Fcb., 1048
ss 1
COMMUNICATIONS TO THE EDITOR
order OT a t different rates. However, the induced exchange was reproducible when conditions were held constant so we were able to correct for it using the equation % exchange == % exchange (measured) - % exchange (induced) (100) 100 - yo exchange (induced) The corrected values (always three or more excluding the value a t zero time) obeyed the exponential exchange law.',' The half-times for the exchange rates are summarized in Table I. As expected the exchange rate is not dependent on the method of separation when proper account is taken of the induced exchange.
Hg' are each bound to two boron atoms. Each boron atom has three boron neighbors at 1.74 f 0.04 kX and one hydrogen neighbor a t 1.34 * 0.04 kx. In addition, B4, B4', B4" and B4"' each has a boron neighbor a t 1.96 * 0.04 kx and another hydrogen neighbor a t 1.54 * 0.04 kX; B1, B:, Bz, Bz', each has two boron neighbors a t 1.96 * 0.04 kx: BI and Ba', each has another hydrogen neighbor a t 1.34 * 0.04 kX. 45'
TABLB I TI(1)-TI(II1)EXCHANGE RATES 0.0244 f . Tl(I), 0.0244 f . Tl(II1) Acid
Temperature
Method of separation
Exchange half-tim; hr.
1.0 f. "01 1 . 5 f. HNOr 1.5 f . "0% 1 . 5 f . HClO, 1 . 5 f . HClO' 1 . 5 f . HClO, 2 . 5 f . HClO, 3 . 5 f . HClO,
25°C. 24.8 * 0.2" 24.8 * 0.2" 24.8 * 0.2" 24.8 =t0.2" 24.8 * 0.2" 24.8 * 0.2" 24.8 f 0.2"
Bromide Bromide Hydroxide Hydroxide Bromide Bromide Bromide Brornitlc
2.5 * 0.2 1.8 * 0. 1.6 * 0 . 2 36 * 4 35 * 4 33 * 4 45 * 4 67 * 5
M.
O=BORON I 0 :HYDROGEN
Fig. 1.
We are extending this work to determine the effect of temperature, ionic strength, and concentrations of the reactants on the exchange rate.
Each boron atom is bound to five or six other atoms, but the bonds are not all equivalent. In(3) H. A. C. McKay, Nalurc, 119, 997 (1938). asmuch as a bond distance of 1.96 k X has about (4) R. B. Duffield and M. Calvin, TaIs JOURNAL, 68, 557 (1946). half the "bond number"2 of a bond distance of DEPARTMENT OF CHEMISTRY 1.74 kX, one can say that each boron forms five WASZINGTON UNIVERSITY REN&J. PRESTWOODbonds of bond number 0.60. The corresponding Sr. LOUIS,M i s s o m ARTHURC. WAHL radius, R(0.60) = 0.87 kx. Consequently, R(l) RECEMX DECEMBER I 10, 1947 = 0.80 kX in agreement with Pauling.2 This structure for BloHlr gives excellent agreement with the observed X-ray diffraction intensiTHE STRUCTURE OF THE DECABORANE MOLECULE ties and also with the electron diffraction observasir: tions of S. Bauer.a A detailed discussion of the determination of the We axe studying the structure of crystalline decaborane, BIOHl,, by single crystal X-ray dif- structure of crystalline decaborane will be pubfraction methods. We have established the ap- lished soon. (2) L.PaUbg, THIS JOURNAL, 69, 542 (1947). proximate positions of the ten boron atoms and (3) S. Bauet, ibid., TO, 116 (1948). four of the hydrogen atoms, and have assigned probable positions to the remaining ten hydrogen RESEARCH LAFIBORATORY JOHNS. KASPER ELECTRICCOMPANY C. M. LUCHT atoms. (Hydrogen atoms are well resolved in GENERAL NEW YORK. DAVIDHARKER SCHENECTADY, fourier sections.) RECEIVED JANUARY 21,1948 The BlOHI4 molecule has the symmetry &mm2. The bond distances are as follows (see figure): B1-Bl', Bl-BI, B2-Bs, Bn-B,, BrB4, are HETEROGENEITY OF CRYSTALLINE BETA-LACTOGLOBULIN all 1.74 t0.04kX; Bl-Bz and B4-B4' are 1.96 f 0.04 m; B4-H4 is 1.34 f 0.04 kX,l and all other Sir : B-H distances axe assumed the same, except B4That crystalline &lactoglobulin is not a homoHs which is assumed to be 1.54 0.04 kx. (B4- geneous protein was indicated by the solubility B4'" and BI'-B~' are 2.76 * 0.04 k X and are not measurements of Gronwall' and by the electrobond distances.) Each hydrogen atom, except HII phoretic results of Li.% Our experiments with and He' is bound to a single boron atom; HIIand (1) Grbnmll, Compt. mend. haw. lab. Carlsbw;. S4, no. 6-11. 188 f
and H P ' werelocated on an electron density map; (1) Hd, Ed', the position. of the other hydrogen atom are aasumed.
