‘ _ . Ivan-u'l""‘.““ _!¢a.;.1-‘-:‘9‘.'v""l ‘ V . .A" . . 1E . I 7 3% ‘1 ‘A {6.0%. v. t .5312? ... . ..Mfii. .... QM» v A q 13......ucaawwf. . fi .. ,3. 1.9.23.3; .. F . ...: ....Mwnm It . an ..d . $ . . a .. V0. 9 o. $.mfl F. 2 Q iuwnnsoyznsfl. $3... turf. . 2. 2.1%..» ...w 0 1 Hi .- 92: 1 7' . 5...fo .azyh E. y '0‘...) . . . . 2. . . I . « .15.: r..I.....:.. . . . . . . . 53...“: a &. §E§v!,4.m?ut+ t: y. . m . . .z t MMMMfiMM yyMy .m. . .. IV. 9. . .- ...-rm“ . {$37 1T3.th?..+ l '9 . y '05 ..A. I . 374.353.0490.. 5..-!Whflv 4pmyvo....§ip.‘k..lfiflb.w ‘rtfirzg . . o? .ivyfszyfiibrtte ’rxl : .. .I! 7.0. f... biz]: ...; I. 9 .595Lulv.a7‘! . . n? .h’rfit.‘ $10 .32....11133 ..7 . viabifilhfi . z s ’0‘.‘ . 4W2 Fur-£34... 2.01.6 I“. .z: ¢ :IUJN‘ILAYVV . ... .2 . . 1 1’ like __ .. ..., .h: . B .y o {grit-.- .n. . z § y I. r . 3mm -. . 2.17193". kfihlbmwwrflvag. V . y 9.. Mining. . z. 2, .. is. i on»? - t. .. i v I + . . if. . {my ...:L .. augefiidq . . . fi.‘ 2an 1~€.n~.?.v?‘f.tro .- y a»??? y 2,. zrhvzyzyzsa . $1.41... _ ... u . a. 7:1 .. h... . . L 2.. r 2 . «hwy: q. 5..“ . .33... . . . ~LJ . . Q. 1+ yuan . . _ .. .. . 4 he. in“ '4 y " “human-Ru. miflzcfifl cud? 3.. 5:7... . ”V153”. ‘ IS! «#27. N. z. twhortck‘lloflzl a ENHAJ 1 I 40 5.11 301‘}: P L.....zyxnn§§nm? . :nwy an: ..mum z - 3 .. . At . .2 mu .y. 1 ...t. . y .41.”! . ..umynfinwflfiuués r. ; . $A A c . ' . 5 Pg . «wxmehAs M 55.... . IL at, .... mm»? . {.7355: .1 . . ts.) Agra} 1K .. FA.» . . ......ééfl. ....er3, 51.. . ecu». an... 95.3. . v.9!itfili!) . y it" 3.5,“. 3v . l 1.23;. . ...: a .44.... wattsmf writ; v.3: ... Eh..." F. If,“ ‘ I ig- {fiffixgflffi N‘- C is Z 4 p weakyflfi fag”... Al . . . y. 5~A_L. Iv fliflWflWWfiMWMNqMHm‘.‘ .. . . -mfimfla 3M2...” in: ya. . .. ‘ 4 cm» .Sflfififlmwfl ,. A, 73*}. . t .C. . . .....a £3 yank . . £91.”.i) ..t. . y . ... ..ra. . . . ...} . . .. . 4 ; M . . ... . . y LVfimL. Ln<flri “raw? . . . . , . . . , i 2,1: i .b... 1 1.7m; y . . . .. .u 7.)..th fiaflrflfi. y. Edna. Juu _ . . . u. r a . : :2 .u rat! , .kugf. ....” y .. ..u. :1. . , M. M? .r . “.7 J y y y 1.4.. J. “#7 ... s, . 26$... .4 And... 4 t 5.3». Quint».- q . y . . y. . .11“. .4 4 . g 39‘. M t.n.w§wn..:§.wyy1 . .y.:...ymyis...y.i . ”flit “v.93 . {Ian I x . #01?ny :1: .1fiffl. Jfiyh , . l .VKUdmflukh , a. < ...Wfl-afux nvaioif wt: y .. licAJmDUIA .m [JAE-VIM”? ... . . firmxhuufiwr #14:}. 5.5:: 41.7.. 1.?! . . A ' . ”5‘ ' ..M. r. (.31 $21} him. > fa ‘Vk. 1...... 3.. fut"? ... t HQ... .1 y. . S} .. I; . :\. .uu‘aflh with... ... 3 3.9mm... bar; u! .M.».( fish... «35...? and“ r‘ ... Lynn” v .1 . “a .r J ...I.L~..a.m‘:: s L. ”in. V‘ 4: 3 . :2 s , . .91.? 2 iii. 9.". . . a}. Enhannfiiw . . “numb. khaki. ... 9r“. MSU LIBRARIES n. RETURNING MATERIALS: P1ace in book drop to remove this checkout from your record. FINES wil] be charged if book is returned after the date stamped be10w. in mm I. SYNTHESIS AND PHYSICAL CHARACTERIZATION OF SELECTED POLYSTYRENE SUPPORTED PHTHALDCYANINE COMPLEXES II. ELECTRON TRANSFER STUDY BETWEEN TITANOCENE DICHLORIDE AND REDUCED TITANOCENE DICHLORIDE ANION By Michael Danie] Gebler A DISSERTATION Submitted to Michigan State University in partiai fulfiTTment of the requirement for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1982 ABSTRACT I. SYNTHESIS AND PHYSICAL CHARACTERIZATION OF SELECTED POLYSTYRENE SUPPORTED PHTHALDCYANINE COMPLEXES II. ELECTRON TRANSFER STUDY BETWEEN TITANOCENE DICHLORIDE AND REDUCED TITANOCENE DICHLORIDE ANION By Michael Daniel Gebler For the past four decades the rather weak catalytic activity of phthalocyanine complexes toward oxidation process has been known. As a class, the compounds are inexpensive and widely available due to their use as fabric dyes. The limitation to their catalatic activity was claimed to be caused by dimer formation between phthalocyanine groups. Recently due to interest in the isoelectric porphyrin system role in biologic oxidation reactions, work has been renewed on the structually similar phthalocyanines. To be active in oxidation processes these systems rely on site isolation of individual metal ligand complexes. Such a system is reported here. Nickel, vanadyl, cobalt, iron, and manganese phthalocyanine compounds have been attached to polymers by binding the chlorosulfonated metal phthalocyanines to styrene divinyl benzene (20% and 8% divinyl benzene) copolymers. Attachment by direct sulfonylation and by Michael Daniel Gebler formation of the sulfonamide linkage were carried out. Physical characterization of the resulting complexes was accomplished by electron spin resonance analysis, transmittance electron microscopy, scanning electron microprobe analysis, and differential scanning calorimetry. The effect of the presence of the polymer resin based complexes in contact with cyclohexene during its oxidation are reported in detail. Additional oxidation reactions with these complexes were attempted with l- and 2- octene, toulene, and cumene. Biscyclopendadiehyl titanium dichloride (titanocene dichloride) has been reduced with aluminum to give a reduced Ti(III) compound with a pronounced electron spin resonance spectrum. The reduced Ti(III) complex was treated with the parent titanium (IV) complex of titanocene dichloride in tetrahydrofuran. Linewidth broadening of the ESR spectrum allowed calculation of the rate of electron transfer between the Ti(IV) and Ti(III) metal complexes as k=2.3 x 108M‘lsec'l. ACKNOWLEDGMENTS The author would like to thank Dr. C. H. Brubaker and the graduate students of his research groups. Special thanks go to Beth McCulloch, Fred Batzer, and T. Tomi Li. The assistance of the mass spectrotometry, electron microscope, and microprobe analysis laboratories of Michigan State University is gratefully acknowledged. Very special appreciation is expressed to my family for their continuing support in my many years of graduate studies. The editorial and typing skills of Mrs. Christine French earn a special note of thanks for the many trials and revisions made in this manuscript. I. INTR A. B. C. D. Research II. EXPE A. B. TABLE OF CONTENTS ODUCTION . . ......................... 1 Introduction to Catalysts ................... 1 Introduction to Phthalocyanine Compounds .......... .2 Supported Phthalocyanine Compounds .............. 4 Introduction to Specific Polymer Supported Phthalocyanine Carried Out .................... 8 RIMENTAL .......................... 11 Preparation of Chlorosulfonated Metal Phthalocyanines. . . . 11 Preparation of Polystyrene-DVB Copolymer Beads ....... 12 Sulfonylation of Polystyrene-DVB Copolymer Beads with Metal Phthalocyanines .................... 13 Preparation of Substituted Nickel and Vanadyl Phthalocyanine Standards .......................... 14 Preparation of Chlorosulfonated Nickel, Iron, Cobalt, and Manganese Phthalocyanines for Sulfonamide Reactions ..... 15 Preparation of Aminated Polystyrene-Divinylbenzene Copolyer. 15 Preparation of Copolymer Attached Nickel, Iron, Cobalt, and Managnese Phthalocyanines by Sulfonylamide Linkage ..... Equipment for Analysis and Physical Characterization . . . . 17 Elemental Analysis ..................... 17 Oxygen Uptake of Various Organic Compounds Catalyzed by Metal Phthalocyanines .................... 19 Catalytic Oxidation Studies Without Measurement of Oxygen Uptake ........................... 21 III. PHYSICAL CHARACTERIZATION OF POLYSTYRENE SUPPORTED METAL PHTHAOCYAINES ........................ 23 A. Differential Scanning Calorimetry ............. 23 8. Electron Spin Resonance .................. 25 C. Transmittance Electron Microsc0py ............. 28 D. Scanning Electron Microprobe Analysis ........... 30 IV. RESULTS AND DISCUSSION .................... 34 A. Result of Treatment of Metal Phthalocyanines with Culorsulfonic Acid ..... . . . . . . . . ...... . 34 B. Results of Sulfonylation of Polystyrene-0V8 Copolymer . . . 36 C. Results of Sulfonamide Linkage Formation Between Polystyrene-DVB Copolymer Beads and Metal Phthalocyanines . 40 D. Result of Oxygen Uptake Studies ........... . . . 43 E. Result of Oxidation Studies Measured as Total Percentage of Substrate Conversion ........... . ...... . 50 F. Non Attached Phthalocyanines Standard Results ....... 56 G. Other Experiments Carried Out Toward Oxidation with Supported Phthalocyanines ................. 58 V. CONCLUSIONS ........................... 60 PART II I. INTRODUCTION AND EXPERIMENTAL ................. 65 II. RESULTS AND DISCUSSION ..................... 71 APPENDICES APPENDIX A ................. . ........... 84 A.1. Iron Phthalocyanine Supported on 20% DVB- Polystyrene SEM Trace (Sulfonamide linkage) ....... 85 A.2. Nickel Phthalocyanine Supported on 20% DVB- polystyrene SEM Trace (Sulfonamide linkage) ....... 86 A.3. Nichel Phthalocyanine Supported on 20% DVB-polystyrene. SEM Trace (Sulfone linkage) ...... 87 iv APPENDIX B ............................. 88 B.1. Polymer Supported NiPc (used) In Oxidation of Cyclohexene at 7800 (Graphs start after incuba— tion period of 16.75 h) ................. 89 B.2. Polymer Supported NiPc (ground) In Oxidation of Cyclohexene at 780C .................. 90 8.3. Polymer Supported VOPc (new) In Oxidation of Cyclohexene at 78 C ................ . . 91 B.4. Polymer Supported VOPc (used) In Oxidation of Cyclohexene at 78 C (Graph starts after incuba- tion period of 15.5 h) ................. 92 8.5. Polymer Supported NiPc in Oxidation of Cyclohexene with 1,2 Dichloroethane Carrier ............. 93 8.6. Polymer Supported MnPc in Oxidation of Cyclohexene, Sulfonamide linkage, New and Used ............ 94 8.7. Polymer Supported CoPc in Oxidation of Cyclohexene, Sulfonamide linkage, New and Used ......... . . . 