METHYLCHLOROSILANE AMINES
Feb., 1946
The limiting conductances were also evaluated by the mathematical extrapolation of the curves for A vs. .\T. The method of least squares applied to the data for dilute solutions gave the following values for the limiting conductances Salt
A#
Slope
Theor. slope
KIOk KReOd
127.92 128.20
-94.08 -97.08
-89.61 -89.68
The extrapolated value obtained by Pushin and Tutundziel was 125.7 for potassium perrhenate. The data were also plotted according to a method proposed by Shedlovsky6in which an extrapolation function of the form l/h = (l/Ao) kdZ is used. The data for solutions more dilute than 0.01 molar fell on straight lines and yielded values of A, of 127.95 for potassium metaperiodate and 128.25 for potassium perrhenate. These values are somewhat higher than those obtained by either of the other two methods but the extrapolation is rather long and steep. The agreement between the various methods of evaluating A0 is considered satisfactory. The values accepted for AO are 127.90 for the metaperiodate and 128.20 for the perrhenate. These are believed accurate to within ~ 0 . 0 7conductance unit. Mobilities of the Anions.-Since the mobility of the potassium ion at infinite dilution and 25' is well known (73.52 on the Jones and Bradshaws standard), it is possible to calculate the mobilities of the metaperiodate ion and the perrhenate ion using the limiting conductances of the potassium salts measured in this investigation. This would yield the following values: metaperiodate ion, 54.38 * 0.07 and perrhenate ion, 54.68 * 0.07.
129.5
128.5
127.5
+
Summary The electrical conductances of aqueous solutions of potassium metaperiodate and potassium (6) T.Shedlovsky, J. Franklin I n s f . , 116, 739 (1938).
[CONTRIBUTION FROM THE
24 1
c
lZ7.0
0
0.01
0.02
0.03
C.
Fig. 1.-Plots of Shedlovsky function.
perrhenate were measured over the concentration range 0.0004 m to approximate saturation. The limiting conductances were determined by three independent methods with satisfactory agreement. The densities of the solutions were determined with an accuracy of approximately 1part in 15,000. The limiting conductances of the anions were determined using the recorded value for the potassium ion and the measured values for the salts from this investigation. BLOOMINGTON, INDIANA
RECEIVED NOVEMBER 30, 1945
RESEARCH LABORATORY, GENERALELECTRIC COMPANY]
Derivatives of the Methylchlorosilanes. IV, Amines BY ROBERT 0. SAUERAND R. H. HASEK'
In the first paper2 of this series which described the hydrolysis and alcoholysis products of trimethylchlorosilane attention was directed toward the unstable nature of trimethylsilanol as compared with triethylsilanol and its higher homologs. We now report a parallel study of the ammonolysis and aminolysis products of this chlorosilane. Stock and Somieski3treated chlorosilane vapor (1) Present address: Tennessee Eastman Corp., Kingsport, Tenn. (2) S u e r , Ttus JOURNAL, 66, 1707 (1944). (3) Stock and Somieski, B e . , 64, 740 (1921).
with ammonia and isolated the tertiary amine (HsSi)aN, b. p. 52'. With excess ammonia presence of the primary and secondary amines was assumed ; neither, however, was isolated. Aminosilane presumably condensed with itself producing ammonia and disilazine, HsSiNHSiHa; the latter decomposed slowly with the formation of silane and the non-volatile polysilazine, (H2SiNH),. Chlorosilane with methylamine and with ethylamine yielded4 N-methyldisilazine, b. p. 32O, and N-ethyldisilazine, b. p. 6 6 O , respectively. (4) Emeleus and Miller, Nature, 14S, 996 (1938);
819
(1939).
