Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The catalytic decarbonylation of aldehydes using iron porphyrin complexes Belani, Ramesh 1985

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

Item Metadata

Download

Media
831-UBC_1985_A6_7 B44.pdf [ 3.93MB ]
Metadata
JSON: 831-1.0060600.json
JSON-LD: 831-1.0060600-ld.json
RDF/XML (Pretty): 831-1.0060600-rdf.xml
RDF/JSON: 831-1.0060600-rdf.json
Turtle: 831-1.0060600-turtle.txt
N-Triples: 831-1.0060600-rdf-ntriples.txt
Original Record: 831-1.0060600-source.json
Full Text
831-1.0060600-fulltext.txt
Citation
831-1.0060600.ris

Full Text

THE CATALYTIC DECARBONYLATION OF ALDEHYDES USING IRON PORPHYRIN COMPLEXES By RAMESH M. BELANI B.Sc. (Hons.), BOMBAY UNIVERSITY, 1977 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE i n THE FACULTY OF GRADUATE STUDIES DEPARTMENT OF CHEMISTRY We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA JULY 1985 © RAMESH M. BELANI, 1985 In presenting t h i s thesis i n p a r t i a l f u l f i l m e n t of the requirements for an advanced degree at the University of B r i t i s h Columbia, I agree that the Library s h a l l make i t f r e e l y available for reference and study. I further agree that permission for extensive copying of t h i s thesis for scholarly purposes may be granted by the head of my department or by h i s or her representatives. I t i s understood that copying or publication of t h i s thesis for f i n a n c i a l gain s h a l l not be allowed without my written permission. Department of 1 The University of B r i t i s h Columbia 1956 Main Mall Vancouver, Canada V6T 1Y3 i i ABSTRACT The aim of this project was to investigate the use of iron porphyrin complexes as potential homogeneous catalysts for the decarbonylation of aldehydes. Complexes of the type Fe(TPP)L2 (where L = n-Bu3P, PPh3 or piperidine) were prepared and reacted with CO gas, or with aldehydes as sources of CO. Since the loss of coordinated CO from the Fe(TPP)(CO)(n-Bu3P) complex was more facile, the bis(n-Bu3P) phosphine system was studied in more detail. The X-ray structure of FeTPP(n-Bu3P)2 Is described, and this includes the first determination of an Fe^-P bond distance for a metalloporphyrin. The study using Fe(TPP)L2 complexes as decarbonylation catalysts was somewhat hindered by the extreme air-sensitivity of the porphyrin complexes in solution. UV/visible spectroscopy and gas chromatography were used to monitor the decarbonylation reactions. The reaction mixtures were analysed by GC/MS. The decarbonylation reactions were characterised by inconsistent turnover numbers and lack of reproducibility; during the decarbonylation of phenylacetaldehyde, bibenzyl was detected. Such factors are indicative of a free radical mechanism, similar to that proposed earlier for related Ru(II) porphyrin systems. The carbonylation of FeTPP(n-Bu3P)2 by CO gas was of interest with respect to the catalytic reaction, which must involve formation of a carbonyl complex. The reaction, FeTPP(n-Bu3P)2 + CO K -» FeTPP(n-Bu3P)(CO) + n-Bu3P i i i was found to have a K value of 0.72 at 29°C, while the temperature dependence of K was studied to obtain the thermodynamic parameters AS and AH for the equilibrium. iv Table of Contents ABSTRACT i i Table of Contents iv List of Figures • v i i List of Tables x Table of Abbreviations x i i Acknowledgements xiv Chapter I INTRODUCTION 1 1.1 General introduction 1 1.2 The choice of phosphine ligand 5 1.3 The use of metalloporphyrins in decarbonylation reactions 9 1.3.1 Ruthenium (II) porphyrin complexes 10 1.3.2 Mechanism of decarbonylation using the Ru(II)TPP(PPh3)2/n-Bu3P system 16 I. 3.2 Iron (II) porphyrin complexes 19 Chapter II EXPERIMENTAL 21 II. 1 Techniques 21 II.2 Gases 21 V 11.3 Solvents 22 11.4 Other chemicals 22 11.5 Spectroscopic measurements 23 11.6 Decarbonylation procedures 26 11.7 Tetraphenylporphyrin and complexes 27 11.8 Program for gas chromatographic analysis 32 II. 8.1 Turnover numbers 34 Chapter III STRUCTURE OF FeTPP(n-Bu3P)2 35 III. l Structural analysis 35 111.2 X-ray structural analyses of metalloporphyrins .... 39 111.3 Characteristics of the Fe-Np bond distance 40 111.4 Characteristics of the M-L bond distance 45 111.4.1 Steric interactions 45 111.4.2 Degree of n-backbonding 48 111.4.3 Spin-state of metal ion 49 111.5 Model Fe-L (axial) bond distance of 6-coordinated low-spin Fe(II) tetraphenylporphyrin complexes .... 50 111.6 Conclusions 52 Chapter IV THE REACTION OF FeTPP(n-Bu3P)2 WITH CO AND 0 2 53 IV. 1 Reaction with CO 53 IV. 1.1 Aldehydes as sources of CO 69 IV.2 Reaction of FeTPP(n-Bu3P)2 with 0 2 72 v i Chapter V THE CATALYTIC DECARBONYLATION OF ALDEHYDES USING FeTPP(n-Bu 3P) 2 75 V . l Choice of FeTPP(n-Bu 3P) 2 75 V.2 Preliminary reactions with aldehydes 76 V.3 Factors influencing decarbonylation 81 V.3.1 Influence of added phosphine 81 V.3.2 E f f e c t of CO 83 V.3.3 E f f e c t of 0 2 83 V.3.4 Solvents 84 V.3.5 Varying the r a t i o of aldehyde to porphyrin ........ 87 V.3.6 Influence of temperature 88 V.3.7 Radical i n h i b i t o r s 89 V.3.8 Rate of gas flu s h i n g 89 V.3.9 Control tests 90 V.4 Decarbonylation mechanism 90 V.4.1 The role of the phosphine ligand 94 V.5 Comparison of FeTPP(n-Bu 3P) 2 system with other decarbonylation systems 94 V.6 Conclusions 95 V.7 Suggestions for further studies 96 v i i LIST OF FIGURES Figure 1.1 A molecular o r b i t a l picture of the bonding of a t r a n s i t i o n metal to CO and a phosphorus ligand 6 Figure 1.2 Mechanism of decarbonylation using [Rh(P-P) 2] + as ca t a l y s t 8 Figure 1.3 Meso-tetraphenylporphyrin 10 Figure 1.4 V i s i b l e spectrum t y p i c a l of sol u t i o n no longer active for c a t a l y t i c decarbonylation; that shown i s for sol u t i o n of Ru(TPP)(PPh 3) 2/(n-Bu 3P) a f t e r decarbonylation of PhCH2CH0; ( ) same sol u t i o n In presence of hydroquinone; i n a c t i v e for decarbonylation 14 Figure 1.5 E.S.R. signals at l i q u i d nitrogen temperature i n 5:1 CH 2C1 2/CH 3CN: A, Ru(TPP)(CO)( tBu 2POH)/cyclohexen-4-al system; B, the Ru(TPP)(CO)( tBu 2POH)/pyridine-2-aldehyde system; C, the Ru(TPP)(PPh 3) 2/n-Bu 3P/2-phenylacetaldehyde system ... 15 Figure 1.6 Proposed decarbonylation mechanism using Ru(TPP)(PPh 3) 2/ n-Bu3P system 18 Figure II.1 Evacuable c e l l for Optical Density Measurements 25 Figure I I I . l Stereoscopic view of FeTPP(n-Bu 3P) 2 structure 36 Figure III.2 A diagram i l l u s t r a t i n g the s t e r i c i n t e r a c t i o n s of an a x i a l ligand of a metalloporphyrin with the porphinato core. The dihedral angle <t> i s between the plane of the ligand and the plane defined by porphinato nitrogen atom, the metal atom and the ligand nitrogen atom (N ) 45 v i i i Figure IV.1 UV/visible spectrum of FeTPP(n-Bu3P)2 in toluene (~ 10-ltM) 57 Figure IV.2 UV/visible spectrum of FeTPP(n-Bu3P)2 in toluene (~ I0~h M) + CO gas (0.5 - 1 atmosphere) 58 Figure IV.3 Spectral changes observed for the reaction FeTPP(C0)(n-Bu3P) + n-Bu3P ^ FeTPP(n-Bu3P)2 +. CO at 29°C 59 Ao~As Figure IV.4 Plot of log ( A e_ A ) versus log [n-Bu3P], 29°C. For the CO reaction FeTPP(CO)(n-Bu3P) + n-Bu3P - — * FeTPP(n-Bu,P)0 + CO 66 1 Figure IV.5 Van't Hoff plot, LnK versus T < i m n p r a , „ r p 68 Figure IV.6 UV/visible changes observed after addition of aldehyde to FeTPP(n-Bu3P)2 in toluene (or CH2C12). Spectral changes are reversible on vacuum pumping the optical cell 71 Figure IV.7 UV/visible spectrum of FeTPP(n-Bu3P)2 in toluene (or CH2C12) after exposure to 0 2 73 Figure IV.8 UV/visible changes on addition of n-Bu3P to oxidised porphyrin solution shown in IV.7 74 Figure V.l GC trace for decarbonylation of phenylacetaldehyde (~ 10_3M) using FeTPP(n-Bu3P)2 (~ 10_1+ M) in refluxing CH2C12 (23°C) after 8 minutes 79 ix Figure V.2 GC trace of phenylacetaldehyde (~ 10 - 1 M) with FeTPP(n-Bu3P)2 (~ 10" 3 - 10 - 4 M) in refluxing CH2C12. Product collected in cold trap 80 Figure V.3 Bibenzyl detected during decarbonylation of phenylacetaldehyde. Identification by GC/MS and comparison with computerized MS library 86 Figure V.4 Tentative decarbonylation mechanism using FeTPP(n-Bu3P)2 as catalyst 92 X L I S T O F T A B L E S Table 1.1 Decarbonylation of aldehydes using a Ru(TPP)(PPh3)2/n-Bu3P catalyst system 12-13 Table II.1 GC retention times (mins) for standards using 0V101 column 32 Table II.2 GC retention times (mins) for standards using 0V17 column 33 Table III.l Bond lengths (A) in FeTPP(n-Bu3P)2 with estimated standard deviations in parentheses 37 Table III.2 Bond angles (deg) in FeTPP(Bu3P)2 with estimated standard deviations in parentheses 38 Table III.3 AN^  values reported for iron porphyrin complexes 41-42 Table III.4 Fe-Np bond distances in Fe-tetraphenylporphyrins 43 Table III.5 M-L bond distance changes due to steric factors 47 Table IV.1 Data used to calculate equilibrium constant for the reaction FeTPP(n-Bu3P)2 + CO —FeTPP(n-Bu 3P)(CO) + (n-Bu3P) 60-65 Table IV.2 Values of [n-Bu3P] for log [ A e_^ 1 = 0, at 18-40°C 67 CD Table IV.3 Solubility of 1 atmosphere CO in toluene, corrected for vapor pressure of toluene 67 Table IV.4 Equilibrium constant (K) values for the reaction FeTPP(n-Bu3P)2 + CO =~ FeTPP(C0)(n-Bu3P) + (n-Bu3P).. 67 xl Table IV.5 Equilibrium data for reaction of Fe and Ru porphyrin complexes with CO in toluene solvent. M( porphyrin)L2 + CO M(porphyrin)(CO)L + L 69 Table IV.6 Aldehydes used as source of CO. Time required to completely form CO adduct, hours. Reaction carried out in evacuated optical cell (Figure II.1) at 23°C. Carbonyl adduct formation confirmed by UV/visible spectrum. Completion checked by extinction coefficient 70 Table V.l Decarbonylation of aldehydes with FeTPP(n-Bu3P)2 in 30 ml CH2C12 at 24°C 78 Table V.2 Effect of excess phosphine on decarbonylation of benzaldehyde in 30 ml CH2C12 at 24°C 82 Table V.3 Effect of CO on decarbonylation of benzaldehyde in 30 ml CH2C12 at 24°C 83 Table V.4 Effect of varying ratio of aldehyde to porphyrin in 30 ml CH2C12 at 24°C; argon flush 88 Table V.5 Decarbonylation of phenylacetaldehyde at 40°C in 30 ml CH2C12; argon flush 88 x i i TABLE OF ABBREVIATIONS atm atmosphere A absorbance at any time, t Ae absorbance at the equilibrium position Ao absorbance of the reactant metalloporphyrin at time zero A absorbance of the metalloporphyrin product at the end CO of reaction CHjClj methylene chloride or dichloro methane CH3CN acetonitrile °C degrees centigrade cm centimetre CPO chloroperoxidase (enzyme) DDQ 2,3-dichloro-5,6-dicyanobenzoquinone Esr Electron spin resonance EXAFS Extended X-ray Absorption Fine Structure FeTPPL2 Bis-ligated iron(II) tetraphenyl porphyrin (L = ligand) GC/MS Gas chromatography/Mass spectroscopy AH enthalpy change for the reaction Him Imidazole °K degrees Kelvin K equilibrium constant k kinetic rate constant x i i i M molarity M(porphyrin)L five-coordinate metal porphyrin complex M(porphyrin)L 2 six-coordinate metal porphyrin complex M(porphyrin)L(CO) carbonylated six-coordinate metal porphyrin complex l-Melm 1-methylimidazole mg milligram n-BugP n-butylphosphine nm nanometre OEP octaethylporphyrin dianion OMBP octamethyltetrabenzoporphyrin Pc phthalocyanine dianion pip piperidine PpIX protoporphyrin IX py pyridine AS entropy change for the reaction S spin state s second T temperature t time TPP tetraphenylporphyrin dianion e molar e x t i n c t i o n c o e f f i c i e n t v frequency (cm - 1) xiv ACKNOWLEDGEMENTS I am g r a t e f u l to Professors David Dolphin and Brian James for t h e i r guidance at every stage of t h i s project. I thank Mr. Peter Borda for elemental analyses, Dr. S. Rettig for the X-ray structure determination of FeTPP(n-Bu 3P) 2 and Dr. Eigendorf's group for the GC/MS an a l y s i s . I am indebted to several members of the Bio-inorganic research group. Joanne Crocker i s complimented for her swift typing of this manuscript. A teaching a s s i s t a n t s h i p (1982-85) i s acknowledged. 1 CHAPTER I INTRODUCTION 1.1 General Introduction This thesis is concerned with the homogeneous catalytic decarbonylation of aldehydes using iron porphyrin complexes as catalysts. Decarbonylation of aldehydes [1], acyl halides, aroyl halides, alcohols and ketones has been reported [2-4] in the literature. This chapter reviews this field with emphasis on the decarbonylation of aldehydes using transition-metal complexes. Such reactions are useful in organic synthesis and are characterised by the use of mild conditions, minimal side-reactions and good stereoselectivity. Other decarbonylation methods include the use of non-transition metals, and thermal and photochemical reactions [5,6]. Systems used to decarbonylate aldehydes include those using ruthenium trichloride hydrate [7-10], iridium halides [11], OsBr2(PPh3)3 [12], Ru 2Cl 3(PEt 2Ph) 6 +Cl - [13.1(h)], RhCl(PPh3)3 [14], trans-Mo(N2)2~ (Ph2PCH2CH2PCH2)2 [15], and palladium [16]. Of these, Rh ICl(PPh 3) 3 (Wilkinson's complex) has been the most widely studied for homogeneous stoichiometric decarbonylation of aldehydes under mild conditions (eqs. 1.1, 1.2). 