Cobalt(III)-promoted hydrolysis of chelated glycine esters. Kinetics

Cobalt(III)-promoted hydrolysis of amino acid esters and peptides and the synthesis of small peptides. Paul A. Sutton and David A. Buckingham. Account...
0 downloads 0 Views 1023KB Size
Download PDF
6032

cyclic transition state with t-butyl alcohol. This conclusion is consistent with that derived from the H-isotope effects in the related aqueous medium.

Acknowledgment. We wish to thank Dr. J. Fildes and the analytical unit of the John Curtin School of Medical Research, A.N.U.,for the microanalyses.

Cobalt (111)-Promoted Hydrolysis of Chelated Glycine Esters. Kinetics, Anion Competition, and O’*-ExchangeStudies D. A. Buckingham, D. M. Foster, and A. M. Sargeson

Contribution from the Research School of Chemistrj,, Australian National Unicersity, Canberra, A.C.T. 2600, Australia. Received M a y 6 , 1968 Abstract: Kinetic data for the hydrolysis of [Co(en),(glyOC3H~)](C1O4)3 in aqueous solution have been obtained, and the results confirm the proposal made by Alexander and Buschl that a chelated ester intermediate [Co(en)z(X = C1, Br; (glyOR)I3+is formed following the Hg?+-inducedremoval of halide ion from [C~(en)~X(glyOR)]~+ R = CHI, C2H5,C3H7). The same intermediate is generated following the HOC1 oxidation of coordinated Br-. Anion competition results indicate that the chelated ester intermediate is formed directly in the Hgzf-assisted reaction but suggest that some [Co(en),(HzO)(glyOC3H7)] 3+ is generated in the HOC1-promoted reaction. 0 l8 studies establish that hydrolysis of [C~(en)~(glyOCH~)]~+ proceeds without opening of the chelate ring and that the coordinated ester is bound to Co(II1) through the carbonyl oxygen.

I

n a recent publication Alexander and Busch have discussed the cobalt(lI1)-promoted hydrolysis of coordinated glycine esters (reaction 1). Similar, although less detailed information has been obtained for the hydrolysis of small and relatively simple peptide molecules (reaction 2).2 Both reactions result in a final cobalt(II1) product containing the chelated Namino acid anion, and a marked enhancement in rate, compared to the uncatalyzed reactions, was found for hydrolysis of the ester and peptide bonds, respectively.

+

+

[ C ~ ( e n ) ~ X ( g l y O R ) ] ~Hg2+ ~ HzO + [C~(en)~gly]~+ HOR

+

+ H+ + HgX+

(1)

X = C1, Br; R = CHI, CzHa, i G H 7

+

[Co(trien)OH(HzO)]*+ NHzCHRCONHCHR’. . COZR” + [ C O ( ~ ~ ~ ~ ~ ) ( N H Z C H2+R C O Z ) ] NHzCHR‘. .COzR” Hz0 (2)

+

I

+

Our interest in the above reactions has been stimulated following the isolation and characterization of chelated amino acid ester, amino acid amide, and dipeptide ester complexes of general structure I a 3 Coordination complexes of this structural type have been invoked in the metal ion catalyzed acceleration in the 3+

hydrolysis of such substrates, but, with the exception of [Co(trien)(glyg1yOEt)l3+ which was isolated from reaction 2,5 the reported evidence to support the formation of such intermediates either prior to or as a step in the hydrolysis reaction is largely of an indirect kind. For example, Alexander and Busch proposed a chelated ester intermediate [C~(en)~(glyOR)] 3+ as the reactive intermediate (structure I) in the Hg2+-promoted hydrolysis (reaction 1). The evidence for this species was obtained primarily from the change in the >C=O stretching frequency as the monodentate ester (1735 cm-l) was first chelated (1610 cm-I) and then hydrolyzed to [Co(en),gly12+ (1640 cni-’). As the [C~(en)~(glyOR)](ClO~)~ compounds have now been isolated,3 the hydrolysis rates can now be compared with those attributed to this intermediate by the previous authors. Other mechanistic aspects of this reaction which merit consideration are the following. (a) Does opening of the chelate ring precede hydrolysis? (b) Does removal of halide ion from the monodentate ester complex lead to initial incorporation of solvent at the vacated coordination site prior t o chelation? (c) What properties does the monodentate ester aquo complex have? Some of these questions are potentially soluble using 0 ls-tracer techniques, and this paper presents some results obtained from such investigations. Experimental Section

O d

Analar reagents were used throughout without further purification. O’a-Labeled glycine methyl ester hydrochloride was prepared as follows. Glycine hydrochloride was enriched by refluxing the 0’8-enriched water.6 salt for 24 hr at pH 0.5 (HCI) with 1.5 atom

‘R I R = OR, NRR’, SHRCHR’C0,R” (1) M. D. Alexander and D. H. Busch, J . A m . Chem. SOC.,88, 1130 (1966). (2) D. A. Buckingham, J. P. Collman, D. A. R . Happer, and L. G. Marzilli, ibid., 89, 1082 (1967). (3) D. A . Buckingham, L. G.Marzilli, and A. M. Sargeson, ibid.,89,

Journal of the American Chemical Society

90.22

1

(4) (a) H. Kroll, ibid., 74, 2036 (1952); (b) M. L. Bender and B. W. Turnquest, ibid., 79, 1889 (1957); (c) an excellent review of this subject is given by M. L. Bender, Advances in Chemistry Series, No. 37, American Chemical Society, Washington, D. C., 1963, p 19. (5) J. P. Collman and E. Kimura, J . A m . Chem. SOC., 89, 6096

