592 P I N C H A N G AND
59
values for both nickel surfaces are compared in
Fig. 4. These values were one to two entropy
units lower for the unreduced surface and supported
the contention based on previous data that this
unreduced surface was the most heterogeneous of
the two t,ypes studied. Furthermore, the con-
figurational entropies for both surfaces tended
toward zero as 0 + 0, a fact already deduced from
the shapes of the integral entropy curves at low 0.
The calculated Sc curves increased with in-
creasing coverage and eventually exceeded the
maximum possible configurational entropy, S,.
For the reduced sample, this took place at 0
and undoubtedly was evidence for the onset of
0.2 - coverages. These values are very near the entropy
adsorbate interactions at unusually low coverages.
The equations developed by Hill2 to treat adsorp-
tion on a heterogeneous surface with nearest neigh-
SOME MEASURE,MENTS OF DIFFUSION I N LIQUIDS
BY P I N C H A N G .4ND
Division of Chemical Engineering, University of Calijornia, Berkeley, Calif.
Received November 26, 1954
Diffusion coefficients for iodine and toluene in a series of saturated hydrocarbons and for benzoic, acet,ic and formic
acids in a number of organic solvents are reported. Data were obtained by the diaphragm cell method.
Introduction
At the present time diffusion in liquids is not
fully understood. Besides inadequate theoretical
description of the liquid state, insufficiency of
experimental data has constituted one maiii
difficulty in its investigation. Not all the data in
the literature are sufficiently accurate, and some
pertain to relatively high concentrations of the
solute thereby requiring a knowledge of activity
coefficients for interpretation. In addition there
has been more extensive study of the diffusion of
various solutes in a given solvent than of a single
solute in various solvents. The latter is desirable
for consideration of the role of solvent properties.
This investigation was undertaken to provide
additional data for dilute solutions to serve as a
possible basis for the development of improved
correlations of liquid diffusion.
Experimental
A . Solvents Used.-All solvents except normal hydro-
carbons were of C.P. grade and were used without further
purification.
The normal hydrocarbons were first passed through a
silica gel column, and then distilled. The portion of dis-
tillate having a boiling point within one degree of the boiling
point of the pure substance was recovered for use in the
diffusion studies. The viscosities of the normal hydro-
carbons were measured with an Ostwald viscosimeter.
Their viscosities are compared with those reported in
Timmermans' as follows.
Viscosity of
Viscosity of pure solvents at
solvent used 25O reported in
Solvent at 2 5 ' , (2.1). Timmermans
n-Decane 0,860 0.853
n-Dodecane 1.416 1.353
n-Heptane 0.384 0.389
%-Hexane 0.304 0.294
n-Octane 0,509 0.508
n-Tetradecane 2.12 ...
(1) T. Timmermans, "Pliysico-chemical Constants of Pure Organic
Compounds," Elsevier Publishing Co., Inc., New York, N. Y.
C. R. WILKE Vol.
bor interaction were too complex to treat mat'he-
matically even if a proper distribution function had
been available. Thus it was likely that a model of
localized adsorption on a heterogeneous surface
without adsorbate interaction was not valid
beyond 0 = 0.2 and 0 = 0.5 for the reduced and
unreduced surfaces, respectively.
In the presence of interactions a division into a
non-configurational and configurational- entropy
becomes arbitrary. Thus, it is difficult to interpret
the essentially constant entropy values at higher
of liquid argon for both samples.
Acknowledgment.-The authors gratefully ac-
knowledge the financial support provided by the
Office of Naval Research, Project NR-057-186,
Contract N8onr-74300.
c. R. ~ ~ I L K E
I t is seen that except for n-dodecane and n-hexane viscosity
values agree very closely with those reported in the litera-
ture.
B. Apparatus.-The diaphragm cell method was used
for measurement and its principles and technique have
been described previouslg.2 The diaphragms had a pore
diameter of two to five microns so that stirring had no
detectable effect on the transport within the diaphragms.
Each of the cells consisted of a vertical glass diaphragm
approximately 3 mm. in thickness and 5 cm. in diameter
separating the cell into two compartments having a volume
of 80-100 ml. each. Short iron rods encased in glass were
placed in each compartment to be operated as magnetic
stirrers. Magnets were located suitably beneath the cells
on pulleys which were rotated by a variable speed belt
drive. This arrangement provided adequate stirring t o
maintain uniform concentrations in each compartment. The
entire apparatus was built into an air-bath containing eight
rotating magnets capable of simultaneous operation of four
cells. Bath temperature was controlled to f.0.1'.
C. Procedure.-In treating the diffusion experiment tht,
standard equation2 was used
where
p' = the cell constant
Di = the integral diffusion coefficient
t= time during an experiment
ACi and ACr = the initial and final concentration dif-
ferences
Equation 1 is based on the assumption of concentration-
independent diffusion coefficients and quasi steady-state
conditions. Therefore the diffusion coefficient obtaine!
is an integral diffusion coefficient. The cell constant, p ,
has to be determined with a standard substance whose diffu-
sion coefficient is known as a function of concentration.
For this purpose, 0.1 N KCl aqueous solution was used.
For the diffusion of 0.1 N KCl solution in water until con-
centrations in the two compartments are, respectively,
0.075 and 0.025 N the integral diffusion coefficient is 1.870
X cm.Z/sec. a t 2503 and 2.093 X cm.2/sec.
a t 30". The value a t 30' is based on diffusion data ob-
tained partly from Harned and Nuttall3 and partly from the
extrapolation by theory of Onsager and F u ~ s s . Since~
(2) A. R. Gordon, N . Y . Acad. Sci., 46, 285 (1915).
(3) H.S. Harned and R. L. Nuttall, ibid., 61, 781 (19491.
July, 1953 MEASUREMENTS
OF DIFFUSION
IS LIQCIDS 593

