7 AN ANALYSIS (IF THE SELECTIYITY OF THE OLEFIN ' HYDROGENATION CATALYST M TRISTRlPHENYLPHOSPHINECHLORORHODIUM (I) SUPPORTED 0H 2% DIVINYLBENZENE CROSSLINKED POLYSTYRENE BEADS A Disserfoflon for ”10 Degree of pk. D. MICHIGAN STATE UNIVERSITY Edward Marsh Sweet I976 LIBRARY ‘- IV , ”we U11 five; slt‘] This is to certify that the ‘ thesis entitled AN ANALYSIS OF THE SELECTIVITY OF THE OLEFIN HYDROGENATION CATALYST . TRISTRIPHENYLPHOSPHINECHLORORHODIUM (I) SUPPORTED ON 2% DIVINYLBENZENE CROSSLINKED POLYSTYRENE BEADS presented by Edward Marsh Sweet has been accepted towards fulfillment of the requirements for Ph. D. deg”? in Chemistry Major professor Date /’ / 7 0-7 639 ; ’m“ 72 An ' factors tion se POlyst} Pu 96ndan1 ’1 giving Equili': g defici triphe hydrOg itY fr RhCl (g eXCess attri] Produ tion ABSTRACT AN ANALYSIS OF THE SELECTIVITY OF THE OLEFIN HYDROGENATION CATALYST TRISTRIPHENYLPHOSPHINECHLORORHODIUM(I) SUPPORTED ON 2% DIVINYLBENZENE CROSSLINKED POLYSTYRENE BEADS BY Edward Marsh Sweet An investigation was conducted to determine what factors cause the observed alteration from homogeneous solu- tion selectivity when RhCl(PPh3)3 is supported on polystyrene—2% divinylbenzene copolymer beads (1,2). An initial study used beads on which a portion of the pendant phenyl groups were substituted with CHZPPh2 groups, giving a pendant group equivalent to benzyldiphenylphosphine. Equilibration with [RhCl(COE)2]2 (COE=cyclooctene) in deficiency or in excess, followed by 1/3 equivalent of triphenylphosphine per equivalent of rhodium, gave an active hydrogenation catalyst. Significant alteration in selectiv- ity from an otherwise similar catalyst prepared with RhCl(PPh was noted for the catalyst prepared from an 3)3 excess of [RhCl(COE)2]2. Alteration in selectivity was attributed to different ratios of phosphine to rhodium produced by different methods of preparation. Interpreta— tion of overall selectivity was not possible since several species phosphil A 1 groups : equivale an exce: hydroge1 catalys1 indicat< eXperiel There '1; volume. rES’cric- Edward Marsh Sweet species containing mixtures of bead—phosphine and triphenyl- phosphine were possible. A bead support was made containing pendant phenyl groups substituted with Pth, giving a supported phosphine equivalent to triphenylphosphine. Equilibration with either an excess or a deficiency of RhCl(PPh3)3 gave an active hydrogenation catalyst. Comparison of the supported catalyst selectivity with homogeneous catalyst selectivity indicates that the active site of the supported catalyst experiences a small significant phosphine concentration. There is evidence for selectivity with changing molecular volume. At least some of the selectivity is caused by restricted diffusion of alkene to the active site. REFERENCES l. R. H. Grubbs and L. C. Kroll J. Am. Chem. Soc., _3, 3062 (1971). 2. R. H. Grubbs, L. C. Kroll, and E. M. Sweet J. Macromol. Sci; Chem., AZ, 1047 (1973). AN ANALYSIS OF THE SELECTIVITY OF THE OLEFIN HYDROGENATION CATALYST TRISTRIPHENYLPHOSPHINECHLORORHODIUM(I) SUPPORTED ON 2% DIVINYLBENZENE CROSSLINKED POLYSTYRENE BEADS BY Edward Marsh Sweet A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1976 Fe‘ sons ef for whi F0 my rese Project past, w thank D the ele H. Su f and the Chemist the dig the dis I DUrham project Michiga ACKNOWLEDGMENTS Few successful undertakings are the result of one per- sons efforts alone. Many have contributed to this work, for which I am grateful. Foremost among those mentioned must be Dr. R. H. Grubbs, my research preceptor, who envisioned and directed the project, and the members of my committee, both present and past, who provided much insight and many ideas. I wish to thank Dr. P. Rasmussen and Mr. V. Shull for assistance with the electron microprobe work, Mr. Bob DeVries and Mrs. S. C. H. Su for assistance with some of the experimental work, and the members of the Michigan State University Analytical Chemistry Consulting Service for design and construction of the digital time logger, and Mrs. Shirley Goodwin for typing the dissertation. I also thank the United States Army Research Office, Durham and Michigan State University for support of the Project, and the Dow Chemical Company, Inc., Midland, Michigan, for the gift of polystyrene beads. ii II. RE III- cc IV- E) TABLE OF CONTENTS I. INTRODUCTION..... ...... ...... ............. . ..... General. ..... .................... ... ....... The Nature of the Support.... ............. .. The Hydrogenation of Olefins With Tristri- phenylphosphinechlororhodium(I).. ......... Mechanism....... ....................... .. Consideration of a Rate Expression... ..... II. RESULTS AND DISCUSSION................. ......... Bead Preparations ........................... Characterization of Beads... ........ ... ...... Swelling of Ratios. .... ................... Density........ ............... .. ..... ..... Microprobe Analysis............ ..... . . Studies on Benzyldiphenylphosphine Equivalent Beads........... ....... . .......... ........ Beads Hydrogenations ......... ............. Homogeneous Catalyst Hydrogenations ....... Analysis of Results on Benzyldiphenylphos— phine Equivalent Beads. ........ ........... Studies on Triphenylphosphine Equivalent Beads ......... ........ ............ ....... Bead Hydrogenations. ................... .. Determination of Constants Within the Rate Expression for the Homogeneous Catalyst... Analysis of Results for Triphenylphosphine Beads.. ................ ........... ........ III. CONCLUSIONS...... ............. . .......... ....... Suggestions for Further Work... ..... ......... IV. EXPERIMENTAL....... ....... ..... ......... . ....... General.. ........ .... ........................ Analytical.......... ........... ........... Materials ....... .... ........ .............. 40 44 44 47 56 73 74 76 76 76 . 7 r;;(‘ 3* _ TABLE OF CONTENTS--cont1nued Page Special Equipment..... ........ .............. 79 Preparations ............................... 80 Tristriphenylphosphinechlororhodium(I)... 80 Benzyldiphenylphosphine. ...... ........ 81 Biscyclooctenerhodium(I) Chloride Dimer [RhCl(cyclooctene) 22] or [RhCl(COE) 22] 83 Preparation of Beads......2 ....... ... ...... 84 Batch 1 (Benzyldiphenylphosphine equiva— lent beads) ..... ....... ..... .......... 84 Chloromethylation. ...... ..... ..... .... 84 Phosphination........ ....... .... ..... . 85 Batch 1A...... ....... ... ....... ..... ..... 86 Batch 1B......... ...... . .......... . ...... 87 Batch 2....... ............. . ...... .. ..... 88 Batch 3 ...... .... ...... ..... ......... .... 89 Batch 4 ..... .................... ......... 89 Batch 5................... ..... ... ....... 91 Bromination..... .............. ........ 91 Phosphination... ........... ........... 92 Batch 5A........ ................ ......... 93 Batch 5B...... ........................ ... 93 Densities................. ......... . ....... . 94 Swelling Ratios....... .......... ............ 95 Microprobe Analysis............ .......... ... 95 Preparation of Beads Sectioning in a Matrix .......................... ...... 95 Polymerization of Beads into Epoxy.... 95 Polymerization of Beads into Styrene.. 96 Sectioning ......... ........ ......... 97 The Half— bead Method .................. 98 Determination of Elemental Radial Distribu- tion..... ................... .......... 98 Attempted Determination of the Phosphorous to Rhodium Ratio. ................. ....... 99 Hydrogenation Procedure ..... . ...... . ........ 101 General.. ............. .. .......... . ..... lOl Bead Hydrogenations ..... . .............. .. 102 Hydrogenations Using Homogeneous Tristri— phenylphosphinechlororhodium(I).. ..... 105 Comparison of Rates for Phosphorous to Rhodium Ratios Less than Three ........... 107 Evaluation of Data ..... .. ........... . ..... .. 107 Calculation of Relative Rates ............ 108 Homogeneous Catalysis Data ............... 109 iv TABLE 01 REFEREM APPENDIl TABLE OF CONTENTS--continued REFERENCES ............................ . ..... . ........ APPENDICES I. SAM—-An Automated Hydrogenation Apparatus ...... II. Unpublished Computer Programs Used............. Page 112 115 126 LIST OF TABLES TABLE Page 1. Synopsis of Bead Preparations................... 18 2. Swelling Ratios ..... . .......... . ............ .... 19 3. Phosphorous to Rhodium Ratios by Microprobe Analysis ...... .... ........ ... ........ . ..... ..... 30 4. Relative Rates of Hydrogenation Using Benzyl— diphenylphosphine Equivalent Beads of Various Preparations... ............ ... ....... ........... 32 5. Rate versus Run Sequence-—Batch 1B After Adding Triphenylphosphine............. ..... ............ 35 6. Rates of Hydrogenation for Variation in Phos— phine to Rhodium Ratio............... ...... ..... 37 7. Effect of Loading on Relative Rates............. 46 8. Relative Rates of Reduction in Several Solvents of Different Swelling Ratio for Batch 5A ..... ... 46 9. Rate versus Catalyst Concentration for 1.0 M Cyclohexene.............. ................ ....... 49 10. Observed Values of k6 and K Obtained from Alkene Dependence of Rate With Small Quantities of Added Triphenylphosphine......... ...... ...... 51 ll. Inverse Rate versus Phosphine Concentration Data for Cyclohexene........... ........ .............. 57 12. k xK5 Values. From the Dependence of Rate on Triphenylphosphine ...... .. ................ ...... 59 13. k6 and K5 Values ..... ....... ...... . ...... ....... 60 vi LIST OF TABLE LIST OF TABLES-~continued TABLE 14. 15. 16. 17. Page Relative Rates of Reduction for Homogeneous Solutions of RhC1(P¢3)3 in Toluene.............. 65 Corrected Relative Rates for Batch 5A (Deficient Rh)...... ...... . ...... ............... 66 Corrected Relative Rates for Batch 5B (Saturated Rhodium).... ...... .. ...... ........... 67 The Ratio of Relative Rates of Alkene Reduction 69 for Batch 5B Compared to Batch 5A............... \ FIGURE LIST OF FIGURES FIGURE Page 1. Probable mechanism of alkene hydrogenation catalyzed by RhC1L3. ........... ......... ....... 9 2. Scheme I--bead functionalization............... 15 Lu 0 Microprobe spectrum of Batch 1A; Rh L a’ 15 kV, 0.050 HA, half bead Section... ......... ........ 22 4. Microprobe spectrum of Batch 1B; Rh La , 25 kV, 0 050 HA, section.............................. 23 5. Microprobe spectrum of Batch 5; Br L , 16 kV, 0.023 uA, half bead (bromine distribution prior to phosphination).............................. 24 6. Microprobe spectrum of Batch 5A, E’; P La 15 kV, 0.032 uA, section........ ............. . ........ 25 7. Microprobe spectrum of Batch 5A, Rh; Rh La , 15 kV, 0.032 uA, section....................... 26 8. Rate versus run for Batch 1B after adding triphenylphosphene............................. 36 9. Rate versus phosphine to rhodium ratio for benzyldiphenylphosphine ......... ............... 38 10. Rate versus phosphine to rhodium ratio for triphenylphosphine.. ....... .. ......... ......... 39 11. Rate versus RhCl(P¢3)3 concentration for 1.0 M cyclohexene.................................... 5o 12. K5(observed) vs [Lo] from alkene dependence of rate for cyclohexene........................... 52 13. k6(observed) vs [L0] from alkene dependence of rate of cyclohexene...... ..... ..... ..... ....... 53 viii W LIST OF FIGURE 14. k, LIST OF FIGURES-—continued FIGURE 14. 15. l6. 17. 18. k6K (observed) versus [LO] from alkene dependence of rate for cyclohexene............ Inverse rate versus [L ] for 1.0 x 10_3 M RhC1(P¢3)3 and 1.0 M cgclohexene.............. Ratio of relative rates of Batch SE to Batch 5A versus molar volume of the alkene.... SAM—-Gas control and measurement section...... SAM—~electronic control section............... ix Page 54 58 70 118 119 m Ow intereS‘ to inso catalys ous ana Ported recentl them he A cataly: the re; comple elimin the Ca reused P has 5( suFPO] Selee rhodi. Red , fiF I. INTRODUCTION General Over the past several years there has been growing interest in attaching homogeneous transition metal catalysts to insoluble polymer supports. This technique gives a catalyst with properties generally similar to the homogene- ous analogue. A large number of catalysts has been sup— ported in this manner. As several reviews have appeared recently covering this topic (1,2,3), I will not enumerate them here. A primary advantage of supported over homogeneous catalysts is the ease with which they may be removed from the reaction mixture by filtration once the reaction is completed. The product is obtained free of catalyst, eliminating a sometimes difficult separation step, while the catalyst is retained in an active form which may be reused. Activation (4) and alteration in selectivity (l,5,6,7) has sometimes been noted when a homogeneous catalyst is supported. It is toward determining factors affecting the selectivity in the system of tristriphenylphosphinechloro- rhodium(I), [RhCl(PD3)3], attached to polystyrene copolymer- ized with 2% divinylbenzene that this work is directed. 1 Rh support styrene a suppo transit allows support linked this St St an effj ture ar of reac monitoi the kiI mechan; Should Cataly: ATE‘TE D llSed i in the Statio T p°r0us RhCl(P¢3)3 was one of the first homogeneous catalysts supported on polymers (6,8,9). The 2% divinylbenezene- styrene copolymer substituted with phosphine equivalents is a support that has been widely used for this and other transition metal catalysts. The presence of phosphine allows a large number of transition metal species to be supported. A study of the system RhCl(P¢3)3 on 2% cross— linked polystyrene may be applicable to other systems using this support. Supported RhCl(P¢ is, like the homogeneous species, 3)3 an efficient olefin hydrogenation catalyst at room tempera- ture and atmospheric pressure (6,10). Therefore, the rate of reaction can be measured with reasonable precision by monitoring the uptake of hydrogen with time. Studies of the kinetics of the homogeneous catalyst have provided much mechanistic information. A comparison of these two systems should offer some insight into the nature of the supported catalyst. The Nature of the Support Divinylbenzene crosslinked polystyrene beads have been used in diverse ways; such as, ion exchange resins, supports in the Merrifield solid phase peptide synthesis, and the stationary phase in gel permeation chromatography. Two broad classifications of beads exist. The macro— porous (macroreticular) resins are produced in a manner that Is onran which wt geneous on degr content ranges the swe limit ( the bea In gene crossli conside PIOpert Pc the sta its inc believE the sta Volume equath Where V outsid< that leaves small channels throughout the bead which remain on removal of solvent from the bead. Microporous beads, which were used throughout this work are essentially homo~ geneous. Microporous beads can be further subdivided based on degree of crosslinking, expressed as the divinylbenzene content of the polymerization mixture, which generally ranges from 1 to 8%. The degree of crosslinking affects the swelling ratio in a given solvent, and the exclusion limit (the largest molecular weight species that will enter the bead), both decreasing with increasing crosslink density. In general, observations of properties of beads made at one crosslink density should be applicable to others with due consideration made for a change in magnitude of that property. Polystyrene beads of several kinds have been used as the stationary phase in gel permeation chromatography since its inception (10,11). The primary cause of separation is believed to be an exclusion of large volume molecules by the stationary phase, which leads to the highest molecular volume species being eluted before to the lower. The basic equation for GPC interpretation is V V = Vo + KGPC s’ r where Vr is retention volume, V0 is the volume of solvent outside of the stationary phase, and VS is the volume of solvent station: 0t] diffusiv advance a statil between ing use an equi Wh to poly is capa origina from pe Indeed, Cules t into 11 the n01 distri] G1 Here t: or We'll Vr e811 and so solvent within. KGPC represents the fraction of the stationary phase accessible to the molecule (10,13). Other factors, such as flow effects and restricted diffusion have some influence, and have at times been advanced as the primary factor in separation (14). However, a static partitioning experiment shows good correlation between the distribution coefficient and K for the pack— GPC ing used (15), indicating that the separation is essentially an equilibrium process. While gel permeation chromatography is normally applied to polymers, it has been known from its inception that GPC is capable of separating small molecules. Moore, in his original paper, separated xylenes (M.W. 106. diameter 8.9 A) from perchloroethylene (M.W. 166, diameter 7.7 A) (11). Indeed, the technique is more selective toward small mole— cules than large ones. Thus, oligomers may be separated into individual components of (monomer)n length (16), while the normal GPC of a full polymer gives a molecular weight distribution not resolved into components. GPC is one basis for the concept of an exclusion limit. Here the exclusion limit is the smallest molecular volume, or weight, species that will pass through the column with Vr equal to V0. Below this point, Vr is greater than V0, and some separation occurs indicating that the species is entering the stationary phase. Tl tograpl microp< that t] they a phase. L One of distri made t at 155 tion j Systex Ion e: hydro mer bi hYdro of hy inher that magni rang. the; EISSQ‘ The results and concepts of gel permeation chroma- tography should be directly applicable to the 2% crosslinked microporous beads used in this study. With the exception that the beads used are somewhat larger than usual for GPC, they are identical to those sometimes used as a stationary phase. Less directly related are the ion exchange resins. One of the few instances of determination of the pore size distribution of a microporous resin in the swollen state was made by Krska and Dusek for a sulfonated polystyrene resin at 15% crosslink density. They found a pore size distribu- tion in the range of 5 to 10 A (17). Comparison of their system with the one of interest must be made with caution. Ion exchange resins are so highly functionalized with hydrophylic groups that they become hydrophylic. The poly— mer backbone and major portion of the pendant groups remain hydrophobic. The observed pores could be due to association of hydrophobic portions of the polymer, and not due to any inherent structure. The results from gel permeation chromatography indicate that size selectivity will be observed on the support. The magnitude for a given size range, particularly for the size range of 6 to 12 A where selectivity has been observed in the reduction of alkenes (4,18), cannot accurately be assessed based on GPC. T] 0f6t< crossl to 2% distri 6 to l furthe RhCl(R compar is nee rates rates insigh SUPPOI