2 0 il R-C-H + RhCl(PPh3)3 > RH + RhCl(CO)(PPh3)2 + PPh3 (eq. 1.1) 0 II RiCHCR^C-H + RhCl(PPh3)3 > RjC = CKl + H 2 + RhCl(CO)(PPh3) 2 + PPh3 (eq. 1.2) Olefin is produced (eq. 1.2) when a B-hydrogen is present, e.g. heptanal is decarbonylated to give 86% hexane and 14% hexene [17]. A mechanistic study [18] has indicated that oxidative addition of the aldehyde to Rh* is the rate determining step. An Intermediate formed on oxidative addition of aldehyde has been observed by "chelate trapping"; i.e. using the ability of 8-quinoline carboxaldehyde to form a chelate after oxidative addition to Rh1. This resulted in the first stable, isolable Rh(III) acyl hydride complex [19] (1) CD 3 which on prolonged heating i n xylene at 130°C yielded the expected product, quinoline. In an e a r l i e r study, Lochow and M i l l e r [20] attempted to trap an intermediate acyl-rhodium hydride by addition of the Rh-H bond to a double bond of an o l e f i n i c aldehyde. M i l s t e i n [21] has synthesised a cis-hydridoacylrhodium complex not s t a b i l i s e d by chelation, using Rh(PMe 3) 3Cl. S t a b i l i t y of these compounds (2) i s due to the very slow rate of PMe3 d i s s o c i a t i o n . I ' I RhL,Cl + RCHO > "<r,^ 3 o<\ V L (eq. 1.3) L = PMe3; R = CH 3, Ph Heating above 60°C y i e l d s the alkane and small amounts of aldehyde formed by re v e r s i b l e reductive e l i m i n a t i o n . If the product aldehyde i s removed, the acyl-hydride reductive elimination process i s the major one. This i s claimed to be the f i r s t d i r e c t observation of aldehyde reductive elimination from a hydridoacylmetal complex. Another study [22(a)] using RhCl(PPh 3) 3 showed that the intermediate a l k y l and aryl-rhodium hydride complexes formed during aldehyde decarbonylation can be trapped by intramolecular addition to a double bond, and that the intermediate a c y l hydride complex forms and decomposes with retention of stereochemistry. An 4 investigation [22(b)] for the enantioselective catalytic decarbonylation of racemic aldehydes using [Rh(P-P ) 2]X complexes (X = Cl, BF^, and P-P represents a chiral di-tertiary phosphine ligand) reported the formation of optically active hydroacylated products formed via kinetic resolution of the precursor racemic aldehyde. Stoichiometric decarbonylation of aromatic aldehydes with RhCl(PPh3)3 is thought to involve oxidative addition of aldehyde to give an aroyl hydride, followed by reverse of the CO insertion reaction and reductive elimination of alkane (eq. 1.4). Catalytic decarbonylation requires loss of CO from the metal carbonyl. H I ITX Rh + RCHO > Rh H I I L ^ I -> R h ^ > Rh (CO) + COR RH (eq. 1.4) CO Free-radical intermediates in the decarbonylation of aldehydes using Rh*Cl(PPh3)3 have been excluded on the basis of a study by Kampmeier et al. [23], who reported that exo- and endo-5-norbornene-2-carboxaldehyde gave, on decarbonylation, norbornene and nortricyclane, respectively; citronellal was decarbonylated to only 2,6,dimethyl-2-heptene; and a mixture of CgH5CH2CDO and p-CH3C6H5CH2CHO on decarbonylation gave no D crossover in the product. 5 The catalytic decarbonylation of aldehydes using RhCl(PPh3)3 or RhCl(CO)(PPh3)3 at high temperatures (300 °C) has been reported [24]. Lower temperatures (110°C) for catalytic decarbonylations have been claimed [25] for 3 related catalysts: Rh 2(n-Cl 2)(CO) 2(PPh 3) 2; Rh(^-Cl2)(PPh3)^ and RhCl3(CO)(PPh3)2. The drawback noted with the RhCl(PPh3)3 system is that the loss of coordinated CO from RhCl(CO)(PPh3)2 is not observed even at 200°C under vacuum [26], upon treatment with molten PPh3 at 100°C [27], or with UV irradiation [28]. Excess phosphine decreased the catalytic activity of RhCl(PPh3)3 [1(c)]. 1.2 The choice of phosphine ligand A key step in the catalytic decarbonylation of aldehydes using transition-metal complexes is the loss of coordinated CO from the interme-diate transition-metal carbonyl complex. The use of phosphine ligands to labilise CO has been reported [1(f)]. Other studies indicate the use of phosphines in catalyst systems [29]. Although one to three monophosphine ligands can readily coordinate to a metal centre, the high trans-effect exhibited by phosphine ligands, combined with their steric bulk, result in easily dissociable complexes, whereby metal sites become available for further reactions. A study of steric and electronic effects of phosphorus ligands on the chemistry of transition metal complexes [31] has indicated that steric effects of phosphorus ligands are usually more important than electronic effects. 6 The bonding between a transition metal, trans disposed CO and phosphorus ligand is as shown In Figure 1.1. Bonding by CO Involves a donation from the carbonyl carbon and back-donation from a filled metal d orbital of appropriate symmetry with an unfilled antibonding it orbital of CO. The bonding by phosphine is similar except that the unfilled orbital on phosphorus is a d orbital. When both CO and P are bonded to M, both ligands compete for n backdonation from the filled metal d n orbital. A weaker it-accepting phosphine allows more donation to the carbonyl n orbital, thus lowering the carbonyl stretching frequency. A stronger it-acceptor phosphine should form a stronger M-P bond. However, steric factors come into play, so that 'thin' ligands compete more favorably for the metal site despite electronic considerations Figure 1.1 A molecular orbital picture of the bonding of a transition metal to CO and a phosphorus ligand. [30]. The high temperatures required for catalytic decarbonylation of aldehydes using RhCl(PPh3)3 or RhCl(CO)(PPh3)2 make the system impractical 7 for routine synthetic use. Since the CO is not photolabile [28], ligand modifications were designed in attempts to labilise the CO in complexes similar to RhCl(C0)(PPh3)2. One such approach was the use of solvent stabilised cationic complexes [31] such as [Rh(PPh 3) 2] + and [Rh(PPh3)(CO)]+. The lower effective basicity of Rh in the last two mentioned complexes is partly responsible for a noted increased lability of CO. Complexes of the type [Rh(P-P)2]X and [Rh(P-P)(solvent)2]X, where X = Cl~ or BF^" and P-P = Ph2P(CH2)nPPh2 (called dppm and dppp for n = 1 and 3 respectively) bind CO reversibly. The [Rh(dppe)2]+ complex (n » 2) does not react with CO while the [Rh(dppb)2]+ complex (n = 4) absorbed > 1 mole of CO per Rh [32]. The cis phosphine stereochemistry is responsible for a major increase in decarbonylation reactivity as compared to the trans-triphenylphosphine analogues [1(c)]. For example, the catalytic decarbonylation of benzaldehyde by [Rh(dppp)2]BF1+ at 150°C was 200 times faster than with trans- RhCl(CO)(PPh3)2 [1(f),33]. The catalytic activity of the Rh(P-P)2+ complexes was found to be dependent [33] on the chelate ring size (dppp > dppe > dppb > dppm), the bis-phosphine complex [Rh(P-P)2]+ found to be twice as active as [Rh(P-P)]+ at a l l temperatures. Operating temperatures as low as 100°C for the [Rh(dppp)2]+ system were effective, and the type of counterion did not affect the reactivity. During decarbonylation runs using [Rh(dppp)2]+ at 150°C no Rh(III) intermediates were detected, and quenching the reaction with cold pentane gave only [Rh(dppp)2]+ and [Rh(dppp)2C0]+, either of 8 which could be reused as catalyst. The loss of CO from [Rh(dppp)2CO]+ is facile [1(f),32]. A possible mechanism of decarbonylation using [Rh(P-P)2]+ as catalyst, where (P-P) • dppp or dppe is reported [1(c)] (Figure 1.2). Figure 1.2 Mechanism of decarbonylation using [Rh(P-P)2]+ as catalyst [1(c)] A mechanism involving a n 2 (n-bonded) carbonyl has been suggested by Rauchfuss [34] for an unusual ruthenium complex which on heating undergoes facile intramolecular decarbonylation (eq. 1.5). 9 Ph Ph Ph P (eq. 1.5) Ph Ph Ph Oxidative addition is considered unlikely as i t would involve an uncommon IV Ru intermediate. 1.3 The Use of Metalloporphyrins in Decarbonylation Reactions Porphyrins are versatile macrocyclic ligands which chelate with most metals via the 4 pyrrole nitrogens; metalloporphyrins can exhibit additional ligand binding at axial coordination sites above and below the porphyrin plane to form 5- and 6-coordinated species. The use of ruthenium porphyrins in the decarbonylation of aldehydes [1(a)], has been reported. The structure of the meso-tetraphenylporphyrin, a frequently used ligand, is shown in Figure 1.3 Figure 1.3 Meso-tetraphenylporphyrin 1.3.1 Ruthenium (II) porphyrin complexes Studies in these laboratories have shown that phosphine ligands in an axial position of a metalloporphyrin labilise a trans CO [35]. The use of RuTPP(PPh3)2 [1(b)] and RuTPP(CO)(t-Bu2POH) [1(a)] as catalysts for the 11 decarbonylation of aldehydes has been reported (Table 1.1). Decarbonylation of aldehydes occurred at ambient temperatures using either of the above catalysts (~ 10~3 M); addition of n-Bu3P (~ 10~5 M) and pretreatment of the reaction mixture with CO to prevent formation of inactive RuTPP(n-Bu3P)2, resulted in rapid decarbonylation; for example, a turnover of 5 x 104 h - 1 at 60°C was obtained in the case of phenylacetaldehyde using a CH2C12/CH3CN solvent system (Table 1.1). The CO pretreatment was not necessary for the case of RuTPP(CO)(t-Bu2POH) as catalyst. UV/visible spectral studies were carried out to monitor the decarbonylation reaction using RuTPP(PPh3)2/n-Bu3P as catalyst to identify possible species present during catalysis. From the spectral data obtained, the active catalytic solution was considered to contain two main species (i) a bis-phosphine, either a mixed phosphine complex, or a bis(tri-n-butylphosphine) complex, and (i i ) a solvated species like RuTPP(CO)(S) (where S = solvent). Inactive solutions showed a Soret at 410 nm which was attributed to a RuTPP(CO) species. Addition of hydroquinone, a radical inhibitor, inhibited the decarbonylation reaction. The added hydroquinone did not Interact with the Ru(II). Figure 1.4 shows the UV/visible spectrum of an 'inactive' decarbonylation reaction mixture, and also the inactive solution that resulted from addition of hydroquinone. 12 Table 1.1 Decarbonylation of aldehydes using a Ru(TPP)(PPh3) ^ " B ^ P catalyst system [la] Substrate Major Product(%)- Conversion (time)-c Turn-over— C6H5CHO Benzene (100) 10(5) 10 C6H5 CH=CHCHO(trans) Styrene (100) 20(10) 20 C6H5CH2CH0 Toluene (95)- 30(1),90(4) 10 3^ £-CN-C6HHCHO Benzonitrile (100) 15(12) 20 n-C6H13CHO n-C6H14 (65)- 10(1) 10 2 2-Ethylbutanal n-C5H12 (85)- 30(1) 10 3 ^ ^ C H O C H , 0 0 (60); (35) n-C6Hltf (5) 30(1), 50(18), 90(50) 2 x l 0 2 13 Table 1.1 - cont'd. Substrate Major Product(%)- Conversion (time)-Turn-over— or (Q (70); Q (35) 10(1), 20(12) 102 10(1), 30(36), 90(150) 102 20(5) 102 T OMe 1 OMe _ - CHO a g 0 ( 1 0 0 > 20(3) 4xl0 2 — Identified by g.c.-m.s. and/or n.m.r.; % refers to amount of major species in the decarbonylation products at the highest conversion noted. — % Conversion of aldehyde (time in h). — For the first hour at ambient temperature, based on loss of aldehyde and/or formation of product as detected by G.C. — Small amounts of benzene also detected. ® At ~ 60°C, turnover 5 x 104 h - 1 . — Other products not yet identified; may be decomposition products. & Using Ru(TPP)(C0)(tBu2P0H) in toluene. 1A tu u z < CO o to DQ < 400 • 500 600 Figure 1.4 Visible spectrum typical of solution no longer active for catalytic decarbonylation; that shown here is for solution of Ru(TPP)(PPh3)2/nBu3P after decarbonylation of PhCH2CH0;( ) same solution in presence of hydroquinone; inactive for decarbonylation 1 Some evidence for Ru(III) intermediates was obtained by cyclic voltammetry performed on the Ru(TPP)L2 systems (L = n-Bu3P; PPh3) and also on the RuTPP(PPh3)2/n-Bu3P catalytic system. During decarbonylation of the indane aldehyde, (Table 1.1), waves at -0.1V, and in the phenylacetaldehyde system, waves at -0.08V and +0.08V were thought to result from Ru(III) species [1(a)]. Infrared measurements during the decarbonylation of the indane aldehyde using the RuTPP(PPh3)2 catalyst system revealed a small peak at 2015 cm-1 which was tentatively attributed to a Ru(III) hydride 15 [36]. ESR studies performed on the very slow catalytic decarbonylation of pyridine-2-aldehyde and cyclohexene-4-al detected the presence of organic free radicals (g = 2.00) in both systems (Figure 1.5), no signals being observed in the absence of the Ru complexes. During decarbonylation of phenylacetaldehyde, a low temperature broad signal at g = 2.20 was assigned to a low-spin d 5 Ru(III) species [1(a)]. g values *-00 Figure 1.5 E.s.r signals at liquid nitrogen temperature in 5:1 CH2C12/CH3CN: A, the Ru(TPP)(CO)(tBu2POH)/cyclohexen-4-al system; B, the Ru(TPP)(C0)(tBu2P0H)/pyridine-2-aldehyde system; C, the Ru(TPP)(PPh3)2/n-Bu3P/2-phenylacetaldehyde system. The RuTPP(PPh3)2 complex, unlike RuTPP(n-Bu3P)2, at the concentrations used, rapidly dissociated one phosphine according to eq. 1.6 [1(a)] 16 K RuTPP(PPh3)2 =» RuTPP(PPh3) + PPh3 (eq.1.6) •>«: The 5-coordinate species added a carbonyl ligand instantaneously if CO gas was used, while stoichiometric carbonyl formation from an aldehyde occurred less rapidly. Tri-n-butylphosphine used with RuTPP(PPh3)2 in the catalyst system was known to displace CO more readily than triphenylphosphine; how-ever, excess n-Bu3P inhibited the catalytic decarbonylation reaction [1(a)]. Nitrile solvents such as benzonitrile or acetonitrile were used for the catalytic system using RuTPP(PPh3)2/n-Bu3P. Two effects from the use of these solvents were considered, namely, prevention of dimerization or aggregation of unsaturated intermediates and the formation of solvated species like RuTPP(phosphine)(CH3CN) for which there was some spectral evidence [1(a)]. It was found that nitrile solvents were not essential for the catalytic decarbonylation reaction as the reactions occurred in CH2C12 solvent but with lower yields. 