Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

An investigation of metal carbonyl complexes including 59Co and 35Cl nuclear Quadrupole resonance studies Chia, Lian Sai 1974

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

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
831-UBC_1974_A1 C49_5.pdf [ 12.77MB ]
Metadata
JSON: 831-1.0060069.json
JSON-LD: 831-1.0060069-ld.json
RDF/XML (Pretty): 831-1.0060069-rdf.xml
RDF/JSON: 831-1.0060069-rdf.json
Turtle: 831-1.0060069-turtle.txt
N-Triples: 831-1.0060069-rdf-ntriples.txt
Original Record: 831-1.0060069-source.json
Full Text
831-1.0060069-fulltext.txt
Citation
831-1.0060069.ris

Full Text

cl AN INVESTIGATION OF METAL CARBONYL COMPLEXES INCLUDING 59 35 Co AND C l NUCLEAR QUADRUPOLE RESONANCE STUDIES BY LIAN SAI CHIA B.Sc, Nanyang University, Singapore, 1965 M.Sc, University of B r i t i s h Columbia, 1971 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY i n the Department of CHEMISTRY We accept this thesis as conforming to the required s t a n d a r d A THE UNIVERSITY OF BRITISH COLUMBIA February, 1974 In presenting th i s thesis in pa r t i a l fu l f i lment of t h e requirements f o r an advanced degree at the Univers i ty of B r i t i s h C o l u m b i a , I a g r e e t h a t the L ibrary shal l make it f ree ly ava i lab le for reference and s t u d y . I further agree that permission for extensive copying of t h i s t h e s i s for scho lar ly purposes may be granted by the Head of my Department o r by his representat ives. It is understood that copying or p u b l i c a t i o n o f th i s thesis for f inanc ia l gain sha l l not be allowed without my written permission. Department of The Univers i ty of B r i t i s h Columbia Vancouver 8, Canada Date - i i -ABSTRACT An improved method for preparing (L-L)Fe 2(CO)g complexes (L-L i s a fluorocarbon-bridged d i t e r t i a r y arsine or phosphine) has been developed. The reactions of these complexes with monodentate L and bidentate L-L ligands give r i s e to eight types of products. These include L m ( L - L ) F e 2 ( C O ) 5 and ( L - L ) m ( L - L ) F e 2 ( C O ) 5 i n which the mono-dentate ligand replaces one carbonyl group from F e A ; ( L m ) 2 ( L - L ) F e 2 ( C O ) 4 c A and (L-L) (L-L)Fe 2(CO) 4 i n which two carbonyl groups on Fe have been replaced; ( L m ) 2 ( L - L ) F e 2 ( C O ) ^ , (L-L) b(L-L)Fe 2(CO) and (dppp)k(L-L)Fe 2(CO)^ i n which one carbonyl group on each i r o n has been replaced; and (L m) 3(L-L)Fe 2(CO) i n which the ligands replace two A B carbonyl groups from Fe and one from Fe . The spectroscopic properties and some reactions of these d e r i v a t i v e s are described. The reactions of Co 2(CO)g and (ir-alkyne)Co 2(CO)g with some monodentate and bidentate ligands and that of [LCo(CO) 3] 2 with bidentate ligands have been investigated. Several new compounds have been i s o l a t e d ; these include bridged derivatives ( L - L ) 2 C o 2 ( C O ) 2 ( C 0 ) 2 and L(L-L) CCo 2(CO) 3(CO) 2, non-bridged d e r i v a t i v e s (L-L)Co 2(CO)^ and [(L-L)Co(CO) 3] 2, i o n i c products [ L 2 ( L - L ) C o ( C O ) ] 2 + ( X ~ ) 2 and 2+ -L.(L-L)Co„(CO). (X )„, and s u b s t i t u t i o n products L (RC = CR)Co„(CO), (n = 1-4), (L-L)°(RC = CR)Co 2(CO) 4 > (dppp)°(RC = CR)Co 2(CO) 4 > (L-L) 2(RC = CR)Co 2(CO) 2 and (L-L) 2(RC = CR)Co 2(CO) 2 < The proposed molecular structures are based on chemical and spectroscopic information. - i i i -59 In p a r a l l e l with the preparative work, Co nqr studies on cobalt carbonyl complexes have been c a r r i e d out. The e f f e c t s of Y on the nqr spectra of the [ Y 3 P C o ( C O ) ( Y = o p h » 0 M e » 0 E t > B u t l ) compounds are found to be both inductive and conjugative. The coupling constants of the L 2Co(CO)^X (L = Ph.jP; X = CoCCO)^, BPh^) compounds appear to be i n s e n s i t i v e to the formal oxidation state of the Co atom. The spectra of some substituted derivatives of Co„(C0) o have been interpreted z o i n the l i g h t of t h e i r structures. The nqr data of (RC = CR')Co„(CO)-z o (R = R' = H, CH2OH, Ph, CF 3; and R = Bvf, R* = H) are found to be s e n s i t i v e to the inductive and conjugative e f f e c t s of the R and R' groups. 35 ' 1 Some C l nqr studies on f l u o r o a l i c y c l i c YC = CC1(CF„) z n (Y = substituent, n ~ 2 to 4) compounds were also undertaken. The 35 C l nqr frequencies of these compounds are roughly re l a t e d to the e l e c t r o n e g a t i v i t y of the substituents Y and to the r i n g s i z e , suggesting that the inductive e f f e c t i s the main factor responsible f o r the change. I t also seems that the substituents F, OMe, NMe2 show both inductive and conjugative e f f e c t s . - i v -Acknowledgements I am extremely g r a t e f u l to Professors W.R. Cullen and M.C.L. Gerry for t h e i r patience, invaluable guidance and assistance throughout the course of t h i s work and during the preparation of t h i s manuscript. I would also l i k e to express my appreciation to Professor J.R. Sams and Dr. J.C. Scott f o r t h e i r i n t e r e s t i n Mossbauer spectra of some b i i r o n carbonyl complexes. My thanks are extended to the s t a f f s of the E l e c t r o n i c , Glass, and Mechanical workshops f o r t h e i r t e c h n i c a l assistance Mr. P. Borda for prompt and accurate microanalyses Dr. E. Koster, Miss P. Watson and Mr. W. Lee for running the nmr spectra Mrs. Y.S. Choo for the typing of this thesis Drs. R.K. Pomeroy, A.E. Fenster and A.W. Wu for reading parts of the manuscript. The f i n a n c i a l support of the Chemistry Department, the U n i v e r s i t y of B r i t i s h Columbia and the International Nickel Company of Canada, Limited, i s g r a t e f u l l y acknowledged. TABLE OF CONTENTS Abstract Acknowledgements L i s t of Tables L i s t of Figures General Introduction L i s t of Abbreviations Chapter 1. B i i r o n Carbonyl Complexes I. Introduction I I . Experimental 1. Techniques 2. Materials 3. New Ligands B. C. U l t r a v i o l e t I r r a d i a t i o n Using a 450 Watt Lamp Al t e r n a t i v e Way of Making ( L m ) 2 ( L - L ) F e 2 ( C 0 ) 4 and ( L m ) 3 ( L - L ) F e 2 ( C O ) 3 Complexes Reaction of ( L m ) 3 ( L - L ) F e 2 ( C O ) ^ with CO Page i i i v x i v x v i i 1 3 6 11 11 12 12 A. fgfars 12 B. fgfars 13 4. (L- L ) F e 2 ( C 0 ) 6 Complexes 13 5. Preparation of Compounds Derived from (L-L)Fe 2(CO) 6 Complexes 15 A. U l t r a v i o l e t I r r a d i a t i o n Using a 100 Watt Lamp 15 28 28 2^uu; 3 witn CO 28 6. Reaction of (Ligand) b(L-L)Fe 2(CO) 4 with I and HC1 29 - v i -TABLE OF CONTENTS (Contd.) 6, ( L m ) 2 ( L - L ) F e 2 ( C 0 ) 4 (Type VI) and (L m)^(L-L)Fe,,(CO) (Type VII) Complexes 50 7. ( L m ) 3 ( L - L ) F e 2 ( C O ) 3 (Type VIII) Complexes 51 8. Cleavage of Fe-Fe Bond by I 2 and HC1 52 IV. General Discussion 53 1. Preparative 53 2. Infrared Spectra 57 3. E l e c t r o n i c Spectra References Chapter 2. B i c o b a l t Carbonyl Complexes I. Introduction Page I I I . Results and Discussion 31 1. (L-L)Fe 2(CO) 6 Complexes 31 2. L m(L-L)Fe 2(C0) (Type I) and ( L - L ) m ( L - L ) F e 2 ( C 0 ) 5 (Type II) Complexes 36 3. ( L - L ) C ( L - L ) F e 2 ( C O ) 4 (Type III) Complexes 39 4. ( L - L ) b ( L - L ) F e 2 ( C O ) 4 (Type IV) Complexes 44 5. (dppp) (L-L)Fe 2(CO) 4 (Type V) Complexes 48 59 62 64 1. Review of the Reaction of Lewis Bases with C o o ( C 0 ) o 64 A. S u b s t i t u t i o n Derivatives with Bridging CO Groups 64 ( i ) LCo 2(CO)^(CO) 2 64 ( i i ) L 2 C o 2 ( C 0 ) 4 ( C 0 ) 2 65 - v i i -TABLE OF CONTENTS (Contd.) Page ( i i i ) Higher Substituted L Co„(C0)^ (CO)„ 67 n z t>—n JL B. Sub s t i t u t i o n Derivatives without Bridging CO Groups 70 ( i ) L C o 2(C0) ? 70 ( i i ) [LCo(CO) 3J 2 70 ( i i i ) L n C o 2 ( C 0 ) ^ _ n 71 (iv) Other Cobalt Carbonyl Complexes 73 C. Disproportionation Derivatives 75 ( i ) L Co(CO)" X" 75 n 5-n ( i i ) L n C o 2 + [ C o ( C O ) ^ ] 2 77 D. Cluster Complexes 77 (i ) Sulfur-Containing Compounds 77 ( i i ) Cullen's Cluster Compound 79 E. ( T r-Alkyne)Co 2(CO) 6 and Related Complexes 80 F. Complexes Isolated from High Temperature and High CO Pressure Reaction 82 2. Statement of Problems 83 I I . Experimental • 85 1. Reaction of Co 2(C0)g with Various Ligands 88 A. With fgfars and fgfars 88 B. With f.AsP 88 C. With f.PS 90 D. With f^AsS and Related Unsymmetrical Ligands 90 E. With dppm 93 - v i i i -TABLE OF CONTENTS (Contd.) Page 2. Reaction of f^fosCo 2(CO) g with f^fos 94 3. Reaction of ( f o s ) ^ Co^CO)^ with CO 94 4. Reaction of [ (T r-C 7H g) CCo(CO) 2] 2 with f^fos 95 5. Reaction of (f^fos) 2Co 2(CO) 4 with CO 95 6. Reaction of (TI-C-H0) CCo o(C0), with f,fos, f 0 f o s and dab 96 7. Reaction of [LCo(CO> 3] 2 with (L-L) 97 A. Thermal Reaction 97 (i ) Reaction of [Bu^PCo^O) 3 ] 2 with f^fos 97 ( i i ) Reaction of [(MeO) 3PCo(CO) 3] 2 with f^fos 98 ( i i i ) Reaction of [Ph 3PCo(CO) 3] 2 with dppe 98 B. Photolytic Reaction 99 8. Reaction of (RC = CR)Co 2(CO) 6 with Monodentate Ligands 103 A. Preparation of (MeO)„P(RC = CR)Co?(C0)c. Complexes (R = Ph, CH20H) 103 B. Preparation of [(MeO) J?-]~(RC = CR)Co 9(C0), Complexes (R = Ph, CH20H) 1 4 104 C. Preparation of [(MeO) 3P] 3(RC = CR)Co 2(C0) 3 Complexes 104 D. Preparation of [(MeO) 3P] 4(PhC = CPh)Co 2(C0) 2 105 9. Reaction of L (RC = CR)Co„(C0)^ with CO 106 n 2 6-n A. Reaction of [(MeO)-P],(PhC = CPh)Co„(CO)0 with CO 106 B. Reaction of [(MeO) 3P] 3(PhC E CPh)Co 2(CO) 3 with CO 106 C. Reaction of [(MeO) 3P] 2(PhC = CPh)Co 2(C0) 4 with CO 107 - ix -TABLE OF CONTENTS (Contd.) P a 8 e 10. Reaction of (PhC = CPh)Co2(C0) g with Bidentate Ligands 107 A. Preparation of (L-L)(PhC E CPh)Co2 (co)4 107 B. Preparation of (L-L) (PhC = CPh)Co 2(co) 2 110 11. Reaction of (L-L) 2(PhC = CPh)Co 2(C0) 2 with CO 112 12. Reaction of (PhC = CH)Coo(C0), with Bidentate Ligands 112 Results and Discussion 116 1. (L-L)Co 2(C0) 6 Complexes 116 2. (L-L) 2Co 2(C0) 4 Complexes 120 A. (f 4fos) 2Co 2(C0) 4 120 B. (f 4fos) 2Co 2(CO) 4 121 3. LCo 2(CO) 7 122 4. [(L-L)Co(CO) 3] 2 Complexes 123 5. Reaction of [LCo(CO) 3] 2 with (L-L) 127 6. L(RC E CR)Co2(CO)5 Complexes 131 7. L 2(RC = CR)Co2(CO)4 Complexes 133 8. L 3(RC = CR)Co2(CO)3 135 9. L 4(RC = CR)Co2(CO)2 135 10. (L-L)(PhC = CH)Co2(CO)4 137 A. (L-L) C(PhC = CH)Co2(CO)4 137 B. (L-L) b(PhC = CH)Co 2(C0) 4 140 11. (L-L)(PhC = CPh)Co 2(C0) 4 140 A. (L-L) b(PhC = CPh)Co 2(C0) 4 140 B. (L-L)C(PhC = CPh)Co2(CO)4 142 TABLE OF CONTENTS (Contd.) P a g e C. (L-L) C'(PhC E CPh)Co2(CO)4 U 2 D. (L-L)^'(PhC E CPh)Co2(CO)4 12. (L-L) 2(PhC = CPh)Co2(CQ)2 IV. General Discussion A. . The Nature of the Donor Atom of the Ligand B. The Nature of the Substituents on the Donor Atom of the Ligand C. The Bite of the Ligand D. Temperature Effect E. Solvent Effect F. Effect of CO Pressure G. General Comments Possible Reaction Paths A. Reaction of (RC = CR')Co2(CO)g with Mono- and Bi-dentate Ligands B. Reaction of Co2(CO)g with Mono- and Bi-dentate Ligands References 145 145 148 1. Factors Affecting the Formation of Cobalt Carbonyl 1 / Q Complexes ±c*° 148 149 149 150 150 151 151 152 152 153 (i) The Disproportionation of Co 2(C0) g 1 5 5 ( i i ) Substitution Reaction of Co 2(C0) g 1 5 6 160 - x i -TABLE OF CONTENTS (Contd.) Page Chapter 3. Introduction to Nuclear Quadrupole Resonance spectroscopy I. Theory of Nuclear Quadrupole Resonance 165 1. Introduction 165 2. The Nuclear Quadrupole Moment 166 3. The Electric Field Gradient (EFG) 167 4. Quadrupole Energy Levels 168 5. Factors Affecting the EFG 176 6. The Intramolecular EFG 176 II. Instrumentation and Procedure 179 1. The Spectrometer 179 2. Factors Affecting the Signal Strength 180 3. Experimental Techniques 182 Appendix I 184 References 188 35 Chapter 4. Cl Nuclear Quadrupole Resonance Spectra of 1-Substituted 2-Chloro Polyfluorocycloalkenes I. Introduction 190 II. Experimental 193 III. Results and Discussion 197 1. Assignment of Signals 197 198 2. Effect of Substituents on v ( 3 5 C l ) - x i i -TABLE OF CONTENTS (Contd.) 3. E f f e c t of Ring Size 35 4. Temperature Dependence of v( Cl) IV. Summary Appendix I. Cpmputer Program for Multiple Regression (I) I I . Computer Program f o r M u l t i p l e Regression (II) I I I . Substituent E f f e c t s References 59 Chapter 5. Co Nuclear Quadrupole Resonance Spectra of Some Cobalt Carbonyl Complexes I. Introduction I I . Experimantal I I I . Results and Discussion 1. [Y 3PCo(CO) 3] 2 Complexes A. Comparison of Experimental Parameters B. The E f f e c t of Co-P Back Donation 2. L 2Co(CO) 3X~ Complexes 3. Substituted Complexes of Co 2(CO)g A. f 4 f a r s b C o 2 ( C O ) 6 B. ( T T - C 7 H 8 ) C C O 2 ( C O ) 6 C. t(7T-C 7Hg) CCo(CO) 2] 2 4. (RC = CR')Co 2(CO) 6 Complexes A. Forbidden Transitions - x i i i -TABLE OF CONTENTS (Contd.) B. Spectra of (HOCH.C = CCHo0H)Co„(C0), Z z z o C. The Assignment of Transition Frequencies of Other Compounds D. Lattice Effects i n (RC = CR')Co2(CO)6 E. Substituent Effects in (RC = CR')Co2(CO)6 F. Effect of Ligand Substitution IV. Summary References - xiv -LIST OF TABLES T a b l e , Chapter 1 XII Maximum v c of Some Representative Iron Carbonyl Complexes XIII Page I Preparative Details, Analytical Data and nmr Spectra of New (L-L)Fe„(C0), Complexes 14 II Preparative and Analytical Data for Derivatives of (L-L)Fe 2(CO) 6 16 III Spectroscopic Data for Derivatives of (L-L)Fe2(CO)g 24 IV Infrared CO Bands of (L-L)Fe 2(C0)g Complexes 32 V MOssbauer Data at 80°K for (L-L)Fe„(C0), Complexes 33 VI Mossbauer Parameters at 80°K for (Ligand) m(L-L)Fe 2(CO) 5 Complexes 36 VII hi NMR Data for (Ligand) m(L-L)Fe 2(C0) 5 Complexes 38 VIII Mossbauer Parameters at 80°K for (L-L)°(L-L)Fe2(CO)^ Complexes 43 IX Mossbauer Parameters at 80°K for (L-L) b (L-L)Fe 2(C0) 4 Complexes 46 X Bites of the Fluorocarbon-bridged Ligands 55 XI Bites of Some Hydrocarbon-based Ligands 56 57 Electronic Spectra of (L-L)Fe 2(CO) g Complexes 59 Table Chapter 2 I Preparative Details for the New Ligands II Analytical and NMR Data for the New Ligands III Mass Spectrum of [f 4AsSCo(C0)^] 2 93 IV Photolytic Reaction of [LCo(CO) 3] 2 with (L-L) 1 00 Page 86 87 - XV -LIST OF TABLES (Contd.) T a b l e Chapter 2 XI XII T a b l e Chapter 3 T a b l e Chapter 4 Previously Reported 3 5C1 NQR Data for Chloro-II polyfluoroalkenes at -196° 19 i ' r III References for the Preparation of YC = CC1(CF0) (n = 2 to 4) 1 n IV JC1 NQR Data for Chloropolyfluoroalkenes at -196° 37 V Cl NQR Frequencies for Chloropolyfluoroalkenes at -196° Page V Infrared Data for (Ligand)Co 2(CO) g 117 VI Infrared CO Bands of Some (L-L) 2Co 2(CO) 4 Complexes 121 VII Infrared Data for LCo 2(CO) 7 Complexes 123 VIII Infrared Data of Some [(Ligand) Co (CO) Complexes 125 IX \ NMR Data for Some L (RC = CR)Co_(C0), Complexes 132 n I o-n X Infrared CO Bands of Some L n(RC = CR)Co2(CO)g_n Complexes 133 Infrared CO Bands of (L-L)(RC = CR')Co2(CO)4 Complexes 143 Favorable Conditions for the Formation of Bridged and Non-bridged Substituted Derivatives of Co 2(CO) g 156 Page I Quadrupole Moments of Nuclei Investigated Here 167 II Averaged Transition Probabilities, W, of Various Transitions for 1=7/2 1 7 5 Page 190 F NMR. Data for YC = CF(CF 2) 2 Compounds 192 194 195 196 - xvi -LIST OF TABLES (Contd.) Table VI VII VIII IX Chapter 4 Inductive and Conjugative Parameters of Substituents One Parameter Correlation Equations for YC = CClfCF ) (n - 2, 3) 2 B Two Parameter Correlation Equations for Some YC = CC1(CF_) (n= 2, 3) z n Temperature Coefficients of Some YC = CCl(CF-) (n = 2 to 4) Compounds 2 n Page 199 204 207 215 Table II III IV V VI VII VIII IX X Chapter 5 SQ + -Co NQR Data for L 2Co(C0) 3X and [LCo(CO) 3] 2 Complexes 59 Co NQR Data for (RC = CR')Co 2(C0) 6 and Related Compounds Molecular Parameters for [Y 3PCo(C0) 3] 2 Compounds Inductive and Conjugative Parameters of Y in the [Y 3PCo(C0) 3] 2 Compounds The a - and iT-Bonding Parameters 59 Co NQR Data for Some Co(I) Complexes Some Bond Distances and Angles for f^farsCo 2(C0)g Forbidden Frequencies of (RC = CR1)Co,(CO), (R = R' = H; R = Bu t, R1 = H) Possible Assignments of v^, and v 3 for (HC = CH)Co 2(C0) 6 at 0° ^9Co NQR Data of (RC = CR*)Co„(C0)- and Va T and la of R and R* 2 6 ^ I Page 231 233 240 242 246 249 252 259 262 266 - x v i i -LIST OF FIGURES Figure Chapter 1 1. 2. 6. IR Spectra of Representative Iron Carbonyl Complexes F 1 g u r e ChaDter 2 2^ ;8 7. Reaction Scheme for Substitution of Coo(C0) 2 8 1. Two Possible Types of Distorted Nuclei 4. The Transition Frequencies as a function of the Asymmetry Parameter for 1=7/2 Page Crystal Structure of f^farsFe 2(C0> 6 6 Crystal Structure of f 4AsP°f^AsPFe^CO)^ 40 3. Two Types of Infrared CO Bands for (L-L) c(L-L)Fe 2(CO) 4 Complexes 41 4. MBssbauer Spectrum of f,AsP Cf,farsFe.(C0). 42 H o 2 4 5. MQssbauer Spectrum of f^fos^f^fosFe 2(CO)^ 45 58 Page 1. Crystal Structure of [ (TT-C,H., _)Co(C0) „]„ 68 O 1U 2. 2 2. Crystal Structure of As 2Co 2(CO) 5PPh 3 75 3. Crystal Structure of (f^fars) 2Co 4(CO) g(2H?) 79 4. Crystal Structure of As 3Co(C0) 3 83 5. Reaction of (RC = CR')Co2(CO)6 with Mono- and Bi-dentate Ligands 154 6. Disproportionation of Co ?(C0) R 157 159 USHEf. Chapter 3 Page 166 2. Quadrupole Energy Levels and Transitions for I = -| 170 3. Quadrupole Energy Levels and Transitions for I = y 172 174 - x v i i i -LIST OF FIGURES (Contd.) Figure Chapter 3 5. Schematic Diagram of an NQR Spectrometer 6. Decca NQR Spectrometer Block Diagram Figure Chapter 4 Page 179 181 Page 1. v ( 3 5 C l ) of YC = CC1(CF 2) 2 vs a of Y 202 v ( 3 5 C l ) of YC = CC1(CF 2) 3 vs aj. of Y 203 35. 2. 3. 6. v( Cl) of YC = CC1(CF ) vs a T o (after dividing Y into three groups) f Y groups) 205 4. v( Cl) of YC = CC1(CF ) vs o of Y (after dividing Y into three groups) '35„,N ,. „ £ . _i . 19 7. Temperature Dependence of EtC = CCl(dFn) F i g u r e Chapter 5 206 v( Cl) of YC = CC1(CF ?) ? vs 6( 1 9F) of Vinylic Fluorine of YC~= CF(CF 2) 2 (Y = F, OMe, NMe2) Temperature Dependence Curves of YC = CC1(CF ) 211 2 n 210 2 ) 2 213 page 1. Molecular Structure of Co o(C0) o 250 2. Crystal Structure of (TT-C,H 0) CCO„(CO), 253 / o 2. o 3. Crystal Structure of [ ( 7 r-C 6H g) CCo(C0) 2] 2 255 4(a). The Bond Lengths of (PhC = CPh)Coo(C0), 257 2 O 4(b). The Bond Angles of (PhC = CPh)Co0(00). 257 2 5. A Plot of e Qq vs n for (RC E CR')Co 2(C0) 6 Complexes 268 GENERAL INTRODUCTION Owing to their interesting structures and bonding characteristics, metal-metal bonded compounds have been the subject of extensive studies during the last two decades. As a result there are now many such compounds including metal carbonyl derivatives, metal halides and a variety of cluster species.- Among these, biiron and bicobalt carbonyl complexes are particular attractive for the following reasons. F i r s t l y they are susceptible to investigation by well known spectroscopic techniques such as vibrational, electronic, nuclear magnetic resonance, Mo'ssbauer or nuclear quadruple resonance spectroscopy. Secondly, they serve as simple models for elucida-tion of metal-metal bonding in more complicated systems such as metal clusters and alloys. For these reasons the investigation of these bimetal carbonyl complexes were carried out. This forms the basis of Chapters 1 and 2. Although the phenomenon of nuclear quadrupole resonance (nqr) was discovered much earlier than nuclear magnetic resonance (nmr) and Mossbauer effects, the nqr technique i s s t i l l under developed compared with the well developed nmr and Mbssbauer spectroscopies. This is an unfortunate situation because nqr is potentially a very powerful tool for elucidation of structure and bonding. This prompted an exploration of nqr spectroscopy. However, when this work was started, the spectrometer (a home made manual type) was only good for the region 30-40 MHz and for this reason the work 35 was initiated by studying Cl nqr spectra of 1-substituted 2-chloro-polyfIt'orocycloalkenes. With the purchase of a commercial Decca spectrometer - 2 -59 results came more easily and the more interesting project, Co nqr studies of some cobalt carbonyl complexes, was undertaken. The result of these studies are presented i n Chapters 4 and 5. For the ease of under-standing the contents of Chapters 4 and 5, an introduction to nqr spectroscopy i s given in Chapter 3. - 3 -Lists of Abbreviations The following abbreviations w i l l be used in this thesis. f^AsF : dimethylarsinopentafluorocyclobutene f^AsCl : l-dimethylarsino-2-chlorotetrafluorocyclobutene f^SF : methanthiopentafluorocyclobutene f^fars : 1,2-bis(dimethylarsino)tetrafluorocyclobutene f^AsP : l-diphenylphosphino-2-dimethylarsino-tetrafluorocyclobutene f^fos : 1,2-bis(diphenylphosphino)tetrafluorocyclobutene fgfars : 1,2-bis(dimethylarsino)hexafluorocyclopentene fgAsP : l-diphenylphosphino-2-dimethylarsino-hexafluorocyclopentene f^fos : 1,2-bis(diphenylphosphino)hexafluorocyclopentene fgfars : 1,2-bis(dimethylarsino)octafluorocyclohexene fgfos : 1,2-bis(diphenylphosphino)octafluorocyclohexene dab : 2,3-dimethylarsinohexafluorobutene f^PS : l-diphenylphosphino-2-roethanthio-tetrafluorocyclobutene f^AsO : l-dimethylarsino-2-methoxy -tetrafluorocyclobutene f^AsS : l-dimethylarsino-2-methanthio-tetrafluorocyclobutene f^AsN : l-dimethylarsino-2-dimethylamino-tetrafluorocyclobutene diars : 1,2-bis(dimethylarsino)benzene dppm : 1,2-bis(diphenylphosphino)methane dpam : 1,2-bis(diphenylarsino)methane dppe (Pf-Pf, diphos) : 1,2-bis(diphenylphosphino)ethane dpae (Asf-Asf) : 1,2-bis(diphenylarsino)ethane arphos (Asf-Pf) : l-diphenylphosphino-2-diphenylarsinoethane - 4 -dppp : 1,3-bis(diphenylphosphino)propane DPPA : 1,2-bis(diphenylphosphino)acetylene cis Pf=Pf : cis-l,2-bis(diphenylphosphino)ethylene Pf-Pf-Pf : bis(2-diphenylphosphinoethyl)-phenylphosphine P(-Pf) 3 : tris(2-diphenylphosphinoethyl)-phosphine Pf-Pf-Pf-Pf : 1,1,4,7,10,10-hexaphenyl-l,4,7,10-tetraphosphadecane bipy : bipyridine phen : 1,10-orthophenanthroline TP : l,l,l-tris(diphenylphosphinemethylene)ethane CO*5 : bridging CO group C O : terminal CO group C-,HQ : norbornadiene C^ Hg : isoprene CgHg : cyclohexa-l,3-diene C^H^Q : 2,3-dimethylbuta-l,3-diene L : monodentate ligand L-L : bidentate ligand i r : infrared uv : ultraviolet nmr : nuclear magnetic resonance nqr : nuclear quadrupole resonance V^Q : infrared carbonyl stretching frequency \ : wavelength J : nmr coupling constant ppm : parts per million - 5 -Mbssbauer isomer shift MBssbauer quadrupole sp l i t t i n g f u l l width at half maximum about, approximately effective atomic number - 6 -Chapter 1 Biiron Carbonyl Complexes I. Introduction In 1969 Cullen et. a l . [1] reported that the fluorocarbon-bridged ligand f^fars reacted with Fe^iCO)^ *-n boiling cyclohexane solution to give an orange solid. The X-ray analysis [2] revealed that this com-pound has the structure shown in Figure 1. The molecule consists of two A B A * inequivalent iron atoms Fe and Fe . Fe is approximately octahe-s B drally coordinated by two As atoms, three CO3 groups, and Fe ; while B * Fe is coordinated by the C=C bond of the cyclobutene group, three CO groups, and Fe (which is acting as a donor). Thus, the coordination around Fe can be regarded as either a distorted trigonal bipyramid with the C=C bond occupying one site or a distorted octahedron with the two C atoms occupying two sites. In either case, the symmetry at Fe is lower than that at Fe . Subsequently, the related compounds f nfosFe2(CO)g (n = 4, 6) [2], f AsPFe0(CO), (n = 4, 6) [3,4] and dabFe„(CO), [5] were simi-n z o Z D l a r l y prepared. However, a l l attempts [4] to prepare fgfosFe2(C0)g failed. This type of compound which contains two different iron atoms, seemed to be an interesting subject for Mbssbauer and other spectroscopic * A , „ B ^ We w i l l always refer to Fe and Fe as the "hexa-coordinated" and penta-coordinated" iron atoms, respectively (as indicated in Figure 1). § In some places in this thesis common atoms, groups and compounds w i l l be designated by symbols. - 7 -- 8 -studies. With this in mind, the preparation of the remaining complexes of the series, f farsFe„(CO), (n = 6, 8) was carried out. Also, because n L o of the large demand for this type of compound for studies on their reactions with, group V ligands, an improved method for the synthesis was developed. Although many dinuclear iron carbonyl compounds with Fe-Fe bond are now known, l i t t l e work has been done on the reactions of these compounds with Lewis bases. To our knowledge substitution reaction of only the following four systems have been reported, v i z . (a) [(ir-C,-H,.)Fe(CO)2]2, i.» [6-11]; (b) [Fe(CO) 3SR] 2 (R = Me, Et, Bufc, Ph), 2, [12-15]; (c) Tr-C 5H 5Fe 2PPh 2(CO) 5, 3, [16] and (d), [Fe(CO) 3PR 2] 2 (R = Me, Et, Ph), 4_, [28], a l l of them contain bridging groups (CO, SR or PR0) . - 9 -In the preliminary study [3] of this work, i t was found that when an acetone solution of f4AsPFe2(CO)g and f^AsP was irradiated with uv light, two products of the same molecular formula (f 4AsP)f 4AsPFe2(CO) 4 were isolated. The Mossbauer and i r data of the red isomer are consis-tent with a structure in which the second f^AsP ligand replaces one CO group from each iron atom, 5_ (L-L* = f AAsP) . On the other hand, the 5 6 corresponding data for the dark brown isomer suggest a structure in which the second f^AsP ligand chelates to Fe (after replacement of two CO groups from Fe ), 6_. Under the same conditions, f^AsP and f g A s P react with f 6AsPFe 2(CO) 6 to afford the bridged f 4AsP bf 6AsPFe 2(CO) 4 and the chelated f 6AsP Cf 6AsPFe 2(CO) 4 > respectively. The monodentate ligand Ph3P also reacts [3] with f 4AsPFe 2(CO)g to give an orange-red complex. Based on i t s i r and MSssbauer data, the structure ]_ in which the Ph^P ligands replace one CO group from Fe A was proposed. In view of the ease of substitution of CO in the (L-L)Fe 9(C0) < ; - 10 -(CO) 3 7 compounds, i t seemed desirable to extend this work i n the hope of isolating new compounds for MSssbauer and other physical studies. Furthermore, the ligand bridged complexes such as (f 4AsP) bf 4AsPFe2(CO) 4, _5, could conceivably be useful for homogeneous catalysts. This prompted a preliminary investigation of i t s reaction with some simple molecules such as I 0 and HC1 . - 11 -II. Experimental 1. Techniques A l l volatile chemicals were manipulated using a conventional vacuum system. The Carius tubes used in this work were constructed of thick-walled Pyrex glass with a volume of about 80 ml. Ultraviolet irradiations were carried out i n two ways : (a) The reactants and the solvent (ca. 20 ml.) were sealed in a Carius tube. This was shaken and irradiated with a 100 Watt Hanovia lamp at ca. 20 cm distance. (b) A 450 Watt Hanovia lamp was placed i n a water cooled quartz jacket which was inserted into a large glass jacket containing the solution (ca. 300 ml.) to be irradiated. Nitrogen was bubbled through the solution during the course of the reaction. A l l chromatography was carried out under a nitrogen atmosphere using nitrogen saturated solvents. The petroleum ether mentioned below i s the 30° - 60° boiling fraction. Infrared spectra were recorded on a Perkin-Elmer 457 spectropho-tometer and calibrated using polystyrene and cyclohexane. Unless otherwise stated, a l l nmr spectra were run on a Varian T-60 spectrometer with chemical shifts given in ppm downfield from internal SiMe^ for 19 and upfield from internal CFCl^ f° r F spectra. Ultraviolet and visible spectra were run in methylene chloride solution using a Cary 15 spectrophotometer. Mass spectra were measured with an AEI MS-9 mass spectrometer with direct introduction of solid samples. The Mossbauer data were obtained by Dr. J.C. Scott of this department [21]. The spectrometer - 12 -and attendant experimental details have been described elsewhere [1, 21] A l l molecular weights were measured with a Mechrolab. Vapor Pressure Osmometer using benzene as a solvent. Melting points were determined in evacuated capillaries and are uncorrected. Most microanalyses were performed by Mr. P. Borda of this department, and some by Pascher Mickroanalytisches Laboratorium, Bonn, Germany. 2. Materials The iron pentacarbonyl and the ligands diars, dpam, dpae, arphos, dppm, dppe, dppp, (Ph0).jP and (E = P, As, Sb) . were obtained commercially and used without further purification. The fluoro-carbon-bridged ligands f^fars, r n A s P ( n = >^ *>) and r n r o s ( n = 4, °> 8) were prepared using published procedures [1, 17-20]. The other ligands were prepared as described next. 3. New Ligands A. fgfars Perfluorocyclopentene (16 g, 75 mmol.) and excess dimethylarsine [22] (20 g, 189 mmol,) were sealed in a Carius tube. The tube was heated to 100 for 6 days. It was cooled In liquid nitrogen and then opened. The volatile contents were removed in vacuo. The less volatile portion was d i s t i l l e d under vacuum to give a colorless liquid (17.5 g, 61%), b.p. 84/0.1 mm., identified as f^fars. Anal. Calc. for C9 H13 F8 A S2 : C ' 2 8 , 1 ; H ' 3 * 2 ; F* 2 9 ' 7 ' F o u n d : C> 2 8-°> H ' 3-1> F> 2 9 ' 6 "hi nmr(CDCl ): singlet at 1.33 ppm. 1 9 F nmr(CS?) : tr i p l e t (J = 5.5 Hz) 3 ^ - 13 -at 103.8 ppm (area 4), quintet (J = 5.5 Hz) a t 130.0 ppm (area 2), B. fgfars Perfluorocyclohexene (17.0 g, 75.0 mmol.) and excess dimethy-larsine [22] (20.0 g, 189 mmol.) were sealed in a Carius tube. The tube was heated to 150° for 8 days then opened in the usual way. After removing the volatile contents i n vacuo, the non-volatile residue was extracted with petroleum ether. The extract was concentrated and cooled to yield pale yellow crystals of f g f a r s (19.5 g, 59%), m.p. 51°. Anal. Calc. for c 1 0 H i 2 F 8 A s 2 : C ' 2 7' 7> H» 2 , 8 » F» 3 5 - ° * Found : C, 27.5; H, 2.8; F, 34.7. 1H nmr(CDCl3) : singlet at 1.38 ppm. 19 F nmr(CS2): doublet (J = 0.12 Hz) at 102.0 ppm (area 4), doublet (J = 0.24 Hz) at 133.4 ppm (area 4). 4. (L-L)Fe 2(CO) 6 Complexes In this work an improved way of making this type of complex was found as outlined next. A benzene solution (10-15 ml.) containing 2.0 mmol. of ligand and 5 to 10 g. of Fe(CO),. was heated at 150° in an evacuated Carius tube for 1 to 2 days. At the end of the reaction period the i n i t i a l l y yellow solution had become deep red. The tube was cooled i n liquid nitrogen and opened. The volatile contents were removed at reduced pressure. The resulting red residue was dissolved i n a minimum amount of methylene chloride and chromatographed on a F l o r i s i l column. A 5% diethyl ether - 95% petroleum ether mixture eluted the more soluble iron carbonyl complexes - 14 -(e.g. (L-L)Fe(CO) 4, (L-L)Fe(CO) 3, etc.). The desired product was eluted with a 10% diethyl ether - 90% petroleum ether mixture. The preparative details, analytical data and nmr spectra of the two new (L-L)Fe„(CO), 2. O complexes (L-L = fgfars and fgfars) are shown in Table I, while their infrared, electronic and Mb'ssbauer spectral data are listed in Tables IV, XIII and V, respectively. * 1 TABLE I. Preparative Details , Analytical Data and H nmr Spectra of New (L-L)Fe : ?(C0) f i Complexes L-L Color m.p. (°C) Reaction Time** (days) Yield (%) Anal. Calc. (%) Found 1 § H nmr (CDClg) ppm. fgfars orange-red C 27.5 C 26.8 S : 1.55 (area 6) 225° 1 70 H 2.3 H 2.3 S : 2.40 (area 6) fgfars orange-red 210° 1 45 C 26.9 H 1.7 C 27.0 H 1.9 S : 1.57 (area 6) S : 2.53 (area 6) Other known (L-L)Fe 2(C0)g complexes could also be prepared by the new method, e.g. f^farsFe 2(C0) 6, reaction time 1 d., yield 20%; f 4AsPFe 2(CO) 6, 2 d., 90%; f 4fosFe 2(C0) 6, 2d., 80%; fgAsPFe2(CO)g, I d . , 50%; f 6fosFe 2(CO) 6, 2 d., 10%. ** Prolonged heating should be avoided because some (L-L)Fe„(C0), / o complexes (L-L = f 4 f a r s , fgfars, fgfars) decompose to yield (Me 2As) 2Fe 2(CO)g. § S = singlet - 15 -5. Preparation of Compounds Derived From (L-L)Fe,(C0) 6 Complexes Eight types of iron carbonyl complexes have been obtained from the (L-L)Fe 2(CO) 6 complexes, namely, (i) L m(L-L)Fe 2(CO) 5 , ( i i ) (L-L) m(L-L)Fe 2(CO) 5, ( i i i ) (L-L) C(L-L)Fe 2(CO) 4, (iv) (L-L) b(L-L)Fe 2(CO) 4, (v) dppp b(L-L)Fe 2(CO) 4, (vi) (L m) 2(L-L)Fe 2(CO) 4, (vii) (L m) 2(L-L)Fe 2(CO) 4 and ( v i i i ) (L m) 3(L-L)Fe 2(CO) 3; where m, c and b indicate whether the ligand acts as a monodentate, chelating, or bridging group, respectively; and s and d indicate whether the two L m ligands are on the same or different iron atoms, respectively. These complexes were prepared by two general methods. A. Ultraviolet Irradiation Using A 100 Watt Lamp The preparative details are li s t e d in Table II. The (L-L)Fe o(C0), 2 o (1.0 mmol.), appropriate amount of ligand, and acetone (15—20 ml.) were sealed under vacuum in a Carius tube. After irradiation for several days (see column 4) with the 100 Watt uv lamp, the tube was opened and the volatile contents were removed. The residue was dissolved in a small amount of methylene chloride and chromatographed on F l o r i s i l . The excess ligand and the unreacted (L-L)Fe 2(C0)g were removed by eluting with a 10% diethyl ether - 90% petroleum ether mixture. The f i n a l product was eluted with the solvent indicated in column 5. The analytical data for these complexes are also given in Table II and their i r CO bands and electronic spectral data are li s t e d i n Table III. * ' For types (i) to (v) complexes, equimolar quantities of the (L-L)Fe~(CO), 2 6 and ligand were used, while for types (vi) to ( v i i i ) complexes, 1:2 or 1:3 molar ratios were used. TABLE II. Preparative and Analytical Data for Derivatives of (L-L)Fe-(CO) Complex Color 5 m.p. Reaction Time 4* Eluent yield Anal.(%) Mol. Wt. (°C) (days) (%) Found (Calc.) Found (Calc.) Ph 3P mf 4AsPFe 2(CO) 5 d.r. 208 7 A 24 C 52.9 H 3.4 (53.0) ( 3.4) 915 (928) (PhO) 3P mf 4AsPFe 2(CO) 5 o.r. 190 . 4 A 20 C 50.8 H 3.4 (50.4) ( 3.2) 934 (976) Ph 3As mf 4AsPFe 2(CO) r. 180 20 A,B 5 C 50.7 H 3.4 (50.6) ( 3.2) Ph 3Sb mf 4AsPFe 2(CO) 5 o.r. 110 40 A 5 C 48.5 H 3.2 (48.3) ( 3.1) Ph 3Sb mf 4fosFe 2(CO) 5 o.r. 145 30 A 30 C 57.4 H 4.8 (55.6) ( 3.2) (PhO) 3P mf 4farsFe 2(CO) d.b. 175 15 A 40 C 41.9 H 3.4 (41.5) ( 3.0) (PhO) 3P mf gfarsFe 2(CO) 5 o.r. 140 10 A 18 C (39.8) H ( 2.7) TABLE II. (Contd.) Complex Color^ m.p. (°C) Reaction Time (days) + Eluent Yield (%) Anal.(%) Found (Calc.) Mol. Wt. Found (Calc.) f 4fars mf 4AsPFe 2(CO) o. r. 105 20 C,D ~5 C 37.0 H 2.8 (37.2) (2.8) 949 (1000) dpae mf 4AsPFe 2(CO) o.r. 160 12 C,D 6 C 50.8 H 3.7 (51.0) (3.5) 1170 (1152) arphos mf 4AsPFe 2(CO) 5 o. r. 140 31 C,D 7 C — H — (53.1) (3.7) dppp mf 4AsPFe 2(CO) 5 o.r. 155 4 C,D 18 C 56.1 H 4.2 (55.7) (3.9) 1004 (1078) f 4AsP mf 4farsFe 2(CO) 5 o.r. 168 4 C,D 30 C 37.0 H 2.8 (37.2) (2.8) 976 (1000) f 4AsP mf 8farsFe 2(CO) 5 o.r. 100 4 C,D 15 C — H — (36.0) (2.6) c C 41.4 (41.6) diars f 4AsPFe 2(CO) 4 d.b. 213 10 C,D 15 H 3.3 (3.1) TABLE II. (Contd.) Complex Color 5 m.p. (°C) Reaction Time (days) + Eluent Yield (%) Anal.(%) Found (Calc.) Mol. Wt. Found (Calc.) f 4fars Cf 4AsPFe 2(CO) 4 d.b. 140 20 A 5 C — (37.0) H — (2.9) f 4AsP Cf 4AsPFe 2(CO) 4 d.b. 190 7 C,D 25 C 45.9 (45.7) H 3.5 (3.2) 1040 (1052) f 4fos Cf 4AsPFe 2(CO) 4 d.b. 195 11 C,D 25 C 53.0 (53.0) H 3.3 (3.2) 1176 (1132) f 6fars Cf 4AsPFe 2(CO) 4 d.b. 190 11 C,D 22 C 36.3 (36.4) H 2.7 (2.8) 1051 (1022) f 6AsP Cf 4AsPFe 2(CO) 4 d.b. 215 8 C,D 18 C 44.2 (44.6) H 2.8 (2.9) 1085 (1102) f 6fos Cf 4AsPFe 2(CO) 4 d.b. 90 12 C,D 15 C — (51.8) H — (3.1) fgfars Cf 4AsPFe 2(CO) 4 d.b. 180 4 C,D 18 C 36.1 (35.8) H 2.8 (2.6) TABLE II. (Contd.) Complex Color 5 m.p. Reaction Time + Eluent Yield Anal.(%) Mol. Wt. (°C) (days) (%) Found (Calc.) Found (Calc.) fgfos Cf 4AsPFe 2(CO) 4 d.b. 183 31 C,D 10 C 50.7 H 3.2 (50.7) (3.0) 1203 (1232) dppm Cf 4AsPFe 2(CO) 4 d.b. 215 4 C,D 15 C 55.2 H 4.2 (55.2) (3.8) 1068 (1022) i dpae Cf 4AsPFe 2(C0) 4 d.b. 150 31 C,D 10 C 51.7 H 3.8 (49.1) (3.6) i arphos Cf 4AsPFe 2(C0) 4 d.b. 170 31 B,C 25 C 53.0 H 3.9 (53.3) (3.7) 1014 (1080) dppe Cf 4AsPFe 2(C0) 4 d.b. 150 4 C,D 50 C 56.4 H 4.3 (55.6) (3.9) dppp Cf 4AsPFe 2(C0) 4 d.b. 145 i o | C ~5 C H (56.0) (4.0) f 6AsP Cf 6AsPFe 2(CO) 4 d.b. 210 7 C,D 20 C 43.8 H 2.7 (43.8) (2.8) 1135 (1152) TABLE II. (Contd.) Complex Color 5 m.p. Reaction Time + Eluent Yield Anal. Mol. Wt. (°C) (days) (%) Found (Calc.) Found (Calc.) f 4AsP Cf &farsFe 2(CO) 4 d.b. 189 20 C,D 30 C 36.5 H 3.0 (36.4) (2.8) 978 (1022) f 6fars°f 6farsFe 2(CO) 4 d.b. 200 8" B 10 C 26.2 H 2.5 (26.6) (2.4) f 4AsP Cf gfarsFe 2(CO) 4 d.b. 171 4 B 15 C 35.8 H 2.8 (35.8) (2.6) f gfars CfgfarsFe 2(CO) 4 d.b. 205 4 C,D 20 C 26.2 H 2.3 (26.4) (2.2) 1115 (1092) diars°f.fosFe„(CO), 4 2 4 d.b. 194 10 C,D 40 C 50.6 H 3.9 (50.2) (3.6) f 4fars bf 4AsPFe 2(CO) 4 r. 190 20 B 15 C 37.2 H 2.9 (37.0) (2.9) 998 (972) f 4AsP f 4AsPFe 2(CO) 4 r. 205 4 B 8 C 45.6 H 3.2 (45.6) (3.2) 1101 (1052) TABLE II. (Contd.) Complex Color 5 m.p. (°C) Reaction Time (days) t Eluent Yield (%) Anal.(%) Found (Calc.) Mol. Wt. Found (Calc.) dpam f^AsPFe^CO)^ r. 160 B 26 C 51.3 H 3.9 (50.8) (3.5) 986 (1110) fo dpae f 4AsPFe 2(CO) 4 r. 155 31 B 12 C 50.8 H 3.6 (49.1) (3.6) arphos bf 4AsPFe 2(CO) 4 r. 161 31 B 10 C 52.7 H 4.1 (53.3) (3.7) 997 (1080) f 4AsP bf 6AsPFe 2(CO) 4 r. 175 7 B 20 C 44.6 H 3.0 (44.7) (2.9) 1126 (1102) f 4AsP f 4fosFe 2(CO) 4 r. 220 7 B 30 C 53.0 H 3.2 (53.0) (3.2) 1102 (1132) f 4 f o s b f 4 f o s F e 2 ( C O ) 4 r. 215 20 B 10 C 59.0 H 3.4 (59.4) (3.3) 1124 (1212) dppp f 4AsPFe 2(C0) 4 p. 172 l o f B 20 985 (1050) H 4.3 (4.0) - 22 -•u cd O fl o fe cd s o fe 0 CU 3 rH fe-ci <~-o co •H OJ ^ w e cd O ' r l TJ cd H ^ P i c j u o o CJ cu rH !• o u y—\ o O CO m H CM rH rH CO m rH o o o rH rH VO rH cn vo oo m cn r~- m r-» VO CM ~cr m cn r-» cn rH cn m cn rH CO cn vO ^ i n v—' m ^ i n m ^ v—* v—• v r-» r-. cn oo o r-» • • • • 1 1 1 1 • • 1 1 1 o s t s t cn 1 1 1 1 m s t 1 1 1 vO m m c j x u w O 33 Cj 33 u 33 CJ 33 CJ o o m r-. o m CM rH rH rH PP PP PP PP PP PP *> < <! < 00 m o o o o o rH vO 00 s t cn i n s t O o O o CO 00 CM rH cn s t rH rH rH rH rH rH & U u U X- JO TJ St St •—V o <? s t s t o o CJ —y o o O CM St CJ a o CM cu —s •—y '—y cu fe o CM CM CM fe CO u CU CU CD P i u fe fe fe CD cd CM p-l CO fe < m cu CO o CO St oo fe < < <+-! <4H CO s t St s t CO CM CO CM o <u <W M-l i i i— i T3 CM T3 CM T3 CM P-l P-l St y \ cn cn M-l CO -O £. *—s <3 C/l C/l o o ft cn CO cn Si ,£! ft si X! si S ft ft PH PL, PM •—y T3 ^ > 1—• '—' TABLE II. (Contd.) Complex Color 5 m.p. Reaction Time + Eluent Yield Anal.