RAYMOND E. DESSS
1580
containing a few drops of water a t reflux for 12 hours. Dilution with 250 ml. of water yielded a small recovery of the amide (0.21 g.), and acidification and cooling of the filtrate gave after purification 0.80 g. (40%) of yellow needles, m.p. 201-204'; mixed melting point with an authentic sample of ferrocenoic acid (m.p. 210" dec.), 204-206'. I-Acetyl-I '-carbamylferrocene was prepared by a procedure essentially identical t o that described above for 1acetyl-I '-diphenylcarbamplferrocene. From 4.6 g. (20 [ C O N T R I B U T I O S P R O M TIIE D E P A R T M E S T O F
Vol. 82
millimoles) of carbaniplferroceiie was obtained 3.2 g. (61(,&) of the acetylcarbamylferrocene as orange-red crystals (from benzene) melting at 151-152'. Attempted hydrolysis of l - a c e t ~ l - l ' - c a r b a ~ n ~ l f e r r o c c ~ ~ ~ in 10-15y0 ethanolic potash, acetic-hydrochloric acid mixtures, concentrated hydrochloric acid, and sulfuric acid containing sodium nitrite gave only intractable tars.
CHAPELHILL, S C CHEMISTRY, ~ S I V E l C S I Yl
OF C I N C I S N A Y " ]
The Constitution of Organoberyllium Halides BY RAYMOND E. DESSY RECEIVED ACGVST3, 1959
+
(CsH6)?Be Be*Br2 ~2 2CetIjBe*13r has bceu irivertigateti usirig Be' as Dxcliaiige i u thc s)'stetii (C&)2Bc.13e*Br2 tag. Results sliow that, there is no exchange uf Be between the two species (CaHj)2Bc and BeBr,, and iiidicate t h a t there is no sucli structure as CeHjBeBr.
ii
The normal reaction between ail alkyl halide and beryllium metal apparently cannot be used to form an organoberyllium ha1ide.j Gilman has succeeded6 ReM f MX2 2RhIX (1) in preparing ether solutions of substances described used to represent the structure in solution of organo- as organoberyllium halides by reaction of K X commetallic compounds of the Group I1 metals are not pounds and Be metal in the presence of HgC12 as universal. By use of radioactive tracers i t has a catalyst under high pressure and temperature. been shown1 t h a t where LI = Mg, and R = E t , No analytical data were reported. C6H6, there is no exchange of Mg between Gilman also has pointed out7 that solutions of R2Mg species and Mg*X2 species, thus indicating dialkylberyllium react with an excess of beryllium the non-existence of RMgX. Similarly, Garrett2 chloride to give a solution "with the characteristics" has reported t h a t where M = Cd and R = E t of "organoberyllium halides." Apparently all there is no evidence for exchange of Cd between t h a t is known about the differential behavior of I12Cd species and CdXz species. dialkylberyllium compounds and "organoberyllium The existence of the species R2Cd.CdXz receives halides" is t h a t the former react more rapidly with support from equilibrium constant measurements hIichler ketone.8 I t also has been shown8 t h a t and molecular weight determinations. Prelimi- thermal decomposition of the materials ternied nary data2 involving measurements of equilibrium organoberylliuni halides leads to the formation and constants in the system AI = Zn and R = E t also evolution of dialkylberyllium compounds. X11 of indicate the non-existence of the RZnX species. this evidence for the existence of RBeX is speculaOn the other hand, there is little doubt t h a t in the tive. system hI = Hg, and R = alkyl, aryl t h a t a species CoatesYreports that although (CH3)2Be reacts such as RHgX does exist-it is isolable, and with methanol explosively, the reaction with HCI molecular weights indicate a monomer. In most requires long periods of heating The gas evolved solvents the equilibrium appears to be almost was identified as methane, and the white solid quantitatively to the right. Rate measurements remaining was assumed t o be inethylberyllium have been made on the system, and a mechanism for chloride (logPmm = 6.52 - 2614/T; cf. (CH&Be the reaction has been proposed. 3 , 4 where log P m m = 13.29 - 51OO/T). It thus appears that there is only one demonI n light of the scant knowledge as to the existstrable case of the existence of an R M X com- ence of organoberyllium halides, i t was felt that an pound. Beryllium seemed to provide the most investigation of exchange in the system important additional test case.