1.3.2 Mechanism of decarbonylation using the Ru(II)TPP(PPh3)2/n-Bu3P system. A free radical mechanism was proposed for the decarbonylations (Figure 1.6). The reactions were characterised by poor kinetic data, irreproducibility of the reactions, and a wide range of turnover numbers was obtained for similar conditions; trace 0 9 was also found necessary to 17 initiate the reaction; further, frozen catalyst solutions gave e.s.r. signals due to free radicals, cyclic voltammetry indicated the presence of Ru(III) intermediates, radical inhibitors prevented decarbonylation without destroying the catalyst and, in the case of cyclohexane-aldehyde, methylcyclopentane was one of the products, a plausible route for its formation being the rearrangement of the 5-hexane-l-yl radical [37] (eq. 1.7). Such acyl radical species can be stabilised by a metal complex in a cage reaction [38]. (Figure 1.6) Decarbonylation of the acyl r a d i c a l was thought to be possibly metal a s s i s t e d , giving a Ru(II) carbonyl that was subsequently decarbonylated by nu c l e o p h i l i c attack by n-Bu3P (eq. 1.8). RuTPP(CO)(n-Bu 3P) + n-Bu3P — R u T P P ( n - B u 3 P ) 2 + CO (eq. 1.8) This reaction [1(b)] occurs thermally i n toluene at 30°C with an equilibrium constant (K) equal to 18.5. Oxidative addition of the aldehyde to the c a t a l y s t , CO migration (with subsequent elimination) and reductive-elimination of product would, 19 in the case of metalloporphyrins, require coordination numbers higher than 6, and involve Ru(IV) intermediates, both of which are unusual [1(b)]. 1.3.3 Iron(II) porphyrin complexes The present work deals with the study of bis(phosphine)tetraphenyl-porphinatoiron(II) as a potential decarbonylation catalyst. The reaction of 6-coordinate iron(II) porphyrins with CO and 0 2 proceeds via a dissociative mechanism and is represented [39] by eq. 1.9 k x - L k 2 + X FePL2 _ ^ FePL -=* FePL(X) (eq. 1.9) lc_ 1 + L k_2 - X (P = porphyrin; X = CO or 02) The extreme air-sensitivity [40] of simple Fe(II) porphyrins has resulted in the use of 'slowed down' ruthenium analogues [41] as the reactive intermediates, for the ruthenium complexes are considered more stable and easier to handle when compared to the iron porphyrin complexes. Since the use of Ru(II) porphyrin complexes as catalysts for the decarbonylation of aldehydes was successful, the use of iron porphyrins seemed a logical extension of the decarbonylation studies. At the onset of the present work early in 1982, iron(II) porphyrin complexes containing axial phosphine ligands had not been synthesized. The discovered ease of preparation of 20 such iron porphyrin complexes soon led to their being found effective for catalytic aldehyde decarbonylation; the mild operating temperatures found (25-40°C) and the low cost of Fe versus Ru or Rh could lead to cheap, effective catalysts for this decarbonylation reaction, particularly i f the potential use of polymer-bound phosphines is considered. 21 CHAPTER II EXPERIMENTAL 11.1 Techniques Elemental analyses were carried out by Mr. P. Borda of this department. Gas chromatography was performed on a Hewlett-Packard 5830A or a Carle 113 instrument using 0V101 and 0V17 packed columns. Infrared spectra were recorded on a Nlcolet 5DX spectrometer. Gas-liquid chromatography and mass spectroscopy were carried out on a Carlo-Erba Fractovap 4160-Kratos MS80RFA instrument. Nuclear magnetic resonance spectroscopy was recorded on Bruker WH-80 and WH-400 instruments. Visible spectra were recorded on Cary 17 or Perkin-Elmer 552A UV/visible spectrometers, both being fitted with thermostatted cell compartments. Quartz cells of 1 mm or 1 cm path length were used. 11.2 Gases CP. grade carbon-monoxide, nitrogen and purified argon were obtained from Canada Liquid Air Ltd. or Union Carbide Canada Ltd. 22 11.3 Solvents All solvents were distilled, stored and handled under argon using Schlenk techniques [42]. Purification methods outlined in the literature [43] were used: acetonitrile, dichloromethane and toluene (Aldrich, spectral grade) were distilled from CaH2• Benzene (Eastman Kodak, spectral grade) was distilled from a solution containing the blue ketyl formed by the reaction of sodium-potassium alloy with a small amount of benzophenone. Prior to use, the solvents were degassed by 4 freeze-pump-thaw cycles. Propionic acid (Mallinckrodt, reagent grade) was dried with NagSO^ , and then distilled after refluxing with a few crystals of KMnO^ . 11.4 Other chemicals Liquid aldehydes were distilled under nitrogen, at reduced pressures, while solid aldehydes were recrystallised from appropriate solvents [43]. Triphenylphosphine (E. Merck) was recrystallised from n-hexane. n-Butylphosphine (Strem Chemicals, Inc.) was fractionally distilled under vacuum, the purity being checked by 3 1P nmr (CgD6 solution, in vacuo, 6 -31.8 ppm w.r.t. H3P0lt). 23 Piperidine (Fisher Scientific Co.) was dried with KOH and fractionally distilled under nitrogen. Pyrrole (Fisher Scientific Co.) was dried with NaOH and fractionally distilled under reduced pressure from CaH2 and stored under nitrogen. The pyrrole was redistilled immediately before use. II.5 Spectroscopic Measurements Because of the extreme air-sensitivity of the iron porphyrins in solution, an evacuable optical cell (Figure II.1) was used for a l l optical density measurements. The cell was maintained at constant temperature in a thermostated cell compartment, which was attached to a constant-temperature circulating bath. The solid iron porphyrin complex was placed in the quartz path of the cell through A (Figure II.1), and the solvent added Into B. The solvent was degassed by repeatedly freezing i t in the flask of the cell, pumping off any uncondensed gas and then thawing. This freeze-pump-thaw cycle was repeated 4 times. The degassed solvent could then be added to the solid sample. The cell solution was then allowed to equilibrate in the thermostatted .compartment. After an i n i t i a l spectrum was recorded, CO at a known pressure of 1 atm was admitted into the cell which was then shaken in order to ensure complete mixing of the gas in the solution. The resulting spectrum was recorded. Since the concentrations of the iron porphyrin complexes used were very low (~ 10"5 M) and the solubility of CO 24 is of the order of 10" 3 M atm-1, the gas concentrations in solution and the partial pressure of the gas remained essentially constant throughout the experiment. Gas pressure was measured using a mercury manometer. For experiments using aldehydes as sources of CO, a similar procedure was followed. In the case of liquid aldehydes, a carefully degassed aliquot was injected into the cell through C (Figure II.1) using a gas-tight syringe, whereas solid aldehydes were placed along with the iron porphyrin complex in the quartz part of the cell through A (Figure II.1) 25 A (<-4 S i l i c o n e septum VACUUM LINE FITTED WITH A HIGH VACUUM TEFLON STOPCOCK FITTED WITH B-14 CAP \ QUARTZ CELL - Figure (II.1) Evacuable C e l l for Optical Density Measurements 26 II.6 Decarbonylation Procedures Three methods were used, method 3 being the most successful Method 1; An evacuable optical cell (Figure II.1) under argon was used. Liquid or solid aldehyde was introduced into the cell containing a solution of Fe^TPPLj in 10 mL CH2C12« Concentrations of aldehyde and iron porphyrin were ca. 10" 3 M and 10""'* M, respectively. The optical cell was evacuated and sealed, and reaction followed by monitoring the increase of porphyrin carbonyl peak at 438 nm. (These reactions with CO gas or aldehydes as sources of CO are described in detail in Chapter IV). Method 2; Concentrations of aldehyde and iron porphyrin complexes comparable to those in method 1 were used. Aldehyde was added to a CH2C12 solution of FeTPP(n-Bu3P)2 in a Schlenk flask under argon. After stirring the reaction mixture for a few minutes, the Schlenk flask was connected to another empty flask and the system evacuated. The mixture was then bulb-to-bulb distilled, and the distillate and residue analysed for decarbonylation products by G.C. Method 3: The reaction apparatus consisted of a side-armed flask fitted with a reflux condenser that was connected to a cold trap (dry ice and acetone, -79°C). The system was continually flushed with argon. Aldehyde (~ 10 _ 1 to 10~2 M) was added to a refluxing solution of FeTPP(n-Bu3P)2 in CH2C12 27 (~ 10 _ ; i-10 - b M). The reaction mixture was monitored by G.C. analysis of samples withdrawn every 5 minutes. Rapid decarbonylation of phenylacetaldehyde was obtained using this method. Argon gas used for the three experiments was deoxygenated by passing i t through a BASF deoxygenation catalyst maintained at 45°C and dried by passing i t through P 20 5, KOH and D r i e r i t e columns. II .7 Tetraphenylporphyrin and complexes H2TPP was prepared by a method sim i l a r to that of Adler et a l . [44]. Freshly d i s t i l l e d pyrrole (56 mL, 0.8 mole) and benzaldehyde (80 mL, 0.8 mole) were refluxed for 30 min i n 3 L of propionic a c i d . Cooling the mixture, f i l t e r i n g , washing with methanol and hot water and f i n a l l y drying, yielded 26 g (20% y i e l d ) of purple c r y s t a l l i n e TPP. To remove the c h l o r i n impurity [45], samples of TPP were dissolved i n re f l u x i n g toluene and DDQ was added. The mixture was cooled and extracted with NaOH solu t i o n containing Na 2S0 l |. The organic layer was washed with water and dried with Na 2S0 1 +. Toluene was removed under reduced pressure and the product c r y s t a l l i s e d from CH 2C1 2/methanol. Elemental analysis for C^H^N^ calc u l a t e d : C 85.90; H 4.92; N 9.12 found: C 85.20; H 4.99; N 9.10 UV/Vis Spectrum: benzene solvent \ (e,M - 1cm - 1) nm 419(e = 4.67 * 10 5); 5l4(e = 1.86 x 101*); 549(e = 7.58 * 10 3); 591(e = 5.37 * 10 3); 647(e = 3.3 x 10 3) 28 The data agree well with those given In the l i t e r a t u r e [44]. Mass spectrum: (m/e) H 2TPP +(614); H 2TPP 2 +(307) Fe'I"I"^(TPP)Cl was prepared following the method of Kobayashi et a l . [46]. Thus, H2TPP (1.5 mmole) and FeCl 2*4H 20 (.25 mraole) were dissolved i n 500 mL of DMF and refluxed for 3 h. The solution was cooled and d i l u t e (1:1) HC1 (20 mL) added. The dark purple c r y s t a l s obtained were r e c r y s t a l l i s e d twice from a mixture of 1,2-dichloroethane and hexane (1:1 v/v). Y i e l d based on H2TPP = 86%. Elemental analysis for C H 4H 2 8N l +FeCl cal c u l a t e d : C 75.06; H 4.01; N 7.96 found : C 73.96; H 3.91; N 7.71 UV/Vis spectrum: CH 2C1 2 solvent, X (e, M^cm - 1) 501 nm (e = 1.04 x 101*); 418 nm (e = 3.02 x 10 4). Ext i n c t i o n c o e f f i c i e n t s measurements are i n CH 2C1 2 since t h i s solvent i s used i n the decarbonylation reactions (Chapter V). Ex t i n c t i o n c o e f f i c i e n t s of Fe***(TPP)Cl i n benzene agree well with those given i n the l i t e r a t u r e [46]. Mass spectrum: (m/e) FeTPP +(668); FeTPP 2 +(334) [ F e I I I ( T P P ) ] 2 0 was obtained by shaking a benzene sol u t i o n of Fe l : t I ( T P P ) C l ( l g , 10 mL) with aqueous KOH (30 mL) [47]. The r e s u l t i n g benzene s o l u t i o n was evaporated to dryness. The residue was dissolved i n 10 mL CH 2C1 2 and then methanol (15 mL) was added. The product was r e c r y s t a l l i z e d twice from a mixture of 1,2 dichloroethane and n-hexane (1:1 v/v). Y i e l d based on F e m ( T P P ) C l = 93%. 