NOTES
560
T-ol. 65
plexes these reactions would be expected t o have This substance could then react rapidly with hya type S N 1 mechanism. There would then be a drogen peroxide to form oxygen competition between coordination of the remaining fast carboxylate group on the ligand and solvation YCOOO-~ HzOz YCOOH-2 t Oz OH- (3' by a mater molecule for the cobalt atom with Reaction 3 is analogous to one proposed for the coordination number five. decomposition of hydrogen peroxide by Reactions 2a and 2b are the acid and base hy- and to the reaction between CO(KH3)&204+ drolysis of tlhe intermediate, YCOOOCOY-~. If and hydrogen peroxide." Reaction 3 also satis(2) is ignored or if the rate constants for ( 2 ) and fies the observed stoichiometry ( n = 1.5) in the (2a) are of the same order of magnitude, then the of excess hydrogen peroxide. observed rate constants, ka and lib, are of the same presence the other hand the stoichiometric reaction orders of magnitude as the acid and base hy- (nOn = 0.5) between cobalt(I1) and hydrogen perdrolysis consi;ants of other ethylenediamine comin the presence of a large excess of cobalt(I1) plexes of cob,ilt(III) and reaction 2b may also be oxide suggests the occurrence of of type S N ~ C B . ' ~ fast Reaction 2c satisfies the observed kinetics with .OH respect to OOH- for the formation of oxygen and YCOOOII-2 YCoH20-2 -+2YCoOOH-* (4a) the second step in the formation of cobalt(II1). fast Although some doubt has been raised as to whether YCOHZO-' .OH -+YCOOH-' Hz0 (4b) substitution hy reagents other than hydroxyl ion occurs in reactions with cobalt (111) c~mplexes,'~The formation of free radicals in reaction 4% aiid there had been a recent report of such a substi- their subsequent reaction with hydrogen peroxide tution hy cyanide Also observations in this would also account for the variations in the amount Laboratory indicate that C O ( ~ \ ; H ~ ) ~can C ~be + ~ of hydrogen peroxide decomposed at different formed directly from (NH3)6CoOOCo(NH3)6+4. pH's. Acknowledgment.-The author wishes to acAn alternate reaction to (2c) is knowledge support of this project by Research YCOOOC'OY-' f OOH- + Hz0 + Grant RG-4458 from the U. S. Public Health ~YCOOH-' 02 OH- Service. The author also wishes to thank the Fels However, experiments in this laboratory show that Institute, Yellow Springs, Ohio, for the use of the ( N H ~ ) & o O O C O ( N H ~ is ) ~ +not ~ involved in the Warburg apparatus and the Kettering Institute, catalytic devomposition of hydrogen peroxide Yellow Springs, Ohio, for the use of the polarograph. Others who have contributed to the eyby ammoniacal solutions of cobalt(I1) sulfate. In reactions (2)-(2c) the reactive product is perimental work include Margaret Warga and YCOOO-~which is in equilibrium with YCOOOH-~. Roger Bakeman.
+
+
+
+
+
+
+ +
(14) F. Basolo and R. G. Pearson. "Mechanisms of Inorganic Reaotions," John Wiley and Sons, Inc., New York, N. Y., 1958 p 125 and 130. (15) D.A L. Hope and J. E. Prue, J . Chem. SOL., 2752 (1960).
(16) B. Chance and R. R. Ferguson in "The hlechanisin of Enzyme Action," edited by W.D. MoElroy and B. Glass, The Johns Hopkins Press. 1954, p. 359. (17) P. Saffir and H. Taubc, J . A m . Chem. Soc , 8 2 , 13 (1960).
NOTES ICE
1711
[--AN ACETOKE HYDRATE?
B Y .kRTlIU
s. Qursr .4ND
I ~ E N Rs. YFRAYK
Departmen' of Chemzitrw, Universaty of Pittsburgh, l'zttsburgh Pennsylvanza Received August IO. 1060
IS,
In 1938 Cohcn and van der Horst' described a solid material prepared by cooling acetone-water solutions to temperatures near -35'. X-Ray powder patterns, density determinations and chemical analyses of their finely divided crystals led these authors to conclude that they had obtained a new solid phase of water, which they called Ice VIII. They reported that this ice belonged to the cubic system, with a = 9.68 A. at -35' and a density of 1.015 g./cc. a t -30". f l ) E. Cohen and C.J. G van der Horst, L.physik Chem., B40,231
(1938).
