4432
NOTES
methoxide for one-half to one hour. The steroid was isolated uia a digitonin precipitation a n d obtained free in the usual manner by pyridine cleavage of thr digitonide. From the non-digitonin precipitable fraction the a-naphthylurethan of methyl alcohol was isolated and recrystallized twice from ligroin; m.p. 121-122', mixed melting point with authentic urethan, 122-124'. A model DU Beckman spectrophotometer was used in the determination of the ultraviolet spectra. The infrared spectra of the steroid derivatives, mulled in nujol, were recorded with a Perkin-Elmer model 21 spectrometer equipped with a sodium chloride prism. We are grateful to Roger A . Pickering for obtaining the infrared data. ARGOSNBCASCERRESEARCH HOSPITAL I~BPARTMEST OF MEDICINE ~ 7 S I V E R S I T VOF CHICAGO C H I C A ( > O , ~ l , l , l N O I S ,A N D L)EPAK1.MENI. U P C H E M I S I ' K Y ILLIYOISI s s r r i u m OF TE~HWJLO(;Y
CHICAW,II.LINOI::
Evidence for Carbonium Ion Formation from Crystal Violet in Concentrated Sulfuric Acid 'r!-I(JMPSOS A N D MI1,I'ON
CKIVEIJ \ r A R C H
-4 SMITI1
28, 19\75
I t has been well established that triphenylcarbiarid many substituted triphc.ii\.lcarbinc,Is ionize iri sulfuric acid t o give stably c~arb(~iiiiiiii ioiis, LICcording to the cquatioii ~ C ~ f , I ? , , ~ . , C4O I2li2Sfj, ~ --f (Ctill.,' c' ' i-211so; 4- 1 1 , O ' The solutions are intensely colored, give ifactors of 4.0 by cryoscopic measurements, arid uiidergo various reactions typical of carbonium ions. U'hen the carbinol contains three dimethylamino groups, as in crystal violet. the ionization in sulfuric acid would be expected to be accompanied by protonation of one or more of the dimethylamino groups. Newman and Deno' obtained an i-factor of (i.0 for crystal violet in lOOe& sulfuric acid ; the solution was orange-brown in color. T h c ionization was assumed to be ii(i1
+
C ~ T ~ , X O T T4rr,soI ----f
+
+
((Cli:~i.~.IICijF~risCCI,IiIN(CII,,i I I 0 I-!11S05- (i' =
-+
Vol. 77
should be noted that the formation of the benzenoidal ion in sulfuric acid also leads to an i-factor of (i.0, so t h a t cryoscopic measurements aloize cannot indicate whether a carbonium or benzenoidal ion is present. The different interpretations of the behavior of crystal violet in sulfuric acid solution as presented by Newman and Deno on the one hand and Branch and Warba on the other appear to stem from coiiflicting observations. The orange-brown color observed by the former experimenters is hardly conipatible with the complete lack of bands throughout the range of 273-TO0 nip as reported by the latter workers. Newnian and Deno synthesized their trip-dimethylaminophenylcarbinol by previously described r n e t h o d ~while ,~ Branch and Warba purified commercial crystal violet by recrystallization. No melting points or analytical data are recorded in either case. It is possible that the commercial sample may have been the triphenylmethane derivative which is also known as crystal violet. In view of these possibilities it seemed advisable to obtaiu :I sample of crystal violet identified by analytical iiieans, arid to study its behavior in sulfuric acitl. Results The spectruiii of crystal violet in coricentrated sulfuric acid is shown iri Fig. 1. The spectrum is analogous to other triphenylcarbonium ion spectra studied by Newman' and by Branch2; both the wave lengths of maximum absorption and the value of the extinction coefficients a t maximum absorption are comparable. The color of the solution was orangebrown in agreement with Newman's observations. An identical spectrum was obtained for the methyl ether of crystal violet. I t has been shown3 that a solution of triphenylcarbonium ions in sulfuric acid reacts immediatelJwith methanol to form the corresponding methyl ether, according to the equations
+ ( c ~ H ~ ) ~ c+; IH?O o c--+ H ~(c,ITaxoCir.-t- 1 1 . 0 IC~H,),C+ C H ~ O I I-+ (C611ajlCTIOCIf,
1
The ordinary acid-catalyzed etherification o f triphenylcarbinol or similar hindered alcohols, ;is well .417 i-factor of (5.U2 was obtained for the xilalogous as bisulfates of the type proposed by Branch, would tri-p-aminopheriylcarl)iiiol,with the saiiie type o f be expected to give practically 110 ether in a short ionization assumed. reaction time. A solution of crystal violet in conBranch and Walba? have reported measurement centrated sulfuric acid gave the methyl ether i l l of the spectrum of crystal violet in concentrated approximately 50c& yield for a reaction time of not sulfuric acid3 and reported t h a t no bands were pres- more than 5 minutes. This is an excellent confirinaent from 2i.S-700 mp. The explanation offered is tion of the assumption t h a t the solution contained that all three dimethylamino groups are protoncarbonium ions. ated, thereby increasing the positive charge on the The above data support the original conclusion oi molecule t o the extent t h a t any carbonium ion, if Yewman that only two dimethylamino groups of formed, reacts with a bisulfate ion to give a henzecrystal violet are protonated in sulfuric acid, and noidal ion the resulting species is a carbonium ion. Newrnaii (ICFI,jUSCbfll)!COF[ 51fUS04 --f also suggested that the electron distribution in this ion was represented more accurately by the ( I u i i l irc,;ir4),;cFiso4 i~ I I S O -, ir.,o+ inoid structure This benzt.rioida1 ion should possess no color. I t fi.rii
+
+
(1) hl. S. N e w m a r l uiid N. C'.