direct h0moge homoge exPres reaCt: believ The presence of a pore size distribution in the range of 6 to 12 A is shown by an ion exchange resin with a crosslink density of 15%. Reduction of crosslink density to 2% for the support would indicate that such a pore size distribution would occur at somewhat larger sizes than 6 to 12 A. The Hydrogenation of Olefins With TristrIphenylphosphinechlororhodium(I) Mechanism The primary purpose of this work is not to provide further mechanistic study of olefin hydrogenation by RhCl(R¢3)3. A consistent method for evaluation of, and comparison among, the rates of reduction of various alkenes is needed. These rate comparisons, denoted as relative rates under given conditions, may be compared to relative rates from the bead supported catalyst. This may provide insight into the environment of the catalytic site on the supported catalyst. The validity of this insight depends directly on the validity of the rate comparisons made in homogeneous solution. Thus a detailed discussion of the homogeneous solution mechanism and its relation to the rate expression seems in order. The original investigation of the kinetics of this reaction was by Wilkinson and co—workers (10). They believed the catalyst dissociated completely in solution to give Rh hydroge product (olefin inhibit various the ab: of alke (19-23 D by Sie Tolman consid basis of the ment c 0f hyé large] Observ by th. large A mec Figur give RhCl(P¢3)2, which underwent oxidative addition of hydrogen to this followed by fast addition of alkene to give products. The formation of an olefin complex RhCl(P¢3)2 (olefin) was postulated to account for observed olefin inhibition of the reaction. Subsequent investigation by various workers showed that the primary species present in the absence of hydrogen was RhCl(P¢3)3 while in the presence of alkene and hydrogen, RhClH2(P¢3)2(olefin) was formed (19—23). Detailed reinvestigation of the kinetics has been made by Siegel and Ohrt (23), and Halpern and co-workers (22,24). Tolman has investigated certain aspects of the reaction not considered by others (25). These three studies will be the basis for the following consideration of a mechanism. Many of the conclusions of Siegel and Ohrt, drawn from measure- ment of overall rate as a function of various concentrations of hydrogen, alkene, catalyst, and phosphine, have been largely displaced by instrumental observation, or lack of observation, of intermediates and their formation constants, by the latter two workers. Their work still provides a large body of experimental evidence for interpretation. A mechanism based on these three studies is outlined in Figure l. Fun— I l Non.nun a rune \ .mQOHuflcsoo quE Mots: soauflccm comoncwc mo cofluuom Hanan m waco How muccooom oudou meme Am G A .maoxaaa: who cowuomou mHSD an UoEnom mofloomm .cuomawm SQ tocSHUCH Do: me coauomou mace At .mcofluflccoo oEOm unmoa um Hops: Oomm Ucm mm Sum H on mDQSOEm DamoHMNcmHm CH “commum mum moaommm cocflauoccb 11L? .cumm Mo ucoocomoocfi EDHHQHHHdwo owfimchpofinomp Hmauom.fimll M U A N N m N .N ~93an Nm no . A maocm . A maocm tumsou oumu copoHOOmmm osu we ax sung coauomou ca EDHHDHHHcvo owadcmcofiuoneaAfll.An M .moocmumsdouflo oEOm Hoods camouflammm Emwcmcoofi oauofiflpmfcmfiaoallm cued .Emficmcoofimhuomammula comm Am .mqaonm an commamumo coaumcomoupmfi ocoxam mo EmficmcooE manmnoum .H ousmflm l Nam ... .m + fix H whflwfim NHNAHOAMHiNm 1A , NHNAHUAMH Na + 3s 1 US 31.3 N a H. .N + .NN I Na...» 1 N N N: u N as a A 53m .393 x N m + m x + as .N N . . vqufi s+ a: A: 9+ \ + w. r A n .m + w .. um \ WM M N .l m Tl m .l m A 305.. 592 .II m 48.3 N: + m + N Eh Emu HUHNHwH H fiumm Th1 the prel one atm (L = PT RhClL3 due to and 0hr RhClL3S in the which c Tolman tions c the prj except Tolman dichloi and qu; T( can be Cumsta; genati used, large dinIer. is ind 10 There is general agreement that, starting with RhClL3, the predominant species in solution during hydrogenation at one atmosphere pressure, are RhClH2L3 and RhCleLzs (L = P¢3, S = olefin). Marginally significant amounts of RhClL3 (1 to 5%) may also exist at low alkene concentrations K due to the equilibrium H + RhClL ——345 RhClH L . Siegle 2 3 ~r——- 2 3 and Ohrt also report evidence for a species identified as RhClL3S (S = cyclohexene); Tolman, however, sees no change in the P31NMR spectrum of RhClL on addition of cyclohexene, 3 which casts extreme doubt on its existence. Halpern and Tolman both find evidence for formation of small concentra— tions of RhClL2 and RhClH2L2. These are intermediates in the principle route for addition of hydrogen to RhClL3 except in the limit of very high phosphine concentration. Tolman sees P31NMR spectroscopic evidence for RhCleL2 in dichlormethane at 26°C, though the concentration is small and quite uncertain. Tolman presents evidence that [RhClL and [RhClL212H2 212 can be present in significant concentrations under some cir— cumstances. In the case of L = P(p—tolyl)3, it is a hydro— genation catalyst, although under some of the conditions used, the results and kinetics were obscured by presence of large quantities of RhClHZL S formed by dissociation of the 2 dimer. He indicates that the rate of dimer hydrogenation is independent of alkene concentration above 0.3 M alkene, indicati by this H2 [RhClI the 0V8] tions ca increas: increas< the two and mag alter t Present but do hydroge the rel Rh ClHZI Tl follows v a . 11 indicating that most dimer exists as [RhClL212 and the rate by this route then becomes the rate of formation of HZIRhClLZJZ. While the importance of this dimeric form to the overall kinetics is unclear, some qualitative observa— tions can be made. Given an initial set of conditions, increasing hydrogen pressure or phosphine concentration will increase the relative amount of monomer present, based on the two formal equiliberia, [RhCle]2 + 2L ————¥’ 2RhClL3 [RhClL2]2 + 2H2 + 2L ————¥ 2RhClH2L3 and mass action. The presence of HZ'RhClL2 and RhClHZLZS alter the relative amounts of monomeric and dimeric species present from the values predicted by the above equiliberia, but do not affect the basic argument. In the presence of hydrogen, an increase of olefin concentration will increase the relative amount of monomer present by formation of RhClHZLZS. The primary path for hydrogenation of Olefins is as follows: a) Oxidative addition of hydrogen to minute quantities of RhClL2 gives RhClH verted to RhClH2L3. Equilibrium concentrations of RhClH 2 2, which is rapidly con— b v 2L3 and RhClHZLZS are established and maintained. 95h It tically concent phosphi rate co RhClHZL these I monomer are pre exPress Where Phosph; which : ttatio Phosph tion 0 A Sidera Thine. indie; 12 c) The alkene complex RhCleLZS decomposes to giVe products. Consideration of a Rate Expression It is clear from the preceding that several mechanis— tically important species are present in sufficiently low concentrations that they do not affect either rhodium or phosphine concentration. Halpern's work shows that all rate constants prior to the rate limiting decomposition of RhClHZLZS to products are sufficiently fast to consider these reactions as thermodynamic equiliberia within the monomer system. Assuming that only RhClH2L2S and RhClH2L3 are present in significant quantity, he gives the rate expression: _ kSKSICOJTS] dt_W , (1) Where [CO] is total rhodium concentration. [L] is total phosphine in solution, both catalyst generated and added, which reduces the utility of this form since at low concen— trations of added phosphine, or in the absence of added phosphine, the quantity generated by the catalyst in forma— tion of RhCleLZS can be substantial. A more useful form can be obtained by explicit con- sideration of catalyst generated phosphine and added phos- Phine. Solution of the equation for the equilibrium indicated by K with such consideration leads to the 5 expressi Taking I gives ( Another present inelude these t to tote order t SeVeraJ tomple: purpose beham consid l3 expression—— ds _ k6([LO] + KSIS] 1+4K5 S][Co 'E—U(—l+ ——————-—-——2). (2) ([Lo] + KSISD Taking a two term Taylors series expansion of (2) about 1 gives (1), showing that it is a special case of (2). Another special case of (2) is when no added phosphine is present. This gives—- ds k6K5[S] 4[C ] _ _ O ..d_E___.2 (1+ 1+K5[S] ). (3) Terms for dimeric catalyst formation have not been included. Any attempt to derive a rate expression including these terms, even without including catalyst contributions to total phosphine explicitly, produces equations of higher order than quadratic, which, coupled with the inclusion of several more terms, would produce an expression of such complexity that it would be practically useless for the purpose at hand. The course chosen is to assume monomeric behavior as described by Halpern. Where deviations occur, consideration will be given to the possibility that they are caused by dimer formation. E§§§.iii Th divinyl electrc stituti a rhodi [RhCl(< Scheme equiVa Produc Phosph A C manner method inert aPPfire SPeCie tiallj II. RESULTS AND DISCUSSION Bead Preparations Two broad routes to functionalizing the initial styrene— divinylbenzene copolymer were used, chloromethylation and electrophylic bromination (18). This was followed by sub- stitution of diphenylphosphide for halogen and addition of a rhodium(I) chloride species, either as RhClL or 3 [RhCl(cyclooctene)2]2. These two routes are summarized in Scheme I (Figure 2). Initial chloromethylation produces a catalyst where Rh(I) is attached by one or more phosphine groups roughly equivalent to benzyldiphenylphosphine. Initial bromination produces a catalyst where Rh(I) is attached to one or more phosphines roughly equivalent to triphenylphosphine. A few comments on the two procedures are in order. Chloromethylation followed by phosphination in the manner of Grubbs and Kroll (6) is the simpler of the two methods, requiring fewer steps and less operation under inert atmospheres. A disadvantage, for my purpose, became apparent during the work. Equilibration of the Rh(I) species and triphenylphosphine with the beads will poten- tially produce several species equivalent to RhClL L‘ 3-n n 14 ClCH2 OCH2 CH3 Brz, catalyst, solvent, N2 SnCl4 LiP¢2 THE/Reflux N2 2 [\fifi o m N w E; Rh(I)Can ¢H,N2,R.T. Rh(I)ClLT . LiC4H9, ¢H, N2 2. ¢2PC1,THF ,N2 @cnzwz amncnmm) @sz Rh(I)ClL(n_m) m m Figure 2. Scheme I-ebead functionalization. (L = P91 attache attache between and sel be proc' Se of bror inatiox bromin: due to iron C1 fluori. versus reason mer us of bro POlyme thg an amount the de thEOre batch as 001 leadil hhle j 16 (L = P¢3, Ll = ¢CH2P¢2, n = 1,2,3). At least some of the attached Rh(I)Cl units have two or more bead phosphines Iattached (8,18). The inherent differences in reactivity between these species could affect the observed reactivity and selectivity. Only two species, RhClL and RhClL' 3 3, can be produced in solution with certainty. Several routes to brominated polymer were tried. Use of bromine and anhydrous ferric chloride gave moderate brom- ination (3.8% substitution versus 9.5% possible based on bromine). The polymer was appreciably darkened, probably due to decomposition; however, the presence of residual iron compounds was not ruled out. Bromine and borontri— fluoride in nitromethane gave little substitution (0.43% versus 9.8% possible) although this procedure has given reasonable results on a 20% crosslinked macroreticular poly— mer using identical material and conditions (18). The use of bromine and borontrifluoride in nitrobenzene produced a polymer relatively colorless (very light brown) and contain— ing an acceptable amount of bromine. Twice the theoretical amount of bromine and borontrifluoride was used to obtain the desired degree of substitution (5.2% versus 10.5% theoretical for batch 4, 11.5% versus 17.4% theoretical for batch 5). It is possible the reaction becomes quite slow as concentration of bromine and borontrifluoride decrease, leading to the requirement for an excess to achieve reason- able reaction rate. Tl methan¢ swell 1 little PI plishe n-buty Additi for ex did no Eviden react Probe very c This j of the was p] i Phenyi indie. When light ligan in be 17 The cause for low reactivity of the beads in nitro- methane is uncertain. Nitromethane does not appreciably swell 2% crosslinked microporous beads; it may be that little of the reactants enter the beads. Phosphination of the aryl brominated beads was accom— plished in two steps. Initial lithium exchange using excess n-butyl lithium in benzene produced reddish brown beads. Addition of a two molar excess, based on bromine available for exchange, of chlorodiphenylphosphine followed by reflux did not completely remove the color; addition of water did. Evidently insufficient phosphine was present to completely react with the lithiated phenyl groups. Subsequent micro— probe analysis showed that phosphorous was absent from the very center of the bead but present to a significant depth. This indicates that the wash procedure did not remove all of the excess butyl lithium. Since substantial phosphine was present, these beads were used for further preparations. A potentially useful observation from this is that Phenyl-lithiated beads provide an internal colorimetric indication for completion of reaction with an electrophile. When reaction is complete the beads become significantly lighter in color. Rhodium(I) was attached to the phosphinated beads by ligand exchange from RhCl(P¢3)3 and [RhCl(cyclooctene)2]2 in benzene. RhCl(P¢3)3 requires several weeks to assure complet several shorter ciency solutit [RhCl (4 hours. A in thi Table 18 completion, while [RhCl(cyclooctene)2]2 is complete in several days. The use of a deficiency of either markedly shortened the time required to reach equilibrium. A defi- ciency of RhCl(P¢3)3 was almost completely absorbed from solution in about two days, while a deficiency of [RhCl(cyclooctene)2] was completely absorbed in several hours. A summary of preparation for each batch of beads used in this study is given in Table 1. Table 1. Synopsis of Bead Preparations Bead Equilibrated Phosphene Species and Batch Type Conditions Analysis Deficient 2.08% Rh 1A i—<—O_:>—CH2P¢3 [RhCl(COE)2]2 1.77% P Excess 2.62% Rh lB t*<::>‘CH2P¢3 [RhCl(COE)2]2 1.76% P Deficient 0.32,0.42,0.29% 5A T*<::>‘P¢2 RhCl(P¢3)3 Rh 0.86,0.87% P Excess 0.47,0.76% Rh SB t*:::>“P¢2 Rhc1(P¢3)3 0.93% P Charac IU) beads ratio of bea Measur wet vc except equili T ficat: for sv stitu‘ Table Beat l9 Characterization of Beads Swelling of Ratios Swelling ratios for 32 to 35 mesh, unfunctionalized beads are presented in Table 2. As used here, swelling ratio is defined as the apparent relative increase in volume of beads on adding solvent (Swelling ratio = V(wet)/V(dry)). Measurement of volumes was done in a graduated cylinder, wet volume was measured four hours after adding solvent except for cyclohexane which required two days to reach equilibrium. This is intended as a rapid, simple method for quanti— fication of the observation that some solvents are better for swelling beads than others, and not as a general sub~ stitute for methods employing gain in weight on swelling. Table 2. Swelling Ratios Solvent Swelling Ratio [V(wet)/V(dry)] Benzene 3.90:0.20 Toluene 3.45:0.18 Tetrahydrofuran 3.65:0.12 Ethylether 2.05:0.11 Nitrobenzene 2.35:0.l3* Nitromethane 1.10:0.07* Cyclohexane 1.95:0.10 * Beads float in this solvent. DE T1 was me; 1.021i T solven weighe T ent de solven ing ra volume l tive 1 a smaj of hi sampl emite res01 hents 0h wk rate, 20 Density The density of 32 to 35 mesh, unfunctionalized beads was measured in ethanol and benzene. Values found were l.021:0.051 g/ml for benzene and l.005:0.007 for ethanol. These values are based on the difference in volume of solvent required to fill a volumetric flask containing a weighed quantity of beads and the volume of the flask. There is no significant difference between the appar— ent density in the solvent of high swelling ratio and the solvent for low swelling ratio, indicating that swell— ing ratio is a reasonably good indication of internal volumes accessible to solvent. Microprobe Analysis The electron microprobe provides a rapid, nondestruc— tive technique for elemental analysis of materials within a small volume area (26,27). In this technique a narrow (0.5 micron diameter) beam of high energy (10 to 25 kV) electrons is focused on the sample. X—rays characteristic of each element present are emited from a 1 to 10 micron diameter volume. These are resolved, detected and counted. This technique was used to obtain the location of ele— ments within beads from each batch used. Count rate is plotted on a recorder while the stage, on which the bead section is held, is moved at a constant rate. The recorder trace then corresponds to elemental densitj are pr¢ batche c with t bromin locate indica added '1 consis Preswe °Pera1 The CI to de‘ Press] hate the s the s hetho elect quire Skill Casti 21 density as a function of location. Representative spectra are presented in Figures 3 through 7. The spectra of batches 5A and 5B are identical. Only batch 5A is shown. Comparison of the scan of batch 5 brominated beads with the batch 5A phosphorous spectrum shows that while bromine was distributed throughout the bead, phosphorous was located primarily in the outer sections. This confirms the indication that insufficient chlorodiphenylphosphine was added during the synthesis of this batch. Two methods of sample preparation were used. The first consists of cutting a bead in half with a razor blade. Preswelling of the bead in benzene is helpful, and the operation is best conducted under a low power microscope. The cut bead may readily be examined under the microscope to determine how near a centering cut was made. The method presents some difficulty in actually obtaining a spectrum. Preferably the surface to be scanned should be parallel to the stage. It is very difficult to align the cut surface to the stage with the needed degree of precision when this method of bead cutting is used. Constant refocusing of the electron beam is required as the scan is taken. This re— quirement does not present insuperable difficulty for a skilled operator, and adequate spectra are obtained. The second method of sample preparation starts with casting the beads in an epoxy matrix. When set, the matrix 1,) -- .aofluomm amen Lama .aa omo.o .>x ma .aq am Ada coumm mo Ednpoomm oQOMQOHOHS .m ousmflm 13324.55 8b owe vmm mmN N9 8 o a low 22 1.0m 399/ SINOOO 23 8030mm 53 omoé .i 3 .sq SN ma coumm mo Esnuoomm onoumouowz .v ousmwh 3.: 32455 Gm» ONM mmm mm. mm. V0 0 _ _ _ q _ J o ION m n Ca. 0 m. S 100 V .9 a E Ofln\m.r2300 Figur 24 COUNTS/sec J5 0) OD <3 :3 <3 no (3 l () l I l T’ I l I o 96 192 288 384 480 576 DISTANCE ([1) Figure 5. Microprobe spectrum of Batch 5 Br La, 16 kV, 0.023 uA, half bead (bromine distribution prior to phosphination). s .GOHDoom .4: mmo.o .