Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Photochemistry of medium and large membered ring diketones in both the solution and crystalline state Lewis, Thillairaj Johnathan 1993

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

Item Metadata

Download

Media
831-ubc_1993_spring_phd_lewis_thillairaj.pdf [ 9.87MB ]
Metadata
JSON: 831-1.0061803.json
JSON-LD: 831-1.0061803-ld.json
RDF/XML (Pretty): 831-1.0061803-rdf.xml
RDF/JSON: 831-1.0061803-rdf.json
Turtle: 831-1.0061803-turtle.txt
N-Triples: 831-1.0061803-rdf-ntriples.txt
Original Record: 831-1.0061803-source.json
Full Text
831-1.0061803-fulltext.txt
Citation
831-1.0061803.ris

Full Text

PHOTOCHEMISTRY OF MEDIUM AND LARGE MEMBERED RING DEKETONES IN BOTH THE SOLUTION AND CRYSTALLINE STATE by  THILLAIRAJ JOHNATHAN LEWIS B.Sc., University of Jaffna, Sri Lanka, 1985 A THESIS SUBMITTED IN PARTIAL FULFILMENT 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 March 1993 © Thillairaj Johnathan Lewis, 1993  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 my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.  (Signature)  Department of  ^H  y ASTY  The University of British Columbia Vancouver, Canada  Date  DE-6 (2/88)  0%141 AP i2.1 L  1(193  ABSTRACT  The type II photochemistry of nine diametric cyclic diketones (ten, twelve, fourteen, sixteen, eighteen, twenty, twenty-two, twenty-four and twenty-six membered rings), two non-diametric cyclic diketones (sixteen and seventeen membered rings) and two cyclic keto-alcohols (sixteen and eighteen membered rings) was studied in both solid and solution media in order to investigate the relationship between conformation and photoreactivity. Upon irradiation, most compounds investigated underwent smooth y-hydrogen abstraction. The stereoelectronic requirements for the y-hydrogen atom abstraction and the reactivity differences observed in the solid and solution reactions have been investigated with the help of X-ray crystallography and molecular mechanics calculations. Stereoselective cyclization was observed during the solid state photoreactions of all diketones investigated, with the exception of the ten, fourteen and seventeen membered ring compounds. The stereochemistry of the major solid state product correlated well with the hydrogen abstraction geometry found in the solid state conformation. Upon irradiation in solution, in rings larger than fourteen membered, the stereoselectivity is largely lost, but a slight preference for trans over cis cyclobutanol formation together with fairly large amounts of cleavage product formation were observed. In smaller rings, however, the product distributions in solution were quite similar to those observed in the solid state. In the above series of diketones, two groups of y-hydrogen atoms can be distinguished with respect to their stereoelectronic dispositions in the solid state conformations. The y-hydrogens having close 0.-H contacts (<2.78 A) with the carbonyl oxygen atoms make a boat-like abstraction geometry, whereas the other type of y-hydrogens with longer 0.-H contact distances (close to 3 A) make a chair-like abstraction geometry. The geometric parameters for abstraction of all closest y-hydrogen  atoms on each diketone remain close to their average values of d = 2.72 A, 01) = 52.4°, A = 82.4°, 0 = 114.7°. In non-diametric diketones the chemically non-equivalent 7-hydrogen atoms lead to the possibility of forming regioisomeric pairs of cyclization and cleavage products. In the solid state conformation of the sixteen membered non-diametric diketone, the nonequivalent 7-hydrogen atoms differ further in their stereoelectronic dispositions. The regioselectivity observed in this diketone reveals an efficient abstraction of the y-hydrogen atoms having close 0...H contacts (boat-like abstraction geometry), and almost no abstraction for the 7-hydrogen atoms with 0...H contacts close to 3 A (chair-like abstraction geometry). Interestingly, four diketones exhibit solid-solid phase transitions at elevated temperatures. In an attempt to investigate such solid state behavior, photochemical reactions were performed at various temperatures above and below the transition temperature, together with solid state NMR (13C CPMAS, deuterium wide line NMR techniques), solid state FTIR and X-ray powder diffraction studies. The phase transitions of three of the four diketones are reversible in nature, and above the transition points the molecular behaviour with regard to the photochemistry resembles that in isotropic fluid media. The X-ray crystal structure analysis of the two crystal modifications of the twenty six membered ring diketone, cyclohexacosane-1,14-dione, reveals the existence of "conformational polymorphism". The solid state photochemistry of these dimorphs shows a divergent photochemical behaviour, and such differences have been shown (with the help of the X-ray crystal structures), to be the result of conformational factors rather than packing effects. Investigations also reveal an irreversible solid-solid phase transformation of one of these two crystal modifications into its dimorph.  iv  In diketones, substitution of at least one of the a-hydrogen atoms with a methyl group (one methyl group at each a-carbon atom) completely changes the photoreaction pathway to yield type I products exclusively. Photolysis of three diametric diketones (sixteen, eighteen and twenty membered rings) included in zeolites afforded both Norrish type I and type II photoproducts.  TABLE OF CONTENTS pages ABSTRACT^ LIST OF FIGURES  ^ix  LIST OF TABLES  ^xvi  ACKNOWLEDGEMENTS  ^xviii  INTRODUCTION CHAPTER I  1.0. General  ^1  1.1. The Topochemical principle  ^2  1.1.1. Bimolecular reactions  ^3  1.1.2. Non-topochemical photodimerization  ^6  1.2. The concept of the reaction cavity  ^10  1.3. Unimolecular reactions  ^14  1.4 Type II reaction 1.4.1. Geometrical relationship for hydrogen abstraction  ^17 ^19  1.5. The Norrish type I reaction  ^22  1.6. Polymorphism and solid-solid phase transitions in crystals  ^24  1.7. Conformational analysis of medium sized ring and macrocyclic compounds 1.8. Objectives of current research  ^26 ^32  RESULTS AND DISCUSSION CHAPTER II  2.0. Preparation of starting materials  ^  36  vi  pages CHAPTER III  3.0. Photochemistry of medium and macrocyclic diketones that undergo ^46  type II reactions  ^49  3.1. Photochemistry of diametric diketones 3.1.1. Structural determination of the photoproducts  ^51  3.1.2. Photochemistry of non-diametric diketones  ^54  3.1.3. Photochemistry of cyclic keto-alcohols  ^57  3.2. Diketones that give cis-cyclobutanol derivative as the major photoproduct in the solid state  ^62  3.2.1. Solid state conformation and photochemistry  ^62  3.2.2. Hydrogen abstraction geometry and biradical geometry  ^65  3.2.3 Structure reactivity correlation  ^68  3.2.4 Singlet and triplet reactions.of diketones  ^70  3.3. Diketones that give trans-cyclobutanol as the major product in the solid state  ^84  3.3.1. Solid state conformation and photochemistry  ^84  3.3.2. Hydrogen abstraction geometry and biradical geometry  ^86  3.3.3. Structure reactivity correlation  ^88  3.4. Diketones in which the stereoselectivity is significantly lowered in the solid state  ^91  3.5. Chemoselectivity in the solution and solid state photoreactions of diketones 3.6. Photochemistry of the ten membered ring diketone  ^99 ^105  3.7. The best geometrical requirements for 7-hydrogen abstraction in diketones  ^109  vii  pages CHAPTER IV  4.0. Polymorphism, solid-solid phase transitions and solid phase order ^  dependent photochemistry of diketones  113  CHAPTER V  ^  5.0. The photochemistry of alkylated cyclic diketones  ^  5.1. The photochemistry of cyclic mono- and diketones in zeolites  148 156  EXPERIMENTAL. CHAPTER VI  ^  6.0. General information  165 ^ 169  6.1. Synthesis of starting materials 6.1.1. Synthesis of diametric diketones by ozonolysis of bicyclic  ^ 169  olefins  6.1.2. Synthesis of diametric diketones by Blomquist's high dilution ^  technique  171  6.1.3. Synthesis of non-diametric diketones by Blomquist's high ^  dilution technique  179  6.1.4. A miscellaneous reaction from pimeloyl chloride under high ^  dilution technique  181  6.1.5. Synthesis of diametric diketone by Dieckmann condensation reaction  ^  182  6.1.6. Synthesis of cyclic keto-alcohols by partial reduction of diketones 6.1.7. Synthesis of tetramethylated diketones 6.1.8. Deuteraction of diketones  ^ ^ ^  184  187  189  viii  pages 6.2. Photochemical studies  ^191  6.2.1. General  ^191  6.2.2. Photochemistry of diametric diketones  ^194  6.2.3. Photochemistry of non-diametric diketones  ^213  6.2.4. Photochemistry of keto-alcohols  ^220  6.2.5. Photochemistry of tetramethylated diketones  ^220  6.3. Quantum yield studies  ^223  6.4. Quenching studies  ^226  6.5. Computational generation of diketone geometries  ^227  BIBLIOGRAPHY  ^230  ix  LIST OF FIGURES Figures^ Caption^ pages ^3 1. Compounds displaying medium dependent photoreactivity 2.  Solid state photochemistry of the three crystal modifications of trans-cinnamic acid ^4  3.  A few examples of solid state photodimerization  ^7  4.  Three crystal modifications of anthracene derivatives  ^9  5.  (1) Favorable and (2) unfavorable solid state reactions in the reaction cavity ^11  6.  Photochemistry of a diacyl peroxide  ^12  7.  Steric compression control of compound 4  ^13  8.  The cobaloxime complex  ^14  9.  Photochemistry of (a) ene-diones, (b) cyclohexenones, (c) 13, 'yunsaturated ketones, (d) arylketones and (e) 1,4-dienes ^17  10.  Type II reaction  11.  Geometrical parameters for the abstraction of hydrogen atom by the carbonyl oxygen atom ^20  12.  Theoretically ideal values of the geometrical parameters for hydrogen abstraction by the carbonyl oxygen ^22  ^18  13.^The a-cleavage processes (Norrish type I reaction) of cyclic ketones ^23  pages 25  14.  Types of phase transitions ^  15.  The^solid^state^conformation^of^cis-1,6-diaminocyclodecane dihydrochloride^  27  The diamond lattice conformation of the ten membered carbon frame ^  27  17.  The solid state conformation of cyclododecane ^  28  18.  Strain-free diamond lattice carbon frame of the fourteen membered ring^  29  19.  The corner positions of cyclic compounds ^  30  20.  Transformation of the cyclohexadecane molecular conformation from square [4444] (diamond lattice) to rectangular [3535] ^  31  21.  Diametric diketones ^  32  22.  (a) Cisoid,^(b) gauche (pre-cis),^(c) gauche @re-trans) and (d) transoid or anti conformational arrangements of the biradical intermediates ^  34  Synthesis of the ten and twelve membered ring diketones via ozonolysis of bicyclic olefins ^  36  24.  Synthesis of diketones from diacyl derivatives ^  37  25.  A possible mechanism for the formation of diketones ^  39  26.  A possible mechanism for the formation of y-pyrone ^  41  27.  Fate of adipoyl chloride under different synthetic conditions ^  42  16.  23.  xi  28.  Synthesis of non-diametric diketones  pages ^44  29.  Methylation of diketones  ^45  30.  Photolysis of monoketones in cyclohexane  ^46  31.  A UV-VIS spectrum of 10-2 M diketone 9 in cyclohexane  ^48  32.  Photolysis of diametric diketones in cyclohexane  ^50  33.  Newman projection of cis and trans-cyclobutanol derivatives down the ring junction ^53  34.  1H ntrir spectrum of allyl moiety of ene-dione 13e  ^55  35.  Photolysis of non-diametric cyclic diketones  ^56  36.  Photolysis of cyclic keto-alcohols  ^58  37.  Solid state conformations of (a) twelve membered (MM2), (b) sixteen membered, (c) sixteen membered (non-diametric), (d) twenty membered, (e) twenty-four membered and (f) twenty-six ^ membered ring diketones 64  38.  Newman projections of boat-like and chair-like abstraction geometries down the C2-C3 carbon-carbon bonds ^65  39.  (a) Pre-cis biradical intermediate of diketone 9. (b) Biradical intermediate reaction center ^68  40.^(a) Ene-dione / trans-cyclobutanol versus quencher concentration plot. (b) Ene-dione / cis-cyclobutanol versus quencher concentration plot ^72  xii  41.  pages Singlet and triplet pathways of type II reactions in diketones ^73  42.  Diagrammatic representation of the formation of cis-cyclobutanol derivative 9c by the concerted cyclization of the solid state ^ conformation of diketone 9 74  43.  Possible low energy conformations of diketone 9 generated by MM2 ^78  44.  Possible low energy conformations of diketone 7 generated by MM2 ^84  45.  Solid state conformations of (a) eighteen, (b) twenty two, (c) twenty six (needles) membered ring diketones ^86  46.  Newman projection of the boat-like and chair-like abstraction geometries down the C2-C3 carbon-carbon bonds ^88  47.  (a) Biradical intermediate from diketone 10, (b) Biradical intermediate reaction center ^90  48.  Solid state conformation of the fourteen membered ring diketone 8.^91  49.  Five low energy conformations of diketone 8 generated by MM2....^96  50.  Conformational change from 8A to 8B  ^97  51.  A diagrammatic representation of the biradical intermediate reaction center showing 01, 02 and 0 dihedral angles  ^100  52.  Conformational interchange (pseudorotation) of cyclododecanone biradical intermediate from gauche to anti conformation ^105  53.^The X-ray crystal structure of diketone 6  ^106  pages ^107  54.  y and 6-Hydrogen abstraction pathways of cyclodecanone  55.  Three possible low energy conformations of diketone 6 generated by MM2 ^108  56.  The ORTEP stereodiagram of unit cell packing of (a) Plate (viewed from "a" axis), and (b) Needle (viewed from "b" axis) crystals of diketone 14 ^115  57.  Solid state photochemistry of a-adamantyl-p-chloroacetophenone.. ^116  58.  Differential scanning calorimetry thermograms of (a) Plate and (b) Needle crystal modifications of 14 ^118  59.  The powder diffraction patterns obtained for the (a) Plate, (b) Needle and (c) Annealed plate forms of 14 at room ^ temperature 119  60.  The solid state FTIR spectra of (a) Plates at 20°C, (b) Plates at 60°C and (c) Needles at 20°C in KBr matrices. (d) A solution ^ spectrum of diketone 14 recorded in CC14 122  61.  trans-1-(4-Pentanoylpheny1)-4-pentylcyclohexane (46) and trans-1-  hepty1-4-(4-pentanoylphenyl) cyclohexane (47)  ^125  62.  13C CP/MAS Spectra of Plates (a) at 27°C and (b) at 57°C, (c) 13C NMR spectrum of diketone 14 in hexane ^129  63.  The aliphatic region of diketone 14 in 13C CP/MAS spectra from (a) Plates at 27°C, (b) Plates at 57°C, (c) Annealed plates at 27°C ^ and (d) Needles at 27°C 130  64.^Differential scanning caloiimetry thermograms of (a) Diketone 9, (b) Diketone 10, and (c) Diketone 15 ^132  xiv  pages 65.  The solid state FTIR spectra of diketone 9 recorded at (a) 20°C, (b) 35°C, (c) 40°C in KBr matrices. (d) A solution spectrum of ^135 diketone 9 recorded in CC14  66.  13C CP/MAS spectra of diketone 9 at (a) 27°C, (b) 37°C, and (c) a solution 13C NMR recorded in CDC13 ^141  67.  The expanded methylene region of diketone 9 at 27°C (13CP/MAS) ^141  68.  2H NMR spectrum of diketone 9 (a) at 27°C and at37°C  69.  13C CP/MAS spectrum of the carbonyl region of diketone 9 at 32°C ^144  70.  13C CP/MAS spectra of diketone 10 (a) at 27°C and (b) 92°C. (c) A solution 13C NMR spectrum of diketone 10 recorded in ^ CDC13 146  71.  21-1 NMR spectra of diketone 10 (a) at 27°C and (b) 82°C  72.  Photoepimerization of tetramethylated diketones in the solid state ^148  73.  The ORTEP stereodiagrams of the sixteen membered ring diketone 2R*,8S*,10R*,16S*-tetramethylcyclohexadecane-1,9-dione^(31) and^the^twenty-four^membered^ring^diketone 2R*, 12S *,14R*,24S *4etramethylcyclotetracosane-1,13-di one (32)^151  74.  Type I and type II products from the photolysis of 2-methyl cyclododecanone ^153  75.^Photolysis of iii- and tetramethylcyclodecanone  ^142  ^147  ^153  XV  pages ^154  76.  Photolysis of 2,n-diphenylcycloalkanones.  77.  A diagrammatic representation of possible diastereomers of tetramethylated sixteen membered diketone (31), at the photostationary state ^155  78.  The two modes of photoreaction of mono and diketones in zeolites ^157  79.  Enhancement of type I products in zeolite NaY with respect to the solution reaction ^158  80.  Dependence of type I to type II product ratio on the cation  81.  Illustration of the [SiO414- and [A104]5- tetrahedra that are the primary building blocks of zeolites. Also shown are representations of the sodalite cage, zeolites (A, X and Y) and the unit cell compositions of X and Y zeolites 160  82.  Cation locations inside the faujasite cages  83.  Electronic interaction of lithium cation with the carbonyl chromophore impedes type II hydrogen abstraction sterically^163  84.  Diagram illustrating the partitioning of the type 11 (65) and type I (68) biradical intermediates to products and to their starting ketones ^163  85.  Apparatus used for the analytical photoreactions at elevated temperatures ^192  ^158  ^161  86. The flow diagram showing the "Monte Carlo Multiple-Minimum" conformational search ^228  xvi  LIST OF TABLES Tables^  Caption^  pages ^38  I.^Percentage yields of diketones and 7-pyrones  Product percentages from the irradiation of monoketones in cyclohexane ^47 Methine carbon chemical shifts of the cyclobutanol derivatives  ^52  IV.  Product percentages at zero percent conversion obtained at 20°C ^  V.  The geometrical parameters of the 7-hydrogen atoms  VI.  Type II product ratios of diketone 9, from direct photolysis and photolysis with the quencher ^72  VII.  Quantum efficiencies of type II photoproduct formation from diketones 9, 10 and 12 ^80  VIII.  The 7-hydrogen geometrical parameters of diketones that give trans-cyclobutanol derivatives as the major solid state photoproducts ^87  IX.  Cyclization / cleavage ratios of diketones both in solution and solid state ^100  X.  Biradical parameters of diketones  XI.  Geometrical parameters corresponding to 7-hydrogens having the closest 0...H contacts ^110  61 ^67  ^101  xv ii  pages XII.  The geometrical parameters of the biradical intermediates corresponding to the y-hydrogen atoms having closest 0...H contact distances ^112  XIII.  FTIR band frequencies (cm-1) of annealed plates, needle crystals and low and high temperature solid phases of plate crystals as a ^ function of temperature 123  XIV.  Product percentage percentages of diketone 14 from (a) Plates, (b) Needles, (c) Annealed plates and (d) isotropic media (hexane) as a ^ function of temperature 126  XV.  Photochemical results of 13-form cinnamic acid as a function of ^128 temperature  XVI.  Product percentages of diketones (a) 9, (b) 10, and (c) 15 as a function of temperature and medium ^136  XVII.  Solvent systems and their corresponding boiling points used for analytical irradiations at elevated temperatures ^192  XVIII.^Internal standards  ^225  xviii  ACKNOWLEDGEMENT I am very grateful to Professor John. R. Scheffer for the valuable guidance and encouragement he gave me throughout the period of my study. His understanding and patience helped bring this thesis to a successful conclusion. My special thanks go to Professor James Trotter and Dr. Steve Rettig for the X-ray crystallographic analysis. I also wish to express my sincere appreciation to Professor Trotter for letting me use the computer to print out the MM2 structures. My appreciation is also extended to Professor Larry Weiler, who allowed me to use the Silicon Graphics work station for MM2 calculations. The generous help with these calculations I received from Leonard Lerner is also gratefully acknowledged. I sincerely thank Professor Colin A. Fyfe and Dr. Leslie H. Randell, who performed the solid state NMR studies reported in this thesis, and I would also like to thank Mark Eade, Tony Fu, Ed Graziani and Mylvaganam Murugesupillai for proofreading my thesis. Finally, I would like to acknowledge the dedicated and friendly help of the NMR and Mass spectrometry laboratory staff and the service of the micro-analyst, Mr. P. Borda.  CHAPTER I. 1.0. General. Chemical reactions induced by sunlight have been a primary factor in the evolution of life on our planet.1 It is believed that within the first billion years of evolution, plant life began to produce oxygen into the atmosphere. Photolysis of oxygen in the stratosphere generated ozone, which provided a protective layer by filtering high energy radiation from sunlight, and was an essential factor in the survival of animal life forms, including man.2,3,4,5 Despite the fact that chemical changes caused by light have played such a  significant role in the evolution of life, only during the last two centuries have lightinduced chemical reactions, so called "photochemical reactions", been systematically investigated in the laboratory. In recent years, photochemistry has made immense contributions to the field of chemistry. Large numbers of complex molecules that are essentially unavailable by alternative synthetic methods or which could only be made via tedious synthesis procedures, can now be elegantly synthesized by photochemistry. Due to a long-standing notion among chemists that chemical reactions require mobility, molecules in crystals which have very small translational and vibrational motions were considered unsuitable for chemical reactions. Any studies of chemical reactions were therefore mainly focused on liquid and gaseous states, which were thought to possess the necessary molecular freedom. However, the molecular and atomic motions in the solid state are not as restricted as originally thought.6 The history of solid state chemistry can be traced back to the beginning of the nineteenth century. The thermal transformation of crystalline ammonium cyanate to urea by Friedrich Wohler (1828) could be considered as the first example of a solid state reaction.7 A few years later, in 1834, H. Trommsdorff discovered the first organic solid  2  state photoreaction, in which crystals of santonin, when exposed to sunlight, turned yellow and cleaved.8 During the course of the late 19th and early 20th centuries, work on solid state photochemistry increased dramatically, and a wide variety of photochemical reactions were investigated by the pioneers of photochemistry.9,10,1 1,12,13 These studies came to a halt mainly due to a lack of understanding of the nature and the structure of crystals. During the last forty years spectacular progress has been made in the field of solid state chemistry. Development of direct methods of crystal structure determination, and the advent of relatively inexpensive fully automated X-ray diffractometers with digital computers, enable one to obtain pictorial details of the molecular structure and the crystal lattice. A crystal structure obtained immediately prior to reaction led to the possibility of investigating structure-reactivity correlations with a deeper understanding of the molecular packing and the topochemistry in the crystal lattice. Unlike isotropic fluid phases, the molecules in crystals are arranged in a constraining environment; therefore the photochemical reaction pathways in the solid state tend to be controlled by the crystal lattice. In the majority of cases, the intrinsic reactivity of the molecule is modified in the solid state to yield the lattice-controlled photoproduct with high selectivity. Some examples of such lattice-controlled reactions, which differ from the intrinsic solution reactivity, are depicted in Figure 1.  1.1. The Topochemical Principle. Kohlschutter17 (1918) first proposed this principle, which is the basis of the entire field of organic solid state chemistry. He stated that the nature and the properties of the solid state reaction products are governed by the constraining influence of the three dimensional periodic environment present in the crystals, and thus the reactions tend to occur with a minimum of atomic and molecular motion.  3  hv  solution (Ref. 14)  hv  solution 0  dimers 0 (Ref. 15)  hv  hv  solution  crystal (Ref. 16)  OH  Figure 1: Compounds displaying medium dependent photoreactivity.  1.1.1. Bimolecular Reactions: In the early 1960s, after the advent of modern X-ray crystallography, Schmidt and co-workers systematically studied the factors that govern reactions in the organic solid state, especially the photo-induced [2+2] dimerization reactions of the trans-cinnamic acid system,18,19 and confirmed the postulate proposed by Kohlschutter. The photodimerization of cinnamic acids in crystals was first observed in 1889 by Liebermann20, and this incomplete work was later reinvestigated by Bernstein and Quimby2i in 1943, who interpreted the formation of a-truxillic and P truxinic acids from -  two types of cinnamic acid crystal modifications as lattice-controlled reactions. Schmidt and co-workers, as a result of their extensive studies on these systems, demonstrated a correlation between the crystal structure and the nature of the products obtained and also a correlation between the crystal structure and the photoreactivity or photostability of organic substances. These observations support the statement proposed by Kohlshutter, namely that reactions in the solid state proceed with minimum atomic and  4  molecular displacement22,23,24 with the preservation of configuration from reactants to products. However, as opposed to the solid state reactions, the molecules in the isotropic media were found to undergo photoisomerization to the cis-form. From the studies on trans-cinnamic acid and its derivatives, Schmidt reported an excellent illustration of the topochemical principle. The trans-cinnamic acid compounds generally show three crystal modifications, namely a, 13 and 7-types (Figure 2).18,19,22,24  a-type Ph \ C—C  \  HOOC\^COON C  \  hv crystal  COOH  HOO  a-truxillic acid  Ph  Ph  p.type  Ph Ph  \ ^ C^C \  \^ C=—C  \  COOH  hv --Ø, crystal  COON  Ph  COOH  COOH  p-truxinic acid  y-type ^ /COON C^=C  hv  Ph"^COOH  crystal  no reaction  C^ C/ Ph"  COOH ^ /^hv solution Ph/^ trans^  Ph/^COON cis  Figure 2: Solid state photochemistry of the three crystal modifications of trans-cinnamic acid  5  As shown in Figure 2: a)  In the a-type; the intermolecular center-to-center distance between the  overlapping double bonds of adjacent molecules is between 3.6-4.1 A, and the adjacent molecular pairs are related by a centrosymmetric anungement. b) In the 13-type; the adjacent molecules are parallel and slightly translated. The neighboring olefinic double bonds have considerable face-to-face overlap with a center-to-center distance of 3.9-4.1 A. c) In the y-type; the adjacent molecules are related by a translation but the double bonds are offset in such a way that they do not overlap. The center-to-center distance between the double bonds is between 4.7-5.2 A. Irradiation of a-type and I3-type crystals caused a photochemical dimerization reaction to yield centrosymmetric a-truxillic acids and mirror symmetric 13-truxinic acids, respectively, whereas 7-type crystals were found to be photostable. As a result of the above observations, Schmidt drew the following conclusions for solid state [2+2] photocycloaddition reactions: a)  The product formed is governed by the nature of the packing of the  neighbouring molecules around the reactant rather than the intrinsic reactivity of the reactant. b) Proximity (<4.1 A) and a parallel alignment of the potentially reacting centers are crucial for the dimerization. c) Dimerization between nearest neighbor molecules occurs with a minimum of molecular and atomic movement. d) There is a direct correlation between the configuration and symmetry of the products and reactants in the solid state photodimerizations.  6  Therefore a knowledge of the closest neighbor disposition as well as the determination of the distance between double bonds could allow one to predict the geometry of the products obtained by photoreaction. Photoreactive a and 13 type crystal modifications have intermolecular distances of 3.6 to 4.1 A, and so dimerization should not occur beyond this limit. Schmidt23 suggested an upper limit of 4.1 A for the distance between the reactive double bonds for a [2+2] photocycloaddition reaction. However, in the case of m-bromo-cinnamic acid, in spite of a 3.9 A distance between the potentially reactive double bonds, irradiation does not lead to any photodimerization.25 Here the photoinertness was suggested to be due to the non-parallel arrangement of the reactive double bonds, which leads to poor overlap between them. In the case of the photoreactive a and 13 crystal modifications, both distance and parallel alignment of double bond criteria are satisfied, but they are not fulfilled by the y-modification,2426 where the large distance (4.7-5.1 A) and poor overlap between the double bonds makes them photostable. In recent years, these topochemical postulates have provided a landmark in organic solid state photochemistry and have been used as rules to rationalize a large number of [2+2] photochemical dimeiization reactions of compounds with widely varying structures, such as the dimerization of fumaryl derivatives,27 butadiene derivatives,2829 coumarins,30-34 and benzilidene cyclopentanones.35,36 An example for each case is shown in Figure 3. 1.1.2. Non-topochemical Photodimerization. Although the majority of solid-state photodimerization reactions strictly follow the topochemical rules, there are reports of dimerization reactions which deviate significantly from the accepted topochemical postulates put forward by Schmidt. One example of such behaviour was observed by Ramamurthy et al. on coumarin derivatives.37 Irradiation of 7-chlorocoumarin causes a [2+2] photodimerization reaction even though the distance between the reactive double bonds (4.5 A) exceeds Schmidt's proposed upper limit of  7  4.2 A. Likewise, in 7-methoxycoumarin dimerization occurs in spite of the reactive double bonds being 65° with respect to each other with a center-to-center distance of 3.8  A. The above exceptions indicate that a certain amount of flexibility is required in applying these rules, depending on the system under investigation.  fumaryl derivatives /CN NC  /  hv  C^ C  -6,  crystal (Ref. 27)  butadiene derivatives  HOOC.,....e7/..  hv  HOOC  crystal  HOO  COOH^ --", HOOC.,,,,,,e.„,-  COOH  ..,7*COOH .7COOH (Ref. 28, 29)  coumarins: OMe ()me Me0  hv --, crystal  (Ref. 33)  benzilidenecyclopentanones  hv --I.  o  crystal^  X=H = Br (Ref. 35)  Figure 3: A few examples of solid state photodimerization.  8  Other examples of such non-topochemical photodimerizaion have been found for some 9-cyanoanthracene derivatives,38 in which the type of products formed cannot always be predicted from the X-ray structure of the starting materials. In the mid 1960s, Craig and Sarti-Fantoni,38 who first observed such non-topochemical dimerization, showed that these reactions take place at defects or surfaces or zones that are disordered. Schmidt and co-workers later investigated a series of anthracene derivatives and found many cases behaving in a non-topochemical fashion. 39,40 They have also shown that anthracene derivatives crystallize in three different crystal modifications, a, 13 and y, having three different packing arrangements as depicted in Figure 4. Molecules in a-type crystals are packed in head-to-tail (centrosymmetric) fashion and upon irradiation become topochemically controlled head-to-tail dimers. In the 13-type crystals, the molecules (R= Cl, CHO, CN) are arranged in a head-to-head (noncentrosymmetric) fashion, however, upon irradiation give head-to-tail photodimers instead of the topochemically controlled head-to-head products. In both the a and 13-types the neighboring molecules are arranged in pairs with considerable overlap between them. In the a-type the C(9')--C(101) contacts between the neighboring molecules is less than 4.1  A.  In the 13-type the distances between the meso-carbons of the nearest neighboring  molecules are less than 4.1  A. Whereas in the y-type, the non-parallel alignment and large  distance (> 5 A) between the meso-atoms of the neighboring molecules makes them photostable. Almost all 9-substituted anthracene molecules, upon irradiation hi solution, dimerize to yield only the head-to-tail dimers and not the head-to-head dimers. From these results it can be concluded that head-to-head dimers are not a stable form. In crystals, dislocations and other defect sites are commonly present, but they are not detectable by X-ray diffraction methods, which give the averaged structure of the molecules at regular sites. Since the number of molecules located at the defect sites is usually a small fraction of the molecules in the regular lattice sites, the photodimerization should be governed by the regular lattice sites. But in crystals where unusual  9  photochemical reactions are observed, the defect sites seem to govern the reaction, even though they are only present in small amounts. a-type:  hv  head-tail dimer R = Cl, Br, Me, CO2 Me, CONH2  head-head arrangement R = CI, CHO, CN.  mirror symmetric dimer 7-type  no reaction  R- OMe, Cl, CN  Figure 4: Three crystal modifications of andracene derivatives.  When crystals are irradiated three events can occur from the excited state molecule: deactivation by radiative or radiationless processes, reaction (dimerization in the  10  present example) and transfer of excitation to another site. If we assume that the deactivation process is independent of the nature of the site, then when energy is absorbed by molecules in a crystal the reaction takes place at the site where the absorption occurred. When the dimerization process is very slow, the process of transfer of excitation to a neighboring site has a higher probability of occurring. Since the normal symmetry of the sites is disrupted at the dislocations, molecules at these sites are likely to act as trapping centers for excitation. It has been suggested38,41 that the excitation energies of the anthracene molecules are slightly reduced when they are displaced from the regular lattice sites. Defect sites can therefore function as favoured areas for reaction. In other words, if the light energy can be transferred rapidly within the crystals after absorption, then the photochemistry of the ideal lattice need not be important. Instead, photoreaction would become more probable in regions where excitation energy can preferentially migrate. Although the number of molecules at the defect sites is small, as the reaction progresses the defect sites will multiply in order to give an appreciable yield of products. 1.2. The Concept of the Reaction Cavity Topochemical principles and defects can be used to rationalize chemical reactions in the crystalline state, regardless of whether molecules in the lattice are suitably arranged for a reaction or not. However, in some examples, even though molecules in the lattice are favorably arranged for a chemical reaction, they remain photostable.42 To explain this Cohen43 introduced a new qualitative concept into the topochemical principle known as the "Reaction Cavity". The cavity or the cage is the space in the crystal lattice occupied by the molecules which are directly involved in the chemical reaction. This space, with well defined dimensions and shapes, is limited or governed by the contact surfaces of the internal molecules with the surrounding molecules. Any molecular rearrangement or other  11  changes within the cavity during the process of a reaction can exert pressure on the cavity wall. In particular, the formation or removal of empty space within the cavity is energetically unfavorable, since these involve great changes in attractive and repulsive forces (Figure 5.). Therefore reactions that are controlled by the crystal lattice will occur with minimum deformation of the reaction cavity.  Figure 5: (1) Favorable and 2) unfavorable solid state reactions in the reaction cavity (transition state in dotted lines).  The "Reaction Cavity" concept has become very useful in explaining the photostability of several compounds where the molecules in the crystal are arranged (topochemically) ideally for photodimerization.42 Several examples of the unusual photobehaviour of crystals which do not obey the topochemical postulates have been reported from time to time, and a wide range of concepts has been put forward to provide more accurate information about the reaction cavity. McBride et a/.44 introduced the "local stress" concept to explain the mechanisms by which diacyl peroxides decompose in the solid state (Figure 6). Solid state irradiation of bis (3,3,3-triphenyl propanoyl) peroxide (1) was analyzed by ESR spectroscopy to identify the phenyl group (all three phenyl groups are non-equivalent by virtue of their different environments in the anisotropic medium) of the neophenyl radical (2) which migrated in the solid state to give radical (3). An anisotropic "local stress", developed  12  inside the crystal lattice by a recently liberated carbon dioxide molecule, was postulated to be the controlling factor in the selection and migration of the phenyl group. Moreover, the amount of stress was measured by FTIR, since a particular stretching frequency of carbon dioxide depends on pressure. It was suggested that the stress was transmitted to one side of the migration terminus, thus inclining the radical carbon in the opposite direction, more towards the migrating phenyl group. By comparing the behaviour of a variety of diacyl peroxides, it was shown that stress is a more important factor in these reactions than the topochemical postulate, including the shape of the cavity. Ph  0^0  II^II  Ph3CCH2 C —  -  0  —  0  (1)  —  C  —  CH2CPh3  hv crystal -CO2  ^  Ph  —  C 61 12 —I,. Ph—O—CHVPh -  -  I ^I  Ph^Ph (2)^(3)  Figure 6: Photochemistry of a diacyl peroxide.  Gavezzotti6,45 takes into account the volumes occupied by molecules and empty spaces in the crystalline arrangement. He developed a quick and precise method to calculate the volumes of the empty and filled spaces in the lattice. Packing density diagrams calculated from a computer program allowed him to locate the void zones or holes or channels in the crystal structure. Using these maps he analyzed a variety of solid state reactions and concluded that a prerequisite for crystal reactivity is the availability of free space around the reaction site. The free spaces around the reacting partners, the size of which may vary from system to system, can favour reaction between less than ideally oriented pairs. Scheffer, Trotter and co-workers46 took into account some specific steric interactions between the reactant molecules and surrounding molecules that prevent  13  chemical reactivity in the solid state. They termed this effect "Steric Compression Control". Compound (4) (Figure 7) undergoes unimolecular photorearrangement in the solution state, but in the solid state, even though the reactive double bonds in the neighboring molecules are arranged in a topochemically favorable distance and orientation, the crystal remains photostable. It has been proposed that the unreactivity is related to the arrangement of the neighboring molecules in the crystal lattice. If the [2+2] photocycloaddition reaction were to proceed, the steric compression of the two methyl groups of the reacting molecules with those of the surrounding molecules would increase, and thereby prevent the photodimerization. neighboring molecule NVVV\A. I^I Me, iio  0  neighboring molecule  (4)  ^(Ref. 46)  Figure 7: Steric compression control of compound 4.  A quantitative relationship between a solid state reaction rate and a crystal lattice parameter has been developed recently by Ohashi and co-workers.47,48 They investigated several optically active cobaloxime complexes (5) (Figure 8) which undergo racemization at the cyanoethyl chiral center upon the action of X-rays. Because of their single crystalto-single crystal (topotactic) nature, these reactions could be followed by crystal structure analysis (X-ray) of the reaction intermediates, and their relative rates determined. These rates were then con-elated with the reaction cavity volumes, which were calculated from  14  the X-ray crystal structure data. It was found that the larger the reaction cavity, the faster the reaction rate. 113 C.03.^CN HIPIN■C-  X = (S) - NH 2CH(CH3)C6H5 [(R) -1-cyanoethyl] = (S) - NH 2CH(CH3)C6H5 [(S) -1 -cyanoethyl] = C5H5 N [(S) -1-cyanoethyl] (5)  Figure 8: Cobaloxime complexes used in the experiment.  1.5. Unimolecular Reactions: As described earlier, the close proximity and proper alignment of the reactive sites of neighboring molecules are important prerequisites for a successful bimolecular photoreaction in the solid state. When these conditions are fulfilled, topochemically controlled dimers are formed in which the intrinsic and relative geometries of the reactants are preserved. Thus the intermolecular packing arrangements and the crystal lattice constraints play an important role in controlling solid state bimolecular reactions. In the last 30 years, studies of chemical reactions in the organic solid state have been concerned mainly with bimolecular processes. Correlation of the [2+2] photocyclization reactions with X-ray crystal structure data has provided valuable insights into the requirements for feasible reactions.  15  A major problem with studying bimolecular solid state reactions is the unpredictability of the preferred packing arrangement of organic molecules in crystals. In unimolecular reactions (such as intramolecular hydrogen abstraction, electrocyclization, fragmentation reactions, etc.), conformational factors usually determine the success and the type of the reaction, whereas intermolecular packing effects play only a secondary role. Therefore, X-ray crystal structure analysis provides an opportunity to study the intrinsic reactivity of a single conformer. Another advantage in dealing with unimolecular solid state reactions is that the molecules that make up the crystal lattice are generally found in their lowest or near to their lowest energy conformations. 49 These conformations, which may also be predicted using the principles of conformational analysis, provide a certain amount of information regarding possible unimolecular chemical reactivity in advance. Unlike bimolecular reactions, the majority of unimolecular photorearrangements proceed through relatively large conformational and configurational changes along the reaction coordinate. Favorable reaction pathways involving low energies of activation are often observed in unrestricted systems, such as isotropic fluid media. But in the crystalline phase, because the molecular motions are limited by the external physical restraints exerted by the lattice environment, and despite relatively high activation energies (in the isotropic medium), molecules tend to undergo alternate, less motion pathways to yield topochemically controlled products. These products are often different from the ones observed in isotropic liquid media, where they are usually formed by a greater motion pathway. Solution products are often topochemically forbidden in the solid state, since the conformational changes required for their formation are too large to be permitted by the constraints of the crystal lattice. Although sporadic reports on unimolecular photorearrangements are to be found in the literature of the last 100 years, such medium-dependent unimolecular photoreactivity has not been clearly explained owing to the absence of structural  16  information about the crystalline phase.50-53 At present, with the help of X-ray crystallography, which can accurately determine the structure as well as the environment of the starting materials, solid state reactivity patterns can be rationalized by means of structure-reactivity correlation studies. Such studies provide a clear understanding of the range and nature of the motions of atoms permitted during the rearrangement process. In most solid state photorearrangements it is assumed that the intermediates and the transition states in the reaction under study resemble the starting materials. In the last 20 years Scheffer and co-workers have made major contributions to the study of unimolecular photorearrangements in the solid state by using structure-reactivity correlations. They have systematically investigated a series of closely related systems (cyclohexenones54, ene-diones55, 13,7-unsaturated ketones56, aryl ketones57 and 1,4dienes58) and have elucidated how the ground state conformations influence the excited state behaviour of these systems. Examples of the above systems are depicted in Figure 9.  Ph  Ph  (a) Ph  Ph Ph  Ph  solid state  (Ref. 55c)  ^  solution reaction  ^  100% 75%  ^  ^  hv  hv  crystal  solution OAc  00% 25%  0  17  (c)  hv_/,. Ito solution^0  (Ref. 56)  (d)  hy  +  OCH3 cis solid  ^  soution (C6 H6)  60%  ^  28%  ^  ^  ^  trans 34% 72% (Ref. 57d)  • de  CH302  +  0 +  CO2CH3  ,....„...^  crystal  ^  solution  85%  ^  small amounts  ^  -50%  ^  small amounts  ^  -50%  (Ref. 58E)  Figure 9: Photochemistry of a) ene-diones, b) cyclohexenones, c) 13, 7-unsaturated ketones, d) arylketones and e) 1,4-dienes.  1.4 Type II Reaction. The Type II reaction59,60,76,115 is the most commonly known and well studied example of a photochemically-induced intramolecular hydrogen abstraction reaction. As  18  shown in Figure 10, upon irradiation the y-hydrogen atom of a ketone is abstracted by the (n,it*) excited carbonyl oxygen atom through a cyclic six membered transition state to produce a 1,4-biradical intermediate. These biradical intermediates can cyclize to cyclobutanol derivatives (Yang reaction)61 or cleave (Norrish type 11)62 to an alkene and an enol (isolated as the ketone, but detectable spectroscopica1ly)63a or undergo reverse hydrogen transfer to regenerate the starting materia1.63b R' cyclization  biradical intermediate  ^ R' + Rh,^‘‘‘Il HO  cis  1---‘ HO R. trans  CH2  OH enol  11 CH2 alkene  Figure 10. Type II reaction.  