4740 Photoreduction of CO(NH&N~~+ at 2537 A. Photolabilization of an Ammonia Ligand’
balt(I1) was determined spectrophotometrically6 in comparison to uranyl- and ferrioxalate actinometry.7 At the higher values of I,, the rate of formation of Co2+ Sir: was not zero to first order (depending on whether 100% It has been reported2 that irradiation of the ligandor less light was transmitted) in substrate concentration to-metal charge-transfer region of aqueous C O ( N H ~ ) ~ - as is generally observed: but rather ( p ~ ~ was ( ~ ~ very ) N32+leads to 100% photoreduction ( k . ,Coz+). We often found to increase with time. Typical kinetic can now report compelling evidence that C O ( N H ~ ) ~ - behavior in this region of I, is shown in Figure 1 where a OH2N32+is actually the predominant initial product of distinct increase in the rate of cobalt(I1) production is this reaction and that this latter complex is itself subseexhibited after 20-25 % of reaction; this corresponds to quently photolyzed. ( ~an~ initial ) value of 0.17 to a a change of p ~ ~from value greater than 0.35 (I, decreases only about 20% over about 75 % of reaction in this case). I I A careful examination of the cobalt(I1) yields and the spectra of irradiated solutions in a very large number of experiments has revealed a number of unexpected ( ~ the~ap) kinetic features: (1) the value of p ~ ~and pearance of the break in the kinetics curve appeared to be dependent on the rate at which light was absorbed; Le., at high I,, c p ~ ~ ‘v ( ~0.2 ~ )initially, becoming 0.3-0.6 after 20-25% of reaction while, at low I,, ~ c ~ ( I apI ) proached an apparent limit of -0.6 with no break in the curve;g (2) in some experiments at very high I,, the appears to depend on the ratio of irravalue of (oco(r~) diation to dark time during the exposure and sample removal; (3) the apparent change in [ C O ( N H ~ ) ~ Nas~ ~ + ] measured at its charge-transfer band maximum (301 nm) was considerably larger than that calculated at 254 nm, but disappearance of [CO(NH~),N,~+] (as determined by the absorbance change at 301 nm) was found to be stoichiometric in the appearance of Co2+ as a 0 1 2 3 4 5 6 7 was found photolysis product; (4) the value of EXPOSURE T I M E IN MIN. to be very insensitive (Ad 5 10%) to the presence of Figure 1. Co(I1) production in the 2537-A photolysis of COpotential radical scavengers, such as H+, NOz-, NO3-, M ; [H+] = 0.1 M ; (NH3)sN32’: [ C O ( N H ~ ) ~ N = ~ ]9.88 ~+ X N3-, H202,02,I-, Br-, or C O ( N H ~ ) in ~ ~the + solution 1. = 3.3 x 10-4 einstein 1.-1 min-1; lo = 5.5 x 10-4 einstein during irradiation; (5) the yield of Nz (in vacuo) relative 1.-1 min-1. to Co2+ was found to be independent of the period of irradiation. These observations imply the formation of a species Some evidence has been reported recently3 which from the photolysis which is very similar to C O ( N H ~ ) ~ suggests that C O ( N H ~ ) ~ O H is ~ Cformed ~ ~ + as a product N32+both with respect to its near-ultraviolet and visible in the uv irradiation of C O ( N H ~ ) ~ Cand ~ ~ +the , photoabsorption spectra and the qualitative features of its labilization of a Crrrl-ammine bond seems to be fairly Table I. Spectral and Photochemical Data €254
Complex
CO(NH3)aNa
’+
C O ( N H ~ ) ~ ( O H ~ (equilibrium )N~~+ mixture)
A,
(log
e),
nm
r-
-
cpcocX1)-
Intermediate
pep- N, - apparent
0.63 1.35
0.17 0.27
0.36 0.34
0.17-.0.7 0.25 - 0 . 3 5
301 (3.89) 308 (3.91)
(1) This work was supported by the National Science Foundation (Grant GP 7048) and was presented in part before the Division of Physical Chemistry, 155th National Meeting of the American Chemical Society, San Francisco, Calif., April 1968. (2) A. W. Adamson, Discussions Faraday SOC.,29, 163 (1960). (3) L. Moggi, N. Sabbatini, and V. Balzani, Garz. Chim. Ital., 97,980 (1967). (4) A. W. Adamson, J. Phys. Chem., 71, 799 (1967). ( 5 ) Prepared as described by M. Linhard and H. Flygare, Z . Anorg. Allgem. Chem., 262, 328 (1950).
