Nov., 1962
HEATSOF NEUTRALIZATION AND RELATIVE STRENGTHS OF AMINESIN BENZENE
2149
with certainty, has a minimum value of 1.0 and probably does not exceed 1.4. The average molecular weight of polypropylene chain between crystallites therefore is about 250 to 350, or approximately 6 to 8 monomer units. M , of similar magnitude was found for PE.Q (ARl)Borp = 210Ovz2 cal./mole With hexane and heptane as solvents,. plots of for benzene-isotactic PP (13) (XI h 4- eh/RT) us. 1/T give straight lines with Similar value can be calculated for the ben- slopes of 7.0 f 0.2 X lo2. The slopes of the zene-PE system from Rogers’ data. Subtraciion of straight lines correspond to eh/R, from which Eh can the ( A H I ) ~term , ~ from (AH1)eorpyields (AH1)eI of be calculated to be 1400 cal./mole, in agreement 1490 f 100 cal./mole a t v2 = 1.0. The agreement with eq. 12. The intercepts of these straight lines with the ordinate gives (xl A), from which X can of this value in eq. 12 is satisfying. Typical plots of the vapor sorption data according be calculated. Table I summarizes the results of these calculations. to eq. 10 yield straight lines. The quantities (xl The quantities eh/RT and X are comparable in X q,/RT) and M o / ( ~ ) 0obtained 2 from these magnitude but opposite in sign. The deviation of straight lines are listed in Table I. The quantity M , / ( C Y )remains ~~ constant within both the energy and the entropy contributions from the temperature range of this study. Moreover, the conventional expressions of polymer solutions and value obtained with aliphatic hydrocarbon solvents rubber elasticity appear to be important considerais in excellent agreement with that obtained with tions in dealing with the sorption of vapors by benzene. The presumption that crystallite “melt- polycrystalline materials. It is of interest t o out” does not play a role in our experiments there- speculate that, a t -5”, eh/RT and X cancel each X E ~ / R Tis)identical with XI. fore is substantiated. The term CY)^^, not known other and (xl The sorption of hexane by the isotactic PP film at (15) “International Critioal Tables,” MoGraw-Hill Book Co., New -5” may then become “pseudo-ideal.” York, N. Y., Vol. 5 , p. 151.
heptane or octane, reported in the l i t e r a t ~ r e ’to~ be 690 cal./mole. The partial molar enthalpy of sorption of benzene by isotactic PP, containing both the elastic contribution (AHI),~and the (AH1)mix term, was found to be
+
(Anl),,,,.,
+
+
+
+ +
HEATS OF NEUTRALIZATION AND RELATIVE STRENGTHS OF AMINES IN BENZENE BYTHOMAS E. MEAD Central Research Division, American Cyanamid Company, Stamford, Connecticut Receined April 6 , 198.9
Heats of neutralization of amines have been obtained in benzene at 28’ from thermometric titrations with trichloroacetic acid, A plot of these AH values us. base strength in water for aliphatic amines shows three parallel lines corresponding to I, 2, and 3” amines. This splitting has been ascribed to differeyes in solvation by benzene and water of both amines and This order is believed due to “F strain” between amine reference acid. Base strength order in benzene was 1’ > 2’ > 3 and undissociated acid and to solvation of the protonated amines by the benzene x-electrons. Primary and secondary aromatic amines showed a weakening of base strength in benzene relative to the aliphatic and tertiary aromatic amines. Heats of reaction were estimated for 1:2 and 1:3 molar amine: acid associations.
.
Introduction The dissociation constants of amines in water are used commonly as a measure of base strength. It is well known that discrepancies occur in comparing base strength values (dissociation constants) with molecular structure of amines. Solvation of an amine moleculle or its conjugate acid with water molecules limits direct comparison of dissociation constants of primary, secondary, and tertiary amine~.~A J non-solvating solvent would appear the best choice for obtaining these base strength values. Forman and Humee have correlated Taft Zcr* values with heats of neutralization (AH,) of aliphatic primary and straight chain secondary amines in acetonitrile. They found AH, for (1) R. P. Bell and J. W. Bayles, J . Chem. Soc., 1518 (1952).
