray Jones, Dr. Bozena Borecka and Dr. Gunnar Olovsson, for the X-ray crystallographic analysis. My special thanks go to Dr. Tai Y. Fu for most of the X-ray crystallographic analysis in this thesis. I would like to thank and express my appreciation to Professor Brian R. James and his group members, especially Kenneth MacFarlane. With their assistance, I have done my research on asymmetric hydrogenation and gained additional knowledge on organometallic and catalytic chemistry. Thanks to all my friends for everything they have done for me, and special thanks to Janet Gamlin, Mardy Leibovitch, Brian Patrick and Joe C. H. Wu for proof-reading my thesis. Finally, I would like to express my appreciation for all the help from the NMR, Mass Spec and Elemental Analysis laboratories, and also from the Chemistry Department staff in various aspects.  xx  DEDICATION  To my parents, my wife and my newborn son  xxi  INTRODUCTION  l  As optically active materials continue to gain importance in the biological and physical sciences as well as in the technologies of molecular electronics and optics, so does the need for the efficient synthesis of chiral molecules. Over the past decade or so, chiral drugs have emerged as an important issue in the pharmaceutical industry as well as in various regulatory agencies. 1  The demand for single enantiomers rather than  racemates has stimulated the development of asymmetric synthesis in chemistry. There are a number of options to produce single enantiomers. Resolution has been the method of choice for isolating many chiral compounds, but it is labor-intensive and rather limited, even though modern chromatographic techniques can give very good resolution for many compounds. Nature provides an abundance of chiral compounds, which are 100% one enantiomer in many instances. These compounds are either isolated from natural sources or manufactured by enzymatic processes.  More enantiopure  compounds can be obtained by asymmetric synthesis.2  I. Asymmetric Synthesis An asymmetric synthesis is a reaction in which an achiral unit in an ensemble of molecules is converted by a reactant into a chiral unit in such a manner that the stereoisomer ic products are produced in unequal amounts. 3 The reactants include not only the usual chemical reagents but also solvents, catalysts and physical forces such as circularly polarized light. Enantiomeric excess (e.e.) is used to evaluate the efficiency of an asymmetric reaction.  In contrast, diastereomeric excess (d.e.) is used when  diastereomers are produced. Both are defined below:  2  e-e-% = T 1 ^ ! x 100 = %R- %S = - ^ ^ [*] + [S] [a]pUre de  [ - - % =[A] H}+ [B] fi, x 100 = %A - %B  Asymmetric syntheses can be carried out in solution as well as in the solid state. In the former case, catalytic asymmetric transformations attract a great deal of attention because of their economic use of optically active compounds. The solid state, with its high reaction selectivity, is also a very good medium for asymmetric synthesis. Previous chirality transfer approaches, such as classical optical resolution of racemates and transformation of chiral compounds, are stoichiometric in terms of chirality and require at least one equivalent of a chiral source in order to create a new chiral compound. Biological enzymatic reactions are excellent for their stereoselectivity, but rather limited in their application scope. More recently, however, certain metal complexes with chiral organic ligands have been shown to act catalytically and to multiply chirality.4 These organometallic catalysts are endowed with functionality and chirality which allow differentiation of diastereomeric transition states. Such molecular catalysts not only accelerate reactions of associated substrates, but also control the stereochemical outcome of reactions in an absolute sense. The catalysts are used in various organic reactions such as hydrogenation. With the breakthrough in asymmetric hydrogenation, Sharpless epoxidation and asymmetric isomerization of allylamines, other types of organic transformations are under intensive investigation.4 Asymmetric syntheses via photochemical reactions have been investigated both in fluid phases and in the solid state. Asymmetric induction in fluid phases usually gives photoproducts with low optical yields.5  In contrast, topochemically controlled  phototransformations in the solid state generally occur with good to excellent optical yields.^ Most of these reactions make use of chiral substituents that are chemically  3  bonded to the reagents.  For reactions in the solid state, not only the chirality of  molecules, but also the chirality of the crystal affects the stereochemical outcome of the reaction. By using crystal chirality, absolute asymmetric transformations in the solid state are possible." In these instances, prochiral molecules that form chiral crystals have been transformed into optically active products by using the asymmetric environment of the crystals as the sole source of chirality.  II. The Topochemical Postulate The topochemical postulate for solid state chemical reactions was first proposed by Kohlschutter in 1918.7 According to the postulate, reactions in crystals proceed with a minimum of atomic and molecular movement.  This postulate was further modified  many years later as a result of the studies of different solid state reactions. In the 1960s, Schmidt and his coworkers^ carried out pioneering work on solid state [2 -I- 2] photocycloaddition, and established that the reactivity of each substrate in the crystalline state is governed by its molecular packing in the crystal lattice. For example, trans-cinnamic acid derivative 1 reacted differently upon photolysis in solution as well as in three polymorphic crystalline forms, a, |3 and y, as shown in Figure 1. ° Photolysis of trans-cinmmlc acid derivative 1 in solution gives cw-cinnamic acid derivative 2, and no dimerization reaction is observed. In the a crystalline form, the molecules are packed in a head-to-tail arrangement, and the center-to-center distance between the two nearest double bonds is 3.8-4.1 A. Irradiation of an a form crystal afforded a [2 + 2] photocycloaddition reaction product, a-truxillic acid derivative (3), which is a centrosymmetric dimer. The molecules in the metastable (3 form are packed head-to-head with a nearest-neighbor contact between two double bonds of 3.9-4.1 A.  The mirror-symmetric (3-truxillic acid derivative (4) was formed upon  photolysis of the P form trans-cmnamic acid derivative in the solid state. The y form, in  4  which the nearest double bonds make contact at 4.7-5.1 A, is photostable.  This  photostability was interpreted as being due to lattice constraints which do not permit the potentially reactive centers to move sufficiently close together to form a photodimer. This example demonstrates both the topochemical control of solid state reactions and the reactivity differences among polymorphs. After their systematic studies on the [2 + 2] photocycloaddition reactions of trans-cinnam'ic acid and its derivatives, Schmidt and coworkers proposed that the center-to-center distance between two neighboring double bonds should be less than 4.2 A in order for them to take part in a photocycloaddition reaction. 8  Ar  Ar.  hu COOH  COOH  solution  Ar  Ar  HOOC  COOH  hu form  HOOC,  Ar  COOH Ar  Ar Ar COOH  COOH COOH  P form  COOH Ar, HOOC  V  hu COOH  y form  no reaction  Ar  Figure 1. The Photochemical Reactions of rra/zs-Cinnamic Acid Derivatives.  Cohen extended the topochemical postulate by introducing the reaction cavity concept to interpret the course of a variety of solid state reactions.9 In this approach, a reacting molecule in a crystal is considered as a substance lying in a cavity formed by the  5  presence of its adjacent molecules, and the shape of the cavity is set by the packing of the crystals (Figure 2). The atomic and molecular movements necessary for a reaction cause pressure on the cavity wall which may become distorted, but any distortion in the shape of the cavity will be restricted by the closely packed environment. Cohen has redefined the topochemical postulate to mean that those reactions which proceed under lattice control do so with minimal distortion of the surface of the reaction cavity. In cases where more than one reaction pathway is possible, the pathway leading to least disruption of the cavity would be favored. Re a eta nt  Transition  Products  I i  +*  )~£>  n  Figure 2. The Reaction Cavity Concept: Before reaction (full line), the transition state (broken line).  In Figure 2, there are two reaction pathways that occur via the transition states B' and C . For reaction I, the shape and size of the transition state B' resemble those of the reaction cavity, but for reaction II, there is a large shape change in the transition state C.  According to the reaction cavity concept, the formation of B is topochemically  feasible and thus more favorable than the formation of product C. Ramamurthy et al. 10 expanded the concept of the reaction cavity to include a number of organized and constrained media. The reaction cavity was redefined as an enclosure 6  that reduces the mobility of reactant molecules in at least one dimension and provides a boundary which reactant molecules may not cross without overcoming an energy barrier. Therefore, the size and shape of the reaction cavity can be stiff or flexible.  In the  former case, none of the guest molecules can diffuse out, and the cavity walls do not bend, as in the case of crystals and some inclusion complexes; in the latter case, some of the guest molecules may exit the cavity and the walls of the cavity are sufficiently mobile to allow considerable internal motion of the enclosed molecules, as in the case of micelles and liquid crystals. 10a, 10b interactions between the cavity wall and the guests are also taken into account. The active cavity wall may serve as a template for the guest as it proceeds to products; when there is no significant interaction it is considered to be passive. 10a, 10b Cavities are categorized as (1) "initial reaction cavity", defined by the space in which the excited states of reacting molecules are generated; (2) "effective reaction cavity", which encompasses the space the excited states and their intermediates explore from the time of their inception to the moment of their final product formation; and (3) "final reaction cavity", which includes only the space where the product determining steps occur. 10a T n e above concept has been useful in analyzing reactions carried out in different media. 10a The topochemical principle states that reactions in the solid state prefer to occur with a minimum amount of atomic and molecular movement. This implies that a certain amount of motion in the crystal lattice is tolerable. Many solid state organic reactions involve considerable molecular movement, and as a result, the crystal lattices are destroyed at the end.  However, there are a few examples of single crystal-to-single  crystal photochemical [2 + 2] dimerizations which represent the ultimate topochemical reactions with minimum molecular motions. H The crystallographic data show that the dimerization process in these "topotactic" crystal transformations requires very little motion of the atoms.  As a result, not only can the starting substrates convert  quantitatively, but the whole crystal remains single throughout the process. 7  Scheffer, Trotter and coworkers^ have systematically investigated unimolecular reactions in the solid state. They suggested that low energy ground state conformations of the reacting substrates in the solid state are mainly responsible for the photochemical reactions taking place there. They also developed a semi-quantitative model called steric compression to analyze specific intermolecular interactions and steric effects. 13 These ideas are based on the di-7t-methane rearrangement of dibenzobarrelenes and Norrish type II reactions. An example of solid state conformational control of stereoselectivity is given in Figure 3.14  O  (CH2)]D J ^ .  (O^B  ° 5  /3Q.„ (QJ2)B  ».,.X ^  (CH2)8 + ( C J I 2 ) B  O 6  O 7  ^ + (C£2)K)  *  j  ^  O 8  Figure 3. Stereoselective Norrish Type II Reactions of Macrocyclic Diketone 5.  Diketone 5 was found to crystallize in two different crystal forms with different conformations, and striking differences in the reactivity between the dimorphs were demonstrated in the solid state Norrish type II reactions.  As shown in Figure 3, the  photolysis of 5 in solution led mainly to the formation of Norrish type II cleavage product 8. In the solid state irradiation, the plate-like crystals gave rise to stereoselective formation of cw-cyclobutanol 6, and the needle-like crystals afforded £ra/w-cyclobutanol 7. Very little cleavage product 8 was detected in the solid state reactions. This was explained as being the result of conformational differences between the dimorphs and the geometric requirements of the reactions. 14  8  HI. The Di-7t -Methane Rearrangement The di-7t-methane rearrangement, one of the most thoroughly studied organic photoreactions, was developed conceptually by Zimmerman in the 1960s. ^  The phrase  "di-7t-methane" denotes a general reactant structure of two 7t-bonds that are separated by a methane or sp3 hybridized carbon atom. The simplest di-7t-methane system is 1,4pentadiene (9, Figure 4).  The proposed di-7i-methane rearrangement mechanism is  shown in Figure 4, and diene 9 is used to depict the skeletal change during this rearrangement.  v^  hv  9  10  11  12  Figure 4. The Postulated Di-7i-methane Rearrangement Mechanism. As depicted in Figure 4, Zimmerman postulated that the excited state of 1,4pentadiene (9) bridges to form a new sigma bond and afford a cyclopropyldicarbinyl 1,4diradical 10, which proceeds onward to 1,3-diradical 11. The final ring closure yields vinylcyclopropane (12). The diradicals 10 and 11 were first proposed by Zimmerman as approximations of species along the reaction pathway, but recently Zimmerman et al. showed that the type of diradical 10 is a true, thermally equilibrated intermediate. l" a These radical species have proved very useful in understanding, rationalizing and even predicting the course of reactions that will be discussed later. (i) Reaction Mechanism Although the di-7t-methane rearrangement has been studied extensively for three decades, the reaction mechanism remains somewhat uncertain. The first mechanism was  9  postulated by Zimmerman and is shown in Figure 4. Two diradical intermediates were proposed in this mechanism, and many efforts have been made to confirm their existence by theoretical calculations and experimental approaches. Good evidence came from the independent  generation  of  these diradical  species.  In one example,  the  cyclopropyldicarbinyl diradical (diradical 10 in Figure 4) was generated by Zimmerman via nitrogen extrusion from the appropriate azoalkane (Figure 5). 16  y.  14  R  =H,H =benzo = 2,3-naphtho R=H,CH3  Figure 5. The Denitrogenation Reaction of Azoalkane 13. It was observed that the ground state diradical species 15, which was presumably generated by the thermolysis of azo compound 13, led quantitatively to Grob fragmentation and cycloreversion to the corresponding barrelene 14 (R = H).16b The direct photolysis (via the singlet excited state) of compound 13 (R = H) afforded mainly two compounds, 14 and 16.16b i n the triplet-sensitized irradiation of 13, the di-7tmethane product semibullvalene(s) (16 and 17 when R = CH3) was formed  10  exclusively. 16 This example provided evidence that diradical 15 could lead to both the starting material and the product of the above di-7t-methane rearrangement.  The  identical regioselectivity in the triplet-sensitized photolysis of compounds 13 and 14 (R = CH3) indicated that the diradical 15 is a thermally equilibrated intermediate. 16a However, it was questioned by Adam et a l . ^ whether the diradicals generated by nitrogen extrusion were indeed the same as the ones involved in the di-7c-methane rearrangement.  They suggested that the di-71-methane rearrangement and the  photochemical denitrogenation of azoalkanes were disjointed chemical events. Turning to the second type of diradical proposed in the Zimmerman mechanism (diradical 11 in Figure 4), Zimmerman suggested that for the triplet di-71-methane rearrangement of bicyclic barrelene systems, these species are true intermediates. This conclusion was based on photochemical studies of the deuterium labeled barrelene system. 1° The photolysis of barrelene 18 is outlined in Figure 6, where the bridgehead deuterium labels are depicted by filled circles. It was found that the triplet-sensitized irradiation of barrelene 18 led to the formation of semibullvalenes 22 and 23 in a 1:1  hv  13  20  19  21  C£)  CB  22  23  Figure 6. The Photolysis of Deuterium Labeled Barrelene 18.  11  ratio. This was considered to be the result of radical coupling reactions occurring in the diradical resonance structures 20 and 21. Therefore, this diradical was taken to be a true intermediate of modest lifetime. ^  Furthermore, photochemical studies by Paquette on  deuterium labeled dibenzobarrelenes and benzonorbornadienes also strongly suggest the direct intervention of the second type of diradical species during the di-7t-methane reaction process. 19 Schaffner and co-workers^O studied the di-Ti-methane reaction process using spectroscopic methods.  The low temperature di-7i-methane rearrangement of a  naphthobarrelene-like compound was monitored by ESR and IR spectroscopy, and the experimental results suggested the existence of diradical intermediates.™ A question raised from studies of the aryl di-7i-methane rearrangement is whether there is any need for the formation of the first type of cyclopropyldicarbinyl diradical in this reaction. As shown in Figure 7, the formation of diradical 27 results in the loss of aromaticity in the aryl moiety. Alternatively, a 1,2-aryl shift mechanism, with the direct  ^-""^  1,2-aryl shift  —  ^"^  24  25 A  27 Figure 7. The 1,2-Aryl Shift Mechanism.  12  26  formation of the second type of 1,3-diradical, was proposed by Paquette and coworkers. 21 This mechanism is depicted in Figure 7. Both of the above-mentioned mechanisms have explained some di-7t-methane reactions, but the Zimmerman mechanism is generally acceptable, as the diradical species have been shown to be true intermediates. 16a (ii) Reaction Multiplicity One of the striking facets of the di-7c-methane rearrangement is its dependence on reaction multiplicity. l^ a Generally speaking, acyclic systems undergo the di-7c-methane rearrangement via singlet excited states, whereas cyclic systems react via triplets. This dependency is a result of possible alternate reaction pathways and how the structure of the molecule, in particular the diene unit, is suited for each of the available reactions. The most important of the competing pathways are cis-trans isomerization22 and cycloaddition.23  A comparison of the rate constants for each of these processes is  shown below: l^b Singlet state lk£ A > lkDpM> *kcn Triplet state 3 kc T I > 3 k DPM > 3 kc A Where: CA = Cycloaddition reactions DPM = Di-7u-methane rearrangement CTI = Cis-trans isomerization The comparison demonstrates that, in the triplet state, an acyclic compound will undergo cis-trans isomerization before the di-7t-methane rearrangement, whereas a cyclic compound, being unable to cis-trans isomerize, can readily rearrange. Thus the di-7tmethane rearrangement of cyclic diene systems generally occurs from the triplet excited state and is often in competition with cycloaddition reactions. On the other hand, acylic diene systems undergo the di-7i-methane rearrangement from the singlet state, often in competition with the cis-trans isomerization.  13  Multiplicity has been found to control the course of the di-rc-methane rearrangement. A dramatic example is shown in Figure 8.24  j  nm  j s example, compound 28 gave  photoproduct 29 on direct irradiation, and photoproduct 30 from sensitized photolysis. This behavior has been rationalized from a quantum mechanical point of view.  Ph -Ph direct Ph  / ^h  -C02Me C 0 2Me  Ph  C02Me 29  28  Ph  —  30  C02Me  p h  Ph  C02Me  31  Figure 8. Multiplicity Dependence of the Di-7i-methane Rearrangement of 28.  (iii) Regioselectivity of the Di-7i-methane Rearrangement The di-7r-methane rearrangement is very susceptible to the control of its regioselectivity by substituents. As an important part of di-7i-methane photochemistry, regioselectivity studies have played a significant role in understanding and establishing the reaction mechanisms.  Regioselectivity has been correlated with the different  stabilities of the two possible diradical species (10 and 11 in Figure 4) that can be produced in the initial and the following step of the reaction mechanism. The regioselectivity of the di-7t-methane rearrangement of acyclic 1,4-dienes has been investigated systematically by Zimmerman and co-workers. ^ a As an example, the di-7r-methane rearrangement of unsymmetrically substituted diene 39 is shown in Figure 9. Zimmerman and Pratt^ reported that direct irradiation of l,l-diphenyl-3,3,5-  14  trimethyl-l,4-hexadiene (39) afforded compound 38 as the only photoproduct.  As  depicted in Figure 9, Zimmerman et al. suggested that there are two alternative ringopening processes, "a" and "b", that convert the initially formed diradical 34 into diradical species 35 and 37, respectively. Process "a" utilizes the odd electron at the benzhydryl center, and the resulting formation of diradical 35 leads to the loss of benzhydryl delocalization. In contrast, process "b" uses the less stabilized odd electron and goes on to form diradical 37, which still possesses the delocalization energy. 15a  Ph  Ph  Ph  Ph  35  36  ^1 Ph  Ph  k  Ph  37  Ph  38  Figure 9. Delocalization Control of Regioselectivity of the Di-u-Methane Rearrangement of 39.  (iv) Photochemistry of Dibenzobarrelene and Derivatives Dibenzobarrelenes and its derivatives have been found to undergo three types of photochemical  reactions:  the  di-7t-methane  rearrangement,  the  tri-7i-methane  rearrangement and an intramolecular [2 + 2] photocycloaddition reaction. The study of the barrelene class of compounds has been extended to include photoreactions of benzobarrelene,26 dibenzobarrelene, and notably to various  15  derivatives. 12c,27,28,30 j n 1966, Ciganek studied the solution photochemistry of the dimethyl dibenzobarrelene diester derivative 40.27 The mechanism postulated for this transformation is consistent with Zimmerman's proposal and involves bond formation between a bridging vinyl carbon and an adjacent aromatic ring carbon, referred to as vinyl-benzo bridging (Figure 10). Subsequent studies revealed that the di-7t-methane rearrangement of dibenzobarrelenes is a triplet-specific reaction. 15a j n contrast, the direct irradiation (via the singlet excited state) of some dibenzobarrelene derivatives has been reported to give dibenzocyclooctatetraene (COT) type photoproducts.29 Based on well-established benzo- and naphthobarrelene reaction mechanisms, it has been proposed that this reaction involves an initial intramolecular [2 + 2] photocycloaddition through the singlet excited state followed by thermal reorganization of the resulting cage compound (Figure 11).30 The direct thermal conversion of the intramolecular [2 +2] cycloadduct to COT is symmetry-forbidden, and, although a non-concerted process is possible, Scheffer et al. suggested that the process involves a retro Diels-Alder reaction, followed by a symmetry-allowed electrocyclic ring opening.28  41  42  43  44  45  45  Figure 10. Photorearrangement of Dimethyl Dibenzobarrelene Dicarboxylate (40).  16  not isolated 46  47  48  a. R2 = H, R2 = H. b. Rx = H, R2 = C(Me)2OH Figure 11. The [2 + 2] Photocycloaddition Reaction of Dibenzobarrelenes.  The term "tri-7i-methane rearrangement," which was proposed by Zimmerman et al., was used to describe the interactions among three n moieties within one molecule during a photorearrangement.31 An example of a tri-7i-methane rearrangement was reported by Scheffer et al.; the postulated reaction mechanism is depicted in Figure 12.32 The tri-7t-  solid state E=CQ2Me 49  tri-7r-methane mechanism  t E Me  Me 53  54  Figure 12. The Tri-7t-methane Rearrangement of Compound 49.  17  methane rearrangement was considered to be a singlet photochemical reaction. Steric crowding was suggested as the reason why these dibenzobarrelenes undergo the tri-7tmethane rearrangement rather than the [2 + 2] photocycloaddition reaction. Because there are four substituents on the barrelene unit, the [2 + 2] photocycloaddition of compound 49 will lead to a more sterically congested intermediate than the tri-Ti-methane rearrangement intermediate 54.  IV. Asymmetric Photoreactions in the Solid State Chiral crystals are crucial for asymmetric photochemical syntheses in the solid state. There are 230 possible ways to pack molecules into a crystal lattice and these are called space groups. Of the 230 possible space groups, 65 are chiral. Crystal chirality arises from the dissymmetric spatial arrangement of the molecules in the crystal lattice. Consequently, all resolved chiral molecules must crystallize in chiral space groups. This provides a way to create chiral crystals for prochiral molecules by forming salts or complexes with resolved chiral molecules. Racemic compounds will either crystallize in an achiral space group that contains equal amounts of each enantiomer, or they will spontaneously resolve into chiral crystals of each enantiomer. It is far from common that racemic compounds resolve spontaneously upon crystallization.33 However, this phenomenon allowed Pasteur in 1848 to perform the first optical resolution utilizing racemic sodium ammonium tartrate. Achiral molecules can also spontaneously deposit chiral crystals.  If a system reaches supersaturation under racemizing conditions,  intentional seeding, stirring and accidental seeding from the environment could result in the formation of enantiomorphously pure crystals.34 The methodology of using crystal chirality to generate molecular chirality was recognized by Schmidt, and the first example of a topochemical asymmetric synthesis using chiral crystals was reported by Penzien and Schmidt in 1969 (Figure 13).35 They  18  demonstrated that the gas-solid reaction between single, enantiomorphously pure crystals of 4,4'-dimethylchalcone (55) (space group P2\2\2\) and gaseous bromine gave the corresponding dibromide product 56 in up to 25% optical yield. The first absolute asymmetric synthesis via a solid state photochemical reaction was also reported by Schmidt and co-workers in 1973. They conducted [2 + 2] photodimerizations in the chiral space group P2\2\2\ to afford up to 70% enantiomeric excesses.36  Ar  ft  Br I  Br2 vapor  Ar^-tolyl  55  Tf  Ar  II 56  Figure 13. Solid State Asymmetric Bromination. Crystallization followed by solid-state reaction has been recognized as a route to asymmetric synthesis from achiral starting materials.6,37  Solid state photochemical  reactions are enviable for their regio- and stereoselectivity because of the steric interactions between the molecules in the crystalline lattices.38 Chiral molecules usually give higher diasteroselectivity when reacted in the solid state than in solution. Prochiral molecules, whether spontaneously crystallized in a chiral crystal or cocrystallized with other resolved chiral molecules, could react to give optically active product(s). (i) Absolute Asymmetric Photochemical Synthesis Asymmetric dimerization was one of the first photoreactions applied for asymmetric induction°\36  an( j  has been most intensively studied.39 The absolute stereochemical  reaction pathways were analyzed by X-ray crystallography.40  More recently, an  absolute asymmetric synthesis by [2 + 2] photocycloaddition of a charge-transfer complex was followed by X-ray powder diffraction and claimed to be a single crystal-to-  19  single crystal transformation (Figure 14).41 The enantiomeric excesses for this type of reaction range from moderate up to a high of 95%.  Intramolecular [2 + 2]  photocycloaddition was also shown to give high enantioselectivity. Sakamoto reported that an achiral acyclic imide underwent [2 + 2] cycloaddition to give a chiral oxetane in >95% e.e. (by polarimetry).42  Figure 14. Asymmetric Single Crystal-to-Single Crystal [2 + 2] Photoaddition.  Unimolecular photoreactions other than cycloaddition have been investigated for the purpose of absolute asymmetric induction. Two examples of Norrish type II reactions in chiral crystals have been reported, which are shown in Figure 15. Scheffer, Trotter and co-workers discovered that the adamantyl ketone 57 crystallizes in a chiral space group Pl\2\2l,  and solid state photolysis yielded cyclobutanol 58 in 80% enantiomeric excess  (Eq.l, Figure 15).43 The other example comes from Toda et al. Compound 59 in space group Fl\l\l\  was photoisomerized to afford P-lactam 60 in 93% optical yield  (Eq 2, Figure 15). 44 The di-7t-rearrangement of dibenzobarrelenes has been thoroughly investigated in our laboratory. 12,28  Several dibenzobarrelenes have been shown to undergo absolute  asymmetric photochemical rearrangements in the solid state. The enantioselectivity is high when the photolysis is carried out to low conversion. dibenzobarrelene 61 crystallizes in the chiral space group Fl\l\l\,  20  For example, and the solid state  di-7i-methane rearrangement of the crystal gives photoproduct 62 in > 95 % optical yield at low conversion.43, 45 (gq 3^ Figure 15).  (1)  hu, Fl\2\2\  Me  e  Ar=/7-aC6Hr  ^ Ar  7lt^2^ OH 58  hu,P2i2i2i (2)  ph  .  OH CH3 -CH3  PhCO-CON(/-Pr)2 -N  59 0  60  R0 2 C  CO2R  ^  ^  C0 2 R  hu, P2{L{2X * •  (3)  R = i-Pr, Et  61 Figure 15. Unimolecular Absolute Asymmetric Photoreactions in Chiral Crystals.  (ii) Photochemical Asymmetric Induction in Enantiopure Hosts Toda's group has succeeded in controlling the stereochemical courses of photoreactions of guest molecules in crystalline inclusion complexes of resolved host molecules.46  As shown in Figure 16, Toda et al. carried out different reactions  including inter- and intramolecular [2 + 2] electrocycloadditions (Eq 3 and 4), electrocyclic ring closure (Eq 1) and Norrish Type II photocyclization (Eq 2). Different resolved host molecules were used, and two typical examples, 71 and 72, are shown in Figure 17.  21  o .OR hu  (1)  Hectrocyclic reaction R=Me,Et  Host  e.e.%  71  100  63  O  O  hu  (2)Ph -y^ X N Me'  OH Ph- T  _^  NorrishtypeH  Host  e.e.%  Chemical yield  71  100  Host  e.e.%  Chemical yield  100  57%  Host  e.e.%  Chemical yield  71  78  -N  'Me  O"  Me  82%  66  65  O (3)  hu [2+2]  MeO MeC>2  MeO MeO^C  67  72a 68  O (4)  P  Q  hu [2 + 2] 69  55%  70  Figure 16. Asymmetric Reaction in Inclusion Complexes.  Ph2C —OH  i  CI-  O  €1 Ph" C - C = C - C = C - C ^ P h OH  Ph 2 C-OH  (CH2)n  O  OH 72a 72b  (R,R>(->71  n=2 n=3  Figure 17. Typical Resolved Host Molecules.  22  A wide range of host and guest molecules was studied, some of which are shown in Figures 16 and 17. It is noteworthy that most of the products obtained within the inclusion complexes are formed in good to quantitative enantiomeric excesses.47,48 Other groups have used inclusion complexes for asymmetric inductions in solution. For example, Weber et al. carried out photocatalytic enantiodiscriminating oxidations of a-pinene by using cyclodextrin-linked porphyrines and molecular oxygen.49 Given that cyclodextrin forms a more favorable inclusion with one enantiomer than the other of apinene, enantioenriched products were formed. The enantiomeric excesses ranged from near zero to 40%, depending on the solvents used. (iii) Photochemical Asymmetric Induction in Chiral Salts Prochiral acids and amines can be forced into chiral space groups by formation of salts with resolved amines and acids respectively. The prochiral compounds are "preresolved" in these chiral crystals, and upon photoreaction, asymmetric induction can occur. Moderately high enantioselectivities have been obtained in the solid state di-7imethane rearrangement and Norrish Type II reactions. Three examples are given in Figures 18.50,51 and 19.52 in addition to the high optical yields observed in the di-7trearrangement of dibenzobarrelenes 73 and 78,50 information was obtained on the absolute stereochemical pathways of these rearrangements (Figure 18).51 Enantiomeric excesses were achieved as high as 97% for the Norrish type II reaction of a-admantantyl derivative 79 when a salt with prolinol was used (Figure 19).  The absolute  stereochemical reaction pathways in the solid state were rationalized according to topochemical principles.51>52 Qn the other hand, the solution photolysis of all the above salts gave racemic products.  23  XO2C  CQzX  HOzC  CQ2B  CO2X  hu »•  ciystaJ  COOH X=  +/  H2N - > - ,"] \ /  EnantiomerofX  e.e.%  RS-  Yield (%)  (->80 (+>76  e.e.% 0 0  96 94  4 6  NH 3 A hu E=CP2Et ciystal  E  +  O  A=  Yield (%)  Enantiomer of A  e.e.% of 76  (+) (-)  (+>64 (->68  SOsH  Figure 18. Photoasymmetric Induction of Salts via the Di-71-methane Rearrangement.  H hu AT  ciystal  "Ar  Z±T7  79  a^j 81  80 HOCH,  Ar =  OH  -@-°* »x£j  Enantiomer of Amine S<+) S-(-)  81/80 Ratio 5.5/1 5.5/1  e.e.% of 81 (+>97 0-97  Figure 19. Asymmetric Induction of Salt via the Norrish Type II Reaction.  24  V. Catalytic Asymmetric Hydrogenation Catalytic asymmetric hydrogenation has become one of the most powerful tools for the synthesis of optically pure amino acids, and has had a broad general impact on asymmetric catalysis. The efficiencies of those transition metal based catalysts are often comparable or even superior to those of biocatalysts. This field has been reviewed quite often since the late 1970s.53a (i) Mechanism The mechanism of asymmetric hydrogenation of dehydroamino acids catalyzed by cationic chiral rhodium complexes of cis-chelate. diphosphine ligands has been revealed though the work of Halpern and his coworkers.53 Figures 20 and 21 illustrate the reaction and the mechanism respectively of the [Rh(R,R-Dipamp)]+-catalyzed hydrogenation of methyl-Z-(a)-acetamidocinnamate (MAC).  COOCH3 /P^\  <Q  NHCCH3  &  H  2 » catalyst  Figure 20. Asymmetric Catalytic Hydrogenation of MAC.  The oxidative addition of molecular hydrogen to [Rh(/?,/?-Dipamp)(MAC)]+ was found to be the rate-determining step, and two diastereomeric catalyst-substrate complexes, major and minor, are formed, which proceed to produce R and S products respectively.  The minor intermediate has been shown to have a greatly enhanced  reactivity over the more stable major component, and this reactivity then determines the configuration of the hydrogenated product.  25  H.  OChMe C=C' ^NHCOMe Ph MeOzC^NH  HN __.C0 2 Me Me  h^Ph/  \/  fV  P\  I ""  H7  Me H  ^O  Me  H  Me  S  ~Rh  ) w r t „ \ ^ R h v MeQzC—Y | Hy*  Me-^-NH-^pCOjMe  °  MeCy^?^ Ph^  '-Ph  7? minor product  Me °  major product  Figure 21. Mechanism of Asymmetric Catalytic Hydrogenation of Methyl Z-(cc)-Acetamidocinnamate.  (ii) Chiral Ligands In the early 1970s, Kagan, Knowles and others developed Rh(I) complexes of chiral chelate bisphosphines (diop and dipamp respectively) that had great success in the asymmetric hydrogenation of a-acylaminoacrylic acids.54,55  Since then, a large  number of enantiomerically pure bisphosphines have been synthesized and used as chiral ligands.56 Some typical bisphosphines are listed in Figure 22, which include diop (82)54, dipamp (83)55, chiraphos (84)57, nor phos (85)58, BPPFA (86),59binap (87), 60 BPPM (88)^1 an d duphos (89).62 Besides bisphosphines, chiral bidentate ligands such  26  as diphosphinites, bis(aminophosphine), and aminophosphine-phosphinite have also been developed. 63  ..,L- PPh 2  84 Chiraphos  83Dipamp  85 Norphos Rn.  Ph2P.  NMe7  tX  I PPh2 COtf-Bu)  87Binap  88 BPPM  R  x>  R=Me,Et  89 Duphos Figure 22. Bisphosphine Ligands for Asymmetric Catalysis.  Although there are still many unanswered  questions regarding  asymmetric  hydrogenation and associated ligand-substrate effects, several trends have emerged from the numerous reactions carried out over the years using different ligand-metal-substrate combinations. ones.64,53b  Chelate phosphines seem to be more effective than the monodentate The bisphosphines that form a more rigid five-membered chelate ring  generally seem to have a higher enantioselection over the ones that form more flexible six- and seven-membered chelate rings.64 More precisely, it has been suggested that a chiral "face-edge" array of phenyl groups, induced by the chelate ring conformation, can be the origin of asymmetric bias, which discriminates the enantiotopic faces of the prochiral substrate in the binding step.64  27  (iii) Chiral Catalysts Rhodium complexes of chiral bisphosphines were the catalysts commonly used during the developing stage. These complexes are prepared in situ by mixing [RhCl(diene)]2 or [Rh(diene)2]Y (Y = BF4 and CIO4, etc.) and a bisphosphine, and then using it directly as a catalyst. However, one must be careful in the preparation of the catalyst, since rhodium species without phosphine ligands often work as catalysts for hydrogenation. Isolated complexes could be obtained from methanol or other solvents. In the late 1970s, James et al. reported pioneering work on the use of chiral Ru(II) complexes as catalysts for asymmetric hydrogenation. 65  The Ru complexes  Ru2Cl4(diop)3, RuHCl(diop)2 and RuCl2(diop)2 were synthesized by a ligand exchange reaction of RuCl2(PPh3)3 with excess diop, and used as catalysts for asymmetric hydrogenation of olefins. Binap-Ru(II) complexes were developed in Japan as the most efficient catalysts among ruthenium based catalysts for asymmetric synthesis. Two of those complexes, 90 and 91, whose structures are given in Figure 23, were often used and gave very high enantiomeric excesses for asymmetric hydrogenation of olefins and other substrates. Complex 90 could be prepared directly or indirectly from different sources such as [RuCl2(COD)]2, [RuCl2(benzene)]2,66 and complex 91 was prepared from the corresponding ruthenium-arene complexes."? Other mononuclear Ru(II) complexes with bisphosphines have been prepared from different sources, but few were fully characterized or shown to be efficient catalysts for asymmetric catalytic reactions. Genet and coworkers formulated a general synthesis from Ru(2-methylallyl)2(COD) and various bisphosphines to reach complex 92 (Figure 23).^^  Subsequently, complex 92 could be transformed easily into other  catalytic precursors by reaction with HC1, HBr, HI or other carboxylic acids.  28  90  91 Figure 23. Ruthenium-Based Chiral Catalysts.  92  (iv) Substrates Rh, Ru and Ir complexes with chiral ligands have been prepared as catalysts for the asymmetric hydrogenation of C=C, C = 0 and C=N bonds. With the breakthrough in the hydrogenation of C=C bonds, the asymmetric hydrogenations of C = 0 and C=N moieties are of current interest. oc-Acylaminoacrylic acids were the first olefinic substrates successfully used in the homogenous asymmetric hydrogenation.  A number of Rh-bisphosphines catalyst are  known to be efficient catalysts for this type of asymmetric hydrogenation.63a it  was  revealed that olefins with additional coordinating functionalities such as amido, carboxyl, amidomethyl, carbalkoxymethyl and hydroxycarbonylmethyl groups, are a requisite for getting higher enantioselectivities. Employment of bisphosphine-Ru(II) complexes^ as catalysts, however, has greatly expanded the kinds of olefinic substrates applicable for the reaction, and now the asymmetric hydrogenation of olefins catalyzed by chiral transition metal complexes is one of the most practical methods for preparing optically active organic compounds. Chiral cyclopentadienyl complexes of titanium™ have also opened a new direction for asymmetric hydrogenation of simple olefins. The asymmetric hydrogenation of functionized ketones catalyzed by chiral transition metal complexes has attracted much interest.4 Bisphosphine complexes of Rh and Ru have been used as catalysts for this type of reaction. Bisphosphine-Ru complexes are  29  also excellent hydrogenation catalysts for kinetic resolution of chiral u-substituted 13-keto ’ 7 esters.  High enantioselectivity in the hydrogenation of simple ketones has been  difficult to attain with conventional chiral bisphosphine complexes.  However, some  complexes of Rh and Ir with chiral nitrogen ligands were shown to be excellent catalysts b 63 for enantioselective transfer hydrogenation of alkyl aryl ketones. Relatively few examples of the enantioselective hydrogenation of imines have been 72 Most of the studies have involved empirical testing for matching of metal, published. ligand and substrates, and it was reported that cyclic or aryl-containing imines generally gave products of higher enantiomeric purity.  VI. Organic Clathrate Chemistry The term clathrate was introduced by Powell 73 in 1948 to describe a particular form of a molecular compound in which one component (host) forms a cage structure imprisoning the other (guest). This type of compound is also called crystal inclusion, as its formation results from crystal lattice forces. hydrogen bonding also contribute.  However, coordinative forces such as  74 termed the crystal clathrates as Weber  coordinatoclathrates and clathratocomplexes. The former demonstrate a certain degree of coordinative participation but have a dominant clathrate character, and the latter ones behave the other way around. There are several reviews on this subject. 75 Different types of clathrate hosts have been developed, and most of the classical ones have been discovered by accident and not via directed synthesis. Figure 24 gives a brief review of some examples. Individually designed elements involve molecular bulkiness, rigidity, hydrogen bonding, Coulomb attraction as well as the use of particular molecular architectures, among them those featuring a “wheel-and-axle” ,76 a “roof”, or “scissors” (Figure 24).77  30  95 roof-shaped  96 scissor-Eke  Figure 24. Typical Inclusion Host Molecules.  One of the important applications of host-guest chemistry is directed to chemical analysis and molecular separation processes. Chemically different species are separated as well as constitutional isomers, regioisomers, stereoisomers^^ and even isotopic isomers. In the case of enantioselective guest inclusion, a chiral host lattice is required. The solid state chemistry of clathrates leads to many synthetic uses. The crystal lattices of clathrates are in a strict periodical arrangement of structurally identical molecular units and contain holes, channels, layers etc., which may include guest molecules in an oriented fashion. One of the early studies in this area was the inclusion polymerization of dienes in the channels of thiourea, leading to stereoregular polymers. '4  An enantioselective reaction is expected when optically active host  compounds are used. Toda's group has applied these principles successfully to solid state asymmetric induction.46,79  31  VII. Research Objectives and Outline of the Thesis This thesis is an integrated study of asymmetric reactions in the solid state as well as in solution.  Specific organophosphorus compounds were chosen to serve different  purposes. The chemistry and photochemistry of the phosphine oxides were investigated with the emphasis on the solid state photochemistry. Specific phosphine oxides were designed for crystalline host-guest complexes that would undergo the solid state di-Ttmethane rearrangement reaction. Much work has been done on the di-7t-methane rearrangement of dibenzobarrelenes in the solid state, 12,50,51,80,81 5ut uncertainties in their solid state photobehavior still exist. In order to further understand this reaction, a very specific reactant was aimed to be photolysed in a series of crystalline host-guest assemblies in which the photoreactant was kept constant and a variation was only made in the unreactive counterpart. The required crystallinity of the assemblies allowed their structures to be analyzed by X-ray crystallography. Due to the constant reactant in different assemblies, specific factors that affect the reaction could be identified according to the X-ray crystal structures. Etter and Baures^Sb showed that triphenylphosphine oxide forms large, high-quality crystals when cocrystallized with a variety of hydrogen-bond donors. With our interest in the design of the above guest-host assemblies, triphenylphosphine oxide (TPPO) as well as bisphosphine dioxides (VII) were selected to cocrystallize with dibenzobarrelene carboxylic acid derivatives (VDI) (Figure 25). Of particular interest was the possibility of utilizing P-chirality to prepare enantiomorphously pure complexes for solid state asymmetric induction studies. We expected that these cocrystalline complexes would be high-quality crystals, which could be analyzed by X-ray crystallography, and that the dibenzobarrelenes could undergo di-7i-methane rearrangement while the phosphine oxides are photochemically inactive. This project is included in Chapter 3.  32  o  o  II  II  Ar  Ar  -p— CH2CH2 — p  vn Ar=  CH3  ~CH -© ,CH3  CH3  V, T CH3  vin  'r~S\  R = —CH2CH3 -CH(CH3)2  (DP^O)  Figure 25. Donors and Acceptors of Cocrystalline Complexes.  In the above cocrystalline complexes, the phosphine oxides are used as the variable hosts, the dibenzobarrelenes as the constant photoreactive guests. It may be possible to design a dibenzobarrelene host that can not only include variable guest molecules but also undergo the di-u-methane rearrangement. The inclusion complexes of the same host but with different guests would provide information about the solid state photochemical reactions. Weber and Czugler^^ have delineated some of the features of good hosts: they should be bulky and pack inefficiently with voids, and they should contain appended sensor groups that will coordinate to the guests. One of the most important types of coordination is hydrogen bonding, Etter and co-workers have carried out important studies in this area, including the demonstration that triphenylphosphine oxide, an excellent hydrogen-bond acceptor, forms crystalline complexes with a wide variety of hydrogen-bond donors.78b  T n e concepts of Weber and Etter were combined in the  design of the dibenzobarrelene host: a bulky "roof-shaped" coordinatoclathrate with appended diphenylphosphinoyl as sensor groups. Figure 26 illustrates two examples; we expected  that  the  dibenzobarrelene  skeleton  would  undergo  the  di-rc-methane  rearrangement^ and that the compounds could crystallize in a chiral space group for the  33  solid state absolute asymmetric synthesis. Several cases of solid state photochemical absolute asymmetric induction have been investigated, but few can be used for synthetic purposes.42,43,45 7 ^ above hosts could create a series of inclusion complexes with different guests. Thus, the chances for the hosts to crystallize in a chiral space group are enhanced, and an absolute asymmetric synthesis is possible in these supramolecular assemblies. This project is included in Chapter 4. O  O  Figure 26. Phosphine Oxide Derivatives of Dibenzobarrelene.  The di-71-methane rearrangement reaction is very susceptible to the control of its regioselectivity by substituents. ^ Two compounds shown in Figure 27 were designed in order to study substituent effects by the phosphinoyl group. This project was aimed at determining the regioselective directing ability of phosphinoyl and ester/acid groups. The striking character of the phosphinoyl group is its ability to form strong hydrogen bonds, and these two compounds may serve as good models to study hydrogen bonding effects in the rearrangement. The photochemistry of these dibenzobarrelenes is included in Chapter 5.  R=CH2CH3 R=H  Figure 27. Phosphine Oxide Derivatives of Dibenzobarrelenes.  34  Chapter 1 is concerned with the synthesis of the required starting materials. Methodologies were developed for preparing phosphine oxide derivatives of ethenoanthracene (dibenzobarrelene) and ethanoanthracene.  