1426
voi. 74
ALBERTL. HENNEAND RALPHL. PELLEY
These bonds are weak, being of the order of 5 kcal. per mole.16 Thermal agitation, ie., increased temperature, disrupts them leading to the disappearance of metachromasy. Conversely, lowering the temperature to near 0” induces metachromasy in solutions which are orthochromatic a t room temperature. In order for the dye “non-polar” residues to polymerize, it is necessary to secure the correct ratio of dye to substrate concentration. In the presence of a large excess of substrate, a small number of dye inolecules may form the usual polar link a t random, but, being widely separated, the aromatic rings of the dye cannot polymerize to yield metachromasy, In view of the findings of Rabinowitch and Sheppard15,14with regard to the spacing requirement for aggregation of thiazine dyes, it should be noted that nucleic acid! have been reported to have a spatial unit of 3.5 A.I7
The effect of solvents on the van der Waals binding also has been shown. Alcohol abolished metachromasy of toluidine blue induced by gum arabic. We have found that alcohol also abolishes metachromasy induced by nucleic acids.g These findings supplement the observations of Sheppard that aggregation of thiazine dyes may involve the bonding of a molecule of water between neighboring resonating dye ions1* and those of Rabinowitch and Epsteinlj that alcohol prevents the dimerization of the thiazine dyes methylene blue and thionine. Acknowledgment.-This work was aided by a grant from the American Cancer Society to the Department of Preventive Medicine, the Johns Hopkins University School of Medicine, recommended by the Committee on Growth of the National Research Council.
(16) I,. Pauling in K. Landsteiner, “The Specificity of Serological Reactions,” Harvard University Press, Cambridge, Mass., 1945, p. 275. (17) W. T. Astbury, “Symposia of the Society for Experimental Biology,” Vol. I, Cambridge, 1947. p , 66.
(18) S E Sheppard and .4 (1944)
[CONTRIBUTIOSFROM
THE
BALTIMORE, MARPLAND
L.
Geddes, THIS JOURNAL, 66, 2003
RECEIVEDMAY26, 1951
DEPARTMENT O F CHEMISTRY, THEOHIO STATE UNIVERSITY]
Aeidity of Trifiuorinated Alcohols and Saponification Rates of their Acetates BY
LiLBERT L.
HENNE.4ND RALPHL.
PELLEY]
hleasured by an electrometric procedure, the ionization constants for CF2CHPOH,CFsCH(CH3)OH and CF3C(CH&OH were found to be not more than In 50% aqueous acetone 0.1 N in NaOH at 25”, the saponification of CF3CH20Ac and CF3CH(CH3)OAc showed second order kinetics with K 0.6 and 0.15 1. mole-’ sec.-I. The rate constants for 0.1 A’ acid-catalyzed reactions expressed as K X lo6 set.-' were found to be: CHFaCH20Ac 2.0, CF3CH20Ac 1.4, CFICH(CH3)OAc 1.2, C F ~ C H ~ C H ~ O2.6 AC and C8HbOAc 4.0. These values are discussed.
Fluorinated alcohols are known to be acidic. They are reported to form alcoholates on standing with alkali carbonates, alkaline-earth carbonates, but not bicarbonates,2 not to form complexes with calcium chloride, tenaciously to resist dehydration14and one of them, CF3CHOHCH3, is reported to have an ionization constant of K = lo-’, intermediate between phenol and acetic acid. The following alcohols were synthesized : CF3CHzOH,* CHF2CHzOH by reduction of difluoroacetic acid obtained in much improved yield, CF&!HOHCH3 by reduction of CF3COCH%obtained in improved yield, CF3C(CH3)20H7 and CF3CHzCH20H.’ Ionization constants were determined by measuring the PH a t the half-equivalence point, which is equal to the pK between the limits of 10 and 4J and, for acids with K , in the range of l o d 3 to a good approximation of it slightly beyond these limits.l0 The observed results a t 25’ were: CFaCHzOH 4.0, CF3CH(CH3)OH 6.3 and CF3C(1) Socony Vacuum Fellow, 1949-19ZO. (2) F. Swarts, Bull. Classe sci., Acad. Roy. Belg., i31-760 (1902). (3) F. Swarts, Bull. doc. chim. Belg., 45, 471 (1934). (4) F. Swarts, ibid., 56, 191 (1927), and J. V. Schmitz, Ph.D. dissertation, The Ohio State University, 1949. (5) F. Swarts, ibid., 58, 99 (1929). ( 6 ) A. L. Henne, R. M . Alm and M. A. Smook, THISJOURNAL, 70, 1968 (1948). ( 7 ) E. Gryszkiewicz-Trochimowski, Rec. irav. chiin., 66, 427 (1947). (8) E. T. McBee and A. Truchan, THISJOURNAL, 70, 2910 (1948). ( 9 ) G . hl. Bennett, G . L. Brooks and S. Glasstone, J . Chem. S o c . , 1521 (1935). (10) A . L. Ilenne a n d C. J. Fox, THIS JOURNAL, 73,2323 ( l ! l > l ) ,
