FORMATION OF HOI
June 5 , 1955
AND
HYDRATED IODINE CATIONBY HYDROLYSIS OF IODINE 2957
substantially first order in total iodine, first order in iodide ion and zero order in Mn(I1) as appears from the summarized results, Table 11. If we postulate a mechanism analogous to reactions 2-5 above it would follow that kobs
= ko(1
- wm)
13-
and log
kob.
= log
l b . One actually finds such a linear dependence (not shown) but pthermal = 11 whereas llphoto = 1.4 by extrapolation to GOo. An alternative interpretation of these results employs the primary process
ko - p f
px1-'/2
where kobs is the first-order rate constant with respect to iodine, ko is the "actual" rate of dissociation and 1 - W , is the iodide ion dependent probability of non-recombination described by equation
evidently can be simplified by ignoring the term ( 12- . ) ? and
ill
rate = k7(Ia-)(I-) AND
Mn(I1)
4.4 2.6 2.7 2.6 2.5 1.3 0.77 .79 .75
.40 .40
.40 .40 .20 .10 1.2 3.0 .10 1.2 2.8 .049 0.76 Stoichiometric concentration.
1.1 1.1 1.1
a
0.59
3.5 5.8 2.9 3.0 6.1 3.1 6.0
2.3 2.1 2.1
(7)
--+ k7 ( I ~ - ) ( I - )= ~ ( I * - . ) + Z k4(r2-.)(Mn(II))
TABLE I1
THERMAL REACTIONBETWEEN 1 3 ADDEDI- AT 60"
+ I - +I*-. + I*-'
If the reaction of Iz-. with Mn(I1) is as efficient as has been postulated for the photochemical reaction, then the steady-state relation
WITH
7.3 6.4 6.8 6.6 6.2 6.7 7.7 7.9 15
The test of this rate law appears in the fifth column of Table 11. Except that the point a t 0.049 M I - is discordant for each of these treatments, either method furnishes an adequate empirical description of the results. Since we are unable to account for the observed difference between slopes of the photochemical and thermal diffusion functions we prefer mechanism (7). Acknowledgment.-We are grateful to Dr. R. hi. Noyes for helpful criticism, and for an opportunity to read an unpublished manuscript. We greatly appreciate several helpful discussions with Professor J. L. Magee and Dr. L. Monchick. NOTREDAME, INDIANA
[CONTRIBUTION FROM THE
DEPARTMENT O F CHEMISTRY,
UNIVERSITY OF CALIFORNIA,
DAVIS]
The Formation of Hypoiodous Acid and Hydrated Iodine Cation by the Hydrolysis of Iodine BY T. L. ALLENAND R. M. KEEPER RECEIVED JANUARY 10, 1955 By means of spectrophotometric analyses of aqueous solutions of iodine it has been possible t o determine the equilibrium constant for t h e reaction Ip(aq) -I- HnO(1) = HOI(aq) H*(aq) + J - ( a q ) . This constant is 5.40 X at 25.0' and 0.49 x at 1.6'. It has also been possible t o set a n upper limit of 1 X for the equilibrium constant at 25.0' of the reaction Iy(aq) HsO(1) = HzOI*(aq) I-(aq). The thermodynamic properties of HOI(aq) are calculated.
+
+
+
Aqueous solutions of iodine exhibit absorption maxima in the neighborhood of 288, 3-52 and 462 i ~ i p . ~The . ~ first two of these peaks are attributed to triiodide ion and the longest wave length peak is due to solvated iodine. The triiodide ion in these solutions is a result of the reaction (eq. 4) of iodine with the iodide ion produced by impurities or by hydrolytic reactions of iodine (eq. 1, 2, 3). The rate of production of hypoiodous acid is known to be
+ + $- 51- + OH'
In(aq) f H?O = H,OI+ II?(aq) f E120 = H01 f H + I3Iz(aq)
+ 3Hn0
103-
12(aq)
+ 1-
= 1s-
(1) (2)
(3) (4)
very rapid.3 I t has been postulated that the hydrated iodine cation (H201i-)is a precursor4 of hypoiodous acid. In any case it would be possible to form the hydrated iodine cation by a rapid addi1. 1. Custer and 5. Natelson, Anal. Chem., 81, 1005 (1919). (2) I,. I. Katzin, J . Chem. Phrs., 21, 490 (1953). (3) H. A. Liebhafsky, THISJOURNAL. 63. 2074 (1931). (4) K . J. Morgan, Quarf. K u o , 8 , 123 (195-1).