J. Chcm. Soc..
ROBERT0. SAUER AND R. H. HASEK
242
Vol. 68
I n our experiments trimethylchlorosilane has cleaved quantitatively by dilute acids, giving given only one ammonolysis product, hexamethyl- ammonium salts and the silanol (as the primary disilazine.2 Under similar conditions we ob- product). However, the compound is remarktained triethylaminosilane and hexaethyldisil- ably resistant to hydrolysis by boiling water or azines from triethylchlorosilane. Hence, it ap- dilute alkalies. Such behavior is apparently a pears that the stability of the RsSiNHz com- consequence of its water insolubility, for upon pounds is of about the same order as that of their addition of methanol to the amine-water mixR3SiOH analogs. Ready elimination of ammonia ture a homogeneous solution results accompanied or water (yielding the disilazine or disiloxane, by vigorous hydrolysis of the compound. In addition to the aminosilanes noted above respectively) appears to characterize the lower we have prepared impure trimethyl-N-ethylmembers of both series. Hexamethyldisilazine was found unexpectedly aminosilane, b. p. 90°, and trimethyl-N,N-diethylstable to attack at the nitrogen-hydrogen bond. aminosilane, b. p. 126’. We have based our assigned structural formulas Although an 88-94% yield of methane is obtained2 in the Zerewitinoff determination (equa- for the four new compounds prepared from trition l),no reaction was observed with metallic so- methylchlorosilane on the method of synthesis and on the recognized stability of the carbonsilicon and carbon-nitrogen bonds under the dium at 125’ after sixteen hours. Neither did conditions of synthesis and isolation. Deterthis amine discharge the blue color of sodium dis- mination of neutral equivalents was used to consolved in liquid ammonia. Another indication of firm the composition of the expected products. the inertness of the nitrogen-hydrogen bond was The only product isolated from the reaction of our failure to isolate from the ammonolysis of trimethylchlorosilane and ammonia was hexatrimethylchlorosilane the tertiary amine, [(CH&- methyldisilazine, although one might predict a SiI3N, a compound which appears sterically pos- priori that three are possible. The structure sible by inspection of Fisher-Hirschfelder models assigned t o this product was confirmed by eleand chemically probable in the light of Stock’s mentary analyses for carbon, hydrogen, silicon results. This point was checked by our inability and nitrogen and further supported by satist o effect a reaction between purified hexamethyl- factory neutral equivalent and “active hydrogen’’ disilazine and trimethylchlorosilane (or silicon determinations. From trimethylchlorosilane and methylamine only two silicon-containing prodtetrachloride) at about 60°. ucts can result assuming inviolability of carbonWith methylamine trimethylchlorosilane yields trimethyl-N-methylaminosilane, b. p. 71O , which silicon, carbon-nitrogen and carbon-hydrogen upon further treatment with the chlorosilane bonds; neutral equivalents on the two products obtained confirmed our expectations. From diyields heptamethyldisilazine (equation 2). ethylamine and trimethylchlorosilane one pre3(CH&SihTHCHa f (CH&?Cl ----f dicts (on the assumptions listed above) only a 2 [(CH3)&]ZhCH3 CH3NHz.HCl (2) single product, trimethyl-N, N-diethylaminosilane. The significant stages in the reaction appear to be The neutral equivalent of the product found represented by indicated that the expected structure was obtained. The structure of the compound from (CH&SiNHCH3 + (CH3)&X1 [(CH~)~S~]ZNCH HC1 I (3a) trimethylchlorosilane and ethylamine was as(CH&SiNHCH3 2HClsigned primarily on the basis of the yield of tri(CH3),SiCl + CH3NH2.HC1 (3b) methyl-N-methylaminosilane obtained in the reTwo factors may suppress the analogous se- action with methylamine. The neutral equivaquence of reactions with hexamethyldisilazine; lents on our product, trimethyl-N-ethylaminoeither the formation of hydrogen chloride is silane, indicated the presence of appreciable inert blocked by the “inertness” of the hydrogen at- impurity. tached to nitrogen, or else the degradative cleavExperimental age by hydrogen chloride in the sense of equation Method of Analysis.-All neutral equivalents reported in 3b does not proceed. The latter factor is pre- this experimental part with the exception of those on cluded by the demonstration that hydrogen the triethylsilyl derivatives were obtained by adding chloride (in accordance with the generalization the sample to excess 0.1 N hydrochloric acid. After vigorous shaking the excess acid was back-titrated with of Stock3) readily cleaved hexamethyldisilazine. 0.1 N sodium hydroxide solution to the methyl orange The hydrolysis of hexamethyldisilazine is un- end-point. usual in several respects. As noted earlier2 it is Trimethylsilyl Derivatives
+
+
+
( 5 ) Rraus and Nelson (THISJOURNAL, 66, 195 (1934)) report the Hexamethyldisi1azine.-( a) Preparation of this subpreparation of this compound by the reaction of potassium amide in stance was carried out in ethereal solution as previously liquid ammonia with triethylsilane, followed by neutralization of the described.2 It was noted that care must be taken to keep potassium salt with ammonium bromide. Their yield of [ ( C ~ H ~ B the reaction mixture anhydrous. In one experiment in Si IzN-K was nearly quantitative indicating the condensation which water was added to the reaction mixture to dissolve B(CzHrf8SiNH[(CsHs);Si]zNNHathe ammonium chloride hydrolysis of the amine resulted, under the conditions of their experiment. yielding trimethylsilanol and hexamethyldisiloxane. Our
+
+
Feb., 1946
METHYLCHLOROSILANE AMINES
Droduct had the followinn DroDerties: b. D. 125.7-126.2' 1758 mm.) ; n20~I .4078;-dio4 1vac.j ami; neut. equiv. 160.2, 160.5 (theory, 161.3). (b) Reaction of Trimethvlchlorosilane with Liauid An%onia.-A 2-liter unsilver