(CH&OH 2.5 X lo-”. The order of magnitude contradicts that given by swart^,^ but agrees with those obtained by Richter” for HOCH2CF2CF~CHZOH, and 10-l2 and by McBee12 for HOCH&FzCFCHzOH and HOCH2(CF2)L!H20H, 10-l2. To test the results, U.S.P. phenol was measured in like fashion, and found to have K = 1X while the accepted value1* for specially repurified phenol is 1 X 10-lo. Since glass dectrode measurements tend to give values which are too low in solutions of pH > 10, and since the actual difference of p H before and after adding the alcohols to the sodium hydroxide was very small, it may be said that the ionization constants of these alcohols has been shown to be not more than and that a plausible value might be taken as 4 >( lo-’*. This shows that the fluorinated alcohols are a t least l o 4times more acid than ethanol.I4 The hydrolysis of esters by hydroxide ions in aqueous solvents is known to show second order kinetics. The accepted explanation is 0
R’C-OR
0-
1; + OH- IfR’C-OH, I
Slow
OR (11) S. B. Richter, PhD. dissertation, The Ohio State University, 1951. (12) E. T. McBee, W. F. RIarzluff and 0. R. Pierce, Verbal, 118th A C.S. Meeting, Chicago, Ill., September, 1950. (13) M. Mizutani, 2. p h y s . Chcm., 118, 318 (1925). (14) 1. H. Hildebrand and P.S . Denner, THIS JOURNAL, 44, 2824 ( I 922).
ACIDITY OF TRIFLUORIWTED ALCOXOLS
March 20, 195’L 0-
0
II
I R’C-OH I
+ OR-, fast
R’C-OH
OR
and R’COZH 4-RO- J_ RCOZ- f ROH, fast
The slow step is facilitated by recession of electrons from the seat of reaction. Since in alkaline solution the probability factor PZ in the Arrhenius equation K = P Z O e - W R T remains fairly constant with changes in the activation energy E , the mechanism is largely independent of steric influences and consequently, since there is only one main parameter, the effect of an electronegative group in the ester can be easily measured. In good agreement with expectations, it was found that the base hydrolysis rate of CF3CH20Ac( K = 0.6 1. mole-’ set.-') was six times faster than that of CH3CH20Ac (0.10816), and that of CF&!H(CH3)0Ac (0.155) six times faster than that of CH&H(CHg)OAc (0.0269. For acid-catalyzed hydrolysis, the assurnedl6 mechanism is a bimolecular, pseudo-first order one ~
0
0
I1
ti
f H + I_ R’C-?HR,
R’C-OR
HzO
11
+ R’C-OHR Jr 0
1;
fast
0
0
+ I1 H20-C-R’
f ROH, SIOW
+
R’C--OH2
-
K X lo6 SIX.-’ were: CFaCH2CH20Ac 2.6, CHFzCHzOAc 2.0, CFzCHzOAc 1.4, CFa(CH3)CHOAc 1.2 and CsHsOAc 4.0. Among the measured rates, phenylacetate agrees with the value 3.9 reported for a homogeneous systemz0 and considered more reliable than the values 7.7 and 6.6 observed in heterogeneous system^'^^^^; the value for CHF2CHzOAc also agrees with the value of 1.7 reported by Swarts2 in an heterogeneous system. It can also be said that our experimental result3 are consistent. CFaCHzOAc and CFs(CH3)CHOAc have rates which differ only slightly; in the other two compounds, where the fluorine content is smaller (CHF2CH2OAc) or more remote (CFaCH2CH20Ac),the rates are faster. €lowever, all rates are slower than those of unfluorinated esters, while they should have Been faster if nucleophilic acceptance of H2O is the dominant factor. Patently, induction is not the main factor influencing the rates of acid hydrolysis, or else the accepted mechanism is inadequate. All rates appear in Table I, with the literature values recalculated for K X lo6 sec.r’, log e, in 0.1 N acid where needed. TABLE I HYDROLYSIS RATESO F ACETATES AT 25.OoZ2 Ester
.
