C L ~ ~ LI. R KL ~ E L C I I.\ND HILTON X.SMITH
1312
the reported m.p. of the substance prc[)ared by chlorosulfonation of ethyl N-phenylcarbamate is lOnYo.'3 A mixed m.p. of the products prepared by the two methods showed no depression.
_-(13) R. Adams, 1'. 11. Long a n d A. J. Johansun, ' h i s -
-
\Vhen a reaction mixture of ethyl S-phenylcarbamate and 15% oleum was heated to higher temperatures than previously, evolution of carbon dioxide started at 75' and became vigorous at 90'. Sulfanilic acid precipitated when the rextion mixture was pourcd on ice.
J U U K N A I . ,61,
SEW YORK, N. Y.
2342 (lQ3!1).
[CONTRIBUTION NO.
Vol. 7 5
109 FROM T H E
DEPARTMENT O F CHEMISTRY, UNIVERSITY OF T E N K E S S E E ]
Reactions of Carboxylic Acids in Sulfuric Acid BY CLARK hl.
II'ELCH AND
HILTON A.
Sh'IITH
RECEIVED SEPTEMBER 25, 1952 The reactions of several organic acids in sulfuric acid have been investigated. The "i" factors have been determined by cryoscopic methods, the evolution of carbon monoxide studied, and the carbonium ions present when the acids are dissolved in sulfuric acid have been allowcd to react with various reagents. The influence of certain substituents on the stability of the carbonium ions also has been investigated.
4 n interesting and peculiar preparation of stable triphenylcarbonium ions is by the reaction of triphenylacetic acid with sulfuric acid. Carbon mon oxide is quantitatively evolved. (CsHS)3COOOH $- HzSOa + (CaH,)3CC
+ CO + IIiOi + 2HSO4-
TABLE I THEYAN'T HUFFFACTORS A N D YIELDSOF CARBUN hloxOXIDE FROM CARBOXYLIC ACIDS IN 1 0 0 ~ oSULFURIC ACID AT 25" Yield
t,
Color of solution
Organic acid
Dip henylacetic
Yellow-orange
$2
4 0 4 9 3.3 3.5 3.7 1.7 2.3 5 4 58 3.7 4.8 2 0 2 0 2 0 3 0 ti3
min
ofpcp /C
loh 40 180 50 98
If the deeply colored solution is subsequently poured into water, triphenylcarbinol is precipitated,' or if the solution is poured into an alcohol E&-( p-chloropheny1)- Orange loob acetic 'the corresponding ether is obtained.2 The well 1200 known decomposition of formic acid in sulfuric acid Pale yellow 20 0 Bis-(p-nitropheny1)is similar to this reaction. 1400 acetic The behavior of other carboxylic acids in 100% Red 30b 15d sulfuric acid has been studied in order to deter- Benzilic 1300 mine how generally the loss of carbon monoxide oc15 0 Red curs in this medium. Several m-ethods were used to 4,4'-Dichlorobenzilic 80 determine the course of the reaction in each case. Colorless 13 0 The van't Hoff i factors were determined cryoscop- Isobutyric 1 100 ically. The principal constituent of any gas Colorless 20 15 evolved was identified and the total volume de- Ilicyclohcsylacetic Yellow 250 52 termined. The organic products obtained by pourRed-brown (300 65< ing the reaction mixture into various reagents were 10 5 determined in several instances. X portion of the 2,4,G-Trimethylcyclo- Colorless 1x0 44 hexanecarboxylic results appears in Table I, the remainder following Red-brown 1100 7 Y in the discussion of the individual compounds. a The average of several determinations. The time at Diarylacetic Acids.-The value of i that should which gas evolution had become sufficiently slow so as not theoretically be obtained after the loss of carbon to interfere with the freezing point determinations. Asmonoxide has taken place is 4.0, since the carbon suming the gas to be pure carbon monoxide (see Experimonoxide formed does not remain in the solution. mental). d F o r the most dilute solution used (0.011 Af The van't Hoff factors actually observed for di- in benzilic acid). The gas was contaminated with SOZ. phenylacetic acid increased rapidly with time owing (CeHs)*CHCOOH HzSO~+ to the occurrence of sulfonation and are inconclu(CaHs)gCH + CO $. H30' + 2HSOasive. The low yield of carbon monoxide indicates The same reaction was probably observed by that competing reactions predominated. When the colored solution was poured into water, Bistrzycki and Siemiradzki, who isolated bis-(diphenylmethyl) ether from the reaction of fuming Lt 13% yield of bis-(diphenylmethyl) ether was obtained. Alternatively, when the solution was sulfuric acid with diphenylacetic acid. The quanpoured into glacial acetic acid and the product titative conversion of a,a-diarylpropionic acids to saponified, a lOy0yield of diphenylcarbinol was ob- 1,l-diarylethylenes by treatment with sulfuric acid tained. These reactions are characteristic of di- and then with water6 undoubtedly proceeds by conphenylcarbonium ions, which have been shown to version of the acids to stable methyldiarylcarbonium ions. This type of ion reacts with water to be stable in 100Tl sulfuric acid.3 It is clear that a form an olefin6rather than a carbinol. reaction analogous to that of triphenylacetic acid (4) -4.Bistrzycki a n d B. v. Siemiradzki, Bcr., 39, 51 (19OG); 41, did occur to a rniiior cxterit. lGG5 (1908).