In conjunction with the  asymmetric reactions, different methods were used in Chapter 2 to determine the optical purities. A new complexing chiral shift reagent, dibenzoyltartaric acid, was studied for measuring the enantiomeric excesses of the phosphine oxides. Chapter six deals with asymmetric catalytic hydrogenation.  Two bisphosphine  ligands (Figure 28) that were prepared in Chapter 1 and Chapter 4 were used. Both bisphosphines were examined as ligands for the asymmetric hydrogenation of Z-(a)-Nacetamidocinnamic acid.  Further investigation was carried out on the rhodium and  ruthenium complexes of anthraphos. O  O  anthraphos  Figure 28. Bisphosphine Ligands.  35  RESULTS AND DISCUSSION  36  CHAPTER ONE GENERAL ORGANIC SYNTHESIS  1.1 Synthesis of Optically Pure Phosphine Oxides with P-Chiral Centers There are a number of methods for the preparation of single enantiomers of tertiary phosphine oxides with P-chiral centers.82-86  Qne of the most common methods  involves stereospecific synthesis via displacement reaction at the P-center. This method consists of the preparation of a diastereomeric mixture of compounds with P-O, P-S or P-N bonds from available optically pure compounds and phosphorus compounds, followed by the separation of the diastereomers and the displacement of the OR, SR or NR heteroatom group with Grignard reagents or organolithium compounds. The two diastereomers differ in their configuration at the P-chiral center (7?p or SW), and either one could be obtained stereoselectively during their synthesis. In this thesis, the optically pure monophosphine oxides 108-113 were prepared from the corresponding aryl Grignard reagents and menthyl methylphenylphosphinate (ftp- or Sp-107) by modifying the literature method (Figure 1.1). 87 The diastereomeric mixture of menthyl methylphenylphosphinates (107) was obtained according to the sequence of reactions as shown in Figure 1.1.87 T n e separation of diastereomers Rn- and Sp-107 has been achieved by fractional recrystallization of the mixture from hexanes according to Mislow's method87 or from 25% pinene in hexanes.88  Pure seeds of these two  diastereomers were obtained by fractional recrystallization from hexanes. By carefully seeding a hexane solution of the mixed diastereomers with both or either seed of Rp- and Sp-107 at about 5 °C, 7?D-107 formed large blocks up to a size of 6 x 4 x 2 mm and a  37  -Pd2  /OCHs  CH3OH/Py Hexanes  ^OCH3  O  CH 3l  ll  -P—OCH 3  100 °C  CH3  Ipcyccu o )) '/  O  Menthol  P—OMen | CH3  n  p—a  Py/Et 2 0  CH3  107 Fractional reciystallization and/or manual separation of crystals  1 P OMen  p*, lOMen i CH3  CH3  V107  V107  ArMgBr  ArMgBr  5-  /?-  CH3 Ar=  108  109 -  CH3  -<^)-CH3  ^  CH3 110  111 CH3  CH3  -o<  112 CH3  Figure 1.1. Preparation of Monophosphine Oxides.  38  113  weight of 45 mg, and .Sp-107 formed only small needles from the same solution. The crystals were filtered, and the mother liquor was further concentrated and seeded. This cycle was repeated several times. Manual separation of the crystals combined with fractional recrystallization gave a very good separation of the two diastereomers. The purity of the two diastereomers was further confirmed by ^H NMR spectroscopy and by the measurement of their optical rotations.89 Grignard displacement reactions of Rp- or Sp-107 yielded (R)- or (S)-108-lll and 113 respectively with inversion of configuration. The chemical and optical yields are shown in Table 1.1. Good chemical yields were obtained along with optical purities as high as 99%, except in the case of oxide 112, where a 24% chemical yield and a 42% e.e. was found (Table 1.1). Compound 110 was prepared from a mixture of Grignard reagents composed of 2,3-dimethylphenylmagnesium  bromide (30%) and 3,4-  dimethylphenylmagnesium bromide (70%) without any detectable product formation from the former.  Table 1.1. Chemical and Optical Yieldsa of Aryl Grignard Reaction with Phosphinate 107 reaction conditions compound  t (°C)  (5)-isomer  time (h) yield (%)  (/?)-isomer  e.e.%  yield (%)  e.e.% 88 81  108 109  80 80  16 5  93 65  >99 86  -  110° 111 112  80 80 95 80  16 3 70 3  58 72  >95 >95  64  -  -  24  86  -  -  113  -  -  94 >95 42 -  a  Optical purity was determined by chiral HPLC (Chiralpak OP, see Chapter 2 for details). "A  mixed Grignard reagent composed of 2,3-dimethylphenylmagnesium bromide (30%) and 3,4dimethylphenylmagnesium bromide (70%) was used in this reaction.  39  The displacement of the MenO group of the menthyl phosphinate (107) by the Grignard reagent is very sensitive to structural factors, and judicious matching of the starting phosphinates with the Grignard reagent is often a prerequisite for the overall success.°°a A sterically hindered o-methyl-substituted phenylmagnesium bromide can decrease both the chemical and optical yields because the methyl group is likely to retard sterically the nucleophilic attack of the Grignard reagent on the phosphorus center of the phosphinate. In this way, only compound 110 was observed and the other possible product did not form, when a mixed Grignard reagent prepared from a mixture of 3- and 4-bromo-o-xylenes was used. The non-observed product could be attributed to a much faster reaction of 3,4-dimethylphenylmagnesium bromide than 2,3-dimethylphenylmagnesium bromide, an o-methyl-substituted aryl Grignard reagent. Similarly, a low yield of compound 112 resulted from another o-methyl-substituted aryl Grignard reagent, 2,4,5-trimethylphenylmagnesium bromide, which required the most severe reaction conditions shown in Table 1.1. The Grignard displacement reaction was carried out in the aromatic solvents benzene or toluene.  Diethyl ether or tetrahydrofuran introduced in from Grignard reagents  decreased the rate of the reaction in benzene or toluene and thereby lowered the yield of the phosphine oxide. Careful replacement of ether with benzene or toluene was required for a faster reaction as well as for a higher yield, especially for the sterically hindered Grignard reagents. This may be attributed to the increased nucleophilicity of the poorly solvated nucleophile in benzene or toluene compared to the better solvated species in ether. A greater yield was obtained when the ether solution of the Grignard reagent was replaced by benzene or toluene before the phosphinate was added in. Freshly distilled anhydrous ether was preferred for preparing the Grignard reagents; and the undistilled anhydrous diethyl ether from Fisher Chemicals stopped the displacement reaction completely for unknown reasons.  40  Bisphosphine dioxides 114-117 were prepared through oxidative coupling of the corresponding monophosphine oxides 108-111 (Figure 1.2).90 Bases including w-butyllithium, sec-butyllithium and lithium diisopropylamide (LDA) were used to deprotonate the methyl group of the monophosphine oxides, but no obvious improvement for the coupling reaction was observed with the different bases.  \  O  )\-i!-Ar L i•*•»«*•-, a  7 V  *  /-—\  O  Ol  /fys-p-<"**~  X^ML,  L CH3  Ar =  ffl CH33  a  yCH3  117 CH3  Figure 1.2. Oxidative Coupling of Monophosphine Oxides.  Some of monophosphine oxides were not optically pure, but any presence of the opposite enantiomer in the monophosphine oxide would result in the formation of a meso bisphosphine dioxide along with the desired chiral bisphosphine dioxide.  A better  optical purity is predicted to be obtained in the dimeric product as the meso compound can be separated simply by recrystallization or chromatographic methods. The theory here was recently developed for the enrichment of optically active compounds."! Assuming that there is no discrimination for the coupling reaction among the enantiomers, the percent increase in optical purity, i.e. (e.e. % of the dimer) - (e.e. % of the monomer), would be:  41  % Increase = } ~ ^*%l^ 2 x (e.e. %) 1 + (e.e. %) where e.e. % is the optical purity of the monomer. For example, an 80% e.e. of monophosphine oxide will give a 97.6% e.e. of the corresponding chiral bisphosphine dioxide, a 17.6% enrichment according to the equation above. An enantioenriched sample of chiral compound 114 was resolved on a chiral HPLC column (Chiralcel OD, see Chapter 2 for details). The chromatographic analysis showed that a >95% optical yield was obtained for (R,R)-114 by coupling (R)-10S of 88% e.e., showing an enrichment of the optical purity during the coupling reaction. The results also showed that the racemization at the P-center of both monomer and dimer of the phosphine oxides did not occur during the deprotonating process and coupling reaction; otherwise a lower optical yield would have been obtained. The optical yields of other oxidative coupling products shown in Figure 1.2 were not determined but could be calculated from the optical purities of the corresponding monomers.  Since all the  monomers used had optical purities of >80%, a more than 95% e.e. was supposed to be afforded for the dimeric phosphine oxides and even higher after the recrystallization. One should be aware that the racemization at P-center of phosphine oxides does not occur in refluxing sodium hydroxide solution, but it does in concentrated hydrochloric acid over several days.92,93  1.2 Preparation of Substituted Anthracenes and Acetylene Derivatives Anthracene derivatives 1,5-dichloroanthracene (118) and dimethyl anthracene-1,5dicarboxylate (122), were synthesized from the commercially available 1,5-dichloro9,10-anthraquinone by following the literature procedures (Figure 1.3).94-96 Reduction of l,5-dichloro-9,10-anthraquinone by zinc powder in pyridine with a continuous  42  Zn/H*  » Pyridine  118  o  a CuCN  |  H^O  COOH  COOH Zn,NH3 •<  HOOC  121 I CH 3 CH,H f CH2OH  COOCH, IiAIHt  CH3OOC  122 Figure 1.3. Preparation of Substituted Anthracenes.  43  addition of acetic acid yielded 1,5-dichloroanthracene (118, Figure 1.3).  When 1,5-  Dichloro-9,10-anthraquinone was reacted with CuCN in 7V,Af-dimethylacetamide, compound 119 was obtained. 95 This cyano compound was hydrolyzed in acid and reduced by zinc in ammonia solution to give carboxylic acid 121. Compound 122 was synthesized by the esterification of acid 121 in methanol with a small amount of sulfuric acid as a catalyst.  Reduction of ester 122 with lithium aluminum hydride in  tetrahydrofuran afforded anthracene-1,5-dimethanol (123, Figure 1.3). Ethynediylbis(diphenylphosphine oxide) (125) was prepared according to Figure 1.4. Ethynyl-dimagnesium bromide, prepared by passing acetylene through a diethyl ether solution of ethylmagnesium bromide, was reacted with chlorodiphenylphosphine in ether to afford ethyndiylbis(diphenylphosphine). This phosphine was subsequently oxidized by hydrogen peroxide to its dioxide 125.97-98  D™  r^-r^ »„ n Ph 2 Paa 2 o  BrMg-C=C-MgBr —  ».  H202 p^p-c^c—ppj^  ft ^  124  i?  Ph 2 P-OsC-PPh 2  125  Figure 1.4. Preparation of Ethynediylbis(diphenylphosphine oxide) (125). Ethyl 3-(diphenylphosphine)-2-propynoate was synthesized for the first time via a nucleophilic substitution reaction between the conjugated base of ethyl 2-propynoate and chlorodiphenyphosphine at -78 °C. The product, which was not stable enough to store in the atmosphere, was oxidized with hydrogen peroxide to afford ethyl 3-(diphenylphosphine oxide)-2-propynoate (131, Figure 1.5).  EtC^C-C^C-H  1. Butyllithhim >• RC^C-CSC-PPh;, 2Ph2PCl  H^  ft *» EtCbC-CEC-PPh-, i 3 i  Figure 1.5. Preparation of Ethyl 3-(Diphenylphosphine oxide)-2-propynoate. 44  1.3 Synthesis of Dibenzobarrelene Derivatives The 9,10-dihydro-9,10-ethenoanthracene skeleton is shown in Figure 1.6 along with its numbering. The origin of this nomenclature stems from the anthracene ring system; however, this skeleton is known more commonly as "dibenzobarrelene". Throughout the thesis, the two names will be used interchangeably.  Figure 1.6. The Dibenzobarrelene Skeleton. The synthesis of dibenzobarrelene derivatives was first reported by Diels and Alder in 1931.^9  This short and productive method was used to synthesize substituted  dibenzobarrelene 102, from which compounds 103-106 were derived (Figure 1.7). The  CO2CH3 CQ2CH3  + CQ2CH3  /%°  oII IIo a—c-c-a 103  R= —CH2CH3 —CH(CH3)2  105 106  Figure 1.7. Preparation of Dibenzobarrelene Ester-Acids.  45  acid anhydride 104 is stable and could be stored at room temperature under anhydrous conditions, and this compound proved to be a good precursor for the preparation of compounds 105 and 106 in almost quantitative yields simply by alcoholysis. Bis(diphenylphosphinoyl)acetylene (125) is formally termed l,2-ethynediylbis(diphenylphosphine oxide). This compound has been used, along with other phosphoryl substituted acetylenes, as a dienophile in the Diels-Alder reaction shown in Figure 1.8.^0-101  A yield of around 60% for the phosphoryl or phosphinoyl  substituted benzene was obtained.  o II PPh 2  + PPh2 n O  r^Y°  -002 >•  k^°  n  o II PPh 2  PPh2 O  125 Figure 1.8. Dienophile of l,2-Ethynediylbis(diphenylphsophine oxide).  The diphenylphosphinoyl group Ph2P(0)- has been measured to be a moderately strong electron acceptor and is comparable in inductive effect with ester group. 102 On the basis of the electron-withdrawing property of the phosphinoyl group and the ester group, the acetylene derivative 125 would be expected to have a reactivity in the DielsAlder reaction more or less as that of dimethyl 2-butyne-l,4-dioate (Figure 1.7). On the other hand, the bulkiness of the phenyl rings on the two phosphinoyl groups would decrease the reactivity of phosphine oxide 125 towards the Diels-Alder reaction if steric factors arise as a major problem. In this thesis, acetylene derivative 125 was used for the first time to prepare Diels-Alder adducts with anthracene and some of its derivatives. Acetylene derivative 125 proved to be an excellent dienophile in the Diels-Alder reaction with anthracene and its derivatives. Adducts 126-129 were obtained in very  46  high yields when the reactions were run in a melt at 170-200 °C (Figure 1.9), and the isolated yields are included in Table 1.2. CWzo-dichlorobenzene (b.p. 179-180 °C) was  Table 1.2. Isolated Yields of Diels-Alder Adducts between Anthracene or Its Derivatives and l,2-Ethynediylbis(diphenylphosphine oxide) anthracene derivative used  adduct formed  anthracene 118 (1,5-dichloro-) 122 (1,5-diester) 129 (1,5-dimethanol) 9-anthrone a  yield (%) 87  126 127  47a  128 129 130  85 94 71  Low yield was due to a small scale reaction, and the actual yield for the reaction  may be higher as a nice crystalline material was afforded after reaction.  O  o  9  R  a2P  PPh II 2  1 PPh 2 ii O  +  r ^ ^ ^ ^ 1r^i 1  PPh2  vf  ^ r ^ \ ^ \-s^ R  125 R=H  a COJCHJ  CH2OH  =H  a  118 122 123  C02CH3 CH2OH  126 127 128 129  Figure 1.9. Diels-Alder Reactions between Anthracene or Its Derivatives and l,2-Ethynediylbis(diphenylphosphine oxide).  used as solvents in the case of the reaction with anthracene-1,5-dimethanol (123). Since the melting point of compound 123 is relatively high (234-236 °C) and the reactivity of its hydroxyl groups at this temperature may be substantial, heating compound 123  47  together with compound 125 failed to form adduct 129. outlined in Table 1.2 were run without any solvents.  The rest of the reactions  An incomplete reaction was  observed when anthracene was refluxed with compound 125 in o-dichlorobenzene overnight; in contrast, heating them together at 190 °C for 20-40 min gave an isolated yield as high as 87%. Although other phosphoryl substituted acetylenes such as phosphinates were not tested as dienophiles in the Diels-Alder reaction with anthracene and its derivatives, they should work in the same way as the oxide 125 or better, as shown by Kyba et al. with the diene of Figure 1.8.101 Potentially, an asymmetric Diels-Alder reaction could be accomplished by using a chiral acetylenic phosphinate derivative or a chiral anthracene derivative.  The adduct,  an ethenoanthracene,  can be further  reduced to  ethanoanthracene. Both ethenoanthracene and ethanoanthracene can be derivatized into phosphines, which are ligands for asymmetric catalysts. The preparation of such ligands will be described later on in this Chapter and the catalysts in Chapter 6. The anthrone shown in Figure 1.10 has been reported as a diene for Diels-Alder reactions in pyridine and DMF, since it can equilibrate with 9-anthrol, a reactive species (Figure 1.10). ^ ^ J^Q presence of a hydrogen bond acceptor (e.g. pyridine) shifts the equilibrium to the right and speeds up the reaction. Phosphine oxides are known as a strong hydrogen-bond acceptors and should shift the equilibrium to the 9-anthrol side when present.  In addition, the hydrogen bonded phosphinoyl group is expected to  OH  Figure 1.10. Equilibrium between 9-Anthrone and 9-Anthrol.  48  become a better electron-withdrawing group, making the alkyne more reactive towards dienes. The hydrogen bond between diene and dienophile also promotes their approach together for the reaction.  Despite the unfavorable steric interaction presented in  anthrone, a 9-substituted anthracene derivative, a good yield of adduct 130 was formed when the reaction was carried out in a melt of compound 125 and excess 9-anthrone (Figure 1.11). The effect of hydrogen bonding may be also demonstrated by the DielsAlder reaction between the hydrogen donor (CH2OH) diene of anthracene-1,5dimethanol (129) and the acceptor (P=0) of phosphine oxide 125, in which a milder reaction condition afforded the highest yield in Table 1.2. The reactivity of the diene and the dienophile could be enhanced by the hydrogen bonding between anthracene 129 and phosphine oxide 125 although the hydrogen bond is weaker than that between anthrol and the same oxide.  Figure 1.11. Preparation of 9-Hydroxyethenoanthracene 130.  Dienophile 131 bearing electron withdrawing ester and phosphinoyl groups was reacted with anthracene to form the adduct 132 shown in Figure 1.12. This alkyne has the advantage of having one less bulky group and an activation provided by phosphinoyl and ester groups. Upon hydrolysis of adduct 132 in a basic alcoholic/aqueous solution, carboxylic acid 133 was formed (Figure 1.12).  49  O II PPh 2 PPh 2  GQ2CH2CH3  131  Figure 1.12. Preparation and Hydrolysis of Diels-Alder Adduct 132.  We found that rraw-l,2-ethenediylbis(diphenylphosphine oxide) does not undergo a cycloaddition reaction with anthracene in high boiling point solvents such as odichlorobenzene as well as in the melt, although it has been used as a dienophile in the synthesis of Norphos dioxide. 104 However, l,2-ethynediylbis(diphenylphosphine oxide) was shown to be a reactive dienophile toward anthracene, anthrone and anthracene derivatives except 9-substituted anthracenes. The results can be explained by comparing the  structures  of  the  two  dienophiles.  The  linear  structure  of  ethynediylbis(diphenylphosphine oxide) 105 allows the two bulky diphenylphosphinoyl groups enough space away from the anthracene molecule during its approach to achieve the geometry that the reaction requires. In the case of /ra«5-l,2-ethenediylbis(diphenylphosphine oxide), the two large diphenylphosphinoyl groups are closer to the anthracene ring, and the steric effect is apparently sufficient to inhibit the approach of the dienophile.  Although l,2-ethynediylbis(diphenylphosphine oxide) turns out to be a  better dienophile, it does not react with 9-substituted anthracenes such as 9methoxyanthracene, a relatively reactive diene. These anthracenes own less room at the  50  9-, 10-positions to allow the dienophile approach and will have a greater steric interference between the anthracenes and the diphenylphosphinoyl groups.  1.4 Magnesium/Methanol Reduction Reaction Magnesium in methanol can reduce carbon-carbon double bonds conjugated to ester, amide, and nitrile substituents. 106-109 However, it has not been reported that a carboncarbon double bond "conjugated" to a phosphinoyl group could be reduced by magnesium in methanol. Ethenoanthracenes 126 and 129 were successfully reduced by magnesium in methanol to ethanoanthracenes 134 and 135 respectively at room temperature (Figure 1.13) in isolated yields as higher as 90%.  9  However, attempts to run the same reaction for  O  O PPh2  9 H  Ph 2 P R \  ^2  Mg/CH3OH  tzfzdt-^s* R  R  = H CH 2 0 H  PPh2  R=  126 129  H  134  CH2Ce  135  Figure 1.13. Reduction of Double Bond Conjugated to Phosphinoyl Group.  ethenoanthracenes 128 and 130 yielded a substance that showed a 31p NMR spectrum having a continuum of peaks in the region corresponding to phosphine oxide. This may indicate that a polymer or oligomer formed. Ethanoanthracene 134 could not be isolated from the final reduction mixture of ethenoanthracene 126 when a 3 M rather than 6 M HC1 solution was used for the work-up, and the 31p NMR spectrum showed that this  51  mixture probably contained mainly two dimeric diastereomers of the ethanoanthracene and that it also contained a small amount of ethanoanthracene 134 and other trace products. Dissolving metal reductions are known to occur via radical anion type mechanisms, and these species are stabilized by electron-withdrawing functional groups such as carbonyls.!06>ii0  j t has been accepted that the P = 0 double bond of pentavalent  phosphorus compounds may be represented by the resonance forms of P = 0 and P -CT.lll The interactions of phosphinoyl group with an adjacent anionic center have been found to be dominantly inductive and possibly to be moderately resonant. 111 We may conclude that phosphinoyl groups are not as good as carbonyl or cyano groups at stabilizing radical anions in terms of conjugative interaction. The electron-withdrawing property and probably together with certain degree of resonance interaction from diphenylphosphinoyl group will be part of the driving force for the reduction of its "conjugated" carbon-carbon double bond.  On the other hand, the very polar  phosphinoyl group may coordinate to the surface of the metal or the reactive metal cation clusters, and this could also be accounted for the successful reduction. The mechanism of the reduction can be proposed as shown in Figure 1.14, somewhat similar to Birch reduction. 106,110 The radicals in Figure 1.14 are reactive species and could combine together to give a dimer, oligomer and even polymer when the situation allows. For instance, molecules having other functional groups in certain positions, such as ethenoanthracenes 128 and 130, give no corresponding ethanoanthracenes when they subjected to magnesium/methanol reduction but probably a polymerization. Even in the case of ethenoanthracene 126, without any functional groups, improper workup procedures will not result in the expected product 134, but probably the dimers, as shown by 31p NMR spectroscopy. At this stage, we have not determined why this occurs. Nevertheless, almost quantitative yields could be obtained for compounds 134 and 135 under the conditions described in the experimental part. 52  135  R=CH2OH  Figure 1.14. Mechanism for the Reduction of a Phosphinoyl "Conjugated" Double Bond.  Reduction of double bonds "conjugated" to phosphinoyl groups with magnesium in methanol provides a chemoselective reduction method.  Like other dissolving metal  based reducing agents, this method gives a trans hydrogenation of the double bond. The trans conformation of ethanoanthracene 134 is confirmed by X-ray crystal structure analysis of [Rh(Anthraphos)(COD)]BF4.  The structure of ethanoanthracene 135 is  established as it shown in Figure 1.13 according to its ^H, 31p and l^C NMR spectra; the trans- conformation is obviously shown by its C2-symmetry-fit NMR spectra and also by the almost identical Jpjj and Jpp coupling constants to those of ethanoanthrance 134 (Table 1.3). The almost identical coupling constants also indicate that compound 135 is an anti isomer. Otherwise, a syn isomer should have different coupling constants from those of ethanoanthrace 134, as the steric interaction between the bulky diphenylphosphinoyl and methanol groups of the molecule causes the conformation of ethanoanthracene 135 to be different from that of ethanoanthracene 134. The most  53  similar conformations of these two ethanoanthracenes are also shown in Jp^ coupling constants whose values are almost identical in every case (see experimental part for details) and in the almost identical ABX pattern of l^C NMR spectrum that could be assigned to C4a and C8a of ethanoanthracene 135 or CSyn of ethanoanthracene 134. On the other hand, an anti ethanoanthracene 135 is the most possible reduction product, where this anti product is thermodynamic stable and kinetically favorable. According to the mechanism shown in Figure 4.14, the proton could transfer directly from the nearby methanol group within the molecule to give the anti isomer; and the anti isomer diminishes the steric interactions between the bulky diphenylphosphinoyl and methanol groups.  Table 1.3. NMR Spectroscopic Data of Ethanoanthracenes 134 and 135. lH NMR, 5 ppm (JPH, Hz)  a c  3 1 P NMR,  comp'd  bridgehead-H  ethano-H  5 ppm (JPP, Hz)  134a 135  4.40(6) 5.05 (6)b  3.90(17) 3.83(17.5)b  29.69 (6.0)d 31.15,(8.1),M 29.16c  The spectra were recorded in CDCI3; ^the spectra were recorded in —2% CD3OD in CDCI3;  The spectrum was recorded in DMSO-dg. ''The coupling constants of J PP were estimated from NMR spectra (see Appendix I for the calculation).  Ethanoanthracene 134 was resolved by dibenzoyltartaric acid (DBT), a reagent that has been used for the resolution of chiral phosphine oxides. 112-115 The resolution of ethanoanthracene 134 was quite efficient by using the above reagent; and precipitation of ethanoanthracene 134 by one equivalent of L-DBT afforded an almost optically pure diastereomeric solid of (J?,J?)-(-)-134-(L)-DBT and left a solution of 64% d.e. of (5,5)(+)-134-(L)-DBT. The e.e. was determined by 31p NMR spectroscopy (see Chapter 2  54  for details). In this way, 1.8 g of optically pure (-) and 1.5 g of optically pure (+)-134 were obtained from 4.3 g of the racemate. By following the well documented reduction procedure of phosphine oxides to phosphines, H2-115 phosphine oxide 134 was reduced with trichlorosilane in anhydrous benzene to phosphine 136 in 76% yield after recrystallization from a mixture of diethyl ether and pet ether (Figure 1.15). This phosphine, for which the name Anthraphos is proposed from its ligand family of Chiraphosl^  anj  Norphos,H7 j s air-stable as a  solid, but slowly oxidizes into its oxides in solution. The use of Anthraphos as a ligand for asymmetric catalysis, will be discussed in Chapter 6. O  tfH  Ph2  H  PPh2  0£  SiHO, »• Benzene  o>, 136  134  Figure 1.15. Preparation of Anthraphos.  55  ™fc  o  CHAPTER TWO ANALYSIS OF OPTICAL PURITY  Optical purity is defined as [a] ^ _ ^mixture of enantiomers ,e. % = — Mpure enantiomer  e e  or  ,™  The determination of optical purity can be performed by numerous analytical procedures. H° Polarimetry is usually used for the comparisons with literature data, but not considered sufficiently reliable if any optically active impurities may present. Modern methods for the separation and/or determination of enantiomers include: (1) derivatization using chiral reagents and analysis by HPLC, GLC, TLC, NMR, etc.;118 (2) chiral stationary phases for HPLC, GLC, TLC; 119 (3) nuclear magnetic resonance spectroscopy by using chiral shift reagents; 120 (4) enantiomer specific immunoassays. In order to monitor asymmetric syntheses in this thesis, two of the above methods were used: NMR chiral shift reagents and HPLC with chiral stationary phases.  2.1 NMR Chiral Shift Reagents NMR spectroscopic analyses can be performed with chiral lanthanide shift reagents 120 an( j chiral solvating reagents 121 or after derivatization with optically pure reagents. 11^ In this thesis, a lanthanide shift reagent tris[3-(heptafluoropropylhydroxylmethylene)-(+)-camphorato], europium (III) derivative, (+)-Eu(hfc)3, was used for the enantiomeric analyses of the dibenzosemibullvalene derivatives, and a new chiral complexing shift reagent, dibenzoyltartaric acid, was developed for the optical purity determinations of the phosphine oxides studied in Chapter 1 as well as in Chapter 4. 56  2.1.1 Optical Purity Determination of the Dibenzosemibullvalene Derivatives Complex (+)-Eu(hfc)3 has been used previously as a chiral shift reagent to determine the optical purities of dibenzosemibullvalene derivatives by *H NMR spectroscopy  in  our  laboratory.43,80b,81a,81c  Some  of  the  signals  of  dibenzosemibullvalene 138 and 139 were resolved by Eu(hfc)3 in CDCl3.81a,81c Further separation of the shifted peaks was achieved by using a mixed solvent of CDCI3 and CCI4 and/or by performing the measurement under anhydrous conditions. Figure 2.1 shows portions of the *H NMR spectra arising from the methyl ester groups of a mixture of the enantioenriched dibenzosemibullvalene 138 and 139.  CH3CH202C  .CQzCHs  138  CHjQzC  C02CH2CH3  139  —1  «.«5  1  1  1  1  «.B8  3.9S  5.98  3.85  1  1  1  1  3.38  3.'5  3.'8  3.6S  1 —  3.68  PPtt  Figure 2.1. *H NMR Spectra of the Methyl Ester Groups of a Mixture of Enantioenriched Regioisomers 138 and 139 in CDCI3. The spectra were recorded in the presence of (+)-Eu(hfc)3 (upper spectrum) and without (lower spectrum). Enantiomeric excesses: 138, 64.6% (-M4bS,8bS,8cS,8dS); 139, 17.8%. 57  The singlets at 5 3.86 and 3.71 ppm in the lower spectrum in Figure 2.1 were assigned to the methyl ester groups of compounds 138 and 139 respectively. Both singlets were resolved by (+)-Eu(hfc)3 in CDCI3 into an almost baseline separation (upper spectrum in Figure 2.1). The resolved two peaks at 5 3.95 and 3.93 ppm were assigned to (+)(4bR,8bR,8cR,8dK)-138 and (-)-(4bS,8bS,8cS,8dS)-13S respectively by using an authentic sample with a known enantioenriched enantiomer. For regioisomer 139, the assignment was not done due to lack of such an enantioenriched sample. The integration of the resolved peaks gave the enantiomeric excesses of both compounds.  The  enantiomeric excesses of dibenzosemibullvalene 138 and 139 were 64.6% and 17.8% respectively according to the upper spectrum in Figure 2.1. Similarly, the ^H NMR signals of the methyl ester groups of dibenzosemibullvalene 141 and 142 (Figure 2.2) were also resolved by the same shift reagent in the same way as those of compounds 138 and 139. The enantiomeric excesses of all the four compounds were determined by the lH NMR spectroscopy while the chiral shift reagent of (+)-Eu(hfc)3 was used in Chapter 3.  N02  97  HOOC  n  I  V  CH-O-Cn Ph-C-O-CH CDOH 98  O  R = P(0)Ph2 149  O PPh2  126 Rj=H 129 Ri=CH 2 OH 130 Ri=H  141 Ri=0O2Me R^CO^Pr* 142 R^CQzPri R2 = C02Me O O II  143 R!=PPh 2  R2=H R2=H R2 = 0]  Figure 2.2. Chiral Shift Reagents and Substrates. 58  134  II  R2 = PPh2  2.1.2. Dibenzoyltartaric Acid as a Chiral Complexing Shift Reagent for Phosphine Oxides Development of methods to determine the optical purity of phosphine oxides spectroscopically has attracted relatively little attention.  Small 31p and ^H NMR  chemical shift differences have been observed for chiral phosphine oxides when they are in the presence of an optically pure solvent. Thus Mislow and coworkers used (+)-lphenyl-2,2,2-trifluoroethanol (Pirkle's reagent) to determine the optical purity of tertbutylmethylphenylphosphine oxide. 122 Also Pirkle and coworkers^ observed small NMR peak separations of 1.4-3.2 Hz in the other chiral phosphine oxides by using the same solvents. More recently, Moriyama and Bentrude^4 have demonstrated that terf-butylphenylthiophosphinic acid gives better separation (3.0-10.6 Hz at 90 MHz) of enantiomeric groups in the ^H NMR and suggested that this compound may be generally useful for this purpose. Two groups have used the chiral shift reagents (R)-(-)-N-3,5dinitrobenzyl-1-phenylethylamine and (+)- or (-)-l-phenylethylamino-3,5-dinitrobenzamide (97, Figure 2.2) to determine optical purity by proton NMR on the basis of chemical shift differences, usually of a methyl group. 125 A complete resolution for several phosphine oxides was obtained by ^H NMR spectroscopy at 400 MHz when shift reagent 97 was used, and a partial resolution was also observed for other phosphine oxides, depending on the structure of the compounds being investigated. 125a Chiral shift reagents, such as 97, can be considered as complexing agents that interact with phosphine oxides through hydrogen bonds between the acidic hydrogen atoms of the substrates and the oxygen atoms of the phosphoryl groups. Dibenzoyltartaric acid (DBT, 98 in Figure 2.2) has been used as a resolving reagent for chiral phosphine oxides based on the formation of diastereomeric hydrogen bonded complexes between chiral phosphine oxides and (+)-£>- or (-)-L-DBT. 112-115 Qn the basis of this strong hydrogen bond, we decided to investigate the use of DBT as a chiral  59  complexing 31p NMR shift reagent for phosphine oxides. As hydrogen bonds between the carboxylic acid groups of DBT and the phosphine oxide, DBT could cause some signal shifts in the 31p NMR spectrum of the phosphine oxide, especially of the bisphosphine dioxide in view of possible double hydrogen bonding. In principle, either enantiomer of DBT may give a peak separation of 31p NMR spectra of the phosphine oxides. 31  DBT was found to be a very effective  P NMR shift reagent for some bisphosphine  dioxides. Figures 2.3 to 2.6 show typical examples of spectra before and after adding  cJ  •"ni'm^  R  J  ^"•"-•••IJ—.  V»-H*-»A.  H  R=PCO)Ph2 149  JAm^^ —-,  11.0  '  •' 'I  42.0  I  40.0  <~*  ^_^JvL,  1  |  I  |  i  1  38.0  36. 0  34.0  32.0  30.0  28.0  >——r—  2 6 . fl  Figure 2.3. 3 1 P NMR Spectra (81 MHz) of Racemic Phosphine Oxide 149 in CDCI3. The lower spectrum was recorded on compound 149 only, and the upper spectrum was recorded in the presence of one equivalent of D-DBT.  60  **--*«*^l<Hi**J-*K*'  >>TI\>. *Y>* « i'i»*»'iN><t<K^>^^^nii' ^^w m>i* i M i l y ^ ^ ^ ^ i i * ^ . ^ ^ , ! ^ S***" m # . " W V ^ ^ i < * »  134 _w—' 1  i •  36.0  1  i •  35.0  34.0  1 I ' 33.0  —i—i—1—r—i  32.0 PPH  31.0  r-  -r—i—r—r I  30.0  29.0  '  '  '  ' I  28.0  Figure 2.4. 3 1 P NMR Spectra (81 MHz) of Racemic Phosphine Oxide 134 in 3:1 CDC^/o-dichlorobenzene. The spectra were recorded in the presence of one equivalent of D-DBT (upper spectrum) and in the absence of D-DBT (lower spectrum). D-DBT, and Table 2.1 summarizes the results by using the racemic phosphine oxides. Deuterated chloroform as the solvent gave the best resolution.  Nitrobenzene or o-  dichlorobenzene could be added without loss of resolution when solubility was too low to carry out the measurement. As in the case of ethanoanthracene 134 (Figure 2.2), odichlorobenzene was used to increase the solubility of the two diastereomers that presumably form. Either enantiomer of DBT could be used for the measurements. As commercially available L-DBT comes as a water adduct (I-DBT-H20), and the presence of water lowers the resolution, the use of D-DBT is preferred.  61  OfeOH CH2OH  ^t**^***^****^^ 34.0  33.0  32.0  31.0  30.0  29.0  5 (ppm) Figure 2.5. 3 1 P NMR Spectrum (81 MHz) of Racemic Phosphine Oxide 129 in CDCI3. The spectrum were recorded in the presence of 1.5 equivalents of Z)-DBT.  Table 2.1. Resolution of 31p NMR Signals (81 MHz) of Racemic Phosphine Oxides in the Presence of £>-DBTa P  ?,  resolved signals (8 in ppm)  resolution  oxide  143[  (+)-143: 32.33 (d, J PP = 11.5 Hz), 30.45 (d, J PP = 11.5 Hz)e (-)-143: 32.65 (d, J PP = 11.4 Hz), 31.33 (d, J PP = 11.4 Hz)  0.32 ppm (25.7 Hz) 0.88 ppm (70.9 Hz) respectively  149c  37.38: (d, J PP = 24.5 Hz), 32.33 (d, J PP = 24.5 Hz). the other enantiomer: 37.00 (d, J PP = 24.4 Hz), 32.57 (d, J PP = 24.4 Hz)  0.38 ppm (30.9 Hz) 0.24 ppm (19.4 Hz) respectively  134°  R,R-(-)-134: 34.61 S,S-(+)-134: 34.37  0.24 ppm (19.4 Hz)  129c  31.86 the other enantiomer: 31.73  0.13 ppm (10.0 Hz)  ^DCly was used as the solvent except for compound 134, for which 3:1 CDC^/o-dichlorobenzene was used. ''Phosphine oxide (0.03-0.1 M) and one equivalent of D-DBT were used; c 1.5 equivalents of DDBT were used; dThe correlations of the chemical shifts and the absolute configurations were done by using a sample with an enantioenriched isomer of known absolute configuration; e(->(4hS',8bS',8c.S,,8dS> and (+)-(4b/?,8b/?,8c/?,8d#)-143 were obtained.  62  v_  u 143  1  i — i — i — i —1i — i — i — i — i — i — i — i — i — i — i — i — i — i — r -  34.0  33.0  32.0  i  31.0 PPM  i—i—i—p  30.0  -T  1  1  1  29.0  1-1  1  1  T-  28.0  Figure 2.6 3 1 P NMR Spectra (81 MHz) of Phosphine Oxide 143 in CDCI3. The spectra were recorded on a racemic mixture (bottom spectrum), a sample of 10% e.e. (4bR,$bR,8cR,SdR) in the presence of 0.5 (spectrum 1), 1.0 (spectrum 2), 1.5 (spectrum 3) and 2.0 (spectrum 4) equivalents of D-DBT.  63  As shown in Figure 2.6 and in Table 2.1, a resolution of up to 0.88 ppm (71 Hz at 81 MHz) was obtained for compound 143 (spectrum 2 in Figure 2.6). As far as we know, this is the highest resolution ever achieved by a chiral shift reagent on a 31p NMR spectrum. Well resolved peaks were also observed for compounds 149 (Figure 2.3) and 134 (Figure 2.4). Although one equivalent of D-DBT gave a partial resolution to compound 129, an almost complete resolution was afforded when 1.5 equivalents of D-DBT were used as shown in Table 2.1. The resolved spectrum of compound 129 is shown in Figure 2.5; compound 129 gives only a singlet at 5 30.39 before resolution. These results show that DBT is a powerful shift reagent for the bisphosphine dioxides. As an example, phosphine oxide 143 (Figure 2.2) was studied for further details. Figure 2.6 shows a series of 31p NMR spectra in the presence of different amounts of D-DBT. A sharp shift in the spectra takes place between zero and one equivalent of DDBT (spectra 1 and 2). As the amount of D-DBT increases, the shift increases smoothly (spectra 3 and 4) and stays at about 4 ppm from the original two doublets when two equivalents of D-DBT are added (spectrum 4). The maximum resolution is reached at one equivalent of D-DBT, and both increasing and decreasing the amount of D-DBT lowers the resolution, which is demonstrated by the plots in Figure 2.7. The changes caused by this 1:1 molar ratio of substrate to shift reagent indicate that they most likely form a complex with 1:1 stoichiometry. A larger resolution is observed for the nonbenzylic phosphorus (upfield group of peaks) than for the benzylic one (downfield group). Attempts to apply the same shift reagent to some of the bisphosphine dioxides of Figure 1.2 as well as the monophosphine oxides of Figure 1.1 failed to give any resolutions to 31p NMR and ^H NMR spectra. Molecular flexibility may reduce the ability for a bisphosphine dioxide to form a hydrogen bonded complex with shift reagent DBT.  On the other hand, a mono(phosphine oxide), as a whole, would have less  interaction with the acid by means of its monodentate feature. 64  In contrast, all the  phosphine oxides listed in Table 2.1 are rigid bidentate molecules, and they can apparently interact with DBT more effectively by hydrogen bonding than the flexible bisphosphine dioxides and monophosphine oxides.  -  ^/•v  0.80 -  A5 (ppm)  0.60 0.40  ^^-®  1  0.20 : 0.00 <f<  /  ~^-  /  , , i  0.00  i  0.40  i  i  i  i  i  0.80  i  i  1.20  i i  I  i  i  1.60  ^ i  2 I  i  2.00  Equivalents of D-DBT  Figure 2.7. Resolution (A8 = 8. - 8+) of 3 1 P NMR Spectra of Phosphine Oxide 143 as a Function of Equivalents of D-DBT. Plot 1 (non-benzylic-P) and Plot 2 (benzylic-P) are corresponding to the two doublets of the compound.  2.2 ChiralHPLC Analysis of Phosphine Oxides Virtually complete enantiomeric separations have been obtained chromatographically for phosphine oxides using a column with a bonded stationary phase of 3,5(N02)2C6H4CO-NHCHPhCONH(CH2)3,126 with normal stationary phases on a cyclodextrin columnl27  o r us j n g  pirkle's chiral stationary phase. 128  \ range of  organophosphorus compounds can be resolved on a column of optically active poly(tritylmethacrylate). 129  65  Two chiral columns have been used in this thesis, a CHIRALCEL OD and a CHIRALPAK OP(+).  The packing material for the CHIRALCEL OD is cellulose  tri(3,5-dimethylphenyl carbamate) on a 10 urn silica gel substrate, which is shown in Figure 2.8. The packing composition of the CHIRALPAK OP(+) is poly(diphenyl-2pyridylmethyl methacrylate) coated on a 10 urn silica gel substrate as depicted in Figure 2.9.  R  Figure 2.8. Packing Composition of the CHIRALCEL OD Column.  NR  \  /  Figure 2.9. Packing Composition of the CHIRALPAK OP(+) Column.  66  2.2.1 Optical Purity Determination on the CHIRAPAK OP(+) Column The CHIRALPAK OP(+) column, which has poly(methacrylate) as its stationary phase, is known for its ability to separate aromatic enantiomers in the viewpoint of its producer.  With their aromatic features, monophosphine oxides 108-112 (structures in  Figure 1.1) and bisphosphine dioxide 143 (structure in Figure 2.2) were resolved on this column. The chromatographic results are shown in Table 2.2 with the chromatographic resolution, Rs, which is calculated based on the equation: 130  Table 2.2. Chromatographic Data for Phosphine Oxides on the CHIRALPAK OP ( + ) Column (25 cm X 0.46 cm I.D.) at Room Temperature  comp'd  a  a so]vents  HPLC conditions fl ow rate detector (ml /min) UV (nm)  108  MeOH  0.45  254  109  93:7 MeOH/H 2 0  0.35  110  MeOH  0.45  254  111  MeOH  0.45  112  90:10 MeOH/H 2 0  143  93:7 MeOH/H2Q  retention time (min) b  Rs  ($)-(-): 8.9 2.8 (*)-(+): H.O  comments well separated  1.4  nearly baseline separation  (S)-(-): 8.8 (*)-(+): 9.5  1.1  nearly baseline separation  254  OS)-(-): 8.9 (/?)-(+): 9.5  1.3  nearly baseline separation  0.35  230, 254  14.6, 17.2C  2.7  well separated  0.35  230, 254 (-): 31.8 d (+): 40.0  1.3  nearly baseline separation  230, 254 (5)-(+): 13.5 (*)-(-): 14.5  The structures of these compounds are shown in Figure 1.1 (108-112) and Figure 2.2 (143). "The  correlations of the retention times with the absolute configurations were obtained by using only one optical enantiomer of known absolute configuration or a sample with an enantioenriched isomer of known absolute configuration. cThe correlation of the retention time with the absolute configuration was not done; ^(-)(4b.S',8bS',8GS,,8dS> and (+)-(4bfl,8bfl,8cK,8dK)-143 were obtained.  67  __ 2(t 2 -ti)  R S  twi + tw2  where: t^ and t 2 are the retention times of the two components; t w i and t w 2 are the band widths of the two components. The correlation of the retention times with the absolute configurations was obtained by using only one optical enantiomer of known absolute configuration or a sample with an enantioenriched isomer of known absolute configuration. For phosphine oxides 108-112, sharp peaks were observed and resolutions of Rs > 1.0 were obtained.  The peak broadening and tailing of bisphosphine dioxide 143  decreased its chromatographic resolution, but a nearly baseline separation with a resolution of 1.3 was still obtained (last two peaks of Figure 2.10). The resolutions given by the CHIRALPAK OP (+) column to the compounds in Table 2.2 were enough to undertake the measurements of their optical purities. This chromatographic method was used in Chapter 1 as well as in Chapter 4 to monitor the asymmetric syntheses. 2.2.2 Optical Purity Determination on the CHIRALCEL OD Column Bisphosphine dioxide 114 (Figure 1.2) was resolved on the CHIRALCEL OD column (Table 2.3). However, attempts to resolve other bisphosphine dioxides such as 117 (Figure 1.2) on this HPLC column were not successful, although a chromatographic resolution of 1.2 was observed for compound 114 (Table 2.3). One important feature suggested by the producer of this column is its ability for resolution of compounds with aromatic, carbonyl, nitro, cyano or hydroxyl groups. yV-Acetylphenylalanine methyl ester (157), which bears a carbonyl and a phenyl group, was demonstrated to be well resolved on this column (Rs = 1.8, Table 2.3). This chromatographic method was used to measure the enantiomeric excesses of A'-acetylphenylalanine methyl ester. This ester was prepared from product A-acetylphenylalanine (156) from the catalytic asymmetric  68  hydrogenation described in Chapter 6.  In this way, the enantiomeric excess of the  hydrogenation product 156 was obtained.  Table 2.3. Chromatographic Results with the CHIRALCEL OD Column (25 cm x 0.46 cm I. D.) at Room Temperature HPLC conditions3 detector soivents UV (nm)  comp'd 114  97:3 231 hexane/EtOH  157 a  9:lhexane/ /-PrOH  254,230  retention time (min)b  Rs  comments  R,R-(-): U.2 S,S-(+): 17.3  1.2  nearly baseline separation  #-(-): 11.4 S-(+): 13.4  1.8  well separated  A flow rate of 1.0 ml/min was used. bThe correlations of the retention times with the absolute  configurations were obtained by using only one optical enantiomer of known absolute configuration or a sample with an enantioenriched isomer of known absolute configuration.  2.3 Measurement of the Optical Purities of the Crude Photolysed Samples All methods described in this chapter were developed to measure the optical purities of the specific compounds used in this thesis. The compounds, usually from photolysis, were present as a mixture, and direct analysis of the optical purities without separation was preferred.  Chiral HPLC may separate impurities as well as resolve the chiral  compound(s), and 31p NMR usually gives a simple spectrum from only a few 31p nuclei on an organic molecule. These two methods have the advantage that chemically pure samples may not be required for the measurement of optical purity. Following are examples where the measurements of optical purity were performed on the crude samples directly.  69  As shown in Figure 2.10, a HPLC chromatogram of a photolysed sample of compound 126 (t = 26.82 min) on the Chiracel OP(+) column indicates that the starting material as well as the two enantiomers of photoproduct 143 [(-)-(4b1S,,8b.S',8clS',8(LS), t = 32.06; (+)-(4bR,SbR,ScR,SdR), t = 40.96 min] are all separated. The analysis of the conversion of a photolysis and the optical purity of photoproduct 143 could be done on the same chromatogram.  This chromatographic method was used to monitor the  o CO <N  o.oo  ^JL l.OO  3.OO  4.OO  S.OO  6.00  Retention Time (10 min) Figure 2.10. Chromatogram of a Photolysed Sample of Ethenoanthracene 126 on a Chiracel OP(+) Column (25 cm x 0.46 cm I.D.). Eluent, 93:7 methanol/water; flow rate, 0.35 ml/min; detector, UV at 230 nm; room temperature. Compound 126, t = 26.82. (-)-(4b£8b,S',8c£8(LS>143, t = 32.06; (+H4bR,SbR,ScR,SdR)-143, t = 40.96; 26.8% e.e. (4b<S,,8b£8GS,,8dS).  70  photoreaction of ethenoanthracene 126 as described in Chapter 4. conversion was calibrated by *H and  31  The extent of  P NMR spectroscopy due to the different  sensitivity of these two compounds to the UV detector. Remarkably,  31  P NMR spectroscopy can also measure the enantiomeric excess of  photoproduct 143 in the presence of ethenoanthracene 126 by using the chiral complexing shift reagent D-DBT (or L-DBT). As shown in the spectra in Figure 2.11,  M  / U  I  34.0  33.0  I  r  32.0  -i—i—i—r  31.0 PPM  30.0  1  I ' 29.0  1  I ' 28.0  Figure 2.11. 3 1 P NMR Spectra of a Sample from the Photolysis of Ethenoanthracene 126. The sample contained both 126 and photoproduct 143. The spectra were recorded in the presence of one (spectrum 1) and two (spectrum 2) equivalents of D-DBT based on the moles of (126 + 143) as well as without D-DBT (bottom spectrum, here only a singlet observed for 143). Photolysis conversion, 45%; enantiomeric excess of 143, 67% e.e. (4b6',8b5,8c5',8dS). 71  two unresolved peaks corresponding to the two compounds were observed without DDBT (bottom spectrum).  