95 REFERENCES ............................. 96 LIST OF TABLES Elemental Analysis of Bound Metal Phthalocyanine-Polymer Complexes. ........... . ............. 18 Rate of Oxygen Uptake for .Phthalocyanine Catalyzed Cyclohexene Oxidations. . . . . . . . .............. 47 Oxidation Product conversions of Cyclohexene Catalyzed by bound Phthalocyanine Complexes. . ............ . . . . . 52 vi CHAPTER I INTRODUCTION A. Introduction to Catalysts Classically catalysts are grouped into two categories, distinguished from each other by the physical state of the catalytic compound and the substrate upon which it acts. Those which dissolve in a reaction medium in which a substrate also dissolves are termed homogeneous and those which remain undissolved in the solvent are called heterogeneous or "contact" catalysts. Each group has certain advantages and disadvantages when used in reaction mixtures. Homogeneous compounds tend to be more efficient and Operate at lower temperatures and pressures,1 and tend toward more selectivity to particular substrates. Heterogeneous catalysts are dependent upon mechanical factors such as how well the system exchanges fresh substrate and brings it into contact with the undissolved catalyst surface. The number and type of true heterogeneous catalysts available are limited to a small set of expensive metals, such as platinum and palladium. The contact catalysts show similar reactivity toward substrates and, therefore, display distinct selectivity limitations. However, the physical nature of the heterogeneous catalysts results in isolation of the compounds in a physical phase distinct from that of the substrate upon which it acts. Thus, it facilitates separation of reaction product from the catalytic reagent. The desire to produce a hybrid-phase catalyst, that is a complex which displays advantages of both heterogeneous and homogeneous systems, has led to the development of compounds where an insoluble support replaces one or more ligands on an active metal center}:2 Grubbs1 gives an extensive review of various types of supports and catalysts available and suggests a classification of hybrid-types based upon mode of ligand attachment to the metal centers. He also discusses matrix imposed selectivity changes which result from mechanical limitation of the insoluble support. Thus, in theory, judicious choice of support material, catalyst, and hybridization mode (attachment) should result in production of specific catalysts to use with specific substrates. 8. Introduction to Phthalocyanine Compounds Phthalocyanines are man-made macrocyclic compounds which are isoelectronic with the naturally occuring porphins (Figure 1). Early investigators recognized that metal phthalocyanines displayed the ability to slightly enhance oxidation processes,3’4 and for some time much research effort was put into their use. However, due in part to Phthalocyanine CH=CH | | CH: — C WC CH l N U , CH== C-—CH l >NH HN< H CH==C C—-CH | || CH=C /KC—CH l l CH=CH povphin or PMPhyrin (ll) Figure 1. Comparison of Structures of Phthalo— cyanine and Porphyrin the low reactivity which contemporary researchers attribute to the formation of oxygen bridged metal phthalocyanine dimers,5 these efforts waned in all but a few specific areas. Review of recent literature reveals that much of the current interest in the catalytic behavior of phthalocyanines involves their use as "sweetening" agents in petroleum. In that role they are used to remove mercaptans and other sulphur containing compounds by oxidation.5‘11 Other areas of current interest are vapor phase reactionslz’13 and biologic enzyme modeling by manganese phthalocyanine.14 C. Supported Phthalocyanine Compounds The expansion of work performed on the oxygen carrying ability of the metal porphine systems15 and their known "site" isolation in biological systems gives new impetus to study methods of support of the phthalocyanines.5 Without attachment to a non-dissolving support it is possible to use the phthalocyanines as homogeneous or heterogeneous catalysts by simple choice of reaction solvent. Likewise, aromatic substitution of the macrocycle will result in a change of its solubility characteristics. Such choice of reaction conditions do not, however, result in site isolation of the metal complex. Therefore, it is better to examine the procedure of supporting phthalocyanines for subsequent reactions not as simply "heterogenizing" them but rather as attempting to control the environment about the metal centers in order to increase reactivity toward a specific system. Such systems would require rigid covalent attachment to bind the phthalocyanine groups to an inert support at specific locations. In doing so, not only is heterogeneity preserved, but also it should then be possible to control environmental factors which result in dimer formation of the metal phthalocyanine complex5 (Figure 2). Zwart and co-workers describe such a system in their preparation of covalently attached cobalt phthalocyanine-polymer systems.5 Their method of attachment to the polymer involves chloromethylation of the polymer and then amination so that cyanuric chloride can be used as a coupling agent to aminated cobalt phthalocyanine (Figure 3). The degree of specificity which a particular attached catalyst system will show toward any given substrate should, in theory, be controlled to an even greater extent by the matrix environment which the substrate must penetrate to reach the catalytic metal centers. Ideally such systems would contain insoluble supports, which, by their physical nature, prevented or enhanced the ability of the substrate to pass through it to specifically located catalytic centers. 0f necessity such conditions could only be provided by custom synthesis of the support. In fact, in order to prepare systems in which the number and placement of active metal centers were exactly known, it would be /U'U/M 3 _M/U“U “M/ 0‘0 4 \\\\\\\\\\\\ Figure 2. Dioxygen Bridged Phthalocyanine Dimer and Dioxygen Monomeric Phthalocyanine on Polymer Support / \ + Cl--Cl‘12-O-CI-I3 + SnCl4——> / \ Hz-Cl ... 1) K-phthalimide, oMF.1oo° , A 2) NHZNHz, EtOH. reflux CH1CI CHzNH: 1 cyanuric chloride coupling of amines: (Aminated polymer + aminated phthalocyanine) Cyanuric 3 R,NH 0! Chloride + 3RNH’ (R or R,)H,N [4N] NH,(R or 12,) x N \/ NH,(R or R.) ”“2 H O“""””‘<}./" w. ~ l. | ‘ O H N—(t'o-N N“: N- N N ”t Figure 3. Cyanuric Chloride Coupling Method to Support Phthalocyanines on Polymers necessary to produce well defined polymers from monomers to which the catalyst had previously been attached. Such systems are known for the phthalocyanines, research being provided on similar systems for the manufacture of dyes. However, such intricate work with the nature of the polymerization process itself is beyond the scope of the study undertaken here, which has as its primary thrust the attachment of phthalocyanines on commerically obtained polymers and subsequent characterization of the resulting compounds. 0. Introduction to Specific Polymer Supported Phthalocyanine Research Carried Out The original intention of the research reported here was to compliment ongoing research performed in the past in Dr. Brubaker's group515:17’18 and Dr. Grubb's grouplg’20 by supporting a new class of] compounds on polystyrene support which would later be used as catalysts. Since it was desired to anchor the phthalocyanines solidly to the polystyrene matrix the methods of sulfonylation and sulfonamide formation were chosen so that a covalent linkage would form between the ligand and the polymer support. The former method results in a sulfone linkage which is a rather stable bond and both methods result in a linkage which has a shorter side chain than that of the cyanuric coupling method used by Zwart.5 The length of this side chain seemed critical since prevention of dimer formation was one of the goals of the research. In addition, since the phthalocyanine group was multiply substituted in the chlorosulfonation step, it was thought necessary to keep the side chain short so that the degree of additional cross linkage through multiple attachment of each phthalocyanine molecule was held at a minimum. After the initial experiments it was clear that the degree of "loading" of phthalocyanine onto polymer was low and that the surface of the beads seemed to react but the interior of the beads remained unreacted to sulfone or sulfonamide linkage. Therefore, many experiments followed in which reaction time, temperature, and solvent were adjusted in an attempt to increase the linkage formation. Surprisingly solvent was of little effect and time and temperature change the loading only very Slightly. By accident, in one experiment, a sample of beads was crushed by the stir bar and so, thus exposed to additional reactive surface, the sample was seen to take up much more phthalocyanine reactant. Therefore, it was reasoned that some blockage was occuring during the initial reaction of the bead at its surface with the phthalocyanine reactant. Concurrent experiments with oxidation reactions of the prepared phthalocyanine beads with organic substrates were disappointing because the reactions showed little or no increase in reaction rate above that of the unsupported phthalocyanine-organic substrate mixtures, and so apparently the site isolation of the metal centers was not being fully realized. Thus, because of the difficulties with synthesis and use of these 10 compounds, a detailed characterization of the beads was carried out using scanning electron microprobe analysis, electron spin resonance, transmittance electron microscopy, and differential scanning