+
+
( 1 ) A Bistrzycki and E Reintke, Beu.,
(2) (3)
38,839 (1905)
H A. Smith and R J S m i t h , THIS JOURNAL, 70,2401 (1948). C hl Welch and II. A Smith, rbid., 72, 4748 (1950)
(.5) A. Bistrzycki and E. Reintke, ibid., 38, 840 (1905). (6) M. S. Newman a n d N. C. Deno, THISJ O U R N A L73, , 3644 (1951).
March 20. 1953
REACTIONS OF CARBOXYLIC ACIDSIN SULFURIC ACID
1413
A striking example of the carbonium ion forma- to the deactivation of the benzene rings. No carbon tion is afforded by bis- (p-chloropheny1)-acetic acid. monoxide whatever was evolved. This result is The van’t Hoff factor was 3.5 and the yield of car- confirmation that the a-hydroxyl group is attacked. bon monoxide was nearly quantitative. Addition It is opposite to the effect observed in the diarylof the colored reaction mixture t o water resulted in acetic acids, in which the resonance effect of the an 88% yield of bis-(p,p’-dichlorodiphenylmethyl) chlorine atoms facilitated the loss of carbon monether. Addition of the reaction mixture to acetic oxide. The van’t Hoff factors observed for 4,4‘-dichloroacid and saponification resulted in a 52% yield of bis- (p-chloropheny1)-carbinol. These reactions are benzilic acid increased rapidly with time, indicating characteristic of stable bis-(9-chloropheny1)-carbo- that the highly colored solute was undergoing rapid n i u q ions.6 secondary reactions. When the reaction mixture By contrast, bis-(p-nitropheny1)-acetic acid un- was allowed to stand just long enough to reach its derwent no loss of carbon monoxide whatever. Its most intense coloration before it was poured into van? Hoff factor places i t with di- and trichloroace- water, most of the original compound could be retic acid7 as a very weak proton acceptor, after allow- covered. The highly colored solute initially formed ance is made for the partial protonation of the nitro was probably a resonance hybrid of carbonium and groups such as occurs in nitrobenzene.8 The oppo- oxonium ion structures. site effects that p-chloro and p-nitro groups have ArtCCOOH + H2SOa--f on the reaction of diphenylacetic acid in sulfuric I OH acid are the ones that would be predicted on the [Ar&OOH C+ Ar2C-C=& ‘0’ J basis of the resonance effects of these groups. Benzilic Acids.-The intense and characteristic color which benzilic acids exhibit in sulfuric acid The acid-catalyzed condensation of benzilic acid could be the result of two modes of reaction, one in- with benzene derivatives to yield triarylacetic volving the a-hydroxyl group or one involving the acids1° very probably proceeds via the formation of carboxyl group. The observed van’t Hoff factors such ions as the reactive intermediates. Aliphatic Acids.