R’COzH
+ H+, fast
1427
CHaCHzOAc
K(8cid)
see. 1-106
11.3*’ 11.023
(CH3)zCHOAc
5. SI9
K(base) 1. mole-’ see. - 1
0 . 10W6
.02616
It has been shown17 that the energy of activation 6 . 224 decreases as substituents on the acid side of the 13.319 .0O15l6 (CHa)aCOAc ester become more strongly electron attracting. 13.219 Since the slow step is the nucleophilic acceptance n-CaH.rOAc 11.210 of a molecule of water, it might be expected that 11.324 electropositive groups on the alcohol side of the CBH~CH~OAC 10.819 . 19716 ester would slow down the reaction, and it is genCaHbOAc 7.7’9 1.2716 erally accepted that isopropyl acetate is saponi6 . 620 fied a t half the rate of ethyl or normal propyl 4.0 acetate. Tertiary butyl acetate is, however, saponi3.9’8 fied just as fast as ethyl acetate, and this reversed CFaCHzOAc 1.4 0.6 trend is explained by calling for an additional CFsCH( CH3)OAc 1.2 .15 mechanism.18 Furthermore, Hin~helwood~~ points CF~CHZCH~OAC 2.6 out that, in acid hydrolysis, changes in electronic CHFzCHzOAc 2.0 ( E )and steric (PZ)factors with change in structure 1.7’ are roughly equal and opposed, and that this causes close grouping of the rates. Experimental In the present study, the rates were determined CHFzC0zH.-At Dry Ice temperature, anhydrous amin a homogeneous solvent of 50% aqueous acetone monia (130 g. or 7.7 moles) was liquefied into a 500-ml. steel by weight and in the presence of an unvarying 0.1 vessel equipped with the customary gage and needle valve, N acid concentration, by following the change of and containing a few grams of cupric acetate: tetrafluoro(65 g . or 0.65 mole) was sucked in. The reaction ester concentration with time. First order kinetics ethylene was allowed to warm up t o room temperature. were observed, from which the hydrolysis constant mixture An exothermic addition took place which was kept under was obtained as 2.303 times the slope of a log (a - control by intermittent immersion in a Dry Ice-bath to prex ) versus t plot,19 where a is the initial ester con- vent the pressure from exceeding 500 p.s.i. The cupric moderates the reaction, which must nevertheless be centration, and x the hydrolyzed amount a t time salt treated with adequate I n 20 minutes, the t; the customary check on accuracy was made by reaction was over and the pressure fell to 220 p.s.i. The plotting ester concentrations a t time t and obtaining excess of ammonia (110 g . or 6.3 moles) was vented off. a smooth curve. The rates measured, expressed as (20) W. A. Waters, J . Chcm. Soc., 1014 (1926). (15) A. Skrabal and A. M. Hugetz, Monalsh.. 47, 17 (1926). (16) J. PIT. E. Day and C. K.Ingold, Trans. Favaday Soc., ST, 686
(1941). (17) E.W.Tim and C. N. Hinshelwood, J. Chem. Soc.. 862 (1938). (18) S. Cohen and A. Schneider, THIS JOURNAL, 6S, 3382 (1941). (18) W. E.Roseveare, ibid., 53, 1651 (1931).
(21) R.Lowenherz, 2. physik. Chem., 16, 389 (1894). (22) A. Hemptinne, ibid., 18, 561 (1894). (23) Average of twelve literature values (24) M. H Palomaa, Ann. Acad Sci. Fenn Scr. A . Tome, 4, N o 2 (1913). (25) G W Rigby, U S Patent 2,484,528 t o du Pont.