(2) A. F. Trotman-Diokenson, ibid., 1293 (1949). (3) H. K. Hall, Jr., J. Phys. Chem., 60, 63 (1956). (4) H. K. Hall, Jr., J . A m . Chem. Soc., 79, 5441 (1957). (5) M. M. Davis and H. B. Hetaer, J. Res. Natl. Bur. Std., 60, 569 (1958), RP 2871. (6) E. J. Forman and D. N. Hume, J . Phus. Chem., 63, 1949 (1959).
rneta- and para-substituted aromatic amines correlated with pKb(H20) values, but “no obvious correlation” of AHn with pKb(HZO) values for non-aromatic amines was obtained. Deviation of branched chain secondary and tertiary aliphatic amines was explained by the B strain hypothesis. Limited solvation of amines and ammonium ions in acetonitrile explained these results. Preliminary thermometric titrations of amines in benzene with trichloroacetic acid by the author indicated a greater spread of AHn values than those obtained by Forman and Hume in acetonitrile. Solvation of amine molecules and cations should be less in benzene. This study concerns the heats of neutralization of amines in benzene with a reference acid, and the correlation of these reaction heats with base strength values in various solvents. Experimental Materials.-Reagent grade benzene and trichloroacetic acid from Baker and Adamson Go. were used throughout.
THOMAS E. MEAD
2150
Both were employed without further purification. A Karl Fischer titration indicated 0.03% water in the benzene. Titrant solutions consisted of trichloroacetic acid in benzene of approximately 1.0 M with a water content of 0.06% by Karl Fischer titration. Titrant solutions were standardized daily by diluting an aliquot with acetone and titrating potentiometrically with 0.1 N aqueous sodium hydroxide. Solutions of amines in benzene were 0.02 M , and were placed in ground glass-stoppered Erlenmeyer flasks prior to thermometric titration. Most of the amines were Eastman Kodak White Label grade. Eastman Practical grade diallylamine was distilled and analyzed 99+ % by gasliquid chromatography. Mass spectroscopy found no apparent impurities. Eastman White Label grade N,Ndiethylaniline was distilled and analyzed 99 % pure by gas-liquid chromatography. Triethylenediamine, tetramethylguanidine, and bis-( cyanomethy1)-amine were prepared and purified in our Laboratories. Anhydrous grade tetramethylethylenediamine was obtained from Rohm and Haas Co. Apparatus.-The automatic thermometric titration apparatus consisted of the following. A Sargent Model C motor-driven 10-ml. capacity buret delivered titrant a t a constant rate of 1 ml./min. to the adiabatic titration cell, a dewar flask of approximately 175-ml. capacit . Situated on top of the neck of the flask is a three-holed $at cork disk which admits the capillary buret delivery tube, a glass stirrer powered by a 500 r.p.m. synchronous motor (Bodine Electric Co.), and a temperature detector. A reservoir (approximately 7-ml. capacity) placed in the titrant delivery tube inside the dewar flask allowed for initial temperature equilibrium between total titrant to be delivered and solution to be titrated. The buret tip was fitted with thin-wall polyethylene tubing drawn out to a capillary to minimize heat transfer and decrease diffusion between titrant and titrated solution. A Veco No. 32824 bead thermistor available from Victory Engineering Corp. and having a resistance of 2000 ohms f 10% at 25' was the temperature detector. Thermistors were calibrated over an approximate temperature range of 4". A maximum temperature sensitivity of approximately could be obtained on a 0-1 mv. Speedomax recording potentiometer with the thermistor connected as one leg of an a.c. rectified Wheatstone bridge. A constant temperature box enclosed all apparatus except the recorder. Temperature regulation to approximately 0.1' was obtained with a blower, 200-w. heater, and thermistor electronic controller (Model 63 from Yellow Springs Instrument Co.). A flexible hose connected at diagonal corners outside the box provided for continuous recycling of the temperature-controlled air. All thermometric titrations were performed in the constant temperature box a t 28.0'. All infrared studies were performed with a Beckman IR-4 infrared spectrophotometer. Technique .-The initial slope method of Keily and Hume7 was used to determine heats of reaction from the expression
+
Vol. 66
where c, d, and V are the specific heat, density, and volume, respectively, of the solution. C,' then was calculated from eq. 3