29 Elemental analysis for C88H56Ng0Fe2 calculated: C 78.14; H 4.14; N 8.28 found: C 77.40; H 4.25; N 8.25 UV/vis spectrum: CH2C12 solvent, ^ ^ ( E . M~1cm~1) 571 nm (e - 3.71 x 103); 403 nm (E - 3.63 x I0k) Extinction coefficients of [Fe***(TPP)]20 in benzene agree well with those given in the literature. [47]. Mass spectrum: (m/e) FeTPP+(668); FeTPP2+(334) Fe I X(TPP)(pip) 2 [48,49] Piperidine (~ 4 mmoles) was added to a refluxing solution of Fe I I : t(TPP)Cl (1.2 mmole) in CH2C12 (100 mL) under argon . After evaporating half the CH2C12 solution, methanol (30 mL) was added dropwise. The deep blue crystals obtained on filtration were washed with methanol, ether and dried in air. Yield based on FeII:t(TPP)Cl « 92%. Elemental analysis for Cm +H 2 8N 4Fe»2C 5H 1 1N calculated: C 77.25; H 5.96; N 10.02 found: C 77.09; H 6.10; N 10.13 UV/vis spectrum: Piperidine solvent. ^ n m( e» M^cm-*) 655 nm (e - .4.91 x 102); 525 nm (e = 1.63 x 103) 420 nm ( e - 7.10 x 103) Mass spectrum: (m/e) FeTPP+(668); FeTPP2+(334) FeIi;TPP(PPh3)2 To Fe I I I(TPP)Cl (~ 2 x 10~2 mmoles) dissolved in 20 mL CH2C12 under 30 argon was added 4 molar equivalents of triphenylphosphine (0.08 mmoles) and 2 molar equivalents of Na 2S 20 4 (dissolved in degassed water). Addition of a reducing agent like Na2S20^ or NaBH^  was necessary for the reaction to proceed. The reaction mixture was stirred and mildly heated (~ 30°C) for 30 mins. The organic layer was syringed out and evaporated to a third of its original volume, and methanol then added. The compound crystallised out and was dried under vacuum. Yield based on Fe***(TPP)Cl = 61%. Elemental analysis for CggHjgN^FePj* — CH30H 4 calculated: C 80.28; H 4.91; N 4.66 found: C 79.39; H 5.66; N 3.84 UV/vis spectrum: CH2C12 solvent, added PPh3 to prevent dissociation. X (e, M-lcm-1) nm 590 nm (e = 1.6 x 101*); 550 nm (e = 2.1 x lO 4); 450 nm ( E = 2.6 x 105) Mass Spectrum: (m/e) FeTPP+(668); FeTPP2+(334); CH30+(31) FeI]:TPP(n-Bu3P)2 To Fe i : t I(TPP)Cl (~ 0.2 mmoles) in 30 mL CH2C12 under argon was added n-Bu3P (3 ml) and the solution was refluxed for 20 min. No reducing agent was needed since n-Bu3P itself reduces the Fe(III) porphyrin [49,50]. The volume was reduced to a third by bubbling argon through the reaction mixture, and methanol was added. The lustrous purple crystals were washed with methanol and dried under argon. Yield based on Fe I I I(TPP)Cl = 96%. Elemental analysis for C68H82Ni+FeP2 calculated: C 76.14; H 7.65; N 5.22 found: C 75.86; H 7.70; N 5.14 31 UV/vis spectrum: CH2C12 solvent; X (e, M_1cm_1), spectrum is independentof added phosphine. 358 nm (e = 4.88 x 101*); 455 nm (e = 1.286 x 105); 603 nm (2.34 x 101*) These extinction coefficients agree with those in the literature [50]. Mass spectrum: (m/e) FeTPP+(668); FeTPP2+ (334) NMR nmr, C 6D &, in vacuo, TMS external standard. -0.87 (br, CH2CH2P, 12H), -2.0 (br, CH2P, 12H), 0.5 (t, CH2CH3, 18H), 0.6(t, CH2, 12H), 7.5-7.6 (m, m-H, p-H, 12H), 8.32 (m, o-H, 8H), 8.74 (s, pyrrole-H, 8H) 3 1P nmr, CgD6. singlet, 5.65 ppm with reference to H3P01+. The crystal structure is described in Chapter III and is the first reported for an iron porphyrin complex containing a Fe-P bond. Inactive Species Solutions of FeIITPP(n-Bu3P)2 in CH2C12 (~ 10~2 M) were exposed to oxygen and the oxidised species was isolated by reducing the volume of CH2C12 by a third and adding hexane (10 mL). The compound thus precipitated was recrystallised twice from CH2Cl2/hexane (1:1 v/v). Elemental analysis obtained: C 73.03; H 4.30; N 7.42 UV/visible spectrum: CH0C1, solvent; \ (e, M~1cm~1), 415nm (e = 9.01 x *• nm 10 \ based on FeI1:TPP(n-Bu3P)2) Mass spectrum: (m/e) FeTPP+(668); FeTPP2+(334) 32 II.8 Program for Gas Chromatographic Analysis Two programs were found useful in identifying the components present in the decarbonylation reaction. They are given along with the standard retention times (Table II-l;II-2) determined by using authentic compounds. Program 1. Hewlett-Packard 5830A instrument equipped with a thermal conductivity detector using helium carrier gas flow. The column used was a 6 ft * 0.125 in, 6% 0V-101 (methyl-silicone support) packed column. Temperature program: 60°C/5 min, increase temperature by 5°C/min to 180°. Maintain at 180°C for 5 min. Time (min) Table II.1 GC retention time (min) for standards. STANDARD RETENTION TIME(min) CH2C12 1.09 Benzene 2.16 Toluene 4.32 Benzaldehyde 12.18 Phenylacetaldehyde 15.36 33 Program 2: Hewlett-Packard 5830A, 0V17 column instrument equipped with a thermal conductivity detector using helium carrier gas flow. The column used was a 6 ft x 0.125 in, 6% 0V-17 (methyl-phenyl-silicone support) packed column. Temperature program: 80°C/5 min, increase temperature by 5°/min to 220°. Maintain at 220°C for 5 min. Time (min) Table II.2 GC retention time (min) for standards. STANDARD RETENTION TIME(min) CH2C12 1.03 Benzene 1.93 Toluene 4.09 Benzaldehyde 11.98 Phenylacetaldehyde 14.64 34 II.8.1 Turnover numbers The GC area integrator gave a constant area count for a fixed amount of aldehyde (whether neat or diluted in CHjd^)* The decarbonylation reaction mixtures, being well stirred, were assumed to be homogeneous. The G.C. area count was thus a direct measure of the amount of aldehyde present in the reaction mixture of known volume. This method of extrapolating the GC area count was confirmed by testing it for 6 different aldehydes in CH2C12« In each case, the extrapolated amount was found equal to the known amount of aldehyde present in solution. moles of aldehyde consumed in 1 minute Turnover min - 1 - ( ) moles of porphyrin 35 Chapter III STRUCTURE OF FeTPP(n-Bu3P)2 III.l Structural Analysis The X-ray analysis of the title compound was performed by Dr. S. Rettig of this department. Crystals of FeTPP(n-Bu3P)2 are tric l i n i c , £ = 12.499(3), b - 12.528(2), c_= 12.039(2)A, a - 116.30(1), p - 109.79(1), x -98.13(1)°, _z = 1, space group PI. The structure was solved by conventional heavy-atom methods and was refined by full-matrix least-squares procedures to R - 0.060 and Rw = 0.070 for 3551 reflections with 1^  > 36(_I) collected at 22°C with Mo Ka radiation on an Enraf-Nonius CAD4-F diffractometer. Hydrogen atoms were fixed in idealised positions. The ri-butyl groups a l l display relatively large degrees of thermal motion, especially the C(27)-C(30) group. Anomalous geometric parameters involving these atoms result from thermal motion and/or unresolved disorder. The structure of FeTPP(n-Bu3P)2 is shown in Figure III.l. Bond lengths ( A ) and bond angles (Deg) with estimated standard deviations are displayed in Table III.l and III.2, respectively. 36 ure I I I . l Stereoscopic view of FeTPP(n-Bu 3P) 2 37 Table III.l Bond Lengths (A) with estimated standard deviations in parentheses Bond Length( A) Bond Length( A) Fe - P ( 1 ) 2 . 3 4 5 7 ( 1 1 ) C O D - C ( 1 2 ) 1 . 3 5 8 ( 6 ) Fe - N ( 1 ) 1 . 9 9 8 ( 3 ) C O D - C O 6 ) 1 . 3 6 4 ( 6 ) Fe - N ( 2 ) 1 . 9 9 3 ( 3 ) C ( 1 2 ) - C ( 1 3 ) 1 . 4 0 0 ( 7 ) P(1 ) - C ( 2 3 ) 1 . 8 0 9 ( 5 ) C ( 1 3 ) - C O 4 ) 1 . 3 5 4 ( 9 ) P ( 1 ) - C ( 2 7 ) 1 . 8 8 5 ( 6 ) C O 4 ) - C ( 1 5 ) 1 . 3 3 3 ( 9 ) P ( 1 ) - C ( 3 1 ) 1 . 8 2 3 ( 5 ) C ( 1 5 ) - C ( 1 6 ) 1 . 3 7 4 ( 8 ) N ( 1 ) - C O ) 1 . 3 7 6 ( 5 ) C ( 1 7 ) - C 0 8 ) 1 . 3 7 3 ( 6 ) N ( 1 ) - C U ) 1 . 3 7 2 ( 5 ) C ( 1 7 ) - C ( 2 2 ) 1 . 3 7 4 ( 6 ) N ( 2 ] - C ( 6 ) 1 . 3 8 4 ( 5 ) C O S ) - C ( 1 9 ) 1 . 3 8 9 ( 7 ) N ( 2 ) - C O ) 1 . 3 8 3 ( 5 ) C ( 1 9 ) - C ( 2 0 ) 1 . 3 7 7 ( 8 ) C O ) - C ( 2 ) 1 . 4 3 1 ( 5 ) C ( 2 0 ] - C ( 2 D 1 . 3 4 8 ( 8 ) C(1 ] - C O O ) ' 1 . 3 9 7 ( 5 ) C ( 2 1 I - C ( 2 2 ) 1 . 3 8 8 ( 7 ) C ( 2 ] - C ( 3 ) 1 . 3 4 3 ( 5 ) C ( 2 3 - C ( 2 4 ) 1 . 5 2 8 ( 6 ) C ( 3 ] - C ( 4 ) 1 . 4 4 0 ( 5 ) C ( 2 4 > - C ( 2 5 ) 1 . 4 7 6 ( 7 ) C ( 4 ) - C ( 5 ) 1 . 4 0 0 ( 5 ) C ( 2 5 l - C ( 2 6 ) 1 . 5 2 3 ( 8 ) C ( 5 ) - C ( 6 ) 1 . 3 9 5 ( 6 ) C ( 2 7 - C ( 2 8 ) 1 . 3 9 5 ( 1 1 ) C ( 5 1 - C O D 1 . 4 9 7 ( 5 ) C ( 2 8 - C ( 2 9 ) 1 . 7 1 8 ( 1 4 ) C ( 6 J - C ( 7 ) 1 . 4 2 8 ( 6 ) C ( 2 9 I - C O O ) 1 . 2 6 2 ( 1 4 ) C ( 7 l - C ( 8 ) 1 . 3 2 6 ( 6 ) C ( 3 1 l - C ( 3 2 ) 1 . 4 3 9 ( 7 ) C ( 8 - C O ) 1 . 4 3 3 ( 6 ) C ( 3 2 ) - C ( 3 3 ) 1 . 5 5 5 ( 8 ) C ( 9 l - C O O ) 1 . 3 9 5 ( 5 ) C ( 3 3 ) - C ( 3 4 ) 1 . 4 1 7 ( 1 0 ) C ( 1 0 ) - C ( 1 7 ) 1 . 5 0 0 ( 5 ) 38 Table III.2 Bond angles (deg) with estimated standard deviations in parentheses Bonds An g l e ( d e g ) Bonds Angle(deg) p(1) -Fe -NO ) 90.20(9) C( 5 ) - C ( 6 ) - C ( 7 ) 124.71 4) p(1) -Fe -N(2) 90.42(9) C( 6 ) - C ( 7 ) - C ( 8 ) 107.91 4) p(1) -Fe - P ( 1 ) 1 180 C( 7 ) - C ( 8 ) - C ( 9 ) 107.21 4) p(1) -Fe - N O ) ' 89.80(9) N< 2 ) - C ( 9 ) - C ( 8 ) 110.3 4) p(1) -Fe - N ( 2 ) ' 89.58(9) N( 2 ) - C ( 9 ) - C O 0 ) 124.8 4) N( 1 ) -Fe -N(2) 90.16(11) C( 8 ) - C (9)-C(10) 124.9 4) N{ 1 ) -Fe - N O ) ' 180 C( 9 ) - C(10)-C(17) 118.1 3) N( 1 ) -Fe - N ( 2 ) ' 89.84(11) ' CI 9 ) - C ( 1 0 ) - C ( 1 ) ' 124.21 3) N(2) -Fe - N ( 2 ) ' 180 CI 17) - C ( 1 0 ) - C ( 1 ) ' 117.71 3) Fe -P(1 ] -C(23) 1 1 6 . 4 8 0 5 ) CI 5 ) - C O 1 )-C(l2) 122.31 4) Fe -P(1 ] -C(27) 113.4(2) CI 5 ) - C(11)-C(l6) 119.91 4) Fe -P(1 ] -C(31) 1 17.4(2) CI 12) -CO 1 )-CO 6) 117.8 4 ) C(23)-P(1 )-C(27) 102.5(2) c 1 1 ) -C(12)-C(13) 121.2 5) C(23)-P(1 )-C(31) 108.2(2) c 12) -C(13)-C(14) 119.2 6) C(27)-P(1 )-C(31) 96.0(3) c 13) -C(14)-C(15) 119.8 5) Fe -N( 1 I -CO ) 127.4(2) CI 14) -CO 5)-C( 16) 121.2 6) Fe -N( 1 1 -C(4) 127.4(2) c 1 1 ) -C(16)-C(15) 120.81 5) C(1 ) -N( 1 -C(4) 105.0(3) c 10) -C(17)-C(18) 119.81 4) Fe -N(2] -C(6) 127.5(3) c 10) -C(17)-C(22) 122.51 4) Fe -N(2] -C(9) 128.0(3) c 18) -C(17)-C(22) 117.7 4) C(6) -N(2] -C(9) 104.5(3) c 17) -C(18)-C(19) 121.0 ,5) N( 1 ) -C(1 ! -C(2) 110.4(3) c 18) - C O 9 ) - C ( 2 0 ) 120. 1 i5) N( 1 ) - c o ] -COO) ' 125.6(3) c 19) -C(20)-C(21) 119.3 ,5) C(2) - c o : -COO) ' 124.0(4) c ,20) -C(21)-C(22) 120.5 r5) C(1 ) -C(2] -C(3) 107.4(4) c h7) -C(22)-C(21) 121.3 ,5) C(2) -C(3] -C(4) 106.5(4) P 1 )- C(23)-C(24) 119.6 3) N( 1 ) -C(4, -C(3) 110.7(3) c r23) -C(24)-C(25) 114.2 r5) N( 1 ) -C(4> >-C(5) 125.6(3) c (24) -C(25)-C(26) 114.7 (6) C(3) -C(4] »-C(5) 123.7(3) p (1 )-C(27)-C(28) 119.2 (5) C(4) -C(5, >-C(6) 124.1(3) c (27) -C(28)-C(29) 119.2 (9) C(4) -C(5 I-CO 1 ) 118.4(4) c (28) -C(29)-C(30) 106.7 (10) C(6) -C(5 »-C(11 ) 117.5(4) p (1 )-C(31)-C(32) 117.9 (4) N(2) -C(6 l-C(5) 125.3(4) c (31 ) -C(32)-C(33) 113.7 (6) N(2) -C(6 l-C(7) 110.1(4) c (32) -C(33)-C(34) 117.5 (7) 39 III.2 X-ray structural analysis of metalloporphyrins Biologically important compounds like hemes, cytochromes and some enzymes contain at their active site an iron porphyrin surrounded by a protein which provides axial ligand(s) to the fifth (and sixth) coordination site(s) above and below the porphyrin plane. Variation of ligand(s) leads to definite structural changes [51] which in turn result in a specific physiological function for that structure. For example, cytochrome, involved in electron transfer, is an iron porphyrin which is axially ligated with histidine and methionine provided by the surrounding protein, whereas for hemoglobin, involved in oxygen binding, the protein provides an axial histidine ligand to the iron porphyrin and protects the vacant sixth coordination site. The relationship between the axial ligand field strength, the spin state of the metal ion and stereochemistry of the porphyrin macrocycle, first recognized by Williams [52] and by Hoard [53(a,b)], means that X-ray structural data for model protein-free iron porphyrin systems can be used to predict the structure of corresponding complex natural systems. The choice of tetraphenylporphyrin (TPP) and octaethylporphyrin (OEP) as model porphyrin systems has been rationalized [51,54] on the basis that peripherial substituents on the porphyrin play a minor role in active site chemistry, whereas the axial ligands on the porphyrin play an important role in heme reactivity. Two examples involving the use of model bond distances to study natural enzyme systems 40 are given later in Section III.5 by way of illustration. Structural parameters for porphyrin are: (i) N-Ct: The radius defined by a pyrrole nitrogen to the porphyrin centre; the radius indicates the flexibility of the porphyrin ligand. N-Ct is restricted by the macrocyclic porphyrin, except in cases of high-spin 6-coordinated metalloporphyrins where the metal, being in-plane with the macrocyclic ring, is accommodated by a radial expansion of the porphryin core. (ii ) M-Np: The metal-porphinato nitrogen bond distance. ( i i i ) AN^ : The displacement of metal with respect to the mean plane of the 4 pyrrole nitrogens (Table III.3) (iv) A Core: This represents metal displacement with respect to the 24 atom core. (v) M-L: The metal-axial ligand bond distance. Of these parameters, the M-Np and M-L bond distances, being important determinants of porphyrin structure, will be discussed in this chapter with reference to FeTPP(n-Bu3P)2' III.3 Characteristics of the Fe-Np bond distance For tetraphenylporphyrin, the undistorted N-Ct distance of 2.01A [56] can accommodate low-spin Fe(II) or Fe(III) without strain or distortion. The high-spin Fe(II) or Fe(III) ions have an electron in the d 2 2 orbital (Table III.4), which increases the Fe-Np bond distance, x —y This increase is accommodated by an out-of-plane displacement of the Fe for 41 Table III.3 AN4 values reported for iron porphyrin Complexes [51] d 6 configuration low (S=0) high (S=2) interned. (S=l) Fe(II) 2 2 _x -y * -z 2 + x 2-y 2 i z 2 2 2 -x z-y z t l +4-xz,yz t + txz,yz + txz,yz t4-z2 tixy -f-ixy t+xy coord, no. 5 or 6 5 or 6 4 AN^ , A 0.21-0.23 0.0-0.11 0.45 0.00 0.00 42 Table III.3 (cont'd) d 5 configuration Fe(III) Coord. No AN4, A low (S = l / 2) high (S = 5 / 2) admixed interned. ( S = 3 / 2 , 5 ' 2 ) interned. ( s = 3 / 2 ) 2 2 -x -y + x 2-y 2 T x 2-y 2 2 2 -x z-y ^  -z 2 + z 2 + z 2 t z 2 -t-4- +xz,yz + ixz,yz t +xz,yz t -t-xz,yz +4-xy + xy •i + xy •Hxy 6 5 or 6 5 6 0.0 - 0.9 [0.47] 0.0 0.26 - 0.28 0.0 43 Table III.4 Fe-Np bond distances in Fe-tetraphenylporphyrins [55(a),57] oxidation state spin state 5 coordinated 6 coordinated high-spin 2.085 A 2.060 A Fe(II) low-spin 2.001 A 1.997 A high-spin 2.069 A 2.045 A Fe(III) low-spin 2.015 A 1.986 A 5-coordinated species and a radial expansion of the porphinato core in case of 6-coordinated species; in both cases the following trends are observed: (Table III.4) (i) Fe-Np bond distances in 5-coordinate Fe(II) or Fe(III) porphyrin species are longer than those in corresponding 6-coordinated species. (ii) Fe-Np bond distances in high-spin Fe(II) or Fe(III) porphyrin species are longer than those in corresponding low-spin species. This is due to high-spin metal ion, being larger than low-spin metal Ion, tending to be out-of-plane in the direction of the axial ligand; this motion is defined as AN^  (Section III.2 ) . Frequently, the porphyrin atoms are displaced towards the axial ligand, this resulting in 'doming' defined as A Core - ANU. The increase in M-Np bond distance on changing the spin 44 state is general for metalloporphyrin and a typical example is given below [55(a)]: Low-spin d 8 Co[(TPP)(l-MeIm)] has AN^  = 0.14A; M-Np = 1.977(3)A [55c] and high-spin d 5 Mn[(TPP)(l-MeIm)] has AN^  = 0.56A; M-Np - 2.128(7)A [55c]. Table III.4 illustrates the Fe-Np bond distance for different cases of spin state, oxidation state and coordination number. Values for intermediate spin complexes are expected to be intermediate between those of corresponding high and low-spin species [55(a)]. Values of Fe-Np (1.995 A) and ANi+(0) for FeTPP(n-Bu3P)2 (Tables III.l and III.3) compare favorably with those expected for low-spin 6-coordinate Fe(II) porphyrin species (Table III.4 and III.3, respectively). 45 III.4 Characteristics of M-L bond distance III.4.1 Steric interactions [60] Steric interactions between ligand atoms and atoms of the porphyrin core set a lower limit on the metal-axial ligand (M-L) bond distance (Figure III.2) Figure III.2 [60] A diagram illustrating the steric interactions of an axial ligand of a metalloporphyrin with the porphinato core. The dihedral angle is between the plane of the ligand and the plane defined by a porphinato nitrogen atom, the metal atom and the ligand nitrogen atom (N^). [The diagram is for Fe(TPP)(NCS)(py).] 46 Maximum steric hindrance occurs when H is directly above the nitrogen atom. Steric hindrance is diminished by a more favorable orientation of the ligand and/or ruffling of the porphinato core. For the above illustration (Figure III.2), the H ...N contact is considerably smaller than the normal packing distance (2.9 A ) . Packing contacts as l i t t l e as 2.5 A occur in metalloporphrins, Implying greater strain in the systems. [Normal packing distance is defined as the sum of the Van der Waals radii of hydrogen ( 1 . 2 A) and half thickness of an aromatic ring ( 1 . 7 A ) ] . Smaller internal angles of 5-membered rings of axial ligands increase the perpendicular distance between the ligand ring hydrogen atoms and the porphyrin core; hence, imidazole will have a larger H ...N contact distance than an 6-membered ring like pyridine, and a saturated ring like piperidine will have the greatest steric hindrance and the longest M-L bond. These effects are illustrated in the following series [55(a)] of Co(III) complexes (Table III.3). 47 Table III.5 Compound M-L bond distance Reference [Co(TPP)(RTm)2]+ 1.93(2) A [61] Co(TPP)(py)(Cl) 1.978(8) A [62] Co(TPP)(py)(N02) 2.036(4) A [63] [Co(TPP)(pip)2]+ 2.060(3) A [64] For the case of FeTPP(n-Bu3P)2 (Figure III.l), the closest distance between a hydrogen on the Bu3P chain and the porphyrin in some cases was found to be less than 2.9 A (the normal packing distance). The hydrogens on carbons 27, 28, 31 and 32 were 2.5, 2.49, 2.71 and 2.44 A , respectively from the porphyrin; thus steric hindrance does play a role in the increased bond distance of 2.347(11) A found for Fe-P as compared with the standard Fe-P bond distance of 2.243 A [58]. 48 III.4.2 Degree of Tt-backbondlng The single-bond covalent radii bond distance for a Fe-P a bond is 2.243 A [58]; there may also exist a it-backbond between Fe and P. In general, the degree of it-bonding in a metal complex depends oh valency state and coordination number of the metal cation, isomeric form, electronegativity effects from neighbouring atoms, and packing considerations [58]. Steric hindrances which prevent the axial ligand from approaching the metal atom also influence the bond length (see Section III.2.1). The collective effect of these factors, in the case of the Fe-P bond, is reflected in the large variations of the Fe-P bond distances reported [58] (2.13-2.36 A ) . Comparison of the Fe-P (terminal) bond distance of 2.35 A observed for Fe^TPP(n-Bu3P) 2 shows i t to be longer than that reported for Fe 3(CO) uPPh 3 [58a] (Fe-P(terminal) - 2.25 A) or cis-FeI2(CO) 2(PH 3) 2 [58(b)] (Fe-P (terminal) - 2.27 A ) . Since the two trans phosphines are good it-acceptors which compete for the metal d-orbitals, the decreased it-bonding could result in longer Fe-P bond distances. This effect, plus the differences in oxidation state (greater it-bonding with Fe°), and combined with steric factors discussed earlier (Section III.4.1), may account for the longer Fe-P bond distance. 49 III.4.3 Spin-state of metal ion Literature data show that changing the spin-state of the metal ion results in large differences in Fe-P bond distances. X-ray structural determination of the compound FeCl2(dppen)2»2(CH3)2CO [65] shows Fe-P (average) to be 2.584 A for the high-spin (295°K) but only 2.3005 A for the low-spin form (130°K) [dppen = cis-l,2-bis(diphenylphosphino)ethylene, (Ph2PCH=CHPPh2)2]• The high-spin complex FeCl2[(Ph2PCH2CH2)2PPh]2»2(CH3)2CO [66] has Fe-P distances in the range of 2.66-2.71 A , whereas the Fe-P distances are in the range of 2.23-2.24 A for the low-spin complex. The longer bond distances in the high-spin complexes were attributed to ligand rigidity, non-bonding interactions and packing effects in the complexes. The Fe-P bond length in FeTPP(n-Bu3P)2 (2.35 A) (Table III.l) compares favorably with the low-spin form of the above two compounds• For the case of metalloporphyrins, populating the d^2 orbital results In a significant lengthening of the M-L bond distance [67]. Thus, an increase of 0.163 A in the Fe-N (axial) bond distance for [Fe(0EP)(3-Clpy)2](C10l4) is observed on going from the low-spin form (98°K) to the high-spin form (293°K). Increases in M-L distances on going from low-spin to high-spin state have been reported [68] for Cr(II), Mn(III) and Co(II) porphyrin systems. 50 III.5 Model Fe-L(axial) bond distances of 6 coordinate low-spin Fe(II)  tetraphenylporphyrin complexes The Fe(II)-N (axial) bond distance of 2.014(5) A observed in FeTPP(l-MeIm)2 [69] is that expected for low-spin species. Severe steric hindrance between the piperidine a-hydrogen atoms and porphyrin core results in the long Fe-N (axial) bond distance of 2.127(3) A reported for FeTPP(pip)2 [70]. The Fe-S (axial) bond distances reported for a range of spin states, oxidation states and coordination number of iron, and variation in the nature of the sulfur ligand, are within a narrow range. Thus, for 6-coordinate low-spin thioether derivatives, Fe(III)-S (2.34 A) [71(a)] and Fe(III)-S (2.33, 2.35 A) [71a] were reported. The values for high-spin 5-coordinate thiolate complexes, Fe(II)-S (2.360 A) [72] and Fe(III)-S (2.324 A) [73] are s t i l l quite close. Changing the spin-state and coordination number for Fe(II) thiolate does not affect the Fe-S bond distance (high-spin complex Fe(II)-S = 2.360(2) A, low-spin complex Fe(II)-S = 2.352(2) A). The data suggest that hemoproteins containing sulfur ligands have Fe-S bond distances in the same corresponding range (2.32-2.36 A), and the use of model Fe-L (axial) and Fe-Np bond distances should assist in the characterization of hemoproteins. An EXAFS study [74] on the heme iron site of bacterial cytochrome P-450 has reported Fe-S bond lengths for low-spin ferric, high-spin ferric, and ferrous and ferric carbonyl states of the enzyme. By comparing Fe-Np and Fe-S bond distances 51 with those of a model system, the presence of a thiolate sulfur donor in each of the P-450 stages examined has been suggested. A X-ray study of chloroperoxidase (CPO) [57] has, on the basis of Fe-S bond distances and AN^  values (see Table III.3), suggested thiolate ligation for the resting state of the enzyme. The Fe(II)-C bond distance of 1.90 A for FeTPP(t-BuNC)2 [75] complex is abnormally long, crystal packing effects being considered responsible for the nonlinear Fe-N-C angle reported for this compound; the Fe-N-C angle Is expected to be linear In the corresponding isocyanide hemoprotein adduct. A normal Fe-C bond distance (1.78 A) is reported for FeTPP(SC2H5)(C0) [72] and for Fe(TPP)(py)(C0) [76] . (Fe-C = 1.77 A ) . The CO groups in these model compounds are linear, unlike those for hemoproteins, where the CO groups are tilted due to steric hindrance. The resulting lowered CO affinity in the hemo-proteins is believed to retard CO poisoning [77]. Dioxygen complexes of Fe using 'picket-fence' porphyrins [78] to prevent dimerization to u-oxo species have yielded Fe(II)-0 bond distances in the range of 1.745(18) A-l.898(7) A . Although these measurements were hampered by large thermal motion of the picket fence atoms, they do provide an estimate of expected Fe-0 bond distances for oxygenated heme. For example, an Fe-0 bond distance of 1.83(6) A has been reported for oxymyoglobin [79]• The Fe(II)-P bond distance of 2.3457(11) A in FeTPP(n-Bu3P)2 is longer than that for the single Fe-P bond distance (~ 2.243 A ) . The 52 Increase being attributed to steric hindrance and diminished n-backbonding. III.6 Conclusion The title compound is the first example of a crystallographically determined Fe-P bond length within a metalloporphyrin. Phosphines play an important role in transition-metal chemistry by helping to stabilize low oxidation states. Because of their high field strength, phosphines are used as co-ligands to stabilise hydride, alkyl, aryl and CO complexes of transition metals especially group VIII elements. These properties have been utilised in catalysis using transition-metal complexes [80]. The potential of metalloporphyrin phosphine complexes for catalysis remains to be fully explored; the findings described in this thesis suggest considerable scope via free radical, atom transfer processes (see Chapter V). 53 CHAPTER IV THE REACTIONS OF FeTPP(n-Bu 3P) 2 WITH CO AND 0 2 IV.1 Reaction with CO Toluene was used as solvent to study the t i t l e reaction as previous studies on related systems [1(b),35,81] had used the same solvent. On addition of one atmosphere of CO to a so l u t i o n of FeTPP(n-Bu 3P) 2 (~ 10"1* M) (Figure IV.1), an instantaneous spectral change was observed and the reaction went to completion to form FeTPP(CO)(n-Bu 3P) as judged by the v i s i b l e s p e c t r a l change (Figure IV.2) and i . r . s o l u t i o n data ( i n CH 2C1 2, ~ 10~ 3 M v = 1993 cm - 1). The reaction proceeds according to the following equilibrium (eq. IV.1) K FeTPP(n-Bu 3P) 2 + CO ^- FeTPP(CO)(n-Bu 3P) + n-Bu3P ^ (eq. IV.1) In the absence of added phosphine, the p o s i t i o n of the forward reaction i s independent of the CO pressure used between 0 . 5 - 1 atmosphere, i e . the reaction i s complete. Excess phosphine i s not necessary to generate the correct spectrum for the precursor 6-coordinate FeTPP(n-Bu 3P) 2 species, i n d i c a t i n g the phosphine complex remains 6-coordinated. This species has the Soret band 54 at 455 nm (e = i . 