calorimetry. Results appear in Chapters III and IV. CHAPTER II EXPERIMENTAL A. Preparation of Chlorosulfonated Metal Phthalocyanines Following the chlorosulfonating procedure outlined by Moser and Thomas4 or Booth3 in their reviews of phthalocyanine compounds, the tetrasubstitution of the metal phthalocyanine compounds was carried out simply by dissolving the materials in chlorosulfonic acid and heating them to 125°C. The reaction products were then separated from the excess chlorosulfonic acid solvent by vacuum distillation and the substances thus recovered were used in subsequent polymer sulfonylation steps without further purification. Although the amounts of phthalocyanine and chlorosulfonic acid solvent varied throughout the research, a general procedure was developed as follows: 2.25 g of Nickel phthalocyanine (Eastman lot number A2E) or vanadyl phthalocyanine (Eastman lot number A7A) was placed in a 200 mL three-neck round bottom flask fitted with a water condenser, magnetic stirring bar, and thermometer. Thirty mL of chlorosulfonic acid was pipetted into the flask and the mixture heated to 125-13500 overnight. The flask was continuously flushed with nitrogen to maintain an inert 11 12 atmosphere. The excess chlorosulfonic acid was removed by vacuum distillation leaving behind a thick, highly colored oil of chlorosulfonic acid contaminated tetrasulfonated metal phthalocyanine complex. The ratio of metal phthalocyanine to chlorosulfonic acid varied for the experiments in which cobalt phthalocyanine and iron phthalocyanine were the complexes upon which substitution was carried out. Unfortunately due to their lack of adequate stability in the strong, acidic solvent, the resulting products proved at best to be marginally successful in the following sulfonylation steps. (See Chapter III). Finally it should be noted that the length of time which was allowed to complete the chlorosulfonation was in great excess of the two to three hours suggested by Moser and Thomas4. The reaction time was so chosen to insure that the subsituted metal phthalocyanine was indeed tetra substituted for, as Booth points out, 3 control of time and temperature results in selective substitution of the macrocycle at from one to four positions. Further, Booth makes the point that chlorosulfonation carried out by the preceding method results exclusively in 3-substitution. B. Preparation of Polystyrene-0V8 Copolymer Beads The 20% divinylbenzene-polystyrene cepolymer beads used as the solid support are commercially available and were gifts from the Dow Chemical Company. Prior to their use in sulfonylation reactions with 13 Chlorosulfonated phthalocyanines they were sieved through standard 28-32 mesh screens (0.0234-0.0197 in.). Then they were washed with the following solvents: 10% HCl, 10% NaOH solution, water, 1:1 water/methanol, methanol, 1:1 methanol/dichloromethane, dichloromethane. The beads were then dried under vacuum. C. Sulfonylation of Polystyrene-0V8 Copolymer Beads with Metal Phthalocyanines The choice of sulfonylation conditions resulted from suggestions by Olah in his review of Friedel-Crafts type reactions.21 Nitromethane was chosen as the Solvent for the reactions and aluminum chloride as the sulfonylation catalyst. The beads resulting from the first experiment varied so much in color and, thereby, their presumed loading of metal complex on polymer support, that an entire series of reactions were undertaken in which amounts of chlorsulfonated metal phthalocyanine, aluminum chloride, or polymer were varied. (See Chapter III). Finally a general procedure was developed for the two sulfonylation products which were later tested for catalytic activity. The following procedure was typical: The 200 mL round bottom flask that contained the Chlorosulfonated metal phthalocyanine (either nickel or vanadyl phthalocyanine) was refitted with a mechanical stirrer in addition to a condenser and thermometer. Ninety mL nitromethane and 2.25 9 aluminum chloride were added. The solution was stirred until most of the oily Chlorosulfonated metal phthalocyanine had dissolved. Then 15 g 20% 14 DVB-polystyrene c0polymer beads was added and the solution heated to 45-500C. The solution was stirred for 48 h under nitrogen. The nitromethane was removed by suction filtration and the beads were then washed with 1) nitromethane, 2) methanol, 3) water, 4) methanol, and 5) dichloromethane. The beads were then floated in a separatory funnel containing dichloromethane 4 to 8 times and thosebeads and other extraneous materials that sank were removed. The beads were allowed to dry in the air. The beads were transferred to a Soxhlet extraction apparatus and extracted with methanol for 70-72 h. The final extraction solvent was checked by visible spectrosc0py and revealed no trace of excess unreacted metal phthalocyanine. The beads were dried in a vacuum at 55-6700. 0. Preparation of Substituted Nickel and Vandayl Phthalocyanine Standards So that oxygen uptake and catalytic oxidation studies of the polymer attached metal phthalocyanine compounds could be compared to identical, non-supported metal complexes, tetrasubstituted nickel and vanadyl phthalocyanine-sodium sulfonate were made by the following prodedure: 0.2 9 metal phthalocyanine was placed in a round bottom flask fitted with condenser, magnetic stirrer, and thermometer. 2.66 mL of chlorosulfonic was added. The mixture was stirred under nitrogen at 115-1180C for 19 h. The solutions were hydrolysed by pipetting them over ice. The resulting water solution was neutralized with sodium 15 bicarbonate until production of carbon dioxide ceased. The solution was filtered out then placed in dialysis tubing (Union Carbide size 36 DM), which had been previously prepared by boiling it in water for 15-20 min and then rinsing it with water. The filled tubing was placed in a 2000 mL beaker filled with water and the water was stirred for 3 h before the dialysis tubing was changed and the process repeated for an additional 3 h. The water was changed twice during the next 11 h but the dialysis tubing was not changed. A small sample of the solution 2- remaining in the dialysis tubing was checked for 802' and CO with barium chloride. The solution contained in the dialysis tubing was then allowed to air dry in an open beaker to yield the final product. E. Preparation of Chlorosulfonated Nickel, Iron, Cobalt, and Manganese Phthalocyanines for Sulfonamide Reactions These reactions are nearly identical to those carried out for the standards above. Typically, the appropriate metal phthalocyanine (l.5 g) was treated with chlorosulfonic acid (20 mL) at elevated temperatures 120-13006 for 1 1/2 h. Then the solution was hydrolyzed- over ice and neutralized with Nazco3 (40 g). The resulting solution was filtered into dialysis tubing (Union Carbide 36 DM) and so purified overnight in about 6 L of distilled water. The solution remaining in the dialysis tubing was evaporated to dryness. Before use, the compounds so formed were treated with thionyl chloride at reflux for 4 h and then isolated by vacuum distillation. 16 F. Preparation of Aminated Polystyrene-Divinylbenzene Copolymer7 10 g of 20% (or 8% in other reactions) divinylbezene-polystyrene beads was treated with 38 mL acetic anhydride, 8.4 mL acetic and 3 mL nitric acid and the solution stirred under N2 for 5-14 h depending on amount of reaction desired. They were isolated and washed by stirring in five portions of acetic acid for 15 min each. Then 40 mL acetic acid was added with 12 mL of HCL and 11 g of SnCl2°2H20 for the reduction 0f “"02 groups to -NH2. The solution was stirred at 400C for 2 d, 22.5 h. The aminated beads were washed by stirring with a 3:10 mixture of HClzacetic acid (5 portions, 15 min each). Then they were washed with a mixture of 3:5:5 HClzTHF: Methanol (15 min each, 3 times) and then twice with methanol. The beads were dried at room temperature and then treated before use by washing two times with 10% methanolic KOH, then 5 times with methanol and twice with methylene chloride. Finally, the beads were vacuum dried at least 4 h. G. Preparation of Copolymer Attached Nickel, Iron, Cobalt, and Manganese Phthalocyanines by Sulfonamide Linkage 0.175 g of the appropriate Chlorosulfonated metal phthalocyanine was stirred in a chlorinated solvent (chloroform, methylene chloride) or nitromethane along with 0.5-1 9 of aminated beads. Some reactions were carried out at higher temperatures (80-900C) although there seemed to be little effect of temperature on the final product. In a few 17 cases triethylamine was added to the mixture and some effect was seen upon its addition (see results of SEM below). ’The beads were isolated and washed with, typically, chloroform, methylene chloride, methanol, water, and methylene chloride. They were floated in methylene chloride in a separatory funnel to remove any heavy unreacted particulate matter and then Soxhlet extracted with methanol before being vacuum dried. H. Equipment for Analysis and Physical Characterization Thermal analysis by differential scanning calorimetry was carried out on a DuPont Model 990 thermal analyzer. Electron spin resonance was done by use of a Varian 4502 X-band instrument equipped with a Micro-Now gauss meter and a Hewlett Packard X5328 frequency meter. The transmittance electron microscopy was performed with a Philips 300 electron microscope and the scanning electron microprobe analysis utilized an Allied Research Laboratory (ARL) EMX-SM spectrometer. I. Elemental Analysis The elemental analysis of supported phthalocyanines was carried out by the scanning electron microprobe as this method proved more precise