(1)
tion of hydrogen ion to hypoiodous acid, SO that the rate of establishment of the equilibrium represented by eq. 1 must be rapid. Keactioii 4 is also quite fast. Although the equilibrium constant for the formation of iodate ion (eq. 3) is extremely small,6 the reaction would go far enough to produce appreciable amounts of iodide ion in solutions of PH > 2. However, the reaction is slow in acid solutions.6 From the rate law given by Bell and Gelles,6 it may be shown that in the presence of the iodide ion formed by reaction 2 the rate of reaction 3 is inversely proportional to the square of the hydrogen ion concentration. Since equilibria (l), (2) and (4) are established in freshly prepared iodine solutions and equilibrium (3) may be neglected, then, if there is no iodide ion present because of impurities, the concentration of iodide ion produced must be equal to the sum of ( 5 ) W. 0.Lundberg, C. S. Vestling and J. E. Ahlberg, THISJ o t J R 69, 264 (1937). ((9 K. P. Bell and E. Gelles, J . Chem. Soc , 2734 (1951).
NAI.,
the concentrations of hypoiodous acid' and the hydrated iodine cation
partment was swept out coiitiiiually with ZL streann of dry, chilled nitrogen. It was noted t h a t the conceiitration of I., - iiicrcabcti with (1-1 (18-1 = (HzOI') 4- (HOI) time, as evidenced by a n increase in t h e optical densities at and 352 mp. The rate of increase was very low except By utilizing the equilibrium constants for equations 288 in the solutions of PH > 4.8. In the less acid solutions the 1, 2 and 4, it is possible t o eliminate the terms rate of change increased m:irkedly with incrrasing p H . I t (HzOI+), (HOI) and (I-) from equation 5 and to is prohahle that the incrcasc in 1:.- concentration \mi caused by a slow reaction of iodiiic ivith trace.; of rctlucing agent.; rearrange the equation to give equation c i in the buffer solutioiia, it-hich in the region of higher p H waq accompanied by the hl-drolysis of iodine to produce iodate (reaction 3 ) . For solution.; of PH > 4.8, the optical t1en.ities mere extrapolated to the time of mixing the buffer solti.I11 concentrations are expressed in moles per liter irhon tetrachloride solutioii. entratioiis werr calculated from t h e and >)*is the mean activity coeficient of the ions krlowri concentrations of perchloric acid, acetic acid ami and is assumed to be the same for all the univalent acetate ion, and the ionization constant of acetic acid a t the ions present. The concentration of iodine may be temperature of the s o l u t i ~ n .The ~ activity coefficients of the determined from the optical density of the freshly ions were taken t o be the same as t h e activity coefficients prepared iodine solutions a t 462 m p where the ab- of hydrochloric acid in a solution of 0.01 ionic strength a t the appropriate t e m p e r a t ~ r e . ~As a check on the calculated sorption is almost entirely due to solvated iodine. hydrogen ion concentrations, t h e ,bH of each solution wa.: The concentration of triiodide ion may then be de- measured with a Beckman pH meter. Calculated and ohtermined from the optical density at 288 or 352 m p served values were in good agreement, with a n average (leafter subtracting the absorption due t o iodine. It viation of 0.02 pH unit. the wave lengths of the absorption maxima atit1 is then possible to evaluate the right side of eq. (i t h eAlthough extinction coeficients a t these wave lengths for aqueouq and plot the values obtained against the reciprocal iodine and I:,- have been determined by ilwtrey and Conof the hydrogen ion concentration. The resulting nick,'" it was considered desirable t o check the values they plot should be a straight line whose intercept would obtained. Essentiall>-the same procedures were followed, except t h a t one cm. cells were used. In determining the he K l and whose slope would be K2.