3046
NOTES The measurement of reaction rates may be useful as the third method of investigating the molecular state of strong electrolytes, in addition to Raman intensities and nuclear magnetic resonance. Acknowledgment. This work was supported by the United States Atomic Energy Commission under AEC Contract No. W-7405-eng-48.
C,,ld tmoles/liter 1
Figure 1. The acidity function, H,,logarithms of the activity, C ~ E ~ S Oand ~ , logarithms of the initial reaction rate: b, HzSOc (Monger and Redlich); 0, HzS04 (present work); 0,&SO4 (present work).
HzS04 is the intermediate between an instantaneous and a subsequent slow reaction. It is, of course, true that in a case of this type the kinetic formalism, based on activities, is the same for a trimolecular reaction of H+, HSOd-, and HzOz and for a bimolecular reaction of Hi304 and H2.02. However, a classical bimolecular reaction progress with a specific reaction rate increasing almost by lo6 for an increase of the sulfuric acid concentration from 5 to 12.5 mol/l. appeared to be so interesting that some further experimental support was considered desirable. The results are shown in Figure 1. The logarithms of the initial reaction rate with 1 mol of HzOz/l.a t 25.0" as found now are compared with the results of Monger and Redlich. The slope of the logarithm of the activity of sulfuric acid2 and the slope of the acidity functiona -Ho are also shown. The kinetic results (logarithmic slope value, 0.64) are in reasonable agreement with the slope of 0.71 of the activity. The acidity function with a slope of 0.52 between c = 5 and 12.5 fits the data less well. More recent data4 for Ho (slope = 0.58 between c = 9 and 12.5) would fit in the smaller range about as well as the activity. Results obtained in heavy solutions (D/(H D) = 0.86) agree within the experimental error limits with those from the light solutions. This result supports the assumption that the hydrogen ion is not rate controlling. (The activity function of D2S04 has the same values between c = 5 and 12.5 as HzS04.' Resultsa for DNOa indicate' that deuterium substitution does not appreciably change the strength of any strong acid.) Naturally one may call the variation of the initial rate by a factor of 106 a solvent effect. However, it appears to be preferable to avoid this term in an example which can be explained by a simple and perfectly reasonable model and to reserve the term "solvent effect" to cover those cases for which it has been invented, namely, where no simple model is applicable.
+
The Journal of Physical Chemistry
(2) (a) J. I. Gmitro and T. Vermeulen, A.1.Ch.E. J.,10, 740 (1964); (b) J. I. Gmitro and T. Vermeulen, Lawrence Radiation Laboratory Report UCRL-10886 (1963). (3) M. Paul and F. A. Long, Chem. Rev., 57, 1 (1967). (4) M. J. Jorgenson and D. R. Hartter, J . Amer. Chem. Soc., 85, 878 (1963). (6) E. Hogfeldt and J. Bigeleisen, ibid., 82, 16 (1960). (6) 0. Redlich, R. W. Duerst, and A. Merbach, to be submitted for publication. (7) G . N. Lewis and P. W. Schutz, J. Amer. Chem. Soc., 56, 1913 (1934); 0. Halpern, J. Chem. Phys., 3, 466 (1936).