(%) Mol. Wt. (°C) (days) (%) Found (Calc.) Found (Calc.) (Ph 3Sb) 3f 4fosFe 2(CO) 3 r. 115 90 B 5 C 68.5 (73.1) H 5.2 (4.7) 1573 (1749) [(PhO) 3P] 3f 4AsPFe 2(CO) 3 r. 140 50 B 3 C (58.4) H (4.0) * In general a bidentate ligand (L-L') gave two or three of the following types of complexes (i) (L-L»)m(L-L)Fe2(CO) 5, ( l i ) (L-L') b(L-L)Fe 2 (C0) 4 and ( i i i ) (L-L') C(L-L)Fe 2(C0) 4. The derivatives of a certain (L-L') were prepared and isolated from the ^ ame reaction, although they are listed separately according to their classes in this table. d.b. = dark brown, d.r. = dark red, r. = red, o.r. = orange red, p. = purple. t Solvent key : A : 10-20% diethyl ether - 90-80% petroleum ether mixture; B : 20-50% diethyl ether - 80-50% petroleum ether mixture; C : CH2C12; D : acetone. The chromatographic column was eluted with the solvent i n the following order : petroleum ether, A, B, C, and then D. - 24 -TABLE III. Spectroscopic Data for Derivatives of (L-L)Fe„(CO), 2 6 Complex v C 0(CH 2Cl 2) cm -1 * * a,,(CH0Cl0)mu(e . ) maxv 2 2' K V molar Ph 3P mf 4AsPFe 2(CO) 5 2034(10) 1979(9) 1960(8) (PhO) 3P mf 4AsPFe 2(CO) 5 2044(10) 1990(9) 1970(9) Ph 3As mf 4AsPFe 2(CO) 5 2035(10) 1977(9) 1958(9) Ph 3Sb mf 4AsPFe 2(CO) 5 2036(10) 1976(9) 1954(9) Ph 3Sb"*f 4fosFe 2C(0) 5 2037(10) 1984(10) 1969(10) (PhO) 3P mf 4farsFe 2(CO) 5 2035(10) 1987(8) 1965(8) (PhO) 3P mf gfarsFe 2(CO) 5 2038(10) 1988(9) 1967(9) f 4fars mf 4AsPFe 2(CO) 5 2037(10) 1977(9) 1957(8) m dppm"'f4AsPFe2(C0)5 2043(10) 1980(7) 1960(8) dpae mf 4AsPFe 2(C0) 5 2035(10) 1976(9) 1957(9) arphos mf 4AsPFe 2(C0) 5 2035(10) 1975(8) 1955(8) dppp mf 4AsPFe 2(CO) 5 2036(10) 1979(9) 1960(8) f 4AsP mf 4farsFe 2(C0) 5 2035(10) 1980(9) 1958(9) f 4 A s p m f 8 f a r s F e 2 ( C 0 ) 5 2036(10) 1979(8) 1959(8) 386(10922) 382(9610) 387(3900) 385(9782) 388(6170) 391(8272) 385(12390) 386(8467) 387(5649) 382(11232) 380(7978) 386(8842) 397(15100) 391(5892) TABLE III. (Contd.) Complex v_.1(CH0Cl0) cm X mu(e , ) CO 2 2 max molar diars Cf 4AsPFe 2(CO) 4 2002(10) 1942(8) 1902(7) 385(5267) f 4fars Cf 4AsPFe 2(C0) 4 2008(10) 1948(8) 1903(6) 396(8111) f 4AsP Cf 4AsPFe 2(CO) 4 2011(10) 1952(9) 1917(8) 1907(7)sh. 403(8857) f 4fos Cf 4AsPFe 2(C0) 4 2012(10) 1950(8) 1922(7) 412(8256) f 6fars Cf 4AsPFe 2(CO) 4 2007(10) 1945(9) 1907(8) 392(9012) f 6AsP Cf 4AsPFe 2(C0) 4 2011(10) 1951(7) 1906(6) 401(8816) f 6fos Cf 4AsPFe 2(CO) 4 2013(10) 1952(9) 1921(7) 1907(5)sh, 400(~5000) f 8fars Cf 4AsPFe 2(CO) 4 2010(10) 1950(9) 1907(7) 393(10205) fgfos Cf 4AsPFe 2(CO) 4 2012(10) 1951(9) 1908(6) 403(9486) dppm Cf 4AsPFe 2(CO) 4 2004(10) 1942(9) 1904(8) 401(8657) dpae Cf 4AsPFe 2(CO) 4 2004(10) 1939(8) 1913(7) 405(10116) arphos Cf 4AsPFe 2(C0) 4 2004(10) 1939(9) 1916(8) 406(9144) dppe Cf 4AsPFe 2(C0) 4 2007(10) 1944(9) 1921(8) 413(7675) dppp Cf 4AsPFe 2(CO) 4 2006(10) 1941(8) 1916(7) 406(13936) - 26 -TABLE III. (Contd.) Complex v C ( )(CH 2Cl 2) cm"1 * A mu(e max mo f 6AsP Cf 6AsPFe 2(CO)^ 2011(10) 1951(8) 1909(7) 393(8273) f 4AsP Cf gfarsFe 2(CO) 4 2005(10) 1944(8) 1912(7) 403(10084) fgfars cf 6farsFe 2(CO) 4 2006(10) 1945(9) 1910(8) 393(8432) f 4AsP Cf gfarsFe 2(CO) 4 2002(10) 1945(9) 1912(8) 400(10124) fgfars CfgfarsFe 2(CO) 4 2009(10) 1948(9) 1913(8) 392(8339) diars Cf 4fosFe 2(CO) 4 2009(10) 1943(9) 1921(8) 372(5881) f 4 f a r s b f 4AsPFe 2(CO) 4 1982(9) 1944(10) 1922(7) 1894(6) 431(5742) f 4AsP bf 4AsPFe 2(CO) 4 1988(9) 1947(10) 1922(6) 1901(5) 430(7013) dpam bf 4AsPFe 2(CO) 4 1983(8) 1946(10) 1918(4) 1898(3) 430(4052) dpae f 4AsPFe 2(CO) 4 1981(9) 1939(10) 1915 (6) 1890(4) 442(11662) arphos bf 4AsPFe 2(CO) 4 1982(9) 1944(10) 1918(7) 1902(6) 448(3279) dppp bf 4AsPFe 2(CO) 4 1982(9) 1932(10) 1914(6) 1870(3) 522(7070) f 4AsP bf 6AsPFe 2(CO) 4 1985(9) 1948(10) 1926(5) 1905(4) 416(8820) TABLE III. (Contd.) Complex 'C0 ( C H2 C 12 ) -1 * cm X my (e max molar f 4AsP bf 4fp SFe 2(CO) 4 1988(9) 1950(10) 1920(3) 1902(4) 406(1415) f 4 f o s b f 4 f o s F e 2 ( C O ) 4 1994(9.5) 1954(10) 1928(6) 1910(8.5) 411(9532) dppp bf 4fosFe 2(CO) 4 1986(9) 1932(10) 1917(4,sh) - 1881(3) 504(7985) (Ph 3As) 2f 4AsPFe 2(CO) 4 1985(9) 1948(10) 1923(7) 1904(6) 432(5813) (Ph 3Sb) 2f 4AsPFe 2(CO) 4 1985(9) 1948(10) 1921(5) 1899(6) 423(6160) (Ph 3Sb) 2f 4fosFe 2(CO) 4 1993(9.5) 1953(10) 1930(3) 1911(5) 438(9890) [(PhO) 3P] 2f 4AsPFe 2(CO) 4 2013(10) 1966(9) 1907(5) 402(11762) [(PhO) 3P] 2f 8farsFe 2(CO) 4 2008(10) 1961(9) 1902(5) 400(9215) (Ph 3As) 3f 4AsPFe 2(CO) 1969(9) 1938(10) 1908(9) 500(1892) (Ph 3Sb) 3f 4fosFe 2(CO) 3 1968(6.5) 1944(10) 1913(7) 515(3210) [(PhO) 3P] 3f 4AsPFe 2(CO) 3 1970(7) 1946(10) 1915(6) 510(4350) ** Integers in parentheses refer to relative peak height. emolar = m o l a r extinction coefficient. - 28 -B. Ultraviolet Irradiation Using a 450 Watt Lamp This method is better than method A described above because i t reduces the reaction time from several weeks to several hours, and also the reaction carried out in this way can be easily monitored by the use of i r spectroscopy. An acetone solution (ca. 300 ml.) of (L-L)Fe 2(C0)g (1.0 mmol.) * and the appropriate amount of ligand were irradiated with the 450 Watt uv lamp for a few hours (ca. 1-3 h.). The reaction was monitored by i r techniques. The desired product was isolated as In Section A above. C. Alternative Way of Making (L m) 2(L-L)Fe 2(C0)^ and (L m) 3(L-L)Fe 2(C0) 3 Complexes Equimolar quantities (1.0 mmol.) of L m(L-L)Fe 2(C0)^ or (L m) 2(L-L)Fe 2(C0) 4 and the ligand L™ were dissolved in acetone (300 ml.) and the solution was irradiated with the 450 Watt uv lamp for 1-3 h , the reaction being monitored by i r techniques. The fi n a l product was purified as in Section A above. D. Reaction of (L m) 3(L-L)Fe 2(CO) With CO The complex (Ph^Sb)^fosFe^CO) (0.1 g) was stirred in For types (i) to (v) complexes, equimolar quantities of ligand and (L-L)Fe 2(CO) 6 were used, while for types (vi) to ( v i i i ) complexes, 2:1 or 3:1 molar ratios were employed. - 29 -benzene (50 ml.) at 60°, whilst CO was slowly bubbled into the solution. After l y days approximately one-half of the tris-substituted compound was converted into the bis-substituted (Ph^Sb^f^fosFe^CO^ as shown by the i r spectrum. Under the same conditions, the tris-substituted [(PhO) 3P] 3f 4AsPFe 2(CO) 3 was about 35% converted to i t s bis-substituted [(PhO) 3P] 2f 4AsPFe 2(CO) 4 as indicated by the i r data. 6. Reactions of (Ligand) b(L-L)Fe 2(CO) 4 with I 2 and HC1 A. Reaction of dpambf.AsPFeo(C0). with I„ 4 2 4 2 To a methylene chloride solution (25 ml.) of dpam bf 4AsPFe 2(C0) 4 (0.5 mmol.) was slowly added I 2 (0.5 mmol.) in methylene chloride (25 ml.) while s t i r r i n g . After complete addition of I 2 , the i n i t i a l l y orange-yellow solution had changed to deep red, and the i r spectrum showed new CO bands at -2060, -2050, -2020 and -2000 cm"1. After 2h at room temperature, the solution was evaporated to dryness to give a dark red residue. This was dissolved i n a minimum volume of CH2C12 and chromatographed on F l o r i s i l . Petroleum ether eluted a trace of dpam f 4AsPFe 2(CO)^. Diethyl ether eluted a red band which yielded a red o i l . Repeated chromatography and recrystallization failed to afford an isolable solid. The spectroscopic data of the oily substance were as follows : v C Q(CH 2Cl 2) : 2061 (6), 2048(8), 2018 (10), 1998 (9) cm"1. A (CH Cl ) : no strong bands characteristic - 30 -of a a* transition were observed between 300 and 700 my. Methylene chloride followed by acetone eluted a brown band which yielded a dark brown powder which did not show any CO bands. B. Reaction of f.AsP f.AsPFe„(CO), with HCl 4 4 2 4 Equimolar quantities (1.0 mmol.) of f^AsP bf^AsPFe^CO)^, HCl and acetone were sealed in a Carius tube. The tube was l e f t at 0°. The orange color solution had changed to deep red within — h. The tube was then opened and excess NaBPh^ in acetone was added. The deep red solution immediately changed color to pink and then to orange. The resulting solution was evaporated to dryness and extracted with diethyl ether. The combined ether extracts were evaporated to dryness and chromatographed on F l o r i s i l . fo Petroleum ether eluted a trace of f.AsP f.AsPFe„(CO),. A 4 4 2 4 petroleum ether - diethyl ether mixture (1 : 1) eluted an orange band which give a brown o i l . Attempts to obtain crystals were not successful. The brown o i l had v c 0(CH 2Cl 2) at 2063 (10), 2011 (10), 1960 (8, br.), 1903 (7, br.) cm ^ and did not show any strong bands between 300 and 700 my characteristic of a -> a* transition (CH2C12 solution). - 31 -III. Results and Discussion 1. (L-L)Fe 2(CO) 6 Complexes The new ligands fgfars and fgfars react with Fe(CO),. at 150° i n a Carius tube to give high yields of the (L-L)Fe 2(C0)g complexes I and II. These compounds have spectroscopic properties very similar to those of the reported (L-L)Fe 2(C0)g complexes [1] and are assumed to have analogous structure (Figure 1 ). The infrared carbonyl bands of I and II together with other (L-L)Fe 2(C0)g complexes are listed in Table IV. As expected in a l l (L-L)Fe 2(C0)g complexes there are six carbonyl stretching frequencies. Three of these, ~2061, ~2023 and -1983 cm \ are relatively insensitive to ligand substitution, the variation in frequency (Av ) being less than 7 cm ^ in a l l the compounds listed in Table IV, while the other three, at ~1997, -1966 and -1954 cm \ vary markedly with ligand substitution (Av = 12 to 22 cm "S . Provided no coupling occurs between carbonyl groups A B on Fe and those on Fe , then the three V ,^Q bands at -2061, -2023 -1 B and -1983 cm can be assigned to CO groups on Fe , while the other three bands probably arise from CO groups on Fe . The v of the (L-L)Fe 2(C0)g complexes increase i n the order (L-L) = f n f a r s < f nAsP < f^fos. This order is expected because phosphine is a better ir-acceptor than arsine [24]. The MOssbauer data for the (L-L)Fe 2(C0)g complexes are tabulated in Table V. The assignment of these data has been given elsewhere [1, 4, 21] TABLE IV. Infrared CO Bands of (L-L)Fe 2(C0) 6 Complexes Complex V CO ( C6 H12 ) -1 * cm Reference f 4farsFe 2(C0) 6 2059(s) 2022(vs) 1992(s) 1982(s) 1965 (w) 1950(m) [1] f 4AsPFe 2(CO) 6 2062(7) 2024(10) 1996(8) 1982(5) 1969(4) 1952(3) [4] f 4fosFe 2(CO) 6 2062(s) 2024(vs) 2001(s) 1984(m) 1971(m) 1965(m) [1] f f ifarsFe 2(CO) 6 2060(8) 2019(10) 1992(9) 1980(9) 1953(6) 1942(7) a f 6AsPFe 2(CO) 6 2062(8) 2026(10) 2000(8) 1986(7) 1973(4) 1955(4) [4] fgfosFe 2(CO) 6 2063(s) 2026(vs) 2004(m) 1985(m) 1973(m) 1968(m,sh) [1] f gfarsFe 2(CO) 6 2061(7) 2020(10) 1994(8) 1985(7) 1958(4) 1946(4) a Symbols or figures in parentheses refer to relative peak height, s = strong, m = medium, w = weak, sh = shoulder, v = very. a : This work. - 33 -TABLE V. Mossbauer Data At 80°K For (L-L)Fe 2(CO) 6 Complexes L-L 6(mm/sec)^ A(mm/sec) * r(mm/sec) Iron Site Ref. fgfars 0.32 1.44 0.25 B [1] 0.28 0.64 0.25 A f 4 f o s 0.32 0.23 ** +1.32 -0.66** 0.26 0.26 B A [1] f 6 f o s 0.32 1.19 0.23 B [1] 0.22 0.65 0.23 A f 4AsP 0.31 . , ** +1.45 ** -0.83 0.23 B [3,4] 0.27 0.23 A f 6AsP 0.32 1.19 0.24 B [3,4] 0.26 0.73 0.30 A fgfars 0.33 ' ,** +1.41 ** -0.67 0.28, 0.26 B [21] 0.30 0.24, 0.28 A fgfars 0.32 1.38 0.26, 0.28 B [21] 0.28 0.62 0.28, 0.28 A * Experimental uncertainty +0.01 mm/sec. t Relative to sodium nitroprusside , experimental error +0.01 mm/ sec. ** Sign determined with source and absorber at 4. 2°K [21]. The isomer s h i f t (6) measures the total s-electron density at the iron nucleus. The quadrupole splitting (A) arises from the interaction between the electric f i e l d gradient at the iron nucleus and the nuclear quadrupole moment of the 14.4 KeV level in 5 7Fe. The parameter r . i s the f u l l - 34 -width at half-maximum of the resonance line. As noted from this table the A 6 of Fe varies markedly as the ligands are changed from f nf° s t o f n f a r s , while that of Fe hardly varies. This result i s expected A because substitution of CO groups on Fe would have only a second B order effect on the parameters of Fe . In a l l the (L-L)Fe2(C0)g complexes B A studied, the 6 of Fe is greater than that of Fe . This indicates that A B the s-electron density on Fe i s greater than that on Fe . From the same A Table one can see that the 6 of Fe decreases in the order : f farsFe o(C0), > f AsPFe-(CO), > f fosFe o(C0), . n / o n z o n z o A This indicates that the s-electron density of Fe increases i n the order li s t e d . Two factors could be responsible for this change, namely (i) an increase of 4s electron density, or ( i i ) a decrease of 3d electron density (which would cause the contraction of the s-electron wave functions). However, i t i s not easy to differentiate the relative importance of these two factors. Some interesting trends in A values are also noted from Table V. A Fir s t of a l l , the magnitude of the A at Fe i n compounds of the type fnAsPFe2(C0)g is greater than that of the symmetrical derivatives f nfarsFe2(C0)g and f nfosFe2(C0)^. This type of behavior i s expected because the a-donor and ir-acceptor a b i l i t i e s of arsines and phosphines are different. Secondly, the A of Fe decreases as the ring size increases. This trend correlates well with the decrease of ring strain as the ring size increases [1]. However, changes in the hybridization at the C atoms B bonded to Fe may also affect the A i n such cases. - 35 -Previously [1] the (L-L)Fe2(C0)g type complexes (L-L = fgfars, f^fos and f^fos) were prepared by refluxing the respective ligand with Fe^CO)^ in cyclohexane. In this work i t was found that pyrolysis of a benzene solution of the ligand and excess Fe(CO),. at 150° i n a Carius tube gives f a i r l y high yield of the desired complexes. The new method has many advantages over the old one : f i r s t l y the starting material Fe(CO),. ($15/lb) is much cheaper than that of Fe^CO)^ ($45/100g); secondly, the isolation of the desired product i s easier; and thirdly the yield of the product i s , in general, higher. During the i n i t i a l studies on the preparation of the (L-L)Fe 2(CO) 6 complexes of f n f o s [1] and f n A s P [3], i t was found that the yields decrease with increasing of the ring size (compounds with n = 8 are not known). This was originally taken as an indication that as the ring became less strained there is less tendency for the double bond to coordi-B nate to Fe [4]. However, the good yields of fgfars and fgfars derivatives (70% and 45% respectively) suggest that ring strain i s not the only factor and that steric hindrance (possibly by bulky PPl^ groups) plays some part. A l l the above (L-L)Fe2(C0)g complexes react with various arsines and phosphines under uv irradiation to give a number of derivatives which can be classified into eight groups, as described below. - 36 -2. L m(L-L)Fe 2(CO) 5 (Type I) And (L-L) m(L-L)Fe 2(CO) 5 (Type II) Complexes The elemental analysis and molecular weight measurement of these compounds suggest the formula (Ligand) m(L-L)Fe 2(CO) 5. The Mossbauer data for this complex type are listed in Table VI. By comparing Tables V and VI, TABLE VI. MSssbauer Parameters At 80°K For (Ligand) m(L-L)Fe 2(CO) Complexes (Ligand) m(L-L) 6(mm/sec) A(mm/sec) r(mm/sec) Iron Site Ref. Ph.Pmf.AsP 3 4 0.30 1.38 0.27, 0.31 B [4] 0.35 0.56 0.27, 0.32 A dppp mf 4AsP 0.31 1.28 0.25, 0.39 B [21] 0.32 0.59 0.33, 0.35 A f. AsP mf . f ars 4 4 0.30 1.35 0.22, 0.26 B [21] 0.34 0.44 0.24, 0.26 A (PhO) 3P mf 4AsP 0.29 t +1.51T 0.31, 0.30 B [21] 0.29 -0.58+ 0.29, 0.32 A (PhO) 3P mf gfars 0.31 1.26 0.26, 0.25 B [21] 0.29 0.50 0.23, 0.26 A (Ph),Sb mf.fos 3 4 0.30 1.27 0.24, 0.23 B [21] 0.38 0.17 0.27, 0.28 A Relative to sodium nitroprusside. Experimental uncertainty, +0.02 mm/sec. t Sign determined with source and absorber at 4.2°K . - 37 -one can see that replacement of a CO group from a (L-L)Fe o(C0) £ to 2 6 in B give (ligand) (L-L)Fe 2(CO) 5 > the 6 of Fe hardly varies, but the 6 of Fe increases considerably. This implies that mono-substitution by A the ligand takes place on Fe . However, since there are three CO groups A * , * on Fe , two equatorial CO s, one axial CO, i t is of interest to know which has been replaced. In the (L-L) b(L-L)Fe 2(CO) 4 complexes (see below), where substitution of equatorial CO group is assumed to occur, the A of Fe is dramatically increased and i s positive in sign. However, the A of Fe i n the mono-substituted compounds remains small and negative (Table VI). Furthermore, a l l the fourteen (ligand) m(L-L)Fe 2(CO)complexes obtained show three bands in the carbonyl stretching region at -2037, -1 t -1981 and -1961 cm (Table III), suggesting a more symmetrical structure such as 7_. The highest frequency CO band at -2037 cm ^ is relatively insensitive to ligand substitution, the change in frequency (Av p n ) being -1 less than 10 cm ^ in a l l the compounds lis t e d , while the lower frequency bands at -1981 and ~1961 cm ^ vary markedly. Thus the CO band at -2037 cm ^ apparently The CO groups trans to the Fe-Fe bond are always referred to as the axial CO groups, and the ones cis to the Fe-Fe bond as equatorial; for both Fe^ and Fe^. t Since the solid state i r spectra are very similar to the corresponding solution spectra, i t i s thus apparent that the stereochemistry around each iron atom is retained in the solid state. - 38 -B arises from the CO group on Fe . The "Hi nmr spectra for some of the (ligand) m(L-L)Fe 2(C0) 5 complexes and their parent compounds are given in Table VII. TABLE VII. hi NMR Data for (Ligand) m(L-L)Fe 2(CO) 5 Complexes Compound 6 (CDC13) ppm. f 4farsFe 2(C0) 6 S : 1.50 (area 6) S : 2.21 (area 6) S : 1.60 (area 6) (PhO) P mf farsFe (CO) S : 2.07 (area 6) M : 7 ' 2 0 t ^ (area 15) M : 7.30 f 4AsP mf 4farsFe 2(C0) 5 S S S M M 1.01 (area 6) 1.03 (area 6) 2.35 (area 6) 7.35 7.40 (area 10) The compounds (PhO) 3P mf 4farsFe 2(CO) 5 and (f 4AsP) mf 4farsFe 2(CO) 5 show two and three singlets, respectively, i n the methyl proton region indicating that the axial CO group has been replaced. If the monodentate ligand (PhO)3P and f 4AsP m had replaced one of the equatorial CO groups, then the XB nmr spectrum of the resulting compounds would be more compli-cated (probably four and five absorptions, respectively, in the methyl - 39 -proton region). Thus, the data so far obtained a l l point to the symmetrical structure 7_. In the hi nmr spectrum of (f^AsP^f^farsFe 2(C0),_ one of the singlets at ca 1.0 ppm can be assigned to the two methyl groups on f^AsP. Since the chemical shift of this singlet i s not greater than that of the free ligand (1.33 ppm) i t is concluded that the ligand f^AsP i s A coordinated to the Fe via P atom (coordination via the As atom w i l l shift the As-Me peak downfield considerably). 3. (L-L) C(L-L)Fe 2(CO) 4 (Type III) Complexes Almost a l l the bifunctional arsines and phosphines investigated react with the (L-L)Fe 2(C0)g complexes to give dark brown complexes (L-L) (L--L)Fe2(C0)4. The crystal structure of a compound of this type c *!* f^AsP f^AsPFe^CO)^ has been established by an X-ray study [23]. This i s illustrated i n Fugure 2. In this molecule, Fe has approximately octa-hedral geometry, coordination being from the As and P atoms of each f^AsP ligand and a CO group, while Fe li e s in the sixth position. The geometry about Fe is similar to the parent compound (cf. Figure 1). The i r data of this complex type are listed in Table III. The majority of compounds show only three CO bands at -2007, -1946 and -1 c ~1912 cm . A few compounds (including f^AsP f 4AsPFe 2(CO)^ ) have four CO bands at -2007, ~1946, -1919 and -1907 cm"1. At f i r s t glance, one would conclude that the three band spectrum indicates the more symmetrical structure, j$, while the four band spectrum suggests the less symmetrical - 40 -- 41 -one such as Figure 2. From Figure 3, one can see that the pattern of these two types of spectra i s very similar i f one regards the two lowest frequency bands in (b) as due to the spli t t i n g of the lowest frequency stretching mode. The same pattern of CO bands in (a) and (b) implies that a l l these compounds probably have the same type of structure. 2030 (a) Figure 3. Two Types of Infrared CO Bands for (L-L)°(L-L)Fe2(CO)4 Complexes (b) 2000 1950 1900 1870 v c o ( a n " 1 ) - 42 -Since arsine and phosphine are known to be worse ir-acceptors than A CO, as the result of ligand replacement on Fe , the vr of CO on A Fe must be considerably lowered. The highest frequency CO band at -1 A -2007 cm is evidently not due to the unique CO on Fe ; i t therefore must arise from CO'(s) on Fe . The MSssbauer spectra [21] of most of these complexes show three component peaks of relative area 1 : 1 : 2 . A typical spectrum is i l l u s -trated in Figure 4 while the Mbssbauer parameters are tabulated in Table VIII. VIII. o oo CO CO < cr h-(f4AsP) ifarsFeJCOL 1 . 0 0 - f ^ % « ^ o « o .92H ° Of, o o -1 0 1 2 VELOCITY (mm/sec) Figure 4. Mo'ssbauer Spectrum of (f.AsP) cf ,farsFe„(C0), 4 o 2 4 - 43 -TABLE VIII. Mossbauer Parameters At 80°K For (L-L)°(L-L)Fe2(CO) Complexes (L-L)°(L-L) 6(mm/sec) A(mm/sec) ** T(mm/sec) Line Assignment Iron Site Ref. f 4AsP Cf 4AsP 0.28 1.07 0.28, 0.28 1,3 B [3] 0.50 0.61 0.27, 0.28 2,3 A 0.28 +1.05? 0.30, 0.30 1,3 B [21] 0.50 -0.61T 0.29, 0.30 2,3 A f 6AsP Cf 4AsP 0.29 1.07 0.27, 0.27 1.3 B [3,4] 0.49 0.67 0.27, 0.27 2,3 A f 6AsP Cf 6AsP 0.28 0.97 0.28, 0.33 1,3 B [3,4] 0.47 0.60 0.28, 0.33 2,3 A 0.29 0.96 0.23, 0.28 1,3 B [21] 0.47 0.59 0.27, 0.28 2,3 A f 4 A s P C f 6 f a r s 0.30 1.15 0.24, 0.29 1,3 B [21] 0.52 0.73 0.26, 0.29 2,3 A fgfars f 4AsP 0.27 0.91 0.31, 0.36 1,3 B [21] 0.53 0.40 0.26, 0.36 2,3 A fgfars Cfgfars 0.32 0.99 0.25, 0.28 1,3 B [21] 0.49 0.65 0.22, 0.28 2,3 A diars f 4AsP 0.28 0.86 0.29, 0.32 1,3 B [21] 0.50 0.45 0.32, 0.32 2,3 A dppm f 4AsP 0.24 0.80 0.25, 0.30 1,3 B [21] 0.53 0.51 0.22, 0.22 2,4 A f.AsP°fDfars 4 8 0.30 1.01 0.23, 0.26 1,3 B [211 0.46 0.64 0.24, 0.26 2,3 A diphos f 4AsP 0.27 1.08 0.34, 0.46 1,3 B [21] 0.51 0.59 0.36, 0.46 2,3 A fgfos°f4AsP 0.28 1.06 0.21, 0.32 1,3 B [21] 0.49 0.64 0.27, 0.32 2,3 A f4fos°f4AsP 0.25 1.10 0.26, 0.36 1,3 B [21] 0.49 0.62 0.24, 0.36 2,3 A •fc Relative to sodium nitroprusside. Experimental uncertainty, +0.02 mm/sec. Sign determined with source and absorber at 4.2°K. - 44 -As can be seen, the Mb'ssbauer parameters of these complexes l i e in the range <5 = 0.46 - 0.53 mm/sec. and A = 0.4 - 0.67 mm/sec. for A B Fe , while for Fe , 6 = 0.25 - 0.32 mm/sec. and A = 0.80 - 1.15 mm/sec. It i s known that the CO groups, which are very effective ir-acceptors, can act as electron "sinks" ;Ln withdrawing electron density from the metal. When CO is replaced by a less effective -rr-acceptor, this electron density w i l l be localized on the metal atom. Since there i s only one carbonyl A A group l e f t on Fe a considerable increase in the isomer shift of Fe over that of the corresponding iron atom in the related (L-L)Fe2(C0)g and mono-substituted derivatives is expected. g The large decrease in the A at the Fe in these compounds relative to those in equivalent complexes of the type (L-L)Fe2(C0)g and (ligand) m(L-L)Fe2(C0),. is expected because A is positive for Fe^, and i f the z-axis is more or less in the direction of the Fe-Fe bond then any increase in electron density along the Fe-Fe bond w i l l tend to g decrease the A at Fe . 4. (L-L) b(L-L)Fe 2(C0) 4 (Type IV) Complexes Some of the bidentate ligands react with (L-L)Fe2(C0)g to give red complexes with the molecular formula (L-L)'(L-L)Fe2(C0) 4 as suggested by analytical data and molecular weight measurements. These complexes have i r CO bands at ~1985(vs), ~1947(vs), ~1921(m), ~1900(m) cm"1. Previously we have concluded that the highest v C Q in (L-L)Fe 2(C0)g complexes arises g from the carbonyl groups on Fe . We also note that substitutions of one or A. TCi two carbonyl groups from Fe to give L (L-L)Fe 9(C0) c. and - 45 -(L-L) (L-L)Fe 2(CO) 4 complexes lower the vC()(max) by 25 and 54 cm"1, respectively. However, substitution of two CO from (L-L)Fe2(CO)g to give (L-L) b(L-L)Fe 2(CO) 4 causes the v c o(max) to lower by ca. 74 cm - 1. A larger decrease in vC Q(max) in this complex type suggests that at least B one carbonyl group on Fe has been replaced. Thus two possible cases R need to be considered : (a) The ligand is chelated to Fe or (b) the B A ligand is bridging both Fe and Fe . The MBssbauer spectra of this class of compounds (Table IX ) are also very different from those of any other derivatives described above. .Most of the spectra consist of two broadened lines (e.g. Figure 5), although occasionally three or four lines can be resolved. Comparison of Tables V, VI, VIII and IX, leads to the conclusion that (i) the A of Fe in this class of compounds are much i T 4 T O S ; y o s r e 2 i u j / 4 - O O c O o f 9 0 ° c £ r o O P o co CO CO < cr -1 0 1 - 2 VELOCITY (mm/sec) Figure 5. MSssbauer Spectrum of f.fos f.fosFe^(CO), ^ 4 2 4 - 46 -TABLE IX. MSssbauer Parameters At 80°K For (L-L) b(L-L)Fe 2(CO) 4 Complexes (L-L) b(L-L) 6(mm/sec) A(mm/sec) T(mm/sec) Line Assignment Iron Site Ref. f.AsPbf.AsP 4 4 0.41 0.30 1.48 1.26 0.28, 0.26 0.28, 0.29 1,3 2,3 B A [21] 0.36 0.36 1.21 1.21 0.32, 0.45 0.32, 0.45 1,2 1,2 B A [3] f 4AsP bf 6AsP 0.35 0.35 1.30 1.30 0.34, 0.44 0.34, 0.44 1,2 1,2 B A [3] f 4AsP bf 4fos 0.39 0.32 1.48 1.35 0.29, 0.26 0.29, 0.30 1,3 1,2 B A [21] dppp bf 4AsP 0.39 0.31 1.84 1.28 0.23, 0.24 0.25, 0.25 1,4 2,3 B A [21] f 4 f a r s b f 4 A s P 0.38 0.38 1.09 1.09 0.28, 0.31 0.28, 0:31 1,2 1,2 B A [21] arphosbf.AsP 4 0.43 0.30 1.43 1.18 0.24, 0.23 0.24, 0.26 1,3 2,3 B A [21] b dPPP f^fos 0.42 0.29 1.80 1.54 0.25, 0.23 0.25, 0.25 1,3 1,2 B A [21] dpam f.AsP 4 0.38 0.38 1.22 1.22 0.39, 0.33 0.39, 0.33 1,2 1,2 B A [21] f 4 f o s b f 4 f O S f 4 f o s b f 4 f O S + + 0.34 0.34 0.38 0.30 1.31 1.31 +1.35 1.26 0.29, 0.38 0.29, 0.38 1,2 1,2 B A B A [21] [21] t t Relative to sodium nitroprusside. Experimental uncertainty : +0.02 mm/sec. From magnetic perturbation results. - 47 -A larger and opposite i n sign to those of Fe i n the other derivatives, ( i i ) the A of Fe shows much more variation in magnitude than i t does B i n other complexes. Since chelation at Fe should have only a minimal A effect on the parameters of Fe , the f i r s t conclusion above indicates that this possibility, (a), can be ruled out. The second conclusion substantiates the second type of structure (b) namely bridging between _ A B Fe and Fe . From a molecular model constructed using the known bond angles and bond lengths of f^farsFe 2(CO) g [2] and f 4AsP Cf^AsPFe^CO)^ [23], A B one can estimate that for a ligand (L-L) to bridge Fe and Fe at an axial-axial or axial-equatorial position, the bite of (L-L) should be o about 5.1 A. Since the maximum bites of the ligands studied here (except o dppp) are a l l smaller than 5.1 A (see Tables X and XI), bridging from either axial position to any equatorial position on the other iron atom, and from one axial position to another are not l i k e l y . Thus, there remain only four possible bridging structures - those between the various equato-r i a l positions : (a) 1-1', (b) 2-2', (c) 1-2', and (d) 2-1' (9). The latter two poss i b i l i t i e s are less l i k e l y . However, they cannot be discounted with certainty. Keeping in mind that Fe-Fe and Fe-P bond lengths are 2.87 and o 0 2.27 A respectively [2, 23] and assume that Fe-Fe-P angle = 178 (cf. Fe-Fe-C axial = 178° [23]), the minimum distance spanned by a diphosphine i f i t bridges from one axial position to any equatorial position on the other Fe is about 2.87 +2.27 =5.1 A . - 48 -9 The MSssbauer results of (L-L)Fe 2(C0)g, (ligand) m(L-L)Fe 2(CO) 5 and (L-L) (L-L)Fe 2(CO)^ complexes have revealed that the change in 6 value on substituting an arsenic atom should be somewhat greater than that on substituting a phosphorus. That i s , A6(As) > A6(P) on the same iron atom. On this basis, i t can be concluded from the Mossbauer data of b b (L-L) (L-L)Fe 2(CO) 4 compounds that for cases where (L-L) contains both As and P the <5 data lead to a formulation in which the phosphorus A B atom is bonded to Fe and the arsenic to Fe [21]. 5. (dppp) b(L-L)Fe 2(CO) 4 (Type V) Complexes The ditertiary phosphine dppp also reacts with (L-L)Fe2(CO)g to give three types of derivatives, i.e. dpppm(L-L)Fe2(CO),-, c b dppp (L-L)Fe 2(CO) 4 and dppp (L-L)Fe 2(CO) 4. The f i r s t two have similar structures to types II and III, respectively. The i r spectra of the violet dppp b(L-L)Fe 2(CO) 4 complexes show four CO bands with a pattern very similar to those of type IV complexes. However, the frequencies of the second and fourth bands (Figure 6(d), (e)) are considerably lower. Moreover, the electronic spectra of the dppp complexes show a strong band at 504-522 mp whose wavelength is much higher than the corresponding band of type IV compounds (at ~425mu). Thus, i t is believed that these complexes - 49 -must have a slightly different structure. Their Mossbauer spectra show three or four lines, and comparison of data with those of the parent A B (L-L)Fe^(CO)g complexes reveals that carbonyl groups from Fe and Fe have been replaced. Hence, a structure in which dppp is chelating at Fe can be ruled out. Since the color of these compounds i s very similar to those of cluster compounds such as (L-L)Fe.j(CO)g (L= fgfars, f^AsP) , i t was f i r s t thought that they might have a dimeric structure such as jLO (L-L = dppp). However, molecular weight measurements indicate a monomeric L L CO OC CO OC L L 10 1' 2 D N D'' formula. The possible structures remaining are then (a) bridging the two Fe atoms at axial-axial positions. (This is unlikely because the maximum bite of o dppp is only ca. 5.7 A, while the minimum bite required to bridge at o 3, 3 should be ca. 7.4 A) (b) bridging the two Fe atoms at axial-equatorial positions (at least two p o s s i b i l i t i e s ) , (c) bridging at 1 and 2' and (d) bridging at 2 and 1' positions. Unfortunately i t does not seem possible to differentiate these. 3* Fe Fe 3 11 - 50 -6. (L m)2(L-L)Fe 2(CO) 4 (Type VI) and (L m) 2(L-L)Fe 2(CO) 4 (Type VII) Complexes Some of the monodentate ligands are found to give bis-substituted (L m) 2(L-L)Fe 2(CO) 4 derivatives when large ligand to (L-L)Fe 2(C0)g mole ratios are used. The red complexes derived from Ph^As and Ph^Sb have i r CO bands very similar to those of (L-L) b(L-L)Fe 2(C0) 4, suggesting that the two L m ligands are bonded to two different atoms. Since there i s no connection between the two ligand molecules, any one of the nine possible orientations should be considered. However, the high similarity in both frequency and pattern of the CO bands between the (L m) 2(L-L)Fe 2(C0) 4 and their corresponding (L-L) b(L-L)Fe 2(CO) 4 compounds points to the same configuration in both systems. When excess (PhO)^P i s used, dark red bis-substituted deriva-tives of the formula [(PhO)^P] 2(L-L)Fe 2(C0) 4 result. However, unlike the former complexes, these compounds have i r CO bands resembling those of (L-L) (L-L)Fe 2(C0) 4. Hence they are assigned to a structure in which the two (PhO)3P ligands are coordinated to Fe A. - 51 -7. (LU1) (L-L)Fe (CO) (Type VIII) Complexes A l l the above mentioned (L m)2CL-L)Fe2(CO) 4 complexes react with excess ligand to give the tris-substituted (L m)^(L-L)Fe2(C0)^ compounds in low yield. Their i r spectra show three CO bands at ~1969(s), ~1942(vs) and ~1910(s) cm \ Since they are readily converted to their respective bis-substituted complexes by reaction with CO, a structure such as 12 in which two L m ligands replace two carbonyl groups from Fe^ and one L m replaces one carbonyl group from Fe^ can be proposed for them. Owing to the low yield and their in s t a b i l i t y in solution and solid state, no H^nmr and MHssbauer spectra or reliable microanalytical data are available. They are thus mainly characterised by means of i r spectroscopy. - 52 -8. Cleavage of Fe-Fe Bond by 1^ and HC1 The ligand bridged complex dpara.bf ^AsPFe2 (CO)^ reacts with 1^ to give a red o i l ( I I I ) whose infrared spectrum show four CO bands at 2061 (6), 2048 (8), 2018 (10) and 1998 (9) cm"1. Similarly, f 4AsP bf 4AsPFe 2(C0) 4 reacts with HC1 to yield a red product which reacts with NaBPh^ to afford a brown o i l (IV) with four CO bands at 2063 (10), 2011 (10), 1960 (8, br.) and 1903 (7, br.). The v (max) of ( I I I ) is considerably higher than that of the parent complex suggesting the addition of electron-withdrawing iodo group. The electronic spectra of ( I I I ) and (IV) do not show the intense a •+ a* bands indicating the cleavage of the Fe-Fe bond, but l i t t l e can be said about the structure of these derivatives on the basis of the information at hand. - 53 -IV. General Discussion 1. Preparative The results of the reactions of monodentate ligands with (L-L)Fe2(C0)g complexes show that the yi e l d and ease of formation of substitution compounds decreases in the order : mono-substituted > bis-substituted > tris-substituted. This is expected because phosphites, phosphines, arsines, and stibines are known to be worse ir-acceptors than carbon monoxide and as the result of ligand substitution, the strength of the remaining Fe-CO bonds w i l l be increased considerably [24], making further substitution more d i f f i c u l t . Another interesting point is that (PhO),jP replaces two carbonyl groups from the same iron atom (Fe ) while Ph^E (E = P, As, Sb) A B substitute at both Fe and Fe . The phosphite i s a worse cr-donor but a better n-acceptor than the Ph^E ligands and i t is tempting to rationa-liz e the difference in reactivity on this basis. However, the successful isolation of (L-L) (L-L)Fe2(C0)^ complexes of dppm, dppe, etc., which have the similar a-donor and ir-acceptor capacities as the Ph^E ligands, suggests that electronic effect is not the only factor for directing substitution. Since the size of the donor atom increases in the order P < As < Sb, the steric repulsion between the phenyl groups and the other ligands on the same Fe atom decreases in the order : Ph^P > Ph^As > Ph^Sb. However, the insertion of one oxygen between the Ph groups and P atom should tend to decrease the steric requirements of the resulting in. s ligand (PhO),jP, and may account for the formation of (L ) 2(l-L)Fe 2(CO) 4 - 54 -from (PhO)3P and not from the Ph 3E (E = P, As, Sb) ligands. Both dppm and dppe which have bulky Pl^P groups also give (L-L) (L-L)Fe2(C0) 4 compounds but this i s probably due to the chelate effect. The results of the reactions of bidentate ligands with (L-L)Fe2(C0)g (Table II) show that the four membered ring fluorocarbon-bridged ligands fgfars, f^AsP and f^fos can form (L-L) b(L-L)Fe2(C0)^ compounds, while their higher homologs fgfars, fgAsP, fgfos, fgfars and fgf° s give (L-L) (L-L)Fe2(C0)^ exclusively. This result i s not unexpected because the bite of these ligands (Table X) decreases in the order l i s t e d . As the bite decreases, i t s chelating a b i l i t y increases but bridging power decreases. In particular fgfars seems very reluctant to form chelate complexes [25]. However, i f this effect i s the only factor affecting the formation of products, then since the bite of dpam i s smaller than that of fgfars, f Q f a r s , etc., (cf. Tables X and X I ) , dpam would be expected to give exclusively o (L-L) (L-L)Fe2(CO)4 complexes. From Table II one can see that dpam forms exclusively (L-L) b(L-L)Fe2(CO) 4 complex, thus bite alone cannot explain this result. However, since the electron-withdrawing groups can enhance the ir-accepting a b i l i t y of the fluorocarbon-bridged ligands, i t i s probable that the chelating a b i l i t i e s of f,fars, f_fars, etc., are o o greater than dpam. - 55 -TABLE X. Bites of the Fluorocarbon-bridged Ligands Ligand Calculated Normal B i t e * Experimental Bridging B i t e Chelating B i t e f . f a r s 4 4.00 °A 4.02 °A [29] f ,AsP 4 3.89 °A 3.28 °A [30] f. fos 4 3.83 °A f , f a r s 6 3.43 °A f ,AsP 6 3.41 °A f , f OS 6 3.35 °A 3.11 °A [31] dab 3.29 °A 3.18 °A [32] diars 3.27 °A 3.12 °A [33] fgfars 3.23 °A f 8 f o s 3.13 °A * The normal b i t e of these ligands i s defined as the distance between two donor atoms D, E i when each CD bond i s located on the b i s e c t o r the exocyclic C ^ angle (see f i g ;ure) . C ^  ^ . C "c = C ' § Calculated by means of the known bond lengths [1, 23, 29-33] : C-P = 1.84 A, C-As = 1.94 A, C=C : 1.30 A, and assuming that the exocyclic angles of these ligands are : 270° f o r the four membered, 252° f or the f i v e membered and 240° f o r the s i x membered r i n g compounds and dab and d i a r s . - 56 -TABLE XI. Bites of Some Hydrocarbon-Based Ligands Ligand Calculated Normal Bite Experimental Bridging Bite Chelating Bite dppm dpam 2.99 °A 3.15 °A 2.65 °A [37] 2.79 °A [38] 2.88 °A [36] dppe 3.38 °A 3.05 A [34] arphos dpae dppp 3-45 ° A 3.50 °A 5-70 °A (max. bite)' 3.14 °A [39] § The normal bites of dppm and dpam are calculated on the assumption that C-C =1.54 °A, C-P = 1.84 °A, C-As = 1.94 °A and natural tetra-hedral angle = 109.5°. For dppe, arphos and dpae, the ligand molecule is assumed to be arranged in a cis configuration with the two donor atoms in a staggered conformation. The maximum bite of dppp can be calculated from the data in the f i g u r e - 57 -2. Infrared Spectra The i r spectra of some representative iron carbonyl complexes are illustrated in Figure 6, while their maximum v_ are list e d in Table XII. By comparing the v (max) of these compounds, one can see that the highest energy CO bands for (L-L)Fe o(C0), should be due to the CO z o B group(s) on Fe , a result also expected from consideration of the ligand A A on Fe . Thus replacement of one CO on Fe of (L-L)Fe2(CO)g to give (ligand) (L-L)Fe 2(C0) 5 > lowers v c o(max) by -25 cm . Further replace-A -1 ment of another CO from Fe , decreases v^Q(max) by another 27 cm However, on substitution of one CO from , Fe B of (ligand) m(L-L)Fe 2(CO)^ to give (L m) 2(L-L)Fe 2(CO) 4 or (L-L) b(L-L)Fe 2(CO) 4, the vCQ(max) -1 A decreases by ~52 cm . Further substitution of one CO on Fe of this compound to give the tris-substituted (L m) 3(L-L)Fe 2(CO) 3 results in a decrease of vr,n(max) by 17 cm \ TABLE XII. Maximum v C Q of Some Representative Iron Carbonyl Complexes Complex v c o(max) cm 1 ( C H ^ i p (L-L)Fe 2(CO) 6 ~ 2062 L m(L-L)Fe 2(CO) 5 ~ 2037 (L-L) C(L-L)Fe 0(CO). ~ 2010 (L-L) b(L-L)Fe 2(CO) 4 ~ 1985 dppp b(L-L)Fe 2(CO) 4 ~ 1985 (Lm)_(L-L)Fe„(C0)o ~ 1968 - 58 -2000 1950 1900 1870 IR Spectra of Representative Iron Carbonyl Complexes (a = (L-L)Fe 2(CO) 6; b = (ligand) m(L-L)Fe 2(CO) 5; c = (L-L)°(L-L)Fe2(CO)4; d = (L-L) b(L-L)Fe 2(CO) 4; e = dppp b(L-L)Fe 2(CO) 4; f = (L m) 3(L-L)Fe 2(CO) 3 ). 59 -This conclusion is consistent with our previous deduction that the vCQ(max) of (L-L) Fe 2 (CO) g i s due to the CO group(s) on Fe . In summary a prediction can be made that (i) replacement of one B —1 CO from the Fe w i l l cause the v^(max) to decrease by -50 cm , A while ( i i ) replacement of one CO group from Fe w i l l decrease v^(max) by -25 cm 1 . This i s because (i) has a f i r s t order effect on the v^Q(max), while ( i i ) has a second order effect. 