>x ma . q m . m .mm coumm mo Esuuoomm ohoumouofiz m mucosa is 82590 Ono mhm 00¢ mvn 0mm mt o 0 fl _ _ _ _ . m . ion . m 100. m Ii as [09 V a a room 26 .s .coapoom .m: mmo.o .>x ma g cm cm .mm coumm mo Esupoomm onoumonoflz .h ondoem 13324.55 one new one new 08 m. _ o _ _ _ _ _ ...o :,. .1 . O— , O A. _ ”low 0 _ n N [on N... l W a , 3 IOV [Om is cut gives name method taine< overc< the l; and c With with becam swell break these and t WOulc styrg bead: Polyi the; nude hanz Both Heat 27 is cut at 5 to 10 micron thickness on a microtome. This gives thin sections of beads which can be mounted and scanned with ease. There are several disadvantages to this method. It is difficult to guarantee that the section ob— tained is from a point near the center of the bead. To overcome this, a great many bead sections were cut and only the largest were used. The process of examining sections and comparing their size under a microscope is quite tedious. With this method it is not possible to preswell the beads with benzene. When swelling was attempted, the epoxy matrix became too soft to cut at the desired thickness. Without swelling, the beads are quite brittle and often shatter, break, or come out of the matrix. The probable cause of these problems is a difference in hardness between beads and the epoxy matrix. It seemed that linear polystyrene would be an ideal polymer matrix for polystyrene beads. Polystyrene beads were successfully polymerized into styrene by using benzoylperoxide. Attempts to polymerize beads containing attached rhodium, while producing a good polymer, always showed evidence of rhodium diffusion into the matrix. Polymerization was attempted in a sealed tube under vacuum after freeze thaw degassing using either benzoylperoxide or azobisisobutrylnitrile as initiator. Both showed some indication of rhodium within the matrix. Heating beads with styrene under the same conditions also produc forms presex satis attac that quite it is above the e the t rhodi inter dist: local ratit USefl and 3Uch of u IQIa 28 produced a light red styrene solution. Evidently styrene forms a sufficiently stable complex with some of the rhodium present to remove it from the beads. The use of a polystyrene matrix could prove quite satisfactory for other species which are more firmly attached than this one. Preliminary evaluation indicates that the matrix can be cut at one to five micron thickness quite well. The embedded beads must be colored; otherwise, it is impossible to distinguish between bead and matrix. For qualitative work, the most satisfactory of the above techniques is cutting the bead in half, because of the ease of preparation and the greater certainty of seeing the true center cross—section. An attempt was made to obtain elemental ratios of rhodium to phosphorous on several beads. This would be of interest particularly for batch 1A where phosphorous was distributed uniformly throughout the bead while rhodium was localized. The gross analysis cannot reflect the local ratios. A series of such local analyses would also be a useful supplement to the gross analysis on both batch 5A and 5B where the rhodium gross analysis showed some variance. Such ratios can be obtained with relative ease by comparison of unknown relative count rates for each element with the relative count rates from a standard of known composition. l RhCl(I unider conta< stand; tion 1 as th the t solid noted lytic satur est i has t to tl SOHS 13): hetW1 tion indi ratl as p the 29 In the attempt to obtain this information the standard, RhC1(P¢3)3, decomposed producing small quantities of an unidentified yellow material where the electron beam made contact. The results from batch 5A were obtained using this standard. RhCl(P¢3) was included in a styrene polymeriza— tion mixture, which was cut at 2 micron thickness and used as the standard for batches 1A, 1B, and 5B. It was hoped the thin section would dissipate heat better than the bulk solid. While no physical evidence of decomposition was noted, the results are in gross disagreement with the ana- lytical data, even on beads of batch 1B which are completely saturated with rhodium. The results are reported for inter- est in Table 3. Comparison between different batches, which was the purpose, should not be made since each is compared to the standard separately. Within each batch some compari- sons can be made. Those batches made by initial chloromethylation (1A and 13) are rather more uniform, both within each bead and between different beads, than those made by initial bromina— tion (5A and 5B). If the analysis of perrhodiated batch 1B is considered indicative of its true phosphorous to rhodium ratio, the ratio is 2.23 = P/Rh. Compared to that based on the standard as P/Rh = 3 of 7.2, the standard had actually decomposed to Rh Po.93' Table Batch Batch Batcl an 30 Table 3. Phosphorous to Rhodium Ratios by Microprobe Analysis Bead l Bead 2 Bead 3 Batch 1A Average 9.6 lO.l4:0.83 10.17:0.81 10.26:O.81 9.66:0.79 9.76:0.78 9.74:0.80 8.10:0.63 9.26:0.74 9.76:0.77 Batch 1B Average Analysis 7.08:0.56 7.43:0.60 7.42:0.67 7.50:0.61 7.47:0.60 6.90:0.54 7.26:0.56 6.82:0.55 6.61:0.52 Batch 5A l3.35+l.35 2.20:0.17 9.25:0.80 10.81¥0.87 3.41:0.18 12.20:1.09 12.15£0.94 8.18:0.44 103410.69 Batch 5B 12.68+l.09 17.35:1.27 15.23:1.27 15.75I1.25 18.23:1.53 13.42:1.05 15.34$l.24 18.00:1.50 l6.25:l.36 If factor Studie Eguiva l l equil: rates bexenl were l brati by ad. RhCl( are e rhodi along on th throu are l ihell 05 of these 31 No extensive effort was made to find a more satis- factory standard. Studies on Benzyldiphenylphosphine Equivalent Beads Bead Hydrogenations From the results of Kroll on catalysts of this type equilibrated with RhCl(P¢3)3, the maximum difference in rates of reduction were seen to occur using l-hexene, cyclo— hexene, and cyclooctene (18). Therefore, these alkenes were used in this study. Two batches of beads were employed, both made by equili- bration of [RhCl(COE)2] with phosphinated polymer followed by addition of triphenylphosphine. In contrast to RhCl(P¢3)3 equilibration, this procedure gives beads that are extremely oxygen-sensitive in the dry state. Batch 1A was shown by microprobe analysis to contain rhodium primarily in the outer half of the bead, as observed along a diameter (corresponding to roughly 2/3 of the volume on the bead). Batch 1B contained rhodium of equal density throughout. The results of hydrogenations using batches 1A and 13 are presented in Table 4; data from Kroll Batch E (18) are included for comparison with RhC1(P¢3)3 equilibrated beads of otherwise similar preparation. Correction factors are those of reference 18. 32 owuum>coo “wH.H wcwuoooHomo .oo.H u 0cmxon¢H Eoum oo.H u mcmxwfioaowo ou “mm.o ocmxosua loo.a mcoxwonomo "Ama mucoummouv wuouomw coauomuuoo I 0:0“ 00 ma H sm.o mm.o vo.o+av.o mo.o+mm.o so.QH~m.o oo.qus~.o uoaoso mm.o MH.H FN.H mo.aumm.o no.qwmn.o mo.quo.H mo.qHH~.H mcwxmmua oo.H oo.H oo.H oa.QHoo.H oa.quoo.n mo.QHoo.H mo.qnoo.a mcoxw: Ioaumo H u xmu mumm mummm mumm mummm ovum momma oumuumndm msomcmeEom mm H u w>wumawm wumm 0>Humem ovum mbwumawm wumm Houomm coHuomuHoo vmuomuuoo w>wumamm vouowuuoo w>wumamm @wuowuuoo w>aumamm «Amao muse ma en mcowumummmum msofium> mo momma ucme>flswm wcwnmwonmawconmflcamncmm wCHmD cowumcomouchm mo moumm w>wumamm . q 0.3.68 showec rhodi creas relat and L l-hex LK-E most beads for 1 elenu ous Chan is p betw ent 0ft had Pho: rfim 33 The results were rather unexpected. Batch 1A and LK-E showed similar activity despite a difference in location of rhodium within the polymer. Batch lB showed a much de- creased relative rate for l—hexene and a somewhat increased relative rate for cyclooctene when compared to batches 1A and LK—E. The change in order of relative rates from l—hexene greater than cyclohexene in the cases of 1A and LK—E to cyclohexene greater than l—hexene for 1B seemed most significant. The phosphorous/rhodium mole ratio (RP/Rh) on these beads was somewhat lower than that on beads using RhCl(P¢3)3 for rhodium equilibration. For comparison, ratios based on elemental analysis were: LK—E, RP/Rh = 3.2; 1A, RP/Rh = 2.8; lB, RP/Rh = 2.2. It has been shOWn that changing the ratio of phosphor— ous to rhodium ratio at a given rhodium concentration will change the rate. The absolute rate and magnitude of change is phosphene dependent (28). The relative rate differences between the several batches might be caused by their differ— ent ratios of phosphine to rhodium. In a partial test of this hypothesis, one equivalent of triphenylphosphine was added to beads of batch 1B which had been used for hydrogenations. This should increase the phosphine to rhodium ratio on the beads. Some rhodium was removed from the beads in the process. After thoroughly washix of hY( Absoh phosp tion with as th hexer rates falls ! befo: rate unde was rhod Phir is E ram 1‘H1 34 washing the beads free of soluble rhodium, relative rates of hydrogenation for l-hexene to cyclohexene were measured. Absolute and relative rates for reductions after adding phosphine are presented in Table 5, and depicted as a func- tion of run sequence in Figure 8. Relative rates are taken with the average cyclohexene rate equal to 1.0. The rate of reduction of 1-hexene falls dramatically as the run sequence progresses, while the rate for cyclo- hexene remains constant within the reproducibility of the rates. l—Hexene is initially faster than cyclohexene and falls toward a relative rate in the vicinity of that seen before adding phosphine. Homogeneous Catalyst Hydrogenations An investigation of the effect of phosphine ratio on rate, similar to that of Wilkinson and co~workers (28), was undertaken for benzyldiphenylphosphine. Triphenylphosphine was included for comparison. The rate for 20 ml of a solution, 1.01 x 10‘3 M in rhodium, as a function of substrate, phosphine, and phos— phine to rhodium ratio, is presented in Table 6. This data is plotted in Figures 9 and 10. For benzyldiphenylphosphine, cyclohexene and l-hexene rates are nearly identical above a ratio of 2.25 P/Rh. l—Hexene shows a lower rate below a P/Rh ratio of 1.75. Table 35 Table 5. Rate versus Run Sequence——Batch 1B After Adding Triphenylphosphine Run Substrate Rate Belative Rate (ml/min) RCHX = 1.00 l cyclohexene 0.517 1.25 2 cyclohexene 0.399 0.97 3 cyclohexene 0.364 0.88 4 cyclohexene 0.389 0.94 5 cyclohexene 0.447 1.08 6 cyclohexene 0.302 0.73 7 l—hexene 0.576 1.40 8 1-hexene 0.506 1.22 9 -—————~— Temperature of reaction out of bounds 10 cyclohexene 0.421 1.02 11 cyclohexene 0.391 0.95 12 l-hexene 0.392 0.95 13 1—hexene 0.339 0.82 14 cyclohexene 0.443 1.07 15 l—hexene 0.300 0.73 16 cyclohexene 0.424 1.03 EBHX = 0.4127:0.056 0' = 0.018 m Rate (ml/mi Fig 36 0.6- Rate (ml/min) 005‘ 0. l I l 0 5 10 15 Run Number Figure 8. Rate versus run for Batch 13 after adding triphenylphosphene. A l-Hexene, O Cyclohexene. CHUNNH Fad—H005“ 0“ QCHCAWMOHHW “an COHUNHHMNV .HOmn COHHMCMWQHUWM N0 Wmvuvmm. um @HQNH 37 .w>fiumucmmonmmuss,msmmm oaumu mfisu am wusmflm Eouma as.a mm.s m~.w oo.m wm.a mm.m ms.a an.~ mo.H mm.a oa.~ m¢.m ~N.~ ma.” m~.~ .sa.o vw.m HH.~ mo.~ ms.H mm.w am.“ oo.N om.m om.H mm.o ma.~ mo.m ma.H 05.0 mm.H mm.H mm.a Hm.a NH.H vm.o mm.a mm.o so.a No.H Ho.H mme ammumme mo.ounm\m oanmm mcmxmcoHowo mcwxmmla wcoxmfioaomo mqummla mumuqusm mcwxmcoHomo\mcmxmmIH Hmcmcmflua Hadwcmmoawncwm mewsmmonm oumm m>aumamm wo.nwicna\aao oumm coo.mm .mquam z o.H .am zmuoaxo.fl pm He o.o~ Oeumm Esfiponm 0p msflcmmonm CH :oHuMHHm> How cowumcwwoupmm mo mmumm .m manna (ml/ 38 3.0- E) Rate (ml/min) E] I 76% 2.0.. //;? 1.0.. O I I. I 1.0 2.0 3.0 P/Rh Ratio Figure 9. Rate versus phosphine to rhodium ratio for benzyldiphenylphosphine. C)l—Hexene, C] Cyclohexene. 39 Rate (ml/min) Figure 10. Rate versus phosphine to rhodium ratio for triphenylphosphine. Q l—Hexene, E] Cyclohexene. phine than ratic trip] P/Rh the As; M Prel loca of- sin par Pax am 40 The qualitative results obtained using triphenylphos~ phine indicate that lvhexene is reduced at a rate greater than cyclohexene over a wide range of phosphine to rhodium ratios. Both alkenes show substantially greater rates with triphenylphosphine than with benzyldiphenylphosphine. There is variation of the relative rates with changing P/Rh ratio. The nature of this variation is dependent upon the phosphine. Analysis of Results on Benzyldiphenyl- phosphine Equivalent Beads Use of deficient rhodium, as [RhCl(COE)2]2, allows the preparation of beads containing rhodium preferentially located in the outer portions of the bead. Use of an excess of this complex gives a uniform distribution of rhodium similar to use of excess RhCl(P¢3)3. The selectivity of bead catalysts prepared from RhCl(P¢3)3 and a deficiency of [RhCl(cyclooctene)2]2 is similar, while the selectivity of the bead catalyst pre- pared from excess [RhCl(cyclooctene)2] is quite different particularly with respect to the comparison between l-hexene and cyclohexene. The gross and local P/Rh ratios are dif- ferent for each of the three methods of preparation. Any analysis of this system requires some assumptions be made on the nature and number of phosphines associated with each rhodium atom. The P/Rh local ratio, while more indicative than gross analysis of this characteristic, is not 1 with effe phin chlo poly avai phir bEt) shm by? fire dip ple abc CE] 41 not necessarily the true number of phosphines associated with each rhodium atom. It does set an upper limit. Steric effects could reduce the availability of some pendant phos- phine groups. Chemical reaction of pendant phosphines with chloromethyl groups during preparation of phosphinated polymer has been suggested as another source for loss of available phosphine (2). Both benzyldiphenylphosphine (beads) and triphenylphos- phine are employed in each method of syntheses. The order of addition and quantities of the latter differ substantially between the methods of preparation. While studies have shown that bidentate and tridentate "chelation“ of rhodium by bead phosphine occurs (4,7,18), the conditions employed in synthesis from RhCl(P¢3) produces large quantities of free triphenylphosphine (one P¢3 per bead phosphine used). It does not seem unreasonable that a mixture of bead benzyl- diphenylphOSphine and unattached triphenylphosphine com- plexes could exist under these conditions. In the case of beads prepared from [RhCl(COE) the 212 above arguments do not apply. Chelation would seem to be certain if bead phosphines are available. The coordination of rhodium with bead phosphine would seem to be a purely kinetic process. Once a rhodium attaches a phosphine it is limited in mobility to a rather small area of the bead, enough to coordinate with one or possibly two additional phos: and t rhod: attr rele l-he rho< tiOI inh dec 42 phosphines. Sufficient equilibrium between two coordinate and three coordinate species may be present in the deficient rhodium beads to allow the majority of species to achieve a three—coordinate state. For beads made with excess [RhCl(COE) the ability 212 to achieve high coordination is not present as all phosphines in the vicinity will be occupied by rhodium. Thus, some rhodium of low phosphine coordination will be present. In light of the analytical and microprobe results, the relative rates may be rationalized. The reduction of l-hexene is faster than cyclohexene by beads on which rhodium is primarily three coordinate. When the coordina— tion of rhodium is decreased, to the vicinity of two, the inherent reactivity of the species changes, in particular decreases, toward l-hexene relative to cyclohexene. The addition of triphenylphosphine to the beads sup— ports this view, if one assumes some species such as (bead- benzyldiphenylphosphine)2RhC1P¢3 is formed from the two coordinate species. This restores the selectivity to at least that of RhCl(benzyldiphenylphosphine)3. Since the triphenylphosphine is not attached to the beads, as it is labelized by the reaction, it diffuses off of the beads, restoring the catalyst to its inherent reactivity before addition. the 1 tion func the of h pol} the ObS( res cyc phc and am m1} 43 Change in the phosphine to rhodium ratio upon changing the method of synthesis of catalyst provides some explana- tion for alterations in selectivity. The study of rate as a function of the phosphine to rhodium ratio does not explain the observed relative rate of l-hexene reduction for beads of batch 1A or batch LK-E. If all phosphine present is polymer bound, and thus equivalent to benzyldiphenylphosphine, the 1-hexene relative rate should be about 1.0, while the observed relative rate is significantly greater. The observed relative rate for l-hexene could be the result of steric exclusion by the support being greater for cyclohexene than for l—hexene. It is also possible that the inherent relative rate of a species containing triphenyl- phosphine is higher for l—hexene. The difference in rates and relative rates between benzyldiphenylphosphine species and triphenylphosphine species is so great that even a small number of species containing triphenylphosphine as a ligand could account for this effect. Note also that any tri— phenylphosphine containing species must be of the low coordination form, (bead—Ch2P¢2 ) RhCl (M3) , since a third phosphine, if P¢3, will be removed during hydrogenation and difuse away. Where the species is (bead— CHZ-P¢2)3RhCl or (bead-Chz-P¢2)2, the phosphine will not leave. for the pres a be gem ben sol pol gre ti) tht Gt 44 Studies on Triphenylphosphine Equivalent Beads Bead Hydrogenations Much of the uncertainty in interpretation of results for the benzyldiphenylphosphine equivalent beads stems from the possibility that more than one type of phosphine is present. It is, of course, impossible to exactly duplicate a bead attached phosphine in solution. It is possible to generate a polymer phosphine-solution phosphine pair that is far more similar than the polymer bound equivalent of benzyldiphenylphosphine—triphenylphosphine pair used. Two possibilities present themselves--either make the solution phosphine benzyldiphenylphosphine, or make the polymer bound phosphine equivalent to triphenylphosphine. The latter pair was chosen, partly because of its greater activity and partly because of the lower air sensi- tivity of triphenylphosphine. The ready availability of the well—characterized triphenylphosphine based homogeneous catalyst was also a consideration. Beads containing a diphenylphosphide group directly attached to the pendant phenyl group were successfully synthesized as batch 5, giving a close equivalent of tri— phenylphosphine. Two batches of beads containing rhodium were made by equilibration with RhCl(P¢3)3 in benzene. diun almc 8X01 5011 con ora was WEI are EXE SW Cm 45 Batch 5A contains less than the maximum possible rho- dium content. In synthesis of this batch, the solution was almost completely decolorized, changing from dark maroon to a faint yellow. Batch 5B was made in a similar manner except that sufficient RhCl(P¢3)3 was added to