$ extinction coefficients of iodine, it was necessarl- t h a t the concentration of I { - be inade suKicientl>-small t h a t its light Experimental absorption was negligible compared t o t h a t of iodine. This In order t o remove any iodide impurities in the iodine, was accomplished by using solutions containing apprecinble concentrations of H ~ and - IOa-, which removed I - and 1, solutions of iodine in carbon tetrachloride were prepared by the reverse reactions of equations 3 and 4 . From the and washed repeatedly with conductivity water. The most probable values"6 for K l and K.3, it may be shown that iodine solutions and conductivity water were placed in glass for the reaction stoppered flasks, and t h e flasks were rotated in a thermostat a t 25' for several hours. Then the aqueous phases IiH' 10321. 2H2O = SH,OI(5') were pipetted off and fresh conductivity ivater was added. t h e equilibrium constant, Kj, is 6 X For the range This washing process served t o remove both iodide and of concentrations of 12, IO3- and H-employed in these soluother water-soluble impurities in the iodine and also to retions, most of the IOs- was reduced t o HsOI'. In calcumove iodide formed by reaction of iodine with a n y traces of lating the extinction coefficients of IP. it was necessarl- to reducing agents in t h e carbon tetrachloride. neglect any light absorption by the H t O I I , as its extiiictioii The aqueous phase in each flask was then replaced irith coefficients are not known. In order t o determine whether one of ten buffer solutions which were of the same ionic or not this procedure was introducing appreciable errors, the strength ( p = n . O l ) , and which ranged in pH from 2.04 t o initial concentration of 10,- w : t ~ varied over a fourfold 5.67. The buffer solutions were prepared from perchloric range. This resulted in no appreciable change i l l the calacid, sodium perchlorate, sodium acetate and water. In the case of the more acid solutions (pH of 2.04 and 3.01) it1 culated extinction coetiicieiits. T h e water used in all experiments was redistilled from which the rate of oxidation of iodide ion by oxygen might alkaline permanganate solution. Eastman Spectro Grade have been appreciable, t h e air was displaced by bubbling carbon tetrachloride ~ r a sredistilled, and a middle fraction nitrogen through t h e solutionr. The flasks were covered of b.p. range 75.9-76.2' (uncorrected) \viis used. Othrr with aluminum foil t o exclude light, shaken vigorously for chemic:ils were of a C.P. or Reagent grade. several minutes, and then rotated in the thermostat for approximately one hour. During this period the iodine Results and Discussion distributed between the two phases, and t h e equilibrium concentration of IJ- in the aqueous phase, resulting from reThe wave lengths of the absorption maxima :m(1 ;Letions 1, 2 and 4, was established. the molar extinction coefficients a t these maxima arc Samples of the aqueous phase were withdrawn by forcing tabulated in Table I, along with the results of Xwthe sample into a pipet with compressed air or nitrogen. trey and Connick. ''I The agreement between the The samples were placed in one-cm. ground-stoppered quartz cells, and t h e optical densities of the solutions were measured two sets of results is well within the probable esperiwith a Beckman type D U spectrophotometer. Optical mental error. tlensities were corrected for light absorption by the cells TABLE 1 and buffer solutions. The cell compartment of the spectrophotometer was MOLAREXTISCTIOS COEFFICIESTS OF Ia-(aq) ASD I~(:Ic]) equipped with a w&ter jacket through which water from t h e This w r k h w t r e y :in(! Cotillirk 2A.O' thermostat was circulated. I n three exyriments Wa\-e a'avc after measurement of the optical densities a t 25.0 , the cell length length Id imp) !mi& I. I, I? temperature was reduced to 1.6" by circulating ice-water through the jacket, and t h e optical densities were determined 270 .... 121 2'70 17,200 121 a t t h a t temperature. T o prevent condensation of moisture 288 40,000 CJ? 287.5 40,000 95 oil the cell windows at t h e lower temperature, the cell com352 26,400 20 333 26,100 18 (7) 'The ionization constant of HOI is so small (4.5 X lo-'? t h a t in 462 1,030 '742 4ti0 973 7%
+
( s i )
+
acid solutions the concentration of 0 1 - is negligible See 4.Skrabal, H e r . , 7SB, 1.570 (19421. (81 Katzin? has determined K I from spectrophotometric measureiiierits of aqueous iodine solutions in which sufficient acid was present i o repress reaction 2. However, he obtained evidcnce t h a t hi.; values were too large because of iodide impurities.