3. Electronic Spectra The electronic spectral data of the (L-L)Fe 2(C0)g complexes studied are shown in Table XIII. A l l the compounds li s t e d show two strong TABLE XIII. Electronic Spectra of (L-L)Fe 2(C0) 6 Complexes (L-L) A (CH0C1„) max 2 2 my(e - ) molar f.fars 4 382 (8235) 232 (12500) f .AsP 375 (8536) 229 (16500) f 4 f O B 361 (6482) 226 (9850) fgfars 382 (8549) 233 (25050) f ,AsP 6 375 (5952) 229 (11040) f 6 f o s 360 (5521) 226 (10500) fgfars 381 (7701) 232 (15000) emolar = m o l a r extinction coefficients. - 60 -bands in the region 200-400 my. Gray et. a l . [26, 27] have assigned the very intense band at 280 and 243.7 my in Cr(C0)g and Fe(CO),., respectively, as the lowest allowed M -> L charge transfer transitions. By analogy, the very strong band at -230 my of the compounds listed here can be assumed to be the same type of charge transfer band. The other strong band at -375 my is ca. 14000 cm 1 (131 my) lower in energy than would be expected for a M ->• L excitation. Furthermore, Gray et. a l . [35] have also made a rigorous assignment of the band at 345 my i n the electronic spectrum of Mi^CO)^ to a a -*• a* transition involving orbitals from the metal-metal bond. Since the strong band at ~375 my of the compounds studied here is very similar, both in energy and e m o ^ a r » to the o -+ o* transition band in M^CCO).^, i t can reasonably be assigned to the same type of transition. By analogy, the corresponding strong band of the derivatives of (L-L) Fe '(CO), (Table III) can be z o similarly assigned. From Table XIII one can see that A--,, of the o -+ a* band of the (L-L)Fe2(C0)g compounds decreases in the order : f farsFe„(C0), > f AsPFeo(C0), > f fosFe o(C0),. n z o n . z o n z o This suggests that the strength of Fe-Fe bond increases in the same order: f farsFe 0(CO), < f AsPFeo(C0), < f fosFe o(C0), n 2 6 n 2 6 n 2 6 which might be an indication that the better donors increase the electron A B density on Fe allowing i t to become a better donor in turn to Fe . However, a comparison of Tables III and XIII reveals that the apparent strength of the Fe-Fe bond decreases as the degree of substitution - 61 -increases (except dpppb (L-L) Fe 2 (CO) 4 ) : (L-L)Fe 2(CO) 6 > (ligand) m(L-L)Fe 2(CO) 5 > (L-L) u(L-L)Fe 2(CO) 4 ~ (L m) 2(L-L)Fe 2(CO) 4 > (L-L) b(L-L)Fe 2(CO) 4 ~ (L m) 2(L-L)Fe 2(CO) 4 > (L m) 3(L-L)Fe 2(CO) 3 ( > dppp b(L-L)Fe 2(CO) 4 ) It i s remarkable that in this criterion the di-substituted complexes dppp (L-L)Fe 2(CO) 4 show a very weak Fe-Fe bond. The reason for these results i s not clear. - 62 -References 1. W.R. Cullen, D.A. Harbourne, B.V. Liengme and J.R. Sams, Inorg. Chem., 8, 95 (1969). 2. F.W.B. Einstein and J. Trotter, J. Chem. Soc, (A), 824 (1967). 3. L.S. Chia, M.Sc. Thesis, University of British Columbia, (1971). 4. L.S. Chia, W.R. Cullen and D.A. Harbourne, Can. J. Chem., 50, 2182 (1972) . 5. J.P. Crow, W.R. Cullen, J.R. Sams and J.E.H. Ward, J. Organometal. Chem., 22, C29 (1970). 6. R.B. King, K.H. Pannel, CA. Eggers and L.W. Houk, Inorg. Chem., _7, 2353 (1968). 7. R.J. Haines, J. Organometal. Chem., 21, 181 (1970). 8. R.J. Haines and A.L. du Preez, Inorg. Chem., 8, 1459 (1969). 9. R.G. Hayter, J. Am. Chem. Soc, 85, 3120 (1963). 10. W.R. Cullen and R.G. Hayter, J. Am. Chem. Soc, 86, 1030 (1964). 11. M. Cooke, M. Green and D. Kirkpatrick, J. Chem. Soc. (A), 1507 (1968). 12. J.A. De Beer and R.J. Haines, J. Organometal. Chem., 36, 297 (1972). 13. J.A. de Beer, R.J. Haines, R. Greatrex and N.N. Greenwood, J. Chem. Soc. (A), 3271 (1971). 14. J.A. De Beer and R.J. Haines, J. Organometal. Chem., _3_7, 173 (1972). 15. J.P. Crow and W.R. Cullen, Can. J. Chem., 49, 2948 (1971). 16. R.J. Haines and CR. Nolte, J. Organometal. Chem., 36, 163 (1972). 17. W.R. Cullen, P.S. Dhaliw.al and G.E. Styan, J. Organometal. Chem., j$, 633 (1966). 18. W.R. Cullen, D.S. Dawson and P.S. Dhaliwal, Can. J. Chem., 45, 683 (1967). 19. W.R. Cullen, D.F. Dong and J.A.J. Thompson, Can. J. Chem., 47, 4671 (1969). - 63 -20. L.S. Chia and W.R. Cullen, Can. J. Chem., 50, 1421 (1972). 21. J.C. Scott, Ph.D. Thesis, University of B r i t i s h Columbia (1973) 22. R.D. Feltham and W. Si l v e r t h o r n , "inorg. Syntheses", V o l . 10, 159 (1967). 23. F.W.B. E i n s t e i n and R.D.G. Jones, Inorg. Chem., 12, 255 (1973). 24. E.W. Abel and F.G.A. Stone, Quart. Rev., 23, 325 (1969). 25. W.R. Cullen, Adv. i n Inorg. Chem. Radiochem., 15, 323 (1972). 26. N.A. Beach and H.B. Gray, J. Am. Chem. S o c , 90, 5713 (1968). 27. M. Dartiguenave, Y. Dartiguenave and H.B. Gray, 28. D.T. Thompson, B r i t i s h Patent, 1,096,404; CA. 68, p. 95985 r. 29. J.P. Crow, W.R. Cullen, F.L. Hou, L.Y.Y. Chan and F.W.B. E i n s t e i n , Chem. Comm., 1229 (1971). 30. F.W.B. Ei n s t e i n and J.D.G. Jones, Inorg. Chem., 12, 255 (1973). 31. F.W.B. E i n s t e i n and C.R.S.M. Hampton, Can. J. Chem., 49, 1901 (1971). 32. A. Mercer and J. Tro t t e r , Submitted for pu b l i c a t i o n . 33. D.S. Brown and G.W. Bushnell, Acta Cryst. , 22., 296 (1967). ' 34. J.A. McGinnety, N.C. Payne and J.A. Ibers, J. Am. Chem. S o c , 91, 6301 (1969). 35. R.A. Levenson, H.B. Gray, and G.P. Ceasar, J . Am. Chem. S o c , 92, 3653 (1970) . 36. M.G.B. Drew, J.C.S. Dalton, 626 (1972). 37. F.A. Cotlon, K.I. Hardcastle and G.A. Rusholme, J. Coord. Chem., _2, 217 (1973). 38. K.K. Cheung, T.F. L a i , and K.S. Mok, JCS(A), 1644 (1971). 39. G.J. Palenik, W.L. Steffen, M. Mathew, M. L i and D.W. Meek, Inorg. Nucl. Chem. Le t t e r s , 10, 125 (1974). - 64 -Chapter 2 Bicobalt Carbonyl Complexes I. Introduction 1. Review of the Reaction of Lewis Bases With Co„(C0) o 2 o The X-ray structural studies [1] have shown that i n the solid state Co2(CO)g consists of dimeric units with the two Co atoms bonded together through the two bridging CO groups (Figure 1 of Chapter 5). However, i r studies [2, 3] have revealed that in solution Co_(C0)o contains two tautomeric forms, one corresponding to the bridged solid state structure, the other a non-bridged form, v i z . (0C)4Co-Co(C0)^ . This tautomerism is related to a solvent and temperature dependent equilibrium [4]. If Co2(CO)g is sublimed on to a probe at liquid nitrogen temperature, paramagnetic Co(CO). i s formed [5]. However, i f Co o(C0) Q is heated to 60°, i t is converted to Co 4(CO> 1 2 [6]. (Under high CO pressure, Co2(C0)g i s stable [7] with respect to Co^CO)^ even at 200°). Dicobalt octacarbonyl reacts with various Lewis bases to give a large number of derivatives as described in the following sections. A. Substitution Derivatives With Bridging CO Groups (i) LCo 2(CO)^(CO) 2 - 65 -Poilblanc et. a l . [8] f i r s t reported the solution i r spectrum of LCo2(CO)^(CO)2 ^ L = E t3 P)» b u t t h ^ v f a i l e d t° isolate this compound. Carty et. a l . [9] have recently found that a reaction of DPPA with Co2(CO)g in benzene gave a red-brown solid. On the basis of i t s analytical and spectroscopic data, structure 1 was suggested. However confirmatory X-ray studies are needed. o OC Co Co CO «' V \ 9 rO c OC—•"Co"- - "Co—CO i ( i i ) L 2Co 2(CO)J(CO) 2 Although the existence of di-substltuted L 2Co 2(CO)^(CO) b (L = Et 3P [10], Et 3As [11], Me2PPh [12]) compounds was detected by i r spectroscopy several years ago, none of them were physically isolated. In 1970 Crow et. a l . [13] reported the f i r s t series of stable (L-L)Co-(CO), 2 6 complexes (L-L = fgfars, f 4 f o s , fgfos, f g f o s , fgAsP, and dab, etc.) which were prepared by stirring the ligand L-L with Co o(C0) D in 2 o - 66 -petroleum ether. The i r spectra of these complexes show the presence of bridging and terminal CO groups. However, unlike Co2(CO)g and L2Co2(CO)4(CO>2 (L = Et^As, Et 3P etc.), these complexes maintain their bridged form structure in solution and do not appear to exist as an equilibrium mixture of both bridged and non-bridged forms. In the fgfars, f^fos and fgAsI? derivatives the two bridging CO bands are separated by ~ 45 cm \ while in the f^fos, fgf° s a n <* dab derivatives, the separation between the two bridging CO bands is only ~ 15 cm 1 . Using the solid state structure of Co2(CO)g [1] as a model, three possible structures 2^  — k_ should be considered. Structure k_ contains a unique terminal CO group on the substituted Co atom which could be expected to have i t s v C Q at lower energy than other terminal CO groups. Such a band is not observed in the i r spectrum of any of the complexes and structure 4- was consequently ruled out. From 2 and 3 one can see that - 67 -the bridging CO groups i n 1_ are not located symmetrically with respect to the ligand, while i n 3^, they are both c i s to the ligand. Because of thi s energy d i f f e r e n c e between the asymmetric and symmetric bridging CO stretching bands w i l l be much la r g e r i n complexes with configuration 2. On this b a s i s , they assigned 2_ to the f i r s t class of complexes, and _3 to the second class of complexes. The f i r s t assignment has been confirmed by the determination of the s o l i d state structure of f . f a r s Co 0(CO), [14]. 4- z o The second assignment now appears to be in c o r r e c t (see discussion on page 120) . When equimolar quantities of Co o ( C 0 ) o and a diene (norbornadiene z o (C^Hg), 2,3-dimethylbuta-l,3-diene (C^H^) , cyclohexa-1,3-diene (C^Hg), isoprene (C^Hg) ) are refluxed i n pentane, compounds of formula (ir-diene) Co2(C0)g [15, 16] are found. An X-ray d i f f r a c t i o n study [17] has shown that (iT-C-,Hg)Co2(C0)g has a structure v i r t u a l l y i d e n t i c a l with that of Co2(C0)g [1] but with the two CO groups trans to the bridging CO groups replaced by the chelating, bidentate diene (see Figure 2 of Chapter 5). There are no i n d i c a t i o n s that the ( i T-diene) Co„(C0), complexes are tauto-z b meric as t h e i r i r spectra are independent of the solvent used [15]. This i s unexpected i n view of the behavior of Co2(C0)g [2,3], LCo 2(CO) 7 [8, 18], [LCo(CO) 3] 2 [11, 18], and [ ( i r-diene)Co (CO) 2 ] 2 [4]. t b ( i i i ) Higher Substituted L Co„(C0), (C0) o n 1 o—n / When a diene to Co o(C0)„ mole r a t i o of 20:1 i s used, the z o (7T-diene)Co2(C0)g complex f i r s t formed undergoes further CO replacement to give [(ir-diene)Co(C0) 2] 2 (ir-diene = C 7Hg, CgH^, CgHg and C^Hg) [4]. - 68 -These complexes are isoelectronic with [ (ir-C^H^)Fe(CO) ^ ] 2 , and, like the latter compound, also exhibit cis-trans isomerism in solution [19, 20]. Evidence from the solid state [4] and solution [4, 19] i r spectra indicate that [ (ir-CgH10)Co(C0) 2 ] 2 exists solely as the trans-species. Recent X-ray results [21] have confirmed this suggestion (Figure 1). In contrast, the i r spectrum of [ (ir-CgHgCo(CO) 2] 2 suggests that although i t exists as a mixture of c i s - and trans-isomers in solution, only the c i s -isomer is found in the solid state [4]. This result has also been confirmed by X-ray studies [22] (see Figure 3 of Chapter 5). Figure 1. Crystal Structure of [ (rr-C^^) Co(CO) £] [21] Behrens et. a l . [23] treated [ (TT-C 7H 8)CO(CO) 2] 2 with chelating, bidentate ligands such as bipy, phen, dppe and TP at 80° and isolated the [(L-L) Co(C0) 2] 2 complexes. They assigned 5_ and 6_ for these - 69 -compounds. cis-form, C„ trans-form, C„, zv Zn. 5 6 Recently Poilblanc et. a l . [24] have found that If tertiary phosphines (Me^, Et^P) and Co 2(CO) g (10:1 mole ratio) are refluxed in benzene, the orange to red tris-substituted (R 3P) 3Co 2(CO)^(CO) 2 (R = Me, Et) are formed. While the Me3P complex exists in solution mainly as the bridged form, the Et-^ P analog has more non-bridged form in the equilibrium mixture. However, both complexes seem to exist as bridged form i n the solid state. In the case of the reaction between Me„P and Co o(C0) o j Z o mentioned above, a tetra-substituted derivative L^Co^CO) 2(C0) 2 was i n i t i a l l y formed. This then decomposed to give, the tris-substituted compound. The same red compound (Me 3P) 4Co(CO) 2(C0) 2 has been obtained by heating [ (ir^Hg)Co(CO) ^] 2 and Me3P in benzene at 40°. - 70 -ler Capron-Cotigny [25] has also claimed the detection of a highe substituted derivative (Ph 3?) 6Co 2(C0) b but no discussion of the possible structure was presented. B. Substitution Derivatives Without Bridging CO groups (i) LCo 2(C0) ? The reaction of Ph3P with Co 2(C0) g in nujol gives a brown solid Ph 3PCo 2(CO) 7 [18]. Its i r spectrum shows the absence of bridging CO bands, and the presence of five terminal CO bands. On this basis, the following structure _7 has been assigned (L = Ph 3P). CO ( i i ) [LCo(CO) 3] 2 Many phosphites, tertiary phosphines, arsines and stibines react with Co 2(C0) g readily to form the non-bridged complexes [LCo(C0) 3] 2 [10-12, 18, 26-28] where the two Co atoms are held together by a Co-Co - 71 -bond as in 8 or 9_. Recent X-ray studies on [Bu"pCo(CO)3] [29, 30], [(PhO) 3PCo(CO) 3] 2 [31] and [ P h 3 P C o ( C 0 ) [ 3 2 ] have revealed that a l l the three compounds have the same type of geometry 8 . O Like Co 2(C0) g [2,3] and LCo 2(CO) 7 [8, 18], [LCo(CO) 3] 2 (L = R3P, R3As or R3Sb) [11, 18] also exist in solution as solvent-[3, 4, 11] and temperature-[2, 4] dependent equilibrium mixtures of bridged and non-bridged isomers. Although derivatives of the type [LCo(CO) 3] 2 could be expected for phosphines of the type R n p ^ 3 n (X = halogen, n = 0, 1, 2) stable products were not isolated [26, 33]. ( i i i ) L Co (CO)!: n z a—n When solutions of fgfars Co2(C0)g are heated an unstable compound of the formula f^farsCo,, (CO) ,- i s formed [13]. On the basis of - l i -the, i r data, the structure 10 similar to that of f,farsFe„(C0), [34] H Z D has been suggested. Perfluorocyclopentadiene reacts with Co 2(C0)g to give an orange compound of the formula (C,-F 6) 2Co 2(CO) 4 for which structure 11 has been proposed. - 73 -Capron-Cotigny [25] has recently claimed the detection of [(EtO) 3P]g C o2 which is one of the previously unknown class of derivatives L n C ° 2 ( C O ) 8 - n ( n ±- 3 )' (iv) Other Cobalt Carbonyl Complexes Tetraphenyldiphosphine reacts with Co2(CO)g to give a red compound which probably has the structure 12_ or 13_ [35]. A Ph Ph Ph ^ P h (OQ.Co Co(CO)3 V_ r-^f t P ^ N P h ( O C ) 3 C o / L _ _ A o ( C O ) 3 v 12 13 The reaction of (CF^PH with Co2(C0)g can lead to either the related product [Co(CO) 3P(CF 3) 2] 2 or a polymeric red solid, depending on the presence of solvent [36]. Tetramathyldiphosphine reacts in boiling benzene to give a dark green crystalline solid. A complete elemental analysis and molecular weight measurement suggest the unexpected formula (Me 2P) 2Co 3(CO) 7 [35]. Under the same conditions, tetramethyldiarsine gives an intense red solution - 74 -with evolution of gas [35]. The only product isolated was a red-jelly like material of unknown composition and structure. Hieber et. a l . [37] reported that ethyl mercaptan and thiophenol reacted with Co2(C0)g to give air-stable crystalline solids [Co(CO) 3SR] n (R = Et, Ph, and n being probably 2). Their solubility i n nonr-polar organic solvents suggests that they are presumably similar in structure to 12. Wilkinson et. a l . [28] found that solutions of L2Co(CO)3Co(CO)~ or [LCo(CO) 3] 2 (L = P(CH 2SiMe 3> 3 ) in CC14 or CHCl 3 evolve CO at room temperature. The major product i s the emerald green, diamagnetic complex [LCo(CO) 2Cl] 2 for which 14_ has been proposed. CO CO CO CO 14 Reaction of Co2(CO)g with AsCl 3 in THF gives a redsolution from which As 2Co 2(CO) 6 can be isolated at -20°. This compound reacts with Ph3P to yield a mono-substituted compound whose X-ray structure i s - 75 -illustrated in Figure 2 [32]. Figure 2. Crystal Structure of As 2Co 2(CO) 5PPh 3 [32] C. Disproportionation Derivatives (i) L Co(CO)t X" n 5-n Tertiary phosphines [10, 27, 28, 38], arsines [10] and stibines [10] react with Co 2(C0) g to produce ionic L2Co(CO)3Co(CO)~ (15) as well as substituted [LCo(CO) 3] 2-OC—-Co' . . .CD' ' 0 Co(CO)4 15 - 76 -Recently, Poilblanc et. a l . [39] have reported the synthesis of L^ C o C C O^BPh^ complexes (L = (MeO)3P, n = 3 — 5; L = R3P (R = Me, Et , n = 3). The i r and nmr data indicate stereochemical non-rigidity in the cations. Previously, i t has been established that isocyanides RNC (R = Me, Ph) [40, 41] react with Co 2(C0) g to yield L 5Co +Co(C0)~ (L = RNC) complexes. The reaction of polydentate phosphines and arsines with Co 0(CO) 0 have been reported [42, 43]. At room temperature, polydentate ligands such as dppe, dpae, arphos, cis Pf = Pf, Pf-Pf-Pf, P ( - P f ) 3 and Pf-Pf-Pf-Pf yield cationic substituted products and Co(CO), anion. The cations appear to be dinuclear species of general formula 2+ [Co^(CO)^(ligand) 3] (16) and mononuclear species of formula [Co(CO)(ligand) n] + (17). The complexes can be prepared by both thermal - 77 -and photochemical reactions, with the former route favoring 16_ and the latter route favoring 17. Very recently i t has been reported [44] that Co 0(CO) 0 reacts Z o with bidentate ligands R2E(CH2>nER2 (R 2E = Ph2P, (p-tol) 2P, n = 2; R2E = Me2As, n = 2 — 6, 8) and diars to give [(L-L)Co(CO) 2] 2 or (L-L)Co 2(CO)g complexes via the ionic intermediates [(L-L) 3Co 2(CO) 4][Co(CO) 4] 2 and [(L-L)Co(CO) 3][Co(CO)^ , respectively. In the [(L-L)Co(CO) 2] 2 complexes the ligands are believed to be chelating, while the structures of (L-L)Co„(CO), complexes could be 2, 3, or 4 Z o — — — depending on the values of n. ( i i ) L nCo 2 +[Co(CO)~] 2 Wender et. a l . [45] showed that independent of the basicity, the reaction between nitrogen and oxygen bases, L, and Co_(CO)Q gives rise z o 2+ -to disproportionation products, ^•n^ ° [Co(CO) 4] 2, where n i s 6 for monodentate ligand. These authors suggested that the f i r s t step of this reaction i s disproportionation to Co(I) and Co(-I) : Co_(CO)o + L > [Co(CO).L] + .Co(CO)7 • Z o H 4 This has recently been confirmed by a study of the reaction between Co o(C0) o and alcohols or water [46]. Z o D. Cluster Complexes (i) Sulfur-Containing Compounds At room temperature sulfur can react with Co o(C0) o i n 2 8 - 78 -hydrocarbon solvent to give sulfur-substituted Cobalt Carbonyl derivatives [47-49]. The isolated products are [Co 2(CO> 5S] n (n being probably 2), 18, Co3(CO)gS, 19, and Co 3(CO) 7S 2, 22/ These structures were proposed on the basis of i r data. (OCLCo-y - H K ° C C O L ( 0 ° 2 W ^ ( C 0 , 2 s 3 X Co (CO) y-o ( C O ) o / • 18 19 Carbon disulfide reacts with Co 2(CO) g [50] to give a range of products. A compound isolated in very low yield by chromatography has the formula (CS2) Co^CO) 1 Q . Based on i r data a butterfly structure, similar to that of (EtC = CEt)Co 4(C0) 1 ( ) (21) [51] has been suggested. When Klumpp et. a l . [52] reinvestigated the reaction of alkyl mercaptan (EtSH, PhCH0SH) with Co o(C0) o, they found that at least three types of cluster compounds ( a l l black crystalline solids) could be isolated, viz. (EtS) ?Co 4(CO) 5, (EtS) 3Co 4(CO) ? and (PhCR^S)SCo3(CO) . While the f i r s t two compounds have bridging CO groups, only terminal CO bands are observed in the i r spectrum of the third compound. Similarly, reaction of phenyl mercaptan with Co 0(CO) 0 gives an unexpected product SCo3(CO)g whose X-ray structure is identical to 19 with a triply bridged S atom and nine terminal CO groups [53]. ( i i ) Cullen's Cluster Compound When solutions of f 4farsCo2(C0)g are heated, in addition to the formation of f^farsCo2(C0),., 10, another new compound (f 4fars) 2Co 4(C0) 9(2H?) i s obtained in low yield [54]. The X-ray structural data reveal that this molecule contains four Co atoms and a bicyclobutenyl Figure 3. Crystal structure of (f.fars)„Co.(C0)Q(2H?) - 80 -system derived from two molecules of fgfars after displacement of one AsMe2 from each fgfars (Figure 3). In order to satisfy the EAN rule, i t is proposed that the sixth coordination site of each CO-bridged Co atom is occupied by a H atom. E. (ir-Alkyne) Co2(C0)g and Related Complexes In 1953 Wender et. a l . [55] discovered that many alkynes can react with Co2(C0)g to give (,!T-alkyne)Co2(CO)g complexes. Since then, this reaction has been the subject of extensive studies [56-74]. The i r spectra [55-74] of these compounds do not show CO bands attributable to bridging CO groups. An X-ray investigation [75] of (PhC = CPh)Co2(CO>6 has ascertained that i t s structure (see Figure 4 of Chapter 5) is derived from Co2(C0)g [1] simply by replacing i t s two bridging CO groups by an alkyne. Perfluoro-2-butyne also reacts with Co o(C0) o to form the similar 2 o type of compound [76], Octafluorocyclohexa-1,3-diene [73, 74] reacts with Co2(C0)g with elimination of fluorine to give a red complex (OC)„Co.C,F,.Co(CO). O D D J whose X-ray structure contains a perfluoro-cyclohex-l-yn-3-ene group, with the two Co atoms bonded to the triple bond of the unsaturated cyclic system [73]. The alkynylarsine (HC = C)-As reacts with Co,,(C0)o to yield j 2 o [Co2(CO)6(HC = C)] 3As. The X-ray results [77] show that the alkyne is coordinated but not the As atom. - 81 -The re a c t i o n between alkynylphosphines RR'PC = CR" and Co2(C0)g are more complicated. At l e a s t four types of complexes have been i s o l a t e d [9, 78] : ( i ) (RR'PC = CR")Co 2(CO) 6, ( i i ) (RR'PC = CR") 2Co 4(CO) ( i i i ) (RR'PC s CR")_Co o(C0),, and (iv) (RR'P(O)C = CR")Co 0(CO), . These complexes were characterized by a n a l y t i c a l and spectroscopic data. In types ( i ) and (iv) complexes, only the alkyne i s bonded to give the (ir-alkyne) Co 2(CO)g type structure 22. In type ( i i ) complexes, both the alkyne group and phosphorus atom are coordinated to form the dimeric structure 2^ 3, which has been confirmed by X-ray data for R = R' = Ph , R" = CF^ [79]. In type ( i i i ) complexes, only the phosphorus atom of each t b alkynylphosphine i s bonded to y i e l d the L 0Co 0(CO),(CO) 0 structure. 22 23 Hiibel et. a l . [61] reported that one or two CO groups i n (PhC = CPh)Co 2(CO)g can be replaced by Ph^P to give mono- and b i s -substituted d e r i v a t i v e s . However, the possible structures of these compounds were not considered. - 82 -Crow et. a l . [13] found that the ligands fgfars and f^fos can replace two CO groups from (PhC = CH)Co_(CO), to give black complexes (L-L)(PhC a CH)Co2(CO)4 (L-L = fgfars, f^fos). On the basis of spectroscopic data they assigned structure 24_ for these compounds (R = Ph, R' = H). Other ligands such as f^AsP, f^fos, fgf° s ^ d dab reacted with (PhC = CH)Co2(CO)g to yield only liquids or tarry solids which could not be purified [13]. 24 F. Complexes Isolated From High Temperature And High CO Pressure Reaction The reaction of Co 2(CO) g with (AsMe)5 in hexane at 200° under 200 atm. CO gives, in addition to intractable products, an un-expected compound As 3Co(CO) 3 [32]. The X-ray structure of this compound is illustrated in Figure 4. Figure 4. Crystal Structure of As„Co(CO)_ [32]. 2. Statement of Problems From the above review, i t is clear that the products of the reaction between Lewis bases and Co2(C0)g depend very much on the nature of the donor atom(s) of the ligand. Thus, nitrogen and oxygen bases favor the formation of ionic compounds. Phosphites, i n general, give substituted [LCo(C0).j]2 complexes. While phosphines, arsines and stibines can form both ionic L 2Co(CO) 3Co(CO) 4 and substituted [LCoCCO)^ complexes. The sulfur-containing ligands are unique in that they usually give rise to cluster compounds. In view of these facts, i t seemed interesting to investigate the reaction of Co2(C0)g with various unsymmetrical ligands f 4AsP, f 4PS, f^AsS, f 4AsO, f 4AsN, f 4AsF and f^AsCl. The fluorocarbon bridged ligands are chosen because previous studies [13, 54] have shown that fluorocarbon-bridged ligands (e.g. f 4 f a r s ) can not only stabilize the - 84 -resulting simple cobalt carbonyl complexes but can also react to afford novel compounds in some instances. Since the proposed structure 3 of the (L-L)Co„(CO),(B) — Z b complexes (L-L = fgfos, fgf° s a n& dab) is uncertain [13], the reaction between (^(CCOg and some other related new ligands fgfars and fgfars has been carried out i n hope of shedding more light on the problem. Although many [LCo(C0).j]2 complexes are now known, no bridged derivatives of the structure 25(a) or (b) have been isolated or detected. As this type of complexes could be of use as a homogeneous hydrogenation catalyst, i t was decided to try to synthesize examples. 25(a) 25(b) (where D E is a bidentate bridging ligand) A limited number of previous studies [13, 61] have revealed that one or two CO groups on (ir-alkyne)Co2(C0)g can be replaced by phosphines or arsines. However, the exact configuration of the resulting complexes was unknown. This prompted us to further explore this area. - 85 -II. Experimental The techniques employed are same as those described i n Chapter 1. The chemicals Co2(CO)g, FC = CF(CF 2) n (n = 2-4 ), Et^N, Me2NH, MeOH, and MeSH, the hydrocarbon-based ligands R^P (R = Bu11, Ph, OMe), dppm, dppe and dppp were obtained commercially and used without further purification. The starting materials Me2AsH [80] and Ph2PH [81] used for making other ligands were synthesized by the l i t e r a -ture methods. The fluorocarbon-bridged ligands fgfars, f^AsP, f^fos, fgAsP, fgfos, fgfos, dab, f^AsF, f^AsCl were prepared by the published procedures [34, 82-87]. The preparation of fgfars and fgfars has been described in Chapter 1. The f^SF required for making f^AsS and f^PS was synthesized from perfluorocyclobutene, MeSH and Et^N [88] . The other new ligands f^AsS, f^AsO^ f^AsN and f^PS were prepared by a general procedure outlined below. The preparative details are lis t e d in Table I. The reactants were sealed in an evacuated Carius tube and l e f t at the temperature indicated in column 2 of Table I for the time listed i n column 3. The tube was then opened and the volatile reactants removed in vacuo. The less volatile contents of the tube contained the desired product which was purified by the method shown in column 7. The analytical and nmr data of these ligands are tabulated in Table II. - 86 -TABLE I Preparative Details for the New Ligands Reactants Reaction Temp. (°C) Reaction Time Product Yield % b.p.(°C) mm. P u r i f i -cation 5 f^AsF (38.3m mol.) KOH (53.6m mol.) Me OH (50ml.) 0° 1 h. * f 4AsO 65 34°/0.1 Vd f 4SF (52.6m mol.) Me2AsH (141m mol.) 120° 4 d f 4AsS 70 57°/0.1 Vd f 4AsF (30.2m mol.) Me2NH (33.2m mol.) 25° 2 h f.AsN 4 67 67°/0.1 Vd f 4SF (23.2m mol.) Ph2PH (20.2m mol.) 25° 1 d f 4PS 55 mp. 48° c & r The crude product was washed with water, dilute hydrochloric acid and water, and dried with anhydrous MgS04 before d i s t i l l a t i o n . Vd = d i s t i l l a t i o n under vacuum c = chromatographed on F l o r i s i l using ethyl ether - petroleum ether mixture (1:20) as eluent r = recrystallized from petroleum ether. - 87 -TABLE II Analytical and NMR Data for the New Ligands Ligand Color Calc. Anal. Found hs. nmr(CDCl3)* f 4AsO colorless C 32.31 32.40 S : 1.30 (6) H 3.50 3.55 S : 4.00 (3) f.AsS pale C 30.43 30.32 S : 1.30 (6) yellow H 3.30 3.50 S : 2.50 (3) f ,AsN pale C 35.16 35.13 S : 1.35 (6) f yellow H 4.44 4.43 S : 3.05 (6) f 4PS white C 57.30 55.93 S : 2.40 (3) H 3.69 3.40 M : 7.40 (10) * Chemical . shifts are given in ppm down f i e l d from TMS. Numbers in parenthesis are the integral area. S = singlet, M = multiplet - 88 -1. Reaction of Co o(C0) o with Various Ligands Z o A. With fgfars and fgfars Dicobalt Octacarbonyl (0.95g, 2.78 mmol.) and fgfars (1.05g, 2.73 mmol.) were stirred in petroleum ether (50ml) at room temperature. After l-j h, reaction was complete as indicated by the i r spectrum. The reaction mixture was fi l t e r e d to remove a blue precipitate. The solvent was removed under reduced pressure and the red residue was dissolved in about 5ml. of methylene chloride and chromatographed on a F l o r i s i l column. A red band was eluted from the column with a mixture of petroleum ether - ethyl ether (20:1). The solution obtained was evaporated to dryness and the residual red solid was recrystallized from petroleum ether to give orange-red crystals of I, fgfarsCo 2(CO)g (l.Og, 59%), m.p. 105°(dec). Anal. Calc. for C 1 5H 1 2FgAs 2Co 20g : C, 26.86; H, 1.81. Found : C, 26.80; H, 1.84. v fC,H 1 0) : 2067 (10), 2004 (9), 1995 (10), CO o lz 1813 (5), 1800 (8) cm"1. % nmr(CDCl3) : singlet at 1.75 ppm. The related red solid II, f gfarsCo 2(CO)g, m.p. 90° (dec), was similarly prepared in 25% yield. Anal. Calc. for C^gH^2FgAs2Co20g : C, 26.7; H, 1.67. Found : C, 26.4; H, 1.59. v^(C,H 1 0) : 2068 (8), C O b I z 2005 (9, sh), 1997 (10), 1817 (3), 1802 (6) cm"1, hi nmr(CDCl3) : singlet at 1.80 ppm. B. With f.AsP 4 The reactants f^AsP (1.2g, 3.0 mmol.) and Co 2(C0) g (1.05g, 3.0 mmol.) were stirred in petroleum ether (100ml) for 2— h at - 89 -20°. The solvent was removed under reduced pressure and the residue was extracted with ethyl ether. The combined extracts were fi l t e r e d and evaporated to give a dark red o i l . This was dissolved i n 5ml. of methylene chloride and chromatographed on a F l o r i s i l column. An ethyl ether - petroleum ether mixture (1:20) eluted a brown-orange band which yielded brown-red crystals of f 4AsPCo 2(CO) 6 (0.25g, 12%), III, m.p. 92° Anal. Calc. for C^H^F^AsPCo^ : C, 41.14; H, 2.31; mol. wt. 700. Found : C, 41.23; H, 2.41; mol. wt. 714. v C 0(CH 2Cl 2) : 2068 (9), 2019 (9), 2000 (10), 1996 (10), 1823 (7), 1797 (8). hi nmr(CDCl3) : singlet at 1.40 ppm (area 6), multiplet at 7.35 ppm (area 10). The second band, violet-red, was eluted by 5-50% ethyl ether -95-50% petroleum ether to yield violet crystals of IV, [f 4AsPCo(C0) 3] 2 (0.90g, 30%), m.p. 117°. Anal. Calc. for C 4 2 H 3 2 F 8 A s 2 P 2 C o 2 ° 6 : C' 4 5 * 2 5 ' H, 2.90; mol. wt. 1114. Found : C, 45.00; H, 2.95; mol. wt. 1070. v C 0(CH 2Cl 2) : 1991 (3, sh), 1969 (10). hi nmr(CDCl3) : singlet at 1.10 ppm (area 12), and multiplet centered at 7.50 ppm (area 20). Ethyl ether followed by methylene chloride eluted the third band to yield an unstable orange powder V, [f 4AsPCo(C0) 3] 2 (0.25g, 8%), m.p. 110°. v C 0(CH 2Cl 2) : 2040 (0.5), 2004-1997 (10, br.). Anal. Calc. for C42 H32 F8 A s2 P2 C o2°6 : C ' 4 5* 2 45 H> 2 « 9 0 ; m o 1 - w t - 1114. Found : C, 44.85; H, 3.20; mol. wt. 888. -^H nmr(CDCl3) : singlet at 2.05 ppm (area 12), multiplet centered at 7.40 ppm (area 20). - 90 -C. With f.PS 4 Dicobalt Octacarbonyl (0.75g, 2.20m mol.) and f^PS (0.85g, 2.39m mol.) were stirred in petroleum ether (120ml) at room temperature for 3 h. The solvent was removed and the dark-brown residue was chromatographed on F l o r i s i l . Petroleum ether eluted a dark-brown band to give black crystals of VI, f 4PSCo 2(CO) 7 (0.2g, 14%), m.p. 105° . Anal. Calc. for C^H^F^PSCo^ : C, 42.99; H, 1.96; mol. wt. 670. Found : C, 42.77 H, 2.20 mol. wt. 597. v^CR^l^ : 2070 (9), 2025 (9), 2000 (10), 1968 (8) cm - 1 . ha. nmr(CDCl3) : singlet at 2.25 ppm (area 3), multiplet centered at 7.55 ppm (area 10). Ethyl ether followed by methylene chloride eluted a second band which afforded brown crystals of VII, [f 4PSCc(C0) 3] £ (0.50g, 23%), m.p. 152°. Anal. Calc. for C 4 0 H 2 6 F 8 P 2 S 2 C o 2 ° 6 : C y 4 8 , 1 0 ; H» 2 , 6 3 ; mol. wt. 998. Found : C, 48.21; H, 3.00; mol. wt. 950. v C ( )(CH 2Cl 2) : 1990 (4, sh), 1968 (10) cm"1, ""n nmr(CDCl3) : singlet at 2.40 ppm (area 6), multiplet centered at 7.50 ppm (area 20). D. With f^AsS and Related Unsymmetrical Ligands Dicobalt Octacarbonyl (0.90g, 2.63m mol.) and f^AsS (l.Og, 3.65m mol.) were stirred in petroleum ether (100 ml) for 1 h. at room temperature. After removal of the solvent, the orange residue was chromatographed on F l o r i s i l . Petroleum ether - ethyl ether mixture (4:1) eluted an orange band to give orange crystals which were recrystallized from chloroform - petroleum ether to yield orange needles of VIII, - 91 -[f 4AsSCo(CO) 3l 2 (0.70g, 50%), m.p. 70° ( d e c ) . Anal. C a l c for C20 H18 F8 A s2 S2 C°2°6 : C ' 2 8 , 6 4 ; H» 2 ' 1 7 ; A s » 1 7- 9 05 C o » 14.08; S, 7.64; Cl, 0.00; mol. wt. 838. Found : C, 28.70; H , 2.10; As, 18.10; Co, 14.08; S, 7.74; Cl, 0.00; mol. wt. 713. v C ( )(CH 2Cl 2) : 2066 (0.5), 2045(2.0), 1990 (10, br.) cm-1, hi nmr(CDCl3) : broad singlet at 2.10 (area 12), 19 two broad singlets at 2.50 and 2.70 (area 6) ppm. F nmr(CH 2Cl 2) : singlets at 108.0 (area 2), 108.4 (area 2) ppm. Mass spectrum : see Table III. Attempts to convert VIII into i t s BPh^ and PFg salts by the method of Hieber [10, 38] were unsuccessful. This compound is stable in the solid state for more than one Q year at ~0 , but i t decomposes in solution after a few min. The other related ligands f^AsF, f^AsCl, f^AsO and f^AsN reacted with Co2(C0)g to give the same type of products, but these are a l l unstable even in the solid state : (i) [f 4AsFCo(C0)^] 2, IX, orange-red crystals, yield 55%, m.p. 90°. v c (CH2C12) : 2055 (1), 1992 (10, br.) cm"1, hinmr(CDCl3) : broad singlet at 2.05 ppm. ( i i ) [f 4AsClCo(C0) 3] 2, X, dark-red crystals, yield ~40%, m.p. 106°. v (CH2C12) : 2070 (1), 1991 (10, br.) c m " 1 , 2H nmr(CDCl3) : broad singlet at 1.90 ppm. ( i i i ) [f 4AsOCo(CO) 3] 2, XI, orange crystals, yield 35%, m.p. 150°. v c (CH2C12) : 2078 (1), 1988 (10, br.) cm"1, hi nmr(CDCl3) : - 92 -TABLE III. Mass Spectrum of [f.AsSCo(CO)„]„ * Nominal Nominal mass Assignment Assignment 443 highest peak 181 f,AsS-SMe-2Me-F 4 400 f4AsSCo(C0)3-Me-4H 171 f.AsS-AsMe-4 2 386 f4AsSCo(C0)3-2Me-3H 161 f4AsS-SMe-2F-2Me 372 f 4As.SCo(C0) 2-Me-4H 153 f4AsS-4F-SMe or f4AsSCo(CO)3-SMe 152 f,AsS-AsMe„-F 4 2 344 f4AsSCo(CO)-Me-4H 142 f,AsS-3F-SMe-2Me 4 316 f.AsSCo-Me-4H 4 134 As Co 303 f4AsSCo-2Me-2H 133 f4AsS-AsMe2-2F 291 f4AsSCo-3Me-H 124 f.AsS-SMe-AsMe0 4 2 276 f.AsS 4 123 f,AsS-SMe-2Me-4F 4 261 f.AsS-Me 4 118 f,AsS-AsMe„-2F-Me 4 2 257 f.AsS-F 4 115 Co(C0) 2 246 f.AsS-2Me 4 114 f.AsS-AsMe.-3F 4 2 238 f,AsS-2F 4 105 AsMe2 231 f4AsS-3Me 90 As Me 229 f.AsS-SMe 4 87 Co(CO) 219 f,AsS-3F 4 80 AsC or L r J 210 f.AsS-SMe-F 4 75 As 200 f.AsS-4F 4 59 Co 195 f4As3-SMe-F-Me 47 SMe 182 f4AsSCo-SMe-4F-2Me 32 S 28 CO Main peaks and assignable peaks only. - 93 -broad singlet at 2.20 (area 12), broad singlet at 4.50 (area 6) ppm 1 . (iv) [f 4AsNCo(C0) 3J 2, XII, orange crystals, yield 42%, m.p. 100°. Anal. Calc. for c 22 H24 F8 A s2 N2 C o2 06 : C ' 3 1 , 7 3 5 H > 2 - 9 1 J N, 3.37; mol. wt. 836. Found : C, 32.04; H, 3.34; N, 3.54; mol. wt. 756. v C 0(CH 2Cl 2) : 2040 (2), 1993 (10, br.) cm"1. lE nmr(CDCl3): broad singlet at 2.00 (area 12), broad singlet at 3.10 (area 12) ppm. E. With dppm Equimolar quantities of dppm and Co2(C0)g (0.7g) were stirred in benzene (50ml) at room temperature for 3 h. The solvent was removed, the residue was dissolved in a minimum amount of methylene chloride and chromatographed on F l o r i s i l . The product was eluted with ethyl ether -petroleum ether (1:1). After removal of solvent, the orange powder was recrystallized from methylene chloride - petroleum ether to give orange crystals of XIII, dppmCo2(C0)6 (yield ~60%), m.p. 160°. Anal. Calc. for C31 H22 P2 C o2°6 *' C ' 5 5' 5 5> H» 3«31; mol. wt. 670 . Found : C, 55.00; H, 3.26; mol. wt. 613. v n 0(C 6H 1 2) : 2050(5), 2016(9), 1990(10), 1840(4), 1794(3). hi nmr(CDCl.) : triplet ( J L . , = 10 Hz) at 3.05 ppm. 3 r n - 94 -2. Reaction of f^fosCo 2(CO) g With f^fos The f 4fosCo 2(CO) 6 (A) [13] (0.5g, 0.64 mmol.), f^fos (0.5g, 1.01 mmol.) and acetone (25 ml) were sealed i n a Carius tube. The tube was irradiated with a 100 watt uv lamp at _ca. 20 cm. from the source while shaking. After 2 d. the tube was opened. The mixture was evaporated to dryness and then chromatographed on F l o r i s i l using methylene chloride as eluent to give a dark brown solution. After concentration and cooling dark brown crystals of XIV, f 4fosCo 2(C0) 6 (B) (0.2g, 40%), m.p. 115° were obtained. Anal. Calc. for C„.H0„F.P_Co00, : C, 52.31: 34 20 4 2 2 6 ' H, 2.60; mol. wt. 782. Found : C, 51.90; H, 2.69; mol. wt. 743. v C Q(CH 2Cl 2) : 2070 (9), 2018 (9.5), 1994 (10), 1815 (6), 1790 (7) cm"1. The mother liquor was evaporated and recrystallized from ethyl ether - petroleum ether mixture (1:4) to give black crystals of XV, (f/fos)^Co„(C0). (0.2g, 25%), m.p. 164°. Anal. Calc. for C 6 0H 4 0F gP 4Co 20 4 : C, 56.52; H, 3.17; mol. wt. 1218. Found : C, 55.05; H, 4.03; mol. wt. 1155. v^CK^Cl^) : 1972 (8), 1952 (10), 1765 (6) 2C1 2) : singlet at 107.8 ppm. and 1732 (9) cm"1. 1 9 F nmr(CH 9ClJ : singlet at 3. Reaction of (f 4fos) 2Co 2(C0) 4 With CO Complex XV (0.2 mmol.) was stirred i n benzene (50 ml.) at 50 for 2 h. whilst CO was slowly bubbled into the solution. The reaction mixture was evaporated to dryness and then chromatographed on F l o r i s i l . Petroleum ether followed by ethyl ether eluted a brown-red band. Concen-tration and cooling afforded dark brown crystals of XIV of known i r spectrum. - 95 -4. Reaction of [ (Tr-C^Hg) °Co (CO) ^] 2 The reactants f.fos (1.0 mmol.) and [ (TT-C,H0) Co (CO) „] _ [16] 4 l o l l (0.5 mmol.) were stirred i n benzene (50 ml.) at 60°. After 1/2 h the red solution had changed to dark brown. After 1 h the solution was evaporated to dryness to give a dark brown o i l . This was dissolved i n methylene chloride (ca.5 ml.) and chromatographed on F l o r i s i l . Petroleum ether eluted trace of [ (TT-^HQ)CO(CO) ^ ] ^. Ethyl ether eluted a dark brown band which gave dark brown crystals. These were recrystallized from methylene chloride - petroleum ether mixture (1:20) to give pure (f 4fos) 2Co 2(C0) 4, XVI (50%), m.p. 165°. Anal. Calc. for CggH^FgP^o^ : C, 56.52; H, 3.17; mol. wt. 1218. Found : C, 55.82; H, 4.00; mol. wt. 1165. v C ( )(CH 2Cl 2) : 1950 (10, br.), 1735 (6, br.) cm"1, hi nmr(CDCl3) : 19 multiplet centered at 7.40 ppm. F nmr(CH2Cl2) : singlet at 108.0 ppm. 5. Reaction of (f^fos) 2Co 2(C0) 4 With CO Complex XVI (0.2 mmol.) was stirred in benzene (50 ml.) at 50° for 2 h while CO was slowly bubbled into the solution. The deep red solution was evaporated to dryness and then chromatographed on F l o r i s i l . Petroleum ether followed by ethyl ether eluted a dark red band which yielded dark red crystals whose i r spectrum is identical with that of f 4fosCo 2(CO) 6 (A) [13]. - 96 -6. Reaction of ( i T-C^g) Co2(CO)g with f gf os, fgfos and dab Excess norbornadiene ( C7 Hg) (100 mmol.) and Co2(C0)g (10 mmol.) were stirred in petroleum ether (80 ml.) at room temperature. After 2 h, a red precipitate had formed. The reaction mixture was l e f t at this temperature for another hour and filtered to give a red residue and a dark brown f i l t r a t e . The red residue was dissolved in a minimum amount of ethyl ether and fil t e r e d to give a red solution which yielded red crystals after evaporation and recrystallization from petroleum ether. The melting point and i r CO bands of the red crystals were identical with those reported for (7r-C 7H g)Co 2(CO) 6 [22, 23]. The dark brown f i l t r a t e was concentrated and cooled to give red crystals, again identified as (TT-C 7H 8)CO 2(C0) 6. The total yield was ca. 80%. Equimolar quantities (10 mmol.) of (TT-C^H 0)CO„(C0) , and the I o 2 O ligand (f^fos, fgfos, or dab) were stirred in petroleum ether (50 ml.) at room temperature. After a few hours (ca. 3 h) most of the (TT-C_H0) Co„ (CO) , had reacted as indicated by the i r spectrum of the / o 2. 