Skip to main content

Full text of "A N. China ploughman [graphic]"

See other formats


EXCHANGE 




UNIVERSITY OF PENNSYLVANIA 



The Role of Ionic Activities in Catalysis in 
Liquid Systems 



Acetyl-Chloramino-Benzene 



P. Chloracetanilide 



A THESIS 

PRESENTED TO THE GRADUATE SCHOOL AS PARTIAL 
FULFILMENT OF THE REQUIREMENTS FOR THE 
DEGREE OF DOCTOR OF PHILOSOPHY. 

BY 

HARRY SELTZ 



Philadelphia, Pennsylvania 
1922 



UNIVERSITY OF PENNSYLVANIA 



The Role of Ionic Activities in Catalysis in 
Liquid Systems 

Acetyl-Chloramino-Benzene ^ P. Chloracetanilide 



A THESIS 

PRESENTED TO THE GRADUATE SCHOOL AS PARTIAL 
FULFILMENT OF THE REQUIREMENTS FOR THE 
DEGREE OF DOCTOR OF PHILOSOPHY. 

BY 

HARRY SELTZ 



Philadelphia, Pennsylvania 
1922 



ACKNOWLEDGMENT 

The author wishes to express his sincere thanks to Dr. Herbert 
S. Earned, under whose direction and guidance this work was 
done. 




INTRODUCTION 

The catalytic effect of specific ions on the velocities of certain 
reactions, in liquid systems, has been the subjects of much in- 
vestigation. The velocity of hydrolysis of esters and the inver- 
sion of sucrose are accelerated by the presence of either the 
hydrogen or hydroxyl ions. Benzaldehyde is converted to 
benzoin by the catalytic influence of the cyanide ion. Hydrogen 
peroxide decomposes into water and oxygen upon the intro- 
duction of the iodine ion. Acetyl-chloramino-benzene is con- 
verted to p-chloracetanilide by the simultaneous presence of 
the hydrogen and chlorine ions. 

It was early noticed that, in very dilute solutions, the velocities 
of such reactions were proportional to the concentrations of 
the catalyzing ions, as computed from conductivity ratios, a 
relation which would be predicted by the classic " Dissociation 
Theory." As the concentration of the catalyzing ion increases, 
however, this proportionality no longer holds. The velocity 
increases more rapidly than would be expected from conductivity 
measurements. To explain this discrepancy the "Dual Theory" 
of catalysis was advanced by Senter (Trans. Chem. Soc. 91, 
467, 1907); Acree (Amer. Chem. Journ. 37, 410; ibid. 38, 258, 
1907); and others. This theory attributed a catalytic effect 
not only to the ion in question, but also to the undissociated 
molecule. For example, in the hydrolysis of esters by hydro- 
chloric acid if we denote the catalytic effect of the undissociated 
molecule of the acid by k m , and that of the hydrogen ion 
by kh, and the total observed velocity constant by k, then: 

k = k h + (1 a) k m 

in which a is the degree of dissociation of the acid at the given 
concentration. This theory has considerable applicability, 
but there is no experimental evidence to prove the existence 
of any such catalytic property of the undissociated molecule. 
Arrhenius, as early as 1899, indicated a possible relation 
between the osmotic pressures of the ions and their catalytic 
influence on certain reactions. Apparently the first suggestion 

3 



494211 



that the ion activities, as defined by G. N. Lewis (Proc. Amcr. 
Acad. 37, 49 (1901), etc.), should be used in velocity calculations 
in catalysis in liquid systems, was made by Harned (J. Amer. 
Chem. Soc., Vol. 37, 11, 2460 (1915) ). Such a conclusion can 
be reached from thermodynamic considerations. 

The "Free Energy Function" (F) is defined by the following 
equation : 

F= U TS + PV ' (1) 

in which, U = Internal Energy 

T = Absolute Temperature 
S = Entropy 
P = Pressure 
V = Volume 

The differential equation relating these quantities, pressure 
and temperature constant, is: 

d F] PfT = dU TdS + PdV. . . v (2) 

For a reversible process the following relation exists between 
the entropy, internal energy, and work (W), 

dU + W 
~T~ 

Substituting this value of dS in equation (2): 

[ dF] PjT = W PdV (4) 

for any reversible process. 