19 * 10 5M - 1cm - 1). The spectrum of FeTPP(CO)(n-Bu 3P) has the Soret band at 438 nm (e = 2.23x 105 M - 1cm - 1). The ext i n c t i o n c o e f f i c i e n t for the l a t t e r compound i s determined from spectral data using one atmosphere of CO, when the carbonyl complex i s found to be f u l l y formed. The equilibrium was studied by f i r s t forming the carbonyl complex. Successive additions of n-Bu3P regenerated the bis-phosphine species. The reaction was studied by observing the changes i n absorbance at 455 nm (Figure IV.3). The equilibrium constant K for eq. IV.1 i s [FeTPP(CO)(n-Bu 3P)][n-Bu 3P] K = (eq. IV.2) [FeTPP(n-Bu 3P) 2][CO] which can be written as (Ae - A<= ) Log K = Log [ ] + log [n-Bu 3P] - log [CO] (Ao - Ae) (eq. IV.3) where A Q - absorbance of the FeTPP(n-Bu 3P) 2 species at 455 nm, = absorbance of the carbonyl product at 455 nm, estimated by ex t i n c t i o n c o e f f i c i e n t s , and Ae = absorbance at equilibrium using various added phosphine concentrations at a fixed CO pressure (1 atm). Rearranging eq. IV.3 gives 55 (Ao - Ae) Log [ ] = Log [n-Bu3P] - log [CO] + log K (Ae - A ) (eq. IV.4) (Ao - Ae) A plot of log [ ] versus log [n-Bu3P] should give a straight (Ae - A ) CO line of slope + 1 (Figure IV.4). (Ao - Ae) The value of K is obtained from the graph at log [ ] = 0 (Ae - A») [n-Bu3P] This gives K = for each temperature. [CO] Values of [n-Bu3P] obtained at various temperatures are listed in Table IV.2. Solubility of CO In toluene [82] between 10-40°C, corrected for vapour pressure of toluene [83] is listed in Table IV.3. The K equilibrium values obtained over a range of 18-40°C are presented in Table IV.4. A Vant Hoff plot (Figure IV.5) using the data in Table IV.4 yielded the thermodynamic data AH (9.3 ± 2 KJ/mole) and AS (28 ± 8 J/mole). The low value for AH (~ 9 KJ/mole) suggests that the Fe-CO and Fe-(n-Bu3P) bond strengths are similar. Equilibrium data for reaction of Fe and Ru porphyrin and related systems with CO in toluene solvent are presented in Table IV.5. The K equilibrium values in Table IV.5 reflect the fact that TPP, being less basic [84] than OEP, prefers n-Bu3P, a stronger a donor than CO, 56 hence the higher affinity for CO binding by RuOEP(n-Bu3P)2 versus RuTPP(n-Bu3P)2. Comparing the K values obtained for FeTPP(n-Bu3P)2 with that reported for RuTPP(n-Bu3P)2, Fe binds CO or n-Bu3P nearly equally whereas Ru binds n-Bu3P more strongly than CO. The ability of the trans phosphine ligand to labilise the coordinated CO can be judged by the low K equilibrium values of complexes 2, 3 and 4 (Table IV.5) when compared to the K values reported for the other bis-ligated porphyrin complexes in Table IV.5. 57 3.0-, Figure IV.1 UV/visible spectra of FeTPP(n-Bu 3P) 2 ( toluene using c e l l of path length 1 cm 58 WAVELENGTH, nm Figure IV.2 UV/visible spectrum of FeTPP(n-Bu3P)2 (~ 2.3 * 1CT5 M) in toluene + CO Gas (0.5 -1 atmosphere). Cell path length is 1 cm. 59 3.0-, WAVELENGTH, nm Figure IV.3 Spectral changes observed for the reaction FeTPP(CO)(n-Bu 3P) + n-Bu 3P •* FeTPP(n-Bu 3P) 2 + CO at 29°C 60 Table IV.1 Data Used to Calculate Equilibrium Constant for the reaction IC FeTPP(n-Bu3P)2 + CO ^ FeTPP(n-Bu3P)(C0) + (n-Bu3P) <sr Spectroscopic Data Ao = Absorbance of FeTPP(n-Bu3P)2 species at 455 nm. A = Absorbance of the carbonyl product at 455 nm. CO Ae = Absorbance at equilibrium using various added phosphine concentrations at a fixed CO pressure (1 atm). Aliquots of n-Bu3P in toluene (2 x 10~5 moles n-Bu3P/10 uL toluene) were used. Log [n-Bu3P] values in Table IV.1 include phosphine that results from dissociation from the precursor bis(phosphine) complex. 61 Table IV.1 (cont'd) Set 1 455 nm Ao = 0.262 A = 1.341 CO Temp = 18°C Obs. No. (Ao-Ae) (Ae-A ) T ,Ao-Ae. L°S Ae-A > CD Log [n-Bu3P] 1 -0.321 -0.758 -0.37316 -2.64174 2 -0.506 -0.573 -0.054004 -2.36838 3 -0.614 -0.465 40.12071 -2.20192 4 -0.686 -0.393 +0.24193 -2.08188 5 -0.798 -0.281 +0.45329 -1.91072 6 -0.871 -0.208 +0.62195 -1.78830 7 -0.924 -0.155 +0.77534 -1.69289 8 -0.943 -0.136 +0.83213 -1.65205 Ao-Ae, Results of least-square plot of Log ( A e_ A ) v s L o8 [n-Bu3P] 1. slope = 1.18 2. correlation =0.99 ,Ao-Ae 3. x - intercept at Log ( A e_ A ) = 0 is -2.32 = Log [n-Bu3P] 62 Table IV.1 (con't) Set 2 455 nm Ao = 0.462 A - 1.986 CD Temp = 23°C Obs. No. (Ao-Ae) (Ae-A ) / T Ao-Ae. ( L°8 Ae-A > CO Log [n-Bu3P] 1 -0.708 -0.816 -0.061656 -2.35457 2 -0.973 -0.551 +0.24696 -2.07468 3 -1.122 -0.402 +0.44577 -1.90588 4 -1.208 -0.316 +0.58239 -1.78463 5 -1.273 -0.251 +0.70515 -1.68994 6 -0.096 -0.190 +0.76105 -1.57807 Results of least-square p l o t of Log ( A e _ A ) versus Log [n-Bu 3P] CO 1. slope = 1.09 2. correlation = 0.99 3. x - intercept at Log (|^~) = 0 is - 2.304 = Log [n-Bu3P] 63 Table IV.1 (con't) Set 3 455 nm Ao = 0.293 A = 1.516 CO Temp = 29°C Obs. No. (Ao-Ae) (Ae-A ) T .Ao-Ae. L°S Ae-A > OO Log [n-Bu3P] 1 -0.548 -0.675 -0.090523 -2.3406 2 -0.767 -0.456 +0.22583 -2.0673 3 -0.893 -0.330 +0.43233 -1.9008 4 -0.973 -0.250 +0.59017 -1.78082 5 -1.046 -0.177 +0.77155 -1.64656 Results of least-square plot of Log ( A e_ A ) versus Log [n-Bu3P] CO 1. slope = 1.16 2. correlation = 0.99 3. x - intercept at Log ( ^ I ^ ) = 0 Is -2.26 = Log [n-Bu3P] 64 Table IV.1 (con't) Set 4 455 nm Ao = 0.380 A = 1.737 GO Temp - 35°C Obs. No. (Ao-Ae) (Ae-A.) T .Ao-Ae. L°S Ae-A > CO Log [n-Bu3P] 1 -0.719 -0.638 +0.051908 -2.1970 2 -0.969 -0.388 +0.39749 -1.90819 3 -1.085 -0.272 +0.60086 -1.73628 4 -1.149 -0.208 +0.74225 -1.61344 Results of least-square plot of Log (^ -_A ) versus Log [n-Bu3P] CO 1. slope = 1.1 2. correlation = 0.99 3. x - intercept at Log (™e_k ) = 0 is -2.24 = Log [n-Bu3P] 65 Table IV.1 (con't) Set 5 455 nm Ao = 0.219 A = 1.038 CO Temp = 40°C Obs. No. (Ao-Ae) (Ae-A ) T ,Ao-Ae. L o g Ae-A > CO Log [n-Bu3P] 1 -0.393 -0.426 -0.035017 -2.22741 2 -0.542 -0.277 +0.29151 -1.96163 3 -0.627 -0.192 +0.51396 -1.79795 4 -0.675 -0.144 +0.67094 -1.65409 5 -0.709 -0.11 +0.80925 -1.55399 Results of least-square plot of Log ( A e_ A ) versus Log [n-Bu3P] CO 1. slope = 1.1 2. correlation = 0.99 A n — A P 3. x - intercept at Log i^e-k ) = 0 is -2.20 = Log [n-Bu3P] 66 00 o log ln-Bu3Pl .Ao-Ae Figure IV.4 Plot of log ( A g _ A ) versus log [n-Bu 3P], 29°C 67 Table IV.2 Values of [n-Bu3P] for log [^°~^e^] = 0, at 18-40°C Temp°C 18 23 29 35 40 [Bu3P] 4.78xl0-3 4.97X10-3 4.94x10-3 5.75x10-3 6.3x10-3 Table IV.3 Solubility of 1 atmosphere of CO in toluene, corrected for vapour pressure of toluene Temp°C 18 23 29 35 40 solubility in moles/litre 7.66x10-3 7.66x10-3 7.66x10-3 7.64x10-3 7.71xl0-3 Table IV.4 Equilibrium constant (K) values for the reaction FeTPP(n-Bu3P)2 + CO K ~» FeTPP(C0)(n-Bu3P) + (n-Bu3P) Temp°C 18 23 29 35 46 K equilibrium 0.625 0.648 0.717 0.753 0.818 68 VAN'T HOFF PLOT Figure IV.5 Plot of Ln K versus 1/Temperature (°Kelvin) 69 Table IV.5 Equilibrium data for reaction of Fe and Ru porphyrin complexes with CO in toluene solvent M(porphyrin)L2 + CO M(porphyrin)(C0)L + L Complex K Temp °C Ref. 1 RuOEP(CH3CN)2 4 x 1014 30 1(b) 2 Ru0EP(n-Bu3P)2 0.677 31 K b ) 3 RuTPP(n-Bu3P)2 0.054 26 [35] 4 FeTPP(n-Bu3P)2 0.65 23 present work 5 FeOMBP(pip)2 2340 23 [81] 6 FeOMBP(py)2 209 23 [81] 7 FePpIX(pip)2 230,000 23 [81] 8 FeTPP(pip)2 150,000 23 [81] IV.1.1 Aldehydes as sources of CO. Aldehydes, both liquid and solid, were also tested as a potential source of CO. The reaction was carried out using an evacuable optical cell (Chapter II, Figure II.1) at 23°C, using CH2C12 solvent, aldehyde (~ 10_3M) and FeTPP(nBu3P)2 (~ 10"1* - 10~5 M). The formation of a carbonyl complex over a period of time (depending on the aldehyde) was detected by UV/visible spectroscopy (Figure IV.6). This spectral change was reversible on vacuum-pumping the optical c e l l . Table IV.6 lists the various aldehydes used and the time required to completely form the CO adduct. 70 Table IV.6 Aldehydes used as source of CO (times required to completely form CO adduct, hours). Reaction carried out in evacuated optical cell (Figure II.1) at 23°C. Carbonyl adduct formation confirmed by UV/visible spectra. Completion indicated by extinction coefficient. LIQUID CH CHO 2 Benzaldehyde Phenylacetaldehyde Salicylaldehyde (4-6) (4-6) (12-14) SOLID N(CH3)2 LOCH, CH30 -OCH, ?CH, p-d ime thylamino-benzaldehyde 3-methoxy, 4 hydroxy-benzaldehyde (16-18) 2,4,6-trimeth-oxyaldehyde (20) p-nitrobenzaldehyde (3 days) 2,6-dichlorobenzaldehyde (3-4 days) piperonal (6-7 days) 71 WAVELENGTH . nm Figure IV.6 The UV/visible spectral changes observed when aldehydes (Table IV.6) were used as sources of CO. The spectrum due to FeTPP(n-Bu3P)2 ( ) changes to that of the carbonyl complex ( ) over a period of time. Spectral changes are readily reversible on vacuum pumping. 72 IV.2 Reaction of FeTPP(n-Bu3P)2 with 0 2 This reaction was not studied in detail but is reported here because of its relevance to the decarbonylation reaction studied in Chapter V. Solutions of FeTPP(n-Bu3P)2 in CH2C12 were extremely air-sensitive and exposure to 02 (or air) resulted in an irreversible 'oxidation' to give a final solution which absorbed at 415 nm (Figure IV.7). The solutions exhibiting the Soret at 415 nm were found inactive for decarbonylation of aldehydes (see Chapter V). Excess phosphine (n-Bu3P) was found to prevent the oxidation of the Fe** species and in an interesting reaction, n-Bu3P was found to reconvert the inactive species (415 nm) to the 'active* species (455 nm) (Figure IV.8). The implication of this reaction was that a slight excess of n-Bu3P should help prevent the deactivation of the porphyrin catalyst by 02. Decarbonylation reactions using excess phosphine did not, however, result in detectable decarbonylation over 3 hours (Chapter V, Section V.3.1). 73 3.0-, WAVELENGTH, nm Figure IV.7 UV/visible spectrum obtained when FeTPP(n-Bu 3P) 2 i n toluene (or CH 2C1 2) i s exposed to 0 2 . 74 3.0 Figure IV.8 UV/visible spectral changes observed on adding n-Bu3P to oxidised 415 species in Figure IV.7 75 CHAPTER V THE CATALYTIC DECARBONYLATION OF ALDEHYDES USING FeTPP(n-Bu3P)2 V.l Choice of FeTPP(n-Bu3P)2 While a decarbonylation catalyst can abstract CO from an aldehyde, the ease of removal of CO from the intermediate metalloporphyrin carbonyl complex is a key consideration in the choice of a suitable catalyst. Preliminary experiments with CO gas and three bisligated iron(II) porphyrins systems in CH2C12 indicated that FeTPP(PPh3)2 and FeTPP(n-Bu3P)2 formed carbonyl complexes readily and that the coordinated CO could be removed on vacuum pumping the optical cell used to monitor these reactions (Chapter IV Section 1.1). Since the loss of coordinated CO appeared more facile in the case of FeTPP(n-Bu3P)2 i t was decided to investigate the decarbonylation reaction with aldehydes using this complex. In the case of FeTPP(pip)2, the formation of the carbonyl complex on reaction with CO gas was instantaneous, as for the two phosphine complexes; however, the loss of coordinated CO from the piperidine carbonyl complex was extremely difficult. The presence of a suitable it-acceptor phosphine ligand clearly labilises the trans-coordinated CO. The reactions with CO were monitored by UV/visible spectroscopy. 76 V.2 Preliminary Reactions with Aldehydes Aldehydes were tested as sources of CO (Table IV.1). Since preliminary r e s u l t s with various aldehydes were encouraging (Table V . l ) , phenylacetaldehyde was chosen to study the decarbonylation reaction in more d e t a i l . P u r i f i e d phenylacetaldehyde was obtained on vacuum d i s t i l l a t i o n and the purity checked by *H nmr. It was necessary to obtain pure aldehyde since impurities were c i t e d as being responsible for side-reactions and eventual destruction of the porphyrin c a t a l y s t [1(b)]. The GC retention times (Tables 11.1,2) for phenylacetaldehyde and the decarbonylation product, toluene, were s u f f i c i e n t l y d i s t i n c t to permit unambigious i d e n t i f i c a t i o n ; also, toluene could be detected r e a d i l y by *H nmr. A t y p i c a l decarbonylation reaction was conducted as follows: aldehyde (ca. 10""2 M) was added to a so l u t i o n of FeTPP(n-Bu 3P) 2 i n CH 2C1 2 (ca. 1 0 - 3 to 10 - 5 M) placed i n a side-armed reaction f l a s k (Chapter II.6, Method 3). The reaction mixture was s t i r r e d , warmed or refluxed i n some cases and continually flushed with argon. Samples were withdrawn using gas-tight syringes and injected into o p t i c a l c e l l s (Figure II.1) which were previously sealed under argon. The reaction mixture exhibited absorption peaks at 455 nm (the bis-phosphine species) only i f the system had been continually flushed with argon, while a spectral peak at 438 nm (the carbonyl species) was observed when argon flushing was discontinued for 20-30 minutes or longer but with the reaction mixture under argon. This 77 438 nm peak would eventually disappear along with the re-formation of the 455 nm peak under argon flushing. Addition of air or oxygen caused an irreversible change of the reaction mixture (dull red —> greenish red) which absorbed at 415 nm (Figure IV.7). Samples withdrawn from reaction mixtures displaying 455 nm or 438 nm peaks or both, usually analysed by GC for toluene and unreacted phenylacetaldehyde. No toluene was ever detected for those reactions which displayed a Soret at 415 nm nor was any loss of initial phenylacetaldehyde detected. Reaction mixtures were analysed by GC and components were identified by either identical retention times with authentic samples (Tables 11.1,2) and/or by GC/MS. GC traces for typical decarbonylation reactions are shown in Figure V.l and V.2. 78 Table V.l Decarbonylation of aldehydes with FeTPP(n-Bu3P)2 in 30 mL CH2C1? at 24°C FeTPP(n-Bu3P)2 (moles) Aldehyde (moles) Ratio aldehyde Decarbonylation product detected on GC (3 hours) Turnover min - 1 UV/VIS spectra porphyrin 3.7 x 10 - 7 1.45 x l O - 4 Benzaldehyde 390 Benzene 0.167 438 nm 455 nm 1.02 x 10 - 5 1.45 x 10 - 3 Benzaldehyde 140 none - 415 nm 8.1 x 10~6 1.27 x 10"3 phenylacet-aldehyde 150 Toluene 17.5 455 nm 8.4 x 10"7 1.29 x I0~h salicyl-aldehyde 150 none 415 nm* 1.02 x 10"5 1.45 x 10"3 piperonal 140 none -* 415 nm Similar results were obtained using the other 5 aldehydes shown in Table IV.6. Trace impurities and/or trace oxygen are probably responsible for the 'oxidised' spectra of 415 nm (see section IV.2). 