Gas-Phase Far-Ultraviolet Absorption Spectrum of Hydrogen Bromide and Hydrogen Iodide'
by B. J. Huebert2 and R. M. Martin Department of Chemistry, University of California, Santa Barbara, California 08106 (Received February 18, 1068)
Quantitative data for the continuous absorption spectrum of HBr and H I are of interest for tho determination of excited-state potential energy curves and transition strengths, as well as for analytical applications.3~~ Extinction coefficients for these absorptions have been determined previously by photographic spectrophotometrya4,6 I n the course of photochemical studies with HBr and HI, we have measured th,e extinction coefficients for HBr from 2300 to 1700 A and for HI from 3000 to 1800 A, using a Cary 15 double-beam photometric spectrophotometer. The results are given in Table I and Figure 1, and comparisons wit,h previous data are shown in Table 11. HI was prepared from 55% aqueous hydriodic acid (Merck reagent grade) using standard vacuum techniques6 and was stored in the dark at - 196". HBr was prepared from 48% aqueous HBr (Mallinckrodt reagent grade) and was stored in the same way. Spectra were taken in an Aminco optical cell with Suprasil quartz: (1) Partial support of this work by a grant from the National ScienafP Foundation is gratefully acknowledged. (2). National Science Foundation Undergraduate Research Participant, 1966. (3) R. S. Mulliken, J . Chem. Phys., 8, 382 (1940). (4) J. Romand, Ann. P h m . (Paris), 4, 627 (1948). (6) (a) C. F. Goodeve and A. W. C. Taylor, Proc. Roy. XOC., A152, 221 (1936); (b) C. F. Goodeve and A. W. C. Taylor, ibicZ., A154, 181 (1936). (6) R. M. Martin and J. E. Willard, J. Chem. Phys., 40, 2999 (1964) a
3047
NOTES windows which had a path length of 10,000 cm. The cell was connected to the vacuum line via a stopcock greased with Kel-F 90, and the gas pressure was measured with a Wallace and Tiernan FA141 diaphragm gauge. A comparison of the calibrated gauge readings with a mercury manometer gave agreement to within
dynode setting at 4 and thz sensitivity at 4, it was possible to go down to 1700 A with slit widthsoof 2 mm, corres onding to a band width of about 5 A. Above 1800 these conditions gaveonarrower slit widths and band widths of less than 1.2 A, so that for the continu-
w
Table I : Molar Extinction Coefficients for HBr and HI" -,-HBr-
HI
, €,
8,
a
1. mol-1 cm -1
1. mol-1 cm -1
2300 2250 2200 2150 2100 2050 2000 1950 1900 1850 1800 1760b 1700
20 -I: 3 30 =t3 62 =k 3 102 z!z 3 154.5 228 319 428 534 633 694 710 665
X,
e,
0.8 0.9 2.0 2.2 3.2 3.5 2.0 1.3 5.0
--r
6,
x, B
1. mol-1 om-'
1. mol-] om-'
3000 2900 2800 2700 2600 2537 2500 2400 2300 2215a 2100 2000 1950 1900 1877c 1850 1800
7.4 13.6 24.9 45.3 78.1 106.8 124.3 170.3 203.5 212.9 196.0 164.1 146.1 133.5 130.8 134.2 207.0
0.1 0.1 0.2 0.1
0.8 0.6 1.2
0.9 1.0 0.5 0.7 0.4 0.9 0.5 0.8 1.2 2.0
" The 2150-2300-A values for HBr are based on a single determination and maximum errors are given; the other absorbancy indexes are mean values and 6 is the average deviation from the mean. * Maximum. ' Minimum.
Table I1 : Comparison of Absorption Data for HBr and H I
-----------
HBr 1. mol-1 om-1-This Ref 5a Ref 4 work
----
,-e,
X,
A
2300 2150 2000 1850 1785" 1760"
23 102 324 576
Maximum.
78 240 463 5 25
20 102 319 633 710
HI 7 1. mol-1 om-l-This Ref 5b Ref 4 work
-e,
X,
b
3000 2700 2500 2215" 2150" 2000 1890b 1877b
6 41 102
182
7.4 45.3 124.3 213 174 140 118
164
131
' Minimum.
.t0.05 torr over the range 0-30 torr, with an error of less than 0.02 torr below 10 torr. The spectrophotometer was purged with nitrogen having specified impurity levels of less than 5 ppm (total), less than 2 ppm of oxygen, and less than 2 ppm of water. With the Cary 15
260
240
220 A,
200
180
mw
Figure 1. Absorption of gas-phase HBr and HI.