October 23, 1968

6033 The dry product (21 g) was methylated by refluxing until dissolved content of the COZ was determined by the mass spectrometer. in methanol (200 ml) and thionyl chloride (12 ml). The methyl In a similar manner, glycine recovered from the complexes deester hydrochloride obtained on evaporation was redissolved in scribed below was decomposed with HgClz and Hg(CN)2 usually warm methanol and crystallized by cooling in an ice bath. Anal. by heating overnight at 400". The 0l8content of the solvent was Calcd for C3H802C1: C, 28.70; H, 6.42: N, 11.16. Found: C, obtained by distilling the water (1 ml) on the vacuum line and 28.79; H, 6.26; N, 11.45. Glycine isopropyl ester was prepared equilibrating it with C o n(-0.1 mmole) overnight at 70°.12 in an analogous manner. O18-Labeled cis-[Co(en)~Br(glyOMe)]Br~was heated to 250" HOCl was prepared by freshly extracting a solution of chlorine(1) under 0.03 mm pressure, whereupon the free glycine methyl ester oxide in carbon tetrachloride into an equal volume of water (or distilled over. The ester was collected in a liquid nitrogen cold dilute HC104).7 The acid so prepared was consistently 1.8-2 M trap and decomposed to COPas described above. (estimated by iodometric titration). Hg2+-Induced Hydrolysis of 018-Labeled cis-[Co(en)%Br(glyPmr spectra were recorded on a Perkin-Elmer R-10 spectrometer OMe)]z+. Carbonyl-018-labeled cis-[Co(en)~Br(glyOMe)](ClO~)~ using sodium trimetliylsilylpropaiiesulfonate (TPSNa) as an internal (0.5 g) was dissolved in 10 ml of 0.6 M Hg2+-0.1 M HCI04 solureference (DrO) or tetramethylsilane (TMS) (acetone-&). Infrared tion. Hydrolysis of the purple bromo ester to the orange glycinato spectra were taken on SP-200 Unicam or Perkin-Elmer 257 instrucomplex was complete within 30 sec at 20". Excess NaI was added was followed ments. Hydrolysis of [C0(en)~(glyOC~H,)](Cl0~)~ until all the initially precipitated HgI2 had redissolved, and on coolon the Perkin-Elmer 257 in 0.658 N DCI using a 0.025-mm lead ing anhydrous [C~(en)~gly](HgI~) precipitated. This was collected, spacer. Spectrophotometric rates were measured on a Cary 14 washed with water and methanol, and dried in an evacuated desicspectrophotometer. Bio-Rad analytical Dowex 50W X 2 (200cator for 2 hr. It was dried again in the vacuum line at 10-3 mm for 400 mesh) cation-exchange resin was used to analyze products from 0.5 hr to remove all traces of solvents. the competition experiments. To recover the glycine, [C~(en)~gly](HgI~) was heated slowly to The 01* content in the COYrecovered from the labeled complexes 200" at 0.03 mm whereupon free glycine sublimed. This was were determined using an Atlas M-86 mass spectrometer. collected on a cold-finger condenser, and the glycine decomposed Preparation of Complexes. cis-[Co(en)~Br(gIyOCH3)]Br2 and to COPwith the HgC1P-Hg(CN)2mixture. cis-[Co(en)?Br(glyOC3H7)]BrPwere prepared from rmns-[Co(en)PAcid Hydrolysis of OlS-Labeled [C0(en)~(glyOMe)]3+. CarBrs]Br.HBr and the appropriate glycine ester hydrochloride as b0nyl-O~~-[Co(en)~(glyOMe)](CIO4)3 (0.9 g) was dissolved in 0. l M described by Alexander and Busch.8 These compounds were conHC104 (3 ml). After 5 min HgI?, followed by excess NaI, was verted to their perchlorate salts by dissolving in hot 10-3 MHC104, added. The [C~(en)~gly](HgI~) product was collected and treated followed by addition of excess NaC10, and cooling in an ice bath. as above. HOC1-Induced Hydrolysis of 018-Labeled [Co(en)2Br(glyOMe)]z+. The products were twice recrystallized from hot dilute HClOl Carb~nyl-O~Ws-[Co(en)~Br(glyOMe)](CIO& (0.5 g) was treated by adding NaC104. They were washed with ethanol and dried in an evacuated desiccator. Anal. Calcd for [Co(en)~Br(glyOCH~)]- with 5 ml of HOCl solution. (The solution was made by extracting a C120xarbon tetrachloride solution into an equal volume of (c104)2: C, 15.38; H , 4.24; N, 12.80. Found: C, 15.28; H, C, 1.8 M AgCI04-0.2 M HC104 solution and filtering to remove 4.41 ; N, 12.93. Calcd for [C0(en)~Br(glyOC,H~)](Cl0~)~: 18.80; H , 4.73; N, 12.18. Found: C, 18.84; H, 4.99; N, 12.36. AgC1.) The reaction mixture turned orange within 30 sec at 20". cis-[Co(en)~Cl(glyOC~H~)](C10~)2 was prepared in a similar manner SOPwas bubbled through the solution for 1 min to reduce unreacted from tr~ns-[Co(en)~Cl~]CI and glycine isopropyl ester hydrochloride. HOCI, and excess NaI was added, whereupon the product precipiAIZNI. Calcd for [ C O ( ~ ~ ) ~ C ~ ( ~ ~ ~ O C ~ H ~ C, ) ] 19.70; ( C I O ~ )tated P ~ HasP O[C0(en)~gl?](AgI2)~. : This was washed with water and H , 5.33; N, 12.77. Found: C, 19.93; H , 5.33; N, 12.89. methanol and dried in a vacuum desiccator overnight. The