Cpsolnfor benzene was calculated from eq. 2 using literature specific heat d a h 8 The sum of C,' (12.18 cal./deg.) and C, (17.89 cal./deg.) for benzene was the total heat capacity value for the amine solutions in benzene (30.07 cal./deg.), and was used for all calculations of A H . The specific heat of the 0.02 M amine in benzene solutions was assumed to be equal to the specific heat of pure benzene. The average C, value obtained from aqueous sodium hydroxide-hydrochloric acid thermometric titrations agreed within experimental error with an average C, value obtained by the method of Keily and Hume,'i.e., addition of heat electrically t o known amounts of water. The general operating procedure consisted in adjusting the temperature of the solution to be titrated as close to the titrant temperature (28.0') as possible. This was accomplished by placing the benzene solution of the amine in an Erlenmeyer flask and adjusting to the desired temperature by heating the flask under a stream of warm water. Next a 50-ml. aliquot was pipetted into the dewar flask (pipet and flask set at 28') and the solution was titrated thermometrically. The volume of titrant added to effect a 1:1molar reaction was calculated by measuring the calibrated recorder chart length along the time (volume) axis from the start of the titration (point B on Fig. 1)to the projected end point F. Figure 1 illustrates the calculation of the initial slope (d T / dV). The product of the length RC and the appropriate A T per chart deflection was dl'. The corresponding differential volume change, dV, was the chart length (CD) times the appropriate calibration factor. Thus the heat change for the 1:l molar reaction ( A H t ) was determined by substituting the above calculated values in eq. 1. The heat of reaction (AHz) for the 1:2 amine:acid association was estimated by calculating slope EG (Fig. 1) and substituting this value in eq. 1 along with the values of C, and M used in calculating AH,. This calculated value of AHz was only approximate, for the initial temperature for the second association was greater than the titrant temperature. However, this should not introduce an error greater than about 2%. The pooled standard deviation of all the thermometric titration determinations of 0.02 iM amines in benzene (124 trials with 34 amines) was 0.37 kcal./mole. Effect of Concentration on AH.-To determine the effects of concentration on A H values, benzyl-, isobutyl-, and tributylamine were treated with 0.5 and 1.5 M titrant concentrations. Amine concentrations were 0.004 and 0.04 iV! (or 0.02 M ) , respectively. No change in AH was noted at these concentrations.
Results Experimental heats of neutralization obtained AH = from thermometric titration of amines with triwhere C, in cal./deg. is the total heat capacity of the dewar chloroacetic acid in benzene are listed in Table I along with corresponding base strength values in flask and its contents and of the liquid to be titrated. M is the concentration of titrant in moles/l., d T is the tempera- various solvents. ture change in 'C., and d V is the corresponding volume Aliphatic Amines.-Sharp 1:1 molar end point change in ml. The dT/dV change is assumed equal to breaks were obtained for all aliphatic amines (conAT/AV. The total heat capacity (C,) for 5O-.ml.. volumes was de- taining no other heteroatoms) with the exception termined by performing a thermometric titration of known of sterically hindered tribenzylamine. QuantitaAH, and calculating C, from eq. 1. Fifty-ml. volumes of tive titrations are possible at 0.001 M amine conaqueous 0.02 M hydrochloric acid were titrated thermocentration. The 1:2 and 1: 3 amine:acid associametrically with 1.0 M sodium hydroxide (AH = -13.5 kcal./mole). The total heat capacity consists of the heat tion end points were rounded and semiquantitacapacity of the solution to be titrated (Cpsoln),and the heat tive. A representative thermometric titration capacity of the dewar flask and contents (C,'). Using curve is presented in Fig. 1. Primary and secondspecific heat data listed in the International Critical Tables,8 ary aliphatic amines had lower 1 : 2 and 1: 3 molar C, Boln for the known reaction was calculated as heats of association than shown in Fig. 1. (See Table I.) A plot of AH values obtained for aliphatic amines (7) H. J. Keily and D. N. Hume, A n d . Chem., 28, 1294 (1956). in benzene under the conditions described vs. (8) "International Critical Tables," Vol. 5 , MoGraw-Hill Book Co., p K b ( H 2 0 ) values obtained from the literature Inc.. New York. N. Y . , 1929, p. 115.