ALBERTL.HENNEAND RALPHL. PELLBY
1428
VOl. 74
Ionization Constants.-Ten ml. of 0.0446 N CF3CHaOH mixed with 7.40 ml. of 0.0303 N NaOH (half neutralizatlon point) showed pH 11.4; hence K = 4 X 10-12. Ten ml. of 0.039 N CF3CH(CH3)0H mixed with 6.40 ml. of 0.0303 N S a O H showed PH 11.2; hence K = 6 X 10-12. Ten ml. of 6°F h=C( CHF2)-N=C( CHF&-G=C( CHF?). 0.440 N CF~CfCH1)~OH mixed with 7.30 ml. of 0.0303 N This mixture was refluxed with aqueous sodium hydroxide NaOH showed'PH"il.60; hence K = 2.5 X 10-l2. Ten until the ammonia evolution ceased to be perceptible (about ml. of 0.064 N phenol (U.S.P.) and 10.50 ml. of 0.0303 N six hours), acidified and continuously extracted with ether. NaOH showed bH 8.90: hence K = 1 X lo*. The meRectification of the extract gave 61 g. of wet CHF2COzH cision of the pH reading was 3~0.05unit. u hich was dried by distilling from P206t o yield 49.7 g. (0.52 Base-catalyzed Hydrolysis.-Known amounts of ester mole or 80% over-all) of good acid, b.p. 134O, 7 t a ~1.3428. were sealed in thin wall bulbs placed in 50% aqueous aceCHF~COZH(alternate).-In a 500-ml. steel vessel, ab- tone by weight, 0.01 N in NaOH. After holding 20 minutes colute ethanol (46 g. or 1 mole) was treated with sodium in a thermostat held a t 25.00 i 0.05', the bulb was crushed (2 g. ); after cooling in Dry Ice, tetrafluoroethylene (67 g. a t time zero; a t time t , 1 i d . of 0.1 N HCl was added, and or 0.67 mole) was added. The vessel was sealed and shaken the excess of acid was back-titrated with 0.01 N NaOH to a a t 55' for four hours. The pressure rose to 350 p.s.i. then phenolphthalein end-point. Separate bulbs were used for fell to zero. Working up gave 78 g. (77%) of CHFzCH2- t l , t?, t 3 , etc. The rate constant K = [2.303/t(a $1 log OEt, b.p. 54'. Fuming nitric acid (42 ml.) was added drop- [b(b - x ) / a ( a - x ) ] , where a and b are the concentrat:on of wise to this fluorinated ether (42 g. or 0.3 mole) and re- base and ester. and x the hydrolyzed ester, is reduced to fluxed for 8 hours. Much etching occurred. Pouring into K = [l/t][x/a(a - x ) ] when a and b are approximately water permitted decantation of 23 g. (0.16 mole) of re- equal. covered fluorinated ether. Continuous extraction of the Hydrolysis of CF3CH(CH8)0Acat 25.Oo.-u = 0.00907; aqueous layer with ether, desiccation of P2O6 and distilla- b = 0.00875; stopped with 1 ml. 0.1013 N HCl and back tion gave 15 g. (0.15 mole) of ethyl difluoroacetate, b.p. titrated with 0.00907 N NaOH. Time in sec.: 0, 30, 60, 100°. This is 50% conversion and almost quantitative net 120, 180, 240, 360, 480; corresponding x: 0, 0.0004.5, yield. 0.00089, 0.00187, 0.00205, 0.00245, 0.00301 and 0.00350; CHP2CH~OH.-DiAuoroacetic acid (85.5 g. or 0.89 mole) K = 0.155 1. mole-' read from the l / ( a x ) os. t slope. was slowly dropped into stirred benzoyl chloride (140 ml. CF3CHzOAc measured under comparable conditions gave, or 1.2 moles.); progressive heating to 170" caused the dis- for two sets of measurements, K = 0.6 1. mole-' sec.-l. tillation of CHFzCOCl(75 g., 0.65 mole or 74%) and CHF2Acid-catalyzed Hydrolysis.