Infrared Studies of Amine Complexes. II. Formation Constants and

Infrared Studies of Amine Complexes. II. Formation Constants and Thermodynamic Properties of Amine-Chloroform Complexes1. Kermit B. Whetsel, and J. H...
0 downloads 0 Views 706KB Size
KERMITB. WHETSELAND J. H. LADY

1010

Infrared Studies of Amine Complexes. 11. Formation Constants and Thermodynamic Properties of Amine-Chloroform Complexes'

by Kermit B. Whetsel and J. H. Lady Tennessee Eastman Company, Division of Eastman Kodak Company, Kingsport, Tennevsee (Received July 19, 1963)

The association of chloroform with primary and secondary amines in cyclohexane solution was studied by measuring the effects of solvent composition and temperature on the nearinfrared N-H bands of the amines. Formation constants, K C , were determined at 35' for the 1: 1 complexes of chloroform with the following amines: aniline, p-phenetidine, p-chloroaniline, m-chloroaniline, o-chloroaniline, p-aminoacetophenone, X-methylaniline, K-ethylaniline, and cyclohexylamine. With the primary aromatic amines, a linear relation exists between log Kc and pK, (HZO). Thermodynamic constants for the aniline and cyclohexylamine complexes were calculated from equilibrium data obtained at temperatures bekween 10 and 50'. The heats of formation of these complexes are -1.7 3t 0.2 and -3.6 =k 0.3 kcal. mole-', respectively. The results are related to those obtained previously in a general study of the effects of chloroform on the near-infrared N-H bands of primary aromatic amines.

The effects of chloroform on the N-H stretching bands of aromatic amines have been studied in both the fundamental and the first overtone regions of the spectrum. 2 - 4 Although the observed solvent effects have been interpreted in terms of hydrogen bonding of chloroform with the amino group, no values have been reported for the formation constants and heats of formation of the complexes. In the present paper, formation constants are reported for the chloroform complexes of aniline, five ring-substituted anilines, N-methylaniline, N-ethylaniline, and cyclohexylamine. The thermodynamic properties of the aniline and cyclohexylainine complexes are also given. The results were obtained by studying the first overtone and combination N-H bands of the amines in the cyclohexane-chloroform solvent system.

tions of p-phenetidine containing chloroform developed a haze upon being exposed to the undispersed radiation of the source. The instrument operating conditions described previously were also used, except that the spectral slit width was kept a t about 3 cm.-l. Fivecentimeter cells were used in the overtone region. Because of the relatively strong absorption of cyclohexane, 1-cm. cells were used in the combination region. Ch,emicals. The amines were purified by recrystallization and/or distillation through a 1.4-m. spinning band column. Center cuts from the distillations were analyzed by gas chromatography, usually on two columns of different polarity, and judged to be more than 99% pure. Reagent grade chloroform was passed through a chromatographic column packed with partially deactivated alumina to remove the ethanol used

Experimental Equipment. The spectrophotometer and the ternperature-regulating device described previously6 were used with one modification. Corning y o . 7-56 filters were used between the source and the cells in order to remove the bulk of the ultraviolet and visible radiation

T h e Journal of Physical Chemistry

(1) Presented at the 14th Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, Pittsburgh, Pa., March, 1963. (2) A. '2. Moritz, Spectrochim. Acta, 17, 365 (1961). (3) K. B. Whetsel, ibid., 614 (1961). (4) F. H. Lohman and W. E. Norteman, Jr., Anal. Chem., 35, 707 (,Qfi2)