(1042). (2) Li,
JOUZlNAb,
68, 2746 (1946).
this protein conb-m these reports and indicate a relationship bet.ween the solubility of the protein and its composition as defined by the elcctrophoretic patterns obtained a t PH 4.7. @-Lactoglobulinwas prepared by the method of Palmer3which involves the fractionation of milk whey with ammonium sulfate after the removal of casein with acid. The isolated @-lactoglobulin was recrystallized four times by dialyzing away the salt from sodium chloride solutions a t the isoelectric point. Electrophoretic studies on a 1% solution of this material in acetate buffer of ionic strength 0.1, p€t 4.8, showed the presence of a two component system. The same preparation was electrophoretically homogeneous on the alkaline side of the isoelectric point. Fractionation experiments on the whey proteins from the filtrate after the partial removal of crystalline @-lactoglobulin gave crystalline fractions with a mobility comparable to that of the usual @-lactoglobulina t PH 8.3. Electrophoresis pH 4.5, however, indicated a variation in the concentration of the components in each of the fractions. These new crystalline fractions also varied from the standard preparation in that they showed a greater solubility in water and in dilute salt solutions. Crystallization of the standard preparation by dialysis from an acetate buffer solution, vaned with respect to PH, gave fractions with a partial separation of t'he components as indicated electrophoretically. Partial separation of P-lactoglobulin has also been obtained by fractionation with alcohol a t low temperatures. The fractions obtained by alcohol have properties in agreement with the data reported in the table below for preparations obtained by the other methods. These results will be reported subsequently. The solubility experiments were made as dcscribed by Gronwall.' In every case a suspension of the crystalline material containing 7.6 * 0.1 mg. protein nitrogen per ml. was equilibrated for a twenty-four hour period. The protein nitrogen was then determined on the supernatant liquid after centrifugation. The table illustrates the variations in the electrophoretic components of a number of preparations at pH 4.8 and the solubility of these crystalline fractions a t pH 5.2 * 0.1. Camposition. Preparation
Ir
1.4-1.5
x
lo(
2.2-3.2
Solubility (mg. N/cc.) a t 25' in 0.02 A4 in Ha0 NaCl
0.08 1.1 72 .10 1.3 38 62 .12 1.5 40 60 .18 1.7 46 54 .13 1.6 L' 54 46 .19 1.9 59 41 Mb a Preparations made from the standard preparation by means of acetate buffer of varying PH. b From filtrate of standard p-lactoglobulin preparation.
Ba AS Standard Kb
Vol. 70
COMMUNICATIONS TO THE EDITOR
882
28
(3) Palmer, J . Biol. Chem., 104, 359 (1934).
The variation in solubility in salt solutions of preparations with different electrophoretic compositions is an adequate explanation for the divergent solubility data reported by Palmera and Gronwall.' EASTERN REGIONAL I~ESEARCIILAB. T. L. MCMEEKIN U. S. DEPARTMENT OF AGRICULTURE B. D. POLIS AGRICULTURAL RESEARCII ADMINIS. E. S. DELLAMONICA 18, PA. * J. 11. CUSTER PHILADELPIIIA I