(-)
C, dT ilt dV
HEATSOF NEUTRALIZATIOR' AND RELATIVE STRENGTHS OF AMINESIN BENZENE 2151
Nov., 1962
TABLE I CORRELATION5 O F BASESTRENGTHS, Std. A H i , dev., kca1.i kcal./ mole mole
NO.
-1"
Amines-
1 n-Butylamine 2 sec- Butylamine 3 &Butylamine 4 Allylamine 5 Benzylamine 6 p-Anisidine 7 Aniline 8 m-Nitroaniline 9 m-Bromoaniline 10 p-Chloroaniline
29.3 29.6 29.1 27.4 26.6 18.8 16.4 0.3 6.2 9.8
0.0 .6 .3 .2 .1 .3 .2
29.4 33.3 28.5 27.2 24.5 28.3 25.0 22.5 13.8 23.8 31.2 16.5 29.8 12.7 0.8
.1 .4 .3 .4 .2 .3 .5 .2 .I 2 .2 .2 .3 .4
.2 .3
2 U * , AND
HEATSO F REACTION O F AMINES
No.
- AHz,
of detns
koal./ mole
3 8
0.7 .3 .7 .4 .6
3.36 20 3.44 6 3.58 6 4.51 4 4.66 4 8.71 6 9.42 6 11.54 21 10.49 3 10.00 6
1.2 1.9
2.95 6 2.75 4 2.99 6 3.50 4 4.71 4 2.88 20 5.64 4 5.57 9 9.15 6 4.00 6 -0.06 9 8.85 9 3.00 4 8.70 22 13.8 23
4 4 4 3 3 1 2 3
.o
pKb
Lit.
(HzO)
ref.
AHNP in nitromethane (mv.)
3 42 43 72 105 375 445b 597b 529 48'7
AHNP in ethyl acetate (mv.)
-33 -35 -21 11 21 267 338 503 443 416
Z&
Lit. ref.
0.79 .77 .79 1.11 1.20
4 4 4 4 4
0.11 .23 .08 .23 .75 0.35 1.16
4 4 4 4 4 4
0.25
4
2' Amines
11 12 13 14 15 16 17 18 19 20 21 22 23 24 34
Di-i-propylamine Di-n-butylamine Di-sec-butylamine Di-i-butyla,mine Diallylamine Piperidine Morpholine Dibenzylarnine N-Methylaniline 1,3-Diphenylguanidine Tetramethylguanidine N-Ethylaniiline Di-n-propylamine Isoquinoline Bis-.(cyanomethyl)-amine
4 3 8 4 4 5 4 4 4 4 4 3 4 4 1
1.3
2.9 1.8 2.3 2.3 2.6 1.9 3.3
2.0
- 19 -24 - 19 4 87 -47 79 154 434 0 (std.) - 297 411 - 19 299 386
-33 -31 0 21 61 - 67 23 108
4
0 (std.)
- 186
338 -40 289 597
3" Amines
4 42 23.0 3.11 6 -36 -0.39 6 .5 4.9 Tri-n-butylamine 2 612 13.5 7.10 9 274 .3 Tribenaylamine 4 82 19.8 5.07 4 .2 4.8 78 0.22 4 Dimethylbenzylamine 4 11.3 8.94 6 348 374 .6 24 .5 N,N-Dimethylaniline 15.3 280 286 .41 24 3 N,N-Diethylaniline 7.48 6 .4 3 13.6 307 286 .4 8.81 6 Pyridine -34 5.32 9 3 21.6 Triethylenediamine .2 Tetramethylethylene22.2 3 1 -65 diamine .I 4.90 9 2 11 27.8 49 33 Triethanolamine 6.24 20 .8 3 343 35 p-Chloro-K,N-diethylaniline 12.7 .6 333 8.25 9 5 - AH, is the heat of reaction for a 1:1 trichloroacetic acid: amine reaction in benzene. - AHz is the heat of association for a 2 : l trichloroacetic acid:amine association in benzene. AHNP values in nitromethane and ethyl acetate were obtained from Dr. C. A. Streuli, with the following exceptions. AHNP values in nitromethane for aniline and mnitroaniline were obtained from ref. 25. 25 26 27 28 29 30 31 32
gave three parallel lines corresponding to 1, 2, and 3 O amines (Fig. 2). The lines shown in Fig. 2 represent least, squares solutions of the appropriate data. T o our knowledge, no such correlation has been noted in the literature for other non-aqueous solvents in which potentiometric data were obtained. If these same AH values were plotted vs. half-neutraliza tion potential values (AHNP) obtained for these amines in ethyl acetate (Table 11)) a similar but less obvious splitting was observed. AHNP values were obtained from Streulig and are a potentiometric measure of base strength in nonaqueous solvent^.^ The separation of a given group of compounds on the basis of 1, 2, and 3' amine as illustrated in the plot of AH in benzene us. plib(H,O) appears to be due to differences in (9) C. A. Streuli, private communioation.