The thermodynamic criterion, therefore, for equilibrium in 
any system where only mechanical work is done, is that: 

dF= O 

The " activity" (a) of any component of a system is related to 
the free energy of that component by the equation (G. N. 
Lewis, Proc. Amer. Acad. Art. Sci. 43, 259 (1907); Bornsted, 
J. Am. Chem. Soc. 42, 761 (1920) ) : 

F = RT In a + i (5) 

in which "i" is the free energy at some zero state. Let us 
consider the following reaction: 

A + B + Ct^D + E + H 
in which A, B, C, D, E and H represent molal quantities. Let 



5 

FA> F B , F c etc., represent the corresponding free energies. 
Then, for equilibrium (dF = o) : 



or, from equation (5) : 

(RTln a A + i A ) -f (RTlna B + i B ) + (RTln a c + i c ) = 

(RTlna D + in) + (RTlna E + i E ) + (RTlna H + in) 
and, 

In a D + In a E -f In a H In a A In a B In a c = - - = KI 

RT 

or, = K (6) 

aA X a B X a c 

We thus arrive at an expression for equilibrium similar in 
form to the "Mass Action Law," except that the usual con- 
centrations are replaced by activities. This is the thermodyna- 
mic expression for equilibrium in any system. If we consider 
this equilibrium to be a dynamic one, the opposing velocities 
being equal, it follows that the velocities can be related: 
V = Ka A X a B X a c 



V = Ka D X a E X a H i ' 
Such a deduction leads directly to the idea that, in the con- 
sideration of reaction velocities, catalyzed by specific ions, 
the proportionality is not to the concentrations of such ions, 
but rather to their activities. 

Harned (J. Amer. Chem. Soc., Vol. XL, 1461 (1918) ), in a 
discussion of catalysis in the presence of neutral salts, pointed 
out some interesting relations between reaction velocities and 
activities. 

Jones and Lewis (J. of Chem. Soc. 117, 1120 (1920) ), studied 
the inversion of sucrose by sulphuric acid from the view point 
of the activity of the catalyzing hydrogen ion. They calculated 
this activity, in the presence of various concentrations of 
sucrose, from electromotive force measurements of the cells: 



H 2 SO 4 (c) Sucrose (c) 



Saturated 
KC1 



HgCl 



Hg 



They found that the reaction could be explained on the basis 
of the hydrogen ion activity, providing the "water displacement 



6 

effect" be taken into consideration. They obtained good agree- 
ment for K w from the relation: 

K u - Kuni 



[H 2 0] x a H 

in which K w is the true bimolecular velocity constant at any 
temperature, K un i the observed unimolecular constant, [H 2 O] 
the concentration of the water, and an the activity of the 
hydrogen ion. The objection to this work is the fact that the 
calculation of the hydrogen ion activities from the cell indicated 
above, requires the elimination of the liquid junction potential 
by saturated potassium chloride solution. Furthermore, in 
this reaction the water molecule enters into the stoichiometrical 
equation, introducing a complicating factor. 

Akerlof (Z. physik. Chem. 98, 260 (1921) ) has carried out 
an extensive investigation of the hydrolysis of ethyl acetate 
by acids in the presence of neutral salts; with corresponding 
measurements of hydrogen ion activities. Here again the 
liquid junction difficulty enters. Moreover, the introduction 
of neutral salts complicates the problem considerably. 

It is the purpose of this paper to attempt to establish the 
activity relation free from any such complicating influences 
as those mentioned above. A reaction apparently ideal in this 
respect is the conversion of acetyl-chloramino -benzene to p- 
chloracetanilide, catalyzed by hydrochloric acid. This reaction 
was first studied in aqueous solution, from a dynamic stand- 
point, by Rivett (Z. physik. Chem. 82, 201 (1913) ). He deter- 
mined the velocity constants of the conversion at 25 C. for 
concentrations of hydrochloric acid varying from 0.1 to 1.0 m. 
This reaction is catalyzed by the simultaneous presence of the 
hydrogen and chlorine ions, but, as shown by Rivett, the velocity 
is not proportional to the product of the concentrations of these 
ions, as calculated from conductivity measurements, the devia- 
tion being over 30% in the range studied. 