FORMATIOX COXSTAN'TS OF AMINE-CHLOROFORM COMPLEXES

as a stabilizer. The eluent was dried over calcium chloride and distilled. The center cut, b.p. 60.5", was collected and used within a few days. Chloroform prepared in this manner gave amine solutions that remained clear rand showed constant absorption for a t least 1 week. The cyclohexane was Eastman Kodak Spectro grade. Sample Preparation. The aromatic amines were studied at concentrations between 0.07 and 0.15 M , except that more dilute solutions of acetophenone were required because of the limited solubility of the amine. The concentration of cyclohexylamine was in the range 0.25 to 0.35 M . All concentrations were corrected for the variation of solvent density with temperature. Reference solutionis were prepared by mixing the appropriate volumes of chloroform and cyclohexane with a volume of carbon tetrachloride approximately equal to the volume of the amine in the sample solution. At temperatures other than ambient, differential spectra of the reference solutions were measured to obtain correction factors for the solvent unbalance which resulted from the temperature differential between the two cells. Formation Constants. Preliminary work showed that, self-association of the amines could not be neglected without introducing significant errors into the determination of the complex formation constants. A method described by Grunwald and Coburn6 for treating systems of this type was modified so that most of the calculations could be performed on an electronic corn-. puter. If dimerization of the amine and formation of the chloroform complex are the only significant inter-, actions occurring in the mixtures, the systems can be described by the equations

where K D and Kc are the formation constants of the amine dimer and the amine-chloroform complexes, respectively; CS, CM,CD, and CC are the equilibrium concentrations in moles per liter of chloroform, amine monomer, amine dimer, and complex, respectively. If C A o and CSO represent the total initial concentrations of amine and chloroform, respectively

and

1011

Substituting eq. 1, 3, and 4 into eq. 2 and rearranging gives 2KCKDcM3

+ (2KD f

[KC(CSo-

+

KC)cM2 cAo)

f 1]cM -

cAo

=

0

(5)

If Beer's law holds for the individual species AI1

=

+

+

EMCM EDCD ECCC

(6)

where A is the measured absorbance, l is the path length in centimeters, and E M , ED, and EC are the molar absorptivities of amine monomer, dimer, and complex, respectively. Substituting for CD and CC in eq. 6 and rearranging gives EC =

AI2 - E D K D C M - ~EMCM (7) (c,o K ~ [ ~ K ~ c ~ ~- ~cAo)chl1

+

+

The values of EM, ED, and K D required for the solution of eq. 5 and 7 were obtained from separate studies of the amines in cyclohexane solution. The methods described in the preceding paper5 under the discussion of the limiting slope treatment were used to obtain all of the constants except ED and K D for cyclohexylamine. Since ED of the cyclohexylamine dimer a t 2006 mp appeared to be close to zero, dimerization constants were determined at this wave length and used to calculate values of ED a t 2030 inw Since the values of C M in this work were always much less than unity, the cubic term of eq. 5 could be neglected without altering the final results. An assumed value of Kc was substituted into the quadratic form of eq. 5 to calculate a set Qf values of CM for a series of solutions containing known total concentrations of amine and chloroform. These CM values and the experimentally determined values of A/2 were substituted into eq. 7 to obtain a corresponding set of values of EC.' This procedure was repeated several times using different assumed values of KC. The standard deviation and the average of each set of EC values were calculated and plotted against the assumed value of KC. The best estimate of Kc was obtained from the minimum of the standard deviation plot. The value of EC corresponding to this value of KC was obtained from the average EC plot. In order to determine the range of KC values to be substituted into eq. 5 , an approximate value of KC was determined by a graphical method in which selfassociation of the amine was neglected and the eqiiilibriuni concentration of chloroform was assumed to (6) E. Grunwald and W. C. Coburn, Jr., J . Am. Chem. SOC.,80, 1322 (1958). (7) An electronic computer was used t o obtain solutions of eq. 6

and 7.

Volume 68, Number 6

M a y , 1,964

KERMITB. WHETSELAND J. H. LADY

1012

be equal to the initial concentration. and 7 then reduce to

Equations 5

Table I : Aniline-Chloroform Complex in Cyclohexane Solution" CHCla, mole/l. at 26'

and (9)

which, after combining and rearranging, can be written as

If A/ZCAO is plotted against ( E M - A/ZCAo)/CSot approximate values of KC and EC can be obtained from the reciprocal of the slope and the intercept, respectively.