U r n o , 'I'HIS
JUUKSI.,
73, 3ti44
f 1951).
(2) G . Brduch and H . Walba, i b i d . , 76, IJH4 rlH5.l). (3) A solutio" of Z parts acrtrc acid a n d !i8 parts r u i i r r i i l r a f r d firric acid.
y111-
COMMIJNIC.\TIOK'S TO T r m EDITOR
.lug. 20, 1955
4433
than by the usual carbonium ion structure 45
The rapid formation of the ether suggests that the latter formulation is of considerable importance. Experimental The spectra were run on a Beckman model DU spectrophotometer; 1-cm. thick quartz cells were used. Crystal Violet Preparation.-The carbinol was prepared by dissolving the chloride salt of crystal violet (a product of the National Aniline Division, Allied Chemical and Dye) in water and precipitating the carbinol by addition of solid sodium hydroxide pellets. After filtration, the solid c u binol was blue in color. Attempts t o purify the product by recrystallization from ligroin and benzene were futile. The method of purification was that followed by Heertjes, et I Z ~ . who ,~ also reported that simple recrystallization from organic solvents failed. The carbinol was dissolved in acetone and precipitated by the slow addition of an aqueous sodium hydroxide solution. The carbinol, collected by filtration, was much less colored than originally. Repetition of this process several times yielded a white solid. 10 order t o remove traces of sodium hydroxide from the carbinol, it was dissolved in a small volume of anhydrous ether, and the solution filtered. The carbinol, after evaporation of the ether, was stored under vacuum since exposure to air muses the compound t o turn blue slowly. The compound melted a t 200" dec.; a melting point of 205" has been reported.6 .4nal. Calcd. for C2&lONa: C, 77.08; H, 8.02. Found: C, 77.02; H , 8.11. Methyl Ether Preparation.-A 3-g. sample of crystal violet was dissolved in 10 ml. of concd. sulfuric acid, giving an intense orange-brown solution. This solution was added dropwise t o 50 ml. of absolute methanol which was cooled in a ( 6 ) P. hl. Heertjes, J. C. Vat1 Kerkhof and K .4 friru. chim., 62, 737 (1943).
De Vries, Rec.
4.0 w
8
A
3.5 3.0 \
I
250
300
-
I
L-
I
1
I
I
I
360 400 450 500 Wave length, nip. Fig. 1.-Absorption spectra of triphenylcarboiiium ions in H2SO4: -, tri-p-dimethylaminophenylcarbinol;. , . . , tri- -, triphenylcarbino1.1 Crystal p-nitrophenylcarbinoll; 14. Spectrum for methyl violet solution was 2.5 X ether of crystal violet was identical with that for crystal x-iolet.
--
Dry Ice-acetone-bath during the addition. The resulting solution was immediately added to 200-300 ml. of a n icewater mixture and sufficient sodium hydroxide solutioii added to make the solution basic. The methyl ether wa\ extracted from this solution with two 100-ml. portions of ether. Evaporation of this solution yielded a blue solid which was purified in the same manner as that describetl for the carbinol. The yield of crude material was allproximately 50%: The purified ether melted with deconiposition a t 145-lsOo. Anal. Calcd. for C28H330N3: C, 77.38; H, 8.91. Found: C, 77.16; H, 8.19. DEPARTMENT OF CHEMISTRY OF TENNESSEE UNIVERSITY KNOXVILLE, TESNESSEE
-
COMMUNICATIOhTS T O T H E EDITOR TRACER STUDY O F T H E OXYGEN IN PRECIPITATED HgO AND Agz0 ; 0" ISOTOPE EFFECT I N THE REACTION OF OH- WITH H g + + AND Ag+
Sir: 0 I 8 experiments have been used to investigate the source of the oxygen in HgO and 4g20 precipitated by OH-. Further, an 0 I 6 / O l 8 effect was found in the reactions of H g + + and Ag+ with OH-; 016concentrates preferentially in the precipitated oxide. Tracer experiments were as follows. One ml. of a solution of AI NaOH, enriched in 0 I 8 (1.4 atom %) was added to 100 ml. of ca. M Hg(NO& in normal water (0.20 atom % 0l8);the precipitated HgO was washed, dried and pyrolyzed. The 0l8'0l6ratio of the liberated oxygen was 1% lower than the ratio for normal water. Thus the oxygen in the HgO is derived exclusively from the original aquo-cation, which apparently exchanges only slowly with solvent water. Proton transfer from hydrate water to OH- is a satisfactory explanation. However, in similar experiments with
AgN03 (1-9 M ) , the AgzO was found to be substantially enriched in 018,ranging from 0.67 atom % (no stirring, high [Xgi]) to 0.26 atom % (vigorous stirring, low [Ag+]).l A reason for the difference between H g + + and A g t may be the tighter binding of the coordinated water around the more polarizing Hg++.? The 0l6/'Ol8 isotope effect was studied with normal 0 I 8 abundance materials. Oxygen from the pyrolysis of the oxides was compared mass spectrometrically with reference oxygen which ratio from distilled differed by +2.3y0in 018,/016 water. Fractionation factors with respect to water were computed: cr = ( 0 1 6 / 0 1 8 ) o x i d e / (016/018) watet . When a solution of Hg++ was added to a solution containing excess OH-, a was 1.018. For the (1) Auxiliary experiments showed negligible exchange between water and HgO or A g i o in 1 hr. ( 2 ) 13. Tanbe, el ~ l !discusged . in J . Phys. Chen., 68, 523 (1954)) studied exchange between hydrated cations and solvent water and found differences in lability for various (other) cations.
