5846
A . J . CHALKA N D J. HALPERN
VOl. s1
PHYSICAL AND INORGANIC CHEMISTRY [ C O N T R I B U l ION
F R O M THE
DEPARTMENT O F C R E M I P T R Y , IJNIVERSITY
O F BRITISH
COLUMBIA]
Medium Effects in the Homogeneous Catalytic Activation of Molecular Hydrogen by Metal Salts. I. Cupric and Cuprous Heptanoates in Heptanoic Acid' BY A. J. CHALKAND J. HALPERN RECEIVED APRIL 22, 1959 The kinetics of the hydrogen reduction of cupric heptanoate (CuHp2) to the cuprous salt were examined in heptanoic acid solution over the temperature range 125 t o 155". The reaction is homogeneous, and proceeds autocatalytically, the rate law a t low cuprous heptanoate concentrations being, -d[H,]/dt = kl[H~][CuHp2] k 2 [ H t ][CuHp], where k l = 1.4.X l O I 3 exp[-30,200/RT] and k z = 1.0 X 1O1O exp[--21,000/RT] set.-'. The two rate constants were identified with separate reaction paths, in which the rate-determining steps involve the heterolytic splitting of H Z by CuH,p, and CuHp, respectively, giving rise t o unstable hydride intermediates which react rapidly with CuHpP or other reducible substrates such as quinone With increasing cuprous heptanoate concentration, the kinetic dependence of the Cu( I)-catalyzed reaction on the total Cu(1) concentration, was found t o shift from first toward one half order. This was attributed t o association of CuHp t o form an inactive dimer. Sodium heptanoate was found t o decrease the rate of reaction, apparently by forming inactive higher cupric and cuprous heptanoate complexes. The deuterium isotope effect arid exchange reaction were also examined and found t o be consistent wit!i the prnposcd mechanism.
+
Sodium heptanoate ( N a H p ) was prepared by neutralizing a Introduction of reagent grade sodium hydroxide with heptanoic The homogeneous reduction of cupric salts by solution acid, washing with benzene and drying at 110". Cupric hepmolecular hydrogen previously has been examined tanoate (CuHpz) was prepared by mixing in stoichiometric in two solvent systems, (i) aqueous solution^^^^ proportions, concentrated aqueous solutions of cupric suland (ii) organic amines especially q ~ i n o l i n e . ~fate and sodium heptanoate; the precipitated product w r y recrystallized from 1,2-dichloroethane and dried at 100 . Profound differences in the kinetics and mecha- 2,2'-Biquinoline (cuproin), a n Eastman white label product, nisms of the reaction in the two types of solvents was purified by vacuum sublimation. Hydrogen, obtained were noted. Perhaps the most striking is the ob- from Canadian Liquid Air Co., was passed through a "deservation, as yet unexplained, that in aqueous OXO" catalytic purifier and dried over P20s. Deuterium w s from the Stuart Oxygen Co. and similarly purified. solution hydrogen is activated by Cu(I1) but not obtained Mass spectrometric analysis showed i t t o contaiu 0.7% by Cu(1) while the reverse is the case for the amine HD.6 Kinetic Measurements.-The reaction kinetics were folsolvents. To obtain further insight into this phenomenon, lowed by measuring the volume of hydrogen gas taken up a t pressure using conventional apparatus and techthe study of hydrogenation of cupric salts has been constant niques.E The reaction vessel and gas buret were separately extended to other solvent systems notably car- thermostated, the former in a silicone bath maintained a t boxylic acids and hydrocarbons, the latter being the reaction temperature (+0.08'), the latter in a waterconsidered as approaching most closely to a truly bath maintained at 30.0 & 0.05'. Prior to commencing the the solution was dehydrated by heating at about inert medium. The present paper describes a reaction, 145' under reduced pressure and degassed by repeated kinetic study of the hydrogenation of cupric hepta- freezing and melting under vacuum. The shaking rate of the noate in heptanoic acid solution. Since heptanoic reaction vessel generally exceeded 250 cycles per minute and acid is one of the products of the reaction, its choice this, combined with the use of small volumes of solution in a large vessel (-2 ml. in 30-50 ml.) with a n indented as solvent minimizes complications such as have relatively surface, ensured absence of physical control (i.e., due t o been encountered in earlier related slow dissolution of hydrogen). Deuterium Exchange.