maintain the solution as a dark maroon. The analytical data for both batches are quite variable. The color, which is of some use in determining the rhodium content, was quite different. In solution, batch 5A was orange, which changed to dark yellow on drying; batch 5B was maroon in solution and became dark orange on drying. Relative rates of reduction for a number of alkenes were measured in toluene for both batches. These results are presented in Table 7. The effect of swelling ratio on relative rate was examined for batch 5A using three solvents of different swelling ratio. These results are presented in Table 8. Alkenes were all 1.0 molar in solvent. They were chosen since they represent the size (molecular volume) range over which the maximum change in relative rate occurred using benzyldiphenylphosphine equivalent beads. A second requirement was that the molecule be relatively rigid, and roughly spherical. B-pinene showed a constant pattern of increasing rate in time. in contrast to other alkenes, which showed a Tab] Alkt Cycl Cyc‘ Cyc l-H Cyc Bet Ta): 46 Table 7. Effect of Loading on Relative Rates Relative Rate in Toluene Alkene Batch 5A Batch 5B Cyclopentene 1.80 (0.15) 1.75 (0.03) Cyclohexene 1.00 (0.04) 1.00 (0.05) Cycloheptene 0.97 (0.06) 0.81 (0.05) l-Hexene 1.26 (0.11) 0.92 (0.04) Cyclooctene 0.64 (0.05) 0.43 (0.08) Beta-pinene 0.35 (0.02) 0.08 (0.003) Table 8. Relative Rates of Reduction in Several Solvents of Different Swelling Ratio for Batch 5A Solvent Alkene Benzene Toluene Cyclohexane SR 3.90 SR 3.45 SR 1.95 Cyclopentene 2.10 (0.20 1.80 (0.15) 1.73 (0.38) Cyclohexene 1.00 (0.07) 1.00 (0.04) 1.00 (0.15) Norbornene ---- 1.31 (0.09) -—-- Cycloheptene 0.95 (0.07) 0.97 (0.06) 0.54 (0.08) l-Hexene ~——— 1.26 (0.11) ---— Cyclooctene 0.68 (0.04) 0.64 (0.05) 0.105 (0.080) Camphene ---- 0.29 (0.01) 0.038 (0.001) Beta-pinene 0.14 (0.01) 0.35 (0.02) 0.059 (0.007) decr the shou cate unct sigI alke @ BEE E by eva (2) an 001 T_____i decrease of rate in time. All rates were extrapolated to the time of introduction of alkene, so this abnormality should have no major effect on the relative rate. There is some increase in selectivity with increased catalyst loading. The selectivity remains essentially unchanged on going from toluene to benzene. It does change significantly toward higher relative rates for smaller alkenes on going from toluene to cyclohexene. Determination of Constants Within the Rate Expression for the Homogeneous Catalyst Halpern‘s rate expression for hydrogenation of alkenes by RhCl(P¢ and its various modified forms were used for 3)3 evaluation of constants, k6 and K5, throughout this section. Consideration of the most general form, 4K51COIIS] ‘2)“aa=—t——7—r—"l+/1+nm2’ o 5 shows that the same constants describe the phosphine, alkene, and catalyst dependence. Thus, determination of constants by observing rate as a function of any one or a combination of these variables should be possible. The initial attempt to determine both constants was by variation of catalyst concentration at 1.0 M alkene concen— tration. The rates obtained vary with catalyst concentra- tion in a manner described by equation 3. (3) for Figt Whi] higl lar; tio: ove alk out hyd prc est cor ti< em 48 ”[81 w d (3) -d_f:=_6_§___(_l+‘/1+K5[g] ). The results from variation of catalyst concentration for cyclohexene are presented in Table 9, and plotted in Figure 11. The data were fit by KINFIT (29) to equation 3. While the fit was satisfactory, the parameters were so highly coupled that the estimated deviation was almost as large as the parameters. Observation of rate as a function of alkene concentra- tion was next undertaken. Hydrogen uptake was monitored over the entire course of a run until all of the initial alkene had been used. One option of program SAMV (30) is output of rate versus volume of hydrogen used. Volume of hydrogen may be converted into alkene concentration within program KINFIT and the data from a single run used to obtain estimates of the required constants. Four runs were made in this manner using catalyst concentrations of 1.0xlO-3M and initial alkene concentra— tions of 0.5 M. Phosphine concentration was varied for each run and ranged from 0.1x10"3 M to 0.5x10"3 M. Values of k6, K5, and their product k6K5 are presented in Table 10. These values are plotted versus phosphine concentratiOn in Figures 12, 13, and 14 respectively. There is a clear variation of parameters concomitant with Phosphine concentration, which indicates that the rate Tabj 49 Table 9. Rate versus Catalyst Concentration for 1.0 M Cyclohexene Cx104(moles/1) Rx105(moles—£'1-sec‘1) Value 0 Value 0 10.03 0.091 14.76 0.126 6.32 0.076 11.28 0.036 3.99 0.026 7.61 0.032 3.98 0.023 8.00 0.097 2.51 0.016 5.56 0.026 1.58 0.021 4.23 0.018 1.002 0.013 3.10 0.015 0.500 0.013 1.66 0.008 0.443 0.005 1.22 0.007 Rate (mol 50 Rate x 105 (moles-azml-sec"l ) //£D 4.0—- (3;? 0.0 I I I I I 0.0 2.0; 4.0 6.0 8.0 10.0 [01x104 (M) Figure 11. Rate versus RhCl(P¢3)3 concentration for 1.0 M cyclohexene. C) experimental4 —— calculated, k6 = 0.303, K5 = 4.72 x 10" . Tab] 51 Table 10. Observed Values of k6 and K Obtained from Alkene Dependence of Rate With Small Quantities of Added Triphenylphosphine [RhCl(P¢3)3] = 1.00::10‘3 M '3 ‘1 s - k K 5 —1 [P¢3]X10 UM k6(sec ) K5x10 (unitless) 6 5x10 (sec ) 0.1119 0.879 4.109 3.61 (0.0055) (0.311) (0.737) (0.64) 0.2000 0.636 7.497 4.77 (0.010) (0.045) (0.956) (0.27) 0.2927 0.623 8.093 5.04 (0.015) (0.057) (0.727) (0.01) 0.5014 0.367 15.25 5.60 (0.025) (0.039) (2.03) (0.15) K x] (uni 52 stlo5 (unitless) 18.. 16" 14-- 12'- 10" 0 I I I I I I 0.0 0.1 0.2 0.3 0.4 0.5 0.6 [Lo]x103 (M) Figure 12. K (observed) vs [Lo] from alkene d pendence of rate for cyclohexene. 53 l.2—— k6 (sec‘l) 1.0—— 0.8 0.6—— i °'°IIIIII 0.0 0.1 0.2 0.3 0.4 0.5 0.6 [LOIx103 (M) Figure 13. k (observed) vs [Lo] from alkene dgpendence of rate for cyclohexene. kK (se 54 -1 k6K5x10 (sec- ) <3 I I ITI I- 0.0 0.1 0.2 0.3 0.4 0.5 0.6 [L°]x103 (M) Figure 14. k K (observed) versus [L ] from alkene dgpgndence of rate for cyglohexene. exp: for val cat act gre phi f0] ra‘ iI 55 expression does not properly describe the alkene dependence for the phosphine concentrations used. The decrease in the value of k6K5 with decreasing phosphine concentration indi- cates dimer formation is removing catalyst from the more active monomer system. This removal of catalyst should be greater at low phosphine concentrations than at high phos- phine concentrations since added phosphine reduces dimer formation. The final method of obtaining the constants of the rate expression was addition of large quantities of tri- phenylphosphine. Concentrations of triphenylphosphine 3 M to 100.0x1o’3 M. The nominal catalyst 3 ranged from 5.0x10- concentration was 1.0x10- M and alkene concentration was 1.0 M. The rates were normalized to 1.0 2 of solution and inverted by program EVLT (30). These were then entered into KINFIT versus phosphine concentration. The data were fit to the inverse of equation 1. NH“ 1 1 (m) [L] + W (4) (A = k6K5) Inclusion of catalyst and alkene concentrations, as calcu- lated by EVLT, gave a slightly smaller error. While 3 nominally 1.0x10- M and 1.0 M respectively, there were slight variations. for Fig man ori be< cm 8‘1] Mr“. 56 The inverse rate versus phosphine concentration data for cyclohexene are presented in Table 11 and plotted in Figure 15. Values of A and k6 for other alkenes obtained in this manner are given in Table 12. Having the values of A, the original rate versus catalyst concentration experiments become useful. With the product k6K5 = A, either k6 or K5 can be obtained independently. k6 was obtained using equation 3 in KINFIT as /" "—' '4"k6Tcoj \ R = (— 1 + / 1 + __7i"__ ) . an» K5 was then calculated from A and k6. Values obtained in this manner are given in Table 13. Analysis of Results for Triphenylphosphine Beads An analysis of relative rates obtained on the supported catalyst is undertaken to answer two questions. First, how does the inherent reactivity of the active site within the supported catalyst compare with the reactivity of the homogeneous catalyst? Second, are factors other than inherent reactivity of the active site required to explain the observed relative rates? Two assumptions are made throughout the analysis; triphenylphOSphine and the supported phosphine are suffi- cently similar to produce catalysts which would have similar Tab 57 Table 11. Inverse Rate Versus Phosphine Concentration Data for Cyclohexene 3 1.00x10' M RhCl(P¢3)3 1.00 M Cyclohexene [L1x103 (M) 1/R.x10'4 (sec—z—mole'l) 4.016 (0.055) 9.780 (0.484) 5.056 (0.126) 8.276 (0.221) 10.00 (0.14) 21.11 (0.79) 21.08 (0.17) 34.72 (1.67) 49.91 (0.17) 76.33 (5.41) 59.01 (0.12) 95.67 (3.16) 69.12 (0.15) 120.4 (4.4) 78.61 (0.16) 133.7 (1.7) 87.75 (0.16) 149.5 (6.8) 97.62 (0.18) 173.2 (7.4) (seI 58 %x104 1) (sec-z-mole- 180-— 160—— 140- 120- lOO“ 80— 60— 40— 20— C) I l l T__ o 20 4o 60 80 100 [L0] x 103 (M) 3 Figure 15. Inverse rate versus [L ] for 1.0 x 10- M RhCl(P¢3)3 and 1.0 M cyclohexene. 59 Table 12. k xK Values. From the Dependence of Rate on Tgipgenylphosphine 1 = [Lo] + 1 " "“T‘J‘I‘I’ I 1 R k6K5 CO S k6 Co 6 Alkene k K x10 k 6 5_1 6 —1 (sec ) (sec ) Cyclopentene 93.25 1.57 ( 1.47) (3.63) Cyclohexene 61.40 0.110 ( 2.24) (0.043) Cycloheptene 63.37 -0.42 ( 4.42 (1.98) Cyclooctene 30.44 0.130 ( 2.96) (0.189) Norbornene 80.77 —0.0055 ( 7.96) (0.0012) Camphene 8.82 0.056 (0.48) (0.172) l-Hexene 94.47 0.118 ( 7.22) (0.420) Tab 60 Table 13. k6 and K5 Values Alkene k6(sec_l) K5x104(unitless) Cyclopentene 0.555 1.68 (0.101) (0.30) Cyclohexene 0.209 2.94 (0.017) (0.26) Cycloheptene 0.365 1.74 (0.012) (0.13) Cyclooctene 0.463 0.657 (0.021) (0.070) Norbornene 0.361 2.24 (0.017) (0.24) Camphene 0.108 0.817 (0.0090) , (0.081) 1-Hexene 0.839 1.13 (0.054) (0.11) act hyd clc cat siI wh: be. t1 a] 61 activity under similar conditions, and the concentration of hydrogen within the supported catalyst is sufficiently close to the concentration in solution to have no effect on catalyst activity. The former assumption seems justified since the supported phosphine is aryl-diphenyl substituted while the latter is reasonable since the hydrogen molecule, being much smaller than the alkenes employed, is less sub- ject to steric constraint and has a greater rate of diffu- sion than any alkenes used. In addition, the catalyst is almost zero order in hydrogen concentration at one atmos— phere pressure. These assumptions lead to the relative rates of alkene reduction being a function of alkene and phosphine concen- tration at each active site integrated over the fraction of active sites available to each alkene. The inherent reac— tivity of the active site is dependent on both phosphine and alkene concentration. It is useful to define a reactiv— ity based upon phosphine concentration alone at an alkene concentration of one molar, since the concentration of alkene at the active site is, through the diffusion con— stant, related to the size of the alkene, as is the fraction of active sites available to the alkene. The phosphine concentration at the active site is assumed to be independ— ent of alkene size. Absolute rates for homogeneous solution hydrogenations with 1.0 M alkene concentration can be calculated from equ Dix 011 he] 62 equation 1 for a given phosphine and catalyst concentration. Dividing the rate of the compared alkene by the rate of the standard (cyclohexene) gives a relative rate dependent Only on phosphine concentration. The expression for calculating homogeneous solution relative rates is: k' K' kK RR: 6 5 / 65 [L] + K; [L] + RS where RR is the relative rate, [L] is phosphine concentra- tion, primed constants are for the compared alkene and unprimed are for the standard. The comparison between the supported catalyst and the homogeneous solution catalyst is made through a corrected relative rate (CRR)= CRR = RRB/RR, where RRB is the relative rate obtained with the supported catalyst. If the homogeneous solution relative rates are obtained for conditions adequately describing the supported catalyst active site conditions and if all alkenes have equal accessibility to all active sites, then all corrected rela— tive rates should be 1.0 within the associated error. Obviously there are an infinite number of possible phosphine concentrations with corresponding homogeneous solution relative rates and corrected relative rates. ca: pr] 63 In the face of this number of possibilities the two limiting cases of very high and very low phosphine concentration present themselves. These limits are approximated when [L] is either much greater or much less than K5[S] in equation 1 for all alkenes considered. In either limit, the relative rate becomes independent of phosphine concentration. In that they are limits, the true phosphine concentration must lie at or between these limits and likewise the homogeneous solution relative rates. The corrected relative rates for these limits must then bracket 1.0 if phosphine concentra- tion is the only factor affecting the selectivity of the supported catalyst. If both limiting corrected relative rates are less than (or greater than) 1.0 for a given alkene the relative rate on the supported catalyst cannot be accounted for based on phosphine concentration alone. Equilibration studies have suggested that the polymer supported phosphine has substantial mobility (8,18). Assuming a polymer density of 1.0 and 0.86% phosphorous (the lowest analysis), a uniform distribution of pendant phosphine groups would give a phosphine concentration of 0.28 M within the polymer. Even an order of magnitude less effective phosphine would give relative rates closely approximating the limit of high phosphine concentration. Equilibration studies show phosphine mobility on a rather long time scale (hours to weeks) which may not be signifi— cent to the hydrogenation reaction. Since the reactions we! 64 were conducted well below the glass transition temperature of polystyrene (about 100°C)(40), substantial short term immobility of the pendant phosphine groups is possible, leading to a low effective phosphine concentration. Relative rates of reduction for homogeneous catalyst solutions at 1.0 M alkene concentration and the limiting phosphine concentrations are presented in Table 14. The low phosphine concentration limiting relative rates are designated RBI and the high phosphine concentration limiting relative rates are designated RRII. The corresponding corrected relative rates are CRRI and CRRII respectively, and are presented in Table 15 for batch 5A and Table 16 for batch 5B. The observed relative rates for batch 5A are consistent with the inherent reactivity of the active site being the primary factor determining relative rates. All corrected relative rates (plus or minus two standard deviations) bracket 1.0 for this batch. Effects due to molecular size are obscured due to uncertainty in the absolute reactivity of the active site. Neither the high or low phosphine concentration limits describe the activity of the active site leading to the requirement for a small, significant ~phosphine concentration to be present at the active site. It should be emphasized that this is a kinetically required phosphine concentration at an average active site and does 65 Table 14. Relative Rates of Reduction for Homogeneous Solutions of RhC1(P¢3)3 in Toluene Alkene Relative Rate (Cyclohexene - 1.0) Infinite dilution High Phosphene Cyclopentene Cyclohexene Cycloheptene Cyclooctene Norbornene Camphene l—Hexene 2.65 (0.53) 1.000 (0.081) 1.75 (0.15) 2.21 (0.21) 1.72 (0.16) 0.517 (0.060) 4.01 (0.42) 1.52 (0.06) 1.000 (0.036) 1.03 (0.08) 0.496 (0.051) 1.32 (0.13) 0.143 (0.009) 1.54 (0.13) 2AM ¥C¢.r(. wuwQC\ flu I-llilfu .Houomm soapoouuoo n m0 .mumu 0>meamu Uo>ummno u mmo .wumu 0>Humawu pouoouuoo u «mo «m0\mmo u mmo menu .vH wanna cw ansm msomsmmOEon new mum“ 0>Humeu man mH uouomm sowuumuuoo usaa .oo.H n mamstoaomo so tench mum mmumu 0>HumHmH Haas AeH.ov Amoo.ov Aho.ov “coo.ov AHo.ov No.~ mva.o wm.o sam.o m~.o ~.Hwa mamamsmu AhH.ov Aamo.ov Aeo.ov Adm.ov Amo.ov mm.H omv.o mm.o HN.N v0.0 H.0ma ocmuoooHoho AoH.oV Ama.ov Avo.ov Amv.ov AHH.oV mm.o vm.a Hm.o Ho.v om.a o.mma 0cmxmmlm Amo.ov Amo.ov Awo.ov Ama.ov Aoo.ov ”w vm.o mo.H mm.o mn.a no.0 m.oaa scoummcoHomu Ama.ov Ama.ov Ammo.ov Ama.ov Amo.ov mm.o mm.H Nme.o mn.a Hm.a w.moa unocHOAuoz “mo.ov Aeo.ov Amo.ov Amo.ov Amo.ov oo.H oo.H oo.H oo.H oo.a on.vm ocwxwuoaowo AHH.ov Aoo.ov AmH.ov Amm.ov AmH.ov mH.H mm.a mw.o mo.~ om.H mm.wm msmucmmoHOho mumm, uouomm mums nouomm mums imaos\asv maoxaa w>fiumaom cofluowunoo m>flumHmm sowuowuuoo m>wumawm mcwxas wusm mo wouoouuou b cwuowuuoo n mumbuombo ossao> undo: monasmmosm swam monasmmoam sou ism unmaoaumae am noumm you wanna 0>HumHmm nouowunoo .ma menus .HOpomm cofluomuuoo H mm .oumu m>ammHou pm>uomno u mmo .oumu 0>Humaou pouowuuoo n mmo «m0\mmo u «MO was» .va manna cH macaw msomsomofion Mom mush 0>Humaou wau ma HouOMM coauoouuoo snag .oo.H n ocmxosoaowo so woman was mwumu 0>Humflou HHfluwamm Acowuomunoo o>aumamm GOHuomHHoo o>wucamm msmxafi ousm mo pouumnuoo pouowuuoo n mwo>uwmno masao> umaoz mwcflammonm swam mewsmmoam BOA Aesflconm coumusuva mm soumm How wwumm o>wumamm tmuomuuoo .mH manna 1101 ac 68 not represent the phosphine concentration available to an active site over a long period of time, nor does it repre— sent the phosphine concentration at any particular active site. Limits may be set on this effective phosphine concen- tration. Assuming deviations from limiting relative rates would be detected only if the effective phosphine concentra- tion is greater than ten times the largest K5 or less than one-tenth of the smallest K5, the effective phosphine concentration lies within the range of 6.6x10"6 M and 2.9x10m3 M. The upper limit of this range is clearly much less than the calculated minimum concentration on the poly— mer which indicates substantial immobility of pendant phosphine groups on the time scale of the reaction. In contrast to batch 5A, the relative rates of the more heavily leaded batch 5B cannot be due to phosphine concentration at the active site. The corrected relative rates are less than 1.0 when taken at either limit for all alkenes larger than cyclohexene. There is some decrease in relative rate associated with increasing molecular size. A more dramatic indication of the alkene size dependence of the relative rate is obtained from comparison of relative rates between the two batches of catalyst. The ratio of relative rates obtained with batch SB to those of batch 5A is given in Table 17. A plot of the ratio versus molecular volume of the alkene is presented in Figure 16. Table 17 Alkene Cyclope Cyclope Cyclohe l-Hexen Cyclooc —._...__ 69 Table 17. The Ratio of Relative Rates of Alkene Reduction for Batch SB Compared to Batch 5A Alkene Molar Volume Relative Rate Ratio (ml/mole) Batch 5B/Batch 5A Cyclopentene 88.25 0.972 (0.083) Cyclopentene 94.76 1.00 (0.071) Cycloheptene 116.9 0.835 (0.073) 1—Hexene 125.0 0.738 (0.072) Cyclooctene 130.1 0.672 (0.135) I" 1‘1. 