solvation of the amines in the two solvents. Less splitting was observed in the plot of AH in benzene V S . AHNP in ethyl acetate because of the greater similarity in solvating power of these two solvents. These results substantiate the importance of solute-solvent interaction in comparing relative base strengths. Aniline and 1 and 2' aryl amines with alkyl substitution on the nitrogen were included with the respective 1 and 2' aliphatic amines in the calculation of the least squares line, as little change in slope or standard deviation df the lines was observed when these data were included or excluded (Table 11). Di-n-butylamine was excluded from all least squares calculations, as the experimental value of AH exceeded the 99% confidence limits for line numbers 6 and 10, and exceeded the 95%
215'2
THOMAS E. MEAD
Vol. 66
TABLE I1 Eq. no.
Type of amine
Std. dev. of line (koal./moIe)
Least squares lines
9 10 11 12
BENZENE US. PKb (HpO) 1' Aliphatic - A H = -2.05pKb 36.40 1' Aliphatic and aniline -AH = 2.17pKb f 36.89 2' Aliphatic, N-methylaniline, and N-ethylaniline -AH = -2.32pKb f 35.84 3" Aliphatic and aromatic -AH = -2.04pKb f 29.52 1' and 2" Aromatic (including substituted) -AH = -6.44pKb 74.42 LEASTSQUARES LINESFOR -AH IN BENZENE us. AHNP (ETHYL ACETATE) 1' Aliphatic and aniline LAH = -0.0367 AHNP 28.12 2" Aliphatic, N-methylaniline, and N-ethylaniline -AH = -0.0383 AHNP f 27.95 3" Aliphatic and aromatic 23.09 -AH = -0.0331 AHNP 1O and 2" Aromatic (including substituted) -AH = -0.0829 AHNP 43.31
0.57 0.96 1.69 1.53
13 14 15 16
LEASTSQUARESLINESFOR -AH I N BENZENE US. AHNP (NITROMETHANE) l oAliphatic and aniline -AH = -0.0306 AHNP f 30.00 2 " Aliphatic, N-methylaniline, and N-ethylaniline, -AH = -0.0315 AHNP 28.10 3" Aliphatic and aromatic -AH = -0.0282 AHNP f 22.12 1' and 2" Aromatic (including substituted) -AH = -0.0866 AHNP 52.23
0.62 1.oo 0.78 1.37
LEASTSQUARESLINESFOR -AH
4 5 6 7 8
IN
0.26 .26 .79 96 1.49
+
I
+ + +
+
+
LEASTSQUARES L I N FOR ~ -AH 17
+
1O Aliphatic
IN
BENZENE v8. ZO* -AH = -6.39 X U *
+ 34.36
0.18
TABLE I11 EXPERIMENTAL AND CALCULATED HEATSOF REACTION FOR HETEROCYCLIC AROMATIC AMINES,GUANIDINES, AND AMINESCONTAINING HETEROATOMS Exptl.
No.