than trace elemental analysis by commercial atomic absorption. The method compared the x-ray count rate of the metal in the compound of interest to the count rate of the pure element for the metal in question. Results appear in Table I. The beads were not uniform in loading (see section on scanning 18 TABLE I Elemental Analysis of Bound Metal Phthalocyanine Polymer Complexes Supported Polymer % Metal Measured 5%Metal avg. nmmle of Metal Complex at Surface of Bead Entire Bead for Entire Bead in Reaction per 0.l q beads Ni Sulfone _4 linkage 0.06 0.0ll l.87 x 10 V0 Sulfone Same as _5 linkage whole bead* 0.0027 5.30 x l0 Co Sulfonamide _4 linkage 0.l65 0.024 4.07 x l0 Fe Sulfonamide _4 linkage 0.l63 0.035 6.63 x l0 Ni Sulfonamide linkage 0.03 ** ----- Mn Sulfonamide ** ** linkage *Very low loading **Count rate a background level. (See Text) 19 electron microprobe below), therefore, it is necessary to average the count rate to determine the average percentage of metal in any particular sample. J. Oxygen Uptake of Various Organic Compounds Catalyzed by Metal Phthalocyanines An oxygen uptake manifold was constructed which provided a means of supplying oxygen to a vessel containing the catalyst and organic compound upon which the oxidations were to be carried out, Figure 4. As attending apparati, there were a mercury manometer, which actually was of little use in the experiments reported here, a mercury filled gas buret with leveling bulb, an oil bubbler device fitted with a marked dip tube with which it was possible to maintain manually the pressure inside the manifold at atmospheric pressure, and a gas drying tower through which oxygen was introduced into the system. A typical oxygen uptake experiment was conducted as follows: 0.1 g of supported metal catalyst was placed in the reaction vessel with a magnetic stirring bar and the vessel, connected to a water condenser, was attached to the oxygen manifold. An oil bath heated by a nichrome immersion coil and stirred by a mechanical stirrer was placed about the reaction flask so that a chosen temperature for the reaction could be maintained (Either 50-600C or more commonly 75-8500). 3 mL of organic solvent-reactant was injected into the reaction vessel which had previously been evacuated and refilled with oxygen through the 20 o Q-l mmfivsam oxapmb nowhxo now vaomflqmz umHaH smwmxo . Hm3ou wafihuw mmw . nonmaoama husouma amusn mmw «3| 0| «I ll [II lllll‘ib H NIH ML: \A Dr— (\J Hlllll Hmwamfifiou HQUG3 . SSDUG> . .5! mamsmm neaeese Hue ml .q onsmflm Q.D U'U Manifold for Oxygen Uptake Studies Figure 9. 21 manifold. The solution was stirred throughout the course of the oxygen uptake study which varied from approximately 6 h for the more effective catalysts at higher temperatures to 48 h for the less reactive and lower temperature studies. In some cases the catayst beads were isolated from the reaction mixture washed with fresh portions of solvent and quickly dried under vacuum. The reaction conditions were reestablished and a second run carried out using fresh organic reactant. (Chapter III). The tetrasubstituted metal phthalocyanine-sodium sulfonated standards were studied in a similar manner with the exception that the amount of catalyst used in the experiments was determined by computing the amount of catalyst present in the beads from elemental analysis (as weight of metal in grams), performed commerically, by Schwartkopf Microanalytical Laboratory; and taking a corresponding amount of non-supported catalyst. Since these typically were very small amounts, it was necessary to take up a given sample of non-supported catalyst in water, make a standard solution in a volumetric flask and pipet an appropriate amount into the reaction flask. The water was allowed to evaporate from the solution leaving the standard catalyst which was then used in the oxygen uptake studies. K. Catalytic Oxidation Studies Without Measurement of Oxygen Uptake 1. Reaction Conditions A less eleborate oxygen manifold was constructed that allowed 22 three reaction vessels to be used at once. It maintained internal oxygen pressure by allowing an excess flow of oxygen to pass continuously through its oil bubbler. Once again an oil bath-heater system was used to control reaction temperature. Reaction conditions were chosen so that 0.1 g of supported catalyst or its equivalent in standard non-supported catalyst was stirred in 3 mL of reactant-solvent which were the organic compounds (cyclohexene, 1-octene, 2-octene, toulene, cumene) upon which oxidation was attempted. The vessels were attached to water condensers, which were hooked to the oxygen supply manifold. The reactions were long-term, lasting for a typical time of 47 h. The reaction mixtures resulting from the procedure were separated from the catalyst; the catalyst, were washed in iSOpropanol and stored. 2. Analysis of Products The reaction products were analyzed by gas chromatography by using a Varian Aerograph Model 920 Chromatograph with thermal conductivity detector. An SE-30 column proved to be effective in separating the components of the mixture. It was 10 ft long by 1/4 in in diameter and contained 10% SE-30 supported on WHP. Column temperatures varied from 50-1000C depending on the particular use of the device and the nature of the reaction mixture. Amounts of each component were determined by the integration device on the recorder. In one case, that of cyclohexene, the nature of one component was confirmed by Michigan State University's Biochemistry Department Mass Spectrometry Facility by using a coupled gas Chromatograph/Mass spectrometer. CHAPTER III PHYSICAL CHARACTERIZATION OF POLYSTYRENE SUPPORTED METAL PHTHALOCYANIENES A. Differential Scanning Calorimetry Due to failure of the thermal analyzer only the first samples of the sulfonylation reactions were analyzed by DSC. The results appear in Figure 5. The figure shows that the bead has changed thermal characteristics after treatment with a metal complex. The low loading (very low due to sulfonylation attachment, see next chapter) of the cobalt species is only slightly different from the parent polystyrene but does show a shifting of its exothermic transitions. The difference for the nickel complex is dramatic. The transition between l00-l750c is greatly broadened from that of the parent. The transition corresponds to the decomposition of the polystyrene (Polystyrene does not melt cleanly even in the nitrogen atmosphere of the 050 cell). The most likely process responsible for such an observation is that additional crosslinkage between polymer chains is taking place with the availability of multiple sulfonylchloride groups on each of the metal 23 24 I’ }.Copc ]-NiPc Thermal Energy Release 1 1 1 1 1 50 IOO ISO 200 250 '°C Figure 5. Differential Scanning Calorimetry Traces of 20% Divinyl- Benzene Polystyrene, Cobalt and Nickel Phthalocyanine Bond 20% Divinylbenzene Polystyrene. 25 centered phthalocyanine molecules. In fact, a qualitative difference between highly reacted polystyrene beads and their unreacted counterparts was noted in the cutting process in sample preparation necessary for scanning electron micr0probe analysis. An increased surface hardness was noted in the reacted beads. This also is most likely due to a degree of cross linking by the metal complex during reaction. 8. Electron Spin Resonance ESR data were unsurprising. Figure 6 shows the spectra of the vanadyl complex made by the sulfonylation method and the manganese complex with the sulfonamide linkage. Both spectra are as expected. Vanadyl has a spin number, I = 7/2, and shows the expected 8 line spectum. Likewise the Mn(II) complex is 5/2 and displays 6 lines in its spectrum. Spectra for iron and nickel were not obtained indicating Fe(II) and Ni(II) were present. Cobalt displays somewhat unexpected behavior. Lack of an ESR signal suggests that the cobalt has possibly been oxidized to Co(III) from the expected Co(II) state during complex formation. However, cobalt spectra are difficult, at best, to obtain with supported species on the polystyrene matrix. The only other feature of note is a strong singlet at a Q value near free spin. This line is seen in all samples prepared by sulfonamide linkage and is also present in unreacted amino polystyrene beads made by the method outlined in the preceding chapter. Multiple attempts were made to rid the samples of the extra signal but in all 26 Figure 6. Electron Spin Resonance Spectra of 20% Divinylbenzene Polystyrene Bound Complexes of Manganese and Vanadyl Phthalocyanine. 27 9; = l. 972 A H g" = 1963 (First order) 1- .L A J. J. J. II INN 11 II II J. .1. Room Temp. Scan Range = 1000 G ]-VOPc Field Set = 3l95 G 8 g . 2.054 , Temp. = 77K ]-MnPc Scan Range = IOOO 6 Field Set = 3|OO Figure 6. 28 preparations the signal persisted apparently due to the same radical formation on the amino polystyrene during its preparation. C. Transmittance Electron Microscopy The samples were prepared by embedding the polymer beads in epoxy resin and taking ultra-thin sections. Figure 7 shows a photomicrograph of both 8% crosslinked divinylbenzene-polystyrene and 20% divinylbenzene-polystyrene at 40,000X magnification. The 20% sample contains cobalt phthalocyanine. In addition Figure 7 shows two other photographs at the edge of a phthalocyanine reacted bead at 32,000X and 52,000X magnification. From the micrograph the size of the open structure of the bead can be seen, hence, because samples were all 20% DVB-polystyrene c0polymer in the catalytic experiments, it is interesting to note that in these cases the channel size is approximately 500 times the diameter of a phthalocyanine molecule. However, as is seen in the reacted sample in Figure 7c the beads surface appears practically occluded at the surface. Unfortunately, the fine resolution of the electron microscope would not allow the certain determination that high surface loading would preclude entrance of other phthalocyanine reagent into the interior of the bead. The TEM evidence coupled with the scanning electron microprobe data (see below) indicates that the phthalocyanines attach themselves at available sites at the polymer surface and that such attachment then interferes with penetration of the remainder of the reactant to the interior of the bead. Figure 7. 