70 Ratio 0.0—(M, I I I I 80 100 120 140 Molar Volume (ml/mole) Figure 16. Ratio of relative rates of Batch SB to Batch 5A versus molar volume of the alkene. Two : dependencI steric ex from a gi concentra the extei a point I molar v01 and deri is quite an incre upon the alkene 1 has dent of a sh geneous not pos On the alk 0f cat; cannot the pr. rate 0 Change at lea 71 Two factors could produce the observed alkene size dependence; diffusion of the alkene to the active site and steric exclusion of alkenes above a certain critical size from a given active site. Diffusion serves to reduce the concentration of alkene within the polymer below that of the external solution (31). The concentration of alkene at a point within the polymer is inversely proportional to the molar volume and rate of use. Precise calculation of rate and derivatively, relative rates, in a variable order system is quite complex (31). An alteration of selectivity upon an increase in loading is indicative of diffusion influence upon the rate. The possibility of steric exclusion of the alkene from the active site is also present. A recent study has demonstrated a substantial increase in steroselectivity of a similarly supported catalyst compared to the homo- geneous catalyst in a system where diffusional effects were not possible (41). Qualitatively, diffusion reduces the concentration of the alkene while steric exclusion reduces the concentration of catalyst available to an alkene. These two alternatives cannot be unambiguously distinguished within the limits of the present study since both serve to lower the relative rate of reduction with increasing molecular volume. The change in selectivity with change in loading indicates that at least some of the selectivity of batch 5B is due to diffusio are an i 72 diffusion of the alkenes to the active site at rates which are an inverse function of the molecular volume. l. T tially di portion < reactiviI with add selectiv selectiv some p01 diffusic 2. crease ing rat 3. 0f IRh< Prepan select the ca 4 detern bead. III . CONCLUSIONS 1. The supported catalyst shows selectivity substan- tially different from the homogeneous analogue. Some portion of the observed selectivity is due to the inherent reactivity of the tested alkenes in the presence of catalyst with added triphenylphosphine. Superimposed upon this is a selectivity based on molecular volume. The increase in selectivity with increased loading suggests that at least some portion of the size selectivity is caused by restricted diffusion of alkenes to the catalytic site. 2. Given solvents of equal polarity, there is an in- crease in selectivity in going to a solvent of lower swell- ing ratio. 3. The bead supported catalyst prepared from an excess of [RhCl(COE)2]2 shows a different selectivity than that prepared from an excess of RhCl(P¢3)3. This change in selectivity is due primarily to the inherent reactivity of the catalytic site at different phosphene concentrations. 4. The electron microprobe is a useful instrument for determining the location of elements within the polymer bead. The heat generated by electron beam contact with the 73 material circumsta 5. W of the pe hydrogen an electI tinue un' color ch Suggesti l. polymer. subject 2. cules m molecul phine 0 TN ment. long 1. Other availa 3 ligand 74 material being analyzed may alter its composition under some circumstances. 5. When lithiated polymer (i.e., polymer on which some of the pendant phenyl groups have lithium substituted for hydrogen and are equivalent to phenyl lithium) reacts with an electrophile, addition of the electrophile should con— tinue until the polymer ceases to be red or brown. The color change indicates completion of reaction. Suggestions for Further Work 1. RhCl(P¢3) should be supported on 20% macrorecticular polymer. This should produce a very active catalyst not subject to major diffusion effects and phosphine inhibition. 2. Further studies of selectivity should include mole— cules much larger than the ones used here. Preferably, all molecules should have very similar reactivity at all phos— phine concentrations. Two types of compounds might fit the latter require- ment. Straight chain l-alkenes are reasonably available to long lengths; allyl ethers of straight chain alcohols, or other alcohols, are fairly easily prepared from the readily available alcohols. 3. A catalyst is needed which is not dependent on its ligand environment. Its kinetics should be well-defined in homogeneous solution. The catalyst should preferably be first 0rd large cor Witl diffusiOI centrati( lyst com 4. sults in satisfac conditic 75 first order in alkene concentration, or zero order over a large concentration range. With such a system it might be possible to separate diffusional effects, which lower the effective alkene con- centration, and steric effects, which lower effective cata— lyst concentration. 4. The microprobe work should be continued. The re- sults indicate that x-ray fluorescence might also be a satisfactory method of gross analysis under somewhat milder conditions than the microprobe. angel. 5223 All All gas Model 90 taken us tus and Mic Inc. Mj Researc] “.3; Al from va Univers Without A] hYdrOge benzOpI solven. to two notice distil IV. EXPERIMENTAL General Analytical All NMR spectra were run on a Varian T-60 spectrometer. All gas chromatographs were taken on a Varian Aerograph Model 90-P using 1/4 inch columns. Melting points were taken using a Thomas—Hoover capillary melting point appara— tus and are not corrected. Microanalyses were performed by Galbraith Laboratories, Inc. Microprobe spectra were obtained on an American Research Laboratories EMX-SM Microprobe. Materials All solvents were reagent grade. These were obtained from various major manufacturers through the Michigan State University chemistry department. They were used as received without any attempt to remove organic impurities. All aprotic solvents used for inert atmosphere work or hydrogenations were distilled from sodium or potassium benzophenone ketyl under a nitrogen atmosphere. Kontes solvent stills were employed. Solvents were stored for up to two days in type 2 flasks on some occasions, with no noticeable effect on reactions. Use immediately after distillation was preferred, but not always possible. 76 Prot purged WI tube plac 15 minutI All cyclohex Company obtained Chemical 9 Exc from soc invaria] pinene I stirred filtere nitroge A1 use. c L materia throng] Purifi. Stated column two Ca 77 Protic solvents used for inert atmosphere work were purged with nitrogen, introduced through a gas dispersion tube placed near the bottom of the container, for at least 15 minutes. All alkenes were of at least 95% purity. Except for cyclohexene, these were purchased from Chemical Samples Company or Aldrich Chemical Company. Cyclohexene was obtained from Matheson, Coleman & Bell or J. T. Baker Chemical Company. Except for beta-pinene, all alkenes were distilled from sodium or potassium under nitrogen. Potassium was invariably used for alkenes boiling below 80°C. Beta— pinene was fOund to be unstable to this procedure. It was stirred over activated aluminia for at least one hour, filtered through a fritted funnel, and distilled under nitrogen. All liquid alkenes were distilled immediately prior to use. Solutions of solid alkenes, made from distilled material, were freeze thaw degassed prior to use. All gases were obtained from Air Reduction Corporation through Michigan State University Stores. Argon and pre— purified nitrogen were used as received. Hydrogen (analysis stated as 99.95%) was passed through two 40 mm by 0.9 m columns of BASF—BTS catalyst heated to 1400C. Between the two catalyst tubes and after the final tube, it was passed through a temperatI methods. These prI flow rat washes c quite sa Rhc from Eng n-Butyl Triphenj Pany ant use. T‘ an iner diately was obt tilled in a 1:} PI Cal CO] A major chemis 78 through a similar tube of 4A molecular sieves at room temperature. Final purification was by either of two methods. Altech “oxytrap” tubes were placed in the line. These proved quite efficient and convenient, but their low flow rate at low pressure presented some problems. Gas washes containing sodiumbenzophenone—ketyl in toluene proved quite satisfactory. Rhodiumtrichloridetrihydrate (RhCl3-3H20) was obtained from Engelhard Industries Incorporated and used as received. n-Butyl lithium was purchased from Alpha Inorganics. Triphenylphosphine was obtained from Pressure Chemical Com— pany and was recrystallized twice from 95% ethanol prior to use. The twice recrystallized phosphine was stored under an inert atmosphere or recrystallized a third time imme— diately prior to use. Chlorodiphenylphosphine (technical) was obtained from Aldrich Chemical Company. It was dis- tilled under vacuum (0.5 torr) and stored under nitrogen in a type 2 flask. Polystyrene beads were obtained gratis from Dow Chemi— cal Company. Any other chemicals were reagent grade obtained from major manufacturers by the Michigan State University chemistry department. Special I Sevr For brev. name whe flasks. Typ neck rou attached 3} flask e: oppositI A . as reli Large v Ciency. I! This is frit at frit is stoPoo: g 24/40 I either 1 0r t Pletic 79 Special Equipment Several special flasks were used for various purposes. For brevity these are described here and referred to by name where they are used. They are modified round bottom flasks. , Type 1 Flask. The type 1 flask is a T 24/40 single neck round bottom flask. A standard taper 2 mm stopcock is attached at an angle of about 450 to the neck. Type 2 Flask. This flask is identical to a type 1 flask except that an additional stopcock is attached opposite the first. A type 2 flask, when sealed with a glass stopper, is as reliable for inert atmosphere work as a Schlenk tube. Large volumes of liquid may be stirred with greater effi— ciency. Type 3 Flask (Bead Hydrogenation Reaction Vessel). This is a type 1 flask fitted with a 40 mm extra coarse frit attached to the side opposite the 2 mm stopcock. The frit is covered on the outside by a bubble to which a 2 mm stopcock is attached. Egggtion Flask Connector. This device consists of a 24/40 male and 24/40 female ground-glass joint connected to either end of a 4 mm stopcock. It is used to attach a type 1 or type 3 flask to the hydrogenation apparatus. On com— pletion of the reaction, the stopcock is closed and the reaction under a Press: 3;} Rh< co-workt it was I A condens two gla mantle. was pla with ni SE round 1 dissolx OI dissolw ethanoj tainin. T nitrog was st U filter 80 reaction flask with adaptor removed, leaving the beads under a nondestructive atmosphere. Preparations Tristriphenylphosphinechlororhodium(I) RhC1(P¢3)3 was prepared by the method of Wilkinson and co-workers (10,32). The work—up differed slightly in that it was accomplished under nitrogen. A 1 liter 3-neck round bottom flask was fitted with a condenser topped by a nitrogen inlet system with bubbler, two glass stoppers, a magnetic stirring system and heating mantle. Two and eight—tenths g RhCl3-3H20 (0.0106 mole) was placed into the flask which was then thoroughly purged with.nitrogen. Seventy ml of oxygen free ethanol was placed into the round bottom flask. This was stirred until the RhCl3-3H20 dissolved. One hundred and twenty g of triphenylphosphine was dissolved in 350 ml of warmed (about 50°C) oxygen free ethanol. This was poured into the round bottom flask con- taining RhCl-BHZO in ethanol. The round bottom flask was closed except for the nitrogen inlet system atop the condenser, and the mixture was stirred at reflux for at least three hours. Upon cooling to room temperature, the mixture was filtered through a frit under nitrogen. The solid was washed wi‘ and two 1 The for 6 how rapid eva be drawn Yie] microcry: Thi: scales u] I_3_e_rl Ben lar to t All oper using ox Th1 “1018) we ma(Inetic tion fIII nitr0ge; the fla 34-0 ml Th lithiun bath Wa below a 81 washed with three 100 m1 portions of oxygen—free ethanol and two 100 ml portions of ethyl ether. The solid was dried under vacuum (0.1 torr or less) for 6 hours. Caution: vacuum must be applied slowly or rapid evaporation of ether will cause some of the product to be drawn into the vacuum system. Yield: 6.84 g (70% based on RhCl -3H20) of maroon 3 microcrystalline solid. This preparation was repeated several times on various scales up to 5.0 g of initial RhCl3-3H20. Benzyldiphenylphosphine Benzyldiphenylphosphine was prepared by a method simi- lar to the direct method of Tamborski and co—workers (33). All operations were performed under a nitrogen atmosphere using oxygen—free solvents. Three and one-half g of clipped lithium ribbon (0.507 mole) was placed into a 500 ml type 2 flask equipped with a magnetic stirrer and under a nitrogen atmosphere. An addi- tion funnel was attached to the flask and purged with nitrogen. One hundred ml of tetrahydrofuran was placed in the flask, and 80 ml of THF was added to the funnel with 34.0 ml (0.188 mole) of chlorodiphenylphosphine. The phosphine solution was added to the stirred lithium/THE mixture dropwise over a two hour period. An ice bath was periodically used to keep the reaction temperature below about 50°C. The completiI solids a The flask by The two 50 I second : funnel, chlorid content of the closely maimed hundre the re M to a E calciI 75 m1 solut the c 500 11 tion 82 The reaction mixture was stirred for four hours after completion of the addition. Stirring then ceased and the solids allowed to settle. The liquid was transferred to a second 500 m1 type 2 flask by syringe. The solid remaining in the first flask was washed with two 50 ml portions of THF which were then added to the second flask. The second flask was fitted with an addition funnel, into which 80 ml THF and 270 ml (0.235 mole) benzyl- chloride was placed. This solution was added to the stirred contents of the flask over a two hour period. Near the end of the addition, the contents of the flask were observed closely. Addition ceased when only a faint red color re— mained. The mixture was stirred for 30 minutes. One hundred ml of saturated aqueous ammonium chloride was added; the resulting mixture was stirred for 12 hours. Most of the organic layer was transferred, by syringe, to a 500 ml type 2 flask containing about 50 g of anhydrous calcium chloride. The aqueous layer was washed with two 75 m1 portions of benzene and this was added to the THF solution. The organic portion was stirred two hours with the calcium chloride. The dried solution was transferred in portions to a 500 ml type 1 flask set up as the pot of a vacuum distilla- tion apparatus. The solvent was removed by distillation 83 under nitrogen. As solvent was removed, more of the solu— tion was added and distillation continued. Much foaming was noted during distillation. The calcium chloride was washed twice with 50 ml portions of benzene, which were added to the pot. When 100 to 150 ml remained, vacuum was applied and continued until bubbling ceased in the pot (pot temperature about 50°C). The system was filled with nitrogen and the pot removed. The residue in the pot was recrystallized from warm absolute isopropanol in four portions. Yield: 28.3 9 (first crop of crystals) of a white solid. M.P. 72-73.5°c (sealed evacuated tube). H'NMR: (resonances from internal TMS in CC14) 63.37 (2H), 67.07 (5H), 67.25 (10H). The signals at 67.07 and 67.25 are broad and not well separated. IR: A thin film was prepared by melting the compound between two NaCl plates on a hotplate inside of a glovebox. This showed no evidence of P = O stretch by comparison with a sample similarly prepared which had been exposed to air in THF. Biscyclooctenerhodium(I) Chloride Dimer TREEITEyclooctene)2l2 or [RhCl(COE)2_I__2 This preparation is similar to that of Porri and co- workers (34). All operations were performed under a nitro— gen atmosphere using oxygen—free solvents. 84 Two and one-tenth grams of RhCl3'3H20 and 50 ml of abso- lute ethanol were placed intoa Schlenk tube. Six m1 of cyclooctene, freshly distilled from sodium under nitrogen, was added. The solution was stirred for four days. The resulting precipitate was filtered from the solution on a frit, washed twice with ethanol, and dried under vacuum for six hours. Yield: 1.12 g of a yellow solid (42.4% yield). Preparation of Beads Batch 1 (Benzyldiphenylphosphine equivalent beads) Chloromethylation The method used was that of Pepper, Paisley and Young (35). Two hundred seventy m1 of chloromethylethylether was placed in an Erlenmeyer flask which was placed in an ice bath. The contents were stirred until cool (about 5°C) and 2.7 ml (4.0 g, 0.024 mole) of stanic chloride was added. Stirring continued for 5 minutes. This solution was added to 30.7 g of polystyrene beads (2% corsslinked with divinylbenzene, 30—80 mesh, 0.284 mole as styrene) in a 500 ml round bottom flask. A drying tube containing calcium chloride was placed onto the flask and the mixture stirred at room temperature for five hours. The beads were removed from the solution by vacuum filtration on a sintered glass fritted funnel and washed twice, while in the funnel, with 1:1 dioxane:water. 85 They were transferred to an Erlenmeyer flask and stirred for five minutes with a 1:1 mixture of 10% hydrochloric acid and dioxane. Upon filtration, they were washed twice with dioxane, transferred to a flask and stirred with dioxane for five minutes, filtered, and washed twice more with dioxane. The beads were dried under vacuum for one day. Phosphination This procedure is similar to the method of Grubbs and Kroll (4,18), based on the method of Tamboriski et a1. (33). All operations were performed under nitrogen with solvents prepared as previously described. Four and two-tenths g (0.61 mole) of clipped lithium ribbon was placed into a 400 ml Schlenk tube and 200 ml of tetrahydrofuran added by syringe. An addition funnel was attached and 100 ml of THF was mixed with 24 ml (0.133 mole) chlorodiphenylphosphine therein. This solution was added to the stirred lithium-THF mixture over a period of 20 minutes. The reaction mixture rapidly turned dark red during the addition. On completion of addition, the Schlenk tube was sealed and the mixture stirred for 24 hours. Twelve g of chloromethylated beads were placed into a 500 m1 type 1 flask, a magnetic stirring bar was added and a condenser attached. Three hundred ml of the red solution in the Schlenk tube was transferred to the flask using a 50 ml syringe, and a 18 gauge needle. The stirred mixture '5’}?le . :3, I.) U ‘..