We have also prepared crystals belonging to the cubic system by the slow cooling of GO weight yo solutions of acetone in water. In contrast to the tiny crystals obtained by Cohen and van der Horst, our procedure yielded relatively large, well-formed, octahedral crystals (usually obtained in a range of sizes, up to approximately 1 mm. in edge length) which were suitable for single crystal X-ray studies. Preparation of Sample and Procedure for X-Ray Measurements.-The octahedral crystals were best prcpnr cd when the acetone-water solution was contained in a triplewalled glass vessel (two air spaces) which was immersed in a cold bath, the temperature of which usually was kept n w r -40". Suitable crystals seemed to be obtained more readily if the acetone-water solution was first frozen solid and thtm melted ticforc the slow cooling to near -40" was carried out. These crystals could be isolated by pouring the suspenfiionof the crystals in the mother liquor onto an aksorbent paper in :t cold box held to a temperature near -35 Single crystals
.
March, 1961
?;OTES
were placed (without being permitted to warm up) in glass capillary tubes which were then mounted on the goniometer head of an X-ray camera. Weissenberg photographs were taken, using copper Ka radiation. The X-ray camera was eiiclosed in a specially constructed, double-walled, Plexiglasj box which could be cooled to approximately -35 to -40’ by the use of solid carbon dioxide. Several different single crystals were studied, all of them freshly prepared and most of them approximately 0.5 mm. in diameter. Measurements on one crystal usually required 10 to 12 hours. Unit Cell and Space Group.-Weissenberg photographs obtained with the morphological twofold and fourfold as rotation axes showed diffraction patterns that were consistent with m3m Laue symmetry. When indexed on the basis of a cubic cell, these data contained only reflections of the type h k = 2 3 , k I = 2n, and h 1 = 2n for the general ( h k l )class, and only reflections of the type IC I = 4n for the (Okl) zone, thereby uniquely fixing the space group as Fd3m within the assumed octahedral system. All observable powder diffraction lines were consistent with the above extinctions and were indexable on the basis of a cubic cell with the dimensions 17.16 which was determined from oscillation and Weissenberg photographs. These diffraction data indicate that the structure under investigation is isostructural, a t least in its water framework, with one of the “liquid hydrates” described by von Stackelberg.2 3 The cell dimensions, which are presumably determined by the water structure alone, are identical within experimental error in the two cases. In order to extend the comparison between this structure and the “liquid hydrate” of von Stackelberg, representative three dimensional intensity data for the crystal were qualitatively estimated and compared with the spectral intensities calculated from the model described2 by von Stackelberg for the oxygen framework of the “liquid hydrates.” Lorentz and polarization corrections were applied to the square of the calculated structure factor amplitudes, in order that the calculated quantity might be more realistically compared t o the observed values. The intensity distribution for the calculated and observed spectra were remarkably similar, further substantiating the isostructural relationship that appears to exist between this crystal and the hydrates of chloroform, dichloromethane, ethyl chloride, etc., that were studied by von Stackelberg. The unit cell of these crystals contains 136 water molecules, which form a framework of slightly deformed pentagonal dodecahedra. These pentagonal dodecahedra share faces to form a unit cell containing 8 large voids (in the form of hexadodecahedra, with 12 pentagonal faces and 4 hexagonal faces) and 16 small voids (spaces inside the pentagonal dodecahedra). If each large void contains a molecule of acetone, the formula of the hydrate will be (CH&C0.17Hz0. Density and Composition of the Crystals.-The crystal density was determined by flotation in organic liquids. The measured density was found to be quite variable, decreasing as the period of time increased between the isolation of the crystal and the density measurement. The maximum observcd value of the density, found when freshly prepared crystals were used, was 0.95 g./cc., which corresponds to what would be observed if all eight molecules of the acetone were present per unit cell. Crystals which h t l been allowed to stand for several hoiirs in a partial varuum nc:w temperatures of -30 to -40” gave a densit77 of 0.80 g./cc., corresponding to thedensity of the framework, i.e., to the complete “evaporation” of acetone. ‘Lhe composition of the crystals was determined hy t h r use of a gas chromatograph (Burrell Kromo-Tog, Model K2). A column consisting of Tween on Celite separated the acetone and water peaks completely, with only a slight tailing of the wjter pclak. For complete fillin2 of the large voids in the crystal, a value of 15.9 weight 7 0 acetone would be expected. KO value as large as this was observed, whether the rrvstals were d a t i v e l y fresh, or had been standing for some time a t Dry Ice temperatures, the usual result being between 9 and 14 weight q0 scctone. This result is to be expected for well-drained crystals, considering thr fact that the crystals lose ac*cLtoncreadily, ,and that whrn thc crystal is melted for :inalj pis, acetone is thc morr volatile component and woiild tend to he lost prefcwntially. Therefore,