diffusion data 01 IiCl up to 0.1 A' solution at 4 and 25" and gives reproducible cell constaiits. Furtherinore, the
the existing data a t 30" agree well with the theory, the measurement of diffusion coefficient of potassium chloride
values obtained by extrapolation were e\pected to be a t 50.9" gave a value of 3.18 X cm.2/sec. i n agreement
correct within their experimental error. This was evi- wit11 the estrapolated value from the data of Harned and
denced by agreement between the values of cell constant Kuttall a t 25 and 30" (Table 11).
obtained a t 25 ant1 30".
For a nieasurcment, the solutions weie kept a t constant TABLE
I1
temperature in a water-bath. Either concentrated or OF 0.1 AT ICC1 INTO WATER
DIFFUSION
dilute solution was charged into one compartment and Tepp.,
application of vacuum to the other comp:wtment enabled C. D X 105, ciii.?/sec.
the solution to wet and fill ul, the glass diaphragm com-
pletely. This compsrtment \vas then filled. The other 25 1.870 (Haimed and Sut,tall)
compartment was similarly filled with the other solution, 30 2.093 (Harned and Nuttall)
and preliminary diffusion started in order to obtain a 50.0 3.18 (This refiearch)
continuous concentration gradient of solute across the 50.9 3.20 (Estrapolnted v:~Iiic)
diaphragm. For estimating the time necessary to do so the
following rule2was used 2. Measurement of the diffusion copfic4ent of hydro-
Dt = 1.2
- chloric acid a t initial concentrations of 0.07 1V and zero a t
L2 80" yielded a value of 3.40 X cm.*/sec. Interpolation
where of data of James and Gordon4at 25 and 35" gave a value of
3.30 X 10-6 cm.2/sec. On the other hand, Fdinski's
1 = time required datum6 of 0.1 N HC1 diffusing into water a t 16" is also
D = diffusion coefficient of the solute higher than the interpolated value from Gordon's dntn
L = effective length of the diaphragm (Table 111).
After sufficient tinip hnd elapsed, thc solution in each com- T.4BLE 111
partment was replaced by fresh solutions and the normal
diffusion period stwted. The noi,mnl diffusion period DIFFUSION OF 0.07 HCI I S T O WATER
A \

