J. Phys. Chem. 1985,89, 3302-3304
3302
TABLE VIII: Distances and Asymmetry Parameters for CO$ (298
K) 0-- -0
C=O
1.15998 1.16249 1.16525 1.16421
re/A rz/A rg/A
rJA ,q10-7
~3
5.1 (2)
(3) (4) (5) (5)
2.31996 2.32498 2.32512 2.32379
(6) (8) (9) (9)
shrinkage 0 0 0.00538 0.00463
5.0 (2)
“See Appendix and also ref 42 for definition of the parameters. equilibrium are calculated from the potential constant^,^' us,kb, and kdd, by using the first-order perturbation based on the wave functions of harmonic oscillators. The Fermi resonance between v I and 2u2 is taken into consideration by the method described in ref 20. The local z axis for the C = O pair is taken along the equilibrium positions of the atoms, and its p axis perpendicular to z . The average values of these Cartesian coordinates, ( Az) etc., are derived from those of the dimensionless normal coordinates by making a linear transformation. On the other hand, the average values of the curvilinear coordinates are obtained from the expansion series of the Cartesian coordinatesZoas listed in Table VII. The uncertainties in the average values are estimated to be less than 0.5%, judging from the mean deviations of the experimental data used in the determination of the potential constants in ref 37. The average values reported in ref 20 are found to be essentially correct.
The r,, r,,, and r, distances and K parameters are calculated from the above average values and the equilibrium C=O distance according to their definitions given in ref 42. A small contribution from the centrifugal distortion effect is also included in the calculation.22 The results are listed in Table VIII. The present calculation shows that the cubic and quartic mean values of the displacement coordinates, listed in Table VII, do not contribute significantly to the final average structures. In fact, the r, distance of the C = O bond calculated by use of the linear and quadratic mean values only is 1.16421 A, the difference from A. Therefore the the correct value being only less than 1 X fifth- and higher-order terms are negligible. The results of the present calculation are essentially identical with those reported by Hilderbrandt and The r,(C=O) distance reported by Murata et ai.,” 1.1646A, is slightly too long. Nevertheless, the structural parameters of molecules calibrated by the r, distance reported by Murata et al. are not biased significantly, because the difference is well within the limit of experimental error of gas electron diffraction. Registry NO. CjH6, 75-19-4; C3D6,2207-64-9.
Supplementary Material Available: Experimental data of the total intensity for cyclopropane obtained by gas electron diffraction (2 pages). Ordering information is given on any current masthead page. (42) K. Kuchitsu, Bull. Chem. SOC.Jpn., 40, 498 (1967).
Composition of Bromine Hydrate George H.Cady Department of Chemistry, University of Washington, Seattle, Washington 98195 (Received: March I, 1985)
Samples of bromine hydrate have been produced by slow condensation of a known weight of water vapor at constant pressure in the presence of bromine vapor, at constant pressure, upon a surface at a constant temperature near 0 ‘C.The hydration number, n, in the formula, Brz.nH20, has then been calculated from the weight of water and the weight of bromine in each sample. Many values have been obtained, most of them within the range 7.8-8.6. I am of the opinion that the higher values have resulted from incomplete conversion of water to hydrate. I believe that the correct value is somewhat less than 7.9.