although the experimental results do not give the expected value, one may conclude that the value 15.9% probably would be approached under the most favorable conditions. From the density determinations and the chromatographic analyses, it was concluded that the crystal structure under consideration was that of an acetone hydrate, (CH&CO. 17H20. Because the crystals seemed to lose acetone so readily, it was of interest to determine whether the loss of the acetone had any effect upon the structure of the crystal. Crystals that had lost most of their acetone (as determined from density measurements) retained their sharp, octahedral appearance, but were then opaque. This opacity could be due either to frost coating the crystal, or to changes in the interior of the crystal, or to both factors. In order to determine the effect of the loss of acetone upon the crystal structure, the following experiments were carried out: Dry, cold nitrogen gas was passed slowly over a collection of these crystals held at a temDerature between -35 and -40”. After 7 hours, the density and composition of some of them were measured, and X-ray photographs taken of others. For 10 crystals, the range of densities corresponds to 0 to 1 acetone molecule per unit cell. Gas chromatographic analyses on another sample corresponded to 1% by weight acetone present, or less than 0.5 molecule of acetone per unit cell. The X-ray photographs of the crystals, however, showed the powder pattern of hexagonal ice, and contained only a few spots corresponding to the acetone hydrate structure, indicating that the framework structure was not stable at -35’ when most of the acetone molecules had been lost. This does not necessarily mean, however, that a small amount of acetone cannot be lost without destroying the structure. As stated above, the X-ray measurements which showed good hydrate structure usually required a single crystal to stand for 10 to 12 hours, and other crystals which had stood as long as this typically showed densities corresponding to the loss of, say, 20% of their acetone. This may have been responsible for the small amount of diffraction from hexagonal structure always present on the X-ray photographs, but the latter also might have come (as we imagined) from the unavoidable frost on the capillaries. We are therefore unable either t o affirm or t o deny that the structure can survive small losses of acetone. The fact that fresh crystals seem to have their full complement of acetone is in keeping with Glew’s finding,4 that the hydrate of bromochlorodifluoromethane, which has the same 17 A. framework, displays essentially complete occupancy of the interstitial sites.
+
+
+
+
w.,
( 2 ) M . von Pt:ickelberi: a n d 11. R. AIuiIer. J . Chem. I’hys., 19, 1319 (1951); 2. Elektrochem., 6 6 , 25 (1954). 13) W. F. Cluussen, J . Chem. Phys., 19,259 (1951).
561
Discussion X-Ray powder photographs obtained in this Laboratory for the acetone hydrate correspond quite closely to the powder photographs published by Cohen and van der Horst. Since the acetone hydrate crystal has a face-centered cubic lattice, if its X-ray powder pattern were interpreted as resulting from a primitive cubic latticz, one could possibly obtain a value of a = 0.7 A. The analytical results of Cohen and van der Horst, indicating the absence of acetone in the crystal, were probably the result of the ease with which acetone escapes from the crystal preparations. The high density they reported may have arisen frcm some replacement of acetone by CC14, which mas a component of the mixtures they used for density comparisons (we succeeded in producing a little replacement of this kind, and the much larger surface-to-volume ratio in their material would have favored such a process). The fact that they also obtained cubic crystals from solutions of acetaldehyde, propionaldehyde or pyridine doubtless is accounted for by thc generality of the phcnomenon of hydrate formation. I t is also possible that the isotropic. cubes, r r t n hedra and tetrahedra obtained by Rau5 on wn4) D. N. Glew, Can. J . Chem., 88, 208 (1960).
562
KOTES
Yol. 65
densing water vapor onto a metal mirror at -72” are “liquid hydrate” type crystals, since other researchers6’ who sought to repeat Rau’s results could do so only when they used water contaminated with organic compounds, such as ethyl alcohol. Although double hydrates of acetone with scme of the inert gases have bgen prepared,s giring cubic crystah with a = 17.3 A., this is the first time, SO far as we can find, that the acetone hydrate has been reported. Acknowledgment.-Ke are grateful to Dr. Richard K. 3fcMullan of the Crystallography Laboratory for his help with the X-ray diffraction studies. -1S.Q. wishes to thank the Kational Scieiice Foundation for financial support.