usually required a few days. At the end of the experiments, Diffusion coefficient, D X IO$ DIII.~/ISPC,
Inter-
samples were nithdrrtnw from each compaitment for anal- polation
ysis. Geneid practice byas to run the experiment long Temp., This froni
enough so as to malie log, ( A C i ) / ( A C f ) larger than 1.2. OC. rcsearcli Gordon Falinsk
I n this case 1 % error in titration would give about 470 error 30 3.40 3.30 , . .
in diffusion coefficient.
D. Analyses.-The analyses of solute concentrations
IG ... 2.39 2.52
were pe,rfornied by t,itration with a microburet, graduated to 3. 1Ieasurement of diffusion corKcic:iit of iodine i t i
0.02 ml. Standard solution of sodium thiosulfate of con- bromobenzene a t 30" yielded 1.394 X 10-6 cni.z/sec. while
ceiitration 0.05 N was used for analysis of iodine, and 0.1 N the ext,rltpolated value from RIillcr's data6 cnnir out to be
standard solution of sodium hydroxide was used for the 1.39 X 10-5 cin.2/sec.-in rsorllent :tgrceinent. Diffusion
analysis of the acids with phenolphthalein as indicator. of iodine i n carbon tetrachloride was also inpasured a t 30"
The reproducibility of this titration was usually =t0.15%. and compared with Rliller's data. Table I V shows that the
For the dcterminatioii of potassium chloride concentration, present values a t mean concentrattioils of 0.0501 and 0.0921
standard silver nitrate solution was used. The error in the ' \ i a t 30" are, respectively, 1.58 X l o + and 1.51 X 10-j
value of the diffusion coefficient introduced by the titration cm.Z/sec. compared with an extrapolat,~dv:tlue of 1.53 X
was found to depend on the time of the experiment.. I t wns 10-5 cni.*/sec. from RIillei,'s data for diffusion from 0.095
important to titrate solutions in one run with the same N to zero conceiit,ration in his apparatuw. Since? Miller
standard solution so that the error i n stand:irtlization of the used a different method of measurement, this integral
solution, if any, would be cancelled out, producing no coefficient would not correspond exactly to one obtained
effect on the value of diffusion coefficients. in this study with initial concentmtions of 0.005 A' and
The analysis for toluene in normal hydrocarbons was zero.
made by means of D Beckman spectrophotometer. The
extinction coefficients were determined and found to obey TABLEIV
Beer's law. However, these extinction coefficients may not DIFFIWOS O F IODINEI N CCI,
be those for the pure substances, for although the solvents D x 10i rlii.?/nec. A I'.
had been refined, there was no proof that they were chemi- Temp., This concn.,
cally pure. The diffusion coefficient was considered to be "C. I enenrch Niller !V
independent of the extinction coefficients as long as Beer's 30 1.58 ,.. 0.0501
law was obeyed. Values of extinction coefficients for the
solvents used are shown in Table I. The reproducibilit,y I .50 , . . 0.0921
of the iipect,rophotometric reading is within 1.5%. 1,5:3 0.048

TABLE I Results
EXTINCTION
COEFFICIENTS O F TOLUESE:
IS NORRIAL Data for diffusion of benzoic acid, acetic acid,
HYDROCARBONS formic acid, cinnamic acid, toluene and iodine are
Wave Molal tabulated in Tables V through X. The repro-
Temp., length, extinction
Solvent OC. A. coefficient ducibility of data depended upoti the duration of
n. Hexane 23.5 2600 185.0 f 0 . 8 the experiment and the technique of titration.
2620 2 0 G . 8 3 ~ .7 The acids and iodine were titrated by standard
n-Heptane 23.5 2GOO 1 0 9 . i f .1 solutions with a microburet calibrated to 0.02 of a
2620 1 1 2 , S f .2 milliliter. Usually three milliliters of standard
n-Decane 23.5 2GOO 1 0 5 . 2 f .4 solutions were required to titrate a sample of larger
2620 1 0 6 . 3 f .1 volume, and it was possible to read the buret to
n. Dodecane 23.5 2600 1 4 8 . 2 2 ~.5 0.01 milliliter. The reproducibility was therefore
2G20 1 6 6 , 4 f .6 within 0.3y0. For the diffusion measurement, the
values of A C i / A C f were usually around 1.4. Since
E. Reliability of the Method for Measurements Used in the diffusion coefficients depended 011 the logarith-
This Investigation.-The reliability of the present method of
measurement was examined by the following measures: (4) W. A. James a n d A. R. Gordon, J . Chem. P h y s . , 7 , 9G3 (1439).
1. Calibrat,ion of the diaphragm cells a t 25 and 30°, (5) AT. Falinski, Cornpl. rend., 218, 754 (1944).
using data of Harned and Nuttal13 for potassium chloride, (6) C. C. hliller, Proc. Rou. Soc. ( L o n d o n ) , 106, 724 (1024).
594 PIXCHAR'GA X D C. R. WILKE T'ol. 59