Photochemically-induced y-hydrogen abstraction reactions in aliphatic carbonyl compounds are known to involve both singlet and triplet (n-e) excited states,116 where the non-bonding orbital of the electron deficient oxygen atom in the excited carbonyl group abstracts the y-hydrogen atom. Photochemically-induced y-hydrogen abstraction reactions have been widely investigated in the solid state in order to study the favoured57 transition state geometries for hydrogen abstraction by the carbonyl chromophore. In solid state photochemical reactions, owing to the constraining crystalline environment, the possible transition state geometries for abstraction are limited to conformations which closely resemble the ground state reactant. Thus the X-ray crystal structure of the reactant obtained prior to the reaction can provide valuable information  19  regarding the transition state geometric requirements for hydrogen abstraction. Using this information, structure-reactivity correlations in solid state photochemical reactions can be investigated. Furthermore, since the probable conformations of the 1,4-biradical intermediate of the reactions can be predicted from the X-ray crystal structures, we also learn the partitioning behaviour of the biradical intermediates to cyclization, cleavage and dispropotionation products.  1.4.1. Geometrical Relationship for Hydrogen Abstraction. The preferred geometry for a successful 7-hydrogen abstraction in a conformationally mobile system was suggested by Wagner as a strain free chair-like six membered abstraction geometry.64 Houk et (11,65,66 from their force-field models and ab initio calculations, showed that the preference for regioselective hydrogen transfer  through a six-membered ring transition state (7-hydrogen abstraction) over a sevenmembered (8-hydrogen abstraction) is the result of the favorable entropy of activation, rather than an unstrained chair-like transition state geometry. The geometrical relationship between the abstracting oxygen atom and the hydrogen atom being abstracted can be described by four parameters (d, 0, A and co), as shown in Figure 11. d = the distance in angstroms between hydrogen and oxygen, 0...H.  co = the angle formed between the 0...H vector and its projection on the mean plane of the carbonyl group. = The C=0.-H angle.  A = The C-1-1-0 angle.  20  e)*"..mil\ A  C^  ■•••••  Figure 11: Geometrical parameters for the abstraction of a hydrogen atom by the carbonyl oxygen atom. To find the optimum geometrical requirements for hydrogen abstraction, both theoretical and experimental aspects have been investigated. On theoretical grounds, an interatomic distance of 1.8  A has been proposed as the upper limit67 for the Type  II  y-hydrogen abstraction. Djerassi and co-workers have also suggested an upper limit of 1.8 A for d, based on their work on the McLafferty rearrangement of steroidal ketones.68 Similarly, a distance of 2.1  A has been estimated for the well known Barton reaction for a  series of conformationally rigid steroid systems.69 In the above two cases, the estimate is based on measurements obtained from molecular models. A few years ago Scheffer et al. studied a variety of ketones that undergo intramolecular hydrogen transfer in the crystalline state, and obtained valuable information regarding the geometrical parameters. In analyzing their reactivities, they considered the ground state parameters of the reactant obtained from the X-ray structure. The solid state hydrogen abstraction distance was found to be much greater than had previously been considered feasible. In an initial investigation of a series of ene-dione55 compounds, it was found that the successful intramolecular hydrogen abstraction by oxygen occurred with 0-•H contact distances less than or equal to 2.6  A.  Based on these results he  suggested that hydrogen abstractions can occur over a distance that is less than or equal to the sum of the van der Waals radii of the atoms involved70 (van der Waals radius of oxygen = 1.52 A; hydrogen = 1.20 A).7133 Furthermore, his studies in the a-cycloalkylp-substituted acetophenone57f and a-adamantyl-p-substituted acetophenone57a-e series  21  (Figure 9d), which undergo Norrish Type II y-hydrogen abstractions, indicate that the 2.7 A is not an absolute upper limit for hydrogen atom abstraction. Five out of the seventeen compounds studied had an abstraction distance greater than 2.7 A. Scheffer also pointed out that whenever the ground state geometrical relationship between the abstracting and the abstracted atoms is close to the ideal values (vide infra), minimal molecular motion is required for reaction. The oxygen atom of the carbonyl group cannot abstract hydrogen with ease in all directions, because the non-bonding orbital involved in the abstraction lies in the nodal plane of the it-bond. Therefore, a successful hydrogen abstraction is generally observed when the hydrogen atom approaches the carbonyl group in a way that allows it to have substantial overlap with the n-orbital of the oxygen atom. Almost two decades ago, Turro suggested that, in the case of a hydrogen abstraction by a carbonyl group, the hydrogen atom should lie along the axis of the carbonyl non-bonding orbital (co = 00).72 Since then several examples have been reported in which efficient intramolecular hydrogen transfer has taken place despite the developing 0—H bond having a fairly large co angle.64 The abstraction is, however, expected to be more facile when co = 00; several examples report no abstraction when co = 90°,74,77 but there are some exceptions. Efficient intermolecular hydrogen abstractions have been observed in the photochemistry of the solid complexes of acetophenone and deoxycholic acid, even when co = 90°.75 It was suggested that the molecular motions within the crystal lattice permitted low values of co. Wagner76 proposed a cos2c0 dependency of the abstraction rate on the dihedral angle associated with the developing 0—H bond and the carbonyl nodal plane. Recently, the investigation of the preferred angular relationships for hydrogen abstraction (in the n,n* excited states) has been approached using semiempirical and ab initio computations, which have provided a firm theoretical foundation for this problem.  Dorigo and Houk undertook ab initio studies on various systems in order to investigate the geometric requirements for the intramolecular 'y-hydrogen abstraction of triplet  22  butanal.66 Their investigation revealed that large deviations from 0 =180° or co = 00 values increase the AH# (enthalpy) dramatically. A linear arrangement of C—H-0 (0 = 1800) is thought to be preferred for a favorable abstraction, but several examples indicate that the angle 0 can vary significantly from 180°.64,76,70,117 With regard to the C=0-..H angle A, since the singly occupied nonbonding orbitals are involved in the hydrogen abstraction, the best value is believed to vary from 90° to 1200,78 depending on the orbital hybridization. The theoretically ideal values79 of these geometrical parameters for an efficient hydrogen abstraction by the carbonyl oxygen are summarized in Figure 12. d (A)^co (0)^A (0)^0 (0) <2.7^0^90-120^180 Figure 12: Theoretically ideal values of the geometrical parameters for hydrogen abstraction by the carbonyl oxygen. 1.5. The Norrish Type I Reaction. Photochemically-induced reactions involving a-bond cleavage of saturated ketones are termed Norrish type I reactions. Although photochemical a-cleavage reactions of cyclic ketones were first observed by Ciamician and Silber80 at the beginning of this century, it was thirty years before Norrish and co-workers initiated mechanistic studies81,82a of these reactions, and referred to them as type I. These reactions were  found to take place through both singlet and triplet excited states.82b The radical pair formed during the primary a-cleavage process can undergo one or more reactions, depending on the electronic, structural and stereochemical properties of each transition state of the subsequent reactions and the conditions under which the reaction is carried out. The most common reactions derived from the radical pair of a ketone are radical recombination, disproportionation and decarbonylation. In the case of saturated cyclic  ^ 23  ketones, the biradical intermediate formed from the ring cleavage generally undergoes one or more intramolecular reactions, as illustrated in Figure 13.  ^ro 1^  (C H2)n— C I-12  CH2 CHO  I (CF12)n-l—CH=C1-12 t^ reclosure / ^disproportionation 0  1^ (C H2)n—C H2  CH2 —C, 22_0. 1  —CO  p.  4612  I^•  (C142)n —OH2  (CH)— l-L2  cf==c)  biradical intermediate  I  ROH ^•^ --• (CI)—cH3  o  I (CI-)n —cH2  A  !  '0  ROH --0.  (c HAT— CH2 CH2 co, R I (CH2)n—CH3  '.''.C.* ()R 0 (C 112)n  I GH2  Figure 13: The a-cleavage processes (Nonish type I reaction) of cyclic ketones.  Much mechanistic work has been done on cycloalkanones, and it has been directed particularly at the effects of structure on the reactivity of n-n* singlet and triplet states. The results indicate that the triplet is more reactive than the singlet.83,84 However, in cyclic ketones, the rate constant for a-cleavage increases with the relief of ring strain and the presence of radical-stabilizing a-substituents.84 Photochemically-induced a-cleavage reactions of ketones are frequently observed in gas and liquid phases, but to our knowledge a solid state reaction of this type has not been reported to date.  24  1.6. Polymorphism and Solid-Solid Phase Transitions in Crystals By definition, polymorphism85 is the ability of a substance to crystallize in different crystal forms under various conditions. These crystal forms (also known as varieties or phases) correspond to different periodic arrangements of the constituent elements, such as atoms, molecules and ions. The conformation of a molecule is not necessarily constant from one polymorph to another, and different polymorphs may exhibit different molecular conformations. The existence of different conformers of the same molecule in different polymorphic modifications is termed "conformational polymorphism".86 Therefore, the molecular conformation observed in crystals is usually but not always necessarily the one with the lowest energy. An important factor that determines the existence of conformational polymorphism is concerned with the interplay between the crystal forces (intermolecular forces) and the molecular geometry (intramolecular forces). Polymorphs of a given compound differ in structure and physical properties, such as solubility, melting point, density, hardness, crystal shape, optical and electrical properties, as generally observed in the crystals of two different compounds. Solid-solid transformation between the phases or the polymorphs can be achieved by variations in temperature or pressure, and this phenomenon is known as a "phase transition" or "polymorphic transition"87. Polymorphs can be classified into enantiotropic and monotropic systems.88 During phase transitions an enantiotropic system may become a monotropic and vice versa. Enantiotropic crystals have transition temperatures lower than their melting points, whereas, monotropic crystals theoretically should have transition temperature above the melting point. However, the absence of a transition point below the melting point cannot absolutely confirm a monotropic crystal, since the transition point may be found below room temperature or may not be observed owing to slow transition. The reversibility of phase transitions depends on the stability of the crystal lattice. A metastable form of a crystal may either persist for years or undergo spontaneous transformation to the stable form. Mnyukh89 pointed out the necessity of crystal  25  imperfections, such as holes, dislocations or defects, to start phase transformations. Many of the transformations he studied seem to occur or be initiated at crystal defects, where the arrangement of the molecules in the lattice is interrupted. It is, therefore, difficult to initiate a phase transition in a single crystal without defects. He also suggested that when suitable defects of the right size and shape are absent, or when their concentration is low, a crystal modification can sometimes be kept indefinitely at a temperature well outside its normal range of stability. Phase transitions are generally accompanied by the deformation of one structure to another, involving the translational or rotational motions of molecules. In some cases, the basic crystal lattice geometry and symmetry are preserved, but the molecules may be related by simple rotations or conformational isomers. Some general examples, which represent different modes of transitions, are depicted in Figure 14. 6  a  Figure 14: Types of phase transitions.  a) Phase transitions involving simple rotations of the molecules, without disruption of the lattice geometry and symmetry.  26  b) The complex transition of molecules without a simple relationship between the lattices. c) The phase transition accompanying a conformational change, also known as "conformational polymorphism". d) The simple rotation of the molecules, changing the symmetry by increasing the lattice parameters (doubled).87  1.7. Conformational Analysis of Medium Sized Ring and Macrocyclic Compounds The ideas of conformational analysis are widely used in the interpretation of chemical transformations and reaction mechanisms in modern chemistry. The stable chair form of cyclohexane, with its distinguishable substituents at axial and equatorial positions, was first demonstrated by Hassel et a/90,91 in 1943. Since then a remarkable development in the conformational analysis of aliphatic six membered ring systems has taken place,92,93 although only during the last twenty years or so have the conformational arrangements of the larger cycloaLkane ring systems (containing ten or more members) and their simple derivatives been investigated. From their investigation of the properties of medium sized ring and macrocyclic hydrocarbons, alcohols and ketones, Ruzicka and Prelog94,9514 showed that the melting point versus ring size curve did not rise steadily, as with aliphatic hydrocarbons. They also observed that several other physical and chemical properties of the compounds depended on the ring size. The conformational details controlling the physical and chemical properties of the medium sized ring and macrocyclic compounds were not clear until Dunitz96 (1961) demonstrated the unique conformation of the ten membered ring skeleton of cyclodecane and a variety of its derivatives by X-ray diffraction (Figure 15). Dunitz concluded that the conformation of the ring skeleton common to these different ten membered cyclic  27  compounds must represent a potential energy minimum. Thus the conformational properties of the cyclic compounds seem to depend more on ring size than on the substituents or simple heteroatoms present in the ring.  Figure 15: The solid state conformation of cis-1,6-diaminocyclodecane dihydrochloride.  In 1918 Mohr first described the construction of "strain free" molecules by using diamond lattice templates. 97 The three dimensional carbon framework of the diamond molecule was considered a unique way of extending the ideal tetrahedral carbon bond lengths, bond angles and dihedral angles infinitely in a chair-like manner. Thus any conformation that is superimposable on the diamond lattice should retain a form of chair arrangement with a minimum of angle and torsional strain. Dale 98 in 1963 pointed out that the solid state conformations of the ten membered ring derivatives closely follow the diamond lattice, as does the chair conformation of cyclohexane (Figure 16). This was a breakthrough in the conformational analysis of medium and large rings.  Figure 16: The diamond lattice conformation of the ten membered carbon frame.  28  Dale qualitatively investigated a series of saturated cycloalkanes, using space-filling molecular models, in order to classify all strain-free conformations with ring sizes varying from six to thirty.98,99 From his studies, Dale made two important conclusions: 1) Only even membered carbon rings can have skeletons without torsional strain, i.e., any cyclic systems having an odd number of carbon atoms cannot fit on the diamond lattice arrangement, and hence cannot have a strain-free conformation. 2) No ring size between cyclohexane and cyclotetradecane can have a strain-free diamond lattice conformation. He found severe transannular hydrogen interactions within the medium size rings, (less than fourteen membered) due to the short contacts between the hydrogen atoms pointing into the ring. According to his studies, a variety of molecular models of cyclododecane without angle strain can be constructed; however, he suggested that in order to avoid as much torsional strain and transannular interaction as possible the confirmation of the twelve membered ring has to accommodate a non-diamond lattice square conformation. Evidence to support his findings come from the non-diamond lattice square solid state conformation of cyclododecane100 (Figure 17) and its derivatives.101  Figure 17: The solid state conformation of cyclododecane.  29  In the case of the ten membered ring compound, cis-1,6-diaminocyclodecane dihydrochloride, its stable existence in the diamond lattice solid state conformation (Figure 15)96 is rather surprising; however, it clearly indicates a compromise between angle strain, torsional strain and transannular repulsion. Dale also recognized that the fourteen membered ring would be the first large ring to exist in a strain-free diamond lattice conformation (Figure 18). A typical feature of the majority of the strain-free conformations of the medium and large rings is that there are four "corners" in the conformations, each consisting of carbon atoms that are flanked by two gauche C—C bonds.  Figure 18: Strain-free diamond lattice carbon frame of the fourteen membered ring.  Dale noticed that the hydrogen atoms at the corner carbon atoms are directed away from the ring. From his qualitative studies on cyclic compounds carrying gem-dimethyl groups (C14-C24) 102 and ketal derivatives,103 he concluded that the two geminal substituents would occupy the corner positions in order to avoid severe transannular interactions. As shown in Figure 19, the two adjacent gauche bonds at the corner are flanked by bonds with anti-dihedral angles (+1800 -600 -60° +180°). The C—C bonds of the rest of the molecule are anti to one another. Dale's notation' 04 of a ring conformation consists of a series of numbers within brackets, each number representing the number of bonds on each side between two corner  30  atoms. The sequence of numbers in the square brackets starts with the smaller number of bonds between corner atoms, followed by the larger number. For example the diamond lattice rectangular conformation of cyclotetradecane is [3434].  5^6^7 *0^0^40—moss0 _60o ^1800  Figure 19: The corner positions (*) of cyclic compounds.  As experimentally demonstrated by Dunitz,96 the introduction of functional groups on the cycloalkane ring has little effect on the preferred conformations. The replacement of a methylene unit by a carbonyl group (sp2) or a heteroatom should reduce the transannular H/H interactions existing in cycloalkanes, and consequently lead to a stabilizing effect on the conformation. Generally, carbonyl substituents in medium and large rings are found a to the corner positions, but not at the comers.105, 118 Since the methylene group at the corner position is almost unhindered, the substitution of a corner methylene group by a carbonyl group does not relieve any non-bonded interactions. Furthermore, the carbonyl oxygen atom at a corner position is eclipsed with two hydrogen atoms on the a and a' carbons, and this geometry is known105 to be higher in energy than the one where the oxygen atom is eclipsed with a methylene group and a hydrogen atom, as observed when the carbonyl is a to the corner atom. In large rings, many of the strain-free skeletons that can be constructed using models are unstable owing to their having either more gauche bonds than the theoretical minimum or structures that are too open and do not fill the space efficiently. Dale98a  31  suggested that in large rings, rectangular conformations will be reasonably stable, as opposed to square ones, because the square has a large hole in the centre and lacks the van der Waals attraction for the hydrocarbon chain on the opposite side. He considered that the stable rectangular conformations would contain two long parallel chains linked by a pair of short chains containing two carbon atoms (C2 bridges) which would fill the space efficiently with substantial van der Waals attraction between them. X-ray structures106 of several large membered rings have been reported to have conformations with two long parallel chains joined by four carbon segments. Dale concluded that only the 14, 18, 22, 26 etc. membered rings or the (CH2 )4n+2 series can have this ideal rectangular, diamond lattice conformation [3x3x]. The (CH2)4n series can also have rectangular diamond lattice conformations, which would contain two long chains linked by C3 bridges [4x4x]. According to Dale, however, the [4x4x] conformations of the (CH2)n series would prefer to collapse to a non-diamond lattice [3x3x] rectangular conformation, in order to adopt better internal van der Waals contacts. Such compact but strained conformations are thought to be favoured for the (CH2)n series. A pictorial representation of the above effect on sixteen membered ring is depicted in Figure 20.  [4444]  [3535] Figure 20: Transformation of the cyclohexadecane molecular conformation from square [4444] (diamond lattice) to a rectangular [3535] (joining the dotted lines would lead to a non-diamond lattice arrangement).  32  1.8. Objectives of Current Research The present study is an extension of previous research work from our laboratory on type II reactions in the solid state, with the aim of improving our knowledge of the empirical guidelines on hydrogen abstractability by the carbonyl oxygen. Our interest in macrocyclic diketones was triggered by the reported X-ray crystal structure of cyclooctadecane-1,10-dione (6), an eighteen-membered ring diametric diketone,107 in which four of the eight y-hydrogen atoms of the solid state conformation were situated in what appeared to be an ideal geometry for abstraction by the carbonyl oxygens. To begin the investigation, a series of nine diametric cyclic diketones, with ring sizes ranging from 10 to 26, and having the general structure depicted in Figure 21, were synthesized108-112 and their X-ray crystal structures obtained. All diametric diketones investigated were known compounds.113,114  n = 4 (6), n =5 (7), n =6 (8 ), n =7 (9), n =8 (10), n =9 (11), n= 10 (12), n= 11 (13), n= 12 (14). Figure 21: Diametric diketones.  Examining the X- ray crystal structures of these diketones revealed that, in each case at least one y-hydrogen atom had an intramolecular cyclic C=0---H contact of less than 3 A. This has been reported in the studies on a-cycloallcyl acetophenones and  33  a-adamantylacetophenones79 as being the highest tolerable distance for a feasible hydrogen abstraction in the solid state. Furthermore, several other characteristic properties of these cyclic diketones, such as high melting points, good quality crystals, solid-solid phase transitions above room temperature (in few examples) and conformational polymorphism (twenty-six membered ring diketone) prompted us to investigate their photochemical behaviour with multiple objectives. Our major goal was to investigate the type II reactions of these diketones, and to perform structure-reactivity correlation studies by examining the X-ray data, in order to gain more knowledge of the preferred spatial relationship between the carbonyl oxygen and the abstracted 7-hydrogen atom for a successful abstraction. Knowing the limits of such geometrical parameters, which are directly linked with the favoured geometry necessary for the initial y-hydrogen abstraction, would serve as a useful guideline to predict the success of the type II reaction. Our second major goal was to study the partitioning of the biradical intermediates into products. A ketone molecule in a collision free medium can attain a variety of conformations owing to C—C bond rotations. Since the most preferred orientation for hydrogen abstraction is the nodal plane of the carbonyl group, 7-hydrogen abstraction by the excited ketone would afford biradical cisoid (iBR) as the initial biradical intermediate; however, iBR will soon equilibrate with other possible biradical conformations such as pre cis (BRi), pre trans (BR2) and transoid or anti (t-BR) as shown in Figure 22.59a, 60, -  -  115,116  It is generally thought that efficient cleavage of the biradical intermediates requires a parallel arrangement of the central bond with the singly occupied p-orbitals.115 Therefore the transoid or anti t-BR biradical intermediate would exclusively give elimination (cleavage) product, whereas the cisoid and gauche biradicals can lead to both  34  cyclization and cleavage products. However, the gauche or cisoid conformations are generally known to undergo mainly cyclization rather than cleavage.115  Figure 22: a) Cisoid, b) gauche (pre-cis), c) gauche (pre-trans) and d) transoid  conformational arrangements of the biradical intermediates.  or anti  Based on these considerations, we were interested in studying the effect of the crystal lattice on product selectivity (cleavage:cyclization and cis:trans product ratios), which would be influenced by the rigid conformations of the 1,4-biradical intermediates enforced by the solid state. The restricted mobility of the 1,4-biradical intermediate in the constraining solid state medium may lead to the formation of topochemically controlled products. For comparison, product ratios in isotropic media were also explored, by solution state irradiations carried out in hexanes.  35  Our third major goal was the investigation of the solid state photochemistry of diketones in which the basic structure is slightly modified. We synthesized a nondiametric sixteen-membered ring diketone, cyclohexadecane-1,8-dione (15), in which the two carbonyls are separated by six methylene groups on one side and eight on the other, so that the abstraction of two non-equivalent 7-hydrogen atoms present in the unsymmetric conformation would lead to regioisomeric pairs of cis, trans cyclobutanol derivatives and cleavage products. Through this, the regioselectivity of the solid state hydrogen abstraction can be investigated. A seventeen membered ring diketone, cycloheptadecane-1,9-dione (16) was also investigated, using a similar strategy. Another structural modification was achieved by substituting a-hydrogen atoms with methyl groups, with the idea of promoting Norrish type I processes in macrocyclic diketones. As explained in the Introduction section, the initial step of the type I reaction involves a-bond cleavage, and so, by introducing alkyl groups at the a-carbon atom, the carbon-carbon a bond can be weakened, since a-cleavage leads to a more stable secondary biradical intermediate. The photoreactions of diketones were also investigated in zeolites (a collaboration with Dr. V. Ramamurthy at Du Pont), where, interestingly, both Norrish type I and type II reactions were observed. Our fourth major objective was the investigation of the solid-solid phase transitions observed in four diketones, primarily by photochemistry (at different temperatures), together with solid state NMR (13C NMR, deuterium wide line NMR), FTIR and DSC techniques. MM2 calculations (energy minimization and steric energy calculations), of several diketones were also performed in order to determine the low energy conformations. Our main interest was directed towards finding any conformations having lower energy than the solid state conformation.  36  CHAPTER II 2.0. Synthesis of Starting Materials. The diametric diketones cyclodecane-1,6-dione (6) and cyclododecane-1,7-dione (7) were synthesized via ozonolysis (Figure 23) of the corresponding bicyclic  olefins. 108,109,110 The precursor (17) for the synthesis of diketone (6) was purchased from Aldrich Chemical Co., while for the preparation of diketone (7), the corresponding bicyclic olefin (20) was synthesized from cyclododecane-1,5,9-triene (18), as reported by Dale et a!.110 (1)03/Me01 1/ CH2Cl2 -  (CI42)2^(C H2)2  (2) Me2S  (17)  ^  (6)  Na/Al203 heptane  (18)  ^  (19)  heptaneI H2/Na/Al203  (CH2)3^(cI-I2)3  (1) 03/Me0H/ CH2C12 4^  (2) Me2S  (7)  ^  (20)  Figure 23: Synthesis of the ten and twelve membered ring diketones via ozonolysis of bicyclic olefins.  37  Preparations of all other diketones were achieved by bimolecular cyclization of their corresponding diacyl derivatives (Figure 24). The 20 membered ring diketone (11) was synthesized by the Dieckmann condensation reaction112 using diethyl dodecanedioate (21) in the presence of potassium tert-butoxide. The diametric 14 membered (8), 16  membered (9), 18-membered (10), 22-membered (12), 24-membered (13) and 26membered (14) ring diketones were synthesized using Blomquist's high dilution technique,111 which involves the bimolecular cyclization of the corresponding diacid dichlorides (22) in the presence of triethylamine.  / CO2C2H5  (1)tBuO -K4- I (C142)11  \CO2C2H5  (2)H30+  xylene ^*^(C142)9^( HA  (11)  (21)^  /COO (1) NEt3 / C6H6 (CF12)n+2  ^■ (CI-12)n^(CH2)n + (C  \coa^(2) KOH/Me0H (3) H20  o /.7 NV\ other )n-1^I^I^  (22)  n =4 (8), n =5 (9), n =6 (10) ,^n = 3 (23) n =8 (12) , n =9 (13) , n= 10 (14)  ^  Figure 24: Synthesis of diketones from diacyl derivatives.  n = 4 (24)  (C1 12)n 1±products -  -  38  All diacid dichlorides were made from their corresponding &acids, which were chlorinated with thionyl chloride.111 The yields of the diketones are summarized in Table 1. Although the yields are quite low because of the polymerization of the acid chlorides, they are comparable to the reported values.111,112 The purification of the diketones was successfully carried out by column chromatography and recrystallization, instead of the reported distillation or vacuum sublimation methods.111 Table I: Percentage yields of diketones and y-pyrones .  diketone (ring size) 10 12  diketone (%) -  pyrone (%) -  56  14  09  21  16  28  <1  18  25  20  14  22  41  24  19  26  29  -  -  -  -  -  A possible mechanism for the formation of diketones from diacid dichlorides is illustrated in Figure 25. In the presence of base, diacid dichlorides undergo dehydrohalogenation to yield bifunctional ketenes (25).119 The bimolecular cyclization of the diketenes, followed by base hydrolysis and decarboxylation, leads to the final product. During this process tertiary amine acts as a dehydrohalogenating and condensing agent. 111  39  / C = C=0  Et3N CsH6  Et3N C6H6  (CI-)n  (1)  (CF12)n+1  O-==0  (2)  (25)  (Ref. 119)  0 ./(CH2)r4TCH==.0  ^ Nc/  ^• ^ C\ 'i0 4\ (CH2)n+1( OK  (4)I  °  (2), (3) -5\ (CH2)n —cH==0  (Ref. 111, 120)^0  KOH/Me0H (c H2)r4.1  0  (CF12)n+1  %\  HO  (C)n1  0  (CF-2)rio  OH (6)  -  (C1 12)„  0(CH2)n±i  - \ (CH2)n+i  -CO2 0  (7)  (C142)n+i 0  H20 I (8) A <(cHA+ 1 \ _ c*0  (CH2)n+ ,.  (4), (5), (6), (7), (8) ^■  0  Figure 25: A possible mechanism for the formation of diketones.  C  40  As indicated in Table I, the percentage yields of the diketones drop dramatically at the 14 membered ring diketone (8) (n=6), and approach zero in smaller rings. These results may be due to an unexpected side reaction which leads to the formation of a byproduct, tricyclic y-pyrone. During the synthesis of the fourteen membered ring diketone (8), large quantities of the by-product tricyclic y-pyrone (24) were obtained along with the  diketone. When the twelve membered ring diketone (7) synthesis was attempted by this method, the corresponding tricyclic y-pyrone (23) was obtained exclusively. Synthesis of the ten membered ring diketone from adipoylchloiide gave neither the diketone nor the corresponding y-pyrone; however, during the workup procedure, large quantities of polymeric material were formed. These polymeric materials were found to be soluble only in chloroform. A possible mechanism for the formation of y-pyrone is depicted in Figure 26. The competition between the intramolecular and intermolecular reactions of the diketene intermediate (25) may be mainly controlled by the following two factors: (1) the length of the diketene molecule (entropy factor). (2) the ring strain (enthalpy factor) When^the two ketene functionalities within a molecule are far apart, and therefore the intramolecular cyclization is less favoured. In these cases, the intramolecular cyclization involves a large unfavorable entropy loss along with severe transannular interactions, which are prominent in medium size rings. When n = 4, 5 and 6, the transition states associated with intramolecular cyclization will have 5, 6 and 7 membered rings respectively. In such cases the entropy loss during intramolecular cyclization is much lower than during their intermolecular reaction (ie. bringing two individual molecules together);123 furthermore, the small rings are relatively unstrained compared to the medium size rings, therefore intramolecular cyclization prevails.  ^  41  ^/  E t3N / C6H 6 A  ^\  C=C=0  ^Et3N / C 6H6 (CH2)n^______•,. ^A  c-=c=o  (Ref. 120, 121)  i  .7,.......,,,,, (CF)-2^  Et3N / C6H6 (C1-2)n-2^  (CH2)n-1  0  C (C1-12)n-1  OH-  C%0  IKOH/Me0H  (Ref. 122)  it  (CI-12)n-1  -CO2  ^■  0  /N/4._____),\V\  r4  ) -1^rik^(C  )n-1  1  ?  .  H20 A  a) protonation b) loss of water  Figure 26:  A possible mechanism for the formation of y-pyrones.  Based on these arguments, the formation of symmetric tricyclic y-pyrones (when n = 5, 6) could be rationalized as an intramolecular diketene cyclization, followed by the intermolecular dimerization of the cyclic a-keto ketenes. This eventually leads to the formation of tricyclic y-pyrone in the presence of base.  42  It is, however, both interesting to consider and necessary to explain the absence of tricyclic rpyrone during the synthesis of the ten membered ring diketone (6). This absence can be explained by reference to the work of J. C. Sauer,121 who confirmed the formation of the a-keto ketene (26) from adipoyl chloride (n = 4) by trapping it with ethanol at room temperature (Figure 27). When Sauer attempted to isolate this five membered cyclic a-keto-ketene (the postulated intermediate for the formation of the corresponding rpyrone) from the final reaction mixture by distillation, the attempt failed, and instead large amounts of polymeric compounds were isolated. Et3N / CEH6 /c pi-12)2 R.T. (25)  c2HSOH  ck  1= c,Hso  0  (26)  —^II o C II^II  polymer  n  Figure 27: Fate of adipoylchloride at different experimental conditions.  It has been suggested that the thermally unstable five membered keto-ketene opens up and undergoes polymerization at high temperatures. Since this five membered a-ketoketene was not our target compound, reaction was never attempted at room temperature to reconfirm the formation of the cc-keto-ketene. However, Sauer's observations clearly explain our experimental results, since the high temperature used during the reaction prevents the formation of the corresponding rpyrone. During the synthesis of the sixteen  43  membered ring diketone (n = 7), the final reaction mixture on GLC showed a trace of a peak corresponding to its pyrone (< 1%) (recognized only on its retention time relative to that of diketone), but the compound could not be successfully isolated or characterized. Two other synthetic routes for the preparation of symmetrical tricyclic y-pyrones are known in the literature. 1) Bimolecular condensation of eneamines (27) in the presence of phosgene.124  (1) COCl2 ^■ (2) HI7 H20  n=3 or 4  ^  yield 42%  (27)  2)^Condensation of eneamines (27) with P-ketoesters (28) at elevated temperatures. 125,126  0  II  z.,C\ (CI-)n  OEt  \\,--•o  (27) (28)  n = 3 or 4  ^  yield 70%  The spectral data for y-pyrone 23 correlate well with the reported126 values. To our knowledge, the synthesis of y-pyrone 24 (obtained as a side product during the preparation of the fourteen membered ring diketone) has not been reported in the literature.  44  The non-diametric diketones cyclohexadecane-1,8-dione (15) and cyclohepta decane-1,9-dione (16) were synthesized by Blomquist's method, using two different diacid dichlorides as starting materials (Figure 28). The desired products were separated from the by-products (diametric diketones A and B) by column chromatography.  /COG!^/ COCI  (1) NEt / (CF12)n+2^+^(C E42)n, +2 \ COCI^\ COCI  co6 ^.^  (CF,^(CH2),I,  (2) KOH / Me0H (3) 1-120  15 (n = 6, n' = 8) 16 (n = 7, n' = 8)  +^(01-(2)n^(CI-12)„^+^(cF)n,^(Cl-).  A  ^  B  Figure 28: Synthesis of non-diametric diketones.  Synthesis of the keto-alcohols 9-hydroxycyclohexadecanone (29) and 10-hydroxy cyclooctadecanone (30) was achieved by the partial reduction of diketones 9 and 10 with sodium borohydride.127 The tetramethylated (all cis) sixteen membered ring diketone, 2R*, 8S*, 10R*, 16S*-tetramethylcyclohexadecane-1,9-clione (31) and the twenty four membered ring  45  diketone 2R*, 12S*, 14R*, 24S*-tetramethylcyclotetracosane-1,13-dione (32) were prepared by the alkylation procedure reported for cyclohexanone,128 which treats a mixture of diketone and methyl iodide (— 1:4 molar ratio) with potassium hydride (Figure 29). The purification of the tetramethylated diketones was achieved by recrystallization and high performance liquid chromatography (HPLC). The stereochemistry of the diketones was determined by X-ray crystallography.  ill H  CH3I /KI (CF-{2)n^(CI-)n  DME  n = 5, 9.  (CI-12)n^(CH) +^  n = 5 (31) n = 9 (32)  Figure 29: Methylation of diketones.  other  products  46  CHAPTER ILL 3.0. Photochemistry of Medium and Macrocyclic Diketones that Undergo Type II Reactions. The photochemistry of cycloalkanones with ring sizes ten membered and above is well known for hydrogen abstraction reactions by carbonyl oxygen atoms.129-134. Upon irradiation in cyclohexane, these cyclic monoketones 34-39 are reported to undergo mainly intramolecular y-hydrogen abstraction to afford varying amounts of cyclization 40c and 40t (Yang reaction)61 and cleavage 41 (Norrish type 11)62 products together with small amounts of the photoreduction product cyclic alcohol 42 depending on the ring size (Figure 30).130 Photolysis of cyclodecanone (33) in cyclohexane, however, exhibits a unique e-hydrogen abstraction to afford 9-decalol exclusively.  135,136  The product  percentages of the cycloalkanone photochemistry in cyclohexane, reported by K. H. Schulte-Elte et al 130 are shown in Table II. HO IIH  hv C6H12  (CI-12)n+1 (CI-12)n-2^(CI-12)n-2  n = 6 (34)  (40c)  (41)  (40t)  n = 7 (35) n=8  (36)  n = 9 (37) n = 10 (38)  (CI-)n  n = 11 (39) .  (42)  Figure 30: Photolysis of monoketones in cyclohexane.  47  Table.!!: Product percentages from the irradiation of monoketones in cyclohexane.  monoketone (ring size) 11 12 13 14 15 16  cis-cyclotrans cyclo cleavage (%) butanol (%) butanol (%) 14 40 08 11 64 08 23 45 18 39 12 30 17 11 52 13 09 58 -  -  mono-cyclic Unknown alcohol (%) products (%) 12 26 10 07 05 09 20 <1 14 <1 <1 20  Two notable observations can be seen in the above product distributions from monoketones: (1) The fraction of cleavage products increases with ring size. (2) A stereoselective cyclization is observed in smaller rings. As described by Wagner,115 the above results clearly indicate that a competition between the cyclization and cleavage processes accompanies a change in ring size. The preference for cleavage product formation and the loss of stereoselectivity in the cyclobutanol formation in the larger membered rings suggest that the 1,4-biradical intermediate conformations can easily undergo conformational rearrangements in the isotropic medium to achieve the preferred orbital overlaps to afford cleavage product 41 as well as both cyclobutanol stereoisomers. In smaller rings, however, such conformational changes in the biradical intermediates may be restricted due to intramolecular structural considerations. The above variation in the product distributions seems to reflect the ability of the ring to adopt various conformations which consequently control the partitioning of the biradical intermediates into products. In the solid state, because the molecules are restricted to a single conformation and their structure can be obtained by X-ray crystallography, the effect of ring size on the  48  preferred solid state conformation can be determined along with the favorable hydrogen abstraction geometries and the effect of conformation on the partitioning of the biradical intermediate into products. The carbonyl excitations of aliphatic ketones are generally known to associate with an elongation of the C=0 bond and pyramidalization at the carbonyl carbon,231 and therefore the stereoelectronic dispositions of the y-hydrogen atoms with respect to the carbonyl oxygens in the excited molecules will be altered. However, the structure-reactivity correlation studies conducted in our laboratory on a wide range of compounds indicate79 that the ground state geometrical parameters obtained from the X-ray structures provide valuable information regarding the reactivity of the molecules. Such analysis can not be successfully achieved with medium or large ring monoketones, as most of them are either liquids or low melting solids at room temperature. Therefore, the diketones, having relatively high melting points and stable solid state conformations, are recognized as ideal candidates for the above structurereactivity correlation studies. As generally observed in aliphatic carbonyl compounds, all medium and macrocyclic diketones investigated in our laboratory showed a broad, low-intensity UV absorption maximum attributable to the n—nr* transition in the range of 270-290 nm, but mostly close to 280 nm. A UV-VIS absorption spectrum of 10-2 M diketone 9 (e = 41) in cyclohexane, typical for all other diketones in this study, is shown in Figure 31. 088 A  THRESHOLD^8 810  WL^OPD^WL^OFT^WI^DPO^WI.^OFD 2 6 2^8.4114 1 4 3^1.0263 E=41  38818 —Wave length  Figure 31: A UV-VIS spectrum of 10-2 M diketone 9 in cyclohexane.  4880  49  3.1. Photochemistry of Diametric Diketones. Irradiations of all diametric diketones were carried out in the crystalline phase and as 104 M solutions in cyclohexane (for details see Experimental Section). All nonalkylated diketones underwent a smooth type II photoreaction, except the ten membered diketone, cyclodecane-1,6-dione (6), which remained photostable in both media. The Xray crystal structures of both alkylated and non-alkylated diketones reveal that at least one of the y-hydrogen atoms in each compound is located in a position favorable for abstraction by the carbonyl oxygen. Interestingly, however, the alkylated diketones afforded only Norrish type I photoproducts and will be discussed later in chapter V. As outlined in Figure 32, the irradiation of the diketones to low conversions (10-15%) led to three major photoproduct types; a cis-cyclobutanol derivative, a trans-cyclobutanol derivative, and the cleavage product ene-dione. Low conversions were maintained for all diketones in order to avoid any secondary photoreactions from the reactive carbonyls in the photoproducts and also to avoid any serious damage to the lattice of the original crystals. In accordance with custom138, cis-cyclobutanol here refers to the isomer in which the hydroxyl and the hydrogen atom at the ring junction are cis to each other, and trans-cyclobutanol refers to the isomer where the hydroxyl and the hydrogen atom at the ring junction are trans to one another. All photoreactions were monitored and analyzed by gas-liquid chromatography (GLC) using a DB-17 capillary column. As observed by several others138,139, the ciscyclobutanol derivatives have longer retention times than their corresponding diastereoisomeric trans-cyclobutanol derivatives; however, cleavage product ene-diones have the shortest retention time. The major photoproducts formed in all cases were isolated by column chromatography and fully characterized, except the cis-cyclobutanol derivative 10c from the eighteen membered diketone 10, which could never be separated from the reaction mixture, and could only be identified by its relative GLC retention time.  0^  HO^  HO  HO  o hv^ (CH2)n (CH2)n --• (CH2)n  (CH2)n.2  -■.-.11110.  (CH2)n (CH2)n.2  /\ A  +^(CH2)n  (CH2)n+2^(CH2)n  diketone  biradical intermediate  cis  trans  n=37  7c  7t  7e  n=48  8c  8t  8e  n=59  9c  9t  9e  n = 6 10  10c  10t  10e  n = 7 11  11c  11t  11e  n = 8 12  12c  12t  12e  n = 9 13  13c  13t  13e  n=10 14  14c  14t  14e  Figure 32: Photolysis of diketones in cyclohexane.  ene-dione  Llt  c)  51  In column chromatography, the cleavage product ene-dione was eluted first, followed by the trans-cyclobutanol derivative and finally the cis-cyclobutanol derivative. The unreacted starting materials invariably had the longest retention times on the GLC, but were eluted first in column chromatography. The above elution orders in both the GLC and the column chromatography are general for all diketones investigated.  3.1.1. Structural Determination of the Photoproducts. The cyclobutanol ring junction stereochemistry assignment could not be determined by 1H nmr, as the methine protons are masked by the methylene protons, but was accomplished by obtaining the X-ray crystal structures of the photoproducts ciscyclobutanol derivative 9c and trans-cyclobutanol derivative lit. With the known configurations of 9c and lit, the 13C num and attached proton test (APT)140 correlations enabled the stereochemical assignments of the cyclobutanol photoproducts from other diketones to be made. The methine carbon at the ring junction of the cis-cyclobutanols, identified by the APT, appears at 48.6 - 50.1 ppm, whereas the analogous signal for the trans-cyclobutanol is shifted upfield to 40.4 - 43.7 ppm. These chemical shift differences have been observed in all cis/trans pairs, as indicated in Table III, and this trend appears to be general. The chemical shift difference between cis and trans-cyclobutanol derivatives (M) from the fourteen membered ring diketone 8 is close to 9 ppm; however, smaller values for M are observed in the larger membered rings, as shown in the Table III. Steric compressions between the 13C carbon atom and a 7-carbon atom at the gauche position are generally known to lower the chemical shift.141 As indicated in Figure 33, similar gauche interactions by the 7-methylene groups in the larger ring of the cyclobutanol  52  derivatives may account for the chemical shift differences observed for the bridgehead methine carbon atoms of the cis and trans-cyclobutanol derivatives. Methine carbon chemical shifts of the cyclobutanol derivatives.  