79 I 1 4 . 6 3 PHENYLACETALDEHYDE S T O P fcp 5 S 3 B f l A R E A 5 i R T fiRER A R E A V. 4 . B 9 2 2 4 3 B 5 1 . 6 9 4 1 4 . 6 3 2 B 9 6 B 4 8 . 3 B 6 X F : l . B B B B E + B Figure V.l G.C. trace for decarbonylation of phenylacetaldehyde (~ 10 - 3 M) using FeTPP(n-Bu3P)2 (~ I0~h M) in CH2C12 (23°C) after 8 minutes. 80 1 S T A R T CH 2C1 2 2B.43 4.32 TOLUENE fc» 583BA AREA ?S RT 4.32 2B.43 AREA 4B36BB 112 X F : AREA V. 99.972 B.B28 l . B B B B E + B Figure V.2 G.C. trace of phenylacetaldehyde (~ 10 - 1 M) with FeTPP(n-Bu3P)2 (~ 10"3 M - lO"4* M) in refluxing CH2C12. Product collected in cold trap. 81 V.3 Factors Influencing Decarbonylation Several factors were studied for their effect on the decarbonyl-ation reaction. V.3.1 Influence of added phosphine Addition of 1, 10 or 40 molar equivalent excess phosphine (n-Bu3P or PPh3) inhibited the decarbonylation reaction completely. No decarbonyl-ation product was detected by GC nor was any aldehyde consumed. The UV/visible spectrum of such a reaction showed a Soret at 455 nm (typical of 6 coordinate Fe(II)TPP(n-Bu3P)2) which did not change on addition of aldehyde; the results obtained for the case of benzaldehyde using no excess phosphine and excess phosphine are given in Table V.2. 82 Table V.2 Effect of excess (n-Bu3P) on decarbonylation reaction in 30 mL CH2C12 at 24°C. 0 FeTPP(n-Bu3P)2 (moles) Benzaldehyde (moles) Product detected on G.C. (1-3 hours) 1 3.7 x 10~7 1.45 x 10-lt Benzene 3.7 x 10~7 1.45 x I0 _ l t none 2 + 1 molar equivalent n-Bu3P 3.7 x 10~7 1.45 x 10"1* none 3 + 10 molar equivalent n-Bu3P 3.7 x 10~7 1.45 x 10_lt none 4 + 40 molar equivalent n-Bu3P Similar results obtained using PPh3. 83 V.3.2 Effect of CO Pretreating the porphyrin in CH2C12 with CO did not affect the rapid decarbonylation reaction; however, flushing the system with CO instead of argon gas resulted in complete inhibition of the reaction. (Table V.3). The UV/visible spectrum of such a reaction mixture displayed the characteristic carbonyl Soret at 438 nm (Figure IV.6). Table V.3 Effect of CO on decarbonylation reaction in 30 mL CH2C12 at 24°C FeTPP(n-BuP)2 (moles) phenylacetaldehyde (moles) ratio of aldehyde to porphyrin turnover min _ 1 8.1 x 10 - 6 1.2 x 10"3 150 0 (if CO flush) 17.5 (if argon • flush) V.2.2 Effect of 0 2 Trace oxygen appeared to initiate the decarbonylation reaction catalysed by RuTPP(phosphine)2, though its exact role is unclear [1(b)]. For the Fe-porphyrin system, excess 0 2 (or air) deactivated the catalyst 84 which could be reconverted to the bis-phosphine species with excess n-Bu3P. This reaction was followed by UV/visible spectroscopy (see Section IV.2). V.3.4 Solvents Benzene, CH2CI2, CH3CN and toluene (in the case of benzaldehyde) were tested. Dichloromethane, a low-boiling solvent (~ 40°C), was found suitable since the porphyrins and aldehydes readily dissolved in i t ; further, the GC retention times of CH2C12 and toluene were sufficiently apart to facilitate identification (Tables 11.1,2). Benzonitrile and acetonitrile were tested since they are reported [3] to stabilise intermediates in decarbonylation reactions and prevent dimerization of the active species; however, no aldehyde decarbonylation was observed using these solvents. Toluene could be considered a suitable solvent for the decarbonylation of benzaldehyde since the GC retention times for toluene and benzene are sufficiently apart; (Tables 11.1,2). The decarbonylation reaction for benzaldehyde occurs smoothly in toluene or CH2C12. Neat aldehydes were also tested as solvents. In the case of phenylacetaldehyde, it was necessary to use extremely high temperatures (>200°C) and longer reaction times (1-3 hours), as compared to CH2C12 solvent system, before decarbonylation products were detected by GC. This was in contrast to the low temperatures required for decarbonylations in CH0C10 solvent. 85 During the decarbonylation of phenylacetaldehyde, whether neat aldehyde or in CH2C12 solvent, a dimer (PhCH2CH2Ph, M.W=182, identified by GC/MS) (Figure V.3) was detected infrequently. Such a product is indicative of a free-radical coupling of two benzyl radicals. 8 6 Figure V.3 Bibenzyl detected during decarbonylation of phenylacetaldehyde using FeTPP(n-Bu3P)2. Bibenzyl was identified by GC/MS. 87 V.3«5 Varying the ratio of aldehyde to porphyrin The ratio of aldehyde to porphyrin was varied from 1 to 600. In some cases neat aldehyde was tested as solvent. No trend was apparent between the decarbonylation rate and the ratio of aldehyde to porphyrin due to inconsistent turnover numbers obtained. In general, the decarbonylation reaction in the temperature range 24-40°C was extremely rapid with detectable product formation within several minutes. In many cases, more than 90% of the aldehyde was consumed within 8 minutes for a l l ratios of aldehyde to porphyrin. All decarbonylation reactions were completed within 3 hours as indicated by no further production of decarbonylated product. Reactions were monitored by GC. The UV/visible spectra of the decarbonylation solutions in many cases s t i l l indicated an 'active' species (455 nm); however, addition of further aldehyde to this 'active' species via gas-tight syringes resulted in the oxidation of the solutions as shown by the disappearance of the 455 nm peak and the appearance of a Soret peak at 415 nm, typical of oxidised, 'inactive' catalyst solutions. Table V.4 shows typical results obtained for two ratios of aldehyde to porphyrin. 88 Table V.4 Effect of varying ratio of aldehyde to porphyrin in 30 ml CH2C12 at 24°C; argon flush FeTPP(n-Bu3P)2 phenylacetaldehyde * ratio of aldehyde turnover min - 1 (moles) (moles) to porphyrin 2.39 x 10~5 1.2 x 10 - 3 50 6.33 1.3 x 10~5 2.6 x 10"3 200 10 inconsistent turnovers obtained using neat aldehydes. V.3.6 Influence of temperature Rapid decarbonylation at ambient temperatures was observed. Increasing the temperature (40°C, refluxing CH2C12) increased the turnover min - 1. A maximum turnover min - 1 of 100 was obtained in refluxing CH2C12 (Table V.5) Table V.5 Decarbonylation of phenylacetaldehyde at 40°C in 30 ml CH2C12; argon flush FeTPP(n-Bu3P)2 phenylacetaldehyde ratio of aldehyde turnover min - 1 (moles) (moles) to porphyrin 4.7 x 10 - 6 1.27 x 10 - 3 270 100 89 V.3.7 Radical inhibitors Hydroquinone and 2,6 di-tert-butyl-p-cresol [1(b)], which can act as radical inhibitors, were used to investigate whether the decarbonylation reactions followed a radical mechanism. No decarbonylation of phenylacetaldehyde occurred when either of the above radical inhibitors was used. The dull red colour of the Fe(II) species (455 nm peak in the UV/visible spectrum) remained unchanged over one week for samples of phenylacetaldehyde (~ 10"3 M), porphyrin (~ 10_l* M), hydroquinone (~ 10-1+ M) In CH2C12 at room temperature sealed in evacuable optical cells (Figure II.1). V.3.8 Rate of gas flushing Rate of decarbonylation was affected by the rate of argon flushing through the system. For a typical decarbonylation reaction studied in CH2C12 at 24°C, under an argon atmosphere, phenylacetaldehyde (~ 10~3 M) and porphyrin (~ 10 - 5 M), turnover numbers of 0.017 min - 1 and 6.3 min - 1 were obtained. The larger number was obtained when argon was continually flushed through the system. (Flow rates were judged by bubbling rates.) Such a mechanistic correlation Is reasonable since any build-up of CO evolved would inhibit the decarbonylation reaction. 90 V.3.9 Control tests A blank reaction containing phenylacetaldehyde (~ 10 - 3 M) in 30 mL CH2C12 but no Fe porphyrin complex was monitored by GC. No loss of the aldehyde occurred over 3 hours. Another control test using n-Bu3P instead of FeTPP(n-Bu3P)2 in CH2C12 solvent did not result in loss of in i t i a l phenylacetaldehyde. These tests indicate that the iron porphyrin with phosphine ligands is responsible for the decarbonylation reaction. V.4 Decarbonylation Mechanism Studies on the decarbonylation of aldehydes using RuTPP(PPh3)2 indicated a radical mechanism and Ru(III) intermediates [1(a),1(b)]. A similar mechanism could be operating in the case of the FeTPP(n-Bu3P)2 system. The irreproducibility of the reactions, the very large variations In the turnover numbers, the inhibition of the reaction in the presence of radical inhibitors and the detection of bibenzyl (PhCH2CH2Ph) (Figure V.3) obtained during the decarbonylation of phenylacetaldehyde, are a l l consistent with a free-radical mechanism. A tentative mechanism is outlined in Figure V.4. Hydride abstraction from aldehyde by the Fe porphyrin leads to formation of the acyl radical. Rapid decarbonylation of the acyl radical (k ~ 5 x 107 s - 1 for the case of phenylacetyl radical [85]), metal assisted 91 if necessary, gives a Fe(II) carbonyl 4 (UV/visible spectra 438 nm) which is subsequently decarbonylated by nucleophilic attack by n-Bu3P from 2. This phosphine can displace coordinated carbonyl as shown by eq. V.l K FeTPP(CO)(n-Bu3P) + n-Bu3P ^ FeTPP(n-Bu3P)2 + CO 92 Figure V.4 Proposed decarbonylation mechanism of aldehydes using FeTPP(n-Bu 3P) 2 93 This reaction occurs thermally in toluene at 29°C with a K value of 1.4, determined in Chapter 4, Section 1. The bis-phosphine Fe(II) porphyrin species 1 thus re-formed (UV/visible spectra 455 nm, Fig. IV.1) continues the catalytic cycle. (2 > 4). The decarbonylation as shown in Figure V.4 is intramolecular in aldehyde and is probably assisted by a 'solvent cage'. It is likely that a 'secondary'reaction may also be occurring, viz R* + RCHO > RH + R*CO (species b) The R free radical species formed in the reaction b > c may recombine. This product in the case of bibenzyl was detected and identifed by,GC/MS (Figure V.3). The role of argon or nitrogen flushing is that by assisting the removal of free CO, it diminishes (or prevents) formation of the 'carbonyl' species 4; rather, the 'active' 5-coordinate species 2 is obtained, hence argon flushing accelerates the decarbonylation reaction. Oxygen appears to deactivate the catalyst; the inactive decarbonylation mixture has a Soret at 415 nm. 94 V.4,1 The role of the phosphine ligand Tri-n-butylphosphine, a good it-acceptor, presumably labilises the trans CO in the intermediate carbonyl-metalloporphyrin complex 4 (Figure V.4). The UV/visible spectral changes on vacuum pumping confirm this. In contrast, the FeTPP(pip)2 system binds the CO tightly and the bound CO is not lost on vacuum pumping; only the use of piperidine as solvent labilises the CO in the latter system. The dissociation of a phosphine ligand from RuTPP(PPh3)2 was considered essential for the decarbonylation reaction and the non-dissociation of the phosphine ligand in RuTPP(n-Bu3P)2 was considered a possible reason for the non_reactivity of the latter porphyrin towards aldehydes [1(a)]. In the case of FeTPP(n-Bu3P)2, however, no spectral changes were observed over the 10~3 - 10 - 5 M range in CH2C12, showing that phosphine dissociation Is negligible even at these low concentrations. V.5 Comparison of FeTPP(n-Bu3P)2 system with other decarbonylation systems Operating temperature for decarbonylation catalysts [1(c)] RhCl(PPh3)3 (178°C), trans-Rh(C0)Cl(PPh3)2 (178°C), [Rh(dppm)2]+ (178°C) and [Rh(dppm)2]BFlt (150°C) are high when compared with those reported for RuTPP(PPh3)2 (40°C) and FeTPP(n-Bu3P)2 (24-40°C). The ability of the porphyrin-related catalysts to decarbonylate aldehydes at ambient temperatures has only recently been matched according to a recent 95 literature report [86] by a 'decarbonylase' enzyme which decarbonylates aldehydes stoichiometrically at room temperature. Another factor to be considered is that the turnover min - 1 for the porphyrin systems are larger than those reported for non-porphyrin systems by a factor of 10 2. V.6 Conclusions The FeTPP(n-Bu3P)2 complex in solution forms a carbonyl complex FeTPP(CO)(n-Bu3P) with CO, or with aldehydes used as a source of CO. The coordinated CO could be easily displaced by vacuum pumping or by flushing argon through the reaction solution. Excess n-Bu3P also displaced CO rapidly. UV/visible spectroscopy was used to follow the reaction of FeTPP(n-Bu3P)2 with CO gas and with aldehydes. Equilibrium constants were determined for the reaction of CO with the porphyrin and thermodynamic parameters (AH and AS) were obtained. The decarbonylation of phenylacetaldehyde using iron porphyrin complexes proceeds rapidly at ambient temperatures in CHjClj as solvent. Reactions were monitored by UV/visible spectroscopy and by GC, and products identified by GC/MS. The study was hindered by irreproducibility of the reaction, with wide variation in turnover numbers for similar sets of reactions; the system was extremely oxygen sensitive. Whether trace oxygen is necessary for the decarbonylation to proceed has not been established; however, excess oxygen irreversibily destroyed the catalyst. Excess phosphine (PPh3, n-Bu3P) inhibited the reaction as did the use of a CO gas 96 atmosphere. The decarbonylation almost certainly occurs via a free radical mechanism; such a mechanism was postulated in an earlier investigation [1(a)] using a related ruthenium catalyst system. No dissociation of a phosphine ligand (as judged by spectral changes) was detected in the case of FeTPP(n-Bu3P)2. V.7 Suggestions for further studies The extreme air-sensitivity of the FeTPP(n-Bu3P)2 catalyst system was a drawback and while excess phosphine prevented the deactivation of the catalyst, no detectable decarbonylation of aldehydes occurred in such solutions. The possible use of modified porphyrins which are less air-sensitive is worth considering, although the modified catalytic system may not retain its original activity. Intermediate species formed during the course of the decarbonylation reaction should be isolated and identified; for example, the 'carbonyl' species which absorbs at 438 nm. The use of esr and cyclic voltammetry detected the presence of free radicals and Ru(III) species in an earlier study on a related system [1(a)]; similar studies could be conducted on the present system. Some labelling studies, similar to those reported by Kampmeier et al. [23], should confirm a free-radical mechanism, ie., decarbonylating a mixture of aldehyde and deutero-aldehyde should give a 'D' crossover if free radicals are involved. 