Even though the crystalline substance, bromine hydrate, has been known since 1829,’ its composition is still a controversial matter. Many determinations have been made by several methods, and values ranging from 6 to 12 have been reported for the hydration number n in the formula, Brz.nH20. The subject has been reviewed by Dyadin and Aldako2 who themselves used Schreinemaker’s method to study the system B2-Hz0-KBr at 0 OC and concluded that their data together with those of others who have used the method indicated the existence of different hydrates having the hydration numbers 12.0, 10.0, 8.3, and 7.0. Others who have determined the composition have been of the opinion that the crystals are of only one type. I agree with this and feel that the variety of hydration numbers found by direct analysis of the crystals or by analysis of mixtures of crystals with aqueous solutions using Schreinemaker’s method results from errors in analysis. The analysis involves several difficulties which include both the volatility and the reactivity of bromine. A method which has been used to obtain approximately correct values for hydration numbers of many gas hydrates is that of de F ~ r c r a n d . ~To use this method, Mulders4 made careful mea(1) Lbwig, C. Ann. Pogg. 1829, 16, 376. (2) Dyadin, Yu.A.; Aldako, L. S . Zh. Strukt. Khim. 1977, 18, 51-57; Rum. J. Struct. Chem. 1977, 18, 41-47. (3) de Forcrand, R. C. R. Hebd. Seances Acad. Sci. 1902, 134,835, 991. (4) Mulders, E. M. J. Thesis, University of Delft, 1937.
0022-3654/85/2089-3302$01.50/0
surements of pressure vs. temperature for the system Br2-H20 at temperatures somewhat below 0 OC when the phases present a t equilibrium were ice, solid hydrate, and vapor (Br, + H 2 0 ) . This information gave the enthalpy, AH, for the reaction Br2-nH20(s) Br2(g) nH,O(s)
-
+
She also made similar measurements at somewhat higher temperatures when the phases at equilibrium were solid hydrate, liquid solution of Brz, and vapor (Br2 H20). This information gave AH for the reaction Br2.nH20(s) Br2(g) nH20(l)
-
+
+
When the difference between the two values of A H was divided by the enthalpy of fusion of a mol of ice, a value of 8.36 was found for n. On the basis of X-ray crystallography of many gas hydrates Stackelberg and Miilles classified these substances into two types, structure I and structure 11. Bromine hydrate was thought to be a structure I solid with the cubic unit cell having an edge length of about 12.0 A.5 A subsequent study of the structure by Allen and Jeffrey showed the substance to be tetragonal with the cell constants, a = 23.8 A and c = 12.2 A.6 A detailed structure determination was not made, but Allen and Jeffrey considered (5) Stackelberg, M.v.;Mailer, H. R. Z . Elektrochem. 1954, 58, 25. (6) Allen, K. W.; Jeffrey, G. A. J . Chem. Phys. 1963, 38, 2304.
0 1985 American Chemical Society
Composition of Bromine Hydrate certain properties of the crystals to be “good presumptive evidence” that the host lattice of bromine hydrate was isostructural with that of tetra-n-butylammonium salt hydrates. From this they reasoned that the hydration number of bromine would be 8.6. Persons working with bromine hydrate in recent years have favored this value as probably correct. Additional evidence showing that bromine hydrate is not cubic, and therefore not structure I or structure 11, is found by using polarized light. When the crystals are observed while held between crossed Polaroid plates they appear bright, like ice, not dull like structure I or structure I1 hydrates. The observation is made in a darkened room while looking directly into a beam of light passing through the crossed polaroids and the vessel containing the crystals. I have been determining the composition of various gas hydrates over a period starting in 1976.7-10 I have used procedures in which a known weight of water vapor condenses in the presence of the gas at substantially constant pressure to form the solid gas hydrate upon a surface at a constant temperature near 0 OC. The solid hydrate forms directly from the gaseous phase. From time to time during this period, attempts were made to determine the hydration number of bromine, but in general the results did not agree well. Values ranging from about 7.8 to 8.6, sometimes higher, were obtained. It was not possible to have a sufficiently high pressure of bromine vapor to cause the method to work satisfactorily a t -0.3 “C, the quadruple point, a pressure of 44 mm is required to make the hydrate stable. The vapor pressure of bromine at this temperature is only 66 mm. In many of the experiments, conversion of water to the hydrate was incomplete. An additional potential source of error was the possible presence of water soluble impurities formed by the reaction of bromine with Halocarbon stopcock lubricant or other material in the vacuum line for handling bromine and water. An impurity such as hydrogen bromide could cause part of the water to remain in the liquid state and not be converted to bromine hydrate. Many variations of the usual procedure were tried, and numerous determinations were made by quantitative synthesis of the hydrate from bromine and liquid water or ice. The latter method usually gave hydration numbers in the range 8.4 to 8.6. Eventually it was found that synthesis of the hydrate from water and bromine vapors at temperatures several degrees below 0 “ C did give substantially reproducible hydration numbers.