DIKETONE (ring size) 12 14 16 18 20 22 24 26  cis  -  Chemical shift in benzene (ppm) trans  CYCLOBUTANOL  48.6 49.2 49.8 50.1 49.5 50.1 50.1  -  AS  CYCLOBUTANOL  40.4 41.6 42.2 43.1 43.4 43.5 43.7  8.8 8.2 7.0 6.1 6.6 6.4  A careful analysis of the molecular models of the cyclobutanol photoproducts reveals that in the minimum energy conformation of the trans-cyclobutanol derivatives, both 7-methylenes of the methine carbon atom tend to be located at the gauche or close to gauche positions, especially when the ring size is smaller.  With the increase in ring size, the 7-methylenes tend to be located farther away from the gauche positions. The gradual increase in the chemical shift of the methine carbon atom of the trans-cyclobutanol with increasing ring size is evidence of this observation. In the case of the cis-cyclobutanol derivatives, however, these interactions seem to be smaller than their corresponding diastereoisomers, irrespective of the ring size. Further evidence to support this argument comes from the X-ray crystal structures of the cis-cyclobutanol derivative 9c and the trans-cyclobutanol derivative lit. In the case of 9c, one 7-methylene is at the anti position to the methine carbon atom, and the other  53  y-methylene is quite far away from the gauche position (torsional angle close to 120°). In case of the trans-cyclobutanol lit, although the ring size is larger than for 9c, both ymethylenes are at gauche positions. Therefore, the larger y-gauche steric compressions in the trans-cyclobutanol derivatives, when compared to those of the corresponding ciscyclobutanol derivatives, correctly account for the lower chemical shift of the former. The gradual decrease in the AS ppm values and the increase in the chemical shift values that accompany the increase in ring size suggest that the larger the ring size, the smaller the effect.  OH  7C ......--'  CIS  p H2)n TRANS  Figure 33: Newman projection of cis and trans-cyclobutanol derivatives down the ring  junction. The mass spectra of both cis and trans-cyclobutanol derivatives show a fragment at m/e (M-28), corresponding to the loss of an ethylene group by a [2+2] cycloreversion of the cyclobutanol ring. The signal corresponding to this fragment, although quite weak compared to the base peak found in all cyclobutanol photoproducts, is evidence of cyclobutanol formation.  54  The ene-dione cleavage products were characterized mainly by the 1H nmr pattern of the vinyl end of the molecule. The spectrum in Figure 34b corresponds to the -CH2-CH=CH2 moiety of compound 13e, which is typical for all cleavage products isolated. The spectrum, decoupled at 2.01 ppm (Figure 34c), simplifies the peak at 8 = 5.75-5.85 ppm to a doublet of doublets. The coupling constants J = 10 Hz and 18 Hz denote the cis and trans coupling of Hx by HA and Hs respectively. 141 Therefore the peak at 8 = 5.75-5.85 ppm attributable to Hx, a doublet of doublet of triplets (J = 18,10 and 7 Hz), is the result of coupling by HB, HA and HD protons, as represented in Figure 34a. The peaks at 8 = 4.90 ppm and 5 = 4.95 ppm in the spectra, decoupled at 5.79 ppm (Figure 34d), are attributable to HA and Hs respectively. The observed small splitting of these peaks would be the result of geminal coupling between HA and HB and the longrange coupling by HD methylene protons. The two doublets in Figure 34b, attributable to HA (8 = 4.90 ppm, J = 10 Hz) and Hs (8 = 4.95 ppm, J=18 Hz), are therefore the result of  cis and trans coupling by Hx respectively. The peak at 8 = 2.0 ppm (JDx - JDE = 7 Hz),  corresponding to the methylene protons HD, is two overlapping triplets.  3.1.2. Photochemistry of non-diametric diketones. Photolysis of two non-diametric diketones, the sixteen membered ring diketone 15 and the seventeen membered ring diketone 16, in cyclohexane afforded a complex mixture of six type II photoproducts, as indicated in Figure 35. The unequal lengths of the methylene chains on either side of the two carbonyl functionalities lead to two nonequivalent 7 positions in the molecule, and thus the abstraction of the non-equivalent 7-hydrogen atoms affords regioisomeric pairs of cis-cyclobutanol derivatives, transcyclobutanol derivatives and ene-dione cleavage products. The cis-cyclobutanol regioisomers (15c1 and 15c2) from the sixteen membered ring diketone were isolated by column chromatography as solids.  55  HE^/— Hx^HE  Hx  HB  JBX = 18 Hz 1  331  211 3 311 2 1  (b) Z--/  JL_. (c)  NI/  1  JBX = 18 H..%  JAB = 10 Hz  (d)  6 . 0^5.5^5. 0^4.5^2.0 Figure 34:^(a) splitting pattern of the ally! moiety of ene-dione 13e. (b) 1H nmr spectrum of allyl moiety of ene-dione 13e. (c) spectrum decoupled at 8 2.01. (d) spectrum decoupled at 8 5.79.  56  n = 6, n' = 8 (15)  n = 7, n' = 8 (16) Ihv  o n = 6, n' = 8 n = 7, n' = 8  (15c2) (16c2)  (C142)n-2 (CNA'  n = 6, n' = 8 n = 7, n' = 8  (15c1) (16c1)  +^(CH)n (CH2)n._2  ^ n = 6, n' = 8 (15t2) ^n = 6, n' = 8 (15t1) n = 7, n' = 8 (16t1) n = 7, n' = 8 (16t2)  y  H^(CIA-1 (CH)Y+2 CH3  \/  y  4.^  0^01_12  C H2^0^0  n = 6, n' = 8 n = 7, n' = 8  (15e2) (16e2)  n = 6, n' = 8 n = 7, n' = 8  Figure 35: Photolysis of non diametric diketones. -  Y Y  (CFL2)n+2 PH2)n.-1^H  (15e1) (16e1)  57  Even though the stereochemistry at the ring junction was confirmed from the 13C nmr chemical shift values of the methine carbon atoms, the regiochemistry of the products was determined by X-ray crystallography in both cases. The other photoproducts from the sixteen membered diketone, as well as those from the seventeen membered ring diketone, were isolated and identified as a mixture of regioisomeric pairs. The product ratios obtained in the solid state photolysis of the seventeen membered ring diketone 16 were quite similar to those obtained in solution. However, the solid state reaction of diketone 15 afforded the cis-cyclobutanol derivative 15cl almost exclusively. With the exception of 15c1 and 15c2, all the other products gave rise to a single peak on GLC for each regeoisomeric pair, and their relative proportions could therefore not be determined.  3.1.3. Photochemistry of cyclic keto-alcohols. The sixteen membered ring keto-alcohol 9-hydroxycyclohexadecane-1-one (29) and the eighteen membered ring keto-alcohol 10-hydroxy cyclooctadecane- 1-one (30) were irradiated as crystalline solids and in cyclohexane. The photoproducts could not be separated and characterized, but were identified to a certain extent by comparing the GLC retention times with those of the authentic samples made from the photoproducts of the sixteen and eighteen membered ring diketones 9 and 10. As indicated in Figure 36, type II reactions of keto-alcohols can lead to five photoproducts. The irradiation of keto-alcohol 29 led to the development of five new peaks on GLC (DB 17, 190 °C); peak retention times and the corresponding amounts in parenthesis are given below. solid state reaction : RT's 4.5 (32%), 8.7 (22%), 8.8 (18%), 9.2 (18%) 9.5 (10%) min. solution reaction :^RT's 4.5 (40%), 8.7 (15%), 8.8 (20%), 9.2 (17%) 9.5 (8%) min.  58  (CI-)n^(CI-)n  n = 5 (29)  cJ^n = 6 (30)  H OH  OH^  HO  (CH2)n (01-12) _2  HO  n = 5 (29c)  n = 6 (30c)  n = 5 (29t) n = 6 (30t)  (CI42)n-1^(CF)2 CI-13  X  H OH  n = 5 (29e) n = 6 (30e)  Figure 36: Photolysis of keto-alcohols.  The reduction of 9c (the cis-cyclobutanol photoproduct from diketone 9) with sodium borohydride gave two peaks on GLC (DB 17, 190 °C), which were identical to  59  those observed at RT's 9.2 and 9.5 min (in terms of their retention times). These two photoproducts of 29 were assumed to be the diastereoisomers of cis-bicyclo[12.2.0]hexadecane-6,12-diol (29c). Likewise, the reduction of 9t (the trans cyclobutanol -  photoproduct from diketone 9) gave two peaks that were identical to those at RT's 8.7 and 8.8 min, and were assumed to be the diastereoisomers of trans-bicyclo[12.2.0] hexadecane-6,12-diol (29t). The above observations were also confirmed by co-injecting the photoreaction mixture with the reduction products on GLC. The relative position of the peak at RT 4.55 min with respect to the cyclobutanol photoproducts is comparable to the usual location of the cleavage products in other diketones, so this peak could belong to the cleavage product hexadecane-18-ene-2,10-diol (29e). In the case of the cyclic keto-alcohol 30, however, the irradiations in both media showed only three new peaks on GLC, at RTs 5.0, 9.1 and 10.2 min. Product ratios could not be calculated from the GLC traces, since the peak at 10.2 min overlaps with the unreacted starting material, but after the removal of the unreacted starting material from the final reaction mixture by column chromatography, the product ratios were determined and are given below in parentheses with their corresponding retention times. solid state reaction : RT's 5.0 (33%), 9.1 (59%), 10.2 (08%). solution reaction :^RT's 5.0 (56%), 9.1 (32%), 10.2 (12%). The reduction of 10t (the trans-cyclobutanol from diketone 10) with sodium borohydride gave a single peak at RT 9.1 min. The co-injection of the photoreaction mixture and the reduction product increases the intensity of the peak observed at RT 9.1 min. This clearly indicates that the peak at 9.1 min corresponds to the two diastereoisomers of the trans cyclobutanol photoproduct, trans-bicyclo[14.2.0]-  octadecane-1,10-diol (30t). The peaks at RTs 5.0 and 10.2 min could not be analyzed by co-injection experiments, but from the retention times of these two peaks relative to that of trans-cyclobutanol photoproduct, one can assume them to be the peaks belonging to the cleavage product ene-diol 10-hydroxy-hexadecane-10-ene-2-one (30e) and the  60  diastereomers of cis-cyclobutanol derivative cis-bicyclof14.2.0]octadecane-1,10-diol (30c), respectively. The X-ray crystal structures of both keto-alcohols 29 and 30 could not be obtained due to the poor crystal quality. The product ratios calculated (at 20°C) from GLC at various time intervals were plotted as a function of conversion for all ketones studied. The ratios obtained by extrapolating the plots to zero percent conversion and normalizing to 100% are compiled in Table IV. The following important observations are apparent from Table IV: 1) The proportion of type II cyclization (Yang reaction) is greater in the solid state than in solution, except for the fourteen and seventeen membered ring diketones, where the product ratios are independent of the medium and found to be quite similar in both media. 2) In most cases, the stereoselectivity of the cyclobutanol formation increases in the solid state compared to solution. Interestingly, with each two-carbon increment of ring size, the stereochemistry of the major solid state photoproduct changes, with the exception of the fourteen and twenty six membered (plates) (14P) membered rings. 3) In solution, both a greater degree of type II cleavage products and the preference for the formation of the presumably less strained trans-cyclobutanol over ciscyclobutano1162 are observed in rings larger than fourteen membered. 4) In the case of the twelve membered ring diketone, the cis-cyclobutanol derivative is stereoselectively formed as the major product in both media. A similar preference for cis-cyclobutanol in both media is observed from the fourteen membered ring diketone 8; however the stereoselectivity is significantly lowered. 5) The two crystal modifications of the twenty six membered ring diketone plates (14P) and needles (14N) stereoselectively afforded cis-cyclobutanol derivative 14c and  trans-cyclobutanol derivative 14t respectively.  61  Table IV: Product percentages at zero percent conversion obtained at 20°C.  (ring size)  SOLUTION STATE  SOLID STATE  DIKETONE  cis (%)  trans(%) cleavage(%)  cis(%)  trans(%) cleavage(%)  diketones 12  99  00  <01  84  00  16  14  58  29  13  65  25  10  16  89  10  01  22  35  43  18  03  84  13  17  42  41  20  90  04  06  10  23  67  22  04  91  05  10  34  56  24  98  01  01  15  27  58  26needle  09  91  00  14  33  53  26plates  97  03  00  14  33  53  non-diametric cyclic diketones  15d^15c2  15c1^15c2  16*  98^02  00  00  13^13  35  39  17  26  23  51  25  28  47  cyclic keto-alcohols 16  28  40  32  25  35  40  18  08  59  33  12  32  56  16* - Non-diametric sixteen membered ring diketone.  To aid in their solid state conformational analysis, X-ray crystal structures of all diketones, except that of the eighteen membered ring diketone cyclooctadecane-1,10dione (10), were determined. The crystal structure of the eighteen membered ring  62  diketone was reported earlier by Allinger et al.107 The solid state conformation of the ten membered ring diketone 6 (Figure 52) was also reported by G. Germain in a personal communication to J. Dunitz 143 almost 20 years ago. However, to our knowledge, the Xray structure details have never been published. In the case of the twelve membered ring diketone, the X-ray crystal structure could not be refined owing to the molecular disorder observed in the crystal lattice; however, the preliminary studies indicate a basic square [3333] conformation with both carbonyls at non-corner positions pointing towards the same side of the ring (the conformation shown in Figure 37a is generated by MM2). Low energy conformations of the ten, twelve and fourteen membered ring diketones (see Experimental Section for details) were generated by MM2. The force field calculations of diketones with a ring size larger than fourteen membered were impractical, owing to the length of cpu time required on a shared computer. However, for the 16 membered ring diketone, a limited search generated possible low energy conformations.  3.2. Diketones That Give Cis-cyclobutanols As The Major Photoproducts In The Solid State. 3.2.1. Solid State Conformation and Photochemistry. As indicated in Table IV, the twelve membered ring diketone 7, sixteen membered ring diketone 9, non-diametric sixteen membered ring diketone 15, twenty membered ring diketone 11, and twenty six membered ring diketone (plate crystals) 14P stereoselectively gave cis-cyclobutanol derivatives as major solid state products. The solid state conformations of these diketones are depicted in Figure 37. A common feature observed in the solid state conformations of this group of diketones is the [3x3x] carbon frame. Interestingly, with the exception of the twenty six membered ring plates 14P, the others belong to (CH2)4n series, and as suggested by Dale  63  for cycloalkanes,98a they prefer to crystallize in non-diamond lattice rectangular conformations rather than in diamond lattice [4x4x] conformations. However, Dale's qualitative analysis of the sixteen membered ring diketone predicts the [4444] square arrangement for the solid state conformation.103 On the other hand, plate crystals of diketone 14P, which belongs to the (CH 2)4n+2 series, crystallize with their carbon atoms arranged in a diamond lattice frame. It seems that the orientation of the two carbonyls in the solid state conformation is determined by the number of methylene groups separating them. When the number is odd the carbonyl groups are syn to one another, and when it is even they are anti. The carbonyls of all compounds belonging to this group are in the short segments of the rectangular conformations, a to the corner atoms, hence the y methylenes and the carbonyl are placed in gauche positions (ideal for type II reactions), rather than in anti positions.  K2  64  (c)  H  (f)  (e)  HI  C3  Hi  H7  HIO Nb  HI)  Ht4 • HIS HIS HMI 1419  Figure 37: Solid state conformations of (a) twelve membered (MM2), (b) sixteen membered, (c) sixteen membered (non-diametric), (d) twenty membered, (e) twenty-four membered and (0 twenty-six membered ring diketones. (the yhydrogen atoms with short 0....H contacts (d < 2.79 A) are indicated by dotted lines).  65  3.2.2 Hydrogen Abstraction Geometry and Biradical Geometry. In each conformation, the two pairs of symmetry-related y-hydrogens accessible to the carbonyl oxygens can be considered for analyzing the solid state photoreactivity of the diketones. One of these two pairs of y-hydrogen atoms has short 0...H contacts (d <2.76 A) with the carbonyl oxygen, and forms a boat-like six-membered abstraction geometry. The other pair makes a chair-like abstraction geometry; however, their 0...-H contact distances are generally close to 3A. In the case of the twelve membered ring diketone 7, only one pair (H6 and H16) of such y-hydrogen atoms is present and each hydrogen atom makes short contacts with both carbonyl oxygen atoms. The y-hydrogen atom H6 has short contacts with the carbonyl oxygen atoms 01 (d = 2.73 A) and 02 (d = 2.82 A) and makes boat-like and chair-like abstraction geometries, respectively. The 0...H contact distances and the abstraction geometries of the other y-hydrogen atom H16 with the 01 and 02 carbonyl oxygens are reversed. A pictorial representation of the two types of abstraction geometries (which are general for all diketones belonging to this group) and their Newman projections down the C2-C3 bond are given in Figure 38. The distance and the angular parameters of the y-hydrogens are given in Table V.  H (1)  (2)  Figure 38: (a) 1) Newman projection down the C2-C3 carbon-carbon bond (boat-like  abstraction geometry). 2) Pictorial representation of boat-like hydrogen abstraction geometry.  66  (1)  ^  (2)  (b) 1) Newman projection down the C2-C3 carbon-carbon bond (chair-like abstraction geometry). 2) Pictorial representation of chair-like hydrogen abstraction geometry. As described earlier in the Introduction, the distance and angular relationships between the abstracting carbonyl oxygen atom and the 7-hydrogen atom to be abstracted are important for a successful abstraction. A careful inspection of the geometrical parameters reveals that at least one 7-hydrogen atom in each diketone is situated within 2.76 A of a carbonyl oxygen with angles co, A, and 0 within the ranges of 50.4-63.8°, 77.9-84.2° and 113.1-117.5°, respectively. The 7-hydrogen atoms with close 0....H contacts are indicated in Figure 37 by dotted lines. Although these values deviate slightly from the suggested ideal values,79 several examples of successful hydrogen abstractions in these ranges, through boat-like rather than chair-like six-membered abstraction geometries, have been reported from our laboratory on a-cycloalkyl-pchloroacetophenones144 Not surprisingly, therefore, 7-hydrogen atoms with the above geometrical parameters would be subject to smooth hydrogen abstraction. A key assumption in the solid state structure-reactivity correlation studies is that the initially formed 1,4-biradical intermediate has the same basic geometry in the solid state as its ketonic precursor. In the solid state, any large changes in the molecular conformation are energetically disfavored by the constraining lattice medium; these  67  restrictions, together with the reported short life of the biradical intermediates, 145-148 support the above key assumption. Table V: The geometrical parameters of the y-hydrogen atoms.  y-hydrogen (RING SIZE) abstracted 12 H6 H6 H16 H16 DIKETONE  d (A)  co (°)  A (°)  0 (°)  01  2.80 2.73 2.80 2.73  66.3 63.8 66.4 63.8  74.2 77.9 74.2 77.9  116.7 113.4 116.7 113.4  abstraction geometry CHAIR BOAT CHAIR BOAT  abstracting oxygen 01 02 02  16  H6 1110 H20 H24  01 02 02 01  2.74 2.92 2.79 2.92  53.2 60.2 53.4 60.1  81.6 65.9 81.7 65.5  115.4 117.5 111.1 116.0  BOAT CHAIR BOAT CHAIR  20  H5 1413 H23 H31  01 02 02 01  2.70 3.00 2.70 2.99  51.7 59.1 52.1 59.7  82.5 63.0 84.2 63.9  114.8 114.4 114.8 115.4  BOAT CHAIR BOAT CHAIR  24  H5 H17 H27 H39  01 02 02 01  3.12 2.69 3.17 2.67  56.1 49.6 55.0 50.4  59.1 84.0 57.0 83.2  115.2 116.4 115.5 117.5  CHAIR BOAT CHAIR BOAT  26P  H5 (1130) 1120 (H43)  01(02) 02(01)  3.30 2.76  51.7 52.0  52.8 82.1  114.8 113.1  CHAIR BOAT  16*  H6 H10  01 01  2.99 2.72  59.0 53.3  65.5 81.4  116.6 115.7  CHAIR BOAT  16* - non-diametric sixteen membered ring diketone.  The representation of the possible biradical intermediate for the sixteen membered ring diketone and its reaction center depicted in Figure 39 are typical for all other  68  diketones in this group. The singly occupied orbitals in the initially formed biradicals (iBR) are not orthogonal to each other as is generally assumed for non-cyclic allcanones. However, in the gauche 1,4-biradical intermediate, the abstracting oxygen atom at Cl and the unabstracted 7-hydrogen atom at C4 would be in a "syn" relationship to one another (pre-cis biradical intermediate).  3.2.3 Structure Reactivity Correlation. As shown in Figure 39, the biradical intermediate has four modes of closure to give cyclobutanol. Biradical closure through the least motion route "path a", or the largest motion route "path d", would lead to a cis-cyclobutanol derivative, whereas "path b" or "path c" would lead to a trans-cyclobutanol derivative. Nd  Figure 39: a) Pre-cis biradical intermediate of diketone 9. b) Biradical intermediate  reaction center  69  Since the biradical closure to products is likely to be topochemically controlled in the solid state, the least motion closure through "path a" would be expected to be favoured and lead to cis-cyclobutanol derivative. Indeed, from this group of diketones, as shown in Table IV, cis-cyclobutanol derivatives were stereoselectively and almost exclusively obtained in the solid-state. Since the fate of the biradicals depends upon the the flexibility of the reaction cavity and its free volume, the observed stereoselectivity in this group of diketones suggests that the reaction cavity is very stiff, and indicates that the 1,4-biradical intermediate surrounded by the ground state ketone molecules remains in almost the same conformation as in the original crystal. The preference for cis-cyclobutanol formation in the solid state photochemistry of the diketones in this group can be related to the geometry of the biradical intermediate. Therefore, an examination of the diketone solid state conformation can provide valuable information regarding the preferred stereochemistry of the major cyclobutanol photoproduct. Interestingly, in this group of diketones, the abstraction of any of the four accessible 7-hydrogen atoms would generate a pre cis biradical intermediate, irrespective -  of which hydrogen atom is abstracted, and the least motion closure would eventually lead to cis-cyclobutanol derivatives. Since a successful hydrogen abstraction depends to a considerable extent on the 0-..-H contact distance,60 in the twenty-four and the twenty-six membered ring diketones the abstraction of the 7-hydrogen atoms with very large contact distances (> 3.12 A) can tentatively be ruled out. Successful intramolecular 7-hydrogen abstractions in the solid state have experimentally been shown to be feasible over contact distances of up to 3.1  A.79  Therefore, the abstraction of the 7-hydrogen atoms with  distances close to 3 A may be possible in diketones; however, the reported abstractions of the 7-hydrogen atoms with contact distances close to 3 A have better angular parameter values of w and A for abstraction, which are also important in determining hydrogen abstractability. Such abstractions in the diametric diketones cannot be proved experimentally from the photochemical results, since abstraction of all four accessible  70  hydrogen atoms would eventually lead to the same cis-cyclobutanol photoproduct. However, as discussed below, the photochemical results from the non-diametric sixteen membered ring diketone solve this ambiguity. Although there are two pairs of non-equivalent 7-hydrogen atoms present in diketone 15, owing to the C2 symmetry of the solid state conformation, only the nonequivalent hydrogen atoms H6 and H10 need to be considered. The 7-hydrogen atom H10 (d = 2.72 A, co = 53.3°, A = 81.4°, 9 = 115.7°) is more stereoelectronically favored for abstraction by the carbonyl oxygen than H6 (d = 2.99 A, co ,= 59.0°, A = 65.5°, 0 = 116.6°). Interestingly, the angular parameters of H6 are quite similar to those of the 7hydrogen atoms with longer contact distances (3 A) found in all other diketones. The abstraction of H6 and H10 could afford the cis-cyclobutanol derivative regioisomers 15C2 and 15C1 respectively, but the solid state photolysis leads almost exclusively to 15C1 via the abstraction of the stereoelectronically favoured hydrogen atom H10. A mere 2% formation of 15C2 is an excellent experimental evidence for a poor abstraction of the 7hydrogen atoms having contact distances close to 3 A in diketones.  3.2.4 Singlet and Triplet Reactions of Diketones. In solution, except for the smaller membered rings, the stereoselectivity characteristic of the solid state is largely lost, but a slight preference for trans over ciscyclobutanol derivatives is observed. As is well established for the type II reactions of aliphatic carbonyl compounds,  59,229  photochemical reactions of diketones were found to  occur via both the singlet and triplet n,n* excited states. Type II reactions were in fact one of the first known photochemical processes involving both electronic states.  149,150  In an attempt to investigate the multiplicity-dependent photochemistry of diketones, quenching studies of diketone 9 were performed using 2,3-dimethy1-1,3butadiene as a quencher. As described in the Experimental Section, irradiation of 10-2 M  71  diketone 9 in hexane was carried out at various quencher concentrations. The results indicate that all three photoproducts, cis-cyclobutanol (9c), trans-cyclobutanol (9t) and the cleavage product ene-dione (9e), are formed from both the singlet and triplet excited states. The product ratios at 0% conversions were calculated, as described earlier in this section, for different quencher concentrations, and were plotted as a function of quencher concentration (Figure 40). The singlet product ratios were obtained from the plateau of the graphs and normalized to 100% (Table VI). Three notable differences can be seen in the singlet state product distribution compared to that of the direct photolysis (Table VI): (1) The predominant formation of the trans-cyclobutanol derivative is reversed. (2) The value for the cis/trans ratio is increased. (3) The value for the cleavage/cyclization ratio is increased.  (a) 45  05 0 0.00^0.20^0.40^0.60^0.80^11)0^1.20^1.40  quencher concentration (M)  ^  1.80  ^  2.00  72  (b)  0.00^0.20^0.40^0.60^0.80^1.00^1.20^1.40^1.60^  1.80^2.00  quencher concentration (M)  Figure 40: a) Ene-dione/trans versus quencher concentration plot. b) Ene-dione/cis versus quencher concentration plot.  Table VI: Type II product ratios of diketone 9, from direct photolysis and photolysis with the quencher.  medium  excited state  hexane  triplet + singlet  22  35  43  0.63  hexane-quencher  singlet  21  16  63  1.31  cis (%)  trans (%) cleavage (%)  cis/trans  Reactions of both the singlet and triplet excited states are generally known to be qualitatively similar, but quantitatively different. The observed product distributions seem to reflect the quantitative differences in the photochemistry of singlet and triplet excited states. It has been well recognized that triplet type II reactions occur completely via biradical intermediates, while singlet reactions occur through a biradical intermediate,76,151,152 as well as via a concerted mechanism (Figure 41).59,137 The evidence for concerted processes from the singlet excited states comes from the  73  stereospecific formation of the type II products 153,154 and from the radiationless decay back to the starting ketone (S0) with preserved stereochemistry at the 7-carbon atom.155 singlet products  concerted  0 hydrogen -■ singlet products abstraction  BR1  So^Si^ISO^ 04 4 hydrogen  triplet products abstraction BR  3  Ti  Figure 41: Singlet and triplet pathways of type II reactions in diketones.  In aliphatic carbonyl compounds, owing to much slower intersystem crossing (kst = —1.5x108 sec-1)230 from singlet to triplet excited states compared to those of aromatic carbonyl compounds, longer singlet excited state lifetimes are generally observed. The estimated lifetime of an excited singlet state of an aliphatic carbonyl compound has been reported to be approximately 10 ns.156 In such time scale bond rotations may not be allowed in the singlet excited molecules. Even though the carbonyl carbons very likely pyramidalize upon excitation, the carbon skeleton of the ground state molecular conformations will be preserved; thus, the stereochemistry of the product formed by concerted cyclization may reflect the conformations of the ground state molecule. As depicted in Figure 42, the concerted singlet reaction of the solid state conformation of the sixteen membered ring diametric diketone 9A would stereospecifically afford the photoproduct cis-cyclobutanol derivative (9c).  74  diketone  singlet excited state (Si)  cis-cyclobutanol photoproduct  Figure 42: Diagrammatic representation of the formation of cis-cyclobutanol derivative 9c by the concerted cyclization of the solid state conformation of diketone 9.  The possible low energy conformations of diketone 9 generated by MM2 calculations indicate that the solid state conformation has the least energy. Nineteen other low energy conformations were also found within a 10 Umol-1 energy window above the solid state conformation. The first ten low energy conformations are given in Figure 43. Analysis of these conformations reveals that the majority of the conformations have a built-in preference for cis-cyclobutanol formation (0--H contacts in conformations with built in preference for trans-cyclobutanol formation are indicated by dotted lines). Therefore, the relatively large amount of cis-cyclobutanol formation from the singlet state reaction may be partly due to concerted cyclization. On the other hand, the observed results may also be an outcome of the quantitative differences in the product distribution of the singlet and triplet biradical intermediates, due to their different lifetimes. It is well known that C—C bond rotation rates157 are generally much faster than the relatively longer triplet lifetimes of the 1,4-biradicals. Therefore, in diketones, as generally observed in other aliphatic ketones, the triplet biradical intermediate would be likely to reach a conformational equilibration.158,159 However, the singlet biradicals are relatively quite short lived, and hence any significant conformational changes are thought to be generally impossible before they collapse to products.  75  Therefore, the concerted mechanism and the short singlet biradical intermediate lifetimes would account for the preference for cis-cyclobutanol derivative over trans relative to the direct photolysis, which occurs through both singlet and triplet excited states. The larger values for the cleavage/cyclization ratio observed in the singlet reactions compared to the direct photolysis of the diketone 9 are quite surprising, but similar results have been reported for acyclic aliphatic ketones as well. 155,160 However, in smaller membered cyclic ketones (cyclododecanone), singlet state reactions have been reported as giving high degrees of cyclization products.131,134 conformation^relative energy.  9A^0.0 kJ morl  98^2.0 kJ moil  76  conformation^relative energy.  9C  3.6 kJ moll  9D^4.9 kJ moll  9E^5.3 kJ moll  77  conformation^relative energy.  9F  6.3 kJ morl  9G  6.5 kJ morl  9J  7.6 kJ moll  78  conformation^relative energy.  9H^6.7 kJ mo1-1  91  6.8 kJ morl  Figure 43: Possible low energy conformations of diketone 9 generated by MM2.  The diketones with rings larger than twelve membered may be quite flexible and therefore may be able to explore various conformations in the isotropic medium. Therefore, the molecular motions necessary to form the type II products may be allowed without much strain in the ring. During the triplet reactions, since many bond rotations are allowed during the relatively longer biradical lifetime, the triplet biradicals can easily  79  attain conformational equilibrium and may partition to produce larger amounts of transcyclobutanol than cis-cyclobutanol. The formation of larger amounts of the transcyclobutanol derivative may be attributable to the greater strain energy intrinsic to the cisisomer. 162 The triplet sensitization reactions of diketones have been attempted for the sixteen membered ring diketone using acetone as the sensitiser; however, a complete photosensitisation could not be achieved. Although reports indicate successful acetone sensitisation of the photoreactions of saturated aliphatic ketones, the triplet energy (ET = 80-81 kcal 1)163 may not be sufficiently high to sensitize diketone 9. The quantum efficiencies of the triplet type II reactions of medium size cyclic ketones131 as well as those of acyclic ketones are generally reported to be much higher than those of the singlet processes. The product distributions from the triplet biradica1s will therefore dominate in the direct photolysis. In fact, in this group of diketones, with the exception of the twelve membered ring diketone 7, preference for trans-cyclobutanol derivatives over cis is observed during direct photolysis. The direct photolysis quantum yields of the photoproducts of the diketones 9, 10 and 12 were measured using standard procedures164 (see Experimental Section) in benzene at A, = 313 nm. The observed values given in Table VII are the average of three measurements. Due to the partial overlapping of the GLC peaks corresponding to the ciscyclobutanols of diketones 10 and 12, only approximate quantum yields could be estimated from the areas of GLC peaks relative to the other products. The quantum yields for cyclobutanol formation from diketone 9 are quite comparable to the reported values for the cyclobutanol formation from acyclic alkanones in non-polar solvents (e.g., cyclobutanol formation from 2-hexanone,  ocy  = 0.075).160  For the elimination process, significantly higher values are generally reported for the acyclic ketones (e.g., acetone formation from 2-hexanone, 4c1 = 0.252);160 however, quite similar values were reported for the elimination processes of cyclododecanone in cyclohexane (0c1 = 0.043).131 The overall quantum yield for type II product formation  80  observed in diketone 9 is significantly lower than that for aliphatic  acylic160,165  and cyclic  (cyclododecanone)131 ketones reported in non-polar solvents.  Table VII: Quantum efficiencies of type II photoproduct formation from diketones 9, 10 and 12.  diketone (ring size)  cis 0.019 (± 0.004)  trans  cleavage  0.031 (± 0.005)  0.054 (± 0.009)  9  (16)  10  (18)  0.015  0.038 (± 0.008)  0.045 (± 0.007)  12  (22)  0.009  0.032 (± 0.006)  0.037 (± 0.006)  The almost exclusive formation of cis-cyclobutanol derivative 7c from the solid state photolysis of the twelve membered ring diketone 7 can be explained from the X-ray crystal structure by using the same rationale as used for the other diketones. However, the stereoselective formation of the cis-cyclobutanol derivative in solution is interesting to consider. A similar stereoselective formation of a cis-cyclobutanol derivative (cis:trans = —6:1) has been reported in the solution state photolysis of cyclododecanone.130,131,134 As observed in cyclic diketones, photoreactions of cyclododecanone occur via both the singlet and triplet excited states. The singlet state reaction gives cis-cyclobutanol almost exclusively (cis/trans = 25), but the efficiency of the singlet reaction has been reported to be quite small compared to that of the triplet reaction (Ot = 0.53 / 0.03).131 Experimental results indicate that cyclododecane and its derivatives105,161 generally exist in a single square [3333] conformation in isotropic media even at room temperature. In solution, a square conformation [3333], with the carbonyl at the noncorner position, has been confirmed as the single and only conformation for cyclododecanone (monoketone), by Anet and co workers105 from their 13C nmr studies at  81  140 K and by 1Cristiansen and Ledaa1161 from their 1H nmr results at room temperature. The almost exclusive formation of the cis-cyclobutanol from the singlet reaction of the monoketone could be explained by the concerted mechanism or a mechanism proceeding through the short lived singlet biradical intermediate, as suggested by Matsui et al.134 MM2 calculations of diketone 7 reveal six non-diamond lattice conformations (7A, 7B, 7C, 7D, 7E and 7F) within a 10 kJ mold energy window, as depicted in Figure 44. The solid state conformation 7A was found to be similar to the global minimum. Using the calculated strain energies of these conformations, the Boltzmann distribution at 20°C was calculated. The ratio of the conformers 7A:7B:7C:7D:7E:7F is 84:8:3:2:<2:<2. Therefore, as predicted by Da1003 from his qualitative analysis, the solid state conformation should be predominant in isotropic media. As observed in the solid state conformation, the abstraction of any of the accessible y-hydrogen atoms (with the exception of H5 in conformation 7F) by the carbonyl oxygen atoms would lead to a precis biradical intermediate.  The results suggest that, in diketone 7, the biradical intermediate may have difficulties attaining a pre-trans conformations. The majority of the triplet biradical conformers of diketone 7 in the equilibrium distribution may have pre-cis arrangements as observed in the low energy conformations generated by MM2. A careful inspection of the molecular models of the biradical intermediates of the twelve membered ring diketone reveals that the biradical closure through "path b" or "path c" (Figure 39b) to form transcyclobutanol involves severe transannular short contacts. Such conformational changes necessary for the formation of the trans-cyclobutanol may be energetically unfavorable in diketone 7. However, the small amounts of trans-cyclobutanol observed in the cyclodecanone131 suggests that this factor is less important in the case of the monoketone. The results indicate a possibility of limited amounts of interconversions between pre-cis and pre-trans biradical intermediates during the triplet biradical lifetime. Burchill et a1131  82  demonstrated the formation of small amount of trans-cyclobutanol from cyclododecanone by ring rotations, but the molecular model analysis does not seem to correlate with this explanation. An ideal orbital overlap for cleavage may be equally difficult in such a small ring; therefore, the formation of appreciable amounts of the ene-dione cleavage product from the diketone 7 is rather surprising, and will be discussed later in this chapter. conformation^relative energy.  7A^0.0 kJ moll  7B^5.7 kJ moll  83  conformation ^relative energy.  7C^7.86 kJ mai  7D^9.24 kJ mai  7E^9.66 kJ morl  84  conformation^relative energy.  7F^9.67 kJ morl  Figure 44: Possible low energy conformation of diketone 7 generated by MM2.  3.3. Diketones that Give trans-Cyclobutanol as the Major Product in the Solid State. 3.3.1. Solid State conformation and Photochemistry. A stereoselective formation of trans-cyclobutanol derivatives was observed as the major process (Table IV) in the solid state reactions of the eighteen membered ring diketone 10, the twenty two membered ring diketone 12 and the twenty six membered ring diketone (needle crystal modification) 14N. The solid state conformations are depicted in Figure 45. The solid state conformation of the eighteen membered ring diketone 10 has been reported by Allinger et all° to have the carbon atoms arranged in a diamond lattice conformation with C2h molecular symmetry (Figure 45a), but not a rectangular one which might have been expected for a large ring. However, the diamond lattice rectangular conformation [3636] generated by molecular mechanics force fields166 was found to have  85  4.8 k.lmol-1 higher energy than the solid state conformation. In the case of the twenty six membered ring needle crystal modification (a dimorph of the plate crystal modification), the conformation is "zig zag" shaped, where the indentations in the methylene chain along one side of the long molecular axis are matched by extrusions on the other. As predicted by Dale98a, the twenty two membered ring, which belongs to the (CH2)4n+2 series, could adopt a diamond lattice rectangular conformation [3838]; however, nature seems to crystallize it in a more complex conformation as shown in Figure 45b. Interestingly, in this group of diketones the carbonyls are anti to each other, and not surprisingly an even number of methylene carbons is present between the two carbonyl groups. The geometrical parameters of the y-hydrogen atoms are summarized in Table VIII.  (a)  86  (b)  (c)  Figure 45: Solid state conformations of (a) eighteen, (b) twenty two, (c) twenty six (needles) membered ring diketones (the y-hydrogen atoms with short 0-...H contacts (d < 2.82 A) are indicated by dotted lines).  3.3.2. Hydrogen Abstraction Geometry and Biradical Geometry. As observed in the previous section, all y-hydrogen atoms having short 0....H (d < 2.82 A) contacts make boat-like abstraction geometries, whereas the other 'yatoms make chair-like abstraction geometries with contact distances larger than hydrogen atoms 3 A. In the highly symmetrical conformation of the eighteen membered ring diketone 10,  87  all four closest y-hydrogen atoms, which are equivalent by symmetry, make an intamolecular six-membered cyclic 0....H contact of 2.78 A with the carbonyl oxygen atoms. As indicated in Table VIII, the twenty six membered ring diketone (needles) 14N has a pair of y-hydrogens (H6 and H29) that have short 0....H contacts close to the ideal values (2.73 A), whereas in the twenty two membered ring diketone one y-hydrogen atom has an 0...H contact of 2.71 A, and the other one of 2.82 A. A 8-hydrogen atom (H7) of the twenty two membered ring diketone 12 makes a seven membered cyclic 0....H contact of 2.89 A with favorable angular geometry (Table VIII) for abstraction, but a cyclopentanol derivative was not detected during the photolysis. The rate of 8-hydrogen atom abstraction is generally known to be much slower than that of y-hydrogen atom abstraction for equivalent C-H bonds. 64 This difference has been suggested to be due to the greater entropy loss and torsional strain during 1,6-hydrogen transfer, compared to 1,5.60 The other y-hydrogen atoms in both the twenty two membered ring (H26) and twenty six membered (needle) ring (H20 and H43) diketones have 0....H contacts greater than 3.10 A, and make chair-like abstraction geometries. Table.VIII: The y-hydrogen geometrical parameters of diketones that give trans-  cyclobutanol derivatives as the major solid state photoproducts. DIKETONE y-hydrogen  abstracting  (RING SIZP)^abstracted  oxygen  18 22  26N  H5 H7 (8) H16 H26 H35 H6 (H29) H20 (H43)  01 01 02 02 01 01(02) 02 (01)  d (A)  co (°)  A (°)  0 (°) abstraction geometry  2.78 2.89 2.82 3.10 2.71  54.1 63.5 52.1 56.6 44.2  78.3 63.0 80.5 58.6 87.7  112.7 146.6 111.4 114.3 113.6  2.73 3.26  48.6 57.2  84.6 57.5  114.9 100.0  BOAT BOAT CHAIR BOAT BOAT CHAIR  88  The two types of abstraction geometries and their Newman projections down the C2-C3 bonds are depicted in Figure 46.  ..,, C5 CH H  2  C3  C2^Cl  H H^Cn-1 CH2 /  H (2)  (1)  Figure 46: a) (1) Newman projection down the C2-C3 carbon-carbon bonds (boat-like abstraction geometry). (2) Pictorial representation of boat-like hydrogen abstraction geometry.  \ Cn-1 H  H CH2^,  I C5 C4 I,^I ---CH2  H^ci^ C2  (2) b) (1) Newman projection down the C2-C3 carbon-carbon bonds (chair-like abstraction geometry). (2) Pictorial representation of chair-like hydrogen abstraction geometry.  89  From the diagram it is apparent that the abstraction of any y-hydrogen atom (with the exception of H20 and H43 from the twenty six membered ring diketone) that has a short 0 -.H contact with the carbonyl oxygen leaves the unabstracted y-hydrogen atom -  and the oxygen atom in an anti relationship in the biradical intermediate. The biradical intermediate from the eighteen membered ring diketone 10 and its biradical reaction center, shown in Figure 47, are typical for this group of diketones.  