+
+
(9) H. S. Harned and B. B. Owen, "The Physical Chemistry of IElertrolytic Solution-," Iieinhrtl(l Piil,l Corp., S e w York, AT. I-, II143, 11. ,580, 5-17. (101 A . I ) . Awtrey m i d R . li Connick, T H I Sl o i m u i , , 73. lX4:' (l!l.;l)
June 5, 1955
FORMATION OF HOI
AND
HYDRATED IODINE CATIONBY HYDROLYSIS OF IODINE 2950
From these extinction coefficients and the observed optical densities, the concentrations of 1 2 and 11- in the buffer solutions were calculated. At the concentrations of I- present in these solutions, its light absorption is negligible. It was necessary to neglect any light absorption by HOI or HzOI+, since their extinction coefficients are not known. The concentration of Is- was calculated from observations a t each of the two Is- absorption peaks. The values thus obtained agreed within the experimental error, indicating t h a t neglect of light absorption by HOI and H20I+ caused no serious error. The best value of 1 3 - was taken as the average of the two independent determinations. The most probable values for Kq were considered to be those calculated from the equation relating this equilibrium constant and the temperature, which is given in the paper by Davies and Gwynne." The function K1 Kz/(H+) was then calculated from the experimental results at the various hydrogen ion concentrations. Average values for experiments within the range of 1/(H+) from 1 X 102to 12 x l o 3 are shown in Fig. 1, and those a t higher values of l / ( H + ) are shown in Fig. 2. The accuracy of the experimental points is greater for those experiments shown in Fig. 2. At low values of 1/ (H+) (high acidity) reaction (2) is sufficiently repressed that most of the light absorption a t the 1 3 absorption peaks is caused by 1 2 , and against this relatively high 1 2 background it is difficult to determine Is- with accuracy. For those experiments shown in Fig. 2, K, is negligible compared to KZ/(H +). Therefore Kz may be determined by multiplying the value of K, K2/(H+) for each experiment in this range by (H+). This procedure has the merit that the experiments may be equally weighted. The value of Kz as determined in this fashion is (5.40 0.25) X a t 25.0" and (0.49 0.03) X a t 1.6". The slopes of the straight lines in Figs. 1 and 2 were made equal to these values. The intercept of the straight line in Fig. 1 was adjusted to give the closest agreement with the several experimental points. Because of the small size of the intercept (K,) and the uncertainties in the experimental points it is not possible to determine the intercept with accuracy. However, one may set an upper limit to Kl of 1 X a t 25.0". Since in Fig. 2 this intercept is negligibly small, and since i t would be expected to be even smaller a t 1.6", the straight lines were drawn to pass through the origin. It may be seen that the points lie close to the straight lines. I n order to determine K1, solutions of iodine in aqueous perchloric acid of higher acidity were then investigated, using 10-cm. cells for determination of the ultraviolet spectra. Samples were analyzed approximately one-half hour after mixing the iodine-carbon tetrachloride solution and the perchloric acid solution t o decrease errors from the slow increase in triiodide concentration mentioned previously. Results are shown in Table 11. At these low concentrations of 13-, the smaller extinction coefficient of 1 2 at 352 mp makes this wave
71
I
I
I
I
I
+
OY
I
0
2
I
I
I
6
8
I
12 I / ( H + ) x 10-3. dependence of K , K,/(H+) on the re4
10
+
Fig. 1.-The ciprocal of the hydrogen ion concentration at high acidity and at 25.0".
+
*
*
(11) M. Davies and E. Gwynne, THISJOTJRNAL, 74, 2748 (1952).
n
0
*,
I
1
I
I
8
16 24 32 40 48 I/(H+) x 10-4. Fig. 2.-The dependence of K I K,./(H+) on the reciprocal of the hydrogen ion concentration a t low acidity: 0 , 25.0'; 0, 1.6". 0
+
length more favorable than 288 mp for determination of 13-. Values of (I3-) in Table I1 were accordingly obtained from observations a t 352 mp. The perchloric acid solutions were prepared from either Mallinckrodt A. R. 60% HClO4 (M) or Baker Reagent 60% HC104 (B), and were either air-saturated or nitrogen-saturated. The presence of oxygen made no difference in the results, a t least within the experimental error. This is probably due to the low rate of oxidation of iodide ion by oxygen a t these low iodide concentrations. The source of the perchloric acid made an important
T. L. ALLEN.4ND R. hf. KEEFER
2960
difference. The 1.0 A f HC104 solution prepared from Baker HClO4 had a much lower triiodide concentration than the comparable solution prepared from Mallinckrodt HC104. This could be caused by a reducing impurity in the solution prepared from Mallinckrodt HC104, or an oxidizing impurity in the solution prepared from Baker HC104. Since the 0.1 Jd HClOb (hI) had only about one-tenth as high a value of (I3-) as the 1.0 M HCIOl (M) it appears t h a t the former explanation is correct. [Vhether the impurity was present in the original reagent or was introduced in the preparation of the solution was not established. However, the evidence is sufficient to indicate that the triiodide concentration for the first solution of Table TI is in c'rror by a large amount. TABLE I1 ANALYSES OF SOLUTIONS OF HIGHACIDITY (HCIOI), mole:l.