The situation is more complicated than, but not essentially different from, that which is possible in simple substituted ethanes, as pointed out by P ~ p l e . The ~ non-equivalence in question persists when a time average is performed over ail (nine) staggered (or eclipsed) rotational isomers. Phenomena of this type x\-hich depend on the energetic non-identity of rotational isomers have been reported by a number of investigators.6-8 However, it has perhaps not been adequately recognized that when the symmetry of substitution is sufficiently low, the non-equivalence of methylene protons can in principle still persist when the isomers are all accidentally of equal energy, or even when the internal rotation is free.5 (This statement depends entirely on a symmetry argument. ( 5 ) I T Rau, ScTlr Deut. d k a d Luft , 8 , 65 (1944) It seems unlikely that in practice the asymmetry ( 0 ) H \1 Chilong, J Glaciol 1, 5J (1947) with respect to internal rotation required to produce ( 7 ) 1 TV B i c x c r and 11. P Pdiner, Proc I’hys Soc 648,705 (1951) ( 6 ) J. Cr \Talle- haluie, 166, 129 (1960). magnetic non-equivalence n-ould not also be reflected in some angular dependence of potential energy.) IKTERPRE’rATIOX OF THE MAGKETIC If the phenomena observed by Finegold are to RESONANCE SPECTRUM OF THE be explained in this way, it is clearly not necessary METHk’LEK’E GROUP IN CERTAIN that the molecule coiitaiii two methylene groups. In fact, similar compounds containing only one tXlYNT6ETRICALLY SUBSTITUTED methylene group should give ‘‘iiormal” methylene COMPOUKDS resonances if Finegold’s interpretation is correct, BY J. S. WAUGHAND F. A. COTTON but should continue to shorn non-equivalences if our alternative description is valid. To decide Department of Chemistry and Laboratoru f o r Chemzcal and Solzd Stal, Phusics, Massach i s e t t s Institute of Technology, Cambridge, X a s s a between these possibilities, we have observed the chusetfs proton resonance spectrum of C6H5-S(O)-OCHaReceived Sooember 11, 1960 CH3,9n-hich differs from diethyl sulfite only in Finegold has recently reported and commented replacement of one of the ethoxy groups by a on some apparently anomalous features of the phenyl group. The methylene resonance. shown methylene proton resonance in dialkyl sulfitesJ in Fig. 1, is readily recognized as attributable to a and in 0,0’-diethylmethylphosphonothioate,2 and pair of non-equivaleiit protons, in this case split similar peculiarities have been noted recently by to almost exactly the same extent by the methyl other investigators. These may conveniently protons. An analysis of the spectrum shons that be discussed with specific reference to the situation the chemical shift between these protons is 0.434 in diethyl sulfite. Instead of the 1:3:3:1 quartet i0.005 p.p.m. and the spin-spin coupling between expected for the splitting of the methylene reso- them is 10.0 f 0.5 c.p.s. Each of them is coupled nance by the adjacent methyl group, Finegold ob- to the methyl protons with a coupling constant of served h o such quartets, slightly displaced from 7.1 f 0.5 c.p.s. All the coupling constants are one another and showing slightly different cou- within the normal ranges’O l 1 for structures of this pling constants with the methyl group. That is, type. It is to be expected that the coupling coiione pair of methylene protons in the molecule is stants will be approximately the same as in diethyl clearly not equivalent to the other. Finegold sulfite, but the chemical shift between the made the natural assumption that each of these methylene protons is probably quite sensitive to pairs is in fact one of the methylene groups, and the substituents on the sulfur and may well be was thus forced to conclude that the two ethyl different. groups are ;,omehojv chemically different, Le., It is to be noted that a non-equivalence such as that thle two ,idfur-ethoxy bonds are differently Finegold proposed should lead to a simpler hybridized. methylene spectrum, inasmuch as the spin couWe feel it appropriate to point out that the above pling between non-equivalent hydrogens would be unipiiori is not necessary, and that an alterna- vanishingly small because of the large separation int crprel ation can be made and supported ( 3 ) J. A. Poplr, N o Z I’hys , 1 , 1 (1958) which does less violence to accepted theories of I’ XI Nair and J. 1) Rubeits. J . -In& C h t m S U L, 79, 45b2 chemical bonding. I t simply involves the obser- (1957) J N Slioolrrs and B L Crauford J I J 1101 vation that the two methylene protons of thc 270(7)(1967). same methylene group are not stereochemically (8) 1). RI. Graliam and J. b \t augli, J Ciirm i ’ h 7 / 3 , 27, 9bb equivalent, because of the lack of symmetry of the (1957) (9) This coinpound %as hindly prepared by Llr 4 Blake folloning (non-planar) substituted sulfur atom with respect t o internal rotation about the S-0-C linkage. the procedure of R Otto a n d h Rossine Ber , 18, 2493 (1885) and was characterized by C , H and S analsses. ((7)
(1) 1%.Iinezold, Proc. Chem. Soc.. 283 (1960) ( 2 ) H. Einegold, J. Am. Chem. Soc.. 82, 2641 (1900). (3) J. 1) Roherts ( t o be i>itblished). (-1) B. I hliailiro plirate communication.