3.3.3. Structure Reactivity Correlation. In the solid state, as depicted in Figure 47, the least motion closure of the biradical intermediate through "path a", with retention of the orientation of the hydroxy group and the unabstracted y-hydrogen atom at the Cl and C4 carbon atoms, would lead to the trans-cyclobutanol derivative as the major photoproduct. This is precisely the experimentally observed result as indicated in Table IV. In the solid state conformation of the twenty six membered ring diketone (needle) 14N, in addition to the two closest 'y-hydrogen atoms H6 and H29 (d = 2.73  A, co = 49°,  A = 85°, 9 = 115°), the other two y-hydrogen atoms H20 and H43 (d = 3.26 A, co = 57°, A = 58°, 0 = 100°) make a chair-like abstraction geometry which can be considered for analysis. The abstraction of H6 (H29) is obviously preferred on the basis of its abstraction parameters and would lead to a pre-trans biradical intermediate. On the other hand, it is apparent from the solid state conformation of diketone 14N that the abstraction of H20 (H43) should lead to a pre-cis biradical intermediate. However, upon irradiation, the needle like crystals of the twenty six membered ring afforded 91% trans-cyclobutanol and 9% cis-cyclobutanol derivatives. It is interesting to point out that, as described earlier in this chapter, the irradiation of its dimorphic plate-like crystals is highly stereoselective for cis-cyclobutanol. Such widely divergent photochemical stereoselectivity observed from  90  these conformational polymorphs is one of the rare examples found in the literature.57d,167,168,169  HS  ak^H2  CI  • HE  Figure 47: a) Biradical intermediate from diketone 10. b) Biradical intermediate reaction center. The solid state photochemical results of the dilcetones analyzed so far can be explained in terms of the solid state molecular conformations obtained by X-ray. This indicates that the diketones react in a topochemically, conformationally specific manner in the crystal. However, the formation of 9% cis-cyclobutanol from the twenty six membered diketone 14N is explicable as arising from either abstraction of H20 (H43) or the loss of control of the biradical closure following the abstraction of H6 possibly due to the presence of defect sites. Since low temperature reactions afforded similar results,  91  crystal melting during photolysis can be ruled out. A similar argument would correctly account for the formation of the minor products observed in the solid state reaction of other diketones analyzed so far. This suggests that the reaction cavities in these diketone crystal lattices allow limited amounts of motion to the excited state molecules or to the biradical intermediates. In solution, the product ratios are comparable to the solution results observed in the previous section. A similar rationale can therefore be offered to explain the loss of stereoselectivity and the preference for the formation of the trans-cyclobutanol over cis.  3.4. Diketones in which the Stereoselectivity is Significantly Lowered in the Solid State. The fourteen membered ring diketone crystallizes in a rectangular diamond lattice conformation [3434] (Figure 48).  .H17^147 C11^•.^•^CS  H121^.• 1412 CWCs • •. C 1416^ CS^C7 1415^02^HS 14130^01411  1410  Figure 48: Solid state conformation of the fourteen membered ring diketone 8.  92  The two carbonyls in the molecule are placed in the middle of the longer segment (P-to the comer atom), rather than on the short segment as generally observed in the other rectangular conformations of diketones. In the solid state reactions of the diketones discussed in this thesis, it is interesting to note that the general alternating trend observed in the stereoselectivity of cyclobutanol formation with respect to ring sizes is not followed by the fourteen membered 8 and twenty six membered (plates) 14P ring compounds. If the trend were followed, one would expect a stereoselective formation of transcyclobutanol as the major solid state product from both fourteen and twenty-six membered diketones. As discussed earlier in this chapter, the plate dimorph 14P gives cis-cyclobutanol exclusively, whereas in the fourteen membered diketone 8, the corresponding ciscyclobutanol was obtained as the major product. However, the needle dimorph of the twenty six membered ring diketone 14N obeyed the trend observed for type II stereoselectivity. Interestingly, these two examples have rectangular diamond lattice solid state conformations [3x3x], whereas all other diketone rectangular conformations have a non-diamond lattice arrangement. An inspection of the fourteen membered solid state conformation reveals that there are four equivalent 'y-hydrogen atoms (H6, H7, H17, H19) (d = 2.71 A, 0) = 52.5°, A = 83.0°, 0 = 116°) located favorably for abstraction through a six-membered boat-like abstraction geometry. The Newman projection through the C2-C3 bond and the abstraction geometry common to all four hydrogen atoms is depicted in Figure 39b. Upon irradiation, the abstraction of any of these four hydrogen atoms would lead to a pre cis -  biradical intermediate. Therefore, in the solid state, the least motion closure with retention of the orientation of the unabstracted y-hydrogen and the hydroxy group at C4 and Cl respectively should have led to the formation of cis-cyclobutanol. Surprisingly, the stereoselectivity is significantly lowered compared to the selectivity observed in the solid state reactions of the other diketones, although, as indicated in Table IV, preference for  93  cis-cyclobutanol formation over trans-cyclobutanol (cis:trans=2:1) is maintained. The predominant formation of the cis-cyclobutanol derivative in isotropic media is similar to the solution results with the twelve membered ring diketone 7 and the monoketones130 described earlier in this chapter. The formation of trans-cyclobutanol in the solid state reaction from diketone 8 is rather surprising, as it cannot be related to the solid state conformation, and clearly indicates the partial loss of topochemical control. Such abnormal non-topochemical behaviour, which cannot be explained by structure-reactivity correlations, may be the result of one or more of the following situations: 1) Crystal melting and concomitant loss of topochemical control. 2) A fraction of the excited state molecules or the 1,4-biradical intermediates may reside in defect sites. 3) There might be a difficulty in the formation of the twelve membered ring (larger ring in the cyclobutanol derivative) from the fourteen membered ring (dilcetone). Since the melting point of this diketone is fairly high (148-149°C), the highest among all diketones investigated, the loss of stereoselectivity is probably not related to crystal melting during photolysis. Furthermore, lowering the photolysis temperature had no effect on the product distributions. Therefore, the loss of topochemical control due to crystal melting can be ruled out. Any conformational rearrangements in the constrained solid state lattice medium would involve high transition state potential energies. However, excited state molecules or biradical intermediates residing at defect sites could undergo such transformations. The defect sites in crystal lattices, unlike regular lattice sites, are generally known to provide quite flexible reaction cavities, with enough free volume for the molecules to undergo such conformational changes. In the case of the fourteen membered ring diketone 8, during the solid state photolysis the excited molecules or the biradical intermediate molecules at the  94  regular lattice sites may topochemically afford cis-cyclobutanol as the major photoproduct, whereas the molecules residing at defect sites may undergo conformational isomerization or may even reach conformational equilibrium (if the biradical lifetime permits), and consequently afford both cis and trans-cyclobutanol derivatives. A larger than normal value of the residual index R factor (0.073) of the crystal structure of diketone 8 may be an indication of the presence of such defect sites in the crystal lattice. On the other hand, a greater degree of difficulty associated with biradical cyclization to form cyclobutanol photoproducts possessing twelve membered rings may also be partially responsible for the loss of topochemical control. Such difficulty in the cyclization process may increase the lifetime of the biradical intermediate and as a result allow time for the molecules to equilibrate conformationally. Cyclization processes that form medium size ring compounds are generally known to be difficult. Evidence to support this can be seen from the rate constants reported for the ring closure reaction to form cyclic lactones.170 A significant drop in the rate constant for twelve membered ring formation compared to that of the thirteen membered ring analogue has been reported. Although the entropy factor favours formation of the twelve membered compared to the larger membered rings, the unfavorable enthalpy factor in the medium size ring formation, presumably due to transannular interactions, increases the activation energy for the cyclization process. A molecular model analysis also indicates severe transannular contacts during the cyclization process of the diketone 8 biradical intermediate. A similar situation may also exist during the photolysis of the twelve membered ring diketone 7 (the single square conformation has a built in preference for ciscyclobutanol), but the ring size may not be large enough to allow conformational change from the pre-cis conformer into pre-trans, even during the lengthened biradical lifetime; thus, only cis-cyclobutanol derivatives were observed. At this point, investigation of the low energy conformations of fourteen membered ring diketone 8 may provide valuable information. The molecular mechanics calculations  95  of diketone 8 generated five low energy conformations (8A, 8B, 8C, 8D and 8E) within a 5 kJmo1-1 energy window. The conformations are depicted in Figure 49. The solid state diamond lattice conformation 8A was found to be the global minimum. From his qualitative analysis 1°3 Dale concluded that the solid state conformation is the predominant conformation in solution as well. However, our calculation shows the conformation 8B has only 0.3 klmo1-1 higher energy than the global minimum. conformation^relative energy  8A^0.0 kJ morl  8B^0.3 kJ mal  96  conformation^relative energy  8C^2.0 kJ moll  8D^2.8 kJ morl  8E^4.5 kJ morl  Figure 49: Five low energy conformations of diketone 8 generated by MM2.  97  The Boltzmann distribution at 20°C (8A:8B:8C:8D:8E = 35:32:15:12:6) indicates almost equal amounts of 8A and 8B at equilibrium. Conformation 8A can isomerize into  8B by —120° bond rotations around C10-C11 and C12-C13 (Figure 50). In conformation 8B the abstraction of the y-hydrogen atom H19 (d = 2.81 A) by the carbonyl oxygen atom 01 will afford a pre-trans biradical intermediate. A similar pre-trans biradical intermediate is also possible from the abstraction of the y-hydrogen atoms H8 or H19 in conformation 8D. However, the abstraction of the rest of the y-hydrogens in these low energy conformations would afford pre-cis biradicals. Therefore, conformational isomerization of the biradical intermediates as a result of both the flexible reaction cavities at the defect sites and/or the longer biradical lifetimes may contribute to the significant loss of stereoselectivity in the solid state.  H2  H2  H2^  01  H4 C2  I-21  CI  H3  C13 H2 ,'^ HI 02  41^  H19  It  10H20  HS  HS H7^ He  C11 -120° CIO H16  it  CS H14  111 46  HI2  C6  02  H10  11.^'MP C9^C7 HIS  HI3  H9  HII  8A  Figure 50: Conformational isomerization from 8A to 8B.  8B  98  In solution photolyses, the contribution of low energy conformations and the isomerization of the biradical intermediates would correctly account for the loss of stereochemistry. However, the preference for the more strained cis-cyclobutanol formation over its stereoisomer is also observed in other small membered monoketones and in the twelve membered ring diketone. The observed solution results indicate that either the equilibrium conformations of the biradical intermediates possess largely the precis conformations (MM2 calculations further support this) rather than pre-trans, or the pre-trans biradical intermediates may have more difficulty achieving the best overlap for  closure than the pre-cis, due mainly to the size of the ring. The absence of transcyclobutanol formation from the twelve membered ring diketone clearly suggests that the pre-trans biradicals are impossible due to ring size.  The seventeen membered ring diketone does not show any stereoselectivity either in solution or in solid states. The X-ray structure of these compounds could not be obtained due to their poor crystal quality, as is often the case with the odd membered cyclic compounds. The loss of stereoselectivity may be due to: 1) A single solid state conformation that leads to both pre-cis and pre-trans biradical intermediates. 2) The lattice may contain many defect sites as indicated by the poor crystal quality. The solution-like product distribution in the solid state photolysis suggests that the majority of the molecules are residing at defect sites. The reaction cavities at these defect sites may be quite flexible, with large free volumes to allow conformational isomerization to yield almost solution-like product distribution. Although the melting point of the diketone is quite low (68-69°), lowering the photolysis temperature had no effect on the product ratios.  99  3.5. Chemoselectivity in the Solution and Solid State Photoreactions of Diketones. The cyclization to cleavage product ratios from the type II reaction of the diketones at 20°C are compiled in Table IX. From the table it is apparent that, with the exception of the fourteen and seventeen membered rings, all diketones undergo more cyclization in the solid state reaction than in the corresponding solution reaction. In the fourteen and the seventeen membered ring diketones, however, cyclization/cleavage ratios are almost the same in both media. There are several reports of type II reactions where cyclization/cleavage ratios in the solid state and in solution are practically identical as observed in the fourteen and seventeen membered ring diketones.57f,171 The preference for cyclization over cleavage observed in the irradiation of diketones in cyclohexane seems to vary with ring size. In the case of both the twelve and fourteen membered ring diketones, a high degree of cyclization is observed in both media, and the cyclization/cleavage ratios are comparable, as opposed to the larger membered rings, where the amount of cyclization and cleavage are almost equal in the isotropic media. A similar trend in the chemoselectivity has been reported for irradiation of monoketones in cyclohexane (Table II), as mentioned earlier in this chapter.130 The preference for cyclization products in the solid state photochemistry of diketones can be related to the 1,4-biradical geometry. It is assumed that the hydrogen abstraction occurs in the crystalline phase with minimal conformational changes, to produce biradicals with the same basic conformation as their ketonic precursor. The conformation of the biradical intermediate reaction sites can be defined by three torsional angles, 01, 02 and 0. 01 -  Angle of the p-orbital on Cl with respect to the C2-C3 bond.  02  Angle of the p-orbital on C4 with respect to the C2-C3 bond.  -  0- Dihedral angle of Cl-C2-C3-C4.  100  The torsional angles 01 and 02 (Figure 51), are calculated from the X-ray crystal structure data based on two assumptions: 1) The hybridization of the ring carbon C4 bearing the unabstracted y-hydrogen atom changes from sp3 to sp2. 2) The p-orbitals at Cl and C4 lie perpendicular to the planes delmed by 01-C1-C2 and C3-C4-05, respectively.  Table IX: cyclization/cleavage ratios of diketones both in solution and solid state.  DIKETONE CYCLIZATION (%) (ring size) SOLID STATE HEXANE 12 99 84 14 87 90 99 57 16 87 59 18 94 20 33 95 22 44 24 99 42 100 26 (P and N) 47 16* 100 61 17 49 53 8  C2 Cl  Cn-1  CLEAVAGE (%) SOLID STATE HEXANE <01 16 13 10 01 43 13 41 06 67 56 05 01 58 00 53 00 39 51 47  C3  01^02 ^  ^  C4 C5  01  Figure 51: A diagrammatic representation of the biradical intermediate reaction center showing 01, 02 and 0 angles.  101  The 01 and 02 values and the torsional angles around the central sigma bond are given in Table X. Table X: Biradical parameters of diketones.  diketone  7-hydrogen  carbonyl oxygen  01  (0)  02  (0)  C2 C3 dihedral -  angle (°)  (ring size) 10  H3 (H12)  02 (01)  67.5  -30.3  62.6  12  H6 (H16)  01  12.2  85.0  -60.3  H6 (H16)  02  55.9  21.2  -64.6  14  H6  01  -65.6  -25.2  64.2  16  H6  01  -67.2  -24.4  65.6  H10  02  -19.3  -83.5  70.7  H20  02  -68.1  -24.5  68.5  H24  01  -19.3  -86.0  60.5  18  H5  01  -69.1  -89.5  74.2  20  H5  01  71.4  24.4  -67.6  H13  02  22.6  87.5  -59.2  H23  02  69.4  24.9  -65.5  H31  01  22.1  84.9  -58.4  H16  02  -68.7  85.6  67.8  H26  02  -29.2  -82.3  59.8  H35  01  80.6  -87.0  -69.4  H5  01  -27.1  -83.2  61.2  H17  02  -73.8  -25.4  68.2  H27  02  -30.8  -82.2  59.9  H39  01  -71.6  -25.7  64.9  H6 (H29)  01(02)  73.3  -86.4  -69.0  1143 (H20)  01(02)  -32.0  -81.2  58.1  115 (H30)  01(02)  -37.8  -82.6  -59.5  H43 (H20)  01(02)  70.6  24.2  -67.4  H6  01  20.5  84.6  -64.0  HIO  01  66.4  25.8  -63.6  22  24  26N  26P  16*  102  The sign and the magnitude of the angles ei and 02 depend on whether the top or bottom lobe is used for measurement. The lobe which makes the smallest torsional angle with the C2-C3 bond was chosen in each case. From Table X, it is apparent that all biradical intermediate C1-C2-C3-C4 conformations are gauche rather than cisoid or anti, with dihedral angles around the central sigma bond on the order of 65°. As shown in Figures 39b and 47b, the singly occupied p-orbitals at Cl and C4 in the biradical intermediate are in close proximity to one another, but are out of alignment with the central sigma bond C2-C3. In each case at least one torsional angle, either 01 or 02, is close to 60°. As stated in the Introduction, due to orbital overlap considerations, the gauche or cisoid arrangements of the 1,4-biradical intermediates can undergo both cyclization and  cleavage, while the anti arrangement undergoes cleavage exclusively. Furthermore, there is general agreement that the cleavage of a 1,4-biradical requires a significant overlap between the central sigma bond being broken and the singly occupied p-orbitals.172 Evidence for the above statement emerges from Hoffman's calculations, in which an extensive overlap between the p-orbitals and the central sigma bond of a 1,4-biradical intermediate optimizes the mixing of  it and  a levels, which promotes cleavage.173  Therefore, extensive atomic and molecular motions around the Cl-C2 and C3-C4 bonds are required to align the p-orbitals on Cl and C4 with the C2-C3 bond to achieve an ideal arrangement (i.e., the 01 and 02 need to approach 0°) for cleavage. Since a significant overlap of the singly occupied orbitals through space is generally accepted as being necessary for a successful cyclization59,60,115,116,  it  is not surprising that the  preferred least motions in the solid state would lead to a predominant cyclization. In solution, as opposed to the solid state, alternative diketone conformations as well as possible conformational isomerization of the 1,4-biradical intermediates can allow for a better overlap of the singly occupied orbitals with the C2-C3 bond, and as a result can lead to increased values of the cleavage/cyclization ratio. However, the greater  103  degree of cleavage products from large rings, compared to that observed in twelve and fourteen membered rings, seems to indicate that in large rings, conformational rearrangements of the biradical intermediates can more easily lead to the ideal orbital overlap for cleavage, including an anti conformation (which may be generally difficult in cyclic compounds). Whereas in medium rings the Cl-C2 and C3-C4 bond rotations may be limited from achieving significant overlap between the central sigma bond and singly occupied p-orbitals due to the ring size; thus, they stay in gauche or cisoid arrangements, which are rather intrinsic to cyclic ketones and yield larger amounts of cyclization. Analysis of the molecular models of medium rings indicates severe tansannular contacts when the biradical intermediates are ideally placed for cleavage. For non-cyclic ketones, compared to the cyclic ketones, completely reversed cleavage/cyclization ratios are reported in the literature.115 The majority of the biradical intermediates from non-cyclic ketones tend to stay in stable anti rather than cisoid or gauche conformations in isotropic media; not surprisingly, therefore, this leads to large  amounts of cleavage products. In the solid state and in the organized media, however, even larger cleavage/cyclization ratios have been reported for non-cyclic ketones compared to those observed in isotropic media.162,174,175 From a fascinating type II photochemical investigation of a series of conformationally fixed cyclohexanones, Ian Fleming et al. showed a great influence of the orientation of the C2-C3 bond on the cleavage/cyclization ratio.176 Since a continuous orbital overlap with C2-C3 bond is considered necessary for the cleavage reaction, a higher degree of cyclization was observed from a molecule where the C2-C3 bond is held more or less in an orientation unfavorable for the C2-C3 bond cleavage. A similar analysis would clearly explain the larger degree of cyclization observed in the twelve and the fourteen membered ring diketones in the solution photolysis. In the gauche biradical conformation the C2-C3 bond may have difficulty attaining a favorable orientation for cleavage due to the small ring size.  104  The small amounts of cleavage products in the twelve and fourteen membered diketones in isotropic media can be explained by "pseudorotation" 105,177,178 of the rings, a process much like the out-of-plane methylene group "puckering" of the cyclopentane envelop conformation rotating around the ring.179 Similar ring motions have been proposed for the high temperature solid phases of the cyclic compounds that undergo solid-solid phase transitions.98a Such rotations have been well investigated for cyclododecanone (Figure 52) in isotropic media. 105 A ring rotation of the biradical intermediate of the diketone can place both singly occupied orbitals at corner positions from the non-corner positions, and thus the biradicals are no longer gauche but anti, with both p-orbitals considerably overlapped with the central a-bond. The estimated rate constant105 for such conformational interchange in cyclododecanone between conformations having the C=0 group at corner and non-corner positions at room temperature is 2x107 s-1, and this is of the order of the triplet biradical lifetimes.158,159 In isotropic media, such rotations of the biradical intermediates of the smaller rings may account for the formation of cleavage products in appreciable amounts. A similar argument has been proposed by Burchill et al.131 for the formation of small amounts of cleavage products from cyclododecanone. During the solid state reaction of the fourteen membered ring diketone 8, similar rotations of the biradical intermediates at defect sites may partially account for the formation of 13% of cleavage product 8c. The large degree of cleavage product (-50%) observed in the solid state photolysis of the seventeen membered ring is comparable to its solution results. The solution-like cyclization/cleavage ratio is explicable using the same analogy used to explain the loss of stereoselectivity observed in the solid state reaction. A similar explanation would correctly account for the loss of topochemical control in the photochemistry of the ketoalcohols. Crystals could not be grown for either keto-alcohol 29 or 30, but as observed in the seventeen membered ring diketone, they exist in powder or amorphous form. The observed product ratios indicate that the reaction cavities are quite flexible and the free  105  volume present in the cavities are large enough to allow solution-like conformational changes as observed in the seventeen membered ring diketone.  gauche  anti  Figure 52: Hypothetical conformational interchange (pseudorotation) of cyclododecanone biradical intermediate from gauche to anti conformation.  3.6. Photochemistry of the Ten Membered Ring Diketone. As described earlier in this chapter, the ten membered ring diketone 6 crystallizes with the diamond lattice conformation 6C (Figure 53). The conformation has two yhydrogen atoms, H3 and H12, having short 0.-H contacts (d = 2.74  A, co = 52.1°, A =  91.7°, 0 = 113.3°) with favorable angular parameters for abstraction (the abstraction geometries are boat-like); surprisingly, diketone 6 remains photostable both in solution and in crystalline media. The two 8-hydrogen atoms, H8 and H15, also lie relatively close to the carbonyl group, but the stereoelectronic disposition of these atoms with respect to the carbonyl groups is quite different from that of the y-hydrogen atoms (d = 2.81  A, (0 =  98.5°, A = 76.2°, 0 = 134.4'). The unfavorable value of the angular parameter co, which is close to 90°, would make the 8-hydrogen abstraction unlikely.  106  Figure 53: ORTEP stereodiagram of diketone 6.  Based on a low temperature X-ray crystal structure analysis, cyclodecanone101has also been reported to have a conformation similar to that of diketone 6C, with the carbonyl next to the comer position (Figure 53). The same monoketone conformation has also been reported to have the lowest energy by MM2 calculations.136,180,181 This monoketone conformation has a 7-hydrogen lying close to the carbonyl groups, (d = 2.54 A) with co angle = 560.136 However, cyclodecanone has long been known for the unique e-hydrogen abstraction in cyclohexane (non polar) solvent.135 Sauers et al 136 recently reported the divergent photobehaviour of this cyclic ketone, which depends on the polarity of the isotropic medium (Figure 54). As opposed to the situation in cyclohexane, irradiation of cyclodecanone in t-BuOH, gave photoproducts formed via both the 7-hydrogen abstraction (bicyclo[6.2.0]decan-1-ol) and the &hydrogen abstraction (9-decalo1). Lewis bases like tert-BuOH are known to solvate the hydroxy biradical intermediate by forming a hydrogen bond with the hydroxy group of the biradical intermediate and thereby suppressing the reverse hydrogen transfer back to the starting ketone.59,115,116 Such an effect of the solvent molecules has been reported to be  responsible for the increased lifetime of the biralical intermediate in polar solvents.182  107  OH y H-abstraction  e H-abstraction  ^■ -  reverse c H-transfer  cyclization  Si  •  -  reverse y H-transfer  cleavage  hv  •  OH  S0  Figure 54: 7 and e-hydrogen abstraction pathways of cyclodecanone.  During the photolysis of cyclodecanone in cyclohexane, both 7 and e-hydrogen abstractions are presumably feasible from lowest energy and other alternative conformations, but an efficient reverse hydrogen transfer of the y-hydrogen atoms has been suggested to explain the exclusive formation of 9-decalol (e-hydrogen abstraction). However, in polar solvents (tert-BuOH), the reverse hydrogen transfer is slowed down or prevented due to hydrogen bonding and therefore products from both 7 and &hydrogen abstractions have been observed in a ratio of 3:2. MM2 calculations of diketone 6 reveal three low energy conformations (6A, 6B and 6C) within a 20 kJ mold energy window (20 kJ mo1-1 above the global minimum), as illustrated in Figure 55. Interestingly, a conformation similar to the solid state conformation 6C has the highest energy, and is totally different from the other two conformations. Using the calculated strain energies of these three conformers, the Boltzmann distribution at 20°C was calculated. The ratio of the conformations 6A:6B:6C is 97:2:1, so conformation 6A is the dominant conformation in solution.  108  conformation^relative energy  6A^0.0 kJ mori  6B^9.9 kJ morl  6C^11.9 kJ mei  Figure 55: Three possible low energy conformations of diketone 6 generated by MM2.  109  In the MM2-derived solid state conformation 6C, the angular parameters of the 7-hydrogen (c) = 51.5°, A = 93•30, 0 = 113.00) and 5-hydrogen (co = 83.1°, A = 75.0°, 0 = 136.3°) atoms are close to those found by X-ray, but the 0...H contact distances (cty = 2.54 A, ci6 = 2.63 A) are quite short compared to those of the solid state conformer. The highly symmetrical global minimum 6A has a diamond lattice conformation. There are four 5-hydrogen atoms with close 0...H contacts (d = 2.67 A) and angular parameters favorable (co = 63.3°, A = 91.6°, 0 = 101.5°) for abstraction by the carbonyl oxygens. The non-symmetrical conformation 6B has a 7-hydrogen atom H13 with a close contact distance of 2.52 A, and its other three 5-hydrogen atoms have contact distances of <2.79 A with the carbonyl oxygen. However, in both cyclohexane and t-BuOH, as well as the solid state, diketone 6 was found to be photostable. As observed for cyclodecanone, the photostability may be due to an efficient reverse 7 or 5-hydrogen abstraction in the biradical intermediates (even in polar solvents) or an efficient deactivation of the excited molecules by radiative or radiationless processes that are intrinsic to the molecule.  3.7. The Best Geometrical Requirements for y-Hydrogen Abstraction in Diketones. The geometric disposition of the 7-hydrogens that have the closest 0...H contacts in each diketone are compiled in Table XI. Interestingly, in all diketones the transition state abstraction geometry is boat-like rather than a chair-like, which is considered to be a preferred strain-free transition state for 1,5-hydrogen abstraction.59,76,115 The nature of the six membered 1,5-hydrogen transfer abstraction geometry for type II reactions has been investigated both theoretically and experimentally.  110  Table XI: Geometrical parameters corresponding to y-hydrogens having the closest 0...H contacts. ring size g-hydrogen  d (A)  co (°)  A (°)  0 (°)  product  geometry  12  H6  2.73  63.8  77.9  113.4  CIS  BOAT  14  H6  2.71  52.5  83.0  116.0  CIS  BOAT  16  H6  2.73  53.2  81.6  115.4  CIS  BOAT  16*  H10  2.72  53.3  81.4  115.7  CIS  BOAT  18  H5  2.78  54.1  78.3  112.7  TRANS  BOAT  20  H23  2.69  52.1  84.2  114.8  CIS  BOAT  22  H35  2.71  44.2  87.7  113.6  TRANS  BOAT  24  H39  2.67  50.4  83.2  117.5  CIS  BOAT  26N  H6  2.73  48.6  84.6  114.9  TRANS  BOAT  26P  H20  2.76  52.0  82.1  113.1  CIS  BOAT  average  2.72  52.4  82.4  114.7  best  2.71  44.2  87.7  113.6  Boer et al183 suggested a planar (01-C1-C2-C3-C4-7H) transition state for the type II process and its radical cation state equivalent, the Mc Lafferty rearrangement. Lewis and Wagner have, however, disproved this planar structure for type II reactions, because the hydrogen abstraction rate constants in planar transition state geometries are expected to be diminished significantly by sub stituents at the a and (3 positions. In a planar transition state, any substituents at the a and 13 positions would cause unfavorable eclipsing interactions with the methylene hydrogens, resulting in higher activation energies and consequently lower rate for 7-hydrogen abstraction. However, no such effect was  111  found and this led to the postulation of a strain free chair - like transition state by Wagner.59,76,115  In diketones, however, the y-hydrogen atoms which make chair-like abstraction geometries have relatively unfavorable 5, co and A values for abstraction, compared to the y-hydrogen atoms which make boat-like abstraction geometries. Of all the y-hydrogen atoms in Table XI, H35 in the twenty two membered ring diketone has the best values (d = 2.71A, co = 44.2°, A =87.7°, 9 = 113.6°) for abstraction. Interestingly, the parameter values of all y-hydrogen atoms with closest 0.--H contact distances are quite close to their mean values (d = 2.72A, co = 52.4°, A = 82.4°, 0 = 114.7'). From the investigation of ene-diones and enone-alcohols (hydrogen abstraction by oxygen occurred over a range of 2.3-2.7A) conducted in our laboratory, 14 the ideal distance for hydrogen abstraction by an oxygen atom was suggested to be equal or less than the sum of the van der waals radii of the oxygen and hydrogen atoms (d 2.72A); in these compounds, the values of co and A were also reported to be quite close to the ideal values. In our present investigation, efficient y-hydrogen atom abstractions were found at distances as high as d = 2.78A, co = 63.8°, A = 77.9°, 0 = 112.7°. However, these values may not be the conclusive upper limit for diketones, since in each diketone more than one accessible y-hydrogen atom leads to the same photoproduct. Another investigation from our laboratory on a-cycloalkylacetophenones57f revealed that y-hydrogen atom abstractions are feasible with 0.--H contact distances up to 3.1A, which is significantly higher than the ideal value, but the corresponding angular parameters were quite favorable for abstraction compared to those 'y-hydrogens with d values close to 3A of the diketones. Eleven out of fourteen compounds investigated in this series underwent smooth reaction through boat-like abstraction geometries, as observed in diketones. The torsional angles in Table XII describe the geometries of the biradical intermediates corresponding to the closest y-hydrogen atoms in each diketone.  112  Table XII: The geometrical parameters of the biradical intermediates corresponding to the y-hydrogen atoms having closest 0...H contact distances.  ring size y-hydrogen  01 (°)  02 (°)  0 (°)  product  geometry  12  H6  55.9  21.2  -60.3  CIS  BOAT  14  H6  -65.6  -25.2  64.2  CIS  BOAT  16  H6  -67.2  -24.4  65.6  CIS  BOAT  16*  H10  66.4  25.8  -63.6  CIS  BOAT  18  H5  -69.1  -89.5  74.2  TRANS  BOAT  20  H23  69.4  24.9  -65.5  CIS  BOAT  22  H35  80.6  -87.0  -69.4  TRANS  BOAT  24  H39  -73.8  -25.4  68.2  CIS  BOAT  26N  H6  73.3  -86.4  -69.0  TRANS  BOAT  26P  H20  70.6  24.2  -67.4  CIS  BOAT  average  69.2  66.7  An interesting correlation between the biradical geometries and the stereochernistry of the major solid state product formed is apparent from Table XII. For all diketones the 01 values remain close to the mean value 69.2°, with the exception of a slight deviation in the twelve and twenty two membered rings. However, a large difference in 02 values can be seen between the pre-cis and pre-trans biradical geometries. It seems that the solid state photolysis of diketones stereoselectively affords cis-cyclobutanol derivatives and trans-cyclobutanol derivatives as the major products when the values of angle 04 are close to 25° and 87°, respectively.  113  CHAPTER IV.  4.0. Polymorphism, Solid-Solid Phase Transitions and Solid Phase Order Dependent Photochemistry of Diketones. As described in the Introduction, the polymorphic crystal modifications of organic compounds can differ from each other not only in the packing arrangement of the constituent molecules, but also in their molecular conformations as well. Furthermore, polymorphs with different conformers can exhibit molecular packing in the same space group (conformational isomorphism)86 or in a different space group (conformational polymorphism).80a Such variation of molecular arrangements in crystal lattices for a given compound may lead to significant differences in chemical behaviour. The effects of polymorphism on thermal or photochemical behavior cannot be fully understood unless the refined molecular structures of these crystal forms are known. When the molecules that build up the lattice pack differently with different symmetry relationships, they can undergo various intermolecular reactions upon irradiation. Several examples of polymorphs displaying such variations in bimolecular photochemical reactivities in the solid state are recorded in the literature.184 The classic example of polymorphic-dependent photoreactions would be the elegant work of Schmidt and co-workers on the polymorphs of the trans-cinnamic acid system.23 The reactivity differences between the three crystal modifications (trimorphic) of the cinnamic acid derivatives, as described earlier, were explained with the help of X-ray crystal structure analysis as being the outcome of different crystal packing arrangements. An extensive investigation into these systems in fact provided the foundation of the topochemical principle.185  114  Obviously, however, any reactivity differences between conformational polymorphs would be mainly the result of different molecular conformations. Since molecular conformations greatly influence unimolecular reaction mechanisms, irradiations of conformational polymorphs may exhibit dramatically different unimolecular reactivity processes. In such systems, therefore, one can develop an understanding of the effects of conformation on chemical reactivity. A striking example of conformational polymorphs dictating unimolecular reactivities is found in the divergent photo-behaviour of the two crystal modifications of the twenty-six membered ring diketone cyclohexacosane-1,14-dione (14),186 as described earlier in Chapter III. The plates and needle crystals of diketone 14, upon irradiation at room temperature, underwent stereoselective cyclization to yield predominant formation of cis (14c) and trans (14t) cyclobutanol derivatives respectively. The X-ray crystal structure analysis confirmed the existence of the conformational polymorphism. As shown in Chapter III, significantly different molecular conformations are packed with space group P21/n in plates and with space group P21/c in needles. The ORTEP packing stereodiagrams of the plate and the needle dimorphs and their corresponding cell parameters are shown in Figure 56. Only a few examples of conformational polymorphs displaying substantially different unimolecular reactivity are reported in the literature.57d, 167,168,169 To our knowledge, only one other example of this type which is accompanied by a complete X-ray structure analysis has been reported,57d and this study was with a-adamantyl-pchloroacetophenone (43). Compound 43 crystallized in two crystal modifications, needles (P21/n) and plates (C2/c). The irradiation of the needle forms gave 74% transcyclobutanol derivative 44 and 26% cis-cyclobutanol derivative 45, whereas irradiation of the plates gave exclusively the trans-cyclobutanol derivative 44 (Figure 57).  115  a)  a = 5.541 (2) A b = 28.372 (2)  A  c = 8.005 (2) A 13= 98.89 (3)° P2iin monoclinic  b)  a = 8.107 (2) A b = 5.526 (2) A  c = 28.274 (3)A 13 = 97.98 (1)° P2ik monoclinic  Figure 56: The ORTEP stereodiagram of unit cell packing of (a) plate (viewed from "a" axis), and (b) Needle (viewed from "b" axis) crystals of diketone 14.  116  r\L  Ar  C2/c biradical 10/, A r  cis -cyclobutanol formation severely hindered CI  P 21/n biradical  trans  z  ^cis  Figure 57: Solid state photochemistry of a-adamantyl-p-chloroacetophenone.  The reactivity differences were explained based on the X-ray crystal structures. In this example, unlike in the twenty-six membered ring diketone, the geometries of the 1,4biradical intermediates generated would be very similar in both crystal modifications (the unabstracted y-hydrogen and the hydroxyl group in the biradical intermediate are cis to each other), except for the orientation of the plane of the aryl ring, and thus it might be expected that least motion closure would lead to cis-cyclobutanol in both cases. However, in the plate crystals, it was suggested that the motions required to form the ciscyclobutanol product are inhibited due to the substantial steric hindrance developed between the aryl group and the bulky adamantyl moiety. In our present study, the plate crystal modification of the twenty-six membered ring diketone exhibited an irreversible solid-solid phase transition behaviour. When the plate crystals were heated slowly on the hot stage of the Fisher-Johns melting point  117  apparatus (1°C min-1), the colourless transparent crystals cracked between 53-58°C, and became microcrystalline and then opaque. The differential scanning calorimetric analysis (DSC) of the plates (Figure 58) indicated the existence of an endothermic (AH = —6 kJ mol-1) solid-solid phase transition at 54°C (in this experiment the temperature of the crystals was raised gradually from 30°C at a rate of 2°C min-1). A broad endotherm corresponding to the melting point was also observed at 69-70°C. The needle crystals on the other hand showed no solid-solid phase transition on DSC analysis, but a sharp endotherm corresponding to the melting point was observed at 69-70°C. The annealed plate crystals (virgin crystals of the plate form which had been taken through the transition temperature or the melting point at least once and then cooled to room temperature) interestingly showed a DSC thermogram identical to that of the needles. The heating DSC thermogram of the annealed plates in the subsequent scans is reproducible regardless of the duration between each scan. Figure 58 represents the DSC thermograms from the heating of the plates (Figure 58a) and needles (annealed plates) (Figure 58b) of diketone 14. Phase transition temperatures (°C), melting points (°C) and heats of transition (kJ mol-1) in parenthesis are included. The X-ray crystal structure analysis of the annealed plate crystals could not be performed, due to their physical nature (rnicrocrystalline and opaque). The powder diffraction patterns obtained for the plate, needle and annealed-plate forms (Figure 59) at room temperature indicate that the crystal structure of the annealedplates is similar to that of needles but different from that of plates. The diffraction patterns of the needles and plates show peaks at quite different angles, whereas the pattern of the annealed-plates is similar to that of the needle form. With regard to the positions of the peaks, the unit cell of the annealed-plates and that of the needles is identical. As expected, needle crystals were grown from a saturated solution of diketone 14 when seeded with annealed plates.  118  20^30  (a)  40  SO^60^70  60  90  100  10  0  FE  -  -10 —  tp  tp 54°C (6.3 kJ mo1-1) mp 70°C (76.7 kJ mo1-1)  Cr  -  -20 —  -  -30 —  -40 —  Mp Temperature (T)  (b)  I  20^30 AO 50^60^70 SO 100 90 10^tliiIIIIIIIIIIII  nip 70°C (76.2 kJ mol-1) -  -30  -AO Temperature (°C)  Figure 58. Differential scanning calorimetry thermograms of (a) plate and (b) needle  crystal modifications of 14. Phase transition temperatures (°C) and heats of transitions (kJ mo1-1) in parenthesis are included.  119  (a)  •  • 1•1•1•1•11•1•1•1•1•1•1•1•1•1•1•1`1•1•1•1•1•1•1•1•1•1•1•1•1`1•1•1•1•1•1•1•1•1•1•1•1•1.1•1•  5.^10.^15.^20.^25.^30.^35.^40.^45.  50.  (b)  • I'1•1•1•1•1•1•1•1•1•1`1•1•1•1•1•1•1•1•1•1•1•1•1•1•1•1•1•1•1•1•1•1•1•1•1•1•1•1•11•1•1•1•1•  5.^10.^15.^20.^25.^30.^35.^40.^45.^50.  (c)  1 1. .1•1•1•1.1.1•1•1•1•1•1•1•1•1•1•1.1•1•1•1•1•1•1•1•1•1•1•1•1•1•1•1•1'1•1•1•1•1•1•1•1'1•1•1•1•  5.^10.^15.^20.^25.^30.^35.^40.^45.^50. ^  20 ^■  Figure 59: The powder diffraction patterns obtained for the (a) plate, (b) needle and (c) annealed plate forms of 14 at room temperature.  120  A solid state Fourier transform infrared (MIR) analysis of the plate crystal modification of &ketone 14 obtained as a function of temperature in KBr matrices showed a significant difference (especially in the fingerprint region 700-1500 cm-1) between the spectra taken below and above the transition point (54°C) as shown in Figure 60. The FTIR spectrum of the needle crystals and a spectrum recorded for diketone 14 in carbon tetrachloride (20°C) are also included for comparison. The spectra at high temperatures were taken using a special KBr pellet holder, as described in the Experimental Section. The temperature of the sample was gradually increased and the spectra were obtained at 20°C, 55°C, 60°C and 65°C. All the sharp bands observed in the spectra of the plate crystals are completely replaced by a new set of sharp bands at the high temperature solid phase, and interestingly, the spectra obtained above the transition point are practically identical to the spectrum of the needles and annealed-plates (major changes observed in the spectra are illustrated in Table XIII, using the frequency of the bands as a function of temperature). The above analysis explains the existence of a single new conformation in the high temperature solid phase, presumably the zigzag conformation observed in the needle crystal forms. For needles and annealed-plates, the spectra remained unchanged at all temperatures. From these results, it seems that, at the transition point, the metastable low temperature solid phase (plate forms) is irreversibly transformed into a high temperature solid phase, and upon cooling, this phase, which is presumably the needle form, remains stable at room temperature. Although dimorphs are usually known to melt at different temperatures,88 in diketone 14, not surprisingly, both plates and needle crystal modifications have identical melting points, and this becomes an additional evidence for the transformation of plate crystal forms into needles at the transition point.  121  (a) 110  K.  40,  1  1  I;  I  ;11! I  10  y l;  Ift  ■4  I.  ■ ^I  66 43 311  -4 PY 114 -yr 3200^2440^  2000  1 C.:43  (b)  1- cm-1 -  14 •  36  122  (c  )  '1  1  1  014  7f  3200  (d)  2400^  .2see  jure. "3  r^I  tee St. '  96 91 ve#4.96.91 07 27 17.2? -  77 63 77.65  1 1^I  67 99  a) c.) co^se 35 p  67.90  40.71 40.71  C 39 17  26.07 29.43 29 43 19 79 19 70  3209  2440 MS^16,0^1 zee^  sea  Figure 60: The solid state FTIR spectra of (a) plates at 20°C, (b) plates at 60°C and (c) needles at 20°C in KBr matrices. d) A solution spectrum of diketone 14 recorded in CC14  123 Table XIII: FTIR band frequencies (cm-1) of annealed plates, needle crystals and low and  high temperature solid phases of plate crystals as a function of temperature.  