6 solution. The resulting mixture was filte r e d , evaporated and chromatogra-phed on F l o r i s i l . Petroleum ether - ethyl ether (2:1) eluted a brown band which afforded brown crystals whose i r spectra were identical with those reported for the (L-L)Co„(CO), (B) (L-L = f,fos, f Qfos or dab ) 2 O b o complexes [13] • - 97 -7. Reaction of [LCo(CO) 3] 2 With L-L A. Thermal Reaction (i) Reaction of [Bu^PCoCCO)3]2 with f^fos Equimolar quantities (2.0 mmol.) of f^fos and [Bu 3PCo(C0) 3] 2 were refluxed in benzene (50 ml.). After 2 h the reaction was complete. The resulting solution was filtered and evaporated to dryness to give a dark brown o i l . This was extracted with ethyl ether. The combined extracts were evaporated to dryness and then chromatographed on F l o r i s i l . Petroleum ether eluted a trace [Bu3PCo(CO)3]2» Ethyl ether - petroleum ether (up to 1:1) eluted a brown red band which give a dark red o i l . Further chromatography of the dark red o i l using ethyl ether - petroleum ether (1:10) as eluent and recrystallized from methylene chloride - petroleum ether (1;20) yielded dark brown crystals of XVII, f 4fos(Bu 3P)Co 2(C0) 5 (13%), m.p. 165°. Anal. Calc. for C ^ H ^ F ^ C o ^ : C, 57.46; H, 5.29; mol. wt. 954. Found : C, 56.74; H, 5.22; mol. wt. 820. v C ( )(CH 2Cl 2) : 2023 (10), 1967 (9), 1930 (9), 1794 (7) cm"1. "Si nmr(CDCl3) : multiplets with main bands at 1.15 and 1.75 (area 27), multiplet centered at 7.55 1 q (area 20) ppm. F nmr(CH2Cl2) :,singlets at 107.2 and 107.6 ppm of relative ratio 1:1 . Passing CO through a benzene solution of XVII for ca. 2 h at 50°, gave some intractable products plus f^fos and a red o i l whose i r and nmr spectra were identical with those of [Bu^PCo(CO)0]0 . - 98 -( i i ) Reaction of [(MeO) 3PCo(CO) 3] 2 and f^fos The cobalt complex (2.0 mmol.) and the ligand (4.0 mmol.) were refluxed in benzene (50 ml.) for 9 h. The solution was filtered and the solvent removed under reduced perssure. The red residue was dissolved in methylene chloride (5 ml.) and chromatographed on F l o r i s i l . Petroleum ether eluted a trace of [(MeO) 3PCo(CO) 3l 2. An ethyl ether -petroleum ether mixture (1:10) eluted a yellow band which gave yellow crystals identified as [ f ^ f o s c C o ( C 0 ) 2 XVIII (0.25 g, ~10%), m.p. 145°. Anal. Calc. for CggH^FgP^Co^ : C, 59.11; H, 3.32; mol. wt. 1218. Found : C, 59.15; H, 3.89; mol. wt. 1123. v^(c-C,H 1 0) : 1994 (8), CO o Li. 1942 (10) cm"1. 1H nmr(CDCl3) : multiplet centered at 7.50 ppm. 1 Q F nmr(CH2Cl2) : singlet at 107.7 ppm. A mixture of ethyl ether - petroleum ether (1:5) eluted a red band to give red crystals of [ (MeO) 3P] f 4 f o s C o 2 ( C 0 ) X I X (0.10g, -4%), m.p. 130°. Anal. Calc. for C 0,H o r iF,P 0Co 00 0 : C, 49.26; H, 3.34; mol. wt, 36 29 4 3 2 8 877. Found : C, 50.02; H, 3.25; mol. wt. 843. V C 0 ( C H 2 C 1 2 ^ : 2 0 2 0 1970 (8), 1944 (10), 1797 (6) cm"1. \ nmr(CDCl3) : doublet (J = 11 Hz) at 3.20 (area 9), multiplet centered at 7.60 (area 20) ppm. ( i i i ) Reaction of [Ph 3PCo(C0) 3] 2 with dppe Equimolar quantities (2.0 mmol.) of dppe and [Ph 3PCo(CO) 3] 2 were refluxed in THF (50ml.). After l | h the reaction was complete. The resulting solution was filtered and evaporated to give a yellow o i l . This was washed repeatedly with petroleum ether to remove excess dppe o - 99 -and then extracted with diethyl ether (3 x 25 ml.). The combined ether extracts were evaporated to _ca. 25 ml. and cooled in dry ice-acetone to yield yellow crystals of XX, (Ph 3P) 4(dppe)Co 2(C0) 4[Co(C0) 4] 2 (55%), m.p. 235°. v C 0(CH 2Cl 2) : 2012 (3), 1962 (7), 1887 (10) cm"1. hi nmr(CDCl3) : quartet (J = 7 Hz) at 3.50 (area 4), multiplets with strong peaks at 6.90, 7.10 and 7.40 (total area 80) ppm. The same compound, XX, was redissolved i n acetone and treated with excess NaBPh4- Addition of a few drops of water to the concentrated acetone solution initiated formation of yellow crystals of XXI, (Ph 3P) 4(dppe)Co 2(CO) 4(BPh 4) 2 (80%), m.p. 210°. Anal. Calc. for C146 H124 P6 B2 C°2°4 1 C > 7 5 , 7 1 ; H ' 5-89'> m o 1 - w t - 2314. Found : C, 75.05; H, 5.40; mol. wt. 2210 (osmometric i n acetone). v C ( )(CH 2Cl 2) : 2005 (9), 1953 (10) cm"1. "Hi nmr(CDCl3) : quartet (J = 7 Hz) at 3.55 (area 4), multiplets with strong peaks at 7.00, 7.20 and 7.45 (total area 120) ppm. B. Photolytic Reaction In the photochemical reaction, equimolar quantities (2.0 mmol.) of the reactants and the appropriate solvent (Table IV) were irradiated either with a 100 Watt or 450 Watt uv lamp. The resulting solution was f i l t e r e d and the solution was evaporated to dryness to give the crude products. Some of these were converted to tetraphenylborate by the method mentioned above. Unless otherwise stated, a l l the products were purified by recrystallized from the solvent listed in column 7 of Table IV. TABLE IV. Photochemical Reactions of [LCo(CO) ] with (L-L) Reactants Source Reaction Time Product Color? m.p. (°C) Recryst. from* Y i e l d (%) [Ph 3PCo ( c o ) 3 ] 2 dppe acetone 100 Watt 10 d [(Ph 3P) 2(dppe)Co(CO)] 2[Co(CO) 4] 2 (XXII) [ (P*i3P) 2 (dppe) Co (CO) ] 2(BPh 4) 2 (XXIII) y y 225 210 (1) (3) 50 40 [Bu 3PCo(CO) 3] 2 dppe acetone 100 Watt 3 d (Bu 3P) 4(dppe)Co 2(CO) 4[Co(CO) 4 ] 2 (XXIV) y 205 (1) 20 [3u^PCo(C0) 3] 2 dppp acetone 100 Watt (Bu^P) 4(dppp)Co 2(CO) 4[Co(CO) 4] 2 (XXV) (Bu^P) 4(dppp)Co 2(CO) 4(BPh 4) 2 (XXVI) y y 197 208 (1) (3) 25 18 [ 3 U 3 ¥ c o ( C 0 ) 3 ] 2 f , f a r s 4 acetone 100 Watt (Bu 3P) 2Co(CO) 3Co(CO) 4(CH 3) 2CO (XXVII) y 116 (1) 22 [Bu 3PCo(CO) 3] 2 f , f a r s 4 petroleum ether 450 Watt 15 h ( B u ^ P ) 2 f 4 f a r s C o 2 ( C O ) 4 (?) XXVIII d.b. — ** 10 y - yellow, d.b. - d a r k brown. (1) = d i e t h y l ether; (2) = methylene chloride/petroleum eth (3) = acetone ** Attempts to p u r i f y by r e c r y s t a l l i z a t i o n and chromatography f a i l e d to y i e l d i s o l a b l e s o l i d . - 101 -Thus were prepared the following (cf. Table IV) : (i) [(Ph 3P) 2(dppe)Co(CO)] 2[Co(CO) 4] 2, XX I I , v ^ C O ^ C i p : 1930 • cm"1; 1H nmr(CDCl3) : quartet at 3.40 (area 8), multiplets with strong peaks at 7.20 and 7.50 (total area 100) ppm. ( i i ) [(Ph 3P) 2(dppe)Co(CO)] 2(BPh 4) 2, XXI I I : Anal. Calc. for C174 H148 P8 B2 C o2°2 : C ' 78'61'> H» 5 , 6 3 5 m o 1 - wt. 2656. Found : C, 77.29; H, 6.00; mol. wt. 2512 (osmometric in acetone). v co^^2^ 12^ '' l 9 3 ^ c m ^' hi nmr(CDCl3) : quartet at 3.45 (area 8), multiplets with strong peaks at 7.05, 7.20 and 7.45 (total area 140) ppm. ( i i i ) (Bu 3P) 4(dppe)Co 2(C0) 4[Co(C0) 4] 2, XXIV : v C ( )(CH 2Cl 2) : 2007 (6), 1954 (9), 1890 (10) cm"1. (iv) (Bu 3P) 4(dppp)Co 2(CO) 4[Co(CO) 4] 2, XXV : v ^ C H ^ ) : 2000 (8), 1945 (10), 1890 (10) cm"1. (v) (Bu 3P) 4(dppp)Co 2(CO) 4(BPh 4) 2, XXVI : Anal. Calc. for C127 H174 P6 B2 C o2°4 : C ' 72-"> H» 9- 1 75 m o 1 - w t - 2088. Found : C, ; H, ; mol. wt. 1865 (osmometric in acetone) . v (CH.2CI2) : 2002 (7), 1947 (10) cm"1. (vi) (Bu 3P) 2Co(CO) 3Co(CO) 4(CH 3) 2CO, XXVII : Anal. Calc. for C34 H60 P2 C o2°8 ! C ' 5 1 * 8 0 ' H» 7 - 6 ° ; m o 1 - wt. 776. Found : C, 51.20; H, 7.70; mol. wt. 703. v C Q(CH 2Cl 2) : 1997 (9), 1889 (10) cm"1. hi D.mr(CDCl3) : broad singlets at 0.80 (area 18), 1.35 (area 36J and 2.00 (area 6) ppm. - 102 -(vii) f 4fars t >[(Bu 3 V)Co(CO) 2] 2(?), XXVIII; dark brown o i l , soluble in petroleum ether and diethyl ether. Attempts to purify by chromatography or recrystallization failed to yield crystals. v c o(CH 2Cl 2) : 1978 (7), 1958 (6), 1918 (10), 1902 (7) cm"1. *H nmr(CDCl3) : broad singlets at 0.70 (area 18), 1.30 (area 36) and 1.90 (area 12) ppm. - 103 -8. Reaction of (RC = CR)Co 2(C0) 6 With Monodentate Ligands A. Preparation of (MeO)3P(RC ='CR)Co2(C0)5 complexes (R = Ph, CH20H) Equimolar quantities (1.0 mmol) of (Me0).jP and the (RC = CR)Co2(CO)6 were stirred in benzene (50ml.) at 70° for ca 3 1 The solvent was removed under reduced pressure and the residual dark red o i l was dissolved in a minimum volume of methylene chloride and chromato-graphed on F l o r i s i l . The unreacted (RC = CR)Co2(C0)g was eluted with petroleum ether. The product was eluted with an ethyl ether — petroleum ether mixture (1:1) and recrystallized from the same solvent to give dark red crystals of the desired product (MeO)3P(RC'= CR)Co 2(C0) 5 : (i) (MeO)3P(PhC = CPh)Co2(CO)5, XXIX, m.p. 84°; yield 85%. Anal. Calc. for C22 H19 P C o2°8 : C ' 4 7' 1 45 H ' 3 ' 4 3 ? m o i ' w t " 5 6 0 ' Found C, 46.55; K, 3.76; mol. wt. 522. v C ( )(CH 2Cl 2) : 2064 (8), 2015 (10), 2002 (10), 1974 (7) cm"1. % nmr(CDCl-) : doublet ( J M » 11 Hz) at 3.17 (area 9), multiplets centered at 7.21 (area 6) and 7.83 (area 4) ppm. ( i i ) (MeO)3P(HOCH2C = CCH20H)Co2(C0)5, XXX, m.p. 8 5 ° ; yield 26%. Anal. Calc. for c 1 2 H i 5 P C o 2 ° 5 : C ' 3 0 - 7 7 5 H» 3.24; mol. wt. 468. Found : C, 30.76; H, 3.29; mol. wt. 444. v^CCR^l^ : 2068 (7), 2018 (10), 2007 (10), 1980 (6) cm"1. lH nmr(CDCl0) : tr i p l e t ( J O T = 6 Hz) J nn at 2.45 (area 2), doublet ( J p H = 11 Hz) at 3.70 (area 9), another doublet ( J H H = 6 Hz) at 4.75 (area 4) ppm. - 104 -B. Preparation of [ (MeO) 3P] 2(RC .= CR)Co 2(C0) 4 Complexes (R = Ph, CH20H) Trimethyl phosphite (2 mmol.) and the (RC = CR)Co2(C0)g (1 mmol.) or equimolar quantities (1.0 mmol.) of (Me0)3P and (MeO)3P(RC = CR)Co2(CO)5 were refluxed in benzene (50 ml.) for ca 2 h. The resulting mixture was filtered and evaporated to dryness. The residue was recrystallized from ethyl ether - petroleum ether (1:4) to give red crystals of [(MeO)3P]2(RC = CR)Co 2(C0) 4 : (i) [(MeO)3P]2(PhC = CPh)Co2(CO)4, XXXI, dark red crystals, m.p. 140°, yield 80%. Anal. Calc. for C24 H28 P2 C°2 04 : C ' 4 3' 9 05 H, 4.31; mol. wt. 656. Found : C, 45.02; H, 4.98; mol. wt. 615. v C 0(CH 2C.l 2) : 2026 (7), 1971 (10) cm"1, hi nmr(CDCl3) : tr i p l e t (J = 6 Hz) at 3.60 (area 18), multiplets centered at 7.55 (area 6) and 8.10 (area 4) ppm. ( i i ) [(Me0)3P]2(H0CH2C = CCH 20H)Co 2(C0) 4 > XXXII, orange red needles, m.p. 75°, yield 50%. Anal. Calc. for C ^ H ^ F ^ o ^ g : C, .29.79; H, 4.30; mol. wt. 564. Found : C, 28.80; H, 4.63; mol. wt. 575. v (CH2C12) : 2030 (7), 1973 (10) cm"1, h nmr(CDCl3) : tr i p l e t ( J R H = 6 Hz) at 2.40 (area 2), another t r i p l e t (J = 5.5 Hz) at 3.65 (area 18), doublet ( J ^ = 6 Hz) at 4.65 (area 4) ppm. C. Preparation of [(MeO)3P]3(RC = CR)Co 2(C0) 3. Complexes Trimethyl phosphite (3.0 mmol.) and (PhC = CPh)Co 2(C0) 6 (1.0 mmol.) or equimolar quantities (1,0 mmol.) of (MeO),P and - 105 -[(MeO)3PJ2(PhC = CPh)Co2(CO)4 were refluxed in toluene (25 ml.) for 2 h. After removal of the solvent, the residue was dissolved in a minimum quantity of methylene chloride and chromatographed on a F l o r i s i l column. The product was eluted with an ethyl ether - petroleum ether mixture (1:10). Upon evaporation, the dark violet residue was recrystallized from methylene chloride - petroleum ether to give dark red crystals of t(MeO)3P]3(PhC = CPh)Co2(CO)3, XXXIII, yield -30%, m.p. 165°. Anal. Calc. for C 2 6H 3 7P 3Co 20 1 2 : C, 41. 49; H, 4.97; mol. wt. 752. Found : C, 41.98; H, 5.40; mol. wt. 729. v C ( )(CH 2Cl 2) : 1997 (9), 1950 (10) cm"1. 1 * H nmr(CDCl3) : multiplets at 3.21 (area 9), 3.50 (area 18), 7.23 (area 6) and 7.93 (area 4) ppm. decoupling at 4Q.4933 Hz caused the two multiplets at 3.0-3.5 ppm to coalesce into two singlets : 3.06 (area 9) and 3.33 (area 18) ppm. Under the same conditions, (MeO)3P (3.0 mmol.) and (H0CH2C = CCH2OH)Co2(CO)6 (1.0 mmol.) or equimolar quantities (1.0 mmol.) of (Me0)3P and [(MeO)3P]2(H0CH2C = CCH20H)Co2(C0)4 reacted to give a red o i l with v C ( )(CH 2Cl 2) at 1998 (9) and 1953 (10) cm"1. Attempts to purify this compound further were unsuccessful. D. Preparation of [(MeO)3P]4(PhC = CPh)Co 2(C0) 2 Trimethyl phosphite (4.0 mmol.) and (PhC = CPh)Co2(C0)g (1.0 mmol.) or (MeO)3P and [(Me0)3P]3(PhC = CPh)Co 2(C0) 3 were refluxed in toluene (25 ml.) for 3 h. After removal of the solvent, the residue * — • Using Varian HA 100 nmr spectrometer. - 106 -was chromatographed on a F l o r i s i l column. The unreacted [(MeO)3P]3(PhC = CPh)Co2(CO)3 was eluted with an ethyl ether - petroleum ether mixture as in C above. The product was eluted with 20-100% ethyl ether - 80-0% petroleum ether to yield dark brown crystals. These were recrystallized from methylene chloride - petroleum ether to give dark brown needles of XXXIV, [(MeO)3P]4(PhC = CPh)Co 2(C0) 2 (yield ~20%) , m.p. 95°. Anal. Calc. for C28 H46 P4 C°2°2 : C ' 3 9 , 6 2 5 H> 5*48; mol. wt. 848. Found : C, 40.00; H, 5.42; mol. wt. 803. V C 0 ( C H 2 C 1 2 ^ : 1 9 9 6 ^ ' 1942 (8), 1922 (10) cm"1, hi nmr(CDCl3) : singlet broad band at 3.43 (area 36), multiplets centered at 7.21 (area 6) and 7.97 (area 4) ppm. 9. Reaction of L (RC = CR)Co„(C0), With CO n 2 6-n A. Reaction of [(Me0)3P]4(PhC = CPh)Co2(CO)2 With CO Complex XXXIV, [(MeO)3P]4(PhC = CPh)Co2(CO)2 (0.5 g), was stirred i n benzene (100 ml.) at 70° and CO was slowly bubbled into the dark-red solution. After 30 min. the carbonyl bands at 1942 and 1922 cm 1 disappeared and the CO bands at 1997 and 1950 cm 1 became very strong. The resulting solution was evaporated, chromatographed and recrystallized from ethyl ether - petroleum ether (1:10) to give dark red crystals (0.35 g, 80%) with i r and nmr spectra identical with those of complex XXXIII, [(MeO)3P]3(PhC = CPh)Co2(C0)3-B. Reaction of [(MeO) P] 3(PhC = CPh)Co2(CO)3 With CO Complex XXXIII (0.2 g) was stirred in benzene (50 ml.) at 70°. - 107 -After CO was bubbled through this solution for 1 h, a l l the complex XXXIII was converted to the di-substituted complex XXXI which was isolated and purified by recrystallization from petroleum ether. C. Reaction of [(Me0)3P]2(PhC E CPh)Co2(CO>4 With CO Under the same conditions, the di-substituted complex XXXI was only ca. 10% converted to the mono-substituted complex XXIX after 24 h. 10. Reaction of (PhC = CPh)Co.(CO), With Bidentate Ligands Z o A. Preparation of (L-L)(PhC = CPh)Co 2(C0) 4 In general, equimolar quantities (10 mmol.) of ligand L-L and (PhC = CPh)Co 2(C0) 6 were refluxed in petroleum ether (50 ml.) for 5 h. The solution was concentrated, chromatographed and recrystallized to yield (L-L)(PhC = CPh)Co 2(C0) 4 as black or dark green crystals. Thus were prepared the following : (i) fgfars(PhC = CPh)Co 2(C0) 4, XXXV, dark green crystals, m.p. 175°, yield 60%. Anal. Calc. for c 2 6 H 2 2 F 4 A s 2 C o 2 ° 4 : C' 4 1 - 9 8 ; H, 3.12; mol. wt. 742. Found : C, 42.05; H, 3.00; mol. wt. 660. VCO ( C6 H12 ) : 2 0 3 5 ( 8 ) ' 2 0 0 2 ( 1 0 ) ' 1 9 8 3 ( 9 ) ' 1 9 5 9 ( 5 ) c m - 1 - 1 h ™*(CDC1 3) : * singlet at 1.60 (area 12), multiplet centered at 7.37 (area 10) ppm. 19 F nmr(CH 2Cl 2) : singlet at 105.9 ppm. ( i i ) f4AsP(PhC = CPh)Co 2(C0) 4, XXXVI, dark green crystals, * In benzene this i s s t i l l a singlet at 2.00 ppm. - 108 -m.p. 150°, yield 20%. Anal. Calc. for C ^ H ^ F ^ s P C o ^ : C, 52.25; H, 3.44; mol. wt. 822. Found : C, 52.55; H, 3.19; mol. wt. 735. v C 0(c-C 6H 1 2) : 2033 (7), 2004 (10), 1981 (8), 1961 (2) cm"1. 1H nmr(CDCl3) : singlet at 1.45 (area 6), multiplets at 7.40 (area 10) and 7.50 (area 10) 19 ppm. F nmrCCE^C^) : singlets at 105.2 (area 2) and 106.4 (area 2) ppm. ( i i i ) f 4fos(PhC = CPh)Co 2(C0) 4, XXXVII, dark green crystals, m.p. 215°, yield 66%. Anal. Calc. for C.,HonF.PoCoo0. : C, 61.20; 46 30 4 2 2 4 H, 3.36; mol. wt. 902. Found : C, 60.90; H, 3.50; mol. wt. 839. v c o(c-C 6H 1 2) : 2038 (7), 2013 (10), 1987 (7), 1967 (2) cm"1, "^H nmr(CDCl3) : 19 multiplets at 7.40 (area 10) and 7.50 (area 10) ppm. F nmr(CH2Cl2) : singlet at 105.3 ppm. (iv) f 6fars(PhC = CPh)Co2(CO)4, XXXVII, dark green crystals, m.p. 180°, yield 50%. Anal. Calc. for C o^H o oF-As„Co o0. : C, 40.73; 21 22 o 2 2 H H, 2.95; mol. wt. 792. Found : C, 40.91; H, 2.81; mol. wt. 730. v C 0(C 6H 1 2) : 2032 (8), 2003 (10), 1980 (9), 1958 (5) cm"1. 1H nmr(CDCl3) : singlet at 1.60 (area 12), multiplet at 7.40 (area 10) ppm. 1 9 F nmr(CH 2Cl2) : singlets at 104.2 (area 4) and 130.7 (area 2) ppm. (v) f6AsP(PhC = CPh)Co2(CO)4, XXXIX, black crystals, m.p. 65°, yield 40%. Anal. Calc. for C^H^FgAsPCo^ : C, 52.06; H, 3.01; mol. wt. 872. Found : C, 51.50; H, 2.90; mol. wt. 833. v^(C,H._) : 2054 (7), 2034 (7), 2005 (10), 1983 (8), 1969 (4) cm"1, h. nmr(CDCl3) : singlets at 1.30 (area 2), 1.50 (area 1), 1.80 (area 1), multiplets at 19 7.30 (area 10) and 7.50 (area 10) ppm. F nnn^CH^C^) : quite complicated with strong bands at 103.1, 104.6, 128.4 and 131.1 ppm. - 109 -(vi) f 6fos(PhC = CPh)Co 2(C0) 4, XL, black crystals, m.p. 180°, yield 50%. Anal. Calc. for C 4 7H 3 ( )F 6P 2Co 20 4 : C, 60.44; H, 3.60; mol. wt. 952. Found : C, 59.24; H, 3.18; mol. wt. 905. v„ r t(C,H 1 0) : C O 0 Li -1 19 2056 (10), 2004 (9), 1982 (7), 1963 (2) cm . F nmr(CH 2Cl 2) : singlets at 109.4 (area 2), 110.5 (area 2) and 134.8 (area 2) ppm. (vii) f gfars(PhC = CPh)Co 2(C0) 4, XLI, black crystals, m.p. 63°, yield 66%. Anal. Calc. for C O QH 0-F 0As„Co 00, : C, 39.91; H, 2.75; 28 22 8 2 2 4 mol. wt. 842. Found : C, 39.90; H, 2.64; mol. wt. 785. v^(C,H 1 0) : > CO O LZ i 2054 (10), 2005 (10), 2001 (10), 1981 (8), 1973 (9), 1955 (9) cm - 1. nmr(CDCl3): singlets at 1.45 (area 6) and 1.77 (area 6), multiplets at 7.37 and 7.50 (area 10) ppm. 1 9 F nmr(CH 2Cl 2) : singlets at 103.4 (area 4) and 136.3 (area 4) ppm. ( v i i i ) fgfos(PhC = CPh)Co 2(C0) 4, XLII, black crystals, m.p. 176°, yield 75%. Anal. Calc. for C. oH o_F oP„Co„0. : C, 57.01; H, 3.12; mol. wt. 1002. Found : C, 57.60; H, 3.03; mol. w t . 896. v C 0(c-C 6H 1 2) : 2051 (10), 2004 (9), 1984 (8), 1964 (3) cm"1. 19 F nmr(CII2CI2) • singlets at 101.5 (area 4) and 133.5 (area 4; ppm. (ix) (dab)(PhC 5 CPh)Co2(CO)4, XLIII, black crystals, m.p. 80°, yield -60%. Anal. Calc. for C26 H22 F6 A s2 C°2°4 : C ' 4 7 ° 1 3 5 H> 3- 3 65 mol. wt. 780. Found : C, 46.50; H, 3.48; mol. wt. 729. V C 0^ C" C6 H12^ : 2052 (10), 2005 (10), 1999 (10), 1972 (9), 1957 (9) cm"1. \ nmr(CDCl3) : singlets at 1.45 (area 6) and 1.80 (area 6); multiplets at 7.35 and 7.60 19 (area 10) ppm. F nmr(CDCl3) : singlet at 50.5 ppm. - 110 -(x) dppm(PhC = CPh)Co2(CO)4, XLIV, dark brown crystals, m.p. 225°; yield 80%. Anal. Calc. for C43 H32 P2 C o2°4 : C ' 6 5 - 1 7 ; H, 4.07; mol. wt. 792. Found : C, 63.91; H, 4.32; mol. wt. 773. v C 0(c-C 6H 1 2) : 2029 (8), 1993 (10), 1980 (9), 1960 (3) cm"1. *H nmr(CDCl3): t r i p l e t ( J p R = 10 Hz) at 3.20 (area 2), multiplets centered at 7.40 (area 10) and 7.50 (area 20) ppm. (xi) dppe(PhC = CPh)Co2(CO)4, XLV, dark green crystals, m.p. 165°; yield -10%. Anal. Calc. for C ^ H ^ P ^ o ^ : C, 65.51; H, 4.26; mol. wt. 806. Found : C, 64.98; H, 4.51; mol. wt. 812. v C 0(c-C 6H 1 2) : 2031 (8), 1996 (10), 1975 (9.5), 1954 (4) cm"1. hi nmr(CDCl3): multiplet at 2.5-3.5 (area 4), multiplet centered at 7.40 (area 10) and 7.55 (area 20) ppm. (xxi) dppp(PhC = CPh)Co 2(C0) 4, XLVI, dark brown crystals, m.p. 140°; yield 20%. Anal. Calc. for C 4 5 H 3 6 P 2 C o 2 ° 4 : c> 65.85; H, 4.43; mol. wt. 820. Found : C, 64.42; H, 4.63; mol. wt. 798. v C ( )(CH 2Cl 2) : 2036 (8), 1975 (10) cm"1, -hi nmr(CDCl3) : multiplet at 2.4-3.4 (area 6), multiplet centered at 7.35 (area 10) and 7.50 (area 20) ppm. B. Preparation of (L-L) 2 (PhC = CPh)'Co2(C0)2 The (L-L)(PhC = CPh)Co2(CO)4 complex (1.0 mmol.) and the ligand (L-L) (1.3 mmol.) i n 300 ml. of benzene were irradiated with the 450 watts uv lamp for 8 to 12 h. The reaction was monitored by observing the disappearance of the (L-L)(PhC = CPh)Co2(C0)^- The f i n a l solution was dark brown or black. The benzene was removed under reduced - I l l -pressure and the resulting o i l chromatographed and recrystallized to yield (L-L) 2(PhC = CPh)Co2(CO)2 as dark brown or black crystals. (i) (f 4fars) 2(PhC E CPh)Co2(CO)2, XLVTI, black crystals, m.p. 130°; yield -10%. Anal. Calc. for C ^ H ^ F g A s ^ o ^ : C, 37.65; H, 3.37; mol. wt. 1020. Found : C, 37.93; H, 3.52; mol. wt. 1053. v C 0(c-C 6H 1 2) : 1951 (7), 1944 (10); (CH^C^) : 1930 (10, br.) cm"1. nmr(CDCl3): singlets at 1.30 (area 6) and 1.50 (area 6), multiplet at 7.20 (area 10) ppm. 1 9 F nmr(CH2Cl2) : singlet at 107.2 ppm. ( i i ) (dab)2(PhC = CPh)Co2(CO)2, XLVIII, dark brown crystals, m.p. 190°; yield -15%. Anal. Calc. for C32 H34 Fi2 A s4 C°2°2 : C» 3 5 , 4 5 > H, 3.14; mol. wt. 1096. Found : C, 35.04; fl, 3.13; mol. wt. 1131. v C 0(CH 2Cl 2) : 1922 (10), 1898 (10) cm"1. "^H nmr(CDCl3) : singlets at I. 45 (area 6) and 1.80 (area 6), multiplets at 7.35 and 7.60 (area 10) ppm. 19 F nmr(CH 2Cl 0): singlet at 51.3 ppm. The above two (L-L) 2(PhC = CPh)Co 2(C0) 2 complexes could also be prepared by the thermal method described next. The (L-L)(PhC E CPh)Co 2(C0) 4 (1.0 mmol.) and the ligand L-L (1.1 mmol.) were refluxed in toluene (25 ml.) for ~£0 h. The resulting dark brown solution was evaporated to dryness. The residue was chromato-graphed and recrystallized from a methylene chloride - petroleum ether mixture (1:20) to yield the desired product, e.g. ( i i i ) (f 4fos) 2(PhC E CPh)Co2(CO)2, XLIX, dark brown crystals, m.p. 195°; yield -40%. Anal. Calc. for c 72 H50 F8 P4 C o2°2 " C > 6 4 > 6 7 5 - 112 -I I , 3.78; mol. wt. 1340. Found : C, 65.24; H, 4.77; mol. wt. 1384. v C 0(c-C 6H 1 2) : 1928 (10), 1913 (10) cm"1. 1 9 F nmr(CDCl3) : singlet at 108.0 ppm. 11. Reaction of (L-L) 2(PhC = CPh)Co 2(C0) 2 With CO Complex XLIX, (f 4fos) 2(PhC = CPh)Co2(CO)2 (0.5 g) was stirred in hexane (50 ml.) at 50° and CO was passing through the solution. After 6 h, ca. 60% of XLIX had been converted to complex XXXVII, f 4fos(PhC = CPh)Co 2(C0) 4, as indicated by the i r spectrum. 12. Reaction of (PhC = CH)Co 2(C0) 6 With Bidentate Ligands Equimolar quantities (1.0 mmol.) of (PhC = CH)Co„(C0), z o (prepared i n situ) [56] and the ligand L-L were refluxed i n benzene (50 ml.) for ca. 2 hours . The resulting solution was concentrated, chromatographed and recrystallized to give dark green or black crystals of (L-L)(PhC = CH)Co 2(C0) 4. The f 4fars(PhC E CH)Co 2(C0) 4 and f 4fos(PhC = CH)Co 2(C0) 4 had been prepared previously [13]. A.. (f 4AsP)(PhC = CH)Co 2(C0) 4, L, dark green crystals, m.p. 135°, yield -15%. Anal. Calc. for C ^ l ^ ^ A s P C o ^ : C, 48.26; H, 2.98; mol. wt. 746. Found : C, 47.85; H, 3.04; mol. wt. 709. v^C^K^ : 2037 (8), 2006 (10), 1982 (9), 1962 (3) cm"1, hi nmr(CDCl3) : singlets at 1.50 (area 3), 1.75 (area 3), 5.30 (area 0.5) and 5.60 (area 0.5), multiplets at 7.40 (area 5) and 7.50 (area 10) ppm. - 113 -B. (fgfars)(PhC = CH)Co2(CO)4, LI, black crystals, m.p. 110°, yield 50%. Anal. Calc. for C-.H., oF,As oCo o0, : C, 35.20; H, 2.54; 21 18 6 2 2 4 mol. wt. 716. Found : C, 35.16; H, 2.63; mol. wt. 695. V C 0 ( C _ C 6 H 1 2 ^ : 2052 (9), 2003 (10), 1974 (8), 1959 (8) cm"1, hi nmr(CDCl3) : singlets at 1.25 (area 3), 1.60 (area 3), 1.80 (area 3) and 1.85 (area 3) , singlet at 19 5.45 (area 1) and multiplets at 7.40 and 7.60 (area 5) ppm. F nmr(CDCl3) : quintets at 106.6 (area 4) and 125.8 (area 2) ppm. * Variable temperature experiment indicated that the data are independent of temperature from -80° to 30°C. C. (f 6AsP)(PhC = CH)Co 2(C0) 4, LII, black crystals, m.p. 90°, yield 45%. Anal. Calc. for C^H^FgAsPCo^ : C, 46.72; H, 2.79; mol. wt. 796. Found : C, 46.05; H, 2.60; mol. wt. 747. v„rt(c-C£H,_) : 2055 (10), 2004 (10), 1990 (3), 1974 (8), 1958 (5) cm"1. 1H nmr(CDCl3)': complicated, main peaks are singlets at 1.70 and 1.80 (area 6), singlets at 4.80 and 4.90 ( area 1), multiplets at 7.10 and 7.30 (area 15) ppm. 19 F nmr(CH2Cl2): quite complicated with five strong singlets at 104.4, 105.3, 106.7, 107.2 and 108.7 (area 4), and a strong singlet at 128.4 (area 2) ppm. D. (f gfars)(PhC = CH)Co 2(C0) 4, LIII, black crystals, m.p. 100°, yield 50%. Anal. Calc. for C ^ H ^ F g A s ^ o ^ : C, 34.46; H, 2.37; mol. wt. 766. Found : C, 34.67; H, 2.59; mol. wt. 730. V C 0 ( C " " C 6 H 1 2 ) : 2 0 5 0 ^ ' 2002 (10), 1989 (7), 1967 (3) cm"1. \ nmr(CDCl3) : tri p l e t ( J = 1.0 Hz) at 1.25 (area 3), t r i p l e t (J = 1.3 Hz) at 1.60 (area 3), doublet - 114 -(J = 1.7 Hz) of doublet (J = 0.5 Hz) at 1.77 (area 3), doublet (J = 1.6 Hz) of doublet (J = 0.7 Hz) at 1.90 (area 3), singlet at 5.25 (area 1), multiplets at 7.30 and 7.60 (area 5) ppm. H- F decoupling at 94.0866352 H^ caused the four multiplets at 1-2 ppm to coalesce into 19 four singlets. F nmr(CH 2Cl 2) : singlets at 101.7 (area 4) and 133.8 (area 4) ppm. * Using Varian HA-100 nmr Spectrometer. E. fgfos(PhC = CH)Co 2(C0) 4, LIV, black crystals, m.p. 190°, yield -80%. Anal. Calc. for C42 H26 F8 P2 C o2°4 : C ' 5 4 , 3 5 H, 2.84; mol. wt. 926. Found : C, 54.88; H, 3.20; mol. wt. 874. v„-<c-C,H,J : 2050 (9), CO o 1/ 2002 (10), 1989 (7), 1967 (3) cm"1. ^ nmr(CDCl3) : singlets at 4.60 (area 0,5), 4.80 (area 0,5), multiplets at 7.10 (area 5) and 7.50 (area 20) 19 ppm. F nmr(CH 2Cl 2) : singlets at 101.7 (area 4) and 133.8 (area 4) ppm. F. (dab)(PhC = CH)Co2(CO)4, LV, black crystals, m.p. 70°, yield 30%. Anal. Calc. for C ^ H ^ F g A s ^ o ^ : C, 34.09; H, 2.58; mol. wt. 704. Found : C, 33.55; H, 2.30; mol wt. 676. v^(c-C,,H10) : CO o Li. 2057 (6), 2022 (10), 1986 (3) cm"1. Vnmr(CDClg) : singlets at 1.25 (area 3), 1.60 (area 3), 1.80 (area 3) and 1.85 (area 3), singlet at 5.25 19 (area 1), multiplets at 7.30 and 7.50 (area 5) ppm. F nmr(CH 2Cl 2) : singlet at 51.3 ppm. - 115 -o %G. f gfos(PhC E CH)Co2(CO)4, LVI, black crystals, m.p. 155 yield 50%. Anal. Calc. for C 4 1 H 2 6 F 6 P 2 C o 2 ° 4 : C ' 5 6 • 1 6 5 H» 3' 0 05 mol. wt. 876. Found : C, 55.53; H, 2.75; mol. wt. 803. v^(c-C,H 1 0) : CO o 12 2049 (9), 2002 (10), 1990 (7), 1968 (4) cm"1, h. nmr(CDCl3) : singlets at 4.60 (area 0.5), 4.75 (area 0.5), multiplets at 7.20 (area 5) and 7.55 19 (area 20) ppm. F nmr(CH 2Cl 2) : singlets at 108.5 (area 4) and 134.3 (area 2) ppm. - 116 -III. Results and Discussion 1. (L-L)Co 2(CO) 6 Complexes Like fgfars [13], fgfars and fgfars also react with Co2(C0)g to give brown-red solids I and II, respectively, of the formula (L-L)Co 2(CO) 6. However, unlike f^farsCo 2(CO)g (A) which shows two bridging CO bands at 1842 and 1786 cm \ a difference of 56 cm \ both of these new compounds give two bridging CO bands at ~1815 and ~1801 cm \ a difference of only ca. 14 cm 1 . Since the frequencies and pattern of the i r CO bands in these compounds are very similar to those reported for other (L-L)Co2(CO)g (B) complexes (L-L = f^fos, f g r o s and dab) (Table V), i t is reasonable to assume that they a l l have the same type of structure. X-ray studies [13] have confirmed that f 4farsCo 2(CO)g (A) has a structure 2 which is similar to that of Co2(CO)g [1], the ligand fgfars acting as a bridging bidentate ligand replacing one equatorial CO group from each Co atom. Regarding the structure of the other (L-L)Co2(CO)fi (B), Crow et. a l . [13] favored _3 and discounted 4^, because structure h_ "contains a unique CO group on the substituted Co atom which would.be expected to have i t s stretching frequency at lower energy and would manifest i t s e l f as a single isolated peak in the i r spectrum". From group theory a compound with structure having four terminal CO groups may give rise to a maximum of four terminal CO bands. From Table V, i t can be seen that I, II and the other (L-L)Co 2(CO) 6 (B) compounds (L-L = fgfos, fgfos and dab) give only three terminal CO bands. It is possible that the fourth terminal CO band is - 117 -TABLE V. Infrared Data for (Ligand)Co„(CO), 2 o Ligand -1 * vco c m Solvent Ref. f, fars 4 2052(6), 2020(8), 1996(10), 1842(3), 1786(3) a [13] f 4 f o s 2066(1), 2052(5), 2022(8), 1998(10), 1842(2), 1798(2) a [13] dppm 2050(5), 2016(9), 1990(10), 1840(4), a e 1794(3) f .AsP 4 2068(9), 2019(9), 2000(10), 1996(10), 1823(7), 1797(8) b e f.fos 2070(9), 2018(9.5), 1994(10), 1815(6), b 1790(7) f,fars 0 2067(10), 2004(9), 1995(10), 1813(5), 1800(8) a e fgfars 2068(8), 2005(9,sh), 1997(10), 1817(3), 1802(6) a e f g f OS 2072(7), 2006(10), 1998(8), 1818(2), 1800(5) a [13] fg fOS 2072(7), 2004(10), 1998(9), 1818(2), 1802(5) a [13] dab 2068(5), 2006(6), 1996(10), 1812(1), a [13] 1800(3) C-,H0 2076(11), 2024(17), 2012(13), 2009(sh), 7 8 1849(2), 1834(10) c [15] 2057(5.1), 2047(sh), 2012(1.2), 1994(6.7), diars 1989(10), 1821 (sh), 1811 (sh), 1799(2.1), 1791(sh) d [44] 2058(5.0), 2041(0.6), 2006(1.4), 1995(5.4) Me2As(CH2)3AsMe2 1983(10), 1822(0.3), 1812(sh), 1800(1.5) 1781(0.4) d [44] . * Figures in parentheses are relative intensities. a = cyclohexane , b = CH2C12, c = heptane, d = hexane, e = this work - 118 -weak or is degenerate with one of the observed bands. Furthermore, recent studies [17] have shown that (Tr-diene)Cb-(CO) , (Tr-diene = norbornadiene, z o C^Hg) has a structure identical with 4^  in which the diene acts as a chelating bidentate ligand replacing two equatorial terminal CO groups from one Co atom. The i r data of this compound show four terminal CO bands at 2076, 2024, 2012 and 2009 (sh) cm - 1. Since there is no evidence that this compound i s tautomeric i n solution [15], the four observed bands are probably a l l fundamentals. This suggests that the CO band due to the unique terminal CO group on the substituted Co atom in (7r-C-,H„) CO„ (CO), / o z o could have a stretching frequency of at least 2009 cm \ a frequency comparable to those of the other terminal CO groups on the non-substituted Co atom. The present investigation also found (Table V) that f,AsPCo„(C0),, 4 Z D III, which has i r data similar to other (L-L)Co2(C0)g (B) complexes and is not tautomeric in solution, shows four terminal CO bands at 2068, 2019, 2000 and 1996 cm"1. Thus the CO band due to the unique terminal CO group i n (L-L)Co2(C0)g (B) could have a frequency comparable to those of the other terminal CO groups. Hence structure _4 cannot be excluded for the other (L-L)Co2(C0)g (B) complexes. The pattern of CO stretching vibration i n (L-L)Co2(C0)g (B) and (ir-C7Hg)Co2(C0)g are very similar (Table V), and therefore they probably have the same structure (i.e. 4). This conclusion is supported by the finding that the reaction of (TT-C-,H O)CO O(C0) , with L-L gives / o z o (L-L)Co2(C0)g (B) under mild conditions. Since in this work i t was also found that under the same conditions, excess ligand does not react with (ligand) Co2(C0)g (ligand = L-L or C7Hg) , the formation of - 119 -(L-L)Co 2(CO) 6 (B) from ( ir^H^Co^CCOg is evidently simply due to the replacement of C_H by L-L : Other evidence to support this conclusion is (i) the highest v C Q in (L-L)Co2(CO)g (B) is considerably higher than that i n f^farsCo 2(CO)g, ( i i ) other studies on carbonyl derivatives of Cr, Mo, W [84], Mn, Re [89] and Fe [34, 90] indicate that the ligands (L-L) (L-L = fgfos, fgfos, fgfars, .fgfars and dab) with few exceptions always act as a chelating bidentate ligand. Recently Manning et. a l . [44] assigned structure _3 to diarsCo 2(C0)g and Me 2As(CH 2) 3AsMe 2Co 2(C0)g, and structure _4 to Me2As(CH2)^AsMe2Co2(CO)g (n =6, 8). The former two complexes have i r CO bands very similar to those of the (L-L)Co-(CO),. (B) complexes (Table V) which as described above are now believed to have structure 4_. The terminal CO bands of the latter two complexes are similar to those of - 120 -[(MeO)3P]2(PhC = CPh)Co 2(C0) 4 where the two (MeO>3P ligands are believed to occupy axial-axial positions (see page 134). Furthermore the bite of diars does not seem to be large enough to form complex of geometry ^; while that of Me0As(CH0) AsMe0 (n = 6, 8) is sufficient to do so. Thus 2 2 n 2 i t appears that their proposed structures are probably not correct. Previously [13] i t has been reported that f^fos reacts with Co2(C0)g to give f 4fosCo 2(C0)g (A) whose i r CO bands are similar to those of f 4farsCo 2(CO)g (A) and hence has the same type of configuration, i.e. 2. It has now been found that under uv irradiation f.fosCo_(CO), — 4 L o (A) can be isomerized to f^fosCo 2(CO)g (B), XIV. The i r CO bands of this compound are similar to those of the (L-L)Co„(CO)- (B) complexes I o and thus the compound may be assigned the same structure k. Complex XIII, dppmCo2(C0)g, has i r CO bands very similar to those of fgfarsCo 2(CO)g [13] and hence is believed to have the same geometry (i.e. 2). Like the other (L-L) Co 0 (CO) , (A) complexes (L-L = fgfars and f^fos) , XIII also shows three terminal CO bands and two bridging CO bands. For a structure of Cs symmetry (like 2), four terminal CO bands are expected, but one of them might be accidently degenerate. 2. (L-L) 2Co 2(CO) 4 Complexes A. (f 4fos) 2Co 2(CO) 4 Besides the isomerization to f 4fosCo 2(C0)g (B), XIV, uv irradiation of f 4fosCo 2(C0)g (A) in the presence of f 4 f o s gives a black - 121 -solid XV. Analytical data and molecular weight measurement suggest the formula (f^fos^Co^CCO^. The pattern and frequencies of i t s i r CO bands are quite similar to those of (dppe) 2Co 2(CO) 4 and (TP)2Co^(CO)^ (Table VI), and for this reason the same type of structure .5 can be proposed for XV. The successful conversion of XV into XIV (of possible configuration 4), by reaction with CO, lends further support to this point. TABLE VI. Infrared CO Bands of Some (L-L) 2Co 2(CO) 4 Complexes complex v C 0(CH 2Cl 2) cm"1 * Ref. (dppe) 2Co 2(CO) 4 1958(m), 1925(vs), 1753(m), 1725(vs) [23] (TP) 2Co 2(CO) 4 1975(m), 1918(s), 1760(s,sh), 1728(s) [23] (f 4fos) 2Co 2(CO) 4, XV 1972(8), 1952(8), 1765(6), 1732(9) a (f 4fos) 2Co 2(CO) 4, XVI 1950(10,br), 1735(6,br) a a = this work. B. (f 4fos) 2Co 2(CO) 4 Attempts to synthesize the above mentioned (f 4fos) 2Co 2(CO) 4 by reacting [ (ir-C^g) Co (CO) 2] 2 with f 4 f o s lead to the formation of a new compound, XVI, (f 4fos) 2Co 2(CO)^. The 1 9 F nmr spectrum of XVI indicates that the two f 4 f o s ligands are equivalent. However i t s i r CO bands - 122 -are quite different from those of the isomeric XV. Furthermore, complex XVI reacts with CO to give f^fos Co 2(CO) g (A) (of possible structure 2), hence 26_ is proposed for this compound. 3. LCo 2(CO) 7 The ligand f^PS reacts with Co 2(CO) g to give a new compound f 4PSCo 2(CO) 7, VI, in low yield. The 1H nmr(CDCl3) spectrum of this compound shows a singlet at 2.25 and a multiplet centered at 7.55 ppm in the ratio of 3 to 10. The singlet at 2.25 ppm is l i t t l e changed from i t s position in the free ligand f^PS (2.40 ppm ) indicating that the S atom is uncoordinated (coordination would shift the resonance down f i e l d considerably [90]). It is therefore concluded that the ligand is coordinated to one of the Co atoms via a P atom. The i r spectrum of this complex shows four terminal CO bands very similar to those reported for (Ph 3P)Co 2(C0) 7 (Table VII), and, therefore, the same configuration 1_ (L = f^PS) can be proposed for this compound. - 123 -TABLE VII. Infrared Data for LCo 2(CO) 7 Complexes L v c o c m - 1 Solvent Ref. Ph3P 2079(4.4), 2026(4 1996(10), 1964(3) .8), 2010(sh), hexane [18] f 4PS 2070(9), 2025(9), 1968(8) 2000(10) CH2C12 * a * a : This work. 4. [(L-L)Co(CO) 3] 2 Complexes The ligand f^AsP reacts with Co2(C0)g to give three main products : (i) a brown red solid f.AsPCo0(CO),, III, ( i i ) a violet solid H i. O [f 4AsPCo(CO) 3] 2, IV, and ( i i i ) an orange powder [(f 4AsP)Co(C0) 3] 2, V. Compound III has i r CO bands similar to those reported for (L-L)Co2(C0)g (B) complexes (Table V) and so structure k_ can be proposed for i t . Compound IV has an empirical formula f 4AsPCo(CO) 3 as deduced from elemental analyses. However, molecular weight measurement suggests the dimeric formula [f 4AsPCo(C0) 3] 2. Further, i t s i r spectrum in the carbonyl region is very similar to those reported for the [LCo(C0) 3] 2 (L = Bu3P, (Me0)3P, (Ph0)3P, etc. ) complexes, and therefore 8^  can be assigned to this compound. The V nmr spectrum of IV has a single methyl proton peak for the Me2As-groups (at 1.10 ppm ) which is close to - 124 -that of the free ligand f^AsP (1.30 ppm ) and i t is concluded that the As atoms are not coordinated. Therefore, each f.AsP molecule must be 4 coordinated to each Co atom via the P atom. This deduction has been confirmed by a recent X-ray analysis [91]. The molecule has the staggered configuration identical with j5. This was the f i r s t example known where a potential b i - or tri-dentate ligand acts as a monodentate one and affords the [(L-L)Co(CO)^]2 complex. Others are discussed later. An isomer of [f4AsPCo(CO)^]^, complex V, was also isolated from the same reaction. The structure of this isomer i s discussed below (page 127). The ligand f^PS reacts with Co2(C0)g to give, in addition to the mono-substituted complex VI, a di-substituted compound [f^PSCo^O).^' VII. This compound is assigned a non-bridged Co-Co bonded structure similar to complex IV on the basis of i t s i r spectrum. The shift (2.40 ppm.) of the nmr signal due to the S-Me groups is consistent with the S atoms being equivalent and uncoordinated, i.e., each f^PS ligand is coordinated to a . Co atom via the P atom. Another type of unstable derivative, orange to red solids, of formula [(L-L)Co(CO)^]^ is given by the ligands f^AsF, f^AsCl, f^AsO, f^AsN, and f^AsS. The isomeric derivative of f^AsP is also in this class. Their i r spectra (Table VIII) are similar to each other, but slightly different from the afore mentioned [(L-L)Co(CO)^]^ (L-L = f4AsP_, f^PS) type spectra. - 125 -TABLE VIII. Infrared Data of Some [(Ligand)Co(CO)3] Complexes Complex -1 * v co c m Solvent Ref. [Bu3Vco(C0) 3] 2 1970(1.2), 1951(10) a [12] [Ph 3PCo(CO) 3] 2 1977(sh), 1957(10) b [12] t(MeO)3PCo(CO)3] 2 1992(1.3), 1975(10) a [12] [(PhO)3PCo(CO)3] 2 1998(sh), 1979(10) b [12] [f4AsPCo(CO)3].2, IV 1991(3,sh), 1969(10) c d [f 4PSCo(CO) 3] 2, VII 1990(4,sh), 1968(10) c d [f 4As_SCo(CO) 3] 2, VIII 2066(0.5), 2045(2.0), 1990(10,b r.) c d [f 4AsPCo(CO) 3] 2, V 2040(0.5), 2004-1997(10,br.) c d [f 4AsFCo(CO) 3J 2, IX 2055(1), 1992(10,br.) c d [f 4AsClCo(CO) 3] 2 ! , x 2070(1), 1991(10,br.) c d [f 4AsOCo(CO) 3] 2, XI 2078(1), 1988(10,br.) c d [f 4AsNCo(CO) 3] 2, XII 2040(2), 1993(10,br.) c d * Figures i n parentheses are relative heights of bands measured on a linear transmittance scale. a = heptane; b = CHC1.,; c = CH9C19; d = this work. - 126 -Complex VIII, [f ^AsSCoCCO) . j ^ . i s the most stable one i n this series. From a survey of the literature i t s i r CO bands are similar to those of (a) [LCo(CO) 2Cl] 2 [28], (b) I^CoCCCO^X- [38], (c) R3MCo(CO)4 [92], and (d) [LCo(CO) 3] 2 (Table VIII). Possibilities (a) and (c) are ruled out by the analytical data. Possibility (b) can be eliminated since i f X were Co(CO)4 the compound would show v C Q at -1890 cm"1 (which was not observed); furthermore there was no reaction of the compound with either NaBPh. or NH.PF, . 