Results Aromatic Amines. The spectra of 0.107 144 solutions of aniline in the cyclohexane-chloroform solvent system a t 30.5" are shown in Fig. 1. As the concentration of chloroform increases, the symmetric S-H band a t 1493.5 mp is replaced by a weaker and broader one with a maximum extending from about 1494 to 1501 mp. At lower temperatures, a distinct maximum occurs a t 1501 mp. Similar changes occur in the region of the asymmetric band near 14.50 mp. The absorbance changes a t 1493.5 mp were used to determine the formation constant of the aniline-chloroform complex. Data and results obtained a t several temperatures between 10 and 50" are summarized in Table I. In Fig. 2 the standard deviation of ec is plotted against Kc for the aniline-chloroform system a t three temperatures. Values of KC determined from the positions

The Journal of Physical Chemistrg

A/Z at 1493.5 mfib--------30.5" 41.6O

21.6O

49.5O

0 .494lC 0.743lC 1.2336c 1,7271 2.4673 3.4542 4.9346 6.9084 8.8823 12.216OC

0.1698 0.1688 0.1660 0.1650 0.1622 ,1634 ,1630 ,1594 ,1588 ,1568 ,1478 ,1482 ,1458 ,1452 ,1454 ,1374 ,1382 ,1388 .1386 ,1390 ,1260 ,1274 ,1270 ,1276 ,1272 ,1158 ,1172 ,1172 ,1184 ,1192 ,1050 ,1068 ,1056 ,1080 .lo80 ,0970 ,0976 ,0968 ,0988 ,0990 ,0912 ,0920 ,0914 ,0930 ,0930 ,0866 ,0874 ,0870 ,0882 ,0886

€M, l./mole-cm.

1.940 0.80 0.44 1,020

1.905 0.80 0.40 1,006

0.80 0.36 0.994

1.848 0.80 0.33 0.981

1,826 0.80 0.31 0.971

0.579 0.623 0.590

0.526 0.500 0.632 . 0,630 0.522 0.500

0.448 0.640 0.438

0.400 0.625 0.404

ED, l./mole-cm.

KD, . l./mole . Density correction

Kc, l./mole l./mole-cm. Kc (cor.),d l./mole EC,

1.880

a Aniline concentration 0.1074 M a t 26'. b A = Al4Qa.am p AlelOm p ; 1 = 5.00 cm. c These samples not used in calculation Assuming CC is 0.630 l./mole-cm. a t all temperaof Kc and ec. tures.

of the minima in these plots are summarized in Table I. Since the results obtained with solutions containing low concentrations of chloroform are quite sensitive to experimental error and to small errors in EM, only the data from solutions containing between about 1.7 and 9 M chloroform were used to determine Kc. Within the limits of experimental error, the values obtained in this manner were consistent with the absorbance data obtained with samples containing as little as 0.4 M or as much as 12.1 M chloroform.

0.03

Figure 1. First overtone N-H stretching bands of aniline in cyclohexane-chloroform solutions a t 30.5" (5-cm. cells).

loo

r

Figure 2. Formation constants of the aniline-chloroform complex.

FORMATION CONSTANTS OF AMINE-CHLOROFORM COMF~LEXES

The values of EC corresponding to the best estimaters of K c ranged from 0.623 to 0.640 1. mole-1 cm.-l. Since no systematic variation of EC with temperature was evident, the average value of 0.630 was used to determine a corrected KC a t each temperature from the EC us. assumed Kc plots. The corrected values of KC differed from the original ones by no more than j=O.O11 1. mole-' (Table I). A plot of log Kc us. 1 / T for the aniline-chloroform complex gave a AHo value of -1.72 f 0.12 kcal. mole-l. The comparable value obtained with the uncorrected values of K,c was - 1.64 rfr 0.16 kcal. mole-'. These error limits represent one standard deviation of the slope of the line. In a following section it will be shown that errors in EM, ED, and K D introduce LL further uncertainty of about =kO.10 kcal. mole-' into the determination of AH O. When all sources of error are considered, a realistic value for AHo is -1.7 0.2 kcal. mole-l. Formation constants of several other aromatic amine-chloroform complexes a t 35 O are shown in Table 11. The value reported for the p-aniinoacetophenone complex is an approximate one determined by the graphical method. The value of EM for p aminoacetophenone is subject to considerable error, because the solubility of the amine in cyclohexane its quite low. Since any error in E M is reflected in Kc, use of the more refined method was not justified.