-Deuterium exchange measuredue to medium changes as reaction proceeds. The results are reported in some detail since they differ ments involved removing samples of the gas into evacuated bulbs, connected through stopcocks t o the apparatus just significantly from those for both solvent systems above the reaction vessel. Following removal of each samexamined earlier. Subsequent papers in this ple, the apparatus was refilled with D2 to atmospheric presseries describe studies involving other cupric and sure. The gas samples were analyzed mass-spectrometricuprous salts and other solvents, as well as some ally.^ However, because the volume of gas replaced through the removal of each sample was not accurately known and related studies on silver and mercuric salts. The because of the likelihood of non-uniform gas composition heptanoate salts of these several metals did not throughout the apparatus (the reaction vessel and gas buret react detectably with hydrogen in heptanoic acid being connected by a considerable length of capillary), the solution a t temperatures of up to 150": Fe(III), results are only of semi-quantitative significance. Solubility of Hydrogen.-The same apparatus was used Ce(IV), Co(II), Ki(II), Cd(I1). to measure the solubility of hydrogen in heptanoic acid a t various temperatures. After a known volume of heptanoic Experimental acid was equilibrated with H P at about 150 mm. pressure, Materials.-Heptanoic acid ( H H P ) , an Eastman white label product, was purified by distillation, b.p. 220-221'.
(1) Support nf this work through grants from the Research Corporation and t h e h-ational Research Council of Canada is grntcfully acknomledged. (2) R. G. Dakers and J. Halpern, Can. J . C h e m . , 32, 969 (19.54). (3) E. R. Wacgregor and J , Halpern, Trans. Me!. SOC.A.I.B4.E., 212,
244 (1958). (4) (a) M. Calvin, Tronr. Faraday Soc., 34, 1181 (1938); THIS
JOURNAL,61, 2230 (1939); (b) S. Weller and G. A. Mills, ibid., 75, 7F9 (1953); L. Wright and S. W d l e r , ihid., 76,3345 (1954); L.Wright, S. Weller and G. A. Mills, J . Phy.?. Chem., 69, 1060 (19.55); ( c ) hl.
shaking was stopped and the pressure of H2increased t o some new value. Shaking then was resumed and the amount of gas taken up t o saturation was determined. By this procedure it was also possible t o measure the rate of solution of H Z and thus establish directly that the reaction under any giver1 set of conditions was free from physical control. Calvin and W.K. Wilrnarth, TKISJOURNAL, 76, 1301 (1956); W.K. Wilmarth and M. Barsh, i b i d . , 78, 1305 (1956). ( 5 ) We are grateful t o Dr. E. W. C. Clarke for this and other mass spectrometric analyses reported in this paper. (6) J . L. Bolland, Proc. Roy. SOC.(London). A186, 218 (1946); A. J. Chalk ani1 J. I? Suiitli, Trans. Faraday SOC.,63, 1214 (1957).
Nov. 20, 1959
ACTIVilTION OF
Spectra.-Spectra of cupric heptanoate solutions were recorded with a Cary Model 14 spectrophotometer. All data were corrected for changes in density with temperature.
Results and Discussion Stoichiometry and Kinetics.-Typical rate plots depicting the uptake of hydrogen by solutions containing various initial concentrations of CuHpz are shown in Fig. 1. The reaction occurring is 1.
2CuHpp
+ Hz
----f
2CuHp
+ 2HHp
5847
MOLECULAR HYDROGEN BY METAL SALTS
(1)
A
I N I T I A L [CuHpz]
u'
v 0
A 0
0029 M
- 0073 - 0 186 - 0352
s1
*
'.
3
s4
At the reaction temperatures, CuHp is soluble in HHp and stable with respect to disproportionation. Completion of reaction 1 (at which point the ini-
--
15 0
W
1 2 3 4 5 HZabsorbed X loa, mole L-1. Fig. 2.-Initial variation of the rate with extent of reaction (145', 700 mm. H2).