29 Electron Photomicrographs of A) 8% Divinylbenzene polystyrene at 3200K, Cobalt Phthalocyanine Attached to 20% Divinylbenzene polystyrene at B) 40000X, C) 32000X, and 52000X. 30 D. Scanning_Electron Microprobe Analysis In scanning electron micrOprobe analysis a narrow beam (0.3 micron diameter) of high energy electrons (10-25 kV) is swept across a .cross-section of a sample bead. X-rays characteristic of the elements present in the sample are produced from a 1 to 10 micron diameter volume. The instrument provides a detector which is usually set to receive the K“,line of the element in question. Count rate as a function of beam sweep is plotted-typical graphs appear in Figure 8 and the Appendix. Therefore, the graph represents amount of a particular element at any particular position along the sweep of the electron beam. The samples were prepared by slicing the beads with a razor blade while attempting to section them as close to the center of the bead as ' possible. The bead sections were mounted on a carbon target with two sided rubber tape and the target and sample then were coated with carbon in a high vacuum evaporization apparatus. The coating is necessary to insure adequate conductance of the sample. In all cases the results are similar. A high degree of loading of the phthalocyanines is achieved at the bead surface but a sharp and dramatic dr0p off occurs in loading toward the center of the beads. Even in experiments with extended reaction time and in the presence of an amine scavenger (triethylamine) to pick up the HCl produced during the coupling reaction no large improvement is seen (Figure 8b). To be sure, the excess time of treatment and higher temperature at which the reaction was run (27 h and reflux temperature of l,2 dichloromethane) 31 Figure 8. Scanning Electron Microprobe Analysis of Phthalocyanine Reacted 20% Divinylbenzene Polystyrene (A & 8). Comparison SEM of Cobalt Porphyrin Attached 20% Divinylbenzene Polystyrene Analysis (c). 32 }-CoPc Condensed Scale B ]-CoPc (Extended Time, Higher Temp. With Triethylamine) Full Scale Chart ]-CoPor (King) Condensed Scale 2500 Counts/sec. JWMW Figure 80 33 and the triethylamine results in some penetration to some interior sites but the distribution is far from uniform. There appear to be pockets in the interior of higher loading followed by rapid drop off of metal centers much as scanning across the surface boundry and then into the interior of a sample appears. Indeed, these pockets of high concentration of phthalocyanine approach the surface value of metal complex loading but sites surrounding the pockets are unreacted. For comparison a sample was obtained of cobalt porphyrin attached to 20% divinylbenzene and a comparison SEM analysis carried out. This sample also displays the typical loading of the surface of the polymer bead but the interior of the bead remains unreacted. (See next chapter) The SEM method was also used for elemental analysis since count rates at any particular point or an average count rate for a cross section can easily be compared to the count rate of the pure-metal, thus, a facile analysis results. This method proved consistantly more precise than trace elemental analysis by commerical atomic absorption. Results appear in Table I, p.18. CHAPTER IV RESULTS AND DISCUSSION A. Result of Treatment of Metal Phthalocyanines with Chlorosulfonic Acid In all, five different metal phthalocyanines were purchased along with metal free phthalocyanine for use in chlorosulfonic acid treatment steps so that -SOQ-Cl groups would be attached to the aromatic, macrocyclic ring. As expected, only those compounds where the metal is tightly bound (Ni, V0) were able to withstand the highly acidic reaction conditions over a long period in which chlorosulfonic acid acted as reagent and solvent. Nickel phthalocyanine showed the greatest stability while vanadyl phthalocyanine showed marked resistance to attack by sulfuric acid in experiments designed to test their stability toward acid, Figure 9. However, iron phthalocyanine breaks down completely over a long reaction time in such a manner as to allow the acid to attack and degrade the macrocyclic ligand itself. Likewise, the macrocycle of metal free phthalocyanine visibly breaks down in acid as noted by the loss of its characteristically dark color. Experiments to show degradation as a function of time of the Co and Fe complex are displayed in Figure 10. For these compounds which are less 34 35 NiPc \ 20min >2'/1hr 23'/ahr l .J I l 50° 600 700 300 nm VOPc _ 6min lhr 23%hf $00 600 700 800 nm Figure 9. Visible Spectra of Nickel and Vanadyl Phthalo— cyanine in Concentrated Sulfuric Acid 36 stable in acid it is critical to adjust the reaction time so that enough of the macrocylic complex remains for subsequent reaction. That is, a competing reaction exists between chlorosulfonyl substitution and acid degradation of the complexes in question. There are many factors that influence stability of particular phthalocyanine compounds4, among which are covalent character, tendency toward square planner coordination, and ability of a particular metal to undergo oxidation state changes in the macrocyclic environment. Such factors coupled with the necessity of the highly acidic environment in the first step of both the sulfonylation and sulfonamide reaction sequence to produce the polymer supported metal phthalocyanines will, of course, provide a limiting condition which narrows the yield of the entire process. These phthalocyanines complexes with a marked degree of stability (Ni and V0) are favored over those which must be more carefully prepared (Co and Fe) and those which seem to produce no attached complex under controlled conditions (Mn and especially metal free phthalocyanine). 8. Results of Sulfopylation of Polystyrene - DVB Copolymer Beads with Metal Phthalocyanines As mentioned in Chapter II, a sequence of reactions was undertaken at the onset of research into these supported phthalocyanine complexes to determine which reaction conditions would maximize the amount of metal complex loading. In addition, since the initial reaction to support nickel phthalocyanine on beads resulted in a product that contained beads of various hues, reaction conditions were sought that 37 CoPc 10min 3hr 49%!" A l 500 l l l 600 700 .00 nm FePc 14min 3hr \——-I9’/.hr J l l L 500 600 700 300 nm Figure 10.. Visible Spectra of Iron and Cobalt Phthalocyanine in Concentrated Sulfuric Acid 38 would afford uniformly loaded beads. The conditions adjusted in various reactions included time, temperature, amount of aluminum chloride catalyst, and sequence of addition of beads and catalyst to the reaction mixture; each will be considered in turn. Time seems to determine the degree of loading up to a maximum, beyond which additional time does not effect total reaction. Therefore to insure that the time factor is overriden one need only to insure that sufficient time is allowed so that no further reaction takes place. Conversly if one wanted to lower the degree of loading the reaction time during the sulfonylation step could be shortened. Temperature seems to effect both loading and homogeneity of color. Experiments were undertaken in the sulfonylation step at room temperature up to the reflux temperature of the solvent, nitromethane (lOlOC). Again it was found that loading could be increased to approximately 50-600C. Above that temperature range it was apparent that loadings actually were lower. Such conditions might possibly be due to competition between the rates involved of the sulfonylation reaction at higher temperatures and exchange of reactant into and out of the pores of the beads. Indeed, such factors as exchange rate difference due to porosity difference from bead to bead would seem to account for the mottled apparance and, hence, the variable loading of the sulfonylation products. Another factor considered below is the denaturing of the beads by the aluminum chloride sulfonylation catalyst. Such effects are highly magnified at elevated temperatures. The amount of aluminum chloride used as a sulfonylation catalyst 39 has two effects. The catalyst was always used in large excess of stoichiometrically required amounts, thus limitations by ordinary chemical factors were not considered, but in extremely large excess the beads seemed to denature and a brownish colored product was recovered from the reaction mixture. The other effect was mechanical. If too much aluminum chloride was added the resulting complex formed between it and the Chlorosulfonated metal phthalocyanines turned into a thick, semi-solid mass which prevented stirring of the mixture. Additional solvent was required to relieve the situation and allow resumption of uniform stirring. The sequence by which reactants were added to the reaction mixture seemed to have very little effect on the eventual outcome of the overall reaction. At first such considerations were made in an attempt to overcome the multi-shaded bead problem. Since the first sulfonylation