4 4 6 Both the analytical data and mass spectrum of VIII suggest the formula f 4AsSCo(C0) 3- However, molecular weight measurement using an osmometric method gives a value of 713 suggesting the dimeric formula [f 4AsSCo(C0) 3] 2 (Calc. 838). Hence i t is possible that complex VIII has the metal-metal bonded configuration J3. The nmr spectrum of a fresh sample of VIII shows a broad singlet at 2.00 (area 12) and two broad singlets at 2.50 and 2.70 (area 6) ppm. The singlet at 2.00 indicates that the AsMe2 groups are coordinated ,while the other two singlets indicate that the SMe groups are not coordinated. Although the following configuration can be proposed for VIII, - 127 -where AsZ is f^AsS, i t does not account for the two SMe resonances. The formula of the isomeric compound [f^AsPCoCCO)^]^, V, i s substantiated by analysis and molecular weight. The "4l nmr spectrum of V shows a singlet at 2.05 ppm. due to coordinated AsMe2 groups, hence structure 27 (AsZ = f^AsP) can be proposed for i t . Similarly, f^AsN yields [f^AsNCotCO) 3] 2 in which the As is bonded as in 27. Very unstable complexes were obtained from f^AsF, f^AsCl, and f^AsO, which can probably be formulated the same way. These compounds did not give satisfactory analytical data and mass spectra. Inspection of Table VIII indicates that these AsMe2 bonded [(L-L)Co(C0) 3] 2 derivatives have i r frequencies at higher wave numbers than the other compounds with the same skeleton. The reason for this is not clear. 5. Reaction of [LCo(C0) 3] 2 with (L-L) When [LCo(CO) 3] 2 (L = Bu^P or (Me0)3P ) and f^fos are refluxed in benzene, the dark red complexes XVII and XIX of the formula L ( f 4 f o s ) C o 2 ( C 0 ) c a n be isolated. The i r spectra of these derivatives show both terminal and bridging CO bands. The single bridging CO band indicates that both the ligands L and L-L are symmetrical with respect to the two bridging CO groups, hence 28(a) and (b) are possible struc-tures. The CO band at the low frequency of 1930 cm 1 is consistent with configuration J28(a), since a complex of configuration 2_8(b) would be expected to have i t s terminal CO frequencies higher than this. A structure such - 128 -\ ° 8 II II II , / \! \\l, 28(a) 28(b) as 28(a) can also help rationalize the ease of formation of [LCo(CO) ] (L - Bu"p) and f ^ o s from XVII in the CO exchange reaction 3J2 O \ / / — / / -> [(L-L)Co(C0) 2] 2 + [LCo(C0) 3J 2 (L-L) + other products where (L-L) is f.fos 4 Such a process is d i f f i c u l t to envisage for 28(b). - 129 -When [Ph^PCoCCO)^]2 and dppe are refluxed in a polar solvent (THF), the yellow complex XX i s obtained. The infrared spectrum of this compound shows three CO bands at 2012 (3), 1962 (7) and 1887 (10) cm"1, similar to those reported for (dppe) 3Co 2(CO) 4[Co(CO) 4] 2 ( O I possible configuration 16). Reaction of XX with NaBPh^ yields a yellow compound XXI whose molecular weight and analytical data indicate the formula (Ph 3P) 4(dppe)Co 2(CO) 4(BPh 4) 2. Thus the configuration 29(a), similar to J-6_, can be proposed for the cation. Co L L \ / C 0 ov Co c 0 2+ 29(a) In the photolytic reaction between [LCo(CO) 3] 2 and (L-L) two types of products are Isolated. One of these has a formula L 4(L-L)Co 2(CO) 4X 2 (L = Bu^, L-L = dppe, X = Co(C0) 4 > XXIV. L = Bu^, L-L = dppp, X = Co(C0) 4, XXV. L = Bu^P, L-L = dppp, X = BPh4, XXVI.) The infrared CO bands of this complex type are similar to those of the complexes XX and XXI, hence the same basic structure 29(a) can be +2 -assigned. The other complex type has a formula [L 2 (L-L) Co (CO)]^ (X ) 2 (L = Ph3P, L-L = dppe, X = Co(C0) 4, XXII; or BPh4, XXIII) and shows a CO band at -1935 cm \ A band in this region i s reported [42, 43] for L4Co(C0)"*"x compounds (of possible configuration 17), and the related dimeric configuration 29(b) can be suggested for the cation [L 2(L-L)Co(CO)]^ 2+ - 130 -L Co C 0 -L L. L L L Co C 0 2+ 29(b) In polar solvents, the photochemical reaction between [LCo^O).^ (L = Bu^P) and fluorocarbon-bridged ligands (e.g. fgfars, f^fos) gives only the known ionic I^Cc^CO^X salts (of possible con-figuration 15). However, in nonpolar solvents, fgfars reacts with [Bu nJ?Co(C0) 3] 2 under uv irradiation to afford a dark brown o i l . The "4l nmr spectrum of this substance shows three singlets at 0.70, 1.30 and 1.90 ppm. The singlet at 1.90 ppm suggests that both the AsMe2 groups of fgfars are coordinated. The ligand bridged structures 25(a) or 25_(b) are possible for this species (L = Bu^P, D~E = fgfars), while the ligand chelated configuration such as ^9_(c) i s less l i k e l y because fgfars is known to be reluctant to form chelate complexes. The low yield and i n s t a b i l i t y of this species precludes i t s complete characterization. Further studies are needed to cla r i f y the situation. 0 C L L I \ / Co Co — A 29(c) - 131 -6. L(RC = CR)Co2(CO)5 Complexes When equimolar quantities of (MeO)„P and (RC E CR)Co„(CO), J £• 0 (R = Ph, CH2OH) are heated, mono—substituted (MeO),jP(RC = CR)Co2(CO),. is formed. This type of complex gives four terminal CO bands. Using the solid state structure of (PhC = CPh)Coo(C0), [75] as a model, the possible configurations 30_ and _31 can be suggested. Provided the structure i s rigid in solution, then in 30, the two R groups w i l l be inequivalent, and a compound such as 3>0_ (R = CH20H) would be expected to show in i t s nmr spectrum absorptions due to two different CH20H groups. However, a geometry such as 31 would only show absorption due to two equivalent CH20H groups. As can be seen (Table IX), XXX shows a tri p l e t at 2.45 ppm indicating the presence of one type of OH group, and a doublet at 4.75 ppm indicating the presence of one type of CH2 group, i.e. the structure i s consistent with 31. - 132 -TABLE IX. h NMR Data (CDC13) for Some L ^ R C E C R ) C O 2 ( C O ) 6 _ r Complexes Compound nmr data (MeO)3P(HOCH2C=CCH2OH)Co2(CO) tri p l e t (JM - 6 Hz) at 2.45 (area 2), (XXX) doublet ( J p R = 11 Hz) at 3.70 (area 9), doublet ( J R H = 6 Hz) at 4.75 (area 4) ppm. [(MeO)3P]2(HOCH2CECCH2OH)Co2(CO) 4 t r i p l e t ( J ^ = 6 Hz) at 2.40 (area 2), (XXXII) t r i p l e t (J = 5.5 Hz) at 3.65 (area 18), doublet ( J R H = 6 Hz) at 4.65 (area 4) ppm. (MeO)3P(PhCECPh)Co2(CO)5 doublet ( J p R = 11 Hz) at 3.17 (area 9), (XXIX) multiplets at 7.21 and 7.83 (total area 10) ppm. [(MeO)3P]?(PhCECPh)Co2(CO) tri p l e t (J » 5.5 Hz) at 3.65 (area 18), (XXXI) multiplets at 7.55 and 8.10 (total area 10) ppm. [(MeO)3P]3(PhC=CPh)Co2(CO)3 multiplets at 3.21 (area 9) and 3.50 (XXXIII) (area 18), multiplets at 7.23 and 7.93 (total area 10) ppm. [(MeO)3P]4(PhCECPh)Co2(CO) broad singlet at 3.43 (area 36), multiplets (XXXIV) at 7.21 and 7.9 7 (total area 10) ppm. - 133 -TABLE X. Infrared CO Bands of Some L (RC=CR)Co0(CO)£ Complexes N 2 6-n r Compound v C 0(CH 2Cl 2) cm"1 (MeO) 3P(HOCH 2CHCCH 2OH)CO 2(CO) 5 (XXX) 2068(7), 2018(10), 2008(10), 1981(6). (MeO)3P(PhC=CPh)Co2(CO) (XXIX) 2064(8), 2015(10), 2002(10), 1974(7). [(MeO)3P]2(HOCH2CECCH2OH)CO2(CO) (XXXII) 2032(7) , 1974(10 ,br.) [(MeO)3P]2(PhCECPh)Co2(CO) (XXXI) 2026(7), 1971(10,br.) [(MeO)3P]3(PhCECPh)Co2(CO) (XXXIII) 1997(9) , 1950(10). [(MeO)3P]4(PhCECPh)Co2(CO) (XXXIV) 1996(2), 1942(8,sh), 1922(10). 7. L2(RC = CR)Co 2(C0) 4 Complexes When a 2:1 mole ratio of (MeO) _P and (RC = CR)Co„(CO), J 2 b (R = Ph, CH2OH), or equimolar quantities of (MeO)3P and (MeO)3P(RC = CR)Co2(CO)5 are refluxed in benzene, [(Me0)3P]2(RC = CR)Co 2(C0) 4 is formed. The i r spectra of [(MeO) P] 2(RC = CR)Co 2(C0) 4 (R = Ph, XXXI; R = CH2OH, XXXII) complexes (Table X) show only two terminal CO bands at - 134 --2030 (m) and -1970 (vs) cm"1. The simple pattern of the CO bands suggests a symmetrical structure and four arrangements J52 to _3_5 should be considered. The hi nmr data for XXXII (R = CH^ OH) are listed i n Table IX. The t r i p l e t at 2.40 ppm and the doublet at 4.65 ppm indicate that the two R groups are equivalent, hence _34 and 35_ can be ruled out leaving 32_ and _3_3 as poss i b i l i t i e s . However, as this compound can be synthesized from XXX (of possible configuration 31) and can also be converted back to XXX by reaction with CO, 32_ i s favored. R I < o C ^ C o — i C ^ f c o d s I *d =C0 I R I oc co R A 1 32 33 - 135 -From Table IX one can see that the nmr spectra of the [(MeO)3P]2(RC = CR)Co2(CO)4 complexes, XXXI and XXXII, show a 1:2:1 tr i p l e t at 3.65 ppm. Harris [93, 94] has developed the theory of nuclear spin systems of the type XnAA'X^ (in the present case X = H, A = P) and has shown that a simple 1:2:1 tr i p l e t resonance w i l l be observed when ^(A-A')^ > > ^(A-X) + ^(A' X) I' ^ e n c e t^ l e t r i p l e t resonance observed in these complexes can be attributed to strong P-P coupling. It i s known [95] that trans (R 3P) 2M(CO) 4 (R = Me, OMe; M = Cr, Mo, W) have large P-P coupling while the cis analogs do not. The above result suggests that the two (MeO)3P molecules are not cis to each other as in 33. It is remarkable that v i r t u a l coupling i s seen i n a molecule of configuration 32. 8. L 3(RC = CR)Co2(CO)3 The tri-substituted complex [(MeO)3P]3(PhC = CPh)Co2(CO)3, XXXIII can be made by refluxing equimolar quantities of (MeO)3P and complex XXXI i 31 in toluene. The T l nmr spectrum of XXXIII, after decoupled from P, shows two singlets at 3.06 and 3.33 ppm of relative, ratio 1:2 indicating the presence of at least two different sets of (MeO)3P groups. Since this compound can also be converted to XXXI (of possible configuration 32) by reacting with CO, 36 (R = Ph) appears to be most li k e l y . 9. L 4(PhC = CPh)Co2(CO)2 Complex XXXIV, [(Me0)3P]4(PhC a CPh)Co2(CO)2, can be prepared by refluxing (Me0)3P and XXXIII in toluene. Its hi nmr spectrum shows a broad methoxyl proton peak suggesting that the four (Me0)3? molecules - 136 -36 37 are equivalent or nearly so. Thus, possible structures are 37 t o 3 9 • However, this compound can be readily converted to XXXIII (of possible configuration 36) and vice versa, hence 38 or 39 is more lik e l y . L R I ' / f x W R' N \ ! 38 39 In 38 and 39, the two axial ligands and the two equatorial ligands are expected to be inequivalent. But, apart from the acetylene molecule which is quite remote, the two ligands on each Co atom are in fact in quite - 137 -similar environments (cf. molecular models) and the chemical shifts of the proton could well be close accounting for the broad singlet. The same compound XXXIV shows three CO bands at 1996(2), 1942 (8, sh) and 1922 (10) cm"1. Since (MeO)3P is a worse ir-acceptor than CO, replacement of one CO group from XXXIII to give XXXIV, w i l l be expected to lower the considerably. The band at 1996 cm \ which is identical to the v (max) of XXXIII, is evidently due to the presence of trace XXXIII as an impurity. However, i t is s t i l l not certain whether the band at 1942 (sh) cm 1 i s fundamental or not. If i t i s , then a weak-strong pattern of the CO bands would tend to favor 39. 10. (L-L)(PhC E CH)Co2(CO)4 Previously [13] i t was found that f.fars and f.fos reacted 4 4 with (PhC = CH)Co2(CO)6 to give the black (L-L)(PhC = CH)Co 2(C0) 4 complexes believed to have configuration 7A_ (R = Ph, R' = H). However, attempts to isolate the similar compounds by using f,fos, f,AsP, f D f o s O O O and dab were unsuccessful [13]. In this work, by modifying the reported procedure, the (L-L)(PhC E CH)Co2(CO)4 complexes of these and the related ligands f 4AsP, fgfars and fgfars have been isolated. A. (L-L)°(PhC = CH)Co„(C0). 2 4 Except for the derivative of f 4AsP, a l l the new complexes LI to LVI are black solids with four CO bands at -2050, -2000, -1980 and -I960 cm 1. The \ nmr spectrum of fgfars(PhC E C H ) C O 2 ( C O ) 4 , LI, shows - 138 -four equal intensity singlets at 1.25, 1.60, 1.80 and 1.85 (total area 12), a singlet at 5.45 (area 1), and multiplets centered at 7.40 and 7.60 (total area 5) ppm. The singlet at 5.45 ppm ascribed to the = CH group suggests that only one compound is present, while the four singlets in the region 1 to 2 ppm indicate that the four methyl groups on the ligand fgfars are inequivalent. Variable temperature studies (from -70° to +30°) indicate that the H^ nmr spectrum of LI is independent of temperature, and hence i t s structure i s probably r i g i d . Keeping in mind that the bite of the ligands fgfos, fgfos, fgfars, fgfars and dab [34, 84, 89, 90] does not favor the formation of bridged complex, two possible structures 40 and 41_ are suggested for these compounds (R = Ph, R' = H). Or R I v co R I R' 40 41 19 The F nmr spectrum of LI shows a tr i p l e t at 106.6 (area 4) and a quintet at 125.8 (area 2) ppm , very similar to that of the free ligand fgfars [85]. This suggests that the ligand i s coordinated f a i r l y symmetrically so that the original symmetry of the ligand i s preserved. - 139 -1 Q The F nmr data for fgfars(PhC = CH)Co2(CO)4, LIII, f,fos(PhC = CH)Co„(CO)., LVI, f„fos(PhC = CH)Co„(CO),, LIV, and o z 4 o z H (dab)(PhC = CH)Co2(CO)4, LV, further i l l u s t r a t e this point. The preferred structure of this complex type is therefore probably 40. When R = Ph, R' = H, and L-L = fgfars, i t can be seen from molecular models that in 40, the AsMe2 group near the ph group is different from that near the H atom. Furthermore, i t is evident that the two methyl group up (cis to the axial CO) are different from those trans. This accounts for the four methyl proton peaks observed for complex LI. Complex LV, (dab)(PhC = CH)Co2(CO)4 gives the same type of h nmr data. While the hi nmr spectrum of LIII, fgfars(PhC = CH)Co 2(CO) 4 > 19 L, shows four multiplets i n the region 1 to 2 ppm. However, F-TI decoupling reduced these multiplets to four singlets and hence the stereochemistry is probably the same as 40. Complex LII, f&AsP(PhC = CH)Co 2(CO) 4 > shows i r CO bands similar to those of the (L-L)°(PhC = CH)Coo(C0). (L-L = f-fars, f 0 f a r s , Z 4 0 O f,fos, f„fos, dab) complexes mentioned above, hence i t is believed that the ligand fg^sP acts as chelating bidentate ligand. Nevertheless, the hi nmr spectrum of LII is very complicated. The two singlets at 4.80 and 4.90 ppm suggest that the sample contains two isomers of possible configurations 40_ and/or 41 (in each of these there would be two possible ways of arranging the mixed ligand). The complex pattern of i t s 19 F nmr spectrum further indicates the presence of more than one isomer in the sample. - 140 -B. (L-L) b(PhC = CH)Co2(CO)4 The mixed ligand f^AsP gives a black solid, L, whose Tl nmr spectrum shows singlets at 1.50 and 1.75 (total area 12), another group of singlets at 5.30 and 5.60 (total area 1), and two multiplets at 7.40 and 7.60 (total area 5) ppm. The presence of two singlets at 5.30 and 5.60 ppm assignable to the = CH group, indicates that this sample contains two different species. Since the i r CO stretching frequencies of L are very similar to those of fgfars(PhC = CH)Co2(CO)4 whose structure i s believed [13] to be 24, and the bite of f A A s P is quite close to that of fgfars, L is probably a mixture of two isomers with possible configuration 24 (R = Ph, R' = H; R = H, R' = Ph). 11. (L-L)(PhC = CPh)Co 2(C0) 4 The ligands f ^ a r s (n = 4, 6, 8), f RAsP (n = 4, 6), fRf°s (n = 4, 6, 8), dab, dppm, dppe and dppp a l l react with (PhC = CPh)Co2(CO)6 to form (L-L)(PhC = CPh)Co 2(C0) 4 complexes which can be classified into four groups : A. (L-L) b(PhC = CPh)Co2(CO)4 The ligands f 4 f a r s , f 4AsP, f 4 f o s , fgfars and dppe give dark green solids XXXV to XXXVIII and XLV with four CO bands at -2035, -2005, -1980 and -I960 cm-1, similar to those of (f 4fars)(PhC = CH)Co2(C0) for which structure 24 (R = Ph, R' = H) was suggested [13]. The ligand dppm gives dark brown solid XLIV, but i t s i r CO bands are similar to those of the above compounds and i s , therefore, assumed to have the same - 141 -type of configuration. 59 Recent Co nqr studies (Chapter 5 of this thesis) on fgfars(PhC = CPh)Co 2(C0) 4 reveal that the two Co atoms i n the molecule are equivalent. On this basis, i n addition to 24 (R = R' = Ph), configuration 42 (R = R' = Ph) must also be considered for class A complexes. R 42 The hi nmr data of these (L-L)(PhC = CPh)Co 2(C0) 4 (L-L = fgfars, f^AsP, fgfars) complexes show a singlet at -1.5 ppm for the AsMe2 group from -80° to +30°. This supports the more symmetrical configuration ^2. However, the same type of complex is also produced by the ligand dppm whose small bite would not seem to favor the formation of a compound with configuration 42. Furthermore, i t has been concluded above that the red complexes [(MeO)3P]2(RC = CR)Co 2(C0) 4 (R = Ph, CH20H) have a geometry 32^ which i s identical to 42^  (provided that the two L's are replaced by one (L-L)). If class A complexes have the geometry 42, - 142 -then their i r CO bands should be quite similar to those of [(MeO)3P]2(RC = CR)Co 2(C0) 4. The data i n Tables X and XII reveal that this is not the case. Thus, the more probable structure for these compounds is 24. The reason for the singlet -AsMe2 nmr absorption i s not clear especially in the light of the results for complexes of configuration 40. However, inspection of molecular models does indicate a similar environment for each Me group in 24, hence the expected two Me resonances may be degenerate. B. (L-L) C(PhC = CPh)Co 2(C0) 4 The ligand fgfos reacts to yield complex XL, f 6fos(PhC = CPh)Co 2(C0) 4, with four CO bands at 2056, 2004, 1982 and 1963 cm \ similar to f,fars(PhC 5 CH)Coo(C0), discussed before, hence o Z 4. 40 or 41_ can be proposed for complex XL. But, i n contrast to the 19 fgfars(PhC = CH)Co 2(C0) 4 case, the F nmr spectrum of XL shows three groups of resonances at 109.4, 110.5 and 134.8 ppm (intensity ratio 1:1:1) indicating that the ligand i s coordinated less symmetrically. Hence 41^  is favored for this complex. C. (L-L) (PhC = CPh)Co2(CO)4 The ligand f„fos gives complex XLII, f Qfos(PhC = CPh)Co„(CO)., o O Z 4 with four CO bands similar to those of f,fars(PhC = CH)Co„(C0). and O Z 4 fgfars(PhC = CPh)Co 2(C0) 4 > hence 40 and 41 are possible for XLII. - 143 -TABLE XI. Infrared Carbonyl Bands of (L-L)(RCECR')CO 2(CO) 4 complexes Compound v (cm - 1) f 4fars(PhCECH)Co 2(CO) 4"" 2036(9), 2001(10), 1978(10), 1955(3) f 4AsP(PhCECH)Co 2(CO) 4 2037(8), 2006(10), 1982(9), 1962(3) f 4fos(PhCECH)Co 2(C0) 4 2036(8), 2006(10), 1982(9), 1962(3) f 6fars(PhCECH)Co 2(CO) 2052(9), 2003(10), 1974(8), 1959(8) f 6AsP(PhC=CH)Co 2(CO) 4 2055(10), 2004(10), 1990(3), 1974(8), 1958(5) f gfars(PhCECH)Co 2(CO) 4 2050(9), 2002(10), 1989(7), 1967(3) fgfos(PhCECH)Co 2(CO) 2050(9), 2002(10), 1989(7), 1967(3) dab(PhC=CH)Co 2(CO) 4 2057(6), 2022(10), 1986(3) f 4fars(PhCECPh)Co 2(CO) 4 2035(8), 2002(10), 1983(9), 1959(5) f 4AsP(PhCECPb)Co 2(CO) 4 2033(7), 2004(10), 1981(8), 1961(2) f 4fos(PhCECPh)Co 2(C0) 4 2038(7), 2013(10), 1987(7), 1967(2) fgfars(PhCECPh)Co 2(CO) 4 2032(8), 2003(10), 1980(9), 1958(5) fgAsP(PhCECPh)Co 2(C0) 4 2054(7), 2034(7), 2005(10), 1983(8), 1969(4) fgfos(PhCECPh)Co 2(C0) 4 2056(10), 2004(9), 1982(7), 1963(2) fgfars(PhCECPh)Co 2(CO) 2054(10), 2005(10), 2001(10), 1981(8), 1973(9), 1955(9) fgfos(PhCECPh)Co 2(CO) ? 2051(10), 2004(9), 1984(8), 1964(3) dab(PhCECPh)Co 2(C0) 4 2052(10), 2005(10), 1999(10), 1972(9), 1957(9) dppm(PhCECPh)Co 2(C0) 4 2029(8), 1993(10), 1980(9), 1960(3) dppe(PhCECPh)Co 2(C0) 4 2031(8), 1996(10), 1975(9.5), 1954(4) dppp(PhCECPh)Co 2(C0) 4 § 2036(8), 1975(10) (fgfars) 2(PhC=CPh)Co 2(CO) 2 1951(7), 1944(10) (dab) 2(PhCECPh)Co 2(C0) 2 § 1922(10), 1898(10) (f 4fos) 2(PhCECPh)Co 2(CO) 2 1928(10), 1913(10) cyclohexane; ** From r e f . 13;; § CH 2C1 2 - 144 -19 However, the F nmr spectrum shows two absorptions at 101.5 and 133.5 ppm , very similar to that of the free ligand, indicating that the ligand is symmetrically coordinated. Hence 40 i s preferred for XLII. The ligand fgAsP gives XXXIX, fgAsPCPhC = CPh)Co 2(C0) 4 > which shows five CO bands at 2054, 2034, 2005, 1983, and 1969 cm"1. The bands at 2054 and 2034 cm 1 indicate that this compound contains two isomers of possible structures 24_ and 40 (or 41), respectively. The 1 19 complex pattern of H and F nmr spectra of this compound further supports this conclusion. The ligands f_fars and dab give XLI, f_fars(PhC = CPh)Co_(C0)., o O Z 4 and dab(PhC = CPh)Co 2(C0) 4, respectively, with i r CO bands very similar to f 6fos(PhC = CPh)Co 2(C0) 4 and f 6fars(PhC = CH)Co2(CO)4, but revealing more than the anticipated number of CO stretching frequencies that one would predict from group theory. , This indicates that there are probably two isomers of possible configurations 40 and 4J_ present in these solids. 19 However, the F nmr spectra of these complexes show that the ligand i s symmetrically coordinated i n each case. The above two seemingly opposite conclusions can be satisfactorily explained i f we assume that the geometries of these complexes are not rigid in solution; i.e. there i s a rapid exchange between 4^0_ and 41, with a rate much faster than the nmr time scale, but slower than the i r time scale, so that i r can "see" two isomers while the nmr can only detect the average signals. - 145 -D. (L-L) b'(PhC = CPh)Co 2(CO) 4 The ligand dppp reacts to form a black solid XLVI with two terminal CO bands which are different from a l l the above mentioned (L-L)(RC E CR')Co2(CO)4 complexes, but similar to those of [(MeO)3P]2(RC E CR)Co 2(CO) 4 (R = Ph, CH2OH) complexes (cf. Tables X and XI). Moreover, the bite of this ligand, dppp, is probably greater than any other ligands (L-L) studied here, i t is possible that XLVI has the configuration 42. 12. (L-L) 2(PhC E CPh)Co2(CO)2 Some of the above mentioned (L-L)(PhC = CPh)Co2(CO)4 complexes undergo further substitution reactions. Thus, f 4 f a r s reacts with f 4fars(PhC = CPh)Co2(CO)4 (of probable structure 24) to give a dark-green -1 19 solid XLVII with two CO bands at 1951 and 1944 cm . Its F nmr spectrum shows a singlet at 107.2 ppm , similar to that of the free ligand, suggesting that both ligand molecules are coordinated in the similar way or symmetrically. Hence 43 (R = R' = Ph) can be suggested for XLVII. 43 - 146 -Similarly, the ligand f^fos reacts with i t s (PhC = CPh)Co2(CO)4 complex XXXVII (of probable configuration 24) to yield XLIX, (f 4fos) 2(PhC = CPh)Co2(CO)2, with two CO bands at 1928 and 1914 cm"1. 19 Its F nmr spectrum indicates that the two f^fos molecules are symmetrically coordinated, hence 4_3 can be assigned to XLIX. This assignment is supported by the conversion of XLIX back to XXXVII by reaction with CO. The ligand dab reacts with (dab)(PhC = CPh)Co 2(C0) 4, XLIII (of probable geometry 40) to afford a black product, XLVIII, with two CO bands at 1924 and 1900 cm Its "*"H nmr spectrum shows two equal intensity singlets at 1.45 and 1.80 ppm , similar to that of XLIII, and 19 i t s F nmr spectrum shows a singlet at 51.3 ppm , again very similar to that of XLIII. The data indicate that the two ligand molecules in XLIII are identical. The most probable structure for this complex is 44 (R = R' = Ph) : 44 - 147 -The same complex had not reacted with CO in benzene solution at 50° after 2 days. This result is unexpected because other tetra-substituted derivatives (f^fos^CPhC E CPh)Co2(CO)2 and [(MeO)3P]4(PhC E CPh)Co 2(C0) 2 a l l react easily (in 2 hr.) with CO to give the di-substituted complexes XXXVII and XXXI, respectively. It is possible that the small bite of dab result in the formation of very stable chelate complex. - 148 -IV. General Discussion 1. Factors A f f e c t i n g the Formation of Cobalt Carbonyl Complexes There are many factors which can a f f e c t the formation of cobalt carbonyl complexes. These include both the reaction conditions and the nature of the ligand employed. A. The Nature of the Donor Atom of the Ligand The products of the reaction between Lev/is bases and C o 2 ( C 0 ) g depend very much on the hardness of the ligands. Thus, nitrogen and oxygen 2+ + bases, which are hard bases and can s t a b i l i z e the cations L,Co or L rCo o 5 (where L i s a nitrogen or oxygen containing ligand), favor the formation of these i o n i c compounds. Phosphites, being s o f t bases, i n general, prefer the formation of substituted [LCo (C0).j ]2 complexes, whereas phosphines, atsines and s t i b i n e s , which have intermediate hardness, can form both i o n i c L 2Co(CO) 3Co(CO)~ or [ ( L - L ) 3 C o 2 ( C O ) 4 ] 2 + [ C o ( C O ) ~ ] 2 and substituted [LCo(CO) 3] 2 or (L-L)Co 2(CO) 6 complexes, Most [R 3ECo(CO) 3]2 (E = P, As, Sb) complexes tautomerize i n s o l u t i o n to give a bridged and non-bridged mixture. The proportion of the bridged form i n general increases along the series P < As < Sb. Since the Co-E bond length increases i n the series : Co-P < Co-As < Co-Sb, i n t e r -ligand s t e r i c i n t e r a c t i o n s w i l l decreases as the atomic number of E increases. - 149 -B. The Nature of the Substituents on the Donor Atom of the Ligand Electron withdrawing groups such as alkoxyl, halogen or fluoro-carbon groups w i l l decrease the cr-donation of the ligand, whereas electron donating groups such as alkyl groups w i l l increase this. Thus, due to the presence of electron withdrawing OR groups the phosphites (RO).jP are softer than ordinary phosphines R^ P which do not contain electron withdraw-ing substituents. For this reason, the former ligands give the substituted [LCo(CO).j]2 products, while the latter ligands, in general, yield both ionic L2Co(CO)3"Co(CO)4 and substituted [LCo(CO>3]2 compounds. The steric effect of the substituent is also important. For instance, the bridged [R 3AsCo(CO) 3] 2 tautomers are found to decrease in importance in the order Et > Pr 1. This is because the bulkier Pr^As repel each other more in the bridged form. C. The Bite of the Ligand This is defined as the distance between the two donor atoms of the ligand in the complex. In general, ligands with large bites favor the formation of bridged complexes, whereas the ones with smaller bites prefer to form chelate complexes. Thus, since the bite of the fluorocarbon-bridged ligands decreases i n the order : f.fars > f.AsP > f.fos > f,fars > f,AsP > 4 4 4 6 6 fgfos > dab > fgfars > fgfos (Table VIII of Chapter 1), i t is not surpri-sing to find that fgfars and f^AsP give bridged complexes, e.g. f 4 f a r s Co 2(CO) 6 and f 4AsP (RC = CR*)Co2(CO)4 (R = R1 = Ph; R = Ph, R' = H), while the remaining ligands like f f if° s a n d dab give chelate complexes, - 150 -e.g. f gfos CCo 2(CO) 6 and (dab)2(PhC 5 CPh)Co 2(C0) 2 . The reaction between the ligand Me2As (CH2)nAsMe2 (n = 3, 4, 6, 8) and Co2(C0)g [44] gives three types of complexes. In the preceeding discussion i t was suggested that they can be grouped as follows : (i) [Me 2As(CH 2) 4AsMe 2] bCo 2(CO)g (of possible configuration 2); ( i i ) [Me0As(CH_) AsMe«] Co„(C0), (n = 6, 8) (of possible configuration 3); and z Z n Z z o — ( i i i ) [Me0As(CHj „AsMeo]CCo„(C0)(. (of possible configuration 4). The Z Z j Z Z b — distance spanned by the ligand in the above three complex types decreases in the order 3^  > 2_ > 4^  as would be anticipated from chain length considera-tions. D. Temperature Effect Tertiary phosphines, arsines and stibines react with Co2(C0)g to produce ionic L2Co(CO)^Co(C0)~ as well as substituted [LCo(C0) 3] 2 complexes. The relative amounts of the two mainly depend on the temperature used. At low temperature, formation of the ionic compounds is favored, while at higher temperature this i s converted to the substituted compound. Thus, Ph3As and Ph^Sb form L^Co^O^Co^O)" complexes below 0°, but above 0° these complexes decompose to yield [LCo(C0).j]2 complexes. E. Solvent Effect The reaction between Co2(C0)g and phosphines, arsines and stibines also depends on the solvent used. In polar solvent the proportion of ionic products L^Co(C0) 3Co(C0) 4 increases, while in non-polar solvent the formation of simple substituted product [LCo(C0) o] o is enhanced. - 151 -Furthermore, most of the [LCo(CO).j]2 complexes tautomerize in solution to give non-bridged form. Whereas non-polar solvent favors the bridged form, polar solvent usually increases the proportion of non-bridged form in the equilibrium mixture. F. Effect of CO Pressure Dicobalt octacarbonyl is unstable with respect to tetra cobalt dodecacarbonyl at room temperature and pressure, but, under high CO pressure, i t i s stable with respect to the latter species even at 200°. Thus, at room temperature and pressure most tertiary arsenic ligands react with Co 2(C0) g to afford substituted [LCo(CO) 3] 2 or ionic I^CoCCO^CoCCO)" compounds. At 200 atm. CO pressure and 200°, (AsMe),. reacts with Co2(C0)g to yield an unusual compound As^CoCCO)^. It is predicted that high temperature and pressure reactions of Co o(C0) o w i l l give many new compounds unobtainable under less drastic conditions. G. General Comments It i s interesting to note that a l l the fluorocarbon-bridged ligands studied here do not form any ionic complexes. This result is not unexpected because the presence of an electron withdrawing fluorocarbon-bridge C=C(CF 2) n (n = 2, 3 or 4) in these ligands i s expected to decrease the a-donation and hence the hardness of the ligands. Furthermore, m t b none of these ligands give (L-L) 2Co 2(CO) 4(CO) 2 complexes (L-L acts as monodentate ligand). This i s probably because these ligands are a l l much bulkier than Me„As. - 152 -The symmetrical bidentate ligands fgfars (n = 4, 6, 8), ^ n ^ o s (n = 4, 6, 8) and dab a l l give exclusively bridged (L-L)Co2(CO)g complexes (of possible configurations 1_ or 4). However these ligands do not give [(L-L)Co(CO) 3l 2 c o m P l e x e s (° f possible configuration 8) indicating that they prefer to act as bidentate ligands rather than monodentate ones. In contrast a l l the unsymmetrical ligands f^AsF, f^AsCl, f^AsO, f^AsS, f^AsN, f^PS and f^AsP yield [(L-L)Co(CO) 3] 2 complexes. However, except f^AsP, a l l the other unsymmetrical ligands do not afford (L-L)Co2(C0)g complexes (of possible structure 2^, J3 or 4). The f i r s t two do not because they are essentially monodentate. In four of the remaining ligands the second donor group is one of -OMe, -SMe or -NMe2, none of which form cobalt carbonyl complexes by simple displacement of CO groups. 2. Possible Reaction Paths A. Reaction of (RC = CR')Co2(CO)6 with Mono- and Bi-dentate Ligands In the present work i t has been found that ten types of derivatives can be obtained from (RC E CR')Co2(C0)g complexes, vi z . (i) L(RC = CR')Co2(CO)5 (L = (MeO)3P, R = R' = Ph, C^OH), ( i i ) L 2(RC E CR')Co2(CO)4 (L = (MeO^P, R = R» = Ph, CHgOH) , ( i i i ) L 3(RC = CR')Co2(CO)3 (L = (MeO)3P, R = R« = Ph, CH20H), (iv) L 4(RC = CR')Co2(CO)2 (L = (MeO)3P, R = R' = Ph), - 153 -(v) (L-L) b'(RC = CR')Co2(CO)4 (L-L = dppp, R = R' = Ph), (vi) (L-L) b(RC = CR')Co2(CO)4 (L-L = fgfars, f^AsP, etc. R = R' = Ph; R = Ph, R' = H), (vii) (L-L) b(RC E CR')Co2(CO)2 (L-L = f ^ a r s , f^fos. R = R' = PK), . ( v i i i ) (L-L)°(RC = CR')Co2(CO)4 (L-L = dab, f g f o s , fgfars, etc. R = R' = Ph; R = Ph, R* = H), (ix) (L-L) 2(RC = CR')Co2(CO)2 (1-L = dab, R = R' = Ph) and (x) (L-L) (RC E CR')Co2(CO)4 (L-L = fgfos; R = R' = Ph). Seven of these are previously unknown. The probable routes of formation of these derivatives from (RC = CR')Co„(CO), are set out in L o the reaction scheme in Figure 5. For instance, type (iv) complex is formed via steps (T) to <2); type (v) complex via steps (5) and d) ; type (vii) complex via steps (z) to @) ; type (ix) complex via steps (7) and @ - (O) ; and type (x) complex via steps (j) and (R) . It is assumed that the substitution reactions involve i n i t i a l loss of CO followed by reaction with the incoming ligand [96]. B. Reaction of Co„(CO) 0 With Mono- and Bi-dentate Ligands Nine classes of derivatives of Co2(CO)g have been obtained in this work, namely : (xi) [L4 (L-L) bCp 2(CO) 4] 2 +(X )~ (L = (MeO)3P, Bu^, L-L = dppe, X" = Co(CO)4, BPh"), (xii) (L-L)Co2(CO)^ (L-L = f 4PS), ( x i i i ) [(L-L)Co(CO) 3] 2 (L-L = f 4AsP, f 4PS, f 4AsS, etc.), (xiv) (L-L) 2Co 2(CO) 4 (L-L = f 4 f o s ) , (xv) (L-L) bCo 2(CO) 4(CO) b (L-L = dppm), (xvi) (L-L) CCo 2(CO) 4(CO) b (L-L = f 4 f o s , fgfars, etc.), (xvii) (L-L) bCo 2(CO) 2(CO) b ( L - L - f 4 f o s ) , (xviii) (L-L) ^Co^CO) 2(C0) b (L-L = f 4fos) and (xix) L(L-L)Co 2(CO)^(C0) b (L = (MeO)3P, Bu^P ; - 154 -- 155 -(L-L = f^fos). (xx) [L 2(L-L)Co(CO)]2 +(X~) 2; of which classes (xi), ( x i i ) , ( x i i i ) , (xvii), (xix) and (xx) were previously unknown. Since the ionic and the substituted compounds are formed via different paths, i t is instructive to discuss them separately. (i) The Disproportionation of Co2(CO)g The formation of complexes Co(CO)^ nL*Co(CO) 4 can be possibly explained by a heterolytic cleavage of the Co-Co bond under the influence of an increased charge due to the introduction of a Lewis base L : L \ / L ' °C \ / C \ I + > Co^ Co CO v [LCo L;j [Co(CO)A] L' 1 1 oc X where L' can a l l be both L or CO. It is interesting to note that the number n of ligand L present in the cation depends on the nature of L', possibly upon i t s hardness. Thus, with nitrogen or oxygen bases, the +2 formation of l Q0o complexes passes through the following stage : Co 2(CO) g + L > LCo(CO)4Co(CO)4 . This intermediate has been confirmed by a study of the reaction between Co2(CO)g and alcohol or water [46]. With softer bases, such as R^E (R = alkyl, aryl; E = P, As, Sb) the di-substituted derivatives can be obtained [15]. With even softer bases, such as dppe and ( M e O ) i t i s thus not surprising to find poly-i 3Co 2(CO) 2 +][Co(CO) 4] 2 substituted complexes, e.g. [(L-L) Co,(CO)J+][Co(CO).]_ (L-L = dppe), m, [LnCo(CO)^_n]X (L = (MeO)3P, n = 3—5, X = BPh 4), & to i . - 156 -The flow sheet for this type of reaction i s illustrated i n Figure 6. The formation of type (xi) complex mentioned above i s probably via steps (f|) and (||) or via (28) . ( i i ) Substitution Reaction of Co 2(CO) g Besides the ionic compounds, Co2(CO)g can also give rise to two different types of substitution products, i.e. bridged and non-bridged derivatives. This result is presumably due to the tautomerism of Co2(CO)g in solution : \ / \ / ^ I \ / — Co <- -> Co — ^ — Co Co — ' ^ 7 v ^—: /\ i This equilibrium is solvent- and temperature-dependent. Thus a pentane solution of Co2(CO)g at room temperature contains ca. 55% of bridged and 45% of the non-bridged form. Table XII l i s t s the conditions which favor a particular side of this equilibrium. TABLE XII. Favorable Conditions for the Formation of Bridged and Non-bridged Substituted Derivatives of Coo(C0) 2 o Bridged-Form Non-Bridged From (1) non-polar solvent (1) p o l a r s o l v e n t (2) lower temperature (2) h i g h e r temperature (3) less bulkier ligand (3) m o r e bulkier ligand - 157 -^ C \ / L—Co—Co—L I A L [L6Co] (e) +2 [L 5Co +] CO 28) L L (n) \ / I -Co—Co | / \ "CO /I, ® L-L L (17 (i) L (2 [(LCo(CO) 4][Co(CO) 4] L L CO i2+ L—Co y1 Co—L (xi) CO V L (18 [L Co(CO) ] 4 (h) N M L (20) [L2Co(C0)3"] L > (f) CO [L 3Co(CO) 2] (g) L-L L L—CoC (k) 23, + V L \ 1 ^ —Cod . I (j) CO -1+ L-L L L. L—Co 'Co—L L L CO L-L L„ L (xx) L L —Co Co— r y i L L 2+ 2+ L L—Co / I T L (m) Figure 6. Disproportionation of Co o(C0) Q z o - 158 -Some derivatives of arsines and phosphines, e.g. L 2Co 2(CO)g (L = Et^As, Et.jP, etc.) are also involved i n a solvent- and temperature-dependent equilibrium. However, there are cobalt carbonyl complexes which do not tautomerize in solution. For example [Bu^PCo(CO)^],, a n c* [(RO^PCoCCO)^^ (R = Me, Ph, etc.) complexes exist in solution as exclusively the non-bridged form, while f^farsCo 2(CO)g and other (L-L) Co-tCO)- (L-L = f.fos, f/-fos, f,fars, f_fars, f Q f o s , dab etc.) L o 4 b o o o complexes exist exclusively as the bridged form. The possible paths for the reactions between Co2(CO)g and monodentate L and bidentate L-L ligands are shown in Figure 7. As can c be seen, complex (xiv) , [(L-L) Co(CO) 2] 2 (L-L = f^fos) , is probably formed via steps (30) , ($5) to (35) ; complex (xvii) (L-L)^Co^CO) 2 (L-L = f^fos), probably via steps (4^ , (44) and (45) ; complex ( x v i i i ) , (L-L) 2Co 2(CO) 2 (L-L = f^fos), probably via steps (4l) to (43) ; and complex (xix), L(L-L) CCo 2(CO) 3(C0) 2 (L = (MeO)3P, BunP, L-L = f^fos) , probably via steps @ and (40) or via steps (48) , (||) , (38) and - 160 -References 1. G.G. Summer, H.P.Klug and L.E. Alexander, Acta. Cryst., 17, 732 (1964). 2. K. Noack, Spectrochim. Acta., 19, 1925 (1963); Helv. Chim. Acta., 47_, 1555 (1964). 3. G. Bor, Spectrochim. Acta., 19, 2065 (1963). 4. P. McArdle and A.R. Manning, J. Chem. Soc. (A), 2123 (1970); and references therein. 5. H.J. Keller and H. Wawersik, Z. Naturforsch., 20b, 938 (1965). 6. L. Mond, H. Hirtz, and M.D. Cowap, J. Chem. Soc, 798 (1910); Z. Anorg. Allgem. Chem., 68, 207 (1910). 7. I. Wender, H.W. Sternberg, S. Metlin and M. Orchin, Inorg. Syn., _5, 190 (1957). 8. G. Capron-Cotigny and R. Poilblanc, Bull. Soc. Chim. France, 1440 (1967) 9. A.J. Carty and T.W. Ng, Chem. Comm. 149 (1970). 10. W. Hieber and W. Freyer, Chem. Ber. 93, 462 (1962). 11. A.R. Manning, J. Chem. Soc. (A), 1665 (1968). 12. A.R. Manning, J. Chem. Soc. (A), 1135 (1968). 13. J.P. Crow and W.R. Cullen, Inorg. Chem. 10, 2165 (1971). 14. J.P. Crow, W.R. Cullen, W. Harrison and J. Trotter, J. Am. Chem. Soc. 92, 6339 (1970). 15. J. Newman and A.R. Manning, J. Chem. Soc. (Dalton), 1593 (1973). 16. G. Winkhaus and G. Wilkinson, Chem. and Ind., 1083 (1960); and J. Chem. Soc. 602 (1961). 17. F.S. Stephens, J. Chem. Soc. (Dalton), 1754 (1972). 18. G. Bor and L. Marko', Chem. and Ind. (London), 912 (1963). 19. P. McArdle and A.R. Manning, Chem. Comm. 1020 (1968). 20. P. McArdle, A.R. Manning and F.S. Stephens, Chem. Comm., 318 (1969). - 161 -21. F.S. Stephens, J. Chem. Soc. (A), 2745 (1970). 22. F.S. Stephens, J. Chem. Soc. (Dalton), 1752 (1972). 23. von H. Behrens and W. Aquila, Z. Anorg. Allgem. Chem., 356, 8 (1967). 24. C. Pegot and M. Poilblanc, Compt. rend. 268c, 955 (1969). 25. G. Capron-Cotigny, These Zeme' Cycle, Paris, 1967. 26. A. Sacco, Ann. Chim. 43, 495 (1953); C.A., 48, 5012 (1954). 27. R.B. King, Inorg. Chem., 2, 936 (1963). 28. A.T.T. Hsieh and G. Wilkinson, J. Chem. Soc. (Dalton), 867 (1973). 29. J.A. Ibers, J. Organometal. Chem. 14, 423 (1968). 30. R.F. Bryan and A.R. Manning, Chem. Comm. B16 (1968). 31. J.A.M. Case, Ph.D. Thesis, University of Wisconsin (1967). 32. A.S. Foust, Jr., Ph.D. Thesis, University of Wisconsin (1970). 33. Private communication from A. Sacco (1973). 34. W.R. Cullen, D.A. Harbourne, B.V. Liengme and J.R. Sams, Inorg. Chem. _8, 95 (1969). 35. R.G. Hayter, J. Am. Chem. Soc, 86, 823 (1964). 36. R.C. Dobbie and D. Whittaker, J. Chem. Soc. (Dalton), 2427 (1973) 37. W. Hieber and P. Spacu, Z. Anorg. Allgem. Chem., 233, 353 (1937). 