- .l 86 of beads and lithiodiphenylphosphine in THF was refluxed for 24 hours. Upon cooling, 100 ml of an oxygen—free aqueous ammonium chloride solution, made from 150 ml satu- rated ammonium chloride solution and 20 ml water, was added. The mixture was stirred three hours. Liquid was removed by forcing it out under nitrogen pressure through a gas dispersion tube inserted through a rubber stopper. A rubber tube leading to a flask was attached to the dispersion tube. The beads were then washed with 150 ml portions of the following solvents or solvent mixtures: 1:1, waterzTHF, twice; 1:1 10% aqueous Hcl:THF; 3:7 water:THF; THF; 7:3 THF:benzene; 1:1, THF:benzene; 3:7 THF:benzene; benzene, twice. Solvent removal was by the method described above. The beads were dried under vacuum (0.1 torr or less) for 24 hours at room temperature. They were stored under nitrogen until used. Batch 1A All operations were under nitrogen. All solvent trans- fers were by syringe. Seven and ninety-four hundredth g of phosphinated beads (Batch 1) were weighed into a Schlenk 3 eq Rh) was added, tube, 0.531 g of (Rhc1(008)212(1.483x10“ a stirring bar inserted, and the vessel was evacuated and filled with nitrogen; 150 m1 of benzene added, the vessel sealed, and the mixture stirred for 24 hours. The liquid was removed and the beads were washed with two 50 ml l 111‘”- .U ...-1 87 portions of benzene, then stirred with a 100 ml portion for 24 hours. As this solvent was quite clear at the end of 24 hours, 3 mole) was added; 0.5296 g of triphenylphosphine (2.0x10- the solution slowly became a light maroon. The beads were stirred in this solution for 24 hours, then repeatedly washed with 100 m1 portions of benzene until the last wash remained clear for 24 hours. They were then dried under vacuum (0.1 torr or less) for 24 hours at room temperature. Batch 1B All operations were under argon and all solvent trans— fers by syringe. Two and one-tenth g of beads (Batch 1) were weighed into a Schlenk tube and 0.532 g (1.48x10-3 eq Rh) was added. The vessel was evacuated and filled with argon, 50 ml of benzene added, the vessel sealed and the mixture stirred for 5 days. This solution was removed and the beads washed once with 100 ml of benzene; 0.320 g of [RhCl(COE) (0.892x10"3 eq Rh) was dissolved in benzene in 212 a separate Schlenk tube. This solution was transferred to the vessel containing the beads; the vessel sealed and the beads stirred with this for two days. The beads were washed repeatedly until no color remained in the solution. They were then dried under vacuum for 24 hours. Argon was introduced and 0.5075 g of triphenylphosphine (1.94::10"3 mole) in 60 m1 of benzene was added, the vessel was sealed, 88 and the contents were stirred for 24 hours. The beads were repeatedly washed with benzene until a wash remained clear for 24 hours; then dried under vacuum for 24 hours and stored under argon. Batch 2 All operations were conducted under nitrogen to exclude atmospheric moisture. Forty and two-tenth g of beads (30- 80 mesh, 2% crosslinked with divinylbenzene, 0.372 mole as styrene) were weighed into a type 1 flask. The flask was then purged with nitrogen; 200 m1 reagent nitromethane from a freshly opened bottle was added followed by 5.4 g of anhydrous ferric chloride. Nineteen ml (0.074 mole) of a 10% V/V solution of bromine in nitromethane was added over a half hour period. The flask was then closed, covered with aluminum foil, and the contents were stirred for 24 hours under nitrogen. The solution was removed from the beads and they were washed consecutively with nitromethane, saturated sodium bisulfite in nitromethane, and nitromethane. They were transferred to a Soxhlet extractor and washed with nitro- methane for 48 hours. Upon removal, they were dried under vacuum for 24 hours. Yield: 38.8 g of yellow—brown beads. Analysis: C, 89.13%; H, 7.48%, Br, 2.72% 3.8% substitution. 89 Batch 3 Fifty-eight and one-half g of beads (28-32 mesh, 0.563 mole) were_placed in a 1000 ml, 3-neck, round bottom lask. The system.was evacuated and filled with nitrogen, an atmos- phere of which was thereafter maintained during the course of the reaction; 300 ml of nitromethane from a freshly opened bottle was added; 3.8 g of borontrifluoride gas (0.056 mole) was bubbled into the liquid and 2.8 ml of bromine (8.79 g, 0.055 mole) in 100 ml of nitromethane was added over a one-half hour period. The flask was sealed, covered with aluminum foil, and the mixture stirred for 36 hours at room temperature. The liquid was removed. Use of a nitrogen atmosphere ceased at this point. The beads were washed consecutively with dichloro— methane:methanol mixtures in proportions 1:0, 9:1, 3:2, 2:3, 1:9, 0.1. Then methanol:water washes in proportions 3:1, 3:2, 2:3, 0:1, were performed followed by water:tetrahydro- furan, 2:1, 1:2, 0:1. They were then washed with dichloro- methane:THF in proportions 1:2, 2:1, 1:0. All washes were done in the reaction flask, the liquid being removed from the beads after each wash using an aspirator attached to a gas dispersion tube frit as a filter. Batch 4 This batch was intended as a test for the use of nitro- benzene as a solvent. During work with the two preceding 90 batches, nitromethane was found to be a poor swelling sol- vent for the beads. Nitrobenzene, while not excellent, is about the same as ethyl ether, and much better than nitro- methane. Five g of beads (32—35 mesh, 425-500 u, 0.0459 moles as styrene) were placed in a 250 ml single type 1 flask. The flask was evacuated and filled with nitrogen, an atmosphere of which was maintained throughout the course of the reaction. Solutions of borontrifluoride and bromine were pre- pared; 3.3 g of BF3(0.0487 mole) was bubbled into 100 ml of nitrobenzene giving a solution 0.487 m, BF3; 1.00 ml Br2 (3.12 9, 0.0195 mole) was added to and mixed with 100 ml nitrobenzene, giving a solution 0.193 M in Brz. Ten ml of the BF3 solution (0.00487 mole BF3) and 25 ml of the Br2 solution (0.00483 mole Brz) were added to the flask contain- ing the beads. The flask was closed, covered with aluminum foil and stirred, 1.00 ml aliquots of solution were taken periodically and analyzed for total bromine content. The aliquot was placed in an Erlenmeyer flask, excess Br2 con- verted to Br- with aqueous Na $03, and Br— determined by the 2 Volhard procedure. After four days no further significant change in bromide concentration was noted. The beads were then subjected to the following work-up in air. The reaction solution was removed and the beads washed twice with nitrobenzene, thence consecutively with nitro- benzenezTHF mixtures in proportions 4:1, 2:1, 1:1, 1:2, 1:4, 91 1:0, and 1:0. They were washed with THF:water mixtures in proportions 10:1, 1:1, 1:0, and 1:0, and thereafter dried under vacuum at 70°C for 24 hours. Analysis: Br 3.83%, 3.87% 5.18% substitution. Batch 5 Bromination Forty and six-tenths g of beads (32—35 dry mesh, 425— 500 p, 0.390 mole as styrene) were placed into a 2-liter 3—neck round bottom flask. The flask was evacuated and filled with nitrogen. Nine hundred ml of nitrobenzene (dried for 2 days over 4A molecular sieves) was added to the flask and 5.2 g (0.077 mole) BF3 was bubbled in. With the contents stirring vigorously, 3.9 ml (0.068 mole) Br2 was added. The flask was placed in an ice bath and the mixture stirred fifteen minutes, then it was removed and stirring proceeded at room temperature. The flask was covered with aluminum foil. Stirring continued for 60 hours. The solution was removed and the beads washed three times with nitrobenzene, then with THanitrobenzene mixtures in proportions 1:10, 1:3, 1:1, 3:1, 10:1, 1:0, and 1:0. The beads were stirred with 1:10, waterzTHF for one hour. The THF:water wash was repeated five times. On the last wash no Br— was detected by addition of a silver nitrate solution to the filtrate. The beads were then washed twice 92 with THF and dried at 120°C for 24 hours. Analysis: 8.10% Br 11.5% substitution. Phosphination This entire procedure was conducted under nitrogen using solvents prepared for inert atmosphere work as described previously. Thirty-eight and one-tenth g of brominated beads were placed in a 500 ml type 2 flask. The flask was stoppered, evacuated, and filled with nitrogen. Two hundred m1 of benzene was added followed by 60 m1 of 2.4 M butyl lithium in hexene. The flask was closed and the mixture stirred for three days under nitrogen. The solution was removed from the beads (gas dispersion tube method), and 50 m1 of 2.4 M n-butyl lithium in hexane with 50 ml of benzene was added-~this was stirred for two days. The solution was removed and the beads washed two times with 150 ml portions of benzene, stirring each time for one hour. The beads were then dried under vacuum until they were free-flowing._ Thirteen m1 chlorodiphenylphosphine (0.0721 mole) in 100 m1 THF was added. The reaction mixture became quite warm and turned red. Two hundred ml additional THF was added and the mixture was stirred one hour at room temperature. It was then refluxed for 24 hours. The solu- tion was cooled and removed from the beads. They were washed five times with THF. The beads retained some reddish 93 brown color after the THF washes-~although they were much lighter than after removing the initial reaction solution. They were washed with a 2:1:1 mixture of THF, water, and saturated aqueous NH Cl which quickly removed the color. 4 The beads were washed twice with 1:1 THF:water, stir— ring each wash for 15 minutes, followed by a 1:1 THF:water wash which was stirred for three hours. They were then washed with 3:7 water:THF (15 min stirring) and pure THF (1 hour stirring). The beads were washed twice with ethyl ether for three hours; after removing this the beads were vacuum dried at room temperature for 18 hours. Batch 5A Twelve and one-hundredths g of phosphenated beads (batch 5) were weighed into a 500 ml type 2 flask; 0.4709 of RhCl(P¢3)3 was placed therein. The vessel was evacuated and filled with argon, three hundred ml of benzene was added and the mixture stirred for 24 hours. At that point, most of the color was gone from the solution and the beads were light red. The beads were washed repeatedly until the wash remained colorless for 24 hours. They were then dried under vacuum (0.01 torr or less) for 24 hours and stored under argon. Batch SB Eleven and thirty-five hundredths g of phosphenated beads were placed in a 500 m1 type 2 flask; 0.80 g 94 3 (3.0x10- mole) of triphenylphosphine was added. The flask was evacuated and filled with argon. Three hundred m1 of benzene was added. One and ninety-five hundredths g of RhCl(P¢3)3 (2.llx 10..3 moles) was added in approximately 0.2 g increments at two day intervals. After completion of addition, the beads were stirred for two weeks. The beads were then washed repeatedly with benzene until the last wash remained color— , less for 24 hours. They were dried under vacuum (0.05 torr) for 24 hours at room temperature, then stored under argon. Densities One to three 9 of beads were weighed into a 10.0 ml volumetric flask. Solvent was added from a burette until about 1 ml remained to full. The flask was stoppered and the burette volume noted. The beads sat with solvent 1 hour. Then solvent was again added until it came to the mark. The sum of the two additions is the volume of solvent required (Vr). The difference of this from 10.0 ml is the volume occupied by the beads. Density (D) then becomes D = w/(l0.0—Vr), where w is the weight of the beads. All measurements were at ambient temperature (25°C:1°C). For benzene it was necessary to use weights near 1.0 9 since more would completely fill the flask after swelling 95 in the solvent. The lower swelling ratio in ethanol allowed more beads to be used. Swelling Ratios About 2 m1 of beads were placed in a 10 ml graduated cylinder and the volume recorded. The cylinder was filled with solvent and the volume of beads was noted periodically (about every hour). The final volume was recorded when no further increase in head volume was noted over a six hour period. The swelling ratio as used here is: SR = V(fina1)/V(initial). Microprobe Analysis Preparation of Beads Sectioning_in a Matrix Polymerization of Beads into Epoxy Microscope slides were washed consecutively in ethanolic potassium hydroxide, nitric acid, and distilled water, then dried. Two slides were taped into a V mold using cellophane tape. The ends of the V mold were closed with cellophane tape. A few beads were placed in the bottom of the V and epoxy resin (Buehler, AB EPO—mix) was placed on top of them. The beads were mixed into the epoxy using a nichrome wire. This hardened for 24 hours at room temper- ature. The polymer was removed from the glass by briefly placing the mold in liquid nitrogen. 96 Polymerization of Beads into Styrene Three hundred ml of styrene was washed four times with 100 m1 of 10% sodium hydroxide and twice with distilled water. It was then dried first over calcium chloride and then over 4A molecular sieves. The dry styrene was distilled under vacuum from a pot held at room temperature into a receiver cooled in dry ice- acetone. That portion of the styrene not used immediately was stored under nitrogen in a refrigerator at ~10°C. Drying and distillation were repeated prior to each subse— quent use. A 12 inch long 6 mm diameter glass tube was sealed on one end and cleaned with ethanolic potassium hydroxide followed by nitric acid and distilled water. It was then dried. A 1% W/V solution of either benzoylperoxide or azo— bisisobutyrlnitrile in the distilled styrene was made. A small quantity of rhodium containing beads were placed into the tube with about 1 ml of the styrene-initi- ator mixture. The contents were then subjected to three cycles of freeze-thaw degassing and the tube was sealed under vacuum. The tube was placed in an oven at 60°C for three days. Use of an oil bath at 60°C was tried. Under the condition of a vacuum in the sealed tube, the styrene will reflux since the top part is cooler than the bottom. "a 1w?" 97 In all cases, some evidence of removal of rhodium from the beads was noted as a light red area around the maroon bead. This was more obvious when using benzoylperoxide. RhCl(P¢3)3 was polymerized into styrene at about 0.002 g/ml using AlBN as initiator. This was carried out exactly like bead polymerizations. Sectioning An epoxy triangle, containing beads, was placed in a microtome. Sections were cut at 10 microns thickness. Only a few sections of many cut were sufficiently intact to be of any use. These sections were examined under a binocular micro- scope. The largest diameter bead sections were identified and the sections containing these were mounted. Only by taking the largest diameter bead sections can the possibil— ity of obtaining sections cut from near the edge be elimi— nated. Sections were usually mounted on graphite disks. In a few instances quartz plates were used. The purpose of mounting is to keep the sections from moving during further preparation and evacuation of the sample chamber. It amounts to gluing the section to the plate. Three sub— stances were used as glue: 1. Alphacyanoacrylate cement. 2. Graphite electron tube coating. 98 3. Adhesive from "Scotch" rubberized adhesive masking tape. The rubber adhesive was found to be the best of the three. There were no problems of having glue get on top of the section, which occurred with the other two. After mounting, the disks (or plates) were coated with graphite from an arc, and placed in the microprobe. The Half—bead Method Beads were placed on the stage of a low power binocular microscope. They were held in place with small tweezers and cut in half with a razor blade, then attached to a graphite disk using alphacyanoacrylate cement. The bead and disk were coated with carbon from an arc and inserted into the micrOprobe. While this method seems relatively crude, it has the advantage of being fast, and the beads are subject to no chemical influence, save atmospheric oxygen, during sample preparation. Determination of Elemental Radial Distribution A selected head was identified in the microprobe micro— scope. Using the secondary electron emission scan at fast scan rate, the head was aligned so that the X axis of the microprobe corresponded to a bead diameter. The X—ray detector was set to obtain a maximum reading on the required wavelength for the desired element. 99 Beads were scanned by moving the stage at constant speed in the X direction. The electron beam location remains fixed. This was deemed preferable to electron beam scan since the beads are usually of larger diameter than the maximum scan width at lowest magnification. Scans were started and ended about 50 microns from the bead edge. X—ray intensity (counts/sec) was plotted on an X—Y recorder or stripchart recorder driven at constant speed. Attempted Determination of the Phosphorous to Rhodium Ratio This technique consists of comparing the relative count rate per element (P/Rh) on beads and a known ratio compound. Tristriphenylphosphinechlororhodium(I) [RhCl(P¢3)3], Pth = 3:1, was chosen as the standard since it should be quite similar to the species on the bead. This was not successful since RhCl(P¢3)3 decomposed to a yellow compound of unknown composition under the electron beam. A small quantity of RhC1(P¢3)3 was polymerized into styrene in the hope that a thin section might provide better heat transfer than the bulk material. No physical evidence of decomposi— tion Was seen; however, the results obtained are in gross conflict with the analytical data. Batch 5A used bulk RhCl(P¢3)3 while 1A, 13 and 5B were analyzed using the polymer trapped RhCl(P¢3)3 as the standard. 100 The plate containing the sample and the one containing the standard were simultaneously placed in the instrument. Total counts were recorded from three twenty second periods at each point. Rhodium and phosphorous were each observed simultaneously, using two different pray detectors. All samples were run at 15 KEV. Three points were observed on the standard. Three widely separated points were observed on each of three beads. Background counts were taken on both the standard plate and bead plate. Each point (background, standard, bead) was reduced to a count rate (counts/sec) with associ- ated standard deviation of the mean (cm). Background was subtracted from each point. The phosphorous count rate was divided by the rhodium count rate giving a P/Rh count rate ratio and associated cm for each point. The standard P/Rh count rate was divided by 3 (3 P per Rh in the standard) giving a count rate per atom ratio (R). R was divided into the count rate ratio for each point on each bead giving the local P/Rh ratio at each point. The data obtained must be regarded as invalid since there is evidence that the standard decomposed. It does demonstrate that this is a viable technique if a suitable standard can be obtained. 