did not allow sufficient time for the aluminum chloride to form a soluble complex with the Chlorosulfonated metal aphthalocyanine before the beads were added, initially it was thought that such a non-constant supply of reactant might be responsible for the non-uniform reaction on the beads. Subsequent experiments, however, showed very little difference in the final product from one experiment to another when the amount of time allowed for aluminum chloride-phthalocyanine intermediate complex formation was varied before addition of the polymer beads to the reaction. Other factors that possibly might be considered to effect the degree of loading from bead to bead, such as stirring rate of the solution and mechanical trapping of the beads by precipitated aluminum 4O chloride-phthalocyanine complex, likewise, had a minimal effect on the outcome of the reaction sequence. Therefore, it seems plausible that the degree of difference in loading from bead to bead depends upon the ability of the pores of an individual bead to carry reactants into its available sites for reaction at any particular temperature. C. Results of Sulfonamide Linkage Formation Between Polystyrene-0V8 Copolymer Beads and Metal Phthalocyanines This series of reactions was undertaken because of the problems encountered with sulfonylation reactions reported above. Additional steps were taken to purify the sulfonated phthalocyanines after the first step of the reaction and to neutralize any excess chlorosulfonic acid. That in addition to a shortened time in contact with chlorosulfonic acid during initial reaction allowed phthalocyanines of iron, cobalt, and manganese to be successfully used with polymer supports. It was not possible to prepare a metal free complex. Also experiments were undertaken to check the effect of solvent on reaction outcome during the subsequent sulfonamide linkage formation. Again degree of loading and uniformity of loading was checked. However, the sulfonamide formation reaction presented two additional problems which could not be solved. The first was caused by the necessity to make amino substituted polymer beads. The procedure used nitric acid to nitrate the polymer and then the nitro groups were reduced with tin chloride to the amine. This step results in a formation of an apparent radical as is shown in 41 the ESR spectrum of the beads. These radicals might have been formed by incomplete reduction of the nitro groups to the amine as it is known that not all of the groups are reduced by the procedure.22 The second problem that arose during the procedure was a complete surprise. It had been known that the reactions with polymer supported phthalocyanines5 and porphyrins22 acting as catalysts seemed diffusion controlled. That is the reactions increased in activity when the beads were ground up. In fact just such a circumstance is shown with the reactions reported in the next section. However, since some members of this laboratory had investigated the interior loading of their beads by using scanning electron microprobe procedures, these compounds were also prepared for the procedure. Upon cutting the beads across its diameter it was clearly evident that the loading was not uniform. They showed a distinct coloration at the surface but virtually no color at the interior positions. Many experiments which involved changes of solvent, time and temperature showed little improvement in the situation. Since this loading problem created difficulty in interpretation of increase of activity with grinding the beads as pure diffusion control of the reaction, a sample of King's reaction product was obtained.23 That sample also showed a marked difference upon sweeping the electron beam across its crossection in SEM analysis. The surface was 148 times more concentrated than the interior (Figure 8c). Thus, it is reasoned that the primary reaction takes place as the groups of reactants come into contact as Chlorosulfonated metal phthalocyanines attach to active groups of amin0polystyrene on the polymer bead surface. Once attached these groups cause crosslinkage as 42 shown by DSC results in the previous chapter. In addition steric and electronic effects likely prevail and the result is a virtually closed surface which prevents any further reaction penetration. The result is for the surface to achieve a high degree of loading while the interior, even the near surface, is essentially unreacted. Experiments to block the surface or to increase penetration by mechanical effects were unsuccessful in greatly changing the overall effect. In some cases a "pocket effect" results, that is, an area of high phthalocyanine loading is surrounded by unreacted surface. One might speculate that these pockets are interstitial sites at channel openings at the surface or interior which become lined with phthalocyanine. In the extreme such loading could approach pure phthalocyanine content as, Dr. King's beads do at their surface. His relatively higher porphyrin to bead ratio at the surface compared to the phthalocyaine to bead ratio at the surface or at the "pockets" might well be attributed to the difference in internal structure shown between his bead and those reported here as shown by their electron micrograph. His are more highly packed in internal structure. If, in fact, loadings approach values of pure phthalocyanine or porphyrin crystals it is necessary to conclude that site isolation does not occur and that catalytic behavior of the supported complexes cannot be expected to be much higher than that of unreacted phthalocyanine. Further, increase of reaction toward oxidations after bead grinding, classically attributed to diffusion control, is most likely just a mechanical effect whereby more active centers are exposed to substrate, much as grinding crystals of unsupported material would increase 43 activity of the compound. D. Result of Oxygen Uptake Studies The oxygen supply manifold showed some difficulty in use especially in those reactions which resulted in very low uptake of oxygen. It became apparent that the air flow about the manifold could affect the readings of gas uptake by as much as plus or minus 5 mL. One had to be certain that the laboratory remained Opened or closed to insure a constant degree of room air exchange throughout the time readings were taken. The mercury manometer proved to be inconvenient to use in measuring the amount of oxygen loss to the chemical system. That task was greatly simplified by maintenance of ambient pressure by adjusting the leveling bulb of the gas buret such that a constant pressure was held in the gas manifold as shown by the oil level in the oil bubbler overpressure device. The chemical systems tested also exhibited certain patterns. For instance, those which were only slowly oxidized did not take up oxygen quickly enough for the data to show any variation other than random pressure increases or decreases ascribable to external factors such as changes of temperature and air pressure. For those systems in which significant oxidation took place there was a usual incubation period of about two hours during which the rate of oxygen uptake increased to a constant value. Such behavior can be attributed to the necessity of the solvent/reactant to establish equilibrum with the active catalytic centers and, hence, the time lag necessary for the metal centers to 44 form an available dioxygen complex intermediate. In some reactions the beads were recovered, washed and dried then immediately reused in a second run. The results were displayed as time vs uptake of oxygen for nickel phthalocyanine complex, Figure 11, and iron phthalocyanine complex, Figure 12. These are typical of the reactive systems. Other graphs are found in Appendix B. Each of these systems, the nickel phthalocyanine complex made by sulfonylate linkage and the iron phthalocyanine complex made by sulfonamide linkage, were used as oxidation catalysts toward cyclohexene which was acting as both substrate and solvent for the systems. These data along with those for the other complexes tested (graphs in Appendix B) are summarized in Table II which also includes data for the oxygen uptake of the standard, non-attached, phthalocyanines. Comparison of rates of uptake for attached phthalocyanine vs unattached compounds can be found in the following section, however, a few comments and conclusions can be drawn from the inspection of the data for attached phthalocyanines. First, note that in all cases the incubation time is approximately two hours, followed by a period of uptake of oxygen which is nearly linear. The period lasts for two to two and one-half hours in which the amount of oxygen uptake increases. Such behavior can be attributed to two competing factors. One is the necessity to form the metal dioxygen intermediate before further reaction with substrate. The second is, of course, the actual oxidation of the organic substrate. Both reactions utilize available oxygen with a corresponding lowering of the gas presure in the supply manifold. Non linearity can be caused by lack of equilibrium between the formation of the dioxygen 45 Reaction Time m .1360 " 330 c: ._.300 -—-270 o —-240 "210 C) r‘l80 - 150 c*120 1 l l l l 50 4O 3O 20 10 Uptake of 0 mL 90 2 Figure 11. Polymer Supported NiPc (new) In Oxidation of Cyclohexene at 780 C .oomb pm maoxmfioaomo mo aoaamvflno 2H A305 omom empaommsm noezaom .Ne mgmfim 5?. om ON 9. )11 .GE mm; 009 mm. Omii - . i . ~ _ d q — A O .. 3 ..NM ...» «xii I. o_. a 0.. b .. 