TABLE \.'
DIFFUSION
O F BENZOIC
,kC'ID
Viscosity Initial cell
D X lo', of solvent coinliartnient concn., DvlT X
9011e n t cin.2/sec. q X 102, poise g. mole/l. 10'0

Carbon 14.8 0.78 1,048 0.07389 0 2.82

tetra- 25.0 0.91 0,907 ,07227 0 2.76
chloride 40.5 1.17 ,732 ,07218 0 2.74
Benzene 15.0 1.17 . 696 , 090GO 0 2.83
25.0 1.88 .503 .00061 0 2.79
40.0 1.76 ,497 ,09052 0 2.78
Toluene 16.0 1.29 ,620 ,03195 0 2.76
25.0 1.49 ,552 .OR195 0 2.77
40.0 I .85 .471 ,03018 0 2.78
.dicetone 13.2 2.37 .342 ,05089 0 2.82
25.0 2.62 ,312 .05099 0 2.75
40.0 3.05 ,268 .05088 0 2.73
Ethylene 25.0 0.043 ( f10%) 17.33 ,05933 0 2.49
glycol 50.0 0.18 (+lo%) 6.62 ,00048 0 3.77

TABLE VI TABLE IX
DIFFUSION
OF ACETICACID^ OF FORMIC
DIFFUSION ACID"
Diffusion Viscosity Diffusion Viscosity
coefficient of solvent coefficient of solvent
Temp., D X 105, 7) X ,102, Dq/T X Temp., D X 106, q X,102, Dq/T X
Solvent OC. crn.z/sec. poise 1010 Solvent OC. cm.2/sec. pome 10'0

Carbon 6.5 1.15 1.201 4.94 Carbon 8.5 1 .61 1.163 (3.05
tetra- 14.8 1.27 1.052 4.63 tetra- 25.0 1.89 0,907 5.74
chloride 25.0 1.49 0,907 4.53 chloride 15.0 1.67 1.048 6.08
40.0 1.78 ,736 4.18 Benzene 6.2 1.99 0.815 5.81
Benzene 25.0 2.09 ,603 4.22 13.9 2.31 .712 5.72
5.9 1.58 ,817 4.62 25.0 2.28 ,603 5.21
Toluene 25.0 2.26 ,552 4.19 Toluene 6.2 2.28 ,706 5.77
15.0 1.90 ,620 4.10 14.1 2.46 .G35 5.44
6.8 1.66 .io1 4.16 25.0 2.65 ,552 4.90
Acetone 40.0 4.04 ,268 3.45 Acetone 25.0 3.77 ,312 3.94
25.0 3.31 ,312 3.46 13.5 3.27 ,346 3.95
15.0 2.92 .337 3.41 6.5 3.13 .370 4.14
Ethylene 1 5 . 0 0.039 (&lo%) 26.09 3.5 Ethylene 25.0 0.094 ( & l o % ) 17.33 5.5
glycol 25.0 0 . 1 3 ( & l o % ) 17.33 7.4 glycol 5 0 . 0 0.22 (=!=lO%) 6 . 6 2 4.4
3 0 . 0 0.20 (kklO%) 13.35 8.8 a Initial cell compartment concentrat'ions approximately
0.003 and 0 mole fraction.
5 Initial cell compartment concentrations approximately