This reaction is particularly adaptable to study from the 
activity standpoint. Since both the hydrogen and chlorine 
ions enter into the catalysis, it should be the product of the 
activities of these ions which determines the velocity at any 
temperature. This activity product can be calculated, as 



shown below, from electromotive force measurements of the 
cells : 

H 2 | HCl ( o) I HgCl I Hg 

with thermodynamic accuracy, without liquid junction potential 
difficulties. This reaction, also, appears to be free from complica- 
ions of the water molecule. 

1 . The reaction Acetyl-Chloramino-Benzene -^P-Chloracetanilide. 

This reaction, according to the investigations of Orton and 
Jones (Report British Asso. for Advancement of Science (1910) 
85), takes place in two steps. 

(1). CflHgNCl.COCHa + H + Cl % CgHsNH.COCHs + C1 2 

(2). C 6 H 5 NH.COCH 3 + C1 2 _> C 6 H 4 C1.NHCOCH 3 + H+C1 
They have shown that, in carbontetrachloride solutions, in the 
case of anilides which do not chlorinate, the reaction 

Ar.NCLAc + HC1 _^ Ar.NH.Ac + C1 2 

is quantitative. (The chlorine was determined by the aspiration 
method). 

Reaction (1) above is a measurably slow reaction, while 
reaction (2) is very rapid, so that the system approximates a 
continuous series of successive equilibria, with the concentration 
of the hydrochloric acid remaining practically constant. 

If equation (7), derived above, 

V = ka A X a B X a c 

is applicable to this reaction, the velocity should be proportional 
to the product of the activities of the acetyl-chloramino-benzene, 
the hydrogen ion and the chlorine ion: 

V = K T a A X a H X a c i 

in which a A is the activity of the chloramine. Since the con- 
centration of the hydrochloric acid remains constant, a H X 
a c i is also constant; and we can write: 

V=ka A (8) 

If the activity coefficient of the chloramine is proportional to 
its concentration throughout the reaction, then k should corres- 
pond to a first order reaction constant. Since the concentration 



8 

of the compound docs not exceed 0.02 M, such an assumption 
is justified. The value of k should be given by the equation 

k = K T X a H X a a 
where KT is constant at any temperature, or 

K T =- - (9) 

an X a c i 

2. Experimental: 

The values of the unimolecular velocity constant k, for 
the reaction acetyl-chloramino-benzene to p-chloracetanilide, 
have been determined at 17.65, 25 and 35 C., for concentra- 
tions of hydrochloric acid from 0.1 to 1.0 M.; all concentrations 
were in mols per 1000 grams of water. The acetyl-chloramino- 
benzene was prepared by the method of Slossin (Ber. Deut. 
Chem. Ges. 28, 3265 (1895) ) and was purified by recrystalliza- 
tion. The hydrochloric acid used throughout was made by 
diluting constant boiling acid, and its concentrations was checked 
by gravimetric analysis. 

The procedure was similar to that used by Rivett. A hot 
saturated aqueous solution of the compound was prepared, 
and after standing for some hours was filtered. A weighed quant- 
ity of this solution was coo'ed slightly below the temperature at 
which the reaction was to be carried out, and a sufficient weight 
of the standardized acid added to bring the concentration to 
the required molality. It is important to note that a considerable 
rise in temperature, amounting to a degree in some cases, may 
take place upon mixing. The temperature was always adjusted 
before a zero reading was taken. The thermostat was regulated 
to - .02 C. A careful regulation of the temperature is essential, 
since a change of 0.1 C. causes a change of about 2% in the 
velocity constant. 