6 ll ON 1cm 4 \ .. cm No as no HE 47 TABLE II. Rate of Oxygen Uptake for ghthalocyanine Catalyzed Cyclohexene Oxidations (78 C) (20% DVB- -Polystyrene Beads) (0.1 g Beads, 3mL Cyclohexene) Table II 48 Catalyst Measured Rate (mL OZ/m) (mL OZ/m/mmole-Metal) Supported NiPc (Sulfone) (New) Supported VOPc (Sulfone) (New) Supported NiPc (Sulfone) (Used) Supported VOPc (Sulfone) (Used) Supported NiPc (Sulfone) (Ground) Supported CoPc (Sulfonamide) (New) Supported FePc (Sulfonamide) (New) Supported MnPc (Sulfonamide) (New) *NiPc (SO3Na)4 Standard *NiPc (SO3Na)4 Standard *NiPc (SO3Na)4 Standard 0.15 0.42 (day 1) 0.26 (day 2) 0.l6 (day 3) l8l4.7 7924.l 1387.7 6980.7 l974.8 712.l 377.3 2900.97 l795.8 ll05.l * . The same reaction monitored over parts of 3 days. 49 intermediate-substrate intermediate and its subsequent breakup to the final oxidation products. Comparison of the "new" polymer supported catalysts, those which had not previously been used in a reaction; with beads which were termed "used" because they had been recovered from one reaction mixture and used again, shows that in each case a small amount of reduction in overall oxygen uptake rate occurs. Such a reduction is usually considered to be caused by poisoning of the catalytic centers. Neither used nickel beads nor used vanadyl beads display deviation from linearity over the portion of the reaction monitored (even though the reaction was allowed to incubate overnight) argues that substantial poisoning of catalytic activity does not take place. Such changes in overall rate might well be assigned to shifts in equilibrium in the reaction mixture due to the buildup or breakdown of the substrate and its oxidation products. Finally, the comparison of values in Table II show that the reaction rates of ground nickel complex beads is slightly higher than the oxygen uptake rate for the whole beads of the same compound. As detailed earlier, the apparent increase in rate for the ground beads is most likely due to the mechanical separation and, hence, availability of a greater amount of active centers rather than strict diffusion control of the reaction. 50 E. Result of Oxidation Studies Measured as Total Percentage of Substrate Conversion As early as 1939 Paquot24 had documented the oxidation of cyclohexene with phthalocyanines acting as catalysts by measuring total conversion of substrate into oxidation products. He reported yields of 3-cyclohexen-1—one, 3-cyclohexene-l-ol, l,3-cyclohexadiene, epoxycyclohexane, and l,1-cyclohexanediol. Of course, at that time he did not have instrumental means of analysis and relied upon chemical means for separation and identification of the products. With the exception of the l,3-cyclohexadiene, gas chromatographic data shows that, indeed, these are the products formed of the reactions reported here. The 10% SE-3O column used for the separations was standardized against authentic samples of each of the possible oxidation products and showed good resolution of all but the diene which proved to be nearly coincident in retention time to the parent cyclohexene. However, by adjusting the chromatograph's flow rate and temperature it was found that the two substances would be separated. Some loss of peak resolution did, however, result. Therefore, while the presence of the diene is not ruled out, the shape of the parent peak in a chromatogram run at normal temperature (100°C) and the absence of anomalies in the chromatogram when standards are run at lower temperatures indicate that very little or no diene is present in the reaction mixtures under question here. The gas chromatOgraphic data for the various polymer supported 51 complexes are given in Table III. All reactions except for the one for the sulfonated standard nickel phthalocyanine were run on a simplified oxygen manifold described in a previous chapter. All reactions were 47 h in duration. A number of trends can be seen in the data. First the systems show a dependence on reaction temperature. Nickel and vanadyl complexes display little oxidative activity toward cyclohexene below 78°C. That temperature was chosen because the boiling point of cyclohexene is very close at 83°C. The dramatic increase in overall conversion of cyclohexene into oxidation products at higher temperatures is most likely due to the activation energy, and possibly increased penetration of the reactant/solvent into a few layers of the polymer bead matrix. Also of note is that, although little product is actually formed, there seems to be a preference for the formation of l, 2-cyclohexandiol as is seen by the small concentration of that particular product in reactions run at 53°C. At higher temperatures other reaction products begin to form. In 1939 Paquot attributed their formation to two modes of oxidative action. One was direct metal-oxygen attack on the double bond of cyclohexene; attack which results in epoxyhexane and the l,2 cyclohexandiol. The other mode of oxidation was attributed to attack of the double bond which was claimed to cause formation of 3-cyclohexen-l-one and the 3-cyclohexene-1-o1. Although the actual mode of oxygen attachment to metal complexes in large macrocyclic ligand systems is still a matter of much controversy,25 current work, which has primarily centered around biological dioxygenases, has suggested possible modes of reaction in 52 .moxoamaoo oqficmmooawnpzm venom an nmumampwo oquozoaomo no m20fimam>qoo nonvoam coapmwfixo .HHH manme m.o w.o --- o.m a. ‘ Powflrcch mtaeeo -o_osa-m._ m.mm c.m x.mm <.om some» It-ll--tll-nl.li.10llllii p< ‘2 8:01_ -mcmxm: -o_u>uim 9.0” M m.©m xx. w.~ m.m memm ms m.© m.o m.om mm -1- -1- m.No mm m.x_ m.o w.Nq xx -1- -1- «.mo mm 111 11- oo_ mm om m.m m.m_ ma om.x_ m.o q.qq ax ounce moat» _.qm mm .1. 1.. oc_ ow H_ & _XV_%L> —01— mcwxwz 1:n:V:: u: -mcoxm; 1o_oxo acmxmc wcsum -o_uxuim izxoam iofiuxu -cmaEmp Acowuummc m_oEE Acowuummc mfioea Acowuucmc m_OEE Aco_gummc mFOEE Am~95= Acowuummc mPOEE Acowuommc w_oEE Acowuummc w—OEE Ampoee VI Aee2momv o_ x me._v oaez ncmucmpm mcmxm; io_o>u xcch meeam eee_a cu_3 xchm Accsocmv om.mv eao> umucoaazm om.mv eao> naucoaazm om.av Sac) kuL0qqzm om.mv oao> nwucoaasm Avcaocmv Rm._v eaez vmgcoaqzm em._v ea_z cmucoqqzm Kw._v aaez omucoaasm aw._v ea_z vmucoaasm .FP .o_ .— Amvemn mcmcxum»_oq m>o fiom are __ ow_o>o .ma negate ”am e, m.50> .mH acumen ..mm tom mmE¢ opowz 69 by itself were tested as possible reducing agents toward titanocene dichloride. In the cases of hydrogen and aluminum attempts were made to treat the reducing mixtures with HCl gas to preclude the formation of dimerized product by loss of chloride from the complexes. All reductions were carried out in THF prepared as above to give approximately 0.05M solutions of titanocene dichloride. In the cases of hydrogen and aluminum with HCl gas no reaction was evident. With zinc there appeared to be a near immediate reaction when the solution was shaken under inert atmosphere with zinc dust (total reaction time less than 15 min.). When filtered the solution produced an ESR signal which persisted for several hours. If shaken for longer periods the red-orange solution produced from immediate reaction with zinc turned more greenish in color. While the solution treated with zinc for these longer periods had an ESR signal it was somewhat diminished in intensity to those signals of the solution treated with zinc for a shorter period of time. Both aluminum powder and aluminum foil were used as reducing agents; however, the aluminum foil was first "activated" after the method of Coutts, Wailes, and Martin.32 Small squares (about l cmz) were treated with hot deoxygenated mercurous nitrate solution in water. After 30-60 seconds the water was poured off under argon and THF was introduced to the flask several times to wash the foil. The aluminum so produced was used in experiments with titanocene dichloride straight, with addition of HCl, and with addition of tetrabutylammonium chloride. All reductions were carried out under inert atmospheres. 4. Mixtures of Titanocene Dichloride and Reduced Titanocene 70 Dichloride Complexes These mixtures were prepared in the dry box in volumetric flasks by using reduced compounds with displayed ESR and the "parent" solutions of titanocene dichloride which had been diluted from the approximately 0.05M concentration to various values down to 10‘4M.' Colors varied from a dark red to a very light pink before addition of reduced components. While the amount of reduced complex varied from one sequence of experiments to another, its concentration within a specific reaction run to measure ESR linewidth was kept constant so that the effect on the spectra would be due to concentration of the parent compound alone. 5. Electron Spin Resonance Spectroscopy and Analysis of Reduced Products All ESR spectra were taken on a Varian Model E-4 Spectrometer. Analyses of the reduced products were made by mass spectrometry and commerical elemental analysis. CHAPTER II RESULTS AND DISCUSSION 1. Electrochemical Although the solutions changed color to give a light yellow green solution which would suggest that the titanocene dichloride was reacting, the solutions so formed displayed no ESR signal and had a cyclic voltammagram different from those literature reported analyses for [szTiClZJ' 31 In addition the yellowish solutions were stable in air and did not go back to their original red color after exposure to oxygen containing atmosphere. These condiions suggest that -OH or -0- groups have replaced chloride in the titanocene dichloride and, thus, water vapor or oxygen is incorporated in the electrochemical cell as the reducing process is attempted. It is important to note that, while in the specific cell used a positive pressure of argon blanketed the solution, it was open to the atmosphere at the top. Murr31 claims his cells were "high vacuum" design which resulted in consistent voltammagrams. In addition the presence of the electric potential might have hastened the abstraction of oxygen from the solvent THF by the titanium complexes in solution. 