The peak at lower chemical shift can be assigned to  ethenoanthracene 126 and the other one to photoproduct 143, according to their relative ratio obtained from ^H NMR spectroscopy. When one equivalent of D-DBT was added, two sets of AB type spectra were obtained (spectrum 1): set 1, 8 30.7, 30.8, 32.5 and 32.6 ppm, which is assigned to minor enantiomer (+)-(4bR,8bR,$cR,8dR)-143; set 2, 8 31.5, 31.6, 32.8 and 32.9 ppm, which is assigned to major enantiomer (-)-(4bS,SbS,ScS,SdS)-143. The oversized peak of ethanoanthracene 126 (8 32.4 ppm) overlaps with one of the peaks from the major enantiomer (-)-(4b»S',8b5',8c1S,8(iS)-143. Further isolation of the overlapping peaks was achieved when two equivalents of D-DBT was added (spectrum 2). A baseline separation of the group of peaks between 31 and 32 ppm of the nonbenzylic-P still remained although the other group of peaks that belong to the benzylic-P was not well resolved; a lower resolution for photoproduct 143 was observed. Both spectra 1 and 2 were integrated to give the conversion of the photolysis of ethenoanthracene 126 and the optical yield of photoproduct 143, which were 45% and 67% (-)-(4bS,8bS,8cS,8(LS) respectively. The 31p NMR spectroscopic method serves as a convenient alternative to the chiral HPLC method.  For example, a photolysis sample of ethenoanthracene 126 was  determined by both the HPLC method and the 31p NMR spectroscopy; this sample gave a 30% photolysis conversion and 97% e.e. [(-)-(4b6',8b5',8c1S',8dS)] for photoproduct 143 by the former method while it gave a 33% conversion and >95% e.e. [(-)(4b.S',8b5',8c6',8diS)] by the latter method, where the minor enantiomer of photoproduct 143 was not detected within the spectral limit upon adding chiral shift reagent D-DBT. The HPLC method was used to analyze the photolyses of ethenoanthracene 126 in Chapter 4, as it required much less sample than 31p NMR spectroscopic method. Meanwhile, 31p NMR spectroscopy was used occasionally when an adequate quantity of photolysis samples were available. 72  A sample from the photolysis of ethenoanthracene 130 was shown by  31  P NMR  spectroscopy (lower spectrum of Figure 2.12) to contain starting material 130 [5 27.4 (d), 41.6 (d)], major photoproduct 149 [8 29.2 (d), 34.5 (d)] and other unknown minor photoproducts [8 25.7 (d), 30.9 (d), 28.1, 31.5, 37.6 (d)].  'LJ^  L-L. 42.0  " 'I ' ' ' 40.0  L  JV • • • I ' ' ' •  38. 0  36.0  '"I 34.0  J  149 R = P(0)Pb2  j  M^wwb^tcy^y  L~A/ Lu***JK.  -M*J  I '' ' 32.0 PPM  The spectroscopic  • • • I • • •  30.0  26.0  24.0  Figure 2.12. 31p N M R Spectra of a Photolysed Sample from the Enantiomorphously Pure Complex of 130/wo-propanol (1:1). The sample contained compound 130, photoproduct 149 and other minor photoproducts (lower spectrum). The upper spectrum was recorded in the presence of one equivalent of D-DBT (calculated from the amount of the mixture and the molecular weight of compound 130). Enantiomer excess of photoproduct 130, 64.2%.  73  measurement of the enantiomeric excess of photoproduct 149 was performed by using the shift reagent D-DBT (upper spectrum of Figure 2.8). Either doublet at 5 29.2 (d) or 34.5 (d) that belongs to photoproduct 149 is well resolved into two sets of doublet centered at 8 32.8 or 37.5 ppm respectively. Either set consists of a major and a minor doublets, i.e., four peaks, among which the two outside peaks are well isolated and the two inside peaks overlaps with each other (upper spectrum of Figure 2.8).  The  integrations (l\ and I2) of the two outside peaks gives the relative amounts of the two enantiomers of photoproduct 149, and the integration (I m ) of the overlapping middle peaks gives the total relative amount of both enantiomers. The enantiomeric excesses are obtained independently from the two groups of peaks according to the equation:  e.e.*.JiL±L Im  where: I m is the integral of the overlapping peaks; l\ and I2 are the integrals of the two peaks outside the overlapping peak. According to the upper spectrum of Figure 2.12, the enantiomeric excesses are 63.9% and 64.3% from the group of peaks at 8 32.8 and 37.5 ppm respectively. This method was used in Chapter 4 to monitor the photoreaction of compound 130.  74  CHAPTER THREE COCRYSTALLINE COMPLEXES AND THEIR PHOTOCHEMISTRY  A cocrystalline complex provides an ordered medium for the study of solid state organic photochemistry.  The cocrystallization process takes its advantage from  introducing a reactive molecule into different crystalline media of complexes that do not change the chemical structure of the molecule being investigated.  Such molecular  complexes have been employed for organic solid state photochemical studies by Toda et al.46 They cocrystallized photoreactive guest molecules with different resolved host molecules, and subsequently demonstrated that solid state photolyses led to an asymmetric induction in the guest photoproduct(s), while the host molecules were not photoreactive and only supplied a chiral environment for the guests to react. Most of the hosts studied by Toda's group were hydrogen-bond donors and the guestreactants were usually acceptors.46 In this chapter, the carboxylic acid derivatives of dibenzobarrelene were used as the photoreactive guests. These dibenzobarrelenes have been shown to undergo di-7i-methane rearrangement in the solid state^0,81  an( j w e r e  used as donor-guests. Their solid state photochemistry in the presence of phosphine oxides as hydrogen bond acceptor hosts is described in this Chapter.  3.1 Cocrystalline Complexes of Phosphine Oxides Triphenylphosphine oxide (TPPO, Figure 3.1), a good hydrogen bond acceptor, is known to form high quality crystals when co-crystallized with a variety of hydrogen bond donors, a procedure that was promoted as a crystallization aid for compounds that do not crystallize well on their own. 131 These donor molecules include carboxylic 75  acids, 1^2 amides, 133 imides,134 hydroperoxides 135 an( ] phenols. 136 Other phosphine oxides are also known to form hydrogen bonded complexes with donor molecules in solution, 13' but such organic cocrystalline complexes have not been prepared as far as we know. Phosphine oxides are potential host-acceptor molecules for the photochemical study of guest donors in the solid state. Initially, triphenylphosphine oxide was examined as a host-acceptor for the preparation of hydrogen-bonded complexes with dibenzobarrelenes 105 and 106 (Figure 3.1). Crystalline complexes of this phosphine oxide with dibenzobarrelenes 105 and 106 were obtained from hydrocarbon solvents benzene, toluene, xylenes and cyclohexane. In the case of dibenzobarrelene 105, two kinds of crystals were prepared, one with solvent molecules of benzene or toluene that were included inside the cocrystals and the other without. All these hydrogen bonded complexes have sharp melting points  OH: ^ Ko> vr=-^^CH3 114 CH3  116  & 115 CH3  -0 117  <^d  CH3 R= —CH2CH3  \ 0 ) vppBj)  —CH(CH3)2  101 Figure 3.1. Phosphine Oxides and ^-Donors.  76  105 106  (Table 3.1), which indicate the formation of the complexes.  They were also  characterized by elemental analysis, IR, *H NMR spectroscopy and, in some cases, by FAB (fast atom bombardment) mass spectrometry.  Table 3.1. Properties of Triphenylphosphine Oxide (TPPO) Complexes complex  composition  m.p. (°C)  crystallization solvent(s)  105t 106t 105tt 105tb  TPPO/105(1:1) TPPO/106(1:1) TPPO/105/toIuene (2:2:1) TPPO/105/benzene (2:2:1)  137-139 148-149 77-79(d) 79-80(d)  ^-xylene, mixed xylenes mixed xylenes, cyclohexane toluene benzene  After the successful preparation of triphenylphosphine oxide complexes with dibenzobarrelenes 105 and 106, the bisphosphine dioxides were also investigated for the same purpose.  Like triphenylphosphine oxide, l,2-ethanediylbis(diphenylphosphine  oxide) (101) forms hydrogen bonded complexes with dibenzobarrelenes 105, 106 and pnitrophenol in a stoichiometry of 1:2, and so do its optically active analogs 114-117 (Table 3.2). The optically pure species create enantiomorphously pure crystals with dibenzobarrelenes 105, 106 and /?-nitrophenol (Table 3.2).  The complexes of  bisphosphine dioxides cocrystallize more effectively with their counterpart to give high quality crystals from mixed xylenes than from toluene or benzene.  In one case, a  mixture of cyclohexane and benzene was also used as the cocrystallization solvent (last entry in Table 3.2).  All of the crystalline complexes were suitable for X-ray  crystallographic analysis except for complex 105xm, whose crystals were too small for analysis. The sharp melting points of the bisphosphine dioxide complexes shown in Table 3.2 indicate the formation of the cocrystals.  These complexes were fully  characterized by elemental analysis, IR and ^H NMR spectroscopy as well as, in some cases, X-ray crystal structure analyses. 77  Table 3.2. Properties of Bisphosphine Dioxide Complexesa complex  //-donors  oxides^  m.p.(°C)  crystallization solvents  105d (5,S)-105p  105 105  101 (S,S)-U4  (S,S)-l05m (S,S)-l05xo (R,R)-105xm (S,5)-105xm 106d (S,S)-l06p (S,S)-l06xo  105 105 105 105 106 106 106  (S,S)-115 (5,5)116 (R,R)-117 (S,S)-117  toluene, mixed xylenes toluene toluene toluene or xylenes xylenes or toluene xylenes or toluene xylenes toluene  (S,S)-U6  209-211 191-193 189-190 173-174 158-159 158-159 179-181 185-187 173-174  /7-nitrophenol /7-nitrophenol  101 (R,R)-117  144-145 125-126  toluene benzene/cyclohexane  lOln (R,R)-117n a  101 (S,S)-1U  toluene  The stoichiometry of these complexes is 1:2 bisphosphine dioxide/substrate. "The structures of the  oxides are shown in Figure 3.1.  3.1.1 FTTR Spectra The stoichiometry obtained for the complexes of triphenylphosphine oxide as well as the bisphosphine dioxides indicates that the phosphine oxides are most likely to coordinate their phosphinoyl group(s) with the carboxylic acid group of the dibenzobarrelenes or with the hydroxyl group of /?-nitrophenol. The hydrogen bonded structures of the complexes were confirmed by their FTIR spectra according to the position of the phosphoryl stretching band v(P=0).  In general, v(P=0) is easily  located in the region 1300-1150 cm~l, which is related directly to the sum of the electronegativities of the phosphorus substituents.138  Hydrogen bonding to the  phosphoryl group causes a significant red-shift of the stretching frequency. 138 This frequency shift is usually accompanied by a broadening and intensification of the band. In the case of bisphosphine dioxides, there are two peaks corresponding to v(P=0), and hydrogen bonding also causes the peaks to coalesce in addition to shifting them to lower  78  frequency. 1^8 The v (p=o) bands of the phosphine oxides and their complexes are given in Table 3.3. All the frequency changes upon the formation of hydrogen bonds are found in the complexes of both triphenylphosphine oxide and bisphosphine dioxides. Red-shifts up to 20 cm~l are observed in the case of the triphenylphosphine oxide complexes. For bisphosphine dioxide complexes, both peaks are red-shifted, and the two peaks are merged, as only one peak can be located. Frequency changes can also be seen in the carbonyl stretches of the dibenzobarrellene derivatives and even from the stretches of their conjugated carbon-carbon double bonds in some cases (see Experimental part for details).  Table 3.3. Stretching Bands v(P=0) (cm~l) of Phosphine Oxides only and Those of the Complexes with Dibenzobarrelenes 105 and 106 as well as pNitrophenol phosphorus compound  phosphine oxides only  TPPO TPPO  1191 1191 1191  TPPO 101 114 115 116 117  1176, 1188 1176, 1187 1183, 1198 1174, 1187 1180, 1195  comiplexes with 77-donors 105  106  1180 1174 (bz.)  1171  -  -  -  1174 (tol.) 1158 1154  -  -  1151 1154  1163  -  -  1149  -  1155 1150 1154  -  /7-nitrophenol  -  1153  3.1.2 Mass Spectra The hydrogen bonded complexes can survive fast atom bombardment ionization to give the mass of the complexes. Table 3.4 shows the complex ion peaks of some typical examples. Electron impact is a destructive ionization method and will normally break  79  down the hydrogen bonds. As an example, complex 105d was ionized at 70 eV and 250 °C and a mass spectrum showed masses centered at 705, 629 and 553 in addition to the masses arising from the two components, dibenzobarrelene 105 (M+ at 320) and phosphine oxide 101 (M + at 430). As expected, no complex ion m/e at 1070 was observed for the 2:1 105/101 complex.  The masses at 705, 629 and 553 can be  explained by benzyne and/or ethoxy fragmentations of the complex as shown in Figure 3.2. The ester group is apparently the most sensitive ionization target within the complex, and it loses the ester group to give a mass at 705. Two consecutive losses of benzyne (C6H4) give the masses at 629 and 553.  Table 3.4. Molecular Ions of FAB Mass Spectra of Phosphine Oxide Complexes complex TPPO/105/benzene (2:2:1) (105tt) TPPO/105(l:l)(105t) TPPO/106(l:l)(106t) DPPEO/105(1:2) a  F.W.  molecular ions of FABMS(MH) +  598a  598 612  599 599 613  1070  1071  Benzene molecules are included in the crystal lattice, and only the complex of TPPO/105 (1:1)  shows up as a complex ion in the spectrum.  Ph  Ph  ^+ (553)  OO2H- - •O=P(CH 2 ) 2 P=0 Ph  0^4(76)  Ph (629) -QH4 (76) Ph  -OCH2CH3 (45)  Ph  (X) 2 H--0=P(CH2)2P = 0 Ph Ph  m  (705)  Figure 3.2. EI Fragmentation of Complex 105d (2:1 105/DPPEO). 80  3.1.3 X-ray Crystal Structure Analysis The molecular and crystal structures of seven of the complexes that are listed in Table 3.1 and Table 3.2, lOln, (R,R)-U7n, 105t, 105tt, (S,S)-105p, (£,S)-105m and (S,S)-105xo were analyzed by X-ray crystallography.  Their molecular structures are  shown in Figures 3.3 and 3.4 as well as in Figures 3.14-3.18. As shown by X-ray crystal structure analysis, the hydrogen bonded complexes indeed formed. In the case of bisphosphine dioxide complexes, the two phosphinoyl groups are  Figure 3.3. Stereodiagram of Complex lOln (1:2 DPPEO/p-nitrophenol).  81  Figure 3.4 ORTEP Drawing of Complex R,R-117n (1:2 /?,/M17//?-nitrophenol). Space group, Fly a = 9.799(2) A, b = 20.961(3) A, c = 9.860(2) A, (3 = 91.19°, V = 2024.7 (6) A3; Z = 2; D c a l c = 1.254 g/cm3; R = 0.047.  82  in an anti conformation, an interaction that may involve a dipole repulsion. Table 3.5 summarizes the distances between the oxygen atom of the phosphinoyl group and the oxygen atom of the hydroxyl group of the carboxylic acid or phenol. The distances range from 2.53 A for 105t to 2.69 A in the case of 5,£-105111 and are shorter than the sum of the van der Waals radii for two oxygen atoms, which is 3.04 A. 139 These distances indicate the presence of hydrogen bonds between the phosphinoyl group and the hydroxyl group. 1^9  Table 3.5. Oxygen-Oxygen Distances of P=0- -H-0 in the Crystalline Complexes complex  oxygen-oxygen distance (A)  105t(l:l 105/TPPO) 105tt (2:2:1 105/TPPO/toluene) 5,5-105p (2:1 105/5,5-114) 5,5-105m (2:1 105/5,5-115) S,5-105x0 (2:1 105/5,5-116 lOln (2:1 />-nitrophenol/DPPEO) R,R-117n (2:1 />-nitrophenol/tf,/M17)  2.53 2.58 2.55, 2.60 2.68, 2.69 2.56, 2.61 2.62 2.56, 2.61  3.1.4 Clathratocomplexes or Coordinatoclathrates According to Weber,74 a coordinatoclathrate is a complex that demonstrates a certain degree of coordinative participation such as hydrogen bonding but has a dominant clathrate character and a clathratocomplex, just the other way round. The hydrogen bonded complexes of the phosphine oxides studied may be considered as clathratocomplexes or coordinatoclathrates. The coordinative participation by hydrogen bonding in the crystalline complexes of triphenylphosphine oxide has been demonstrated in a wide range of complexes. 131-136 But the clathrate character resulting from the steric bulk of triphenylphosphine oxide  83  should not be ignored.  As mentioned by Lynch et al.^2b for the complex of  triphenylphosphine oxide with 2,4-dichlorophenoxyacetic acid, the flexibility of the oxyacetic acid side chain allows the filling of the voids in the structure created by the steric bulkiness of triphenylphosphine oxide.  As seen in Figures 3.5 and 3.6, the  packing of triphenylphosphine complexes 105t$ and 105t shows the clathrate character of both the hydrogen bond donor and acceptor. In complex 105t (Figure 3.5), the donor molecule of carboxylic acid 105 packs with its ester group inside the void created by an acceptor molecule of triphenylphosphine oxide while coordinating to another neighboring acceptor molecule. Unlike complex 105t, complex 105tf includes half an equivalent of toluene in the crystal lattice.  The toluene molecule is packed with disorder in an  enclosure between the six phenyl groups of two triphenylphosphine oxide molecules. Two phenyl groups from two different triphenylphosphine oxides are pointed into the two V-shaped voids on both sides of vinyl bridge of dibenzobarrelene 105. The ester substituent on the vinyl bridge of one dibenzobarrelene molecule is packed just under the V-shaped cleft of a neighboring dibenzobarrelene molecule.  Figure 3.5. Packing Diagram of Complex 105t (1:1 TPPO/105). 84  Figure 3.6 Packing D i a g r a m o f Toluene molecules are omitted.  Comp,ex  TPPO/105/,olue„e)  B.sphosphine dioxides ,01 and 114-117 could be considered to be "wheel-and-axle" shaped molecules, one of the favorite geometric shapes for forming clathrates-76  but  these phosphine oxides lack mo,ecular rigidity because of free rotations about the carboncarbon single bond, so they are not considered to be very effective host molecules for mcluston. On the other hand, dibenzobarrelenes 105 and 106 are -roof-shaped" another geometrical body of host,*-  Obviously, hydrogen bonding is important in the  formation of the crystalline complexes between the bisphosphine dioxides and their hydrogen donor molecules, bu, the shapes of a„ these molecules also play a significant role in complex formation.  85  Figures 3.7 and 3.8 show the packing diagrams of the complexes of/^-nitrophenol with two different bisphosphine dioxides. ethanediylbis(diphenylphosphine  oxide)  are  In Figure 3.7, /7-nitrophenol and 1,2packed  in  a  layered  structure.  /?-Nitrophenol molecules are arranged in the voids created by the phenyl rings of the phosphine oxide molecules and sit at the top of the Y-shaped diphenylphosphinoyl group. The nitrophenol molecules are packed head-to-tail to cancel the large dipole moment  Figure 3.7. Packing Diagram of Complex lOln (1:2 l,2-ethanediylbis(diphenylphosphine oxide)/p-Nitrophenol).  86  while they are hydrogen bonded to the phosphine oxides. A different packing pattern is observed  in  the  complex  of  ^-nitrophenol  dimethylphenyl)phenylphosphine oxide) (Figure 3.8).  with  l,2-ethanediylbis((3,5-  Here the smaller nitrophenol  molecules sit in the voids between the phenyl rings from the phosphine oxides and are also hydrogen-bonded via hydroxyl and phosphinoyl groups. The nitrophenol molecules are paired head-to-tail and face-to-face while the whole unit is intercalated into the two  Figure 3.8. Packing Diagram of Complex 117n, (1:2 l,2-Ethanediylbis((3,5-dimethylphenyl)phenylphosphine oxide)/p-Nitrophenol).  87  phenyl rings from two different phosphine oxide molecules, a situation where the nitrophenol molecules are placed between two "wheels" of the phosphine oxide. Both complexes demonstrate that the phosphine oxides arrange their phenyl rings in a way that voids are created to include the nitrophenol. On the basis of their clathrate properties, these two complexes may be described as coordinatoclathrates or clathratocomplexes according to Weber's definitions,74 depending on whether the hydrogen bonding or the clathrate is considered to be the dominant factor in the complex formation. The crystal structures of three complexes of dibenzobarrelene 105 with bisphosphine dioxides were shown by X-ray crystallography to be isomorphous. Part of the crystal data is included in Table 3.6. These complexes are packed in a Fl\ space group.  Table 3.6. Crystal Data for the Complexes of Dibenzobarrelene 105 with Different Bisphosphine Dioxides  formula  105p  105m  105xo  C  C68H 6 0Ol 0 P2 1099.16  C70H64O10P2 1127.22 monoclinic F1X  68 H 60°10 P 2 1099.16 monoclinic  monoclinic  «1 8.988  «1 9.117  12.448  12.100  c, A  26.452  26.602  12.515 26.567  P, deg  90.74  94.85  91.44  Z  2  2  2  vA3  2959  2823  3052  Scaled > S / c m 3 R(F)  1.233  1.248  1.224  0.053  0.063  0.051  RW(F)  0.043  0.059  0.046  fw crystal system space group a, A b, A  88  9.184  As examples, the packing diagrams of complexes 105p and 105m are shown in Figures 3.9 and 3.10 respectively.  The alternating layers of the phosphine oxide and  dibenzobarrelene 105 are shown throughout the crystal lattices of the three complexes listed in Table 3.6. The vinyl substituents of dibenzobarrelene 105 occupy the space between the phenyl or aryl "wheels" of the phosphine oxides, whereas the phenyl or aryl  Figure 3.9. Packing Diagram of Complex 5,5-105p (2:1 105/5,5-114).  89  groups of the phosphine oxide are half intercalated into the V-shaped anthracene substituents of the dibenzobarrelene. The closely packed structures of these complexes demonstrate that the clathrate character of both the phosphine oxides and the dibenzobarrelene is an important factor in the formation of the crystalline complexes in addition to the hydrogen bonds.  Figure 3.10. Packing Diagram of Complex S,S-105m (2:1 105/5,5-115).  90  3.2 Solid State Photochemistry of the Complexes Dibenzobarrelenes 105 and 106 are known to undergo the di-7t-methane rearrangement in their excited states to give the corresponding dibenzosemibullvalenes (Figure 3.11).80,81 Two regioisomers were obtained for each of the above compounds, and the regioselectivity, i.e. the ratio of the two photoproducts, was found to be medium-dependent. In the present thesis, the same photoreactions were observed for these two compounds in their crystalline complexes with phosphine oxides.  ROOC  COOR  HOOC  ^r£m R=CH 2 CH 3 01(013)2  COOH  hu  105 106  COOR  + R=CH 2 CH 3 CH(CH3)2  138a 141a  R=CH 2 CH 3 139a CH(CH3)2 142a  I'  CH2N2  CH2N2  Diazomethane Adducts R=CH 2 CH 3 CH(CH3)2  HOOC  CH3OOC  COOR  137 140  R = CH2CH3 138 CH(CH3)2 141  R=CH 2 CH 3 139 CH(CH3)2 142  Figure 3.11. Di-7t-Methane Rearrangement of Dibenzobarrelene Derivatives 105 and 106 and the Diazomethane Workup.  The identification and analysis of the photoproducts followed the same procedures established before in our laboratory. 80,81  T1IJS  was based on diazomethane workup of  the starting material and the photoproducts (Figure 3.11). The photoproduct mixture was separated by silica gel column chromatography and analyzed by *H NMR 91  spectroscopy in their methyl ester forms (138 and 139 or 141 and 142), after the starting material was converted into diazomethane adducts 137 or 140. The regioselectivities of the photoreaction of dibenzobarrelene 105 in different complexes are shown in Table 3.7. The results show that the solid state photochemistry of dibenzobarrelene 105 is profoundly affected by cocrystallization with various  Table 3.7. Photochemical Regioselectivitiesa of Dibenzobarrelene 105 in its Crystalline Complexes with Phosphine Oxides low conversion (%)b complex conversion 105 105t 105tt 105t b 105d 105p 105m 105xo 105xm  20 19 21 31 49 25 26  138a/139a 0/100 56/44 3/97 1/99 85/15 65/35 18/82 27/73 44/56  high conversion (%)b conversion  138a/139a 56/44 29/71  97 82  24/76 85/15  100 86  27/73 41/59  comments  constant vs. conversion melted at high conversion melted at high conversion constant vs. conversion  constant vs. conversion  a  The regioselectivities (±2%) were determined by integration of the methyl signals from the *H NMR  spectra of the methyl esters of the photoproducts. ^The conversions (+5%) were determined by GC (DB-17, 230°C, head pressure 15 psi) immediately after diazomethane workup of the photolysis samples.  phosphine oxides. The 138a/139a product ratio ranges from 0/100 in crystals of pure 105 to 85/15 in the crystalline complex with ethanediylbis(diphenylphosphine oxide) (105d). Such variations were also observed for dibenzobarrelene 106, and the results are summarized in Table 3.8. In the complexes with triphenylphosphine oxide (106t) and ethanediylbis(diphenylphosphine oxide) (106d), the predominant photoproduct 141a was  92  obtained; in contrast, in the complexes with chiral bisphosphine oxides 106p and 106xo, photoproduct 142a was formed preferentially.  Table 3.8. Regioselectivitiesa of Dibenzobarrelene 106 in Crystalline Complexes with Phosphine Oxides low conversion (%) a  high conv(jrsion (%) a  complex  conversion  141a/142a  conversion  141a/142a  106  13  0/100  98  25/75  106t  16  85/15  96  89/11  106d  14  97/3  97  75/25  106p  -  -  57  37/63  106xo  -  -  64  28/72  Conversions (+ 5%) and regioselectivities (±2%) were determined by GC (DB-17, 225°C, head pressure 15 psi) immediately after diazomethane workup, and % NMR gave the same regioselectivies (+ 2%) as GC.  3.3 Photochemical Asymmetric Induction of the Complexes Bisphosphine dioxides 114-117 are chiral molecules (Figure 3.1), and optically pure enantiomers of these phosphine oxides were prepared via asymmetric synthesis as described in Chapter One (Figure 1.2).  Enantiomorphously pure complexes with  dibenzobarrelenes 105 or 106 were obtained by using optically pure bisphosphine dioxides. The photolysis of these complexes led to the formation of the corresponding enantioenriched photoproducts.  The optical yields given in e.e. % are shown in  Table 3.9. Enantiomeric excess was determined through analysis of the methyl esters by lH NMR spectroscopy at 400 MHz, using the chiral shift reagent (+)-Eu(hfc)3 (see Chapter 2 for details).  For compound 138a, the sign of rotation and absolute  configuration of the predominant enantiomer could be assigned on the basis of a previous  93  chemical shift-absolute configuration correlation; 81 in the case of the other photoproducts, such information is lacking and the net sign of rotation is unknown. From the ^H NMR spectra, however, it is clear that the stereoisomer of photoproduct 139 favored in (5,5)-105xm is the enantiomer of that favored in (R,R)-105xm.  Table 3.9. Optical Yieldsa in Asymmetric Induction of Dibenzobarrelenes 105 and 106 in the Solid State of the Complexes with Chiral Bisphosphine Dioxides photolysis conditions complex  t(°C)  time (h)  ratio of % 138a/139a conversion,  % e. e. 138ab 1139a| c  (5,5)-105p  25 -40 -40  3 120 23  31 44 5  65/35 67/33 77/23  20(-) 17(-) 45(-)  0 0 3  (5,5)-105m  25 -40 -40  1 48 16  49 90 13  18/82 13/87 19/81  21(+) 31(+) 28(+)  0 5 2  (5,5)105x0  25 25 -40 -40  18 6 16 43  100 25 9 91  27/73 27/73 35/65 26/76  17(-) 36(-) 15(-) 16(-)  35 38 53 49  (5,5)-105xm  25 25 -40  5 1.5 144  86 26 91  41/59 44/56 49/51  16(+) 13(+)  50 33 42  25 25 -40  1 3 43  33 95 61  46/54 53/47 38/62  9(-)  (R,K)-105xm  5(+) 3(+) 17(-)  31 38 55  141a/142a |141a| c |142a| c (5,5)-106p (5,5)-106xo  25 25  57 64  2 3.5  37/63 28/72  12 30  8 33  a  Optical yields ( + 5%) were measured by *H NMR spectroscopy, using the chiral shift reagent (+)-  Eu(hfc)3 to resolve the methyl esters. ^Enantiomer ( + )-138a was assigned as (4W?,8W?,8dl?,8d/?)-(+)138a and (-)-138a as its enantiomer according to a chemical shift-absolute configuration correlation.81 c  Only the absolute value was obtained by NMR spectroscopy; in lack of pure compound, the sign of  optical rotation was not decided, neither was the absolute configuration.  94  In terms of enantioselectivity, the optically pure bisphosphine dioxides have relatively low powers of solid state asymmetric induction (Table 3.9).  Generally  speaking, photolyses done at low temperature and low conversion give higher optical yields. Phosphine oxides 116 (complexes 105xo and 106xo) and 117 (complex 105xm) are the best, but even there only modest enantiomeric excesses could be achieved. By comparing the structures of the phosphine oxides, it can be concluded tentatively that the phosphine oxides that are di-substituted by methyl groups give a higher enantioselectivity than those that are mono-substituted.  As an example, the absolute configuration of  photoproduct 138 is assumed to be the same as its methyl ester 138a and can be determined. 81 Among the four complexes (.S^-lOSp, 105xo, 105m and 105xm, where dibenzobarrelene 105 cocrystallizes with the (S, 5)-enantiomer of different phosphine oxides, the former two give the preference to the photoproduct (-)-(4b5,8b»S',8clS,,8d^)138, while the latter two lead to the selection of (+)-(4bR,SbR,ScR,SdR)-13S. These enantioselectivities are considered to be intrinsic to the solid state photoreactions.  3.4 Stereoselectivity of the Photoreactions of Dibenzobarrelenes 105 and 106 in the Solid State Dibenzobarrelenes 105 and 106 are cyclic multichannel di-Tc-methane substrates, and they presumably react via the mechanism shown in Fig 3.12. The photoproducts that are obtained from initial vinyl-benzo bridging at the ester- and acid-bearing vinyl carbons bear a regioisomeric relationship to each other. Enantiomers of either photoproduct are formed via initial vinyl-benzo bridging at the same vinyl carbon but with the two opposite benzo groups (path avs a' and path b vs b'). Consequently, the regioselective pathways (a + a', b + b') leading to the formation of the major products in the solid state may be considered to be primarily determined by the differences in the electronic  95  A = OO2H  E = CO2CH2CH3 £ = 03201(013)2  Figure 3.12. Mechanism of Di-71-Methane Rearrangement of Dibenzobarrelenes 105 and 106.  96  and steric factors about the vinyl carbon atoms and their attached ester or acid groups, 1^0 whereas the enantioselective pathways (a vs. a' and b vs. b') leading to the formation of the enantiomers in the solid state would be primarily determined by the differences in the steric factors and, perhaps, orbital overlap between the vinyl and benzo groups** 1> 141. From the electronic point of view, the regioselectivity of the rearrangement is suggested to depend on the relative abilities of the carbonyl groups to stabilize the radicals at their respective 'vinyl' carbon atoms following initial vinyl-benzo bridging. If a biradical stabilization mechanism is the dominant factor determining the reaction pathways, then the major photoproduct should result from an initial vinyl-benzo bridging at the vinyl carbon atom bearing the carbonyl group which is less conjugated to the central double bond. Such an initial bond formation would result in an odd-electron at the 'vinyl' carbon atom which bears the more conjugated carbonyl group, a situation which is more favorable for reasons of resonance stabilization of the odd-electron. 142 The regioselectivity of the di-7t-methane rearrangement in the solid state may also be affected by the steric packing around the ester or acid group in the reactive molecule and its 'fixed' lattice neighbors. 140  Because benzo-vinyl bridging requires substantial  movement of the attached vinyl substituents, the free volume around each substituent in the crystal lattice may play an important role in determining the stereoselectivities. Theoretical calculations and experiments have shown that a more crowded environment hinders benzo-vinyl bridging.80b, 140 Strong hydrogen bonding adds another factor to the regioselectivity, 143,80a anc j a n initial bridging that involves movement of the COOH group may be disfavored because it disrupts hydrogen bonding. The absolute stereochemical reaction pathways (a vs a' or b vs b') have been proposed to be determined by the torsion angle, (0)C-C=C-C(0), between the vinyl and carbonyl carbon atoms of the dibenzobarrelenes.^l The torsion angle predisposes the molecule toward either of the two enantiomeric conformers in the solid state 97  (Figure 3.13). Owing to better orbital overlap between the interacting orbitals,^! as well as to a diminution of the steric interaction between the substituents and minimum movement for bridging, the reaction would go in the direction of the topochemically favored pathways. °1 >141 Therefore, vinyl-benzyl bridging would occur as indicated by the dashed lines shown in Figure 3.13.  This conformationally determined absolute  stereochemical reaction pathway will be discussed in Chapter 4 in more detail.  o II  c12  iM  9a  minus torsion angle  plus torsion angle  Figure 3.13. Predisposition of Vinyl Substituents and Absolute Reaction Pathways.  The results summarized in Tables 3.7, 3.8 and 3.9 show that the solid-state photochemistry of dibenzobarrelenes 105 and 106 is dependent on their cocrystallized phosphine oxides. In order to analyze the reaction pathways, the crystal structures of typical complexes of dibenzobarrelene 105, i.e. 105t, 105tf, 105d, 105p and 105xo, were analyzed by X-ray crystallography, and their ORTEP or PLUTO diagrams are shown in Figures 3.14-3.18. Complexes 105tj and 105t of triphenylphosphine oxide are in two different space groups, P2\/r\ and PI, respectively, but complexes 105d, 105p and 105xo are all in the same space group, Fl\. 3.4.1 Electronically Controlled Regioselectivity From the structures of complexes 105t and 105tf (Figures 3.14 and 3.15), we can see that one of the two carbonyl groups is out of conjugation with the vinyl bond. The conjugations are described quantitatively by the torsion angles between the carbonyl 98  groups and the etheno bridges that are summarized in Table 3.10. For complex 105t (Figure 3.14), the acid group is twisted 71° out of conjugation with the double bond, whereas the ester group is fully conjugated. This favors the formation of photoproduct 138a in terms of an electronically controlled reaction.  An analogous argument  rationalizes the reversed regioselectivity observed for the complex 105tt (Figure 3.15). In this case, it is the ester substituent that is 69° out of conjugation in the solid state, while the carboxylic acid group is closer to conjugation (14°). This would favor the formation of photoproduct 139a as observed experimentally (Table 3.10).  Figure 3.14. ORTEP Diagram of Complex 105t (1:1 TPPO/105).  99  Table 3.10. 0=C-C=C-C=0 Torsion Angles of Dibenzobarrelene 105 in the Complexes with Triphenylphosphine Oxide complex 105t 105tt a  <*>A(°)a  3>E(°)b  major product predicted  major product observed (138a/139a)  109.0 -13.8  -179.6 110.5  138a 139a  138a (56/44) 139a (3/97)  O A is the torsion angle of 0=C(OH)-C=C; bS>E is the torsion angle of 0=C(OEt)-C=C.  Figure 3.15. ORTEP Diagram of Complex 105tt (2:2:1 TPPO/105/toluene, toluene is omitted).  100  A steric effect may also play a role in determining the regioselectitivity.  Visual  inspection of the packing diagrams for the complexes of 105t and 105tf (Figures 3.5 and 3.6) reveals that, in both cases, the ester substituent of dibenzobarrelene 105 is in a crowded environment, and benzo-vinyl bridging at the ester-bearing vinyl carbon would be considered as a disfavored pathway. On the other hand, strong hydrogen bonding prohibits the movement of the acid group. However, hydrogen bonding is likely to be a stronger force in the complex than other molecular interactions. Overall, therefore, the bridging at the acid-bearing vinyl carbon (to product 138a, Figures 3.11 and 3.12) should be disfavored more than that at the ester-bearing vinyl carbon (to product 139a, Figures 3.11 and 3.12). Table 3.11 summarizes the different possible factors that affect  Table 3.11. Different Factors Influencing the Photochemical Regioselectivities of Dibenzobarrelene 105 in Triphenylphosphine Oxide Complexes electronic complex 105t 105tt  factor  138a 139a  crowding around  regioselectivity //-bonding  egter g r p u p  138a 138a  obseryed 138a/139a  139a 139a  56/44 3/97  the regioselectivity in the photolysis of complexes 105t and 105t^. In complex 105tf, two of the three factors tend to favor photoproduct 139a as a dominant product, while the opposing electronic factor in the case of complex 105t predicts a lower regioselectivity.  The experimental results show that complex 105tt gives almost  exclusively product 139a and that photolysis of complex 105t leads to a slight excess of product 138a (Table 3.11). The results indicate that both steric and electronic factors are important in determining the course of the photorearrangement. The X-ray crystal structures of complexes 105p, 105m and 105xo are shown in Figures 3.16-3.18. There are two independent molecules of dibenzobarrelene 105 in 101  each of the complexes. The ORTEP diagrams of complexes 105p and 105m reveal the extensive thermal motion of the ester and acid substituents. The disorder in the acid and/or ester substituents of all the three complexes may also indicate a substantial thermal movement of both substituents. Generally speaking, the molecular and crystal structures of all three complexes are similar. They are isomorphous and in the chiral space group P2\.  Their packing patterns are also quite similar as shown in Figures 3.9  and 3.10. This suggests that no obvious differences arise from steric factors that might affect the regioselectivity in the photorearrangement of dibenzobarrelene 105 in complexes 105p, 105m and 105xo. In other words, the regioselectivity may be caused by other factors. Table 3.12 is an attempt to analyze the regioselectivity from the electronic point of view, by following the same approach that was applied to complexes 105t and 105tf earlier in this section. In addition, the function of cos^$ is used in Table 3.12 as an angular dependence of the resonance energy to describe the degree of conjugation; 144 this will be a maximum at 180° and 0° and a minimum at 90°. We specify the two  Table 3.12. Analysis of Electronically Controlled Regioselectivities in the Photorearrangement of Complexes of Dibenzobarrelene 105 with Bisphosphine Dioxides 105p  105xo  105m  0 A (°)a (mol. H2f  170/125, 152  -164/128, 150  173/-127, 162  <DE(°)b (mol. 112)  95/94.5  -89/72  -68/59 0.99/0.36, 0.90  COS2OA  (mol. 112)  0.96/0.33, 0.78  0.92/0.38, 0.75  COS 2 O E  (mol. 112)  0.01/0.01  0.00/0.10  0.14/0.27  65/35  2 7/73  18/82  major product observed (138a/139a) a  d>Ais the torsion angle of 0=C(OH)-C = C; b O E is the torsion angle of 0=C(OEt)-C = C. cThere are  two independent molecules in the unit cell; the carboxylic acid group in molecule 2 is disordered.  102  Figure 3.16. ORTEP Diagram of Molecules in Complex 105p (2:1 105/(5,5)-l 14).  103  Figure 3.17. ORTEP Diagram of Molecules in Complex 105m (2:1 105/(5,5)-115).  104  Figure 3.18. PLUTO Diagram of Molecules in Complex 105xo (2:1 105/(5, S)-l 14).  105  independent molecules in each of the complexes as molecules 1 and 2, of which the acid carbonyl of molecule 2 is disordered and two torsion angles are given. The carboxylic acid groups of molecule 1 are well conjugated with the vinyl bonds in the three complexes (Table 3.12). Although the carbonyl groups in molecule 2 are disordered, a similar situation occurs in their conjugation to the vinyl bonds in the three different complexes. The ester carbonyls are out of conjugation with the vinyl bonds in all the three complexes, among which complex 105p is most completely non-conjugated (95°). Therefore, a very similar electronic effect would be present in the three complexes during the solid state photorearrangements and a similar regioselectivity would be expected. The analysis of the crystal structures of complexes 105p, 105xo and 105m predicts that the three complexes should undergo the di-7t-methane rearrangement to give a very similar regioselectivities.  The prediction does not agree with experimental results  (Table 3.12). The disagreement may arise from the disordered ester and/or acid vinyl substituent, which undermine the above analysis. Other factors such as the conformation of benzo groups may affect the regioselectivity of the reaction. 3.4.2 Absolute Configuration of Dibenzobarrelene 105 and its Enantioselectivity Dibenzobarrelene 105 becomes chiral in crystalline complexes 105p, 105m and 105xo because of the twisted vinyl bond (Figure 3.13). The two vinyl carbons can be considered as stereogenic centers. According to the sign of the C-C = C-C torsion angle (Figure 3.13) of dibenzobarrelene 105, a negative torsion angle is designated (11/?, 12/?)105 in the crystalline state and a positive angle is termed (115,12^-105. The absolute configuration of dibenzobarrelene 105 in these complexes was determined by X-ray crystallography, where the absolute configuration of the bisphosphine dioxides are known to be (S,S) (see synthesis in Chapter 1) and phosphorus is a sufficiently heavy atom to give rise to measurable anomalous dispersion. 145  106  What follows is an attempt to rationalize the enantioselectivity in the photolysis of complexes 105p, 105m and 105xo, according to the steric effect around the vinyl bond. If the steric and orbital overlap factors are considered to control the absolute stereochemical reaction pathways, 141,81a (UR,l2R)-105 and (115,125)-105 would give the 4bR,8bR,ScR,8dR- and 4b5,8b5,8c5,8d5 enantiomer respectively for either regioisomer 138a or 139a (Figure 3.12). Photoproduct 138a is taken as an example since its absolute configuration is known. Table 3.13 summarizes the predicted isomer based on the X-ray crystal structure analyses and the isomer actually obtained. The  Table 3.13. Torsion Angle-Enantioselectivity Correlation complex  $(C-C=C-C) mol 7/mol 2  (5,5)-105p (5,5)-105m (5,5)-105xo  0/3 19/7 0/-3  absolute configuration of 105 (115,125)b (115,125) (ll/U2i?) b  favored enantiomera predicted observed (-)-138a (-)-138a (+)-138a  (-)-138a (+)-138a (-)-138a  a  Enantiomer (+)-138a was assigned as (4W?,8W?,8ftc,8d7?)-(+)-138a and (-)-138a as its enantiomer.  "Molecule J is "achiral", and the absolute configuration is assigned to molecule 2.  predicted enantiomer does not agree with the enantiomer observed for the photolyses in general, and both complex (5,5)-105m and (5,5)-105xo disagree with the prediction; only (5,5)-105p gives the predicted isomer. The largest torsion angle in complex 105m does not give the best enantiomeric excess, whereas 105xo gives a moderate enantiomeric excess, as shown in Table 3.9. Because of the small torsion angles and the thermal motion of the vinyl substituents, the predictions are superficial. The disorder in the vinyl substituents further undermines the predisposition of the absolute configuration. The largest torsion angle that occurs in molecule 1 of complex (5,5)-105m accompanies a large thermal movement on both the carbonyl groups (Figure 3.17).  107  CHAPTER FOUR INCLUSION COMPLEXES AND THEIR PHOTOCHEMISTRY  Crystalline host-guest inclusion complexes have been used in studies of solid state photochemistry,46-49,146-151  an( j m e  bimolecular reaction of a host with its guest in  this condensed medium has been investigated. 149-151 A host alcohol was shown to undergo solid state photosolvolysis to form ether(s).149  j  n  another case, an  intermolecular hydrogen abstraction reaction was achieved between a ketone and its host deoxycholic or apocholic acid. 150,151 An inclusion host that is capable of undergoing a photochemical reaction by itself has not been studied. In this chapter, photoreactive hosts and their photochemistry in host-guest inclusion complexes will be discussed. One of our main goals in developing and studying photoreactive hosts is to control their photoreaction stereospecifically in the solid state. Crystalline inclusion complexes from the same host with different guests could crystallize in more than one space group, and the chances for the host to crystallize in a chiral space group could be enhanced. Such chiral crystals are then used for solid state photochemical asymmetric inductions. Moreover, the same host in a variety of inclusion complexes provides a structurally related system for the study of structure-reactivity relationships in the solid state, where the variations are the unreactive and relative small guests.  4.1 Formation of Inclusion Complexes Weber and Czugler^4 have delineated some of the important features of good hosts: they should be bulky and pack inefficiently with voids, and they should contain appended sensor groups that will coordinate to the guests.  108  "Roof-shaped' molecules such as  ethenoanthracene 103 and ethanoanthracene 95 (Figure 4.1) are known to be inclusion hosts.77 Unfortunately, the former is not very photoreactive in the solid state and also not an efficient host molecule, and the latter is inactive toward photoreactions. In order to design molecules that are efficient hosts and also photoreactive, the skeleton of molecule 103 and diphenylphosphinoyl groups were used, where the phosphinoyl groups were designed as a functional sensor for inclusion and the dibenzobarrelene skeleton as a photoreative moiety.  Such dibenzobarrelene phosphine oxide derivative would be  expected to be good inclusion hosts and to undergo the di-7i-methane rearrangement. 15a  O  O II PPh2  Ri  126 127 128 129 130  Ri=H R!=Q R^OpzCHa Ri=CH2OH Ri=H  R2 = H R2=H R2=H R2=H R2 = OH  Figure 4.1. Inclusion Host Compounds.  The  desired  compound,  9,10-dihydro-9,10-ethenoanthracene-ll,12-bis(diphenyl-  phosphine oxide) (126, Figure 4.1), was synthesized as described in Chapter 1. Ethenoanthracene 126 was found to form inclusion complexes with a wide range of aliphatic alcohols, ketones, ethers and esters.  All these complexes are shown in  Table 4.1. They were prepared by recrystallizing the host from the appropriate solvent and were characterized by IR and *H NMR spectroscopy as well as elemental analysis, and some of them were analyzed by TGA (thermogravimetric analysis) and DSC (differential scanning calorimetry). The inclusion stoichiometry is independent of the  109  size and bulkiness of the guests, and a 1:1 molar ratio of host to guest was invariably obtained from small guests such as ethanol as well larger guests such as ethyl acetate. The host, ethenoanthracene 126, includes not only alcohols, which may form hydrogen bonds with the host, but also other polar compounds, where the directed binding force may be of the dipole-dipole type.  Not surprisingly, ethenoanthracene 126 does not  include non-polar aromatic solvents such as benzene and very small molecules such as methylene chloride.  Table 4.1. Inclusion Compounds of Ethenoanthracene 126 guest included  recrystallization solvent(s)  comments  acetone 2-butanone 2-pentanone 3-pentanone methyl vinyl ketonea ethyl acetate tetrahydrofuran dimethoxyethane methanol0 ethanol n-propanol zsopropanol tert-butanol acetic acid0  acetone/CH2Cl2 2-butanone/CH2Cl2 2-pentanone/CH2Cl2 3-pentanone/CH2Cl2 methyl vinyl ketone ethyl acetate tetrahydrofuran dimethoxyethane methanol ethanol «-propanol wo-propanol  needles needles needles needles needles needles or plates needles or plates needles decomposes upon standing in air prisms or/and plates prisms prisms needles decomposes upon standing in air  a  tert-butanol/CH2Cl2 acetic acid  Prepared by Dr. Tai Y. Fu.  b  The ratio of host to guest was not determined due to the unstable  complex.  110  Thermally, the inclusion complexes in Table 4.1 begin to effloresce at approximately 120 °C, and eventually the empty host melts at 237-238.5 °C. As an example, the acetone complex was submitted to thermal analysis. TGA showed that this complex started to lose acetone at about 100 °C, which was completed at 152 °C (Figure 4.2). The weight loss corresponds to a 1:1 host/guest molar ratio. A similar trend was also observed by DSC. The endothermic loss of the guest took place between 100 °C and 150 °C and then the host melted at 237 °C (Figure 4.3). Unlike acetone, alcoholic guests gave a different pattern for the DSC trace. As shown in Figures 4.4 and 4.5, a higher temperature was required to drive the guest molecules of ethanol or isopropanol out of the host. A two-stage decomposition of the guests was observed. A similar DSC pattern from these two different alcohol guests was also demonstrated.  102139.10°C 100-  98-  96-2  147.B1°C(I) -8.7B0X  94-  -0 92151.57°C  246.76'C  90100  150 Temperature (*C)  200  250  300  Figure 4.2. Thermogravimetric Analysis of Inclusion Complex 1:1 126/acetone. Sample weight, 9.89 mg; heating rate, 10 °C/min.  Ill  234.B1*C  135.00•C  B -?-  -3-  236.75* 50  100  150  200  250  Temperature (*C)  Figure 4.3. Differential Scanning Calorimetry of Complex 126/acetone. Sample weight, 7.16 mg; heating rate, 10 °C.  "I 98. e y e 4.16BJ/9  151.91*C 40.SBJ/g  234.0B°C 53.43J/g  0-  -r-4117.63*C 166.60°C  236.74°C 30  110  150 Temperature  190  230  I 270  (*C)  Figure 4.4. Differential Scanning Calorimetry of Complex 1 126/ethanol. Sample weight, 6.85 mg; heating rate, 10 °C. 112  1-  •1  10B.42°C 5.495J/3  <H \  233.09'C 56.61J/g  146.5B°C 50.99J/g i—H-— 1  123.26*C  ^VT" ~-A T 161.04*C  -1-  -2-  I 236.42"C 53  90  130  170  210  250  Temperature (*C)  Figure 4.5. Differential Scanning Calorimetry of Complex 1:1 126/isopropanol. Sample weight, 7.33 mg; heating rate, 10 °C.  Ethenoanthracene inclusion hosts 127-130 (Figure 4.1) were designed by attaching substituents to ethenoanthracene 126.  The preparations have been discussed in  Chapter 1. Ethenoanthracene 130 differs from 126 in that a hydroxyl group, as a sensor group, is substituted at the 9-position. Two more hydroxylmethyl sensor groups were placed at the 1, 5-position of ethenoanthracene 126 to result in ethanoanthracene 129. With its two phosphinoyl and two hydroxyl groups, ethenoanthracene 129 has the most sensor groups among the ethenoanthracenes shown in Figure 4.1. The hydroxyl sensor group that was introduced into host ethenoanthracene 126 was to enhance its coordinative ability for inclusion. On the other hand, non-coordinative chloride and ester groups were substituted at the 1,5-positions of ethenoanthracene 126 to obtain C2-symmetric hosts, ethenoanthracenes 127 and 128. Ethenoanthracenes 127 to 130 were examined for inclusion complexes, but only ethenoanthracenes 129 and 130 were found to be effective inclusion hosts. Table 4.2 113  summarizes the inclusion complexes with these two hosts, which were characterized by the same procedures as described for ethenoanthracene 126.  The host:guest  stoichiometrics obtained from these inclusion complexes depend on the guest, ranging from 3:2 for the inclusion complex of 129 with ethanol to 1:2 for 130 with acetic acid.  Table 4.2. Inclusion Complexes with Host Ethenoanthracenes 129 and 130 host: guest molar ratio a  host  guest  129 129 130 130 130 130 130  ethanol ethyl acetate acetone tetrahydrofuran acetic acid zso-propanol te/t-butanol  3:2 1:1 1:1 2:3 1:2 1:1 1:1  comments -  needles from acetone/CH2Cl2 -  yellowing upon standing in air fine needles  a  The stoichiometry was determined by *H NMR spectroscopy and elemental analysis.  Host ethenoanthracenes 126, 130, 129, 127 or 128 exhibit a decreasing ability to form inclusion complexes in the number and the type of guests they include. Ethenoanthracene 126 is the most effective host, and includes almost all aliphatic compounds tested.  Ethenoanthracene 130 forms inclusion complexes with a limited  number of alcohols and other aliphatic compounds, but no inclusion is obtained from ethanol and ethyl acetate. Ethenoanthracene 129, with its more abundant sensor groups, turns out to be a poorer host than 130 and much poorer than 126.  No inclusion  complexes have been successfully prepared from the C2-symmetric ethenoanthracenes 127 and 128. The decreasing inclusion ability of the ethenoanthracenes coincides with their decreasing symmetry. The symmetry decreases from ethenoanthracene 126 (C2V) to ethenoanthracene 130 (Cs) and ethenoathracenes 127, 128, 129 (C2). The latter three  114  ethenoanthracenes are at the same level of symmetry in terms of the number of symmetry elements, but ethenoanthracene 129 turns out to be better hosts than 127 and 128 because of the presence of the coordinative hydroxyl sensor groups. The relationship between the symmetry of a host and its ability to form inclusion complex has been noticed, but not fully developed.74,152  4.2 X-ray Structure Analysis of Inclusion Complexes The crystal and molecular structures of the inclusion complexes of ethenoanthracene 126 with ethyl acetate, ethanol, isopropanol, and n-propanol were determined by X-ray diffractometry. The structures of ethenoanthracene 130 and its inclusion complex with isopropanol were also analyzed by X-ray crystallography. The molecular structures are shown in Figures 4.6-4.11. The guest molecules are normally disordered in the crystal lattices and are omitted in Figures 4.6-4.9. Part of the crystallographic data are listed in Table 4.3.  Complexes of Ethenoanthracene 126 with ethanol, isopropanol and n-  Table 4.3. Partial Crystal Structure Data for the Inclusion Complexes inclusions  126-EtOH  126-iPrOH  126/iPrOH  130-iPrOH  126-EtOAc  space group  P2\2i2i  P2i2i2\  Fl\2\2\  Fl\l\2\  P2i/c  a, A  18.076 20.904  18.191  b, A  17.983 20.682  18.223 9.435  c, A  9.384  9.432  18.528 20.676 9.452  20.748 9.436  103.430  P, deg Z VA 3  21.946  4  3563  4 3561.5  3621.0  4 3670.2  4  4  3490  Scaled* g / c m 3 1.238 R(F) 0.038  1.239  1.240  1.249  1.254  0.047  0.040  0.037  0.049  0.040  0.051  0.039  0.032  0.058  RW(F)  115  Figure 4.6. ORTEP Stereodiagram of Complex 1:1 126/EtOH.  Figure 4.7. ORTEP Stereodiagram of Complex 1:1 126/2-Propanol.  116  Figure 4.8. ORTEP Stereodiagram of Complex 1:1 126/7-Propanol.  Figure 4.9. ORTEP Stereodiagram of Complex 1:1 126/EtOAc.  117  Figure 4.10. ORTEP Stereodiagram of Complex 1:1 130/2-Propanol.  Figure 4.11. ORTEP Stereodiagram of Ethenoanthracene 130.  propanol as well as the complex of ethenoanthracene 130 with isopropanol are isomorphous and occupy the chiral space group Pl\l\l\.  Inclusion complex 126/EtOAc  (1:1) crystallizes in an achiral space group Fl\lc. A unique molecular recognition is demonstrated by hosts 126 and 130 in the formation of the inclusion complexes.  The parameters of the unit cells of all the  complexes in Table 4.3 are comparable to each other except for complex 126/EtOAc (1:1), where P = 103.43°, and the b and c axes are switched. For ethenoanthracene 126, the volumes of the unit cells increase as the size of the guests increase from ethanol, propanol and ethyl acetate, and so do the densities of the crystals, but not as much as the volumes in terms of percentage increase. An almost equal cell volume and density are shown by the two structural isomers of propanol, wo-propanol and npropanol.  The host increases the size of the unit cell to provide more room for  accommodating a larger guest. Careful examination of the molecular structures of all the complexes (Figures 4.64.11) reveals an intramolecular dipole interaction between the two phosphinoyl groups of an ethenoanthracene molecule. The oxygen atom of one of the phosphinoyl groups points directly at the phosphorus atom of the other (Species A in Figure 12). The P = 0 • P = 0 contacts are summarized in Table 4.4.  The van der Waals radii of  phosphorus and oxygen are 1.52 and 1.80 A respectively, 153 anc j the van der Waals  A  B  Figure 4.12. Intramolecular Dipole Interactions.  119  contact between these two atoms is 3.32 A. The contacts recorded in Table 4.4 are very close to the 3.32 A van der Waals contact, which assures the dipole attraction present. This is another type of intramolecular arrangement of two phosphinoyl groups in the solid state; in Chapter 2, we saw that they can take an anti staggered conformation on the ethane skeleton, which indicates a dipole repulsion (Species B in Figure 4.12).  Table 4.4. Intramolecular Contacts and Intermolecular Distances in Ethenoanthracene Compounds and Inclusion Complexes compounds/complexes 1 1 126/ethanol 1 1 126/isopropanol 1 1 126/n-propanol 1 1 126/EtOAc 1 1 130/isopropanol 130  P=0"P=0 contacts (A)  P=0-OR distances (A)a  3.34 3.33  2.75 2.88 2.80  3.31 3.41  -  3.23  2.73  3.38  -  P=0-0/P=0-HO distances (A)b -  2.58 2.61/1.74C  a  Intermolecular oxygen-oxygen distance between the phosphinoyl group of a host and the hydroxyl  group of a guest alcohol; ''intramolecular oxygen-oxygen distance between the phosphinoyl and the 9hydroxyl groups of a host; coxygen-hydrogen distance as the hydrogen atom was located.  Evidently, from the variety of inclusion complexes shown in Tables 4.1 and 4.2, hydrogen bonding is not essential for complex formation. However, hydrogen bonding is indicated by IR spectroscopy in those complexes involving hydrogen donor functional groups, where no sharp O-H stretching bands were observed. The hydrogen bonding was analyzed quantitatively by X-ray crystal structure analysis. Table 4.4 shows the distances between the oxygen atom of phosphinoyl group 1 (Figure 4.12) and the oxygen atom of the hydroxyl group of the guest. The intermolecular distances between the two oxygen atoms range from 2.73 to 2.88 A and are smaller than the 3.04 A van der Waals contact between two oxygen atoms. These distances are slightly higher than the average  120  value 2.72 ± 0.04 A of the oxygen-oxygen distance in species of O-H-0,154 indicating a weaker intermolecular hydrogen bonding in the inclusion complexes. In the case of ethenoanthracene 130, an intramolecular hydrogen bond is also observed; the phosphinoyl group is hydrogen-bonded to the alcoholic hydrogens of isopropanol intermolecularly and the 9-hydroxyl group intramolecularly as well, thus a three-center hydrogen bonding between one acceptor and two donors. The packing diagrams of the inclusion complexes in Table 4.3 are shown in Figures 4.13-4.17. The isomorphous structures (Figures 4.13-4.15 and 4.17) adopt the same packing pattern. A channel-like structure along the c-axis is created by the host molecules. These channels provide not only a site for the alcohol molecules, but also for  Figure 4.13. Packing Diagram of Complex 1:1 126/Ethanol.  121  J—J&.  +bz  dk  •#  0^-t^  ^  +a  (fcb  Figure 4.14. Packing Diagram of Complex 1:1 126/n-Propanol.  Figure 4.15. Packing Diagram of Complex 1:1 126/Isopropanol.  122  Figure 4.16. Packing Diagram of Complex 1:1 126/EtOAc.  Figure 4.17. Packing Diagram of Complex 1:1 130/Isopropanol.  123  the polar phosphinoyl groups to form intermolecular hydrogen bonds with the alcohol molecules. A similar channel structure along the b-axis is also presented in inclusion complex 1:1 126/EtOAc (Figure 4.16).  4.3 Photochemistry of Host Ethenoanthracene 126 As expected, ethenoanthracene-ll,12-bis(diphenylphosphine oxide) (126) undergoes the di-7t-methane rearrangement in solution and in the solid state to give a single photoproduct, dibenzosemibullvalene derivative 143, as shown in Figure 4.18. The quantum yield in deuterated chloroform is moderately high (O = 0.28, CDCI3). We assume that the photoreaction follows the standard Zimmerman mechanism shown in Figure 4.22. 1 5 a  Figure 4.18. Photorearrangement of Ethenoanthracene 126.  The photolyses of 126 were carried out in different solutions (acetone, chloroform, benzene and ethanol) as well as in different crystalline inclusion complexes. Photoproduct 143 was always the only product.  The photoreaction in the inclusion  complexes is a crystal-to-glass process. The photolysed crystals were transparent, looked like a single crystal and would lose their imprisoned solvent molecules upon heating at above 100 °C. As examples, the photolysed crystalline complexes of 126 with ethanol  124  and ethyl acetate showed no diffractions when they were submitted to X-ray structure analysis.  4.3.1 Characterization of Photoproduct 143 Photoproduct 143 was fully characterized by ^H, 31p and 1-*C NMR spectroscopy. The lH NMR spectrum indicated two non-aromatic protons at 6 5.00 and 4.50 ppm (C45-// and Cgfi-H, see numbering in Figure 4.19), which couple with the neighboring phosphorus nuclei. These two protons originate from the bridgehead protons of the reaction precursor, ethenoanthracene 126. There are two non-equivalent phosphorus  143  1  • " -  C8d C4b C8c  C8b V^N^Vy^i^W^V^WM^T'-'  >^W*WvvWW^^'^*w^VW^tWM*  VAVs^^yv»t^*M>^VHHK«Av»«Wri,i -1  65  55  50  •15 ppy  1  r-  40  Figure 4.19. Portion of ^ C NMR Spectrum and Assignment of Photoproduct 143.  125  nuclei in the molecule, which are shown by 3 *P NMR spectroscopy usually as an AB type spectrum (8 28.23, 28.66 ppm; J PP = 10.2 Hz). A portion of the  13  C NMR  spectrum is shown in Figure 4.19, where the signals are assigned to the individual nonaromatic carbons and the assignments have been confirmed by APT *3C NMR spectroscopy. Due to the non-equivalent phenyl groups and the coupling between carbon and phosphorus nuclei, an assignment of the spectrum in the aromatic region is not possible. According to the scheme shown in Figure 4.18, photoproduct 143 was reduced to bisphosphine 144,^4 wn ose molecular and crystal structure was determined by X-ray diffractometry (Figure 4.20).  The crystal structure as well as the spectral data for  bisphosphine 144 (see Experimental part for detail) fully support the structure of its dioxide 143.  Figure 4.20. ORTEP Stereodiagram of Bisphosphine (+)-144. X-ray structure analysis data: space group, Fl\2\l\\ a = 17.381(4) A, b = 18.414(3) A, c = 9.591 A, V = 3069(1) A 3 , Z = 4; R = 0.031 for 4b/?,8b/?,8c/?,8dfl-configuration, and R = 0.038 for the opposite configuration. The absolute configuration is thus established as (+)4bR,8bR,8cR,8dR with AR = 0.7%.  126  4.3.2 Asymmetric Induction Among the wide range of inclusion complexes of ethenoanthracene 126, single crystals of the complexes with ethanol, n-propanol and z'so-propanol were found to undergo absolute asymmetric solid state photorearrangement in high enantiomeric excesses (Table 4.5, entries 1-7). The enantiomeric excesses of photoproduct 143, as shown in Table 4.5, were obtained by chiral HPLC (CHIRALCEL OP(+), see Chapter 2 for details).  A nearly optically pure photoproduct could be obtained when the  photoreaction was carried out slowly on an intact single crystal (entries 1, 2, 5 and 7). Reasonably high enantiomeric excesses were still observed when the crystals were ground and photolysed at room temperature without external cooling (entries 3 and 6).  Table 4.5. Optical Yields of Photoproduct 143 from Single Crystal Photolyses (Pyrex) of 126 in its Inclusion Complexes with Different Guests photolysisa  conversion  time (h)  (%)  e.e. % b  titry  guest included  1 2 3 4  ethanol ethanol ethanol, ground crystal ethanol, ground crystal  96 48 18 43  84 21 96 43  91(+)  5 6  isopropanol isopropanol, ground crystal  18 120  43 97  96(+) 69(+)  7  «-propanol  19  32  92(+)  8 9  ethyl acetate acetone 2-pentanone  90 68 26  0  10  18 18 35  89(-) 92(-) 68(-)  0 0  a  Photolysis was carried out at room temperature except entry 4, which was done at -40 °C.  "Enantiomeric excesses (+2%) and photolysis conversions (+2%) were analyzed by chiral HPLC, CHIRALCEL OP (+); (-)-(4b5,8b5,8c5,8dS)- and (+)-(4b/?,8b/?,8cfl,8d/?)-143 were obtained.  127  The heat produced from the photolysis could heat the sample 10° to 30°C above room temperature if an external cooling system was not used. With external cooling, ground samples always gave more than 90% enantiomeric excess.  For example, a ground  sample was photolysed at -40 °C to result in 91 % e.e. at 43% conversion (entry 4). In contrast, no enantiomeric excesses were observed upon photolyses of the crystalline inclusion complexes of ethenoanthracene 126 with ethyl acetate, acetone and 2-pentanone (entries 8-10). 4.3.3 Absolute Asymmetric Photorearrangement of Host 126 X-ray crystallographic analysis of the inclusion complexes of ethenoanthracene 126 with ethanol, wo-propanol and n-propanol indicated that they are isomorphous and crystallize in the chiral space group Fl\l\l\.  The host molecule 126 loses its average  C2V symmetry in the crystalline state and takes on a chiral conformation.  Because  phosphorus is a sufficiently heavy atom to give rise to measurable anomalous dispersion (Af" = 0.434 for Cu K a radiation) during X-ray crystallographic analysis, the absolute configuration of ethenoanthracene 126 in the three inclusion complexes could be determined with a high degree of certainty by the Bijvoet method^  an( ] a r e  given in  Table 4.6 along with the P1-C11-C12-P2 torsion angles (Figure 4.21). Since two chiral centers are created on the 11- and 12-positions, the absolute configuration of molecule 126 in crystalline state can be decided as (115,125) or (11/?, 127?) according to whether  O •)"  11 JTi  (1LS,12S>126  (lLR,12tf>126  Figure 4.21. Absolute Configurations of Molecule 126 in Crystalline State. 128  its P1-C11-C12-P2 torsion angle is positive or negative respectively (Figure 4.21). The stereoisomers of photoproduct 143 that were obtained in each case are also included in Table 4.6. The absolute configuration of photoproduct 143 is correlated with phosphine 144, whose absolute configuration was determined in the same way as for ethenoanthracene 126 (Figure 4.20). 145 The (+)-4b/?,8W?,8c/?,8d/? stereoisomer of 144 was observed, and this enantiomer was obtained by reducing (+)-143 (> 99% e.e.) with trichlorosilane in benzene. We assume that the absolute configuration of photoproduct 143 does not change upon reduction to 144. Solid state photolyses demonstrated that (115,125)-126 (positive P1-C11-C12-P2 torsion angle) gives (-)-(4b5,8b5,8c5,8d5)-143 and that  (ll/?,12/?)-126  (negative P1-C11-C12-P2 torsion angle) gives (+)-  (4W?,8b/?,8c/?,8d/?)-143 (Table 4.6). The following is a description of how the absolute configuration correlation was done. A single crystal of 126/EtOH (1:1) was submitted to X-ray crystallographic analysis and its P1-C11-C12-P2 torsion angle was determined as +24°; this positive torsion angle indicated that the (115,125)-126 absolute configuration was adopted by the single crystal. This very crystal was then photolysed, and the  enantiomeric  excess of the photoproduct  was  measured as 89%  (-)-(4b5,8b5,8c5,8d5)-143 by chiral HPLC (entry 1, Table 4.5).  Table 4.6. Absolute Configuration Coordination in the Photorearrangement of Inclusion Complexes of Ethenoanthracene 126 configuration of photoproduct 143  0(P1-C11C12-P2 (°)  absolute configuration of 126  126/ethanol  +24  115,125  (-)-4b5,8b5,8c5,8d5  126/wo-propanoI  -21  11/?, 12/?  (+)-4b/?,8b/?,8c/?,8dR  126/n-propanol  -23°  11/?, 12/?  (+)-4b/?,8b/?,8c/?,8d/?  inclusion  129  The absolute asymmetric photorearrangement, as shown in Figure 4.22, is assumed to occur via the standard Zimmerman mechanism. 15a From the experimental results, it can be concluded that initial benzo-vinyl bridging occurs between C\\ and Qoa and/or C\i and Cga (path s) for (115,125)-126 (positive P1-C11-C12-P2 torsion angle), and  (4bR,8oR,ScR,SdR)-143  ^-—  rac-126  (4bS,8bS,8cS,8dS>143  (11R,12R)-126  Figure 4.22. Absolute Reaction Pathways of Inclusion Complexes of 126. Ci! and C^ and/or C\i and Cga (path r) for (\\R,\2R)-\26 (negative P1-C11-C12-P2 torsion angle). The situation is represented by structure 126 in Figure 4.23. A positive P1-C11-C12-P2 torsion angle, assigned as (IIS, 125)-126 in Figure 4.23, predisposes the molecule to C12---C9a bridging (or its stereochemical equivalent, Cll---C10a bridging) owing to better orbital overlap between the interacting orbitals as well as to a diminution of the steric interaction between the bulky diphenylphosphinoyl substituents in the  130  transition state. In the alternative pathways leading to the unobserved enantiomer, the diphenylphosphinoyl groups would be driven toward each other during the benzo-vinyl bridging.  In the same way, a negative P1-C11-C12-P2 torsion angle, assigned as  (ll/?,12/?)-126 in Figure 4.23, leads to benzo-vinyl bridging in the opposite direction to give the other enantiomer.  (1LS,12S>126  (11*,12K>126  Figure 4.23. Molecular Predisposition by the P(l)-C(l 1)-C(12)-P(2) Torsion Angle.  The topochemically controlled absolute reaction pathways in the photorearrangement of ethenoanthracenes were briefly discussed earlier in Chapter 3. Ethenoanthracene 126 is an extreme case, in terms of the twisted vinyl bond and the steric hindrance around it as well as the enantioselectivity achieved during the rearrangement reaction. Similar ethenoanthracenes such as 61 (E = z'-PrOH) as well as ethenoanthracene salts 73 and 99 were found to undergo the absolute asymmetric di-7t-methane rearrangement. Garcia-Garibay et al. studied the first case of the absolute asymmetric photorearrangement of ethenoanthracene 61 (E = z'-PrOH, Figure 4.24), and they proposed that the steric effect is mainly responsible for the enantioselectivity.^3,155 The predisposed absolute configuration of ethenoanthracene 61 (Figure 4.24), which was determined by X-ray crystallography, led to a stereospecific di-7i-rearrangement to give (4bS,SbS,ScS,SdS)-62 (E = /-PrOH) at a low conversion (< 10%). The unobserved enantiomer (4b/?,8b/?,8ci?,8d/?)-62 has been suggested to be caused by the severe steric  131  clashing of the ester groups as they are forced toward each other during the initial stages of reaction.  In contrast, the reaction pathway(s) leading to the obtained enantiomer  involves movement of the ester groups away from one another, as indicated by arrows in Figure 4.24. 1 5 5  no  R0 2 C  -C°2R  hu, P2i2i2i * •  R = /-Pr 61  (4hS',8faS,8cS,8dS)-62  Figure 4.24. Absolute Asymmetric Photorearrangement of Ethenoanthracene 61.  The absolute asymmetric photorearrangement of dibenzobarrelenes was further studied by Gudmundsdottir et al.^l They used ionic chiral handles to pre-resolve the dibenzobarrelenes in the crystalline state, where the absolute configurations of the dibenzobarrelenes were described by the sign of the torsion angles depicted in Figure 4.25. As illustrated in Figure 4.25, the negative torsion angle in salt 73 and positive  torsion  angle  in  salt  99 were photolysed  to give  stereoisomers  (4bR,8bR,8cR,8dR)-l38 and (4bS,8b,S,8cS,8cLR)-100 respectively after diazomethane workup. Both salts react presumably in such a way (dotted lines) that the interaction between the vinyl substituents is reduced in the benzo-vinyl bridging transition state rather than increased as it would be if the alternative pathways were followed. These alternative pathways would necessarily involve eclipsing of the substituents at some stage. Orbital overlap, as proposed by Gudmundsdottir, may be another factor that may favor the pathways shown by dotted lines in Figure 4.25. Owing to the negative torsion angle in salt 73 or positive angle in salt 99, the p-orbitals overlap better in the directions  132  indicated by dotted lines, thus favoring formation of (4W?,8W?,8ci?,8di?)-138 or (4b5,SbS,8cS, 8dR)-100 respectively.  99  n  CH3  4hS,8bS,8cS,8d«-100  Figure 4.25. Absolute Asymmetric Photorearrangement of Ethenoanthracene Salts 73 and 99.  4.3.4 Asymmetric Synthesis of Compound 143 The absolute asymmetric photorearrangement of ethenoanthracene 126 in its inclusion complexes has synthetic potential for the preparation of an enantiopure diphosphine. The diphosphine can be used as a ligand for asymmetric catalysts.  This one step  photoreaction produces four stereogenic centers in photoproduct 143, and it can be easily reduced to diphosphine 144 by silanes.H2-115 For preparative purposes, a large amount of enantiomorphously pure crystals were required. A large scale preparation of enantiomorphously pure crystals was achieved by seeding the crystallization solutions with a ground single crystal of known absolute configuration.  Using this method, the size of the crystals could be controlled by the  supersaturation of the solution, the less supersaturated solution resulting in larger crystals  133  and vice versa. The powder provides a tremendous number of enantiomorphously pure crystal seeds and prevents the other enantiomer from crystallizing out of solution, and as a result, a batch of enantiomorphously pure crystals is prepared.  Table 4.7 is a  comparison of the efficiency of preparing enantiomorphously pure samples. The crystals from powder seeding gave comparable enantiomeric excesses as the single crystals, indicating that the crystals from powder seeding are enantiomorphously pure. From a moderately supersaturated stirred solution in ethanol with a single crystal seed, ethenoanthracene 126 only gave partially resolved crystals, although a stirred  Table 4.7. Optical Yields of Photoproduct 143 from Polycrystal Photolysis of 126/EtOH (1:1) Prepared by Different Seeding Methods photolysis conditions3 seeding methods  time (h)  temperature  (%)  e.e.% b  powder seeding  18 10 43  r. t. r. t. -30  95 97 33  15 96  -40 °C -40 °C  28 58  65(+) 67(-) >99(+) >95(+)c >99(-) 94(+)  48 18  r. t. r. t.  21  92(-)  43  -40 C  96 43  68 (-)d 91 ( + ) d  seeding and stirring  19 43  r. t. -40 °C  94 94  11(+) 30(+)  self-crystallization  18  r. t.  90  0  single crystal  a  conversion  The photolysis was done through Pyrex filter on polycrystals except for the single crystal case, in  which only a single crystal was used; r. t., room temperature b(-)-(4bS,8bS,8cS,8&S)- and (+)(4bfl,8W?,8cfl,8dfl)-143 were obtained, and the e.e ( + 2 % ) was determined by chiral HPLC, CHIRACEL OP( + ).  c  The e.e. (±2%) was determined by  31  P NMR spectroscopy with chiral  shift reagent D-DBT. "The photolysis was done on a ground single crystal.  134  crystallization of aqueous NaC103 has been reported to give more than 98% crystal enantiomeric excess.34  i n a special case, self-crystallization from a supersaturated  solution only gave a batch of crystals that no enantiomeric excess was obtained for its photoproduct 143 (Table 4.6). We did not observe a spontaneous resolution. However, some achiral molecules can resolve spontaneously to yield enantiomorphously pure crystals. al  l',l-Binaphthyl is a famous example that was demonstrated by Pincock et  156,157 Rapid equilibrium between the two enantiomers (Figure 4.26) in the molten  phase makes 1',1-binaphthyl maintain its overall symmetry. When the melt crystallizes, random nucleation leads to the growth of enantiomorphously pure crystals.  (5>(+>l',l-binaphthyl  <7?>(->l',l-binaphthyl  Figure 4.26. Equilibrium between the Enantiomers of 1', 1-Binaphthyl.  In order to get high optical and chemical purity, preparative photolysis was carried out at a low temperature and around 30% conversion.  The starting material was  precipitated out in acetone after photolysis and then recycled.  The optically pure  photoproduct 143 was easily obtained from the mother liquor and crystallized from propyl acetate.  As an example, 1 g of the inclusion complex with ethanol was  photolysed at -40 °C for 3 days, and 0.18 g of the photoproduct was obtained from propyl acetate after 0.73 g of the starting material was recycled from acetone. The isolated product, which gave > 99% (+)-(4W?,8W?,8cfl,8dK)-143, was reduced to phosphine 144 in a yield of 0.10 g. Alternatively, 1.0 g of crystals was photolysed to a 40% conversion at -40 °C. After 0.59 g of starting material was recycled from acetone,  135  the oil left after evaporating the filtrate was reduced by 2 ml of trichlorosilane in 5 ml of dry benzene, whereupon 0.20 g of optically pure phosphine 144 was obtained with an overall yield of 50%.  4.4 Photochemistry of Inclusion Host 130 The 9-hydroxydibenzobarrelene 145 has been studied by Wright and coworkers. 158 Instead of giving a hydroxysemibullvalene, irradiation of this material in solution gave almost exclusively one of the keto diesters depending on the photolysis conditions. The rearrangements are shown in Figure 4.27.  Photolysis of ethenoanthracene 145 in  methylene chloride or benzene gave keto diester 148 whereas irradiation in acetone gave  R  ,R hu  R  147  O  148  Figure 4.27. Photochemistry of 9-Hydroxydibenzobarrelene Derivative 145 (R = C0 2 Me).  136  keto diester 147, together with variable trace quantities of ketodiester 148. Wright et al. suggested the formation of the di-7t-methane rearrangement intermediate 146, and its isomerization to two stereoisomers, 147 and 148. The regioselectivity of the observed rearrangement was suggested to be due to hydrogen bonding between the 9-hydroxyl group and its closest ester group; the hydrogen-bonded methyl ester group of diradical 145b is a to a radical site and consequently is less able to stabilize that radical by 7tdelocalization (Figure 4.27). This does not apply to the alternative pathway and hence formation of diradical 145a is preferred over 145b. Instead of ester groups at the 11 and 12-positions, ethenoanthracene derivative 130 was prepared in which the two positions are substituted by diphenylphosphinoyl groups. The preparation was described in Chapter 1. As mentioned earlier in this Chapter, ethenoanthracene 130 is an inclusion host, and this compound and its inclusion complexes were shown to undergo the di-7i-methane rearrangement in a similar way as its analog ethenoanthracene 145. Photolysis of ethenoanthracene 130 in acetone gave exclusively photoproduct 149 (Figure 4.28), while in other solvents and in the solid state of its inclusion complexes, other minor products were also observed by 31p NMR spectroscopy.  OH R=P(0)Ph2 130  hu » • acetone  149a  Figure 4.28. Photorearrangement of Ethenoanthracene 130.  137  4.4.1 Characterization of Photoproduct 149 The structure of photoproduct 149, formed via the photolysis of ethenoanthracene 130 in acetone, was elucidated by *H, 31p and l^C NMR spectroscopy as well as FTIR and mass spectrometry. Both the ^C NMR (8 199.2 ppm) and IR spectra (v 1709 cm~l) showed the presence of a carbonyl group that conjugated to a phenyl group. The 31p NMR spectrum indicates the two nonequivalent phosphorus atoms (5 29.1 and 34.3 ppm) that couple to each other (J = 23.8 Hz). Proton NMR spectroscopy showed two non-aromatic protons (5 5.30, 5.08 ppm) that couple with two phosphorus nuclei. Due to the complication of numerous nonequivalent aromatic carbon atoms and their couplings with phosphorus atoms, a complete assignment of ^C NMR spectrum was not possible, but the non-aromatic carbons could be assigned as indicated in Figure 4.29. The assignment was confirm by APT l^C NMR spectroscopy.  Qb C9  R H R 0 ;o>C V A 9  k'^f'^^'vM^ti 2CC  c  c  10  149 R = Ph2P(0)  9a  v ^ A ^ ^ ^ ^ u ^ y ^ W v ^ ^ ^ Vwyr T W^ft/mti^r*1^^ —1  1  :  •  r~  —i—•—:—I—:—!—^—i  40  •—'  :—t—:  30 ="','  Figure 4.29. Partial ^ C NMR Spectrum of Photoproduct 149 its Assignment.  138  The stereochemistry of photoproduct 149 was deduced from the vicinal coupling constants, JpccH  (JPH)-  The vicinal P-C-C-H coupling has been shown to effectively  follow a Karplus-type variation and is very useful in determining the configuration of organophosphorus compounds. 159  On a typical five-membered ring, cis vicinal  coupling is much less than trans one. 159b Figure 4.30 shows some typical structures and the vicinal coupling constants Jpn and J HH .  From those know compounds, it is  clear that trans coupling constants are several times of cis ones. The coupling constants for photoproduct 149, JPH(9a,4b) and JPH(9a,10), are 14.2 and 1.2 Hz respectively; indicating that Pga is cis to H^ and trans  (CH30)2P(0)  XOH\Q.  H (0)P(OCH3)2  CIS  J ( P 3 A ) = 8.1HZ  trans  J ( P , H B ) = 0-3HZ  R=Ph 2 P(0)  J P H = 17.4-20.0 Hz J P H = 6.0-8 Hz  R = Ph 2 P(0)  150  149  JHH(4b,9a) = 7.1Hz JHH(10,9a)=1.9Hz JPH(10,9a)=17.7Hz  JPH(9a,4b)= 14.2 Hz JPH(9a,10)=1.2Hz JPp(10,9a) = 23.8Hz  Figure 4.30. Vicinal Coupling Constants between Two Nuclei that are Bonded to a Five-membered Ring.  139  Photoproduct 149 was derivatized to compound 150.  According to Warren and  coworkers, 160 a-phosphorylketones undergo dephosphorylation reactions to yield ketones under basic conditions. The dephosphorylation product (150) was obtained from 2% sodium methoxide solution in methanol according to the scheme shown in Figure 4.31. Compound 150 was fully characterized and described in the Experimental section. The stereochemical relationship between H\Q, Hga and H^  in compound 150 were  determined according to their coupling constants on the five-membered ring (Figure 4.30).  For a typical cyclopentene ring, Jcfs is 7.4 Hz and Jtrans *s 4.6 Hz.l"!  Compound 150 shows a coupling constant between H^ and Hga of 7.1 Hz, which indicates that a cis relationship exists between these two protons. The coupling constant between Hg^ and H\Q is 1.9 Hz, which means that a trans stereochemistry is present for these two protons. The vicinal coupling between P\Q and Hga (J = 17.7 Hz, Figure 4.30) shows a cis configuration between these two atoms and confirms the above stereostructure assignments from the proton couplings. The structure of compound 150 as well as its stereochemistry is shown in Figure 4.31. The structure of compound 150 helps to confirm the structure proposed for photoproduct 149.  CH3ONa/CH30H  149  150  R=Ph 2 P(0)  R=Ph 2 P(0)  Figure 4.31. Dephosphorylation of Compound 149.  140  4.4.2 Photorearrangement Mechanism The mechanism of the photorearrangement of ethenoanthracene 130, which is shown in Figure 4.32, is assumed to be the same as that of compound 145 (Figure 4.27). The stereostructure of photoproduct 149 observed is consistent with the standard di-7imethane rearrangement.  R  ,R  R=P(0)Ph 2  Figure 4.32. Photorearrangement Mechanism of Ethenoanthracene 149.  The exclusive photoproduct formed in acetone can be explained by following the same argument as in the case of the photorearragement of ethenoanthracene 145. A hydrogen bonded diphenylphosphinoyl group is conjugated to radical 130b; this conjugation undermines the resonance stability of the phosphinoyl group. In contrast, diradical 130a is not coordinated. Therefore, diradical 130a is more stable than diradical 130b and a dibenzosemibullvalene type of intermediate 149a is presumably formed to give photoproduct 149. But this argument was questioned by Paquette and Bay.162 141  They studied in detail the bridgehead substituent effects, especially phenyl, and suggested that it may be that second biradical formation is product-determining owing to the extraordinary stability of the substituents.  In the di-7t-methane rearrangement of  ethenoanthracene 130, the secondary diradicals are 130a' and 130b1, of which diradical 130a' is stabilized by hydroxyl group in one center. Therefore, 130a' is more stable than 130b'.  Providing the first diradical formation (130a and 130b) is reversible,  intermediate 149a would form to give product 149 (Figure 4.32). Dibenzosemibullvalene intermediate 149a (Figures 4.32 and 4.33) was not observed after photolysis.  The unstable cyclopropanol ring!63  an<j  the sterically crowded  environment may be the main cause of the unobserved intermediate.  Wright and  coworkers 1 " suggested a carbanion mechanism for ring cleavage of dibenzosemibullvalene intermediate 146. This was based on the solvent dependence of the formation of the keto diesters (Figure 4.33). In polar acetone, the carbanion would be stabilized sufficiently for proton attack to occur at either face of C\Q, and the least sterically hindered face prefers kinetically to give a cw-form product with variable trace quantities  Yfsyn attack  /.9 acetone * •  R=C02Me 147 R=P(0)Ph2 149  R=(X>2Me R=P(0)Ph2  146 149a R = CC>2Me 148 Figure 4.33. Cleavage of Cyclopropanol Intermediates.  142  of trans-form (Figure 4.33). Benzene and methylene chloride, being non-polar, would not stabilize a carbanion to any extent, so we suggest that a concerted ring cleavage of intermediate 146 occurs to give trans-form product exclusively (Figure 4.33). 163 Intermediate 149a may cleave via the carbanion mechanism as 146 to give cw-phosphine oxide product 149 in the acetone solution photolysis of ethenoanthracene 130 (Figure 4.33). Because of the bulky diphenylphosphinoyl groups, the overwhelming deform would be expected to form kinetically during the cleavage. 4.4.3 Asymmetric Induction of Inclusion Complex 130/wo-Propanol As indicated by X-ray structure analysis (Table 4.3), the inclusion complex 1:1 13O/z',s0-propanol crystallizes in a Fl\l\l\  chiral space group. As a result, an absolute  asymmetric photorearangement in the solid state of this complex would be expected to occur. Enantiomorphously pure polycrystals were obtained through the powder seeding method described earlier in this Chapter. Single crystals were grown from isopropanol upon slow evaporation. Irradiation of enantiomorphously pure crystals of 130/z-PrOH (1:1) gave enantioenriched photoproduct 149 along with other unidentified minor products. The yield and conversion were estimated by 31p NMR spectroscopy. The enantiomeric excess of photoproduct 149 was measured by 31p NMR spectroscopy by using D-DBT as a chiral shift reagent (see Chapter 2 for details). The results shown in Table 4.8 indicate that moderate enantiomeric excesses (around 65%) were observed for photolyses at both room temperature and low temperature.  Irradiation by using less  intensive light (thus longer irradiation time) or irradiation at low temperature afforded a slightly higher e.e. (entries 4-6), but the maximum enantiomeric excess reached was 70%. In contrast with its isomorphous complexes of 126/iso-PrOH. (1:1), 126/«-PrOH (1:1) and 126/EtOH (1:1), complex 130/wo-PrOH (1:1) undergoes the di-7i-methane  143  rearrangement with a slightly lower enantioselectivity in the solid state. The highest enantiomeric excess, 70%, was observed at -30 °C and at a relative low conversion (Table 4.8). The maximum e.e. does not match that obtained from enantiomorphously pure inclusion complexes of ethenoanthracene 126, where almost enantiomerically pure photoproduct 143 was obtained under the same conditions (Tables 4.5 and 4.7). However, the results obtained from both hosts at room temperature are more comparable.  Table 4.8. Photochemical Asymmetric Induction in 130//-PrOH in the Solid State entry 1  photolysis time (h)a conversion(%)b  yield (%)b  e.e. %c  2  2 2  54 41  53 52  53 65  3  7  55  53  63  4  16  44  56  68  5  16  35  58  70  6  43  48  61  70  a  Photolysis was carried out through a Pyrex filter at room temperature except entry 6,  which was done at -30°C. ^Conversion and yield (± 5%) was estimated by 31p NMR spectroscopy and yield was corrected for unreacted starting material; ce.e. (± 2%) was determined by 31p NMR spectroscopy by using chiral shift reagent D-DBT.  The absolute asymmetric photorearrangement of inclusion complex 130/wo-PrOH (1:1) can be rationalized according to X-ray structure analysis. The ORTEP diagram of complex 130/wo-PrOH (1:1) shown in Figure 4.34 indicates that a static disorder occurs in the crystal, which results in two different conformations for ethenoanthracene 130 with roughly a 4:1 ratio (Figure 4.35). For the major conformer, the 9-hydroxyl group  144  C43A  Figure 4.34. ORTEP Diagram of Inclusion Complex 1:1 130/isopropanol.  ,H'° O JV=</ R  130 minor confermer  149  130 major conformer  Figure 4.35. Absolute Reaction Pathway of Inclusion Complex 130/iso-PrOH (1:1) in the Solid State.  145  points to the outward phosphinoyl group, interacting with it via intramolecular hydrogen bonding. For the minor conformer, the hydroxyl group is located on at the opposite side and points out to an wo-propanol molecule, interacting with it via intermolecular hydrogen bonding.  In both cases, the two diphenylphosphinoyl groups interact via  dipole attraction, and the outward phosphinoyl group is hydrogen bonded to an isopropanol molecule. The P1-C12-C11-P2 torsion angle was measured to be 23.6 (6)°, and assuming the disposition of the diphenylphosphinoyl groups to be the main force directing the photoreaction in the solid state, as it does in the case of the inclusion complexes of 126, the two conformers would be expected to give similar asymmetric induction in their common photoproduct 149. The two species are assumed to follow the vinyl-benzyl bridging as indicated by dash lines in Figure 4.35, and result in the same enantiomer of product 149, which is the major enantiomer. In this case, we do not know the absolute course of the rearrangement. Unlike the inclusion complexes of ethenoanthracene 126, photolysis of inclusion complex 130/wo-PrOH (1:1) could not be driven to reach a complete conversion in the solid state. The latter complex also gave more than one photoproduct in the solid state, but attempts to separate these minor products by silica gel chromatography failed, as these phosphinoyl-bearing compounds are too polar to be eluted from silica gel.  146  CHAPTER FIVE PHOTOCHEMISTRY OF PHOSPHTNE OXIDE 132 AND 133  As discussed in Chapter 4, bisphosphine dioxide derivatives of dibenzobarrelene can undergo the di-7t-methane rearrangement in solution as well as in the solid state. In order to compare the ability of the phosphinoyl and ester groups to direct the regioselectivity of the di-7t-methane photorearrangement, ethenoanthracene 132 (Figure 5.1) was synthesized as described in Chapter 1, and its photochemistry will be discussed in this Chapter. Both ethenoanthracenes 132 and 133 (Figure 5.1) are phosphine oxides. Since the phosphinoyl group is a very strong hydrogen bond acceptor, we want to know how hydrogen bonding affects the photoreaction of these phosphine oxides in solution as well as in the solid state.  O  O  132  133  Figure 5.1. Ethenoanthracenes 132 and 133.  Previously in our laboratory, the photorearrangement of ethenoanthracene carboxylic acid derivatives has been shown to be affected by hydrogen bonding. For example, ethenoanthracene 106 and subsequently 105 were studied (Figure 5.2).80.81 Different media gave different regioselectivies for the formation of photoproducts 138a vs. 139a and 141a vs. 142a (Figure 5.2). Three different hydrogen bonded forms (A, B and C in  147  O . .OR  CO2R/O2H  COjH/Q^  hu  + R=CH 2 CH 3  105  R=CH(CH3)2 106  R=CH2CH3  138a  R=CH2CH3  139a  R=CH(CH3)2  141a  R=CH(CH3)2 142a  Figure 5.2. Photorearrangement of Ethenoanthracenes 105 and 106.  Figure 5.3), were suggested to be responsible for the regioselectivities. Species A is a monomer that is hydrogen bonded to a solvent molecule, B is an intramolecularly hydrogen bonded form, and C is an intermolecularly hydrogen bonded dimer, which also exists in the solid state. The predominance of form A in polar solvents such as acetone  Solvent/,,,  H/,, / 'OvV OR  R = CH2CH3  105  R=CH(CH3)2  R °nr  1  .COOR  106  Figure 5.3. Different Hydrogen Bonded Species of Ethenoanthracenes 105 and 106.  and acetonitrile was suggested to give almost equal amounts of the two photoproducts, since the solvent bonded carboxylic acid group has a similar radical stabilizing ability as the ester group. X-ray crystal structure analysis shows dimer form C is the only species present in the crystalline state. The solid state photolysis of ethenoanthracene 105 or 106 favor the formation of product 139a or 142a respectively, since the carboxylic acid  148  group is held by hydrogen bonding and hinders the motions necessary to form product 138a or 141a. Species B, which exists in benzene solution at a low concentration, has been proposed to photorearrange by a unique mechanism. This mechanism, shown in Figure 5.4, involves an excited state proton transfer, followed by a 1,2-aryl shift to give only one of the possible regioisomers.80,81  R=CH2CH3 1 0 5 R=CH(CH3)2 106  R=CH2CH3 138a R=CH(CH3)2 141a  Figure 5.4. 1,2 Aryl Shift Mechanism.  5.1 Photolysis of Ethenoanthracenes 132 and 133 Phosphorus-31 and proton NMR spectroscopy showed that ethenoanthracene derivative 133 undergoes photorearrangement both in solution and in the solid state. Two photoproducts, 151 and 152, were obtained (Figure 5.5).  Diazoethane workup  converted photoproducts 151 and 152 into their ethyl esters 153 and 154 respectively. Dibenzosemibullvalenes 153 and 154 were also obtained as the only products from the photolysis of ethenoanthrance 132 both in solution and the solid state (Figure 5.5).  149  R=CH2CH3  153  R = CH2CH3 154  Figure 5.5. Photorearrangement of Ethenoanthracenes 132 and 133.  Compound 152 is the major product from the solid state photolysis of ethenoanthracene 133, and pure photoproduct 152 was obtained by crystallizing the photolysed sample from a mixture of chloroform and acetonitrile. The molecular and crystal structure of this photoproduct was analyzed by X-ray crystallography.  The  ORTEP diagram of photoproduct 152 is shown in Figure 5.6. Esterification of 152 by  Figure 5.6. ORTEP Diagram of Photoproduct 152.  150  diazoethane gave dibenzosemibullvalene 154, one of the photoproducts from the photolysis of ethenoanthracene 132. The minor photoproduct 151 could not be separated from the photolysis mixture of ethenoanthracene 133, nor could photoproduct 153 be separated from that of ethenoanthracene 132.  Compounds 151 and 153 were assigned as regioisomers of  dibenzosemibullvalenes 152 and 154 respectively due to their similar chemical properties, and their structures were confirmed by NMR spectroscopy. Table 5.1 shows a comparison of part of the ^H and 31p NMR spectral data. Both photoproducts 151 and 152 show one phosphorus atom with a slight variation in its chemical environment. For dibenzosemibullvalene 152, the two protons #45 and //g^ couple with the phosphorus atom to give two doublets, indicating that the diphenylphosphinoyl group is at the Cgc-position. Unlike 152, dibenzosemibullvalene 151 gives a singlet (//4b) and a  Table 5.1. Chemical Shifts (ppm) and Coupling Constants (multiplicity, Hz) from ^H and 31p NMR Spectroscopy 151 P(0)Ph2 #4b (JPH> #8d(JPH) CH2CH3 (JHH) CH2CH3 (JHH)  152a  153  154^  37.56 36.28 30.13 29.27 - (s) 4.55 (d, 13) 4.88 (s) 4.55 (d, 12) 4.96 (d, 12) 4.35 (d, 11) 4.80 (d, 10) 4.43 (d, 9) 3.92, 3.93 (2 q, 7) 4.09, 4.10 (2 q, 7) 1.08 (t, 7) 1.17 (t, 7)  5 35  a  The structure was determined by X-ray structure analysis; "this compound was obtained by esterification  of compound 152.  doublet (//gd), and only the latter couples with the phosphorus atom, evidence that the diphenylphosphinoyl group is located at Cg^. The same trends are demonstrated in photoproducts 153 and 154 (Table 5.1). These two dibenzosemibullvalenes also show  151  different ethyl ester groups, which confirm their structural variations. A higher chemical shift of the ethyl ester group in dibenzosemibullvalene 154 than that corresponding to 153 indicates that 154 has its ester group attached to the Cgfo benzylic position and the diphenylphosphinoyl group attached Cgc, whereas a switch of these two groups occurs in 153. These chemical shift variations in ethyl ester group are know for other similar regioisomers such as dibenzosemibullvalenes 138 and 139 (see Experimental section for detail).^lt>,81c A c r o s s comparison of acids vs esters (151 vs 153 and 152 vs 154) further confirms the structures shown in Figure 5.4. Dibenzosemibullvalenes 152 (acid) and 154 (ester) show an almost identical chemical environment around H45 in their NMR, since the carbonyl group is further away from the C^-position in both molecules and is isolated by the bulky diphenylphosphinoyl group. In contrast, different chemical shifts are shown by H^ in dibenzosemibullvalenes 151 (acid, S 5.35) and 153 (ester, 5 4.88) as well as by //g<j in all four compounds (Table 5.1).  5.2 Medium-Dependent Regioselectivity The photolysis of ethenoanthracene 132 was carried out in different solvents as well as in the solid state, and a medium-dependent regioselectivity was observed as shown in Table 5.2. The regioselectivity was determined by both ^H and 31p NMR spectroscopy. Higher selectivity (> 70:30 154/153) was observed in protic solvents such as acetic acid, methanol or methanol/water.  Aprotic solvents gave a lower regioselectivity; for  example, a 55:45 154/153 ratio was obtained for the photolysis in acetone. Chloroform, which has a weakly acidic proton, affords a regioselectivity (65:35 154/153) between those of protic solvents and aprotic solvents.  The highest regioselectivity (> 80:20  154/153) was observed in the solid state photolysis (Table 5.2).  152  Table 5.2. Regioselectivities of Ethenoanthracene 132a regioselec tivity (%)  photolysis conditions media  time (h)  benzene acetone ethyl acetate chloroform  conversion (%)  154  153  57  2.5 2 4  100 100 100 100 91  43 45 42  19 2.5 18 19  100 100 49 60  4 2  acetic acid methanol methanol/water (3/2) solid state solid state  55 58 65 70 77  35 30 23 24 15 20  76 85 80  Conversion (+ 2%) and regioselectivity (± 2%) were measured by ^lp and *H NMR spectroscopy.  Due to its insolubility in most common solvents, ethenoanthracene 133 was photolysed only in three different solvents as well as in the solid state.  