PLATES (20°C)  PLATES (65°C)  ANNEALED  NEEDLES (20°C)  PLATES (20°C)  1698 1703  1703  1704  1463  1464  1463  1433  1433  1433  1409  1409  1409  1365  1365  1365  1199  1199  1200  1471  1414  1374  1214  1154  Further evidence to support the nature of the high temperature solid phase and the irreversible solid-solid phase transition observed in diketone 14 came from the type II photochemical investigations of both the plate and needle crystal modifications as a function of temperature. Researchers have widely used the type II photoreactions of ketones to probe the steric and electronic microenvironments provided by a variety of ordered media.157,162, 187-189 Since the conformational requirements of the initial y-hydrogen abstractions and the closure or cleavage of the biradical intermediates into products are specific, an  124  investigation of type II product distributions in various phases provides valuable information regarding the order of the media. In fact, these results can be correlated directly to the motional freedom of the molecules provided by the phase. Because the fate of a biradical intermediate following the y-hydrogen abstraction depends on the flexibility of the cavity wall and the free volume present inside the cavity, information regarding the characteristics of the reaction cavities (size, shape and rigidity) occupied by the molecules (electronically excited ketones or the biradical intermediates) can also be obtained from the photochemical results. Several examples of type II photochemical investigations of guest ketones in various liquid-crystalline phases provided by host compounds are recorded in the literature.157,162,188,189 Liquid-crystals are generally known to exhibit various phases, which can provide a variety of constrained environments capable of directing the motions, the conformations and the reactions of the guest ketone molecules, thus influencing the high selectivity differences in the photochemical reactions of these ketones in various phases. The product distributions can therefore be used to measure the influence of the phase order on the course of the reaction. However, only a few examples of type II photochemical investigations on neat ketones in their various solid phases are known. A recent publication on the type II photoreactions of trans-1-(4-pentanoylpheny1)-4pentykyclohexane (46) and trans-1 -hepty1-4-(4-pentanoylphenyl) cyclohexane (47) (Figure 61) in their various solid, liquid-crystalline and isotropic phases clearly illustrates the influence of the phase order on the product distributions.189a In this study, the different phase orders investigated by deuterium NMR spectroscopy and differential scanning calorimetry correlated well with the photochemical results. The high reaction stereoselectivity for diketone 14 at 20° C has been explained as being the result of the motional restraints imposed by the lattice. Thus, as described earlier in Chapter III, a rigid reaction cavity with a very small free volume in this highly  125  anisotropic medium may limit the motions severely, and therefore the least motion closure leads to either cis or trans-cyclobutanol derivatives, depending on the conformation. The relatively large motions involved with the fragmentation processes would correctly account for the formation of very small amounts of cleavage products.  C H3(C (n = 4)^46 (n = 6)^47 Figure 61: trans-1-(4-Pentanoylpheny1)-4-pentylcyclohexane (46) and trans-1 -hepty1-4(4-pentanoylphenyl) cyclohexane (47)  In addition to the reaction at 20°C, irradiations of the two crystal modifications of diketone 14 were conducted at 0°C, 40°C, 60°C, and 65°C. The product percentages are compiled in Table XIV. The product percentages from the annealed plates (20°C) and hexane (20°C) solution are also included for comparison. Irradiation of plate forms above the transition temperature (54°C), at 60°C and 65°C showed a dramatic reversal in the stereoselectivity of cyclobutanol formation. However, throughout the low temperature solid phase (below the transition temperature), the reaction was quite similar to that observed at 20°C (stereoselective formation of the cis-cyclobutanol 14c). At all temperatures, irradiation of the needle forms stereoselectively gave trans-cyclobutanol derivative 14t, but the selectivity was slightly lowered with increasing temperature. The product ratios of the plates above the transition point were comparable to those of the needle crystals. Interestingly, the photo-behaviour of the annealed crystals was identical to that of the needle forms, at all temperatures, and  126  these results correlate well with the DSC and FTIR analyses. This clearly suggests the irreversible nature of the solid-solid phase transition of the plate crystal form into the needle crystal form at the transition temperature (54°C).  Table XIV: Product percentages of diketone 14 from (a) plates, (b) annealed plates, (c) needles and (d) isotropic media (hexane) as a function of temperature.  , temp (°C)  a  b  c  medium  cis (%)  trans (%)  cleavage (%)  00  Plates  99  01  00  20  Plates  97  03  00  40  Plates  96  04  00  60  Plates  19  81  00  65  Plates  20  76  04  20  Annealed Plates  09  91  00  00  Needles  07  93  00  20  Needles  09  91  00  40  Needles  11  89  00  60  Needles  18  82  00  65  Needle  17  83  00  Solution (hexane)  14  33  53  20  The high stereoselectivity above the transition temperature, although slightly lower than that observed below the transition point, indicates the anisotropic nature of the high temperature solid phase. Presumably the characteristics of the reaction cavities above the transition points may resemble those of the low temperature solid phase. However,  127  inherently larger vibrational and rotational motions of the molecules at high temperature might be the reason for the slightly lowered selectivity. Several years ago Schmidt et al.25 observed a similar irreversible solid-solid phase transformation in the trans-cinnamic acid system and investigated the nature of the transformation with the help of characteristic phase-dependent photobehaviour and powder pattern analysis. Although the photoreaction involved was an intermolecular [2+2] dimerization, the phase-dependent photobehaviour is somewhat similar to that observed in diketone 14. The trans-cinnamic acid underwent an irreversible solid-solid phase transformation around 50°C from the metastable [3-form into a stable a-form; the transformation, however, seems to be quite slow compared to that of diketone 14. As we have described in the Introduction, room temperature irradiation of the a-form causes dimerization to a-truxillic acid and of the 13 form to 13-truxinic acid. However, the type of -  product formed from the 13-form depends on the temperature, due to its phase transformation behaviour. At temperatures higher than —50°C, irradiation of [3-form gave both a-truxillic and 0-truxinic acids. The product ratios were found to depend on the temperature and the duration of exposure of the 0-form crystals to that temperature, as illustrated in Table XV. As suggested by Schmidt, a-truxillic acid is formed via the 13-3a phase change at temperatures higher than 50°C. In an attempt to examine the molecular motions involved in the high temperature solid phase, Fyfe et al.192 have investigated diketone 14, in its various forms, using 13C cross-polarization 'magic angle' spinning experiments (13C CP/MAS). The 13C CP/MAS spectral results have been analyzed and provided by them. Figure 62 (a and b) represents the 13C CP/MAS spectra obtained for the plates at 27°C and at 57°C (below and above the transition temperature). For comparison, the solution 13C NMR spectrum recorded in hexane is included in Figure 62c.  128  It has been indicated that the large spinning side band (-200 ppm, Figures 62a and 62b) pattern for the carbonyl region is a result of the large chemical shift anisotropy (CSA), which denotes the static nature of the carbonyl groups in both the low and high temperature solid phases. Table XV: Photochemical results of p -form cinnamic acid as a function of temperature. Temperature (°C) Irradiation time (hr)  a-form (%)  13-form (%)  25  100  100  00  45  09  100  00  51  50  50  50  64  02  50  50  64  04  00  100  ^ I ^ IN  I II  I^7^1^If -1- ,^,  2119^ PPM  IN  129  .1•2!": PPP^ V  (c)  Ma 220  ^  200  ^  120  ^  150  ^  140  ^  120  ^  100  ^  SISX112  20  Figure 62: 13C CP/MAS spectra of plates a) at 27°C and b) at 57°C, c) 13C NMR spectrum of diketone 14 in hexane. The annealed plates and needle forms also show large CSA for the carbonyl region at 27°C. The restricted motions clearly explain the anisotropic nature of the low and high temperature solid phases. These results correlate well with the photochemical results (high stereoselectivity of the cyclobutanol formation) observed in both the low and high temperature solid phases. Interestingly, the patterns of the peaks corresponding to the methylene region of the spectra, although complicated, look the same in all situations except in the case of the plate form at 27°C (below the transition point), which is quite different from the others (Figure 63). This is additional evidence that the plate form undergoes irreversible transformation into the stable needle form at the transition point.  130  80  40  0  20  40  60  0  20  (d)  PPrn  80  60  40  20  0  PPm 80^60  40^20  0  Figure 63: The aliphatic region of diketone 14 in 13C CP/MAS spectra from (a) plates at 27°C, (b) plates at 57°C, (c) annealed plates at 27°C and (d) needles at 27°C.  131  The sixteen membered ring diametric diketone (9), eighteen membered ring diametric diketone (10) and sixteen membered non-diametric diketone (15) also exhibited enantiotropic solid-solid phase transitions, but unlike the one observed in diketone 14, these phase transitions were reversible in nature. Upon heating, a low temperature stable phase was transformed into a high temperature metastable phase, and upon standing at room temperature the high temperature phase reverted back to the low temperature stable phase. In all cases, when the transparent crystals were gradually heated from room temperature, they became microcrystalline as they passed through the transition temperature and then became opaque. The DSC thermograms shown in Figure 64 illustrate the heating of diketones 9, 10 and 15 at 2°C min-1. The phase transition temperatures (°C) and heats of transition (kJmo1-1) in parentheses are also shown. The phase behaviour of diketones 9 and 10 were reported by Dale et al., but our reproducible transition temperature of diketone 9 (34°C) measured from DSC did not correspond exactly with the published value (28°C).103 Strikingly, diketone 15 has two transition points below the melting point, whereas the others have only one. The heating thermograms in Figure 64 can be reproduced with the same endothermic heats of transition in subsequent scans, irrespective of the time lag between each scan. The heating curves of the annealed crystals were also not dependent on whether the virgin crystal was heated through the transition point or through the melting point. This indicates the reversible nature of these solid-solid phase transitions. The X-ray single crystal structure analysis of the annealed crystals could not be performed, for the same reason mentioned with regard to diketone 14. The infrared spectra of crystals of diketone 9 were recorded in KBr matrices at various temperatures above and below the transition temperature (20°C, 35°C, 40°C, 50°C, 60°C, 70°C).  132  (a)  6  20 30 40 60 60 70 80 90 100 1^1^i^1^1^1^1^1^1^1^1^1^1^1^1^1  tp - 34°C (6.3 la moi-1) mp 86°C (11.7 kJ mol-1) -  -10  tp mp  -15  Temperature (°C)  (b)  20 30 40 SO 60 70 SO 90 100 5  0—  tp  tp 86°C (12.1 kJ mol-1) mp 96°C (29.7 kJ mo1-1) -  -  -15—  -20—  mp Temperature (°C)  (C) 5  20 30 40 50 60 70 80 90 100 1^1^1^1^1^1^1^1^1^1^1^1^1^1^ i^1  Vt 4,_  37°C (23.6 kJ mo1-1) tp 55°C (7.8 kJ mo1-1) mp 77°C (8.1 kJ mol-1) -  -10  lst  tp^mp  -  -15  Temperature (°C)  Figure 64: Differential scanning calorimetry thermograms of a) diketone 9, b) diketone 10, and c) diketone 15. Phase transition temperatures (°C) and the corresponding heats of transitions in paranthesis (kJ moll) are included.  133  The spectra of the low temperature solid phases are significantly different from those of the high temperature solid phases, especially in the fmgerprint region (7001500 cm-1). As the temperature of the crystals passed through the transition point, the bands became broader and the spectra, which became practically identical with that obtained in solution, remained unchanged throughout the high temperature solid phase. The FTIR spectra of diketone 9 obtained at 20°C, 35°C, 40°C are shown in Figure 65. The solution spectrum recorded in carbon tetrachloride at 20°C is also included for comparison. Dale103 observed a similar broadening of the sharp bands for diketone 9 (above the transition point) and suggested a more or less flexible low temperature solid state conformation,103 but he did not rule out the possibility of contributions from other low energy conformations,99a a situation generally accepted for isotropic liquid media. The pattern of the spectra obtained at 35°C, slightly above the transition point, indicates a situation between the low and high temperature solid phase. This rules out an instant phase transition at the transition temperature 34°C. Unlike the diketone 14, the annealed crystals of diketones 9, 10 and 15 behaved exactly like the virgin crystals in all aspects. Although FTIR analysis reconfirmed the reversible nature of the phase transition, the nature of the metastable high temperature solid phase, which produced an infrared spectrum similar to that of the isotropic medium (20°C), intrigued us into further analysis. To explore this, a detailed temperature dependent photochemical investigation of these diketones was undertaken. Virgin crystals were photolyzed as a function of the temperature above and below the solid-solid phase transition temperatures. Irradiations were also performed in the isotropic medium (all reactions were conducted in hexane, except the irradiation of diketone 10 at 90°C, which was carried out in octane). The product percentages, as a function of medium and temperature, are shown in Table XVI.  ^  134  (a)  j . ^;ri^1!;; T';' 11 1 ,:. , 1,^li II!^!III !I' l' '^1.r.  '111.T:' I^'; ri, lir,^Illr II,I 1,^1 Ill^11'11 llil 1111111,1 11,11  : r —1 I  10  ze  I; IIH'I ':; rI I!,^;;;, ,,,^.  ,,^I,^,  ,^,,, atio(1., '' ;,  '  ■^.^,^! ,9'.1. '11 , i 1  I 1  1  , '^ I 1 ^ I^ 1  '^.'  ', ,  I^  I i  ! i  !  ,  if  ,^ ; il 11^.  11 I 1 I I! I, ! :  1 !I  I '!  ;1,',^II 1, 1 t.,I, !^...1 ,! ;^I, 1;^, ,fr  ;  1  I  77 I , 1^'^., I, 1^i,  ,^I^I, ,' ,,op,,I^II! ,  ,,,,f 1^i; ,.  I I^1 ' ^I  1 I^'I III^■ It^I  ; I ! '! 1! : 1  i I 1 I^ ! !! '^ I I^I !^ 1 1  1 ^; 1  1. 1 ^t  I t !,^I 1 I^! I ! ,1!^,', ,^1 I,' !^I! ;!^I^1 r III, I ,^ I lc,' :II 1 !I'il II I^ 1 I^lr III II' !^1 I I ^ .^ i 1 !!!^I ! !^ ! I "^I ; , „,^, •  ''  .  1 :  !  I I  , ! 1 1  1  ;  I  ,  (b)  44'  ,  1  !  ui i^!^i  ,  ,Iii i^' r  24rea^at**  41:.„, ,,,  , ^.  ! '  1 ^.  1600  i  1  I-  ri  li,^,^1. :,^; :  ,-  .^ ,  i^ . „  t til.k  11, i. II : .^ 1 . ,^ I '  II^! HI 1.^l'; II ,'.^I^!III': ;'^'  I I  99  r  , ;^,^. -.1 ,;^' '1 .^, ;^I  I  14 b■■,,,'„6:1^1-  " , ' -  3f  4916  .  ',^, 1.^i , ■^i il^■ 1^'I :r^1, ,i i^'. II^I, 1 [ 1 1^ 111! li,^11! ' H. :1 1 I^; I .^i ' '1^'  r  !^, ]^it .; :r  ! ! 1 I! frif)4S,^I ill I i  i  ima  1 2ilif  i  r  1 .^ 1  .  1  Jill  1  i, 1  , ,  1  , ,i !.' ! 1'  I  1  '11' '^'^11 H Ili! '1 l''''' IH: ', :II'. ,^,1 1^;^.,': 1, 1 ^WI  1 '^;I, , i ,r i^114  'IV ILL,  t ;It': ;It., 'f.1 10^1 It' 71,  'l1', '^11 !II : l' 1,I .' I 11,I ! 1! i' i^l^:! II;  ;'!tr 1i ! '  '^i i. •:, , ;' ',1'''  ,  .: , i-  .  ■ .1  ,-4,00^AMU  lIfIr+—r  ouu #  ,^-f^-r  i  4  -  400  45  135  I  Figure 65: The solid state FTIR spectra of diketone 9 recorded at (a) 20°C, (b) 35°C, (c) 40°C in KBr matrices. (d) A solution spectrum of diketone 9 recorded in CC14.  136  Table XVI: Product percentages of diketones (a) 9, (b) 10, and (c) 15 as a function of temperature and medium. (a)  Temp (°C)  SOLID STATE cis(%)^trans(%)^cleavage(%)  HEXANE cis(%)^trans(%)^cleavage(%)  00  89  08  03  10  90  07  03  20  89  10  01  22  35  43  30  47  16  37  21  32  47  40  17  20  63  19  30  51  60  20  21  59  18  28  54  86  10  04  20 (Annealed) (b) Temp(°C)  SOLID STATE cis(%)^trans(%)^cleavage(%)  HEXANE cis(%)^trans(%)^cleavage(%)  20  03  84  13  17  42  41  30  03  82  15  24  38  38  40  03  81  16  23  36  41  60  04  79  17  23  32  45  80  06  75  19  22  31  47  90  13  49  38  21  30  49  03  81  16  Annealed 20  137  (c)  ,  medium  temp (°C)  cis cl  cis c2  trans^ti + t2  cleavage el+e2  HEXANE  30  13  13  35  39  CRYSTAL  30  98  02  00  00  40  20  09  33  38  60  14  13  26  47  30  98  02  00  00  Annealed  The highly stereoselective cyclobutanol formation observed at room temperature persisted throughout the low temperature solid phases, but interestingly, above the transition points, the stereoselectivity was completely lost and the ene-dione cleavage products predominated accompanied by lesser amounts of cis and trans-cyclobutanol derivatives. The product distributions in the high temperature solid phases are comparable to those in the isotropic medium. Almost constant and temperature-independent product ratios were observed in solution at all temperatures, as observed for diketone 14. Furthermore, the irradiation times required to achieve equal percents of conversions were qualitatively much shorter in the isotropic and high temperature solid phases than in the low temperature solid phases. The influence of the phase on the product distribution seem to be negligible in high temperature solid phases, whereas it is significant in the low temperature solid phases. In the case of diketone 15, only a slight variation in the product ratios could be seen between the two high temperature solid phases; still, the selectivity observed at room temperature remained unchanged up to the first transition point, but beyond this temperature, the selectivity was reduced and the results were quite close to those in solution. In the high temperature solid phase the anisotropic packing order must be  138  largely absent. It appears that the high temperature solid phases mimic the situation in solution. The FTIR results observed for the high temperature solid phase of diketone 9 further support this result. The high stereoselectivity of the cyclobutanol formation in the low temperature solid phases is in fact somewhat similar to that observed for diketone 14. Thus, a similar argument would correctly explain the selectivity. However, the loss of stereoselectivity and the large amounts of fragmentation product in the high temperature solid phases indicates the existence of larger motional freedom. The reaction cavities in the high temperature solid phases must have very flexible cavity walls with large free volume, especially around the reaction center (or the biradical moiety) of the diketone. When large free volumes are available in the reaction cavity, conformational isomerization of the biradical intermediates or the electronically excited diketone is possible. It is generally accepted that the cisoid or gauche 1,4-biradical intermediates afford larger amounts of cyclization products, and the transoid or anti biradical intermediates lead exclusively to elimination products. In the case of the cyclic ketones, the gauche biradical intermediates are topochemically favoured, and therefore, in an anisotropic medium such as the low temperature solid phases, the biradical intermediate formed would prefer to close rather than to cleave, and this was observed experimentally. But in high temperature solid phases, owing to the isotropic nature of the medium, the flexible reaction cavity wall and the large free volume may allow even the gauche biradical intermediate to undergo larger conformational changes, which is generally accepted to be necessary for the fragmentation process. Our insights into the motions involved in the high temperature solid phases of diketones 9 and 10 were investigated by Fyfe et 0.192 using 13C CP/MAS and wide-line deuterium NMR (2H NMR) spectroscopy (the ketones fully deuterated in the methylene positions a to the carbonyl groups). All solid state spectra and their analysis were provided by Fyfe et al.  139  Figure 66 represents the 13C CP/MAS spectra obtained for diketone 9 at 27°C (below the transition temperature) and at 37°C (above the transition temperature). The 13C NMR spectrum obtained in an isotropic medium (CDC13) is also included for comparison. In Figure 66a, from the large CSA (— 200 ppm) associated with the carbonyl signals, it is apparent that the carbonyl groups are either static or their motions are largely restricted at 27°C. A similar large spinning side band pattern is observed at lower temperatures. In the methylene region (the expanded methylene region is shown in Figure 67), the sharp signals with CSA close to 100 ppm suggest that the methylene carbons corresponding to these peaks (likely the a carbon atoms) have very limited motions. Interestingly, just above the transition temperature at 37°C (Figure 66b), the spectrum is significantly different from the one obtained at 27°C, but as expected, it looks quite similar to the 13C NMR obtained in the isotropic medium (Figure 66c). The four sharp signals in the methylene region (as observed in the isotropic medium) correspond to the four different methylene carbons present in an individual molecule. At 37°C the presence of a single carbonyl resonance, as opposed to the two peaks observed at 27°C (consistent with the pseudo C2 molecular axis of symmetry of the solid state conformation), together with the four sharp peaks in the methylene region, indicates the high molecular symmetry, and is due presumably to either the molecular motions or the presence of an inherently highly symmetrical molecular conformation above the transition point. From the reduced spinning side band pattern (-70 ppm) of the carbonyl groups at 37°C (Figure 70b), it has been concluded that they are no longer static, but the finite width of the side bands rules out any isotropic motion.  140  (a)  (b)  r  E, et N  !  141  (c)  •!"^:^■ tPle^V IMMO 104-'  Ppm 220  200^180^160^140^120^100^80^60^40^20^0  Figure 66: 13C CP/MAS spectra of diketone 9 at (a) 27°C, (b) 37°C, and c) a solution 13C NMR recorded in CDC13.  1^  1^•1^1  BO^70^68^58^40^30^20^10^S^—18^—20^-38 PPM  Figure 67: The expanded methylene region of diketone 9 at 27°C (13C CP/MAS).  142  In order to obtain a more detailed picture of the order experienced by the diketone, solid state 2H NMR studies were performed on the a-deuterated ketone 9-d8.  192a  The  deuterium NMR spectra of a sample fully deuterated at the a positions with respect to the carbonyls were obtained at 27°C and 37°C (Figure 68). a)  b) 4  t^  4  1^111 ^1^1^1^1^1^1^1^,^I^-r^I^I^v I^I^i^I^T^ 1 MUM^SIMS^11^-50111111^-1555,5 N ER 7 Z  Figure 68: 2H NMR spectrum of diketone 9 (a) at 37°C and (b) at 27°C.  ^  1  143  The observed quadrupolar splitting provides information on the restrictions that phase order imposes on the motions of the C—D bonds. 192b The classic Pake powder pattern with the quadrupolar splitting (Av – 120 kHz) close to the value of a rigid molecule (127.5 kHz193) observed at 27°C (Figure 68b) and a narrower peak in the high temperature solid phase at 37°C (Figure 68a) have been explained as indicating the fairly static nature of the C—D bonds at the a-methylene groups at 27°C, compared to the high temperature phase. The magnitude of the quadrupolar splitting also suggests that the C—D bonds experience a large degree of phase order. The above two experiments on diketone 9 clearly indicate that the carbonyl groups and the a methylene C—D bonds exhibit a large degree of motion in the high temperature solid phase, compared to that in the low temperature solid phase at 27°C. A 180° ring ffip or internal bond rotations of the methylene groups have been suggested to explain the type of motions involved. These results, correlate well with the observed photochemical results. The large motional freedom observed in the high temperature solid phases, presumably due to the highly flexible nature of the reaction cavity with large free volume, clearly explains the solution-like photobehaviour. As observed in diketone 14, the anisotropic nature of the low temperature solid phase would favour the least motion pathway to afford the stereoselective cyclization. However, in the case of diketone 9 in the low temperature solid phase just below the transition point (30°C), the product distribution observed does not represent the typical ratios of either the low temperature or the high temperature solid phases. The product ratios indicate that limited molecular motions are available, but which are still not comparable to the motions available in the high temperature solid phase or the isotropic medium. The formation of large amounts of the presumably more strained ciscyclobutanol, when compared to its diastereomer, the trans-cyclobutanol, resembles a  144  situation in between the low and high temperature solid phases. Interestingly, the 13C CP/MAS spectrum obtained at 32°C (slightly below the transition point) suggests the presence of small amounts of both phases (Figure 69). Although the photochemical investigation of diketone 9 was performed at 30°C, the temperature variation of ±2°C probably would satisfactorily explain the product distribution observed at 30°C.  I  321^'31O^281^261^241^221^211^III^IBS^lid^121^lie PPM  Figure 69: 13C CP/MAS spectrum of the carbonyl region of diketone 9 at 32°C. 13C CP/MAS (Figure 70) analyses of diketone 10 above and below the transition point reveal that the carbonyl group remains conformationally static in both the low and high temperature solid phases; however, the spectra indicate that the methylene chains undergo rotational motions in the high temperature solid phase. As shown in the 13C CP/MAS spectra (Figure 70), the spinning side band pattern of the carbonyl group remains unchanged in both phases, which indicates that they are not involved in the molecular dynamics of the high temperature solid phase. At 92°C, the slightly lower intensity of the spinning side bands of the methylene carbons compared to that observed at  145  27°C, has been suggested as an indication of slightly motionally averaged methylene groups in the high temperature solid phase. (a)  CSA - 200 ppm  methylene side bands  311-^241  201^t51 PPM  1  50  -SO  (b)  CSA - 200 ppm  methylene side bands  "^  '^1^1^1^'^1^ 1  1 300^250^211^1 SO^III^5^ PPoi  —SO  ^  146  (c)  741 02111401 XL -300 3c OBSERVE 7473 MA SE SECUDICE 100130 DATE 111-05-61 500 005? CDCL 3 FILE C  ^30000  0  ACOu1SITIOM^DEC • vl TN^13 750 DM^I 750 DO^350 .1 0 400 Du^TV, W^16000 DuM^S WY^12 0 OW^6700 TI^0 MO^M Di^0 400 OLP ...." 07  :7  TO^7000 IMIOCESSIMG W 1000 SE^0414 Cl^1000 LO^i 2230 ^ ISO^II 0 EN^05336 IS^500 IMAIN^6 SS^0 IL^hi^DISPLAY IN^1^SP^-703 7 ON^r^V^200000 MS^WI^VS^200 SC^,,,,,,  VC^400 IS^500 WI.^6511 a NET^51106 0 Tor^131 INS^1 000 DC 1•4  1 240  1 I^ ^' '•••I 220^200^lao^I0^ I 140^120^100^80  I•••••"• 1•"•••• '1." •^' 60^40^20^0 PPM  Figure 70: 13C CP/MAS spectra of diketone 10(a) at 27°C and (b) 92°C. (c) A solution 13C NMR spectra of diketone 10 recorded in CDC13. The 2H NMR spectra of the oc-deutrated diketone 10-c18 (Figure 71) indicate that the C-D bonds exhibit larger motions above 82°C, even though the transition temperature is 86°C. This discrepancy may again be due to the temperature gradient. The motional freedom observed in the high temperature solid phase can correctly account for the photochemical results. Diketone 10 seems to exhibit substantial motions in the high temperature solid phase, especially in the methylene region of the molecules. Although solid state NMR spectroscopic analysis of diketone 15 was not performed, in the high temperature solid phases one can presume a situation similar to that observed in diketones 9 and 10. Interestingly, in diketones 9 and 10, the solid state spectroscopic analysis performed for the annealed crystals gave the same results as obtained in the virgin crystals; this provides an additional piece of evidence for the reversible nature of the solid-solid phase transitions in these compounds.  147  a)  b)  I^II^II^1^I^i^1^1-^7^1^1  ^1611800^HOBO^e WERT I  i  -511000  11  1  111^II  -1118188  Figure 71: 2H NMR spectra of diketone 10 (a) at 82°C and (b) at 27°C.  The solid state spectroscopic (13C CP/MAS and 2H NMR) results of dilcetones 9, 10 and 14 correlate well with the photochemical results, as far as the motional freedom  148  CHAPTER V 5.0. The Photochemistry of Alkylated Cyclic Diketones. Irradiation of the tetramethylated sixteen membered ring diketone 2R*,8S*,10R*,16S*.4etramethylcyclohexadecane-1,9-dione (31) and the twenty-four membered ring diketone 2R*,12S*,14R*,24S*-tetramethylcyclotetracosane-1,13-dione (32) in the solid state and in hexane yielded products derived exclusively from the Norrish  type I reaction.10,62 Even though the non-alkylated counterparts, the sixteen (9) and twenty-four (10) membered ring diketones, upon irradiation, undergo efficient intramolecular y-hydrogen abstraction (type II reaction), the substitution of one of the two a-hydrogen atoms by a methyl group (at least four methyl groups in each molecule) completely changes the reaction pathway and leads to photoepimerization.194 As depicted in Figure 72, the photochemistry of the tetramethylated diketones involves an initial a-cleavage generating a type I biradical intermediate (31' or 32'), followed by a reclosure with inversion at the a-carbon atom.  0  )).1  H IIII^. • IPIIIIH  H111^H (C H2) ^(C  hv  (CI-) ^(C142)  (CN)1)^(CHA  H Hi^til1H  n . 5 (31)  (31')  (31a)  n = 9 (32)  (32')  (32a)  Figure 72: Photoepimerization of tetramethylated diketones in the solid state.  149  Unexpectedly, the stereochemistry of diketones 31 and 32 was the one in which all four methyl groups are positioned on the same side of the ring. The ORTEP stereodiagrams of both diketones obtained by X-ray crystallography are shown in Figure 73. In both alkylated diketones, the solid state conformations have rectangular [3x3x] carbon frames, as observed for the corresponding non-aLkylated diketones. The 7-hydrogen atoms with short 0—H contacts are indicated by dotted lines. The 7-hydrogen atoms of diketone 31 H4, H4' (d = 2.89 A, co = 61.5°, A = 63.8°, 0 = 116.6°), H8, H8' (d = 2.71 A, co = 56.8°, A = 78.7°, 0 = 112.5°) and diketone 32 H16,H16' (d = 2.64 A, co = 55.0°, A 83.6°, 0 = 117.2°) are ideally situated for abstraction by the carbonyl oxygen atoms (as observed in the other non-alkylated diketones); however, no type II products were obtained from solid state irradiation. Solid state photolysis of diketone 31 to –30% conversion afforded only one major product, diketone 31a (2R*,8R*,10R*,16S*4etramethylcyclohexadecane-1,9-dione), whereas in hexane, two major photoproducts at retention times 12.3 mins (diketone 31a) and 13.1 mins (product not isolated), were observed on GLC in a ratio of 6:4, with a minor product 31b (RT 10.8 mins). The solid and solution photolysis of diketone 32 to –30% conversion gave only one major product, diketone 32a (2R*, 12R*, 14R*, 24S*tetramethylcyclo- tetracosane- 1,9-dione). On GLC, diketone 32a has a retention time identical to that of the starting diketone 32, and the reaction progress was therefore monitored by analytical HPLC. The structures of the photoproducts were determined by spectroscopic analysis, mainly by comparing their spectra with those of the starting diketones. The 13C NMR and APT spectra of diketones 31 and 32 indicate the presence of four equivalent methyl (16.7 ppm) and methine (44.6 ppm) carbon atoms. The spectral analysis of photoproducts 31a and 31b indicates that both are diastereomers of the starting diketone 31. 13C NMR and APT analysis of 31a reveal that all four methyl and  150  methine groups are in different chemical environments. The only possible structure with four non-equivalent methyl and methine carbon atoms is that of 31a (Figure 72). Photoproduct 31b has all four methyl (16.5 ppm) and methine (45.6 ppm) carbon atoms in the same chemical environment as does the starting diketone, but the chemical shifts are slightly different. Possible structures for diketone 31b are 50 or 51 or 52 (Figure 77), but the stereochemistry of 31b was not determined. As previously mentioned in Chapter III, the photochemistry of non-alkylated medium and large ring cycloalkanones is dominated by y-hydrogen abstraction as the primary photochemical process for both the 1(n n*) and 3(n 70`) excited states. Concomitantly, the products resulting from primary a-cleavage photochemical reactions are virtually absent. In contrast, for small rings up to cyclooctanone, the dominant primary process in solution photolysis is a-cleavage,195 rather than 'y-hydrogen abstraction. Irradiation of cyclononanone, however, affords only the reduction product.130 The preferred a-cleavage reactions of small ring systems have been studied extensively with respect to both the primary process196,197 as well as the factors that influence the behaviour of the acyl alkyl biradical intermediate. 198,199 The evidence indicates that the triplet state is more reactive than the singlet state and that the rate constant for a-cleavage increases with the relief of ring strain and the presence of radical stabilizing a-substituents. 196,200 Cyclododecanone undergoes smooth 'y-hydrogen abstraction upon irradiation in cyclohexane, whereas 2-methylcyclododecanone (53) has been shown to undergo both a-cleavage and 'y-hydrogen abstraction to give type I (54 and 55), and type 11 (56, 57, 58 and 59) products respectively; however, the a-cleavage products dominate (Figure 74).201 The preference for a-cleavage over y-hydrogen abstraction was explained by the faster rate of a-cleavage due to the stabilization of the biradical intermediate by the methyl substituent.201 This explanation has been further supported by Turro et al202 in their recent study of trimethylated and tetramethylated cyclododecanones.  151 H15 C9 1114 1113 H2 115 116 119 C6 CT 1110 1111 CIO 1116 1116  (31)  1122  1122  •••  •  HI  •  CI •  C3  • 1123 to Ca 1121  • : •  113 H16'  C13  HZ  C4  115  114 C5  • 1ilt • 116  HT  ••• C6  C6  H4'  1115  CS  •  1114 41.  •  •••^•-■^ton • ilk^•^' 1•,7  1110 1113  1112  •  •  CT  .^•  • (II  •  119  1117 C11  11, •••40 1119  fa,^to C12  H26  • IF,  1118 C14  1116 C14 •  1125  125 1124  1124  (32)  Figure 73: The ORTEP stereodiagrams of the sixteen membered ring diketone 2R*,8S*,10R*,16S*-tetramethy1cyclohexadecane-1,9-dione (31) and the twenty-four membered ring diketone 2R*,12S*,14R*,24S*-tetramethy1cyclotetracosane-1,13-dione (32)  152  The photochemical reactivity of 2,2,12-trimethylcyclododecanone (60) and 2,2,12,12-tetramethylcyclododecanone (61) in cyclohexane is exclusively type I (Figure 75), with regiospecific a-bond cleavage (on the more substituted side) for the trimethylated compound. These results clearly demonstrate that alkyl substitution facilitates the a-cleavage reaction. In another study,203 irradiation of 2phenylcycloalkanones (62) (Figure 76) and 2,n-diphenyl cycloalkanones (n = ring size) with ring sizes ranging from ten to fifteen also yielded a-cleavage reaction products exclusively. In these arylated cyclic ketones, the biradical intermediate formed by acleavage is much more stable than that from 2-methylcyclododecanone 53. Therefore, a faster rate of a-cleavage has been suggested as the factor responsible for the exclusive formation of type I products from 2-phenylcydoalkanones 62 and both type I and type II photoproducts from 2-methylcycloalkanones 53. As described in the Introduction, the products resulting from photochemical a-cleavage are determined by the competing intramolecular reactions (recombination, disproportionation, decarbonylation) of the type I biradical intermediates. 196,197 It has been found that the recombination process of the type I biradical intermediate is generally efficient and often dominates.203 The reclosure reactions cannot be studied in all compounds, since in most cases the reclosure products cannot be differentiated from the starting ketones. Therefore, the quantum yields of these reactions do not necessarily correlate with the efficiency of the primary a-cleavage process . Unlike the alkylated and arylated cyclic ketones studied in the past, our "all cis" diketones appear to be good models for studying reclosure reactions in cyclic compounds. In larger than twelve membered cyclic ketones, such as the alkylated diketones 31 and 32, where 7-hydrogen abstraction is generally favoured, one would expect both type I and type II processes as was observed in the case of 2-methylcyclododecanone. However, the results from diketones 31 and 32 clearly show that the formation of the type I biradical intermediate is much faster than that of type II biradical intermediate. Even in the non-  153  alkylated medium and large membered cyclic diketones, the lack of type I products may be due to the rapid reclosure of the type I biradical intermediate rather than to the complete absence of a-cleavage primary processes.  0^0\\ CH^ZZI vCH  L  1—°E1 liirl/ jjr+ H  hv  (53)^(54)^(55)^(56)^(57)^(58)^  (59)  Figure 74: Type I and type II products from the photolysis of 2-methylcyclododecanone.  hv (CF  (CI-  F12)7  (R H)^( 60 ) (R = CH3) (61)  (CH2)7^s (C H2C7  Figure75: Photolysis of tri- and tetramethylcyclododecanone.  154  +  62 (n = 10, 11, 12, 15)  Figure 76: Photolysis of 2-diphenylcycloalkanones.  Interestingly, a preliminary photochemical investigation of (R*, S*)-2,12cyclododecanone205 (63) (both methyl groups are on the same side of the ring) in the solid state yielded a single major product, as observed by GLC. This compound has been identified as the type I photoepimerization product by co-injecting the authentic sample (R*, R*)-2,12-cyclododecanone.205 (64) on GLC. However, the solution photoreaction  of (R*, S*)-2,12-cyclododecanone (63) in hexane showed the formation of seven major products (including the solid state product) on GLC; the, isolation and characterization of these products has yet to be achieved. The above experiment clearly suggests that the fate of the biradical intermediate in the solid state is dominated by rapid biradical recombination rather than by rearrangements to other type I products. Irradiation of diketone 31 (RT 11.3 mins) in the solid state for prolonged periods resulted in the appearance of three new peaks in addition to that of 31a (RT 12.3 mins), on GLC at retention times of 10.8, 13.1 and 13.3 mins. After —60 hours of irradiation the four products, along with the starting material, attained a photostationary state. Interestingly, irradiation of the photoproduct 31a also led to a similar product distribution. The photolysis of diketone 31 in cyclohexane also led to a photostationary state after —70 hours of irradiation; however, the photoproduct corresponding to the peak at RT 13.3 min was absent. The ratios of the photoproducts (%) at the photostationary state and their retention times on GLC are given below.  155  Solid state irradiation of 31: RT's 10.8 (12 %), 11.3 (6 %), 12.3 (67 %), 13.1 (4 %), 13.3 (11 %). Solid state irradiation of 31a: RT's 10.8 (14 %), 11.3 (7 %), 12.3 (63 %), 13.1 (5 %), 13.3 (11 %). Irradiation of 31 in cyclohexane: RT's^10.8 (3 %), 11.3 (60 %), 12.3 (20 %), 13.1 (17 %). Apart from 31b, the new photoproducts observed could not be isolated from the final reaction mixture and characterized. Spectral analysis of the product mixture indicated the absence of possible photoproducts from the type II reaction. In the solid state reaction, as shown in Figure 77, at the photostationary state diketones 31 and 31a may coexist with their diastereomers 50, 51 and 52. The solution product 31b with a retention time of 13.1 min could therefore correspond to diketone 50, 51 or 52.  /50  ct)  i'20 ---=''-  31^  31a  N  ^  51  Ct=IFZ=1-0 52  Figure 77: A diagrammatic representation of possible diastereomers of tetramethylated sixteen membered diketone (31), at the photostationary state.  156  5.1. The Photochemistry of Cyclic Mono - and Diketones in Zeolites. Generally, within a constrained or organized medium, both the photochemical and the photophysical properties of organic molecules can be considerably modified.206,207 In this context, the internal structure of zeolites (pores and cages) has attracted recent attention.208 Zeolites are crystalline aluminosilicates with usually well-defuied structures.209,210 Therefore, they provide an ordered host environment for guest molecules under investigation. The photochemistry of a large number of ketones adsorbed by a variety of commonly available zeolites has been reported in the literature,209,210 and shown to exhibit a wide variation in the product ratios compared to irradiation in isotropic media. The "zeolite method" was applied to the cyclic diketones prepared in our laboratory; and to cyclic monoketones prepared by Turro et al (Department of chemistry, Columbia University, New York) in a collaborative study with V. Ramamurthy (Du Pont Company, Wilmington, DE). All photoreactions in zeolites were performed by Ramamurthy et a/.213 The solution photochemical results of monoketones and diketones discussed in this chapter were performed in pentane and hexane respectively. The irradiation of ten (33), eleven (34), twelve (35), thirteen (36) and fifteen (38) membered cyclic monoketones and sixteen (9), eighteen (10) and twenty (11) membered cyclic diketones included in cation (Li+, Na+, K+, Rb+, and Cs+) exchanged X and Y zeolites (faujasites) gave products derived from both type I and type II processes (Figure 78). The major photoproducts obtained in the zeolites were type II cis and transcyclobutanols (66), cleavage product (67) and type I product alkenal (69). A plot representing the type I to type II photoproduct ratios (type I / type II) in NaX zeolite with respect to hexane (diketones) or pentane (monoketones) as solvent is shown in Figure 79. The type I to type II product ratios of compounds 10 and 36 included in various cationexchanged zeolites are also shown in Figure 80 in the form of bar graph.  157  HO^HO  HO  (CI-12^(CH2)n_2  (ci_on_2  (^142)n-2  X  xJ  xJ  65  66c  66f  \■  142)n-^(^c H3 \  X  Type I hv  67 Type II  H^(042)n^(C142)n+2  monoketones:  y  (33) X = CH2; n = 2. (34) X = CH2CH2; n = 2.  \x/  cH2  68  69  (35) X = CH2; n = 3. (36) X =  CH2CH2; n = 3.  (37) X = CH2CH2;  n = 4.  diketones:  (9)  X = CO; n = 5.  (10) X =  CO; n = 6.  (11) X = CO;  n = 7.  Figure 78: The two modes of photoreaction of mono- and diketones in zeolites.  158  5.0 -a- Pentane/Hexane -a- NaY  ruj 4.0 m ,  8. 3.0  = a,  94 2.0 zz  a 1.0 1.0.0  33^34^35^36  38  9  10  11  ketone used  Figure 79: Enhancement of type I products in zeolite NaY with respect to the solution reaction.  ketone 10  ^  ketone 36  Figure 80: Dependence of type I to type II product ratios of compounds 10 and 36 on cations.  The above results clearly indicate a dramatic enhancement of the type I process within zeolites. From Figure 79 it is apparent that, even though type I product formation is enhanced in every cation-exchanged zeolite, the extent of enhancement is higher with smaller cations such as Li+ or Na+, but with the other cations the formation of type I product is dramatically reduced. It is also necessary to direct our attention to the fact that  159  the ketone ring size dictates the extent of the enhancement of the type I products, and that the effect is greater when the ring is smaller. Variations in type I to type II product ratios have been reported for a variety of acyclic ketones adsorbed on cation-exchanged X and Y zeolites.211-213 These investigations also reveal a significant influence of the cations on the selectivity of the photochemical pathways. At this point, prior to analyzing the results, it may be helpful to give a brief description of the structure of zeolites. The primary building blocks of the zeolites are the fSiO414- and [A104]5- tetrahedra units. These units are linked at all corners to form channels and cages or cavities (Figure 81) of distinct size, in such away that no two aluminium atoms sharing the same oxygen atom. Primary cage structures (sodalite cages) are constructed with four and six membered rings of [SiO414- and [A10405- units. These sodalite cage units can construct "supercages" or "channels" in a number of ways (leading to different types of zeolites), two of which are illustrated in Figure 81. These cages and channels are generally interconnected and have access to exterior through a pore or window which cannot be larger than the channels or cages. The pore dimensions therefore determine the size of the molecule that can be absorbed into these structures. As a result of the difference in charge between the [SiO4]4- and [A104]5- units, the total framework charge of an aluminiumcontaining zeolite is negative and it is generally balanced by alkali or alkaline earth metal cations. The cations and water molecules present are located in the cages, cavities, and channels of the zeolites. The position, size and number of cations, as well as the position and number of water molecules, can significantly alter the properties of the zeolites. The X and Y zeolites, commonly known as "faujasite zeolites", used in these experiments have a 7.4 A pore size (diameter) with a supercage diameter of 12 A. The internal structure of these two zeolites is identical except for the difference in the ratio of  160  aluminium to silicon in the framework (the unit cell compositions of X and Y zeolites are given in Figure 81).  Zeolite X & Y  X type M86(A102)86(Si02)106.264 H20 Y type M56(A102)56(Si02)136.253 H20 Figure 81: Illustration of the [SiO4]4- and [A104}5- tetrahedra that are the primary  building blocks of zeolites. Also shown are representations of the sodalite cage, zeolites (A, X and Y) and the unit cell compositions of X and Y zeolites.  In the case of the large membered diketones, all rectangular conformations [3x3xJ,  are narrow enough to enter the pore (7.4 A) regardless of the ring size, since the short segment of the conformation is only — 6.2 A in width. The largest molecule used in this experiment (twenty membered ring diketone) has a molecular length of 11.3  A, which can  fit perfectly within the super cage (cage diameter --12 A). However, the rings larger than the twenty membered ring are too long to fit in the super cages,and therefore, could not be included. All molecules used for the investigation are expected to fit within a single cage of X and Y zeolites.  