5950
COMMUNICATIONS TO THE EDITOR
is estimated from the straight line14 from PH 7.6 to 12 as 1.00 f 0.04 X l o 6 M-' set.-' a t p = 0.10 and pH 10.0. It decreases with increasing ionic strength as predicted by the Bronsted-Christiansen-Scatchard equation (ZAZB = - 1). The value of K H ~ Ois estimated as 2 f 1 x l o p 9 M-l set.-'. The fact that the rate does not become completely independent of pH is ascribed to a small negative salt effect on the borohydride-water reaction. ,4t a constant pH with 0.09 and 1.09 sodium hydroxide solutions, where the water reaction predominates, the addition of sodium and potassium salts steadily decreases the rate. Lithium chloride gives a small acceleration. Variation of the buffer strengths a t constant ionic strength permits M-l assignment of K H ~ P O ~ - as 1 =t 4 X set.-', K H ~ B Oas ~ 1 f 5 X 10-4M-1sec.-1 and &cosas 9 =t4 X 10-5M-1 sec.-l. Thus the Bronsted a is about 0.8. The fact that CY is near unity and KH*o+ is so large means that general acid catalysis is difficult to observe.I5 The hydrolysis in heavy water gives H D Thus no exchange occurs between the hydride and the solvent, and an electronically unlikely BHs intermediate of high symmetry is excluded. A mechanism consistent with the data is
VOl. 82
TABLE J HYDROLYSIS IN AQCEOKSBORATE BCFFERSAT 25.00 $H =k0.02
9.70 9.96 10.20 10.20b
p.
ar
0.10 .05
.lo .10
NaBHd
104kl, sec. -10 1.74f0.02 1 . 0 2 f .02 0.587 f ,008 0.59 i. .01
h7aBD, lO4k1, sec. -1 2.62rtO.02 1 . 3 3 i .03 0.835 i ,007 0 . 8 4 f .01
* 0.01"
kdku 0.67iO.02 .77 rt .03 .70 i .03 .70 i- .03
a Rate = K1[BHa-] where kl = k2[HaO+]. Limits given are standard deviations. Rate constants during a given run were constant t o 85-90% reaction. Experiments where infinite solutions were used again and isotopes crossed.
The purity of each material was tested kinetically for trace metal contamination.4 The remaining solution after a borodeuteride run was used as the initial solution for a. borohydride run. The other combination was also tried. The pH remained constant and the inverse isotope effect was identical; thus the borodeuteride is free from catalytic impurities. The tritium isotope effects for alcoholyses of lithium borohydride and lithium aluminum hydride were reported by Kaplan and Wilzbach as X ? H / ~ T = 0.8-1.2.5 =ipparently experimental problems prevented their estimating the direction of the effect. IVe ascribe the inverse isotope effect to a secondary isotope effect of the other hydrogen or deuterium atoms not undergoing the protonolysis refast action in the rate-determining step. There is a (BH&. ----f products greater reluctance of HB- bonds to change to Diborane can be prepared from borohydrides in HB (uncharged) in aquated borane than that of concentrated sulfuric acid.16 The borane also can DB- bonds to change to DB. The effect is analobe trapped as the trimethylamine adduct when gous to the greater basicity and nucleophilicity of lithium borohydride is treated with trimethyl- DO- compared to HO-.6 Thus borodeuteride should be a stronger base than borohydride. The amine hydrochloride in anhydrous ether.'? non-reacting hydrogen or deuterium atoms have (14) The best line through the points gives an order of 0.92 i. 0.03 in more zero-point vibrational energy in the transihydronium ion. tion state than in the ground state. (15) Thus a t PH 9.24 in a borate buffer solution (0.05 A4 borax), 99% of the reaction is with H30' and about 1% with the H3BOa present. As the tetrahedral borohydride forms a transi(16) H. G. Weiss and I . Shapiro, THISJOURNAL, 81, 6167 (1959). tion state which resembles one-half of the diborane (17) G. W.Schaeffer and E. R. Anderson, ibid., 71, 2143 (1949). molecule, the normal B-H stretching vibrations DEPARTMEST OF CHEMISTRY PURDCE UNIVERSITY ROBERT EARLDAVIS (A1(vl)) a t 2264 cni.-I become more like the B-H stretching frequencies of diborane which occur a t IXDIASA LAFAYETTE, 2524 cm.-l (Ag(v1)).7,8,9 Thus in loose termiDEPARTMEST OF CHEMISTRY C. GARDXER %AIS MASSACHUSETTS INSTITUTE OF TECHNOLOGY nology, the non-reacting B-H bonds stiffen in the CAMBRIDGE 39, MASSACHUSETTS transition state. RECEIVED SEPTEMBER 7, 1960 Some data on the isotope effect of boranes would seem to support this conclusion. The isotope effect AN INVERSE HYDROGEN ISOTOPE EFFECT IN of the hydrolysis of trimethylamine borane in 1.40 THE HYDROLYSIS OF SODIUM BOROHYDRIDE'~2 J4 hydrochloric acid a t 25' is k H / k D = 1.05 = 0.02. Several reactions with diborane have isoSir: tope effects very close to unity. Apparently, it Kinetic study of the hydrolysis of sodium boro- is the placing of the fourth deuterium atom on the hydride in aqueous buffersZ suggested that cleav- boron which leads to an inverse isotope effect. age of a B-H bond is the first step and the rate- The order of magnitude (0.7) is very comparable determining step. I n a further study, NaBDi to the effect found in the hydroxide ions (0.6and NaB- 0.7). of 99% purity was prepared from (OCH8)d on a vacuum line with recrystallization (4) H. I. Schlesinger, H. C. Brown, A . E. Finholt, J. R . Gilbreath, from d i g l ~ m e . Its ~ rate of decomposition was H. R. Hoekstra and E. K. Hyde, i b i d . , 7 5 , 215 (1953). ( 5 ) L. Kaplan and K. E. Wilzbach, i b i d . , 74, 6152 (1952). measured as before2 in borate buffers under nitro(6) C. G. Swain and R. F. W. Bader, Tetraiiadvoil, 10, 182 (19GO); gen. The results are reported in Table I. C. G. Swain, R. F. W. Bader and E. R.Thornton, i b i d . , 10, 200 (1) Supported in part by a research grant and a fellowship to C. I400 >250
0.3 1. z > 25 >lo0
3 3 ,.
..
1
3
.. ..
103.1 1044 102.8 10i.a
101.0 102.~
..
1oi.s
(4) A. E. Martell and G. Schwarzenbach, Rets. Chim. Acta, S9, 853 (1956). (5) R. A. Alberty and R. M. Smith, J. A m . Chem. SOL.,7 8 , 2370 (1966); cf. also L. B. Nanninga, J . Phys, Chcm., 61, 1144 (1967).