(10) R. E Glirh and 4. A . B o t h e r - B y J (1956) (11) H
(1959).
C h ~ m I’hils
25, 382
S Gutoirsky h l Karplus and D R I Grant zhzd 31 1278
March, 19131
563
XOTES
DzO ISOTOPE EFFECTS I N THE CATALYTIC ACTIVATION OF MOLECULAR HYDROGEN BY METAL IONS BY J. F. HARROD AND J. HALPERS Department of Chemistry, The University of British Columbia, Vancouic r Canada Received December 2 f j 1960
With a view to gaining further insight into the role of t'he solvent in the catalytic activation of molecular hydrogen in solution,' t'he rates of reaction of hydrogen with a number of metal ions and complexes, in HzO and DzO, have been compared. Our results are summarized in Table I. TABLE
Metal ion
Fig. 1.
Rate law
1 Ternl,., OC.
kH20/ kD20
(10.1)
Ref.a
k[&] [Cu2+] 110 1.20 k[Hzl [Ag+12 50 1 23 Ag +* 75 1.26 k[Hzl &'+I Hg2 Hgz2+ k[Hz1[Hga2+l iJ 1.33 e Cu(0Ac)z k[H2][CU(OAC)~] 100 0.93 PdCLk[Hz] [PdCl,'-] 80 0.90 0 RhC1c3k[Hz] [RhC1e3-] 80 1.00 hLln04k[Hz][Mn04-] 50 0.99 Ag+ MnOd- k[H~][Bg+][hInO4-] 40 0.93 a Earlier measurements in HaO are described in the quoted reference. The same procedures were used in the present measurements and the results in HaO agree well with the earlier ones. While at higher temperatures, the reaction of Hz with Ag+ follows a rate law and mechanism siniilar to those for Cuz-, the predominant contribution under the conditions of this comparison is from the "termolecular" path which is believed to involve homolytic splitting of Hz according to equation 3 (ref. d, this Table). E. Peters and J. Halpern, J . Phys. Chem., 5 9 , 793 (1955) ; J. Halpern, E. R. Macgregor and E. Peters, ibid., 60, 1455 (1956). d A. H. Rebster and J. Halpern, ibid., 60, 280 (1956); 61, 1239, 1245 (1957). e G. J. Korinek and J. Halpern, ibid., 60, 285 (1956). E. Peters and J. Halpern, Can. J . Chenz., 33, 356 (1955). 0 J. Halpern, J. F. Harrod and P. E. Potter, ibid., 37, 1446 (1959). J. F. Harrod and J. Halpern, ibid., 37, 1933 (1959). A. H. Webster and J. Halpern, Trans. Faradall Soc., 33, 51 (1957). cu2+
+
of the proton pairs from one another. Thus it is at first sight surprising that Finegold mould not have been forced into our interpretation by t'he complexity of his spectrum. Unfortunately, in diethyl sulfite, the chemical shift' seems accidentally to be so small that the additional lines are vanishingly weak, so that' the two possible kinds of nonequivalence are not readily distinguishable from one another. From his figure it would appear that the relevant chemical shift is in the vicinit'y of 0.05 p.p.m. If we assume the spiii-spin coupling between the non-equivalent protons to be 10 c.p.s., as it is in our compound, it) is easy to predict12 that the satellite lines should have intensities of only 1-2%; of t'he central components. 1 3 Similar phenomena are observed in many other molecules.s,4 In diethyl acetal, for example, we have observed a complex met'hylene multiplet n-hich has been analyzed to give a chemical shift difference of 0.152 0.005 p.p.m., an internal spin coupling of 9.2 0.3 c.p.s., and couplings of about 6.7 and 7.2 C.P.S. between the non-equivalent methylene protons and the methyl group.'? It is significant that' the spectrum of ethylal, which differs from acet'al only in the substitution of a hydrogen atom for the central methyl group, contains a perfectly normal methylene quartet. l 4 This circumst'ance is completely in accord with the arguments preseiitcd here, since the ethylal molecule possesses too much symmetry to allow nonequivalence of thr protoiis in the same methylene group. We wish to thank ;he Sntioiial Science I-oundatioii for support of this work ( 1 2 ) J.