161  The supercage of X and Y zeolites contains cations located at three distinct sites as shown in Figure 82.214 The cations at location I are present within the socialite cage, while cations at locations II and III are within the supercages. Since the preferred positions for the guest molecules are supercages, only the cations present within the supercages are expected to influence the reactivity of the guest molecules. The cations at location II are present within the framework of the supercages, whereas the ones at location III occupy the void space of the supercages. Therefore the cations at location III are expected to influence the reactivity of the guest molecules both sterically and electronically.  SodaIle cage, ^ 41 IT Sager cage 11104P1 mg IOW  lipt‘,41.40 SPO  Figure 82: Cation locations inside the faujasite cages.  Turro and co-workers have undertalcen227,228 photochemical investigations of several compounds related to the dibenzyl ketones included in zeolites, where cations were found to have significant influence on the product distribution. In all of these cases, the variations in the product ratios were attributed solely to variations in the free volume of the supercage with respect to the cation. Furthermore, a study by V. Ramamurthy et al. on benzoin alkyl ethers ,212 a-alkyldeoxybenzoins211a, and a-alkylbenzylketones212  162  included in zeolites also illustrated the importance of the size of the cations on the product distribution. Apart from the steric effect, the electronic interaction of these cations on the guest molecules is also known to play a major role in the product distribution in zeolites. The inclusion of guest molecules on X and Y type zeolites has been widely investigated by several spectroscopic techniques.215 These studies indicates that if a molecule with a it-bond or polar functional group is adsorbed on zeolites, it interacts strongly with the exchangeable cations. The interaction energy depends on factors such as the type, radius (ionic) and charge density of the exchangeable cations, as well as the types of zeolites and the structures of the adsorbed molecules. Singly charged cations, such as Li+ and Na+, show stronger interactions than the larger K+, Rb+,and Cs+ cations. Photochemical results similar to those observed for the cyclic ketones have been reported211d for a series of a,a-dialkylphenyl ketones adsorbed on X and Y zeolites. Irradiations of these ketones in benzene and on zeolites (M+ = Li+, Na+, K+, Rb+ and Cs) gave products from both type I and type II processes. It was also observed that the type I process was enhanced significantly within zeolites compared to benzene photolysis. Furthermore, as observed for the cyclic ketones, Li+ and Na+ cations exerted a larger influence than other cations. The above results have been discussed on the basis of the change in the binding ability of the cation with the carbonyl chromophore due to the variation in the electrostatic potential of the cation. A similar argument has been suggested by Ramamurthy213 to explain the results obtained from cyclic mono- and diketones in zeolites. The cations present within the supercage can interact electronically with the carbonyl chromophore and thus severely impede the hydrogen abstraction sterically (Figure 83). The photobehaviour of mono and diketones can be influenced by zeolites, by altering the partitioning of the corresponding reactive intermediates 68 and 65 (Figure 84) between the reverse reaction to starting ketone and decay to products.  163  I H2 C,,,,..,  00  It  1  , /NCH2 H^H  H tl )z.,........ I CH2  HH vx 1 CH2 Type I  --.-----,--0,  I  I FI2 C,. •  H2c,  00  •  C=)  Figure 83: Electronic interaction of lithium cation with the carbonyl chromophore impedes type II hydrogen abstraction sterically.  0  HO type II 4.____  type I -I. 4--  x  ..)^Fl......)  x^X (68)^(69)  / (65) X = CH 2 monoketone X = (C=0) diketone  (66)  Figure 84: Diagram illustrating the partitioning of the type 11 (65) and type I (68) biradical intermediates to products and to their starting ketones.  164  An increased rate of a-cleavage, a decreased rate of hydrogen atom abstraction and an increased rate of intramolecular hydrogen return of 65 should give rise to a greater proportion of type I product, with the last of these being the least likely. Internal hydrogen return should not be favoured over cyclization and fragmentation of the type II 1,4-biradica1,59,216 in a polar medium or a medium containing a large number of oxygens (capable of forming hydrogen bonds), as is the case within the zeolite cavity.217 This suggests that a reduction in the rate of the competing type II process is probably the main cause of the formation of type I products in zeolites. The well established218 fact that molecular motions are restricted within zeolites indicates that reactions involving large segmental motions (type II cleavage) of the reacting molecule will be slowed inside zeolites. This has been supported experimentally by the photochemical investigations of valerophenone (VP) and a,ocdimethylvalerophenone (DMVP)219 in zeolites. The triplet lifetimes of these compounds in various zeolites clearly indicate that the 'y-hydrogen abstraction process is considerably restricted when these ketones are adsorbed on zeolites. For medium and macrocyclic ketones, only a few of the several possible conformational arrangements are suitable for y-hydrogen abstraction. In addition, steric and electronic influences imposed by the cations on guest molecules are expected to restrict conformational interconversions that would be favorable for y-hydrogen abstraction. Therefore, under such conditions, a slow type I process would be able to compete with the type II process. Furthermore, the strong binding ability of Li+ and Na+ cations to the carbonyl chromophore explains the comparatively large effect of these cations. The photochemical study of cyclic ketones in zeolites establishes that type I processes can in fact be induced even in systems that do not show any sign of such reactivity in isotropic media. Such an investigation using zeolites provides an alternate strategy to induce type I processes in cyclic mono- and diketone systems.  165  CHAPTER W 6.0. General information. Melting points (mp): All melting points were taken on a Fisher-Johns hot stage melting  point apparatus and are uncorrected. Transition points (tp): Transition points were obtained from a differential scanning  calorimeter, a Mettler DSC-20 cell interfaced with a Mettler TC 10 TA processor and a Swiss Matrix printer/plotter. Samples (10-15 mg) weighed in an analytical balance were placed in a metal vial with a tiny hole and used for the experiment; the temperature was increased at a rate of 2 °C min- 1. The values reported for AH and temperature were statistical averages of at least three runs. Infrared spectra (IR): The infrared spectra were recorded on a Perkin-Elmer 1710  Fourier transform spectrometer. The positions of the absorption maxima are given in cm-1. Spectra of oily compounds were obtained by placing them between two sodium chloride pellets. Solid samples (2-5 mg) were ground with anhydrous KBr (100-200 mg) and pelleted using a Perkin-Elmer evacuated die 186-0002 and a Carver model B laboratory press at 20,000 psi. Infrared (FTIR) studies of diketones at elevated temperatures: A special apparatus  with a KBr pellet-holder, made by the UBC Electrical Shop, was used for this purpose. A KBr pellet-holder was placed in the middle of an insulated metal coil and the desired temperatures were maintained inside the coil by passing an appropriate current through the coil. The amount of current was controlled by a rheostat. The device was calibrated  166  before use by preparing a standard plot of the temperature inside the coil (measured by a thermometer) versus the rheostat dial reading. Nuclear magnetic resonance spectra (nmr): Proton nuclear magnetic resonance  (1H nmr) spectra were recorded in deuterated chloroform or benzene. The spectrometers used were: a Bruker AC-200 (200 MHz), Varian XL-300 (300 MHz), Bruker WP-400 (400 MHz) and Bruker AMX-500 (500 MHz). Signal positions are reported in units of 5, parts per million downfield from tetramethylsilane (TMS), which was used as an  internal standard. The number of protons, signal multiplicities, coupling constants (in Hz) and assignments are given in parentheses following the signal positions. The abbreviations used to indicate the multiplicities of the proton signals are: s = singlet, br s = broad singlet, t = triplet, q = quartet and m = multiplet. Carbon nuclear magnetic resonance (13C nmr) spectra were recorded on Bruker AC-200 (50.3 MHz), Varian XL-300 (75.4 MHz), Bruker 400 (100.6 MHz) and a Bruker AMX-500 (125.8 MHz) spectrometers using deuterated chloroform or benzene as a solvent and an internal reference (TMS). Spectra signal positions are reported in parts per million (5) relative to the standard, and the signals were assigned based in part on the attached proton test.22° Solid-state deuterium nmr studies and cross polarization magic angle spinning (CPMAS) experiments were recorded on a Bruker MSL-400 spectrometer. Ultraviolet spectra (UV): The UV spectra were recorded on a Perkin-Elmer Lambda-4B  spectrometer. Mass spectra (MS): Both low and high resolution mass spectra were recorded on a  Kratos MS50 mass spectrometer. A Karlo-Erba gas chromatograph coupled with a  167  Kratos MS80 mass spectrometer was used for GC-MS analysis. Relative intensities are recorded as percentages of the most intense peak (base) and are given in parentheses. Microanalysis: Elemental analyses were performed by the departmental microanalyst, Mr. P. Borda. Samples were prepared as follows: Vials containing powdered sample were placed in a special apparatus attached to a container with phosphorus pentoxide, and left on the vacuum line (1-2 mmHg) for 24-48 hr prior to analysis, where the temperature of the samples was maintained using refluxing solvents 15-50°C below their melting points. Gas liquid chromatography (GLC): A Hewlett-Packard 5890A capillary gas  chromatograph attached to a Hewlett-Packard 3392A integrator was used for all gas chromatographic analysis. Fused silica capillary columns (15 m x 0.25 mm) DB-1, DB-17 and carbowax from J&W Scientific, Inc. were used with helium as a carrier gas. Column head pressure was maintained at 15 psi. All retention times reported have limits of ±0.1 min. Thin layer chromatography (TLC): Aluminium sheets coated with 0.2 mm of silica gel  (type 5554) from E.Merck were used for analytical work. Column Chromatography: For purification and separation purposes conventional  column chromatography (gravity) and flash column chromatography221 were performed using silica gel 60, 230-400 mesh (E.Merck) as the stationary phase. A mixture of ethyl acetate and petroleum ether (35-60 °C) solvent system was used as an eluent for most of the compounds. High performance liquid chromatography (HPLC): A Waters 600E system controller  equipped with a Waters 486 UV detector and a Waters fraction collector was used. A  168  Radial-PAK cartridge Oa Porasil, particle size 104), with 8 mm (id) x 100 mm from Millipore (cat # 85720) was used for analytical studies. For preparative scale work, a similar column with an internal diameter of 25 mm (cat # 38504) was used. Crystallographic Analysis: All crystal structures were determined on a Rigaku 4-circle  diffractometer by Dr. Steven J. Rettig and Prof. James Trotter of the UBC Chemistry Department. Powder diffraction patterns: Powder diffraction patterns of the 26-membered ring  diketone were recorded on an X-ray powder diffractometer which uses a Rigaku Rotating Anode X-ray (12 kw rotating anode). Solvents and reagents: Unless otherwise indicated, all reagents were purchased from  Aldrich Chemical Co., and solvents were purchased from BDH Chemicals. Spectral grade solvents were used for photochemical and spectroscopic studies. All diacids and diacid dichlorides were used as received. Valerophenone was purified by reduced pressure distillation followed by recrystallization from pentane. Acetophenone was purified by reduced pressure distillation. Straight chain alkanes tetradecane, tricosane, docosane and tetracosane were purified by recrystallization and used as internal standards for gas chromatographic analysis. Dry solvents and reagents were prepared as follows:222 benzene, xylene and dimethoxyethane (DME) were distilled into a collecting reservoir by heating at reflux over sodium, under a dry nitrogen atmosphere. These solvents were transferred using an ovendried syringe through a stopcock fitted on the reservoir. In some cases solvents were distilled directly into the reaction vessel. Triethylamine was directly distilled and stored over calcium hydride. Deuterated chloroform and benzene used for spectral analysis were stored over Linde 4A molecular sieves.  169  6.1. Synthesis of Starting Materials. 6.1.1. Synthesis of Diametric Diketones by Ozonolysis of Bicyclic Olefins. Cydodecane-1,6-dione (6)  108,109  A solution of 1.5 mL (10 mmol) of 1,2,3,4,5,6,7,8-octahydronapthalene (17) in 40 mL of 3:1 (v/v) MeOH:CH2C12 was placed in a three-necked 100 mL flask and cooled to -30°C in an acetone/dry ice bath. During ozonolysis, the temperature was gradually lowered to -60°C and ozone generated from Welsbach T-23 ozonator was bubbled through the solution. The reaction was monitored by thin layer chromatography (tic). Excess ozone was passed through the solution until the spot on the tic corresponding to the starting material had completly disappeared. The solution was further flushed with oxygen for 10-15 min to remove any traces of ozone and 1.0 mL (13.6 mmol) of dimethylsulphide was added through the middle neck of the flask. The solution was stirred at -10°C for one hour, then at ice bath temperature for one hour and finally at room temperature for one hour. After the removal of solvent in vacuo, the remaining white residue was diluted with water (10 mL) and extracted with diethyl ether (3 x 10 mL). The combined diethyl ether extracts were washed with water, dried over magnesium sulphate and evaporated to dryness to afford a white solid. A major peak found on GLC (DB17, 120°C) at retention time (RT) 7.32 mins (62% by GLC) was later identified as the ten membered ring diametric &ketone 6. Several fractional recrystallizations of the final crude solid from ethyl acetate in petroleum ether solvent system yielded 1.32g (7.8 mmol, 78%) of pure diketone 6 as prisms with mp = 99-100°C(lit. °3 990C). 1NMR: (C6D6, 200 MHz) 62.0-2.1 (8H, m, -CH2-00), 1.6-1.7 (8H, m, methylenes). 13C NMR: (C6D6, 50 MHz) 6 207.4 (C=0), 37.4 (CH), 19.0 (CH2). IR (KBr): 2946, 1688 (C=0) cm-1.  170  MS m/e (rel. intensity): 168 (M+, 5), 150 (31), 111 (36), 97(36), 85(38), 84(56), 83(23), 81(25), 68 (25), 67 (29), 57 (36), 56 (30), 55 (100), 43 (26), 41(44), 39 (21). Calculated mass: 168.1151, found: 168.1153. UV (Cyclohexane) Amax: 292 nm (e, 15) (n-e) The structure of diketone 6 was also determined by X-ray crystallographic analysis. Cyclododecane-1,7-dione (7)109,110  Neutral aluminum oxide (6.5 g) from Fisher Scientific (Brockman activity 1) was dried in the oven at 500 °C for 24 hours and carefully transferred to a 500 mL threenecked round bottom flask containing freshly cut sodium (30 mg) under dry nitrogen (obtained by successively passing the gas through a H2SO4 and then a CaSO4-KOH trap) conditions. The flask was placed in a sand bath and the temperature of the mixture was gradually increased to 200 °C with stirring using a mechanical stirrer. At 200 °C, vigorous stirring was introduced for 20 min. Melted sodium diffused onto the aluminum oxide and formed a fine dark blue powder. After cooling to room temperature, a solution of cyclododeca-1,5,9-triene (18) (7 mL, 38 mmol) in heptane (70 mL) was cautiously added and the mixture was refluxed with stirring for 23 hr, until the GLC analysis (DB17, 95 °C) showed that 85% of the starting material had been consumed. As reported before,110 the final mixture showed five new peaks on GLC. Hydrogenation step :  Refluxing was continued while bubbling hydrogen through the solution until the peaks on the gas chromatogram were simplified into two main peaks with a ratio of 5:1. After 11 hr of reflux the catalyst was filtered off and washed with heptane. Ozonolysis step :  The mixture from the hydrogenation step was subjected to an ozonolysis procedure similar to the one used for diketone 6, with minor modifications, using 60 mL of Me0H, 20 mL of CH2C12 and 4 mL of (CH3)2S. The final mixture with two major  171  peaks on GLC (DB17, 90 °C) at retention times 11.65 and 16.47 min was subjected to flash-column chromatography (silica gel). Elution with 7% (v/v) ethyl acetate in petroleum ether^(35-60°C) gave an oil (RT 11.65 min), a by-product. The peak at retention time^16.47 min was eluted with 9% (v/v) ethyl acetate in petroleum ether as a white crystalline solid (1.32 g, 6.7 mmol, 16.8%) and was later identified as the twelve membered diametric diketone 7. Recrystallization of diketone 7 from the diethyl ether/ petroleum ether solvent system gave needle and prism crystals with mp = 134-136°C (li110 mp = 133-135°C), the IR spectra of these two crystal modifications were found to be identical. 1H NMR: (C6D6, 300 MHz) 5 2.0 (8H, t, J = 7Hz, -CH2-00), 1.3 (8H, m, methylenes),  1.0 (4H, m, methylenes). 13C NMR: (C6D6, 75 MHz) 5 210 (C=0), 40 (CH2), 26 (CH2), 23 (CH2). IR (KBr): 2946, 1699 (C=0) cm-1MS mie (rel.intensity): 196 (Mt, 1), 178 (7), 98 (37), 83 (37), 55 (100). Calculated mass: 196.1464, found :196.1472. UV (Cyclohexane) Xmax: 288 nm (e, 40) (n-R*)  The structure of diketone 7 was confirmed by X-ray crystallographic analysis.  6.1.2. Synthesis of Diametric Diketones by Blomquist's High Dilution Technique. Diketones 8, 9, 10, 11, 12, 13 and 14 were prepared according to the procedure of Blomquist et al.111 The following procedure described for the synthesis of cyclotetradecane-1,8-dione (8) from suberic acid is typical.  172  Preparation of suberoyl chloride from suberic acid:  Ten grams (53 mmol) of suberic acid and slightly more than two equivalents of thionyl chloride (8.5 ml, 116 mmol) were placed in a 100 mL round bottom flask fitted with a magnetic stirring bar and attached to a condenser with a nitrogen inlet. The flask was placed in a water bath and the stirred mixture was refluxed at 55°C until there was no further evolution of hydrogen chloride and sulfur dioxide (4-5 hr). After the reaction was complete, the excess thionyl chloride was removed by vacuum distillation. Trace amounts of thionyl chloride present in the crude acid chloride were removed by refluxing with 3 x 15 mL portions of anhydrous diethyl ether for a few minutes followed by evaporating the ether in vacuo. A pale yellow crude acid chloride (11 mL) was obtained and used for the following cyclization step. Cyclization step (high dilution technique) : The apparatus used in the cyclization step was a 5 litre three-necked round bottomed flask, fittered with an all-glass mechanical stirrer through the central neck. A "Y" tube attached to the second neck had a vertical and an upwardly slanting side arm. An efficient condenser with a drying tube was connected at the upper end of the vertical arm and a nitrogen inlet was attached to the slanting side arm. A dropping funnel was attached at the third neck. Three liters of benzene (freshly distilled over sodium) placed in the 5 litre round bottomed (RB) flask were brought to a gentle reflux under nitrogen atmosphere and —500 mL of benzene was distilled off in order to ensure no residual moisture left inside the flask. Then 85 mL of triethylamine (freshly distilled over calcium hydride) was added using an oven dried syringe and then the mixture was brought to reflux. Crude acid chloride dissolved in 100 mL of benzene was added dropwise from an oven-dried funnel to the above stirred refluxing triethylamine-benzene solution over a period of three hours. After the addition was complete, heating was continued for an hour  173  and the mixture was stirred for another hour while cooling. The mixture was then filtered under suction to remove the precipitated triethylamine hydrochloride. The filtrate was freed of excess triethylamine by washing with water (3 x 25 ml), and concentrated by removing benzene as rapidly as possible in vacuo to yield a reddish-brown residue. Hydrolysis and decarboxylation :  The residue was transferred to a 250 mL round bottom flask to which was carefully added with stirring a solution of potassium hydroxide, made by dissolving potassium hydroxide (10 g) in 50 mL of methanol. The resulting mixture was refluxed on a heating mantle for one hour, and after dilution with 50 mL of water, further refluxing was continued for another two hours. The final reaction mixture was extracted with diethyl ether (3 x 15 mL) and the combined ether extracts were dried over anhydrous magnesium sulphate. After 1-2 hr, the solution was filtered and the solvent was evaporated in vacuo. The final pale-brown colored residue on GLC (DB17, 200°C) gave two major peaks at retention times of 3.41 and 9.86 min which were later identified as peaks corresponding to diketone 8 and a tricyclic y-pyrone 24 respectively. These two compounds were isolated in pure form by flash column chromatography (silica gel, ethyl acetate : petroleum ether = 3:97) , where the y-pyrone has a slightly longer retention time than the diketone. Diketone 8 was eluted as a white crystalline solid (448 mg, 2 mmol, 9.0 % yield). Recrystallization from diethyl ether/ petroleum ether gave plates with mp = 148-149°C (HO 1 1 mp = 146.5-148°C). y-Pyrone 24 was also eluted as a white crystalline solid (1.16 g, 5 mmol, 21% yield). Recrystallization of y-pyrone 24 from ethyl acetate/petroleum ether gave prisms with mp = 108-109°C.  Diketone (8) 1H NMR: (CDC13, 300 MHz) 5 2.4 (8H, t, J = 7 Hz, -CH2-C=0), 1.6 (8H, m, methylenes),  1.2 (8H, m, methylenes).  174  13C NMR: (CDC13, 125 MHz) ö 211.7 (C=0), 41.0, 27.0, 23.1. IR (KBr): 2923,1703 (C=0) cm.-1. MS mile (rel.intensity): 224 (Mt, 4), 206 (16), 167 (14), 139 (14), 125 (26), 113 (46), 112  (100), 111 (38), 97 (40), 84(58), 69(39), 55 (75). Calculated mass: 224.1777, found: 224.1774. UV (Cyclohexane) )t.max: 283.8 nm (c, 41) (nit*)  The structure of diketone 8 was confirmed by X-ray crystallographic analysis. 2,3-5,8 bis(cycloheptano)-4-pyrone (24) 111 NMR: (C6D6, 500 MHz) 5 2.60 (8H, m), 1.85 (4H, m), 1.72 (4H, m), 1.60 (4H, m). 13C NMR: (C6D6, 125 MHz) 5 177.6 (C=0), 166.4 (0-C=C), 124.9 (0=C-C), 34.0 (CH2),  31.8 (CH2), 26.2 (CH2), 24.9 (CH2), 22.4 (CH2). IR (KBr): 2925, 1657 (C=0), 1603 (C=C) cm-1. MS m/e (rel. intensity): 232 (M-I-, 21.1), 204 (33), 203 (23). Calculated mass: 232.1464, found: 232.1462. Anal. calcd for C15H2002: C, 77.55; H, 8.68. found: C, 77.50; H, 8.80.  The structure of y-pyrone 24 was supported by X-ray crystallographic analysis.  Cyclohexadecane-1,9-dione (9)  The procedure used to prepare diketone 8 was modified by using azelaic acid (10 grams, 53 mmol) instead of suberic acid. A final pale-yellow residue showed a single major peak on GLC (DB-17, 180°C) at a retention time of 8.05 min and which was later identified as the peak corresponding to diketone 9. The final mixture was subjected to flash column chromatography (silica gel) and elution with 4% ethyl acetate in petroleum ether (35-60°C) yielded 1.9 g (7.4 mmol, 28% yield) of diketone 9 as a white solid. Recrystallization of diketone 9 from ethyl acetate/petroleum ether gave prisms with mp = 85-86°C (lit103 mp = 83-85°C). Virgin crystals and the crystals annealed to 50°C  175  (3 cycles) were analyzed by differential scanning calorimetry and found to have an endothermic phase transition at 34°C ^= 23 kJmo1-1) (lit103 transition point = 28°C). 1H NMR: (CDC13, 300 MHz) 8 2.4 (8H, t, J = 7, -CH2-C=0), 1.7 (8H, m, methylenes),  1.3 (12H, m, methylenes). 13C NMR: (CDC13, 75 MHz) 8 212.4 (C=0), 41.9, 27.9, 27.6, 23.6. CPMAS 13C NMR: (100 MHz) 8 213.5 and 212.3 (2 C=0), 46.0 (CH2), 44.4 (CH2), 35.5  (CH2), 29.4 (CH2), 22.8 (CH2). IR (KBr): 2979, 1703 (C=0) cm-1.  MS mie (rel.intensity): 252 (Mt, 3), 98 (34), 83 (27), 55 (100), 43 (44), 41(66). Calculated mass: 252.2090, found: 252.2092. UV: (cyclohexane) Amax: 286 nm (e, 41) (nn*).  The structure of diketone 9 was confirmed by X-ray crystallographic analysis. Cyclooctadecane-1,10-dione (10)  Eighteen membered ring diametric diketone 10 was prepared from sebacic acid (15 g, 74 mmol) using the general procedure described for the synthesis of diketone 8. Most of the diketone 10 was precipitated out (with some impurities) from the concentrated solution of the final mixture as a pale yellow mass(-90% by GLC - DB 1, 200°C). Several fractional recrystallizations from ethyl acetate/petroleum ether gave pure diketone as white prisms with mp = 95-97°C (lit223 mp = 96-97°C). The remaining diketone in the final reaction mixture was purified by flash column chromatography (silica gel, ethyl acetate : petroleum ether (35-60°C) = 3:97) and the combined product totaled 2.59 g (9 mmol, 25% yield). Virgin crystals and crystals that had been annealed at 90°C (3 cycles) were analyzed by differential scanning calorimetry and found to have an endothermic phase transition at 86°C = 12 kJmol-  1) (lit103  transition point = 86°C)  176  1H NMR: (CDC13, 300 MHz) 5 2.4 (8H, t, J = 7 Hz, -CH2-C=0), 1.6 (8H, m, methylenes),  1.3 (16H, m, methylenes) 13C NMR: (CDC13, 75 MHz) 5 211.8 (C=0), 42.2 (CH2), 28.7 (CH2), 28.2 (CH2), 23.9  (CH2). CPMAS 13C NMR: (100MHz) 5 220.5 (C=0), 52.6 (CH2), 40.9 (CH2), 39.3 (CH2), 35.4  (CH2). IR (KBr): 2940, 1707 (C=0) cm-1. MS m/e (rel-intensity): 280 (Mt, 27), 223 (36), 183 (23), 112 (43), 111 (32), 98 (45), 97.  (82), 95 (29), 85 (29), 83 (27), 82 (26), 81(41), 71(40), 69 (39), 67 (31), 58 (25), 57 (36), 55 (100), 43 (42), 41(69). Calculated mass: 280.2403, found: 280.2401. UV: (cyclohexane) Xmax: 285 nm (c, 54) (n-e).  Cydodocosane-1,12-dione (12)  The procedure used for the preparation of diketone 8 was followed with only a slight modification by starting with the commercially available acid chloride dodecanedioyl dichloride (4.7 ml, 19 mmol). The final mixture was analyzed by GLC (DB 17, 220°C) and subjected to flash column chromatography (silica gel). Elution with 3% (v/v) ethyl acetate in petroleum ether (35-60°C) yielded 1.3 g (4 mmol, 41% yield) of diketone 12 as a white crystalline solid. Recrystallization from ethyl acetate/petroleum ether gave plates with mp = 54-55°C (lit223 mp = 54-55.5°C).  1H NMR: (CDC13, 300 MHz) 5 2.4 (8H, t, J = 7 Hz,-CH2-C=0), 1.7 (8H, m, methylenes),  1.3 (24H, m, methylenes) 13C NMR:(CDC13, 75 MHz) 5 212.0 (C=0), 42.2 (CH2), 28.9 (CH2), 28.8 (CH2), 28.6  (CH2), 23.8 (CH2).  177  CPMAS 13C NMR:(100Hz) 224 and 222.0 (2 C=0), 54.0, 48.7, 47.8, 45.0, 44.1, 42.9, 42.4, 41.3, 39.8, 37.9, 37.2, 34.3, 33.4. MS m/e (rel.intensity): 336 (M-1-, 22), 211 (22), 168 (23), 153 (24), 135 (27), 125 (43), 111 (30), 109 (27), 98 (26), 97 (31), 95 (41), 83 (34), 81(42), 71(39), 69 (69), 67 (38), 58 (28), 57 (22), 55 (100), 43 (37), 41(69). Calculated mass: 336.3030, found: 336.3036. UV: (acetonitrile) A,max: 282 nm (E, 67) (n-e)  The structure of diketonel2 was supported by X-ray crystallographic analysis.  Cydotetracosane-1,13-dione (13)  The procedure used to prepare diketone 8 was modified by using 10 g (41 mmol) of 1,11-undecanedicarboxylic acid. The final reaction mixture was analyzed by GLC (DB17, 250°C) and subjected to flash column chromatography (silica gel). Elution with 3% (v/v) ethyl acetate in petroleum ether (35-60°C) yielded 1.34 g (4 mmol, 19%) of diketone 13 as a white solid and recrystallization from diethyl ether/petroleum ether gave plates with mp = 62-63°C (Ho 14 mp = 62-63.5°C). 1H NMR: (C6D6, 200 MHz) 5 2.1 (8H, t, J = 7 Hz, -CH2-C=0), 1.6 (8H, m, methylenes),  1.2 (28H, s, methylenes). 13C NMR: (CDC13, 75 MHz) 5 212 (C=0), 42.4 (CH2), 28.9 (CH2), 28.7 (CH2), 28.6 (CH2), 23.9 (CH2). IR (KBr): 2913, 1706 (C=0) cm-1.  MS mie (rel. intensity): 364 (M+,16), 98 (24), 97 (30), 95 (34), 83 (54), 81(40), 71(46), 69 (59), 67 (40), 58 (39), 57 (44), 56 (22), 55 (100), 43 (90), 42 (25), 41(75). Calculated mass: 364.3343, found: 364.3344. UV: (Cyclohexane) Xmax : 285.5nm (e, 41) (n-e)  The structure of diketone 13 was supported by X-ray crystallographic analysis.  178  Cyclohexacosane-1,14-dione (14)  The procedure used for the preparation of diketone 8 was modified by using 10 g (39 mmol) of 1,12-dodecanedicarboxylic acid. The final mixture was analyzed by GLC (DB17, 260°C) and subjected to flash column chromatography (silica gel). Elution with  3% (v/v) ethyl acetate in petroleum ether yielded 2.3 g (6 mmol, 29%) of diketone 14 as a white crystalline solid. Careful recrystallization of diketone 14 from ethyl acetate /petroleum ether at room temperature for a prolonged period gave clear plates. Differential scanning calorimetry of the clear plates (virgin crystals) showed an endothermic phase transition OH = 6.3 kJmo1-1) at 54°C, but the annealed crystals (60°C, 3 cycles) did not show any phase transition except the one at the melting point. When recrystallization was carried out by seeding a concentrated solution with annealed crystals, needle-shaped crystals with a melting point identical to the plates (mp = 70° lit114 69°C) were obtained. Agitation or sudden cooling of the saturated solution also led to needle-shaped crystals. The differential scanning caloiimetry traces, FTIR spectroscopy spectra and the X-ray crystallographic analysis of the plates and the needle crystal modifications confirmed them to be dimorphs.  1H NMR: (CDC13, 500 MHz) 5 2.5 (8H, t, J = 7Hz, -CH2-C=0), 1.7 (8H, m, methylenes),  1.4 (24H, m, methylenes). 13C NMR: (CDC13, 125 MHz) 5 212 (C=0), 42.5 (CH2), 29.05 (CH2), 29.01 (CH2), 28.95  (CH2), 28.90 (CH2), 28.76 (CH2), 28.70 (CH2), 23.8 (CH2). IR (KBr) (see Figure 60):  plates: 2915, 2852, 1698 (C=0), 1471, 1414, 1374, 1214, 1198, 1172, 1004 cm-1. needles: 2919, 2848, 1704 (C=0), 1463, 1433, 1409, 1365, 1200, 1162 cm-1. MS m/e (rel.intensity): 392 (MI-, 13), 254 (23), 111 (22), 109 (21), 98 (23), 97 (43), 96 (24),  95 (44), 83 (57), 71(49), 69 (59), 67 (30), 58 (32), 57 (32), 55 (100), 43 (39), 41(28). Calculated mass: 392.3656, found: 392.3655.  179  UV (cyclohexane) Xmax : 274 nm (e, 88) (n-Tc*)  6.1.3. Synthesis of Non-diametric Diketones by Blomquist's High Dilution Technique. Cyclohexadecane-1,8-dione (15)  Synthesis of the non-diametric diketone 15 was achieved by using a mixture of suberic acid (9 g, 48 mmol) and sebacic acid (5 g, 23 mmol) and employing the same procedure described for the preparation of diketone 8. Repeated trials of the above procedure using different ratios of the starting materials showed the above mentioned ratio to give the optimum yield of diketone 15. The final mixture according to GLC (DB17, 200°C) contained only 20% of diketone 15, due to the formation of the side products diketone 8, diketone 9 and the y-pyrone 24. Diketone 9 and y-pyrone 24 were selectively removed from the the crude reaction mixture by several fractional recrystallisations to increase the proportion of the desired product to 30%. Two elutions of this mixture through a silica gel column (ethyl acetate:petroleum ether (35-60°C) = 3:97) afforded diketone 15 in 90% purity. Several further fractional recrystallizations from diethyl ether/petroleum ether yielded pure diketone 15 (318 mg, —1 mmol, 5.5%). Recrystallization from ethyl acetate/petroleum ether gave plates with mp = 73-74°C. Both virgin and annealed (to 40°C and 60°C) crystals showed endothermic phase transition points at 37°C (AG = 23 kJmo1-1) and 55°C (AG = 8 khno1-1) on the differential scanning calorimeter. 11-1 NMR: (CDC13, 300 MHz) 5 2.4 (8H, m, -CH2-C=0), 1.7 (8H, m, methylenes), 1.3  (12H, m, methylenes).  180  13C NMR: (CDC13, 75 MHz) 8 212 (C=0), 42.2 (CH,), 41.4 (CH2), 27.9 (CH2), 27.5  (CH2), 27.3 (CH2), 23.4 (CH2), 23.3 (CH2). IR (KBr): 2929, 1708 (C=0) cm-1. MS m/e (rel.intensity): 252 (M+, 14), 112 (31), 97 (37), 95 (20), 84(60), 83 (33), 71(25), 69  (49), 67 (25), 58 (25), 57 (26), 56 (27), 55 (100), 43 (73), 42 (38). Calculated mass: 252.2089, found: 252.2090. Analysis calculated for C16H2802: C,76.14; H, 11.18. found: C, 76.30; H, 11.10. UV (cyclohexane) Xmax: 286 nm (e, 44) (nn*)  The structure of diketonel5 was confirmed by X-ray crystallographic analysis. Cycloheptadecane-1,9-dione (16).  Diketone 24 was synthesized by the procedure similar to that used for the synthesis of diketone 15, using a mixture of azeleic acid (5 g, 27 mmol) and sebacic acid (5.5 g, 27 mmol). The final mixture on GLC (DB 17, 200°C) showed three major peaks corresponding to diketone 16 (50%), diketone 9 (24%), diketone 10 (13%) together with traces of several minor peaks. Most of the side products 9 and 10 were selectively precipitated from the crude reaction mixture and removed by filtration. The rest of the mixture was subjected to conventional silica gel column chromatography, and elution with 3% (v/v) ethyl acetate in petroleum ether (35-60°C) completely isolated a mixture of diketones 16 and 10 from the rest of the impurities. Further 3 separate chomatographic elutions of this mixture with 2% (v/v) ethyl acetate in petroleum ether gave diketone 16 in 95% purity and which, on several recrystallizations from diethyl ether/petroleum ether, gave pure diketone 16 (253 mg, 3.5% yield). mp = 38-39°C. 1H NMR: (CDC13, 400 MHz) 8 2.4 (8H, m, -CH2-C=0), 1.6 (8H, m, methylenes), 1.3  (14H, m, methylenes).  181  13C NMR: (CDC13, 50 MHz) 5 217.5 (C=0), 42.45 (CH2), 41.94 (CH2), 28.26 (CH2),  28.17 (CH2), 28.05 (CH2), 23.73 (CH2), 23.70 (CH,). IR (KBr): 2931, 1708 (C=0) cm-1. MS m/e (rel.intensity): 266^15), 111 (24), 98 (76), 84(21), 81(26), 71(28), 67(27), 57  (26), 55 (100). Calculated mass: 266.2247, found: 266.2244. Anal. calcd for C17H3002: C,76.64; H,11.35. found: C, 76.50; H, 11.40. UV: (Cyclohexane) A,max: 285 nm (e, 48) (n-e)  6.1.4. A Miscellaneous Reaction from Pimeloyl Chloride Under High Dilution Technique. 1,2,3,4,5,6,7,8,-Octahydroxanthan-9-one (23):  An attempt was made to synthesize the twelve membered ring diketone 7 from the appropriate diacid. However, the reaction afforded y-pyrone 23 as the only product. Pimelic acid (10 g, 62 rnmol) was used in a procedure similar to that used in the synthesis of diketone 8. Analysis of the final brown-colored crude reaction mixture on GLC (DB 17, 185°C) showed a major peak at a retention time of 10.8 min with several other minor peaks. The compound corresponding to the major peak was purified by flash-column chromatography (silica gel) by eluting with 3.5% (v/v) ethyl acetate in petroleum ether (35-60°C). Recrystallization of the resulting white solid (3.5 g, 17 mmol, 56%) from ethyl acetate/petroleum ether gave compound 23 as prisms with mp 128-130°C (lit126 mp = 131°C).  1H NMR: (CDC13, 500 MHz) 5 2.45 (4H, t, -CH2-C-0), 2.35 (4H, t, J = 6.5 Hz, -CH2-  C-C=0), 1.65 (4H, m, methylenes), 1.60 (4H, m, methylenes).  182  13C NMR: (CDC13, 100 MHz) 5 178.8 (C=0), 161.9 (C=C-0), 120.2 (C-C=0), 27.2  (CH2), 21.7 (CH2), 21.4 (CH2), 20.6 (CH2). IR (KBr): 2934, 1662 (C=0), 1616 (C=C) cm-1. MS m/e (rel.intensity): 204 (M+, 68), 203 (100), 189 (29). Calculated mass: 204.1151, found: 204.1144. Analysis calculated for C13H1602: C, 76.44; H, 7.90, found: C, 76.51; H, 7.97.  The structure of compound 23 was supported by X-ray crystallographic analysis.  6.1.5. Synthesis of Diametric Diketone by Dieckmann Condensation Reaction. Cycloeicosane-1,11-dione (11)112  The synthesis of the twenty membered ring diketone 11 was carried out via a Dieckmann condensation reaction under high dilution conditions with high speed stirring. The apparatus used was similar to the one described for the synthesis of diketone 8 with the exception that instead of a "Y" tube, a cyclic "high dilution apparatus" equipped with a syringe pump (Sage Instruments, model 341A) and an efficient reflux condenser was attached to the second neck. Dry nitrogen was bubbled through the solution via the third neck of the three-necked RB flask. A 60 mL solution of diethyl dodecanedioate (9 g, 30 mmol) in xylene (freshly distilled over Na) was slowly added over a period of 24 hrs (2.5 mL hr-1) from a syringe pump into a refluxing stirred solution of potassium tert-butoxide(24 g, 62 mmol) and dry xylene (1000 mL). Vigorous stirring was maintained throughout the reaction. After all of the ester had been added, stirring and refluxing was continued for one hour and the reaction mixture was allowed to cool to room temperature.  183  The mixture was then acidified with glacial acetic acid, washed with water (3 x 25 mL) and filtered to remove insoluble polymeric ketones. A reddish-brown residue obtained following the evaporation of the solvent in vacuo was placed in a 100 mL RB flask fitted with a reflux condenser and 15 mL of 3N hydrochloric acid was cautiously added with stirring; a small amount of absolute ethanol (3 mL) was also added to promote the solubility. The mixture was refluxed on the heating mantle for 5 hr and cooled to room temperature. The acidic mixture was extracted with ether (3 x 10 mL) and the combined ether extracts were washed with 10% NaHCO3 (3 x 15 mL), dried over magnesium sulphate, filtered and evaporated to dryness in vacuo. Gas chromatographic analysis (DB17, 220°C) of the final mixture showed a major peak corresponding to the diketone 11 at a retention time of 10.83 min with several minor peaks. Purification by flash column chromatography (silica gel, ethyl acetate : petroleum ether (35-60°C) - 3:97) afforded a colourless solid (647 mg, 2.1 mmol, 14%) with mp = 54°C (lit103 mp = 54°C). Recrystallization from ethyl acetate/petroleum ether solvent system gave plates. 1-H NMR: (CDC13, 300 MHz) 5 2.4 (8H, t, J = 7 Hz, -CH2-C=0), 1.4 (8H, m, methylenes),  1.2 (20H, m, methylenes). 13C NMR: (CDC13, 75 MHz) 8 212.0 (C=0), 42.2 (CH2), 28.5 (CH2), 28.3 (CH2), 23.7  (CH2). IR (KBr): 2924, 1699 (C=0) cm-1. MS m/e (rel.intensity): 308 (Mt, 57), 251 (37), 212 (24), 197 (30), 155 (27), 154 (40), 139 (34),  121 (21), 111 (62), 98(31), 81(36), 71(40), 69 (71), 67(32), 58(29), 55 (100), 41(81). Calculated mass: 308.2717, found: 308.2722. UV: (cyclohexane) Xmax: 281 (e, 53) (n-e)  The structure of diketone 11 was confirmed by X-ray crystallographic analysis.  184  6.1.6. Synthesis of Cyclic Keto-alcohols by Partial Reduction of Diketones. 9-Hydroxycyclohexadecanone (29):127  A solution of sodium borohydride (95 mg, 2.5 mmol) in absolute ethanol (30 mL) was slowly added from a dropping funnel to a stirred solution (70 mL) of cyclohexadecane-1,9-dione 9 (2.52 g, 10 mmol) in absolute ethanol (70 mL) at room temperature. The reaction was followed by GLC (DB 17, 170°C) analysis using 0.1 mL aliquots taken from the reaction mixture at different time intervals (samples were quenched with water and ether extracts were injected on GLC). As the amount of keto-alcohol started to decrease, addition was stopped and the mixture was quenched with water and extracted with diethyl ether (3 x 15 mL). The combined ether extracts were dried over anhydrous magnesium sulphate, filtered and evaporated to dryness in vacuo. Analysis of the final mixture on GLC showed 3 major peaks at retention times 5.89, 6.47 and 7.20 min, corresponding to the unreacted diketone 9, kern-alcohol 29, and cyclohexadecane-1,9-diol 30 (both cis and trans) respectively.  Both products were separated from the unreacted starting material in pure form by conventional column chromatography (silica gel). The polarity of the solvent was gradually increased from 2% to 5% (v/v) ethyl acetate in petroleum ether (35-60°C). Unreacted starting material was eluted with 3% solvent, followed by the keto-alcohol 29 as a white crystalline solid (940 mg, 3.7 mmol, 37%) and the diol 67 (mixture of both cis and trans) as a white solid (204 mg, 8%) with mp = 107-109°C. Recrystallization of ketoalcohol 29 from hexane gave needle-shaped crystals with mp = 73-74°C.  9-Hydroxy cydohexadecanone (29) 1H NMR:(C6D6, 300 MHz) 8 3.5 (1H, m, -CH-OH), 2.0(4H, m, methylenes), 1.1-1.5 (24H,  m, methylenes).  185  13C NMR: (C6D6, 75 MHz) 6 209.8 (C=0), 70.0 (=CH-OH), 41.9 (CH2), 35.1 (CH2),  27.8 (CH2), 27.6 (CH2), 27.5 (CH2), 23.5 (CH2), 23.2 (CH2). IR (KBr): 3400-3100 (OH), 2930, 1704 (C=0) cm-1. MS m/e (rel.intensity) 254 (M+, 11), 236 (51), 211 (27), 183 (31), 155 (38), 142 (44), 127  (64), 111 (36), 98(48), 83(61), 71(62), 57(84), 55 (100), 43 (87), 41(87). Calculated mass: 254.2247, found: 254.2243. Anal. calcd for C16H3002: C, 75.40; H, 12.00. found:C, 75.54; H, 11.89. UV: (cyclohexane) Xmax: 275nm (e, 46) (n-e).  Cyclohexadecane-1,9-diol (67) (cis and trans) 1H NMR: (CDC13, 300 MHz) 5 3.75 (2H, m, -CH-OH), 1.5 (8H, t, J = 6.5 Hz, methylenes),  1.4 (20H, m, methylenes). 13C NMR: (CDC13, 75 MHz) 8 70.6 (=CH-OH), 35.0 (CH,), 27.1 (CH2), 27.0 (CH2), 26.9  (CH2), 26.8 (CH2), 23.2 (CH2), 22.8 (CH2). IR (KBr): 3500-3100 (OH), 2933 cm-1. DCI MS m/e (rel.intensity): 274 (M++1+NH3, 31), 257 (M++1, 85), 221 (M++1-2H20,  100). Calculated mass: 256.2404, found: 256.2394. Anal. calcd for C16H3002: C, 74.94; H, 12.58. found: C, 75.13; H, 12.65.  10-Hydroxy cyclooctadecanone (30)  The procedure used for the preparation of keto-alcohol 29 was modified by using cyclooctadecane-1,10-dione (10) (2.24, 8 mmol) and almost three times the stoichiomenic amount of sodium borohydride (228 mg, 6 mmol) used for the synthesis of keto-alcohol 29. Analysis of the final mixture on GLC (DB17, 180°C) showed three major peaks as  observed in the synthesis of keto-alcohol 29. Purification by conventional column chromatography (silica gel, ethyl acetate:petrolium ether (35-60°C)- 2:98 to 6:94) gave  186  keto-alcohol 30 as a white solid (587 mg, 2.1 mmol, 26% yield) plus a white crystalline by-product, cyclooctadecane-1,10-diol 68 (772 mg, 2.7 mmol, 34% yield). Recrystallization of keto-alcohol from hexane gave plates with mp = 78-79°C. Recrystallization of diol 68 from pure CHC13 gave needles with mp = 118-120°C. The second crop of crystals grown from the mother liquor gave flakes with mp = 133134°C. Infra red spectra of these two crystal modifications show some dissimilarities in the fingerprint region. X-ray crystallographic analysis of the needles revealed that both hydroxy groups are on the same side of the ring (cis). The morphology of these two batches of crystals were independent of the solvent of recrystallization. This suggests that these two crystal modifications could be stereoisomers and not dimorphs.  10-Hydroxy cyclohexadecanone (30) 1H NMR: (C6D6, 300 MHz) 5 3.6 (1H, m, -CH-OH), 2.0 (4H, m, methylenes), 1.2-1.6  (28 H, m, methylenes). 13C NMR: (C6D6, 75 MHz) 5 210.0 (C=0), 69.1(=fH-OH), 41.9, 36.0, 28.2, 27.6,  24.5, 23.7. IR (KBr): 3575-3475(OH), 2927, 1708 (C=0) cm-1. MS m/e (rel.intensity): 282 (Mt, 2), 264 (26), 141 (39), 109 (26), 108 (28), 98 (27), 97 (37),  96 (43), 95 (44), 83 (39), 82 (34), 81(63), 71(41), 69 (45), 67 (55), 55 (100), 43 (33), 41 (79). Calculated mass: 282.2560, found: 282.2556. Analysis calculated for C18E13402: C,76.54; H,12.13, found: C, 76.49; H, 12.17. UV (cyclohexane) Xmax: 283 nm (n-e).  Cyclooctadecane-1,10-diole (68) 1H NMR: (CDC13, 300 MHz) 5 3.6 (2H, m, -CH-OH), 1.5 (8H, m, methylenes), .1.3  (24H, bs, methylenes).  187  13C NMR: (CDC13, 75 MHz) 5 70.7 (=CH-OH), 35.3, 27.7, 27.4, 23.6. IR (1(13r):  needle: 3500-3200 (OH), 2931, 1461, 1436, 1301, 1103, 1015, 998, 978 cm-1. plates:^3500-3200 (OH), 2925, 1460, 1338, 1157, 1047, 1018, 995, 978cm-1. DCI MS m/e (rel.intensity): 302 (M++1+NH3, 44), 285 (M++1, 100), 249 (M++1-2H20,  72). Analysis calculated for C1 8H3602: C, 76.00; H, 12.75, found: C, 75.96; H, 12.82.  6.1.7. Sythesis of Tetramethylated Di ketones. Tetramethy1cyc1ohexadecane-1,9-dione-(2R*,8S*,10R*,16S*) (31)128  A three-necked 200 mL round bottom flask containing 1.9 gams (48 mmol) of potassium hydride (purified under nitrogen from 22.4% (w/w) potassium hydride in mineral oil using anhydrous ether) and 80 mL of dimethoxy ethane (DME) (freshly distilled over sodium) was kept in an ice bath and to this, a 50 mL solution of cyclohexadecane-1,9-dione (9) in DME was slowly added with stirring under nitrogen. To the resulting suspension 2 mL of methyl iodide (purified by passing through a column of basic alurnia) was added dropwise with continuous stirring and cooling. After the mixture had been stirred for 15 min, it was filtered and the residual potassium hydride was washed with ether. The filtrate was then washed with water (3 x 15 mL), dried over magnesium sulphate and rotatory evaporated to give a white solid. The final mixture on GLC (DB17, 190 °C) showed three major peaks with five other minor peaks. Recrystallization from ethyl acetate/petroleum ether solvent system enabled the selective crystallization of two of the three major compounds in an almost 1:1 ratio. According to GC-MS analysis, both have parent masses corresponding to the tetramethylated diketone 31.  188  Purification by HPLC (preparative column, ethyl acetate:hexane = 5:95, flow rate = 5 ml min-1) yielded 431 mg (1.4 mmol, 15% yield) of diketone 31 as a white solid. Recrystallization from diethyl ether/petroleum ether solvent system gave needles with mp = 120 °C. The stereochemistry of diketone 31, was confirmed by X-ray crystallographic analysis. 1H NMR: (C6D6, 400 MHz) 8 2.4 (4H, m, =CH-C=0), 1.7 (4H, m), 1.1 (16H, m, methylenes), 1.0 (12H, d, J = 6.5, methyls). 13C NMR: (C6D6, 50 MHz) 8 215 (C=0), 44.6 (CH2), 32.5 (CH2), 29.5 (CH2), 26.9 (CH2), 16.7 (-CH3). IR (KBr): 2929, 1700 (C=0) cm-1. MS m/e (rel.intensity): 308 (M+, 28), 223 (30), 154 (30), 125 (30), 112 (70), 97 (49), 86 (91), 83 (40), 81(24), 71(42), 69 (82), 57 (63), 56 (44), 55 (100), 43 (75), 41(83). Calculated mass: 308.2717, found: 308.2722. Analysis calculated for C20143602: C, 77.86; H, 11.76. found: C, 77.59; H, 11.67. UV (cyclohexane) Amax: 282 nm (e, 56) (n-Tc*)  Tetramethylcyclotetracosane-1,13-dione-(2R*,12S*,14R*,24S*) (32)128 The procedure used for the preparation of diketone 31 was applied to 1.5 g (4 mmol) of cyclotetracosane-1,13-dione 13. The final mixture on GLC (DB 17, 240°C) showed three major peaks. Recrystallization of the final mixture from ethyl acetate/petroleum ether enabled the selective crystallization of diketone 32 along with an impurity (12% by GLC). Purification of diketone 32 by HPLC (preparative column, ethyl acetate : hexane=2:98, flow rate = 5 ml min-1) yielded 336 mg (0.8 mmol, 21% yield) of white solid. Recrystallization of diketone 32 from ethyl acetate/petroleum ether gave needles with mp = 106-107°C. The stereochemistry of diketone 32 was confirmed by X-ray crystallographic analysis.  