The course of the reaction was followed by removing 25 c.c. 
of the solution in a pipette, at definite time intervals, and in- 
troducing it into a solution of potassium iodide. The following 
reaction takes place: 

C 6 H 5 NC1, COCH 3 + 2H + 21 = C 6 H 5 NHCOCH 3 + I 2 + H + Cl 
The p-chloracetanilide does not react with the iodide. The 



9 



free iodine liberated was titrated with a dilute (.015 M.) solution 
of sodium thiosulphate, using starch as an indicator. The 
value of k was calculated from the usual first order equation: 

1 A 
k= ~ log - 
t A! 

in which AQ is the zero titer and Aj the titer after "t" minutes. 
Readings were taken at such time intervals that six to ten titers 
could be made before the reaction reached the half-way point. 
These time intervals, naturally, varied greatly, depending on 
the temperature and concentration of acid. At 18 C. for 
0.1 M. acid, readings were taken every two hours; while at 35 
C., for 1.0 M. acid, it was necessary to make them every thirty 
seconds. 

In Table No. 1 are given the final meanvalues of k for each 
concentration of acid; the maximum variation in a series was 
1%. The concentration of acid (c)'is given in mols per 1000 
grs. of water. 

TABLE 1 

OBSERVED VELOCITY CONSTANTS* 



TABLE 1 




17.65 




25.00 




35.00 


C 


K 


( 
C 


k 


C 


k 


0.1 


0.0002160 


0.1 


0.000467 


0.1 


0.001500 


0.2 


0.000781 


0.2 


0.00171S 


0.2 


0.00562 


0.3 


0.001710 


0.3 


0.003756 


0.3 


0.01200 


0.4 


0.003040 


0.4 


0.00666 


0.4 


0.02070 


0.5 


0.00472 


0.5 


0.01040 


0.5 


0.03254 


0.6 


0.00694 


0.6 


0.01505 


0.6 


0.0473 


0.7 


0.00952 


0.7 


0.02104 


0.7 


0.0645 


0.8 


0.01276 


0.8 


0.02800 


0.8 


0.0868 


0.9 


0.1660 


0.9 


0.0366 


0.9 


0.1126 






1.0 


0.0465 


1.0 


0.1431 



* NOTE: The data by Rivett (loc. cit.), at 25 C. were determined on 
a volume normal basis, at odd concentrations. When corrected to weight, 
molal they correspond fairly well to the values given above. 



10 

3. The Activity Values: 

Activity has been denned above by the equation: 

F = RTln a + i (per mol) 
or in general: 

Fi F 2 =RTm- 
a 2 

in which R is the gas constant, (Fi F 2 ) the decrease in free 
energy of a system resulting from the transfer of one mol of 
the substance, or ion, from a concentration in which its ac- 
tivity is "ai". to a concentration in which its activity is "a 2 ," 
at the absolute temperature T. The free energy change, (Fi 
F 2 ) or ( A F) for a solution of hydrochloric acid in going from 
a concentration Ci to C 2 can be determined by measurements 
of the electromotive* force of the cells: 

H 2 | HCl(c) | HgCl Hg 

at concentrations Ci and C 2 . The difference in electromotive 
forces at these two concentrations, multiplied by the Faraday, 
gives the free energy change directly in joules. 

Ellis (Jour. Amer. Chem. Soc. 38, 737 (1916) and 39, 2532 
(1916) has accurately determined these free energy changes for 
hydrochloric acid at 18 ,25 and 35 C., over a large range of 
concentrations. 

From the equation: 



a H a C i 

the ratio of the products of the ion activities can be calculated. 
Calling F H and F C i the activity coefficients of the hydrogen 
and chlorine ions, respectively, then 

(F H X C) X (Fa X C) = a H X a cl 

in which C is the concentration of the acid. The value VF H XF C i 
denoted by F' a , is the activity coefficient of the acid. 

Ellis assumes this activity coefficient for hydrochloric acid 
to equal the conductance ratio at .001 M., having a value at 
this concentration of 0.985. He claculates from this the values 
of the activity coefficient for the higher concentrations. 



11 



The activity coefficients used in this paper were not taken 
directly from a plot of Ellis 's data, but were calculated by a 
method suggested by Harned (Journ. Amer. Soc. 44, 252 (1922) ) 



JlL 



CutWE 



J4k_ 
K. 

JJfcJL 



-136 




7 



(f.)* 



llfc 




.580 



0- t 



CoNt^TafNTi 

A 



d) HO? 