Table 11: Aromatic Amine-Chloroform Complexes in Cyclohexane Solution a t 35"

1013

Wavenumblr, cm-'

66W

67W

610

6500

610

4950

5050

1.0 0.5

P

f

0.2

1 0.1

0 Warslanglh, m / r

Figure 3. Near-infrared N-H bands of cyclohexylamine in cyclohexane-chloroform solutions a t 30".

In the overtone region the solvent effects are rather small, but in the conibination region the addition of chloroform to a cyclohexane solution of the amine results in the progressive disappearance of the band a t 2006 mp and the appearance of a well-resolved new band a t 2030 mp. The conibination N-H bands of aromatic amines do not show this pronounced solvent e f f e ~ tbut , ~ similar results have been reported by Lohman and Norteman4 for various aliphatic primary amines. Formation constants of the cyclohexylaniine-chloroform complex calculated from the absorbance changes observed a t 2006 and 2030 nip are summarized in Table 111. The constants calculated a t 2006 nip are consistently higher than those obtained a t 2030 nip. The error analysis which follows indicates that this difference represents bias resulting from errors in E N , E D , and K D . The values obtained a t 2030 nip are less sensitive to these errors and are thus the preferred ones.

fCI

Amine

Aniline p-Phenetidine p-C hloroaniline m-Chloroaniline o-Chloroaniline p-Arninoacetophenone' N-Methylaniline N-Ethylaniline

A, mr

l./mole-

KC,

PK a (Hz0)

om.

l./mole

1493 5 1499 5 1492 0 1490 5 1491 5 1485 0

0.63 .76 .64 .59 .43 .92

0.47 .65 .33 .25 .I6 .14

4.58" 5 25"

1485 0 1495 0

.47 .38

25 23

4.85d 5.1Id

Table 111: Cyclohexylamine-Chloroform Complex in Cyclohexane Solution

4.05b 3.5Zb 2.71* 2.1gb

a N. F. Hall and M. R. Sprinkle, J. Am. Chem. SOC.,54, 3469 (1932). J. M. Vandenbelt, C. Heinrich, and S. G. Vanden Berg, Anal. Chem., 25, '726 (1954). Data for this compound treated See ref. by graphical method using approximate value of E M . 12.

Cyclohexylamine. The first overtone and combination N-H bands of cyclohexylamine in the cyclohexane-chloroform solvent system are shown in Fig. 3.

7 -

--

2006 mp---

2030 mr---

fC!

fC>

KC,

KC,

OC.

l./moleom.

l./mole

1. i m oleom.

1. /mole

12.0 17.5 28.0 30.0 38.0 42.5 51.5

0.07 .06 .06 .06 .05 .06 .06

1.74 1.50 1.22 1.08 0.94 0.89 0.77

1.71 1.65 1.62 1.59 1.57 1.51 1.49

1.41 1.31 1.04 1.03 0.84 0.84 0.64

lemp.,

Plots of log K c us. 1/T for the cyclohexylamine complex gave AHovalues of -3.9 f 0.2 and -3.6 f 0.2 kcal. mole-' a t 2006 and 2030 inp, respectively. These values are subject to additional uncertainties of apVolume 68,Xumber 5