I t has been made increasingly clear, over the last few years, t h a t equilibria of the Schlenk type
R,Bc Be*&
( 1 ) Raymund E. Ilessy and (3. 5 . H m d l e r . 'TllIS
JoUKNAL,
80,
3521 ( l % j X j . ( 2 ) A . B. Garrett, Arthur Sweet, U', I,. Slarsliall, David Riley and Anis l'ouma, Rrc. Clzcm. Progress, 155 (1932). ( 3 ) Raymond E. Dessy aud Y. K. Lee, T H I S J O U R N A L , 82, 689 (1960). (4) In a paper b y I. B. Johns and R.M Hixon ( J . Phys. ( ' h e m . , 34, 2 2 2 6 (1930)) in which measurements of the equilibrium constant for the above reaction were measured by e.m.f. studies on the cell
IIgI: 0.005 .I/
RzHg
+ HgT, If 2RIIg-I
i t was concluded t h a t in t h e wries R = A l e , E t , w B u the equilibrium constant was of the order of 100, b u t when R = CeHs of the order of 1. I t seems apparent t h a t their assumptions t h a t t h e junctions I i I I g I ~ I I gand RzHg/Hg arc z e r o are invalid. .Uthough t h e exiierimental technique looA5 promising, further work apparently is necessary.
2R2Be + Be*&
2RDc*X
(2)
would be of value. For practical reasons, I< was chosen to be phenyl and X bromine.
Experimental Beryllium Metal.lo-Cyclotron target material lia\ring the analysis (spectrographic) given was used: AI, ( 5 ) Henry Gilman, Tms J O U R N A I , , 46, 2G93 (1933). ( 6 ) Henry Gilman and F. Schulze, ibid , 49, 2901 (1927). (7) Henry Gilman and F. Schulze, J . (hem. S O L .2G63 , (1927). (8) Henry Gilman and F. Schulze, T H I S J O U R N A L , 49, 2904 (1927). (9) G. E. Coates, F. Glockling and N. I). Huck, J. C h e m . .SOC., 4512 (1952). (10) All handling of beryllium metal should follow t h e suggestions laid down in t h e pamghlet "Some Notes on Safe Handling Practice, for Beryllium," distributed by the Defense Aletals Informution Cent e r , Uattelle hlelnorial Institiitr,
April 5, 1960
1581
CONSTITUTION OF OKGANOBERYLLIUM HALIDES
0.05; Cr, 0.03; Cu, 0.001; Fe, 0.05; Ca, 0.03; Ge, 0.005; K , 0.05; Li, 0.01; Mg, 0.01; and M n , Ni, and Si, lanilinium Difluorenyl T:lui~re~i~-ltrimeth~lammonium ~-Plien?-leth~-ltrimethylam2,3-I)iphenylbut~rnc nionium
%
31.9 35.0
l5,O 6.0 26.1 30 0
(1) R I . Finkelstein, 11. C . Petersen and S. D. Ross, THISJUUKNAI.,
81, 2361 ( l g i i l ) .
The problem of mechanism for thcse reactions resolves itself into the question of whether the cathodic reaction involves a onc -electron change and the formation of radical intermediates or, alternatively, a two-electron transfer and carbanion intermediates. The two possibilities may hc formulated as
++
R1>;(;
KP
4
I
+2e
,,A
+
Ib,
> X P
+
Iy
:
It2,'
,
R1
L----+
Rz
"
The coupling product could result either from dimerization of the radicals or a nucleophilic displacement reaction by the carbanion on the quaternary ammonium ion.2 Some qualitative evidence which favors the freeradical mechanism has been presented,' but definitive proof is lacking. An experiment with an optically active quaternary ammonium salt mould permit a clear choice between the alternativcs provided that the cleavage occurred at the asymmetric carbon atom and the coupled product was not formed in only the meso form. a-Phenylethyltrimethylammonium nitrate proved suitable for this purpose. It could be obtained optically acti gave both meso- and dl-2,3-diphe1iylbutanc 011 electrolysis. In the electrochemical decomposition of active a-phenylethyltrimethylammonium nitrate, that portion of the hydrocarbon product which was not meso would be expected to be totally inactive, if free radicals are the intermediates. Since the radicals, if formed, are formed singly rather than in pairs, this is a more straightforward case than the decomposition of a diacyl peroxide where (2) hfechanisms involving hydrogen abstraction from the quaternary ammonium ion by the carbanion, Rp: -, have not been C U I I sidered, since the hydrocarbon RaH was not found in t h e p r ( ~ I u c l i under the present experimental conditions. Hydrogen abstractiim is, in fact. a inure proljal,!e reaction f o r these species than is displ;rcctneiit, and this may be c u n d c r c d ;is cvirlence against the c a r l ~ ~ n i o r i niechanism.