-The evolution of carbon monare inconclusive owing to the occurrence of extensive sulfonation and polymerization. The yields oxide from such acids as dicyclohexylacetic acid of carbon monoxide were found to vary consider- acid and 2,4,6-trimethylcyclohexanecarboxylic ably with the proportions of benzilic acid and sul- when these were dissolved in 100% sulfuric acid furic acid used. In 0.011 M solution, the yield was was unexpected. Trimethylacetic acid is known 15% while in 0.44 M solution i t rose to 39%. The to undergo this reaction a t elevated temperatures’l same trend was observed using 96% sulfuric acid. but is stable in 100% sulfuric acid a t room temperaThis is in contradiction t o the statement of Klinger ture.6 Isobutyric acid is also stable, its van’t and Standkeg that the reaction is characterized by Hoff factor of 2.0 indicating the carboxyl group t o exactly a 50% yield of carbon monoxide. The in- be protonated. It is evident that branching a t the crease in yield with increasing concentration of ben- P-carbon atom of the carboxylic acid is as important zilic acid can be accounted for if the a-hydroxyl as branching a t the a-carbon in promoting the loss group is preferentially attacked to give a carbonium of carbon monoxide. Attempts to isolate and identify the organic products of the reaction were ion which initiates polymerization. unsuccessful. Synthesis of 4,4’-Dichlorobenzilic Acid.-Montagne12 has described the rearrangement of 4,4’OH dichlorobenzil to 4,4‘-dichlorobenzilic acid and the conversion of the crude acid to its silver salt, which (CsHs)zCCeH4COHCOOH (CeH~)&CsH4&OOH can be recrystallized readily, in contrast to the I I -+COOHCBH~ LOOH (!!6HT crude acid. The acid obtained from the purified The polymers thus formed would be poly- (triaryl- silver salt in this Laboratory was rather impure, acetic) acids which would readily lose carbon mon- probably due to a slow decomposition of the silver oxide to yield highly colored polymeric ions. The salt observed to occur when that compound was in polymerization (and carbon monoxide evolution) solution. Another method of synthesis was developed in would be favored by increasing concentrations of which bis-(p-chloropheny1)-acetic acid was conbenzilic acid. If, on the contrary, the carboxyl group of benzilic verted to its acid chloride, chlorinated under irraacid were preferentially attacked b y the sulfuric diation and the resulting bis-(p-chloropheny1)-aacid, the yield of carbon monoxide should decrease chloroacetyl chloride was hydrolyzed in aqueous slightly with increasing concentration of organic sodium carbonate. The acid prepared in this way solute (and decreased strength of sulfuric acid). A was readily purified without being converted to its slight decrease was actually observed with diphenyl- silver salt. The over-all yield was 48y0. The starting material is available from the alkaline hyacetic acid. drolysis of DDT.I3 This synthesis is the first reI n 4,4‘-dichlorobenzilic acid, the probability of (10) For examples, see A. Bistrzycki and L. Mauron, ibid., 40, 4060 polymerization by the sulfuric acid is reduced owing (7) Cf.L. P. Hammett, “Physical Organic Chemistry.” McGrawHill Book Co., Inc., New York, N. Y.,1940, p. 46. (8) H. P. Treffers and L. P. Hammett, Tnrs JOWXNAL. 19. 1708 (tr137). (D) € Klinaer I. and 0. Stradks, Be*., 12, 1214 (1880).
(1907); A. Bistrzycki and W. Niederberger, Hclu. Chim. Acta, 11, 263 (1928). (11) A. Bistrzycki and L. Mauron, BGY.,40,4870 (1907). (12) P. J. Montagae, Rcc. trao. chim., P i , 19 (19021. 113) 0. C r u . u d t t , A. D.l& and R, Eg.8, drd, S Y I N I D P SSC, S , 21
(10461.