0.003 and 0 mole fraction. TABLE X

DIFFUSION
OF IODINE
I N VARIOUS ORGANICSOLVENTS
TABLE VI1 D X Viscosity Initial cell
DIFFUSION
O F CINNAIUC
ACID 105, of solvent compartment
Temp., cm.z/ q X.102, concn., Dv/T
D X Viscosity Initial cell Solvent OC. sea. poise g. mole/g. X loio
105 of solvent compartment
Teinp., crn.i/ q X, 102, concn., Dv/T n-Tetra- 25 0.96 2.120 0.007745 0 6.81
Solvent '(2, see. poise g. mole/l. X 1010 decane
Carbon tetra- %-Octane 15 2.43 0.579 0.04220 0 4.89
chloride 25 0.70 0.907 0.03501 0 2.30 25 2.76 0,508 ,04210 0 4.71
Benzene 25 1.12 ,599 .03922 0 2.25 40 3.22 0.417 ,04231 0 4.29
Toluene 25 1.18 .552 ,03230 0 2.18 n-Hexane 25 4.24 0.304 .05600 0 4.31
Acetone 25 2.41 ,312 ,04915 0 2.52 Cyclo- 15 1.54 1.050 ,06389 0 5.65
hexane 30 1.92 0.820 ,06334 0 5.21
TABLE VI11 40 2.30 0.703 ,06383 0 5.17
O F TOLUENE
DIFFUSION hZet,hylcg- 40 2.71 0.549 .04422 0 4.74
Viscosity clohexane 30 2.30 0.639 ,04421 0 4.86
of sol\-ent
Tenip., D X 105, v X,102. Dq/T X Et,hj,l 40 1.77 0,821 ,1164 0 4.65
Solrent "C. CIII.~/S~C. poise 10'0 alcohol 25 1.32 1.065 ,1105 0 4.71
n-Hexane 25.0 4.21 0.304 4.13 Broino- 30 1.30 0.085 ,03043 0 4.53
n-Heptane 25.0 3.72 0.384 4.80 benzene
n-Decane 25.0 2.09 0.860 0.04 Carbon 30 1.63 0.839
,01450" 4.57
n-Dodecane 25.0 1.38 1.416 6.54 tetra- 1.61 ,03622"
n-Tetra- chloride 1.58 05005"
decane 25.0 1.02 2.120 7.23 ,09205"
n-Hept ane G.9 2.05 0.469 4.94 Q Average concentration over both compartments during
40.0 4.33 0.333 4.60 the diffusion measurement.
July, 1955 ME.4SlTREMENTS O F I N LIQUIDS
DIFFLTSIOS 595

mic values of ACi/ ACf, a reproducibility within Although this equation has no quantitative ttlieo-
1.5% frequently was obtained, in fact, within retical basis it represents a large body of data Ivitli
lYc in many measurements. an average deviation of about 10%.
For diffusion in very viscous solutions, for Figure 1 sh0n.s experimental data for diffusion of
example the diffusion of acids in ethylene glycol, acetic and Ixnzoic acids in several solvents plotted
the data, were not espected to be accurate because according t o equatioii 2 . Practically all the data
of the slon diffusioii rate and resulting small con- fall withiii 10%) of the correlatioii line indicating
tration changes during an experimeiit. They that satisfactory results are ohlained nssumiiig the
were expecbed t o be good wit'liiii 20:7,, however. acids to lie ill monomeric form.
The diffusion of toluene at 23" in varioiis normal
hydrocarbons was studied. -1nalyses by the Beck- 20 I I l
man spectrophotomet'er were reproducible to mith-
i n 1.5%. With a value of ACi/ACf equal to 1.5,
t,liis reproducibility should g5ve an accuracy of
*2y0 for the diffusion coefficient. No duplicate
runs were made in normal hydrocarbons because
of t8helimited amount of solvents.
Discussion

I L - I I - 1
c
D
Data \\.ere talmi a t various temperatures for x 4
most of the systems. It was of particular interest BENZOIC A SOLVENT ACETIC A
t o examine the group Dq/T which according t'o *
\
0 cc14 a
bot8hthe Stokes-Einstein equation and the Eyring 2 0 BENZENE
theory should be essentially independent of A
I
TOLUENE A
t'emperature. 2 I 1

The Dq/T groups for benzoic acid diffusion in 0.2 0.4 0.6 0.8 1 2 1
carbon tetrachloride, benzene, toluene and acetone 7 P . 5 / n f V 9 .