glycine was recovered in the same manner as for the HgI4'- salt. Directions for preparing the chelated ester intermediates [Co018-Labeled HOCI-Induced Hydrolysis of [Co(en)~Br(glyOCH)~]will be (en)P(glyOCH3)](C104), and [Co(er~)~(glyOC~H~)](ClO~)~ described shortly in a paper describing peptide syntheses.9 The above procedure was repeated using 018-labeled Bromoglycinatobis(ethylenediamine)cobalt(III) Bromide. Powwater,I3and the complex was collected as the HgIa2-salt. Immediately after the addition of H O T 1 to the bromo ester, a dered [C~(en)~Br(glyOCH~)]Br? (11 g) was added to 7.6 N HBr (100 ml) and the solution warmed at 50" for 1 5 min and then shaken sample of the solution was withdrawn and frozen in liquid nitrogen. at room temperature for 76 hr. The solution was cooled in an ice The water from this sample was subsequently distilled under vacuum bath and the fine red precipitate was collected, washed with ethanol and equilibrated for 12 hr (70") in a sealed tube with carbon diand acetone, and dried. This material was dissolved in the minioxide. Kinetic Measurements. The acid hydrolysis of [Co(en)z(glymum volume of water at 80", the solution was filtered, HBr was added (2 ml, 7.6 N ) , and the solution was allowed to stand overnight OC,H,)](ClO,), was followed spectrophotometrically at 487 mp. was dissolved in a t 25". The red crystals were collected and washed with dilute A weighed quantity of [C0(en)~(glyOC~H~)](Cl0~)~ HC104 of known ionic strength at 25" and quickly transferred to a ice-cold HBr and methanol and air-dried (5.6 g). Anal. Calcd for [ C ~ ( e n ) ~ B r ( g l y O H ) ] B r ~C,: ~ ~14.59; H, 4.28; N, 14.18. thermostated cell. The HOCI-induced hydrolysis of [C0(en)~Br(glyOC~H~)](Cl0~)~ Found: C, 14.37; H,4.32; N, 14.48. Preparation of 0 l*-Labeled Complexes. Carbonyl 018-labeled was followed at 48g490 mp. A weighed quantity of ciJ-[Co(enhcis-[Co(en)2Br(glyOCH3)]Brp was prepared from trans-[Co(en)sBr(glyOC3Hi)](C104)Pin water at 25.0 f 0.1" was rapidly mixed BrP]Br.HBr and 0 18-labeled glycine methyl ester hydrochloride.8 with an equal volume of HOC1 and filtered into the thermostated cell. The HOC1 was prepared immediately before use by extracting Anal. Calcd for [Co(en)sBr(glyOCH3)]Br~: C, 16.55; H, 4.56; chlorine(1) oxidexarbon tetrachloride solution directly into an equal N, 13.79. Found: C, 16.35; H, 4.73; N, 13.70. This salt was volume of either 1.12 M NaCD-0.2 M HCIO4 solution or 1.32 M converted to the perchlorate as described above. Anal. Calcd ~)THP H~, :4.46; N, HCIO4. for [ C O ( ~ ~ ) ~ B ~ ( ~ I ~ O C H ~ ) ] ( CC,I O14.89; [C0(en)s(glyOC~Hi)](CIO4)~ was also treated with 2,4,6-collidine 12.39. Found: C, 15.04; H, 4.44; N, 12.67. Carbonyl-0 18-[Co(en)?(glyOCH3)I0, was prepared from carbuffer (0.05 M), pH 8.51 at 25.0". From the visible spectrum of the solution (e4gi 98), it was concluded that hydrolysis to [C~(en)~gly]z+ bonyl-0 18-[Co(en)PBr(glyOCH3)](C104)2.g was complete within the time of mixing ( t i / * < 1 min). Recovery of C o o l 8 . 018-Labeled glycine methyl ester hydrochloride was heated for not less than 2 hr at 400" in an evacuated Measurement of 0l8Exchange in Carbonyl- and Carboxyl-Lasealed tube with an equimolar mixture of HgCL and Hg(CN)2,11 beled [C~(en)~gly]*'. Carbonyl-018-labeled[Co(enz)Br(glyOCH3)]and the COYwas distilled under vacuum, purified by passing through (C104)P(3 g) was dissolved in 0.1 M HClOa (20 ml) containing Hga gas chromatograph using He as the carrier, collected in a liquid (NO& (7 g), and the solution was filtered after 2 min. Excess NaI was added until the HgIz which initially precipitated had redisnitrogen trap, and finally distilled into a Urey tube.llb The 0'8 solved. [C~(en)~gly](HgI~) precipitated on cooling. This was filtered off and washed with water and methanol, yield 6 g. (6) W . H. Mears, J . Chem. Phj,s., 6, 295 (1938). This material was shaken in water (20 ml) with excess AgCl for (7) G .H. Cady, Inorg. SJ,~., 5, 158 (1957). approximately 2 min. The precipitate of AgCI, AgI, and HgIz (8) M. D. Alexander and D. H. Busch, Inorg. Chem., 5 , 602 (1966). was filtered off, and the filtrate was made up to 50 ml, 0.1 M in (9) D. A Buckingham, D. M . Foster, and A. M. Sargeson, to be submitted for publication. (IO) glyOH = monodentate NHnCH.CO?H, gly0 = monodentate (12) F. 4 . Posey and H. Taube, J . A m . Chem. Soc., 79, 255 (1957). NH2CH?CO?-; gly = chelated NH?CHsCO>-. (13) Under these conditions oxygen exchange between HOCl and (11) (a) M . Anbar and S. Guttman, J . A p p l . Rnd. Isotopes, 5, 223 HzO is rapid: M. Anbar and H. Taube, ibid., 80, 1073 (1958). (1959); (b) A. Sargeson and H. Taube, Iiiorg. Chem., 5, 1094 (1966).