CLARKM. WELCHAND HILTONA. SMITH
1414
Vol. 75
ported conversion of a diarylacetic acid to the corresponding benzilic acid. The intermediate bis(p-chloropheny1)-a-chloroacetylchloride is a new compound, Experimental
6% aqueous sodium hydroxide, 50 ml. of ether being added prior to the second washing. The ether solution was dried over anhydrous magnesium sulfate, filtered and evaporated to dryness. There was obtained 0.15 g. of a liquid residue This product was treated with a solution prepared from 1 g. of sodium and 30 ml. of 95% ethanol. The mixture was Freezing Point Measurements.-The apparatus, tech- refluxed for two hours. I t was poured into 100 ml. of water nique and solvent have been described el~ewhere.~The and shaken with 50 ml. of ether. The ether layer was freezing point depressions measured in duplicate runs separated and dried over anhydrous magnesium sulfate. The solution was filtered and evaporated t: drynes7. The checked within 3% of the total magnitude of the depression. residue weighed 0.09 g. and melted a t 62-64 . Eastman Kodak Co. diphenylacetic acid was recrystallize? The solid was dissolved in methanol. The hot solution from benzene prior to use. It then melted a t 147.4-148.4 , Eastman benzilic acid was used without further purification. was treated with Norite A and then filtered. The firtrate Eastman isobutyric acid was redistilled prior to use. Bis- was evaporated to dryness. The product melted a t 64.5(p-chloropheny1)-acetic acid, m.p. 167.2-168.2', was pre- 66.1". A mixture of this with diphenylcarbinol melted The known sample of diphenylcarbinol pared from DDT.13 I t was recrystallized from byzene. a t 64.9-66.1'. Bis-(p-nitropheny1)-acetic acid, 1n.p. 172.3-173.3 , with melted a t 65.3-66.2". The Conversion of Bis-(b-chloroDhenvl)-aceticAcid to Bisdecomposition, was prepared by nitrating methyl diphenylacetate.14 Dicyclohexyiacetic acid, ni .p. 139.4-140.0", was (p,p(-dichlorodiphenylmdhyl)Etlier.2-A 0.300-g. quantity prepared by the hydrogenation of diphenylacetic acicl over of bis-(p-chloropheny1)-acetic acid was added to 20 ml. of Addms platinum catalyst in acetic acid.'& I t was recrys- 100% sulfuric acid. The mixture was stirred for 40 minutes tallized from ethanol-water . 3,4,6-Trimethylcyclohexane- and then poured onto 100 g. of crushed ice. The mixture carboxylic acid, m.p. 105.0-106.0°, was prepared by was treated in the same manner as for diphenylacetic acid. A transparent gum was obtained which crystallized when the hydrogenation of 2,4,6-trirnethylbenzoic acid over .kdanis platinum catalyst.I6 I t was recrystallized from pe- stirred with 1 ml. of methanol for 5 minutes. The yield of troleum ether. Thc preparation of 1,4'-dichlorobenzilic solid was 0.23 g., melting a t 121-125'. It was recrystallized acid will be described separately. The above melting points three times from methanol, after which it melted a t 129.2129.9'. An authentic sample of the etherlg melted a t 1029.4are corrected. 129.9'. A mixture of the two melted a t 129.4-129.9 . Analysis of the Gases Evolved.-The gas evolved from The Conversion of Bis-(p-chloropheny1)-acetic Acid to the given compound in 100% sulfuric acid was collected and a lighted splint was introduced into the gas. If the gas was Bis-fp-chloropheny1)-carbinol.-The solution prepared as observed to burn with a visible flame, i t was considered to above was poured after 40 minutes into 50 ml. of glacial acebe chiefly carbon monoxide, since other gases which could tic acid which had been cooled to its freezing point. The be evolved would not be combustible. This behavior was reaction was carried out and the crude ester saponified in the same way as for diphenylacetic acid. There was obactually observed in all caSes in which a gas was evolved. After tJ:o The volume of gas evolved was determined in an appara- tained 0.143 g. of a solid melting a t 89.2-92.6'. tus similar to that used in the Tschugaeff-Zerewitinoff recrystallizations from ligroin, it melted a t 92.8-93.3 . method for determining active hydrogen, the gas being An authentic sample of bis-(p-chlorophenyl)-carbinolzO collected over mercury. Check runs were made using so- melted a t 92.5-93.3'. A mixture of the two melted a t dium formate as the solute, since this is known to react with 92.8-93.4 O . The Behavior of 4,4'-Dichlorobenzilic Acid in 100% Sulsulfuric acid to give quantitative yields of carbon m0n0xide.l~ 0.497-g. quantity of the compound was Apparent yields of 101-102% were obtained in collecting furic Acid.