5 



o-t 



F.S.! 



from the equation: 



(10) 



in which a 1 ,^ 1 and m 1 are constants at any temperature, and 
C is the concentration of the acid. The values of these con- 
stants were determined graphically to give the best mean 
values of Ellis 's results. They are: 



12 



18 

a 1 0.200 
g 1 0.285 
m 1 0.455 



25 

0.200 

0.286 
0.434 



35 

0.1SO 
0.277 
0.414 



4- Discussion of Results: 

The values of k (Table No. 1), at each temperature, increased 
rapidly with the increase in concentration of hydrochloric acid. 
As a first approximation it was noticed that these velocity 
constants were roughly proportional to the square of the con- 

k 

centration of the acid. An examination of the values of - 

C 2 

(Table No. 2 ) shows, however, a pronounced minimum for each 
temperature, at 0.5 M. acid. Column No. 4, Table No. 2, 
shows the corresponding values of the activity coefficients, 
calculated as described above. It is immediately apparent 
that the minima for these values occur at precisely the same 
concentration, namely 0.5 M. This point is well illustrated 
by the curve (Fig. 1), in which the values of (F^) 2 are plotted 
against concentration of acid! The form of the curve is similar 

k 
to that of , with a perfect coincidence of the minima. 

C 2 

TABLE NO. 2 

Temp. 17.65 C. 



(1) 


(2) 


(3) 


(4) 


(5) 


(6) 






K 






K 


C 


k 


C 2 


*% 


(F x a ) 2 




0.1 


0.0002160 


0.0216 


0.832 


0.692 


0.0314 


0.2 


0.000781 


0.01953 


0.800 


0.640 


0.0311 


0.3 


0.001710 


0.01900 


0.785 


0.617 


0.0308 


0.4 


0.003040 


0.01900 


0.781 


0.610 


0.0313 


0.5 


0.00472 


0.01888 


0.780 


0.608 


0.0311 


0.6 


0.00694 


0.01928 


0.783 


0.614 


0.0314 


0.7 


0.00952 


0.01940 


0.790 


0.624 


0.0313 


0.8 


0.01276 


0.01990 


0.799 


0.639 


0.0314 


0.9 


0.1660 


0.02050 


0.810 


0.655 


0.0310 



Mean 0.0312 



13 



Temp. 25.00 C 



0.1 


0.000467 


0.0467 


0.822 


0.605 


0.0692 


0.2 


0.001718 


0.0430 


0.790 


0.623 


0.0690 


0.3 


0.003756 


0.04175 


0.777 


0.604 


0.0691 


0.4 


0.00666 


0.0416 


0.773 


0.598 


0.0695 


0.5 


0.01040 


0.0416 


0.773 


0.598 


0.0695 


0.6 


0.01505 


0.0418 


0.778 


0.605 


0.0692 


0.7 


0.02104 


0.04295 


0.785 


0.616 


0.0697 


0.8 


0.02800 


0.0438 


0.795 


0.632 


0.0693 


0.9 


0.03660 


0.0448 


0.807 


0.651 


0.0694 


1.0 


0.0465 


0.0465 


0.820 


0.762 


0.0692 



Mean 0.0693 



Temp. 35.00 C 



0.1 


0.001500 


0.1500 


0.815 


0.664 


0.226 


0.2 


0.00562 


0.1405 


0.783 


0.613 


0.229 


0.3 


0.01200 


0.1333 


0.769 


0.591 


0.226 


0.4 


0.02070 


0.1294 


0.763 


0.582 


0.223 


0.5 


0.03254 


0.1302 


0.762 


0.581 


0.224 


0.6 


0.0473 


0.1315 


0.765 


0.585 


0.225 


0.7 


0.0645 


0.1317 


0.771 


0.594 


0.222 


0.8 


0.0868 


0.1358 


0.779 


0.606 


0.224 


0.9 


0.1126 


0.1390 


0.788 


0.620 


0.224 


1.0 


0.1431 


0.1431 


0.800 


0.640 


0.224 



In Column No. 6 are given final values of 



Mean 0.224 
k 



The 



Vr a XC) 2 

constancy of these figures, at each temperature, over the entire 
range of concentration, is well within the limit of experimental 
error. 