Buckingham, Foster, Sargeson 1 Hydrolysis of Chelated Glycine Esters

6034 HClO4 and 0.65 M in NaCIOt. The solution was thermostated HOC1-promoted hydrolysis of [C~(en)~Br(glyOC~to 25", and 5-ml aliquots were periodically withdrawn. [CoH7)I2+,Table 11. The constants were obtained from (en)2gly](Hg14) was recovered and the 01*content of the glycine linear plots of log ( D , - D J against time, and they was determined as described above. are similar to the values obtained by Alexander and The rate of 01* exchange in [ C ~ ( e n ) ~ g l ylabeled ] ~ + at the carbonyl oxygen was determined by following the above procedure using Busch for the Hg2+-promoted reaction (also Table 11) dissolved with Hg(NO& unlabeled [C0(en)~Br(glyOCH~)](Cl0~)~, and are independent of p H in the range of 0.18-4.' in 0.1 MHCIO, made from 1.5% OIQnriched water. The visible spectrum of [C0(en)~(glyOCH~)](Cl0~)~ Competition Experiments. Hydrolysis Promoted by Hgz-. in 0.1 M HClOd was identical with that of [Co(en),[Co(en)?Br(glyOH)]Brz, [C0(en)~Br(glyOC~H~)](C10~)~, and [Cog1yI2+( € 3 4 6 107, €487 97) in the time of mixing (2 min) and (en)zgly]Brz~ H 2 0 (-0.1 mmole) were separately dissolved in solutions (10 ml) containing H N 0 3 (0.1 M ) , Hg(NO& (0.1 M ) , and showed no change with time. N a N 0 3 (5 M ) . After 30 min the visible spectra were recorded. The solutions from [C~(en)~Br(glyOH)]Br~ and [Co(en)zBr(glyOC3H,)](C104)2were sorbed on an ion-exchange column and eluted with Table I. First-Order Rate Constants for the Hydrolysis of 1 M KOAc, pH 5. The recovery of [Co(en)yglylz+was better than [C~(en)~(glyOC,H~)]~+ in Aqueous Solutiona 99%, and no singly positively charged species were detected. For comparative purposes [Co(en),NH3C1]CI2(1 mmole) was treated in [Complex], k X lo3(*0.05), the same manner and elution gave [ C O ( ~ ~ ) ~ N H ~ (using N O ~ ]1 ~M+ M (NaC10,) PH M x 103 sec-l KCI at pH 3) and [ C O ( ~ ~ ) ~ N H ~ H(using ? O ] ~3+M HClOa). The 1 . 0 0 8 . 2 1.13 concentrations were estimated spectrophotometrically, Table IV. 1.0 4.2 1.18 Hydrolysis Promoted by HOCI. [C0(en)~Br(glyOC~H~)](Cl0~)~ 0.66 1.0 8.51b 9.2 >11.5 (0.1 mmole) in water ( 5 ml) was treated with an equal volume of HOCl (1.8 M ) in 1.12 M NaC10,-0.2 MHCIO.,. After 10 min the a 25.0 =t0.1 ', X 487 mp. 2,4,6-Collidine buffer (0.05 M). visible spectrum was recorded from which the [Co(en)2glyl2+/ [ C ~ ( e n ) ~ C l ( g l y O C ~ H ratio ~ ) ] ~ +was calculated, Table 1V. The product distribution was confirmed using ion exchange with 1 M Table 11. First-Order Rate Constants for the Hydrolysis of the NaCIOl as eluent, 99% recovery. A separate HOC1 preparation Reaction Product from the Reaction of HOCl with gave a similar product distribution, and repeated washing of the cis-[C~(en)~Br(glyOC~H~)]*+ in Aqueous Solutiona HOC1 solution Nith cc14immediately prior to use (to remove CL) also gave the same result, Table IV. When the above experiment [Complex], k X lo3 (10.05), was repeated following extraction of CJyO into 0.2 'VI HClOapH M X 103 sec-l 1.8 M AgCIO?, no [Co(en).CI(glyOC3H7)]*' was detected either 1.0 9.0 1.02 spectrophotometrically or by ion exchange, Table IV. Similarly, 1.0 8.9 1.18 addition of NaCl to a 0.2 M HC10n-1.8 M AgCIOt solution of 0. 18b 9.0 1.18 HOC1 (10 ml) sufficient to make the final solution 1.5 Min chloride, 1, o c 9.0 1.13 followed by immediate addition at 5' of [Co(e~i)~Br(glyOC~Hi)]~+ 4.0d 0.7 1.08 (1 mmole) in water (10 ml), gave e j O O 69 indicating 100% [Co(en),Cl(glyOC3Hj)]2- formation, Table 1V. [Co(e~i)~Br(glyOC,H~)]- a 2 5 . 0 i 0 . 1 " , [ H O C I ] = 0 . 9 M , A 4 8 0 m p , p = 0.66. bX490 (ClO& (0.1 mmole) was treated with HOCl (10 mi) made by exmp. Reaction with 0.1 MHgZ". Reference 1 (acetate buffer). tracting CIyOin CCl, (10 ml) into 3 MHzSOa containing 2 MLiaSOr at 5'. After 2 min the solution was diluted (100 ml) and eluted from a cation-exchange resin (2 M NaClO,), pH 7). The [Co(en),(S04)(glyOCnHi)]', [C~(en)~gly]~-, and [C~(en)?Cl(glyOC~Hi)]~+ Pmr and Infrared Studies. Table I11 includes chemical and [Co(en),shift data for [C0(en)~(glyOCH~)](Cl0~)~ were estimated spectrophotometrically, Table IV. When first sorbed on the column, an orange band was observed above the (glyOC3Hi)](C104)ain acetone-d6. Addition of water purple [C~(en)~Cl(glpOC~H;)]*band, indicating the presence of a 3 ion, presumably [Co(en)2(glyOC3Hj)13+. This band rapidly 21 which moves faster than [Co(en)2faded forming [C~(en)~gly] Table 111. Chemical Shifts and Coupliiig Constants for Some Cl(glyOCnHj)]?' on elution. In a separate experiment using 3 M Glycine Ester Complexes in Dilute DCk H.S04 containing 2 M Li2SOa, no reaction was observed after 10 min and no + I species were detected by ion exchange. Similar Coupling (0.1 mmole in water, 5 treatment of [C~(en)~Br(glyOC~H;)l(clo,), conml)with HOCl (5 ml, 0.1 MHC104, 1.9 MAgCIOd, fi1tered)containGlycine ester stants, ~ suggesting the ing N a N 0 3 (5 M ) gave a red solution with e ; 92, absorptions, ppm CPS [Co(en)2NOs(gly0C3H7)]*'.Table IV. However, presence of -33 this species could not be separated from [Co(en)?gly]*' on the exchange resin. Treatment of [Co(en)2Br(gljOC3H7)lZ+with CI:. [Co(endBrte (20 ml, pH 2) was treated (rliOC,H-)1(C104)2(0.1 g ) in d i l ~ ~HCIO, Giih Cli f&'2 min. The visible spectrum of the resulting solution (e,.; 80; 85) indicates quantitative conversion to [Co(en)tCI(glyOC3Hi)]2+,Table IV. This was verified using the ion-exchange technique. In the presence of added Sotz-some 23-25z [Co(e1i)~SO~(glyOC3Hi)]" is formed, Table 1V. Hydrolysis of [Co(en)~(glyOCH,)](ClO~)~ and [Co(en)dglyOCsHj)](C104)3in the Presence of Added Anions. [Co(en)2(glyOCHa)](C104)3(0.1 mmole) was hydrolyzed in 7.6 N HBr (10 ml) for 10 min. The visible spectrum (eta7 99) and ion-exchange properties indicate quantitative conversion to [Co(eii)ngIy]*+. Similarly to the former results in an immediate shift of the glycine hqdrolqsis of [C0(en)~(glyOC~H~)](Cl0~)~ (0.1 mmole) in 0.1 M CH2 triplet from 4.50 to 3.71 ppm and the appearHCIO, (1 0 ml) in the presence of added C1- (1 M), HSOa- (-2 M ) , ance of a methanol CH3 signal at 3.42 ppm. A similar and NO,,- ( 5 M ) results in no anion entry, Table IV.

+

'

Results Kinetics. The first-order rate constants for the acid hydrolysis of [ C ~ ( e n ) ~ g l y O C ~ H are ~ ]given ~ + in Table I. Also the same rate constants were obtained for the Jvurnul of' the American Cheniicnl Society

90.22

but slower change was observed with [Co(en)2(glyOC3Hi)](C104)3. The final spectra were identical with those of [ C ~ ( e n ) ~ g l y ]and ~ + methanol or 2-propanol, respectively. In dimethyl sulfoxide-ds containing a trace of residual water, a slower hydrolysis rate for [C0(en)~(glyOCH~)](C10~)~ was observed, Figure I .