S.n'awli. "1'rocerdin:rs of the IV International Meeting on
1H;Y. Perganion Press, London, in press. (13) I h . Finegold fpiivate communication) has agreed t o the probable correctness of the interpretation proposed here. He has stated that i n many other phosphorus compounds he had studied ( e . q . .
(EtO)zP(O)Me. (F,t,O)zPMe. (ETO)zP(S)Cl, (EtO)zP(O)H and (Et0)zPiO)SJIe all methylene protons appeared t o be equivalent and that these obseryations caused him to choose his previously published explanation for his observations on (ETO)zP(S)Me rather than the one proposed here. (IS) C . S. Johnson. Jr.. J . P. Fackler, Jr.. J. S. Waugh and F. A . Cotton, unpublished work.
C F
'
+
J
It was hoped, in particular, that these measurements would provide a criterion for distinguishing those mechanisms2 in which Hz is split het'erolytically, transferring a proton to a water molecule, e.g. CU'+ -t HP
+ HzO +CuH+ + HaO"
(1)
from those in which a ligand other than water is believed to serve as the proton acceptor, e.g. Cil(OAc)z
+ Ha +CuH+ + HOAC+ OAC-
(2)
or in which H2is split homolytically without proton traii.qfer 2,4g+
+ Ha +2AgH+
3)
The results in Table I fail, however, to provide any clear-cut indication of such mechanist,ic differences. All the aquo ions included in the comparison (Cu2+, Ag+, Hg2+ and Hg2?+) shorn modest reductions in rate, ranging from 20 to 30%, on passing from HzO to D20, while the rates for complexes containing ligands other than water in (1) J. Halpern, J. P h y s . Chem., 68, 398 (1859): Aduances in Catalys i s , 9, 302 (1957); 11, 301 (1959). (2) Refs. r d Table I.
564
Vol. 65
SOTES
their inner coordination shells are nearly equal in the two solvents. This pattern, including the similarity of the isotope effects for Cu2+ and Ag+, suggests that differences in the coordinating properties of HzO and D2O (for which other evidence has been ciited3), rather than specific participation of water in the reaction, are largely responsible for the isotope effect in the case of the aquo ions. This is of interest in view of the many other studies* of H20-D20 isotope effects aimed a t elucidating the role of solvent in the mechanisms of various reactions of metal ions. Grateful acknowledgement is made to the Kational R,esearch Council of Canada and to the donors of the Petroleum Research Fund administered by the American Chemical Society for support of this work. (3) J. Halpern and A. C. Harkness, J . Chem. Phys., 31, 1147
The solution (0.5 ml.) was electrolyzed in millicoulometric cell. The reference electrode was a layer of mercury on the bottom of the cell connetted by a plastic tuhc to the mercury resewor. The anodic mercury w m moved slowly up and down at. one minute intervals by movement of the mercury reservoir. Owing to this agitation of the mercury layer, to the large diameter of the cathode capillary, and to the narrowness of the electrolytic cell, the solution was eontinuously but gently stirred during the electrolysis. Before electrolysis the solutions were freed from dissolved oxygen by bubbling hydrogen through them, and during the electrolysis hydrogen gas wa8 passed very slowly over the surface of the electrolyte. The hydrogen was partially saturated with ammonia by treatment with ammonium chlorideammonia buffer of the same concentration as the electrolyte solution. Polarographic curves were recorded before and after electrolysis
.
Results The effect of electrolysis on the diffusion current of Co(I1) ions is shown in Fig. 1. The solutions
(1959). ( 4 ) J. Hudis and R. W. Dodson. J . A m , Chem. Sac., 7 8 , 911 (195G); F. B Baker and T. 1%‘. Newton, J . Phys. Chem., 61, 381 (1957); A . Znirkel and IT. Taube. J . A m . Chem. Soc., 81, 1288 (1959).