Amine
Cal~d.~
-AH1
-AH1
(koal./ mole)
(kcal./ mole)
Difference (koal./male)
pdnisidine 18.8 18.0 +0.8 Morpholine 25.0 22.7 f2.3 Il3-Diphenylguanidine 23.8 26.4 -2.6 Tetramethylguanidine 31.2 35.7 -4.5 Isoquinoline 12.7 11.8 f0.9 Pyridine 13.6 11.6 f2.0 Triethylenediamine 21.6 18.7 4-2.9 Tetramethylethylenediamine 22.2 19.5 3-2.7 33 Triethanolamine 27.8 16.8 fll.0 34 Bis-(cyanomethy1)amine 0.8 4.0 -3.2 a Calculated -AHl values were obtained by substitution of the corresponding pKb value in the appropriate least squares equation. 6 17 20 21 24 30 31 32
proof was obtained of the existence of 3 lines rather than 1 line with greater scatter. The statistical proof involved an "F" test. P-values of 54.5 and Fig. 1.-Representative thermometric titration curve for 11.8 were obtained for AH vs. pKb(H20) and AH reaction of tertiary aliphatic amines with trichloroacetic vs. AHNP (ethyl acetate), respectively. In both acid in benzene. Ne cases these F-values corresponded to 4/(N1 confidence limits for line number 14. From NI - 6) degrees of freedom (NI N2 N8 6 AHNP data obtained by StreuliB*10 in non-aqueous equaled 14), and were well above the 99% point solvents, guanidines, amines containing hetero- (F equals 5.03 a t the 99% level) of the F distribuatoms, and heterocyclic arqmatic amines were tion. Table I1 includes for comparison the least known to deviate from their respectlve groups squares lines for the plot of AH in benzene us. (ie., 1, 2, or 3'); therefore, these amines were ex- AHKP values in nitromethane. cluded from least sqpares calculations (Table 111). The linear relationship obtained between AH A comparison was made of the pooled residual and pKb within the three groups of amines (1, 2, mean square (r.m,s.) solutions for the three lines and 3') indicated that entropy changes (AS) were with the r.m.s. solution for one line with a greater constant, zero, or proportional to the free energy scatter of points. In both the AH 2)s. pKb(H20) change (AF) for each amine. The latter two proand AH us. AHNP (ethyl acetate) cases statistical posals are not logical in the light of analogies which may be drawn from other systems.6 Therefore, if (10) C. A. Streuli, Anal. Chem., 31, 1652 (1989). T ("C).
+
+ + +
Nov., 1962
HEATSOF NEUTRALIZATION AND RELATIVE STRENGTHS OF AMINESIN BENZENE
A S were constant for the different groups of amines, solvation and steric effects are constant within the various groups. I n benzene, the order of base strength observed for a group of amines such as 1, 2, and 3' n-propyl derivatives as well as the negatively substituted allyl or benzyl derivatives is 1' > 2' > 3' (see Table I). This order was observed in ethyl acetate from spectrophotometric measurements by Pearson and Vogelsong" and from potentiometric titrations by S t r e ~ l i . I~n nitromethane, the base strength order observed by Streuli was 3' > 2' > lo,while in water the order is 2' > 3' > I ( F1) strain is believed responsible for the order observed in benzene and ethyl acetate with slight contributions from solvation. F strain (face-to-face strain) involves all strains arising from a frontal attack of two reacting species.l2 Since the reacting entities in benzene are presumably the amine and a trichloroacetic acid ion pair or undissociated molecule, the bulky, trichloroacetate anion possibly could cause F strain in this acid-base reaction in benzene. Pearson and Vogelsongll observed (spectrophotometric measurements) secondary amines in benzene to be the strongest bases of the three types toward the reference acid 2,bdinitrophenol. Varying degrees of F strain and solvation involving the different reference acids are believed responsible for the differences in base strength order. "B" strain (back strain) should be independent of vents,13 and since different orders of base strength are observed in the various solvents, B strain probably i s not important in this case. Some solvation of the protonated amines by the n-electrons of benzene might occur which would favor the order obtained in benzene.14 Primary aliphatic amines should exhibit very limited steric strains and only slight solvation in benzene; therefore, one would expect a good correlation between AH values in benzene and Zu* for 1' aliphatic amines. Taft u* values are constants expressing the inductive effects of a sub~tituent.A ~ very low standard deviation was obtained for the least squares solution from the correlation of the A€€ in benzene for 1' amines us. ZU* values available in the literature (Table 11). Insufficient Zu" values were available for least squares solutions for 2 and 3" amines, but from Zu" data that a,re available, greater deviations appeared present in the more sterically hindered 2 and 3' amines. A greater spread of -AH values (approximately 6.5 kcal./mole) was obtained in the benzene: trichloroacetic acid system than the literature indicated for the acetonitrile :hydrobromic acid (about 2 kcal./mole)6 or benzene :2,4-dinitrophenol systems (about 1.5 kcal./mole).16 These differences illustrate the importance of the reference acid in determining base strengths in low dielectric media where the acid anion is intimately associated with the reactant species. (11) R. G. Pearson and D. C. Vogelsong, J . Am. Chem. Soc., 80, 1038 (1958). (12) R . Spitzer and K. S Pitzer, %bid.,TO, 1261 (1948). (13) R. G. Pearson rtnd F. V. Williams, 7bzd., 76, 258 (1954). (14) Private communication wlth F. A. Cotton. (15) J. W. Bayles rtnd A. F. Taylor, J . Chem. Soc., 417 (1961).