/ October 23, 1968

6035 I

1

I

I

I

h

I

I

I

1

I/

I

1 0

I

I

I

1600 CM'

'

1400

I

1: x)

Figure 2. Infrared spectrum of [C0(en)~(glyOCH~)](Cl0~)~ in anhydrous acetone (broken lines represent acetone absorptions).

bony1 absorption, and another at 1310 cm-I, which is assigned to the skeletal C-0 stretching vibration. The low frequency of the former and the high frequency of the latter relative to the uncoordinated ester are consistent with >C=O coordination.j The infrared spectrum of [C0(en)~(glyOC~H~)](C10~)~ has strong absorptions at similar frequencies. Acid hydrolysis of [C0(en)~(glyOC~H~)](Cl0~)~ in 0.66 N DC1 was accompanied by a marked broadening in the >C=O stretching frequency at 1620 cm-I and a slight shift to a higher frequency. The shift was not as clearly defined as that given by Figure 6 of ref 1, although qualitative agreement was found. Competition Studies. The results of competition studies for the Hg*+- and HOC1-assisted hydrolysis of [C~(en)~Br(glyOC~H~)]*+ and [C~(en)~Br(glyOH)]*+ in 8 7 6 SFf,M4 3 2 1 the presence of added NOs-, HSOd-, and C1- (HOC1 Figure 1. Pmr spectra of the hydrolysis products from [Co(en>sonly), and for the reaction of CIZ on [Co(en)*Br(gly(glyOCH3)](C104)3in dimethyl sulfoxidede containing a trace of OC3H7)I2+and acid hydrolysis of [C~(en)~(glyOR)] 3+ residual water after 5 min, 35 min, 85 min, 8 hr, and 8 hr (water (R = C3H7, CH,) in the presence of HS0,- and C1-, added), respectively (internal reference, tetramethylsilane). Br-, HS04-, and NO3-, respectively, are given in Table IV. These reactions were carried out in a similar manner t o the nitrosation and Hg2+-assisted reactions of The H O D signal slowly moves downfield as the soluthe ions [Co(NH3);X]*+(X- = N3-, OCONH2-; C1-, tion becomes more acidic, and the CH3 absorption at Br-, I-). l 4 The competition products were analyzed 4.03 ppm (chelated ester) concomitantly disappears as spectrophotometrically both before and after separamethanol is formed (3.25 ppm). N o other signals of tion on an ion-exchange column. At least 9 8 % reshort duration appear during hydrolysis. The final covery of products was obtained from the ion-exchange spectrum was identical with that of [Co(en),gly]'+ and experiments except in experiments 16 and 17 when 95 methanol. and 93 recoveries, respectively, were achieved. The infrared spectrum of [C0(en)~(glyOCH~)](Cl0~)~ in anhydrous acetone, Figure 2, shows a sharp absorp(14) D. A . Buckingham, I. I . Olsen, A . M . Sargeson, and H. Satrapa, tion at 1630 cm-', which is assigned to the ester carInorg. Chem., 6, 1027 (1967). Buckingham, Foster, Sargeson

Hydrolysis of Chelated Glycine Esters

6036

l l

d

r!

+

h

;

-L) N

Journal of the American Chemical Society

90:22

1 October 23, 1968

+$

6037 Table V. OI8 Enrichment of Reactants and Products for the Experiments 18-21 establish that no C1-, Br-, NOa-, Hydrolysis of Carbonyl-Ol8-Labeled and SOr?- competition occurs during the acid hydrolycis-[C~(en)~Br(glyOCH~)lO, sis of [ C ~ ( e n ) ~ ( g l y O C ~ Hand ~ ) ] ~[Co(en)2(glyOCHs)l3+. + --Ea---, Other results (Table IV) show that, if competition did Reaction Solvent Productd r b occur, the expected products [C~(en)~X(glyOR)]~+*+ 1 glyOMe 0.633 or [C~(en)~X(glyOH)]~+*+ (X = C1, Br, SO,) are stable 0.622 under the experimental conditions and are readily deHzO-[Co(en)2(glyOCH3)]3+ 0.580 93 2 tectable by the ion-exchange technique. z+ 0.595 95 3 Hgz+-[Co(en)2Br(glyOCH3)] N o competition by SO4?-or NOa- for the intermedi4 H0C1-[C0(en)~Br(glyOCH~)]~+ 0.596 95 ate generated by the Hg2+-assisted removal of Br- in HOCI-[CO(~~)~B~(~~~OCH~)]~+ 1.393 0.707 102 [C~(en)~Br(glyOC~H,)] ?+ and [C~(en)~Br(glyOH)]~+ was a Represents the 0 ' 8 enrichment in atom of the species, less observed (experiments 2-4, Table IV). This is conthe atom % 0 1 8 in COP of normal isotopic composition. * Repretrary to the established competitive properties of NO3sents the 0 1 8 retention of the product species as a percentage of and SO4?-in the reactions of Hg2+with [Co(NH3);XI2+ 01*enrichment in the initial complex (taken as 0.627 atom %) or solvent. Labeled HOCI, unlabeled complex. [C~(en)~gly](X = C1, Br, 1)l4 and with experiment 6 (Table IV) (HgIa) in experiments 3-5. H ~ ]formed ~+ from where 47 % [ C O ( ~ ~ ) ~ N O ~ Nwas [ C O ( ~ ~ ) ~ C ~ in N the H ~presence ]~+ of 5 M N a N 0 3 . Oxidation of [C~(en)~Br(glyOC~H~)]*+ and [Co(en>,ever, the enrichment of the uncomplexed glycine ester Br(glyOH)]?+ with HOCl in 0.1 M HClO, gave [Co(en),is consistent with the enrichments found for glycine gly]*+ quantitatively (experiment 7, Table IV). N o isolated from the products of the hydrolysis reactions. [ C ~ ( e n ) ~ C l ( g l y O C ~ Hor ~ > ][C~(en)~Cl(glyOH)]~+ ~+ was Entries 2-4 establish that [Co(en),gly12+ recovered formed during or subsequent to the removal of Br-. after acid hydrolysis of carbonyl-0 18-labeled[Co(en),These species do not react with HOCl over similar (glyOCH3)](C10q)3 and Hg2+ and HOCl treatment of periods of time (experiment 13). Also, the quantitative carbonyl-O1s-labeled [C~(en)~Br(glyOMe)]Br~, respecformation of [ C ~ ( e n ) ~ C l ( g l y O C ~ H ~on )]~+ treating tively, retains almost complete enrichment. In ex[C~(en)~Br(glyOC~H,)]~+ with Clz (experiment 14) is in periment 5 , unlabeled complex was treated with 0lsgeneral agreement with the similar results obtained labeled HOC1, and the result shows that only one oxywith [ C O ( N H ~ ) : B ~ ] ~ + and [ C O ( ~ ~ ) ~ N H ~l 6B The ~]~+. gen atom from the solvent is incorporated. presence of C1- in the HOCl oxidation results in apKinetics of Oxygen Exchange in the [ C ~ ( e n ) ~ g l y ] ~ + preciable amounts of [C0(en)~Cl(glyOC3H7)]*+being Ion. The results of oxygen exchange between HzO formed (experiments S-lo). This could arise either and [ C ~ ( e n ) ~ g l y ]in~ +0.1 M HCIO? at p = 1 and 25" from Cls oxidation (produced by HOCl oxidation of are given in Table VI. The two sets of results were Cl-) or by a mechanism involving more direct C1- participation. That added anions do compete in the HOCl reaction is shown by entries 11 and 12 in Table Table VI. Kinetic Data for Oxygen Exchange between Water IV. Although a quantitative estimate of the amount and the Carboxyl (A) and Carbonyl (B) Oxygen of [C0(en)2NO~(glyOC~H~)1~+ formed was not possible of [Co(en),gly]z+ since it could be separated from [Co(en),gly]?+ by the Days Rb Atom % 0 ' 8 ion-exchange technique, the low molar absorptivity of the resulting solution (eJB7 92) suggests -33% of [CO(en)2N03(glyOC3H7)]2+. l7 A more conclusive result was obtained with added SO4?-( 5 M , experiment l l ) , where [C0(en)~SO~(gly0C~H~)]+ was separated from [ C ~ ( e n ) ~ g l yby] ~ion ~ exchange at pH 7. A 99% re0 0.014928 0.540 covery was obtained and 20 [C0(en)~SO~(glyOC3H7)]+ 1 0.014731 0.530 was formed assuming 61 as found.for [Co(NH3)50.509 3 0.014301 SO,]+. l 9 This result establishes that Sod2- and HS0,4 0.014308 0.509 compete favorably with the carbonyl grouping in the 0.014054 0.497 9 0.014330 12 0.510 HOC1-assisted removal of Br-. 0.013899 16 0.489 OI8-Tracer Experiments. Results obtained for the 018-labeling experiments are given in Table V. The first entry represents the 0 1 8 content of the carbonyl(B) Co, labeled glycine methyl ester used for the complex 0-C \@* preparations. Recovery of the ester by pyrolysis (250") of [C~(en)~Br(glyOCH~)]Br~ was only partly 0.014937 0.540 successful and was not reproducible ( E = 0.269, 0.445). 0.015326 1 0.559 2 0.014545 0.521 This may have resulted from contamination of the dis0.012725 5 0,431 tilled ester with unlabeled methanol or water. How~