1 90:17
10-3,
Initial
commonplace;4 however, this kind of process has not heretofore been established as the dominant reaction path in the photochemistry of cobalt(II1). We have irragiated acidic solutions of Co(NH3)SN32+ at 2537 A under a nitrogen atmosphere using mercury resonance lamps of various intensities ( 10-5 einstein 1.-’ min-I). The quantum yield of co-
JournaI of the American Chemical Society
x
1. mole-’ cm-1
photochemistry. Since both the photochemistry and absorption spectra of C O ( N H ~ ) ~ N are~ uniquely ~+ characteristic of the presence of the CorlI-Na- bond, the product species must be an azide complex of cobalt(II1). Since the equilibrium mixture of C O ( N H ~ ) ~ O H ~ N ~ ~ + (6) D. Katakis and A. 0. Allen, J. Phys. Chem., 68, 1357 (1964). (7) J. G. Calvert and J. N. Pitts, Jr., “Photochemistry,” John Wiley and Sons, Inc., New York, N. Y., 1966, p 783. (8) J. F. Endicott and M. Z . Hoffman, J . Am. Chem. Soc., 87, 3348 (1965). (9) These observations are consistent with a significant absorption by C O ( N H & O H Z N ~after ~ + -25% of reaction (high I,) and a photostationary state in [Co(NHa)rOHzNaZ+]. Note that e254 for Co(NHa)rOHzNa*+ is about twice ~ Z U of Co(NH&N32+ (Table I). A detailed kinetic analysis will be presented elsewhere. 10 (10) L. S . Beres, J. F. Endicott, M. Z. Hoffman, and R. W. McQuigg, manuscript in preparation.
J August 14, 1968
4741
meets all the requirements listed above, we have also investigated the photochemistry of this material. Pertinent spectral and photochemical data for C O ( N H ~ ) ~ N32+and C O ( N H ~ ) ~ O H ~are N ~compared ~+ in Table I. After a sample of C O ( N H ~ ) ~ was N ~ irradiated ~+ in the cell holder of a Cary 14 spectrophotometer for 1 min, Figure 1. A diagramatic drawing of the coordination polyhedron in [Ni(diars)(triars)](ClO~)~. Crystals are monoclinic, space the absorbance of the solution at 254 nm slowly degroup P2&; a = 18.41, b = 20.21, c = 12.61 A, /3 = 125.80”,Z = creased over the course of about 5 min. This transient 4 formula units per unit cell. At the present stage of least-squares absorbance exhibited a half-life for the decay of 2 4 0 refinement, the residual R = 0.18. The basal plane Ni-As dissec. This is comparable to the half-life of the trans tances are 2.26, 2.27, 2.30, and 2.32 A. The apical Ni-As distance is 2.39 A. The site below the square plane appears t o be unoccis isomerization of C O ( N H ~ ) ~ O H ~reported N ~ ~ + by cupied. Haim.” The lifetime of this transient (which is presumably due to t r a n ~ - C o ( N H ~ ) ~ 0 H z N compared ~ ~ + ) to the time required for photolysis and sampling at high a deep maroon color developed and crystals were de1, is such that the solution containing Co(NH&0H2posited within 30 min at room temperature. In this N3*+will approach equilibrium only as the ratio of way a quantitative yield was obtained of a substance irradiation to dark time becomes very large. which had an X-ray powder pattern identical with that Our observations, summarized above, of the photoof the original material. As further evidence for this chemistry of C O ( N H ~ ) ~ Na~cobalt(II1) ~+, complex conformulation it is possible, after breaking down the taining a strongly “reducing” ligand, show quite clearly original complex with sodium cyanide solution, to isothat the photolabilization of a coordinated ammine has late the triarsine (11) from the reaction mixture. at least as great a quantum yield as photoreduction and A reexamination of the proton magnetic resonance that both of these processes are far more