Experimental Section My best series of experiments with bromine hydrate will now be described. Bromine of high commercial quality was held in contact with phosphorus pentoxide in a glass flask equipped with a Fischer Porter valve having a Teflon stem. Many portions of bromine had previously been withdrawn. Impurities more volatile than bromine should not have been present. Water was added to the reactor as liquid, and air was removed by vacuum before the reactor containing water was weighed. The reactor and container for bromine were then attached to a short and clean vacuum line using only two ground joints. These were lubricated with Halocarbon stopcock lubricant which is largely polymer of CF2= CFCl. The line was evacuated using a diffusion pump in which the oil was a polymer of CF,=CFCl. During transfer of the bromine, the short vacuum line was closed off from the main line with a Fischer Porter valve having a Teflon stem. The amount of bromine added to the reactor was sufficient to convert all of the water in the reactor to hydrate and to also fill the reactor with bromine vapor to a pressure of about 66 mm. In addition to this, there was an excess of about 0.05-0.1 g of bromine. This caused bromine vapor to be present in each run at the vapor pressure corresponding to the coldest spot in the reactor. All weighings were made using a counterpoise of the correct volume and with patience to wait for constant weight. As materials (7) Cady, G. H. J . Fluorine Chem. 1978, 11, 225. (8) Cady, G. H. J . Phys. Chem. 1981, 85, 3225. (9) Cady, G. H. J. Phys. Chem. 1983,87,4437. (10) Cady, G. H. J . Chem. Educ. 1983, 60, 915.
The Journal of Physical Chemistry, Vol. 89, No. 15, 1985 3303
I I
8 cm
Figure 1.
were added, the following weighings were made: (1) empty reactor, (2) reactor plus water, and (3) reactor plus water, plus bromine. Three reactors of the type shown in Figure 1 were used so that runs could be made in triplicate. The valve on each reactor was a Fischer Porter metering valve. During a run, at the time when the reactors were being chilled, the valve was tightened frequently to prevent leakage. Such adjustments were made over a period of about 30 min. The materials were first condensed as solid at the low end of each reactor and the vessels were than placed in a box in a household freezer or refrigerator to give an approximately constant temperature of 0 OC or below. A small electric heater was also placed in the box to warm the bottom portion of each reactor. This caused the bromine and water to evaporate slowly. The vapor condensed as crystals of bromine hydrate near the top of each reactor. Part of the bromine also condensed either as liquid or as solid, depending upon temperature. The distillation required a week or more. In some of my experiments a tiny portion of liquid remained at the bottom of the reactor at the end of the distillation. It had the color of bromine water, but was reluctant to evaporate. This behavior suggests the presence of a water soluble impurity, perhaps hydrobromic acid. In the series of runs now being reported, the droplet of liquid, if present, was too small to be seen. At this stage of the run, the experimental objective was to measure the weight of bromine hydrate. This objective was accomplished by removing the excess liquid bromine from the reactor and then weighing the material remaining. The weight of vapor was then subtracted from the total to obtain the weight of hydrate. To remove the excess liquid bromine, the reactor was held at 0 OC while the liquid distilled into a receiver containing a supply of bromine and water at about -2.5 O C . Before opening the valve of the reactor, the vacuum line was filled with vapor from the receiver. A period of only a few minutes was sufficient to allow the excess liquid bromine to distill from the reactor. A gas density bulb at 0 OC and of about 100 mL volume was now filled with vapor from the receiver. Pressures were equalized in the bulb and
3304 The Journal of Physical Chemistry, Vol. 89, No. IS, 1985 TABLE I Hydration Number of Bromine wt of wt of temp of formation, O C water, g bromine, g -22 -22 -22 -6 -6 -6 -1 -1 -1 -22 -22 -22
0.5405 0.5195 0.7562 0.5405 0.5195 0.7562 0.5405 0.5195 0.7562 0.5405 0.5 195 0.7562