do not vary by more than .57, over a fairly wide
range of temperature. Benzoic acid has an
anomalous behavior in ethylene glycol, as showii
by the appreciable c,hange of DqlT with tempera-
t'ure (Table TT). This does not seem to be at-
tributable t o inaccuracy of the results as mentioned
in the preceding section. However, it is not sur-
prising i n view of the complicated hydrogen bond-
ing structure in ethylene glycol which may change
i n character with temperature.
Escept for the syst'ems of acetic acid in acetone
and toluene and formic acid in acetone, formic acid
and acetic acid do not give constant values of
D v / T . This is possilily due to some complicat.ion
from molecular a'ssociatjon of formic and acetic
acids.
I n an adjunct paper8 the authors lia\,e esamined
a large body of data and proposed an empirical
equation for estimation of diffusion coefficients
where
D = diffusion coefficient, cm.Z/sec.
T = temperature, "I<.
ill = solvent molecular weight
7 = solution viscosity, poise
Fig. 1.-Diffusion of orgnnic acids.
The diffusion rate of foi,mic: acid in et,hyleiie
glycol is lower tQan that of acetic acid, whereas the
converse is t,rue in other solvents. This is evi-
dently due t o the formation of ester. In the
determinat,ion of the formic a.cid by titimation u,ith
sodium hydroxide solution, a fen- drops of t,he latter,
which is expect,ed to be much less thaii t,he quantity
required to neutralize the acid, suffice to turn the
solution red v,it'h pheiiolphthalein, t'he color
vanishing in ahout 30 seconds. This iiidicates
that only a small fractioii of t,he formic acid re-
mains as formic acid in equilibrium ivith t'he estter.
Comparison of data in Tables X ant1 SI indi-
cates that diffusion of iodine i i l saturated hydro-
carbons gi\res coiisiderahly higher T x l u e b of D q / T
than diffusion i n aromatic hydrocai*boiis of similar
molecular weight'. This is in agreement \\.ith the
concept of complex formation i l l iotliiie-aromatic
hydrocarbon sj-stems R S postiilntctl h,v Reiiesi niicl
TABLE
SI
DIFFLWOS
OF I O D I N E IK .4RnAIATIC f I I . I ) K 0 P . 4 R B o S h
Viscniity
Ttiiip, Dq'T Iklisities x ,109,
Solvent C. X 10'0 of solyents ~ O I W

T i = Rolrite molar volume, cc./g. mole Benzene" 8.8G 4 38 ... 0,765

The molar \rolume of the solute at' its iiormal boiling 19.91 4.29 0.888 .G48
point is estimated by the atomic cont,ributions of Toluene" 8.FG :3.98 , . . . G75
19.01 3.93 0.878 ,588
Le Bas9s10for comples molecules. n: is an associa- iii-Syleiie" 8.90 3.71 , . . 745
tion parameter for the soh-ent having a value of 1.0 19.01 3.T4 0.872 G49
for unassociated liquids and va,lues of 2.6, 1.9 and Benzene' 25.0 4.33 ... . GO5
1.5 for water, methanol and ethanol, respectively. Tolueneb 25.0 3 . 9!) , . . 559
(7) C . R . M'ilke, C/ietit, Eriq. P r o y . , 46, 219 (194g). ni-xylellch 25.0 3.73 ... ,589
(8) C . R . \ V i I k e niirl Pin Clinrir, i)resenterl a t Nen T o r k Iiieeting,
.\inericnn I n s t i t u t e of Cheniical Engineers, Decenther, 19.54. RIesit>.Iene* 2 5 ,o :3 . 3!) , . . ,081
(9) Pei,ry, "('lirinical Engineer*' IIantll)ook," RlcGrsrv-Hill Book C. C. AIiller, Proc. Roy. SOC.(Lotidon),106, 724 (1924).
Co.. Inc.. Nt-n,T o r k . S. Y.,19.50. * B. R. Hnnimond and R . H . Stokes, Trans. Faraday SOC.,
(10) .I. H. .41noIcl, I n d . E n g . Chpm., 22, 109 ( l ( 1 3 0 ) . 49,880 (1953).
596 K. BRIL,S. BRILAND P. KRUMHOLZ Vol. 59

Hildebrand” and previously suggested to be a factor
in iodine diffusion by Hammond and Stokes.12
From the foregoing discussion it is evident that
a further theoretical approach is needed which
adequately accounts for specific interactions of
solvent and solute molecules to serve as a basis for
improvement of empirical relations such as equa-
(11) H. .4. Bcnesi a n d J. H. Hildebrand, J . A m . Chem. SOC.,71,
2703-2707 (1040).
(12) H. R. Hsinmond a n d R. H. Stokes, Trans. Faraday Soc., 49,
886 (1953).
tion 2. Although not entirely satisfactory, the
assumption of constanry in Dv/T appears to pro-
vide a fair representation of the temperature de-
pendence. Equation 2 is considered to be useful
as a first approximation for engineering design and
certain other purposes for u-hioh moderate pre-
cision may be sufficient.
Acknowledgment.-Assistance of Research Cor-
poration through a grant-in-aid is gratefully
acknowledged.