(1%)x 103, tnolc/l.
1.0 (M)
1.13
0 . 1 (M)
1.15 1.09
1.0 (B)
Unfortunately the triiodide concentrations for the second and third solutions are too small t o be determined with accuracy. The iodine background makes up 95 and 90%, respectively, of the light absorption a t 352 mp. Therefore it may be concluded that K , is too small to be determined by this method. It is possible t o set an upper limit to K I d I x lo-'" from these experiments. Values12J..9previously reported for K1 a t 25" are 1 X lo-*, 1.2 x 10-11, and a group of values ranging upwards from 0.52 X lo-*. The only one of TABLE I11 TEEEQUILIBRIUM CONSTANT FOR REACTION 2 K 2 x 10"
1, o c .
Method
5.4
25.0
This investigation This investigation Conductivity of aq. iodine s o h . " Conductivity of aq. iodine so1ns.b Kinetic studiesc Iodometric titration error.d Results probably in error due t o neglecting formation of 11p H of iodine s01ns.'~
0.19
1.6
3 0.09 11
25
1.5
20
1.6
25
25
0
these which is less than the upper limit given above and i t is therefore taken as the most is 1.2 X probable value of IC1. The value of 5.4 X for K z obtained in this investigation is in good agreement with the value13 calculated from the fiH of aqueous iodine solutions. Reported values of K , are summarized in Table 111. 'The dissociation constant of the cationic acid H201+,according to the reaction II?C)I.'
TV. C. Brav anti E. L. Connolly, THISJOURNAL, 33, 1186 (1911). b G. Jones and M . L. Hartmann, ibid., 37, 211 (1915). c A . Skrahal, Z . Elektrochem., 17, 665 (1911). d E. Angelescu and Y. D. Popescu, 2. p h y s i k . Chenz., A156, 301 (1931). (12) 11. D. XIiirray. J . Chem. SOL.,127, 882 (1925).
=
FI+
+ NO1
is given by the ratio h72!Kl. &kt95.0" this is 4.5 X 10-2, Froin the equilibrium constant z t 23.0 and 1.6" the thermodynamic constants of reaction 2 have been calculated assuming that A I P is temperature independent in the range from 1.(ito 2 5 " .
+ +
~ f % = ~ ~ (16.71 . ~ ~i 0.03) kcal./mole AH0zss.,6 = (16.7 =t0.6) kcal./mole AS02as.ls = (0 i 2 ) cal./deg. mole
Ua~o~,~,O'p
4.9 0.5 0.9
Vol. 77
The standard state for all substances is taken as the hypothetical ideal state of unit molarity. By combining these constants with known thermodynamic properties14 of other substances which enter into reaction 2 one may obtain the thermodynamic properties of HOI(aq.1 AFt029a,lti= -23.67 kcal./mole
. I I I ~ " ~=~ ~-33.2 . ~ G kcd./111ole A . S ~=~ -32 ~ ~cal./deg. . ~ ~ i~iole SOuqs.ls
=
22 cal./deg. mole
The heat of formation may be compared with the results of Skrabal'j and Skrabal and Buchta16 who determined the heat of the reaction 12-
f OH-
KO1
+ 21-
Combination of their results with known thermodynamic properties of the other substances14 gives heats of formation for HOI(aq) of -34.7 and - 36.3 kcal./mole, respectively. Acknowledgment.-It is with pleasure that the authors acknowledge the financial assistance of a research grant from the Sational Science Foundation. L)avrs,
CALIFORNIA
(13) G. Horiguchi and H. Hagiwmn, H u l l . I N b f . P'hys. Chem. K r search ( J a p a n ) , 22, 6Gl (1943). (11) F. D. Rossini, el a l . , "Selected Values of Chemical Thermodynamic Properties," Xational Biireau of Standards Circular ROO i1952). (15) A. Skrabal, h f o n o f s h . , 39, 99 (1912). (le) A. Skrabnl and I? R u c h i n , i h i , l , 55, 697 (1914).