189  1H NMR: (C6D6, 300 MHz) 5 2.4 (4H, m, =CH-C=0), 1.8 (4H, m), 1.2 (32H, bs, methylenes), 1.0 (12H, d, J = 6.5, methyls).  13C NMR: (C6D6, 75 MHz) 5 216 (C=0), 45.1 (=H-CH3), 33.2 (CH2), 30.3 (CH2), 30.2 (CH2), 27.6 (CH2), 17.1 (-CH3).  IR (KBr): 2919, 1699 (C=0) cm4. MS m/e (rel.intensity): 420 (Mt, 83), 336 (25), 335 (100), 97 (47), 86 (99), 83 (54), 71(33), 69 (85), 57 (50), 56 33), 55 (99.5), 43 (50), 41(48).  Calculated mass: 420.3969, found: 420.3967. Analysis calculated for C28115202: C, 79.94; H, 12.46. found: C, 80.37; H, 12.65.  UV (Cyclohexane) Xmax: 282 nm (e, 54) (n-it*)  6.1.8. Deuteration of Diketones. Cyclohexadecane-1,9-dione-2,2,8,8,10,10,16,16-d8 (69) To a stirred refluxing solution of cyclohexadecane-1,9-dione 9 (1 g, 3.97 mmol) and 25 mL of Me0D, 6.0 mL of a suspension of MeOK in Me0D (potassium hydroxide pellets were added to Me0D until the solution turned turbid) was added along with 50 mL of D20. After refluxing for two hours, the reaction mixture was cooled to room temperature, neutralized with conc HC1, diluted with water (10 mL) and then extracted with diethyl ether (3 x 10 mL). The combined ether extracts were dried over anhydrous magnesium sulphate and evaporated to dryness to yield 99 mg (3.8 mmol, 96% yield) of diketone 69. Recrystallization from ethyl acetate/petroleum ether gave plates with mp = 86°C.  1H NMR: (CDC13, 300 MHz) 5 1.6 (8H, m, methylenes), 1.2 (12H, br s, methylenes). 13C NMR: (CDC13, 75 MHz) 6 212.6 (C=0), 27.9 (CH2), 27.6 (CH2), 23.4 (CH2).  190  MS m/e (rel.intensity) 260 (Mt, 26), 131 (32), 130 (43), 129 (28), 101 (42), 100 (100), 85 (59), 69 (29), 57 (40), 56 (34), 45 (32).  Cyclooctadecane-1,10-dione-2,2,9,9,11,11,18,18-4 (70) The procedure used for the preparation of deuterated diketone 20 was applied to the eighteen membered ring dilcetone 10 (560 mg, 2 mmol). Recrystallization of the deuterated diketone 70 from ethyl acetate/petroleum ether gave plates (530 mg, 1.84 mmol, 92%) with mp = 96-97°C. 1H NMR: (CDC13, 300 MHz) 8 1.6 (8H, m, methylenes), 1.3 (16H, br s, methylenes). 13C NMR: (CDC13, 75 MHz) 8 212.4 (C=0), 28.7 (CH2), 28.1 (CH2), 23.6 (CH2). MS ink (rel.intensity): 288 (Mt, 51), 227 (49), 189 (29), 145 (38), 144 (55), 127 (62), 126 (36), 114 (68), 99 (100), 87 (32), 84 (29), 83 (39), 82 (27).  Cyclohexacosane-1,14-dione-2,2,13,13,15,15,26,26418 (71) The procedure used for the preparation of diketone 69 was applied to twenty-six membered ring diketone 14 (1 g, 2.55 mmol). Recrystallization of diketone 71 from ethyl acetate/petroleum ether gave both plates and needles (756 mg, 1.89 mmol, 74% yield) with mp = 67-68°C.  1H NMR: (CDC13, 200 MHz) 8 1.2 (8H, m, methylenes), 1.0 (32H, br s, methylenes). 13C NMR: (CDC13, 50 MHz) 8 28.9 (CH2), 23.6 (CH2). MS tn/e (rel.intensity): 400 (Mt, 4), 392 (14), 254 (27), 196 (28), 181 (26), 125 (42), 114 (26), 111 (39), 109 (37), 98 (45), 97 (70), 95 (64), 85 (53), 83 (72), 81(58).  191  6.2. Photochemical studies. 6.2.1. General. Photochemical reactions of diketones were explored in solution and in the solid state. All reactions were performed with the Pyrex-filtered output (A, > 290 nm) of a 450 W medium pressure Hanovia lamp, placed in a water-cooled Pyrex immersion well. Irradiations were carried out at various temperatures, especially in the case of diketones with transition points; the solid state reactions were conducted above and below the transition points. All solution phase reactions were performed in reagent grade hexane, with the exception of diketone 10, where octane was used as solvent at 90°C. Photoproduct ratios measured for the two different solvents at room temperature suggests no apparent solvent effects. Analytical photolysis (general procedure):  Solution and solid state photolyses were conducted using a minimum of three samples, and for each sample, irradiations for different lengths of time were explored. Aliquots were routinely injected on GLC at different time intervals and the peak area ratios of the major photoproducts were noted. For each analysis at least three sample injections were made and the statistical average of the photoproduct ratios were recorded. Finally the detector response for each photoproducts was calculated to obtain the actual product ratios. Total percent conversion to products was limited between 10-15 % to prevent the formation of secondary photoproducts, and especially in solid state reactions to prevent the effect of sample melting on product ratios. The reported product ratios (Table 1V) correspond to 0% conversion to products, which were obtained from extrapolating the product ratios versus percentage conversion plots to 0% conversion. Analytical photolyses performed at 20 °C and above were carried out using the apparatus shown in Figure 86. Two specially designed Pyrex test tubes containing the  192  solution (15-20 mg m1-1) of the compound under investigation were placed inside a Pyrex refluxing apparatus. condensor  Figure 86: Apparatus used for the analytical reactions at elevated temperatures.  The desired temperatures were maintained by refluxing appropriate solvents or azeotropic mixtures. The solvent or the solvent system used and their corresponding boiling points are given in Table XVII.  Table XVII: Solvent systems and their corresponding boiling points.  SOLVENT / SOLVENT SYSTEM  BOILING POINT  dichloromethane  —40°C  acetone : hexane = 59:41  50°C  chloroform  —60°C  cyclohexane  —80°C  Isobutylalcohol : water = 70:30  90°C  193  Irradiations below 20°C were carried out by using a bath (ethanol in a Pyrex container) controlled by a Cryocool CC-100-II immersion cooling system from Neslab Instruments, Inc. and the temperature was maintained within ±2 °C.  Solid state reactions:  Approximately 5 mg of the compound under investigation was dissolved in — 0.5 mL of diethyl ether and a thin film was carefully formed on the inner surface of a test tube (10 mL) by slow solvent evaporation. Any remaining traces of solvent were removed under vacuum and the test tube was sealed with parafilm film after the air was exchanged by nitrogen. Irradiations of single crystals and powdered crystals were also investigated in order to compare the results, but no significant differences in product ratios were observed. Solution reaction:  A 10-1 M solution of the compound (1.0 mL) under investigation in a pyrex photolysis tube (3 mm) was degassed by freeze-pump-thaw cycles (at least three times) and sealed under nitrogen atmosphere prior to the photolysis. All irradiations were performed at a constant distance (12") from the lamp.  Preparation and photolysis of annealed crystals:  Diketones that exhibit solid-solid phase transitions were photolyzed above and below the transition points. Crystals annealed above the transition points were also irradiated at room temperature. Temperatures were gradually raised 5-15°C above the transition point (but below the melting point) on the hot stage of the Fisher-Johns melting point apparatus and slowly cooled to room temperature. Usually this cycle was repeated at least three times prior to irradiation. Usually the transparent crystal cracked while passing through the transition point and finally became opaque. Irradiations of powdered annealed crystals were also performed.  194  Preparative scale photolysis (General procedure) :  For solid state photolysis, powdered crystals (0.5 - 1.0 g) were sandwiched between Pyrex glass plates (2.5" x 1.0"), and the edges of these plates were taped with 3M magic tape (Scotch 810). Taped plates were then sealed in polyethylene bags filled with N2 and photolyzed. During irradiation the reaction progress was monitored by GLC (DB 17 or DB 1) and the total percentage conversion to products generally maintained between 10% and 15%. After the reaction had been stopped, the final mixture was scraped off the plates and the photoproducts were purified by column chromatography and recrystallization techniques. The fractions collected from the column were analyzed by either thin layer chromatography or by GLC and the appropriate fractions were combined and the solvent was removed in vacuo. Photoproducts were then characterized by spectral and X-ray crystallographic analysis. Unless otherwise indicated, for the cyclobutanol photoproducts, the stereo chemistry at the ring junction was determined by comparing the spectral characteristics with those of the photoproducts (9c and 11t) and with known stereochemistry determined by X-ray, based mainly on the chemical shift of the methine carbon (Table III).  6.2.2. Photochemistry of Diametric Diketones. The following procedure described for cyclododecane-1,7-dione 7 is typical for a preparative scale solution reaction. A 250 mL solution of diketone 7 (750 mg, 3.8 mmol) in reagent grade hexane was placed in a water-cooled Pyrex immersion well setup and deoxygenated by bubbling nitrogen through the stirred solution for 1/2 hr. A steady flow of nitrogen and stirring was maintained throughout the irradiation. The progress of the reaction was analyzed at  195  different intervals by GLC (DB 17, 155°C). Three new major peaks were observed on GLC at RT's 1.58, 3.27 and 7.94 min. Reaction was stopped when the total conversion to products was 15%. The white semisolid mixture obtained after evaporating the solvent in vacuo, was subjected to column chromatography (silica gel, Et0Ac:petroleum ether (35-  60°C) = 5:95 to 9:91). The polarity of the solvent was gradually increased from 5% to 9% (v/v) Et0Ac in petroleum ether. The compounds at 1.58 and 3.27 min on GLC were eluted together as a mixture with 8% (v/v) Et0Ac in petroleum ether. The third product at RT 7.94 min was eluted with 9% (v/v) Et0Ac in petroleum ether as an oil (45 mg, 6% yield) and was later identified as Bicyclo[8.2.0]dodecan-4-one, 10-hydroxy-(1R*,10R*)-(±) (7c) The compounds at 1.58 and 3.27 min were purified by HPLC (preparative column, Et0Ac:hexane = 15:85, flow rate = 5 mL min-1). Compounds 7u (oil, 32 mg, 4% yield) and 7e (white solid, 94% pure by GLC) respectively were obtained. Several recrystallizations of 7e from Et20/petroleum ether gave pure 7e as flakes (23 mg, 3% yield) with mp = 37-38°C. Compound 7e was later characterized as 11-dodecene-2,8dione. According to spectral analysis, 7u appeared to be a mixture of compounds. Several attempts at purification were unsuccessful. Due to the complexity of the spectra, the compounds in the mixture could not be characterized. The product ratio at 0% conversion is given in table III.  11-Dodecene-2,8-dione (7e) 1H NMR: (C6D6, 300 MHz) 8 5.8-5.6 (1H, m, vinylic), 4.9 (2H, m, vinylic), 2.2 (2H,  two overlapping triplets, -CH2-CH=CH2), 2.0 (2H, t, J=7 Hz, CH2-C=0), 1.9 (4H, t, J=7 Hz, -CH2-C=0), 1.7 (3H, s, methyl), 1.5-1.3 (4H, m, methylenes), 1.1-1.0 (2H, m, methylene).  196  13C NMR: (C6D6, 100 MHz) 8 207.7 and 206.2 (2C=0), 137.7 (-CH=CH2), 115.0 (-  CH=CH2), 43.1 (CH2), 42.2 (CH2), 41.6 (CH2), 29.3 (-CH3), 28.9 (CH2), 23.6 (CH2), 23.4 (CH2). IR (KBr): 3082, 2934, 1704 (C=0) cm-1. MS: m/e (rel.intensity) 196 (Mt, 0.5), 95 (52), 83 (31), 55 (76), 43 (100). Calculated mass: 196.1464, found: 196.1460. Anal. calcd. for C12H2002: C, 73.43; H, 10.27. found: C, 73.45; H, 10.36.  Bicyclo[8.2.0]dodecan 4 one, 10 hydroxy (1R*,10R*) (±) (7c) -  -  -  -  -  13C NMR: (C6D6, 50 MHz) 8 214.2 (CO), 78.5 (C-OH), 48.6 (methine carbon), 43.3  (CH2), 36.9 (CH2), 36.0 (CH2), 33.7 (CH2), 27.4 (CH2), 24.7 (CH2), 23.4 (CH2), 21.1 (CH2), 17.0 (CH2). IR (neat): 3600-3200 (OH), 2970, 1702 (C=0) cm4. MS: mie (rel.intensity) 196 (Mt, 5), 178 (50), 168 (15), 150 (35), 149 (67), 135 (64),  125 (29), 121 (35), 111 (40), 109 (29), 108 (31), 107 (57). Calculated mass: 196.1464, found: 196.1467.  Solid state photolysis of diketone 7  An analytical solid state reaction was carried out as described in the general experimental section. A major peak at RT 7.94 min with traces of peaks at 1.58 and 3.30 mm on GLC (DB 17, 155°C) were observed. A preparative scale solid state reaction was not performed and the major product was not isolated, but the peaks at RT 7.94 and 3.27 min were assumed to be compounds 7c and 7e isolated in the solution state reaction, based on co-injection experiment on GLC.  197  Photolysis of Cyclotetradecane-1,8-dione (8):  Irradiation of dilcetone 8, both in solution and in the solid state, showed three major new peaks on GLC (DB 17, 160°C) at RT's 6.04, 11.79 and 12.91 min. Preparative scale solid state reaction was carried out with 950 mg of starting material as described in the general procedure. Photolysis was stopped after 16% of the starting material had been reacted. All three photoproducts were purified by silica gel column chromatography (Et0Ac:petroleum ether (35-60°C) = 4:96 to 10:90) and were later identified by spectroscopic analysis. 13-tetradecene-2,9-dione (8e) (RT 6.04 min) was eluted with 4% (v/v) Et0Ac in petroleum ether as a white solid (19 mg, 2% yield). Recrystallization of 8e from Et20/petroleum ether gave flakes with mp = 43-44°C. A trans-cyclobutanol, bicyclo[10.2.0]tetradecan-5-one,12-hydroxy-(1R*,12S*)-(±) (8t) (RT 11.79 min) was eluted as an oil with 9% (v/v) Et0Ac in pet-ether (38 mg, 4%). A cis-cyclobutanol, bicyclo[10.2.0]tetradecan-5-one,12-hydroxy-(1R*,12R*)-(±) (8c) (RT 12.91 min) was eluted as an oil with 10% (v/v) Et0Ac in petroleum ether (86 mg, 9%). 13-tetradecene-2,9-dione (8e) 1H NMR: (C6D6, 200 MHz) 5 5.4 (1H, m, vinylic), 5.0 (2H, m, vinylic), 2.0-1.8 (8H, m,  methylenes), 1.7 (3H, s, -CH3), 1.4-1.6 (6H, m, methylenes). 13C NMR: (C6D6, 50 MHz) 5 218.9 and 216.5 (2 C=0), 138.4 (-QH=CH2), 115.1  (-CH=CH2), 43.2 (-CH2), 42.5 (-CH2), 41.7 (-CH2), 33.4 (-CH2), 29.3 (-CH3), 29.2 (-CH2), 23.8 (-CH2), 23.7 (-CH2), 23.1 (-CH2). IR (KBr): 3080, 2933, 1715 (C=0), 1645 cm-1. MS: m/e (rel.intensity) 224 (M+, 10), 155 (30), 152 (33), 113 (100), 109 (83), 97 (68),  95 (40), 94 (33), 84 (33), 69 (32). Calculated mass: 224.1777, found: 224.1785. Anal. calcd for C14H2402: C, 74.95; H, 10.78. found:C, 74.88; H, 10.75.  198  bicyclo[10.2.0Jtetradecan-5-one,12-hydroxy-(1R*,12S*)-(±) (8t) -13C NMR: (C6D6, 50 MHz) 5 210.6 (C=0), 77.3 (C-OH), 42.1, 40.4 (methine carbon), 39.5 (CH), 38.8 (CH2), 33.9 (CH2), 28.5 (CH2), 27.4 (CH2), 25.9 (CH2), 23.1 (CH2), 22.2 (CH2), 22.1 (CH2), 20.4 (CH2). IR (neat): 3600-3300 (OH), 2936, 1703 (C=0) cm-1. MS: nVe (rel.intensity) 224 (Mt 3), 222 (29), 206 (47), 196(15), 165 (28), 163 (37), 149 (44), 139 (29), 137 (33), 136 (25), 135 (41), 133 (26), 131 (27), 125 (44), 124 (30), 123 (44), 122 (42), 121 (59), 119 (29), 113 (46), 112 (79), 111 (67), 110 (81), 109 (78), 108 (46), 107 (83). Calculated mass: 224.1777, found: 224.1783.  bicyclo[10.2.0]tetradecan-5-one,12-hydroxy-(1R*,12R*)-(±) (8c) 13C NMR: (C6D6, 50 MHz) 8 210.0 (0=0), 77.2 (C-OH), 49.2 (methine carbon), 42.3 (CH2), 41.4 (CH2), 36.3 (CH2), 34.5 (CH2), 29.9 (CH2), 28.3 (CH2), 26.4 (CH2), 24.7 (CH2), 23.9 (CH2), 22.4 (CH2), 18.8 (CH2). IR (neat): 3600-3200 (OH), 2928, 1703 (0=0) cm-1. MS: m/e (rel.intensity) 224 (Mt 2), 222 (22), 206 (50), 196 (8), 188 (29), 163 (42), 149  (46), 145 (32), 135 (43), 133 (32), 131 (33), 125 (33), 124 (29), 121 (63), 119 (37), 112 (55), 110 (72), 107 (100), 95 (96), 93 (92), 81(57), 79 (76), 67 (35). Calculated mass: 224.1777, found: 224.1772.  The photoproducts from the solution state reaction were not isolated but were assumed to be the same as their solid state counterparts based on their GLC retention times.  199  Photolysis of Cyclohexadecane-1,9-dione (9): Irradiation in hexane:  Preparative scale solution state reaction of diketone 9 (800 mg) was carried out using the procedure described for diketone 7. Three major peaks were observed on GLC (DB 17, 190°C) at RT's 4.67, 8.64 and 9.56 min. Products were purified by silica gel  column chromatography (Et0Ac:petroleum ether (35-60°C) = 2:98 to 9:91) and were later identified by spectral analysis. An ene-dione, 15-hexadecene-2,10-dione (9e) (RT 4.67 min), was eluted as a white solid with 4.5% (v/v) Et0Ac in petroleum ether (54 mg, 6.8%). Recrystallization of 9e from EtOAC/petroleum ether gave flakes with mp = 53-54°C. Elution with 7% (v/v)  Et0Ac in petroleum ether afforded pure trans-cyclobutanol, bicyclof12.2.0] hexadecan-6one,14-Hydroxy-(1R*,14S*)-(±) (9t) (RT 8.64 min), as an oil; further elution gave several fractions of a mixture of compounds 9t and the compound at RT 9.56 min (9c) as an oil (9c:9t = 29:71 by GLC). When the mixture was kept overnight in the refrigerator, a white  solid formed in the oil. Upon GLC analysis the solid was found to contain compound 9c as the major component. The solid was washed with ice cold petroleum ether and subjected to several fractional recrystallizations from Et0Ac/petroleum ether to give prisms (35 mg, 4.4%) with mp = 91-92°C. Compound 9c was identified as bicyclo [12.2.0]hexadecan-6-one,14-Hydroxy-(1R*,14R*)-(±) (9c). The trans-cyclobutanol 9t present in the mother liquor (> 92% by GLC) was further purified by column chromatography and the combined fractions weighed 27 mg (3.4%).  Solid state irradiation:  Analytical solid state reaction on GLC showed a major peak at RT 9.60 min together with two minor peaks at RT's 8.66 and 4.67 min. The solid state products were not isolated but assumed to be the same as the solution counterparts, based on their GLC retention times.  200  Analytical photoreactions at 0°C (solid), 10°C (solid), 20°C, 30°C, 40°C, 60°C were explored, and the product ratios at 0% conversions are tabulated in Table XVI (a).  Irradiation of annealed crystals:  Crystals annealed to 50°C (above the transition point of 34°C) were powdered and photolyzed at 20°C on an analytical scale as described in the general experimental section. Product ratios at 0% conversion are tabulated in Table XVI(a).  15-hexadecene-2,10-dione (9e) 1H NMR: (CDC13, 300 MHz) 8 5.8 (1H, m, -CH=CH2), 5.0 (2H, m, -CH=CH2), 2.4  (6H, m, -CH2-C=0), 2.2 (3H, s, -CH3), 2.1 (2H, two overlapping triplets, -C112CH=CH2), 1.6-1.5 (6H, m, methylenes), 1.4-1.2 (8H, m, methylenes). 13C NMR: (CDC13, 75 MHz) 8 211.4 and 209.3 (2C=0), 138.5 (-CH=CH2), 114.6  (-CH=CH2), 43.7 (CH), 42.7 (CH2), 42.6 (CH2), 33.5 (CH2), 29.9 (CH3), 29.2 (CH2), 29.0 (CH2), 28.9 (CH2), 29.4 (CH2), 23.7 (CH2), 23.3 (CH2). IR (KBr): 3081, 2932, 1704 ((2=0), 1645 cm-1. MS: ink (rel.intensity) 252 (Mt, 7), 184 (39), 169 (31), 111 (33), 109 (65), 108 (28), 98  (87), 83 (84), 71(56), 69 (32), 68 (49), 67 (30), 58 (34), 55 ( 100), 43 (37), 41(51). Calculated mass: 252.2089, found: 252.2094. Anal. calcd for C16H2802: C, 76.14; H, 11.18. found: C, 76.30; H, 11.10.  bicyclo[12.2.0Thexadecan-6-one,14-Hydroxy-(1R*,14S*)-a) (9t) 13C NMR:(C6D6, 75 MHz) 8 210.6 (C=0), 78.3 (C-OH), 42.3, 41.6 (methine), 40.9  (CH2), 33.5 (CH2), 29.0 (CH2), 27.5 (CH2), 26.8 (CH2), 26.7 (CH2), 26.4 (CH2), 23.7 (CH2), 23.4 (CH2), 22.7 (CH2), 22.0 (CH2). IR (neat): 3600-3300 (OH), 2933, 1703 (C=0) cm-1.  201  MS: m/e (rel.intensity) 252 (Mt, 5), 224 (1), 206 (10), 126 (37), 111 (35), 108 (35), 98 (75), 97 (26), 95 (27), 83 (61), 81(29), 71(33), 69 (30), 67 (30), 55 (100), 43 (43), 41 (41). Calculated mass: 252.2089, found: 252.2085.  bicyclo[12.2.0]hexadecan-6-one,14-Hydroxy-(1R*,14S*)-(±) (9c) I-3C NMR: (C6D6, 75 MHz) 8 209.1 (C=0), 76.3 (f-OH), 49.8 (methine), 40.6 (CH2),  40.4 (CH2), 34.6 (CH2), 33.3 (CH2), 29.5 (CH2), 27.0 (CH2), 26.9 (CH2), 25.6 (CH2), 25.1 (CH2), 23.3 (CH2), 22.5 (CH2), 20.1 (CH2), 19.1 (CH2). IR (KBr): 3600-3300 (OH), 2927, 1703 (C=0) cm-1. MS: m/e (rel.intensity) 252 (M+, 4), 224 (1), 127 (26), 126 (47), 111 (44), 109 (35), 108  (41), 98 (100), 97 (32), 95 (30), 84 (29), 83 (83), 81(32), 71(47), 69 (39), 68 (29), 67 (35), 58 (32), 57 (26), 55 (88), 43 (71), 41(75). Calculated mass: 252.2089, found: 252.2080. Anal. calcd for C16H2802: C, 76.14; H, 11.18. found:C, 76.25; H, 11.06.  For cis-cyclobutanol 9c, the stereochemistry at the ring junction was confirmed by X-ray crystallographic analysis. Photolysis of cyclooctadecane-1,10-dione (10): Irradiation in hexane :  Preparative scale solution state reaction of diketone 10 (750 mg) was carried out using the procedure described for diketone 6. Three new peaks were observed on GLC (DB 17, 200°C) at RTs 6.19, 10.97 and 12.54 min. Products were purified by silica gel  column chromatography (Et0Ac:petroleum ether (35-60°C) = 3:97 to 9:91) and were later identified by spectral analysis. An ene-dione, 17-octadecene-2,11-dione (10e) (RT 6.19 min) was eluted as a white solid with 4% (v/v) Et0Ac in petroleum ether (54 mg, 7.2%). Recrystallization of  202  10e from Et20-petroleum ether gave flakes with mp = 58-59°C. Elution with 7% (v/v) Et0Ac in petroleum ether afforded pure trans-cyclobutanol, bicyclo[14.2.0] octadecan-7one-16-Hydroxy-(1R*,16S*)-(±) (100 (RT 10.97 mm) as an oil (40 mg, 5.4%). Several attempts made to isolate the compound at RT 12.54 (10c) min by conventional column chromatography were unsuccessful. Based on its GLC retention time relative to the other two photoproducts, compound 10c was assumed to be the cis-cyclobutanol, bic yclo [14.2.0] octadecan-7-one-16-Hydroxy-(1R*,16R*)-(±). Solid state reaction Analytical solid state reaction showed a major peak at RT 10.97 min and two minor peaks at RT's 6.19 and 12.54 min. The solid state products were not isolated but assumed to be the same as the solution counterparts, based on their GLC retention times. Analytical photoreactions at 20°C, 30°C, 40°C, 60°C, 80°C and 90°C were also explored and the product ratios at 0% conversions are tabulated in Table XVI(b). Photolysis of annealed crystals : Crystals annealed to 90°C (above the transition point of 86°C) were photolyzed on an analytical scale as described in the general experimental section. Product ratios at 0% conversion are tabulated in Table XVI(b).  17 octadecene 2,11 dione (10e) -  -  -  1H NMR: (CDC13, 400 MHz) 8 5.8 (1H, m, -CH=CH2 ), 4.9 (2H, m, -CH=CH2 ), 2.4  (6H, m, -CH2-C=0), 2.15 (3H, s, -CH3), 2.05 (2H, two overlapping triplets, -Cf_12CH=CH2), 1.55 (6H, m, methylenes), 1.5-1.2 (8H, m, methylenes). 13C NMR:(CDC13, 75 MHz) 8 211.5 and 209.3 (2 C=0), 138.8 (-H=CH2), 114.3  (-CH=CH2), 43.7 (CH2), 42.7 (CH2), 42.6 (CH2), 33.5 (CH2), 29.8 (CH3), 29.6 (CH2), 29.2 (CH2), 29.1 (CH2), 29.0 (CH2), 28.7 (CH2), 28.6 (CH2), 23.7 (CH2), 23.6 (CH2). IR (KBr): 3081, 2931, 1704 (C=0), 1645 cm-1. MS: mie (rel.intensity) 280 (M+, 0.3), 97 (20), 58 (26), 55 (100), 43 (96), 41(42).  203  Calculated mass: 280.2403, found: 280.2397. Anal. calcd for C18143202: C, 77.14; H, 11.43. found:C, 77.19; H, 11.57. bicyclo[14.2.0] octadecan-7-one-16-Hydroxy-(1R*,16S*)-(±) (10t) 13C NMR: (C6D6, 75 MHz) 8 210.0 (C=0), 76.7 (C-OH), 42.2 (methine carbon), 42.0 (CH2), 41.7 (CH2), 41.5 (CH2), 34.4 (CH2), 30.2 (CH2), 28.6 (CH2), 28.5 (CH2), 28.1 (CH2), 28.0 (CH2), 27.0 (CH2), 26.9 (CH2), 24.1 (CH2), 22.9 (CH2), 22.7 (CH2), 21.1 (CH2).  IR (neat): 3600-3300 (OH), 2929, 1708 (C=0) cm-1. MS: nVe (rel.intensity) 280 (Mt, 1.0), 252 (4), 97 (28), 83 (28), 69 (26), 55 (100), 43 (45), 41(44).  Calculated mass: 280.2403, found: 280.2396. Photolysis of Cycloeicosan-1,11-dione (11): Irradiation in hexane : Compound (11) (1.9 g) was preparatively photolyzed in hexane using the standard procedure described for compound 7. Analysis of the reaction mixture on GLC (DB 17, 200 °C) showed three new peaks at RT's 5.29, 9.05, 9.87 min. Products were purified by silica gel column chromatography (Et0Ac: petroleum ether (35-60°C) = 2:98 to 9:91) and were identified by spectral analysis. An ene-dione, 19-eicosene-2,12-dione (11e) (RT 5.29 min) was eluted as a white solid with 2.5% (v/v) ethyl acetate in petroleum ether (115 mg, 6.2%). Recrystallization of lie from Et20/petroleum ether gave flakes with mp = 68-69°C. Elution with 6% (v/v) Et0Ac in petroleum ether gave a trans-cyclobutanol, bicyclo[16.2.0]eicosan-8-one, - (11t). Evaporation of the solvent initially gave an oil, but 18-Hydroxy-(1R*,18S*)-CL) this slowly solidified at room temperature to give a white solid (50 mg, 2.7% yield). Recrystallization of lit from EtO/pet-ether gave prisms with mp = 31-32°C. A cis-  204  cyclobutanol, bicyclo[16.2.0]eicosan-8-one,18-Hydroxy-(1R*,18R*)-(-) (11c) was eluted as a white solid with 7% (v/v) Et0Ac in petroleum ether (15 mg, 0.8% yield) with mp = 54-55°C. Solid state reaction:  Analytical studies indicated the development of only one major peak on GLC (DB17, 200°C) at RT 9.94 min with slight traces of other peaks. Preparative solid state photolysis of diketone 11 (1.53 g) was carried out as described in the general experimental section. After 8% total conversion, the product was purified by silica gel column chromatography (Et0Ac:petroleum ether (35-60°C) = 7:93) as a white solid (67 mg, 44%) and was later identified as compound lit. For the trans-cyclobutanol photoproduct llt, the stereochemistry at the ring junction was confirmed by an X-ray crystal structure. 19-eicosene-2,12-dione (11e) 1H NMR: (C6D6, 200 MHz) 8 5.6 (1H, m, vinylic), 4.9 (2H, m, vinylic), 2.0 (8H, m, methylenes), 1.7 (3H, s, -CH3), 1.5 (6H, m, methylenes), 1.2 (16H, bs, methylenes). 13C NMR: (C6D6, 50 MHz) 5 213.0 and 218.0 (2 C=0), 139.1 (-CH=CH2), 114.5 (-CH=CH2), 43.4 (-CH2), 42.63 (CH2), 42.61 (CH2), 34.1 (CH2), 29.8 (CH3), 29.7 (CH2), 29.6 (CH2), 29.5 (CH2), 29.46 (CH2), 29.41 (CH2), 29.3 (CH2), 29.2 (CH2), 29.1 (CH2), 24.1 (CH2), 24.03 (CH2), 24.02 (CH2). IR (KBr): 3078, 2931, 1705 (0=0), 1645 cm4. MS: tnie (rel.intensity) 308 (M+, 0.5), 71(47), 69 (74), 58 (43), 55 (93), 43 (100), 41 (54). Calculated mass: 308.2717, found: 308.2714. Anal. calcd for C20H3602: C, 77.87; H, 11.76. found: C, 77.84; H, 11.76.  205  bicyclo[16.2.0]eicosan-8-one, 18-Hydroxy-(1R*,18S*)-(±,) (11t) 13C NMR: (C6D6, 50 MHz) 5 210.0 (C=0), 77.3 (C-OH), 43.1 (methine carbon), 42.3  (CH2), 42.1 (CH2), 42.0 (CH2), 33.9 (CH2), 29.9 (CH2), 28.97 (CH2), 28.96 (CH2), 28.6 (CH2), 28.32 (CH2), 28.28 (CH2), 28.23 (CH2), 27.5 (CH2), 27.1 (CH2), 24.0 (CH2), 23.7 (CH2), 22.9 (CH2), 21.3 (CH2). IR (KBr): 3600-3300 (OH), 2932, 1708 (C=0) cm-1. MS: mie (rel.intensity) 308 (Mt, 2), 280 (2), 71(25), 69 (37), 58 (27), 55 (100), 43 (63),  41(64). Calated mass: 308.2717, found: 308.2724. Anal. calcd for C201-13602: C, 77.87; H, 11.76. found: C, 77.71; H, 11.81.  bicyclo[16.2.0]eicosan-8-one, 18-Hydroxy-(1R*,18R*)-(±) (11c) 13C NMR: (C6D6, 50 MHz) 5 210.0 (C=0), 76.7 (C-OH), 50.1 (methine carbon), 42.4  (CH2), 42.1 (CH2), 33.8 (CH2), 33 5 (CH), 29.8 (CH2), 29.6 (CH2), 29.2 (CH2), 28.5 (CH2), 28.4 (CH2), 28.2 (CH2), 28.0 (CH2), 27.7 (CH2), 27.3 (CH2), 24.2 (CH2), 23.6 (CH2), 22.1 (CH2), 19.1 (CH2). IR (KBr): 3400-3600 (OH), 2922, 1709 (C=0) cm-1. MS: mie (rel.intensity) 308 (M-I-, 5), 280 (2), 149 (25), 111 (46), 98 (71), 81(46), 69  (43), 55 (100), 43 (50), 41(83). Calculated mass: 308.2717, found: 308.2763. Anal. calcd for C201-13602: C, 77.87; H, 11.76. found: C, 77.81; H, 11.80.  Photolysis of cyclodocosane-1,12-dione (12)  Irradiation in hexane:  Diketone 12 (820 mg) was preparatively photolyzed in hexane using the standard procedure described for compound 7. The reaction mixture on GLC (DB 17, 235°C)  206  showed three new peaks at RT's 5.92, 9.53, 10.58 min. Products were purified by silica gel column chromatography (Et0Ac:petroleum ether = 3:97 to 7:93) and identified by spectral analysis. An ene-dione 21-docosene-2,13-dione (12e) (RT 5.92 min), was eluted as a white solid with 3.5% (v/v) Et0Ac in petroleum ether (87 mg, 10.6% yield). Recrystallization of 12e from Et20/petroleum ether gave flakes with mp = 72-73 °C. A trans-cyclobutanol, bicyclo[18.2.01docosan-9-one,20-Hydroxy-(1R*,20S*)-(±) (12t) (RT 9.53 mm) was eluted as an oil with 6% (v/v) Et0Ac in petroleum ether (29 mg, 3.6% yield). A ciscyclobutanol, bic yclo [18.2.0] docosan-9- one,20-Hydroxy-(1R*,20R*)-(±) (12c) (RT 10.58 min) was eluted as an oil with 6.5% (v/v) Et0Ac in petroleum ether (21 mg, 2.5% yield).  Solid state irradiation: Preparative solid state photolysis of diketone 12 was carried out as described in the general procedure. GLC analysis indicated the development of one major peak (RT 9.37 min) and traces of several minor peaks. After 15% conversion to products the final mixture was subjected to silica gel column chromatography and the major product was isolated (64 mg, 8.5% yield) and identified as cyclobutanol 12t. 21 docosene 2,13 dione (12e) -  -  -  111 NMR: (CDC13, 400 MHz) 5 5.8 (1H, m, vinylic), 4.9 (2H, m, vinylic), 2.4 (6H, m, -CH2-C=0), 2.1 (3H, s, -CH3), 2.0 (2H, two overlapping triplets, -C112-CH=CH2), 1.6 (6H, bs, methylenes), 1.3 (20H, bs, methylenes). 13C NMR: (C6D6 75 MHz) 5 208.8 and 206.3 (2 C=0), 139.2 (-fH=CH2), 114.6 (-CH=1H2), 43.4 (CH), 42.7 (CH2), 34.2 (CH2), 29.9 (-CH3), 29.8 (CH2), 29.7 (CH2), 29.63 (CH2), 29.57 (CH2), 29.4 (CH2), 29.3 (CH2), 24.2 (CH2), 24.1 (CH2). JR (KBr): 3081, 2917, 1705 (C=0), 1645 cm4.  207  MS: m/e (rel.intensity) 336 (Mt, 2), 125 (31), 109 (26), 95 (39), 83 (48), 81(39), 71  (45), 69 (71), 67 (33), 58 (31), 55 (83), 43 (100), 41(58). Calculated mass: 336.3030, found: 336.3026. Anal. calcd for C22H4002: C, 78.51; H, 11.98. found: C, 78.50; H, 11.92.  bicyclo[18.2.0]docosan-9-one,20-Hydroxy-(1R*,20S*)-Ct) (12t) 13C NMR: (C6D6, 75 MHz) 5 209.7 (C=0), 76.6 (C-OH), 43.4 (methine carbon), 42.4  (CH2), 42.2 (CH2), 42.1 (CH2), 34.2 (CH2), 30.7 (CH2), 29.4 (CH2), 29.1 (CH2), 29.0 (CH2), 28.8 (CH2), 28.73 (CH2), 28.68 (CH2), 28.3 (CH2), 27.9 (CH2), 27.3 (CH2), 24.0 (CH2), 23.5 (CH2), 22.9 (CH2), 21.0 (CH2). IR (neat): 3600-3300 (OH), 2928, 1709 (0=0) cm-1. MS: m/e (rel.intensity) 336 (Mt, 7), 308 (14), 71(25), 69 (37), 58 (27), 55 (100), 43  (63), 41(64). Calculated mass: 336.3030, found: 336.3027.  bicyclo[18.2.0]docosan-9-one,20-Hydroxy-(1R*,20R*)-a) (12c) 13C NMR:(C6D6, 75 MHz) 5 210.0 (0=0), 76.7 (C-OH), 43.4 (methine carbon), 42.5  (CH2), 42.2 (CH2), 42.1 (CH2), 34.2 (CH2), 30.7 (CH2), 29.5 (CH2), 29.1 (CH2), 29.0 (CH2), 28.8 (CH2), 28.7 (CH2), 28.3 (CH2), 28.0 (CH2), 27.3 (CH2), 24.1 (CH2), 23.5 (CH2), 23.0 (CH2), 21.1 (CH2). IR (neat): 3600-3400 (OH), 2922, 1704 (C=0) cm-1. MS: m/e (rel.intensity) 336 (Mt, 1), 308 (4), 83 (31), 69 (42), 67 (31), 58 (31), 55 (100),  43 (54), 41(58). Calculated mass: 336.3030, found: 336.3022.  208  Photolysis of cyclotetracosane-1,13-dione (13)  Irradiation in hexane :  Diketone 13 (1.2 g) was preparatively photolyzed using the standard procedure described for compound 7. The reaction mixture on GLC (DB 17, 245 °C) indicated three new peaks at RT's 6.33, 10.07 and 11.01 min. Products were purified by silica gel column chromatography (Et0Ac:petroleum ether = 3:97 to 7:93) and characterized by spectral analysis. An ene-dione, 23-tetracosene-2,14-dione (13e) (RT 6.33 nun) was eluted as a white solid with 3% (v/v) Et0Ac in petroleum ether (107 mg, 8.9% yield). Recrystallization of 13e from Et20/petroleum ether gave a powdery solid with mp = 7880°C. A trans-cyclobutanol, bicyclo[20.2.0]tetracosan-10-one, 22-hydroxy-(1R*,22S*)(±) (13t) (RT 10.07 min) was eluted as an oil with 6.5% (v/v) Et0Ac in petroleum ether (50 mg, 4.2% yield). A cis-cyclobutanol, bicyclo[20.2.0]tetracosan-10-one, 22-hydroxy(1R*,22R*)-(±) (13c) (RT 11.01 min) was eluted as an oil with 7% (v/v) Et0Ac in petroleum ether (30 mg, 2.5% yield).  Solid state reaction  Analytical solid state photolysis indicated the development of one major peak on GLC (DB 17, 245 °C) at RT 11.05 min with traces of minor peaks. Preparative solid state  photolysis of compound 13 (600 mg) was carried out as described in the general procedure upto 18% conversion to products. The major product was purified by silica gel column chromatography (Et0Ac: petroleum ether = 7:93) as an oil (63 mg, 10.5% yield) and was later identified as the cis-cyclobutanol 13c. The product ratio at 0% conversion is tabulated in Table IV.  209  23-tetracosene-2,14-dione (13e) 1H NMR: (CDC13, 200 MHz) 8 5.9 (1H, m, vinylic), 4.95 (2H, m, vinylic), 2.4 (6H, m,  -CH2-C=0), 2.1 (3H, s, -CH3), 2.0 (2H, two overlapping triplets, -CE2-CH=CH2), 1.5 (6H, bs, methylenes), 1.1 (24H, bs, methylenes). 13C NMR: (CDC16, 50 MHz) 8 211.5 and 209.1 (2C=0), 139.1 (-fH=CH2), 103.1  (-CH=fH2), 43.7 (CH2), 42.7 (CH2), 33.7 (CH2), 29.8 (CH3), 29.5 (CH2), 29.4 (CH2), 29.3 (CH2), 29.25 (CH2), 29.21 (CH2), 29.12 (CH2), 29.01 (CH2), 28.8 (CH2), 23.8 (CH2). IR (KBr): 3080, 2931, 1705 (C=0), 1646 cm-1. MS: m/e (rel.intensity) 364 (Mt 2), 125 (28), 109 (27), 97 (30), 95 (41), 83 (62), 81  (45), 71(50), 69 (48), 67 (34), 58 (32), 57 (33), 55 (100), 43 (85), 41(52). Calculated mass: 364.3343, found: 364.3330. Anal. calcd for C24H4402: C, 79.06; H, 12.16. found:C, 79.00; H, 12.19.  bicydo[20.2.0]tetracosan-10-one, 22-hydroxy-(1R*,22S*)-(±) (13t) 13C NMR: (C6D6, 50 MHz) 8 210.0 (0=0), 77.0 (f-OH), 43.5 (methine carbon), 42.3  (CH2), 42.2 (CH2), 42.1 (CH2), 34.0 (CH2), 30.2 (CH2), 29.6 (CH2), 29.4 (CH2), 29.1 (CH2), 29.0 (CH2), 28 71 (CH2), 28.67 (CH2), 28.63 (CH2), 28.60 (CH2), 27.4 (CH2), 24.1 (CH2), 23.9 (CH2), 23.4 (CH2), 21.1 (CH2). IR (neat): 3650-3400 (OH), 2925, 1698 (C=0) cm-1. MS: ink (rel.intensity) 364 (Mt 7), 336 (12), 111 (25), 97 (29), 95 (43), 83 (49), 81  (44), 71(39), 69 (47), 67 (37), 58 (29), 57 (34), 55 (100), 43 (33), 41(48). Calculated mass: 364.3343, found: 364.3355.  bicyclo[20.2.0]tetracosan-10-one, 22-hydroxy-(1R*,22R*)-(±) (13c) 13C NMR: (C6D6, 50 MHz) 5 209.5 (C=0), 76.7 (a-OH), 50.1 (methine carbon), 42.5  (CH2), 42.2 (CH2), 34.0 (CH2), 33.9 (CH2), 30.2 (CH2), 29.83 (CH2), 29.79 (CH2),  210  29.54 (CH2), 29.2 (CH2), 29.14 (CH2), 29.10 (CH2), 29.02 (CH2), 28.85 (CH2), 28.79 (CH2), 28.76 (CH2), 28.69 (CH2), 28.1 (CH2), 24.1 (CH2), 24.0 (CH2), 22.7 (CH2), 19.2 (CH2). IR (neat): 3600-3440 (OH), 2923, 1708 (C=0) cm-1. MS: m/e (rel.intensity) 364 (Mt, 3), 336 (6), 97 (27), 95 (41), 83 (45), 81(46), 71(36),  69 (44), 67 (39), 58 (27), 57 (31), 55 (100), 43 (27), 43 (37), 41(51). Calculated mass: 364.3343, found: 364.3353.  Photolysis of cyclohexacosane-1,14-dione (14)  Solid state irradiation : Photolysis of the two crystal modifications of diketone 14, plates and needles, indicated the development of two new peaks corresponding to the photoproducts on GLC analysis. Plates upon irradiation showed a major peak on GLC (DB 17, 260°C) at RT 10.30 min and a minor peak at RT 9.47 min, where as needles gave the above peaks at RT's 9.43 and 10.26 min respectively.  Plate crystals (14P) Preparative solid state photolysis was carried out with 600 mg of diketone as described in the general procedure. The major photoproduct was isolated by column chromatography(silica gel, Et0Ac: petroleum ether = 9:91) as a white solid (73 mg, 12.2% yield, mp = 42-43°C) and was later identified as a cis-cyclobutanol derivative, bicyclo[22.2.0Thexacosan-11-one, 24-Hydroxy-(1R*,24R*)-(_±) (14c).  Needle crystals (14N) Irradiation was carried out with 750 mg of diketone and the major photoproduct was isolated by column chromatography (silica gel, Et0Ac:petroleum ether = 8:92) as an oil (47 mg, 6.3% yield). The product was identified as a trans-cyclobutanol derivative, bicyclo[22.2.0Thexacosan-11-one, 24-Hydroxy-(1R*,24S*)-(±) (14t).  211  Irradiation in hexane:  Diketone 13 (800 mg) was preparatively photolyzed using the standard procedure described for diketone 7. The reaction mixture on GLC (DB 17, 260°C) indicated the development of three new peaks at RT's 6.03, 9.43 and 10.40 min. After 17% total conversion, products were purified by silica gel column chromatography (Et0Ac:petroleum ether = 3:97 to 9:91) and characterized by spectral analysis. Elution with 6% (v/v) Et0Ac in petroleum ether yielded 58 mg (7.2%) of 25-hexacosene-2,15-dione (14e) (RT 6.03 min) as a white solid. Recrystallization from Et20/petroleum ether gave a powder with mp = 82-83°C. Compounds corresponding to peaks at RT's 9.43 and 10.40 min were eluted together with 8% (v/v) Et0Ac in petroleum ether, and were shown by co-injection experiments to have retention times identical to those of 14t and 14c, respectively. A solution of this mixture in a Et20/petroleum ether solvent system afforded a white solid upon cooling in ice. After filtration, GLC analysis of this solid showed it to contain 82% of the compound corresponding to RT 10.40 min. Several fractional recrystallizations from Et20/petroleum ether gave 97% pure white solid (6 mg, 1.2%), which was later identified as cis-cyclobutanol 14c. Spectral analysis of the rest of the mixture obtained from the combined mother liquors revealed the compound with RT 9.43 min to be the trans-cyclobutanol 14t. Analytical photolysis :  Plate crystals annealed to 60°C (5 cycles) were powdered and subjected to irradiation at 20°C; development of a single major peak at RT 9.43 min together with a minor peak at RI 10.25 min were observed on GLC (DB 17, 260°C). Analytical photoreactions of virgin crystals (plates and needles) were also explored at 0°C, 20°C, 40°C, 60°C and 65°C. Product ratios of both the plates and the needles at 60°C and above were found to be close to that observed for needles and annealed plates at 20°C. The peaks at RT's 9.43 min and 10.25 min were assumed to be compounds 14t and 14c  212  respectively, based on their GLC retention times. Product ratios at 0% percent conversion for all the above reactions are tabulated in Table XIV.  25-hexacosene-2,15-dione (14e) 1H NMR: (CDC13, 200 MHz) 8 5.8 (1H, m, -CH=CH2), 4.9 (2H, m, -CH=CH2), 2.4  (6H, m, -CH2-C=0), 2.1 (3H, s, -CH3), 2.0 (2H, two overlapping triplets, -CLI2CH=CH2), 1.6 (6H, bs, methylenes), 1.2 (24H, bs, methylenes). 13C NMR: (CDC16, 75 MHz) 5 211.6 and 209.3 (2 C=0), 139.1 (-CH=CH2), 114.0  (-CH=QH2), 43.8 (CH2), 42.8 (CH2), 33.7 (CH2), 29.8 (-CH3), 29.5 (CH2), 29.3 (CH2), 29.21 (CH2), 29.17 (CH2), 29.13 (CH2), 29.0 (CH2), 28.9 (CH2), 23.8 (CH2). IR (KBr): 3074, 2917, 1705 (C=0), 1645 cm4. MS: m/e (rel.intensity) 392 (M-F, 13), 254 (54), 239 (25), 196 (35), 125 (39), 111 (27),  109 (29), 97 (48), 96 (35), 95 (51), 83 (58), 81(47), 71(57), 69 (52), 67 (28), 58 (41), 55 (100), 43 (81), 41(39). Calculated mass: 392.3656, found: 392.3653. Anal. calcd for C26H4802: C, 79.53; H, 12.32. found: C, 79.40; H, 12.40.  bicyclo[22.2.0]hexacosan-11-one, 24-Hydroxy-(1R*,24S*)-(±) (14t). 13C NMR: (C6D6, 75 MHz) 8 209.6 (C=0), 76.7 (.-OH), 43.7 (methine carbon), 42.5  (CH2), 42.2 (CH2), 42.1 (CH2), 34.2 (CH2), 30.4 (CH2), 29.9 (CH2), 29.5 (CH2), 29.4 (CH2), 29.3 (CH2), 29 2 (CH2), 29.13 (CH2), 29.10 (CH2), 29.0 (CH2), 28.8 (CH2), 27.6 (CH2), 24.1 (CH2), 23.8 (CH2), 23.4 (CH2), 20.9 (CH2). IR (neat): 3600-3300 (OH), 2926, 1713 (C=0) cm1. MS: m/e (rel.intensity) 392 (Mt 5), 364 (11), 125 (25), 109 (29), 97 (37), 96 (25), 95  (51), 83 (46), 81(51), 71(39), 69 (50), 67 (40), 58 (27), 57 (36), 55 (100), 43 (32), 41 (46). Calculated mass: 392.3656, found: 392.3649.  213  bicyclo[22.2.0]hexacosan-11-one, 24-Hydroxy-(1R*,24R*)-(±) (14c). 13C NMR: (C6D6, 75 MHz) 5 204.2 (C=0), 76.7 (C-OH), 50.1 (methine carbon), 42.2  (CH2), 42.16 (CH2), 34.2 (CH2), 34.0 (CH2), 30.1 (CH2), 29.9 (CH2), 29.31 (CH2), 29.27 (CH2), 29.12 (CH2), 29.08 (CH2), 29.04 (CH2), 28.99 (CH2), 28.9 (CH2), 28.73 (CH2), 28.66 (CH2), 28.0 (CH2), 24.0 (CH2), 23.9 (CH2), 22.8 (CH2), 19.2 (CH2). IR (KBr) 3560-3300 (OH), 2927, 1713 (C=0) cm-1. MS: m/e (rel.intensity) 392 (Mt, 4), 364 (11), 109 (25), 97 (36), 83 (47), 81(41), 71  (42), 69 (51), 67 (51), 58 (32), 57 (33), 55 (100), 43 (33), 43 (28), 41(45). Calculated mass: 392.3656, found: 392.3642. Anal. calcd for C26H4802: C, 79.53; H, 12.32. found:C, 79.26; H, 12.40.  6.2.3. Photochemistry of Non-diametric Diketone. Photolysis of Cyclohexadecane-1,8-dione (15) Irradiation in hexane:  Diketone 15 (1.5 g) was preparatively photolyzed using the standard procedure described for diketone 7. The reaction mixture on GLC (DB 17, 175°C) showed four new peaks at RT's 6.39, 11.90, 13.10 and 13.51 min. Products were purified by silica gel column chromatography (Et0Ac:petroleum ether(v/v) = 3:97 to 12:88) and characterized by spectral analysis. The product corresponding to the peak at RT 6.39 min was eluted as a white solid with 3.5% (v/v) Et0Ac in petroleum ether (35-60 °C) (68 mg, 4.5% yield). Recrystallization from Et20/petroleum ether solvent system gave flakes with mp = 4142°C. Spectral analysis (1H NMR and 13C NMR) revealed this solid to be a mixture of two regioisomers, 15-hexadecene-2,9-dione (15e1) and 15-hexadecene-2,11-dione (15e2). The peak at RT 11.91 min was eluted as an oil with 7% (v/v) EtOAc in petroleum ether  214  (35-60°C) (72 mg, 5.0% yield), and was later identified as a mixture of two  trans-  cyclobutanol regioisomers, bicyclo[12.2.0]hexadecan-7-one,14-Hydroxy-(1R*,14S*)-(±) (15t1) and bicyclo[12.2.0] hexadecan-5-one,14-Hydroxy-(1R*,14S*)-(±) (15t2). Attempt  made to separate the regioisomers of the cleavage products (15e1 and 15e2) or the transcyclobutanols (15t1 and 15t2) were not successful. Elution with 7.5% (v/v) Et0Ac in petroleum ether gave pure cis-cyclobutanol, bicyclo[12.2.0]hexadecan-7-one,14-Hydroxy(1R*,14R*)-(±) (15c1) (RT 13.51 min) as a white solid (23 mg, 1.5% yield). Recrystallization of 15c1 from hexane gave prism and needle shaped crystals in the same batch. Manually separated needle and plate crystals showed identical melting points (122123°C) and infrared spectra. Further elution (75 fractions) gave a mixture of 15c1 and the compound (15c2) corresponding to the peak at RT 13.10 min. In the last few fractions, pure 15c2 was isolated and was later identified as a regioisomer of compound 15c1, bicyclo[12.2.0]hexadecan-5-one,14-Hydroxy-(1R*,14R*)-(±) (15c2). Recrystallization of 15c2 gave colorless plates (18 mg, 1.2% yield) with mp = 82-83°C.  The regiochemistry and the stereochemistry at the ring junction of the ciscyclobutanols were determined by X-ray crystallographic analysis. Solid state reaction : Preparative solid state photolysis of compound 15 (650 mg) was carried out as described in the general procedure. GLC (DB 17, 175°C) analysis indicated the development of a single major product at RT 13.50 min with traces of minor peaks. The major product was purified (81 mg, 13% yield), and was later identified as the ciscyclobutanol 15c1. Analytical irradiation: Photoreactions of virgin crystals were also explored at 0°C, 10°C, 20°C, 30°C, 40°C and 60°C. Crystals were also annealed to 40°C (above the first transition point at 37°C) and 60°C (above the second transition point at 55°C), and were subjected to  215  irradiation at 20°C. Product ratios at 0% conversion were calculated and compiled in Table XVIc.  15-hexadecene-2,9-dione (10e1) and 15-hexadecene-2,11-dione (15e2): 1H NMR: (C6D6, 200 MHz) 5 5.8-5.6 (2H, m, -CH=CH2), 5.0 (4H, m, -CH=CE2), 2.0 (16H, m, methylenes), 1.7 (6H, two singlets, -CH3), 1.5 (12H, m, methylenes), 1.1 (16H, m, methylenes). 13C NMR: (C6D6, 100 MHz) 5 208.7, 208.5, 206.3 and 206.2 (4 C=0), 139.0, 138.4 (2-CH=CH2), 115.1 and 114.6 (2 -CH=CH2), 43.3 (CH), 43.2 (CH2), 42.6 (CH2), 42.53 (CH2), 42.47 (CH2), 41.7 (CH2), 35.0 (CH2), 33.9 (CH2), 33.4 (CH), 29.62 (CH2), 29.61 (CH2), 29.5 (CH2), 29.4 (CH2), 29.3 and 29.2 (2 -CH3), 29.03 (CH2), 29.0 (CH2), 24.03 (CH2), 23.98 (CH2), 23.89 (CH2), 23.86 (CH2), 23.81 (CH2), 23.1 (CH2), 23.0 (CH2). IR (KBr): 3082, 2932, 1705 (C=0), 1645 cm4. MS: ink (rel.intensity) 252 (Mt, 2), 109 (25), 97 (25), 69 (30),55 (60), 43 (100), 41 (44). Calculated mass: 252.2089, found: 252.2097. Anal. calcd for C16H2802: C, 76.14; H, 11.18. found:C, 76.19; H, 11.19.  bicyclo[12.2.0]hexadecan-7-one,14-Hydroxy-(1R*,14S*)-(1) (15t1) and bicyclo [12.2.0] hexadecan-5-one,14-Hydroxy-(1R*,14S*)-(±) (15t2) : 13C NMR: (C6D6, 50 MHz) 5 210.24 and 210.21 (2 C=0), 77.4 and 77.3 (2 -0H), 41.8 (CH2), 41.31 (CH2),41.29 (CH2), 40.92 (CH2), 40.84 (methine), 40.8 (CH2), 40.7 (CH2), 40.6 (CH2), 40.5 (CH2), 34.4 (CH2), 34.2 (CH2), 29.6 (CH2), 28.4 (CH2), 27.8 (CH2), 27.42 (CH2), 27.37 (CH2), 27.30 (CH2), 26.9 (CH2), 26.5 (CH2), 26.4 (CH2), 24.2 (CH2), 23.9 (CH2), 23.8 (CH2), 22.6 (CH2), 22.0 (CH2), 21.2 (CH2), 21.1 (CH2), 20.2 (CH2).  216  IR (neat): 3600-3200 (OH), 2936, 1708 (C=0) cm-1. MS: mile (rel.intensity) 252 (Mt 9), 224 (20), 125 (37), 113 (31), 112 (66), 111 (38), 97  (70), 95 (37), 85 (28), 84 (92), 83 (44), 81(37), 71(40), 69 (63), 67 (34), 58 (42), 57 (29), 55 (87), 43 (100),41 (79). Calculated mass: 252.2089, found: 252.2092.  bicyclo[12.2.0]hexadecan-7-one,14-Hydroxy-(1R*,14R*)-(±) (15c1) 13C NMR: (C6D6, 100 MHz) 5 210.0 (C=0), 76.2 (C-OH), 50.0 (methine), 43.2 (CH2),  38.0 (CH2), 34.3 (CH2), 31.2 (CH2), 29.2 (CH2), 27.6 (CH2), 26.9 (CH2), 25.7 (CH2), 25.3 (CH2), 23.1 (CH2), 21.4 (CH2), 19.4 (CH2), 19.0 (CH2). IR (KBr): 3417, broad (OH), 2927, 1692 (C=0) cm-1. MS: m/e (rel.intensity) 252 (M1-, 5), 224 (3), 125 (40), 113 (26), 113 (53), 112 (65), 109  (42), 107 (25), 98 (43), 97 (66), 95 (52), 94 (40), 85 (42), 84 (58), 83 (58), 82 (40), 81 (48), 79 (34), 71(49), 69 (65), 67 (49), 58 (51), 55 (100), 53 (29), 43 (70), 41(73), 39 (34). Calculated mass: 252.2089, found: 252.2097. Anal. ca1cd for C16H2802: C, 76.14; H, 11.18. found: C, 76.01; H, 11.21.  bicyclo[12.2.01hexadecan-5-one,14-Hydroxy-(112*,14R*)-(±) (15c2) 13C NMR: (C6D6, 50 MHz) 5 209.8 (CO), 76.6 (C-OH), 50.0 (methine), 42.0 (CH2),  40.6 (CH2), 35.2 (CH2), 35.0 (CH2), 29.9 (CH2), 26.64 (CH2), 26.59 (CH2), 26.0 (CH2), 24.7 (CH2), 24.3 (CH2), 21.9 (CH2), 20.7 (CH2), 18.6 (CH2). IR (KBr): 3417, broad (OH), 2928, 1692 (C=0) cm4. MS: m/e (rel.intensity) 252 (Mt 5), 224 (2), 125 (44), 113 (26), 112 (51), 95 (38), 85  (30), 84 (79), 83 (51), 81(39), 79 (25), 71(44), 69 (69), 67 (32), 58 (33), 55 (100). Calculated mass: 252.2089, found: 252.2087. Anal. calcd for C16H2802: C, 76.14; H, 11.18. found: C, 76.00; H, 11.05.  