The  regioselectivity, which was obtained by 31p and ^H NMR spectroscopy, is shown in Table 5.3. Low selectivity was observed in the aprotic solvents DMSO and acetone, high selectivity in the solid state, and intermediate selectivity in chloroform.  Table 5.3. Regioselectivities of Ethenoanthracene 133a photolysis condition media DMSO-d 6 acetone chloroform solid state solid state  time (hours) 50 2.5 5 17 42  conversion(%) 82 100 100 94 100  regioselecth cities (%) 152  151  55 65 76 90 90  45 35 24 10 10  Conversion (± 2%) and regioselectivity (± 2%) were measured by 3lp and *H NMR spectroscopy.  153  5.3 Hydrogen Bonding and Regioselectivity The photolysis of ethenoanthracenes 132 and 133 shows that hydrogen bonding may be mainly responsible for the photochemical regioselectivity. In aprotic polar solvents, photoproduct 152 is slightly favored over 151 for the photolysis of ethenoanthracene 133; similarly, photoproduct 154 is slightly favored over 153 in the photolysis of ethenoanthracene 132 in aprotic solvents.  Assuming that the standard di-7i-methane  mechanism applies to both photoreactions,15a initial benzo-vinyl bridging on the phosphinoyl-bearing carbon (path b) is only slightly disfavored in comparison with the carbonyl-bearing carbon (path a, Figure 5.7). The results suggest that the stability of both type of radicals (151b vs 152a and 153b vs 154a) is comparable, if an electronically controlled rearrangement occurs (see Chapter 3). However, the steric effect may not be negligible due to the incomparable size of the ethyl ester and the diphenylphosphinoyl groups. Especially, if the phosphinoyl group is hydrogen bonded by solvent molecules, the steric effect would become a major factor to effect the regioselectivity.  R=H 151b R = CH2CH3 153b O II PPh2  R=H 151 R=CH2CH3l53 PPh, PPz8-  R=CH2CH3 132 R=H 133 R=H 152a R=CH2CH3 154a  R=H 152 R=CH2CH3 154  Figure 5.7. Di-7i-methane Rearrangement Mechanism of Ethenoanthracenes 132 and 133  154  Hydrogen bonding could produce a significant energy change due to the steric effect. The conformation analysis of cyclohexanol showed that hydrogen-donor solvents such as alcohols could stabilize the equatorial conformation of cyclohexanol, 164 s i n c e hydrogen bonding to axial oxygen was considered to be difficult due to the sterically encumbered environment. For example, a AG°25 difference of 0.6 kcal/mol was observed for 3,3,5trimethylcyclohexanol equilibrium in cyclohexane and in isopropanol (Figure 5.8). 164  H.C  solvent cyclohexane isopropanol  - AH° (kcal/mol)  - AS° (cal/deg mol)  - AG°25 (kcal/mol)  1.6 2.5  0.3 1.2  1.5 2.1  Figure 5.8. Thermodynamic Parameters for 3,3,5-Trimethylcyclohexanol Equilibrium.  In the photorearrangement of ethenoanthracenes, initial benzo-vinyl bridging on the diphenylphosphinoyl-bearing vinyl carbon (path b, Figure 5.7) would experience greater steric clashing to benzo group than that on the ester-bearing vinyl carbon (path a). Such steric interaction that path b involves could alleviate if the two phenyl groups point away from the bridging side with the phosphoryl oxygen atom pointing in (A, Figure 5.9). This situation could change dramatically when the phosphoryl oxygen is hydrogen bonded. The bridging would whether disrupt the hydrogen bond to avoid the steric interaction between the phenyl and the benzo groups (B, Figure 5.9), or experience the interaction (C, Figure 5.9). In either case, path b would be disfavored sterically, in other words, photoproduct 152 or 154 will be formed preferentially via path a (Figure 5.7).  155  H—R  Figure 5.9. Steric Effect in the Vinyl-benzo Bridging of Ethenoanthracenes 132 and 133.  Phosphinoyl group is a strong hydrogen bond donor, and it can form a weak hydrogen bond with chlorform.165 In the solution photolysis of ethenoanthracene 132, the hydrogen bonding effects are well demonstrated (Table 5.2).  The product ratio  154/153 is increasing from aprotic solvents (benzene, acetone and EtOAc), regardless whether they are polar or non-polar, weak proton donor (chloroform) to protic solvents (acetic acid, methanol and its mixture with water). A reversed situation is observed in the solution photolysis of ethenoanthracene 133, where the product ratio 152/151 is decreasing from DMSO, a comparable hydrogen bond acceptor with phosphine oxides, 1"5 acetone to chloroform.  X-ray crystal structure analysis shows that an  intramolecular hydrogen bond is present between the carboxylic acid group and the phosphinoyl group of ethenoanthracene 133 (Figure 5.13); we assume that this species (I, Figure 5.10) exists in chloroform, but it would dissociate partially to intermolecularly  (CH3V Y=0  (CH3)2-Y=0—H-O-C.  Y=C,S  Figure 5.10. Dissociation of Intermolecular Hydrogen Bond of Ethenoanthracene 133.  156  bonded species II in acetone. The equilibrium would be driven to species II when DMSO is used (Figure 5.10). For species I, the phosphinoyl group is hydrogen bonded intramolecularly and away from the bridging site; this would favor the formation of product 152.  From acetone to DMSO, the species I dissociate into II and the  phosphinoyl group is released, so is the steric interaction via path b. On the other hand, the hydrogen bonded carboxylic acid group of species II would probably increase the steric interaction via path a. In either circumstance, product 152 would no longer be formed as preferentially from species II as from species I. As the relative amount of species II increases from chloroform, acetone to DMSO, the product ratio 152/151 would decrease (Table 5.3). The species I, which is shown in Figure 5.10, may photorearrange via a mechanism of photochemical proton transfer followed by a 1,2 aryl shift (Figure 5.11).80,81 According to this mechanism, the proton is transferred intramolecularly from carboxylic acid group to phosphinoyl oxygen in the excited species I (Figure 5.10), and then the  152  ^  ^ ^  Figure 5.11. Excited State Proton Transfer and 1,2-Aryl Shift for the Intramolecular Hydrogen Bonded Species of Ethenoanthracene 133.  157  aryl shift would give only product 152 (Figure 5.11). Similar to species I, species B (or Q shown in Figure 5.9 may also rearrange via the same mechanism to give product 154, where a proton could be abstracted from donor solvents (Figure 5.12).  Since the  phosphinoyl group is a very strong hydrogen bond acceptor, an excited-state proton transfer to this group is most likely to occur. The mechanisms shown in Figures 5.11 and 5.12 avoid the steric interaction that the bulky diphenylphosphinoyl group involves in Zimmerman mechanism (Figure 5.7).  132 R-H--Q II  R—H^o  COOEt  ///  PPh7 £°&  Figure 5.12. Excited State Hydrogen Transfer and 1,2-Aryl Shift in the Intermolecular Hydrogen Bonded Species of Ethenoanthracene 132.  The amount of hydrogen bonded phosphinoyl species (I, B or Q are medium dependent, and so is the amount of their correspondent photoproducts (152, 154) via the proton transfer mechanism (Figures 5.11 and 5.12). We assume that the species with a "naked" phosphinoyl group undergo the normal di-7t-methane rearrangement to give two products (Figure 5.7).  Therefore, in the photolysis of ethenoanthracene 132, the  regioselectivity 154/153 would increase from aprotic solvent to protic solvent  158  (Table 5.2), as the amount of hydrogen bonded phosphinoyl species increases. On the other hand, the regioselectivity 152/151 would decrease from stronger proton acceptor (DMSO) to weaker one (CHCI3) (Table 5.3), as the amount of intramolecular bonded phosphinoyl group dissociated, which is shown in Figure 5.10.  5.4 Solid State Conformation and Regioselectivity The molecular and crystal structures of  ethenoanthracenes 132 and 133 were  analyzed by X-ray crystallography. The ORTEP diagrams are shown in Figure 5.13 (132) and Figure 5.14 (133).  Figure 5.13. ORTEP Diagram of Ethenoanthracene 132.  The electronic argument that has been discussed in Chapter 3 can be used to analyze the photorearrangement of ethenoanthracene 132 in the solid state. The X-ray structure of ethenoanthracene 132 shows that the carbonyl group is out of conjugation more than the phosphinoyl group (Table 5.4).  Assuming the di-rc-methane rearrangement  mechanism is applied in this case (Figure 5.7), an initial bond formation will occur at the carbonyl-bearing vinyl carbon to lead to the photoproduct 154, where a more stable  159  radical is formed, which is conjugated by the phosphinoyl group. On the other hand, a steric effect might be important here. As shown by the Ortep diagram in Figure 5.13, the diphenylphosphinoyl oxygen atom points away from either bridge side, and the initial bonding at phosphinoyl-bearing vinyl carbon would involve the steric interaction between phenyl and benzo groups. Therefore, path b to produce product 153 would be disfavored in the solid state. The predicted major regioisomer 154 from both electronic and steric factors agrees with experimental observation (Table 5.2).  Table 5.4. Torsion Angles of 0 = C - C = C and 0 = P - C = C of 132 and 133 ethenoanthracene  0>l(O:=C-C ==C)  cos^Oj  o2(o==P-C ==Q  cos2<3>2  132  77  0.051  133  -178  0.998  16 -8  0.928 0.980  The crystal structure of ethenoanthracene 133 (Figure 5.14) shows that its carboxylic acid and phosphinoyl groups form an intramolecular hydrogen bond and that both groups are well conjugated to the vinyl bond (Table 5.4).  The overwhelming formation of  Figure 5.14. ORTEP Diagram of Ethenoanthracene 133.  160  photoproduct 152 (Table 5.3) indicates that the mechanism of excited-state proton transfer and 1,2-aryl shift may be the major pathway followed in the solid state photolysis of ethenoanthracene 133.  The regioselectivity can also be rationalized  according Zimmerman mechanism shown in Figure 5.7.  As both acid carbonyl and  phosphinoyl groups are well conjugated to the vinyl bond, we could assume no obvious difference arise from the electronic effects for both pathways. But difference in the steric effects for both pathways could be seen from the ORTEP structure (Figure 5.14). The initial bridging would require the phenyl groups of the phosphinoyl substituent move toward benzo groups, thus path b would be disfavored, a similar situation that occurs in ethenoanthracene 132.  161  CHAPTER SIX ASYMMETRIC CATALYTIC HYDROGENATION  Asymmetric hydrogenation has been one of the most studied among the asymmetric organic reactions catalyzed by chiral transition metal complexes in particular, rhodium(I) complexes with chiral bisphosphine ligands have become very successful catalysts. However, there has been an increasing interest in ruthenium-bisphosphine catalysts.4 While the rhodium-based systems seem to be limited to the hydrogenation of dehydroamino acid type substrate, the ruthenium catalysts give good to excellent results with a much broader group of hydrogenation substrates, which are useful for a variety of organic syntheses.^ In both cases, optically active bisphosphine ligands have played an important role in enantioselective catalytic reactions. There are continuing efforts to develop newer, easier to synthesize and more efficient chiral ligands for asymmetric catalysts. Yet many ligands have been synthesized from the natural chiral pool, 166 a c c e ss to a wider spectrum of ligands is limited. Optically active bisphosphines and other chiral auxiliaries when they have Ci symmetry have been shown to be particularly effective in asymmetric inductions. 167 j C2  symmetric  bisphosphine  ligand,  anthraphos  nm  (136,  j s chapter, a new  9,10-dihydro-9,10-  ethanoanthracene-ll,12-bis(diphenylphosphine)), will be described as a chiral ligand for use in asymmetric hydrogenation.  6.1 Preparation of Possible Catalyst Precursors from Anthraphos Anthraphos was prepared and resolved as described in Chapter 1. From anthraphos, a cationic complex [Rh(COD)(anthraphos)]BF4 was prepared according to the equation  162  shown in Figure 6.1. Reaction of (ll/?,12/?)-(+)-anthraphos with [Rh(COD)Cl]2 in methanol followed by addition of NaBF4 solution afforded an 80% yield of crystalline (-)-[Rh(COD)((ll/U2#)-anthraphos)]BF4. This complex decomposed at about 250 °C without melting and exhibited [cc]^ = -45° (c = 0.5, CHCI3). The complex appears to be quite air-stable, showing no change in its 31p NMR spectrum over 4 days in CDCI3. Beautiful orange-red bipyramid crystals were deposited after recrystallization of this complex from methanol.  1. Methanol 2. NaBF4 (lU?,12/?>Anthraphos  H  IU  [Rh(OOD)Cl]2  136 [Rh(CODX(l l i U ^ A n t h r a p h o s ^ F -  Figure 6.1. Preparation of [Rh(COD)((l IR, 12#)-anthraphos)] +BF4~  The lH NMR spectrum of (-)-[Rh(COD)((ll/?,12/?)-anthraphos)]+BF4" in CDCI3 at room temperature shows clearly two types of vinylic protons at 8 5.01 and 4.32. Both are broad multiplets due to the complicated couplings, but a partially resolved triplet and a quartet can still be recognized for the signals at 8 5.01 and 4.32 ppm respectively. The four vinylic protons are assigned to the coordinated diene moiety, and the two types of protons are caused by the presence of (lli?,12/?)-anthraphos within the complex. The two coordinated phosphorus nuclei, which couple to rhodium (JpRh = 146.8 Hz), show a doublet at 8 30.00 in the 3 1 P NMR spectrum in CDCI3. Two ruthenium complexes with anthraphos were prepared by the following ligand exchange reaction (Figure 6.2). 168  Thus, rra«s-RuCl2((HS,125)-anthraphos)2 was 163  prepared from a 2:1 molar ratio of (115',125)-anthraphos and RuCl2(PPh3>3 in CH2CI2, while ?ra«s-Ru(H)Cl((HS,12,S)-anthraprios)2 was prepared from RuHCl(PPh3)3 in refluxing hexanes and in the presence of two equivalents of (ll^^^-anthraphos. Both complexes are air-stable in their solid state, but the solution of trans-RuHC\((llS,12S)anthraphos)2 is air-sensitive and slowly turns brown after exposure to air.  H  PPh2 + Rua2(PPh3)3  (llS,12S>Anthraphos  136 trans-RiiCl2((llS,12S)-Anthmphos)2  H  PPh2 Ph 2  2  <SSzt^p;  P  RuHCKPPh3)3  H  \  I /P  H S^l  (llS,12S>Anthraphos  Ph  ^2 a  2  ^P' ph 2  136 fra«s-RuHa((lLS,12S>Anthraphos)2  Figure 6.2. Preparation of Ruthenium Complexes with Anthraphos.  The 31p NMR spectrum of RuCl2((ll^,125)-anthraphos)2 at room temperature shows a singlet at 17.66 ppm, indicating the complex is almost certainly trans rather than cw. 168  The phosphorus-31 NMR spectrum of Ru(H)Cl((ll,S',125)-anthraphos)2 in  CDCI3 reveals two triplets (5 52.32 and 22.04, J = 33.4 Hz) as shown in Figure 6.3 and also the hydride gives a septet at -18.25 ppm in the ^H NMR spectrum (Figure 6.4). The  31  P and lH NMR spectra of Ru(H)Cl((ll£,12S)-anthraphos)2 resemble those of  164  „J  Vv*1k<**^<^mtlt>W'vt*iih*>i*1^^  ss.a  Figure 6.3.  so.rf  3l  4b. a  4a. B  ^S*N*w\w^W«r>»V*«i»*Y^W  I*,.! PPM  isa.a  25.a  2a. a  is.a  P NMR Spectrum of mw«-Ru(H)Cl((l 15,125)-anthraphos)2.  W  "'Jl,  •-f^\'W*wt'{*"M*''  r'  • ' •  13.03  -13.53  Figure 6.4. ! H NMR Spectrum of the Hydride in trans-RuRC\((l 15,125)-anthraphos)2.  165  rran5,-Ru(H)Cl(diop)2, whose structure has been analyzed by X-ray crystallography and shown to be a distorted octahedron. 169  Therefore, we concluded that trans-  Ru(H)Cl((ll1S',125)-anthraphos)2 is probably obtained under the same reaction conditions as those used for the preparation of ?ra/w-Ru(H)Cl(diop)2. The two types of phosphorus atoms in fra«s-Ru(H)Cl((ll£,12iS)-anthraphos)2 give two sets of triplets in its 31p NMR spectrum. Because the coupling constant between the one type of phosphorus and the hydride is accidentally twice that between the other type of phosphorus and the hydride, the resulting ^H NMR spectrum of the hydride will be a septet with an intensity ratio of 1:2:3:4:3:2:1 (Figure 6.4).  The coordinated hydride can also be seen from the  stretching band at 1986 cm~l in the IR spectrum of frww-Ru(H)Cl((HS,12£)anthraphos)2Attempts to prepare Ru[r|3-CH3C(CH2)2]2(anmraPnos) failed to give any pure compound by following the literature method,"^ which is shown in Figure 6.5.  [Ru(OOD)a2]2  +  ph 2 CHs—^ N CH 2 M g a  RU(COD)[T 1 3 -CH3C(CH 2 ) 2 ] 2  +  2Anthraphos  THF/S 2 O +>  RU(COD)[T 1 3-CH3C(CH 2 ) 2 ] 2 81%  »> R u ^ - ^ Q C H ^ H A n t h r a p h o s ) reflux24h  Figure 6.5. Attempted Preparation of Ru[r|3-CH3C(CH2)2]2(anmraPh°s)  6.2 Asymmetric Hydrogenation of Z-(a)-Acetamidocinnamic Acid The asymmetric hydrogenation of (Z)-(a)-acetamidocinnamic acid (155) has been extensively investigated (Figure 6.6), and relatively high enantioselectivities have been achieved by using chiral bisphosphine-Rh catalysts.4 Thus enamide hydrogenation has become a standard test reaction for new chiral phosphine ligands.  166  OOOH /7=A  NHCCH,  H, catalyst  155  156  Figure 6.6. Catalytic Hydrogenation of Z-(oc)-acetamidocinnamic Acid (155).  Two ligands, anthraphos and bisphosphine 144 (Figure 6.7) were evaluated by using the standard substrate 155. For anthraphos, the hydrogenation of 155 was catalyzed in methanol by the isolated catalyst, (-)-[Rh(COD)((ll#,12J?)-anthraphos)]BF4. A catalyst was prepared in situ from bisphosphine 144 by mixing [Rh(COD)Cl]2 with a slight excess of bisphosphine 144 in methylene chloride, and the catalytic reaction was then carried out in methanol/methylene chloride.  The hydrogenated product N-acetyl-  phenylalanine (156) was derivatized to its methyl ester 157 by diazomethane (Figure 6.8), and the optical purity was determined by chiral HPLC (Chiracel OD, see Chapter 2 for details).  H  pa,  PPh,  136 144  anthraphos  PPh,  PPh2 noiphos  Figure 6.7. Bisphosphine Ligands  167  CH2N,  Figure 6.8. Diazomethane workup of A/-acetylphenylalanine (156)  A non-optimized enantiomeric excess of 90% was observed in the hydrogenation of Z-(a)-acetamidocinnamic acid (155) by the anthraphos complex (Table 6.1).  This  catalytic reaction was carried out at an initial pressure of 4 atm in a Schlenk tube. Both high and low ratios of substrate to catalyst were tried, and no obvious change was observed in the enantioselectivity.  During the hydrogenation, the orange complex  dissolved slowly to give a red-orange solution. This solution remained orange until all the  substrate  was  hydrogenated  and  then  turned  straw-yellow.  (S)-(+)-  Acetylphenylalanine was obtained from the substrate by using [Rh(COD)((ll/?,12#)anthraphos)]BF4. The enantioselectivity obtained by anthraphos is comparable to those obtained by using chiraphos (89% e.e., Figure 6.7)^7 and norphos (95% e.e., Figure 6.7). 1 7 0  Table 6.1. Asymmetric Hydrogenation of 155 in Methanol Catalyzed by [Rh(COD)((l IR, 12/?)-Anthraphos)]BF4 substrate/Rh(I)a 203 1000 a  conditions*3 4 atm (initial), 2.5 h 4 atm (initial), overnight  e.e. %c 89.4 (S) 89.5 (S)  Molar ratio; "hydrogenation was carried out at room temperature to reach a complete  conversion, which was determined by *H NMR spectroscopy; HPLC.  168  c  determined by chiral  Bisphosphine 144, whose structure is shown in Figure 6.7, is a cw-bisphosphine. This is the first time that a ds-bisphosphine has been examined for asymmetric hydrogenation. A dimeric precursor of the catalyst was assumed to be prepared in situ according to the equation shown in Figure 6.9. This bisphosphine complex gave much lower enantioselectivities than anthraphos in the hydrogenation of dehydroamino acid 155 (Table 6.2).  The in situ catalyst also showed a lower catalytic activity than  "normal" rhodium-bisphosphine based catalysts; even at a low turnover number (substrate to catalyst), complete conversion was only observed in high pressure hydrogenation. Interestingly, high pressure hydrogenation gave higher enantioselectivity than low pressure.  [g*Ogi • ^ p  p  — (XX)  = bisphosphine 144  Figure 6.9. Preparation of the Dimeric Pre-catalyst in CH2CI2.  Table 6.2. Asymmetric Hydrogenation of 155 by Using an in situ Catalyst from Bisphosphine (4bi?,8W?,8c/?,8d/?)-144 and [Rh(COD)Cl2]2 144/Rh (I) a  substrate/Rh(I)a  conditions'3  conversion (%)c  e.e.%d  1.08 1.06  98 92  19 atm, 24 h 4 atm (initial), 20 h  100 70  34(5) 22 (S)  a  Molar ratio; "hydrogenation was carried out at room temperature and in methanol; cdetermined by *H  NMR spectroscopy; ^determined by chiral HPLC.  Bisphosphine 144 turns out not to be an effective ligand with Rh for asymmetric hydrogenation.  Unlike most of the effective bisphosphine ligands, bisphosphine 144  169  does not have C2-symmetry, which may be the main reason why relatively low enantioselectivity was observed in the hydrogenation of dehydroamino acid 155.167 Bisphosphine 144 was obtained via absolute asymmetric induction (Chapter 4), and here it was used as a ligand for asymmetric catalytic reactions.  These two steps  demonstrate how optically active compounds can be produced from achiral materials and multiplied via a catalytic reaction, a key argument of the origin of asymmetry in life and its amplification. 171 The  two  ruthenium  complexes,  RuCl2((ll£,12iS)-anthraphos)2  and/or  Ru(H)Cl((115,125)-anthraphos)2, showed no detectable catalytic activities in the hydrogenation of dehydroamino acid 155 (C=C), acetophenone (C=0) and A/-phenyl-2phenylethylimine (C=N) at room temperature and at an elevated pressure (up to 1000 psi). The lack of catalytic activity of Ru(H)Cl((l IS, 125)-anthraphos)2 may be attributed to the presence of a five-membered chelate ring. Such a chelate ligand is less likely to dissociate in solution (to give a monodentate bisphosphine moiety) than the sevenmembered one as in Ru(H)Cl(diop)2, which is an active catalyst for the asymmetric hydrogenation of alkenes. 168  6.3 X-ray Structure Analysis of (-)-[Rh(COD)(anthraphos)]BF4 The molecular and crystal structure of (-)-[Rh(COD)(anthraphos)]BF4 was analyzed by X-ray crystallography. The ORTEP diagram of the complex is shown in Figure 6.10 (Space group, P4y, a = 10.200(7) A, c = 39.97(5) A, V = 4158(3) A 3 , Z = 4, D c a l c = 1.394 g/cm3; R = 0.064). The R value for the enantiomorphous space group P43 of (ll£12S)-anthraphos was 6.53%. With AR = 0.9%, (-)-[Rh(COD)-(anthraphos)]BF4 was concluded to have the (11/?, 127?)-configuration ligand according to the Bijvoet method. 145 The Rh atom is coordinated in a distorted square-planar geometry and its chelate phosphine ring adopts the X conformation, which presumably accounts for the  170  product of (S) absolute configuration upon hydrogenation of (Z)-a-acetamidocinnamic acid (155). 172  Figure 6.10. ORTEP Diagram of (-)-[Rh(COD)((l 1R, 12i?)-anthraphos)]+  With the available crystal structures of [Rh(NBD)(norphos)]ClO4170b and [Rh(COD)(chiraphos)]C104,173 a comparison of them with (-)-[Rh(COD)((ll/?,12/?)anthraphos)]BF4 is made in Table 6.3. In anthraphos, the PCCP dihedral angle of the five-membered chelate ring is 61°, which is between that of norphos (64°) and chiraphos (52°), but closer to the former. Similarly, the P-Rh bond lengths in anthraphos (2.31  Table 6.3. Comparison of Structure of Diphosphine Complexes Rh(COD)(anthraphos) Rh(COD)(chiraphos) Rh(NBD)(norphos) PCCP dihedral angle (°) P-Rh bond lengths (A) C-P-Rh angles (°) C=C length of diene (A)  61  52  64  2.31,2.29  2.28, 2.27  2.32, 2.32  105, 106 1.31, 1.37  110, 110 1.36, 1.36  103, 104 1.38, 1.36  171  and 2.29 A) lie between those of norphos (2.32 and 2.32 A) and chiraphos (2.28 and 2.27 A), and the C-P-Rh angles in anthraphos are, 105° and 106°; in norphos, 103° and 104°; and in chiraphos 110°, 110°. The chelate ring of the Rh(I)-anthraphos complex is, therefore, somewhat less strained than that of the norphos complex, a finding that presumably reflects the slightly greater flexibility of the bicyclo[2.2.2]octadiene carbocyclic system of anthraphos compared to the bicyclo[2.2.1]heptene framework in norphos.  The two coordinated C-C double-bond lengths are 1.31 and 1.37 A in  Rh(COD)(anthraphos); only one is close to those in norphos (1.36 and 1.38 A) and chiraphos (1.36 and 1.36 A). This indicates that the 1.31 A double bond portion of the COD ligand within the anthraphos complex is hardly coordinated to Rh(I); or if any, the coordination is more weakly than the other and those in the norphos and chiraphos analogs. The transfer of chirality from the catalyst to the substrate during the hydrogenation process within the chiral bis(diphenylphosphino) ligand system has been considered to arise as a consequence of the chiral array of the four phenyl groups, which occupy pseudo axial and pseudo equatorial positions on the chelate. 1^4 The substituents on the supporting carbon framework enforce a chiral conformation which in turn generates the chiral arrangement of phenyl groups, an approximately C2 symmetrical edge-face array as shown Figure 6.11.  Figure 6.11. An Approximately C2 Symmetrical Edge-Face Arrangement of Phenyl Groups in Rh(I)-Bisphosphine Complexes.  172  Figure 6.12 shows the perspective view of (-)-[Rh(COD)(ll/?,12/?-anthraphos)]BF4 (B), [Rh(NBD)CS,S-norphos)]C104 (C) and [Rh(COD)(5,5-chiraphos)]C104 (A), along with the bisector of the P-Rh-P bond. Note that B is a mirror image of A or C in terms of the -CH(PPh.2)-CH(PPh2)- moiety because OR,Z?)-anthraphos is used and (S,S)-isomers are shown for both norphos and chiraphos.  A  B  Like the other two complexes,  C  Figure 6.12. Perspective View of [Rh(COD)(S,S-chiraphos)]+ (A), [Rh(COD)((ll/U2/?)-anthraphos)]+ (B) and [Rh(NBD)(5,5-norphos)]+ (C).  (-)-[Rh(COD)((l 1/?, 12/?)-anthraphos)]BF4 possesses an approximately C2-symmetrical edge-face array of the four phosphine phenyl rings. Its pseudo-axial and equatorial positions are the most pronounced among the three. The diene ligand in the complex of anthraphos is rotated away from an orientation that would place the double bonds perpendicular to the PRhP plane, in a way almost the same as that of [Rh(COD)(S, Schiraphos)]C104 (A in Figure 6.12). 173  EXPERIMENTAL  174  I. GENERAL PROCEDURES Infrared Spectra (IR). Infrared spectra were recorded on a Perkin Elmer 1710 Fourier transform spectrometer. The positions of the absorption maxima (Vmax) ^"^ reported in cm~l.  Solid samples were ground in KBr (1-5%) and pelleted in an  evacuated die (Perkin Elmer 186-0002) with a laboratory press (Carver, model B) at 17,000 psi. Liquid samples were run neat as a thin film between two potassium bromide plates. Melting Points (m.p.). Melting points were determined on a Fisher-Johns hot stage apparatus and are uncorrected. Nuclear Magnetic Resonance Spectra (NMR). Proton nuclear magnetic resonance spectra ( ! H NMR) were recorded in deuterated chloroform and other deuterated solvents as noted. The spectrometers used were a Bruker AC-200 (200 MHz) and a Bruker WP400 (400 MHz).  Signal positions (d) are given in parts per million (ppm) with  tetramethylsilane as an internal or external reference. The number of protons, signal multiplicities, coupling constants in Hertz (Hz) and assignments are given in parentheses following the signal position. The multiplicities of the signals have been abbreviated as follows: s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet and br. = broad. Carbon nuclear magnetic resonance (^C NMR) spectra were recorded at 50.3 MHz on a Bruker AC-200, at 75.4 MHz on a Varian XL-300 and at 125 MHz on a Bruker AMX-500 spectrometers using deuterated solvents as indicated; the middle peak of the solvent signal was used as internal reference. All spectra were run under broad band 13C~{1H}  decoupling. Chemical shifts (d) are reported in ppm. Assignments, where  given, were supported by APT (attached proton test) ^C NMR spectra.  175  Phosphorus nuclear magnetic resonance (31p NMR) spectra were recorded at 81 MHz on a Bruker AC-200 instrument using an external reference of 85% phosphoric acid. All spectra were run under broad band 31p-{lH} decoupling. Chemical shifts (d) are given in ppm. Assignments, where made, are included in parentheses. Mass Spectra (MS). Low and high resolution mass spectra were obtained from a Kratos MS 50 instrument operating at 70 eV. Fast atom bombardment (FAB) mass spectra were recorded on an AEI MS-9 mass spectrometer with xenon bombardment of a 3-nitrobenzyl alcohol matrix of the sample. Mass to charge ratios (m/e) are reported with relative intensities in parentheses. Molecular ions are designated as M + . Ultraviolet and Visible Spectra (UV-Vis). Ultraviolet and visible spectra were recorded on a Perkin Elmer Lambda-4B UV/Vis spectrometer in the solvents indicated. Wavelengths (1) in nanometers (nm) are reported, and extinction coefficients (e) are given in parentheses. Spectral grade solvents available from BDH or Fisher were used without further purification. Microanalyses.  Elemental analyses were performed by the departmental  microanalyst, Mr. Peter Borda. Gas Liquid Chromatography (GLC). Gas chromatographic analyses were run on a Hewlett-Packard 5890 A gas chromatograph, using the following 15 m X 0.25 mm fused silica capillary columns: DB-1, DB-5, DB-17 (J & Scientific Inc.) or 20 m x 0.21 mm Carbowax 20 M (Hewlett Packard) and with a column head pressure (carrier gas: helium) of 15 psi unless specified.  The signal from a flame ionization detector was  integrated by a Hewlett-Packard 3392 A integrator without response correction. Silica Gel Chromatography. Analytical thin layer chromatography was performed on commercial pre-coated silica gel plates (E. Merck, type 5554) and the plates were developed in the indicated solvent system. The developed plates were observed with UV light. Column chromatography was carried out by using 230-400 mesh size silica gel  176  (Merck 9385) slurry packed with the eluting solvent.  Column size was determined  according to the amount of crude material to be purified. High Pressure Liquid Chromatography (HPLC). High pressure liquid chromatography was performed on a Waters 600E system controller connected to a tunable absorbance UV detector (Waters 486) or programmable photodiode array detector (Waters 994). Chiral columns (Chiralcel OD, 250 mm x 4.6 mm; Chiralcel OP(+), 250 mm x 4.6 mm; Chiral Technologies Inc.) were used to determine enantiomeric excesses (optical purities). X-ray Analyses. All X-ray crystal structures were determined on a Rigaku AFC6S 4-circle diffractometer using single crystal X-ray analysis by Dr. Tai Y. Fu, Dr. Ray Jones, Dr. Bozena Borecka and Dr. Gunnar Olovsson in Professor James Trotter's laboratory in the UBC Chemistry Department. Stereoscopic diagrams were drawn with a locally modified version of the ORTEP program at a 50% probability level. Optical Rotations.  Optical rotations were measured on a Jasco-J710/ORD-M  instrument at room temperature. Specific rotation, [a]^, was calculated by the following equation:  where  t is the temperature at which the optical rotation was measured D is the sodium D line, 589 nm a (+ 2) is the recorded optical rotation in degrees / is the path length in decimeters c is the concentration of the sample solution in grams per 100 ml.  Differential Scanning Calorimetry (DSC) and Thermogravimetric Analysis (TGA). DSC and TGA were performed on a TA instruments Thermal Analyst 2000 thermal analyzer equipped with a DSC 91 OS differential scanning calorimeter and a TGA 51 thermogravimetric analyzer.  Data analyses were done on an IBM PS/1 personal  177  computer connected to the above instrument running DSC Calibration Data Analysis Program Version 5.0. Enthalpies (AH) from DSC are reported in J/g with specified temperature ranges. TGA results are reported as percent weight losses with T\, T2 and T m a x as the temperatures at the start, end and maximum points. Solvents and Reagents. Spectral grade solvents were used for spectroscopic and photochemical studies. Further purification was not carried out unless otherwise noted. Reagents for the syntheses of starting materials were purified according to methods reported in the literature as indicated.  178  H. SYNTHESIS OF STARTING MATERIALS l,2-Ethanediylbis(diphenylphosphine oxide), DPPEO (101)175 Ethanediylbis(diphenylphosphine) (5.0 g, 1.25 mmol, Aldrich) in 40 ml of acetone was placed in a 150 ml round-bottom flask. To it 10 ml of 30% H2O2 in 40 ml of acetone was added slowly. After addition was complete, the solution was refluxed for 1 h. The solid compound (5.0 g, 93%) was obtained after removing acetone. Purer product could be obtained by recrystallization of the crude product from a mixture of benzene and ethanol, m.p. 271-273 °C; lit.1 m.p. 266-268.5 C. iH NMR (CDCI3, 200 MHz): 5 7.12-7.78 (20 H, m, Ar-H); 2.50 (4 H, br. s, PCH2). 31  P NMR (CDCI3, 81 MHz): 8 32.43.  IR (KBr): v m a x , 1439, 1188 (P=0), 1176 (P=0), 1123, 764, 755, 742, 731, 696, 534,512 cm"1.  Dimethyl 9,10-Dihydro-9,10-ethenoanthracene-l 1,12-dicarboxyIate (102)" Anthracene (5.00 g, 28.1 mmol, Eastman Organic Chemicals) and dimethyl 2-butyne-l,4-dioate (4.30 g, 30.3 mmol, Aldrich) were placed in a 50 ml round-bottom flask. The mixture was heated at 190-200 °C for 2.5 h in an oil bath. The flask was cooled down to room temperature and the brown solid was recrystallized from a mixture of chloroform and ethanol to yield 6.5 g (20.3 mmol, 72% yield) of the pure compound, m.p. 160-161°C; lit. 176 m.p. 160-161 °C. iH NMR (CDCI3, 200 MHz): 5 7.48-6.98 (8 H, m, Ar-H); 5.50 (2 H, s, H9 and H10); 3.80 (6 H, s, -COOCH3).  179  13  C NMR (CDCI3, 50 MHz):  8 165.95 (C=0);  147.04, 143.80, 125.46 and  123.85 (Ar-C and vinylic-C); 52.48 (COOCH3); 52.41 (C9 and CIO).  9,10-Dihydro-9,10-ethenoanthracene-ll,12-dicarboxylic acid (103)99 Diester 102 (6.5 g, 20.3 mmol) was suspended in 42.5 ml of 30% NaOH and 25.5 ml of ethanol. The mixture was refluxed for 15 h. The final solution was cooled to room temperature and then washed with diethyl ether (2 x 100 ml) to eliminate traces of unreacted starting material. The aqueous fraction, containing the acid salt, was acidified by dropwise addition of concentrated HCl while stirring in an ice-water bath. The diacid was extracted with diethyl ether (3 x 100 ml). The combined organic fractions were then dried over anhydrous sodium sulphate and evaporated to dryness. The diacid 103 (5.4 g, 91% yield) was obtained as a white solid after recrystallization from a mixture of petroleum ether and acetone, m.p. 239-241 °C; lit. 1 7 7 m.p. 215-216 °C. The melting point of the compound obtained is different from its literature value, but the following NMR spectra show it is compound 103, indicating a dimorph formed in this case. ! H NMR (CD3OD, 200 MHz): 5 7.48-6.97 (8 H, m, Ar-tf); 5.63 (2 H, s, H9 and H10). 13  C NMR (CD3OD, 50 MHz): 8 165.89 (COOH);  146.39, 142.44, 123.47 and  121.77 (Ar-C and vinylic-C); 51.24 (C9 and C10).  9,10-Dihydro-9,10-ethenoanthracene-ll,12-dicarboxylic Acid Anhydride (104) To a suspension of 7.9 g (27 mmol) of diacid 103 in 75 ml of CH2CI2, 6.7 g (52.8 mmol) of oxalyl chloride (Aldrich) in 10 ml of CH2CI2 was dropped in slowly with stirring. The mixture was refluxed for 24 h whereupon a homogenous yellow solution was obtained. The solvent and excess oxalyl chloride were evaporated in vacuo and the  180  resulting solid was recrystallized from ethyl acetate to give light brown prisms. Yield: 6.3 g, (83%), m.p. 253-255 °C; l i t . " m.p. 247 °C. *H NMR (CDC13, 200 MHz): 5 7.04-7.56 (8 H, m, Ar-H), 5.56 (2 H, s, H9 and H10).  9,10-Dihydro-9,10-ethenoanthracene-ll,12-dicarboxylic Acid, 11-Ethyl Ester (105) A suspension of 3.20 g (10 mmol) of anhydride 104 in 30 ml of ethanol, freshly distilled over CaH2, was refluxed for 2 h to result in a homogenous solution. The reflux was continued for another 4 h. The excess of alcohol was evaporated in vacuo and the solid so obtained was recrystallized from acetonitrile to yield 3.34 g (89%) of colorless crystals, m.p. 220-221 °C; lit. 8 1 c m.p. 224-226 °C. *H NMR (CDCI3, 200 MHz): 5 7.34-7.00 (8 H, m, Ar-H); 6.18 and 5.85 (2 H, 2 s, H9 and H10); 4.48 (2 H, q, J = 7 Hz, -COOCH2CH3); 1.50 (3 H, t, J = 7 Hz, COOCH 2 C#3). IR (KBr): v m a x , 3400-2200 (OH of acid and C-H), 1734 ( C = 0 of acid), 1678 ( C = 0 of ester), 1629 (C=C), 1460, 1282, 1221 cm" 1 . MS: m/e (relative intensity), 320 (3.1, M + ) , 276 (22.5), 247 (16.1), 230 (13.5), 219 (8.0), 203 (100), 178 (44.5), 151 (7.7), 101 (29.8), 88 (13.0), 75 (7.6).  9,10-Dihydro-9,10-ethenoanthracene-ll,12-dicarboxylic Acid, ll-(Methylethyl Ester (106) Anhydride 104 (2.9 g, 10.6 mmol) in 30 ml of isopropanol was refluxed for 2.5 h to reach a homogenous solution and continued for another 2.5 h. The excess alcohol was evaporated in vacuo, and the solid so obtained was recrystallized from 10 ml of acetonitrile to yield 3.0 g (85%) of large prisms, m.p. 181-183 °C; lit. 8 0 m.p. 176177 °C. 181  l  U NMR (CDCI3, 200 MHz): 8 7.52-7.00 (8 H, m, Ar-tf); 6.19 and 5.82 (2 H, 2  s, H9 and H10); 5.22 (1 H, hep, J = 7 Hz, -COOC#(CH 3 ) 2 ); 1.42 (6 H, d, J = 7 Hz, COOCH(CHj) 2 ). IR (KBr):  v m a x , 3400-2200 (OH of acid and C-H), 1723 ( C = 0 of acid), 1683  ( C = 0 of ester), 1624 (C=C), 1474, 1460, 1420, 1321, 1282, 1225, 1155, 1101, 1043, 755 cm" 1 . MS: m/e (relative intensity), 334 (5.0, M + ) , 316 (2.2), 290 (26.7), 275 (4.1), 248 (40.6), 230 (7.9), 203 (100.0), 178 (31.0), 153 (6.1), 101 (15.3), 88 (7.1), 43 (20.0).  (R)p and (5)p-Menthyl Methylphenylphinates (/?p-107, S p -107) 8 7 A mixture of two diastereomers was prepared according to Mislow's method. 87 Pyridine (475 ml, distilled from barium oxide), phenyldichlorophosphine (500 g, Aldrich) and hexanes (1500 ml) were combined in a 5 L three-neck flask equipped with a mechanical stirrer, and a solution of methanol (225 ml) and hexanes (80 ml) was added dropwise under nitrogen and at 0 °C over a period of 2 h.  After stirring for one  additional hour, the pyridine hydrochoride was removed by suction filtration under a stream of dry nitrogen, and the filtrate was concentrated on a rotary evaporator. residual dimethyl phenylphosphonite was not further purified.  The  A small amount of the  crude phosphonite was added to a few drops of methyl iodide which were contained in an 1 L three-neck flask equipped with a dropping funnel, reflux condenser, thermometer, nitrogen inlet, and magnetic stirrer, and the mixture was warmed until a violent exothermic reaction began (extreme caution!). The phosphonite was then added at a rate sufficient to maintain the temperature at about 100 °C. periodically to ensure a continuous reaction.  Methyl iodide was added  Stirring continued overnight, and the  reaction mixture was distilled at 104-110 °C in vacuum (1 mm) to yield 225 g of methyl methylphenylphosphinate (lit. 87 b.p. 94 °C at 0.05 mm; lit. 1 7 8 b.p. 119 °C at 3.5  182  mm). The phosphinate (218 g) was dissolved in carbon tetrachloride (1000 ml) and phosphorus pentachloride (271 g) was added under nitrogen at a rate sufficient to maintain the temperature at 40 °C. The mixture was stirred overnight, solvent was removed, and the residue was distilled, b.p. 122-126 °C (1 mm), to afford 199 g of methylphenylphosphinoyl chloride (lit. 87 b.p. 105-110 °C at 0.05 mm; lit. 179 b.p. 155 °C at 11 mm). The acid chloride (154 g, 0.88 mol) was dissolved in anhydrous ether (150 ml), and the solution was added slowly to a solution of pyridine (68.5 g) and (-)-menthol (130 g, 0.83 mol; [<x]$ = -50°, c = 10, ethanol; Aldrich) in ether (200 ml).  Stirring continued overnight, the mixture was filtered to remove pyridine  hydrochloride, and the filtrate was evaporated to afford a colorless oily mixture of Rpand Sp-107 (293 g in total). The two isomers, Rp and Sp, were separated by combining selective seeding with fractional recrystallization from an approximately 15% solution in hexanes at 5 °C. Rp107 was grown as brick-like crystals and Sp-107 as needles. Separation was done by hand. The pure seeds were obtained by fractional recrystallization from hexanes.87 Typically, 293 g of the crude oily mixture was dissolved in hexanes, washed with sodium carbonate and dried over sodium sulphate and magnesium sulphate; 112 g of Sp and 92 g of Rp were separated out as blocks and needle crystals respectively. After three recrystallizations from hexanes, 85 g of Sp-107 and 76.5 g of/?p-107 were obtained. *H NMR (CDCI3, 200 MHz) showed both compounds were pure within the limits of NMR spectroscopic detection (+ 1%).89 Sp-107: m.p. 79.5-80.2 °C, [a]2D° = -92° (c = 1.18, benzene); lit. 87 m.p. 7980 °C, [a]D = -94° (c = 1.0-3.0, benzene, 23-24 °C). *H NMR ( CDCI3, 200 MHz): 89 5 7.90-7.40 (5 H, m, Ar-H), 3.98 (1 H, m, OCH), 2.38 (1 H, m, cyclohexyl-tf), 1.92 (1 H, m, cyclohexyl-//), 1.72 (3 H, d, J PH = 17 Hz, PCH3), 1.60 (2 H, m, cyclohexyl-tf), 1.25 (3 H, m, cyclohexyl-tf and Ci/(CH3)(CH3)), 0.90 (3 H, d, J HH = 7 Hz, -CH3), 0.88 (2 H, m, cyclohexyl-^, 183  0.80 (3 H, d, J H H = 7 Hz, CH(C#3)(CH 3 )), 0.30 (3 H, d, J H H = 7 Hz, CH(CH 3 )(Cffj)). flp-107: m.p. 88-89 °C, [a]2D° = -13.5° (c = 5.82, benzene); lit. 8 7 m.p. 89 °C, [a]$ = -16° (c = 1.0-3.0, 23-24 °C). *H NMR (CDC13, 200 MHz): 8 9 5 7.90-7.40 (5 H, m, Ar-fl), 4.22 (1 H, m, OCH-), 2.19 (1 H, m, cyclohexyl-#), 1.80 (1 H, m, cyclohexyl-#), 1.60 (3 H, d, J P H = 17 Hz, PCH3), 0.98, 0.88 (6 H, 2 d, J H H = 7 Hz, CH(CH3)2),  1.90-0.90 (7 H, a  complex of multiplets, cyclohexyl-// and CH(CH3)2).  (S)-(-)-Methyl(4'-methylphenyl)phenylphosphine Oxide (5-108) 1 8 ° In a 100 ml round-bottom flask, /?-methylphenylmagnesium bromide was prepared from 2.5 g (0.11 mol) of magnesium and 15 g (0.088 mol) of 4-bromotoluene (Aldrich) in 40 ml of anhydrous diethyl ether under nitrogen. To this Grignard reagent, 5.9 g (0.020 mol) of (>S)p-menthyl methylphenylphosphinate (Sp-107) in 20 ml of benzene was added, and diethyl ether was boiled off until the batch temperature reached 85 °C. Additional benzene was then added to make up the total volume of the reaction solution to about 50 ml. The batch was refluxed under nitrogen overnight, cooled in an ice bath and poured into 80 ml of saturated aqueous solution of ammonium chloride.  The  organic phase was separated and the aqueous phase was extracted three times with 40 ml of chloroform each time. The combined fractions were dried over sodium sulphate and evaporated. The residual oil was submitted to silica gel chromatography (silica gel 60, 230-400 mesh). Menthol and other impurities were eluted with benzene, and the product was obtained with 2% ethanol in chloroform. The crude product (4.3 g, 93% yield) was obtained after removing the solvents. Recrystallization from a mixture of benzene and hexanes yielded 2.9 g of pure product, m.p. 114-115 °C; lit. 1 8 0 m.p. 118-119 °C. Anal. Calcd. for C 1 4 H 1 5 OP: C, 73.02; H, 6.57. Found: C, 73.30; H, 6.64.  184  ! H NMR (CDCI3, 200 MHz): 8 7.70-7.12 (9 H, m, Ar-fl), 2.30 (3 H, d, J P H = 1.2 Hz, Ar-CH3), 1.98 (3 H, d, J P H = 13.2 Hz, PCH3). ! 3 c NMR (CDCI3, 50 MHz): 8 15.86, 17.33 (d, J PC = 70.7 Hz, PCH3); 21.47, 21.49 (unresolved doublet, J P C = 1.2 Hz, Ar-CH3). Methyl substituted phenyl ring, 8 129.65, 131.71 (d, J P C = 103.7 Hz, C-l); 130.37, 130.57 (d, J PC = 10.1 Hz, C-2 and C-6); 129.19, 129.44 (d, J P C = 12.2 Hz, C-3 and C-5); 142.10, 142.16 (d, J PC = 2.7 Hz, C-4). Phenyl ring, 8 133.35, 135.36 (d, J PC = 101.2 Hz, P-C); 130.28, 130.48 (d, J PC = 9.7 Hz, o-C); 128.40, 128.64 (d, J PC = 11.8 Hz, m-C); 131.53, 131.58 (d, JPC = 2.7 Hz, p-C). See Appendix I for assignments. 31  P NMR (CDCI3, 81 MHz): 8 30.03.  MS: m/e (relative intensity), 230 (35.1, M + ) , 229 (87.8), 215 (100), 167 (7.7), 153 (14.9), 139 (10.6), 121 (7.7), 114 (13.3), 107 (3.9), 91 (53.2), 77 (44.8), 65 (28.7). Calculated mass for C14H15OP: 230.0861. Found: 230.0853. [a]24 = . 7 9 ° (C  =  7 06, methanol); [cc]2D4 = -14.4° (c = 11.7, benzene);  lit. 1 8 0  [a]D = -8.6° (c = 7.0, methanol). Optical purity: > 99% e.e.; HPLC, Chiralpak OP(+), 25 cm X 0.46 cm I.D.; eluent, methanol, flow rate of 0.45 ml/min; detector, UV at 254 nm.  (/?)-(+)-Methyl(4'-methylphenyl)phenylphosphine Oxide (R-10S) The i?-isomer was prepared following the same procedure as the ^-isomer, beginning with (tf)p-menthyl methylphenylphosphinate /?p-107, m.p. 108-110 °C.  l  H NMR,  31  P  NMR and IR spectra of the /?-isomer were the same as those of the 5-isomer respectively. [a]2D4 = 15.7° (c = 7.58, benzene). Optical purity: 88% e.e. by HPLC; base line separation for the enantiomers was obtained. This enantiomer was not further purified to a higher optical purity and was  185  used for the preparation of its dimer as the reaction would enrich the optical purity by itself.  (S)-(+)-Methyl(3'-methylphenyl)phenylphosphine Oxide (S-109) A solution of (5)p-menthyl methylphenylphosphinate (Sp-107, 5.31 g, 18.1 mmol) in benzene (25 ml) was added to m-methylphenylmagnesium bromide in diethyl ether, which was prepared from m-bromotoluene (9.7 g, 56.7 mmol, Aldrich) and magnesium (1.53 g, 62.8 mmol) in anhydrous ether (30 ml). The ether was removed by distillation until the pot temperature reached 100 C, then 30 ml of benzene was added, and the solution was heated under reflux for 5 h.  The mixture was cooled to 0 C and  hydrolyzed by pouring it slowly into a saturated aqueous solution of ammonium chloride (50 ml). The organic layer was separated and the aqueous layer was extracted three times with 30 ml portions of chloroform.  The combined solutions were dried over  MgSC>4 and evaporated to yield an oil. This oil was chromatographed on silica gel, eluting with ethyl acetate to remove menthol and impurities, then with 10:1 EtOAc/EtOH to remove the product from the column. The product (2.7 g, 65% yield) was recrystallized from a mixture of ethyl acetate and petroleum ether to yield 1.40 g of needles. These needles, which liquefied readily in air when trace amount of organic solvents presented on their surface, were filtered under nitrogen and dried in vacuo, m.p. 59-61 °C. Anal. Calcd. for Ci 4 H 15 OP: C, 73.02; H, 6.57. Found: C, 72.81; H, 6.60. *H NMR (CDC13, 200 MHz):  8 7.80-7.20 ( 9 H, m, Ar-H), 2.38 (3 H, s, Ar-  CH3), 2.00 (3 H, d, J PH = 13.0 Hz, ?CH3). 13  C NMR (CDCI3, 50 MHz): 6 15.86, 17.32 (d, JPC = 73.5 Hz, PCH3); 21.38  (s, Ar-CH3). Methyl substituted phenyl ring, 5 132.85, 134.86 (d, J PC = 101.0 Hz, C1); 130.96, 131.15 (d, JPC = 9.4 Hz, C-2); 138.41, 138.65 (d, JPC = 11.9 Hz, C-3);  186  132.49, 132.56 (d, JPC = 2.8 Hz, C-4); 128.36, 128.61 (d, JPC = 12.5 Hz, C-5); 127.36, 127.57 (d, JPC = 10.1 Hz, C-6). Phenyl ring, 5 133.26, 135.29 (d, JPC = 101.0 Hz, P-C); 130.37, 130.57 (d, JPC = 9.7 Hz, o-C); 128.49, 128.71 (d, JPC = 11.9 Hz, m-C);  131.61, 131.67 (d, J PC = 2.8, Hz, p-Q.  See Appendix I for  assignments. 31  P NMR (CDC13, 81 MHz): 5 29.73.  MS: m/e (relative intensity), 230 (59.2, M+), 229 (100.0), 215 (77.7), 165 (8.1), 153 (6.6), 139 (6.0), 121 (2.9), 114 (5.7), 107 (2.0), 91 (24.2), 77 (17.2), 65 (18.5), 51 (13.5). Calculated mass for C ^ H ^ O P : 230.0861. Found: 230.0852. [a]2D4 = +3.0° (c = 5.26, benzene). Optical purity: 86% e.e., HPLC; Chiralpak OP(+), 25 cm x 0.46 cm I.D.; eluent: 93:7 methanol/H20 at a flow rate of 0.35 ml/min; detector, UV diode at 230 nm and 254 nm; nearly baseline separation was obtained for the enantiomers. The 7?-isomer was prepared in the same manner and characterized by *H and 31p NMR spectroscopy. Its optical purity was 81 % by HPLC. Both enantiomers were not further purified to a higher optical purity and were used for the preparation of the dimer as the reaction would enrich the optical purity by itself.  (S)-(-)-(3',4'-Dimethylphenyl)methylphenylphosphine Oxide (5-110) (6)p-menthyl methylphenylphosphinate OSp-107, 3.09 g, 10.1 mmol) in 20 ml of benzene was treated with excess of a mixture of two Grignard reagents, prepared from 11.5 g (62.1 mmol) of bromo-o-xylene that contains 30% of l-bromo-2,3dimethylbenzene and 70% of l-bromo-3,4-dimethylbenzene (Aldrich) and 1.51 g (62.2 mmol) of magnesium in ether.  The ether was replaced by benzene until the batch  temperature reached 80 °C. The mixture was refluxed in about 40 ml of benzene overnight, cooled in an ice-water bath and poured into 60 ml solution of saturated  187  ammonium chloride. The organic phase was separated, and the aqueous phase was extracted with 3 x 30 ml of chloroform. The combined organic phase was dried over MgSC>4 and evaporated. The residue was chromatographed on silica gel (1. EtOAc, 2. 