POLAROGRAPHIC AND COULOMETRIC 1KVESTIC;ATIOKS OK T H E REDUCTION RATE OF COBALT(I1) IN T H E PRESENCE OF CYSTINE BY EMILIAN B. WERONSKI Department of C‘hemtstry. Unzverstty of Warszawa, Warszawa, Poland Recezved June 88,1060
Enhancement of the diffusion current has been found in “Brdicka’s electrolyte” and other cobalt solutions in the presence of some proteins and amino acids containing sulfhydryl or disulfide groups. This phenomenon has been attributed to the catalytic reduction of hydrogen,’ but the mechanism has not been elucidated.2 In the case of copper(I1) or bismuth(II1) ions in 1 N sulfuric acid, the author observed a similar “catalytic currelit” in the presence of very small concentrations of hydrocarbons, e.g., benzene, toluene, the xylenrs, n-pentane, n-hexane and cyclohe~ane.~ Since the lrttter “catalytic current” is due to an increase in the reduction of copper(I1) or bismuth(II1) ions, not to catalytic reduction of hydrogen, the following experiments were carried out to re-exmiine the current of cobalt ions in the presence of cystine. Experimental The solution studied was a modified Brdicka’s electrolyte: 0.0025 Jf cobalt(I1) chloride and 0.00001 M cystine in a 0.1 A i ainmonium chloride-0.1 M ammonium hydroxide buffer . Polarographic measurements, which followed standard practice, were made with a Cambridge Polarograph (No. C 466573). The capillary had an m-value of 2.06 mg. sec.-1 and a drop time of 4 5 sec. in twice distilled wster a t a h2ight of the mercurj reservoir of 60 cm. (closed circuit) a t 18
.
(1) R Brdickrr, Coll. Czechoslou. Chem. Communs., 5, 148 (1933). (2) R. Brdicka, abad., 11, 614 (1939); Chem. &%sly,34, 59 (1940); Klumpar, Coll. 17zechoslov. Chem. Communs., 13, 11 (1948); A. G.
Strombeia, Zhu7. F Z Z . Khzm.,20, 409 (1946); J. Heyrovsky, Coll. Czechoslov. Chem Communs., 9, 273 (1937); 111. von Stackelberg and 11. Fassbender, 2. Eleklrochem , 6 2 , 834 (1958). (3) Some of the experimental results submitted in partial fulfillrnent of the reiiuircinents for the degree of Doctor. E B. Weronski, Department of ~Cherriistrp,Cni\ersity \$ arsza\ca. 1957.
I
-1
1
-1.4
1
-1.8
V.
Fig. 1,-Curves 1 and 3 arc polarograms of Brdicka’s electrolyte before electrolysis; S denotes the sensitivity of the galvanometer. Curve 2 is the po!arogram of aliquots of Brdicka’s electrolyte after t - l . a s min. of electrolysis at -1.35 v., or after L1.6 min. of electrolysis u t -1.65 v.
were electrolyzed a t either - 1.35 v. (where the diffusion current of cobalt is not distorted) or at -1.65 v. (where the enhancement of current is maximal) until the diffusion current measured at -1.35 v. decreased by about 30%. illiquots electrolyzed a t -1.35 v. for t--1.35 minutes, or a t -1.65 v. for t--1.65 minutes gave identical polarograms (curve 2). We found that the amounts of electricity in-
March, 1961 volved in the two electrolyses are approximately the same: Le., h-l.sa X t--1.35 is about equal to h-1.66 X t - - l . ~where ~, h--1.35 and L 1 . 6 6 are the heights of the polarographic steps measured a t - 1.35 and 1.65 v., respetstively, before the electrolysis. The decrease in the diffusion current of Co(I1) ions at - 1.35 v. (Ah) is proportional to this quantity. If 1--1.35 is determined experimentally, can be estimated within a 10% error. These results suggest that the second wave (at -1.65 v.) of Rrdicka's electrolyte is due to an increase in the reduction rate of cobalt(I1) ions rather than to the catalytic reduction of hydrogen. It follows that the reaction is no more diffusion controlled and probably cobalt ions are provided to the electrode surface in a transport process similar to that ohserved in the case of polarographic maxima. SECONDARY PROCESSES IN GAS PHASE RADIOLYSIS OF HYDROCARBOSS BYJEAN H. FUTRELL Aeronautical Research Laboratoraes, Wmght-Patterson A F B , Ohao Receaved August 16, 1960
I n recent publications Back and Miller' and Back2 have shown that irradiation studies of hydrocarbon vapors must be conducted a t very low conversion if meaningful results are to be obtained. They have demonstrated that unsaturated products formed with very high yield initially in a-particle and y-radiolysis act as internal scavengers; thus a very low steady-state concentration of unsaturates is measured at conversions of a few per cent. This is very reasonably attributed to hydrogen atom scavenging by the unsaturated products. JTTehave confirmed these observations in recent studies in this Laboratory of gas phase radiolysis of n-pentane with cobalt-60 Such behavior, however, is distinctly different from that observed in an earlier study of normal hexane with 2 MeV. electrons from a Van de Graaff a~celerator.~Several exploratory experiments were performed in an attempt, to resolve this discrepancy and the results are reported in this communication. In y-radiolysis of gases in glass ampoules of modest dimensions, the major process of photon degradation is the ejection of secondary electrons from the walls of the container. Hence in our experiments using cobalt-60 y-rays the gas was exposed to a homogeneous flux of high energy electrons of 1 Mev. energy or less. Thus in both series of experiments the gases were irradiated with electrons, and the principal difference in the radiation sources employed is that the electrons from the Van de Graaff are initially mono-energetic while those from cobalt-60 have a broad distribution of energies. We are therefore reluctant to attribute the difference in radiolysis behavior to a simple difference in radiation quality. The electron and y-ray irradiations differed significantly with respect to two other experimental (1) R. A (2) R. h ( 3 ) .I. H. (4) J. €1.