32
L
28
-
24
-
2153
e
a'
20-
ii
-c 1
f
16-
-$
12-
f
a V
I
8-
0
PKb
(e).
Fig. 2.-AH1 in benzene vs. pKb in water for amines: line 5 = primary aliphatic amines and aniline [ 01 ; line 6 = secondary aliphatic amines, N-methylaniline, and N-ethylaniline [e]; line 7 = tertiary aliphatic and tertiary aromatic amines [a]; line 8 = primary and secondary aromatic amines (m- and p-substituted) [ 01.
Aromatic Amines.-Indistinct end points obtained for aromatic amines indicated that only semiquantitative determinations are possible. A weak base such as m-nitroaniline (pKb = 11.5) represented the limit in detectability (AH = -0.3 kcal./mole). Additional heats of association (AHz and AHa) were indistinguishable due to the rounded end points. A separate relation of A H with pKb (Fig. 2) obtained for 1 and 2' aromatic amines (meta or para) indicated a weakening of base strength in benzene relative to the aliphatics and the tertiary aromatics. In nitromethane or ethyl acetate, AHKP : pK, plots are linear over both aliphatic and aromatic amines. Hummelstedt and Hume observed similar phenomena in photometric titrations of acids with aromatic amines in glacial acetic acid and benzene solvent^.^^^^^ They suggest the formation of a hydrogen-bonded complex between protonated and unprotonated amine species. If this were true, a thermometric titration break should have been obtained at the 1 acid concentration. Actually, the titration slopes for all aromatic amines were rounded and could be ascribed to this phenomenon. The 3' aromatics would not be expected to form such a complex, due to steric (16) L. E. I. Hummelstedt and D. N. Hume, Anal. Chem., 32, 576 (1960). (17) L. E. I. Hummelstedt and D. N. Hume, J . Am. Chem. Soc., 83, 1564 (1961).
2154
J. R. SAMS,JR.,G. CONSTABARIS, S N D G. D. HALSEY, JR.
considerations. However, the negative inductive effect of the aromatic ring should destabilize any H-bond formed. KOexplanation was offered as to why the aliphatics v7ould not also form a similar complex. Guanidines, Heterocyclic Aromatic Amines, and Amines Containing Heteroatoms.-As occurs in other non-aqueous solvents,3 bases containing 0 or additional N atoms showed increased basicity in benzene relative to amines of comparable strength in water (Table 111). In water the reaction center of a given group of amines (1, 2, or 3") is solvated to approximately the same extent. These amines in a non-polar solvent are solvated less; therefore, amines containing heteroatoms can solvate intermolecularly and thus enhance their base strength relative to monofunctional compounds. Diamines and guanidines appeared to form a separate line between the 3 and 2' amines from a plot of AH in benzene vs. pKb(H20) (see Table 111), but insufficient data were available to merit calculating a least squares relation. Amine :Acid Associations Greater Than 1 :1.In addition to the 1: 1 molar reaction observed, 1 :2 and 1:3 amine :acid associations were obtained. The order of AH for the second association was the reverse of that for the first, i e . , the AH2 order is 3' > 2' > 1' (Table I). The sum of AH, and AH2 for each series of 1, 2 , or 3' amines is approximately equal. An infrared investigation1*indicated the point of attack for the second association still occurred a t the amine nitrogen rather than the chlorine atom of the acid. The reaction of trichloroacetic acid with nbutylamine and tri-n-butylamine was studied by infrared spectroscopy. The prevalent reaction product a t the 1: 1 end point and in the presence of (18) R. B. Hannan, private communication.