101 Hydrogenation Procedure General All hydrogenations used hydrogen purified in one of the manners described in Materials-~Gases. Hydrogenation rates were monitored using either gas burette measurement of hydrogen volume versus time as measured by a timer reading to 0.01 minute, or by use of the automatic hydrogenation apparatus, SAM (fully described in Appendix I). Batch 1A and 1B rates were measured by the gas burette method. Rate versus rhodium to phosphine ratio and Batch 5A in toluene data was obtained using SAM in the stripchart output mode using a Sargent-Welsh recorder operating at l inch/minute. All other reactions were monitored by SAM using the digital time output mode, with 1 sec being the time unit. In all cases the reaction was set up using the atmospheric hydrogenation apparatus. SAM was used only for the data collection phase of the reaction. Temperature was maintained by placing the reaction vessel in a water bath maintained in equilibrium with water circulating from a thermostated water bath. All reactions were run at 25°C:0.2°C. With one exception, all hydrogenations were conducted at a nominal 1.0 M alkene concentration, as calculated assuming the volumes of alkene and solvent are additive. 102 Given a certain desired total volume of solution, the quantity of alkene required is calculated from its molar volume. Molecular weights and densities used in this calcu- lation were taken from the Handbook of Chemistry and Physics (36), with the exception of norbornene which was taken from Reagents for Organic Synthesis (37). The quantity of sol— vent was taken as the difference of total volume required and volume of alkene required. Solid alkenes differed slightly. A 2.00 M solution in solvent was made after distillation into a tared flask. This solution was used in a manner identical to liquid alkenes. Prior to each use the alkene solution was subjected to three cycles of freeze thaw degassing followed by introduction of argon into the flask. Bead Hydrogenations From 0.5 to 3.0 g of beads were weighed into a type 3 flask or a type 1 flask. A stirring bar was inserted and a reaction vessel adaptor attached using two Kontes "Kem Klamps". This total apparatus is referred to as the reac— tion vessel. The reaction vessel was attached to the hydrogenation apparatus. The apparatus was evacuated to about 0.1 torr and filled with hydrogen three times. The stopcock on the reaction vessel adaptor was opened and the reaction vessel evacuated and filled with hydrogen three times. 103 Sufficient solvent was then introduced so that upon adding the required amounts of alkene, a 1.0 M solution in alkene would produce a total volume of 30.0 ml. All transfers of solvent and alkene were by syringe. The solvent and beads were stirred under hydrogen for 1 hour at 25°C:0.2°C. The required amount of alkene was introduced, and data collec— tion started. At the end of the reaction the stopcock on the reac- tion vessel adaptor was closed and the reaction vessel transferred to a vacuum-argon line. The connection to the vacuum-argon line was evacuated and filled with argon three times. Then the stopcock on the reacton vessel adaptor was opened allowing argon to enter the reaction vessel and work-up proceeded. Work-up consisted of removing the reac— tion solvent and washing the beads three times for 15 minutes each with fresh solvent. When a type 3 reaction vessel was used, solvent was removed through the frit. For a type 1 reaction Vessel solvent was removed using a syringe with a flat tiped 18 gauge needle. This needle was held against the edge of the flask in order to prevent beads from being drawn into the syringe as solvent was removed. For a type 3 flask, it was necessary to apply fresh stopcock grease to the drain stop- cock after each work-up. Both input and drain stopcocks were regreased after every three runs. Any removal of the WW, , F 104 stopcock was accompanied by a high rate of argon flow to maintain an inert atmosphere over the beads. After removal of excess solvent and such maintenance as was necessary, the reaction vessel was attached to the hydrogenation apparatus, and set up for another run. A notable problem in this procedure is that beads are never fully dried between runs. The reasons are twofold. First, complete drying requires 24 hours under a vacuum of 0.1 torr at room temperature, although a close approximation is obtained after six hours. Second, and most important, the apparatus is not completely air tight under vacuum for extended periods. Any attempt to repeatedly dry beads after runs caused substantial loss of activity of the standard alkene over a relatively few runs. The presence of solvent may account for the relatively large deviations of rates observed, since the solvent remaining causes some error in the concentration of alkene. Volumes of total solution much larger than the bead volume were used to minimize the error. The practical problems of solvent transfer and alkene cost somewhat limit how large a total volume of solu— tion one can use. Another possible solution is to use very small quantities of beads, on the order of 0.1 to 0.5 g. The latter procedure presents problems also since this will reduce the rate and make the system relatively more oxygen sensitive. 105 Alkenes were run alternately with cyclohexene—~the standard. Each relative rate is from a comparison of at least two runs of the alkene with the standard. Each time a new charge of beads was used, at least 10 runs with standard were made with no comparisons, which allowed absolute rates of the standard to stabilize about an average. There was initially a change in rate of the standard with each run. Hydrogenations Using Homogeneous Tristriphenylphosphinechloro— rhodium(I) The purpose of these experiments was to obtain con- stants within Halpern's rate expression. The experimental procedure used in all is quite similar. Three types of experiments were used: variation of rate versus catalyst concentration, variation of rate with phosphine concentra— tion, and variation of rate with alkene concentration. For all reactions, a quantity of RhCl(P¢3)3 was weighed into a type 1 flask. Solid triphenylphosphine was also added if required and a stirring bar inserted. A reaction flask adaptor was attached and the flask placed on the hydrogenation apparatus. The system, flask and apparatus, was evacuated and filled with hydrogen three times. Toluene, and triphenylphosphine solution if needed, were added in required amounts and the system closed. The solution was stirred for at least 30 minutes in the water bath at 106 25.0°C10.2°C. The required quantity of alkene was then injected and data collection begun. On completion of the reaction the contents of the flask were disposed of. For rate versus catalyst concentration, catalyst was weighed into the flask. The quantity of solvent and alkene were calculated to give the desired concentration of cata— lyst and a 1.0 M concentration of alkene. For rate versus phosphine concentration, catalyst was weighed into the flask. The volumes of toluene and alkene were calculated to give a catalyst concentration of 1.0x10_ M and alkene concentration of 1.0 M. The quantity of tri— phenylphosphine needed to give the desired concentration was made and a quantity near this was weighed into the flask. Since actual concentration is later calculated, this need not be exactly the desired weight. For both of these methods, data was collected to 20 points or 30 minutes run time, whichever was greater. Four runs were made where variation of rate with alkene concentration was observed. For these a quantity of catalyst was weighed into a type 1 flask and this was set up on the hydrogenation apparatus. Since very low concen- trations of triphenylphosphine were used, a solution at 10.0x10"3 M in toluene was made in a manner analogous to that used for solid alkenes. The required amounts of toluene, triphenylphosphine, and alkene were calculated to 107 3 M catalyst, 0.5 M alkene, and the desired give 1.0x10— concentration or triphenylphosphine. Two ml and 10 ml burettes equipped with luer joints and needles, and operating under nitrogen were used for all liquid transfers and measurements. Comparison of Rates for Phosphorous to Rhodium Ratios Less than Three A 4.02x10-3 equivalent/l solution of [RhCl(COE)2]2 in toluene was made under argon, as was a solution of 6.97x10-3 M benzyldiphenylphosphine and 8.07x10—3 M triphenylphosphine. A 250 ml type 1 flask, equipped with a magnetic stir— ring bar and reaction flask connector was attached to the hydrogenation apparatus. The system was evacuated and filled with hydrogen three times. The required amount of toluene and phosphine solu- tion was injected, followed by 5.0 ml of rhodium solution. The system was evacuated until the solution just boiled and then refilled with hydrogen. The solution was stirred at 25°C:0.2°C for 30 minutes. The required amount of alkene was injected and data collection started. At least 20 points were acquired for each run. Evaluation of Data Initially all data was obtained as volume of hydrogen used versus time. In the case of runs which were monitored by the gas burette, this is the form in which the data was 108 recorded. For runs made using time at which the apparatus added an additional known volume of hydrogen to the system, time of the event was recorded. In the former case, volume of hydrogen used versus time was entered directly into program KINFIT (29). Initial rates were obtained by fitting the equation: V = Rt + At2 + B. The initial rate is R. The additional parameters allow deviation of rate in time and deviation of the initial volume from zero. For runs using SAM, the output times were converted to volume versus time data using program SAMV (30) or one of several earlier versions which accomplished the same trans“ formation. The calculated volume versus time data were then entered into KINFIT in the same manner as gas burette data. The one exception to this was four runs of rate versus alkene concentration which were evaluated under a different option of SAMV and will be treated separately. Calculation of Relative Rates The absolute rate for a given alkene run was divided by the absolute rate for the two standard runs immediately preceding and following the run. This gives a total of four comparisons for each hydrogenation, which reduces the effect of the variance in standard absolute rates. Where only one standard was run after the given alkene, only 109 three comparisons were made. The average of these compari- sons and the standard deviation of their mean are reported.. The standard deviation of the mean for the first 8 runs after the breakin is reported for the standard. If more than one set of beads were used, the average standard devi— ation is reported. Homogeneous Catalysis Data The absolute rates are reported for runs where the phosphine to rhodium ratio was varied below 3:1. For those runs where data was used to evaluate con- stants within the rate expression, an average rate was obtained from KINFIT for the first 20 points of volume versus time data for each run using the equation V = Rt + A. The initial and average rates were then entered into program EVLT (30), along with the weight of catalyst, volume of Solvent, volume of alkene, a concentration of alkene used, weight of phosphine, molecular density of phosphine, and associated estimated errors. EVLT then averaged the initial and average rate, took a standard deviation, normalized these to 1.0 l of solution, and calculated concentratiOns of catalyst, alkene, and phosphine. Associated errors were carried through in these calculations. 110 The required data was entered into KINFIT and the con— stants obtained. For runs where phosphine concentration was varied the data were fit to the equation: [L] + 1 AICJIS] k6[C] ' R: where [L] is phosphine concentration, [C] is catalyst con- centration, and [S] is alkene concentration. R is the rate in moles/sec. A is the product k6K5' A and k6 were the parameters adjusted. Using A as a constant, the rate versus catalyst concen— tration data were fit to, R = 0.5 A(—l + / 1 + 4 k6[C]/A[S] ) . k6 is the only adjusted parameter. Errors reported are standard deviations obtained from KINFIT. K5 was calculated from A and k6 since K5 = A/k6. The special case of four runs where alkene concentra- tion was treated as a variable is now considered. Output from SAM is in time. Each time represents the time required to use a certain volume of hydrogen. The volume used between any two times is known and constant. Thus the time interval is a measure of rate by R = V/(tz—tl). 111 The volume of hydrogen used since the beginning of the run is also known. Thus the rate as a function of volume of hydrogen used is available. In practice, the mole equiva- lent of volume was used (n). This form of output from SAMV was entered into KINFIT. The concentration of catalyst, concentration of phosphine, and total volume of solution (VT) were entered as constants. A set of equations were used which converted moles used to concentration and then entered this into the rate expression for fitting, as follows: [5] = [SI] - n/VT v'r~k -([L] + K [5]) ' R = 6 5 (-1 + v/1+- 2 2 ([L] + Kstsn 4KSIS] [C] ‘ o It is necessary to multiply the entire expression by total volume, in liters, in order to normalize rates to l 2. k6 and K5 are adjusted parameters. All computer programs were executed using the CDC 6500 computer at Michigan State University. Much of the inter- mediate resulting data was stored as card images on perma— nent file. This was particularly true for output from SAMV, Where almost all output was stored in this manner. This data was then entered directly into KINFIT. REFERENCES l) 2) 3) 4) 5) 6) 7) 8) 9 v 10) 11) 12) 13) REFERENCES Z. M. Michalsa and D. E. Webster, Chem. Technol., 5, 117 (1975). John C. Bailar, Jr., Cat. Rev.-Sci. Eng., _9, 17 (1974). C. C. Leznoff, Chem. Soc. Rev., 3, 65 (1974). R. H. Grubbs, C. Gibbons, L. C. Kroll, W. D. Bonds, and C. H. Brubaker, Jr. J. Am. Chem. Soc., 95, 2373 (1973). G. Balvoine, A. Moradpour, and H. B. Kagan, J. Am. Chem. Soc., 26, 5152 (1974)p R. H. Grubbs and L. C. Kroll, J. Am. Chem. Soc., 93, 3062 (1971). R. H. Grubbs, L. C. Kroll, and E. M. Sweet, J. Macromol. Sci., Chem., A1, 1047 (1973). _________._~___.________ James P. Collman, L. S. Hegdus, M. P. Cooke, J. R. Norton, G. Dolcetti, and D. N. Marqurdt, J Am. Chem. Soc., 94, 1789 (1972). _;______._.______ M. Capka, P. Svoboda, M. Cerny, and J. Hetflege, Tetrahedron Lett., 1971, 4787. J. A. Osborne, F. H. Jardine, J. F. Young, and G. Wilkinson, J. Chem. Soc. (A), 1966, 1711. J. C. Moore and J. G. Hendrickson, J. Polym. Sci., Part C, 8, 233 (1965). J. C. Moore, J. Polym. Sci., E. P. Otocka, Accounts Chem. Res., 6, 348 (1973). Part A, _2_, 835 (1964). 112 Jr., 14) 15) 16) 17 v 18) 19) 20 v 21 v 22) 23 v 24) 25) 26) 113 E. F. Casassa J. Phys. Chem., _3, 3929 (1971). W. W. Yau, C. P. Malone, and H. L. Suchan, Separ. Sci., 3, 259 (1970). Von W. Heitz, B. Homer, and H. Ulner, Makromol. Chem., 121, 102 (1969). F. Krska and K. Dusek J. Polymer Sci., Part C, 33, 121 (1972). Leroy C. Kroll Ph.D. Thesis, Michigan State University, 1974. A. S. Hussey and Y. Takeuchi J. Org. Chem., 33, 643 (1970). P. Meakin, J. P. Jenson, and C. A. Tolman, J. Am. Chem. Soc., 33, 3240 (1972) H. Arai and J. Halpern Chem. Commun., 1971, 1571. J. G. Atkinson and M. 0. Luke Can. J. Chem., 43, 3581 (1970). (a) S. Siegel and D. Ohrt, Inorg. Nucl. Chem. Lett., 3, 15 (1972). (b) S. Siegel and D. W. 0hr , Chem. Commun., 1971, 1529. (c) David W. Ohrt, Ph.D. Thesis, The University of Arkansas, 1972. (a) J. Halpern and C. S. Wong, Chem. Commun., 1973, 629. (b) J. Halpern, "Mechanisms of Homogeneous Catalytic Hydrogenation and Related Processes," In: Organotransitionmetal Chemistry, Y. Ishi and M. Tsutsui (Eds.) Plenum Press, New York, N.Y., 1975. C. A. Tollman, P. 2. Meakin, D. L. Linder, and J. P. Jensen, J. Am. Chem. Soc., 96, 2774 (1974). ...—- Microprobe Analysis, C. A. Anderson (Ed.). John Wiley and Sons, New York, N.Y., 1973. 27 v 28) 29) 30) 31 v 32) 33) 34) 35) 36) 37 v 38) 39) 40) 41 v 114 G. C. Chen and R. B. Anderson, Ind. Engr. Chem., Prod. Res. Develop., 33, 122 (1973). S. Montelatici, A. van der Ent, J. A. Osborne, and G. Wilkinson, J. Chem. Soc. (A), 1968, 1054. J. L. Dye and V. O. Niceley, J. Chem. Educ., 43, 443 (1971). See Appendix II for a listing of program SAMV and EVLT. Charles N. Satterfield "Mass Transfer in Heterogeneous Catalysis,“ M.I.T. Press, Cambridge, Mass., 1970. J. A. Osborne and G. Wilkinson, Inorg. Syn., 3, 67 (1967). C. Tamborski, F. E. Ford, W. L. Lehn, G. L. Moore, and E. J. Solski, J. Org. Chem., 31, 619 (1962). L. Porri, A. Lionetti, G. Allegra, and A. Immirzi, Chem. Commun., 1965, 336. .__—— K. W. Pepper, H. M. Paisley, and M. A. Uoung, J. Chem. Soc., 1953, 4097. Handbook of Chemistry and Physics, Slst Ed., Robert C. Weast (Ed.), The Chemical Rubber 00., Cleveland, Ohio, 1970. Reagents for Organic Synthesis, Vol. 1. Louis F. Fieser and Mary Fieser (Eds.). John Wiley and Sons, Inc., New York, N.Y., 1967. R. O. Krueger, J. Appl. Polym. Sci., 31, 2305 (1973). D. G. Larsen, Anal. Chem., 45, 217 (1973). _—— Polymer Handbook, 2nd Ed. J. Brandrup and . H. Immergut (Eds.) John Wiley and Sons, New York, N.Y., 1975. Robert H. Grubbs, E. M. Sweet, and Sawit Phisanbut. "Polymer-Attached Catalysts" in Catalysis in Organic S nthesis 1976, edited by: Paul N. Rylander and Harold Greenfield, Academic Press, Inc., New York, N.Y., 1976. APPENDICES APPENDIX I SAM--An Automated Hydrogenation Apparatus 115 APPENDIX I SAM-—An Automated Hydrogenation Apparatus The automated hydrogenation apparatus, SAM, had its ultimate origin in an automated oxygen absorbtion instru— ment described by Krueger (38). This design was modified to give an apparatus which measures volume independent of the reaction system volume, and can be safely operated in a potentially explosive atmosphere. Other changes allow the volume unit to be easily altered and provide simple maintenance. The basic apparatus consists of the gas control and measurement section and the electronic control section. A diagram of the gas control and measurement section is given in Figure 17, while a schematic diagram of the elec- tronic control section is given in Figure 18. This latter section was designed by the electronic shop personnel of the Michigan State University chemistry department. The gas control and measurement section is basically a U—tube manometer. The operating electrodes (I) detect when the mercury within the manometer reaches certain limits. When the mercury breaks contact with the longer operating electrode, solenoid C closes, isolating SAM from 116 117 Figure 17. SAM—-Gas control and measurement section. Gas Control Section A QMMUOUJ Volume 2 0 mmmmzN o0 mmmmZN mo mmvmzN «O mmmmZN m0 mmmmZN NO mmvaN HO mhoum H MGMHB .sowuomm Houucoo oesouuooaolnzoaa m> coumuaflwds .>o.oa m> tmumuaam .