56c5
SOTES
Back a n d N Viller, Trans. Faradal/ Soc., 65, 911 (1959). Bark, J . Phil? Chrm 64, 124 (1900). Futrell, z b z d . , 64, lfr74 (1960). Futrell, J . A m . Chcm. Soc.. 81, 5921 (1959).
1
0 ETHYLENE ACETYLENE 3 PROPYLENE
A
0
I
2
3
4 5 6 7 PER CENT DECOMPOSITION
8
9
IO
Fig. 1.-Yields of selected products from radiolysis of nparaffin hydrocarbons: A, ethane under all experimental conditions; B, ethylene from Van de Graaff electron irradiation using cylindrical reactors; C, inhomogeneous flux yirradiation; D, homogeneous flux irradiation; E, pseudohomogeneous flux electron irradiation.
parameters-radiation intensity and distributionand the presently reported experiments attempted to separate these parameters. The yields of unsaturated products as measured in four different types of experiments are presented in the accompanying figure. The data are presented as the quantity G/Go for convenience in presentation. ?? is the apparent 100 electron volt yield of a given product a t some particular dose, while Go is the yield of that product obtained by extrapolating the yield curves to zero conversion in each case. Curve D presents the data from 7-radiolysis of npentane in the low conversion region and affirms the conclusions of Back and Miller' for a-particle radiolysis of this compound. It illustrates the rapid attainment of a steady-state concentration of unsaturates undetectable by usual analytical methods a t conversions which have been used in many previous studies. Curve A presents the same quantity for saturated products ethane and propane a t these conversion levels for both the radiolysis of pentane and electron radiolysis of hexane. Curve B presents the yield of ethylene from hexane decomposition with Van de Graaff electron~.~ The Van de Graaff studies mere conducted with cylindrical Pyrex vessels 2.5 cm. in diameter and 80 cm. in length with the beam incident on a thin window (1 mm.) a t one end of the vessel. Because of scattering of the incident beam in passago through the window, most of the radiolysis occurred in the first few centimeters of the tube. In an attempt to simulate the flux gradient of the Van de Graaff reactors, a special ampoule 4.5 feet in length by 2.5 centimeters in diameter was constructed. When pointed a t the center of the cobalt-60 source the dose a t the most intense region was a factor of 50 =t20 higher than the dose a t the other end of the reactor compared to the analogous ratio of intensities of the order of 1000 for the electron experiments. The dotted curve C presents the data obtained in this experiment. These data suggested that the inhomogeneous flux might be responsible for the anomalous behavior in the Van de Graaff irradiations. Unfortu-
566
Vol. 65
NOTES
nately, however, size limitations in using the special target required the use of a more intense y-source, so that there still remained the possibility that source intensity is the major factor. Accordingly, spherical one-liter Pyrex reactors with thin bubble windows w r e constructed for use in Van de Graaff irradiations. While the radiation flux is still not homogeneoils in these reactors, this condition is much more closely approached than with the cylindrical reactors. With the same incident current of 10 microamperes used in the earlier hexane experiments, the cluster of points designated E was obtained for n-pentane radiolysis. These experiments indicate that a significant factor in radiolysis experiments is what may be termed target geometry or flux geometry in the radiation vessel. The observations are consistent with the interpretation that in the earlier Van de Graaff experiments using cylindrical reactors unsaturzkt,ed products which are formed in the reaction zone diffuse into the low radiation flux region of the reactor where they are not exposed to reactive intermediates formed in radiolysis. Thus the steady-state concentration of unsaturated products is much higher than under homogeneous irradiation; and