38. W. Hieber and W. Freyer, Chem. Ber. 91, 1230 (1958). 39. S. A t t a l i and R. Poilblanc, Inorganica Chim. Acta., 475 (1972). 40. A. Sacco, Gazz. Chim. I t a l . , J33, 632 (1953). 41. W. Hieber and Sedlmeier, Chem. Ber., 87_, 789 (1954). 42. A. Socc, Gazz. Chim. Ind. 93, 698 (1963). 43. R.L. Peterson and K.L. Watters, Inorg. Chem. 12_, 3009 (1973). 44. D.J. Thornhill and A.R. Manning, J. Chem. Soc. (Dalton), 2086 (1973) 45. I. Wender, H.W. Sternberg and M. Orchin, J. Am. Chem. Soc, 74. 1216 (1952). - 162 -46. E.R. Tucci and B.H. Gwynn, J. Am. Chem. Soc, 86, 4838 (1964). 47. L. Marko', G. Bor. and G. Almasy, Chem. Ber. 94_, 847 (1961). 48. L. Marko', G. Bor. and E. Klumpp, Chem. Ind. (london), 1491 (1961). 49. L. Marko', G. Bor., E. Klumpp, B. Marko' and G. Almasy', Chem. Ber. 96, 955 (1963). 50. L. Marko' and G. Bor., Angero. Chem. 7_5, 248 (1963). 51. L.F. Dahl and D.L. Smith, J. Am. Chem. Soc, 84, 2450 (1962). 52. E. Klumpp, L. Marko' and G. Bor., Chem. Ber. 97, 926 (1964). 53. CH. Wei and L.F. Dahl, Inorg. Chem. 6, 1229 (1967). 54. W.R. Cullen, "Adv. i n Inorg. and Radiochem.", Vol. 15, 368 (1972). 55. H.W. Sternberg, H. Greenfield, R.A. Friedel, J.H. Wotiz, R. Markby and I. Wender, J. Am Chem. Soc, _76, 1457 (1954). 56. I. Greenfield, H.W. Sternberg, R.A. Friedel, J.H. Wotiz, R. Markby and I. Wender, J. Am. Chem. Soc, _78, 120 (1956). 57. H.W. Sternberg and I Wender, International Conference on Coordination Chemistry, London, April 1959, Spec. Publ. No. 13, The Chemical Society, Burlington House, London, 1959, p. 35. 58. M.R. Tipak, CA. Hollingsworth and J.H. Wotiz, J. Org. Chem., 25, 687 (1960). 59. M.R. Tipak, J.H. Wotiz and CA. Hollingsworth, J. Am. Chem. Soc, 82, 4265 (1958). 60. M. Almasi, L. Szabi', I. Farkas and T. Bota, Acad. Rep. Populare Romine, Studii Cercetari Chim., 8, 495 (1960); Chem. Abstr. 55, 19427e (1961). 61. U. Kriierke and W. HUbel, Chem. Ber., 96, 2829 (1961). 62. J.L. Boston, D.W.A. Sharp and G. Wilkinson, J. Chem. Soc, 3488 (1962). 63. H.W. Sternberg, J.G. Shukys, C Delle Donne, R. Markby, R.A. Friedel and I. Wender, J. Am. Chem. Soc, 81, 2339 (1959). 64. W. Hubel and C. Hogzand, Chem. Ber., 93, 103 (1960). 65. G. Peyronel, A. Ragni and E.F. Trogu, Gazz. Chim. I t a l . , 92, 738 (1962). - 163 -66. R. Markby, I. Wender, R.A. Friedel, F.A. Cotton abd H.W. Sternberg, J. Am. Chem. Soc, 80, 6529 (1958). 67. G. Peyronel, A. Ragnl and E.F. Trogu, A t t i Soc Nat. Mat. Modena, 91, 100 (1960). 68. G. Bor , Chem. Ber., 96, 2644 (1963). 69. W. Hubel and R. Merenyi, Chem. Ber., 9(5, 930 (1963). 70. 0. Vohler, Chem. Ber., 91, 1161 (1958). 71. C. Hoogzand and W. Hubel, Tetrahedron Letters, 18, 637 (1961). 72. E.M. Arnett, M.E. Strem and R.A. Friedel, Tetrahedron Letters, 19, 658 (1961). 73. N.A. Bailey, M.R. Churchill, R. Hunt, R. Mason and G. Wilkinson, Proc Chem. Soc, 401 (1964). 74. R.L. Hunt and G. Wilkinson, Inorg. Chem., 4^, 1270 (1965). 75. W.G. Sly, J. Am. Chem. Soc, 81, 18 (1959). 76. J.L. Boston, D.W.A. Sharp and G. Wilkinson, Chem. Ind. (London), 1137 (1960). 77. P.H. Bird and A.R. Fraser, Chem. Comm., 681 (1970). 78. H.A. Patel, A.J. Carty and N.K. Hota, J. Organometal. Chem., 50, 247 (1973). 79. N.K. Hota, H.A. Patel and A.J. Carty, J. Organometal. Chem., 3_2, C55 (1971) . 80. R.D. Feltham and W. Silverthorn, Inorg. Syntheses, Vol. 10, 159 (1967) 81. D. Wittenberg and H. Silman, J. Org. Chem., 23_, 1063 (1958). 82. W.R. Cullen, P.S. Dhaliwal and G.E. Styan, J. Organometal. Chem., 6, 633 (1966). , ^ 83. W.R. Cullen, D.S. Dawson and P.S. Dhaliwal, Can. J. Chem., 45, 683 (1967). 84. W.R. Cullen, D.F. Dong and J.A.J. Thompson, Can. J. Chem., 47, 4671 (1969) . 85. L.S. Chia and W.R. Cullen, Can. J. Chem., 50_, 1421 (1972). - 164 -86. (a) W.R. Cullen, P.S. Dhaliwal and G.E. Styan, J. Organometal. Chem. 6, 364 (1966). (b) W.R. Cullen, D.S. Dawson, P.S. Dhaliwal and G.E. Styan, Chem. Ind. (London), 502 (1964). 87. W. Hieber, F. Muhlbauer and E.A. Ehmann, Chem. Ber., 65_, 1090 (1932). 88. P.S. Dhaliwal, Ph.D. Thesis, University of British Columbia (1966), p. 39. 89. J.P. Crow, W.R. Cullen and F.L. Hou, Inorg. Chem., 11, 2125 (1972). 90. (a) L.S. Chia, W.R. Cullen and D.A. Harbourne, Can. J. Chem., 50, 2182 (1972) ; and references therein. (b) W.R. Cullen and L.M. Mihichuk, Can. J, Chem., 51, 936 (1973). 91. F.W.B. Einstein, R.D.G. Jones and R. Kirkland, to be published. 92. (a) D.J. Patmore and W.A.G. Graham, Inorg. Chem., 6_, 981 (1967); (b) ibid, 5_, 2222 (1966). 93. R.K. Harris, Can. J. Chem., 42, 2275 (1964). 94. R.K. Harris, Inorg. Chem., 5_, 701 (1966). 95. J.M. Jenkins, J.R. Moss and B.L. Shaw, J. Chem. Soc. (A), 2796 (1969). 96. R.J. Angelici, Organometallic Chem. Rev. _3, 173 (1968). - 165 -Chapter 3 Introduction to Nuclear Quadrupole Resonance Spectroscopy I Theory of Nuclear Quadrupole Resonance 1. Introduction Nuclear Quadrupole Resonance (nqr) i s a branch of radio frequency spectroscopy which does not require a magnet to detect the phenomenon. This effect, f i r s t observed in 1949 [1], arises from the interaction of the nuclear quadrupole moment (for nuclei with I >_ 1) with an asymmetric electric f i e l d gradient, which causes a partial l i f t i n g of the (21 + 1) degeneracy of the spin-orientation quadrupole energy levels. The Hamiltonian operator, H^ , for this interaction i s [2] : V 4 I ( 2 i q - 1) P ^ i ' ^ d 2 , * ! 2 ) ] (1) where e = electrostatic charge i n esu, Q = the nuclear quadrupole moment, q = the electric f i e l d gradient, I = the nuclear spin quantum number, I = the nuclear spin angular momentum operator*, I^ = the projection of I along the Z principal axis of the electric f i e l d gradient tensor*, I + = raising or step-up operator, I = lowering or step-down operator, D = the asymmetry parameter of the f i e l d gradient. "2 Since equation (1) involves 1? , i t follows that in an axially The eigenvalue of I 2 is 1(1+1), and that of I 2 is m 2, where m^  is the magnetic quantum number. - 166 -symmetric f i e l d the quadrupole energy levels w i l l be doubly degenerate with respect to m^  (except the energy level due to = 0). 2. The Nuclear Quadrupole Moment The scalar quadrupole moment of the nucleus i s defined by [3] : Q = j P k r 2 ( 3 c o s 2 9 K - 1) dx k (2) where i s the charge density in a small volume element dx^, inside the nucleus at a distance r ^ from the center and 8^ is the angle which the radius vector r ^ makes with the nuclear spin axis, the nucleus being in the spin state defined by the nuclear magnetic quantum number m^ . Q i s a measure of the deviation of the nuclear charge from spherical symmetry. In Figure 1, the two possible types of distorted nuclei are shown. A prolate nucleus i s elongated along the spin axis, while an oblate nucleus spin axis Prolate (Q > 0) spin axis C Oblate (Q < 0) Figure 1. Two Possible Types of Distorted Nuclei - 167 -is flattened along the spin axis. A nucleus having spherical symmetry would thus have Q = 0. The quadrupole moments for the nuclei studied here are listed i n Table I. TABLE I Quadrupole Moments of Nuclei Investigated Here Nucleus Q (barns) Ref. 3 5 c i - 0.0797 [4] 3 7 c i - 0.0621 [4] 5 9Co 0.4044 [5] 3. The Electric Field Gradient (EFG) The efg is a tensor quantity which describes the distribution of electric charge about the nucleus and is expressed in matrix form as follows : efg =-V V V XX xy xz V V V yx yy yz V V V zx zy ZZ where V. . = - i - 5 L (3) V is the potential at the nucleus arising from a l l external charges; i = x, y, z; j = x, y, z. This tensor is symmetric (i.e. = V ^ ) . Furthermore, since we deal only with the potential arising from charges - 168 -external to the nucleus, Laplace's equation applies and V + V + V =0, xx yy zz Though x, y, z can be any arbitrarily chosen mutually perpendicular set of axes, i t i s most convenient to choose the so-called "principal" axis system, for which a l l off-diagonal elements of the tensor vanish. The axes of the efg coordinate system are then labelled so that |V I > |V I > |V I (4) i zz — yy 1 — 1 xx1 ' The efg can then be specified in terms of two independent parameters : eq - V z z (5) V - V • n = _J5E_ XX w i t h o < n < 1 (6) zz If V = V ^ V , the efg has axial symmetry about the z-axis, and xx yy zz 1 J . ' n = 0. I f V ^ V ^ V , then the efg is asymmetric and ri ^  0. A xx yy zz & J special case occurs when V = V = V ; this arises most commonly when xx yy zz J the efg has tetrahedral, octahedral, or cubic symmetry. A l l orientations of the quadrupole moment with the efg have the same energy; consequently the degeneracy of the energy levels is not removed and no resonance w i l l occur. 4. Quadrupole Energy Levels 2 The quantity e Qq in equation (1) is often called the nuclear quadrupole coupling constant. In.practice, this i s directly related to the 3 observed nqr frequency. Furthermore, for a l l systems, except I = y , n may be obtained from the resonance frequencies. The presence of a nonzero - 169 -n leads to the mixing of states with Am^. = + 2 into the overall interaction. The matrix has elements : H , = I ^  * H ip , dx (7) mm' J I,m1 Q I.mJ. v • where the basis functions IJJT and , are the nuclear spin functions, If m = m^. , then the element i s on the diagonal. The resulting matrix may be diagonalized to obtain the secular equations for spin I. For half-1 integer spins these secular equations can be expressed as I + y order polynomials in energy and involve n in the coefficients. 35 37 3 Both Cl and Cl have I = y , so the energy i s given by 2 2 2 e 0 a quadratic equation [6] : E - 3n - 9 = 0 in units of 41(21 - 1) 1 2 or • The solutions of this equation are i o f 2>l/2 E+ | " i e 2 Q q ( l + 5 -1 2, E l = " I e" Qq ±2 1/2 (8) (9) 3 The energy level diagram for I = both with and without the presence of n i s shown i n Figure 2. As can be seen, since only two energy levels occur, only one resonance frequency can appear. From this single frequency the two unknown 2 n and e Qq cannot be determined separately*. However, since chlorine * 3 When I = -, the n may be determined by nqr line splitting in a weak magnetic f i e l d (the Zeeman effect). - 170 -1 2rt 4 e Qq 2„ e Qq AE = v = » 0 \ N m i = ± 2 = 2 2h 1 2 A 4 e Qq n - o e_Qq. 4 2^  1/2 1 + AE = i s n i - 2 l l / 2 2 1 + 2h r 2-1 + f 1/2 e Qq 1 + f 1/2 n * o e Qq > 0 Figure 2. Quadrupole Energy Levels and Transitions for I = -|. - 171 -atoms are usually involved i n only one nominally single bond, n w i l l usually be less than 0.2. In this case, the error caused by assuming 35 2 that Cl e Qq is equal to 2h x (observed frequency) is less than 35 2 0.7%. Thus, to a good approximation, for Cl, e Qq can be obtained simply by doubling the observed nqr frequency. However, the sign of 2 e Qq i s unobtainable from an nqr experiment. 59 For Co, the energy levels can be obtained by solution of the following fourth order polynomial [6] : 2<> „ . . 2 ^ 2 4 E - 42 1 + 3 - E 2 - 64(1 - n 2 ) E + 105 1 + ^ - 1 =0 41(21 - 1) o r 28 e~ Q <*' <10) in u n i t s o f -*L3a _ L. 2 The quadrupole energies can be obtained as the roots of this equation. 2 For small n these can be expressed as a power series in n . On the other hand, Cohen [7] has solved the equation numerically for n = 0, 0.1, 0.2, 1.0. These roots are given for n = 0 and n = 1.0 in Figure 3. A simpler method, however, of obtaining the energy levels for any arbitrary value of n is to bypass equation (10), and instead set up and diagonalize the matrix of H^, using the elements of equation (7). A computer program for this purpose was written* (Appendix I), and used in the present work. If n = 0, the basis functions tyT are also the eigen-functions of H^. Magnetic dipole transitions can be induced between the Professor M.C.L. Gerry is thanked for his assistance in writing this program. - 172 -1 2 Qq-i v 3 = i i h e 2 ^ * = 7 . 2 4 1 5 7 ^ v , = 0.19125^-2^ h 5 V 1 2„ ' ' I = 2 \ v e Qq = 0 ^ ^ \ \ 2 2 V2 = lAh* Q« \ \ v 3 * 3 2 -2" \ \ K ~ 28 s 1 2 V l " Uh* 1 28"V Qq Qq n = 0 e Qq >0 % 1.88669 28 .v_ = 0.13476^4& ! i * 2 n - - 1 . 8 8 6 6 9 ^ 1 28 v. = 0.19125^-31 -1- h £ 2 - 7 . 2 4 1 5 7 ^ 3 -28 h = l . o Figure 3. Quadrupole Energy Levels and Transitions for I = -2 - 173 -four levels, having frequency ratio 1 : 2 : 3 . These are shown in Figure 3, and labelled as v^, v 2, v^, respectively*. This ratio i s not obtained i f r) 4 0; for example, when n = 1, the ratio becomes v l : v2 : v3 = 1 - 4 1 9 1 2 : 1 : 1-41912 . These are also shown in Figure 3 . In Figure 4 the three transition frequencies are plotted as a function of n . and axe only moderately affected by n, but is very sensitive and in fact, i s actually greater than for D > 0.586. 2 From the three frequencies the two unknown n and e Qq can be evaluated and easily checked. The transitions labelled v ^ , v^, are the strongest lines for a l l values of n . When n ^ 0, however, m^. i s no longer a good quantum number, and the eigenfunctions are linear combination of the \p As a result, two further "forbidden" transitions become weakly allowed; these we c a l l f^ and f2» and are related to v^, v 2 , as follows : f l = V l + V2 < U> f 2 = v 2 + v 3 (12) We w i l l always consider v 1 as the transition frequency between the lowest two levels, v 2 as the frequency between the second lowest and third lowest levels, and as the frequency between the top two levels. - 174 -Tr a n s i t i o n Frequencies ( i n e"Qq) 0.20 h 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Figure 4. The T r a n s i t i o n Frequencies as a function of the Asymmetry Parameter for I = - 175 -Hence, in general, ± 2 is greater than f . Table II l i s t s the relative intensities of the three allowed and two "forbidden" frequencies for An = 0.1 in the complete range of n •• Except for n = 1, is always greater than (Figure 4). ^ r e   n .TABLE II Averaged Transition Probabilities, W, of various Transitions for I = | [7]. n v x v 2 v 3 f1 f 2 0.1 4.77 3.98 2.33 0.03 0.2 4.26 3.93 2.34 0.10 0.3 3.71 3.90 2.34 0.17 0.01 0.4 3.27 3.89 2.35 0.22 0.02 0.5 2.96 3.92 2.35 0.24 0.03 0.6 2.74 3.96 2.36 0.24 0.04 0.7 2.60 4.01 2.37 0.23 0.07 0.8 2.52 4.05 2.38 0.21 0.09 0.9 2.46 4.08 2.40 0.18 0.12 1.0 2.42 4.08 2.42 0.15 0.15 - 176 -5. Factors Affecting the EFG In the solid state, the efg tensor i s affected by both intramolecular and lattice effects. The intramolecular effects arise from (a) the valence electrons in the atom, (b) distortions of the core electrons in the atom by i t s valence electron and (c) other intramole-cular charges distribution. The lattice effects include (d) the extra-molecular charges present i n the lat t i c e , such as ions, other nuclei, and electrons i n neighboring atoms and (e) polarization of the valence and core electrons distribution of the atom by the charges i n the la t t i c e . The lattice efg in ionic solids can be estimated from a point-charge la t t i c e sum [ 8 ] , but i t i s not clear how such an estimate should be obtained for molecular solids such as those.studied in this work. 6. The Intramolecular EFG Only the partially f i l l e d p and d orbitals give rise to the efg. The s-orbitals and f i l l e d p and d orbitals are spherically symmetric and cannot directly contribute to the efg. The molecular efg is computed by considering the variations in the populations of the valence p and d orbitals in the atom and can be simply related to the atomic coupling constant by a variable parameter X [6, 9] q - = Xq- (13) THOI. at. A workable expression for X must be developed. - 177 -For chlorine, the valence orbitals are 3s, 3p , 3p and x ry 3p z. Based on the work of Townes and Dailey [9], the following expression can be derived for chlorine atom [6, 10, 11] : where q %iol. q a t . 1 N - + N } n Z K Tl n J X at. 4 = ~5 j * i ~~f ^ i d t f o r p o r b i t a l s and Nr (14) represents the orbital population of the p^ orbital of the chlorine atom. X i n equation (13) has now been defined. Converting equation (14) to a coupling constant gives h mol. h at. N - i f N + N ) p 2 p p ' v2. *x F y (15) where 2n 35 = 109.7 MHz for Cl [4]. The z-axis is taken to be along the M-Cl bond (where M i s the atom bonded to the Cl atom). For organic compounds, such as those considered in this thesis, 2 (e Qq/h) m oi ^ s always negative [12]. The sign i s , however, not deter-mined by nqr measurements (though i t is by microwave spectroscopy). If the Cl atom takes part only in a bonding, then N D = = 2, but x • y N varies according to the nature of the bond (i.e. the hybridization z at chlorine and the ionic character of the bond). The greater the magnitude of N n (i.e. the higher sp hybridization and ionic character 2 of the bond), the smaller i s |(e Q q / h ) m o l I a n d hence v . The p a r t i c i -pation of unshared electron pairs ; of a Cl atom (by ir-back donation) in the formation of a TT bond decreases Np and/or N p and also decreases l ( e 2 ^ / h ) m o l J ' - 178 -59. For Co the problem is somewhat different because both the valence 4p and 3d orbitals are important. However, a similar type of equation is obtainable [10] : where W . = q d , at. d 2 2 z Nd + N d - Nd " N d xz yz-1 xy * 2 2 x. -y + q P. at. ( „ 1 f ) 1 N - N + N 2 P z x yJ (16) q, ^ = — I A * d,at. 7 J * * " i d x Harris [13, 14] has shown that for many cobalt complexes, including some similar to those considered in this thesis, the second! term in equation (16) (due to the 4p electrons) i s very much smaller than the f i r s t (due to the 3d electrons), and can reasonably be ignored in many cases. The coupling constant can be written : h |mol. I h I d , a t . d 2 2 z N d + N d v xz yz' ~ N d " N d , , xy 2 2 x -y J where d.at.; - " 2 7 8 - 7 m z W (17) 59 The asymmetry parameter for Co may be obtained from the population variation i n the d x and d orbitals [10] : X y x -y fe2Qq 1 XX r 2„ 1 e Qq / h mol. h J d,at. 1 N d +1 2 d 2 2 2 x -y Nd + N d xy xz 2 N d " Nd 2 yz J z ] (18) -179 -r 2„ i e Qq fe2Qql h mol. h d,at. 2 d 2 2 2 x -y N d + N d v xy yz 4 N d - N d ' 2 xz z (19) Substitution of equations (17) to (19) into equation (6) gives N d " N d xz yz' n = N d + N d ' N + X Z V 2 - N — N dz 2 2 dxy d 2 2 z x -y (20) II Instrumentation and Procedure !• The Spectrometer NQR transitions between different m^. states are induced by radio frequency radiation. The following i s a schematic diagram of nqr spectrometer [6]. Power from a radio frequency oscillator (SRO) an xs A L T Tr< I SRO Chart Recorder Amplifier Detector Figure 5. Schematic Diagram of an NQR Spectrometer - 180 -supplied to the circuit LCC'. The sample is placed in the volume of an inductance L which i s tuned to the transition frequency by means of a capacitor C. By using the LC circuit as the oscillating element with electronic feed back, the voltage level of oscillation becomes a function of nuclear absorption. The applied frequency is swept by varying the small capacitance C. The resultant output voltage is then detected, amplified and recorded. A l l nqr spectra reported in this thesis were recorded on a Decca nqr spectrometer. This spectrometer consists of nine units ( (1) power supplies, (2) Zeeman Modulator, (3) modulation and phase sensitive detector, (4) quench generator, (5) automatic gain control, (6) frequency marker, (7) dewar housing and Zeeman Modulation c o i l , (8) main cabinet and (9) SRO extended frequency range ) plus a chart recorder, an external attenuator, and a set of coils to cover the frequency range 7-90 MHz. (Figure 6). 2. Factors Affecting the Signal Strength The strength of nqr signals i s , in general governed by the following factors. (a) Crystallinity : Crystallinity is important because strains within the crystal may broaden the signal and certain kinds of disorder can remove the signal. Annealing crystals at temperatures near the melting point can sometimes help. However, occasionally, the reverse effect i s observed. A.6.C. OFF INPUT QUENCH GENERATOR QUENCH LOCK SIGNAL FREQUENCY MARKER QUENCH SIGNAL QUENCH LOCK SIGNAL ZEEMAN MODULATOR EXT INPUT RESONANCE SIGNAL _ i _ Z MODULATION SIGNAL RELAY VOLTAGE MOD&RSJD. F.M. SIGNAL NQR. SIGNAL MARKER A SIGNAL MARKER B SGNAL MARKER SUPER REGENERATIVE OSCILLATOR Z MODULATION SIGNAL ZEEMAN MODULATOR CABINET POWER SUPPLY TO ALL UNITS MAGNETIC MODULATION N.O.R. SIGNAL PEN RECORDER SIGNAL PSaLLOSCOPE CABINET Figure 6. Decca NQR Spectrometer Block Diagram - 182 -It i s therefore best to perform frequency scans before and after annealing. (b) Purity : Purity i s also important, since defects in the crystal, p a r t i -cularly impurities present in solid solution can reduce the signal intensity. (c) Temperature : Temperature gradients across the sample may produce a serious broadening, since temperature coefficient in the order of -3KHZ/deg. are common. (d) F i l l i n g Factor : The proportion of the total magnetic energy of the oscillator tuned ci r c u i t , which i s within the nqr sample i s called the f i l l i n g factor. In general, the intensity of a signal i s proportional to the f i l l i n g factor of the sample. 3. Experimental Techniques Samples used in this investigation were repeatedly purified u n t i l they are spectroscopic pure (checked by infrared and nmr). Each sample was sealed under a nitrogen atmosphere in a sample tube made from 13mm x 100mm thin-walled pyrex test tube. Solid samples were packed as firmly as possible to f i l l e d at least the total c o i l volume (ca 20mm high). A l l nqr spectra were recorded with Zeeman Modulation with sideband suppression and with a time constant of 10 sec. Since in a real SRO the RF phase may be advanced or retarded more than once - 183 -during the quench cycle, the SRO usually produces poor line shapes. However, this effect can be overcome by switching another capacitor into and out of this tuned c i r c u i t during the quench cycle. If this capacitor is made variable i t can be adjusted to produce zero phase shift between pulses. The required capacitor correction i s provided i n the Decca nqr spectrometer by the "centre shift compensation" system which smooths out the poor line shapes and produces symmetrical lines. The resonance frequencies were then measured by means of the frequency marker on the spectrometer and checked by a Hewlett Packard 5246L electronic counter. The frequency was swept i n both directions so that the effects of output time constants could be eliminated i n the measurement. Since temperature gradients across the sample can produce a serious broadening, a l l samples were l e f t in the spectrometer at least 30 min. before data were taken. For low temperature measurements, the samples were frozen rapidly in appropriate cold bath [15] with shaking i n order to ensure polycry-s t a l l i n i t y . Variable temperature studies were done by freezing the sample and c o i l assembly in liquid nitrogen and then pouring out most of the refrigerant. By this means, a slow temperature variation from -196° C to room temperature (~ 22° C) can be obtained in about 4 hours. Tempera-tures were measured with a calibrated copper - constantan thermocouple (inserted through the small hole on the sample can) and a Rubicon potentiometer. - 184 -Appendix I 2 Computer Program for Evaluation of e Qq and n $COMPILE TIME=100 1 IMPLIC IT R E A L * 8 ( A - H . O - Z ) 2 DIMENSION H( 10 , 10) ,U( 1 0 , 1 0 ) , F R O ) , ( 1 ( 3 ! 3 READ(5 ,5 )X l , E«a 4 5 F O R M A T ! F 5 . 1 . F 1 0 . 5 ) 5 WRITE(6, i ) 6 1 FORMAT t1H1) 7 WRITE(6,6)X I ,EQQ e 6 FORMAT! ' S P I N = » , F 4 . 1 , « EQQ= * , F 1 0 . 5 / / 1 9 WRITE16.78) 10 78 FORMAT t • F IRST L INE GIVES FREOS,2ND L INE GIVES RATIOS 2ND TO 1ST 1 3RD TO 2ND AND 3RD TO 1ST F R E Q S ' / ) 11 ETA=0.530DO 12 81 A = E Q Q / ( 4 . * X I * ( 2 . * X I - 1 . ) ) 13 B = A * E T A / 2 . 1* N M = 2 . * X H - l . 0 l 15 DO 10 J=1,NM 16 DO 10 K=J,NM 17 X J = J - l 18 XM=XJ-XI 19 H ( J , K ) = 0 . 0 D 0 20 IF ( K . E Q . J ) H ( J , K ) = A* (3 . *XM*XM-X I * I X I • ! . ) ) 21 I F I K . E Q - J + 2 ) H ( J , K ) = B * 0 S G R T ( I X I * < X H - l . ) - X M * ( X P + l . ) ) * (X I * (X . I + 1.) -(XM 1 + l . ) * ( X M + 2 . ) ) ) 22 H ( K , J ) = H ( J , K ) 23 10 CONTINUE 24 CALL F A S J A C I H , U , 1 . D - 9 , 0 , 0 , N M , 1 0 ) 25 • F R ( 1 ) = H ( 3 , 3 ) - H ( 1 , 1 ) -26 F R I 2 ) = H I 5 , 5 ) - H I 3 , 3 ) 27' F R « 3 ) = H ( 7 , 7 ) - H ( 5 , 5 ) 28 W R I T E I 6 , 8 0 ) E T A , F * ( 1 ) , F R ( 2 > , F R C 3 ) 29 80 F O R M A T ( F 9 . 3 , 3 F 2 0 . 5 > 30 R ( 1 ) = F R ( 2 ) / F R ( 1 ) 31 * { 2 ) = F R { 3 ) / F * ( 2 ) 32 R(3 > = F R ( 3 ) / F R ( 1 ) 33 WRITE! 6 ,79 )R( l ) , R ( 2 ) t ' R ( 31 34 79 FORHAT(S ( X ,3F?0 .5 / ) 35 I F ! E T A . G T . 0 . 5 9 0 D 0 ) GO TO 82 36 ETA=ETA+0 .00 l 37 GO TO 81 38 82 CONTINUE 39 STOP 40 END 41 SUBROUTINE F A S J A C ! A . S . C R I T , N S , N O R D » N , M ) 42 IMPLICIT REAL*B<A-H .O -Z ) 43 DIMENSION A(M,M),S !M,M) 44 NS1=NS 45 NS2=NS 46 NS3=NS 47 DO 250 I =1,N 48 DO 150 J=1,N 49 A ! J , I ) = A ( I , J ) 50 IF ( I - J ) 100, 101 , 100 51 100 S t I , J ) = 0 . 0 D 0 1 52 GO TO 150 53 101 S ! I » J ) = 1.ODOO 54 150 CONTINUE 55 250 CONTINUE 56 I F ( N S l . t O . O ) GO TO 401 - 185 -57 WR ITE (6 ,7 ) 58 CALL FPU INT(A,N,H) 59 401 CONTINUE 60 INDIC=0 61 ITER=0 62 NS0 = 4*NJ*N +20 63 151 V I =0 .OU0 i 64 DO 206 1=1,N 65 DO 106 J = l , N 66 I F ! I - J ) 107, 106, 107 67 107 V1=VI • A U , J > * * 2 68 106 CONTINUE 69 206 CONTINUE 70 325 I F I N S 2 . E U . 0 ) 50 TO 707 71 WR ITE (6 ,1 ) 72 707 CONTINUE 73 I F ( N . E y . l ) GO TO 3060 74 V I=DSORI (V I ) 75 VF = V I * C ( U T 76 AN=N 77 128 V I=V I /AN 78 ITER=ITER+ 1 79 IF ( I T E R . L E . N S 3 ) GO TO 710 80 WR ITE (6 ,8 ) ITER 81 GO TO 1000 62 710 CONTINUE 83 137 10= 1 84 124 Ib=IO+l. 85 IP= 0 86 121 IP= IP+l 87 IFCAt IP,I<J)) 108, 120, 109 88 108 I F t - A I I P , IQ ) -V I I 120, 112, 112 89 109 I F ( A ( I P . I Q l - V I ) 120, 112 , 112 90 112 IN0IC=1 . 91 ALAM = - M IP, 13) 92 AMU = 0 . 5 U 0 0 * ( A ( I P , I P ) — A { I Q , 10)) 93 IF(AMU) 1 1 3 , 114, 114 94 113 SGN=- l .0DOO 95 GO TO 115 96 114 SGN=+1.0D00 97 115 3 M E G A=SGN*ALAM/DSQRT(ALAM**2 *AMU**2) 98 STHT = OMEGA/DSORT( 2.0D00+2.0D00*DSQRTI I .0D00-0MEGA*42) ) 99 CTHT=DSWRT(1.0D00-STHT**2) 100 DO 116 1 = 1,N 101 IF ( I - I P ) 1 1 7 , 1 1 8 , 1 1 7 102 117 I F ( I - I U ) 1 1 9 , 1 1 8 , 1 1 9 103 119 AIP1 = A ( I P , I ) * C T H T - A I I Q , I ) * S T H T 104 AIQI = A ( I P , I ) * S T H T + A ( I Q , I ) * C T H T 105 A ( I P , I ) = A I P I 106 A ( I C , I ) = A I 5 I 107 118 AIP I = S ( I , I P ) * C T H T - S ( I , I Q ) * S T H T 108 AIOI = S ( I . I P ) * S T H T + S ( I , I U ) * C T H T 109 S ( I , I P ) = A I P I 110 116 S ( I , IQ )=A IQ I 111 A I P ! = A< I P , I P ) * C T H T * * 2 + A < I Q , I Q ) * S T H T * * 2 - 2 . 0 0 0 0 * A ( I P , 1 0 ) * S T H T * C T H T 112 A I3 I = A< IP, IP)*STHT**2+A< I Q , I Q ) *C TH T<=*2 *2. 0 0 0 0 * A ( IP, 10 ) *ST HT*CTHT 113 A I P I O = ( A ( I P , I P ) - A ( 1 0 , I U ) ) * C T H T * S T H T « - A < I P , I C ) * ( C T H T * * 2 - S T H T * * 2 ) 114 A ( I P , I P )=A IPI 115 A [ IQ , IQ )=A101 116 A l I P , I Q ) = A I P I Q - 186 -117 A ( I Q , I P > = A < I P , I Q » 118 DO 123 1=1,N 119 A ( I , I P > = A 1 I P , I ) 120 123 A ( I , I Q ) = A ( 1 0 , I ) 121 120 I F ( I P - I 0+1> 1 2 1 , 122 ,122 122 122 I F ( I O - N ) 124, 125, 125 123 125 I F ( I N D I C ) 126, 127 , 126 124 126 INDIC=0 125 GO TO 137 126 127 I F ( V I - V F ) 1 2 9 , 1 2 9 , 1 2 8 127 129 NQ1 = N - l 128 3060 CUNTINUE 129 IF (NORu) 3 1 1 , 3 0 6 , 3 1 1 130 306 I F I N S 2 . E Q . 0 ) GO TO 403 131 WR ITE (6 ,2 ) 132 403 CONTINUE 133 00 230 1 = 1,N 134 DO 130 J = l , N 135 I F t l - J ) 1 3 0 , 1 3 1 , 1 3 0 136 131 MARK.ER = 0 137 I F I N . E 3 . 1 ) 53 TO 3040 138 DO 501 K = l , N O l 139 DO 301 L=1,NQ1 140 IF (K-L ) 301 , 302, 301 141 302 I F ( A ( K , L ) - A ( K * 1 , L + 1 ) ) 3 0 1 , 3 0 1 , 3 0 3 142 303 M A R K E R = 1 143 HOL E = AI K, L ) 144 A ( K , L ) = A ( K + l , L * l ) 145 A ( K + l , L + l ) = H O L E 146 DO 700 IP=1,N 147 HCLE=S ( I P ,X ) 148 S( I P , K ) = S t I P . K + l ) 149 S ( IP,K+1)=HQI_E 150 700 CONTINUE 151 301 CONTINUE 152 501 CONTINUE 153 3040 CONTINUE 154 IFI MARKER) 304, 304 , 131 155 304 I F (NS2 .EU .0> GO TO 402 156 WKITE16.5) I , A ( I , J ) 157 DO 208 IP=1,N 158 208 W R I T E ( 6 . 3 ) J , I P , S ( I P , J > 159 402 CONTINUE 160 130 CONTINUE 161 230 CUNTINUE 162 311 CONTINUE 163 IF (NORO.EO.O) GO TO 472 164 I M N S 2 . E Q . 0 ) GO TO 472 165 WR ITE (6 ,4 ) 166 DO 508 1=1,M 167 WK ITE I6 .5 ) 11AI 1,1) 168 DO 508 JP=1,N 169 508 W R I T E I 6 , 3 » I , I P , S I I P , I ) 170 472 I F ( N S 3 . E 0 . 0 ) GO TO 703 171 WK ITE (6 ,9 ) 172 CVLL FPR INTJA ,N ,M ) 173 708 CONTINUE 174 1 F O R M A T ( / 2 2 X , l l H E I G E N V A L U E S , 2 2 X , 1 2 H F I G E N V E C T O R S ) 175 2 FORMAT(/22X,10HIQRDERED) ) 176 3 FORMAT I 4 0 X , 1 5 , 1 5 , 3 X , T P E 2 0 . 8 ) - 187 -177 4 F O R M A T ! / 2 2 X , l 3 H ( N O T ORDERED! ) 178 5 F O R M A T ! / 1 0 X , I 5 . 2 X , I P E 2 0 . 8 ) 179 7 FORMATI1H1,10X,12HINPJT MATRIX) 180 8 FORMAT!1X,20HN0 CONVERGENCE AFTER , 1 5 , 10HITERATIONS *ff) 181 9 FORMAT I / /10X,19HDIAGOMALI ZED MATRIX , / ) 182 1O0O RETURN 183 END 184 SUBROUTINE FP* I NT IVMAr ,N,M) . 185 IMPL IC IT R E A L * 8 ! A - H , 0 - i » 186 DIMENSION VMATIM»M) 187 DO 1 K=1,N 188 PRINT 5 1 , ( V 4 A T I K , J ) , J = 1 , N ) 189 1 CONTINUE 190 51 F O R M A T ! 1 H / I 5 E 1 7 . 8 ) ) 191 RETURN 192 END - 188 -References 1. H. Dehmelt and H. Kruger , Naturwiss., 37> 111 (1950). 2. CP. Slichter, "Principles of Magnetic Resonance", Harper and Row, New York (1964), p. 160. 3. H.B.G. Casimir, "Interaction between atomic nuclei and electrons", Tayler's Tweede Genootschap Haarlem, 11, 36 (1936). 4. V. Jaccarino and J.G. King, Phys. Rev. 83, 471 (1951). 5. D.V. Ehrenstein, H. Kopfermann and S. Penselin, Z. Physik, 159, 230 (1960). 6. T.P. Das and E..L. Hahn, "Nuclear Quadrupole Resonance Spectroscopy" in Solid State Physics , Supplement 1 (1958). Frederick Seitz and David Turnbull et. a l . ed., Academic Press Inc., Publishers, New York, London. 7. M.H. Cohen, Phys. Rev. 96, 1278 (1954). 8. R.L. Collins and J.G. Travis, "MSssbauer Effect Methodology," Volume III, I.J. Gruverman, Ed., Plenum Press, New York, 1969, p. 123-161. 9. CH. Townes and B.P. Dailey, J. Chem. Phys. 17, 782 (1949). 10. T.B. B r i l l , Ph.D. Thesis, University of Minnesota (.1970). 11. CH. Townes and A.L. Schawlow, "Microwave Spectroscopy", McGraw-Hill Book Co., Inc., New York (1955). - 189 -12. W. Gordy and R.L. Cook, "Microwave and Radiofrequency Spectroscopy" in "Technique of Organic Chemistry", Vol. IX, Part II ( 2 n d rev. ed.), edited by A. Weissberger, Interscience Publisher, New York (1968). 13. C.B. Harris, Inorg. Chem. 1517 (1968). 14. C.B. Harris, J. Chem. Phys. 49, 1648 (1968). 15. A. Weissberger edit. "Technique of Organic Chemistry", Vol. I l l , Part II, p. 166-p. 170 (1957), Interscience Publishers, New York. - 190 -Chapter 4 35 Cl Nuclear Quadrupole Resonance Spectra of 1-Substituted 2-Chloro Polyfluorocycloalkenes I. Introduction 35 Although extensive Cl nqr spectral studies on chlorinated hydrocarbons have been made [1-3], there are few corresponding studies on chlorinated polyfluoroalkenes. To our knowledge, data for only five compounds of this type were reported in the literature (Table I) prior to the present work. TABLE I. Previously Reported Cl NQR Data for Chloro-Polyfluoroalkenes at -196° Compound v (MHz) * Ref. A 38.14, 38.37 Cl Cl 38.58, 38.75 [4] FC1 /\ aliphatic : 36.08(|), 36.34(|) [4] Cl 1 — * Cl vinylic : 38.15(i), 38.38(|), 38.65(1) F F • 2 37.001, 37.113 [5] Cl Cl A 38.039 [6] Cl Cl CF0-C=C-CF_ I I 39.491 [6] Cl Cl * Figures in parentheses are relative intensities. - 191 -In view of the fact that chemistry of fluorocarbon compounds is quite different from that of their hydrocarbon analogs (e.g. the double bond of a fluoroalkene is electron deficient, whereas that of a hydroalkene is 35 usually electron rich [7, 8]), i t seemed desirable to study the Cl nqr spectra of some chloro-polyfluoroalkenes. In recent years, a number of substituted chloropolyfluorocyclo-i 1 alkenes YC = CCl(CF 2) n have been synthesized (Table III). These compounds, which have a vinylic substituent Y and a vinylic chlorine atom on the adjacent carbon atoms, are somewhat related to ortho substituted aromatic compounds, and i t seemed interesting to investigate substituent 35 effects i n this system by means of Cl nqr spectroscopy. 19 In Table II are lis t e d F nmr data for some substituted i 1 perfluorocyclobutenes, YC = CF(CF 2) 2. One interesting feature i s the large variation in chemical shift of the vinylic fluorine atom. Notably when Y is a NR2 group, the resonance of the vinylic fluorine is well upfield [9, 10]. Cullen and Dhaliwal have rationalized this in terms of a shield-ing of the v i n y l i c fluorine atom due to contributions from canonical forms i — 1 such as R2N = C-CF(CF 2) 2 [9, 10]. Since nqr is a powerful tool for © 0 35 elucidating electronic effects, we have carried out Cl nqr studies on i 1 the related systems YC = CCl(CF 2) n with Cl replacement F. It was also hoped that by varying the size of the ring (n = 2 to 4), the effect of 35 ring size on the v( Cl) could be determined. Previous studies have shown that the temperature coefficients - 192 -A ~ ~ \^AT\ for aliphatically bonded chlorine compounds are greater than those of aromatically bonded chlorine compounds [3, 11] . Since some of 35 the compounds studied here show f a i r l y strong Cl nqr signals, some temperature dependence studies were made and the resulting data are compared with those reported previously. 19 TABLE II. F NMR Data (chemical shifts are reported5 in i r ppm upfield from CFC13) For YC = CF(CF 2) 2 compounds Y 6 (vinylic) Ref. F F 122.5 [12] OMe 142.0 [13] NMe2 156.4 [9, 10] SMe 121.8 [14] AsMe2 105.5 [14] Mn(C0)5 118.6 [12] (iT-C 5H 5)Fe(CO) 2 124.1 [12] - 193 -II. Experimental i 1 The 1,2-dichloropolyfluorocycloalkenes Cl-C = CC1(CF„) I n (n = 2, 3, 4) and 2,3-dichlorohexafluorobutene were obtained from Peninsular Chemresearch Inc., Gainsville, Florida. The derivatives of these compounds were prepared by the literature procedures as indicated in Table III. The desired products were purified by d i s t i l l a t i o n or recrystallization, and identified by means of their b.p. or m.p. and 1 19 known infrared, H nmr and F nmr spectra data. The experimental procedure for measuring the nqr frequencies has been described in Chapter 3. A l l linear regression plots were performed by a Hewlett-Packard model 9100 B calculator and model 9125 A plotter. The straight line of each plot i s a least squares line. The ""Cl and "'Cl nqr frequencies observed for the compounds investigated are li s t e d in Tables IV and V*. 35 Attempts to obtain Cl nqr data for the following compounds failed : YC~= CCl(CF 2) n (n = 2, Y = Me, SMe, PPh2; n = 3, Y = Me, Et, SiMe3; n = 4, Y = F, H, OMe), and C1-C(CF3> = C(CF 3)F. A l l line widths are of the order of 5 KHz. - 194 -TABLE III. References for the Preparation of YC = CC1(CF_) (n = 2 to 4) z n F F Cl 2<K 7^2 Y Cl F F Y Cl CF0-C=C-CF0 3 I I 3 Y Cl F OMe N3 H NEt 2 NMe2 Me Et SiMe 3 SMe AsMe2 PPh 2 MnCCO), [15] [16] [17] [18] [19] [20] [20] [2.1] [22] [9] [24] [25] [15] [16] [17] [18] [10] [10, 19] [20] [20] [21] [22] [23] [24] [15] [16] [18] [10] [15] (7r-C 5H 5)Fe(CO) 2 [25] - 195 -TABLE IV. Cl NQR Data for Chloropolyfluoroalkenes at -196° (in MHz) Y F F 2 ,2 Y Cl Y Cl F2 F2 F2 O '2 Y Cl CF„-C=C-CF„§ 3 1 ! 3 Y Cl F 38.311(30) 38.425(30) 38.999(100) t t Cl 37.001(600) 37.113(600) t 5 ] 38.039(350) [6] 38.666(32) 39.157(40) 39.207(45) [6] 39.491(65) OMe 37.101(400) 37.909(60) t N3 36.866(50) 37.844(25) H 36.495(40) 37.487(60) t NEt 2 37.768(20) 37.331(21) 37.222(18) NMe2 36.276(8) 37.232(50) 37.415(80) 37.965(15) Me t t Et 36.030(45) 36.252(33) t SiMe3 35.825(65) t SMe t 36.713(30) AsMe2 35.648(62) 35.675(45) 36.711(60) PPh2 f 36.638(30) Mn(CO)5 34.801(30) C 5H 5)Fe(CO) 2 34.668(100) Experimental uncertainty : + 0.050 MHz. Figures in parentheses are signal to noise ratios. cis-trans + mixture. 1 No nqr signal was obtained. - 196 -37 * TABLE V. Cl NQR Frequencies (MHz) for Chloropolyfluoroalkenes at -196° Y V 2 ^ F 2 Y Cl Y Cl F 2 F 2 F 2 < C > F 2 Y Cl CF,-C=C-CF0§ 3 1 1 3 Y Cl F 30.250(10) 30.724(25) t t Cl 29.162(200) 29.250(200) 29.982(100) 30.446(10) 30.860(15) 30.900(15) 31.125(24) OMe 29.240(130) 29.878(22) f N3 29.052(14) 29.819(10) H 28.764(11) 29.579(20) t NEt 2 29.348(7) 29.424(8) 29.775(6) NMe2 28.602(25) 29.355(20) 29.507(30) Me t t Et 28.404(14) 28.578(11) t ; SiMe3 28.250(20) t . SMe t 28.940(12) AsMe2 28.092(24) 28.119(19) 28.949(18) PPh2 t 28.840(8) Mn(C0)5 27.412(90) C 5H 5)Fe(C0) 2 27.297(30) Experimental uncertainty : + 0.050 MHz. * Figure in parentheses are signal to noise ratios. ^ Cis-trans mixture. ^ No nqr signal was observed. - 197 -III. Results and Discussion 1. Assignment of Signals 35 37 It can be seen from Tables IV and V that the Cl and Cl nqr frequencies of each chlorine site are approximately in the ratio 1.27 : 1 consistent with the expected value v C ' c i ) Q( J /C1) 35 Furthermore, the intensity of the Cl nqr signal is about three times 37 greater than the corresponding Cl nqr signal (cf. natural abundances : 35 37 Cl : 75.4%, Cl : 24.6%). This ensures that the assignments are correct. In nqr spectroscopy, the number of signals usually corresponds to the number of inequivalent chlorine sites. Table IV show that some compounds have more signals than the number of chlorine atoms in the mole-cule. Since lattice effects can be as much as 0.5 MHz [26] for this type of compound, the splittings of less than 0.5 MHz are probably due to nonequivalent chlorine positions in the crystal lattice rather than chemi-cal inequivalence. 35 3 As mentioned in Chapter 3, for Cl (I = ) , the nqr frequency i s related to the coupling constant by the equation : 1 e2Qg 1 + r n ]2 (5) 2h However, since Cl atoms in the compounds studied are involved in only one - 198 -bond, n w i l l probably be less than 0.2, so the error caused by neglecting n i s less than 0.7%. Thus, to a good approximation, the coupling constant can be obtained simply by doubling the observed frequency. 