10:1 EtOAc/EtOH). Crystallization from a mixture of EtOAc and hexanes yielded 1.5 g (58%) of product. The residue may be purified by vacuum distillation to get rid of menthol and other more volatile impurities and crystallized from a mixture of EtOAc and hexanes. The yield so obtained was slightly less than that by chromatography on silica gel, m.p. 96-97 °C. The product from the Grignard reagent of l-bromo-2,3dimethylbenzene was not observed by TLC or NMR spectroscopy. Anal. Calcd. for Ci 5 H 17 OP: C, 73.74; H, 7.02. Found: C, 73.49; H, 7.05. ! H NMR (CDC13, 200 MHz): 5 7.74-7.06 ( 8 H, m, Ar-H), 2.20 (6 H, br. s, ArCH3), 1.92 (3 H, d, J PH = 13.2 Hz, PCH3). 13  C NMR (CDCI3, 50 MHz): 8 15.89, 17.35 (d, JPC = 73.6 Hz, PCH3); 19.68,  19.85 (2 s, Ar-CH3). Methyl substituted phenyl ring, 8 129.89, 131.95 (d, JPC = 103.4 Hz, C-l); 131.44, 131.63 (d, J PC = 9.8 Hz, C-2); 137.06, 137.40 (d, JPC = 12.1 Hz, C-3); 140.87, 140.93 (d, JPC = 3.0 Hz, C-4). 129.71, 129.96 (d, JPC = 12.9 Hz, C-5); 127.84, 128.04 (d, JPC = 10.0 Hz, C-6). Phenyl ring, 8 133.49, 135.50 (d, J PC = 101.0 Hz, P-C); 130.30, 130.50 (d, JPC = 9.9 Hz, o-C); 128.40, 128.64 (d, JPC = 11.8 Hz, w-C);  131.49, 131.55 (d, JPC = 2.7, Hz, p-C). See  Appendix I for assignments. 31  P NMR (CDCI3, 81 MHz): S 29.83.  IR (KBr): v m a x , 3055, 3012, 2970, 2906, 1442, 1381, 1299, 1197, 1170 (P=0), 1115, 1099, 1019, 908, 885, 827, 770, 755, 744, 698, 643, 564, 540, 506, 496, 477, 439,412 cm"1. MS: m/e (relative intensity), 244 (42.8, M+), 243 (71.0), 229 (100), 215 (15.8), 167 (11.8), 151 (4.8), 139 (9.6), 121 (14.0), 114 (11.3), 105 (17.9), 91 (21.1), 77  188  (48.6), 65 (9.5), 51 (8.2). Calculated mass for  C 15 H 17 OP: 244.1017.  Found:  244.1008. [a]|4 = -16.0° (c = 12.8, benzene). Optical purity: > 95% e.e.; HPLC, Chiralpak OP(+), 25 cm X 0.46 cm I.D.; eluent, methanol, flow rate of 0.45 ml/min; detector, UV at 254 nm.  (/?)-(+)-(3',4'-Dimethylphenyl)methylphenylphosphine Oxide (fl-110) Similarly, R-110 was synthesized from (R)p-menthyl methylphenylphophinate (Rp107), purified by silica gel chromatography (yield 64%) and then crystallized from a mixture of cyclohexane and benzene, m.p. 86-92 °C.  ^H NMR,  31  P NMR and IR  spectra of this isomer were identical with those of (iS)-enantiomer respectively. A lower melting point than its enantiomer may be due to its sensitivity to small amount of impurities. This enantiomer was not further purified to a higher optical purity and was used for the preparation of the dimer as the reaction would enrich the optical purity by itself. [a]2D4 = 15.2° (c = 5.5, benzene). Optical purity: 94% e.e. by HPLC. Nearly baseline separation for the enantiomers was obtained.  (S)-(-)-(3',5'-Dimethylphenyl)methylphenylphosphine Oxide (S-lll) (5)p-Menthyl methylphenylphosphinate (5p-107, 8.2 g, 27.9 mmol) in 30 ml of benzene was added to a solution of 3,5-dimethylphenylmagnesium bromide (83.6 mmol) in 50 ml of diethyl ether, which was prepared from l-bromo-3,5-dimethylbenzene and magnesium. The ether was boiled out until the batch temperature reached 100 °C and then 30 ml of benzene was added. The mixture was refluxed for 3 h, cooled to 5 °C, and poured slowly into 60 ml of saturated aqueous ammonium chloride solution. The 189  two clear phases separated at room temperature. The organic phase was separated and the aqueous phase was extracted with chloroform (2 x 30 ml). The combined organic phase was dried over magnesium sulphate. After evaporation of the solvents, the crude oil was purified by chromatography on silica gel 60 (1. EtOAc, 2. 10:1 EtOAc/EtOH) to result in 4.8 g (72% yield) of 5-111. Recrystallization from a mixture of ethyl acetate and petroleum ether yielded the pure phosphine oxide as colorless needles, m.p. 7274 °C. Anal. Calcd. for C 15 H 17 OP: C, 73.74; H, 7.02. Found: C, 73.67; H, 6.97. ! H NMR (CDC13, 200 MHz):  8 7.80-7.12 (8 H, m, Ar-H), 2.28 (6 H, d, J PH =  1.2 Hz, Ar-CH3), 1.98 (3 H, d, J PH = 13.5 Hz, PCH3). 13  C NMR (CDCI3, 50 MHz): 5 15.87, 17.33 (d, JPC = 73.5 Hz, PCH3); 21.26  (Ar-CH3). Methyl substituted phenyl ring, 8 132.68, 134.67 (d, JPC = 100.9 Hz, P-C); 127.99, 128.18 (d, J PC = 9.7 Hz, o-C); 138.23, 138.48 (d, JPC = 12.5 Hz, m-C); 133.43 (overlapped), 133.49 (d, JPC = 3.0 Hz, p-C).  Phenyl ring, 8 133.43  (overlapped), 135.44 (d, JPC = 101.1 Hz, P-C); 130.36, 130.55 (d JPC = 9.7 Hz, oC); 128.44, 128.63 (d, JPC = 11.8 Hz, m-C); 131.54, 131.60 (d, JPC = 2.7 Hz,/>-C). See Appendix I for assignments. 31  P NMR (CDCI3, 81 MHz): 8 29.82.  MS: m/e (relative intensity), 244 (50.6, M+), 243 (100), 229 (68.9), 211 (1.9), 178 (1.6), 167 (6.9), 139 (4.6), 121 (6.1), 114 (4.8), 105 (12.9), 91 (12.9), 77 (39.6), 65 (5.4), 51 (8.0). Calculated mass for C 15 H 17 OP: 244.1017. Found: 244.1009. [a]% = -13.0° (c = 5.5, benzene). Optical purity: >95% e.e.; HPLC Chiralpak OP(+), 25 cm x 0.46 cm I.D.; eluent, methanol, flow rate of 0.45 ml/min; detector, UV at 254 nm. Nearly baseline separation for an enantioenriched sample was achieved.  190  /f-(3' ,5' -Dimethylphenyl)methylphenylphosphine Oxide (R-l 11) The /^-isomer was prepared following the same procedure as for the S-isomer, beginning with (7?)p-menthyl methylphenylphosphinate (i?p-107), m.p. 65-68 °C. *H NMR, ^lp NMR and IR spectra of this isomer were identical with those of its enantiomer respectively. A lower melting point than its enantiomer may be due to its sensitivity to a small amount of impurities. [a]24 = +13.7° (C = 5.5; benzene). Optical purity: > 95% e.e. by HPLC.  Methylphenyl(2',3',5,-trimethylphenyl)phosphine Oxide (112) The trimethylphenylmagnesium bromide was prepared from 8 g (0.040 mol) of l-bromo-2,4,5-trimethylbenzene (Aldrich) and 1.00 g (0.041 mol) of magnesium in 40 ml of anhydrous tetrahydrofuran and the Grignard solution was carefully replaced with toluene until no THF remained.  (i?)p-menthyl methylphenylphosphinate (i?p-107,  1.31 g, 4.47 mmol) in 20 ml of toluene was added into the Grignard solution. The final solution was kept at 95 °C for 70 h until GC showed no phosphinate left. This solution was then cooled in an ice bath and hydrolyzed by pouring it into 75 ml of saturated aqueous solution of ammonium chloride.  The organic phase was separated and the  aqueous phase was extracted with 3 x 40 ml of benzene. The combined fractions were dried over magnesium sulphate and evaporated to afford an oil. This oil was submitted to chromatography on silica gel. The impurity was eluted with EtOAc and the product with 10% ethanol in EtOAc. The compound (0.28 g, 24%) was obtained after the solvents were removed in vacuo. The optical purity of this crude product was 42% e.e. by chiral HPLC (Chiralpak OP(+), 25 cm x 0.46 cm I.D.; eluent, 9:1 methanol/water at a flow rate of 0.35 ml/min; detector, UV diode at 254 nm, well resolved peaks  191  observed). Recrystallization from a mixture of ethyl acetate and petroleum ether yielded 0.1 g of racemic product, m.p. 139-140 °C. Anal. Calcd. for Ci6H 19 OP: C, 74.38; H, 7.42. Found: C, 74.44; H, 7.59. !H NMR (CDC13, 200 MHz):  8 7.72-6.88 (7 H, m, Ar-H), 2.01 (9 H, 3 s  overlapped), 1.98 (3 H, d, J PH = 13.5 Hz, PCH3). 13c NMR (CDCI3, 50 MHz): 5 16.40, 17.86 (d, JPC = 73.9 Hz, PCH3); 19.26, 19.65 (2 s, m-and P-A1-CR3). 20.62, 20.72 (d, JPC = 4.5 Hz, o-Ar-O^). Methyl substituted phenyl ring, 5 127.30, 129.33 (d, JPC = 102.0 Hz, P-C); 138.85, 139.04 (d, JPC = 8.5 Hz, C-2); 132.63, 132.86 (d, JPC = 11.3 Hz, C-3); 141.01, 141.07 (d, JPC = 2.7 Hz, C-4); 133.59, 134.16 (d, JPC = 12.3 Hz, C-5); 133.16, 133.37 (d, JPC = 11.2 Hz, C-6). Phenyl ring, 5 134.16, 136.15 (d, JPC = 99.9 Hz, P-C); 130.20, 130.40 (d, JPC = 11.1 Hz, o-C); 128.40, 128.64 (d, JPC = 11.9 Hz, m-C); 131.32, 131.37 (d, Jpc = 2.7 Hz, p-C). See Appendix I for assignments. 31  P NMR (CDCI3, 81 MHz): 8 31.03.  MS: m/e (relative intensity), 258 (47.1, M+), 257 (100), 243 (7.4), 210 (2.5), 179 (4.5), 165 (3.8), 139 (1.9), 121 (6.8), 114 (4.2), 105 (2.7), 91 (8.9), 77 (7.9), 65 (1.9), 51 (1.8). Calculated mass for Ci 6 H 19 OP 258.1174. Found: 258.1172. (5)-(-)-(A^,Ar-Dimethylaminophenyl)methylphenylphosphine Oxide (S-llS) (5p)-menthyl methylphenylphosphinate (5p-107, 4.0 g, 13.7 mmol) in 30 ml of benzene was added to the Grignard solution prepared from 12.7 g (63.7 mmol) of 4bromo-Af A/-dimethylaniline (Eastman Organic Chemicals) and 1.58 g (63.6 mmol) of magnesium in 30 ml of tetrahydrofuran (dried and distilled over sodium). THF was removed out by distillation until the batch temperature reached 100 °C. The remaining mixture was refluxed in about 40 ml of benzene at 80 °C for 3 h, cooled in an ice water bath and poured into 100 ml of saturated aqueous solution of ammonium chloride. The  192  organic layer was separated and the aqueous layer was extracted with 3 x 50 ml of chloroform. The combined organic phases were dried over MgSC<4 and evaporated to yield a semicrystalline material.  The crude product was submitted to silica gel  chromatography, impurities were eluted with ethyl acetate and the product (3.06 g, 86%) with 10% ethanol in ethyl acetate.  Further purification was done by recrystallization  from ethyl acetate, m.p. 151-152 °C. Anal. Calcd. for Ci 5 H 1 8 NOP: C, 69.47; H, 7.00; N, 5.40. Found: C, 69.63; H, 7.05; N, 5.52. ! H NMR (CDC1 3 , 200 MHz): 5 7.80-7.30 (7 H, m, Ar-fl), 6.72 (2 H, m, Ar-fl), 3.02 (6 H, s, -N(Cffj)2,), 1.95 (3 H, d, J P H = 13 Hz, ?CH3). 13  C NMR (CDCI3, 50 MHz) 5 16.14, 17.61 (d, J PC = 73.9 Hz, PCH3), 39.94  (NCH3). Amino substituted phenyl group, 8 117.35, 119.59 (d, J P C = 112.2 Hz, C-l (P)), 131.76, 131.98 (d, J PC = 11.1 Hz, C-2), 111.26, 111.51 (d, J PC = 12.7 Hz, C-3), 152.29, 152.34 (d, J PC = 2.4 Hz, C-4 (N)). Phenyl ring, 8 134.39, 136.40 (d, J PC = 101.0 Hz, C-l), 130.34, 130.53 (d, J PC = 9.8 Hz, C-2), 128.27, 128.51 (d, J PC = 11.8 Hz, C-3), 131.26, 131.31 (d, J PC = 2.8 Hz, C-4).  See Appendix I for  assignments. 31  P NMR (CDCI3, 81 MHz) 8 30.01.  IR (KBr): v m a x , 2975 (C-H), 2906 (C-H), 1607, 1523, 1443, 1371, 1302, 1167 ( P = 0 ) , 1122, 899, 810, 780, 747, 698, 529, 506 cm" 1 . MS: m/e (relative intensity), 259 (98.7, M + ) , 258 (100), 244 (90.9), 226 (8.4), 214 (8.6), 197 (24.6), 196 (20.2), 183 (11.0), 166 (16.8), 153 (12.2), 136 (10.5), 128 (16.0), 120 (8.32), 104 (8.3), 98 (8.9), 91 (11.0), 77 (24.8). C 1 5 H 1 8 NOP: 259.1126. Found: 259.1116. [ a ] " = -31.5° (c = 3.0, CH3OH). UV spectrum: ethanol, l(e x 10"4), 281 (2.54), 207 (2.26).  193  Calculated mass for  X-ray crystal structure analysis: Space group, P2\, a = 8.608(4) A, b = 28.807(4) A, c = 11.887(3) A, p = 107.38(3) , V = 2813 (3) A3, Z = 8; R = 0.036; absolute configuration was determined as (S).  (5,S)-(+)-l,2-Ethanediylbis((4'-methylphenyI)phenylphosphine oxide) (5,5-114)90 A stirred solution of 2.0 g (8.69 mmol) of (iS)-methyl(4'-methylphenyl)phenylphosphine oxide (£-108, > 99% e.e. by HPLC) in 20 ml of dry tetrahydrofuran was cooled to -78 °C and treated slowly with 7.0 ml of 1.35 M n-butyllithium (9.45 mmol) in hexane under nitrogen. After 30 min, 1.32 g (9.83 mmol) of dry CuCl2 was added. The mixture was kept at -78 °C for 15 min before being allowed to warm to room temperature, and was then saturated with oxygen for 20 min. After quenching with 20 ml of concentrated hydrochloric acid, the mixture was extracted with 3 x 30 ml of chloroform. The combined extracts were washed with dilute aqueous NH3 until almost colorless, then with water, dried over Na2SC>4 and evaporated to afford 1.7 g of crude product. Recrystallization from a mixture of methylene chloride and acetone yielded 1.0 g (50% yield) of colorless needles, m.p. 249-251 °C; lit. 90 247-248 °C. Anal. Calcd. for C28H2802P2: c> 7 3 - 3 5 ; H> 6 1 6 - Found: C, 73.08; H, 6.14. *H NMR (CDCI3, 200 MHz): 8 7.10-7.64 (18 H, m, Ar-H), 2.41 (4 H, br. s, methylene-fl), 2.30 (6 H, s, Ar-CH3). 13  C NMR (CDCI3, 50 MHz): 8 21.09, 21.76, 21.41 (m, JPC = 66.4 Hz, -CH2P);  21.53 (s, Ar-CH3). Substituted phenyl ring, 8 130.62, 130.71, 130.80 (t, overlapped with that of o-C of phenyl ring, JPC = 9.6 Hz, o-C); 129.42, 129.54, 129.66 (t, JPC = 12.1 Hz, m-C);  142.56 (br. s, p-C).  Phenyl ring, 8 130.71, 130.80, 130.90 (t,  overlapped with that of o-C of substituted phenyl ring, JPC = 9.8 Hz, o-C); 128.62, 128.74, 128.86 (t, JPC = 10.8 Hz, m-C); 131.89 (br. s, p-C). See Appendix I for assignments.  194  31p NMR (CDCI3, 81 MHz): 5 32.75. MS: m/e (relative intensity), 458 (1.2, M+), 457 (2.2), 429 (1.0), 381 (100.0), 367 (87.8), 351 (9.4), 339 (4.4), 290 (5.7), 276 (5.2), 243 (76.1), 215 (40.4), 199 (10.6), 183 (5.4), 165 (7.5), 151 (7.7), 118 (4.7), 104 (5.4), 91 (23.0), 77 (13.2), 65 (10.5), 47(11.3). Calculated mass for C28H28O2P2: 458.1564. Found: 458.1557. IR (KBr): v m a x , 1187 (P=0), 1176 (P=0), 1122, 757, 744, 734, 515 cm"1. UV-Vis Spectrum: ethanol, X(s), 272 (1710, sh), 265 (2440, sh), 231 (7390). [a]2D4 = +6.3° (c = 5.60, CHCI3); lit. 90 [a]2J = +6.9° (c = 0.98, CHCI3). Optical purity: >95 % e.e.; HPLC, Chiralcel OD, 25 cm x 0.46 cm I.D.; eluent, 97:3 hexane/ethanol, flow rate of 1.0 ml/min, room temperature; detector, UV at 231 nm, a nearly baseline separation was obtained for an enantioenriched sample.  (/f,/f)-(-)-l,2-Ethanediylbis((4'-methylphenyl)phenylphosphine oxide) (R,R-IU) The R,R-\somer was prepared following the same procedure, starting with (R)methyl(4'-methylphenyl)phenylphosphine oxide (R-10S, 88% e.e.); m.p. 248-250 °C, [cc]2D4 = -8.5° (c = 5.92, CHCI3). Optical purity by HPLC: >95% e.e., and HPLC conditions were the same as those for the S^S-isomer. The R,R-isom&r contained a very small amount of an impurity detected by HPLC, although ^H and 31p NMR showed no impurities present.  (S,S)-(+)-l,2-EthanediyIbis((3'-methylphenyl)phenylphosphine oxide) (S, 5-115) This compound was prepared in a similar way as 114. Monophosphine oxide R-109 (2.26 g, 9.83 mmol, 86% e.e.) was dissolved in 15 ml of dry THF (freshly dried and distilled over sodium) and the solution was cooled to -78 °C under nitrogen. To this, 12 ml of 0.83 M sec-BuLi (10.0 mmol) in cyclohexane was added slowly via syringe. The mixture was stirred for 2 h at -78 °C. After adding 2.0 g (14.9 mmol) of dry CuCl2, 195  the mixture was allowed to warm to room temperature over 1.5 h and saturated with oxygen for 20 min. The mixture was quenched with 20 ml of 6 N hydrochloric acid and extracted with 3 x 25 ml of chloroform.  The combined extracts were washed with  dilute ammonia solution until colorless and then saline, dried over magnesium sulphate and evaporated in vacuo to result in 0.55 g of white solid.  Recrystallization from a  mixture of methylene chloride and acetone yielded 0.31 g of long needles, m.p. 193-194 °C. Anal. Calcd. for C28H28C>2P2: C, 73.35; H, 6.16. Found: C, 73.53; H, 6.28. ! H NMR (CDC13, 200 MHz): 6 7.16-7.68 (18 H, m, Ar-#), 2.40 (4 H, br. s, methylene-^), 2.08 (6 H, s, Ar-CH3). 13  C NMR (CDCI3, 50 MHz): 5 20.98, 21.64, 22.30 (m, J PC = 64.5, 2.3 Hz, J P P  = 56.2 Hz, CH 2 p );  21  -36 (s, Ar-CH 3 ). Substituted phenyl ring, 5 131.19, 131.28,  131.37 (t, J P C = 9.2 Hz, C-2);  138.64, 138.76, 138.87 (t, J P C = 11.7 Hz, C-3);  132.85 (br. s, C-4); 128.55, 128.66, 128.77 (triplet, overlapped with the triplet of m-C of the phenyl group, J PC = 11.4 Hz, C-5); 127.62, 127.72, 127.82 (t, J PC = 9.8 Hz, C-6). Phenyl ring, 8 130.63, 130.73, 130.82 (t, J PC = 9.7 Hz, o-C); 128.66, 128.77, 128.89 (triplet, overlapped with the triplet of C-5 of the substituted phenyl group, J P c = 11.6 Hz, m-C); 131.94 (br. s, p-C). See Appendix I for assignments. 31  P NMR (CDCI3, 81 MHz): 5 32.51.  MS: m/e (relative intensity), 458 (9.4, M + ) , 457 (2.1), 381 (77.4), 367 (71.5), 351 (9.1), 339 (5.5), 290 (5.5), 276 (5.0), 243 (100.0), 215 (54.2), 197 (11.0), 183 (5.2), 165 (8.7), 151 (9.8), 118 (5.3), 104 (9.0), 91 (47.0), 77 (23.2), 65 (16.6). Calculated mass for C28H28O2P2: 458.1565. Found: 458.1569. IR (KBr): v m a x , 1439, 1422, 1223, 1198 ( P = 0 ) , 1183 ( P = 0 ) , 1120, 1092, 743, 701, 534, 478, 437, cm" 1 . UV-Vis spectrum: ethanol, Ms), 278 (1980, sh), 271 (3000), 265 (2850). [a]24 = 2.6° (c = 4.48, CHCI3). 196  (5,5)-(+)-l,2-Ethanediylbis((3' ,4*-dimethyIphenyl)phenylphosphine oxide) (S, S-116) Compound 5-110 (2.5 g, 10.2 mmol, > 95% e.e.) in 20 ml of dry THF was cooled to -78 °C in a 50 ml flask protected under nitrogen and to it, 8.0 ml of 1.4 M butyllithium (11.2 mmol) in hexane was added slowly with stirring. The mixture was kept at this temperature for 30 min. Dry CuCl2 (1.6 g, 11.9 mmol) was added quickly to the stirred solution. The suspension was kept -78 °C for 15 min, warmed to room temperature and then saturated with oxygen for 20 min to result in a homogenous darkblue solution. After being quenched with 5 ml of concentrated HC1, the solution was extracted with chloroform.  The combined organic phases were washed with dilute  ammonia solution until a colorless solution was reached and then with water, dried over sodium sulphate and evaporated in vacuo to yield 2.0 g of white solid. Recrystallization from acetone gave 1.0 g (40%) of needle clusters, m.p. 179-180 °C. Anal. Calcd. for C30H32O2P2: C, 74.06; H, 6.63. Found: C, 74.25; H, 6.60. !H NMR (CDCI3, 200 MHz): 8 7.74-7.14 (16 H, m, Ar-H), 2.44 (4 H, br. s, methylene-H), 2.22 (12 H, s, Ar-CH3). 13  C NMR (CDCI3, 50 MHz): 5 19.71, 19.88 (2 s, Ar-CH3). 21.06, 21.73, 22.38  (m, CH2P). Substituted phenyl ring, 5 127.70, 129.73 (d, JPC = 102.0, C-l); 131.70, 131.79, 131.89 (t, overlapped with the singlet of p-C of phenyl ring, JPC = 9.6 Hz, C2); 137.32, 137.44, 127.56 (t, J PC = 11.8 Hz, C-3); 141.28, 141.30, 141.33 (t, JPC = 2.7, C-4); 129.91, 130.03, 130.12 (t, JPC = 12.5 Hz, C-5); 128.12, 128.25, 128.35 (t, JPC = 9.7 Hz, C-6). Phenyl ring, 8 131.50, 133.49 (d, JPC = 99.8, P-C); 130.60, 130.70, 130.79 (t, JPC = 9.5 Hz, o-C); 128.60, 128.72, 128.84 (t, JPC = 10.8 Hz, mC); 131.82 (br. s, overlapped with the triplet of C-2 of the substituted phenyl ring, pC). See Appendix I for assignments. 31p NMR (CDCI3, 81 MHz): 8 32.68.  197  MS: m/e (relative intensity), 487 (10.0, M + ) , 486 (30.4), 409 (87.6), 381 (100.0), 365 (8.3), 353 (3.8), 318 (3.6), 290 (4.8), 257 (96.7), 243 (11.7), 229 (43.4), 213 (14.6), 183 (12.7), 165 (11.3), 151 (28.7), 133 (9.3), 105 (24.1), 91 (25.3), 77 (36.2), 47(19.8). Calculated mass for C30H32O2P2: 486.1878. Found: 486.1870. IR (KBr): v m a x , 1438, 1202, 1187 ( P = 0 ) , 1174 ( P = 0 ) , 1115, 1099, 756, 744, 726, 696, 557,511cm" 1 . UV-Vis spectrum: ethanol, A,(s), 277 (1550, sh), 271 (2640), 265 (2920), 229 (31100, sh). [cc]2D4 = +2.5° (c = 7.9, CHCI3).  (S,S)-(-)-l,2-Ethanediylbis((3',5'-dimethylphenyl)phenylphosphine oxide) (5,5-117) To a stirred solution of (5)-(3',5'-dimethylphenyl)methylphenylphosphine oxide (S111, 2.18 g, 8.92 mmol, >95% e.e by HPLC) in anhydrous THF (20 ml) at -78 °C under argon, 8.6 ml of 1.11 M sec-BuLi in cyclohexane (9.55 mmol, Aldrich) was added over a period of 10 min. After stirring for another 2 h at -78 °C, dry CuCl2 (2.12 g, 15.8 mmol) was added and the mixture was stirred for 20 min at -78 °C, then at room temperature for 1 h while being saturated with oxygen.  The final mixture was  hydrolyzed with dilute HC1 and extracted with CHCI3 (3 x 20 ml).  The combined  organic layers were washed with dilute ammonia solution three times and then with water, evaporated in vacuo to result in 1.8 g of solid. Recrystallization from a mixture of acetone and diethyl ether yielded 1.0 g of product, and purer compound could be obtained by recrystallization again from acetone, m.p. 95-97 °C. Anal. Calcd. for C30H32O2P2: C, 74.06; H, 6.63. Found: C, 74.06; H, 6.62. ! H NMR (CDCI3, 200 MHz): 6 7.78-7.10 (16 H, m, Ar-H), 2.48 (4 H, br. s, methylene-fl), 2.30 (12 H, s, Ar-Ctfj).  198  13  C NMR (CDCI3, 50 MHz): 5 21.26 (s, Ar-CH 3 ). 20.93, 21.59, 22.24 (m, J P C  = 68.4, 2.0 Hz and J P P = 52.7 Hz, CH 2 p )- Substituted phenyl ring, 5 132.46, 131.30 (d, J P C = 99.8 Hz, C-l);  128.19, 128.29, 128.38 (t, J P C = 9.6 Hz, C-2); 138.49,  138.61, 138.74 (t, J PC = 11.5 Hz, C-3); 133.84 (br. s, C-4). Phenyl ring, 5 133.26, 131.26 (d, J P C = 100.2 Hz, P-C);  130.72, 130.82, 131.27 (t, J PC = 9.5 Hz, o-C);  128.66, 128.78, 128.89 (t, J PC = 11.8 Hz, m-C); 131.899 (s, p-Q.  See Appendix I  for assignments. 31  P NMR (CDCI3, 81 MHz): 8 33.03.  MS: m/e (relative intensity), 487 (4.0, M + ) , 486 (17.1), 409 (33.3), 381 (35.8), 365 (4.6), 257 (100.0), 229 (66.8), 213 (12.7), 183 (8.3), 165 (8.5), 151 (31.1), 119 (15.8), 105 (40.6), 91 (24.2), 77 (41.6). Calculated mass for C30H32O2P2: 486.1878. Found: 486.1881. IR (KBr): v m a x , 1438, 1180 ( P = 0 ) , 1195 ( P = 0 ) , 1129, 1113, 742, 696 cm"1. UV-Vis spectrum: ethanol, 1(e), 283 (2220), 272 (3040), 266 (2690), 225 (11150). [a]2D4 = -6.4° (c = 4.7, CHCI3).  (R,R)-(+)-l,2-Ethanediylbis((3',5'-dimethylphenyl)phenylphosphine  oxide)  (R,R-U7) Starting from (i?)-(+)-methyl-(3',5'-dimethylphenyl)phenylphosphine oxide  (R-lll,  >95% e.e.) and using 1.55 M LDA in THF instead of s-BuLi, following the same procedure as above, a 56% yield was obtained, m.p. 98-99 °C. [a] 2 ^ = +7.6° (c = 4.7, CHCI3).  1,5-DichIoroanthracene (118) 94 A mixture of 1,5-dichloroanthraquinone (22.2 g, 80 mmol, Aldrich) and zinc powder (61.2 g, 0.94 mol) in 300 ml of pyridine was heated to reflux with magnetic stirring, and 199  80% acetic acid (150 ml, 2 mol) was added dropwise over 6 h. The reaction mixture was stirred under reflux for an additional 30 min, and allowed to cool to room temperature. The excess zinc was filtered and the filtrate was poured into 1600 ml of ice-cold 3 M hydrochloric acid to yield a precipitate. After stirring for 15 minutes, the resulting solid was collected by filtration. Recrystallization of the crude solid from a mixture of ethanol and benzene afforded 5.0 g (25% overall yield) of fine needles, m.p. 185-186 °C (lit. 9 4 ' 1 8 1 178-182 °C, 185 °C). l  U NMR (CDCI3, 400 MHz) 5 8.88 (2 H, s, H9 and H10), 8.03 (2 H, m,), 7.63 (2  H, m), 7.43 (2 H, m). IR (KBr): v m a x , 1619, 1302, 1164, 874, 780, 721 cm"1. MS: m/e (relative intensity), 248 (M++2, 66), 246 (M + , 100), 211 (7), 176 (36).  l,5-Dicyano-9,10-anthraquinone (119) Following the procedure of Rogers and Averill,9^ 10 g (36 mmol) of 1.5dichloroanthraquinone and 9.2 g (0.10 mol) of CuCN were slurried in 50 ml of DMA and refluxed under argon for 3 h. The hot brown solution was poured onto 500 ml of ice, and the brown-green precipitate was filtered and washed with water. This copper complex was decomposed with 500 ml of 3 N nitric HNO3 at 65 C for 3.5 h. The resulting brown solid was filtered and washed with water and air-dried to afford 9 g of crude product. This compound was used for the next step without further purification. *H NMR (DMSO-d6, 200 MHz) 8 8.53 (2 H, dd, J HH = 7.8, 1.3 Hz), 8.41 (2 H, dd, J HH = 7.7, 1.3 Hz), 8.12 (2 H, t, J HH = 7.8 Hz). MS: m/e (relative intensity), 258 (M + , 100), 230 (43.7), 211 (13.2), 202 (64.6), 175 (33.7), 150 (8.8), 101 (28.2), 88 (4.8), 75 (34.5).  200  9,10-Anthraquinone-l,5-dicarboxyIic Acid (120)95,96 Crude l,5-dicyano-9,10-anthraquinone (119, 9.0 g) was refluxed in 100 ml of 70% H2SO4 for one hour and then the hot solution was poured onto 400 g of ice to precipitate the crude acid as a brown solid (9.5 g). This compound was used without further purification. *H NMR (CDCI3, 400 MHz) 5 8.25-8.15 (2 H, m), 8.02-7.80 (4 H, m).  Anthracene-l,5-dicarboxylic Acid (121)95,96 Crude anthraquinone 120 (9.5 g) and 30 g of zinc dust were heated at 90 C with stirring in 300 ml of 20% NH4OH for 4.5 h, during which time the color changed from brown to deep-red and then to yellow. The ammonia solution was added periodically during heating. The final solution was filtered to remove excess zinc and acidified with concentrated HC1 to obtain a yellow precipitate. Filtration of the solid and air-drying afforded 9.0 g of crude product. !H NMR (DMSO-d6 200 MHz) 5 9.60 (2 H, s, H9 and H10), 8.40-8.10 (4 H, m), 7.60 (2 H, m).  1,5-Dimethyl Anthracene-l,5-dicarboxylate (122)95 Crude carboxylic acid 121 (0.9 g) was refluxed in 150 ml of methanol with 2 ml of H2SO4 for 44 h. A golden crystalline solid was formed, and the solid was filtered to afford 0.55 g (55%) of the crude product. Recrystallization from benzene yielded 0.41 g of golden needles, m.p. 202-203 C; lit. 95 200-201 C. !H NMR (CDCI3, 200 MHz) 5 9.65 (2 H, s, H9, and H10), 8.27 (4 H, m), 7.53 (2 H, m), 4.07 (6 H, s, CH3).  201  MS: m/e (relative intensity), 295 (M+ + 1, 20), 294 (M + , 100), 263 (52), 235 (32), 220 (18), 203 (20), 176 (22), 116 (30), 88 (39), 75 (12). IR (KBr): v m a x , 2949, 1700 (C=0), 1543, 1437, 1261, 1216, 1143, 1041 cm"1.  Anthracene-l,5-dimethanol (123) Ester 122 (2.52 g, 8.5 mmol) in 25 ml of THF was dropped into 1.2 g of UAIH4 in 20 ml of THF with stirring. The excess UAIH4 was destroyed with ethyl acetate. The final mixture was hydrolyzed in 50 ml of 4N HC1, and the solid was filtered and recrystallized from a mixture of DMSO and ethyl acetate to afford 1.55 g (76%) of light green needles, m.p. 234-236 C. Anal. Calcd. for C i 6 H 1 4 0 2 : C, 80.65; H, 5.92. Found: C, 80.29; H, 5.90. ! H NMR (DMSO-d6, 200 MHz) 8 8.70 (2 H, s, H9 and H10), 7.96 (2 H, d, J HH = 8.3 Hz, Ar-fl) and 7.51 (4 H, m, Ar-#), 5.40 (2 H, t, J HH = 5.5 Hz, OH), 5.10 (4 H, d, J HH = 5.6 Hz, CH2OU). 13  C NMR (DMSO-d6, 50 MHz) 8 61.14 (CH20H)>  122 88  -  (C4)>  123 47  -  (C3)>  125.17 (C2), 128.03 (C9), 128.83 (C4a), 131.43 (C9a), 137.80 (CI). MS: m/e (relative intensity), 239 (M + + 1, 19), 238 (M + , 100), 221 (18), 207 (10), 191 (74), 179 (66), 165 (17), 152 (9), 104 (5), 94 (10), 89 (10), 76 (13). Calculated mass for Ci^H^C^: 238.0994. Found: 238.0996. IR (KBr): v m a x , 3600-2200 (C-H and O-H), 1086, 994, 877, 799, 732 cm"1.  l,2-Ethynediylbis(diphenylphosphine) (124) 1 8 2 Ethylmagnesium bromide in ether (70 ml, 3M, 21 mmol, Aldrich) was put in a 250 ml three-neck flask equipped with a long condenser, a gas inlet and a magnetic stirring bar.  Acetylene (Matheson) was passed through this Grignard solution for 15 h and  stirred for another 2 h to yield a brown oil of ethynyl-dimagnesium bromide. Then 39 202  ml (27.1 mmol) of chlorodiphenylphosphine (Aldrich) in 100 ml of diethyl ether was slowly added dropwise with vigorous stirring. The final mixture was hydrolyzed with 20 ml of glacial acetic acid in 120 ml of water, neutralized with sodium bicarbonate and extracted with diethyl ether (3 x 100 ml). The ether layer was dried over magnesium sulfate and evaporated. The residual oil was precipitated in n-pentane and recrystallized from ethanol to yield 28.5 g (67%) of the product, m.p. 83-84 °C; lit. 182 m.p. 7980 °C.  l,2-Ethynediylbis(diphenylphosphine oxide) (125) This compound was prepared in near-quantitative yield by oxidation of 1,2ethynediyl-bis(diphenylphosphine) (124) with hydrogen peroxide in acetone. Typically, 24 ml of 3% H2O2 (21.2 mmol) in acetone, diluted from commercial 30% H2O2, was dropped into a solution of 2.8 g (7.10 mmol) of phosphine 124 in 25 ml of acetone. An exothermic reaction ensued. The final solution was refluxed for 30 min, and the solvents were removed to yield a glassy solid. Recrystallization of the solid from a mixture of acetone and diethyl ether resulted in 2.9 g (96%) of the oxide, m.p. 165-166 °C; lit. 182 m.p. 164 °C. 31  P NMR (CDCI3, 81 MHz): 5 8.86.  13  C NMR (CDCI3, 50 MHz, ): 5 97.51, 97.85, 110.21, 100.56 (JPC = 136.4,  17.2 Hz, alkyne-C). Diphenylphosphinoyl group, 5 128.78, 128.92, 129.06 (m, J PC « 13.8 Hz, an accurate calculation is impossible due to the partially resolved peaks, m-C), 129.74, 132.16 (JPC = 122.18 Hz, P-C); 130.89, 130.93, 131.04, 131.15, 131.19 (m, Jpc ~ 11.3 Hz, an accurate calculation is impossible due to the partially resolved peaks, o-C); 133.00 (br. s, p-C).  203  7>ans-l,2-ethenediyIbis(diphenylphosphine oxide) 183 r/ww-l^-dichloroethene (0.70 ml, 9.1 mmol, Aldrich) was added slowly into a hot solution of Ph2PLi, prepared from 0.294 g (42 mmol) of lithium shot and 3.5 ml (20 mmol) of dichlorophenylphosphine in 5 ml of dry THF. The mixture was refluxed for 10 min, cooled to room temperature, quenched by 10 ml of water and extracted with chloroform. The extract was evaporated and the residue was crystallized from methanol to yield 1.5 g of ?ra«5-l,2-ethenediylbis(diphenylphosphine), m.p. 122-125 °C; lit.18 m.p. 125-126 °C. Treatment of the phosphine with 3% H2O2 in hot acetone gave the dioxide. The crude product was crystallized from a mixture of CHCI3 and ethanol to afford 1.1 g of the product, m.p. 315 °C; lit. 183 m p - 310-311 °C. 31  P NMR (CDCI3, 81 MHz): 5 21.42.  9,10-Dihydro-9,10-ethenoanthracene-l 1,12-bis(diphenyIphosphine oxide) (126) A ground mixture of 1.05 g (2.45 mmol) of l,2-ethynediylbis(diphenylphosphine oxide) (125) and 0.447 g (2.51 mmol) of anthracene was heated to about 200 °C with stirring to melt the solids together and then kept at 190 °C for about 20 min until the melted mixture crystallized. Heating was continued for another 20 min at 190 °C to sublime a small amount of excess anthracene. The excess anthracene was removed and the product (1.42 g, 2.14 mmol, 87% yield) was obtained after recrystallization from a mixture of chloroform and acetone or a mixture of methylene chloride and acetone. This compound was found to include one equivalent of acetone inside the crystals; the acetone could be driven off by heating the sample above 120 °C, m.p. 237-238.5 °C. Further purification was done by recrystallization from ethanol. Anal. Calcd. for C 4 oH 30 02P2-C3H 6 0: C, 77.93; H, 5.48. Found: C, 77.72, H, 5.45.  204  TGA: Calcd. weight loss for one equivalent of acetone: 8.76%. Found: 8.78%, (Tj - 139.1 °C, T 2 = 151.6 °C, T m a x = 147.8 °C). DSC: AHi = 73.89 J/g at T = 117.7-135.0 °C; AH 2 = 70.09 J/g at T = 234.8236.8°C. ! H NMR (CDC1 3 , 200 MHz): 8 7.48-6.98 (28 H, m Ar-fl); 5.68 (2 H, dd, J P H = 12, 4 Hz, bridgehead-fl); 2.12 (6 H, s, acetone). ! 3 c NMR (CDCI3, 50 MHz, ): 6 Diphenylphosphinoyl group, 132.92, 130.74 (d, J PC = 109.5 Hz, P-C); 131.86, 131.97, 132.07 (t, J P C =10.3 Hz, o-C); 127.95, 128.07, 128.20 (t, J P C = 13.5 Hz, m-C); 131.73 (br. s, p-C).  Dibenzobarrellene  skeleton, 5 143.59 (C-9a, 4a, 10a and 7a); 125.23 (C-1, 4, 5 and 8); 123.68 (C-2, 3, 6 and 7); 158.80, 156.86 (d, J PC = 97.7 Hz, ethenobridge-C); 55.91, 56.10, 56.30 (t, J PC = 10.0 Hz, C9 and C10, i.e. bridgehead-C). 30.91 (acetone-CH3). See Appendix I for detail discussion. 31p NMR (CDCI3, 81 MHz): 5 28.22. IR (KBr): v m a x , 3052 (C-H), 1707 ( C = 0 of acetone), 1459, 1438, 1241, 1204, 1173 ( P = 0 ) , 1118, 1101, 1043, 769, 749, 726, 697, 647, 632, 590, 541, 517 cm"1. MS: m/e (relative intensity), 604 (6.4, M + ) , 527 (100), 511 (64.9), 425 (2.2), 403 (6.9), 201 (23.4), 178 (79.8), 77 (11.7). Calculated mass for C40H30O2P2: 604.1721. Found: 604.1714. UV spectrum: ethanol, X(e x 1000), 270 (8.02), 264 (9.59), 250 (14.0).  1,5-Dichloro-9,10-dihy dro-9,10-ethenoanthracene-l 1,12-bis (dipheny lphosphine oxide) (127) A mixture of 0.427 g (1.10 mmol) of 1,5-dichloroanthracene (118) and 0.264 g (1.08 mmol) of l,2-ethynediylbis(diphenylphosphine oxide) (125) was heated at 190-  205  195 °C for 30 min to afford a crystalline solid. Recrystallization of the solid from a mixture of CHCI3 and acetone yielded 0.34 g (47%) of plates, m.p. 310-312 °C. Anal. Calcd. for C40H28O2P2CI2: C, 71.33; H, 4.19; CI, 10.53. Found: C, 71.37; H, 4.18; CI, 10.33. ! H NMR (200 MHz, CDCI3): 6 7.55-6.70 (26 H, m, Ar-fl), 6.08 (2 H, dd, J PH = 8, 4 Hz). 13  C NMR (CDCI3 200 MHz): 8 52.78, 52.99, 53.19 (t, J PC = 20.4, C9 and C10).  Anthracene substituent, 5 122.54, 126.07 and 126.67 (C2, C3 and C4); 129.48 (CI); 140.76, 145.32 (C4a and C9a). Diphenylphosphinoyl group, 5 127.80, 127.93, 128.06, 128.19, 128.32, 128.45 and 128.58 (m, J PC = 12.8, m-C); 130.32 and 132.44, (d, J PC = 106.8 Hz, P-C); 131.39, 131.53, 131.64, 131.75, 131.81, 132.03, 132.15, 132.25, 132.35 (m, o- and/?-C). Ethenobridge, 5 156.94, 157.07, 158.80 and 158.92 (dd, J PC = 93.3, 6.4 Hz). See Appendix I for detail discussion. 31p NMR (CDCI3, 81 MHz,): 5 27.52. MS: m/e (relative intensity), 674 (M + + 2 , 1.8), 672 (M + , 2.4), 595 (73.3), 579 (57.8), 425 (2.0), 246 (26.7), 201 (30.9), 183 (29.8). c  Calculated mass for  4 0 H 2 8 ° 2 p 2 c l 2 : 672.0942. Found: 672.0939. IR (KBr): v m a x , 3057, 1573, 1455, 1435, 1210, 1178, 1143, 1120, 909, 886, 782,  753, 742, 729, 698, 642, 591, 561, 547, 528, 519, 506, 453 cm"1.  9,10-Dihydro-9, lO-ethenoanthracene-l 1,12-bis(diphenyIphosphine oxide) -1,5dicarboxylic Acid, Dimethyl Ester (128) A mixture of 0.40 g (1.36 mmol) of dimethyl anthracene-1,5-dicarboxylate (122) and 0.58 g (1.36 mmol) of l,2-ethynediylbis(diphenylphosphine oxide) (125) was heated at 200 °C for 1 h to afford a crystalline solid. Recrystallization of the solid from a mixture of CHCI3 and acetone yielded 0.83 g (85%) of colorless crystals, m.p. 303-305 °C.  206  Anal. Calcd. for C44 H34O6P2: C, 73.33; H, 4.76. Found: C, 73.00; H, 4.81. ! H NMR (CDCI3, 200 MHz): 8 7.62-6.80 (28 H, m, Ar-H and bridgehead-//); 3.75 (6 H, s,  -C02CH3).  !3C NMR (CDCI3, 50 MHz): 8 51.99 (-C0 2 CH 3 ), 52.15, 52.36, 52.56 (t, J PC = 20.7 Hz, C9 and CIO). Anthracene substituent, 5 125.32 (C4), 126.00 (CI), 127.05 (C3), 128.17 (C2), 145.07 and 145.12 (C4a and C9a).  Diphenylphosphinoyl group,  8 127.82, 127.94, 128.03, 128.07 and 128.28 (m-C); 130.57, 130.85, 132.69 and 133.01 (2 d, J P C = 108.8, 107.1 Hz, P-C); 131.58, 131.69, 131.79 and 132.16 (0- and p-C); 132.26, 132.35 and 132.51 (t, J PC = 10.0 Hz, o-C). 157.47, 159.22 and 159.34 (dd, J PC = 100.7, 6.4 Hz).  Ethenobridge, 157.34,  166.63 (C0 2 CH 3 ).  See  Appendix I for detail discussion. 31  P NMR (CDCI3, 81 MHz): 8 27.25.  MS: m/e (relative intensity), 720 (M + , 4.0), 689 (3.0), 643 (100.0), 627 (84.8), 519 (5.6), 487 (3.6), 425 (1.7), 360 (4.0), 294 (9.9), 263 (19.3), 201 (18.8), 183 (21.4). Calculated mass for C44H34O6P2: 720.1831. Found: 720.1824. IR (KBr): v m a x , 3052, 1723 ( C = 0 ) , 1591, 1436, 1273, 1186, 1146, 1119, 957, 753, 727, 695, 676, 593, 549, 519 c n r l .  9,10-Dihy dro-9,10-ethenoanthracene-l, 5-dimethanol-l 1,12-bis (diphenylphosphine oxide) (129) A mixture of 1.43 g (6.0 mmol) of anthracene- 1,5-dimethanol (123) and 2.56 g (6.0 mmol) of l,2-ethynediylbis(diphenylphosphine oxide) (125) was refluxed in 15 ml of o-dichlorobenzene for 3 h.  A solid was obtained after solvent was distilled under  reduced pressure and then recrystallized from ethyl acetate to afford 3.8 g (94%) of light pale brown plates. The crystals were shown to be a 1:1 inclusion with ethyl acetate, which decomposed at 130-140 °C overnight to yield a sample for elemental analysis.  207  Anal. Calcd. for C42H34O4P2: C, 75.90; H, 5.16. Found: C, 75.52; H, 5.20. l  H NMR (CDCI3, 200 MHz): 8 7.80-6.88 (26 H, m, Ar-H), 6.20 (2 H, q, J P H =  8, 4 Hz, H9 and H10), 4.50, 3.90 (4 H, AB type q, J H H = 12 Hz, CH2OH), 3.46 (2 H, s, CU2OH). 13  C NMR (CDCI3, 50 MHz): 8 52.21, 52.41, 52.60 (t, J PC = 19.6 Hz, C9 and  CIO); 62.16 ( C H 2 ° H ) ; 157.42, 157.54, 159.30, 159.40 (dd, J PC = 94.0, 5.3 Hz, C l l and C12, i.e. vinyl-C).  Anthracene substituent, 8 123.50, 125.32, 126.24, 136.14  142.56 and 143.65 (C2, C3, C4, CI, C4a and C9a).  Diphenylphosphinoyl group,  8 127.52, 127.65, 127.78 (t, J PC = 13.0 Hz, m-C); 128.54, 128.67, 128.79 (t, J PC = 12.5 Hz, m-C); 129.73, 129.92, 131.86, 132.10 (2 d, J PC = 109.9, 107.3 Hz, C-P); 131.53, 131.64, 131.74, 132.29, 132.39, 132.50 (m, o- andp-C). See Appendix I for detail discussion. 31  P NMR (CDCI3, 81 MHz): 8 30.39.  MS: m/e (relative intensity), 665 (M+ + 1, 22), 664 (M + , 84), 648 (20), 635 (15), 587 (100), 571 (90), 555 (15), 463 (76), 445 (31), 427 (18), 321 (9), 238 (33), 219 (40), 201 (61), 183 (43), 77 (30).  Calculated mass for C42H34O4P2:  664.1932.  Found: 664.1927. IR (KBr): v m a x , 3400-2200 (O-H, and C-H), 1438, 1204, 1142, 1120, 1094, 1076, 747, 725, 691, 548, 541, 522 cm" 1 .  9,10-Dihydro-10-hydroxy-9,10-ethenoanthracene-l 1,12-bis(diphenylphosphine oxide) (130) Ethynediylbis(diphenylphosphine oxide) (125, 5.00 g, 11.7 mmol) and 2.73 g (14.1 mmol) of anthrone (Aldrich) were mixed in a 25 ml pear-shaped flask. The flask was heated in an oil bath to 175 C until all the solid melted and kept at 165 C for another hour. A crystalline solid formed during heating. The solid was recrystallized from a  208  mixture of methylene chloride and acetone to yield 5.63 g (71%) of colorless needles as an inclusion complex with one equivalent of acetone. An analytical sample could be obtained by recrystallizing again from the same solvents, m.p. 241-243 C, with efflorescence at about 100 C. Anal. Calcd. for C ^ t ^ o C ^ ^ H e O :  C, 76.08; H, 5.35. Found: C, 76.05, H,  5.31. l  H NMR (CDC13, 200 MHz): 5 7.80-6.70 (28 H, m, Ar-H), 4.92 (1 H, dd, J PH =  9.4, 3.3 Hz, bridgehead-fl), 2.18 (6 H, s, acetone). 31  P NMR (CDCI3, 81 MHz): 8 41.51 (d, J PP = 13.8 Hz); 27.28 (d, JPP = 13.8  Hz). MS: m/e (relative intensity), 621 (3.8, M+), 620 (8.1), 527 (1.4), 425 (100.0), 408 (23.0), 349 (45.4), 194 (100). Calculated mass for C40H30O3P2: 620.1670. Found: 620.1667. IR (KBr): v m a x , 3400 - 2400 (O-H and C-H), 1705 (C=0), 1457, 1438, 1253, 1224, 1206 (P=0), 1118, 1087, 786, 750, 724, 697, 642, 559, 543, 515 cm"1. UV spectrum: methanol, A,(e x 10"3), 300 (0.48, sh), 274 (0.30, sh), 267 (0.39, sh), 227 (19.3, br.). X-ray crystal structure analysis: crystals from 5'-(-)-2-methyl-l-butanol; formula, c  40 H 30°3 p 2; space group, Fl\ln\ a = 11.557(1) A, b = 21.663(2) A, c = 13.120(2)  A, 0 = 91.01(1), V = 3284.4(6) A 3 , Z = 4; R = 0.041. DSC:  Crystals from (5)-(-)-2-methyl-l-butanol (no inclusion by *H NMR  spectroscopy), AH = 106.8 J/g at T = 240.2-244.2 °C. Crystals from ethanol (no inclusion by lU NMR spectrum), AH} = 5.30 J/g at T = 197.7-203.7 °C; AH2 = 41.44 J/g at 238.8-242.9 °C. DSC indicated that a phase transition of this solid was observed.  209  Ethyl 3-(diphenylphsophine Oxide)-2-propynoate (131) Butyllithium in hexane (1.03 M, 8.91 ml, 9.2 mmol) was added slowly via syringe to a solution of 0.968 g (9.18 mmol) of ethyl 2-propynoate (Aldrich) in 11 ml of dry THF at -78 °C under nitrogen with stirring. The solution was kept at -78 °C for 3 h and then 2.03 g (9.19 mmol) of chlorodiphenylphosophine was dropped in via syringe.  The  reaction mixture was held at -78 °C for 1 h, warmed to room temperature over a period of 1.5 h, quenched with 1 ml of 10% acetic acid and neutralized with sodium bicarbonate. After removing the organic layer, the aqueous layer was extracted with diethyl ether. The combined organic extracts were dried over magnesium sulphate and evaporated.  The oily residue was chromatographed on silica gel.  Ethyl 3-  diphenylphosphine-2-propynoate was eluted with 1:20 EtO Ac/petroleum ether and the solvents were evaporated. The phosphine so obtained as an oil was not characterized, but was oxidized with 10% hydrogen peroxide in acetone to its oxide (131).  An  analytical sample of compound 131 was obtained by chromatography on silica gel (1:1 EtO Ac/petroleum ether). This compound proved to be a colorless oil (overall yield 40%), and attempts to crystallize the oil failed. Anal. Calcd. for C17H15O3P: C, 68.46; H, 5.07. Found: C, 68.19; H, 5.17. !H NMR (200 MHz, CDCI3): 8 1.30 (3 H, t, J HH = 7 Hz, CtfjC^OOC); 4.28 (2 H, q, J H H = 7 Hz, CH3CH2OOC-); 7.40-8.00 (10 H, m, Ar-H). 13c NMR (50 MHz, CDCI3): CH3CH2);  151 71  -  ester group, 5 13.90 (s, CH3CH2); 63.11 (s,  (s. carbonyl). Diphenylphosphinoyl group, 5 128.77, 129.04 (d,  JPC = 13.7 Hz, m-C); 129.76, 132.20 (d, JPC = 122.7 Hz, P-C); 131.00, 131.26 (d, J P c = 11.4 Hz, o-C); 132.94, 133.00 (d, JPC = 3.0 Hz, p-C). 31  P NMR (81 MHz, CDCI3): 5 8.96.  IR (neat): v m a x , 3080-2900 (C-H), 1718 (C=0), 1439, 1240 (P=0), 1124, 727, 694,540 cm"1.  210  MS: m/e (intensity), 298 (29.8, M+), 269 (8.6), 253 (7.8), 242 (10.9), 226 (45.7), 201 (29.2), 179 (96.4), 165 (11.6), 149 (35.9), 129 (97.3), 102 (46.1), 77 (100), 51 (20.0). Calculated mass for C17H15O3P: 298.0759. Found: 298.0751.  Ethyl 9,10-Dihydro-9,10-ethenoanthracene-ll-(diphenylphosphine Oxide)-12carboxylate (132) Anthracene (4.1 g, 23 mmol) was heated with 6.8 g of crude phosphine oxide 131 at 190 °C for 30 min.  The reaction mixture was chromatographed on silica gel.  Compound 132 was eluted with ethyl acetate and crystallized from a mixture of ethyl acetate and petroleum ether to yield 5.8 g of the product, m.p. 183-184 °C. Anal. Calcd. for C31H25O3P: C, 78.14; H, 5.29. Found: C, 77.98; H, 5.36. *H NMR (200 MHz, CDCI3): 8 0.94 (3 H, t, J HH = 7 Hz, C#jCH2OOC); 3.92 (2 H, q, J HH = 7 Hz, CH 3 C#200C); 5.30 (1 H, d, J PH = 9 Hz, bridgehead-^; 5.61 (1 H, d, J PH = 4 Hz, bridgehead-^); 6.80 - 7.38 (18 H, m, Ar-tf). 31  P NMR (81 MHz, CDCI3): 5 27.44  !3c NMR (50 MHz, CDCI3): 8 13.51, 61.52 (CH3CH2-); 54.10, 54.26, 54.32, 54.50 (2 d, J PC = 8.3 Hz, 9.2 Hz, bridgehead-C); 123.82, 125.32, 125.35 (C-l, 2, 3 and 4); 143.50, 143.55, 143.79, 143.83 ( 2 d, JPC = 2.2 Hz, 2.8 Hz, C-4a and 9a); 147.76, 149.62, 155.59, 155.70 (2 d, JPC = 93.6 Hz, 5.5 Hz, C-ll and 12); 165.38, 165.46 (d, J PC = 4.5 Hz, carbonyl). Diphenylphosphinoyl group, 8 128.34, 128.59 (d, JPC = 12.5 Hz, m-C); 131.06, 133.22 (d, JPC = 108.6 Hz, P-C); 131.41, 131.62 (d, JPC = 10.3 Hz, o-C); 131.88, 131.93 (d, J PC = 2.8 Hz, p-Q.  See Appendix I for  detail discussion. IR (KBr): v m a x , 1727 (C=0), 1622 (C=C), 1475, 1460, 1436, 1370, 1282, 1235 (P=0), 1195, 1157, 1119, 1091, 1033, 996, 779, 753, 724, 702, 634, 621, 579, 541, 529 cm"1.  211  UV-Vis spectrum: chloroform, A,(e x 1000)), 279 (2.05), 272 (3.08), 266 (3.71), 243 (6.71). MS: m/e (intensity), 476 (3.6, M + ) , 447 (1.8), 430 (4.6), 230 (7.3), 201 (30.9), 178 (100), 77 (14.1). Calculated mass for C31H25O3P: 476.1541. Found: 476.1533. X-ray crystal structure analysis data: monoclinic; space group, C c ; a = 17.040(2) A, b = 10.354(2) A, c = 14.979 (3) A, (3 = 109.16(1)°, V = 2496.4(7) A3, Z = 4; R = 0.033.  9,10-Dihydro-9,10-ethenoanthracene-ll-(diphenylphosphine Oxide)-12-carboxylic Acid (133) Ethenoanthracene 132 (4.8 g, 10 mmol) in 40 ml of 20% NaOH and 30 ml of ethanol was refluxed for 2 h and then acidified with concentrated HC1 after being cooled to room temperature. The product was extracted with 3 X 60 ml of chloroform. The combined organic extracts were dried over magnesium sulfate and evaporated.  The  white solid obtained was recrystallized from a mixture of chloroform and acetonitrile to yield 3.2 g (66%) of prisms, m.p. 295-298 °C. Anal. Calcd. for C29H21O3P: C, 77.67; H, 4.72. Found C, 77.54; H, 4.74. l  H NMR (200 MHz, CDCI3): 6 4.98 (1 H, d, J P H = 11 Hz, bridgehead-fl); 6.28  (1 H, d, J P H = 4 Hz, bridgehead-/^); 6.68-7.70 (18 H, m, Ar-tf). 13  C NMR (50 MHz, CDCI3): 5 55.03, 55.26, 55.51, 55.68 (2 d, J PC = 8.7 Hz,  11.3 Hz, bridgehead-C); 123.16, 124.32, 125.14, 125.83 (C-l, 2, 3 and 4); 141.88, 141.94, 144.21, 144.21 (2 d, J P C = 3.0 Hz, 2.2 Hz, C-4a and 9a); 144.36, 146.15, 160.28, 160.37 (2 d, J PC = 89.5 Hz, 4.5 Hz, C-ll and 12); 163.93, 164.06 (d, J PC = 6.5 Hz, carbonyl). Diphenylphos-phoryl group, 5 126.35, 128.55 (d, J P C = 110.9 Hz, P-C); 129.09, 129.34 (d, J PC = 12.8 Hz, m-C); 132.39, 132.60 (d, J PC = 10.9 Hz, oC); 133.66, 133.72 (d, J PC = 6.8 Hz,/?-C).  212  31  P NMR (81 MHz, CDCI3): 5 36.29.  IR (KBr): v m a x ,  2000-2800 (w, OH), 1689 (C=0), 1438, 1422, 1161, 1143  (P=0), 1122, 748, 730, 709, 694, 549, 538 cm"1. UV spectrum: chloroform, X(s X 1000), 297 (0.89), 272 (3.66), 266 (4.82), 258 (5.19), 241 (11.30). MS: m/e (intensity), 448 (0.5, M+), 430 (1.7), 404 (100), 325 (3.3), 307 (3.3), 276 (8.0), 226 (9.9), 202 (76.5), 178 (60.8), 149 (28.0), 97 (7.4), 77 (21.6), 57 (16.1), 41 (14.8), 32(11.3). Calculated mass for C29H21O3P: 448.1228. Found: 448.1218.  7>«ns,-9,10-Dihydro-9,10-ethanoanthracene-ll,12-bis(diphenylphosphine oxide) (134) Magnesium tunings (5.58 g, 0.23 mol) were put in a 250 ml of dry flask containing a magnetic stirring bar and flame-dried. To this flask was added 5.00 g (7.54 mmol) of 9,10-dihydro-9,10-ethenoanthracene-ll,12-bis(diphenylphosphine oxide) (126) in 100 ml of methanol that had been dried and freshly distilled over calcium hydride. Effervescence began soon after stirring, and cooling in cold water was sometimes used in order to keep the reaction going smoothly. The mixture was continuously stirred and the magnesium dissolved within one hour. The final solution was hydrolyzed by pouring it into 60 ml of 6 M hydrochloric acid and extracted with 4 x 50 ml of chloroform. The combined organic phase was washed with water and evaporated to result in an almost colorless oil. Crystallization from about 5 ml of ethyl acetate yielded 4.2 g (92%) of prisms, m.p. 271-273 °C. Note 1. The oily product could be used directly for the resolution by dibenzoyltartaric acid. Note 2. Improper reaction and/or workup procedures did not result in the product. For example, hydrolysis of the reaction mixture in 4N HC1 gave a mixture whose 31p  213  NMR spectrum (CDCI3, 81 MHz) showed a number of peaks:  major component  (maybe a dimer), 5 47.15 (d, J PP = 7.48 Hz) and 35.25 (d, J PP = 7.48 Hz); minor component (maybe another dimer), 47.37 (d, J PP = 8.01 Hz) and 33.18 (d, J PP = 8.06); minor component (compound 134), 29.50 (s); and other trace components. The major and one of the minor products were probably dimers of ethanoanthracene 134 via ethano-bridge, a pair of enantiomers and a meso compound. Anal. Calcd. for C40H32O2 P2: C, 79.18; H, 5.32. Found: C, 79.07; H, 5.30. ! H NMR (CDCI3, 200 MHz): 5 7.80 - 6.40 (28 H, m, Ar-H), 4.40 (2 H, d, J PH = 6 Hz, bridgehead-^), 3.90 (2 H, d, J PH = 17 Hz, ethano-tf). !3C NMR (CDCI3, 50 MHz): 8 38.03, 39.38 (d, JPC = 67.1 Hz, ethano-CH); 45.70 (s, bridgehead-CH); 123.01, 125.48, 125.67, 126.44 (4 s, anthracene-CH); 128.00, 128.11, 128.23, 128.31, 128.42, 128.54 (2 pseudo-triplets, JPC = 11.4 Hz, 11.4 Hz, phenyl m-CH); 130.52, 130.60, 130.69 (t, JPC = 8.6, phenyl o-CH); 130.97, 131.52 (2 br. s, phenyl p-CK); 131.63, 131.72, 131.81 (t, J PC = 9.6 Hz, phenyl oCH); 131.88, 132.01, 133.82, 133.94 (2 d, JPC = 96.9, 98.0, P-C); 139.94 (s, anthracene-Cflfltf); 143.20, 143.29, 143.42, 143.55, 143.62 (m, J PP = 6.0 Hz, JPC = 14.9 Hz, -2.6 Hz, anthracene-C™n). See Appendix I for detail discussion. 31  P NMR (CDCI3, 81 MHz): 8 29.69.  MS: m/e (relative, intensity), 608 (1.0, M+), 607 (3.4, M+), 606 (6.6, M+), 428 (0.7), 405 (33.9), 262 (1.6), 227 (10.0), 201 (22.8), 178 (100.0), 149 (5.6), 89 (4.9), 77 (8.1). Calculated mass for C40H32O2P2: 606.1878. Found: 606.1881. IR (KBr): v m a x , 3057 and 3025 (C-H), 1485, 1469, 1460, 1438, 1310, 1196 (P=0), 1115, 748, 720, 702, 594, 566, 545, 528, 503 cnrl. Attempts to synthesize the same compound through the Diels-Alder reaction of 1,2ethenediylbis(diphenylphosphine oxide) with anthracene failed. For example, 0.4843 g (1.13 mmol) of the phosphine oxide and 0.2147 g (1.20 mmol) of anthracene was refluxed in 10 ml of dichlorobenzene for 24 h, and TLC on silica gel, *H and 31p NMR 214  spectra showed that no reaction had occurred.  Due to the high melting point of the  phosphine oxide, heating it with one equivalent of anthracene without solvent yielded no Diels-Alder adduct either.  9,10-Dihydro-a»ft'-l,5-dimethanol-tra«1s-9,10-ethanoanthracene-ll,12-bis(diphenylphosphine oxide) (135) Magnesium (2.0 g, 82 mmol) was flame-dried in a round-bottom flask and stirred with 2.0 g of ethenoanthracene 129 in 75 ml of freshly distilled (over CaH2) methanol until all the magnesium had dissolved (3 h). The final reaction mixture was hydrolyzed in 3N HC1 and extracted with 3 X 75 ml of CHCI3. The combined organic phases were dried over MgSC«4 and evaporated to afford 1.8 g (90%) of light-gray solid.  31  P NMR  spectra showed the crude product to contain about 5% of a by-product (maybe a synform of the product). Recrystallization from a mixture of methanol and CHCI3 yielded an analytical sample, m.p. 271-274 °C (d). Anal. Calcd. for C42H 3 6 0 4 P2-CH30H: C, 73.92; H, 5.77. Found: C, 74.31; H, 5.48. l  H NMR (CDCI3 with - 2 % CD3OD, 200 MHz): 8 8.00-6.80 (26 H, m, Ar-H),  6.12 (2 H, d, J = 6.7 Hz, Ar-H), 5.05 (2 H, d, J = 6.0 Hz, H9 and H10), 4.80 and 4.33 (2 x 2 H, AB q. J = 11.9 Hz, -Ctf 2 OH), 3.83 (2 H, d, J = 17.5 Hz, H l l and H12). 