-Aqueous acetone 50% by COZH (l.0 g., 0.1 mole or 11%). The acyl chloride was weight and 0.1 N in HCl was brought to 25.00 rt 0.05'. dripped into a solution of LiAlH4 (15 g. or 0.4 mole) in an- A weighed amount of ester was diluted in this solvent to hydrous ether (700 ml.) a t the rate permitted by the capacity give 25.0 ml. of solution about 0.1 N in ester. A 2-ml. aliof the reflux condenser. Rapid stirring was required t o quot was withdrawn a t noted intervals and titrated with keep the heavy slurry formed from setting. After refluxing 0.03 N NaOH to phenolphthalein end-point. Notations for one hour, water (75 ml.) was cautiously added, the ether are: t , time in minutes; [HI +,acid concentration a t t min.; was decanted and the residue poured over a mixture of 500 x , amount of acetic aeid formed and a the ester concentraml. of 6 N HzSO4 and ice. This was continuously extracted tion a t time zero. The rate constant was determined by with the decanted ether. The extract gave 34 g. of wet taking 2.303 times the slope of log ( a x ) versus t plot, a CHFzCHzOH (about 69% yield); it is estimated that about procedure which is regardedL8as more accurate than the 10% of product was accidentally lost during extraction. averaging of constants determined from the equation K = CFaCOCHs (improved).-The condensation of CFaCOsEt [ l / t ]2.303 log [al(a.- x ) ] ; however, when the rates were (780 g. or 5.5 moles) with CHsC02Et (526 g. 0';6 moles) was calculated from this equation (Table 11, last column) the performed as usual.% After 6 hours of reflux instead of the results were practically identical. 60 hours used before, ether and ethanol were removed by heating on a steam-bath. Complete removal of ethanol was TABLE I1 then insured by heating under reduced pressure until a solid ACID-CATALYZED IlYDROLYSES A r 25.00 f 0.05' cake was formed. Sulfuric acid of 25% concentration log log a/ lO4K (1500 ml.) was added and allowed t o stand until the cake CIIPICH~OAC 1, min. [HI+ x ( a - x ) (a x ) (a x) m h - 1 was dissolved. Concentrated sulfuric acid (400 g.) was then 0 0.0960 ....% 0,1263 0.1014 .... added t o bring the solution back to 25% acid concentration. 245 .1226 .0998 0.0037 .OS85 0.0129 1.21 The mixture was heated under reflux with continued stirring .lo70 .0110 .1163 725 .Of318 .0396 1.25 until the organic layer had completely disappeared. The .1144 1295 .0331 .OW33 1.22 .0184 .lo79 vapors passing through the reflux condenser were directed .0317 ,0976 .1277 1700 .0099 .0915 1.24 to a Dry Ice trag which collected 511 g. (83%) of CFCOCHa, b.p. 20-21 Meall 1.23 CF3CHOHCH~.-Trifluoroacetone (40 g. or 0.36 mole) 0 0.03241 0.5106 .... 0.09815 .. was bubbled into a solution of LiAlH4 (4 g. or 0.1 mole) in 176 .09897 0.00082 .03158 ,4994 0.0112 1.46 dry ether (200 ml.) under a Dry-Ice reflux condenser. .09959 .00144 .03097 .4910 .0196 1.23 365 After two hours of refluxing, water (50 ml.) was added with 620 . i o n n ,00254 .03007 .4771 .0324 1.21 stirring, and additional ether was supplied when the mixture lwcame too viscous, Working up as before gave 34 g. of Mean 1.30 CIP,CIIOHCHa, b.p. 77', an 85% yield. Acetylation.-All alcohols were coiivciitiorially acctylCF3CI-I~OAc, CFIC€I(CH~)OACand CFICIIZCH~OA~ ; t t d with acetyl chloridc: CFaCHzOAc, b.p. 78", ZmD were subjected to similar measurements and computations. 1.3202; CHFeCHzOAc, b.p. 106-1015.5"; CFsC€IXH*OAc, Expressed in K X 10Esee.-' the experimental values, from b.p. 112.8-113', 7 t Z o ~1.3410, d204 1.219, A R F ld09; F% slope reading and from averaging, respectively, were found 34.5, calcd. 36.5; CFaCH(CHa)OAc, b.p. 85.6 ; CFsC- to be: CHFICH~OAC, 2.05 and 2.10; CF~CHIOAC, 1.45 and (CH3)20Ac, b.p. 94-95', n Z 61.3424, ~ d254 1.14, A R 1.05. 1.56; CF&H( CH*)OAc, 1.25 and 1.36, and CFsCHzCHzOAc, 2.6 and 3.3. (26) A. L.Henne, If.S. Newman, L. L. Quill and R. A. Staniforth, COLUMBUS, OHIO RECEIVED SEPTEMBER 14, 1951 THIYJOURNAL, 69, 1819 (1947). The reaction product could be steamed out, but it was found simpler to saw the vessel open and to scrape the bright rcd crystalline mixture of ammonium fluoride and fluorinated triazine (100 g. or 100% yield): 3C2Ftf9NHs -+
+
-
-
-
-
-
..
-
.....
-