ous spectra studied there is no significant band-width error. The wavelength scale was checked at high resolution with a medium-pressure mercpry lamp and was found to be accurate to within 2.0 A within the 30001800-A wavelength range studied. The absorbancy was also checked with a 0.500 absorbance neutral density filter and was found to be accurate to within 0.003 absorbancy unit over this range. The HBr spectra were measured at 23-25" and pressures of 2.41-3.00 torr, and the concentrations were calculated assuming ideal-gas behavior. The data given in Table I for HBr are based on three to four independent determinations, except for the 2300-2150-A values noted. HI spectra above 2000 A were measured at 23" and 6.00 torr, whereas those below 2000 were measured at -78" after filling the cell with 16.00 torr at 23". The low-temperature measurements were made with Dry Ice packed around the cell in a Styrofoam cell holder. This was done in order to eliminate the small amount of gasphase Izresulting from thermal decomposition after the sample was introduced into the cell. This decomposition did not result in a measurable change in the HI absorption, and the 1 2 concentration was not sufficient to give observable absorption in the 5000-8 Iz-band region, However, the I2 resulted in apparent extinction :oefficients for HI which were high by 10% at 1900 A and 12% at 1850 1. As expected, no significant error due to 1, was observed at 2000 1. 1, has a band spectrum in this region owith peak separations of approximately 60 cm-l, or 2 A. The error due to Izis consistent with our approximate measurements of average Iz extinction coVolume 7 s Number 8 August 1968
3048
NOTES
efficients of 450 1. m21-l cm-I at 4900 8, 1.0 X lo41. Volume Changes on Mixing Solutions of mol-l cm-l at 1900 A, and 1.0 X lo4 1. mol-1 cm-1 at Potassium Halides and Symmetrical 1850 8 in the 0.050-0.500 torr pressure range. There was no observed effect of pressure on the average exTetraalkylammonium Halides. Evidence for tinction coefficient. Based on these figures, only 0.3% Cation-Cation Interaction. A Correction of the H I would ne$d to decompose t o give the 10% error noted in the 1900-A region. The data for HI in Table and Further Comments I are based on three indepenient measurempts above 2000 8 and two below 2000 A. The 2000-A value inby Wen-Yang Wen, Kenichi Nara, cludes all five measurements, which were in excellent Chemistry Department, Clark University, agreement in spite of the temperature and pressure Worcester, Massachusetts 01610 differences. In Table I1 our results are compared with some of the and R. H. Wood values of extinction coefficients from previous work. Chemistry Department, University of Delaware, For HBr there is satisfactory agreement with the Newark, Delaware 10711 (Received February $1, 1068) Goodeve and Taylor results down t o 2000 8, but at lower wavelengths our values are higher. Their lowest measurement was at 1890 8 and we have taken the I n the original article on this subject,’ the volume of 1850-8 value from their curve through the data. mixing, Amvex, was calculated in milliliters per mole, Romand’s results are approximately 25% lower than while the excess volume, Vex,of a pure electrolyte was ours for HBr. The discrepa!cy is slightly increased if calculated in milliliters per kilogram of solvent. The the curves are shifted 25 A to bring the observed comparison of the two volumes based on eq 5 and 6 of maxima into agreement. ref 1 is only correct if both volumes are in milliliters per Goodeve and Tayl2r made only three measurements kilogram of solvent. The correction affects only Table with H I below 2500 A and failed t o observe the maxiVI1 of ref 1 and the revised results are given in Table I. mum and minimum in the absorption curve. Our This correction does not appreciably affect the concluresults are somewhat higher than theirs above 2500 8 sions of Wen and Nara. The agreement is not as good and are l0-20% higher than Romand’s at lower wavea t low concentrations, but it is better at high concentralengths. The better agreement of Romand’s value at tions. However, with the corrected data, their arguthe minimum may be due to the effect of iodine. ment can be made even stronger. Wen and Nara conclude that cation-cation pairs are responsible for the Bayliss and Sullivan’ have reported molar extinction coefficients for I, of 2.9 X lo4,1.3 X lo5,and 1.5 X 1051. volume changes because of the agreement of the comparison of AmVex/[12y(l - y ) ] (eq 5 ) with [V“(AX) mol-’ cm-l at 2040,1960, and 1890 8,respectively, with Vex(BX)]/12(eq 6) at low concentrations and because a band width of 10 8. We have found values of