Since, by definition of F^: (FaXC) 2 =ajiXaoi 

k 
the relation KT "~ is without any question substan- 

a H Xa ci 

tiated. This conclusion is particularly significant in this case, 
when we consider that these results are based entirely on accurate 
thermodynamic activity data, without complications due to 
liquid junction potentials, and without any assumptions as to 
individual ion activities. 



14 

5. The Temperature Coefficient. 

With the development of the radiation hypothesis of reaction 
velocity, it is of considerable importance to obtain the tempera- 
ture coefficients of reaction velocity constants. The temperature 
coefficient should be expressed by an equation of the form 

dlnK T E c 



d RT 2 

where EC may be called the critical increment. According to 
the rule of Arrhenius E is a constant. It is not necessary to 
mention the various meanings attached to E by different 
authors (W. C. McC. Lewis, System of Physical Chemistry, 
Longmans; Perrin, Annales de Physique, Ser. 9, 11, 5 (1919); 
Tolman (Jour. Amer. Chem. Soc. 42, 2506 (1920) and others), 
but it is of some value to point out that in this case the rule 
of Arrhenius is not valid over the short range of temperature. 
The plot of lnK T against VT is not a straight line but has con- 
siderable curvature, showing that E c varies considerably with 
the temperature. Integration of (6) between temperature 
limits T! and T 2 , assuming that E c is constant, gives 

K T 
log 



l! 1 

whence E c is found to be 1.356 X 10 5 Cals. between 25 and 
35 C., and 1.196 X 10 Cals. between 17.65 and 25 C. This 
is sufficient to show that the rule of Arrhenius does not hold. 

The Neutral Salt Effect. 

From measurements of the electromotive forces of cells of 
the type 

H 2 | HCl ( o.iM)+MeCl (C ) | HgCl | Hg 

it is possible to obtain the product of the activities of the hydro- 
gen and chlorine ions in acid salt mixtures exactly, without 
any difficulties arising from liquid junction potentials. Exact 
measurements of these cells have been made by Harned (Jour. 
Am. Chem. Soc. 38, 1986 (1916); 42, 1808 (1920), ) so that the 



15 

required thermodynamic data are available for applying the 
activity theory to the velocity of conversion of acetyl-chloramino- 
benzene to p-chloracetanilide. Velocity measurements on this 
reaction have been made, using hydrochloric-sodium chloride 
mixtures, and hydrochloric acid-potassium chloride mixtures 

k 

which show that the relation K-p = - which holds so beau- 

a H X a c i 

tifully for hydrochloric acid alone fails to hold in the presence 
of the neutral salts. KT instead of remaining constant, de- 
creases considerably with increasing salt concentration. This 
study will be the subject of a future communication. 

Summary : 

1. It has been derived, from thermodynamic considerations, 
that, in reactions in liquid systems catalyzed by specific ions, 
it is the activities of such ions which determine the velocity 
of the reaction at any temperature when the catalysis depends 
on successive states of equilibria. 

2. The velocity constants of the reaction acetyl-chloramino- 
benzene to p-chloracetanilide, catalyzed by hydrochloric acid, 
have been determined at 17.65, 25 and 35 C. for concentra- 
tions of acid from 0.1 to 1.0 M. 

3. It has been shown that these velocity constants at each 
temperatures are proportional to the product of the activities 
of the hydrogen and chlorine ions of the catalyzing acid. 

4. As far as we are aware, this is the first case in which 
homogeneous catalysis may be calculated with exactness over 
a wide concentration range and in concentrated solutions, 
without guess work concerning liquid junction potentials. 

5. The temperature coefficient has been considered and the 
critical increment has beeji roughly calculated and shown to 
vary considerably with rise on temperature. 

k 

6. The relation K T = - - which holds exactly for 

a H X a c i 

solutions of hydrochloric alone, fails to hold when acid-salt 
mixtures are used as catalysts. 
PHILADELPHIA, PA 



Gaylord Bro 

Makers 
Syracuse, N. 
" PAT. JAN. 21J90S 



4942! i 

seltz, Barry 

Hole of ionic activitiei 



S5 



404211 



UNIVERSITY OF CALIFORNIA LIBRARY