Vol. 66
excess acid was a carboxylate anion
[ C] -C'
06presumably ion-paired to a cation or cations. Infrared measurements also indicated the tri-nbutylammonium ion formed a stronger bond with the carboxylate ion than the n-butylammonium ion. In the presence of excess acid the carboxylate stretching frequencies of the tributylamine : trichloroacetic acid complex showed a greater shift from the spectrum of potassium trichloroacetate than the n-butylamine : acid complex. Recently, Marshall and Steigmanlg reported the presence of triple ions in the reaction of dibutylamine with 2,4dinitrophenol in benzene. If a triple ion were formed according to
[R3NH+A-]
+ [H+A-1 Jr
+
[RaKH+A-H+] A- (18) a tertiary amine would form the strongest bonds, for the protonated primary amine probably is more strongly solvated by t'he benzene 7r-electrons. Acknowledgments.-The author is indebted t'o Dr. Carl A. Streuli for helpful discussions during the course of this work, to Mr. R. P. Davis for performing the infrared analyses, and to Dr. R. B. Hannan for interpreting the infrared spectra. (19) Paper by P. Marshall and J. Steigman presented a t the Metropolitan Regional Meeting (New York and n'orth Jersey Sections) of the American Chemical Society. (20) R. G. Bates and H. B. Hetzer, J . Phys. Chem., 66, 667 (1961). (21) A. I. Biggs, J . Chem. Soc., 2572 (1961). ( 2 2 ) I. M. Kolthoff and N. H. Furman, "Potentiometric Titrations," John Wiley and Sons, Inc., New York, N. Y., 1926, p. 329. (23) G. W.Stevenson and D. Williamson, J . Am. Chem. Soc., 80, 5043 (1958). (24) W. A. Henderson, private communication. (25) Tables for "Analytical Handbook," by L. Meites, in publication.
ADSORPTION OF BRGON O N GRAPHITIZED CARBON BLACIC SURFACE AREA AND HEATS AND ENTROPIES OF ADSORPTION' BY J. R. SAMS,J R . ,G. ~ CONSTABARIS, AND G. D. HALSEY, JR. Department of Chemistry, University of Washinyton, Seattle 5, Washington Received April 6,1962
Adsorption isotherms of argon on the highly graphitized carbon black P33 (2700') between 90 and 137°K. are presented. Estimates of the surface area of the adsorbent and heats and entropies of adsorption computed from the data are discussed. The present results are compared with quantities obtained through the virial coefficient treatment of physical adsorption.
Introduction Our recent studies of the interactioiis of gases with the highly graphitized carbon black P33 (2700") employing a high precision adsorption (1) This research was supported in part by the United States Air Force through the AFOSR, and in part by the American Petroleum Institute. ( 2 ) Standard Oil Co. of California Fellow, 1961-1962. (3) 3. R. Sams, Jr., G. Constabaris, and G. D. Halsey, Jr., J . Phys. Chem., 64, 1689 (1960). (4) G. Constabaris, J . R. Sams, Jr., and G . D. I-Ialuey, Jr., ibid.. 66, 367 (1961).
apparatus,' have yielded considerable information on the nature of these interactions. The data have been analyzed in terms of a virial coefficient treatment analogous to the virial expansion for imperfect gases,*-l0 and the terms in the expa,nsion are (5) J. R. Sams, Jr., G. Constabaris, and G. D. Halsey, J r . , J . Chem. Phys., 9 6 , 815 (1962). (6) R. Yaris and .I. R. Sams, Jr., ibzd.,t o be published. (7) G. Constabaris, J. H. Singleton, and G. D. Halsey, J r . , J . I'liys. Chem., 6S, 1350 (1959). (8) 1%'. A. Steele and G. D. Halsey, Jr., J . Chem. Phys., 22, 979 (1.954).