M a y , 1964

KERMITB. WHETSELAND J. H. LADY

1014

proximately f0.3 and f 0 . 1 kcal. mole-l, respectively, because of possible errors in E M , ED, and KD (see following section). Since the results a t 2030 mp are much less sensitive to these errors, the best value for the heat of formation of the complex is believed to be -3.6 f 0.3 kcal. mole-'. Analysis of Errors. The effects of experimental error and of any interactions other than those represented by eq. 1 and 2 are reflected by the precision of the complex constants. Precision limits for EC were obtained directly from the minimum in the plot of standard deviation of EC vs. assumed KC (see Fig. 2). The values varied from =t0.002 to 'f0.012 1. mole-l cm.-l. These values compare favorably with the estimated experimental error of 0.01 to 0.02 in the measurement of the apparent absorptivities. This condition must be met if the model being used for complex formation is a valid one.6 The precision of KC cannot be determined directly from plots such as those shown in Fig. 2. The values of KC corresponding to the minima of these plots can be determined exactly, but they still have errors associated with them that are related to the errors in EC. Precision limits for KC corresponding to the minimum standard deviation of EC were obtained from the plot of average EC vs. assumed KC. For the aromatic amine complexes the errors in KC were generally less than A0.01; for the cyclohexylamine complex the errors ranged from fO.01 to f0.05. These errors are consistent with the standard deviation of KC calculated from the linear AH O plots. Errors in E M , ED, and KD have relatively little effect upon the precision of EC, KC, and AH O, but they introduce bias affecting the accuracy of the results. Reasonable errors for E M , ED, and KD were estimated by considering the results of the self-association studies. The calculation of the complex constants were then repeated to obtain the results summarized in Table IV. Since the self-association of aniline has been studied extensively,6 the estimated errors for E M , ED, and KD were relatively small; their combined effect on the complex constants amounted to only f0.01 in ec, h0.02 in Kc, and =tO.lO in AHo. Somewhat larger errors were estimated for the cyclohexylamine system, since the self-association of this amine was not studied exhaustively. At 2006 nip the uncertainties in E M , ED, and K D correspond to possible errors of fO.01, hO.10, and f 0 . 3 in ec, KC, and AHo,respectively. At 2030 mp the corresponding errors are only 10.01, +0.05, and f O . l . Thus, the set of complex constants determined a t 2030 m p is believed to be the more reliable of the two. The treatment of the data could be refined by corThe Journal of Physical Chemistry

Table IV: Effect of Errors on Change

0.80 to 1.20b *lo% =t0.5%

EC,

KC, and A H o

Aec (30')

AKg (30')

A(AHo)a

Aniline-CHClt +0.005 +0.006

+o. 01

F0.005

F0.004 f0.003

fO.O1l

F O . 04 f0 . 0 3

Cyclohexylamine-CHCls a t 2006 mp 0.00 to +0.001 +0.007 0.10 +50% -0.003 -0.038 f2% f0.003 f0.062

f O .14

Cyclohexylamine-CHCla a t 2030 mp f10% f 0 . 002 F O .008 +50% +0.003 -0.004 *IO% f0.004 T0.034

TO.01 -0.06 F0.04

+0.04 -0.10

a Kegative values in this column indicate increases in the heat of formation. * Correspondingly larger values of K D were also used. See ref. 5 for discussion of the relation between e~ and KD.

recting for the effects of chloroform self-association. A dimerization constant of 0.013 1. mole-' has been reporteds for chloroform in cyclohexane solutiOn a t 25 O. Using this value to correct our 30.5 O data lowered the formation constant only 0.02 1. mole-' and the absorptivity about 0.05 1. mole-' cm.-'. Since dimerization constants a t other temperatures were not available, the correction was not applied to any of the results shown in Tables I through 111.

Discussion The formation constants a t 25' and the associated thermodynamic constants of the chloroform complexes of aniline and cyclohexylamine are summarized in Table V. Table V : Thermodynamic Data for Chloroform-Amine Complexes AF'za,

Complex

Chloroformaniline Chloroformcyclohexylamine5 5

KM, l./mole

kcal./ mole

0.51

+0.41

-1.7 f0.2

1.10

-0.06

-3.6f0.3

From data obtained a t 2030

AH', kcal./mole

AS'za;

e.u.

-7.1 f 0.7 11.9f1.0

rnp.

(8) C. F. Jumper, M.T. Emerson, and B. B. Howard, J . Chem. Phys., 35, 1911 (1961).