Hv-r

O C

A. Haim and H. Taube, J . A m . Chem. SOC.,85, 3108 (1963). (16) J. F. Remar, D. E. Pennington, and A . Haim, I n o r g . Chem., 4, 1832 (1965). (17) This estimate is based assuming similar absorptivities for [Co(en),N03(glyOC3Hi)l'+ and cis-[Co(en)gN03NH3]2+ at 485 mp. This assumption is thought to be a reasonable one since the two systems [Co(e17)~X(glyOC~Hr)]*and [Co(en)zXNHa]*+(X = C1, Br) have similar absorptivities for the first visible absorption band. '8 (18) M. D. Alexander and D. H. Busch, I n o r g . Chem., 5, 1590 (1966). (19) R . Tsuchida, Bull. Chem. SOC.Japan, 13, 388 (1938). (15)

9

14

17 21

0.009631 0.007727 0.007055 0.005993

0.278

0.184

-

0.151 0.098

[Ht] = 0.1 M , 1.1 = 1.0, 25", [Co(en),gly2+] 0.1 M . * R = [46]/([44] [45]). e The H2018 solvent used to prepare this comOI8. d Represents the 0 1 8 enrichplex contained 1.122 atom ment in atom % less the atom % 0 1 8 in C 0 2 of normal isotopic 0 1 8 = 100R/(2 R). composition (0.201). Atom

+

Buckingham, Foster, Sargeson

z z

1 Hydrolysis of

+

Chelated Glycine Esters

6038

I

as an intermediate in 1 therefore a necessary

the latter two reactions and is precursor to hydrolysis. This is consistent with the proposals made by Alexander and Busch for the Hg2+-promoted reaction on the basis of infrared evidence (cide infra). In addition, the lack of competition by Br-, C1-, and NOs- in the acid hydrolysis of [ C ~ ( e n ) ~ ( g l y O C H ~and ) ] ~ +[C~(en)~(glyOC~H,)] 3+ suggests that either the chelate ring remains intact before, during, and after hydrolysis or that the carbonyl group is a much better competitor than the anions or solvent. Hg2+-Assisted Reaction. Hydrolysis of the ester function cannot precede removal of Br- since [Co(en),Br(glyOR)] 2+ hydrolyzes very slowly in acid solution in the absence of Hg2+,' nor can it occur in the resulting short-lived intermediate of reduced coordination number since the ester function is still intact in the subsequent [C~(en)~(glyOR)]~+ complex. There is ample evidence to conclude that the fast Hg2+-promoted reDAYS mokal of Br- leads to the formation of a five-coordinate Figure 3. 0 ' 8 exchange in (A) [ C ~ ( e n ) g l y ] ~prepared + from intermediate. 2 1 - - ? 3 Also, it has been established in the carbonyI-Oi8-[Co(en)~Br(glyOCH,)]zfand normal water; (B) similar reactions with c i s - [ C ~ ( e n ) ~ N H ~ X(X ] ~ += C1, [Co(en)gly]Z+ prepared from unlabeled [Co(en)~Br(glyOCH)~]~+ Br)16,24 and ~ ~ U ~ ~ [ C O ( N H ~ ) ~ ( (X N D = ~ C1, ) XBr, ]~+ with H2Oi8 (0.1 MHC104, ~.r= 1.0, 25.0'). that both the stereochemistry and optical properties are preserved in this type of intermediate. Applying these considerations to the present reaction eliminatesthe symobtained from [C~(en)~gly]?+ prepared by treating ~arbonyl-O~~-[Co(en)~Br(glyOCH,)](ClO~)~ with Hgz+ metrical trigonal bipyramid as a possible intermediate but does not further define its nature. We have chosen in water of normal enrichment, or by similar treatment to represent it as a square pyramid. This species may of the unlabeled complex with H20.l8 A plot of log then react with the solvent to give [Co(en),(OH,)(gly(0l8complex) against time was linear (Figure 3) for OR)I3+, with the carbonyl grouping to give direct forthe complex obtained from the labeled solvent, giving mation of the chelate ester [Co(en),(glyOR)I3+, or by a a first-order rate constant of 1.0 X sec-1 at 25". combination of these two paths (Scheme I). In the The complex obtained cia the labeled ester showed Little exchange in 16 days under the same conditions. Scheme I It is apparent from the results that there is little or n o mixed label in the two instances and that the two oxygen sites in the glycinato chelate can be distinguished. Discussion Each of the reactions