31  P NMR (CDCI3 with - 2 % CD3OD, 81 MHz): 8 31.15. (DMSO-d6, 81 MHz)  8 29.16. !3C NMR ((DMSO-d 6 , 50 MHz): 8 37.28, 38.59 (d, J PC = 66.7 Hz, C l l and C12). Anthracene substituent, 8 40.91 (C9 and C10), 124.01, 124.70 and 125.21 (C2, C3, and C4 or C6, C7 and C8); 135.21 (CI, C5 i.e. C-CH2OH); 139.84 (C4c, C8c); 142.25, 142.36, 142.49, 142.62 and 142.73 (m, J P P = 8.1 Hz, J P C = 15.0, -2.1 Hz,  215  virtual coupling, C4a, C8a). Diphenylphosphinoyl group, 8 127.88, 127.99 and 128.09 (t, J PC = 11.0, m-C); 128.38, 128.49 and 128.60 (t, J PC = 11.3 Hz, m-C); 130.10, 130.18, 130.26 (t, J PC = 8.0, o-C); 130.74 (br. s, p-C); 131.58 (partially resolved multiplets, o- and/>-C); 132.47, 132.72, 134.38, 134.62 (2 d, JPC = 96.2, 95.6 Hz, CP). See Appendix I for detail discussion. MS: m/e (relative, intensity), 667 (M + + 1, 4.0), 666 (M + , 4.8), 665 (4.7), 648 (0.4), 637 (0.9), 590 (0.3), 465 (31.4), 447 (27.6), 429 (17.5), 351 (8.1), 325 (6.5) 238 (80.3), 227 (100.0), 201 (59.8), 178 (24.2), 149 (18.0), 77 (27.30). Calculated mass for C42H36O4P2: 666.2089. Found: 666.2072. IR (KBr): v m a x , 3400-2700 (O-H and C-H), 1439, 1181 (P=0), 1115, 785, 754, 720,701, 549,527 cm"1.  Reduction of Ethenoanthracene 128, 130 by Magnesium in Methanol Attempts to reduce 9,10-dihydro-9,10-ethenoanthracene-ll,12-bis(diphenylphosphine oxide)-1,5-dicarboxylic acid, dimethyl ester (128) and 9,10-dihydro-10-hydroxy-9,10ethenoanthracene-ll,12-bis(diphenylphosphine oxide) (130) with magnesium in methanol failed to give the expected results, i.e. the reduction of the 11,12-double bond. For examples: Compound 128 (0.90 g, 1.25 mmol) and 0.61 g (25 mmol) of magnesium turnings (flame dried) were stirred at room temperature in 40 ml of methanol (freshly dried and distilled over CaH2) for 1 h. During this time the solution became cloudy, then turned yellow and finally yielded a cream colored solution. The final mixture was hydrolyzed in 20 ml of 6 M HC1 and extracted with 3 x 40 ml of CHCI3. The extract mixture showed a -*lp NMR spectrum (81 MHz, CDCI3) having a continuum of peaks between 8 28-36 ppm, which may indicate a polymer formed.  216  A mixture of 0.5 g (0.70 mmol) of compound 130 and 0.57 g (23 mmol) of magnesium turnings (flame dried) was stirred at room temperature in 25 ml of methanol (freshly dried and distilled over CaHj). Effervescence began after a few minutes and the solution turned dark-green. The metal was dissolved within one hour. The workup procedure was the same as that used for compound 128. The mixture obtained showed a 31p NMR (81 MHz, CDCI3) spectrum having a continuum of peaks between 5 28-38 ppm, which may be explained by assuming polymerization happened during the reduction.  Resolution of 9,10-Dihydro-9,10-ethanoanthracene-l 1,12-bis(diphenylphosphine Oxide) (134) Dibenzoyltartaric acid (L-DBT-H^O and D-DBT, Aldrich) was used for the resolution in ethanol. The flow chart below was followed.  The optical purity was  monitored by 31p NMR spectroscopy in 1:3 CDC^/o-dichlorobenzene. The absolute configuration was obtained by correlating it with its bisphosphine-rhodium complex that was determined by X-ray crystallography. (R,R)-134: m.p. 283-285 °C; [a]2D2 = -33.9 (c = 2.00, methanol); optical purity, > 95% ( 31 P NMR shift reagent of D-DBT). (5,5)-134: m.p. 286-287 °C; [a]2D2 = 33.9 (c = 2.00, methanol); optical purity, > 95% ( 31 P NMR shift reagent of D-DBT).  217  Flow Chart for the Resolution of Compound 134: 7.0 mmol of 134 (4.26 g) •<— 7.0 mmol of L-DBT.H20 100 ml of Ethanol \ Solution (64% d.e.)  3.0 g of Complex (>95% d.e.) Reciystallization  5%KOH/Chloroform L-DBT-  r  4.0mmolofD-DBT  30 ml of Ethanol  -75 ml of Ethanol-  1  2.6g,mp. >230°C(d) Solution DBT  3.2 g of Complex (>95% d.e.)  5%KOH/Chloroform  K-1.1 mnclof L-DBTH 2 0  Reciystallization from 50 ml of Ethanol  25 ml of Ethanol 2.9g,mp. >230°C(d) Solution Crystalline Solid 5%KOH/Chloroform Reciystallization I from Ethanol 1 D-DBT-*0.52 g 1  3.1 g 5%KOH/Chloroform . L-DBT-^ 1 Crude (->134 I Recrystalli2ation from 1 Diethyl ether/acetone 1.8gof(->134  Crude(+>134 I Recrystallization from 1 Diethyl ether/acetone 1.5gof(+>134  9,10-Dihydro-9,10-ethanoanthracene-l 1,12-bis(diphenylphosphine) (Anthraphos, 136) The dioxide 134 (2.00 g, 3.0 mmol) and 7.0 ml of SiHCl3 (Aldrich) in 20 ml of dry benzene (freshly distilled over phosphorus pentoxide) were heated at 100 °C for 7 h. The final solution was hydrolyzed with 50 ml of 25% aqueous NaOH. The benzene layer was separated and the aqueous layer was extracted with benzene three times. The combined benzene layer was dried over MgSCty, concentrated and passed through a short 218  silica gel column.  The oily product eluted with benzene from the column was  crystallized from a mixture of diethyl ether and petroleum ether to yield 1.44 g (76%) of a crystalline colorless solid, m.p. 148-150 °C. The compound was stable as a solid, but slowly oxidized in solution. All the solvents used above were degassed and the solution of the compound was handled under nitrogen by using a glove bag (Aldrich). Anal. Calcd. for C40H32P2: C, 83.59; H, 5.62. Found: C, 83.33; H, 5.80. !H NMR (CDCI3, 200 MHz): 5 7.60-6.68 (28 H, m, Ar-fl); 4.45 (2 H, pseudotriplet, J PH = 5 Hz, bridgehead-fl); 3.19 (2 H, m, J PH = 22 Hz, ethano-fl). 31p NMR (CDCI3, 81 MHz): 8 0.00. MS: m/e (relative intensity), 575 (2.2), 574 (4.6, M+), 498 (0.3), 397 (31.7), 397 (31.7), 396 (100), 319 (7.3), 287 (14.6), 262 (16.0), 204 (22.1), 203 (29.1), 202 (16.7), 185 (32.0), 183 (49.6), 178 (26.9), 152 (8.9), 133 (17.2), 108 (27.1). Calculated mass for C40H32P2: 574.1979. Found: 574.1982. IR (KBr): v m a x , 3067 and 3024 (C-H), 1480, 1468, 1433, 764, 742, 695, 507, 477 cm~l. R,R-ior:m: m.p. 157 -158 °C. [a]2D4 = 106° (c = 1.0, CHCI3). ^S-form: m.p. 158.5-159.5 °C. [a]2D4 = -106° (c = 1.0 CHCI3). The absolute configuration was determined through the X-ray crystallography of its complex with rhodium (I), [Rh(Anthraphos)(COD)].  219  HI. SYNTHESIS OF COMPLEXES  (i)  Complexes of Triphenylphosphine Oxide (TPPO) with Dibenzobarrelene Derivatives All these complexes were prepared by dissolving equimolar amounts of  triphenylphosphine oxide and dibenzobarrelene derivative in the solvent(s) as indicated to reach the approximate concentration (0.1 M).  Crystalline complexes were obtained  through slow evaporation of the solvent at room temperature. Crystals were collected by suction filtration.  The composition of all complexes was confirmed by ^H NMR  spectroscopy (CDCI3, 200 MHz), which simply consisted of the components in the molar ratio as noted. TPPO/9,10-Dihydro-9,10-ethenoanthracene-ll,12-dicarboxylic Acid, 11-Ethyl Ester (105)/Benzene (2:2:1) (105tb) Solvent: benzene. Prisms, m.p. = 79-80 °C. Anal. Calcd. for Cg2H680ioP2: c> 7 7 - 2 2 ;  H  > 5 - 3 7 - Found: C, 77.29; H, 5.52.  IR (KBr): v m a x , 3200-2300 (O-H and C-H), 1723 (C=0), 1697 (C=0), 1636 (C=C), 1459, 1438, 1317, 1288, 1244, 1219, 1174 (P=0), 1121, 1049, 754, 724, 697, 682, 543 cm"1. MS (EI): mle (intensity), 320 (9.6, M + of 105), 278 (51.0, M+ of TPPO), 277 (100), 247 (23.9), 202 (31.6). MS (FAB): m/e, 599 (MH) + , i.e. the complex of the oxide with the substrate; 557, 414.  220  TPPO/9,10-Dihydro-9,10-ethenoanthracene-ll,12-dicarboxylic Acid, 11-Ethyl Ester (105)/ToIuene (2:2:1) (105tt) Solvent: toluene. Prisms, m.p. = 80-83 °C (d). Anal. Calcd. for C 8 3H 7 o0 10 P2: C, 77.32; H, 5.47. Found: C, 77.24; H, 5.39. IR (KBr): v m a x , 3400-2100 (O-H and C-H), 1723 (C=0), 1697 (C=0), 1636 (C=C), 1459, 1438, 1319, 1289, 1245, 1219, 1174 (P=0), 1120, 1094, 1050, 755, 723, 696, 544 cm"1. MS (EI): m/e (intensity), 320 (1.5, M+ of 105), 278 (30.5, M+ of TPPO), 277 (100), 276 (12.3), 247 (10.7), 230 (13.8), 202 (48.9), 178 (20.8), 152 (26.0), 101 (13.6), 88 (9.0), 77 (28.6). X-ray crystal structure analysis data: empirical formula, C83H70O10P2, colorless plate (0.050 x 0.200 x 0.350 mm), monoclinic, space group P2j/n, a = 8.546(2) A, b = 28.164(5) A, c = 14.368(2) A, 3 = 98.81(2)°, V = 3417(2) A3, Z = 2, D c a l c d . 1.253 g/cirP, R = 0.052. Toluene molecules are disordered inside the crystal. TPPO/9,10-Dihydro-9,10-ethenoanthracene-ll,12-dicarboxylic Acid, 11-Ethyl Ester (105) (1:1) (105t) Solvent: mixed xylenes. Colorless prisms, m.p. = 137-139 °C. Anal. Calcd. for C38H 31 0 5 P: C, 76.24; H, 5.22. Found: C, 76.49, H, 5.37. IR (KBr): v m a x , 3200-2200 (O-H and C-H), 1717 (C=0), 1703 (C=0), 1635 (C=C), 1438, 1331, 1292, 1257, 1222, 1199, 1180 (P=0), 1156, 1121, 751, 723, 696, 614,540 cm-1. MS (EI): m/e (intensity), 320 (3.8, M+ of 105), 278 (41.9, M + of TPPO), 277 (100), 247 (16.8), 202 (45.2), 178 (19.1), 152 (18.3), 101 (8.8), 77 (26.6).  221  MS (FAB): m/e, 599 (MH) + , i.e. the complex of the oxide with the substrate, 557, 479, 321, 303, 295, 279. X-ray crystal structure analysis data: empirical formula, C38H31O5P; colorless prism, (0.100 x 0.200 x 0.400 mm); triclinic, space group, P\; a = 13.501(1) A, b = 14.845(1) A, c = 8.5190(6) A, a = 102.208 (7) , (3 = 98.676(7), y = 68.697(6) , V = 1549.6(5) A3, Z = 2, D c a i c d - = 1.283 g/cm3, R = 0.039. TPPO/9,10-Dihydro-9,10-ethenoanthracene-ll,12-dicarboxyUc Acid, 11Methylethyl Ester (106) (1:1) (106t) Solvents: mixed xylenes or cyclohexane. Small prisms, m.p. = 148-149 °C. Anal. Calcd. for C3 9 H 5 30 5 P: C, 76.46: H, 5.43. Found: C, 76.26; H, 5.39. IR (KBr): v m a x , 3200-2200 (O-H and C-H), 1718 (C=0), 1698 (C=0), 1631 (C=C), 1459, 1438, 1274, 1221, 1171 (P=0) 1154, 1121, 755, 724, 696, 543 cm"1. MS (EI): m/e (intensity), 334 (5.7, M+), 278 (44.7, M+ of TPPO), 277 (100), 248 (20.0), 247 (22.7), 203 (25.6), 202 (20.6), 178 (13.3), 152 (8.1), 101 (2.5), 77 (13.2). MS (FAB): m/e, 613 (MH) + , 579, 557, 479, 414, 335, 279.  (ii) Complexes with l,2-Ethanediylbis(diphenyIphosphine oxide) (DPPEO, 101) with Dibenzobarrelene Derivatives These complexes were prepared by dissolving an 1:2 molar ratio of compound 101 to dibenzobarrelene derivative in either toluene or benzene.  The crystals formed upon  solvent evaporation were filtered. The composition of the complexes was confirmed by their lU NMR spectra (CDCI3, 200 MHz).  222  DPPEO (101)/9,10-Dihydro-9,10-ethenoanthracene-ll,12-dicarboxyUc Acid, 11Ethyl Ester (105) (1:2) (105d) Solvents: toluene or mixed xylenes. Shiny prisms, m.p. = 209-211 °C. Anal. Calcd. for C66H5 6 O 10 P2: C, 74.00; H, 5.27. Found: C, 74.08; H, 5.27. IR (KBr,):  v m a x , 3200-2200 (O-H and C-H), 1734 ( C = 0 ) , 1698 (C=0), 1636  (C=C), 1459, 1439, 1286, 1212, 1174, 1158 ( P = 0 ) , 1124, 1096, 1045, 749, 731, 696, 531 cm" 1 . MS (EI): m/e (intensity), 705 (2.2), 629 (0.6), 554 (6.1), 553 (15.8), 430 (2.0, M + of DPPEO), 429 (5.5), 353 (100), 320 (1.3, M + of 105), 276 (11.9), 247 (6.7), 229 (64.2), 202 (60), 178 (52.5), 151 (15.6), 77 (34.7). MS (FAB): m/e, 1071 (MH) + , i.e. the complex of the oxide with two equivalents of the substrate; 993, 950, 915, 883, 861, 831, 783, 749, 706, 677, 631, 582, 566, 537, 506,431 (DPPEOH) + . DPPEO (101)/9,10-Dihydro-9,10-ethenoanthracene-ll,12-dicarboxylic Acid, 11Methylethyl Ester (106) (1:2) (106d) Solvents: toluene, or mixed xylenes. Shiny small prisms, m.p. = 180-181 °C. Anal. Calcd. for C 6 8 H 6 0 OioP2: C, 74.31; H, 5.50. Found: C, 74.29, H,5.49. IR (KBr): v m a x , 3200-2200 (O-H and C-H), 1714 (sh) ( C = 0 ) , 1701 (C=0), 1633 (C=C), 1460, 1438, 1292, 1270, 1223, 1182, 1151 ( P = 0 ) , 1124, 1103, 1086, 747, 731 cm" 1 . DPPEO (101)//? Nitrophenol (1:2) (lOln) Solvent: toluene. Large prisms, m.p. = 144-145 °C.  223  Anal. Calcd. for C3 8 H 34 N20 8 P2:  C, 64.41;  H, 4.84; N, 3.95. Found: C,  64.20; H, 4.81; N, 3.84. IR (KBr): v m a x , 3400-2200 (O-H and C-H), 1592, 1499, 1438, 1332, 1294, 1248, 1173, 1163 (P=0), 1110, 859, 729, 694, 532, 511 cm"1. X-ray crystal structure analysis data: empirical formula, C3gH34N20gP2; colorless prism, 0.060 x 0.080 x 0.250; triclinic, space group P\; a = 13.771(3) A, b = 15.188(2) A, c = 9.189(3) A, a = 98.39(2) , (5 = 100.93(2), y = 72.38(1) , V = 1790.5 A3, Z = 2, D ca i cd . = 1.414 g/cm3; R = 0.041.  (iii) Crystalline Complexes of the Optically Active Bisphosphine Dioxides with Dibenzobarrelene Derivatives These complexes were prepared in the same way as their achiral analogs. Their composition, which was determined by 200 MHz ^H NMR spectroscopy, was 2:1 of bisphosphine dioxide to dibenzobarrellene, as expected. (S,S)-(+)-l,2-Ethanediylbis((4'-methylphenyl)phenylphosphine oxide) (S,S114)/9,10-Dihydro-9,10-ethenoanthracene-ll,12-dicarboxyUc Acid, 11-Ethyl Ester (105) (1:2) (5,£-105p) Solvent: toluene. Prisms, m.p. = 191-193 °C. Anal. Calcd. for C68H60O10P2; C, 74.31; H, 5.50. Found: C, 74.58; H, 5.53. IR (KBr): v m a x , 3200-2200 (O-H and C-H), 1729 (C=0), 1692 (C=0), 1635 (C=C), 1460, 1438, 1268, 1218, 1154 (P=0), 1123, 1097, 1049, 745, 728, 518 cm"1. X-ray crystal structure analysis data: empirical formula, C6gH6()OioP2; colorless prism, size, 0.25 x 0.200 x 0.500, monoclinic, space group, P2\\ a = 8.988(4) A, b = 12.448(3) A, c = 26.452(4) A, f5 = 90.74(3) ; Z = 2; R = 0.053.  224  (5,5)-(+)-l,2-Ethanediylbis((4'-methylphenyl)phenylphosphine oxide) (5,5114) 19,10-Dihydro-9,10-ethenoanthracene-l 1,12-dicarboxylic Acid, 11-Methylethyl Ester (106) (1:2) (5,5-106p) Solvent: toluene. Prisms, m.p. = 185-187 °C. Anal. Calcd. for C 7 oH 6 40 10 P2: C, 74.59; H, 5.72. Found: C, 74.76; H, 5.69. IR (KBr): v m a x , 3200-2200 (O-H and C-H), 1719 (C=0), 1692 (C=0), 1634 (C=C), 1460, 1275, 1222, 1154 (P=0), 1122, 1102, 1046, 727, 520 cm"1. (5,5)-(+)-l,2-Ethanediylbis((3'-methylphenyl)phenylphosphine oxide) (5, 5115)/9,10-Dihydro-9,10-ethenoanthracene-ll,12-dicarboxyIic Acid, 11-Ethyl Ester (105) (1:2) (5,5-105m) Solvent: toluene. Prisms, m.p. = 188.5-190 °C. Anal. Calcd. for C 68 H 60 Oi 0 P2: C, 74.31; H, 5.80. Found: C, 74.19; H, 5.80. IR (KBr): v m a x , 3200-2200 (O-H and C-H), 1729 (C=0), 1694 (C=0), 1635 (C=C), 1460, 1290, 1266, 1219, 1155 (P=0), 1119, 742 cm"1. X-ray crystal structure analysis data: empirical formula, C6gH6oOioP2! colorless prism (0.150 x 0.150 x 0.300), monoclinic, space group, Pl\\ a = 9.117(3) A, b = 12.100(2) A, c = 26.602(2) A, p = 94.95(1); Z = 2; R = 0.063. (5,5)-(+)-l,2-Ethanediylbis((3',4'-dimethylphenyl)phenyIphosphine oxide) (5,5116)/9,10-Dihydro-9,10-ethenoanthracene-ll,12-dicarboxylic Acid, 11-Ethyl Ester (105) (1:2) (5,5-105xo) Solvent: toluene or mixed xylenes. Prisms, m.p. = 178-179 °C. Anal. Calcd. for C V Q H ^ C ^ ; C, 74.59; H, 5.72. Found: C, 74.71; H, 5.72.  225  IR (KBr): v m a x , 3200-2200 (O-H, C-H), 1730 (C=0), 1691 ( C = 0 ) , 1635 (C=C), 1459, 1265, 1217, 1201, 1150 ( P = 0 ) , 1113, 758, 728, 555, 512 cm" 1 . X-ray crystal structure analysis data: empirical formula, C70H64O2P2; colorless prism, size, (0.100 x 0.200 x 0.500); monoclinic, space group, P2y, a = 9.184(6) A, b = 12.515(3)A, c = 26.567(5) A, 0 = 91.44(3) ; Z = 2; R = 0.051. (S,5)-(-)-l,2-Ethanediylbis((3',5'-dimethylphenyl)phenylphosphine oxide) (S,S117)/9,10-Dihydro-9,10-ethenoanthracene-ll,12-dicarboxylic Acid, 11-Methylethyl Ester (106) (1:2)  (S,S-l06xo)  Solvent: toluene, m.p. = 173-174 °C. Anal. Calcd. for C 7 2 H 6 8 0 2 P 2 : C, 74.86; H, 5.93. Found: C, 75.02; H,5.95. IR (KBr):  v m a x , 3200-2200 (O-H and C-H), 1720 ( C = 0 ) , 1692 (C=0), 1634  (C=C), 1459, 1274, 1201, 1180, 1149 ( P = 0 ) , 1112, 1045, 758, 727, 554 cm"1. (R,R)- or (S,,S)-(-)-l,2-Ethanediylbis((3',5,-dimethylphenyl)phenylphospblne oxide) (S,S- or 7?,/?-117)/9,10-Dihydro-9,10-etheiioaiithraceiie-ll,12-dicarboxyUc Acid, 11Ethyl Ester (105) (1:2) (£S-105xm or R,R-W5xm) Solvent: mixed xylenes or toluene. Crystalline solid, m.p. = 158-159 °C. Anal. Calcd. for C7oH 6 4 0 2 P2 of 5,5-105x111; C, 74.59;  H, 5.72.  Found:  C,  74.87; H, 5.72. IR (KBr):  v m a x , 3200-2200 (O-H and C-H), 1720 ( C = 0 ) , 1698 ( C = 0 ) , 1635  (C=C), 1459, 1438, 1272, 1218, 1154 ( P = 0 ) , 1109, 1052, 742, 562 cm" 1 . Complex: (/?,i?)-(-)-l,2-Ethanediylbis((3',5'-dimethylpheiiyI)phenylphosphiiie oxide) (S,S-117)//?-Nitrophenol (1:2) (R,R-117n) Solvent: a mixture of benzene and cyclohexane.  226  Light yellow prisms, m.p. = 125-126 °C. Anal. Calcd. for C42H 4 2N20 8 P 2 : C, 65.96; H, 5.54; N, 3.66. Found: C, 65.72; H, 5.56; N, 3.59. IR (KBr): v m a x , 3400-2200 (O-H and C-H), 1592, 1516, 1499, 1468, 1335, 1298, 1252, 1184, 1153 (P=0), 1128, 1110, 1088, 852, 756, 724, 694, 562, 532cm"l. X-ray crystal structure analytical data: empirical formula, C42H42N2O8P2; yellow prism, size, (0.150 X 0.250 x 0.400); monoclinic, space group, Ply a = 9.799(2) A, b = 20.961 A, c = 9.860 A, 0 = 91.19(2), V = 2024.7 A3, Z = 2, D ca i cd- = 1.524 g/cm3; R = 0.047.  (iv) Inclusion Compounds of 9,10-Dihydro-9,10-ethenoanthracene-ll,12bis(diphenylphosphine oxide) (126); 9,10-Dihydro-9,10-ethenoanthracene-l,5dimethanol-ll,12-bis(diphenylphosphine oxide) (129) and 9,10-Dihydro-10hydroxy-9,10-ethenoanthracene-ll,12-bis(diphenylphosphine oxide) (130) These complexes were prepared by recrystallization from the guest solvents or the guest solvents together with methylene chloride, which helps to dissolve the host. The ratio of host to guest was determined by ^H NMR spectroscopy in CDCI3 on a 200 MHz machine; FTIR spectroscopy and elemental analyses were also used to characterize the complexes. In some cases, TGA and DSC data were also collected. 9,10-Dihydro-9,10-ethenoanthracene-l 1,12-bis(diphenylphosphine oxide) (126)/Acetone (1:1): needles from acetone/CH2Cl2Anal. Calcd. for C40H30O2P2-C3H6O: C, 77.93; H, 5.48. Found: C, 77.72, H, 5.45. *H NMR (CDCI3, 200 MHz): 8 7.48-6.98 (28 H, m Ar-#); 5.68 (2 H, dd, J PH 12 Hz and 4 Hz, bridgehead-^); 2.12 (6 H, s, acetone).  227  IR (KBr): v m a x , 3052 (C-H), 1707 (C=0 of acetone), 1459, 1438, 1241, 1204, 1173 (P=0), 1118, 1101, 1043, 769, 749, 726, 697, 647, 632, 590, 541, 517 cm"1. TGA: Calcd. weight loss for one equivalent of acetone: 8.76%. Found: 8.78% (Ti = 139.1 °C, T 2 = 151.6 °C, T m a x - 147.8°C). DSC: AHX = 73.89 J/g at T = 117.7-135.0 °C; AH2 = 70.09 J/g at T = 234.8236.8°C. 9,10-Dihydro-9,10-ethenoanthracene-ll, 12-bis(diphenylphosphine oxide) (126) /2-Butanone (1:1): needles from 2-butanone. Anal. Calcd. for C44H3 8 0 3 P 2 : C, 78.09; H, 5.66. Found: C, 77.91; H, 5.76. ! H NMR (CDC13, 200 MHz): compound 126, 8 7.48-6.98 (28 H, m Ar-fl), 5.68 (2 H, dd, J PH = 12 Hz, 4 Hz, bridgehead-fl); 2-butanone, 5 2.46 (2 H, q, J = 7 Hz, C(0)C# 2 CH 3 ), 2.12 (3 H, s, C(0)CH3), 1.08 (3 H, t, J = 7 Hz, C(0)CH 2 C#j). IR (KBr): v m a x , 3052 (C-H), 1706 (C=0), 1438, 1205, 1173 (P=0), 1118, 750, 726, 698, 632, 538,516 cm"1. 9,10-Dihydro-9,10-ethenoanthracene-ll, 12-bis(diphenylphosphine oxide) (126) /2-Pentanone (1:1): needles from 2-pentanone. Anal. Calcd. for C45H40O3P2: C, 78.25; H, 5.84. Found: C, 78.40; H, 5.85. *H NMR (CDCI3, 200 MHz): compound 126, 5 7.48-6.98 (28 H, m, Ar-fl), 5.68 (2 H, dd, J PH = 12 Hz, 4 Hz, bridgehead-fl); 2-pentanone, 8 2.41 (2 H, t, J = 7 Hz, C(0)C//2CH2CH3), 2.14 (3 H, s, C(0)CH3), 1.60 (2 H, m, J = 7 Hz, C(O)CH 2 C# 2 CH 3 ), 0.92 (3 H, t, J = 7 Hz, C(0)CH 2 CH 2 C#?). IR (KBr): v m a x , 3052 (C-H), 1707 (C=0), 1458, 1437, 1205, 1175 (P=0), 1118, 750, 726, 697, 631, 591, 538, 516 cm"1.  228  9,10-Dihydro-9,10-ethenoanthracene-ll, 12-bis(diphenylphosphine oxide) (126) /3-Pentanone (1:1): needles from 3-pentanone. Anal. Calcd. for C45H40O3P2: C, 78.25; H, 5.84. Found: C, 78.30; H, 5.97. *H NMR (CDCI3, 200 MHz): compound 126, 8 7.48-6.98 (28 H, m, Ar-fl); 5.68 (2 H, dd, J PH = 12 Hz, 4 Hz, bridgehead-tf); 3-pentanone, 8 2.42 (4 H, q, J = 7 Hz, C(0)CH2CH3), 1.08 (6 H, t, J = 7 Hz, C(0)CH2C#3). 9,10-Dihydro-9,10-ethenoanthracene-l 1,12-bis(diphenyIphosphine oxide) (126)/EthyI Acetate (1:1): needles or plates from ethyl acetate. Anal. Calcd. for C 44 H3 8 04P 2 : C, 76.29; H, 5.53. Found: C, 76.26; H, 5.40. *H NMR (CDCI3, 200 MHz): compound 126, 8 7.48-6.98 (28 H, m, Ar-H); 5.68 (2 H, dd, J PH = 12 Hz, 4 Hz, bridgehead-fl); ethyl acetate, 8 3.70 (2 H, q, J = 7 Hz, OC# 2 CH 3 ), 2.18 (3H, s, CH3C(0)), 1.20 (3 H, t, J = 7 Hz, OCn2CH3). IR (KBr): v m a x , 3058 (C-H), 1730 (C=0), 1459, 1438, 1204, 1170 (P=0), 1119, 749, 726, 697, 632, 590, 540, 518 cm"1. X-ray crystal structure analysis data: empirical formula, C44H38O4P2; colorless plates, crystal dimensions, 0.10 x 0.050 x 0.300; monoclinic, space group P2\/c, a = 18.223(2) A, b = 9.435(3) A, c = 21.946(2) A, 0 = 103.430(7)°, V = 3670.2(9) A 3 , Z = 4, D c a l c ( L = 1.254 g/cm3; R = 0.049. 9,10-Dihydro-9,10-ethenoanthracene-l 1,12-bis(diphenylphosphine Oxide (126)/Ethanol (1:1): prisms from ethanol. Anal. Calcd. for C42H36O3P2: C, 77.53; H, 5.58. Found: C, 77.40; H, 5.53. *H NMR (CDCI3, 200 MHz): compound 126, 8 7.48-6.98 (28 H, m, Ar-tf); 5.68 (2 H, dd, J PH = 12 Hz, 4 Hz, bridgehead-^); ethanol: 8 3.70 (2 H, q, J = 7 Hz, C// 2 CH 3 ), 1.88 (1 H, s, OH), 1-25 (3 H, t, J = 7 Hz, C H 2 C ^ ) . IR (KBr): v m a x , 3376 (O-H), 3058 (C-H), 1438, 1206, 1168 (P=0), 1120, 1101, 769, 750, 725, 692, 648, 632, 590, 541, 517 cm"1. 229  DSC: AHt = 4.17 J/g at T = 98.9-117.6 °C; AH2 = 40.28 J/g at T = 151.9166.6 °C; AH3 = 53.43 J/g at T = 234.1-236.7 °C. X-ray crystal structure analysis data: empirical formula, C42H36O3P2, colorless prism; crystal dimensions 0.200 X 0.200 x 0.400 mm, orthorhombic, space group Fl\2\l\; 4  a = 17.983(2) A, b = 20.682(3) A, c = 9.384(6) A, V = 3490(2) A 3 , Z =  ' Dcalcd. = 1-238 g/cm3; R = 0.038. The absolute configuration was (115,125), and  for (11/?, 12/?), the R-value was 0.046. 9,10-Dihydro-9,10-ethenoanthracene-l 1,12-bis(diphenylphosphine oxide) (126) /2-Propanol (1:1): prisms from isopropanol. Anal. Calcd. for C43H3 8 0 3 P 2 : C, 77.70; H, 5.76. Found: C, 77.54; H, 5.73. lH NMR (CDCI3, 200 MHz): compound 126, 5 7.48-6.98 (28 H, m, Ar-tf); 5.68 (2 H, dd, J PH = 12 Hz, 4 Hz, bridgehead-//); wo-propanol, 8 4.02 (1 H, septet, J = 7 Hz, Ctf(CH3)2), 2.20 (s, OH); 1.22 (6 H, d, J = 7 Hz, CH (CH3)i). IR (KBr): v m a x , 3381 (O-H), 3059 (C-H), 2967 (C-H), 1475, 1463, 1438, 1206, 1166 (P=0), 1120, 1101, 959, 768, 751, 725, 692, 648, 632, 590, 541, 517 cm"1. DSC: AHi = 5.50 J/g at T = 108.4-123.3 °C; AH2 = 50.99 J/g at T = 146.6161.0 °C; DH 3 = 56.61 J/g atT = 233.1-236.4 °C. X-ray crystal structure analysis data: empirical formula, C43H3gC>3P2; colorless prism (0.100 x 0.100 x 0.100); orthorhombic, space group Fl\l\l\;  a = 18.076(2)  A, b = 20.904(2) A, c = 9.432(6) A, V = 3563 (1) A 3 , Z = 4, D c a l c d = 1.239 g/cm3; R = 0.047. 9,10-Dihydro-9,10-ethenoanthracene-l 1,12-bis(diphenylphosphine oxide) (126) /1-Propanol (1:1) prisms from n-propanol. Anal. Calcd. for C43H3803P2: C, 77.70; H, 5.76. Found: C, 77.88; H, 5.69. lH NMR (CDCI3, 200 MHz): compound 126, 8 7.52-6.90 (28 H, m, Ar-fl); 5.68 (2 H, dd, J PH = 12 Hz, 4 Hz, bridgehead-//); n-propanol, 8 3.61 (t, J HH = 7 Hz, 230  CH 3 CH 2 C# 2 OH), 1.54 (2 H, m, J HH = 7, CH3C#2CH2OH), 1.26 (1 H, s, CH3CH 2 CH 2 0#), 0.97 (3 H, t, J HH = 7 Hz, C^jCH 2 CH 2 OH). IR (KBr) v m a x , 3369 (O-H), 3058 (C-H), 1459, 1438, 1205, 1168 (P=0), 1120, 1101, 769, 750, 725, 692, 648, 632, 590, 541, 518 cm-1. X-ray crystal structure analysis data: empirical formula, C43H3gC»3P2; space group Fl^iLi,  a = 18.191(1) k,b  = 20.748(2) A, c = 9.436(2) A, V = 3561.5 (6) A 3 , Z  = 4 , Dcalcd = 1-240 g/cm3; R = 0.040. 9,10-Dihydro-9,10-ethenoanthracene-ll, 12-bis(diphenyIphosphine oxide) (126) /2-Methyl-2-Propanol (1:1): needles from terf-butanol. Anal. Calcd. for C44H40C>3P2: C, 77.86; H, 5.94. Found: C, 77.68; H, 5.84. J  H NMR (CDC13, 200 MHz): compound 126, 5 7.48-6.98 (28 H, m, Ar-fl), 5.68  (2 H, dd, J PH = 12 Hz, 4 Hz, bridgehead-/?); terf-butanol, 5 2.72 (1 H, s, OH), 1.28 (9 H, s, CH3). IR (KBr): v m a x , 3392 (O-H), 3058 (C-H), 2970 (C-H), 1463, 1438, 1205, 1165 (P=0), 1121, 1101, 768, 751, 725, 693, 648, 632, 590, 541, 517 cm"1. 9,10-Dihydro-9,10-ethenoanthracene-ll,12-bis(diphenylphosphine oxide) (126) /Tetrahydrofuran (1:1): small needles from a mixture of THF and CH2C12. Anal. Calcd. for C4 4 H3 8 0 3 P 2 : C, 78.09; H, 5.66. Found: C, 77.87; H, 5.70. ! H NMR (CDCI3, 200 MHz): compound 126, 5 7.48-6.98 (28 H, m Ar-#), 5.68 (2 H, dd, 12 Hz, 4 Hz, bridgehead-//); THF, 5 3.78 (4 H, t, J = 4 Hz, -OCtf2CH2-), 1.82 (4 H, t, J = 4 Hz, -OCH2C//2-). IR (KBr): v m a x , 3048 (C-H), 1459, 1438, 1208, 1173 (P=0), 1118, 750, 726, 699, 631, 590,538,516 cm' 1 .  231  9,10-Dihydro-9,10-ethenoanthracene-ll, 12-bis(diphenylphosphine oxide) (126)/1,2Dimethoxyethane (1:1): needles from a mixture of 1,2-dimethoxyethane and CH2CI2. Anal. Calcd. for C44H40O4P2: C, 76.07; H, 5.80. Found: C, 76.08; H, 5.77. lH NMR (CDCI3, 200 MHz): compound 126, 8 7.48-6.98 (28 H, m, Ar-fl), 5.68 (2 H, dd, J = 12 Hz, 4 Hz, bridgehead-tf),  1,2-dimethoxyethane, 8 3.57 (4 H, s, -  CH2), 3.41 (6 H, s, OCH3). IR (KBr): v m a x , 3056 (C-H), 1459, 1438, 1207, 1164 ( P = 0 ) , 1119, 1101, 959, 750, 726, 695, 647, 632, 590, 539, 517 cm" 1 . 9,10-Dihydro-9,10-ethenoanthracene-l,5-dimethanol-ll,12-bis(diphenylphosphine oxide) (129)/Ethanol (3:2) from ethanol. Anal. Calcd. for 3C42H34O4P22C2H5OH: C, 74.84; H, 5.51. Found: C, 74.68; H, 5.43. lH NMR (CDCI3, 200 MHz): compound 129, 8 7.80-6.88 (26 H, m, Ar-fl), 6.20 (2 H, dd, J P H = 8, 4 Hz, H9 and H10), 4.50 (2 H, d, J H H = 12 Hz, CH2OH), 3.90 (2 H, dd, J = 12 Hz, 10 Hz, C# 2 OH), 3.46 (2 H, t, J = 10 Hz, CH 2 Ofl); ethanol, 8 3.70 (2/3 H, q, J = 7 Hz, CH 2 CH 3 ); 1.50 (1/3 H, br. s, OH); 1.25 (1 H, t, J = 7 Hz, CH 2 C// 5 ). 9,10-Dihydro-9,10-ethenoanthracene-l ,5-dimethanol-l 1,12-bis(diphenylphosphine oxide) (129)/Ethyl Acetate (1:1) from ethyl acetate. Anal. Calcd. for C 4 6 H420 6 P 2 : C, 73.40; H, 5.62. Found: C, 73.07; H, 5.32. lH NMR (CDCI3, 200 MHz): compound 129, 8 7.80-6.88 (26 H, m, Ar-fl), 6.20 (2 H, dd, J P H = 8, 4 Hz, H9 and H10), 4.50, 3.90 (4 H, 2 d of AB type, J H H = 12 Hz, CH2OH), 3.46 (2 H, s, CH 2 0tf); OCtf 2 CH 3 ), 2.18 (3 H, s, CH3C(0)), IR (KBr):  ethyl acetate, 8 3.70 (2H, q, J = 7 Hz,  1.20 (3H, t, J = 7 Hz, OCH2CH5).  v m a x , 3400-2200 (C-H and O-H), 1737 ( C = 0 ) , 1438, 1204, 1141,  1120, 1094, 1076, 1002, 748, 725, 692, 584, 541, 522 cm" 1 . 232  9,10-Dihydro-10-hydroxy-9,10-ethenoanthracene-l 1,12-bis(diphenylphosphine oxide) (130)/Acetone (1:1) from acetone/CH2Cl2Anal. Calcd. for C43H36O4P2: C, 76.08; H, 5.35. Found: C, 76.05, H, 5.31. *H NMR (CDCI3, 200 MHz): 8 7.80-6.70 (28 H, m, Ar-fl), 4.92 (1 H, dd, J P H = 9.4, 3.3 Hz, bridgehead-^), 2.18 (6 H, s, acetone). IR (KBr):  v m a x , 3400-2400 (O-H and C-H), 1705 ( C = 0 ) , 1457, 1438, 1253,  1224, 1206 ( P = 0 ) , 1118, 1087, 786, 750, 724, 697, 642, 559, 543, 515 cm"1. 9,10-Dihydro-10-hydroxy-9,10-ethenoanthracene-l 1,12-bis(diphenylphosphine oxide) (130)/Tetrahydrofuran (2:3) from a mixture of tetrahydrofuran and CH2CI2. Anal. Calcd. for C92H84O9P4: C, 75.81; H, 5.81. Found: C, 76.08; H, 5.81. ! H NMR (CDCI3, 200 MHz): compound 130, 8 9.0 (1 H, s, OH), 7.70-6.70 (28 H, m, Ar-fl), 4.92 (1 H, dd, J P H = 9.4, 3.3 Hz, bridgehead-^; THF, 8 3.78 (6 H, t, J = 4 Hz, OCH2CH2); 1.82 (6 H, t, J = 4 Hz, -OCH 2 C# 2 -). IR (KBr): v m a x , 3400-2200, (O-H and C-H), 1438, 1204, 1120 ( P = 0 ) , 750, 692, 541, 515 cm" 1 . TGA: Calcd. weight loss for 3/2 equivalent of THF: 15.2%. Found: 14.5%, (Ti = 100.8 °C, T 2 = 114.1 °C, T m a x = 109.5 °C). DSC: AH! = 80.51 J/g at T = 96.7-105.7 °C; AH 2 = 89.28 J/g at T = 241.1245.7 °C. 9,10-Dihydro-10-hydroxy-9,10-ethenoanthracene-l 1,12-bis(diphenylphosphine oxide) (130)/Acetic Acid (1:2): from glacial acetic acid. Anal. Calcd. for C 4 4 H3 8 0 7 P2: C, 71.35; H, 5.17. Found: C, 71.52; H, 5.35. ! H NMR (CDCI3, 200 MHz): 8 7.70-6.60 (28 H, m, Ar-H), 4.92 (1 H, dd, J P H = 9.4, 3.3 Hz, bridgehead-tf), 2.01 (6 H, s, C//3COOH). IR (KBr): v m a x , 3700-2200, (O-H and C-H), 1719 ( C = 0 ) , 1693 (C=0), 1438, 1250, 1223, 1157, 1122 ( P = 0 ) , 1101, 749, 728, 692, 557, 541, 514 cm" 1 . 233  9,10-Dihydro-10-hydroxy-9,10-ethenoanthracene-ll,12-bis(diphenyIphosphine oxide) (130)/2-propanol (1:1): needles from isopropanol. Anal. Calcd. for C 4 3H3 8 0 4 P2: C, 75.87; H, 5.63. Found: C, 75.60; H, 5.56. l  H NMR (CDC1 3 , 200 MHz): compound 130, 8 7.70-6.70 (28 H, m, Ar-H), 4.92  (1 H, dd, J P H = 9.4, 3.3 Hz, bridgehead-./?); wo-propanol, 8 4.02 (1 H, septet, J = 7 Hz, C#(CH 3 ) 2 ), 2.20 (s, OH); 1.22 (6 H, d, J = 7 Hz, CH (CH3)2). IR (KBr): v m a x , 3400-2200, (OH and C-H), 1438, 1204, 1120 ( P = 0 ) , 788, 751, 724, 692, 557,541,515 cm" 1 . DSC: AHi = 48.19 J/g at T = 127.5-139.8 °C; AH 2 = 74.74 J/g at T = 237.1241.9 °C. X-ray crystal structure analysis data: empirical formula, C4OH30O3P2C3H7OH; space group, Fl\l\l\;  a = 18.528(3) A, b = 20.676(3) A, c = 9.542(1) A, V =  3621.0(7) A3, Z = 4; R = 0.037. 9,10-Dihydro-10-hydroxy-9,10-ethenoanthracene-ll,12-bis(diphenylphosphine oxide) (130)/2-Methyl-2-propanol (1:1) from tert-butanol/CH2Cl2. Anal. Calcd. for C44H40O4P2: C, 76.07; H, 5.80. Found: C, 75.70; H, 5.69. ! H NMR (CDCI3, 200 MHz): 8 7.70-6.70 (28 H, m, Ar-H), 4.92 (1 H, dd, J P H = 9.4, 3.3 Hz, bridgehead-fl), 1-23 (9 H, s, ( C t f ^ C O H ) . IR (KBr): v m a x , 3600-2400, (O-H and C-H), 1438, 1204, 1120 ( P = 0 ) , 751, 724, 692,557, 541,515 cm" 1 . Crystals from ethanol and (6)-(+)-l-methyl-2-butanol gave no inclusion as shown by ^H NMR spectroscopy. One of the crystals was analyzed by X-ray diffractometry.  X-  ray crystal structure analysis data: formula, C40H30O3P2; space group, P2^/n; a = 11.557(1) A, b = 21.663(2) A, c = 13.120(2) A, p = 91.061(1)°, V = 3284.4(6) A 3 , Z = 4; R = 0.041.  234  (v)  Inorganic Complexes The handling of ruthenium compounds was done under argon through the  manipulation of vacuum line equipment. The solvents were distilled under nitrogen or argon before use. (l,5-Cyclooctadiene)-(ll/?,12/f)-9,10-dihydro-9,10-ethanoanthracene-ll,12-bis(diphenylphosphino)rhodium(I) Tetrafluoroborate, [Rh(COD)(R,jR-anthraphos)]BF4 A mixture 49.9 mg (0.10 mmol) of [Rh(COD)Cl]2 (a gift from Prof. M. D. Fryzuk) and 118.1 mg (0.20 mmol) of the diphosphine, 7?,/?-anthraphos (R,R-136), was stirred in 2 ml of methanol under nitrogen for one hour. The mixture turned to a red orange solution. Sodium tetrafluoroborate (65.8 mg, 0.60 mmol) in 0.6 ml of water was added to afford an orange precipitate, and the solution was kept at 5 °C for one hour. The solid was collected and washed with cold methanol. An orange product (0.129 g) was obtained, and this solid was recrystallized from methanol to yield red-orange bipyramid crystals. The mother liquor was further cooled in the freezer to yield 0.011 g of the bipyramid crystals. Total yield: 80%. This complex decomposed above 250 °C and no melting point could be observed. The complex is air stable and no significant change could be seen in CDCI3 after four days by ^Ip NMR spectroscopy. Anal. Calcd. for C48H44BF4P2Rh: C, 66.08; H, 5.08. Found: C, 66.03; H, 5.07. !H NMR (CDCI3, 400 MHz): 5 8.00-6.70 (28 H, m, Ar-H); 5.00 (2 H, br. s, vinylic-# of COD); 4.36 (2 H, m, vinylic-# of COD); 4.25 (2 H, t, J PH = 3.0 Hz, bridgehead-tf); 2.73 (2 H, d, J PH = 14 Hz, ethano-#), 2.50 (2 H, br. s, CH2 of COD), 2.40 (2 H, m, CH2 of COD), 2.05 (4 H, m, CH2 of COD). 31p NMR (CDCI3, 81 MHz): 5 30.00 (d, J PRh = 146.8 Hz). 19  F NMR (CDCI3, 188 MHz): 6 -78.29.  235  IR (KBr): v m a x , 1640 (C=C), 1437, 1084 (BF4), 1054 (BF4), 749, 695, 577, 516 cm"l. UV-Vis spectrum: methanol, A, (e x 10-3), 449 (2.57), 323 (7.10), 205 (232). [cc]2D3 = -45° (c = 0.5, CDCI3). X-ray crystal structure analysis data: space group, PA\; a = 10.200(7) A, c = 39.97(5) A, V = 4158(3) A 3 , Z = 4; R = 0.064; absolute configuration of anthraphos, (11R, 12R)-(-), and the R value for the enantiomorphous space group P4^ , absolute configuration (115", 125) is 0.0653. Trans-chlorohydridobis[(115,12S)-9,10-dihydro-9,10-ethanoanthracene-ll,12bis(diphenylphosphino)]ruthenium(II), Trans-Ru(H)C\(S, S-anthraphos)2 A mixture of 0.20 mmol (0.201 g) of RuHCl(PPh3)3 (a gift from Mr. K. Marfarlane) and 0.40 mmol (0.228 g) of S, S-anthraphos was refluxed in 20 ml of hexanes under argon for 23 h to afford a yellow precipitate. The solid was filtered and washed with hexanes to afford 0.193 g (76%) of the complex. Anal. Calcd. for RuC 8 oH 65 ClP 4 : C, 74.67; H, 5.09; CI, 2.76. Found: C, 74.29; H, 5.09; CI, 2.68. *H NMR (CDCI3, 400 MHz): 6 7.60-5.70 (56 H, m, Ar-H), 4.13, 3.72 (4 H, 2 s, H9 and H10), 3.48, 2.58 (4 H, m, Hll and H12), -18.28 (1 H, septet, J PH = 13.5 Hz, Ru-H). 31  P NMR (CDCI3, 81 MHz): 5 52.32, 22.04 (2 t, J PP = 33.4 Hz).  IR (KBr): v m a x , 3048 (C-H), 1986 (Ru-H), 1482, 1463, 1434, 1097, 738, 693, 651, 510 cm"1. rr««5-dichlorobis[(HS,125)-9,10-dihydro-9,10-ethanoaiithracene-ll,12-bis(diphenylphosphino)]rhthenium (II), RuCl2(S,S-anthraphos)2 RuCl2(PPh3)3 (0.213 g, 0.222 mmol, prepared by Mr. K Marfarlane) and 0.256 g (0.446 mmol) of S, S-anthraphos were stirred in 5 ml of methylene chloride at room 236  temperature for 24 h. The starting brown solution turned to a deep orange. The solution was concentrated in vacuo to about 1 ml and then 20 ml of ethanol was added to precipitate the product. The solid was filtered and washed with hexanes five times and dried for 2 days in vacuo to yield 0.198 g (67%) of the complex. This complex was proved to include half equivalent of hexanes. Anal. Calcd. for R u C g o H ^ C ^ P ^ l ^ C g H ^ :  C, 73.07; H, 5.25.  Found:  C,  72.88; H, 5.52. ! H NMR (CDC1 3 , 400 MHz): 8 7.60-5.90 (56 H, m, Ar-fl), 4.10 (4 H, br. s, H9 and H10), 3.44 (4 H, br. s, H l l and H12); 0.8-1.6 (7 H, m, peaks from hexanes). 31  P NMR (CDCI3, 81 MHz): 8 17.67.  l,5-Cyclooctadienebis(2-methylaUyl)ruthenium(n),  RU(COD)[TI3-CH3C(CH2)2]2  [Ru(COD)Cl2] n (0.28 g, 1.0 mmol), prepared from RUCI3 and cyclooctadiene in ethanol, was stirred in 15 ml of THF and 10 ml of diethyl ether for several minutes, and then was added 6 ml of 2 M 2-methylallylmagnesium chloride in THF that was prepared from 3-chloro-2-methylpropene (Aldrich) and magnesium in THF.  A homogenous  solution was obtained after adding the Grignard reagent. The solution was stirred for another 10 min and the excess Grignard was precipitated out of the solution by adding more diethyl ether and removed by filtration through Celite. The filtrate was hydrolyzed in ice water and the mixture was extracted with 2 x 20 ml of diethyl ether.  The  combined organic layer was dried over magnesium sulfate, concentrated, filtered through a short plug of neutral alumina column ( 5 x 5  cm) and evaporated to yield a  semicrystalline material (0.267 g, 81%). Recrystallization from a mixture of methanol and petroleum ether yielded 0.22 g of pure product, m.p. = 80-85 °C; lit. 184 m p 80-85 °C. Anal. Calcd. for RuC 1 6 H 2 6: C, 60.16; H, 8.20. Found: C, 59.93; H, 8.31.  237  =  *H NMR (C 6 D 6 , 200 MHz): 5 3.98 (2 H, dd, J = 5, 9 Hz, =CH- of COD), 3.52 (2 H, d, J = 2 Hz, CH3-C(C#2)2)' 2.88 (2 H, s, CH3-C(Ctf2)2), 3.00-2.64 (4 H, m, =CH and CH2 of COD), 1.95 (2 H, m, CH2 of COD), 1.70 (6 H, s, C#j-C(CH2)2), 1.56 (2 H, s, CH3-C(C#2)2), 1-70-1.45 (2 H, m, CH2 of COD), 1.26-1.08 (2 H, m, CH2 of COD), 0.20 (2 H, s, CK3-C(CH2)2. !3c NMR (C 6 D 6 , 50 MHz) 5 111.49 (CH3-C(CH2)2), 88.22 and 70.63 ( = CH- of COD), 51.68 and 51.22 (CH2 of COD), 38.34 and 26.26 (CH3-C(CH2)2)> 24.74 (CH3C(CH2)2).  238  IV. PHOTOCHEMICAL STUDIES All analytical and preparative photolyses were performed by using a Hanovia 450 W medium pressure mercury lamp placed in a water-cooled Pyrex immersion well (thickness = 4 mm, transmits A, > 290 nm). For analytical solution phase photolysis, the sample in a Pyrex tube or an NMR tube was degassed by several freeze-thaw-pump cycles and sealed under nitrogen prior to the photolysis. For preparative photolyses, in general, the compound under investigation was dissolved or suspended in the solvent indicated and purged with nitrogen at least half an hour prior to and during the photolysis. For solid state photolyses, the sample was lightly pressed between two glass slides and the whole assemble was sealed in a polyethylene bag under a nitrogen atmosphere. Low temperature photolyses were carried out by maintaining the sample in a solvent reservoir of chloroform and/or ethanol cooled by a Cryocool immersion cooler, CC-100 II, and the temperature was controlled by a Cryotrol from NESLAB Instruments Inc. The temperature was kept within ±2 °C of the designated temperature and the sample was irradiated with a Hanovia medium pressure mercury lamp. The desired output wavelength was achieved by using a Pyrex glass filter.  (i)  Solid State Photolyses of Dibenzobarrelene Complexes with Phosphine Oxides  Solid State Photolyses of 9,10-Dihydro-9,10-ethenoanthracene-ll,12-dicarboxyhc Acid, 11-Ethyl Ester (105) in its Complexes with Phosphine Oxides The photolyses were carried out in an analytical scale at room temperature or at -40 °C. The photolysis time and temperature are indicated in the text. Following was a typical example: 239  A crystalline complex of (5,5)-(+)-l,2-ethanediylbis((3'-methylphenyl)phenylphosphine oxide)/9,10-dihydro-9,10-ethenoanthracene-ll,12-dicarboxylic acid, 11-ethyl ester (S,S-105m, 55 mg ) was photolysed (Pyrex, 1 > 290 nm) between two glass slides for 48 h at -40 °C, and a cracked yellowish solid was obtained. The photolysed complex 105m was dissolved in a minimum amount of methylene chloride, and the solution was titrated with a yellow diethyl ether solution of diazomethane until the color appeared. Before excess diazomethane was added, the titrated solution was analyzed immediately by GC (DB-17, oven temperature 225 C, head pressure 15 psi) to obtain the conversion (90%). The final mixture stood by with excess diazomethane overnight and then was chromatographed on silica gel (5:1 petroleum ether/acetone).  The derivatized  photoproducts were separated as a product mixture from the phosphine oxide and the starting material, which was converted into the diazomethane adduct after being treated overnight with excess diazomethane. This derivatized photoproducts were identified as a mixture of 138 and 139 by comparison of their *H NMR spectra with those obtained from authentic samples.81 !  H NMR (400 MHz, CDCI3): 8 7.40-6.98 (m, aromatic-tf), The non-aromatic  signals can be assigned to each of the isomers. 8c-EthyI 8b-Methyl  4b,8b,8c,8d-Tetrahydrodibenzo[a,/|cyclopropa[afIpentalene-  8b,8c-dicarboxylate (138): 8 5.06 (1 H, s, C4b-tf), 4.69 (1 s, H, Cgd-H), 4.18 (2 H, q, J = 7 Hz, COOC#2CH3), 3.87 (3 H, s, COOCtfj), 1.26 (3 H, t, J = 7 Hz, COOCH2C#3). 8b-Ethyl  8c-MethyI  4b,8b,8c,8d-Tetrahydrodibenzo[a,/|cyclopropa[c<f|pentalene-  8c,8b-dicarboxylate (139): 8 5.05 (1 H, s, C^-H), 4.67 (1 H, s, C%\>-H), 4.35 (2 H, q, J = 7 Hz, COOCH2CH3), 3.73 (3 H, s, COOCtfj), 1.34 (3 H, t, J = 7 Hz, COOCH2CH5). The regioisomeric composition were determined as 13:87 138/139 by *H NMR spectroscopy above.  For enantiomeric excess measurement, the chiral shift reagent 240  (+)-Eu(hfc)3 (Aldrich) was used and the sample being measured was dried in vacuo at 80 °C overnight. The *H NMR spectra (400 MHz) of the derivatized photoproduct mixture of 138 and 139 described above were recorded in CDCI3 upon addition of a 5 % (+)-Eu(hfc)3 in CDCI3 until each singlet at 3.87 and 3.73 ppm resolved into two peaks with an almost baseline separation. CCI4/CDCI3.  Further separation was achieved in 1:3  The final resolved peaks was integrated to give 31% e.e.  (4bR,SbR,8cR,SdR) for compound 138, and 5% e.e. for compound 139. The absolute configuration of compound 138 was obtained by comparing the spectra with an enantioenriched sample of compound 138 with the known absolute configuration. 81 Solid State Photolyses of 9,10-Dihydro-9,10-ethenoanthracene-ll,12-dicarboxylic Acid, 11-Methylethyl Ester (106) in its Complexes with Phosphine Oxides The same procedure as the case of the complexes with compound 105 was applied. The photolyses were carried out both at room temperature and -40 °C for a certain time as indicated in the text. The regioselectivities were also obtained by GC analysis, in which the two isomers could be separated (DB-17, oven temperature 230 C, head pressure 15 psi). The photoproducts were identified as a mixture of 141 and 142 by their *H NMR spectra, which are identical as those reported in the literature.80 !H NMR (200 MHz and 400 MHz, CDCI3): 5 7.35-7.05 (m, Aromatic-#), The non-aromatic signals can be assigned to each of the isomers. 8b-Methyl  8c-(Methylethyl)  4b,8b,8c,8d-Tetrahydrodibenzo[a,/|cyclopropa[a(]-  pentalene-8b,8c-dicarboxylate (141):  8 5.07 (1 H, s, C^-H), 5.05 (1 H, m,  C0 2 C#(CH 3 ) 2 ), 4.49 (1 H, s, C8b-tf), 3.88 (3 H, s, -C02CH3), 1.22, 1.20 (6 H, 2 d, J = 7 Hz, 8c-MethyI  C02CH(CH3)2). 8b-(MethylethyI)  4b,8b,8c,8d-Tetrahydrodibenzo[a,/]cyclopropa[a/]-  pentalene-8c,8b-dicarboxylate (142): 8 5.24 (1 H, m, C0 2 C#(CH 3 ) 2 ), 5.05 (1 H, s,  241  C 4 b-#), 4.45 (1 H, s, C 8b -#), 3.70 (3 H, s, -C02CH3), 1.32, 1.30 (6 H, 2 d, J = 7 Hz, C02CU(CH3)2).  (ii) Photolysis of 9,10-Dihydro-9,10-ethenoanthracene-ll,12-bis(diphenylphosphine oxide) (126) Photolysis of 9,10-Dihydro-9,10-ethenoanthracene-ll,12-bis(diphenylphosphine oxide) (126) in Acetone Suspension Inclusion complex 126/acetone (1:1, 200 mg), suspended in 100 ml of acetone, was photolysed in an immersion well for 2 h with the Pyrex-filtered output of the 450-W Hanovia lamp until a clear solution was obtained. The photolysis was then continued for another one hour. Nitrogen was bubbled through the mixture for 30 min before and during the photolysis. After photolysis was done, acetone was removed in vacuo and a glassy solid was obtained. Crystallization from ethyl acetate or even better from propyl acetate yielded prisms. Compound 143 was characterized as the only photoproduct. 4b,8b,8c,8d-Tetrahydrodibenzo[a,/|cyclopropa[c^pentalene-8b,8c-bis(diphenylphosphine oxide) (143) m.p. = 244-246 °C. Anal. Calcd. for C40H30O2P2: C, 79.45; H, 5.00. Found: C, 79.50, H, 5.04. !H NMR (200 MHz, CDCI3): 6 8.00-6.52 (28 H, m, Ar-tf); 5.00 (1 H, t, J PH = 11 Hz, C 4b -fl); 4.50 (1 H, d, J PH =11 Hz, C8d-tf). 13  C NMR (75 MHz, CDCI3): 8 aromatic carbide (C), 150.72, 150.69, 150.66,  150.62, 149.53, 149.46, 149.42, 149.35; aromatic carbyne (CH), 133.27, 133.14, 131.99, 131.86, 131.42, 131.39, 131.24, 131.12, 130.96, 130.84, 128.60, 128.44, 128.17, 128.11, 127.95, 127.89, 127.72, 127.36, 127.13, 126.45, 125.40, 121.38, 121.12. Non-aromatic carbons, 64.83, 63.69 (d, JPC = 85.6 Hz, Cgb); 59.43, 59.39,  242  59.31, 59.27 (dd, J PC = 9.2, 3.2 Hz, C 4b ); 51.71, 50.52 (d, J PC = 89.2 Hz, C 8c ); 48.52, 48.49 (d, J PC = 2.2 Hz, C 8d ). 31p NMR (81 MHz, CDCI3): 5 28.73, 28.59, 28.30, 28.16 (AB type spectrum, 8 28.23, 28.66, J PP = 10.2 Hz). IR (KBr): v m a x , 3057 (C-H), 1473, 1437, 1198 (P=0), 1117, 961, 758, 723, 703, 572, 559, 531 cm"1. MS: m/e (intensity), 605 (41.9, M+), 604 (79.7), 527 (48.7), 511 (23.4), 403 (72.1), 386 (46.5), 279 (74.8), 202 (100.0), 201 (85.0), 185 (73.7), 183 (81.7), 178 (48.0), 152 (17.1), 124 (11.7), 77 (35.0), 43 (51.9). Calculated mass for C40H30O2P2: 604.1721. Found: 604.1720. Because of the difficulty in obtaining good quality crystals for X-ray crystallography, compound 143 was reduced to phosphine 144 with trichlorosilane. However, suitable crystals of 143 for X-ray analysis were obtained after the structure of 144 had been solved. 4b,8b,8c,8d-Tetrahydrodibenzo[aJ]cyclopropa[c^pentaIene-8b,8c-bis(diphenyIphosphine) (144) Compound 143 (0.103 mg, 0.17 mmol) and 0.80 ml of trichlorosilane (7.9 mmol, Aldrich) in 5 ml of dry benzene (dried and freshly distilled over P2O5) were heated at 110 °C in a sealed tube for 4 h. The final clear solution was diluted with benzene, washed with 10 ml of 25% sodium hydroxide, dried over anhydrous sodium sulphate and evaporated. The residue was crystallized in a mixture of dichloromethane and ethyl acetate to yield 0.055 g (56%) of compound 144, m.p. = 250-251 C. Anal. Calcd. for C40H30P2: C, 83.89; H, 5.28. Found: C, 83.73; H, 5.29. l  H NMR (200 MHz, CDCI3): 5 3.85 (1 H, dd, J PH = 11.7 Hz, 10.5 Hz, Csd-H);  4.46 (1 H, d, J PH = 2.6 Hz, C 4b -//); 6.60-7.80 (28 H, m, Ar-H).  243  31p NMR (81 MHz, CDCI3): 8 3.32, 1.94, 1.73, 0.36 (AB type spectrum, 8 2.23, 1.44, J PP = 111.5 Hz). IR (KBr): v m a x , 3052, 1471, 1434, 776, 743, 698 514, 486 cm"1. MS: m/e (intensity), 572 (24.0, M+), 387 (100), 309 (43.0), 278 (24.9), 262 (6.6), 233 (5.2), 209 (44.9), 191 (23.7), 169 (9.9), 149 (9.8), 122 (7.8), 108 (19.4), 91 (12.9), 77 (9.9), 69 (10.8), 57 (11.8), 43 (21.4), 32 (18.8).  Calculated mass for  C40H30P2: 572.1823. Found: 572.1824. Solid State Photolysis of Inclusion Complexes of 9,10-Dihydro-9,10-ethenoanthracene-ll,12-bis(diphenylphosphine oxide) (126) The photolyses of inclusion complexes of 126 with different guests (single crystal or polycrystals) were carried out at room temperature and/or low temperature. Compound 143 was the only detectable photoproduct. In the case of the single crystal photolysis, the enantiomeric excess, as well as the conversion, was analyzed by chiral HPLC (Chiralcel OP, 93:7 methanol /water, flow rate, 0.35 ml/min, room temperature, detector, UV at 230 nm and 254 nm).  A single crystal could be crushed before  photolysis for a faster photoreaction, but a slow photolysis of a whole single crystal was preferred because it prevented the sample from heating. Asymmetric Photochemical Synthesis of 4b,8b,8c,8d-Tetrahydrodibenzo[a,/|cyclopropa[cd]pentalene-8b,8c-bis(diphenylphosphine oxide) (143) Preparation of Enantiomorphously Pure Crystals.  A 5-10% solution of  compound 126 in boiling ethanol was allowed to cool to a moderately over-saturated solution, and then a freshly powdered single crystal was dispensed in this solution very carefully. The crystallization began soon after seeding. The crystals were allowed to grow for about 4-8 h at room temperature and then filtered. Photochemical Asymmetric Induction.  Enantiomorphously pure small plates  or/and prisms (0.345 g) were photolysed between two Pyrex glass slides at -40 °C for 244  15 h. The conversion and optical purity of the photoproduct were monitored by chiral HPLC as 26% and >97% [(-)-isomer] respectively. The yellowish photolysed crystals were dissolved in hot acetone (100 ml) and concentrated to precipitate the starting material, and 0.242 g of the starting material was filtered out and recycled. The filtrate was evaporated in vacuo and the oil so obtained was crystallized from propyl acetate to yield 0.040 g of needle crystals, m.p. = 268-270 °C (d.); [a]2D2 = -82° (c = 2.0, CHCI3). The optical purity of the final crystals was >99% [(-)-isomer] by chiral HPLC and by 31p NMR spectroscopy by using the chiral shift reagent D-dibenzoyltartaric acid (one equivalent). Starting with the other enantiomorphic crystal, 1 g was photolysed at -40 °C for 3 days. The other enantiomer of the product 143 (0.180 g) was crystallized from propyl acetate after 0.733 g of the starting material was recycled from acetone. The optical purity of the isolated photoproduct was >99% [(+)-isomer] by chiral HPLC and 31p NMR using the chiral shift reagent mentioned above. The crystals so obtained were reduced with 1 ml of trichlorosilane in 5 ml of benzene at 110 °C in a sealed tube, following the same procedure as that of racemic 144. Crystalline compound 144 (0.099 g) was obtained from ethyl acetate, m.p. = 245-248 °C. The crystal was analyzed by X-ray crystallography and the absolute configuration was identified as the (+)(4bfl,8W?,8cfl,8d/?)-enantiomer, [a]2D2 = 174° (c = 0.40, toluene). X-ray crystal structure analysis data: space group, P2i2i2\; a = 17.381(4) A, b = 18.414(3) A, c = 9.591 A, V = 3069(1) A 3 , Z = 4; R = 0.031 for (4W?,8W?,8c/?,8di?)-configuration, and R = 0.038 for the opposite configuration; so the absolute configuration was established as (+)-(4b/?,8W?,8c#,8dft)-enantiomer with AR = 0.7%. The reduction could be done with the crude oily photoproduct, and the overall yield was much higher. For example, 1.0 g of crystals was photolysed to a 40% conversion at -40 °C. After 0.59 g of starting material was recycled from acetone, the oil left after 245  evaporating the filtrate was reduced by 2 ml of trichlorosilane in 5 ml of dry benzene, whereupon 0.20 g of optically pure phosphine 144 was obtained with an overall yield of 50%.  (iii) Photolysis of Compound 9,10-Dihydro-10-hydroxy-9,10-ethenoanthracenell,12-bis(diphenylphosphine oxide) (130)  Photolysis of Compound 9,10-Dihydro-10-hydroxy-9,10-ethenoanthracene-ll,12-bis(diphenylphosphine oxide) (130) in Acetone Inclusion complex 130/acetone (1:1, 0.70 g) ) in 400 ml of acetone was photolysed in an immersion well with the Pyrex-filtered output of the 450 W Hanovia lamp under nitrogen for 2.5 h. The solvent was removed in vacuo to yield a yellowish glassy solid, m.p. = 140-146 °C.  Attempts to obtain its elemental analysis failed because it  contained small amount of other minor photoproducts and solvents and could not be purified by routine purification procedures. However, 31p NMR indicated compound 149 as the only major product (>95% by integration of 31p NMR spectrum). 4b,9,9a, 10-Tetrahydro-9-oxo-fl«ft'-10-indeno[/, 2-a]indene-9a, 10-bis(diphenylphosphine oxide) (149) J  H NMR (200 MHz, CDCI3) 8 8.30-8.20 (2 H, m, Ar-H), 7.90-6.64 (25 H, m,  Ar-H), 5.72 (1 H, d, J = 8 Hz, Ar-H, the coupling of this proton with aromatic proton(s) was confirmed by COSY), 5.30 (1 H, dd, J P H = 14.2, 3.8 Hz, C^-H),  5.08  (1 H, dd, J P H = 19.7, 11.2 Hz, C\Q-H (syn)). 31  P NMR (81 MHz, CDCI3) 5 29.07, 34.31 (2 d, J P P = 23.8 Hz).  !3C NMR (75 MHz, CDCI3): 8 50.13, 50.94 (d, J PC = 61.3 Hz, CIQ); 55.13 (C45); 65.29, 66.17 (d, J PC = 61.9 Hz, Co a ); aromatic carbons, 124.05, 124.13, 124.25, 124.27, 126.13, 126.17, 126.95, 126.98, 127.74, 127.89, 128.04, 128.20, 128.29, 128.40, 128.45, 130.64, 130.76, 131.28, 131.31, 131.55, 131.58, 131.88, 246  131.99, 132.08, 132.16, 132.20, 133.28, 133.38, 134.79, 137.63, 137.69, 137.76, 142.20, 142.28, 151.78, 151.89; 199.12, 199.21 (d, J P C = 6.7 Hz, C 9 = 0 ) . IR (KBr): v m a x , 1709 ( C = 0 ) , 1438, 1191 ( P = 0 ) , 1115, 754, 697, 528 cm"1. MS: m/e (intensity), 621 (4.2, M + l), 622 (9.5, M+) 543 (0.8), 527 (1.0), 419 (34.0), 262 (7.2), 217 (12.8), 201 (100, P(0)Ph 2 +), 183 (13.9), 165 (14.0), 152 (5.5), 124 (7.8), 95 (4.2), 77 (20.7), 51 (8.8). Calculated mass for C40H30O3P2: 620.1670. Found: 620.1671. This compound was further derivatized into compound 150 by following the procedure below: Compound 149 (100 mg) and 200 mg of NaOCH3 in 10 ml of methanol was refluxed for 1 h to yield a brown solution.  Afterwards, the solution was cooled to room  temperature, quenched with water and extracted with CHCI3. The chloroform phase was washed with saturated ammonium chloride solution, dried over MgSC«4 and rotary evaporated. The residue was chromatographed on silica gel. Product 150 was eluted with ethyl acetate and crystallized from a mixture of ethyl acetate and petroleum ether to yield 50 mg of compound 150. 4b,9,9a,10-Tetrahydro-9-oxo-anft'-10-indeno[l,2-a]indene-10-(diphenylphosphine oxide) (150) m.p. = 249-250 C. Anal. Calcd. for C 2 8H2l0 2 P: C, 79.99; H, 5.03. Found: C, 79.62; H, 5.08. ! H NMR (200 MHz, CDCI3) 5 8.10 (17 H, m, Ar-H), 6.39 (1 H, d, Ar-fl), 4.79 (1 H, d, J H H = 7.1 Hz, C 4b -fl), 4.66 (1 H, dd, J P H = 7.4 Hz, J H H = 1.9 Hz, 2nd doublet was partially resolved and the coupling with Cq^-H was confirmed by COSY, CiQ-H(syn)), 3.74 (1 H, ddd, J P H = 17.7, J H H = 7.1, 1.9 Hz, C 9 a -#). 31  P NMR (81 MHz, CDCI3): 5 32.99.  247  !3C NMR (50 MHz, CDCI3): 5 45.66, 47.03 (d, JPC = 69.3 Hz, C 10 ); 50.16, 53.49 (C 9 a , and C 4b ); 124.33, 124.80, 124.85, 126.21, 126.28, 127.18, 127.24 128.14, 128.45, 128.61, 128.68, 128.83, 131.58, 131.75, 131.79, 131.98, 135.63 (aromatic carbons); 144.63, 144.72, 156.72, 156.76 (C4a, C4C, Cga and Cio a ); 206.69, 206.91 (d, J PC = 11.4 Hz, C 9 =0). IR (KBr): v m a x , 3062 (C-H), 1708 (C=0), 1602, 1468, 1438, 1290, 1242, 1201 (P=0), 1118, 1015, 882, 764, 754, 720, 699, 603, 590, 549, 519, 441 cm"1. MS: m/e (intensity), 421 (12.4, M + l), 420 (38.1, M+), 391 (1.0), 265 (1.4), 219 (80.1), 202 (100.0), 189 (53.8), 165 (23.6), 115 (4.7), 95 (4.4), 77 (28.6), 51 (10.4). Calculated mass for C28H21O2P: 420.1279. Found: 420.1278. Solid State Photolysis and Asymmetric Induction of Compound 130 Enantiomorphously pure crystals were prepared and photolysed in the same way as in the case of compound 126. Solid state photolyses were done at both room temperature and -30 °C. Generally, 20-100 mg of the crystals were photolysed for several hours to days as noted in the text. The photolysed solid was dissolved in CDCI3 and its 31p NMR spectrum was recorded at 81 MHz on a 200 MHz machine. The conversion was measured by the integration of the 31p NMR spectrum. After adding one equivalent of D-DBT (or L-DBT), the two  31  P NMR doublets of the product were split if both  enantiomers were presented and the optical purity was measured by the integration of the spectrum.  (iv) Photolysis of 940-Dmydro-9,10-ethenoanthracene-ll-(diphenylphosphine oxide)-12-carboxylic Acid (133) Photolysis (Pyrex >290 nm) of compound 133 was carried out in acetone, CDCI3 and DMSO-dg on an analytical scale and an approximately 0.05 M solution was used. The preparative photolysis was carried in the solid state. Two products were detected 248  (compounds 151 and 152) and their relative ratios as well as the conversions were measured by *H and 31p NMR spectroscopy. Typically, 0.32 g of compound 133 was packed between two glass slides and sealed in a plastic bag under nitrogen. This sample was irradiated (Pyrex, >290 nm) for 28 h. *H and 31p NMR spectra showed that an almost 100% conversion was obtained and the relative ratio of the two photoproduct was 90:10 152/151.  Recrystallization of the photolysis mixture from a mixture of  chloroform and acetonitrile gave 0.15 g of compound 152 as prisms. ! H NMR (200 MHz, CDCI3) of the photolysis mixture: 8 6.50-8.20 (m, Ar-fl) and the rest of the spectrum is assigned to each regioisomer: 4b, 8b, 8c, 8d-Tetrahy drodibenzo [a,f\ cyclopropa[a/] pentalene-8b-(diphenylphosphine oxide)-8c-carboxylic Acid (151). 5 4.96 (1 H, d, JPH = 12 Hz, C8d-fl), 5.35 (1 H, s, C4b-H). 4b,8b,8c,8d-Tetrahydrodibenzo[fl,/|cyclopropa[cd]pentalene-8c-(diphenylphosphine oxide)-8b-carboxylic Acid (152). 5 4.35 (1 H, d, J PH = 11 Hz, C 8d -fl), 4.55 (1 H, d, J PH = 13 Hz, C^-H). 31p NMR (81 MHz, CDCI3): 5 36.28 (152), 37.56 (151). 4b,8b,8c,8d-Tetrahydrodibenzo[«,/|cyclopropa[c(/]pentalene-8c-(diphenylphospbiiie oxide)-8b-carboxylie Acid (152) m.p. = 275-278 °C. Anal. Calcd. for C29H21O3P: C, 77.67; H, 4.72. Found: C, 77.59; H, 4.68. !H NMR (200 MHz, DMSO-d6): 5 4.28 (1 H, d, J PH = 11 Hz, C8d-fl), 4.60 (1 H, d, J PH = 9.0 Hz, C 4b -tf), 6.92-7.60 (18 H, m, Ar-tf). *H NMR (200 MHz, CDCI3): 5 4.35 (1 H, d, J PH = 11 Hz, C 8d -fl), 4.55 (1 H, d, J PH = 13 Hz, C4b-H), 6.50-8.20 (18 H, m, Ar-H). 13  C NMR (50 MHz, DMSO-d6): 5 45.53 (d, partially resolved, JPC = 1.7 Hz,  C 8d ); 55.32, 55.37 (d, JPC = 3.1 Hz, C 8b ); 57.03, 57.19 (d, J PC = 8.5 Hz, C 4b );  249  61.81, 63.67 (d, J PC  = 93.2 HZ, C 8c ); 121.56, 121.68, 125.63, 126.14, 127.03,  127.45, 127.60 (C^ C 7 , C 2 , C 3 , Ci, C 4 , C 5 and C 8 );  149.67, 149.80, 150.44,  150.53 (C 4 a C4C Cga and C 8e ); Ph2P(0) group: 128.58, 128.63, 128.82, 128.86 (2 d, JPC = 11.8, 11.8 Hz, m-C), 130.90, 131.10, 131.30 (pseudo t, JPC = 10.0 Hz, o-C), 132.12, 132.17, 132.274, 132.30 (2 d, J PC = 2.0 Hz, 2.3 Hz, p-C), 168.69, 168.78 (JPC = 4.8 Hz, C=0). 31p NMR (81 MHz, DMSO-d6): 8 27.86. 31  P NMR (81 MHz, CDCI3): 5 36.86.  MS: m/e (intensity), 448 (3.7, M+), 430 (22.1), 404 (5.3), 247 (5.8), 230 (31.8), 202 (100.0), 189 (7.8), 152 (4.5), 125 (4.0), 95 (7.0), 77 (54.7), 51 (15.7), 44 (14.8). Calculated mass for C29H21O3P: 448.1228. Found: 448.1236. IR (KBr):  v m a x , 3200-2200 (O-H and C-H), 1725 (C=0), 1437, 1251, 1180  (P=0), 742,535 cm"1. X-ray crystal structure data: monoclinic, space group P2\/n; a = 15.206(4) A, b = 10.626(3) A, c = 13.661 (2) A, (3 = 90.06(2)°, V = 2207.4 (8)A3, Z = 4; R = 0.038.  (v) Photolysis of Ethyl 9,10-Dihydro-9,10-ethenoanthracene-ll-(diphenylphosphine oxide) -12-carboxylate (132) Photolyses were carried out in different media including the solid state through Pyrex (>290 nm) on an analytical scale. Two products, 153 and 154, were detected, and the relative ratio of the two products was determined by ^H and 31p NMR spectroscopy. Attempts to separate these two products by silica gel chromatography failed.  So  compound 154 was prepared independently from the corresponding carboxylic acid 152.  250  ! H NMR (200 MHz, CDCI3) of the photolysis mixture (an nearly 100% conversion): 8 6.32-7.80 (m, Ar-H), and the rest of the spectrum is assigned to each compound: 8b-Ethyl  4b,8b,8c,8d-Tetrahydrodibenzo[a,/|cyclopropa[c</lpeiitalene-8b-carboxy-  late-8c-(diphenylphosphine oxide) (154). 8 1.17 (3 H, t, J HH = 7, Cfl r jCH 2 0), 4.10, 4.09 (2 H, 2 q (partially resolved), J HH = 7, CH 3 CH 2 0), 4.43 (1 H, d, J P H = 9 Hz, C 8d -tf), 4.55 (1 H, d, J PH = 12 Hz, C 4b -fl). 8b-Ethyl  4b,8b,8c,8d-TetrahydrocUbenzo[aJ]cyclopropa[af]pentalene-8c-carboxy-  late-8b-(diphenylphosphine oxide) (153). 8 1.08 ( 3 H, t, J H H = 7 Hz,  CU2CH3),  3.92, 3.93 (2 H, 2q, J H H = 7 Hz, CH2CU3), 4.80 (1 H, d, J P H = 10 Hz, C 8d -tf), 4.88 (1 H, s, C 4 b -#). 31p NMR (81 MHz, CDCI3): 8 30.13 (153), 29.27 (154). 8b-Ethyl 4b,8b,8c,8d-Tetrahydrodibenzo [aJ] cyclopropa [cd\ pentalene-8b-carboxylate-8c-(dipheny!phosphine oxide) (154) Potassium hydroxide (0.30 g) was dissolved in 0.30 ml of water and cooled with 5 ml of diethyl ether in a refrigerator. To this cold mixture, 0.20 g of l-ethyl-3-nitro-lnitrosoguanidine (Aldrich) was added with stirring.  A yellow ether solution of  diazoethane was afforded, and this solution was decanted into 0.147 g of carboxylic acid 152 in 3 ml of methylene chloride.  The final solution was concentrated and  chromatographed on silica gel. The product was eluted by ethyl acetate and crystallized from a mixture of petroleum ether and ethyl acetate to yield 0.087 g (56%) of leaf-like crystals, m.p. = 169-170 C. Anal. Calcd. for C31H25O3P: C, 78.14; H, 5.29. Found: C, 78.26; H, 5.25. *H NMR (200 MHz, CDCI3): 8 1.17 (3 H, t, J H H = 7, CH3CH20),  4.10, 4.09 (2  H, 2 q (partially overlapped), J H H = 7, CH 3 CH 2 0), 4.43, 4.55 (2 H, 2 d, J P H = 9 Hz, 12 Hz, C 8 d - # , C 4b -tf), 6.90-7.70 (18 H, m, Ar-fl).  251  31  P NMR (81 MHz, CDCI3): 5 29.27.  !3C NMR (50 MHz, CDCI3): 5 14.13 (CH3); 46.55, 46.61 (d, JPC = 2.8 Hz, C8d); 55.76, 55.82 (d, JPC = 3.3 Hz, C-4b); 58.44, 58.60 (d, JPC = 8.1 Hz, C-8b); 62.01 (s, CH2O); 62.57, 64.41 (d, J PC = 92.8 Hz, C-8c); 121.23, 121.41, 126.16, 126.71, 127.30, 127.36, 127.70, 127.87 (C-6, C-7, C-3, C-2, C-4, C-l, C-8 and C-5); 134.45 (s), 135.44 (partially resolved d, JPC = 1.2), 149.41 and 149.52 (d, JPC = 5.9 Hz), 150.23 and 150.32 (d, J PC = 4.6 Hz), (C-4a, C-5a, C-la and C-5b); 168.14 and 168.24 (d, J PC = 4.9 Hz, carbonyl). Ph2PO group: 5 128.58, 128.83 (d, JPC = 12.2 Hz, m-C); 130.66, 131.08, 132.76, 133.17 (2 d, JPC = 105.8 Hz, 105.4 Hz, P-C); 131.48, 131.68, 131.77, 131.97 (2 d, JPC = 9.9, 10.4 Hz, o-C); 132.18, 132.29, 132.31 (pseudo-t, J PC = 3.3 Hz,/?-C). IR (KBr): v m a x , 1725 (C=0), 1704 (C=0), 1475, 1438, 1300, 1239 (P=0), 1194, 1121, 758, 725, 697, 567, 534, 513 cm"1. MS:  m/e (intensity), 476 (16.9, M+), 430 (53.9), 275 (25.6), 246 (8.2), 230  (34.5), 219 (13.0), 201 (100), 152 (3.1), 95 (4.0), 77 (29.7), 61 (9.9), 51 (6.3), 43 (30.7). Calculated mass for C 31 H 2 503P: 476.1541. Found: 476.1541.  (vi) Quantum Yield 185 of Photoreaction of Inclusion Complex 126/ethanol (1:1) A merry-go-around apparatus was used in a large water bath to maintain a constant photolysis temperature (20 ± 3 °C). The 450 W Hanovia medium pressure mercury lamp was used as the light source. The 313 nm line from the lamp was isolated by a filter combination of a 7-54 Corning glass plate and an aqueous solution of 0.002 M K 2 Cr04 containing 5% K2CO3 circulated through a Pyrex cooling jacket. Valerophenone actinometry was used. The quantum yield of acetophenone from valerophenone has been established to be O = 0.3 with an opaque concentration (0.1 M) of valerophenone in benzene. 186 Three 3 ml benzene solutions of 0.1 M valerophenone  252  and 1 mg/ml of tetradecane were degassed by repeating the freeze-pump-thaw cycle three times under nitrogen. These solutions in 10 ml Pyrex phototubes were placed alongside the test samples in the merry-go-round apparatus and monitored by GC (carbowax column, 15 m, average of two injections). Two 3 ml test samples (concentration of 0.10 M) were prepared from inclusion complex 126/ethanol (1:1) in CDCI3. The 0.10 M solution was determined to be opaque at 313 nm. The sample was photolysed alongside the actinometer solutions and monitored by ^H NMR. The amount of photoproduct 143 was measured by integration of the 1H NMR spectrum using ethanol as the internal standard. The quantum yields were calculated from the following equation: _  moles of photoproduct moles of photons absorbed  The quantum yield of 143 from photolysis of 126/ethanol (1:1) in CDCI3 was determined as 0.20 and 0.29 at 9.5% and 24.4% conversion respectively.  253  V. CATALYTIC HYDROGENATION All solvents were dried and freshly distilled under nitrogen before use.  All  procedures dealing with possible air-sensitive compounds were conducted under argon. Catalytic hydrogenation of alkenes employing H2 gas at 4 atm H2 pressure (initial) was performed in a Schlenk tube. Hydrogen was filled to 1 atm pressure at 77 K and the system was sealed and warmed to room temperature. The internal ^-pressure at 298 K was calculated to be 4 atm according to equation PV = nRT. Experiments done at higher H2 pressure were conducted in glass lined high pressure steel vessels.  (i)  Catalytic Hydrogenation of rrans-a-acetamidocinnamic Acid In situ catalyst. Bisphosphine 144 (0.02 mmol) and [Ru(COD)Cl2]2 (0.01 mmol)  was accurately weighed into a 40 ml Schlenk tube, and 2 ml of methylene chloride was added with stirring to give a homogenous solution.  The substrate, trans-a-  acetamidocinnamic acid (2 mmol, accurately weighed, Aldrich) in 10 ml of methanol was then added into the above homogenous solution. The whole tube was cooled in liquid nitrogen, filled with 1 atm of H2 after three vacuum/H2 cycles and then sealed. In the case of high pressure, a glass liner was used instead of a Schlenk tube. This tube was put into the high pressure steel vessel, and the system was purged and pressurized with H2 to the required pressure. The hydrogenation was conducted with stirring at room temperature for the required time. Rh(COD)(S,S-anthraphos)BF4. A 50 ml Schlenk tube was charged with accurately weighed substrate (0.2-0.5 g, 1-3 mmol) and Rh(COD)(£lS,-anthraphos)BF4 (2-5 mg, 0.002-0.005 mmol) as well as deoxygenated methanol (5-10 ml). The tube was cooled in liquid nitrogen, filled with 1 atm of H2 after three vacuum/H2 cycles and sealed. The  254  hydrogenation was conducted at room temperature and 4 atm pressure until the red orange turned to a straw yellow. RuHCl(S,S-anthraphos). The attempted hydrogenation at 1000 psi was performed with substrate (100 mg) and 10% mol catalyst in 10 ml of 1:1 CH2Cl2/methanol. The attempted hydrogenation at 4 atm pressure was performed with substrate (100 mg) and 2% mol of catalyst in THF. The procedure for the other catalysts mentioned above was followed for high or low pressure hydrogenation.  Substrates other than a-  acetamidocinnamic acid were Ar-phenyl-2-phenylethylimine and acetophenone.  (ii) Workup and Characterization of Hydrogenation Product After the hydrogenation of a-acetamidocinnamic acid (155), the final reaction mixture was evaporated and the residue was redissolved in 0.5 M NaOH. The solution was washed with ethyl acetate, acidified by concentrated HC1 and extracted with ethyl acetate five times. The extract was dried over MgS04 and evaporated to give the only product, TV-acetylphenyl-alanine (156). iV-Acetylphenylalanine (156) *H NMR (CD3OD, 200 MHz): 5 7.40-7.20 (5 H, m, Ar-tf), 4.58 (1 H, dd, J = 9, 6 Hz, a-H), 3.18 (1H, dd, J = 14, 6 Hz, benzylic-tf), 2.82 (1H, J = 14, 9 Hz, benzylic-H), 1.86 (6 H, s, C(0)CH3). MS: m/e (relative intensity), 207 (2.5, M+), 189 (6.2), 161 (5.5), 148 (100.0), 120 (23.5), 104 (10.6), 91 (85.0), 74 (19.7), 65 (8.7), 43 (16.6). The conversion in the hydrogenation reaction was determined by *H NMR spectroscopy.  The hydrogenation product (or along with the starting material) was  titrated to the methyl ester (or esters) by diazomethane, and the conversion of the hydrogenation was also monitored by GC (DB-5, oven temperature 160 °C, head pressure 85 kPa). Ester 157 was resolved by chiral HPLC (Chiracel OD, 25 cm x 0.46  255  cm I.D.; eluent, 9:1 hexanes/wo-propanol, room temperature; (/?)-(-), t = 11.4 min, (S)-(+), 13.4 min) and the optical purity of the product was so measured. An authentic sample was used to determine the absolute configuration.  For example, cc-  acetamidocinnamic acid (0.2236 g) was catalytically hydrogenated by Rh(COD)(£,S'anthraphos)BF4 (4.7 mg) for 2.5 h. A sample (about 10 mg) of the residue from the workup was put in ethyl acetate and titrated by diazomethane in ether until the yellow color appeared. The final mixture was analyzed by GC to give a full conversion, and the optical purity of product 157 was determined by HPLC to give 89.4% e.e. of S-(+)-Nacetylphenylalanine.  256  APPENDIX I ASSIGNMENT AND ANALYSIS OF l^C NMR SPECTRA OF PHOSPHINE OXIDES  Al. Assignment of l^C NMR Spectra of Monophosphine Oxides The phosphine oxides studied have similar structures, and their l^C NMR spectra can be related to each other. On the basis of the chemical shifts and coupling constants from the known spectrum of diphenylmethylphosphine oxide 187 and the chemical shifts of the monosubstituted benzenes (Table Al),188 all the spectra could be unambiguously assigned.  Table Al. Incremental Shifts of the Aromatic Carbon Atoms of Monosubstituted Benzenes3 substituent  C-l (JPC)  C-2 (JPC)  C-3 (JPC)  C-4 (JPC)  -P(Ph)(Me)(0)b -N(CH 3 ) 2 -CH 3  +5.8(100.9) +22.4 +8J)  +1.9(9.7) -15.7 +0/7  0.0(11.8) +0.8 -01  +3.1(2.8) -15.7 -219  Chemical shifts in ppm are based on benzene at 5 = 128.5, + downfield, - upfield. bJpc coupling constants in Hz are included in parentheses.  The coupling constant between phosphorus and carbon in each position on a phenyl ring remains almost constant upon substitution and can be used to confirm the carbon being assigned. The calculated chemical shifts from Table Al for the phosphine oxides are given in Table A2. By using APT ^C NMR spectra, complete assignments of the phosphine oxides are listed in Table A3. The calculated spectra are consistent with the observed spectra; the calculated and observed chemical shifts are very close, and so are  257  the coupling constants. These spectra serve to confirm the structures of the phosphine oxides. As an example, the calculation of Cs-2 chemical shift of compound 111 is delineated as follows: 8 (Cs-2) = 128.5 + 8(P=0) + I 5(CH3) where: 8(P=0) is the incremental shift from phenylmethylphosphinoyl group as shown in Table Al, which is +1.9 ppm at C-2 with a coupling constant, JP£ = 9.7 Hz. 8(CH3) is the incremental shift from methyl group, which are 0.7 ppm and -2.9 ppm at the o- and /7-positions of C-2 respectively.  Table A2. Calculated Chemical Shift for Phosphine Oxidesa 108  109  110  111  112  113  oxide  (4-Me)  (3-Me)  (3,4-di-Me)  (3,5-di-Me)  (2,4,5-tri-Me)  4-N(Me)  Cb-/ Cb-2, C b -6  134.3 130.4  134.3 130.4  134.3 130.4  134.3 130.4  134.3 130.4  134.3 130.4  Cb-3, Cb-5  128.5 131.6 131.4  128.5  128.5  128.5  128.5  131.6 134.2  131.6  128.5 131.6  131.6  131.6  131.3  134.1  132.0  119.9  130.3 130.3  131.1 129.7  131.0 127.4  128.2 128.2  136.5 130.9  133.1 133.1  129.2 129.2  137.4 128.4  138.1 129.1  137.3 137.3  129.8 127.2  113.1 113.1  140.5  132.3  141.2  133.0  141.1  154.6  Cb-4 Cs-7 C s -2 Cs-6 Cs-3 Cs-5 Cs-4 a  (Cb)s are carbons of the phenyl rings and (Cs)s are carbons of the substituted phenyl rings of the  phosphine oxides. An example with the numbering is shown in Figure Al (compound 111).  258  Table A3. Assignment of 13c NMR Spectrum of Monophosphine Oxidesa oxide  108  109  110  111  112  113  (4-Me)  (3-Me)  (3,4-di-Me)  (3,5-di-Me)  (2,4,5-tri-Me)  4-N(Me)2  CW  134.4 (101.2)  Cb-2, Cb-6  130.4 (9.7)  Cb-5, Cb-5  128.5 (11.8) 131.6 (2.7) 130.7 (103.7)  134.5 (101.0) 130.4 (9.9) 128.5 (11.8) 131.5 (2.7)  134.4 (101.1) 130.5 (9.7) 128.5 (11.8) 131.6 (2.7) 133.7 (100.9) 128.1 (9.7T 128.1 (9.7) 138.3 (12.5) 138.3 (12.5) 133.4 (3.0)  135.2 (99.9) 130.3 (11.1) 128.5 (11.9)  Cb-4  134.3 (101.0) 130.5 (9.7) 128.6 (11.9) 131.6 (2.8) 133.7 (101.0) 131.1 (9.4) 127.5 (10.1) 138.5 (11.7) 128.5 (12.5) 132.5 (2.8)  135.40 (101.0) 130.4 (9.8) 128.4 (11.8) 131.30 (2.8) 118.5 (112.2)  C s -i  130.5 (10.1) 130.5 (10.1) 129.3 (12.2)  Cs-2 C s -6 Cs-3 C s -5  129.3 (12.2) 142.1 (2-7)  C s -4  130.9 (103.4) 131.5 (9.8) 127.9 (10.0) 137.2 (12.1) 129.8 (12.9) 140.9 (3.0)  131.3 (2.7) 128.3 (102.0) 138.9 (8.8) 133.5 (11.2) 132.7 (11.3) 134.0 (12.3) 141.1 (2.7)  131.9 (11.1) 131.9 (11.1) 111.4 (12.7) 111.4 (12.7) 152.3 (2.4)  Coupling constants in Hz are given in parentheses. (Cb)s are carbons of the phenyl rings and (Cs)s are carbons of the substituted phenyl rings of the phosphine oxides. An example with the numbering is shown in Figure Al (compound 111).  O.  P(PhXMe)  3.1  Cs-2 1.9 - 29 0.7  Me 8.9  1.9  0.7  0.0  -0.1  P(Ph)(Me)(0) p-Me o-Me  •0.3  -2.9  Figure Al. Incremental Shifts (ppm) of the Aromatic Carbons.  259  The data are outlined in Figure Al. The total incremental shift is -0.3 ppm, so the chemical shift of Cs-2 is 128.2 ppm (observed, 8 128.1 ppm) with a coupling constant, JPC = 9.7 Hz. (observed, JPC = 9.7 Hz).  A2 Assignment and Analysis of ^C NMR Spectra of Bisphosphine Dioxides Carbon-13 NMR spectra of bisphosphine dioxides 114-117 are assigned as shown in Table A4, according to the chemical shifts and coupling constants of the corresponding monophosphine oxides. Table A4. Observed ^C NMR Spectra and Their Assignment of Bisphosphine Dioxidesa oxide  114 (4-Me)  C b -7 Cb-2, C b -6 Cb-5, Cb-5 Cb-4  -  130.7 (9.8) 128.6 (10.8) 131.8 -  Cs-1  -  130.7 (9.7) 128.8(11.6) 131.9 -  116 (3,4-di-Me)  117 (3,5-di-Me)  131.5 (99.8)  132.3 (100.2)  130.7 (9.5) 128.7 (10.8)  130.8 (9.5) 128.9(11.8)  131.8  131.9  127.7 (102.0)  131.9(99.8)  130.6 (9.6)  131.3 (9.2)  131.8 (9.6)  128.3 (9.6)  127.7 (9.8) 138.8(11.7)  128.3 (9.7) 137.4(11.8)  128.3 (9.6)  Cs-3  130.6 (9.6) 129.4(12.1)  C s -5  129.4(12.1)  128.7(11.4)  130.0 (12.5)  138.6(11.5)  Cs-4  142.6  132.8  141.3 (2.7)  133.8  Cs-2 C s -6  a  115 (3-Me)  138.6(11.5)  (Cb)s are carbons of phenyl rings and (Cs)s are carbons of substituted phenyl rings of  phosphine oxides. As an example the structure and the numbering of compound 116 are shown in Figure A2.  260  Figure A2. Structure and Numbering of Compound 116.  All CH-peaks in the aromatic region are partially-resolved triplets except for the carbons that are bonded directly to phosphorus, which do not show up clearly in some cases due to their relatively low intensities.  The methylene carbons (CH2P) are  multiplets consisting of 5 symmetric peaks whenever they can be recognized and a typical spectrum is shown in Figure A3. The spectra observed can be explained as a special ABX type. The calculation gives for the ABX system the transition energies and intensities of Table A5,189 which originate from X, where A = B = 31p, X =  Table A5. Transition Energies and Intensities for the ABX System which Originate from X transition  energy  relative intensity  9  1  10  vx-N v x + D + - D.  11  v x - D + + D.  12  vx + N  13 14 15  V  A +  V  B '  V  1/2[1 + cos((t)+ - ())_)] 1/2[1 + cos((|)+ - <f>_)] 1 0  X  v x - D+ - D.  1/2[1 - cos(4>+ - ()).)]  v x + D+ + D.  1/2[1 - cos(<t>+ - <))_)]  D±cos<|>± = 1/2(5^ ± L), D±sin<t>± = 1/2 (J^); i.e. D+ = 1/21(5^ + D 2 + J A B 2 1 1/2 ; N = 1/2 ( J ^ + JBX); L = 1/2 ( J ^ - JBX).  261  ^C.  N  X  S 4-*  a T3 C D O  o, E o U o  E 3  fa  O 4) O, CO  Pi  s Z  u C<1  u  3 OX)  262  With v A = vB = v P , as in the case of the methylene carbons, five lines (e, c, b, a, d) centered around line b are given in Table A6. Furthermore, if J ^ > > J AX > > JBX, lines e and d will disappear, as their intensity tends to be zero. The remaining three lines a, b and c (intensity 1:2:1) form a partially-resolved triplet with line-spacing, 1/2JAX-  This type of coupling is called virtual coupling. 191  Aromatic CHs of  bisphosphine dioxides 114-117 give this type of triplets, where J AX is the coupling constant between l^C and 31p (Figure A3 is an example). As indicated in Table A3, the coupling constants obtained are comparable to those of the corresponding monophosphine oxides.  Table A6. l^C NMR Transition Energies and Intensities for PCH2CH2P, where A = B = 3 l p , X = 13C transition  energy  relative intensity  a =9 b = 10 + 11 c = 12 d = 14 e = 15  vx-N v x  1 [1 + COS(7t - 2(j))]  1  vx + N vx-2D v x + 2D  1/2[1 - COS(TI - 2<|>)] 1/2[1 - COS(TC - 2<|>)]  Dcosfr = 1/2L, Dsiiuj) = 1/2 ( J ^ ) ; i.e. D = 1/2(L2 + im2)ll2\  N=  1/2(JAX + JBX); L = 1/2(JAX-JBX)-  It is a misunderstanding that identical nuclei do not couple each other.  All  bisphosphine dioxides show strong coupling between the two phosphorus atoms. We were able to calculate these coupling constants from the part of the 13C NMR spectra assigned to the PCH2- CH 2 P moiety. The spectra of 115 and 117 at 200 MHz, 300 MHz and/or 400 MHz clearly showed five lines (for example, the peaks in the region of 20-23  263  ppm except the peak at 21.3 ppm in Figure A3), and the integration of these peaks could be carried out. The calculated results are listed in Table A7. The results are comparable with each other and within the range of the literature values. 190  Table A7. Coupling Constants (Hz) of Bis(phosphine oxide)s 115 and 117 oxide  spectrometer  J PP  115  75 MHz 50 MHz 50 MHz 100 MHz  52.1 56.2 52.7 54.7  117  PC  he  67.8 64.5 68.4 67.2  1.8 2.3 2.0 1.2  J  Virtual coupling and ABX type spectra were also observed in etheno- and ethanoanthracene derivatives as well as in rra«j-l,2-ethenediylbis(diphenylphosphine oxide). Figure A4 shows a typical ABX type spectrum from ethanoanthracene 135. The  -1  !  1  142. 5  r  -i  1  -i  r  142. 0 PFM  1  T -i 141.5  r-  1  r  Figure A4. ABX Type Spectrum from ^C NMR of Ethanoanthracene 135 (this part of spectrum could be assigned to C4a and C8a).  264  spectral assignments for these phosphine oxides were done in the same way as described above (see Experimental section for detail). In few cases, the assignment of 1-*C NMR spectra was not possible because there were many different types of aryl or phenyl groups present.  265  REFERENCES (1)  Federsel, H.-J. CHEMTECH1993, 24.  (2) Asymmetric Synthesis, Morrison, J. D., Ed.; Academic Press: New York, 1983; Vol. 1-5. (3)  Morrison, J. D.; Mosher, H. S. Asymmetric Reactions; Prentice-Hall: Englewood Cliffs, New Jersey; Revised Ed. 1976, ACS Books: Washington, D.C.  (4)  Catalytic Asymmetric Synthesis, Ojima, I., Ed.; VCH: New York; 1993.  (5)  Inoue, Y. Chem. Rev. 1992, 92, 741.  (6) Scheffer, J. R.; Garcia-Garibay, M.; Nalamasu, O. In Photochemistry on Solid Surfaces, Anpo, M., Matsuura, T., Eds.; Elsevier: Amsterdam, 1989. (7) Kohlshutter, H. W. Z. Anorg. Allg. Chem. 1918, 105, 121. (8)  (a) Cohen, M. D.; Schmidt, G. M. J. J. Chem. Soc. 1964, 1996. (b) Cohen, M. D.; Schmidt, G. M. J.; Sonntag, F. I. /. Chem. Soc. 1964, 2000. (c) Schmidt, G. M. J. J. Chem. Soc. 1964, 2014.  (9)  (a) Cohen, M. D. Angew. Chem., Int. Ed. Engl. 1975, 14, 386. (b) Cohen, M. D. Mol. Cryst. Liq. Cryst. 1979, 50, 1.  (10) (a) Ramamurthy, V.; Weiss, R. G.; Hammand, G. S.  In Adv. Photochem.,  Volman, D. H., Hammond, G. S., Neckers, D. C. Eds; John Wiley: New York, 1993; Vol. 18, p 67. (b) Weiss, R. G.; Ramamurthy, V.; Hammond, G. S. Ace. Chem. Res. 1993, 26, 530. (11) (a) Wegner, G. Pure Appl. Chem. 1977, 49, 443. (b) Enkelmann, V.; Wegner, G. /. Am. Chem. Soc. 1993, 115, 10390. (c) Iwamoto, T.; Kashino, S. Bull. Chem. Soc. Jpn. 1993, 66, 2190. (12) (a) Trotter, J. Acta Crystallogr., Sect B 1983, B39, 373. (b) Organic Chemistry in Anisotropic Media, Scheffer, J. R., Turro, N. J., Ramamurthy, V., Eds.; 266  Tetrahedron Symposia-in-Print, 29, Tetrahedron 1987. (c) Scheffer, J. R.; GarciaGaribay, M.; Nalamasu, O. In Organic Photochemistry, Padwa, A., Ed.; Marcel Dekker: New York, 1987; Vol. 2, Part 2, Ch. 20. (d) Chen, J.; Scheffer, J. R; Trotter, J. Tetrahedron 1992, 48, 3251. (13) Ariel, S.; Askari, S.; Evans, S. V.; Hwang, C.; Jay, J.; Scheffer, J. R.; Trotter, J.; Walsh, L. Tetrahedron 1987, 43, 1253. (14) Lewis, T. J.; Retting, S. J.; Scheffer, J. R.; Trotter, J. /. Am. Chem. Soc. 1991, 113, 8180. (15) (a) Zimmerman, H. In Organic Photochemistry, Padwa, A., Ed.; Marcel Dekker: New York, 1991; Vol. 11, p 1. (b) Zimmerman, H. E. In Rearrangements in Ground and Excited States; de Mayo, P., Ed.; Wiley: New York, 1980; Ch. 16. (c) Hixson, S. S.; Mariano, P. S.; Zimmerman, H. E. Chem. Rev. 1973, 73, 531. (16) (a) Zimmerman, H. E.; Kutateladze, A. G.; Mackawa, Y.; Mangette, J. E. /. Am. Chem. Soc. 1994, 116, 9795.  (b) Zimmerman H. E.; Boettcher, R. J.;  Buehler, N. E.; Keck, G. E.; Steinmetz, M. G. /. Am. Chem. Soc. 1976, 98, 7680. (17) (a) Adam, W.; Dorr, M.; Kron, J. Rosenthal, R. J. J. Am. Chem. Soc. 1987, 109, 7074. (b) Adam, W.; De Lucchi, O.; Dorr, M. J. Am. Chem. Soc. 1989, 111, 5209. (18) (a) Zimmerman, H. E.; Binkley, R. W.; Givens, R. S.; Sherwin, M. A. /. Am. Chem. Soc. 1967, 89, 3932. (b) Zimmerman, H. E.; Binkley, R. W.; Givens, R. S.; Grunewald, G. L.; Sherwin, M. A. /. Am. Chem. Soc. 1969, 91, 3316. (19) (a) Paquette, L. A.; Bay, E. J. Org. Chem.. 1982, 47, 4549. (b) Paquete, L. A.; Bay, E. J. Am. Chem. Soc. 1984, 106, 6693. (20) Demuth, M.; Lemmer, D.; Schaffner, K. J. Am. Chem. Soc. 1980, 102, 5407.  267  (21) (a) Paquette, L. A.; Varadarajan, A.; Bay, E. /. Am. Chem. Soc. 1984, 106, 6702. (b) Paquette, L. A.; Varadarajan, A.; Bay, E. /. Am .Chem. Soc. 1986, 108, 8032. (22) Saltiel, J.; D'Agostino, J.; Megarity, E. D.; Metts, L.; Neuberger, K. R.; Wrighton, M.; Zefiriov, O. C. In Organic Photochemistry, Chapman, O. L., Ed.; Marcel Dekker: New York, 1973; Vol. 3. (23) Woodward, R. B.; Hoffman, R. The Conservation of Orbital Symmetry; Verlag Chemie: Wienheim, 1970. (24) Zimmerman, H. E.; Factor, R. E. Tetrahedron, 1981, 37, supplement 1, 125. (25) Zimmerman, H. E.; Pratt, A. C. /. Am. Chem. Soc. 1970, 92, 1409 and 6259. (26) (a) Zimmerman, H. E.; Givens, R.. S.; Pagni, R. M. /. Am. Chem. Soc. 1968, 90, 6090. (b) Zimmerman, H. E.; Bender, C. O. J. Am. Chem. Soc. 1970, 92, 4366. (27) Ciganek, E. J. Am. Chem. Soc. 1966, 88, 2882. (28) Scheffer, J. R.; Trotter, J; Yang, J In Handbook of Organic Photochemistry and Photobiology, Horspool, W. M., Ed.; CRC Press, 1994; In press. (29) (a) Rabideu, P. W.; Hamilton, J. B.; Friedman, L. J. Am. Chem. Soc. 1968, 90, 4465. (b) Adam, W.; De Lucchi, O.; Peters, K.; Peters, E; Von Schering, H. G. /. Am. Chem. Soc. 1982, 104, 5747. (30) (a) Zimmerman, H. E.; Givens, R. S.; Pagni, R. M. J. Am. Chem. Soc. 1968, 90, 6096. (b) Zimmerman, H. E.; Bender, C. O. J. Am. Chem. Soc. 1970, 92, 4366. (c) Bender, C. O.; Shugarman, S. S. /. Chem. Soc, Chem. Commun. 1974, 934. (d) Bender, C. O.; Brooks, D. W. Can. J. Chem. 1975, 53, 1684. (31) Zimmerman, H. E.; Zuraw, M. J. J. Am. Chem. Soc. 1989, 111, 7974. (32) Pokkuluri, P. R.; Scheffer, J. R.; Trotter, J. J. Am. Chem. Soc. 1990, 112, 3676. (33) Addadi, L.; Lahav, M. In Origins of Optical Activity in Nature, Walker, D. C., Ed.; Elsevier: New York, 1979; Ch. 14. 268  (34) (a) Kondepudi, D.; Kaufman, R.; Singh, N.  Science 1990, 250, 975.  (b)  Kondepudi, D.; Bullock, K. M.; Digits, J. A.; Hall, J. K; Miller, J. M. J. Am. Chem. Soc. 1993, 775, 10211. (35) Penzien, K.; Schmidt, G. M. J. Angew. Chem., Int. Ed. Engl. 1969, 8, 608. (36) (a) Green, B. S.; Lahav, M.; Schmidt, G. M. J. Mol. Cryst. Liq. Cryst. 1975, 29, 187. (b) Elgavi, A.; Green, B. S.; Schmidt, G. M. J. J. Am. Chem. Soc. 1973, 95, 2058. (37) Green, B. S.; Lahav., M. /. Mol. Evol. 1975, 6, 99. (38) Photochemistry in Organized and Constrained Media, Ramammurthy, V., Ed.; VCH: New York, 1991. (39) Addadi, L.; Lahav, M. Pure Appl. Chem. 1979, 57, 1269. (40) Hasegawa, M.; Chung, C. -M.; Muro, N.; Maekawa, Y. J. Am. Chem. Soc. 1990, 772, 5676. (41) Suzuki, T.; Fukishima, T.; Yamashita, Y.; Miyashi, T. /. Am. Chem. Soc. 1994, 776, 2793. (42) Sakamoto, M.; Takahashi, M.; Fujita, T.; Watanabe, S.; Ida, I; Nishio, T.; Aoyama, H. J. Org. Chem. 1993, 58, 3476. (43) Evans, S. V.; Garcia-Garibay, M.; Omkaram, N.; Scheffer, J. R.; Trotter, J. /. Am. Chem. Soc. 1986, 108, 5648. (44) (a) Toda, F.; Yagi, M.; Soda, S. /. Chem. Soc, Chem. Commun. 1987, 1413. (b) Sekin, A.; Hori, K.; Ohashi, Y.; Yagi, M.; Toda, F. J. Am. Chem. Soc. 1989, 777, 697. (45) Caswell, L.; Garcia-Garibay, M.; Scheffer, J. R.; Trotter, J. J. Chem. Educ. 1993, 70, 785. (46) Toda, F. Synlett 1993, 303. (47) Toda, F.; Miyamoto, H.; Takeda, K.; Matsugawa, R.; Maruyama, N. /. Org. Chem. 1993, 58, 6208. 269  (48) Toda, F. Mol. Cryst. Liq. Inc. Nonlin. Opt. 1988, 161, 355. (49) Weber, L.; Imiolxzyk, I.; Haufe, G.; Rehorek, D.; Hennig H. /. Chem. Soc, Chem. Commun. 1992, 301. (50) (a) Gudmundsdottir, A.; Scheffer, J. R. Tetrahedron Lett. 1990, 31, 6807. (b) Gudmundsdottir, A.; Scheffer, J. R. Photochem. Photophys. 1990, 54, 535. (51) Gudmundsdottir, A.; Scheffer, J. R.; Trotter, J.  Tetrahedron Lett. 1994, 35,  1397. (52) Jones, R.; Scheffer, J. R.; Trotter, J.; Yang, J.  Tetrahedron Lett. 1992, 38,  5481. (53) (a) Takaya, H.; Noyori, R. In Catalytic Asymmetric Synthesis, Ojima, I. Ed.; VCH:New York, 1993; Ch. 1, reference herein, (b) Halpern, J. In Asymmetric Synthesis, Morrison, J. D., Ed.; Academic Press: New York, 1985; Vol. 5, Ch. 2. (54) Kagan, H. B.; Dang, T. -P. /. Am. Chem. Soc. 1972, 94, 6429. (55) (a) Knowles, W. S.; Sabacky, M. J.; Vineyard, B. D. /. Chem. Soc, Chem. Commun. 1972, 10.  (b) Vineyard, B. D.; Knowles, W. S.; Sabacky, M. J.;  Bachman, G. L.; Weinkauff, D. J. /. Am. Chem. Soc. 1977, 99, 5946. (56) Brunner, H. Top. Stereochem. 1988, 18, 129. (57) Fryzuk, M. D.; Bosnich, B. J. Am. Chem. Soc. 1977, 99, 6262. (58) Brunner, H.; Pieronczyk, W.; Schonhammer, B.; Streng, K.; Bernal, I.; Korp, J. Chem. Ber. 1981, 114, 1137. (59) Hayashi, T.; Kumada, M. In Fundamental Research in Homogenous Catalysis, Ishii, Y.; Tsutsui, M. Eds.; Plenum: New York, 1978; Vol. 2, p 159.  (b)  Hayashi, T.; Kumada, M. Ace Chem. Res. 1982, 15, 395. (60) (a) Miyashita, A.; Yasuda, A.; Takaya, H.; Toriumi, K.; Ito, T.; Souchi, T.; Noyori, R. J. Am. Chem. Soc. 1980, 102, 7932. (b) Takaya, H.; Akutagawa, S.; Noyori, R. Org. Synth. 1988, 67, 20. (c) Takaya, H.; Mashima, K.; Koyano,  270  K.; Yagi, M.; Kumobayashi, H.; Taketomi, T.; Akutagawa, S.; Noyori, R. J. Org. Chem. 1986, 51, 629. (61) Achiwa, K. J. Am. Chem. Soc. 1976, 98, 8265. (62) Burk, M. J; Feaster, J. E.; Harlow, R. L. Organometallics 1990, 9, 2653. (b) Burk, M. J. J. Am. Chem. Soc. 1991, 113, 8518. (63) (a) Koenig, K. E. In Asymmetric Synthesis, Morrison, J. D. Ed.; Academic Press: Orlando, FL, 1985; Vol. 5, p 71. (b) Togni, A.; Venanzi, L. M. Angew. Chem. Int.. Ed. Engl. 1994, 33, 497. (64) Asymmetric Catalysis, Bosnich, B., Ed.; Martinus Nijhoff: Dordrecht, 1986. (65) James, B. R.; McMillan, R. S.; Morris, R. H.; Wang, D. K. W. Adv. Chem. Ser. 1978, 167, 122. (b) Ball, R. G.; James, B. R.; Trotter, J.; Wang, D. K. W. /. Chem. Soc, Chem. Commun. 1979, 460. (66) (a) Kawano, H.; Ikariya, T.; Ishii, Y.; Saburi, M.; Yoshikawa, S.; Uchida, Y.; Kumobayyashi, H. J. Chem. Soc, Perkin Trans. 1, 1989, 1571. (b) Ohta, T.; Takaya, H.; Noyori, R.  Inorg. Chem. 1988, 27, 566.  (c) Kitamura, M.;  Tolunaga, M.; Noyori, R. /. Org. Chem. 1992, 57, 4053. (67) Mashima, K.; Kusano, K.; Sato, N.; Matsumura, Y.; Nozaki, K.; Kumobayashi, H.; Sayo, N.; Hori, Y.; Ishizaki, T.; Akutagawa, S.; Takaya, H. /. Org. Chem. 1994, 59, 3064. (68) Genet, J. P.; Pinel, C ; Ratovelomanana-Vidal, V.; Mallart, S.; Pfister, X., De Andrade, M. C. C.; Laffitte, J. A. Tetrahedron Asymmetry 1994, 5, 665. (69) (a) Noyori, R. Science, 1990, 248, 1194. (b) Kitamura, M.; Hsiao, Y.; Ohta, M.; Tsukamoto, M.; Ohta, T.; Takaya, H.; Noyori, R. /. Org. Chem. 1994, 59, 297. (70) Broene, R. D.; Buchwald, S. L. J. Am. Chem. Soc. 1993, 775, 12569. (71) Takaya, H.; Ohta, T. In Catalytic Asymmetric Synthesis, Ojima, I. Ed.; VCH, 1993; Ch. 1. 271  (72) James, B. R. Chem. Industry. 1994, In press. (73) Powell, H. M. /. Chem. Soc. 1948, 61. (74) Weber, E. Top. Curr. Chem. 1987, 140, 1. (75) (a) Inclusion Compounds, Atwood, J. L., Davies, J. E. D., MacNicol, D. D., Eds; Academic Press: London, 1984; Vols. 1-3; University Press: Oxford, 1991; Vols. 4-5. (b) Top. Curr. Chem. 1987, 140. (c) Top. Curr. Chem. 1988, 149. (76) Toda F. Top. Curr. Chem. 1988, 140, 43. (77) Weber, E.; Czugler, M. Top. Curr. Chem. 1988, 149, 45. (78) (a) Kaupp, G. Angew. Chem. Int. Ed. Engl. 1994, 33, 728. (b) Etter, M. C ; Vaures, P. W. J. Am. Chem. Soc. 1988, 149, 45. (79) Toda, F. Top. Curr. Chem.1988, 149, 211. (80) (a) Garcia-Garibay, M. A.; Scheffer, J. R.; Watson, D. G. J. Org. Chem. 1992, 57, 247.  (b) Garcia-Garibay, M. A. Ph.D. Thesis; the University of British  Columbia, 1990. (81) (a) Gudmundsdottir, A. D.; Scheffer, J. R.; Trotter, J. Tetrahedron Lett. 1994, 35, 1397. (b) Gudmundsdottir, A. D.; Li, W.; Scheffer, J. R.; Rettig, S.; Trotter, J. Mol. Cryst. Liq. Cryst. 1994, 240, 81. (c) Gudmundsdottir, A.D. Ph. D. Thesis, the University of British Columbia; 1993. (82) Valentine, D. Jr. In Asymmetric Synthesis, Morrison J. D. Ed; Academic: New York, 1983; Vol. 4. (83) Koide, Y.; Sakamoto, A.; Imamoto, T. Tetrahedron Lett. 1991, 32, 3375. (84) Corey, E. J.; Chen, Z.; Tanoury, G. /. Am. Chem. Soc. 1993, 115, 11000. (85) Gallagher, M. J. In The Chemistry of Organophosphorus Compounds, Phosphine Oxides, Sulphides, Selenides and Tellurides, Hartley, F. R. Ed.; John Wiley & Sons: Chichester, 1992; Vol. 2, Ch. 2.  272  (86) (a) Pietrusiewicz, M. K.; Zablocka, M.  Chem. Rev. 1994, 94, 1375.  (b) Cardellicchio, C ; Fiandanese, V.; Naso, F.; Pacifico, S.; Koprowski, M.; Pietrusiewicz, K. M. Tetrahedron Lett. 1994, 35, 6343. (87) Korpium, O.; Lewis, R. A.; Chickos, J.; Mislow, K. J. Am. Chem. Soc. 1968, 90, 4842. (88) Knowles, W.S.; Sabacky, M. J.; Vineyard, B. D. Homogenous Catalysis II, Adv. Chem. Ser. 1974, 132, 274. (89) Lewis, R. A.; Korpium, O; Mislow, K. /. Am. Chem. Soc. 1968, 90, 4847. (90) Maryanoff, C. A.; Maryanoff, B. E.; Tang, R.; Mislow, K. J. Am. Chem. Soc. 1973, 95, 5839. (91) Fleming, Ian; Ghosh, S. K. J. Chem. Soc. Chem. Commun. 1994, 99. (92) Hays, H. R.; Peterson, D. J. In Organic Phosphorus Chemistry, Kosolapoff, G. M., Maier, L., Eds; Wiley-Interscience: New York, 1972; Vol. 3, Ch. 6. (93) Denney, D. B.; Tsolis, A. K.; Mislow, K. /. Am. Chem. Soc. 1964, 86, 4486. (94) Traxler, J. T. Synthetic Communications 1977, 7, 161. (95) Rogers, M. E.; Averill, B. A. /. Org. Chem. 1986, 51, 3308. (96) Kuritani, M.; Sakata, Y.; Ogura, F.; Nakagawa, M. Bull. Chem. Soc. Jap. 1973, 46, 605. (97) Charrier, C.; Chodkiewicz, W.; Cadiot, P. Bull. Soc. Chim. Fr. 1966, 1002. (98) Von Hartman, H.; Beermann, C ; Czempik, H.; Z. Anorg. Allg. Chem. 1956, 287, 261. (99) Diels, O.; Alder, K. Justus Liebigs Ann. Chem. 1931, 486, 191. (lOO)Fringuelli, F.; Taticchi, A. Dienes in the Diels-Alder Reaction; John-Wiley: New York, 1990. (101)Kyba, E. P.; Rines, S. P.; Owens, P. W.; Chou, S. -S. P. Tetrahedron Lett. 1981, 22, 1875. (102)Martin, D. J.; Griffin, C. J. Org. Chem. 1965, 30, 4034. 273  (103)Koerner, M.; Rickborn, B. /. Org. Chem. 1989, 54, 6. (104)(a) Brunner, H; Pieronczyk, W. Angew. Chem. 1979, 91, 655. (b) Brunner, H.; Probster, M / . Organometal. Chem. 1981, 209, CI. (105)Orama, O.; Karhu, M.; Nasakkala, M; Sundberg, M.; Uggla, R. Cryst. Struct. Comm. 1979, 8, 409. (106) Harvey, R. G. Synthesis 1970, 101. (107)Hudlick, T.; Sinai-Zingde, G.; Natchus, M. G.  Tetrahedron. Lett. 1987, 28,  5287. (108)Minami, T.; Okada, Y; Nomura, R.; Hirota, S.; Nagahara, Y.; Fukuyama, K. Chem. Lett. 1986, 613. (109)Youn, I. K.; Yon, G. H.; Pak, C. S. Tetrahedron Lett. 1986, 27, 2409. (HO)Birch, A. J. Rao, G. S. Adv. Org. Chem. 1972, 8, 1. (lll)Lobana, T. S. In The Chemistry of Organophosphorus Compounds, Phosphine Oxides, Sulphides, Selenides and Tellurides, Hartley, F. R., Ed.; John Wiley & Sons: Chichester, 1992; Vol. 2, Ch. 1. (112)Miyashita, A.; Yasuda, A.; Takaya, H.; Toriumi,K.; Ito, T.; Souchi, T.; Noyori, R. / . Am. Chem. Soc. 1980 102, 7932. (113)Brunner, H.; Pieronczyk, W. Angew. Chem., Int. Ed. Engl. 1979, 18, 620. (114)(a) Brunner, H.; Probster, M. Inorg. Chim. Acta 1982, 61, 129. (b) Brunner, H.; Pieronczyk, W.; Schonhammer, B.; Streng, K.; Bernal, I.; Korp, J. Chem. Ber. 1981, 114, 1137. (115)Okada, Y.; Minami, T.; Yamamoto, T.; Ichikawa, J. Chem. Lett. 1992, 547. (116)Fryzuk, M. D.; Bosnich, B. J. Am. Chem. Soc. 1977, 99, 6262. (117)Brunner, H.; Pieronczyk, W.; Schonhammer, B.; Streng, K.; Bernal, I.; Korp, J. Chem. Ber. 1981, 114, 1137.  274  (118)(a) Asymmetric Synthesis, Morrison, J. D., Ed; Academic Press: New York, 1983; Vol. 1. (b) Jacques, J.; Collet, A.; Wilen, S. H. Enantiomers, Racemates and Resolutions; John Wiley: New York, 1981. (119) (a) ACS Symposium Series 471, Chiral Separation by Liquid Chromatography, Ahuja,  S.,  Ed.;  American  Chemical  Society,  Washington  DC 1991.  (b) Stevenson, D; Wilson, I. D. Chiral Separations; Plenum: New York, 1988. (120)(a) Wenzel, T. J. NMR Shift Reagents; CRC Press: Baca Raton, FL, 1987. (b) Parker, D. Chem. Rev. 1991, 91, 1441. (121)Fulwood, R.; Parker, D. Tetrahedron Asymmetry 1992, 3, 25. (122)Lewis, R. A.; Naumann, K.; DeBruin, K. E.; Mislow, K. /. Chem. Soc, Chem. Commun. 1969, 1010. (123)Pirkle, W. H.; Beare, S. D.; Munta, R. L. /. Am. Chem. Soc. 1969, 91, 4575. (124)Moriyama, M.; Bentrude, W. G. J. Am. Chem. Soc. 1983, 105, 4727. (125)(a) Dunach, E.; Kagan, H. B. Tetrahedron Lett. 1985, 26, 2649. (b) Zhou, Z. Huaxue Shiji 1987, 9, 109; Chem. Abstr. 1987, 107, 108337. (126)Tambute, A.; Gareil, P.; Caude, M.; Rosset, R. /. Chromatogr. 1980, 363, 81. (127)Macandiere, P.; Caude, M.; Rosset, R.; Tambute, A. /. Chromatogr. 1987, 405, 135. (128)Pescher, P.; Caude, M.; Rosset, R.; Tambute, A. /. Chromatogr. 1986, 371, 159. (129)Okamoto, Y.; Honda, S.; Hatada, K.; Okamoto, Y.; Toga, Y.; Kobayashi. S. Bull. Chem. Soc. Jpn. 1984, 57, 1681. (130) Snyder, L. R.; Kirkland, J. J. Introduction to Modern Liquid Chromatography; John Wiley & Sons: New York, 1974; Ch. 2. (131)Etter, M. C ; Vaures, P. W. /. Am. Chem. Soc. 1988, 149, 45. (132)(a) Lynch, D. E.; Smith, G.; Byriel, K. A.; Kennard, C. H. L.; Whittaker, A. K.; Hanna, J. V. Aust. J. Chem. 1994, 47, 1401. (b) Lynch, D. E.; Smith, G.; Byriel, K. A.; Kennard, C. H. L. Aust. J. Chem. 1992, 45, 835. 275  (133)Etter, M. C ; Gillard, R.; Glaeson, W. B.; Rasmussen, J. K.; Duerst, R. W.; Johnson, R. B. J. Org. Chem. 1986, 51, 5405. (134)Etter, M. C ; Reutzel, S. M. /. Am. Chem. Soc. 1991, 113, 2586. (135)Kropf, H.; Munke, S. /. Chem. Res. 1990, 26. (136)Fuquen, R. M.; Lechat, J. R. Acta Cryst. C1992, 48, 1690. (137)Gramstad, T. Acta Chem. Scand. 1992, 46, 1087. (138)Daasch, L. W.; Smith, D. C. Analyt. Chem. 1951, 23, 853. (139)Vainshtein, B. K.; Fridkin, V. M.; Indenbom, V. C. Modern Crystallography II, Structure of Crystals; Spring-verlag: Berlin, 1982; Ch. 1. (140)Chen, J.; Scheffer, J. R.; Trotter, J. Tetrahedron 1992, 48, 3251. (141)Garcia-Garibay, M.; Scheffer, J. R.; Trotter, J. J. Am. Chem. Soc. 1989, 111, 4985. (142)Demuth, M.; Amrein, W.; Bender, C. O.; Braslavsky, S. E.; Burger, U.; George, M. V.; Lemmer, D.; Schaffner, K. Tetrahedron 1981, 37, 3245. (143)Garcia-Garibay, M. A.; Scheffer, J. R..; Watson, D. G. /. Chem. Soc. Chem. Commun. 1989, 600. (144)Dewar, M. J. S. J. Am. Chem. Soc. 1952, 74, 3341. (145)Bijvoer, J. M ; Peerdeman, A. F.; Van bommel J. A. Nature 1951, 168, 271. (146)Ramamurthy, V.; Eaton, D. F.; Caspar, J. V. Ace. Chem. Res. 1992, 25, 299. (147) Borecka, B.; Gudmundsdottir, A. D.; Olovsson, G.; Ramamurthy, V.; Scheffer, J. R.; Trotter, J. J. Amer. Chem. Soc. 1994, 116, 10332. (148) Toda, F.; Tanaka, K.; Yagi, M. Tetrahedron 1987, 43, 1495. (149)Hayashi, N.; Mazaki, Y.; Kobayashi, K. Tetrahedron Lett. 1994, 5883. (150)Tang, C. P.; Chang, H. C ; Popovitz-Biro, R.; Frolow, F.; Lahav, M.; Leiserowitz, L.; McMullan, R. K. J. Am. Chem. Soc. 1985, 107, 4058. (151)Popovitz-Biro, R.; Tang, C. P.; Chang, H. C ; Lahav, M.; Leiserowitz, L. /. Am. Chem. Soc. 1985, 107, 4043. 276  (152)Weber, E.; Csoregh, I.; Stensland, B.; Czugler, M. /. Am. Chem. Soc. 1984, 106, 3297. (153)Dondi, A. /. Phys. Chem. 1964, 68, 441. (154) Jones ten, M. D.; Schaad, L. J. Hydrogen Bonding; Marcel Dekker: New York, 1974; Ch. 1. (155)Garcia-Garibay, M.; Scheffer, J. R.; Trotter, J.; Wireko, F. /. Am. Chem. Soc. 1989, 111, 4985. (156)Pincock, R. E.; Wilson, K. R.; /. Am. Chem. Soc. 1971, 93, 1291. (157)Pincock, R. E.; Perkins, R. R.; Ma, A. S; Wilson, K. R. Science, 1971, 174, 1018. (158)Richards, K. E.; Tillman, R. W.; Wright, G. J. Aust. J. Chem. 1975, 28, 1289. (159)(a) Gorenstein, D. G. Progress in NMR Spectroscopy 1983, 16, 1. (b) Beneara, C. 7. Am. Chem. Soc. 1973, 95, 6890. (c) Evelyn, L.; Hall, L. D.; Steiner, P. R.; Stokes, D. H. /. Chem. Soc, Chem. Comm. 1969, 576. (160)Wallace, P.; Warren, S. /. Chem. Soc. Perkin Trans. 1988, 1, 2971. (161)Abraham, R. J.; Loftus, P.  Proton and Carbon-13 NMR Spectroscopy, An  Integrated Approach; John Wiley: Chichester, 1985; p 44. (162)Paquette, L. A.; Bay, E. /. Am. Chem. Soc. 1984, 106, 6693. (163)Bury, A.; Earl, H. A.; Stirling, C. J. M. J. Chem. Soc. Perkin Trans. II, 1987, 1281. (164)Eliel, E. L.; Gilbert, E. C. J. Am. Chem. Soc. 1969, 91, 5487. (165)Jonesten, M. D.; Schaad, L. J. Hydrogen Bonding; Marcel Dekker: New York, 1974; Appendix. (166)Blaser, H. -U Chem. Rev. 1992, 92, 935. (167)Whitesell, J. K. Chem. Rev. 1989, 89, 1581. (168)James, B. R.; McMillan, R. S.; Morris, R. H.; Wang, D. K. W. Adv. Chem. Series 1978, 167, 122. 277  (169)Ball, R. G.; James, B. R.; Trotter, J.; Wang, D. K. W. /. Chem. Soc. Chem. Comm. 1979, 460. (170) (a) Brunner, H.; Pieronczyk, W. Angew. Chem. Int. Ed. Engl. 1979, 18, 620. (b) Kyba, E. P.; Davis, R. E.; Juri, P. N.; Shirley, K. R.; Inorg. Chem. 1981, 20, 3616. (171)Ponnamperuma, C ; Honda, Y.; Navarro-Gonzalez, R. In Symmetries Sci. [Proc. Sym.], Gruber, B., Yopp. J. H., Eds; 1989 (Pub. 1990): Vol. 4. (172)Kagan, H. B. In Comprehensive Organometallic Chemistry, Wilkinson, G., Stone, G. G. A., Abel, E. W., Eds.; Pergamon Press: Oxford, 1982; Vol. 5, p 463. (173)Ball, R. G.; Payne, N. C. Inorg. Chem. 1977, 16, 1187. (174)Koenig, K. E. In Asymmetric Synthesis, Morrison, J. D., Ed.; Academic Press: Orlando, FL, 1985; Vol. 5, p 71. (175)Maier, L. Helv. Chim. Acta 1968, 57, 405. (176)Ciganek, E. J. Am. Chem. Soc. 1966, 88, 2882. (177)Figueys, H. P.; Dralanis, A. Tetrahedron 1972, 28, 3031. (178)Harwood, H. J.; Grisley, D. W. Jr. J. Am. Chem. Soc. 1960, 82, 423. (179)Gibson, C. S.; Johnson, D. A. J. Chem. Soc. 1928, 92. (180)Baechler, R.; Mislow, K. /. Am. Chem. Soc. 1970, 92, 3090. (181)Schilling, H. Chem. Ber. 1913, 46, 1066. (182)Hartmann, H.; Beermann, C ; Czempik, Herbert Z. Anorg. Allg. Chem. 1956, 287, 201. (183)Aguiar, A. M.; Daigle, D. /. Am. Chem. Soc. 1964, 86, 2299. (184)Powell, J.; Shaw, B. L. /. Chem. Soc. (A) 1968, 159. (185)Murov, S. L. Handbook of Photochemistry; Marcel Dekker: New York, 1973; p. 131. (186)Wagner, P. J.; Kochevar, I.; Kempainen, A. E. /. Am. Chem. Soc. 1972, 94, 7489. 278  (187)Albright, T. A.; Freeman, W. J.; Schweizer, E. E. /. Org. Chem. 1975, 40, 3437. (188)Silverstein, R. M.; Bassler, G. C ; Morrill, T. C. Spectrometric Identification of Organic Compounds; John Wiley & Sons: New York, 1981; p 265. (189)Abraham, R. J.; Fisher, J.; Loftus, P. Introduction to NMR Spectroscopy; John Wiley & Sons: Chichester, 1988; Ch. 4. (190)Gorenstein, D. G. Progress in NMR Spectroscopy 1983, 16, 1. (191) Atta-ur-Rahman Nuclear Magnetic Resonance, Basic Principles; Springer-Verlag: New York, 1986; p 61.  279  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0061500/manifest

Comment

Related Items