FORMATION CONSTANTS OF AMINE-CHLOROFORM COMPLEXES

The heats of formation of the aniline-chloroform complex and of the aniline dimer6 are the same within the limits of experimental error. The AH" value for the chloroform complex is also comparable to the value of -1.93 kcal. mole-' reported by Daviese for the aniline--water complex in carbon tetrachloride solution. In a nuclear magnetic resonance study of the chloroform-triethylamine complex in cyclohexane solution, Creswell and Allred'" found a formation constant OF 0.51 1. mole-' a t 25" and a AH" of --4.2 kcal. mole-'. A value of -4.2 kcal. mole-' has also been reported for the heat of formation of the diethylamine-chloroform complex in the gas phase.'l These AH" values are about 10% higher than the value now being reported for the cyclohexylamine complex. The formation constant of the triethylamine complex, however, is about one-half as large as that of the cyclohexylamine complex. This difference probably reflects steric hindrance by the alkyl groups of triethylamine to the approach of a chloroform molecule. In Fig. 4, log KCof the aromatic amine-chloroforni complexes is plotted against the pK, of the amines in aqueous solution. The primary amines follow the linear relation log Kc

=

0.247pK, - 1.474

(11)

This relation holds for o-chloroaniline as well as for thle meta- and para-substituted derivatives, indicating that steric effects are relatively unimportant in the formation of the o-chloroaniline complex. These results providc direct experimental confirmation for the interpretation of chloroform solvent effects presented p r e v i ~ u s l y . ~For example, much larger effects are observed with p-phenetidine than with p-aminoacetophenone, because the ratio of complexed amine to free amine in chloroform solution is about 8 : 1 for the first compound and less than 2 : l for the second. The N-alkylanilines do not follow the linear relatioin between log K C and pK. that characterizes the primary amines. Both N-methylaniline and N-ethylaniline are slightly rnore basic than aniline,I2 but the formation constants of their chloroform complexes are less than one-half as large as that of the aiiiline complex. The N-alkyl groups appear to offer substantially greater steric hindrance to formation of the chloroform complexes than to protonation of the amines. The steric effects of the two alkyl groups are quite siniilar in the formation of the chloroforin complexes, but in the dimerization of the amines the effect of the ethyl group must be considerably larger than that of the naethyl group. The dimerization

0.5

1015

1 0

0

N-EthylAniline

I 2

I

I pK.(H,O)

I

4

I

-

I 5

Figure 4. Relation between the formation constant of aromatic amine-chloroform complexes in cyclohexane solution and the pK, of the amines in aqueous solution.

constants for N-methyl- and N-ethylaniline are approximately 0.30 and 0.15 1. mole-', respectively. Kagarise recently reported a formation constant of 0.90 1. mole-l (31") and a AH" of -3.5 kcal. mole-l for the CDCl3-acetone complex in n-hexane solution. l a These values are approximately twice as large as the values now being reported for the CHC13-aniline coiiiplex and about equal to those found for the C H C k cyclohexylamine complex. Since the basicity of the amines in water is many times greater than that of acetone, it is evident that the linear relation expressed by eq. 11 cannot be used with widely different types of bases. Using nuclear magnetic resonance spectroscopy, Creswell and Allredlo found that the chlorofornibenzene complex in cyclohexane solution has a formation constant of 0.11 l. mole a t 23" and a AH" of - 1.97 f 0.35 kcal. mole-l. The free N-H bands of the aromatic amines show high frequency shifts of 1 to 2 mp in cyclohexane solutions containing from 10 to 40 vol. % ' chloroform. These shifts are believed to arise from the interaction of chloroform with the n-electrons of molecules which are not coniplexed through the amino group. With higher concentrations of chloroforni, the concentration of free amino groups becomes so low that the secondary solvent effect is no (9) M.Davies, Ann. Reit. Progr. Chem., 43, 5 (1946). (10) C. J. Creswell and A. L. Allred, J . Phys. Chem., 6 6 , 1469 (1962). (11) J. D. Lambert, J. 5. Clarke, J. F. Duke, C. L. Hicks, S. D. Lawrence, D. AI, Morris, and AI. G. T. Shane, Proc. Roy. SQC. (London), A249, 414 (1959). (12) H.C. Brown and A. Cahn, J . Am. Chem. Soc., 7 2 , 2939 (1950). (13) R. E.Kagarise, Spectrochim. Acta, 19, 629 (1963).