0
0c
x r(
0
8000 12000 Time, sec. Fig. 1.-Absorption of hydrogen by solutions of cupric heptanoate in heptanoic acid a t 145'. 700 mm. Hz. (Experimental points shown only for one plot.) 0
4000
tially blue solutions are clear and colorless) is marked by a sharp break in the rate plot, followed by a much slower reduction of CuHp to Cu 2CuHp
+ Hz +~ C +U 2HHp
(2)
0
Ip -
-
I
-
~
Reaction 1 was unaffected by the addition of glass 0 5 10 15 wool, by varying the volume of solution or of the H1 absorbed X 102, mole L-1. reaction vessel, and was clearly homogeneous in Fig. 3.-Variation of the rate with the extent of reduction of character. 0.352 M CuHp, a t 145". 700 mm. Hz. The essential features of the kinetics are revealed by the rate-extent plots' shown in Figs. 2 two systems including (i) the finite initial rates in and 3. These differ markedly from the kinetics heptanoic acid due to activation of hydrogen by in aqueous solution^^^^ but bear some resemblance Cu(II), and (ii) the kinetic dependence on Cu(1) particularly in respect of the autocatalytic nature of which is second order in quinoline but (at least a t the reaction and the sharp breaks in the rate plots, low Cu(1) concentrations, see Fig. 2 ) close t o to the behavior observed in quinoline ~ o l u t i o n s . ~first order in heptanoic acid; (the apparent deAs in the latter case the autocatalysis is attributed parture from first-order dependence a t higher to activation of hydrogen by Cu(1). However, Cu(1) concentrations, reflected in the curvature of there are also important differences between the the rate-extent plot in Fig. 3, will be discussed later). (7) The points in these plots represent instantaneous rates deterThe kinetic behavior in heptanoic acid is inmined by drawing tangents to the rate plots at various times and are precise to within &57& I n determining rates by this method it is terpreted in terms of the rate law essential that the individual points on the rate plots be very accurate -d[Hs]/dt = Ri Rz = R i [ H 2 ] [CuHpz] kz[Hs] [CuHp] and closely spaced. The present procedure which gives nearly con-
+
tinuous rate plots is very satisfactory from this standpoint.
+
(3)
where XI and Rz are the ratcs due to the activation of Hz by CuHpz and CuHp, respectively Introducing the stoichio~netryrelation [CuHpz]
+ [CIITSP] = [Ct~I-Il)?jo
the rate law becomes
(1)
+
-d[H2l/dt = ~ I [ H ] Z [ C U H ~ (~K ,] O - k ~ ) [ H *[CuHp; l (3)
The initial linear portions of the rate-extent plots in Fig. 2 are in accord with this, and their intercepts and slopes, in conjunction with values of [ Hz] computed from the solubility data in Table 11, yield the values of kl and k p which are listed in Table I. Within the limits of experimental scatter, these show no systematic variation over a twenty-fivefold variation in [CuHpz]oand a threefold variation i n [H?] \'ARIA1
TABLE I ION O F 'I HE I N I T I A L CuHl>pC o S C E N l R A i TOY HZPARTIAL PRI-~~U A I R113' C
3. Effect of Temperature.-Kinetic measurements and hydrogen solubility data over the temperature range 12.5 to 155" are suriiinari7ed in 'I'able 11. The values of lzl and k 2 gavc goocl ,Zr rhenius plots fitted by k , = 1.1 X lo1?r x p [ - 3 0 , 2 0 0 / R T ] ; 2 f - 1 s e c (11) and k2 = 1 0 X IOio c\l)[ -3,1,00O/KT] If
-*
sec
'
(12)
respectively. 'The corresponding heats and entropies of activation are listed in Table 11. A&* appears to lie just above and As2* j ~ s tbelow the normal range of entropy values (approximately - 5 to -10 e u,) for simple biniolecular reactions in solution. TABLE IT sITMMAR\ O F r