2. Zinc Reduction Reactions 71 72 The use of zinc readily produced a complex which contains titanium (III) and has a strong ESR signal. Published reports attribute the signal to a titanocene monochloride but in the presence of a bimetallic complex of titanocene dichloride and zinc, (szTiCl2)22n formed.32 These solutions remain stable for periods of hours if care is taken to exclude air from the system. It is reported32 that the bimetallic complex limits the amount of reduced titanium (III) produced by chemical competition. ESR data are reproduced in Figure 16 at both room temperature and 77K. These spectra bear strong resemblence to those reported very recently for a reduced titanocene chloride complex produced by gamma radiation which is attributed to reduced titanocene dichloride anion33 but the author does not completely rule out the possible presence of titanocene monochloride. Use of the reduced complex to make solutions with parent titanocene dichloride and the ESR spectra were very disappointing because the introduction of even very low concentrations of the parent resulted in complete loss of signal. The causes of the loss of the signal include the probable very low concentration of reduced complex owing to the formation of the bimetallic complex or a chemical transfer reaction between parent and bimetallic complex or between reduced complex and parent but on a very fast time scale. The latter cause seems unlikely in view of the results with the aluminum reduction (see below) which results in apparent electron transfer between parent and reduced titanium (III) complexes which are of similar nature to those proposed (titanoxene monochloride, or reduced titanium dichloride anion) for reduction with zinc. The presence of a complex of a higher molecular 73 Figure 16. ESR of Zinc reduced solutions of Titanocene Dichloride in THF (0.05M). 74 Ti(III) Containing Complex Room Temperature g= 1.9783 100 gauss Ti(III) Containing Complex 77K gX = 2.0024 gy = 1.9859 92 = 1.9510 Figure 16. ‘ 75 weight than the parent titanocene dichloride is confirmed by mass spectrometry, Figure 17. 3. Reductions with Aluminum The use of untreated aluminum powder proved ineffective in producing a reduced complex. Likewise, concurrent use of HCl gas or tetrabutylammonium chloride in reduction reactions appeared to prevent the reaction from taking place. In the case of HCl it is clear that the gas competes with the titanocene dichloride in reaction with the aluminum metal. Tetrabutylammonium chloride is hygroscopic and difficult to dry prior to use in reactions. Since the reduced complexes and, indeed, the parent titanocene dichloride are hydrolyzed by water it is probable that hydrolysis reaction limits the formation of d products. When the mercurous nitrate activated aluminum foil was stirred in the THF solution of titanocene dichloride for several hours under argon the red solution turned brownish green. After filtering in the dry box the solution produced a strong ESR signal and it was this solution which was then added to various concentrations of parent titanocene dichloride and produced the line broadening of the ESR signals, Figure 18. In addition, the mass spectrum of the solution shows that no dimerization has occured because no mass peaks appear above the parent complex at 249 mass units, Figure 19. Unfortunately the mass spectrum of the aluminum reduced solution can not be used to unambigously determine the nature of the reduced species because, if’ [CpZTiClZJ' is present, it could produce the peak at approximately 248-249 units or some of the unreduced titanocene dichloride still in solution could also produce the same 76 mme mcho—sopo wcmoocmpwh twosomm uch mo Escpomam 2 .AH magnet 77 III. mu? new QMm wmm mam @mm 0am amt awn awn u : L-Lv.hc_rPWPr>h—bbpp PLEPtP-.~PFF-—P>»-—>thh|PL1PPb1PhP1P1l—IP>>P_b1-rrbhbb-whit?-—P>herL|rL+lh>.bun—T—hnnr1br A 4 I. n ..r L 1 - I '9’1 1 (I1 13‘; < 1‘13 .31- ._ 1 1 r.“ ‘T‘"““"‘_‘T‘"“’""’”‘“ ”T" “’"T‘"“‘ Tufié—f [\ H. Q) S... 3 C? '1— L1. a maul 1 . a.mex OHM mom mmfi mm“ avg “NH mm“ am an av m r . . . —i—bkPh——wub1W‘F-F-IbthPhFLlPtr—bhuh—‘pkaFF—P‘Phh FPHde-"lubF1_P-P‘11—rltlFfl—"fl-mlwuu—J—F—_—hdfl—VI—F"what—huh—mwflw—JCJH——mvrhflld—VAFIPJ—F—~”~w-m“P“mF—MWAIWMHN‘P—dFH-FEWw.—hWJP.—““.—1.P‘WMFFd __ E _. ...: E = E _ l 4 .- L. : .32.. m m: ...... if? _ ... r1 . an .W _ 11 h;_; _ a ;l.” 1 i n41 _ rt __ _“ n __.N _;.:n ._ -.r f. _ _ mm a aw . _fl:w L nun _ _ waq _ _ _ w. 1 _u" - l I _ . ivy _ M _. r . .r _ _ m” r _ .m mm . .- a . _ w w, H in _ it” _ m -. m "_u a w . _" new m m m u w 1.. _ _ .. m m m. ..fl M _ , w .- n 1 H mm“ _ _ .H w ._ fl M. a . .1 e _ ll .V . n wmfi . . m a m w i m _ - 1 - . .1. _ .r. DIEM.“ Nari—fl. r mmmmm mrm .mwwvm HUME m: Jag "Hmmu . m.nu + oonmcuifl mm mm. mm "wax mmqm am: “mum "chic zzupumam mm.z 78 .muvLOFLUwo acmuocmpwp couscous: mo mcowpecpcmocou mewxce> new; mUTLOFLQVQ wcmuoceuwp emuzuwm EzcwE:F< to m—mcmwm mmm .wH at:m_u 79 1 mmzmm N . x . n . NioH m N v muoH x m n .m X H . qiofi m N muoH x m u .H sz muecopcowo memoocmuwh pcmcma .mH mczm_m 80 onwcopcowo mcwoocmpwh emuzvmm Ezcw53_< to Escyomam mmmz .mH messed 8]. ill. .......1 «H . .q_~ a ...___ 1. . r .44.: _ . ....m : an. e.— C LIT, LL.) 1:1 1 1N If": . w: 1 T m. . on: are 1 mm. amammu men: mnmm "maaeam .... . _ . "luau mmuo + oeuamuma mm we.ae . “at: mmam mus mnmm "seam azaeamam mm:: 1' '1 v1: '1“ H 1:1 '1' .3 ' 0—4 CE I") t .J <1 I.) 82 peak. Likewise the peak at 213 can be attributed to a reduced titanocene monochloride or simply to a chlorine atom cleavage from titanocene dichloride or reduced titanocene dichloride. The rate of electron transfer is calculated in the usual mannerzg'30 by utilizing the equation, k = 1.52 x 1011411 0 whereAH is the change of ESR linewidth in gauss and C is the concentration of the parent titanocene dichloride in molarity terms. The constant 1.52 x 107 is derived from the Bohr magneton, Plank's constant, and a standard 9 value compared to the formula for ESR relaxation times. Since H/C is the slope of the line from Figure 20, which equals -15.03 gauss/M, then, k = 2.3 x 108 M"1 sec‘l. 83 Molarity of titanocene dichloride 0.00025 . 0.0025 P 0.025 - 1 1 L_, 0.5 1.0 1.5 Gauss Linewidth Increase Figure 20. Variation of Linewidth of ESR Signals of Aluminum Reduced Titanocene Dichloride Treated with Unreacted Titanocene Dichloride 84 SUGGESTIONS The exact composition of the Titanium (III) containing complex should be determined. This would require isolation of the complex from the reduction mixture of titanocene dichloride, aluminum, and aluminum chloride without further reaction or decomposition of the reduced compound. In addition, substituted biscyclopentadienyl complexes of titanium could be synthesized, reduced, and purified, used in line broadening experiments. Comparions of data obtained by these experiments would result in determination of many thermodynamic values for electron transfer reactions of these complexes by methods previously used an analogous systems.29-30 85 APPENDIX A 86 Ammwxcwa muflewqomasmv momma 7mm xoaonmumzaoarm>a mom :0 copuommsn oaaawmooamnnnm aonH 2.2.1:: 5‘. 2f: h ‘r A. .. $.§.& _&. u a fimh ....mm a“... . .. .w.4 87 AomquHH oufiemcomasmv mocha 2mm .mdohhpmmaogrm>n mom co vapuommsm onflamhooamnpnm Hoxowz .N.< 88 .31 ill (1" its: 1' A. 3. Nickel Phthalocyanine Supported on 20% DVB-poly- styrene. SEM Trace (Sulfone linkage) 89 APPENDIX B 90 Reaction Time m 150 -120 30 20 10 Uptake of 0 mL 2 BJL Polymer Supported NiPc (used) In Oxidation of Cyclohexene at 78°C (Graphs starts after incubation period of 16.75 h) 91 Reaction Time m T 360 - 330 300 270 240 210 180 150 l I l l 40 30 20 10 Uptake of 0 mL 90 2 B.2. Polymer Supported NiPc (ground) In Oxidation of Cyclohexene at 780 C 92 Reaction Time m "270 240 ’210 180 150 ‘ 120 l 1 l L, ! 90 Uptake of 02 mL 3.3, Polymer Supported VOPc (new) In Oxidation of Cyclohexene at 780C 93 Reaction Time m 120 90 60 3O 40 30 20 10 Uptake of 0 mL 2 B.4. Polymer Supported VOPc (used) In Oxidation of Cyclohexene at 780C (Graph starts after incubation period of 15.510 94 \ Reaction Time m ..300 _.l80 120 ..60 I I I I '40 30 20 10 Uptake of 0 mL 2 8.5. Polymer Supported NiPc in Oxidation of Cyclo- hexene with l, 2 Dichloroethane Carrier 95 80 min. Reaction Time 40 60 20 140 100 Uptake of 02 mL 8‘ o‘ 8.6. Polymer Supported MnPc in Oxidation of Cyclohexene Sulfonamide linkage, New and Used Uptake 02 mL 30 20 96 a 1 n I l 30 6O 90 120 lAOmin. Reaction Time Polymer Supported CoPc in Oxidation of Cyclohexene. Sulfonamide linkage, New and Used 22 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 98 R. 8. King, E. M. Sweet, J. Org. Chem., 1979, 44, 385. ~~~~ Personal communication between Dr. C. Brubaker and Dr. R. 8. King. C. Paquot, Compt. rend., 1939, ggg, 171-3. ~~~~ E. I. Ochiai, "Bioinorganic Chemistry", 1977, Chapter 10, Allyn and Bacon. ~~~~ E. I. Ochiai, J. Inorg. Nucl. Chem., 1973, 35, 3375. ~~~~ H. Hock, H. Kr0pf, J. prakt. Chem.,1959,_9, 173-86. ~~~~ S. Stinson Chemical and Engineering_News, 1980, 58, 31. ~~~~ Li, T. T -T.; Brubaker, Jr.; C. H.; J. Organomet. Chem. 1981, 313;, 223. Li, T.T-T.; Weaver, M. J.; Brubaker, C. H. ; J. Amer. Chem.Soc. in press. Murr, N. E., Chaloyard, A., and Tirouflet, J., J.C.S. Chem. Comm., 19§Q. 446. Coutts, R. S. P., Wailes, P. C., and Martin, R.L., J. Organometal. ~~~~ Symons, M. C. R.; Mishra, S. P.; J. C. S. Dalton 1981, £11, 2258. ~~~~ "llllllllllllllllf