35 2. Effect of Substituents on v( Cl) 35 From Table IV one can see that the Cl nqr frequencies of these compounds decrease in the order lis t e d : F > Cl > OMe > > H > NR2 > Et > SiMe3 - SMe ~ AsMe2 ~ PPh2 > Mn(C0)5 > (TT-C 5H 5)Fe(C0) 2 . The order is essentially that of the electronegativities of the substituents. The implication then is that the variations in nqr frequencies with substituent are due chiefly to inductive effects, though other factors, particularly conjugative effects, may also be significant. Following Semin and cox-jorkers [27] we have attempted to put this on a more quan-titative basis by f i t t i n g the data to equations of the following form : v = v + aa T + b a P (1) o I {-Here v is the nqr frequency. The coefficients a^ . and a £ provide a measure of the inductive and conjugative transmission effects of the various substituents; v , a, b are coefficients to be determined in the o f i t . The origins of the o-parameters are outlined in Appendix III; the values used in the present work are given in Table VI, and are due to Exner [28]. Using the computer programs list e d in Appendices I and II least squares f i t s were made to equation (1) for both YC = CC1(CF 2) 2 and YC = CC1(CF 2) 3. The results obtained are as follows : - 199 -TABLE VI. Inductive and Conjugative Parameters of Substituents [28] Substituent M V 0* c F 0.56 - 0.59 Cl 0.51 - 0.35 OMe 0.31 - 0.63 * N3 0.33 - 0.18 H 0 0 NMe2 0.11 - 0.94 Et - 0.06 - 0.07 SiMe3 - 0.11 + 0.07 SMe 0.22 - 0.24 PPh2 0.12 0.04 From ref. 28 , a p and for N 3 are 0.15 and 0-27, respectively. These were converted to and O Q by means of equation (4) of Appendix III. For YC = CC1(CF 2) 2 (Y = F, Cl, OMe, N 3» H, NMe,,, Et, SiMe^ v = (36.211 + 0.223) + (2.745 + 0.716)^ + (0.033 ± 0.518)a (R .= 0.89 ) (2) - 200 -For YC = CC1(CF 2) 3 (Y = F, Cl, OMe, N 3 > H, NMe^ , SMe, PPh2) : v = (36.709 + 0.386) + (2.835 + 1.138)aI - (0.367 + 0.662)a (R = 0.78) (3) — c In both equations (2) and (3) the coefficients of a have standard errors c larger than the value of b i t s e l f , suggesting that the overall dependence on conjugative effects i s not determinable from the data. The coefficients of Oj are rather better determined though that of the cyclopentenes has a rather large standard error. The correlations are only f a i r . Since the dependence on a was indeterminable, we also deter-mined the dependence on alone. The results are as follows : For YC = CC1(CF 2) 2 : v = (36.211 + 0.181) + (2.723 + 0.569)aT (r = 0.89) (4) For YC = CC1(CF 2) 3 : v = (36.793 + 0.335) + (3.018 + 1.024)a (r = 0.77) (5) As with the f i t s to both a^ . and a (equations (2) and (3)) there i s a f a i r l y reasonable dependence on a , although for the cyclopentenes "a" has a rather large standard error. Three noteworthy points emerge. We always use R to denote the correlation of v and both ex and a c, and use r to denote the correlation of v and a T alone. - 201 -(i) In both cases V q and "a" agree to within the standard errors with the corresponding values in equations (2) and (3). The standard errors are rather lower in equations (4) and (5), confirming the indeterminary of the dependence on a . ( i i ) The values of "a" in equations (4) and (5) are the same within their standard errors, suggesting that the dependence on cr is purely a function of the substituent and independent of the rest of the ring. ( i i i ) The correlation of v with i s only f a i r in both cases. This i s shown pic t o r i a l l y in Figures 1 and 2 in which equations (4) and (5), respectively, are plotted. The scatter of points i s f a i r l y great; apparently some effect other than those considered also affects the nqr frequencies. The most lik e l y cause of the scatter i s differences in extra-molecular potentials caused by differences in the crystal structure of the various substances. These typically result in frequency variations of the order +0.5 MHz [26] i n organic molecules. Since the molecules consi-dered here are rather complex, with complex substituents, large variations in crystal structure, and consequent scatter of the nqr frequencies can be expected. Steric effects of the substituents might also give small frequency variations, and there might also be differences in the C = C Cl angle and C-Cl bond length having similar effects. A l l these contributors could well mask any dependence of frequency on a . With this in mind i t sacmed f r u i t f u l to reexamine the data in more detail to try to find some - 202 -v ( 3 5 C l ) Figure 1. v( Cl) of YC =~CO(CF 2) 2 vs of Y (r = 0.89, intercept = 36.205, slope = 2.723) - 203 -v ( 3 5 C l ) (MHz) 39.00 38.70 1 (r = 0.77, intercept = 36.793, slooe = 3.018) - 204 -dependence on both, oj and ac when variations of crystal structure would be expected to be small. On careful examination of Figures 1 and 2, i t appears that i f the'substituents are divided into the following three groups, then the resulting one parameter f i t s are quite good (see Figures 3 and 4 and Table VII) : (i) substituents containing an atom without lone-pair electrons - H, Et, SiMe^; ( i i ) substituents involving second period elements - F, OMe, N^, NR2; ( i i i ) substituents involving third period elements - Cl, SMe, PPh 2 < TABLE VII. One Parameter Correlation Equations for YC = CCl(CF 2) n (n = 2, 3) Y F F 2 1 2 Y Cl F2 F2DF2 Y Cl H v = (36.498 + 0.009) + Et (6.085 + 0.120)CT]. SiMe3 (r = 1.00) (A) F OMe V = (35.628 + 0.287) + v = (36.706 + 0.167) + N3 (4.656 + 0.787)qI (3.939 + 0.458)0^. NMe2 (r = 0.97) (B) (r = 0.99) (C) C 1 v = (36.050 + 0.256) + S M e (3.812+0.780)o PPh 2 (r = 0.98) (D) 205 -35 y>C ci) 39.0 38.0 37.0 36.0 35.0. -0.15 0.10 ,35, 0.35 Figure 3. v( Cl) of YC = CC1(CF 2) 2 (After dividing Y into three groups) A : Y = H, Et, SiMe3; ] vs of Y Y = F, OMe, N3, NMe2 v ( 3 5 C l ) 39.0 r (MHz) - 206 -38.4 37.8 ,0Me 37.2 NMe; 36.6 PPh, SMe 0.2 0.4 Figure 4. v ( 3 5 C l ) of YC~^ CC1(CF2) (after dividing Y into three groups) vs of Y Y = F, OMe, N , NMe2; D : Cl, SMe, PPh, - 207 -The two parameter correlation equations for compounds containing group ( i i ) substituents can also be obtained as shown in Table VIII. These equations show that there i s a f a i r l y good correlation between v and the two substituent parameters and a^. This implies that the TABLE VIII. Two Parameter Correlation Equations for Some YC = CCl(CF 2) n (n = 2, 3) *2 Cl 2W *2 Y Cl F OMe N3 NMe„ = (35.013 + 0.016) + (5.220 + 0.026)a] - (0.736 + 0.016)a R = 1.00 = (36.353 + 0.076) + (4.262 + 0.128)0 - (0.422 + 0.076)a( R = 1.00 v( Cl) of these compounds are affected by both inductive and conjugative 35 effects. Whereas increasing a increase the v( Cl), the reverse w i l l be expected for increasing a , although the contribution from the conjugative effect s t i l l seems to be small. A l l the nqr data can thus basically be rationalized in terms of two simple canonical forms, namely JL and 2_ below, in which Cl is covalently (by a single bond) and ionically bonded, respectively, to carbon. - 208 -Y Cl Y Cl © 2 By analogy with other ethylenic compounds [29, 30], these forms probably contribute roughly 70 and 20 percent, respectively, to the resonance 2 35 hybrid, with the result that e Qq for Cl is approximately 70 MHz, as found in this work. As the electronegativity of substituent Y increases, the s-character of carbon i n the C-Cl bond increases ; this in turn increases the electronegativity of the carbon atom and decreases the ionic character of the C-Cl bond (i.e. decreases the contribution 2 of structure 2) [30]. Since e Qq of Cl in 2_ is zero, the net result should be an increase in nqr frequency with electronegativity (and a^) of Y, as found. As mentioned in the introduction to this chapter electronegativity 19 alone can not be used to rationalize the F chemical shifts of the i r vinylic fluorine atom in the compounds YC = CF(CF2>2 (Table II) and contribution from canonical form of the type 3^  (Z = F) needed to be considered. Because of the electronegativity difference between F and Cl i t i s not inconceivable that 3 would make a greater ^ contribution to the resonance hybrid when ® Y Z Z = F than that Z = Cl. 3 - 209 -Figure 5 is a plot of v ( 3 5 C l ) of YC = CC1(CF 2) 2 (Y = F, OMe, 19 NMe^ ) vs <5( F) of the vinylic fluorine of the corresponding , . — , YC = CF(CF 2) 2 compounds. The good linear relationship (r = 1.00 and small standard errors of the coefficients) between these two parameters suggests that they are affected by the same factors. The correlation i s not good i f other compounds are included and the significance of the plot is not obvious. 3. Effect of Ring Size 35 From Table IV one can see that the v( Cl) of the ring compounds increases as the ring size increases. Two reasons can be suggested for this change. Fi r s t l y , the increase in the number of electron withdrawing CF 2 group(s) which w i l l decrease the contribution from 2_ , i.e. ,(CFJ ( ' f 3 5 Y—C = C Cl , and h ence increase the v( Cl). Secondly the decrease in the exocyclic JZ = C and C = C bond angles w i l l decrease Y Cl the s-character of C in the C-Cl bond, and hence increase the effec-tive electronegativity difference between C and Cl. However, cis-2,3-dichlorohexafluorobutene, which has essentially the same exocyclic bond 35 angles as 1,2-dichlorooctafluorocyclohexene , has higher v( Cl) (39.491 MHz) than the latter (38.666 MHz). This indicates that the f i r s t is probably the main reason. The nqr frequencies of 1,2-dichlorodifluorocyclopropene at -196° are 38.14, 38.37, 38.58 and 38.75 MHz (Table I). Here the - 210 -38.50 Y F from CFC13> Figure 5. v ( 3 5 C l ) of YC = CC1(CF 2) 2 vs 6( 1 9F) of Vinylic Fluorine of YC = CF(CF.)„ CY = F, OMe, NMe„) Z Z JL. v = (45.933 + 0.303) - (0.062 + 0.002)6 (r = 1.00) - 2 i i a -Figure 6. Temperature Dependence Curves of YC = CCl(CF 2) n (n = 2, (a) Y = OMe; (b) Y = SiMe3; (c) Y = Mn(C0)5. n = 3, (d) Y = F; (e) Y = Cl; (f) Y = H; (g) Y = AsMe2. n = 4, (h) Y = Cl.) - 211b-T E M P E R A T U R E ( ° C ) - 212 -vin y l i c chlorine atoms come into resonance at higher frequencies instead of lower than 1,2-dichlorotetrafluorocyclobutene. This may be due in part to a contribution from the ionic form ® F F © Cl Cl to the ground state of the former compound. The canonical form _4_, which * i s favored because of i t s aromaticity, would increase the polarity of the C-F bond and then increase the electronegativity of the C atoms in the ring. This result would decrease the polarity of the bonds to the vinylic chlorine and increase the resonance frequency. 35 4. Temperature Dependence of v( Cl) 35 The temperature dependence of v( Cl) of hydrocarbons has been well documented [2, 26], but l i t t l e is known about fluoroalkenes. Figure 6 shows the temperature dependence curves of some fluorocycloalkenes 35 investigated here. The v( Cl) is decreasing with increasing temperature which i s qualitatively the same behavior expected from Bayer's theory [31]. The shape of each curve is smooth indicating that there are probably no phase transitions in the region investigated. Figure 7 contains the temperature dependence curve of EtC - CC1(CF 2) 2. At liquid nitrogen temperature, this compound shows two * The corresponding canonical forms for the higher homologs. C1C - CCl(CF 2) n (n = 2 to 4), are not important because they do not have aromatic character. - 213 -v ( 3 5 C l ) ( M H z ) 36.30 i Figure 7. Temperature Dependence of EtC = CC1(CF„) -193 -173 -153 -133 -113 -93 -73 -53 -33 Temperature (°C) - 214 -nqr signals. Usually the intensity of the higher frequency line i s greater than that of the other (10 : 1). But, sometimes, by suddenly cooling the sample in liquid nitrogen and taking the spectrum immediately, two lines of equal intensity are observed at the same temperature. This suggests that there are two phases present with the higher frequency species (a phase) more predominant than the lower frequency one (B phase). At -70°, the high frequency signal disappears and the intensity of the remaining signal increases considerably. Thus there is probably a phase transition at -70° where the a phase is converted to the 8 phase. The latter is stable up to about -34°. Table IX l i s t s the temperature coefficients of the compounds studied here together with those of aliphatically bonded and aromatically bonded chlorine compounds. An interesting feature is that these compounds -4 show a f a i r l y large temperature coefficient a =-0.9 x 10 to -4 o -1.8 x 10 / K, which is higher than that found i n aromatically bonded chlorine compounds [11]. This indicates that although the chlorine is bonded to the fluoroalicyclic ring the mean amplitude of thermal vibration of a C-Cl bond is higher than that bonded to aromatic rings . However, the temperature coefficient of these compounds i s , in general, lower than that found in aliphatically bonded chlorine compounds [3]. This suggests that the mean amplitude of thermal vibration of a C-Cl bond in a fluoroalicyclic ring i s lower than that bonded to an aliphatic chain . - 215 -TABLE IX. Temperature C o e f f i c i e n t s of Some YC = C C l ( C F 2 ) n (n = 2 to 4) Compounds 1 a = — v A v AT 2 C l - 1.2 x 10~ 4 / ° K 2 OMe - 1.5 x 10" 4 / ° K 2 NMe2 - 0.9 x 10~ 4 / ° K 2 Et - 1.1 x 10~ 4 / ° K 2 SiMe 3 - 1.0 x 10" 4 / ° K 2 H - 1.0 x 10~ 4 / ° K 2 N 3 ' - 1.0 x 10" 4 / ° K 2 Mn(CO) 5 - 1.3 x 10" 4 / ° K 2 ( 7T-C 5H 5)Fe(CO) 2 - 0.9 x 10" 4 / ° K 3 Cl - 1.0 x 10~ 4 / ° K 3 F - 1.0 x 10" 4 / ° K 3 NMe2 - 1.3 x 10" 4 / ° K 3 AsMe 2 - 0.9 x 10" 4 / ° K 3 II - 1.0 x 10" 4 / ° K 3 OMe - 1.2 x 10" 4 / ° K 3 N3 - 0.9 x 10" 4 / ° K 4 Cl - 1.8 x 10" 4 / ° K A l i p h a t i c a l l y bonded chlorine compounds [3] 1.Ox 10 4 to - 8 . 6 x 10" 4 / ° K Aromatically bonded Chlorine compounds [11] - 0.5 x 10 4 to - 0.9 x 10 4 / ° K - 216 -IV. Summary 35 i 1 The Cl nqr of the ring compounds YC = CCl(CF 2) n (Y varies; n = 2 to 4 but not a l l combinations) has been studied. 35 The v( Cl) of these compounds decrease i n the order : F > Cl > OMe > N 3 > H > NR2 > Et > SUfej •- SMe ~ AsMe2 ~ PPh 2 > Mn(CO)5 > (ir-C^H^)Fe(CO)2 . This i s roughly parallel to the trend of electronegati-vity of the substituent suggesting that the inductive effect i s the main 35 factor responsible for the change of v( Cl). However, the correlations of v ( 3 5 C l ) of YC = CCl(CF 2) n (n = 2, 3) with a of Y are not good. These indicate that inductive effects are not the only factors affecting the v ( 3 5 C l ) . It has been found that the substituents Y can be divided into at least three groups : (i) Substituents containing lone pair electrons -H, Et, SiMe^; ( i i ) substituents involving second period elements - F, 0, N; ( i i i ) substituents involving third period elements - Cl, S, P, etc. The second type substituents show both inductive and conjugative effects. 35 The v( Cl) of the ring compounds studied increase as the ring size is increased. This can be attributed to the increase in number of electron withdrawing CF,, groups. By means of temperature dependence study a phase transition in i 1 EtC = CC1(CF 2) 2 has been observed. The temperature coefficient of the ring compounds investigated - 217 -lies between - 0.9 x 10~ to - 1.8 x 10 /°K which is greater than that found for aromatically bonded chlorine compounds but smaller than that in aliphatically bonded chlorine compounds. - 218 -Appendix I Computer Program for Multiple Regression (I) SCOHPILE BUN=NOLIST »*9ARNING** MIS-PUNCHED JOB OPTION. NOLIS IS INVALID NQR FITS TO SIGBAI 6 SIGMAC IMPLICIT REAL*8 (A-H,0-Z) DIMENSION A(4,20) ,D(3,3) ,DX(3,3) ,DD(3,3) , V (3) , V AE (3) , SE (3) ,VA(3) DIMENSION DY(20) HEAD(5,1)NF FORMAT(15) J=1 READ(5,5.END=15) (A (1 ,3 )#1=2,4) A (1,J)=1. NOBS=J J = J*1 GO TO 10 CONTINUE FORMAT(3F10.5) WRITE(6,20) FORMAT(/' INPUT DATA...') WRITE(6,21) FORMAT ( • SIGHAI SIGMAC FREQUENCY•/) WRITE (6,5) ( (A (I,J) ,1=2,U),J=1,N0BS) DO 90 J=1,NF DO 85 1=1,NF D(I,J) = 0.0 DO 80 K=1,NOBS D(I,J) = D(I,J)+A(I,K)*A(J,K) CONTINUE IF(X.EQ.J)GO TO 85 D(J,I) = D(I,J) CouTIIauE CONTINUE DO 100 1=1,NP VA (I) =0.0 DO 95 K=1,80BS VA (I) = VA (I) +A (I,K) *» («,K) CONTINUE CONTINUE WEITE(6,111) FORMAT(//• HATRIXBEF0BEINVE8SION•/) WHITE (6,203) ( (D (I,J),J=1,NF),I=1,NF) FORMAT (/IX,1P3D15.6) DO 106 1=1,NF DO 106 J=1,NF DX(I,J)= D (I,J) CALL DIt«VBT (D,NF, 3, DET, CONT) WRITE (6,101)DET.COND FORMAT(5H DET=,D26.16,7H C0ND=,D26.16) WRITE(6,112) FORMAT(//• MATRIXAFTERINVERSION*/) WRITE(6,203) ( (D (I,J),J=1,NF),I=1,NF) WRITE(6,109) FORMAT (/* CHECK OF INVERSION V ) DO 108 1=1,NF DO 108 J=1,SF DD (I,J)=O.OD0 DO 108 K=1,NF DD (I,J)=DD (I,J)*DX(I,K)«D(K,J) CONTINOE DO 107 1=1,NF c c 1 2 3 4 5 6 7 10 8 9 10 11 12 15 13 5 14 15 20 16 17 21 18 19 76 20 21 22 23 24 .80 25 26 * 85 28 90 29 30 31 32 33 95 3 4 100 35 3 6 111 37 38 203 39 «0 4 1 106 42 43 44 101 4 5 46 112 47 48 49 109 50 51 52 53 54 55 108 56 - 219 -57 107 DD(I,I)=DD (I,I)-1 . 0 D 0 58 W BITE (6,203) ((DD(I,J),J=1,BF),1=1,HP) 59 WRITE (6,181) FORMAT {/• CORRELATION COEFFICIENTS*/) 60 181 61 DO 180 1=1,NP 62 DO 180 J=1,NP 63 180 DX(I,J)=D ( I , J ) / (DSQBT(D ( 1 , 1 ))*DSQHT(D(J,J))) 64 WRITE(6,203) ( (DX (I,J),J=1,NF),I=T,NF) 65 105 CONTINUE 66 DO 16 1=1,3 67 16 V(I)=0.ODO 68 DO 17 1=1,HF 69 DO 17 J=1,NF 70 V (I)=V (I) • D(I,J)*VA(J) 71 17 CONTINUE 72 WRITE(6,113) 73 113 FORMAT(//• ORIGINALVECTOR ANSWERS') 74 DO 206 1=1,NF 75 WRITE(6,204) VA(I) ,V(I) 76 204 FORMAT (1P2D16.8) 77 206 CONTINUE 78 110 REGB = 0 . 0 79 DO 115 1=1,NP 80 115 REGB = REGB • V (I)*VA(I) 81 SUHSQ = 0 . 0 82 DO 120 1=1,NOBS 83 120 SUMSQ=SUMSQ*A (4 , 1 )**2 84 RESID = SDMSQ - REGB 85 RNOBS=NOBS-NF 86 SDFIT=RESID/RNOBS 87 S DFTT= DSQPT(SDPIT) 88 WRITE(6,139)SDPIT 89 139 FORMAT (//' STANDARD DEVIATION OF THE PIT IS•,1PD30.15//) 90 DO 135 1=1,NF 91 130 VAR(I) = RESID*D(I,I)/HNOBS 92 135 SE(I) = DSQRT(VAR(I)) 93 WRITE(6,170) 94 170 FORMAT (/• NUO,COEFF OP SIGMAI,COEPP OF SIGMAC'/) 95 DO 171 1=1,NP 96 171 WRITE(6,172) V(I) ,SE(I) 97 172 FORMAT (F20.10,' •OB-,,P6.3) 98 WRITE(6,315) 99 315 FORMAT(//• OBS FREQ CALC FREQ DEVIATIOH 1 STD DEV T TEST') 100 DO 305 J=1,NOBS 101 FCAL=0. 102 DO 310 1=1,HF 103 310 FCAL=FCAL*V(I)* A(I,J) 104 DIPF=FCAL-A (4,J) 105 DO 480 1=1,NP 106 DY(I)=0.0D0 107 DO 480 K=1,NP 108 Dl (I) = D Y (I) *D(I,K)*A (K,J) 109 480 CONTINUE 1 10 DZ=0.0D0 111 DO 481 1=1,SP 112 481 DZ=DZ*DY (I)*A (I,J) 113 DZ=SDFIT* (DSQBT (DZ)) 1 14 TT=DIFF/DSQRT(SDFIT**2-DZ**2) 115 WRITE(6,300)J,A (4,J),FCAL,DIFF,DZ.TT 116 305 CONTINUE 117 300 FORMAT (I5,5F15.5) 118 STOP 119 END - 220 -Appendix II Computer Program for Multiple Regression (II) SCOMPILE 1 IMPLICIT REAL*8 (A-H,0-Z) 2 DIMENSION X (100),Y (100) ,Z (100),R(3,3) ,C(3) ,&(3) 3 N=0 « DO 100 1=1,101 5 N=N*1 6 100 R3AD(5,2,END=103) X (T),1(1),Z(I) 7 103 N=H-1 8 XS=0.0D0 9 XYS=0.0D0 10 XZS=0.ODO 11 YS=0.OD0 12 ZS=0.0D0 13 YYS=0.0D0 11 YZS=0.0D0 15 XXS=0.0D0 16 ZZS=0.0D0 17 AN=N 18 DO 101 1=1,N 19 XS=XS+X(I) 20 XYS=XYS+X ( I ) * Y ( I ) 21 X Z S = XZS+X(I) *Z(I) 22 YS=YS+Y (I) 23 ZS=ZS+Z(I) 2U X X S = XXS + X (I)*X (I) 25 YYS = YYS+Y (I)»Y (I) 26 YZS=YZS + Y (I)*Z (I) 27 101 ZZS = ZZS+Z (I)*Z (I) 28 R (1,1)=YYS*ZZS-YZS*YZS 29 R (1,2)=-YS«ZZ5*ZS*YZS 30 8 (2, 1)=H (1,2) 31 h(1 ,3)=YS*YZS-ZS*YYS 32 R (3,1)=R(1,3) 33 R(2,2)=AN*ZZS-ZS*ZS 34 R(2 ,3) =YS*ZS-AN*YZS 35 R(3,2)=R (2,3) 36 R (3,3) =AN*YYS -YS*YS 37 D=AH*YYS*ZZS+2.0*YS*YZS*ZS-YYS*ZS*ZS-AN*YZS*YZS-YS*YS*ZZS 38 C(1 ) = X S/D 39 C(2)=XYS/D HQ C(3)=XZS/D 111 DO 102 1=1,3 «2 A(I)=0.0D0 «3 DO 102 J=1,3 «Ji» 102 A (I) = A (I) *R (I, J) *C (J) U5 WRITE(6,3) (A (I) ,1=1,3) C «I6 SIX=DSQRT (AN*XXS-XS*XS) «7 SIY=DSQRT ( A N * Y Y S - Y S * Y S ) «8 SIZ = DSQRT (AN*ZZS-ZS*ZS) c 1*9 ROXY= (AN*XYS - XS * Y S ) /SIX/SIT 50 ROYX= HOXY 51 ROYZ= (AN*YZS - YS*ZS)/SIY/SIZ 52 ROZY= ROYZ 53 SOZX= (AN*XZS - ZS*XS)/SIZ/SIX 5"» ROXZ= ROZX C 55 DXZ=DSQRT(1. -ROXZ*ROXZ) 56 DZX= DXZ - 221 -DYZ= DSQRT (1. -R0YZ*B0YZ) DZY= DYZ DXY=DSQRT(1. -R0XY*R0XY) DYX= DXY ROXYZ= (ROXY - ROXZ*RGYZ)/DXZ/DYZ ROZXY= (ROZX - ROZY*ROXY)/DZY/DXY ROYZX= (ROYZ - ROYX*ROZX)/DYX/DZX DR= 1. - ROYZ*ROYZ - ROXY* (ROXY - ROYZ*ROXZ) • ROXZ* (ROXY*ROYZ # ROXZ) Q=DSQRT(1. -DR/(1. -ROYZ*ROYZ)) WRITE (6,U) ROXY,ROYZ,ROZX WRITE(6,5) ROXYZ,ROZXY,ROYZX WRITE(6,6) Q FORMAT(3F12.6) FORMAT ("IX, »A = ',F12.6,« , B= »,F12.6,» , C= «,F12.6) FORMAT (»0*,* ROXY=',F7.3,6X,«ROYZ=•,F7.3,6X,»ROZX=•,F7. 3) FORMAT («0',•ROXY.Z=«,F7.3,6X,'ROZX.Y= 1,F7.3,6X,'ROYZ.X=•,F7.3) FORMAT (•0»,'RX.YZ=',F7.3) STOP END - 222 -Appendix III Substituent Effects As early as 1937 Hammett [32] used linear free energy relation-ship and developed sigma (a) functions : log k = log k° + pa (1) where k is the rate constant or equilibrium constant; k° i s the st a t i s t i c constant; p i s determined by reaction and conditions, but independent of substituents; and a is characteristic of the substituent (in a given position, e.g., meta or para) and independent of the reaction. 19 Later Taft [33] used F nmr in substituted fluorobenzenes to study the substituent effects. He has devoted much effort to the interpretation of a constants i n terms of separated inductive, conjugative and steric effects. Since then, the substituent effect has been the subject of extensive research [28, 34]. The separation of a constants into their inductive and conju-gative compounds would provide two new scales of theoretical interest, and make possible correlations. The solution of the problem requires the following steps : (i) determination of the relative inductive effect of substituents by use of a suitable model system, ( i i ) choice and evaluation of convenient values of a constants, containing inductive and conjugative effects, ( i i i ) recalculation of both quantities on a common scale, and (iv) evaluation of the conjugative effect by substraction. The f i r s t step - 223 -of this program has been essentially completed; the inductive constants, Oj, have been determined by use of r i g i d cyclic systems such as 4-substituted bicyclo [2,2,2]octane-l-carboxylic acids [35-38], or by comparing acid- and base-catalysed hydrolysis of substituted acetates [39, 40], or even from pKa values of substituted acetic acids, assuming that steric effects are small [41, 42]. The important result of these studies is an order of substituents with regard to their inductive effect, independent of the method of determination. In the second step the a constants of para substituent, a , are usually chosen, because they contain both inductive and conjugative compounds in comparable amounts. In the third step, Taft assumed as a f i r s t approximation that cr^ values from reactions of bicyclo [2,2,2]octane derivatives [35] and a ( a m constants of meta substituent) and a values from benzene derivatives P are on the same scale, because of the similarity of the systems. In the fi n a l step, Taft calculated conjugative constants, o^, from the following equation [39, 43, 44] : c r = a + a (2) p i c when the constants cr , are once defined, a can also be expressed [43] c m in terms of a-j- and ac as i n equation (3) : a = a + acr (3) m 1 . c where a = 0.33. Hence i f both a and a are known, o _ and cr can p m I c be obtained from equations (2) and (3) : - 224 -o-j = ( a m - a a p ) / ( l - a) CTc = ( 0p ~ CTm)/(1 " a ) • Further values of o ^ and o " c were obtained from nmr measurements [45] and a c from i r results [46]. The values used in the present work are in Table VI of Chapter 4. They are corrected values according to Exner [47], or recalculated according to new measurements [48-55]. - 225 -References 1. T.P. Das and E.L. Hahn, "Nuclear Quadrupole Resonance Spectroscopy", Academic Press (1958). 2. E.A.C. Lucken, "Nuclear Quadrupole Coupling Constants", Academic Press (1969). 3. I.P. Biryukov, M.G. Voronkov and I.A. Safin, "Tables of Nuclear Quadrupole Resonance Frequencies", Israel Program for Scientific Translations, Jerusalem (1969). 4. R.M. Smith and R. West, Tetr. Lett., 26, 2141 (1969). 5. C. Brevard and J.M. Lehn, du Journal de Chimie Physique, 65_, 727 (1968). 6. E.A.C. Lucken, J. Chem. Soc, 2954 (1959). 7. Milos' Hudlicky', "Organic Fluorine Chemistry", Plenum Press, (1971). 8. W.A. Sheppard and CM. Sharts, "Organic Fluorine Chemistry", W.A. Benjamin, Inc., New York (1969). ' 9. W.R. Cullen, P.S. Dhaliwal and G.E. Styan, J. Organometal. Chem., 6, 364 (1966). 10. W.R. Cullen and P.S. Dhaliwal, Can. J. Chem., 45_, 719 (1967). 11. W. Pies, H. Rager and A. Weiss, Organic Magn. Res., _3, 147 (1971). 12. P.W. Jolly, M.I. Bruce and F.G.A. Stone, J. Chem. Soc, 5830 (1965). 13. A.B. Clayton, J. Roylance, D.R. Sayers, R. Stephens and J.C Tattow, J. Chem. Soc, 7358 (1965). 14. W.R. Cullen, Fluorine Chem. Revs., J3> 96 (1969). 15. R.L. Johnson, Ph.D. Thesis, The University of Iowa (1966), page 130-page 133. 16. P.S. Dhaliwal, Ph.D. Thesis, University of British Columbia (1966), page 37. 17. CO. Parker, J. Am. Chem. Soc, 81, 2183 (1961). 18. D.J. Burton and R.L. Johnson, J. Am. Chem. Soc, 86_, 5361 (1964). - 226 -19. Reference 16, page 41. 20. (a) J.D. Park and R. Fontanelli, J. Org. Chem., 28, 258 (1963). (b) Reference 16. 21. (a) W.R. Cullen and G.E. Styan, J. Organometal. Chem., 6J, 633 (1966) (b) J.D. Park and G.G. Pearson, J. of Fluorine Chem., JL, 277 (1971). 22. Reference 16, page 42. 23. Reference 16, page 42. 24. R.F. Stockel, Can. J. Chem., 46, 2625 (1968). 25. R.B. King and A. Efraty, J. Fluorine Chem., JL, 283 (1971). 26. A. Weiss, "Crystal Field Effects in Nuclear Quadrupole Resonance" in "Topics in Current Chemistry", J30, 1 (1972). 27, 28. G.K. Semin, A.A. Neimysheva and T.A. Babushkina, Izvestiya Akademii Nauk SSSR, Seriya Khimicheskaya, 2, 486 (1970). 0. Exner, "The Hammett equation - the Present Position" i n "Advances in Linear Free Energy Relationships", edited by N.B, Chapman and J. Shorter, Plenum Press (1972), Chapter 1. 29 M.C.L. Gerry, Can. J. Chem., 49, 255 (1971). 30. R.G. Stone and W.H. Flygave, J. Mol Spectrosc., 32, 233 (1969). 31. H. Bayer, Z. Physik, 130, 227 (1951). 32. L.P. Hammett, J. Am. Chem. Soc, 59, 96 (1937). 33. R.W. Taft, Jr., "Separation of Polar, Steric and Resonance Effects in Reactivity" in "Steric Effect in Organic Chemistry", edited by M.S. Newman, John Wiley and Son Inc., (1956), Chapter 13. 34. J. Shorter, "The Separation of Polar, Steric and Resonance Effects" in "Advances in Linear Free Energy Relationships", edited by N.B. Chapman and J. Shorter, Plenum Press (1972), Chapter 2. 35. J.D. Roberts andW.T. Moreland, J. Am. Chem. Soc, J7_5, 2167 (1953). 36. H.D. Holtz and L.M. Stock, J. Am. Chem. Soc, 86, 5188 (1964). 37. CD. Ritchie and E.S. Lewis, J. Am. Chem. Soc, 84, 591 (1962). 38. C.F. Wilcow and J.S. Mclntyre, J. Org. Chem., 30, 777 (1965). - 227 -39. R.W. Taft, J. Phys. Chem., 6_4_, 1805 (1960). 40. R.W. Taft, J. Am. Chem. Soc, 74, 3120 (1952). 41. M. Charton, J. Org. Chem., 29, 1222 (1964). 42. K. Bowden, M. Hardy and D.C. Parkin, Can. J. Chem., 46, 2929 (1968). 43. R.W. Taft and I.C. Lewis, J. Am. Chem. Soc, 80, 2436 (1958). 44. R.W. Taft and I.C. Lewis, J. Am. Chem. Soc, 81, 5343 (1959). 45. R.W. Taft, E. Price, I.R. Fox, I.C. Lewis, K.K. Anderson and G.T. Davis, J. Am. Chem. Soc, 85, 709, 3146 (1963). 46. R.T.C. Brownlee, A.R. Katritzky and R.D. Topsom, J. Am. Chem. Soc, 87, 3260 (1965). 47. 0. Exner, Coll. Czech. Chem. Comm., 31, 65 (1966). 48. 0. Exner and Lakomy', Coll. Czech. Chem. Comm., 35, 1371 (1970). 49. A.V. W i l l i , Z. Phys. Chem. (Frankfurt), 27, 233 (1961). 50. A.A. Humffray, J.J. Ryan, J.P. Warren and Y.H. Yang, Chem. Comm., 610 (1965). 51. J.M. Wilson, A.G. Briggs, J.E. Sawbridge, P. Tickle and J.J. Zuckerman, J. Chem. Soc. (A), 1024 (1970). 52. W. Polaczkowa, N. Porowska and B. Dybowska, Roczniki Chem., 35, 1263 (1961). 53. W.N. White, R. Schlitt and D. Gwynn, J. Org. Chem., 26, 3613 (1961). 54. E.N. Tsvetkov, D.I. Lobanov, M.M. Makhamatkhanov and M.I. Kabachnik, Tetrahedron, 25, 5623 (1969). 55. R.A. Robinson and K.P. Ang, J. Chem. Soc, 2314 (1959). - 228 -Chapter 5 59 Co Nuclear Quadrupole Resonance Spectra Of Some Cobalt Carbonyl Complexes I. Introduction 59 7 7 Co has a nuclear,spin of For spin —, there are three 2 allowed nqr transitions, so that the two useful parameters e Qq and n 59 can be evaluted. Furthermore, Co has a f a i r l y large quadrupole moment (0.4044 barn) [1] and a natural abundance of essentially 100% making the 59 detection of Co nqr frequencies relatively easy. Recently a large number of cobalt carbonyl complexes has been synthesized [2-6]. Many of them are believed to possess unusual structures. Nevertheless, in spite of the above-mentioned favorable conditions, 59 very few Co nqr data for this type of compound are known. For this 59 reason, Co nqr studies on some cobalt carbonyl derivatives have been carried out. 59 Previously Co nqr data for the following types of carbonyl compounds have been reported [7-11] : (i) RR'R"MCo(C0)^ (M = Group IV metal), ( i i ) RR'M[Co(C0)^]2> ( i i i ) RM[Co(C0) 4] 3 > (iv) M[(Co(C0) 4] 4, (v) Hg[Co(C0) 4] 2 and (vi) Co 2(C0)g. As an extension of this work, i t was decided to study the following classes of cobalt compounds. (1) The metal-metal bonded complexes [Y3PCo(C0)^]^, in the hope of getting some new data for future structural correlation and to - 229 -59 2 examine the effect of Y on the Co coupling constant e Qq . (2) The ionic complexes L^CoCCCO^X which are interesting because the Co atom here bears a positive charge. It is desirable to see 59 2 i f the Co e Qq is sensitive to the formal oxidation state of the Co atom. The potential use of nqr as a means of determining structures of compounds, particularly in crystals, was an important objective of this work. To this end spectra of the following complexes were obtained : (3) the derivatives of Co2(C0)g containing bridging carbonyl groups, L Co0(CO)!r (CO)^ ; (4) the (RC B CR')Coo(C0), complexes which have n z o—n z i. o structural and chemical properties very similar to those of Co2(C0)g [3]. II. Experimental The (7r-C 7Hg) CCo 2(CO) 6 and the (L-L) b(PhC = CPh)Co2(CO)4 complexes (L-L = fgfars, f^fos) were prepared by the procedures given in Chapter 2 of this thesis. The other compounds were synthesized by procedures already i n the literature as follows : [LCo(CO) 3] 2 (L = (PhO)3P [12, 13], (EtO)3P [12], (MeO)3P [12], (MeO)PPh2 [12], Bu3P [12] and Ph3P, Ph^s, Ph^Sb [12] ); L2Co(CO)3X~ (L = Ph3P, Bu3P; X = Co(CO)4, BPh4) [14]; [(MeO)3P]3Co(CO)2BPh~ [15]; f 4fars bCo 2(C0) 6 [16]; [ ( T T - C ^ ) CCO(CO) ^\2 [17]; (y-alkyne) Co2(CO) g(alkyne = HC = CH [18], B^C = CH [19], HOCH2C = CCH2OH [20], PhC = CPh [20], CF3C = CCF3 [21]). The new complex (IC = CI)Co 2(CO) 6 was made from IC = CI [22] and Co o(C0) o by a method similar to reference [20]. - 230 -Reactions w.re carried out on a scale large enough to provide 3-5 g. of pure products. A l l the investigated compounds had melting points (or boiling points), i r and nmr spectra in good agreement with the known literature data. A l l samples were kept sealed under a nitrogen atmosphere to prevent decomposition and/or oxidation. The experimental procedure for measuring the nqr frequencies has been described i n Chapter 3. A l l the two-parameter correlation equations of the form 2 2 ^ e Qq = e Qq^ + a^o^ + b^cr with multiple correlation coefficient R were obtained by using the computer program list e d i n Appendix II of Chapter 4, 2 while the standard deviations in the coefficients e Qq , a and b were o obtained by another computer program list e d in Appendix I of Chapter 4. The observed frequencies for the compounds studied are tabulated in Tables I and II. 59 Attempts to detect the Co nqr signals in the range 7 to 60 MHz for the following compounds were not successful : [LCo(C0).j]2 (L = (MeO)PPh2, Ph3P, Ph As, Fb^Sb)., (BunP) 2Co (CO)^x" (X = Co(C0) 4, BPh 4), [(MeO)3P]3Co(CO)2BPh~, f 4fos(PhC = CPh)Co2(CO)4 and (IC = CI)Co„(C0), . z o We w i l l always use R to denote multiple correlation coefficient R^  which gives the correlation of X with respect to Y and Z, where 2 X = e Qq, Y = J o and Z = [a . - 231 -TABLE I. 5 9Co NQR Data for L^oCCCO^X- and [LCo(CO) 3] 2 Complexes Compound Temp(°C) Frequency5f(MHz) n e2Qq(MHz) (Ph3P)2Co(CO)2Co(CO)^ * 34.124(9) 33° 22.564(10) 0.136 159.534 12.125(4) (Ph 3P) 2Co(CO) 3BPh 4 30° 3 3' 6 4 1< 8> 0.219 157.754 22.006(7) [(PhO) 3PCo(CO) 3J 2 33.440(30) 0° 22.281(36) 0.0392 156.104 11.219(16) * 33.474(23) 30° 22.295(30) 0.0414 156.230 11.224(14) [(MeO) 3PCo(CO) 3] 2 30.868(100) 0° 20.573(115) 0.0460 144.080 10.372(83) * 30.849(90) 28° 20.550(100) 0.0405 143.982 10.345(70) rrFfO') PfWrn^ i 30.349(80) 0° 20.232(95) 0.0265 141.674 10.169(65) L V " L w o u ( . U L ) ; 3 2 * 30.345(40) 24° 20.228(46) 0.0320 141.655 10.181(28) - 232 -TABLE I. (Contd.) Compound Temp(°C) §t Frequency (MHz) n e2Qq(Mftz) 30.272(12) 0° 20.170(10) 0.0386 141.291 [Bu nPCo(CO) 3] 2 10.163(7) * 30.125(10) 30° 20.080(9) 10.107(6) 0.0333 140.627 * The temperature of the sample was slightly higher than room temperature (22°C) because of heat generated by the modulating current. § The numbers in parentheses are signal to noise ratios. The f i r s t line gives v^, the second line and the third line v t Experimental errors :+ -0.050 MHz . - 233 -TABLE II. Co NQR Data for (RC = CR')Co,(CO)- and Related Compounds Compound Temp(°C) *§ 2 Site Frequency (MHz) n. e Qq(MHz) (HC = CH)Coo(C0), z o n° 17.584(120) I 11.109(95) 0.543 84.655 10.530(100) u 17.593(130) II 11.115(95) 0.559 84.875 10.757(100) -78° 17.754(220) I 11.217(120) 0.544 85.498 10.649(92) 17.788(230) II 11.238(120) 0.561 85.840 10.906(94) -196° 17.856(132) I 11.281(88) 0.554 86.098 10.851(80) 17.909(142) II 11.319(86) 0.580 86.675 11.263(75) - 234 -TABLE II. (Contd.) Compound Temp(°C) & s 0 Site Frequency (MHz) n e Qq(MHz) (BuZC E CH)Co2(CO)6 0° -17.165(220) I 10.850(76) 0.517 82.377 9.945(48) 17.064(220) II 10.805(72) 0.620 82.941 11.243(48) -78° 17.922(106) I 11.342(72) 0.490 85.780 10.018(66) 17.523(100) II 11.088(60) 0.576 84.750 10.958(34) (HOai„ClCCH OH)Co (CO) 30° 18.053(20) a 11.421(17) 0.500 86.518 10.204(15) 17.699(19) B 11.182(15) 0.559 85.368 10.804(10) 0° 18.127(26) a 11.465(27) 0.498 86.856 10.220(30) 17.826(9) B 11.263(18) 0.557 86.043 10.798(20) - 235 -TABLE II. (Contd.) Compound Terap(°C) Site Frequency (MHz) n 2 e Qq(MHz) 13.091(52) I 8.396(40) 10.003(58) 0.738 64.746 13.091(52) II 8.396(40) 0.722 64.684 (CF3C = CCF 3)Co 2(CO) 6 0° 9.747(76) 12.816(46) III 8.250(56) 9.757(76) 0.741 63.523 12.816(46) IV 8.250(56) 9.605(52) 0.731 63.485 14.802(28) I 9.429(20) 0.675 72.596 0° 10.416(15) 15.936(24) II 10.144(16) 0.671 78.101 (PhC = CPh)Co2(CO)6 11.173(11) 14.786(20) I 9.381(14) 0.673 72.257 30° 10.352(10) 15.836(19) II 10.078(10) 11.152(7) 0.673 77.618 - 236 -TABLE II. (Contd.) Compound Temp(°C) §^ 2 Site Frequency (MHz) n e Qq(MHz) f.fars 4 (PhC = CPh)Co2(CO)4 30° I = 12.878(15) 8.480(10) 0.828 64.881 10.780(8) fgfarsCo 2(CO) 6 0° 18.684(14) I 11.830(14) 0.473 89.097 10.114(9) 18.870(18) II 11.940(12) 0.492 90.306 10.564(9) 30° 18.668(16) I 11.832(8) 0.470 89.121 10.140(7) 18.875(17) II 11.947(10) 0.489 90.324 10.552(8) (w-C 7H g)Co 2(CO) 6 0° 17.160(12) I 10.865(7) 0.620 83.463 11.305(7) 17.562(15) II 11.019(5) 0.576 84.923 10.967(6) - 237 -TABLE II. (Contd.) Compound Temp(°C) *§ 2 Site Frequency (MHz) n e Qq(MHz) [(^-C 7H g)Co(CO) 2] 2 0° 15.172(21) I 9.717(9) 0.700 74.684 10.949(8) 14.873(24) II 9.485(12) 0.697 73.212 10.793(7) Co2(CO)g 196° 19.132 I 12.334 0.3149 90.18 8.417 18.671 II 11.818 0.4837 89.30 10.344 a = a phase; 3 = 3 phase * Figures i n parentheses are signal to noise ratio. The f i r s t lines gives v ^ j the second line v 2 and the third line v § Experimental errors :+-0.050 MHz . ** From ref. 3. - 238 -III. Results and Discussion 59 7 2 For Co, which has I = —, both e Qq and r\ can be obtained from the experimental frequencies using a series of approximations for the transition frequencies [27] . The resulting values can be confirmed with a frequency ratio plot [28]. To carry out this procedure a computer program was written (Appendix I of Chapter 3). The program calculates the energy levels and transition frequencies for any values of the nuclear spin, 2 e Qq, and n . 2 The values of e Qq and n were calculated by comparing the measured resonance frequencies with those calculated using the program. 7 59 Since I = for Co each nucleus should give three strong transitions, having frequencies v ^ , v ^ , v ^ . These frequencies are, furthermore, 2 directly proportional to e Qq, so that the ratios v^lv^, v3^ v2 a n c* 2 v ^ / v ^ are independent of e Qq, though dependent on n. Accordingly the program was used to calculate the values for these ratios for a series of values of n differing by 0.0001. The observed ratios were then compared to the calculated ones to obtain n. In a similar fashion, because the 2 • transition frequencies are directly proportional to e Qq, the calculated 2 frequencies were obtained with e Qq = 1. With n obtained,the calculated 2 frequencies were divided into the observed values to evaluate e Qq for 2 the complex. The averaged values of both e Qq and n are given in Tables I and II; the uncertainties represent the spread in their values. - 239 -1. [Y 3PCo(CO) 3] 2 Complexes Recently X-ray structures of three [LCo(C0) 3] 2 complexes (L = (PhO)3P [23], Bu^ [24] and Ph3P [25]) have been reported. Since these results are relevant to our subsequent discussion, i t seems necessary to mention them here. A l l the reported data (Table III) show that these three compounds have the same type of structure 1. Molecules of this complex type are discrete dimers in the solid state. The coordination geometry of each 8 9 half of the molecule is ^ / approximately trigonal bipyramidal O C I with three equivalent CO ligands C occupying the equatorial positions, while the apical sites are occupied by the ligand L and Co(C0)3L. The Co atoms are displaced ca. o 0.1 A out of the planes of the three attached carbons such that the Co-Co-CO angles are ca. 86°. A; Co-Co bond joins the two halves of the molecule such that the Co-Co and Co-P bonds are collinear, and the CO's are arranged i n a staggered conformation. Thus the structure has D3d s y m m e t r y - The Co-Co bond lengths in these compounds are essentially o the same (2.66 - 2.67 A). However, the Co-P bond length in [(Ph0) 3PCo(C0) 3] 2 (2.085(6) A) is considerably shorter than those found in [Ph 3PCo(C0) 3] 2 (2.191(4) A) and [Bu3PCo(C0) ] (2.18(2) A and 2.19(1) A) . - 240 -TABLE III. Molecular Parameters for [Y3PCo(CO)3] Compounds o A. Distance (A) Y = OPh [23] Y = Ph [25] Y = Bu11 [24b]* Y = Bu" [24a]* Co-Co 2.667(6) 2.661(3) 2.665(14) 2.67(1) Co-P 2.085(6) 2.191(4) 2.178(15) 2.19(1) Co-CO 1.74(2) 1.78(1) 1.75(3) 1.77 1.77(2) 1.78(2) 1.76(1) B. Angles (Degrees) Co-Co-CO 85.0(5) 85.8(2) 87.7 86. 