+

OH1

[Co(en)~Br(glyOR)]~+ Hgz-

+ H D --+

+

[ C ~ ( e n ) ~ g l y ] ~ -HgBr+

+

3+

+ HOR

(3)

+ ROH

(4)

Y

I

I I

[C~(en)~Br(glyOR)]~+HOC1 +

+

[C~(en)~gly]~+ BrCl

+

+

[ C ~ ( e n ) ~ g l y ] ~ ROH +

+ HT

(20) D. A . Buckingham, D M Foster, and A M Sargeson, unpublished results

Journal of the American Chemical Society

1 90:22

,

(5)

results in the formation of [C~(en)~gly]?+ as the only cobalt-containing product. This was established spectrophotometrically for (3) by Alexander and Busch and verified for all three reactions by ourselves using the ion-exchange technique. If the aquo glycinato ion had been formed, it would have been detected by the ion-exchange experiments. [C~(en)~(OH)(glyO)]+1s eluted more rapidly than [ C ~ ( e n ) ~ g l y ]using ~+ 1 M NaCIOi at p H 9.20 Since each of the above reactions differs at least in principle, it is desirable to discuss some aspects of each reaction separately. Acid Hydrolysis of [C~(en)~(glyOC~H,)]~+. The agreement between the rate data for this reactlon (Table I) and that for the Hg2+-and HOC1-promoted reactions (Table 11) indicate that [C~(en)~(glyOC~H,)]~+ is formed

+

W

[C~(en)~(glyOR)]~+ HzO --f

31

o=c, 'OR

event of [C~(en>~(H~O)(glyOR)]~f being formed, the kinetic data require that this species lose water in a fast reaction to form [C~(en),(glyOR)]~+.If this aquo product were formed, it would also be expected that added anions would compete favorably for the five(21) F. A. Posey and H. Taube, J . Am. Chem. SOC.,79, 255 (1957). (22) D. A . Buckingham, I . I . Olsen, A . M. Sargeson, and H. Satrapa, Inorg. Chem., 6 , 1027 (1967). (23) D. A . Loeliger and H. Taube, ibid., 5 , 1376 (1966). (24) D. A. Buckingham, I. 1. Olsen, and A . M. Sargeson, J . A m . Chem. SOC.,press. (25) Buckingham, I . I. Olsen, and A. M. Sargeson, Australian J . Chem., 20, 597 (1967).

October 23, 1968

6039

not compete in the ion-pair interchange process, alcoordinate intermediate generated on loss of Br-. Evidence to support this argument is the demonstration though anions such as S04,- do. Also S042- and NO3- do compete in the HOCl reaction, and the comthat SO4,- and NO3- are better competitors than HzO petitive properties of these anions are similar to those on a molar basis for the [ C O ( N H ~ ) ~2]2 ~and + [Co(en),NH3I3+intermediatesz4 (also Table IV, entry 6 ) genfound in the Clz reaction. If these results are now compared to the analogous [ C O ( N H ~ ) ~ I system, ]~+ similar erated in the same way. In the absence of other concompetition values for the incorporation of sulfate are siderations, there is no reason to expect the five-coordinate ion [C~(en)~(glyOR)]~+ t o behave differently. observed. l5 Also for the HOC1 reaction with [Co(en>zThe [C~(en)zNO~(glyOR)]~+ and [C~(en)~SO~(glyOR)l+NH3BrI2+both acido and aquo products are observed. l6 products expected from such anion competition reacThis implies that a similar pattern of events occurs with tions are stable under the experimental conditions. [C~(en)~Br(glyOR)]~+ and that some aquo glycinato Thus, the absence of SO4,- and NO3- competitionz6 ester complex is formed. suggests that HzO is not involved, and that the ester Tracer Studies. The 0l8 tracer results (Table V) grouping, due to its proximity, is a much better comshow that in all cases examined one oxygen is enriched petitor than any other species in solution.26a in the glycinato product. The 0l8exchange rate in Oxidation by C12 and HOCl. The reaction, cis0.1 M HC104(Table VI) shows that the product formed [ C ~ ( e n ) ~ B r ( g l y O C ~ H ~ ) ]C12, ~ + is striking in that cisuiu labeled solvent exchanges considerably faster than [C~(en)~Cl(glyOC,H,)] *+ is quantitatively formed (Table the product formed uiu labeled ester, and this property IV, entry 14). This result may be compared with the is ascribed to the two sites for the glycinato oxygen quantitative formation of [ C O ( N H ~ ) ~ C I from ] ~ + [Coatoms. The OL8oxygen from the enriched ester is (NH3)jBr]2+ C1, l 5 and with c i s - [ C ~ ( e n ) ~ C l N H ~ ] ~bound + to cobalt and the 0l8from the solvent is in the formed from cis-[C~(en)~BrNH~]~+. l6 The remarkable carbonyl position of the chelated glycine. feature of the C1, reaction is the complete lack of comThese results establish both the nature of the chelated petition provided by the ester carbonyl group. Thus ester intermediate and the main features concerning its the intermediate involved cannot be the same as that subsequent hydrolysis. The position of incorporation proposed for the Hg2+-assistedreaction, and it is likely of one solvent oxygen atom clearly eliminates the ester that it is not one of reduced coordination number. This intermediate in which the ether oxygen rather than the supports Haim and Taube's p r o p o ~ a l that ' ~ the product >C=O oxygen is involved in chelation. Also, it is is produced by interchange within the ion pair evident that formation of the chelated ester in the HOCl reaction must involve rapid Co-OH2 bond rupture in Co3+-Br----Cl+ ---+ Co3+-cl----Br+ the aquo ester precursor and that a mechanism such as A similar mechanism was proposed for the HOCl and H z 0 2 oxidation of coordinated Br- and I-, namely