97 The decarbonylation reactions should be studied using an isolated 5-coordinate Fe(porphyrin)(phosphine) complex. The ease of preparation of these iron porphyrin complexes, mild conditions found necessary, and the low cost of Fe versus Rh or Ru, are a positive feature of the present system. The use of polymer-supported phosphines to heterogenise this catalyst seems a logical 'next step' in the study of this fascinating reaction. 98 References (a) G. Domazetis, B. Tarpey, D. Dolphin and B.R. James, J.C.S. Chem. Comm., 939 (1980); ACS Symposium Series 152, 243 (1981). (b) B. Tarpey, M.Sc. Thesis, University of British Columbia (1982). (c) Homogeneous Catalysis with Metal Phosphine Complexes, Ed. L.H. Pignolet, Plenum Press, New York, 1983, pp. 343. (d) H.H. Wang, Ph.D. Thesis, University of Minnesota, 1982, from: Diss. Abstr. Int. B 1982, 42 (12), 4788. (e) F.H. Jardine, Progress in Inorg. Chem., Ed. S.J. Lippard, vol. 28, Wiley-Interscience, New York, 1981, pp. 63-202. (f) D.H. Doughty, Ph.D. Thesis, University of Minnesota, 1979, from: Diss. Abstr. Int. B 1980, 40 (11), 5257. (g) J. Tsuji, K. Ono, Japan JP 46/21603[71/21603], 18 June 1971. Through CA (74(13): 635934). (h) R.H. Prince and K.A. Raspin, J. Chem. Soc. A, 612 (1969). (i) M.F. McGuiggan, Ph.D. Thesis, University of Minnesota, 1984. from: Diss. Abstr. Int. B. 1984, 45(05), 1464. C.W. Bird, Transition Metal Intermediates in Organic Synthesis, Logos Press, London, 1967, pp. 112, 239. J. Tsuji and K. Ohno, Synthesis, I, 157 (1969). A. Kozikowski and H. Wetter, Synthesis, 2 » 5 6 1 (1976). W.M. Schubert and R.R. Kinther, The Chemistry of the Carbonyl Group, Wiley-Interscience, New York, 1966, pp. 695. 99 6. J.N. Pitts and J.K.S. Wan, The Chemistry of the Carbonyl Group, Wiley-Interscience, New York, 1966, pp. 823. 7. J. Chatt and B.L. Shaw, Chem. and Ind., 931 (1960). 8. J. Chatt and B.L. Shaw, Chem. and Ind., 290 (1961). 9. L. Vaska and J.W. Diluzio, J. Am. Chem. Soc, 83, 1262 (1961). 10. L. Vaska, Chem. and Ind., 1402 (1961). 11. L. Vaska and J.W. Diluzio, J. Am. Chem. Soc, 83, 2784 (1961). 12. L. Vaska, J. Am. Chem. Soc, 86, 1943 (1964). 13. R.H. Prince and K.A. Raspin, J.C.S. Chem. Comm., 156, 1966. 14. M.C. Baird, C.J. Nyman and G.W. Wilkinson, J. Chem. Soc (A), 348 (1969). 15. T. Tatsumi, H. Tominaga, M. Hidai and Y. Uchida, J. Organomet. Chem., 215, 67 (1981). 16. J.W. Wilt and W.W. Pawlikowski, Jr., J. Org. Chem., 40, 3641 (1975). 17. K. Ohno and J. Tsuji, J. Am. Chem. Soc, 90, 99 (1968). 18. C.S. Wong, Ph.D. Thesis, University of Chicago, 1973, via ref. 1(c). 19. J.W. Suggs, J. Am. Chem. Soc, 100, 640 (1978). 20. CF. Lochow and R.G. Miller, J. Am. Chem. Soc, 98, 1281 (1976). 21. (a) D. Milstein, Acc. Chem. Res., 17_, 221 (1984). (b) D. Milstein, Organometallics I, 1549 (1982). 22. (a) J.A. Kampmeier, S.H. Harris and I. Mergelsberg, J. Org. Chem., 49, 621 (1984). 100 (b) B.R. James and C.G. Young, J. Organometal. Chem., 285, 321 (1985). 23. J.A. Kampmeier, S.H. Harris and D.K. Wedegaertner, J. Org. Chem., 45, 315 (1980). 24. (a) J. Blum, E. Oppenheimer and E.D. Bergman, J. Am. Chem. Soc, 89, 2338 (1967). (b) J. Tsuji and K. Ohno, Tetrahedron Lett., 4713 (1966). 25. J.B. Philip, Jr., Ph.D. Thesis, University of Rochester, 1980. from: Diss. Abstr. Int. B 42 (10),219. 26. Cited in reference 1(c), pg. 355. 27. J.A. Osborn, F.H. Jardine, J.F. Young and G. Wilkinson. J. Chem. Soc. (A), 1711 (1966). 28. G.L. Geoffroy, D.A. Denton, M.E. Keeney and R.R. Bucks, Inorg. Chem., 15, 2382 (1976). 29. (a) C. Masters, Homogeneous Transition-metal Catalysis, Chapman and Hall, London, 1981, pp. 5-17. (b) D.H. Doughty and L.H. Pignolet, J. Am. Chem. Soc, 100, 7083 (1978). (c) D.H. Doughty, M.F. McGuiggan, H. Wang and L.H. Pignolet, in 'Fundamental Research in Homogeneous Catalysis', Ed. M. Tsutsui, 3, Plenum Press, New York, 1979, pp. 909. (d) B.R. James and D. Mahajan, Can. J. Chem., 57_, 180 (1979). 30. C.A. Tolman in 'Fundamental Research in Organometallic Chemistry', Eds. M. Tsutsui, Y. Ishii and H. Yaozeng, Van Nostrand Reinhold Co., 1980, pp. 483-95. 101 31. R.R. Schrock and J.A. Osborn, J. Am. Chem. Soc, 93_, 2397 (1971). 32. B.R. James and D. Mahajan, Can. J. Chem., _58, 996-1004 (1980). 33. D.H. Doughty, M.P. Anderson, A.L. Casalnuovo, M.F. McGuiggan, C.C. Tso, H.H. Wang and L.H. Pignolet, Adv. Chem. Ser., 196, 65-83 (1982). 34. T.B. Rauchfuss, in 'Fundamental Research in Homogeneous Catalysis', Vol. 3, Ed. M. Tsutsui, 1979, pp. 1029. 35. S. Walker, M.Sc. Thesis, University of British Columbia (1980). 36. B.R. James, A.D. Rattray and D.K.W. Wang, J.C.S. Chem. Comm., 792 (1976). 37. J.W. Wilt, in Free Radicals, Vol. 1, Ed. J.K. Kochi, Wiley, New York, N.Y. (1973). 38. H.M. Walborsky and L.E. Allen, J. Am. Chem. Soc, 93, 5465 (1971). 39. (a) D.V. Stynes and B.R. James, J. Am. Chem. Soc, 9_6, 2733 (1974). (b) C.J. Weschler, D.L. Anderson and F. Basolo, J. Am. Chem. Soc, 97, 6707 (1975). 40. (a) H. Gray, Adv. Chem. Ser., 100, 365 (1971). (b) E.B. Fleischer, Acc. Chem. Res., 3, 105 (1970). (c) A.L. Balch, Y.W. Chan, R.J. Cheng, G.N. LaMar, L. L-Grazynski and M.W. Renner, J. Am. Chem. Soc, 106, 7779 (1984). (d) D.H. Chin, J.D. Gaudio, G.N. LaMar and A.L. Balch, J. Am. Chem. Soc, 99_, 5486 (1977). 41. (a) N.P. Farrell, D. Dolphin and B.R. James, J. Am. Chem. Soc, 100, 324 (1978). 102 (b) H. Masuda, T. Taga, K. Osaki, H. Sugimoto, M. Mori and H. Ogoshi, J. Am. Chem. Soc, 103, 2199 (1981). 42. D.F. Shriver, The Manipulation of Air Sensitive Compounds, McGraw-Hill, New York, N.Y., 1969. 43. D.D. Perrin, W.L.F. Armarego and D.R.Perrin, Eds. Purification of Laboratory Chemicals, 2nd Ed., Pergamon Press, Sydney, 1982. 44. A.D. Adler, F.R.Longo, J.D. Finarelli, J. Goldmacher, J. Assour and L. Korsakoff, J. Org. Chem., 32, 476 (1967). 45. R. Rousseau and D. Dolphin. Tetrahedron Lett., 4251-54 (1974). 46. H. Kobayashi, T. Higuchl, Y. Kaizu, H. Osada and M. Auki, Bull. Chem. Soc Jpn., 48 (11) 3137 (1975). 47. E.B. Fleisher and T.S. Srivastava, J. Am. Chem. Soc, 9_1, 2403 (1969). 48. L.M. Epstein, D.K. Straub and C. Maricondi, Inorg. Chem., 1720 (1967). 49. J. Del Gaudio and G.N. LaMar. J. Am. Chem. Soc, 98, 3014-15 (1976) 50. T. Ohyo, H. Morohoshi and M. Sato, Inorg. Chem., 23, 1303-05 (1984). 51. W.R. Scheldt and C.A. Reed, Chem. Rev., 81, 543-55 (1981). 52. R.J.P. Williams, Fed. Proc, Fed. Am. Soc. Exp. Biol. ,20, 5 (1961). 53. (a) J.L. Hoard, M.J. Hamor, T.A. Hamor and W.S. Caughey, J. Am. Chem. Soc, 2312 (1965). (b) J.L. Hoard, Science, 174, 1295 (1971). 54. M.A. Stanford, J.C. Swartz, T.E. Phillips and B.M. Hoffman, J. Am. Chem. Soc, 102, 4492 (1980). 103 55. (a) W.R. Scheldt and M. Gouterman, in 'Iron porphyrins' vol. 1, Eds. A.B.P. Lever, H.B. Gray, Add!son-Wesley, 1983, pp. 89. (b) W.R. Scheldt, J. Am. Chem. Soc, 96, 90 (1974). (c) J.F. Klrnar, C.A. Reed and W.R. Scheldt, J. Am. Chem. Soc, 99_, 2557 (1977). 56. V.L. Goedken, in 'Coordination Chemistry of Macrocyclic Compounds', Ed. G.A. Melson, Plenum Press, London, 1979, pp. 607. 57. D.R. English, D.N. Hendrickson, K.S. Suslick, C.W. Eigenbrot Jr. and W.R. Scheldt, J. Am. Chem. Soc, 106, 7258 (1984). 58. Phosphorus - An Outline of its Chemistry, Biochemistry and Technology, Ed. D.E.C. Corbridge, Elsevier (Oxford), 1978, pp. 25, 440-42. The a bond or single bond distance is obtained by simple addition of the respective single bond covalent radii which are 1.16 A and 1.11 A for Fe and P respectively. The shortening of the bond length due to the partial ionic character of the bond is estimated by considering the electronegativity difference between the two atoms. The Schomaker-Stevenson empirical correction [59] is used to calculate this effect; for the case of the Fe-P bond, Bond distance Fe-P = Yp + Y F e " °* 0 9( x p" X F e) <e<l' HI.l) where v and v„ refer to single bond covalent radii for P and Fe 'p 'Fe and X refers to electronegativities of the bonded atoms which are X =2.1 and X^  =1.8 respectively; hence the standard Fe-P bond 104 distance is obtained by substituting the appropriate values in equation III.l. Fe-P bond distance = 1.16 + 1.11 - 0.09 (2.1-1.8) - 2.243 A corrected for ionic shortening. (a) D.J. Dahm and R.A.Jacobson, Chem. Comm., 496 (1966). (b) J.L. Birck, Y. Le Cars, N. Baffler, J.J. Legendre and M. Huber, C.R. Hebd. Seances Acad. Sci., Series C, 273, 880 (1971). 59. S.B. Hartley, W.S. Holmes, J.K. Jacques, M.F. Mole and J.C. McCoubrey, Quart. Revs., 17_, 204 (1963). 60. W.R. Scheldt and M. Gouterman in 'Iron Porphyrins' Vol. 1, Eds. A.B.P. Lever, H.B. Gray; Addison-Wesley, 1983, pp 106-7. 61. J.W. Lauher and J.A. Ibers, J. Am. Chem. Soc.,96, 4447 (1974). 62. T. Sakuri, K. Yamamoto, N. Seins and M. Katsuta, Acta Crystallogr., Sect. B. 31, 2514-16 (1975). 63. J.A. Kaduk and W.R. Scheldt, Inorg. Chem., JL3, 1875 (1974). 64. W.R. Scheldt, J.A. Cunningham and J.L. Hoard, J. Am. Chem. Soc, 95, 8289 (1973). 65. F. Cecconi, M.D. Vaira, S. Midollini, A. Orlandini and L. Sacconi, Inorg. Chem., 20, 3423 (1981). 105 66. M.D. Vaira, S. Midollini and L. Sacconi, Inorg. Chem., 20, 3430-35 (1981). 67. W.R. Scheldt, D.K. Geiger and K.J.J. Haller, J. Am. Chem. Soc, 104, 495 (1982). 68. (a) W.R. Scheldt, A.C. Brinegar, J.F. Kirner and C.A. Reed, Inorg. Chem., 18. 3610 (1979). (b) J.F. Kirner and W.R. Scheldt, Inorg. Chem., 14_, 2081 (1975). (c) W.R. Scheldt, J. Am. Chem. Soc, 96, 84 (1974); R.G. Little and J.A. Ibers, J. Am. Chem. Soc, %.> 4440 (1974). 69. M.E. Kastner, C.A. Reed and W.R. Scheldt, unpublished results. Cited in W.R. Scheldt, Acc Chem. Res., 1£, 339 (1977). 70. L.J. Radonovich, A. Bloom and J.L. Hoard, J. Am. Chem. Soc, 94, 2073 (1972). 71. (a) T. Mashiko. J.C. Marchon, D.T. Musser, C.,A. Reed, M.E. Kastner and W.R. Scheldt, J. Am. Chem. Soc, 101, 3653 (1979). (b) T. Mashiko, C.A. Reed, K.J. Haller, M.E. Kastner and W.R. Scheldt, J. Am. Chem. Soc, 103, 5758 (1981). 72. C. Caron, A. Mitschler, G. Riviere, L. Richard, M. Schappacher and R. Weiss, J. Am. Chem. Soc, 101, 7401 (1979). 73. S.C. Tang, S. Koch, G.C. Papaefthymiou, S. Foner, R.B. Frankjel, J.A. Ibers and R.H. Holm. J. Am. Chem. Soc, 98, 2414 (1976). 74. J.E. Hahn, L.A. Andersson and J.H. Dawson. J. Biol. Chem., 257, 10934 (1982). 106 75. G.B. Jameson and J.A. Ibers, Inorg. Chem., 18, 1200 (1979). 76. S-M. Peng and J.A. Ibers, J. Am. Chem. Soc, 98, 8032 (1976). 77. J.P. Collman, Acc. Chem. Res., 10, 265 (1977). 78. (a) J.P. Collman, R.R. Gagne", C.A. Reed, W.T. Robinson and G.A. Rodley, Proc. Nat. Acad. Sci. U.S.A., 71, 1326 (1974). (b) G.B. Jameson, G.A. Rodley, W.T. Robinson, R.R. Gagne, C.A. Reed and J.P. Collman, Inorg. Chem., 17_, 850 (1978). (c) G.B. Jameson, F.S. Molinaro, J.A. Ibers, J.P. Collman, J.I. Brauman, E. Rose and K.S. Suslick, J. Am. Chem. Soc, 100, 6769-70 (1978); 102, 3224-37 (1980). 79. S.E.V. Phillips, J. Mol. Biol., 142, 531-554 (1980). 80. J. Chatt, in 'Homogeneous Catalysis with Metal Phosphine Complexes'. Ed. L.H. Pignolet, Plenum Press, New York, 1983, pp. 5-6. 81. B.R. James, K.J. Reimer and CT. Wong, J. Am. Chem. Soc, 9_9_ 4815 (1977). 82. L.R. Field, E. Wilhelm and R.Battino, J. Chem. Thermodynamics, 6^, 237 (1974). 83. Handbook of Chemistry and Physics, Ed. R.C Weast, CRC Press Inc., Florida, 62nd Ed. D-180 (1980). 84. J.E. Falk, Porphyrins and Metalloporphyrins, Elsevier Publishing Co., New York, N.Y., (1964). 85. G. Brunton, H.C. McBay and K.U. Ingold, J.Am. Chem. Soc, 99_, 4447 (1977). 86. T.M. Cheesbrough and P.E. Kolattukudy, Proc. Nat. Acad. Sci. U.S.A., 81 (21), 6613 (1984). 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0060600/manifest

Comment

Related Items