85.3(5) 83.9(5) Co-P-Y 118.8(4) 114.2 117.1 112 120.3(4) 118.8(4) Y-P-Y 97.1(5) 104.4 100.9 — 98.5(5) 98.7(5) * This compound has two crystalline forms. The one studied by Bryan and Manning is monoclinic, and the one examined by Ibers is cubic. - 241 -From Table I we can see that a l l the four [Y^CoCCO) 3] 2 n 59 complexes (Y = Bu , OMe, OEt, OPh) gave only three Co nqr signals indicating that the two Co atoms in each compound are crystallographically equivalent. Furthermore, since the frequencies are in the approximate ratio 1 : 2 : 3 , the asymmetry parameter for the Co in these compounds is small, suggesting that the Co atoms are axially symmetric. Although X-ray structures of [(MeO)^PCo(CO)3]2 and [(EtO) 3PCo(CO) 3] 2 are not known, other evidence, such as i r data [12], indicates that they should 59 have the same type of structure (1). Thus, the Co nqr data for these complexes are consistent with their D. , structure. The small values of J d n are apparently due to lattice effects. A. Comparison of Experimental Parameters In Chapter 3 we have mentioned that to a good approximation the molecular efg of a Co atom can be expressed as 2 2 e Qq i = (e Qq) ^mol. v ^ H at,d N d + N d N + x z Y z - N — N, d 2 2 dxy d 2 2 z J x -y ) (1) 2 where (e Qq) , = -278.7 MHz, and N, , etc. are the population at, cl 2 z of the Co 3d orbitals. Brown et. a l . [9, 11] have shown that the 2 e Qq m o^ of the Co atom in X3MCo(C0)4 compounds ( X3 M = ligand containing group IV element M) must be positive and that the quantity in the bracket must be negative. They have chosen the molecular axis (C 3) as the z axis. Since the symmetry of the [Y 3PCo(C0) 3] 2 compounds (3) - 242 -is similar to that of the X^MCoCCO)^ compounds (2), and since, probably, 2 the values of e Qq are also very similar (the covalent o-bonding between two Co's being entirely analogous to the metal-metal bonding in the 2 X3MCo(CO)4 species), i t i s reasonable to consider e QQ^ -^  t 0 D e positive and £ to be negative. i An internal comparison of the data given i n Table I now follows. First of a l l , from Table IV one can see that the +1 effect of an OPh group TABLE IV. Inductive and Conjugative Parameters [27] of Y in the [Y^PCo(CO)3]2 Compounds I c OPh 0.46 - 0.32 OMe 0.31 - 0.63 OEt 0.31 - 0.63 Bu11 - 0.06 - 0.07 - 243 -is greater than that of the Bu11 group. As a result, the electron with-drawing OPh groups in [(PhO)^PCo(CO)3 ] 2 should make (Ph0)3P a worse a-donor than Bu^P. This w i l l cause the value of N, of 3 d 2 z [(PhO) 3PCo(CO) 3l 2 to be lower than that of [Bu^Co^O) 3] 2 . From equation (1) i t is clear that, i f possible back donation i s neglected, the 59 2 Co e Qq of these compounds w i l l then be expected in the order : [(PhO) 3PCo(CO) 3] 2 > [Bu 3PCo(CO) 3] 2 as observed experimentally. However, i f the inductive effect i s the sole factor affecting the 2 value of e Qq, then since the +1 effect of an OR (R = Me, Et) group ( O j = 0.31) is also much higher than that of a Bu11 group (o^. = -0.06), 2 the e Qq of [(RO) 3PCo(C0) 3] 2 (R = Me, Et) should be similar to [(PhO) 3PCo(CO) 3] 2 and differ appreciably from that of [Bu 3PCo(C0) 3] 2 . Neither prediction is borne out indicating that other factors such as conjugative effects of the R groups and back donation from the Co might play some part. From Table IV one sees that a of OR (R = Me, Et) and Bu11 are -0.63 and -0.07, respectively. It is evident that the large +1 effect of OR (R = Me, Et) is well compensated by i t s large -M 2 effect. The net result is that the e Qq's of these complexes are approximately equal. This simply means that again in the absence of back donation,the increased n-release from the OMe and OEt groups causes Of the Co 3d orbitals, d 2 can participate only in ligand-*Co z cr-bonding, while (d , d ) and (d , d ) can participate in xy I I yz xz r * x -y J Co->ligand ir-bonding. - 244 -increased cr-donation from P to Co in spite of the decrease anticipated from inductive considerations. The inductive effect of the OPh group 2 being different accounts for the difference i n e Qq i n the (PhCO^P complex. Using a two-parameter correlation equation of the form : 2„ 2 e "Qq = e QqQ + a ^ o l + b £ c>c (2) one can get, after a least squares treatment (Appendice I and II of Chapter 4), the numerical values of the parameters in equation (2) for the substituents OPh, OMe, OEt and Bu11 at 0° : e2Qq = (145.316 + 1.738) + (39.752 + 5.686)£ a + (23.433 + 4.647)Y a (R=0.99). (3) — u c Since the coefficients a and b' are significantly greater than their respective standard deviations, there appears to be a dependence on both £ and \ a . The linearity of this dependence seems to indicate l i t t l e dependence on synergic back bonding from the Co to the ligand. As equation (3) shows the coefficient of \ o"c is almost 60% that of \ a-j. . This indicates the considerable importance of the conjuga-tive effect of Y. Semin et. a l . [8] have reported that the effect of 59 2 conjugation of X on the Co e Qq in X.jSnCo(C0)4 compounds (X^ = Me^ , Ph^, PhC^, Pt^Cl) seemed to be predominate. A numerical exami-nation of the reported data for X3GeCo(C0)4 (X 3 = C l 3 , B r 3 > I 3 , Ph 3 > Pl^Br) [8-11] at room temperature reveals that this supposition i s also - 245 -correct for the germanium - cobalt complexes as i s shown by the following equation : e2Qq = (99.094 + 0.399) + (65.806 + 1.029)J a + (36.066 + 1.745)Y a (R = 1.00). — u c B. The Effect of Co-P Back Donation Although we have seen that there seems to be l i t t l e Co-P back bonding in these complexes, this i s contrary to popular belief and i t i s therefore of interest to try to establish from another approach the validity of the conclusion. Arguments used to justify n-back donation are generally based on the results of crystal structure determinations and in the present examples are invoked to account for the much shorter Co-P distance i n [(Ph0)3PCo'(C0) ] 2 than i n , say, [Bu^PCo(CO) ] 2 (Table III) . On the basis of infrared measurements Graham [26] has established a set of a- and ir-bonding parameters for various ligands. This procedure has been cr i t i c i s e d [29], however, the parameters at least provide a basis for considering the effects of o-donation and n-accepting separately. This separation has always been a problem in systems of this type. Some of the parameters relevant to our discussion are lis t e d i n Table V. From this l i s t i t is apparent that (Ph0).jP and Bu!^ ? have the same u-withdrawing capacity but different a-donor powers. Similarly, (ROJ^P (R = Me, Et) have smaller a-donor powers but larger n-withdrawal capacities than Bu^P . - 246 -TABLE V. The a- and ir-Bonding Parameters (mdyn/°A) [26] Y3P a TT (PhO)3P - 0.13 0.48 (Me0)3P - 0.36 0.58 (EtO) 3P - 0.38 0.55 Bu^P - 0.48 0.48 CO - 0.06 0.74 Ph3P - 0.15 0.27 e Using a least-squares treatment, a correlation equation relating Qq's of various [Y 3PCo(C0) 3] 2 compounds and Graham's a- and TT-parameters of Y 3P can be obtained : e2Qq = (176.225 + 10.777) + (43.183 + 7.224)a - (30.361 + 21.132) ir (R = 0.99). The large standard deviation of the coefficient b suggests there is l i t t l e correlation with the Tr-parameter assuming that this i s a meaningful quantity. 59 Attempts to detect Co nqr frequencies of other [LCo(C0) 3] 2 compounds (L = Ph.jP, Ph3As, Pb^Sb and (Me0)PPh2) were unsuccessful, although these samples were reasonably pure. - 247 -2. L 2Co(CO) 3 +X complexes At 33° , (Ph3P)2Co(CO)3Co(CO)~ showed three signals at 12.125, 2 22.564 and 34.124 MHz. Using the method for calculation of n. and e Qq 2 mentioned before, one gets n = 0.137 and e Qq = 159.547 MHz. Since these three lines give a good f i t on the frequency ratio chart at n = 0.137, they evidently arise from the same Co atom. At 30° , (Ph3P)2Co(CO)3BPh~ gave two nqr signals at 22.006 2 and 33.641 MHz. From these we obtained n = 0.219 and e Qq = 157.754 MHz. 2 Since the e Qq is very close to that of the f i r s t compound, we can unambiguously assign these frequencies to the Co atom in the cation (Ph 3P) 2Co(CO) 3 . 59 -Attempts to observe the Co nqr frequencies for the Co(C0) 4 anion in (Ph3P)2Co(CO)3Co(CO)~ in the region 7 to 60 MHz were unsuccessful. This is probably because the almost tetrahedral symmetry of Co(C0) 4 would cause any nqr signals to be below 7 MHz ( i f not zero), too low to be detected by the present nqr spectrometer. The i r spectra [14, 15] of the (Pt^P) 2Co(C0) 3X - compounds (X = CoCCO)^, BPh^) show only one CO band attributable to the cation, this indicates a structure 4_ for this cation. The small r\ values (0.1-0.2) of these compounds are also consistent with the proposed structure. The nonzero n values observed indicate the greater sensitivity of that parameter to intermolecular interactions in the solid state [30]'. Since both [LCo(C0) 3] 2 complexes and (Ph.jP) 2Co(C0) 3Co(CO)^ gave smaller - 248 -n values, the larger n value of (Ph3P)2Co(CO)3BPh~ i s probably due to the greater interaction between the anion BPh. and the 4 cation. r oo L c 0 As can be seen from Table I, the values of e Qq for L2Co(CO).jX~ — are close to those of the [LCo(CO) 3] 2 complexes which suggests that the values are not very sensitive to the formal oxidation state of the Co atom. However, further investigation of this series with various ligand L and of the polysubstituted complexes LnCo(CO) ,__*X (n = 3, 4, 5; L = (MeCO^ P etc.) [15] are needed to confirm these ideas. Unfortunately, attempts to observe ~^Co nqr frequencies for (Bu^P)„Co(CO)^X (X = 3 'T Co(CO)4, BPh^) and [(MeO)3P]3Co(CO)2BPh4 were not successful. 59 The Co nqr data of some other Co (I) complexes related to the L 2Co(C0) 3X complexes are listed i n Table VI. It seems that replace-2 ment of CO by Ph.jP tends to increase the e Qq of the resulting * 57 complex . The same tends have been observed in Fe M'dssbauer spectros-copy [31]. Thus the quadrupole spl i t t i n g ( A ) for Fe(CO)^ and trans The result that e Qq of Cl 3SnCo(CO) 3PPh 3 is not higher than that of Cl_SnCo(CO), can be attributed to the larger r\ value of the former 59 2 compound, which trends to lower the Co e Qq of this compound (cf. equation 20 of Chapter 3). - 249 -TABLE VI. 59 Co NQR Data for Some Co(I) * Complexes Compound e2Qq(MHz) n Ref. Ph3GeCo(CO)4 110.34 0.046 [8] 3 Ph3GeCo(CO)3PPh3 (113.5)° — [8] a Ph3SnCo(CO)4 105.38 0.022 t8] a Ph3SnCo(CO)3PPh3 114.53 0.095 [8] a Cl 3SnCo(CO) 4 163.45 0. [ l l ] b Cl 3SnCo(CO) 3PPh 3 163.39 0.208 t8 ] a * a : -196°; b : 25° c : Calculated from the observed frequency v„ = 24.22 MHz by assuming that n = 0.1. (Ph3P) 2Fe(C0) 3 are 2.60 and 2.76 mm/sec, respectively. However the monosubstituted derivative Ph 3PFe(C0) 4 has a A of 2.54 mm/sec. These small differences cannot easily be interpreted on the basis of a- and ir-effects. As mentioned above replacement of CO by Ph-jP i n compounds 2 listed in Table VI causes only a minor change in the e Qq. In contrast to this, the substitution of a Ph3Sn or Ph-jGe group by a Cl 3Sn group 2 causes the e Qq of the resulting complex to increase by more than 50 MHz. This is unexpected on the basis that Cl 3Sn is usually believed to be a strong a-donor [11, 26]. - 250 -3. Substituted Complexes of Co (C0) o £ O 59 Before discussing Co nqr data of the t i t l e compounds, i t seems necessary to mention the solid state structure of their parent compound Co2(C0)g. The X-ray structure [32] of Co2(C0)g consists of dimeric units with the two Co atoms bonded together through the two bridging CO groups (Figure 1). The two Co atoms and the two C atoms of the bridging CO groups are almost tetrahedrally arranged. There are two sets of crystallographically independent molecules per unit c e l l . Figure 1. Molecular Structure of Co„(C0) 8 [32] - 251 -A. f 4fars uCo 2(CO) 6 The molecule fgfars 0Co 2(CO)g [16] shows a marked resemblance o to Co2(CO)g [32]. Each Co atom lie s 0.4 A out of the plane formed by the two C atoms of the bridging CO groups, the As atom and the C atom of the appropriate CO group. The two planes of this type meet at an angle of 88°. Thus each Co atom appears to be octahedrally coordinated, with an axial terminal CO group almost perpendicular to the plane and the bent Co-Co bond trans to the axial CO group. Table VII gives a selection of bond lengths and angles in the molecule. As can be seen the two Co atoms are in very similar environments. 59 From Table II, i t is clear that the Co nqr data for the two Co sites in this molecule are indeed very similar, in good agreement with 59 the X-ray results [16]. A comparison of Co nqr data for Co o(C0) Q 2. o 2 and f 4farsCo 2(CO) 6 (Table II) reveals that both n and e Qq of the Co atoms are f a i r l y insensitive to the ligand substitution. Two reasons can be suggested for this : (i) the a-donation and IT-withdrawing powers of fgfars approach those of CO, ( i i ) the bridging CO groups acts as an effective electron sink. It is d i f f i c u l t to distinguish the consequence of these effects. Although RAsMe2 (R = alkyl group) is a better a-donor and worse ir-acceptor than CO, the presence of an electron withdrawing i 1 group (j' = (j'—^ F2^ F2 ^ n f ^ ^ a r s c a n reasonably be expected to decrease the a-donating but increase the Tr-accepting a b i l i t i e s of the As atoms [33] . Infrared studies on substituted metal carbonyl complexes of fgfars have revealed that the a-donating and Tr-accepting capacities of f, fars are TABLE VII. Some Bond Distances and Angles for f 4farsCo 2(CO) 6 a Dond Distance, A Cod)-As( l ) 2.339 (4) Co(2)-As(2) 2.336 (4) Co(I)--Co(2) 2.482(4) Mean C o - C 1.81(2) Mean C o - C (bridging) 1.94 (2) Moan A s - C 1.95(2) Mean A s - C H 3 1.97(3) Mean C - O 1.14(3) Mean C - F 1.32(3) C(1)=C(2) 1.34 (3) C(l)-C(4) 1.54(3) C(2)-C(3) 1.48(4) C(3)-C(4) 1.57 (4) Bond Angle, deg As(l)-Co(l)-Co(2) 110.1 As(l)~Co(l)-C(6) 92.3 As(l)-Co(l)-C(7) 104.1 As(l)-Co(l)-C(9) 85.7 As(l)-Co(l)-C(10) 157.7 As(2)-Co(2)-Co(l) 109.3 As(2)-Co(2)-C(5) 90.6 As(2)-Co(2)-C(S) 101.5 As(2)-Co(2)-C(9) 86.6 As(2)-Co(2)-C(IO) 157.1 Co(l)-As(l)-C(2) 114.8 Co(l) -As(l) -C(ll ) 117.9 Co())-As(l)-C(12) 119.0 Co(2)-As(2)-C(l) 115.5 Co(2)-As(2)-C(13) 117.3 Co(2)~As(2)-C(I4) 119.0 Co(l)-C(9)-Co(2) 78.6 Co(l)-C(10)-Co(2) 80.3 As(2)-C(IO)-C(2) 134.9 C(2)-C(l)-C(4) 94.8 As(l)-C(2)-C(l) 133.3 C(l)-C(2)-C(3) 93.4 Menu C o - C - O (nonb ridging)' 177.5(2.2). Mean C o - C - O (bridg ins) :40.0(1.9) " Standard deviations for bond distances arc given in parentheses. The standard deviations for angles lie in the range 0 .5 -2 .5° . - 253 -approaching to those of CO [44]. On the other hand, X-ray studies on Crr-C^Hg) CCo 2 (CO) g [34] and [(ir-diene)Co(C0)2]2 [37, 43] have shown that the bridging CO groups act as very effective electron sinks. Mossbauer studies on substituted iron carbonyl complexes also support this view [31]. Thus i t is also possible that the bridging CO groups in f^farsCo 2(CO)^ remove most of the excess 2 electrons from the ligand fgfars, making e Qq of this compound essentially identical to that of Co 0(CO) Q. 2 o B. ( T T - C - H 8 ) C C O 2 ( C 0 ) 6 The X-ray structure of (TT-C_H0) CCO„ (CO) , [34] resembles that of / o 2 0 Co2(C0)g [32]. The norbornadiene ligand bonds to one Co atom via the C=C bonds. The Co(CO)2Co bridging system is non-planar. The Co-C distances in the bridge are not equivalent, those to the substituted Co o o atom being 1.871(3) A and those to the other Co being 1.987(3) A (Figure 2). Co / Co \ 0 co Figure 2. Crystal Structure of (TT-C-,H 0) CCO 0(CO), [34] / o 2. O - 254 -59 From the crystal structure of this compound, six Co nqr frequencies are expected, as observed experimentally. However, as Table II 2 shows the e Qq and n values for the two Co atoms are quite close. This is unexpected because norbornadiene .is a stronger Tr-donor than CO. In the absence of other factors the substitution of C-,H0 for two CO's / o would, i f C-,H0 is a better donor than CO, increase N, and/or / o d xy 2 2 N, and hence increase e Qq. However, e Qq actually decreases. 2 2 This could well result because there can be no back bonding to the norbor-2 nadiene (except to a* orbitals of C-.H0), and, as a result, e Qq would be reduced back again. This lack of back bonding to norbornadiene could also put some extra n-electron density in the bridging CO's, as suggested from the X-ray structure. 2 As mentioned before, the sign of e Qq i n Co2(C0)g is believed to be positive [7]. The substitution of two CO ligands by norbornadiene 2 could possibly cause the e Qq of the substituted Co atom to change sign. However, this is unlikely because the ir-donating norbornadiene ligand would increase (but not decrease) the N, and/or N, of the substituted dxy d 2 2 x -y 2 Co atom, which w i l l , in turn, increase (not decrease) the e Qq value. C. [ ( T T - C 7 H 8 ) C C O ( C O ) 2 ] 2 This compound is rather different from the others considered thus far. Its i r spectrum [35] has been interpreted to show that the molecule contains a planar rather than puckered Co(CO)2Co bridge, and hence - 255 -resembles [iT-C^FeCCO) 2 ] 2 [36] rather than Co 2(CO) g [32]. The i r data suggest also that the compound exists as the cis-tautomer in the solid state. Similar structures were deduced for some other bis-diene derivatives. Amongst these was [(TT-GVH0) CO ( C 0 ) o ] o b o z Z (CgHg = cyclohexa-1,3-diene), whose solid state structure [37] does indeed show the planar bridge (Figure 3). Accordingly one could reasonably 59 c expect the Co coupling constants for [(ir-C^H 0) Co(C0) o]_ to be quite b c different from any of Co 2(CO) g, fgfars Co 2(CO) 6 or ( ir-^Hg) Co 2(CO) 6. The nqr frequencies of [(ir-C^Hg) Co(CO) 2] 2 are given in Table II. There are evidently two crystal environments of the Co atoms, 2 for six lines were obtained, resulting in two sets of values for e Qq and 2 n . And, as expected, both e Qq and n are considerably different from those of compounds previously considered. Because of the radical structural changes between this and the other complexes i t i s d i f f i c u l t to draw any further conclusions from the Figure 3. Crystal Structure of [ (ir-C,H0) Co(CO) _] „ [37] - 256 -coupling constant changes. Clearly, however, i t should be useful to obtain the nqr spectra of other bis-(diene) derivative(s) of the same configura-tion (such as [(-rr-CgHg) Co(CO) 2] 2 mentioned earlier) to determine whether these coupling constants are characteristic of this type of complex. 4. (RC = CR')Co2(CO)6 Complexes 59 In the present investigation, Co nqr data for (RC = CR')Co2(CO)g complexes have been measured for the f i r s t time. The 2 observed frequencies and their assignments, as well as n and e Qq values are listed in Table II. These correspond to the three strongest transitions. With the exception of fgfars(PhC = CPh)Co2(CO)4 and (CF^ = CCF3)Co2(CO)g, a l l the acetylene derivatives in Table II showed six 59 allowed Co nqr signals. This indicates that the two Co atoms in these compounds are crystallographically nonequivalent. In the case of (PhC = CPh)Co2(CO)6, this i s in agreement with the X-ray results. Sly [38] has reported that the two Co atoms in the molecule are considerably different, as shown in Figure 4. For instance, the distance between one o Co atom and the two carbon atoms of PhC = CPh are 2.01 and 2.02 A , o respectively, while the other Co atom is 1.93 and 1.89 A, respectively, apart from these two carbon atoms. Although the crystal structures of other (RC E CR')CO 2(C0) 6 complexes (RC = CR' = HC E CH, BufcC = CH, H0CH2C = CCH20H, CF3C = CCF^ are not known, the similarity between the infrared CO bands of these compounds and those of (PhC = CPh)Co„(C0)i, [18-21] suggests that they a l l have the same type of butterfly structure. The possible structure of f 4fars b(PhC E CPh)Co2(C0)4 has been described in - 257 -- 258 -Chapter 2. This compound is believed to have a structure similar to that of f 4fars bCo 2(CO)g (Figure 2) with the two bridging CO groups replaced by the PhC = CPh molecule . A. Forbidden Transitions Since the asymmetry parameters of these compounds are quite large, their ground state wave functions are linear combinations of the basis 1 3 functions (having m_ = + + -j, .. ., + I) (cf. equation (1) of Chapter 3). Thus possible "forbidden" lines could be weakly allowed at (v^ + v 2 ) and ( v 2 + v^)• In the present work, these lines have been observed for (HC = CH)Co 2(C0) 6 and (Bu^ = CH)Co2(CO)6 (Table VIII). These are the 59 f i r s t cases where such transitions in Co have been detected. They are extremely useful in making assignments. At 0°, (HC = CH)Co2(C0)g gives six strong transitions at 10.530, 10.757, 11.109, 11.115, 17.584 and 17.593 MHz. There are 6' altogether -— = 360 ways of assigning the frequencies to the two Co sites. In view of the fact that is usually higher than both and v 2 (cf. Chapter 3), the two higher frequency signals at 17.584 and 17.593 MHz can be unambiguously assigned as v^(s). But there s t i l l remain 12 different ways of assigning the six frequencies. Fortunately, the two forbidden frequencies f^ and f 2 for each site in (HC = CH)Co2(C0)g have been observed (Table VIII) and this simplifies the assignment. The general procedure i s as follows.: (a) assignment of f^ and f 2 , (b) assignment of v^ , (c) assignment of v 2 > and (d) complete assignment of v^ , v 2 and v^ . In the following discussion, the two Co sites are arbitrarily labeled I and II. ; TABLE V I I I § . Forbidden Frequencies (MHz) of (RC= CR')Co 2(C0) 6 (R = R' = H; R = Bu*1, R Compound Temp(°C) Si t e V l V2 V3 * f l * f2 0° I 10.530(100) 11.109(95) 17.584(120) 21.625(10) 21.639 28.685(2) 28.693 (HC=CH)Coo(C0)^ I b II 10.757(100) 11.115(95) 17.593(130) 21.876(11) 21.872 28.716(2) 28.724 -78° I 10.649(92) 11.217(120) 17.754(220) 21.873(16) 21.863 28.956(4) 28.971 II 10.906(94) 11.238(120) 17.788(230) 22.155(18) 22.144 29.006(4) 29.026 196° I 10.851(80) 11.281(88) 17.856(132) 22.170(11) 22.132 29.155(6) 29.137 II. 11.263(75) 11.319(86) 17.909(142) 22.612(12) 22.582 29.224(6) 29.228 TABLE VIII . (Contd.) Compound Temp(°C) S ± t e V l V2 v 3 f i * f 2 * (Bu tCECH)Co„(CO)^ z o 0° 20.795(4) I 9.945(48) 10.850(76) 17.165(220) 2 Q 22.047(4) II 11.243(48) 10.805(72) 17.064(220) 2 2 -78° 21.353(7) I 10.018(66) 11.342(72) 17.922(106) ' 3 5 f ) 22.015(8) II 10.958(34) 11.088(60) 17.523(100) 0 0 n l r 22.046 * F i r s t l i n e gives observed forbidden frequency, second l i n e gives c a l c u l a t e d one. § Figures i n parentheses are s i g n a l to noise r a t i o s . - 261 -By definition (Chapter 3), f^ must be equal to (v^ + v 2) , and f 2 equal to (v 2 + v^). Since is always greater than and v,,, hence f 2 is always greater than f^. On this basis, we can assign the two forbidden frequencies at 21.625 and 21.876 MHz to f^(s) for the two Co sites. The other two higher frequency but lower intensity lines at 28.685 and 28.716 MHz can be unambiguously assigned to f 2 ( s ) . Previously we have assigned the two lines at 17.584 and 17.593 MHz to v^(s). Then since v2 ~ 2^ ~ V3' t b e t W ° v2^ s^ raust l i e between 11.092 and 11.132 MHz. We therefore assign the two lines at 11.109 and 11.115 to v 2 ( s ) . The remaining two frequencies at 10.530 and 10.757 MHz are immediately assigned as v^(s). After this treatment, there are only four possible ways of assigning the six lines as shown in Table IX. Within the experimental errors a l l the four assignments f i t well with the forbidden frequencies and also the frequency-ratio chart, and hence none of them can be excluded. Should the uncertainty i n frequency measurements be reduced to less than 5 KHz, then a more definite assignment would be possible. If one uses the frequency-ratio chart to assign the six allowed frequencies, at least 12 p o s s i b i l i t i e s have to be considered, hence the > treatment is more laborious. Thus, the forbidden frequencies are quite useful for assigning and confirming the allowed frequencies. 2 A l l the above four assignments give identical n and e Qq values - 262 -TABLE IX. Possible Assignment of v r v 2 and v 3 for (HC^CH)Co2 (C0) 6 at 0 o Assig ;nment Site V l V2 V3 V l + V2 V2 + v3 (a) I 10.530 11.109 17 .584 21.639 28.693 II 10.757 11.115 17.593 21.872 28.708 (b) I 10.530 11.115 17.584 21.645 28.699 II 10.757 11.109 17.593 21.866 28.702 (c) I 10.530 11.109 17.593 21.639 28.702 II 10.757 11.115 17.584 21.872 28.699 (d) I 10.530 11.115 17.593 21.645 28.708 II 10.757 11.109 17.584 21.866 28.693 within measurement errors. However, in Tables II and VIII only the b< f i t assignments are li s t e d . For (Bu 'c = CH)Co2(C0)g the same essential procedure was used. In this case, however, the spacing between the lines was considerably greater, and the two sets of three lines could be unambiguously chosen. The appropriate data are in both Table II and Table VIII. B. Spectrum of (H0CH2C = CCH20H)Co2(C0)6 At 0 , this compound shows three strong and three weak signals. Since the signals which arise from two different sites of a nucleus in the same compound should have similar intensities, the presence of two sets of - 263 -different intensity signals suggest that there may be two different solid phases (a phase and (3 phase) in the sample. At higher temperature, the population of g phase increases, such that at 30° the population ratio of the two phases i s a : $ = 1 : 1. C. The Assignment of Transition Frequencies of Other Compounds At 0°, the compound f 4fars b(PhC = CPh)Co 2(C0) 4 gives only 59 three Co nqr signals indicating that the two Co atoms are identical at this temperature. Here, as before, the assignment and the calculation 2 of n and e Qq are based on the frequency-ratio chart. At the same temperature, (CF-jC = CCF 3)Co 2(C0) 6 shows eight strong signals in three well-separated regions, viz. 8.2-8.4 MHz , 9.6-10.0 MHz and 12.8-13.1 MHz suggesting that they are arising from three different transitions. The four signals in the region 9.6-10.0 MHz indicate the presence of four Co sites. However, for four Co sites there should be 12 signals. Attempts to observe other signals from 7 to 60 MHz were not successful. Since the two lines at 8.2-8.4 MHz and the two lines at 12.8-13.1 MHz are quite broad, i t seems reasonable to assume that each broad line is composed of two signals. This would give the 12 signals required for the four Co sites (I to IV, arbi t r a r i l y labeled). Using the frequency-ratio chart, the best assignment can be obtained as shown in Table II. 2 The almost identical values of n and e Qq for the four sites indicate that these sites are very similar. In this table the signals at - 264 -12.8-13.1 MHz and 8.2-8.4 MHz have been assigned to and v 2 , respectively. From Figure 4 of Chapter 3 we see that and v 2 are very insensitive to the change of r\ values and that they are nearly parallel. This lends further support to the above assumption that each line in the 8.2-8.4 and 12.8-13.1 MHz region is accidentally superposed of two signals. From the same figure we also see that i s very sensi-tive to the change of n. This explains the occurrence of four well-separated signals in the 9.6-10.0 MHz region in spite of the high similarity of the four Co sites. D. Lattice Effects in (RC E CR')Co2(CO)6 From Table II one can see that, with the exception of (PhC E CPh)Co2(CO)6, a l l the (RC E CR*)Co2(CO)6 complexes show very small lat t i c e effects. However, the coupling constants for the two sites in (PhC E CPh)Co2(CO)6 are quite different indicating that the two Co sites in this compound are considerably different. In the crystal structure of this compound the unit c e l l has one type of molecule, but the torsional angles around the two C atoms of the C = C bond are quite different. This would explain the quite different values of coupling constants for the two sites. - 265 -E. Substituent Effects in (RC = CR') Co,, (CO)-L o As can be seen from ; Table II, (PhC E CPh)Co2(CO)6 gives 2 smaller e Qq but larger n values than (HC = CH)Co2(CO)6. Two effects should be considered, viz. steric and electronic. To see i f the steric effects of R and R' are responsible for this, some other compounds with bulky R and/or R' groups have been studied. Both (Bu C = CH)Co2(CO)6 and (HOCH2C = CCH2OH)Co2(CO) (whose bulky R and/or R' groups have 2 very small and a values) give e Qq and n values very close to those of (HC = CH)Co2(CO)g suggesting that steric effects are not impor-tant. The fact that (CF0C = CCF„)Co0(CO), (whose R = CF„ has a T = J J L b 3 I 2 0.46 and a = 0) gives very small e Qq and very large n indicates that the inductive effect i s probably responsible for the changes. 2 From Table X i t appears that +1 effect lowers the e Qq and 2 raises the n values. It has been shown that the molecular e Qq of Co can be expressed as equation (1) (Chapter 3). Sheline et. a l . [7] have 2 reasoned that the (e Qq) n for Co„(CO) 0 must be positive. In view of mol. 1 o the structural similarity between (RC E CR')Co„(C0), and Co„(CO) 0, the L b L o 2 (e Qq) of the former compounds can be assumed to be positive, mo ±. r 59 TABLE X . Co NQR Data of (RCECR')Coo(C0), Z O and I. aT & V a I ^ c of R and R' Compound 2 , * e Qq(MHz) * n l°i I a L c (HCECH)CO 2(CO) 6 84.765 0.551 0.00 0.00 (ButCECH)Co2(CO)6 82.659 0.569..... -0.08 -0.09 i (HOCH CECCH.OH)CO„(CO), z Z 6 86.450 0.528 0.03 0.03 ho <y\ a\ (PhC=CFh)Co_(CO), Z 0 75.350 0.673 0.20 -0.20 I (CF3C=CCF3)Co2(CO) 64.110 0.733 0.92 0.00 Average values at 0°. - 267 -2 The lowering of e Qq by +1 groups could be due to (i) the decrease of N, and/or N. , ( i i ) the increase of N, and/or d d „ „ d xy 2 2 yz J x -y 7 N, , or ( i i i ) the increase of N, . Sheline et. a l . [7] have shown that cl d 0 xz 2 z the z axis of Co2(CO)g must be along the axial CO, the Co atom and the bent bond. The same situation i s expected for (RC = CR')Co2(CO)g. Hence, d 2 w i l l be on this axis. As the bonding between alkyne and Co z occurs on the xy plane, the R and R' groups have l i t t l e effect on the d 2 orbital. Therefore ( i i i ) can be excluded. Furthermore, since +I; z groups w i l l decrease (not increase) the electron density on the Co atoms, ( i i ) can be ruled out as well. On the other hand, (i) w i l l decrease the 2 e Qq of these compounds (cf. equation (1)), as observed experimentally. 2 Thus, we conclude that the lowering of e Qq by +1 groups is due to the depopulation of d and/or d 2 2 orbitals of the Co atom. X ^ x -y 59 From Chapter 3 we see that for Co, n can be expressed as ! [ N d - N d , v xz yz' N d 2 + 2 z N d + N d xz yz d d „ « xy 2 2 x -y The increase of n values by +1 groups can be due to (i) to ( i i i ) mentioned above or (iv) the increase of the population difference between 2 d and d orbitals. Figure 5 is a plot of n. against e Qq for the xz yz ° r ° five compounds studied. As can be seen there i s a good correlation between them (r = 0.98). This suggests that the increase of n by +1 groups i s 2 probably due to the same reason as the decrease of e Qq, i.e. the - 268 -2n e Qq 86.00 ^ 83.25 80.50 77.75 75.00 J 72.25 A 69.50 66.75 4 64.00 0.530 0.572 0.614 0.656 0.698 0.740 Figure 5. A Plot of e Qq vs. n for (RCECR')CO 2(CO) 6 (r = -0.98, intercept = 140.990+7.574, slope = -102.035+12.298.) - 269 -depopulation of d and/or d „ „ orbitals of the Co atom, xy 2 2 x -y Using the least squares treatment mentioned before, a two parameter equation relating e2Qq of (RC = CR')Co2(CO)g and £ cr^ and £ of the R and R' groups can be obtained : e2Qq = (85.015 + 0.901) - (22.350 + 1.834)£ a + (30.105 + 7.939)T a (R = 0.99). — u c This implies that there are strong correlation with \ cr^ . and tolerable correlation with £ a . Since the coefficient a i s negative, there should be a negative dependence of e Qq on £ cr , as found (Table X ). This deduction is also consistent with our previous conclusion that increa-sing ) a T w i l l decrease N, and/or N, of Co atom, which in J- d d 0 0 xy 2 2 x -y 2 turn w i l l decrease the e Qq of the compound. On the other hand, since the coefficient b is positive, there should be a positive dependence of 2 r e Qq on I a . To confirm this point, further studies on other (RC H CR')Co2(C0)g compounds containing +M substituents such as F, Cl, Br or I [39-42] are desirable. Unfortunately, the attempts to detect 59 Co nqr signals for (IC = CI)Co o(C0), failed. z o - 270 -F. Effect of Ligand Substitution A comparison of nqr data of Co2(C0)g with those of b c f.fars Co„(C0), and (TT-C-,R0) CO o(C0), (Table II) reveals that both H Z O / o Z o 2 e Qq and n of Co atoms are not very sensitive to the ligand substitu-tion. In contract, from the results of (PhC = CPh)Co2(C0)g and b 2 fgfars (PhC = CPh)Co2(CO)4, i t appears that the e Qq and n of Co atoms in this system is more sensitive to ligand substitution. The exact reason for this change is not clear. However, i t was pointed out above that the torsional angles around the two C atoms of the C = C bond are quite different (Figure 4). Since the AsMe2 groups are bulkier than the terminal CO groups, i t is possible that replacement of two CO groups by a fgfars ligand w i l l cause the torsional angles to change by several degree. This could account for the relatively large ligand substitution effect in fgfars 1 3(PhC = CPh)Co 2(C0) 4. To c l a r i f y this situation, further nqr studies on other (L-L) b(PhC = CPh)Co 2(C0) 4 complexes (L-L = f^fos, 59 etc.) are necessary. Unfortunately, attempts to obtain the Co nqr data for f/,fos (PhC = CPh)Coo(C0), were unsuccessful. - 271 -IV. Summary 59 New Co nqr data for the following types of compounds have been reported : [Y 3PCo(CO)3] 2 > L2Co(CO)^x", (RC E CR')Co2(CO)6 and substituted derivatives of Co o(C0) o and (PhC = CPh)Co„(CO),. 2. o 2 0 The parameters for [Y 3PCo(CO) 3] 2 complexes are very similar to those of the X3MCo(CO)4 complexes. The effects of Y on the 5 9Co nqr data of the [Y 3PCo(CO) 3] 2 compounds are found to be both inductive and conjugative. However, i t has been shown that there i s l i t t l e TT back-bonding between Co and P i n these compounds. 2 In spite of the positive charge on the Co atom, the e Qq's of the L2Co(CO)3X compounds are very close to those of the covalent [Y 3PCo(CO) 3J 2 and X3MCo(CO)4 complexes, indicating that the e Qq of Co is not very sensitive to the formal oxidation state of the Co atom(s). 59 The Co nqr data of some substituted derivatives of Co 2(CO) g have been interpretted i n terms of their known crystal structures. 59 The effects of R and R' on Co nqr data of (RC = CR')Co2(CO)g have also been investigated. Whereas steric effect i s found to be unimportant, both inductive and conjugative effects are probably 2 responsible for the changes in e Qq and n values. The fact that +1 2 substituent groups decrease the e Qq of (RC = CR')Co2(CO)6 suggests that the bonding between alkyne and the Co atom must involve the Co 3d or xy 3d 2 2 orbital, x -y - 272 -59 c Co nqr data for f^f arsCo 2(C0) 6 > (u-C^g) Co 2(CO) 6 and 2 Co2(CO)g reveal that both e Qq and n of Co atom In this system are less sensitive to ligand substitution. In contrast, the results of (PhC = CFh)Co 2(CO) 6 and f^fars^PhC = CPh)Co2(CO)4 show that these parameters are more sensitive to the ligand substitution. The possible reason for this change was discussed. - 273 -References 1. D.V. Ehrenstein, H. Kopfermann and S. Penselin, Z. Physik, 159, 230 (I960). 2. F. Calderazzo, R. Ercoli and G. Natta, "Metal Carbonyls : Preparation, Structure and Properties" i n "Organic Syntheses via Metal Carbonyls", edited by I. Wender and Piero Pino, Vol. 1, Chapter 1 (1968). 3. W. Hlibel, "Organometallic Derivatives from Metal Carbonyls and Acety-lene Compounds" in "Organic Syntheses via Metal Carbonyls", edited by I. Wender and P. Pino, Vol. 1, Chapter 2 (1968). 4. S.D. Robinson, "Transition metal complexes containing P, As, Sb and Bi donor ligands", M.T.P. International Review of Science, Inorganic Chemistry, Series One, Vol. 6, p. 63 (1972). 5. R.D.W. Kemmitt, "Olefin and Acetylene Complexes", ibid, p. 227 (1972). 6. E.W. Abel and F.G.A. Stone (Senior reporters), A Specialist Periodical Report, Organometallic Chem., Vol. 1, Chapters 6 to 8 (1972), The Chem. Soc. (London). . 7. E.S. Mooberry, H.W. Spiess, B.B. Garrett and R.K. Sheline, J. Chem. Phys., 51, 1970 (1969). 8. A.N. Nesmeyanov, G.K. Semin, E.V. Bryukhova, K.N. Anisimov, N.E. Kolobova and V.N. Khandozhka, Bulletin Division of Chemical Sciences, 1792(1969), translated from Izvestiya Akademii Nauk SSSR, Seriya Khimicheskaya, No. 2, 1936 (1969). 9. T.L. Brown, P.A. Edwards, C.B. Harris and J.L. Kirsch, Inorg. Chem., 8, 763 (1969). 10. J.D. Graybeal, S.D. Ing and M.W. Hsu, Inorg. Chem., 9^, 678 (1970). 11. D.D. Spencer, J.L. Kirsch and T.L. Brown, Inorg. Chem., £, 235 (1970). 12. A. Sacco, Ann. Chim. (Rome), 43, 495 (1953); CA. 48, 5012 (1954). 13. G.H. Whitfield and H.W.B. Reed, J. Chem. Soc, 1931, 1940 (1954). 14. (a) W. Hieber and W. Freyer, Chem. Ber., 91, 1230 (1958); 93, 462 (1960). (b) A. Sacco and M. Freni, J. Inorg. & Nucl. Chem., j$, 566 (1954). 15. S. A t t a l i and R. Poilblanc, Inorg. Chim. Acta., 6J, 475 (1972). - 274 -16. (a) J.P. Crow, W.R. Cullen, W. Harrison and J. Frotter, J. Am. Chem. Soc, 92, 6339 (1970). (b) J.P. Crow and W.R. Cullen, Inorganic Chem., 10, 2165 (1971). 17. G. Winkhaus and G. Wilkinson, J. Chem. Soc, 602 (1961) 18. R. Markby, I. Wender, R.A. Friedel, F.A. Cotton and H.W. Sternberg, J. Am. Chem. Soc, 80, 6529 (1958). 19. U. Kruerke and W. Hubel, Chem. Ber., 94, 2829 (1961). 20. H.W. Sternberg, H. Greenfield, R.A. Friedel, J.H. Wotiz, R. Markby and I. Wender, J. Am Chem. Soc, 76, 1457 (1954). 21. J.L. Boston, D.W.A. Sharp and G. Wilkinson, J. Chem. Soc, 3488 (1962). 22. William M. Dehn, J. Am. Chem., 1598 (1911). 23. J.A.M. Case, Ph.D. Thesis, University of Wisconsin (1967). 24. (a) R.F. Bryan and A.R. Manning, Chem. Comm., 1316 (1968). (b) J.A. Ibers, J. Organometal. Chem., 14, 423 (1968). 25. A.S. Foust, Jr., Ph.D. Thesis, University of Wisconsin (1970). 26. W.A.G. Graham, Inorg. Chem., ]_, 315 (1968). 27. G.K. Semin and E.I. Fedin, Zh. Strukt. Khim., JL, 252 (1960). 28. H.G. Robinson, Phys. Rev., 100, 1731 (1955). 29. L.M. Haines and M.H.B. Stiddard, Advan. Inorg. Chem. Radiochem., 12, 53 (1969). 30. M.G. Clark, Chem. Phys. Lett., 13, 316 (1972). 31. M.G. Clark, W.R. Cullen, R.E.B. Garrad, A.G. Maddock and J.R. Sams, Inorg. Chem., 12, 1045 (1973). 32. G. Gardner-Summer, H.P. Klug and L.E. Alexander, Acta. Cryst., 17, 732 (1964). 33. E.W. Abel and F.G.A. Stone, Quart. Rev., 23, 325 (1969). 34. F.S. Stephens, J. Chem. Soc, (Dalton), 1754 (1972). 35. P. McArdle and A.R. Manning, J. Chem. Soc. (A), 2123 (1970). 36. A.R. Manning, J. Chem. Soc. (A), 1319 (1968) - 275 -37. F.S. Stephens, J. Chem. Soc. (Dalton), 1752 (1972). 38. W.G. Sly, J. Am. Chem. Soc, 81, 18 (1959). 39. W.J. Middleton and W.H. Sharkey, J. Am. Chem. Soc, 80, 803 (1958). 40. 0. Wallach, Ann., 203, 88 (1880). 41. K.A. Koffmann and H. Kirmreuther, Ber., 42, 4235 (1909). 42. V. Grignard and K. Teheoufaki, Compt. rend., 188, 357 (1929). 43. F.S. Stephens, J. Chem. Soc. (A), 2745 (1970). 44. (a) J.P. Crow and W.R. Cullen, unpublished results. (b) W.R. Cullen, Adv. i n Inorg. and Radiochem., 15, 368 (1972). 

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}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            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:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0060069/manifest

Comment

Related Items