+

+

Co3+-Br----HO+---Y- +Co8tOH----Br+---Y(Y = C1, OH)

and in acid solution [Co(NH3),(OH2)I3+is quantitatively formed.15 In the case of the [ C O ( N H ~ ) ~ I ] ~ +HzOz reaction, oxygen-tracer experiments established that at least 62% of the aquo product formed by this path derived its oxygen from the oxidizing agent. A similar tracer experiment was not possible for the HOCl oxidation due to rapid oxygen exchange with the solvent, but Haim and Taube clearly demonstrated the similarities between all three oxidants, Clz, HOCl, and HzOz. It appeared then that the HOCl oxidation was an ideal method of generating the aquo complexes [Co(enb(OHn)(glyOR)I3+ from the corresponding bromo derivative. The results of the reactions given in Table I1 clearly demonstrate that the chelated ester [Co(en),(glyOR)I3+ is the species observed in the HOCl reaction. Such a result may be accommodated by a mechanism whereby the coordinated water molecule in cis-[Co(en),(OH,)(glyOR)I3+ is rapidly displaced by the adjacent >C=O group in a concerted manner. An alternative to this is that the carbonyl group is involved in the ion-pair interchange without intervention of [C0(en)~(0H~)glyOR] 3+. This latter possibility seems unlikely. The analogous reaction with Clz shows that the carbonyl group does

+

Co

CH2

I I H

IC) I OR

H-0-C-0

/Nf - '\ /CH, + 0-c

HzO

I

OR

does not obtain. This is required .by the full retention of the isotope label in the glycinate product (Table V). In view of the relatively fast rate of water exchange required here < 30 sec), compared to that found for the analogous [Co(NH&(H20)l3+ion (t1la 35 hr,

-

Scheme LI

0-c \o

I

OR

(26) The added anion concentration was ca. five times that used in ref 24. (26a) NOTEADDEDIN PROOF. (-)~-[Co(en)~Br(glyOCH3)]Brz,[alD - 115", gave (-)D-[CO(en)~gly]~+,[ c r ] ~- 306",when treated with Hg*+ in 0.01 M HClOa. The product is optically pure, indicating full retention of configuration during the reaction. This experiment will be reported in detail elsewhere.

Buckingham, Foster, Sargeson

zJJ@

0-c

0-C

\O

\@

Hydrolysis of Chelated Glycine Esrers

6040

25”),27we suggest that the H 2 0 molecule is displaced in a concerted s N 2 process. The chelated ester intermediate may further react by aquation (path l), by synchronous hydrolysis and aquation (path 2), or by direct hydrolysis without opening of the chelate ring (path 3), Scheme 11. Path 1 leads either to a product containing the label entirely in the carbonyl position (attack by bound water at the ester carbon) or to a product containing a mixed label (following hydrolysis of the monodentate ester), while path 2 leads to a mixed labeled complex or to a product containing half the enrichment entirely in the carbonyl oxygen. Both these possibilities are eliminated by the results. Path 3, in which the chelate ring remains intact throughout hydrolysis, is the only path which satisfies the results. It allows incorporation of the labeled solvent solely in the carbonyl oxygen of the product with full retention of the labeled ester attached to cobalt. Concluding Remarks A summary of the conclusions for some aspects of this investigation is warranted. Clearly the isolated chelated ester intermediate hydrolyzes at the same rate as the species formed from the Hg2+- and HOC1-assisted aquation of the bromo ester complexes. It appears that in each case the chelate skeleton remains intact before, during, and after hydrolysis. Also, the labeling experiments establish that C-OR bond cleavage occurs for the methyl ester. A similar result might be expected to hold for other alkyl groups, except perhaps t-butyl glycinate. Activation for the rapid hydrolysis presumably arises from the effect of the Co(II1) ion on the coordinated carbonyl group. The independence of the rate of hydrolysis on the H+ concentration suggests that OHz is the nucleophile attacking the carbonyl carbon. The absence of a term first order in [H+], which is found in the free ester hydrolysis, can be accounted for by substitution of the cobalt(II1) ion for H+ in this mechanism. The considerable enhancement in rate which is observed in the present instance presumably arises from the fact that all the ester is present in the “protonated” form, whereas in the acid-catalyzed path for the uncoordinated ester only a small proportion is present as protonated ester. Preliminary results appear to indicate that at higher pH values a substantial increase in rate occurs. (27) H. R . Hunt and H. Taube, J. A m . Chem. Soc., 80,2642 (1958).

Journal of the American Chemical Society

1 90:22

This path may coincide with the [OH-]-dependent path observed in uncoordinated ester hydrolysis. In both the acid- and base-dependent paths for hydrolysis of 0 18-labeled alkyl benzoates, the equivalence of the two enol oxygen atoms in the intermediate RC(OH)20R has been established. 28 However, this situation is not possible for the Co(II1) chelated ester, and our experiments do not lead to any conclusions about the presence of the intermediate 111 shown below.

n

N

Similar studies for the more labile Cu(I1) ester complexesZ9indicate that the “enol” oxygens do become equivalent, possibly through one-ended dissociation of the chelate 111. It seems probable therefore that I11 is formed for the cobalt (111) complex but is very labile toward dissociation. Finally there is a similarity between the chelated ester intermediates and complexes involving carbonyl-oxygen chelated amino acid amides and peptide esters30 (structure I). Accelerated hydrolysis of these species has been reported,2 and it is tempting to suggest that the order of events established for the hydrolysis of the chelated esters applies to these related compounds. Some of these complexes are at present being investigated. Acknowledgments. The authors wish to thank the Microanalytical Unit of the John Curtin School of Medical Research, Australian National University, for the N, C, and H analyses, Mr. S. Brown for the pmr spectra, and Dr. F. J. Bergersen of the CSIRO Division of Plant Industry for the mass spectra. (28) M. L . Bender, ibid.,73, 1626 (1951). (29) M . L. Bender and B. W. Turnquest, ibid.,79, 1889 (1957). (30) M . Fehlman, H. Freeman, I. Maxwell, P. A. Marzilli, P. A . Buckingham, and A. M. Sargeson, Chem. Commun., 488 (1968).

October 23, 1968