important spectra of the original and the new complex in hexathan the formation of Co(NH3);OHZ3+. These obdeuteriodimethyl sulfoxide shows, in addition to the two servations are certainly at odds with a “radical-pair” previously reported peaks, now shifted to ‘T 8.23 and model2 for the photochemistry of coordination com8.41, three smaller, broader peaks at ‘T 6.53, 7.32, and plexes of cobalt(II1). Studies still in progress sugge%t 7.80. These peaks, which are of equal area and together that similar photolabilization occurs in the 2537-A constitute about one-third of the area of the strong irradiation of cis-Co(NH3)4(N3)2+ but not for transpeaks, were obscured by the solvent in dimethylformCO(NH~)~(N&+.” amide. Acknowledgment. The authors wish to thank Mr. Thus we have discovered a new reaction whereby, Laszlo S. Beres for many early studies on this system. in boiling diethylene glycol, o-phenylenebis(dimethy1arsine) is converted in some 20% yield to the triarsine, ( 1 1 ) A. Haim, J . Am. Chem. SOC.,86, 2352 (1964). bis(o-dimethylarsinopheny1)methylarsine. Phillips3 reJohn F. Endicott, Morton Z. Hoffman ported the preparation of the so-called “tris complex” Depariment of Chemistry, Bostori Utiicersity in some 40% yield by heating an ethanolic solution of Boston, Massachusetts 02215 [Ni(diars)&l~] in a sealed tube at 200” but was unable Receiced May 27, 1968 to isolate the complex from the reaction mixture. It is of interest to note4 that the reaction products from the sealed tubes always smelt more strongly than diarsine. A Nickel(I1)-Catalyzed Synthesis of a )~ This is probably due to the presence of A s ( C H ~ formed Triarsine from a Diarsine in the disproportionation. This reaction requires the presence of the nickel(I1) salt since, if a solution of the Sir : diarsine in diethylene glycol is refluxed for the same peSome years ago, one of us’ reported the isolation of a riod of time, subsequent addition of nickel(I1) chloride compound which was considered to have the formula to the cooled solution does not yield the requisite [Ni(diar~)ij(ClO~)~ (diars = o-phenylenebis(dimethy1arproduct. sine)), I, and in 1966 we described2 an improved method These results still leave unsolved the puzzle of the of preparation. The compound was of interest in that structures of the diamagnetic nickel(I1) complexes of the it was diamagnetic although believed to bels2 an octatype [Ni(triars)~](ClO~)~, with the terdentate arsines, hedral trisbidentate complex. 11, 111, and IV. These all have spectra very similar to However, X-ray diffraction studies, which will be CH,-As(CH,), reported in detail elsewhere by P. J. Pauling, show that / the substance which was isolated is o-phenylenebis(diCH2 methy larsine)[ bis(o-dimethylarsinopheny1)methylarsinelc\IH2 , C H ~ A S(CH,), nickel(I1) perchlorate. The compound contains C H r C CHrAs (CH3)2 j u e arsenic atoms arranged about the nickel at the CHr? \C H i k (CH,), corners of an almost regular tetragonal pyramid (Figure CHZ I 1). We have now shown that the original compound CH2 can be prepared directly from the diarsine (I) and the \ C H r As (CH, lZ IV triarsine (11) with nickel(I1) perchlorate hexahydrate. Thus, when equal volumes of 0.1 M solutions of each of these three compounds in ethanol-ether were mixed, 1x1 ~
-
(1) R. S. Nyholm, J . Chem. SOC.,2061 (1950). (2) B. Bosnich, R. Bramley, R. S. Nyholm, and M. L. Tobe, J . Am, Chem. SOC.,88, 3926 (1966).
~
(3) D. J. Phillips, Ph.D. Thesis, University of London, 1958.
(4) D. J. Phillips, private communication.
Communications to the Editor