Volume 68, Number 6

May,196.4

G. 0. PRITCHARD AND J.

1016

longer observed. The n-electron interaction may persist after the amino group is complexed, but its effect upon the spectrum is not apparent, Since the Uelectron interaction results in a very small spectral change, it probably has little effect upon the constants

E(.

FOOTE

reported for the amino group interaction. These constants represent the average effects of hydrogen bonding of chloroform to the amino group, irrespective of whether the a-electrons are involved in a secondary interaction.

The Reactions of C,F, and C,F, Radicals with Hydrogen and Deuterium'

by G. 0. Pritchard and J. K. Foote2 Department of Chemistry, University of California, Santa Barbara, Goleta, California (Receiued August 26, 2963)

By use of the perfluoroaldehydes as the sources of CZF5 and C3F7 radicals, the hydrogen abstraction reactions of these radicals with H2 and D2 are reinvestigated. The previous anomalously high values for the activation energies of these reactions are confirmed. New data for the reaction C Z F ~ D2 --.+ CzFsD D are presented.

+

Introduction In a previous article3 the photolysis mechanisms of perfluoroalkylaldehydes were discussed, especially with regard to the H abstraction reactions

Rr

+ RrCHO

--+

RfH

+ RfCO

(1)

relative to the respective radical recombination reactions

Rr

+ Rr

--+

Rf,

+ H2

Rt

+ Dz

-+

RrH

+H

(3)

+

RrD

+D

(4)

and Reactions 3 and 4 have been quite extensively studied using perfluoroalkyl ketones as the radical sources: CF3 with H2 and D2,4CzFL with H2,5 C3F7 with H2 and DZ,6and C3F7 with Dz.7 The reactions of CF3 radicals with Hz and D2 have also been investiThe Soozrrnal of Physical Ch.emistry

gated, using hexafluoroazomethane as the radical source Calvertg has pointed out that the activation energies obtained for the reactions of C3F7 with H2 and D2'jv7 (and subsequently CzFs Hz5)are anomalously high, when compared to the general pattern of alkyl and perfluoroalkyl H abstraction reactions. I t therefore seemed desirable to redetermine values of E3and E4 for Rt = C2F5 and C3F7, using a different photolytic

+

(2)

where Rf = CF3, CzFs, or C3F7. In this work, using C2F&H0 and C3F7CH0 as the radical sources, we have investigated the abstraction reactions of C2F5 and C3F7 radicals with H2and with Dz Rr

+

(1) This work was supported by a grant from the National Science Foundation and is based in part on a thesis submitted by J. K. F. in partial fulfillment of the requirements for the M.A. degree. (2) Department of Chemistry, University of California, Riverside, California. (3) G. 0. Pritchard, G. H. Miller, and J. K. Foote, Can. J . Chem., 40, 1830 (1962). (4) P. B. Ayscough and J. C . Polanyi, Trans. Faraday Soc., 52, 960 (1956). (5) 9. J. W. Price and K. 0. Kutschke, Can. J . Chem., 38, 2128 (1960). (6) G. H. Miller and E. W. R. Steacie, b. Am. Chem. Soc., 80, 6486 (1958). (7) G. Giacometti and E. W. R. Steacie, Can. J . Chem., 36, 1493 (1958). (8) G. 0. Pritchard, H. 0. Pritchard, H. I. Schiff, and A. F. Trotman-Dickenson, Trans. Faraday SOC.,52, 849 (1956). (9) J. G. Calvert, Ann. Rev. P h w . Chem., 11, 41 (1960).