>o.m H> mmmmmmmw mm>am> twoswaom qom nommoa mews NA soomoH mafia HA smouoomm mm “mouooom am weouuomao isumsmc causes me ooouuowam souuom Nd moouuooao ”Hague mos H4 mamcasume .ma magmas 120 ma unseen H H> mm — HO O 121 the reaction system, and solenoid E opens. Hydrogen enters forcing the mercury up the gas burette until it contacts the shorter electrode, which causes solenoid E to close and solenoid C to open. This returns the volume of hydro- gen within SAM to the reaction system where it is used until contact is again broken with the longer electrode. The electronic control gives different output voltage during the fill cycle and operating cycle. The voltage change at the start of the fill cycle is used to record time. This gives a record of times required for the use of each quanti- ty of hydrogen. The usual manner of time measurement is from the start of the reaction. Taking the first fill cycle as the start of volume measurement gives where V is total volume used at time t, n is the number t of fill cycles after the first, V is the volume introduced at each fill cycle, and V0 is the volume used prior to the first fill cycle. The quantity actually measured is a pressurevvolume (PV) product, which, like volume at constant pressure, is a measure of the quantity of hydrogen introduced. The fact that pressure change is the quantity detected has some impact on operation. When SAM is isolated during the fill cycle, the initial and final volumes and pressures are 122 constant and independent of the rest of the system. On com- pletion of the fill cycle, the volume within SAM becomes part of the total volume of the reaction system and the pressure is lower than for SAM alone. The larger the reac- tion system, the closer its full pressure will approach its refill pressure. This does not affect the accuracy and pre- cision of the volume introduced, but it does affect the precision with which the refill point is measured. Thus the total volume of the reaction system is kept as small as possible. SAM is designed to be attached to an atmospheric hydro- genation apparatus. This was accomplished through a glass T with a stopcock on each arm of the T. One arm leads to the reaction, one to the atmospheric apparatus and one to SAM. The reaction is set-up with SAM isolated. Then the atmos— pheric apparatus is isolated and data collected using SAM. The gas control and measurement section is mounted on a plywood panel with the apparatus supported on wood blocks extending from the panel. Glass portions are held on the blocks by small lengths of 1/4" vacuum tubing, split length- wise, and securely clamped to the smaller glass tubing. The mounting system is not shown in Figure 16. For normal operation, a length of vacuum tubing lead- ing to a closed flask is attached at L. This serves as a gas ballast, maintaining a constant pressure with fluctua— tions in atmospheric pressure. It also keeps mercury vapor % 123 isolated and prevents dirt from entering the apparatus. Stopcock K is closed when electrodes are removed, which maintains pressure within the gas ballast. The operating electrodes may be removed and cleaned, then reinserted without changing the volume. The volume may be changed easily by using electrodes of different length. Stopcock Q allows mercury to be easily removed when cleaning the apparatus. Stopcock U is open for operation. When closed, the upper portion of the apparatus may be evacuated. The pressure bulb must be purged by cycling the apparatus for two hours while allowing hydrogen to escape through a bubbler. Metering valve F is set to provide a low flow rate which will allow equilibrium pressure to be maintained between the pressure bulb and gas burette. Constriction O prevents oscillation of the mercury from side to side. Standardization is the process of determining the operating volume of the instrument. Two methods have been used. Standardization by rate comparison involves alter- nately measuring rates by SAM and the atmospheric apparatus for a single hydrogenation run. Rates for Sam are taken as volume units (V) per second while rates for the atmospheric apparatus are taken as ml/sec. A total of four to eight rates are taken using the atmospheric apparatus and compared 124 to those obtained on SAM immediately prior to and following the atmospheric rate. Volume is calculated by V = R (atmospheric)/R(SAM). V is then obtained as ml/volume unit. The average and standard deviation of these measured volume units is then taken. In the second method, a quantity of alkene is measured into the reaction flask by burette. The total number of volume units, including those extrapolated from Vo back to to, is obtained for the entire reaction. Since the total quantity (moles) of alkene is known and the total number of volume units is known, a value for the volume unit is ob— tained. Several volume unit values obtained in this manner are averaged and the standard deviation is taken. Time was originally recorded using a stripchart re- corder driven at l inch/min. When the electronic control output voltage changes, the recorder pen moves, giving a record of the time at which the fill cycle starts. A digital time logger was later constructed by the Michigan State University Analytical Consulting Service. This instrument utilizes a digital clock with a time base derived from 60 cycle line current. On receiving the output voltage change at the start of the fill cycle, current time is latched. This is converted to serial ASCII and output on 125 an ASR teletype in a manner analogous to that described by Larsen (39). Output is in a format of 6 digits, the least significant being 1 sec, followed by a space. After ten points a carriage return, line feed and space are output. The output is simultaneously punched on paper tape which can be converted to punched cards for computer input. APPENDIX II Unpublished Computer Programs Used 126 127 w22wmzoc n m.ooo.iowcu0u.uw ¥########fivtt**$*##$#$*tfit#$#**###¢#t*¢*#*#**##*##fimfim###¢¢####*fi#*$*#%¢$#fi#fi WJHu uc ozm «cu xewro astsssssssasasstsvssnaavvnsusfiwstss leeaeepq>acu nod .r.~uo..sczac.Eo_.owet¢uo H #*#**¢#####¢##***¢t#$**t$¢¢#*$##$#$¢$#$$$$$¢$¢$$$$$t$fi$¢$$$fi$¢t¢flfl$¢$$$fi¢¢¢fl# .mcco rmnccaa m1» aou >aazm omna we m_r» .emw «sac mxe ncu case when» a we vex» #####*#*#¢#t*##$##t###*#¢##¢¢#¢¢¢*$##*tfi#*#*¢$$¢$$fi$v1t$$¢$$$$$$fifi¢fifi¢t¢$¢¢¢¢ .wwweeqsocu co~. . .¢7cawnzcn~ zco~.eo~.uoa.oocuqua ##¢##**¢$##*#*$##$##fitfitfit#####*#*¢¢#$tt*¢#fit*$¢**fl#**fi4##$$#*#@fl¢*#fi¢#fi¢#*$# me much memo: - 1 .mr mad» ua1a3 poz co - 0 "zone pzn»:o > u> a m .»:aa:c a u> a u a .eroeéc ».m> > - 0.. u,cc» oFIO 1mn MCFC 51mm mp Dodo u~1k $##*#$*#$*$$¢##*$**t#¢##$##**#**#*#$*###*##*$¢####*t**#$*###V****###fl#¢¢##### .~I~o»qzoou oom .a»«..t.Mppa3 ##**#¢##*n*¢**###fifififitfitt*#¢t$*$$$*fi*#fififitfifififi$3$¢*W*$#*##fl##fl*w##*##¢¢¢##**fl .oupe:_4wc p2memp44a72 kc; mad ation rzwsweqwm.:wq+amo .c.m zapuaw> cua7mexm 7 + was > a the? omewumen vuzar momauu<> >uqzaoa wyznn> ucqau>< n q> .wrwh m0< u re .mpqd n 1 .w:34c> u > .u:p+ uh .amzeeaxc ma tuna >mr csrno> no wise wruau> when uc mrscu u>ph<>Homo .e~u2~¥ CHZH krezweqcu w.¢qa~:w.qk«c wzw» uDvou> u23.c> Cezm wvur» spewezco c2: :qm scan wu:sp ya. a Umzq» reacton were $$¢¢$¢¢¢Vt*#*#*#***fl$#*fi####*#*$*¢*fi*fi**Vfi ##*$¢$¢#*$###$**#fl¢t$¢fl¢$t$¢$t$fl¢t$$$tfifl****#t$¢$*$¢$$¢t$¢¢4fl#tflVfififlflfifltflflfififi .avze..muycr:>.iaemcm>.innucia> a ..ooicq».,ccmcm>..ocrca.asanc>>.10am.>..,.1.»>..coree rowu2wzac .ezoeocu.rmnch.»:nzuucrmoae.mruna».h:na?o.k:azee>z2¢ cmpomawc m< zwewmcz mM,seqh-p:me:c manta umnecm vmoo<> are >m cmecmawc nooqo mzmenoorom uc mlHaum a .c.ncu usurp assasstscsasvetsnesssstatvswassvsswtnyssssnatsssasntsmsssvassssautmnmwwsmmems Um... 2 ‘ N 4C>ad>$w+NN<>sxiHec4m«_oolvnxuuaw~:> ozHJn>ci~cwenaae> wn.on.wek> erotic:iee»uasee unoz.snw on cc ~¢$02~Jc>uwn<> *nsssnstasteamsssntcstatestamasst:astmstwsussmsmssssmsa.t.rszsaanmnnwsnmmassa FZHCQ 104w QCU mnww affine EsTJO> mknd>.o7w4®..uzwuwcamugoomvuJw7wuabw Va . wkq~23 an 7«I» uvmg ch 4e24: Us atu ox:rw mo_ec .-w o w cmnw - Cucuwv u— w»7acu o: tdrk wmmJ used or in:::u mama. nuy~ iWwi wczcuo whqdu Zurw odrcc mczcwc C» ozone uzijrorzu Upom.um wt n cusp z. w . awn” .ahqc v4 »:nwzc sou M queue Freon I» : >am>u abouluu cupamzc orq .zceh02cwa C? a C24 m .cczpu. romeo:oma .cuOH emu «rec «Hr» 2a 1 , :HP.. mkysrn .eacwuc amaze; .pocz . . z: uz:.c> n?» no ozaflda> .oo>ne> 22a ur» 07H022 Farr irzoc> wtpdqaoae are . utna> .u2~.c> wipw roam roan FUdnhIIw Ch pZTcrEq «uzwuwc .pmm «vac roem won :rcm 4cmtzcu mi» we uMIP $$$$¢¢¢¢$t#¢#*##$$#fifittttt$flfifl#*fi$fififl*$fififi*t$¢$$flfl##fi$#**##?#*#$flfl$#*##¢¢$fl¢# \\\\\\\\\\\\\\\\\\\\\\\\\\ \\\\\\\\\\\\\ guzzwkzcc :qu \\\\\\\\\\ NH #0 s-‘LL! I! I- ”.1 '\ UUUU (JUL) UUU L'L. L‘L’L‘UUUUUL’L‘LNJU 129 \ \ \ \ \ \ \ \ \ \ \ II \ .. g \ .. \ \ \ t, \ x \ \ "I \ \ \ o \ a \ \ o \ o \ w \ a \ w \v‘ .- \C 2 \LL) Lu \W \H 4- A U.) \ v 2 2 <1' \I" ‘- h- b-q 3 \O ? o o \C A 7 2 O \7‘ ' C O C \ 0 2 O. 0. w > C ’- Av— “H .- > (A b— C v C 0 Lin? c a»- 0 (LC LLJC C ' 1) C. 0 C QC KC (\J o LL'C 1" D—‘b— pup-4 [L C' I" ’7 '7: O I o o a :‘Q' Lg) h—H r1 ,. b—-|—— !- f-F- JC 3 o- o v > 'r- {1 [1 5— 0.0.. d 71- b-F— o ’3 CC a'l CO > o 5"" CLO. (3. C. ’2 .' C 22’. H INC) CC" o '2 t o ? o v t- 2 I. 'T 2 ..J v .1’ (I? 0 C"! Q, I '3‘: 0 0* C.‘ .T )3. (I >> '41-“ L, ’t ‘1) :7 be u o 3 o o -“, Iv- >> Ch > QC! > 02C! LILL.’ 2. o o q r o o 0' o o C,Q’_ §'_I ”>> > UNCLE CIFC'TJT L.) <'__-r.1(\ o o ecu—N I~r-- o>q>> 2':- m4 ,n-q— >> m00>> CH0) L)o'\>> cob-m0. 0>Noo C. 2 n—nm o o p~>-rr~qg ~>3y~p Hf? ~'\ Al—fi— >"‘ 'AF't— J “A>> a- .-!K‘VV O F—NVV >N 0") 53 0 0mm: t—r—nv vulm NV atom at 2 \h} “HUI V 0 £110 'QLIJUID \! O4 .. --.. L-EC‘W-IC 2‘4 Odd)?!“ ox \ on} 'C"! hum: '50: [L'EI' 0CD F‘lf‘ in \C‘ oLT'lL‘Q—uc .>CLIJ(_,'~ g>cwu (v* \U'CUJOV IYCF-urru (TFUOWp-V'.-..Z ’ 'Q C' .. . C‘. CY CC: C. LJfIU‘J" \~H~_xmhq*4:.'o;w- W~-4c aw» \vvvd ”| V 'V—J— I”. .2c ..Jv \hkkchdu4I44Chdttc «OCerC C7 \~HHUUCL3HU>‘C'I.L "k-OHL‘UC 3h Uh M ”#5.!" 0 ve- (\f’ Clr‘ 0 F Cf” C Nfllfl. m m ("0" -3~3’ ¢uf ”‘3 g 130 wzzw Czcc : wawhfco ofl 2Hu.H~IH w". «~ n~.p~..~.zw u~ ozwlflozuzp u— ##fififififitfit#$$#*#####$*##$#$$fi$$fl¢¢$fi¢$flfifi$$$$$**##w#Jfi$$$$w$flfit$¥flflfififl$$fi$$fl¢ oh UHUUCQ v» 7chusowa auxpzsm c2 m4 2mXap mH accu oo ycou ppxu .d p. ww 7H uH ..H.x~ m4 cmnc»m ha 44H; » CHM pg Fez-m» 2% up - uszoa 2m: m1» no ,Jazowrwuzcgv uwczH-u1p.mhfl2H zuxqp mm ow pzwcu owhomgmm uxw acu z :oau Upzwcn uc 1mmybz uI» UH L7H **#$¢#*#$#####*$#**t##*######**##**#####fi#*###*$*$$¢$$##fi**fi##**$#**¢fi$#$#¢fi* . NH+OkwuL7~ ,fl,flvk<:.u.~qomvx»mwnnw ##$######*fifit###*#*#$**$¢#*$**#*$#*W#*$*#fl##tfififlflfifi##fifl$*#**$#*t*#$*#%fi#fififlfifi .kaoo kxwz mIp cw hzuzmaozw-uc wmqmouzh uc mkcc m1» Ukuu hp . .chkqoznak no wk mark .Aw.k24 ma CMYHumc mw >4: rm ##fififittfifit#*##*#*#*t$**####fi#####$#$$##$#*¢$$####$##$¢#**##*fi$#*##$#*###N#*** z.~umc— on” . ou92w w *z$¢##$#¢#$¢$*##3fltfi4flttflknunmfittmm . . , - z .uwwuunocan 22a bl» ua va¢uaozw £24 hackw mIk »« _4<$M my 20H+4m+zuuzou 7m wCZaIU 4mZCNHUaau m1» maul: zcfipugazcu cw «22m Ikrz um: «cu euoyupzw up prk .ozw an» aqmz F¢Ip czpzwapma w4H12.zna ark uo paqkm mrh mam: ca.paomtaHvau.H~a> wo\>cu.hva Vm¢>zxqo>c<>n>ca mamkexu.on»ca .thl.yvknho .Hv>n.y.>n>c +Hu¥ buoz.~uw » cc ~uhoczupocr ##fiafitfififit#fiflfifit#$¢###*#*#¢#$###*#*$$$fi$$t$$¢$$*fittvflt#%*$#$¢W$#####Wfi$###¢#fi Ukz~cu uc cyan kZmoqfio< roam 7mm3wmm wpao mT» wwpqg:u_4u m7».:oam:u awlh **¢#*¢fi*#fi####*#*#####*#$#$#$#$$##$*¢#$fi#**$$####*##$$##***$*#$fi$¢*¢#$¢##$#¢# .oo«.o>..oomva..oux.>..na,.»>..mom.» zo—uZmzwc ,huoz.o>.a.4o>m4>.>.h>.pvmpqa m7ww:oaa:v Ciw w:2~»zcu cm =H+nmaxzucunp “chow ma.cn.cuah.uw Mm ~m.«m.om.oouzwuw - w:zwrzcc cm ..zvxwvcu,Cvc 5.x.rwvun.;.o ..vviwvan._va .Axvxqun.g.< ~+4|fnx $$¢fifi$$fi$8#*¢*$****###$t$########$#$$$$#$$fir$fittfifit$¢fi¢*$$#$$$##$##*#$$*#fi##fi ‘. - . . ‘ 7u,_er .Mmy .tu; .rzy .fiux .CHJ .h04pu . .xwozw u:4<> huwzca cw kuuxohx m»H scan cum: ma buzz .MVIH . mzqumo us 0» p7—co q me xmczw puuzog m1» vzwqwrco ,H.IH chzp Pmmwa<4 NIP muzwv ###¢#*#$¢t*###$###*$###*###*$#**##$##$¢##*$$¢#*$*###$¢##*¢#$#fl#####¢####$#¢#* 1.~nqom co #######fi#m##*#¢$#4###t##$¢####¢#$$####$$##fi##t#$$$flfl$$$$$¢$¢fifi$$t¢fl$flfl$fifi¢¢flfl .c.c km<4 nip V» 2 02¢ xuozm k2_co 2m; puqq nth mp : .pzfiou wpxp pc $$¢¢$$$$tfi######*##t#*##*$*#¢*$V###¢##tta#*##$¥fl##m¢*#***####fiflfltflflflfltfififitfififi \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\ rbzzrhzcu u::m» \\\\\\\\\\ UUU L‘ULJ U UUULJU ; NI—TH POINT vv-‘Ov s‘af‘CEin 3:! =‘J :3 l? 1:: fl 2;! 3:! :1: x: t" 3 3? =3 2) a.“ :3 $5 3 tr 1:: t f} :3 2’: {5 :5. fl # f.”- t: t J. "Aéé?£‘-‘$§*#*§§fi-‘.‘fi§§fi4§bi§§éi§i$§é§§fifi‘fi’kfifl’hfififi-‘J v ‘ 7‘1 ~1A ”<1 ’ ' 3 L "_ ’2 (2‘4! 7 r" a; r :1: "' 4‘ 1"}! ‘2: f‘ 14> r t: '1 '>:f' // :f'. \‘ Q- :11: V :f' F ”.l ,- =I'- w. =Z' t: (_ r '4‘ '1 =2 3’ :1‘ :5 M a"! {'1 j: ~- "x-v 1, if ‘brt * C ' 9:1: 7:: J > ’ 3’.‘ (I '7'; ‘- 90;? I» '4'» Y 1: (wilfi‘ 1‘; ' fl: . '1‘ - g ‘1 ,7. -1 L x’} 'f )1.’:‘,5;J‘--:::f\ u- v t 1,: ‘j' 7;: v a h. ' 7?! H Ii! ’7’; “Ch-7 4! ALL“! 2.4!; =x.‘#r—’:C C at: " 1’) (1. H‘. 'J’d '5' "mm r ..r» "I “1" P--";" "4““ V4! ILL”: "1 #1 LL 1.! ’31.!) LU" ' 7716107th 0)! I»— (ft: ." i1! lr-‘lf‘ t-r—fs 2,“..33 C \(" Z‘ l.’ 11' k U ’1 h ‘4' C32 1‘4! .412 CY LU'" nu zirU‘ nth-O": C; . #VO#¥F'O :‘V fi‘IC'i‘ U-T UC tr.- LLi‘V-“I‘LL #5 4: g; at: pal: at: i: (7' ,4 Q7714 x.- 132 ifil ’3‘ I“ ’1‘ 1:! 1“:- l‘.‘ # 1'! $ if t 3“— :0: SC! 2' * $14. it t() U‘ k i‘ (V U 1":- i' '2 u: :3! #C .4 :r ’3! 4 t it LLJ =0 :31 2‘ :'.x *L C # asr<1 1!: fiuJ C’ it t I-v fl lit 7 1’.“ *9. C‘ ’1‘ I6".\. LL! 32! fr’ 7 U It #P‘ r ‘1 flu.) 5‘! t 111’? LL! tr f; F? it =‘r'LL: #1 *2. ‘1? 33* :' LIJLL'fit "r’ ..." LUV"! 11¢ ’7 4* an: > 0 3* .: (\V 2:: dud: C 017$ r‘ “a .' “P33 1' ( —-’l.‘xfr ‘- 7‘ '-' “(:fr‘ I“ 0 ‘P‘ V1713?- (1 ..x. 3" ~ g: C‘ - }'( I17 1‘. 5‘ at >Y c o rv ,— - >1 -. . v . '\ y \ 5. o- ’ a» ‘5'— ,v ,. K ..L‘ ...: .' I. p- v¢ -. Wilv— f‘. YA I»; C *Y ...4 2‘ I 0" _ C ‘ I YX f‘r‘. . N V) >— If U X'VH.‘ ‘— :_ I r-w V”: : , u an; 2»: Ila-2' I" czm :x :1: Orb-“WT (YD. 3'} L O C‘/* ”.V‘VP‘: {72’1‘VF‘Ch (Ill V“flk:" :n—$]UQ :1: t! :2: “ C :3 x" "‘ k‘UUL—‘UL‘L 133 10! a a it m at t D— 0 g: =11 LL (f‘ :1: It 2 C :0! 10! d (I a): It Cr. <1— v: #1 <1 92 tr It 011 1.1.1 :8 find I? t t C on: :0: #02 21— :1: 1111—4 :4 t :01 Q!- 1‘ 151-07 V? # #UJC. m In aura -U 1—l- a fir.’ 03h.) t IO! JJNO} at: #dc'sc-b-H t #t—‘CIUJ'Z‘I #1 #b’b-Q‘Ob— =1- atm-.24 #1 #261—111 ‘1 161-10 04 t 10! >- I1! mmwmwv— It «one: HO .JU‘) :0! A A»- 25* :C‘Ha at; y» .... H~# z..1u.1u.>- :1: o, 0.11 oZtcob- u. 4: c. CC. zuuazm— 7.: 2 22 C' 9* HHZLULL' $ .- 0 PP . . QC’V C2 ’1‘ O - QQ .... H015! r-dk =1! C 11 CC nu OOVI—b—U‘S t 2 fl - 7.7 ‘7 Gift LLJEIZ 9‘ ' O A o p 9.. CW"? OquJv-I r; 2 A '\ 27 A, H>¢Id a f; C '3 v or: 5’) «>11: r-dh.‘ —- n (1 v -,.< oq V- 5- 033412.14 =1: H )- 1 b—AH )—>— aArxcocm > a: . .> A .. >> 00* CLLU‘Cr a? C v ”' CC 0 ' 2010‘ FL. v=o 0.1—32011.: f; - v v"\ (r o o -1, >>—#1—--w-—qu/‘ f; .- r \\‘- u--.. >> 00*311 34 1.1 r— 0 Ar- r—hvsrs .. >‘l"? C7201” 15‘ ' '- a-ac o o v 0 fl” 9019' HC'C IO! -' "‘ “MC xCe—a— "“ «cm in! .10 t‘ 11 v vvzf «~11 n .... >11$uJ 7~>~ -u'fi ‘5. V‘Pchb—l at o >A |+)< y—ooo ;>‘; Kxatwhi—d It! a» 0F .....\ >< .~ tv V>t (KEEP-Ulla”: ‘EU‘ 0 ma» olr“—L qppap LU '* 3H“ 43* ”H.3Lr' VVF‘ “*V 0' 0.3-.3!- CZLIJIQ'LWJ Z Mv-u‘ >->-*-' D. OZZU .37va HOfilgh—‘zm ctr C H oqu w> CUM-not or’f ox a moat Lu znrva ~0er 111111 3.....: v.':.~~— w om¢~mo>cuau .....-— jr':_ . .-,-n-~-A~ mvm- 240. «A —~~—~> o Ora-«fr :1; 3H0! 111/1614.13 CC!) CO. OILCuLLLD—l': 1| 0 C8070 C‘CTCKCI,CL|"' 012*P—h-u't-ZCfILC'UJuELLOUM-«hw31.».3E‘F41L533unu‘ IO! # :6: tr 0 “- n o c .4.» A 11! It: H o o N --4 o». o v— .4 N NN N UUUUUUUU 134 fl\.cu>n.cu> :ru.. p+4uu.a . anxm.un.qn Xcu \\.mza+,33<§hoa qu ..cwu awkkuq C": flaw: uuqfiuwfln: .mmun_cu acnou u>-<4wo ax» v” msq2 uqzq_na> uxb ck cmxHuuot w .ngauwo.> +cTh nCu acacu cachm ru.»42~kmm u:» m_ utsp ugwahza; :» wmxw.*unc u 2c mHZWCq r rmnwu at» 2:; «a qunw>q AX< mh:_,.n an Fmowu u1+ q u aw qupw_i ,yq . .. Era UH .c t_xraaca whi» >a ouu:ccosz coo cam m:»..4<$u pr2o [p CHJcm -2» swan“ m::.c> uxk mquu .>4kHuH4osH omzsqu «w 4; u.“ uc >p»uz.ug rcwmz cywu: and: wcoo_\qucs ac J\vw405 awhzw mZmIqrcra ur Pick»; ac mzr_c> ;> c coopxmmgcz ac 4\mw405 omwzm mzuIQVCJO a? 2u~p40»2mUwcc Cu Uzdoc 0C 4: Quk.m m? IQUCYJ .1, wlzwm. WC “1?.C> _: 4\wch: quk.u u_.ux.n mi rcwkdnpszYCL Us: #2 wu»2h :2 yqa kc wZIJO> u:u> J. nut/m tut/11.}. ..C uZDJC) JCU> ms44pw zaaccon A ##**#**#$*fififififi$$$$t##fi*$tfi#¢#$#¢¢*¢$*f:*V fhmfifit. ## *####$*##$*****##*#fi#¢#*¢fi. *##*##*#¢fl$*# ****#*$#$n#*###**fififi##$¥**###*###*fl$¢$fi# Bfl>w L‘C’C.‘ UL'LIUUL‘LULSL‘L‘L’L‘UUL‘LXJL‘ (JUL) UUU 135 $.n..n.fi.fi.fi.¢fi.*n.n.$¢*fi.$.r ## 4J~J4fi $4141.: .0... . ......z. #3. J: sv.......ffi.$$.n.v .71..” V*.x..u.#*n. ..., .H , ?.pq._..;7...L LI» V» a: armre¢crq Lc :__1<.».-c;:u 11 >le (I ‘1- U 5.: Jr 2 1:111 tm_:.m» mm 19 13).... m.» m0 MTF .Uw ,9: mlLL mw L: .Hv» w» y: at FDQrZC ##r.$fi.a¥$*.¥$*$.nl¥..JnjfifififiH.141...$J.~J....J...........n.....J.r.n..¥.r..J.wf (.n.fi.w.r.n.Jr~*fi~A .n.fi..n.~: fl.»0Cmn»:u sfiuyau1fi> 9.x . "v T... Vpufiwuwnu .ccnvnro> au\.1u1a ,V»n(mu.:u 1..CI...¢I_}.vguw _>JJ:0n.r> 0:1.... 1:.,LHJ:,..W ##LDhUiU U5: ~>Q+ETU¢ucru L»\mas¢a:m>uur 1vhocmucnu w$*ua¢ h>n+xcau\¥¢n <0 .ur>vkccun me vauzxoran L7: .r:¢:A.UwOV\.L3n L: m¢¢Ap>xk>uvnw>a 1nxyztuu+nv:4cu..xntq>mvkocmuh>u _>+40m>.¢cm>np> \n¢.xq\x«uvnxao o a111.\r>,..naa.(:w>\nsmu.ua:vn ar.03\u:wvnura Mw.a.40\;omvngoo Rafi. <\J>Mvn.>a .ooo~\m :wuum2wm .wso_\mr vanmzw> .ruufix4cuuu3cvu .,uo~\4>wu1>w )mnmo.~+4>n< .woo~\4>uq> \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\ CuDT~b7CL »_>m \\\\\\\\\\ L‘k L'ULI UL>UL2U 136 \. \ \ \ \ \ \ \ \ a \ a \ \ \ \ .- \ O \ \ ...) \ * \ U \ \ LL) \ > \ t \ > \ .J t- \ >1 \ U [s \ \( \ o . \ P X \ .J v- \ x \ > H \ >1 \ LL r— \ y \ O o \ K \ —J '- \ ~< \ > H \ o v o o r o c g» \ o v \ ,. \i‘ \ C 2 \ r~r~ FFF-hl‘ \ \ 7 Z: \ o o o c I I \fo \ u; D. \ fiLf‘fiLDlD-DU‘ \X> \ . b—l \ .... p—l ,— \x 9 \ u: ~— \ u.1uu'u'bJLuu 0 ‘U‘ \ Q 0 \ 2.000 by . \ 2 \ . aux—1 \ o (f \ I o o o. o o 0 4x7: \ a‘ 2' \ .- x xx 1‘ >1") \ 3 > \ 'J‘ U’II'f'JLf 1‘5)‘. \ U‘ o \p. vvvv v ,(__; \ “J u‘ \\ «\\.\(\r\'\ "'f r \ 0 f :{r o o o o .- A 2 r 1‘ o. o’- finflrf/teztfl A ‘JJ ’7 L" __J CM! 5 5 1k .- ~.:: 4:: : '- U' I «u a »»- __;;- #P‘ u,— up. .- > > {HP f _;_j\ __rVJ [‘- nJ—‘_r“,'\r—‘3 ~14 VO'_’~’:‘ I: F" 0 Of 0' > 3. 13 3 TL'C’U'CC‘TF’VCJ 1~n~ KL‘ 9" ' ..l 3‘ > 0 D \. n ._Ib-:H;'_' c t 1" a v— or”; 2.- C o o .- o \y b-ijthQI’)‘ . .yyu; CH 1/ a .4 r '1 \U a}? row! KVU‘UV— L‘v :5 {V !‘~ I ‘1 \ a I b n — ~1 . 3" u 2 0 cd v~..J . \r; ‘(f .14.: on . —a:~\-—r1><~-, P: Q F‘U‘UDP‘Q} (J \ .-I CC (I 52:: O on»): at .10“ 3 >ZE'U1‘1L') I \ ’7‘ 0“ 3. 01. t “11‘ 0 fish vy r >"" [U LULIJ)‘ '> 9 ' > \w" L... *- Q D 0 K7 )1)‘ La" 0 Otto-‘4 ~ \‘1 1‘) [-4 o:-— Tc7‘2’b‘7h.’ 4} (=.Jr-“ fiiffll’tik >“L‘” I" ">41 AAA _ Also-AAa-AAJ- of 5 :1 f1, .- ’1 'X‘x v-u-IN flJMqu'oq-Tfl'n \ o o—«y ~\-<'x. Ian/um WM aka-Hagan '51)- \ at x on - o\x v \K’V‘N (”W'Kfi'mmfl,‘ (\Nmm ra\u on c a v u a l v t(\:‘( c I)“ \009 '0 u-vooomI\ nuyy _><" t)? \c—n ...- (\ Muir-I‘M" ~M~-~\\U L‘ U" ”‘11 U‘LF’J‘f‘ \w\(\—- \ \Ef‘fk’fflflfffifofi-Vvvv VV~VVV~~~ vvvvv vvvb—i—v-r: v—v—t—«r—v-r—b—o—ww \uJIIvLLl vmwwmuamwmmmadq «<1. («dd 44de \r—r-r-Hr—r- P 2.22)}: 2’22'22222’, \HP'Hyb'bMo—1Hv—w—hv