217  For photoproducts 15c1 and 15c2 the stereochemistry at the ring junction was confirmed by X-ray crystallographic analysis. Photolysis of Cycloheptadecane-1,10-dione (16) Solid state irradiation:  Diketone 16 was preparatively photolyzed in the solid state using the procedure described in the general experimental section. The reaction mixture on GLC (DB 17, 195°C) showed three major peaks at RTs 5.19, 9.28 and 10.46 min. Products were purified by silica gel column chromatography (Et0Ac:petroleum ether (35-60°C) = 3:97 to 12:88) and characterized by spectral analysis. A mixture of two regioisomers, 16-heptadecene-2,10-dione (16e1) and 16heptadecene-2,11-dione (16e2) (RT 5.19 min), was eluted with 3.5% (v/v) Et0Ac in petroleum ether as a white solid (65 mg, 7% yield). Recrystallization from Et20/petroleum ether gave flakes with mp = 47-48°C. A mixture of trans-cyclobutanol regioisomers, bicyclo[13.2.0]heptadecan-7-one, 15-Hydroxy-(1R*,15S*)-(±) (16t1) and bicyclo[13.2.0Theptadecan-6-one, 15-Hydroxy-(1R*,15S*)-(±) (16t2) (RT 9.28 min), was eluted as an oil with 10% (v/v) Et0Ac in petroleum ether (30 mg, 3.2% yield). A mixture of cis-cyclobutanol regioisomers, bicyclo[13.2.0]heptadecan-7-one, 15-hydroxy(1R*,15R*)-(±) (16c1) and bicyclo[13.2.0]heptadecan-6-one, 15-hydroxy-(1R*,15R*)(±) (16c2) (RT 10.46 min), was eluted as an oil with 12% (v/v) Et0Ac in petroleum ether (35-60°C) (27 mg, 2.9% yield). Attempts made to separate the above regioisomers were unsuccessful. Irradiation in hexane:  Analytical solution state photolysis was carried out using the procedure described in the general experimental section and monitored by GLC (DB 17, 190°C). As observed in solid state reaction, three major peaks were observed at RTs 5.99, 9.20 and 10.36 min.  218  Preparative scale irradiation was not performed but the products observed on GLC were assumed to be the same as the solid state counterparts, based on their GLC retention times.  16-heptadecene-2,10-dione (16e1) and 16-heptadecene-2,11-dione (16e2) 1H NMR: (C6D6, 400 MHz) 5 5.8-5.6 (2H, m, -CH=CH2), 5.1-4.9 (4H, m,  CH=CH2), 2.0 (16H, m, methylenes), 1.7 (6H, two singlets, -CH3), 1.5 (12H, m, methylenes), 1.1 (20H, m, methylenes). 13C NMR: (C6D6, 50 MHz) 5 203.4 (0=0), 201.0 (C=0), 139.0, 138.7 (2 -fH=CH2),  114.7 and 114.6 (2 -CH=fH2), 43.31 (CH2), 43.28 (CH2), 42.6 (CH2), 42.4 (CH2), 33.92 (CH2), 33.87 (CH2), 29.62 (CH2), 29.60 (CH2), 29.5 (CH2), 29.4 (CH2), 29.3 and 29.2 (2 -CH3), 29.01 (CH2), 28.98 (CH2), 29.8 (CH2), 24.0 (CH2), 23.9 (CH2), 23.89 (CH2), 23.87 (CH2), 23.5 (CH2). JR (KBr): 3081, 2932, 1704 (C=0), 1644 cm4. MS: mie (rel.intensity) 266 (M+, 8), 198 (42), 183 (34), 141 (40,140 (37), 128 (30), 127  (35), 126 (53), 123 (44), 111 (40), 109 (25), 98 (66), 97 (49), 95 (34), 85 (34), 84 (43), 83 (65), 82 (30), 81(44), 71(61), 69 (48), 68 (45), 67 (43), 59 (31), 58 (50), 57 (34), 56 (28), 55 (90), 43 (100), 40 (69), 38 (33). Calculated mass: 266.2247, found: 266.2251. Anal. calcd for C17H3002: C, 76.64; H, 11.35. found:C, 76.60; H, 11.45.  bicyclo[13.2.0]heptadecan-7-one, 15-Hydroxy-(1R*,15S*)-(±) (16t1) and bicyclo [13.2.0]heptadecan-6-one, 15-Hydroxy-(1R*,15S*)-(±) (16t2) 13C NMR: (C6D6, 75 MHz) 5 210.9 and 210.4 (2 C=0), 77.6 and 77.5(2 f-OH), 42.8  (methine), 42.0 (CH2), 41.91 (CH2), 41.86 (CH2), 41.7 (CH2), 40.6 (methine), 41.4 (CH2), 41.3 (CH2), 34.53 (CH2), 34.49 (CH2), 29.8 (CH2), 29.5 (CH2), 28.4 (CH2), 28.2  219  (CH2), 28.1 (CH2), 27.8 (CH2), 27.3 (CH2), 27.1 (CH2), 26.9 (CH2), 24.3 (CH2), 23.6 (CH2), 23.4 (CH2), 23.3 (CH2), 22.9 (CH2), 22.8 (CH2), 21.6 (CH2), 21.2 (CH2). IR (neat) 3600-3300 (OH), 2933, 1713 (C=0) cm-1. MS: ink (rel.intensity) 266 (M+, 5), 238 (3), 98 (58), 97 (37), 84 (25), 83 (50), 81(29),  71(39), 69 (35), 67 (32), 58 (34), 57 (30), 55 (100), 43 (81), 41(70). Calated mass: 266.2247, found: 266.2242.  bicyclo[13.2.0]heptadecan-7-one, 15-hydroxy-(1R*,15R*)-W (15c1) and bicyclo [13.2.0]heptadecan-6-one, 15-hydroxy-(1R*,15R*)-(±) (15c2) 13C NMR: (C6D6, 75 MHz) 5 210.4 and 209.8 (2 CO), 77.0 and 76.8 (2 c-OH), 50.0  and 49.8 (methine), 41.7 (CH2), 41.5 (CH2), 41.3 (CH2), 40.7 (CH2), 34.8 (CH2), 34.51 (CH2), 34.47 (CH2), 34.0 (CH2), 30.7 (CH2), 30.3 (CH2), 28.5 (CH2), 28.4 (CH2), 28.0 (CH2), 27.8 (CH2), 27.7 (CH2), 27.4 (CH2), 27.3 (CH2), 27.1 (CH2), 27.0 (CH2), 26.4 (CH2), 24.1 (CH2), 23.3 (CH2), 22.7 (CH2), 22.4 (CH2), 22.1 (CH2), 21.8 (CH2), 19.2 (CH2), 19.0 (CH2). IR (KBr): 3600-3300 (OH), 2941, 1708 (C=0) cm4. MS: mie (rel.intensity) 266 (M+, 10), 238 (4), 140 (28), 127 (26), 126 (41), 125 (39),  123 (29), 122 (26), 112 (37), 111 (50), 109 (35), 108 (28), 107 (28), 99 (25), 98 (75), 97 (61), 96 (28), 95 (47), 94 (25), 93 (31), 85 (36), 84 (53), 83 (72), 82 (39), 81(55), 80 (33), 79 (38), 71(59), 70 (29), 69 (58), 68 (37), 67 (53), 58 (54), 57 (53), 56 (27), 55 (100), 53 (27), 43 (91), 42 (35), 40 (75), 38 (29). Calculated mass: 266.2247, found: 266.2254.  220  6.2.4. Photochemistry of Keto-alcohols. Photolysis of 9-hydroxy-cyclohexadecan-1-one (29):  Analytical solution and solid state photolyses of keto-alcohol 29 were explored at room temperature. In both media, irradiation gave five new peaks on GLC (DB 17, 190°C), at RT's 4.5, 8.7, 8.8, 9.2 and 9.5 min. Preparative scale irradiations were canied out, but several attempts to purify the photoproducts by conventional column chromatography were unsuccessful. The peaks on GLC were identified to a certain extent, as described in Chapter III. Photolysis of 10-hydroxy-cyclooctadecan-1-one (30)  Analytical solution and solid state irradiations were performed. Irradiation gave three new peaks on GLC (DB 17, 205°C) at RT's 5.0, 9.1 and 10.2 min in both media. As in the case for keto alcohol 29, photoproducts could not be separated by preparative scale photolysis. However, the photoproducts were identified using the authentic samples as described for keto-alcohol 29 (see Chapter III for details).  6.2.5. Photochemistry of Tetramethylated diketones. *  *  *  Photolysis of Tetramethylcyclohexadecane-1,9-dione-(2R*,8S , 10R ,16S ) (31).  Solid state irradiation : Preparative solid state photolysis of diketone 31 (200 mg) was carried out using the procedure described in the general experimental section. GLC analysis (DB 17, 190°C) indicated the development of single major product at RT 12.3 min with traces of other minor products. Photolysis was stopped after 30% total conversion to products and the reaction mixture was subjected to HPLC (preparative column, flow rate = 5 ml min 1, Et0Ac:hexane = 2:98). The major photoproduct (RT 45 min - HPLC) was eluded as a  221  white solid (48 mg, 24%) and was later identified as a diastereomer of 31, tetramethylcyclohexadecane-1,9-one-(2R*,8R*,10R*,16S*) (31a). mp = 85-86 °C. Irradiation in hexane : Diketone 31 (300 mg) was preparatively photolyzed using the procedure described for diketone 7. On GLC analysis development of two major peaks (RT's 12.3, 13.1 min) and a minor peak (RT 10.8 min) with traces of several other peaks were observed. After 30% total conversion to products, the reaction mixture was subjected to HPLC. Only the product at RT 10.8 min was purified as a white semi-solid 31b (8 mg, 2.7% yield), but the two major products were eluted together with other impurities. From mass, infrared and 13C NMR spectral analysis, revealed compound 31b was identified as a diastereoisomer of the starting material with high symmetry (four equivalent methyl and methine groups), but the stereochemistry of the molecule has not been fully determined. The compound observed at RT 12.3 min was assumed to be the solid state product 31a based on its GLC retention time. Analytical irradiation: Solution (hexane) and solid state photolyses were performed as described in the general experimental section. Irradiation of the solid for prolonged periods indicated the development of four major peaks at RT's 10.8, 12.3, 13.1 and 13.3 min. After 60 hrs of irradiation the four products along with the starting material 31 (RT 11.3 min) attained a photostationary state. Similar photostationary state was observed in solution after 70 hrs of photolysis, with the absence of peak at RT 13.3 min. The amount of photoproducts (in percentage) at the stationary state with their corresponding retention times are given below. solid : RTs 10.8 (12%), 11.3 (6%), 12.3 (67%), 13.1 (4%), 13.3(11%). hexane : RT's 10.8 (3%), 11.3 (60%), 12.3 (20%), 13.1 (17%).  222  Tetramethylcyclohexadecane-1,9-dione-(2R*,8R*,10R*,16S*) (31a) 13C NMR: (C6D6, 100 MHz) 5 216.0 and 215.9 (2 C=0), 46.2 , 45.5 , 45.2 , 42.2  (4-CH-CH3), 33.4 (CH2), 33.1 (CH2), 32.9 (CH2), 29.2 (CH2), 29.1 (CH2), 27.6 (CH2), 27.19 (CH2), 27.16 (CH2), 26.7 (CH2), 17.9 , 17.8 , 17.0 , 16.5 (4 -CH-H3). IR (neat): 2927, 1703 (C=0) cm-1. MS: rule (rel.intensity) 308 (Mt, 30), 223 (29), 154 (25), 125 (26), 112 (67), 97 (41), 86  (96), 83 (30), 69 (54), 57 (25), 55 (100), 43 (27), 41(68). Calculated mass: 308.2717, found: 308.2715.  Compound (31b) 13C NMR: (C6D6, 75 MHz) 5 216.48 (C=0), 45.57 (-,CH-CH3), 32.57 (CH2), 28.93  (CH2), 26.76 (CH2), 16.5 (-CH-H3). IR (neat): 2927, 1703 (C=0) cm-1. MS: m/e (relintensity) 308 (M+, 32), 223 (45), 195 (19), 155 (22), 154 (25), 153 (31),  125 (43), 112 (75), 111 (22), 99(30), 98(22), 97(65), 96(25), 95 (34), 86(93), 85 (42), 83 (57), 81(31), 71(47), 69 (100), 57 (43), 56 (28), 55 (76). Calculated mass: 308.2717, found: 308.2711.  Photolysis of Tetramethylcyclotetracosane-1,13-dione-(2R*,12S*, 14R*,24S*) (32)  Preparative scale solid state irradiation of diketone 32 (200 mg) was carried out using the procedure described in the general experimental section. Reaction progress was monitored on HPLC (analytical column, flow rate = 0.5 ml min-1, Et0Ac:hexane = 2:98) using the UV detector (X=280 nm). Development of one major peak (RT 14.4 min) with traces of several peaks was observed. After 30% conversion to products, the reaction mixture was subjected to HPLC (preparative column, Et0Ac:hexane, 2:98, flow rate 5.0 mL min-1). The major product was isolated (RT 23.0 min) as a white solid (44 mg,  223  22%) and was later identified as toramethylcyclotetracosane-1,13-dione-(2R*, 12R*, 14R*, 24S*) (32a), mp = 56-58°C. Irradiation in hexane:  Analytical solution reaction of diketone 32 was carried out as described in the general experimental section. A reaction identical to that occurring in the solid state was observed on HPLC (analytical column). A single major product observed was assumed to be 32a, based on it's HPLC retention time.  Tetramethylcyclotetracosane-1,9-dione-(2R*,12R*,14R*,24S*) (32a) 13C NMR: (C6D6, 75 MHz) 8 216.2 and 216.1 (2 C=0), 45.5, 45.4, 44.3, 43.9  (4-fH-CH3), 33.7 (CH2), 33.6 (CH2), 33.5 (CH2), 32.0 (CH2), 30.2 (CH2), 30.1 (CH2), 29.9 (CH2), 29.5 (CH2), 27.6 (CH2), 27.5 (CH2), 23.0 (CH2), 17.4, 17.2, 17.0, 14.4 (4 -CH-H3). IR (neat) 2924, 1701 (C=0) cm-1. MS: mie (rel.intensity) 420 (Mt, 60), 335 (95), 97 (47), 86 (100), 83 (52), 71(33), 69  (80), 57 (64), 56 (28), 55 (87), 43 (42), 41(41). Calculated mass: 420.3959, found: 420.3959. Anal. calcd for C28115202: C, 79.94; H, 12.46. found:C, 79.94; H, 12.56.  63. Quantum Yield Studies. Internal standards and GC detector responses: Quantum yields were measured for the formation of photoproducts from sixteen (9), eighteen (10) and twenty (12) membered diketones. The straight chain alkanes  n-tricosane, n-docosane and n-tetracosane received from Aldrich Chemical Co were  224  purified by several recrystallizations from Et0H/Et20 solvent system and used as internal standards. All samples were analyzed by using gas chromatographs having flame ionization detectors. The detector responses of the photoproducts with respect to the photoinactive internal standards were measured prior to the photolysis and from these studies the exact number of moles of photoproducts formed during the reaction were calculated by analyzing the reaction mixture with a known amount of internal standard. In order to determine the GLC detector response, a solution containing known amounts of a photoproduct and an appropriate internal standard in hexane was injected into the GLC and the ratio of the area under the peaks was compared with the actual ratio of the photoproduct and the internal standard. At least five chromatographic analyses were made for each sample. The following three solutions (table XVIII) containing diketones and internal standards were prepared in hexane using known procedures, where the diketone concentrations were chosen in such away that the solutions are opaque to 313 nm line of the radiation. The GLC detector response of acetophenone with respect to the internal standard n-tetradecane was also determined in order to evaluate the number of moles of acetophenone formed in the valerophenone actinometry. Finally the number of quanta of photons absorbed during the reaction was calculated using the known quantum yield (4)=0.3) of acetophenone formation from 0.1 M valerophenone in benzene (X= 313 nm). A solution containing tetradecane (1 mg mL-1) and valerophenone (0.1 M) in benzene was used as the actinometer. Benzene was used after being freed from thiophene by a procedure described in the literature.222 Apparatus : A merry-go-round apparatus immersed hi a large water bath was used to perform all irradiations.224 The temperature of the apparatus was maintained at 21 ± 2 °C by  225  circulating water through the water bath. A quartz immersion well with a 450 W medium pressure mercury lamp was placed in the middle of the apparatus. The 313 nm line of the radiation from the lamp was isolated by circulating a solution of 2x10-3 M K2Cr04 containing 5% (w/w) K2CO3 through the immersion well and by placing 7-54 Corning filters in the filter holders. Table XVIII: Internal standards.  Solution  Diketone  concentration  Internal standard concentration  (ringsize)  (M)  1  9 (16)  1.2x10-1  tricosane  4.1x10-3  2  10 (18)  9.5x10-2  docosane  2.6x10-3  3  12 (22)  1.4x104  tetracosane  3.9x10-3  (M)  Irradiation of the samples :  The following procedure for diketone 9 is typical. Three Pyrex test tubes (100x13 mm), each containing 3 mL aliquots of solution 1 were degassed by three freeze-pumpthaw cycles and secured with a stopper. These three tubes were placed in the merry-goround apparatus and the irradiation was performed using the 313 nm line of the 450W medium pressure lamp. A parallel irradiation of two valerophenone actinometers, each containing 3 mL aliqots of valerophenone (0.1 M) and tetradecane (1 mg mL-1) were also performed. The percentage conversion of the valerophenone and the diketones was kept below 5%. The total duration of the photolysis was 12 hr, but altogether four sets (two tubes in each set) of actinometers were used one after the other. At the end, each sample was analyzed on GLC at least three times. The peak area ratios of the photoproducts with respect to the internal standards were calculated and the statistical average was taken.  226  The relative detector responses of the internal standard and the photoproducts studied was used to calculate the quantum yield. The accuracy in this measurement is approximately estimated to be ± 15%.  6.4. Quenching Studies. To calculate the product ratios formed through singlet excited states, quenching studies of diketone 9 were carried out in hexane using 2,3-dimethy1-1,3-butadiene as quencher. A 25 mL stock solution of 10-1M diketone 9 and 10-1M, 1M and 2M 2,3-dimethy1-1,3-butadiene were prepared in hexane using volumetric flasks. These stock solutions were used to make the following 25 rnL solutions: 1) 10-2M diketone in hexane. 2) 10-2M diketone + 10-2M^quencher. 3) 10-2M diketone + 10-1M^quencher. 4) 10-2M diketone + 5x10-1M quencher. 5) 10-2M diketone + 9x10-1M quencher. 6) 10-2M diketone + 12x10-1M quencher. 7) 10-2M diketone + 15x10-1M quencher. 8) 10-2M diketone + 18x10-1M quencher. Analytical photolysis was carried out for each solution as described in the general experimental section and the product ratios at 0% conversion were calculated. Finally, the product ratios (at 0% conversion) were plotted versus quencher concentration to obtain the singlet product ratios from the plateau of the graph.  227  6.5. Computational Generation of Diketone Geometries. Macromodel V3. 1X, an interactive molecular modeling software system on a Silicon Graphics 4D (three dimensional operation) work station, was used to generate the possible low energy conformations of diketones. Steric energies were calculated using Allinger's molecular mechanics (MM2) programme.225 To generate the low energy conformations of cyclic compounds, a crude molecular geometry supplied by the chemist is subjected to Allinger's energy calculation program (MM2), where the atoms are repeatedly moved towards positions of lower energy, ultimately to produce a stable conformation. However, the crude starting coordinates will determine to some extent the resultant stable conformation. To find, therefore, all the low energy conformations, including the global minimum, a number of initial geometries must be supplied, and their energies must be minimized. With medium and large rings, since they are very flexible, it is quite difficult to produce all starting geometries manually. To solve this problem, W.C. Still has written a program called "Monte Carlo Multiple-Minimum",226 (Figure 86) a  random variational conformational search. This program operates in the following manner. Initially a crude molecular geometry is drawn on the terminal screen, and this conformation is then minimized in terms of energy to a nearby minimum energy conformer. For this optimized conformation, a bond (chosen by the chemist) which is located far from the presence of any functional group is broken. At this point the computer regards the molecule as an acyclic chain. The program then attempts to join the two ends of the chain by a random variation of selected torsional angles based on the specified transannular contacts, the closure bond distance, the closure bond angles and the torsional angles, which in turn may be varied according to the precision required. Constraint test 1 is used to eliminate energy-minimized structures, both those whose energies lie outside the selected energetic upper limit relative to the instant global minimum (usually < 25 kJ mo1-1 above the instant global minimum), and those whose  228  inter-atomic distances or torsion angles do not match the explicit distance or torsion constraints provided by the user.  Begin with Random Initial Structure  Recover previous starting geometry  Done. Order structures by energy and output to file  Figure 86: The flow diagram showing the "Monte Carlo Multiple-Minimum" conformational search.  229  Constraint test 2 eliminates starting geometries that have poor ring-closure characteristics, high-energy non-bonded contacts and or energies lying outside the selected upper limit (usually 100 kJ mol-1 above the instant global minimum). Conformations that pass constraint test 2 are subjected to Allinger's MM2 program. A minimized conformation which passes constraint test 1 would be compared with the minima found during previous conformational search steps and would be either rejected as a duplicate conformation or stored as new. One such cycle is known as the Monte Carlo step (MC).  230  REFERENCES: 1. Canuto, V. M.; Levine, J. S.; Augustson, T. R.; Imhoff, C. L.; Giampapa, M. S. Nature (London). 1983, 305 , 281.  2. Cloud, P. Am. J. Sci. 1972, 272, 537. 3. Margulis, L.; Walker, J. C. G.; Rambler, M. Nature (London). 1976, 264, 620. 4. Blake, A. J.; Carver, J. H. J. Atoms. Sci. 1977, 34, 720. 5. Levine, J. S.; Hays, P. B.; Walker, J. C. G. Icarus. 1979, 39, 295. 6. Gavezzoti, A.; Simmoneta, M. Chem. Rev. 1982, 82, 1. 7. Wohler, F. Pogg. Ann. 1828, 12, 253. 8. Trommsdorff, H. Ann. Chem. Phar. 1834, 11, 190. 9. Markwald, W. Z. Phys. Chem. 1899, 30, 140. 10. (a) Ciamician, G.; Silber, P. Chem. Ber. 1901, 34, 2040. (b) Ciamician, G.; Silber, P. Ber. Dtsch. Chem. Ges. 1907, 34, 2040. 11. Stobbe, H.; Steinberger, F.K. Chem. Ber. 1922, 55, 2225. 12. De Jong, A. W. K. Chem. Ber. 1923, 56, 818. 13. Senier, A.; Shepheard, F. G. J. Chem. Soc. 1909, 95, 1943. 14. Scheffer, J. R. Acc. Chem. Res. 1980, 13, 283. 15. Matsuura, T.; Sata, Y.; Ogura, K. Tetrahedron Lett. 1968, 4627. 16. Ariel, S.; Askari, S.; Scheffer, J. R.; Trotter, J. J. Org . Chem. 1989, 54, 4324. 17. Kohlshuter, Z. All g. Chem. 1918, 105, 121. 18. Schmidt, G. M. G. Reactivity of the Photoexcited Organic Molecules; Interscience: New York, 1967; p 227. 19. Cohen, M. D.; Schmidt, G. M. J. J. Chem. Soc. 1967, 239. 20. Liebermann, C. Chem. Ber. 1889, 22, 124 and 782. 21. Bernstein, H. I.; Quimby, W. C. J. Am. Chem. Soc. 1943, 65, 1845. 22. Cohen, M. D.; Schmidt, G. M. J. J. Chem. Soc. 1964, 1996. 23. Schmidt, G. M. G. Pure. App!. Chem. 1971, 27, 647.  231  24. Cohen, M. D.; Green, B, S. Chem. Ber. 1973, 9, 490 and 517. 25. Cohen, M. D.; Schmidt, G. M. J.; Sonntag, F. I. J. Chem. Soc. 1964, 2000. 26. Schmidt, G. M. J. Solid State Photochemistry; Ginsburg, D., Ed.;Verlag Chemie:New York, 1976. 27. Sadeh, T.; Schmidt, G. M. J. J. Am. Chem. Soc. 1962, 84, 3970. 28. Lahav, M.; Schmidt, G. M. J. J. Chem. Soc. B 1967, 312. 29. Green, B. S.; Lahav, M.; Schmidt, G. M. J. J. Chem. Soc. B 1971, 1552. 30. Ramasubbu, N.; Gnanaguru, K.; Venkatesan, K.; Ramamurthy, V. Can. J. Chem. 1982,50, 2337. 31. Bhadbhade, M. M.; Murthy, G. S.; Venkatesan, K.; Ramamurthy, V. Chem. Phys. Lett. 1984, 109, 259. 32. (a) Gnanaguru, K.; Murthy, G. S.; Venkatesan, K.; Ramamurthy, V. Chem. Phys Lett. 1984, 109, 255. (b) Ramasubbu, Gnanaguru, K.; Venkatesan, K.; Ramamurthy, V. J. Org . Chem. 1985, 50, 2337. 33. Gnanaguru, K.; Ramasubbu, N.; Venkatesan, K.; Ramamurthy, V. J. Photochem. 1984, 27, 355. 34. Ramasubbu, N.; Guru Row, T. N.; Venkatesan, K.; Ramamurthy, V.; Rao, C. N. R. J. Chem. Soc., Chem. Commun. 1982, 178. 35. (a) Thomas, J. M. Nature (London) 1981, 289, 633. (b) Nakanishi, H.; Jones, W.; Thomas, J. M.; Hursthouse, M. B.; Motevalli, M. J. Phys. Chem. 1981, 85, 3636. (c) Jones, W.; Ramdas, S.; Theocharis, C. R.; Thomas, J. M.; Thomas, N.W . J. Phys. Chem. 1981, 85, 2594. 36. (a) Nakanishi, H.; Jones, W.; Thomas, J. M. Chem. Phys. Lett. 1980, 71, 44. (b) Nakanishi, H.; Jones, W.; Thomas, J. M.; Hursthouse, M. B.; Motevalli, M. J. Chem. Soc., Chem. Commun. 1980, 611. (c) Nakanishi, H.; Jones, W.; Theocharis, C. R.; Thomas, J. M. J .Chern. Soc., Chem. Commun. 1980, 610. 37. Ramamurthy, V.; Venkatesan, K. Chem. Rev. 1987, 87, 433-481.  232  38. Craig, D. P.; Sarti-Fantoni, P. J. Chem. Soc., Chen?. Commun. 1966, 762. 39. Bart, J. C.; Schmidt, G. M. J. Isr. J. Chem. 1971, 9, 429. 40. Heller, E.; Schmidt, G. M. J. Isr. J. Chem. 1971, 9, 449. 41. Thomas, J. M.; Williams, J. 0. J. Chem. Soc., Chem. Commun. 1967, 432. 42. Murthy, G. C.; Arjunan, P.; Venkatesan, K.; Ramamurthy, V. Tetrahedron. 1987,43, 1225.  43. Cohen, M. D. Angew.Chem., Mt. Ed. Engl. 1975, 14, 386. 44. (a) McBride, J. M. Acc. Chem. Res. 1983, 16, 304-312. (b) McBride, J. M.; Segmuller, B. E.; Hollingsworth, M. D.; Mills, D. E.; Weber, B. A. Science. 1986, 234,830-835. (c) McBride, J.M.; Walter, D. W. J. Am. Chem. Soc. 1981, 103, 7069. (d) McBride,J. M.; Walter, D. W. J. Am. Chem. Soc. 1981, 103, 7074. (e) Karch, N. J.; Koh, E. T.; Whitsel, B. L.; Mc Bride, J. M. J. Am. Chem. Soc. 1975, 97, 6729.  (f) Hollingsworth, M. D.; Mc Bride, J. M. J. Am. Chem. Soc. 1985, 107, 1792. 45. (a) Gavezzotti, A.; Simonetta. In Organic Solid State Chemistry; Desiraju, G. R., Ed.;Elsevier: Amsterdam, 1987; pp 391. (b) Destro, R.; Gavezzotti, A. In Structure and Properties of Molecular Crystals; Pienot, M., Ed.; Elsevier: Amsterdam, 1990; pp 161. (c) Gavezzotti, A. J. Am. Chem. Soc. 1983. 105, 5220. (d) Gavezzotti, A. J. Am. Chem. Soc. 1985, 107, 962. (e) Gavezzotti, A. Nouv. . J. Chim. 1982, 6, 443. (f) Gavezzotti, A. Tetrahedron. 1987, 43, 1241.  (g) Gavezzotti, A. Acta. Dystallogr. 1987, B43, 559. 46. Ariel, S.; Askari, S.; Evans, S. V.; Hwang, C.; Jay, J.; Scheffer, J. R.; Trotter, J.; Walsh, L.; Wong, Y. Tetrahedron. 1987, 43, 1253-1272. 47. (a) Kurihara, T.; Uchida, A.; Ohashi, Y.; Sasada, Y.; Ohgo, Y. J. Am. Chem. Soc. 1984, 106, 5718. (b) Ohashi, Y.; Yanagi, T.; Kurihara, Y.; Sasada, Y.; Ohgo, Y. J. Am. Chem. Soc. 1982, 104, 6353.  48. Ohashi, Y.; Uchida, A.; Sasada, Y.; Ohgo, Y. Acta. Crystallogr. 1988, B44, 538.  233  49. Dunitz, J. D. X-Ray Analysis and the Structure of Organic Molecules; Cornell University Press: Ithaca, New York, 1979; pp 312-318. 50. Ergler, C.; Dorant, K. Chem. Ber. 1895, 28, 2497. 51. (a) Rabinovich, D.; Schmidt, G. M. J. J. Chem. Soc. B 1970, 6-10. (b) Jungk,A. E.; Schmidt, G. M. J. J. Chem. Soc. B 1970, 1427-1434. (c) Jungk, A. E.; Luwisch, M.; Pinchas, S.; Schmidt, G. M. J. Isr. J. Chem 1977, 16, 308-310. (d) Curtin, D. Y.; Engelman, J. H. J. Org . Chem. 1972, 37 , 3439-3443.  52. Cohen, M. D. J. Chem. Soc. B 1968, 373-376. 53. Matsura, T.; Sata, Y.; Ogura, K. Tetrahedron. Lett. 1968, 4627-4630. 54. (a) Appel, W. K.; Greenhough, T. J.; Scheffer, J. R.; Trotter, J.; Walsh, L. J. Am. Chem. Soc. 1980, 102, 1158-1160. (b) Appel, W. K.; Greenhough, T. J.;  Scheffer, J.R.; Trotter, J. J. Am. Chem. Soc. 1980, 102, 1160-1161. (c) Jiang, Z. Q.; Scheffer, J. R.; Secco, A. S.; Trotter, J. Tetrahedron. Lett. 1981, 891-894. (d) Appel, W. K.; Jiang, Z. Q.; Scheffer, J. R.; Walsh, L. J. Am. Chem. Soc. 1983, 105, 5354- 5363. (e) Scheffer, J. R.; Trotter, J.;  Appel, W. K.; Greenhough, T. J.; Jiang, Z. Q.; Secco, A. S.; Walsh, L. Mol. Cryst. Liq. Cryst. 1983, 93 ,1-15. (f) Ariel, S.; Askari, S.; Scheffer, J. R.;  Trotter, J. J. Org . Chem. 1989, 54, 4324. (g) Ariel, S.; Askari, S.; Scheffer, J. R.; Trotter, J.; Walsh, L. J. Am. Chem. Soc. 1984, 106, 5726-5728. 55. (a) Scheffer, J. R.; Bhandari, K. S.; Gayler, R. E.; Wostradowski, R. A. J. Am. Chem. Soc. 1975, 97, 2178-2189. (b) Scheffer, J. R.; Jennings, B. M.;  Louwrens, J. P. J. Am. Chem. Soc. 1976, 98, 7040 (c) Scheffer, J. R.; Dzakapasu, A. A. J. Am. Chem. Soc. 1978, 100, 2163-2173. (d) Ariel, S.; Evans, S.; Hwang, C.; Jay, J.; Scheffer, J. R.; Trotter, J.; Wong, Y. F. Tetrahedron. Lett. 1985, 965-968. (e) Ariel, S.; Askari, S.; Scheffer, J. R.;  Trotter, J. Tetrahedron. Lett. 1986, 27, 783-786. (f) Ariel, S.; Askari, S.;  234  Scheffer, J. R.; Trotter, J.; Wireko, F. J. Am. Chem. Soc. 1987, 109, 4623. (g) Ariel, S.; Askari, S.; Scheffer, J. R.; Trotter, J.; Wireko, F. Acta. Crystallogr. 1987, B43, 532. 56. Appel, W. K.; Greenhough, T. J.; Scheffer, J. R.; Trotter, J. J. Am. Chem. Soc. 1979, 101, 213-215. 57. (a) Ariel, S.; Evans, S.; Garcia-Garibay, M.; Harkness, B. R.; Omkaram, N.; Scheffer, J. R.; Trotter, J. J. Am. Chem. Soc. 1988, 110, 5591. (b) Scheffer, J. R.; Trotter, J. Chem. Rev. Mt. 1988, 9, 271. (c) Ariel, S.; Garcia-Gaiibay, M.; Scheffer, J. R.; Trotter, J. Acta. Crystallogr. 1989, B45, 153. (d) Evans, S.; Omkaram, N.; Scheffer, J. R.; Trotter, J. Tetrahedron. Lett. 1986, 27, 1419. (e) Evans, S. V; Garcia-Garibay, M.; Omkaram, N.;  Scheffer, J. R.; Trotter, J.; Wireko, F. J. Am. Chem. Soc. 1986, 108, 5648. (f) Ariel, S.; Evans, S. V.; Omkaram, N.; Scheffer, J.R.; Trotter, J. Mol. Cryst. Liq. Cryst. 1986, 134, 169.  58. (a) Garcia-Gaiibay, M.; Scheffer, J. R.; Trotter, J.; Wireko, F. J. Am. Chem. Soc. 1989, 111, 4985. (b) Garcia-Garibay, M.; Scheffer, J. R.; Trotter, J.; Wireko, F. Tetrahedron. Lett. 1987, 28, 4789. (c) Chen, J.; Garcia-Garibay, M.;  Scheffer, J. R.; Trotter, J.; Wireko, F. Tetrahedron Lett. 1989, 30, 6125. (d) Garcia-Garibay, M.; Scheffer, J. R.; Trotter, J.; Wireko, F. Tetrahedron. Lett. 1988, 29, 2042. (e) Garcia-Garibay, M.; Scheffer, J. R.; Watson, D. G. J. Chem. Soc., Chem. Commun. 1989,600. (f) Pokkuluri, P. R.; Scheffer, J. R.;  Trotter, J.; Wireko, F. Tetrahedron. Lett. 1989, 30, 1601. (g) Garcia-Garibay, M.; Scheffer, J. R.; Trotter, J.; Wireko, F. Tetrahedron. Lett. 1987 28, 1741. (h) Garcia-Garibay, M.; Scheffer, J.R.; Trotter, J.; Wireko, F. J. Am. Chem. Soc. 1989, 111, 4985. (i) Poldailuri, P. R.; Scheffer, J.R.; Trotter, J. J. Am. Chem. Soc. 1990, 112, 3676. 59. Wagner, P. J. Acc. Chem. Res. 1971, 4, 168.  235  60. Wagner, P. J.; Park, B. -S. In Organic Photochemistry; Padwa, A., Ed.; Marcel Dekker New York, 1991; Vol. 11, Chapter 4. 61. Yang, N. C.; Yang, D. H. J. Am. Chem. Soc. 1958, 80, 2913. 62. (a) Norrish, R. G. W. Trans. Faraday Soc. 1937, 33, 1521. (b) Saltmarsh, 0. D.; Norrish, R. G. W. J. Chem. Soc. 1935, 455. 63. a) Henne, A.; Fischer, H. Angew.Chem., Mt. Ed. Engl. 1976, 15, 435. b) Wagner, P. J. J. Am. Chem. Soc. 1967, 89, 5898. 64. Wagner, P. J.; Kelso, P. A.; Kemppainen, Zepp. R. G. J. Am. Chem. Soc. 1972, 94, 7500-7506. 65. Dorigo, A. E.; Houk, K. N. J. Org . Chem. 1988, 53, 1650. 66. Dorigo, A. E.; McCarrick, Loncharich, R. J.; Houk, K. N. J. Am. Chem. Soc. 1990, 112, 7508. 67. (a) Pople, J. A.; Gordon, J. J. Am. Chem. Soc. 1967, 89, 4253-4261. (b) Winnik, M.A. Acc. Chem. Res. 1977, 10, 173-179. (c) Dewar, M. J. S.; Doubleday, C. J. Am. Chem. Soc .1978, 100, 4935-4941. (d) Morrison, H.; Miller, A.; Pandey, B.; Severance, D.; Strommen, R.; Bigot, B. Pure App!. Chem. 1982, 54, 1723-1732. 68. (a) Djerassi, R. Pure App!. Chem. 1964, 9, 159-179. (b) Djerassi, C.; Williams, D. H.; von Muntzenbecher, G.; Budzikiewicz, H. J. Am. Chem. Soc. 1965, 87, 817-826. 69. Hewster, K.; Kalvoda, J. Angew. Chem., Int. Ed. Eng. 1964, 3, 525. 70. Scheffer, J. R.; Garcia-Gariby, M.; Omkaram, N. Org . Photochem. 1987, 8, 249. 71. Bondi, A. J. Phys. Chem. 1964, 68, 441-451. 72. Turro, N. J.; Weiss, D. S. J. Am. Chem. Soc. 1968, 90, 2185. 73. Edward, J. T. J. Chem. Educ. 1970, 47, 261-270.  236  74. Sugiyama, N.; Nishio, T.; Yamada, K.; Aoyama. Bull.Chern. Soc. Jpn. 1970, 43,1879.  75. Chang, H. C.; Popovitz-Biro, R.; Lahav, M.; Leiserowitz, L. J. Am. Chem. Soc. 1943, 65, 1845-1846.  76. Wagner, P. J. Top. Curr. Chem. 1976, 66, 1. 77. Colpa, J. P.; Stehlik, D. J. Chem. Phys. 1983, 81, 163 78. TUITO, N. J. Modern Molecular Photochemistry; Benjamin Cummings: Menlo Park, California, 1972; Chapter 10. 79. Scheffer, J. R. In Organic Solid State Chemistry; Desiraju, G. R., Ed.; Elsevier: Amsterdam, 1987; Chapter 1. 80. Ciamician, G.; Silber, P. Ber. Dtsch. Chem. Ges. 1907, 40, 2415. 81. Bamford, C. H.; Norrish, R. G. W. J. Chem. Soc. 1935, 1504. 82. a) Saltmarsh, 0. D.; Norrish, R. G. W. J. Chem. Soc. 1935, 455. b) Dalton, J. C., J. Am. Chem. Soc. 1943, 65, 1845-1846.  83. (a) Dalton, J. C.; Turro, N. J. Annu. Rev. Phys. Chem. 1970, 21, 499. (b) Turro, N. J.; Dalton, J. C.; Dawes, K.; Farrington, G.; Hautala, R.; Morton, D.; Niemczyk, M.; Shore, N. Acc. Chem. Res. 1972, 5, 92. 84. Furth, B.; Jost, P.; Sabbah, S.; Kossanyi, J. Nouveau J. Chim. 1979, 3 , 199. 85. (a) Mitscherlich, E. Ann. Chim. Phys. 1822, 19, 350. (b) Mitscherlich, E. Ann. Chim.Phys. 1823, 24, 264.  86. (a) Bernstein, J. In Organic Solid State Chemistry; Desiraju, G. R., Ed.; Elsevier: Amsterdam, 1987; pp 471. (b) Corradini, P. Chim. Ind. 1973, 55, 122.  87. Dunitz, J. D. Pure Appl. Chem. 1991, 63, 177-185. 88) McCrone, W. C. Physics and Chemistry of the Organic Solid State; Fox, D., Ed.; Wiley-Interscience: New York, 1965; Vol 2,Chapter 8. 89) Mnyukh, Y. V. Mol. Cryst. Lig. Cryst. 1979, 52, 163-218.  237  90) Hassel, 0. Tidsskr. Kjemi, Bergves. Metallurgi. 1943, 3, 32. 91) Ramsey, O.B. Stereochemistry; Heyden: London, 1981. 92) Barton, D. H. R.; Cookson, R. C. Quart. Rev. 1956, 10, 44. 93) Eliel, E. L. Stereochemisby of Carbon Compounds; McGraw-Hill: New York, 1962. 94) (a) Ruzicka, L.; Stoll, M.; Huyser, H. W.; Boekenooger, H. A. Helv.Chim Acta. 1946, 29, 1611. (b) Kobelt, M.; Barman, P.; Prelog, V.; Ruzicka, L.  Helv.Chim.Acta. 1949, 32, 356.  95) Prelog, V. J. Chem. Soc. 1950,420. 96) Huber, E.; Dunitz, J. D.; Venkatesan, K. Proc. Chem. Soc. 1961, 463. 97. Mohr, E. J. Prakt. Chem. 1918, 98, 315. 98. (a) Dale, J. J. Chem. Soc. 1963, 93. (b) Dale, J. Angew.Chem., Mt. Ed. Engl. 1966, 5, 1000.  99. (a) Dale, J. J. Chem. Soc., Chem. Comm. 1970, 1340. (b) Growth, P. Acta. Chem. Scand., Ser.A 1979, 33, 203.  100. Dunitz, J. D.; Shearer, H. Helv.Chim.Acta. 1960, 43, 18. 101. (a) Groth, P. Acta. Chem. Scand., Ser.A 1979, 33, 503. (b) Dehli, J.; Groth, P. Acta. Chem. Scand., Ser.A 1969, 23, 587. 102. (a) Dale, J. Pure. Appl.Chem. 1971, 25, 469. (b) Growth, P. Acta. Chem. Scand., Ser.A 1974, 28, 808.  103. Alvik, T.; Borgen, G.; Dale, J. Acta. Chem. Scand., Ser.A 1972, 26, 1805. 104. Dale, J. Acta. Chem. Scand., Ser.A 1973, 27, 1130. 105. Anet, F. A. L.; Cheng, A. K.; Krane, J. J. Am. Chem. Soc. 1973, 14, 7877. 106. (a) Growth, P. Acta Chem. Scand., Ser A 1979, 33, 199. (b) Kay, H. F.; Newman, B. A. Acta Crystallogr. 1968, B24, 615. 107. Allinger, N. L.; Gorden, B. J.; Newton, M. G.; Lauritsen-Norskov, L.; Profeta, S., Jr. Tetrahedron 1982, 38, 2905.  238  108. Huckel, W.; Danneel, R.; Schwartz, A.; Gercke, A. Ann. 1929, 474, 121. 109. Pappas, J. J,; Keaveney, W. P.; Gancher, E.; Berger, M. Tetrahedron Lett. 1966, 36, 4273-4278.  110. Alvik, T.; Dale, J. Acta Chem. Scand. 1971, 25, 1153. 111. Blomquist, A. T.; Spencer, R. D. J. Am. Chem. Soc. 1948, 70, 30. 112. Leonard, N. J.; Schimelpfenig, C. W. J. Org. Chem. 1958, 23 , 1708. 113. Ruzicka, L.; Brugger, N.; Seidel, C. F.; Schinz, H. Hely. Chirn. Acta. 1928, 11, 496.  114. Ruzicka, L.; Stoll, M.; Huyser, H. W.; Boekenoogen, H. A. Hely. Chim. Acta 1930, 13, 1152.  115. Wagner, P. J. In Rearrangements in Ground and Excited States; de Mayo, P., Ed.; Academic: New York, 1980; Vol. 3, chapter 20. 116. Wagner, P. J. Acc. Chem. Res. 1983, 16, 461. 117. Lewis, E. S. Isotopes in Organic Chemistry; Buncel, E.; Lee. C. C., Eds.; Elsevier: Amsterdam, 1976; Vol. 2, p 134. 118. Allinger, N. L.; Gorden, B. J.; Profeta, S., Jr. Tetrahedron Lett. 1980, 36, 859. 119. Luknitskii, F. I.; Vovsi, B. A. Russ. Chem. Rev. 1969, 38(6), 493. 120. Elam, E. U. J. Org . Chem. 1972, 60, 27. 121. Sauer, J. C. J. Am. Chem. Soc. 1947, 69, 2444. 122. Horner, L.; Spietschka, E.; Gross, A. Ann. Chem. 1951, 17 573. 123. March, J. Advance Organic Chemistry, 4th ed.; Wiley-Interscience: New York, 1992; Chapter 6. 124. Belsky. Tetrahedron. 1976, 28, 771. 125. Monson, R. S. J. Heterocyclic Chem. 1976, 13, 893. 126. Cordonnier, G.; Sliwa, H. J. Heterocydic.Chem. 1978, 24, 111. 127. Piers, E.; Wall, W.; Britton, R. W. J. Am. Chem. Soc . 1971, 25 , 5114. 128. House, H. 0.; Kramer, V. J. Org . Chem. 1962, 28, 3362.  239  129. Mori, T.; Matsui, K.; Nozaki, H. Tetrahedron Lett. 1970, 14, 1175. 130. Schulte-Elte, K. H.; Willhalm, B.; Thomas, A. F.; Stoll, M.; Ohloff, G. Hely .Chim. Acta 1971, 54, 1759.  131. Burchill, P. J.; Kelso, G.; Power, A. J. Aust. J. Chem. 1976. 29. 2477. 132. Encina, M. V.; Lissi, E. A. J. Photochem. 1978, 8, 131. 133. Mirbach, M. F.; Mirbach, M. J, Liu, K. C.; TUITO, N. J. J. Photochem. 1978, 8, 299.  134. Matsui, K.; Mori, T.; Nozaki, H. Bull. Chem. Soc. Jpn. 1971, 44, 3440. 135. Barnard, M.; Yang, N. C. Proc. Chem. Soc. 1958, 302. 136. Sauers, R. R.; Huang, S. Tetrahedron Lett. 1990, 31, 5709. 137. Dewar, M. J. S.; Doubleday, C. J. Am. Chem. Soc. 1978, 100, 4395. 138. Wagner, P. J.; Kelso, P. A.; Kemppainen, app. R. G. J. Am. Chem. Soc. 1972, 94, 7506.  139. Lewis, F. D.; Johnson, R. W.; Kory, D. R. J. Am. Chem. Soc. 1974, 96, 6100. 140. Patt, S. L.; Shoolery, J. N. J. Magn. Reson. 1982, 46,535 . 141. Silverstein, R. M.; Bassler,G. C.; Morrill, T. C. Spectrometric Identifiation of Organic Compounds, 4th ed.; John Wiley & Sons, Inc.: New York, 1981;  p259. 142. Herzschuh, R.; Epsch, K. J. Magn. Reson. 1980, 322, 37. 143. Dunitz, J. D. and Iberes. J. A. In Perspectives in Structural Chemistry; John Wiley & Sons: New York, 1968; Vol. 2. 144. Ariel, S.; Ramamurthy, V.; Scheffer, J. R.; Trotter, J. J. Am. Chem. Soc. 1983, 105, 6959.  145. Wagner, P. J. Pure. App!. Chem. i977, 49, 259. 146. Scaiano, J. C.; Lissi, E. A.; Encinas, M. V. Rev. Chem. Intermed. 1978, 2, 139-195.  147. Scaiano, J. C. Acc. Chem. Res. 1982, 15, 252.  240  148. Scaiano, J. C. Tetrahedron. 1982, 38, 819. 149. Wagner, P. J.; Hammond, G. S. J. Am. Chem. Soc. 1965, 87, 4009. 150. Dougherty, T. J. J. Am. Chem. Soc. 1965, 87, 4011. 151. Salem,L. J. Am. Chem. Soc. 1974, 96, 3486. 152. Michl, J. Topics. Curr. Chem. 1974, 46, 1. 153. Stephenson, L. M.; Cavigli, P. R.; Parlett, J. L. J. Am. Chem. Soc. 1971, 93, 1984.  154. Casey, C. P.; Boggs, R. A. J. Am. Chem. Soc. 1972, 94, 6457. 155. Yang, N. C.; Elliot, S. P. J. Am. Chem. Soc. 1969, 91, 7550. 156. Ware, W. R.; Lee, S. K. J. Chem. Phys. 1968, 49, 217. 157. (a) Golden, D. M.; Furuyama, S.; Benson, S. W. Int. J. Chem. Kinet. 1969, I, 57. (b) Pitzer, K. S. Discuss. Faraday Soc. 1951, 10, 66.  (c) Lide, D. R., Jr.; Mann, D. E. J. Chem. Phys. 1958, 29, 914. (d) Lide, D. R., Jr. J. Chem. Phys. 1958, 29, 1426. 158. O'Neal, H. E.; Miller, R. G.; Gunderson, E. J. Am. Chem. Soc. 1974, 96, 3351. 159. Encinas, M. V.; Scaiano, J. C. J. Am. Chem. Soc. 1978, 100, 7101. 160. Yang, N. C.; Coulson, D. R. J. Am. Chem. Soc. 1966, 88, 4511. 161. Kristiansen, P.; Ledaal, T. Tetrahedron Lett. 1971, 2817. 162. Furman. I.; Raymond, J.; Catchings, R. M; Weiss, R.G. J. Am. Chem. Soc. 1992, 114, 6023.  163. Brokman, R. F.; Kearns, D. R. J. Chem .Phys. 1964, 40, 1038. 164. Aoyama, H.; Sakamoto, M.; Kuwabara, K.; Yoshida, K.; Omote, Y. J. Am. Chem. Soc. 1983, 105, 1958.  165. Yang, N. C.; Elliot, S. P.; Kim, B. J. Am. Chem. Soc. 1969, 91, 7551. 166. The conformation was calculated by us using the MM2 force fields. 167. Cohen, M. D.; Schmidt, G. M. J.; Flavian, S. J. Chem. Soc. 1964, 2041. 168. Ichimura, K.; Watanabe, S. Tetrahedron Lett. 1972, 821.  241  169. Ichimura, K.; Watanabe, S. Bull.Chern. Soc. Jpn. 1976, 49, 2220. 170. (a) Galli, C.; Illuminati, G.; Mandolini, L.; Tamborra, P. J. Am. Chem. Soc. 1977, 99, 2591. (b). Galli, C.; Illuminati, G.; Mandolini, L.; Tamborra, P. Acc.Chem. Res. 1981, 14, 95. (c) Benedetti, F.; Stirling, C. J. M. J. Chem. Soc., Perkin Trans. 2 1986, 605.  171. Ariel, S.; Ramamurthy, V.; Scheffer, J. R.; Trotter, J. J. Am. Chem. Soc. 1983, 105, 6990.  172. Wagner, P. J.; Kemppainen, A. E. J. Am. Chem. Soc. 1968, 90, 5896. 173. Hoffman, R.; Swaminathan, S.; O'Dell, B. G.; Gleiter, R. J. Am. Chem. Soc. 1970, 92, 7091.  174. Nunez, A.; Weiss, R. G. Bol. Soc. Chil. Quim. 1990, 35, 3. 175. He, Z.; Weiss, R. G. J. Am. Chem. Soc. 1990, 112, 5535. 176. Fleming, I.; Kemp-Jones, A. V.; Long, W. E.; Thomas, E. J. J. Chem. Soc., Perkin Trans. 2 1976, 7.  177. Rawdah, T. N. Tetrahedron. 1991, 47, 8579. 178. Rawdah, T. N.; El-Faer, M. Z. Tetrahedron. 1990, 46, 4101. 179. (a) Willy, E. W.; Binsch, G.; Eliel, L. E. J. Am. Chem. Soc. 1970, 92, 5394. (b) Lipnick, J. Mol. Struct. 1974, 21, 423.  180. Allinger, N. A.; Tribble, M. T.; Miller, M. A. Tetrahedron. 1972, 28, 1173. 181. Rawdah, T. N. Tetrahedron. 1989, 45, 7405. 182. Small, R. D., Jr.; Scaiano, J. C. Chem. Phys. Lett. 1978, 59, 246. 183. Boer, F. P.; Shannon, T. W.; McLafferty, F. W. J. Am. Chem. Soc. 1968, 90, 7239.  184. (a) Byrn, S. R. The Solid State Chemist!), of Drugs; Acadamic Press: New York, 1982. (b) Cohen, M. D.; Schmidt, G. M. J. J. Org . Chem. 1964, 1996. (c) Berkovitch-Yellin, Z.; Lahav,M.; Leiserovitz, L. J. Am. Chem. Soc. 1974, 96, 918. (d) Cohen, M. D.; Elgavi, A.; Green, B. S.; Ludmer, Z.;  242  Schmidt, G. M. J. J. Am. Chem. Soc. 1972, 94, 6776. (d) Cohen, R.; Ludmer, Z.; Yakhot, V. Chem. Phys. Lett. 1975, 34, 5679. (e) Warshell, A.; Shakked, Z. J. Am. Chem. Soc. 1975, 97, 5679.  185. Dunitz, J. D. In Solid State Photochemistry; Ginsburgh, D., Ed.; Verlag Chemie: New York, 1976; p.255. 186. Lewis, T. J.; Rettig, S. J.; Scheffer, J. R.; Trotter, J. J. Am. Chem. Soc. 1991, 113, 8180.  187. (a) Casal, H. L.; de Mayo, P.; Miranda, J. F.; Sciano, J. C. J. Am. Chem. Soc. 1983, 105, 5155.  188. (b) Turro, N. J.; Wan, P. Tetrahedron Lett. 1984, 25, 3655. (c) Hrovat,D. A.; Liu, J. H.; Tuiro, N. J.; Weiss, R. G. J. Am. Chem. Soc. 1984, 106, 7033. (d) Zimmermann, R. G.; Liu, J. H.; Weiss, R. G. J. Am. Chem. Soc. 1986, 108, 5264. (e) Treanor, R. L.; Weiss, R. G. Tetrahedron. 1987, 43, 1371.  (0 Treanor, R. L.; Weiss, R. G.; Nunez. A. Pure. Appl. Chem. 1988, 60, 999. 189. (a) Furman. I.; Weiss, R.G. J. Am. Chem. Soc. 1992, 114, 1381. (b) He, Z.; Weiss, R. G. J. Am. Chem. Soc. 1990, 112, 5535. 191. Nakanishi, H.; Ueno, K. J. Polym. Sci. 1987, 16, 767. 192. (a) All solid state NMR investigations were performed by Prof. Colin Fyfe et al and the spectral interpretations were provided by them. b) Fyfe, C. A. "Solid State NMR for chemists"., C.F.C. Press, Onttario, Canada, 1983. 193. Davis, J. H. Biochem. Biophys. Acta 1983, 117, 737. 194. (a) Butenandt, A.; Wolff, A. Chem. Ber. 1939, 72, 1121. (b) Wehrli, A, Schaffner, K. Hely. Chem. Acta. 1962, 45, 385. 195. (a) Wagner, P. J.; Spoerke, R. W. J. Am. Chem. Soc. 1969, 91, 4437. (b) Dunion, P.; Trumbore, C. N. J. Am. Chem. Soc. 1965, 87, 4211. (c) Srinivasan, R.; Cremer, S. E. J. Am. Chem. Soc. 1965, 87, 1647.  243  196. Dalton, J. C.; Turro, N. J. Ann. Rev. Phys. Chem. 1970, 21, 499. 197. Chapman, 0. L.; Weiss, D. S. Org. Photochem. 1973, 3 , 197. 198. (a) Weiss, D. S.; Haslanger, M.; Lawton, R. G. J. Am. Chem. Soc. 1976, 98, 1050. (b) Agosta, W. C.; Wolff, S. J. Am. Chem. Soc. 1976, 98, 4182. (c) Yang, N. C.; Chen, R. H. -K. J. Am. Chem. Soc. 1971, 93, 530. 199. (a) Coyle, J. D. J. Chem. Soc. B 1971, 1736. (b) Hammond, W. B.; Yeung, T. S. Tetrahedron Lett. 1975, 1169.  200. Turro, N. J.; Dalton, J. C.; Dawes, K.; Farrington, G.; Hautala, R.; Morton, D.; Niemczyk, M.; Shore, N. Acc. Chem. Res. 1972, 5, 92. 201. Weiss, D. S.; Kochanek, P. M. Tetrahedron Lett. 1977, 9, 763. 202. Han, N.; Hwang, K. C.; Lei, X.; Turro, N. J. J. Photochem. Photobiol.,B 1991, 61, 35.  203. Lei, X.; Zimmt, M.; Doubleday, C., Jr.; Turro, N. J. J. Am. Che. Soc. 1986, 108, 2444.  204. Weiss, D. S. In Organic Photochemistry; Padwa, A., Ed.; Marcel Dekker: New York, 1981; Vol 5, Chapter 4. 205. The dimethykyclododecanones, (R*, 5*)-2,12-cyclododecanone and (R*, S*)-2,12-cyclododecanone were obtained from Prof. L. Weiler's laboratory. 206. (a) Ramamurthy, V. Tetrahedron 1986, 42, 5753. (b) Ramamurthy, V.; Scheffer, J. R.; TU1TO, N. J. Tetrahedron 1987, 43, 1197. (c) Kalyanasundaram, K. Photochemistry in Microheterogeneous Systems; Academic: New York, 1987. Weiss, R. G. Tetrahedron 1988, 44, 3413. (d) Scheffer, J. R.; Trotter, J. Rev. Chem Jnt. 1988, 9,271. (e) Fox, M. A. Top. Curr. Chem. 1987, 142, 71. (f) Anpo, M.; Matsura,T. Photochemistry on Solid Surfaces; Elsevier: Amsterdam,1989.  244  207. Ramamurthy, V. In Photochemistry in Organized and Constrained Media; Turro, N. J., Garcia-Garibay, M., Eds.; VCH: New York, 1991;  Chapter 1. Ramamurthy. V. Ibid.; Chapter 10. 208. (a) TUff0, N. J.; Zhang, Z. Tetrahedron Lett. 1989, 30, 3761. (b) Gessner, F.; Olea, A..; Lobaugh, J. H.; Johnston, L.; Scaiano, J. C. J. Org . Chem. 1989, 54, 259. (c) Liu, X.; Iu, K. K.; Thomas, J. K. J. Phys. Chem. 1989, 93, 4120. (d) Ramamurthy,V.; Caspar, J. V.; Corbin, D. R.; Schyler, B.;  Maki,A. H. J. Phys. Chem. 1990, 94, 3391. 209. Barrer, R. M. In Inclusion Compounds; Atwood, J. L.; Davies, J. E. D.; Mac Nicol,D. D., Eds.; Academic press: London, 1984; pp 191. 210. (a) Meier, W. M.; Olson, D. H. In Atlas of Zeolite Structure Types; Butterworths: Cambridge, 1987. (b) Breck, D. W. Zeolite Molecular Sieves, 2nd ed.; Wiley: New York, 1974. 211. (a) Bahnemann, D. W.; Monig, J.; Chapman. J. Phys. Chem. 1987, 91, 3782. (b) Shiragami, T.; Pac, C.; Yanagida, S. J. Chem. Soc., Chem. Commun. 1989, 831. (c) Meissenr, D.; Memming, R.; Kastening, B. J. Phys. Chem. 1988, 92, 3476.  (d) Ramamurthy, V.; Corbin,D. R.; TUITO, N. J.; Sato, Y. Tetrahedron Lett. 1989, 30, 5829.  212. Ramamurthy, V.; Corbin, D. R.; Eaton, D. F. J. Org . Chem. 1990, 55, 5269. 213. Ramamurthy, V.; Lei, X-G.; TUITO, N. J.; Lewis, T. J.; Scheffer, J. R. Tetrahedron Lett. 1991, 32, 7675.  214. (a) Breck, D. W. Zeolite Molecular Sieves; JohnWiley: New York, 1974. (b) Smith, J. V. Chem. Rev. 1987, 88, 149. 215. (a) O'Malley, P. J. Chem. Phys. Lett. 1990, 166, 340. (b) Piimet, M.; Garbowski, E.; Mathieu, M. V.; Imelik, B. J. Chem. Soc, Faraday Trans. I 1980, 76, 1942. (c) de Menorval, L. C.; Raftery, D.; Liu, S. B.; Takegoshi, K.;  245  Ryoo, R.; Pines, A. J. Phys. Chem. 1990, 94, 27. (d) Czjek, M.; Vogt, T.; Fuess, H. Angew. Chem. Int. Ed. Engl. 1989, 28, 770. (e) Jobic, H.; Renouprez, A.; Fitch, A. N.; Lauter, H. J. J. Chem. Soc., Faraday Trans. I 1987, 83, 3199. (f) de Mallmann, A.;  Barthomeuf, D. J. Phys. Chem. 1989, 93, 5636. (g) Demontis, P.; Yashonath, S.; Klein, M. L. J. Phys. Chem. 1989, 93, 5016. 216. (a) Wagner, P. J. J. Am. Chem. Soc. 1967, 89, 5898. (b) Baltrop, J. A.; Coyle, J. D. Tetrahedron Lett. 1968, 3235. 217. Ramamurthy, V.; Caspar, V. Mol. Cryst. Liq. Cryst. 1992, 211, 211. 218. Lechert, H.; Balser, W. D. J. Phys. Chem. Solids. 1989, 50, 497. 219. Ramamurthy, V.; Corbin, D. R. Johnston, L. J. Am. Chem. Soc. 1992, 114, 3870.  220. Patt, S. L.; Shoolery, J. N. J.Magn. Reson. 1982, 46, 535. 221. Still, W. C.; Kahn, M.; Mitra, A. J. Am. Chem. Soc. 1978, 43 , 2923. 222. Perrin, D. D.; Armarero, W. L. F.; Perrin, D. R. Purification of laboratory chemicals; Pergamon press: Oxford, 1980.  223. Blomquist, A. T.; Prager, J.; Wolinsky, J. J. Am. Chem. Soc. 1955, 77, 1804. 224. Murov, S. L. Handbook of photochemistry; Marcel Dekker: New York, 1973. 225. Bowen, J. P.; Pathiaseril, A.; Profeta, S. Jr.; Allinger, N. L. J. Org . Chem. 1987, 52, 5162.  226. Chang, G.; Guida, W. C.; Still, C. J. Am. Chem. Soc . 1989, 111, 4379. 227. Turro, N. J.; Zhang, Z. In Photochemistry in solid surface, Ampo, M.; Matsura, T. Eds., Elsevier: Amsterdam; 1989, p. 197. 228. Turro, N. J.; Zhang, Z. Tetrahedron Lett. 1987, 28, 5517. 229. Wagner, P. J.; Hammond, G. S. J. Am. Chem. Soc. 1966, 88, 1245. 230. Halpern, A.; Ware, W. R. J. Chem. Phys. 1970, 53 , 1969. 231. Chandler, W. D.; Goodman, L. J. Mol. Spectr. 1970, 35, 232.  

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.831.1-0061803/manifest

Comment

Related Items