-A 3.5-19.0 ml. of gas. With two of the compounds studied, added to 30 ml. of 100% sulfuric acid with vigorous stirring. Solution was completed very quickly, resulting in an intense sulfur dioxide was detected by its odor. The Conversion of Diphenylacetic Acid to Bis-( diphenyl- red color. The solution was stirred for one minute and then methvl) Ether.-A 1. 0 0 - ~ auantitv . of diuhenvlacetic acid was poured into 200 ml. of ice-water. The color disappeared was added to 40 ml. of-l06% suljuric aiid Gith stirring. instantly. The mixture was stirred and was then shaken After the solution had been stirred for 5 minutes, it was with 100 ml. of ether. The ether layer was separated and poured into 300 ml. of ice-water. The orange color of the was dried over anhydrous sodium sulfate after which it was solution disappeared immediately. The mixture was filtered and evaporated to dryness. The myeight of the residue was 0.497 g. indicating no sulfonation had occurred shaken with two 40-ml. portions of carbon tetrachloride. during the short period of reaction with the sulfuric acid. The combined carbon tetrachloride layers were shaken The product was stirred with two 75-m1. portions of hot with 50 ml. of 5% aqueous sodium hydroxide, were separated and evaporated to drynezs. There was obtained 0.12 60-90" ligroin. After these portions were decanted a small amount of gummy solid remained. The combined ligroin g. of a solid melting a t 95-104 . The product was dissolved in methanol and the hot solu- solutions were evaporated to a volume of 2 ml., most of the tion was treated with Norite A, which was found to remove organic product separating from solution. The mixture an uncolored impurity. After filtration the solution was was stirred with 75 ml. of hot ligroin, which was again deconcentrated and the solid was allowed to crystallize. canted from a small amount of gummy material. The ligroin was concentrated to 2 ml. The cycle was repeated After an additional recrystallization, it melted a t 108.9109.4'. A mixture of this with an authentic sample of bis- with 30 ml. of ligroin and then with 15 ml. of ligroin. This (diphenylmethyl) etherla showed the same melting point. method of purification is based on the fact that the gummy impurity dissolved in ligroin more slowly than did the main The authentic sample melted a t 109.1-109.6°. product. The Conversion of Diphenylacetic Acid to Diphenylcarbino1.-The solution prepared from diphenylacetic acid and The ligroin solution finally obtained was seeded with 4,4'sulfuric acid as above was poured after 5 minutes into 50 dichlorobenzilic acid and allowed to evaporate at room ml. of glacial acetic acid which had been cooled almost to temperature with occasional stirring. The residue solidiits freezing temperature. The reaction temperature was fied completely after one week. The yield was 0.473 g. of kept below 35" during the additions of the sulfuric acid solu- material melting a t 87.5-94.0'. The product did not detion. The mixture was stirred and was poured into 200 ml. press the melting point of 4,4'-dichlorobenzilic acid when of water. mixed with the latter. I t s neutral equivalent was 330. The aqueous mixture was shaken with 100 ml. of ether. Assuming it to be 4,4'-dichlorobenzilic acid contaminated T h e ether layer was washed with two 50-ml. portions of with an impurity of very high neutral equivalent, its purity was 90%. The extent of recovery of 4,4'-dichlorobenzilic (14) L. Haskelburg and D. Lavie, THISJ O U R N A L , 71, 2580 (1949). acid was therefore about 85%. Since the sulfuric acid solu(15) H. A. Smith, D. M. Alderman a n d F. W. Nadig, i b i d . , 67. 273 tion had reached its full intensity of coloration during the (1945). one-minute reaction period, it appears that the highly col(16) 13. A. S m i t h and J. A . Stanfield, i b i d . , 71, 82 (1949). 117) T.:. R . Shierz. Ibid., 45, 4 4 7 ( l Q 2 3 ) ; X l Wegner, %. ntznl. C'hem..
42, 42' r i w x i . 13) i > + * , m ? m , l 11u9i.r. I
119) Prepared by t h e method of 0. Orummitt and A. C. Buck, THIS 6 7 , 693 (1945). (20) Prepared by the m e t h i d ' r f J. 1'. Nrmric a n d 73. X f , l'ihhetta, iw,,ii,2001 ( 1 9 a w t
JCIVRNAI.. t w
t h e m"tl,,.d ,d
.z
XI
Lt'wd, J : C k r m . :To