THE KINETICS OF DISPLACEME,R;T REACTIONS INVOLVING METAL

COMPLEXES OF ETHI’LENEDIAMINETETRAACETIC ACID
BY I<.BRIL,S. BRILA N D P. KRUMHOLZ
Conti ibution from the Research Laboratovu of Orquinia S.S. S h Paiilo, Btasil
Received January 11, 1966

The kinetics of the reversible exchange reaction PbY- +

Z n f f & ZnY= +
P b + + has been studied in acetate buffered
eolutions, a t an ionic strength of one, a t 25”, by means of the polarographic method using the streaming mercury electrode.
The exchange proceeds via a t least three simultaneous reaction paths. The proton-catalyzed reaction path is discussed
on the basis of the “stationary state approximation.” The equilibrium constant of the exchange reaction calculated from
the kinetic measurements is insa very good niiinerical agreement with the value of this constant as determined by the usual
equilibrium measurements.

Displa.cement reactions of the geiieral type1 perchlorate and the iron(II1) Enta They
Me(l, + Me(2,Y Mep) + MewY (1) found the acid induced dissociation of the complex
t o be the “rate-determining step” of this exchange
where Y represents the tetravalent anion of ethyl- in acid solutions.
eiiediamiiietetraacetic mid ( = Enta), are measur- It seemed to us interesting to investigate further
ably slow processes.2-10 Schwarzenbach and col- the mechanism of the exchange reaction 1 and in
leagues7studied in detail the kinetics of the reaction particular to examine how far the “quasi-equi-
between copper(I1) salts and the cadmium-Enta librium” hypothesis is justified. For this purpose it
complex. They found that this apparently simple is desirable to choose a system where both the for-
reaction proceeds actually via four different reac- ward and the reverse reaction 1 can be conveniently
tion paths measured.I1 The system
2Cd + CUY
ZnY + Pb PLY + Zn (3)

Cd + CuY + H
The reactions enclosed in the parentheses are as-
sumed t o be in a “quasi-equilibrium state” during
the over-all exchange process.
Long and colleagues studied the kinetics of the
eschange reaction 1 between radioactive iron(II1)
(1) I n this a n d in tile following equations ionic charges are omitted
for t h e sake of simplicity.
(2) F. A. Long, 9. Jones and h1. Burke, Brookhaven Conference
Report 2, Upton, N. Y . , Decemher 1, 19-18.
(3) R. Pribil a n d E. Vicenova, Chem. Lbsty, 45, 532 (1951).
(4) G. Schwarzenbach and E. Breitag, Helu. Chim. Acta, 3 4 , 1503
(1951).
( 5 ) C. h l . Cook, Jr., and F. A . Long, J . A m . Chem. SOC.,73, 4119
(1951).
(6) S.8. Jones a n d F. A. Long, THISJ O U R N A L56, , 26 (1982).
(7) H. dckermann a n d G. Schnwzenbacli, Helu. Chim.Acta, 35, 486
(1952).
(8) K. Bril a n d P. Krumhola, THIS JOURNA L 874
67, , (1953).
(9) K. Bril a n d P. Krumhola, % b i d . ,58, 339 (1954).
(10) G. Schwaraenbach, R . G u t and G. Anderegg, Helu. Chim. &la,
37, 937 (1954).
was found convenient for such study, and the re-
sults obtained are reported.
The polarographic technique using the streaming
mercury electrode as an indicator electrode of the
progress of the reaction was applied throughout
this work.
Experimental Part
Materials.-The preparation and the standardization of
the E n t s disodium salt, zinc nitrate and lead nitrate solu-
tions were described in a previous communication.8 The
solutions of the metal-Enta compleves contained an excess
of 0.5% of uncomplexed metal. The pH of these solutions
was adjusted to about 5 . 5 .
Apparatus.-The jet electrode arrangement, as well AS
the auxiliary equipment were described in a previous com-
munication .* For the kinetic measurements a simple time
registering device was used. Instead of taking galvanome-
ter readings a t given time intervals, the time intervals be-
tween suitably chosen galvanometer readings were deter-
mined. A disk of cardboard about 8 em. in diameter is
fixed o n the axis of an electric clock. When making a gal-
vanometer reading, one pushes down an ink writing pointer
(fixed upon a flexible arm), which leaves a mark on the paper.
This device allows the time to be determined with a precision
of about 0.2 see.
(11) I n t h e system studied b y Ackermann a n d Schwarzenbach,?
the reverse reaction 1 is almost negligible and enters only a8 a numeri-
cally very firnall correction term into the over-all rate expression.