0.6155 0.5843 0.8566 0.6115 0.5898 0.8383 0.5960 0.5769 0.8 194 0.61 32 0.5808 0.8458
hydration
no. 7.79 7.89 7.83 7.84 7.8 1 8.00 8.04 7.99 8.19 7.82 7.93 7.93
the reactor, their valves were closed, and the vessels were later weighed at room temperature. The observed vapor density was used to calculate the weight of vapor in the reactor, while a t 0 OC, and this weight was subtracted from the total to obtain the weight of bromine hydrate. It was then possible to calculate the composition of the hydrate. In the series of runs reported in Table I the cold box for the first set of data was at about -22 OC. For the second set the box was at about -6 O C , and for the third set the temperature was about -1 OC. Finally the last set was again at about -22 OC. After each run, a little bromine was added to each reactor before making the next run. Three of the four sets of data are in substantial agreement and give an average value of 7.87. However, the set at about -1 OC has a somewhat higher average, 8.07. I believe that the higher value probably resulted from a small part of the water remaining in the liquid state. If dissolved impurities caused liquid to remain, the effect should become larger with increasing temperature. After completion of the runs, tests were made to learn whether a bromide might be present as an impurity. Before doing this, bromine gas was removed by evacuating each reactor repeatedly by using an aspirator attached to a water faucet. Repeated cycles of admitting air and then removing it with the aspirator were used until no color of bromine could be seen in the reactor. The solution was then poured from the reactor into a 10-mL volumetric flask, and enough rinse water was used to bring the volume of solution to 10 mL. Half of the solution was used to obtain its pH. The other half was treated with silver nitrate solution, and the resulting silver bromide was weighed. In the order in which data for the
Cady three reactors are listed in Table I, the observed pH values were 3.25,3.04, and 3.08, respectively, and the observed weights of silver bromide were 0.003,0.002, and 0.001 1. The last of these numbers is high enough to suggest that a little bromine had not been removed by the aspirator. The data make it appear that a little hydrobromic acid probably was present, but that its concentration in each reactor was probably not over 0.006 M. In another set of runs, using other portions of water and bromine, the materials were caused to distill from the bottom at about 3 O C to the top portion of each reactor a t 0 O C . The data corresponded to hydration numbers of 8.32, 8.36, and 8.27. When the same samples of materials were used, but with the top portions of the reactors a t -22 O C and the bottom portions somewhat warmer than -22 OC, the data corresponded to hydration numbers of 7.93,7.97, and 7.83, respectively. Apparently the conversion of water to hydrate was more nearly complete at -22 O C than at 0 OC.
Conclusion My conclusion is that the correct value for the hydration number of bromine near 0 OC, with the pressure of bromine about that existing over liquid bromine, probably is a little less than 7.9. If bromine were to form a structure I hydrate in which only the larger of the two types of cavities could be occupied, the hydration number expected for the conditions just mentioned would be 7.79. It appears, therefore, that the hydration number for bromine probably is that expected for structure I, but that the structure is not structure I. The tetragonal unit cell found by Allen and Jeffrey6 has essentially the same size and shape as two adjacent cubic unit cells of a structure I hydrate. Since bromine hydrate is unique, it is desirable that a complete structure determination be made. Acknowledgment. Acknowledgment is made to the donors of the Petroleum Research Fund, administered by the American Chemical Society, for the partial support of this research. I also thank D. W. Davidson of the National Research Council of Canada for his friendly and helpful interest. Since this is likely to be my last scientific publication, I wish to thank those persons who were very influential in getting me started as an experimental chemist. The two most influential were my father, Hamilton P. Cady of the University of Kansas, and my thesis professor, Joel H. Hildebrand of the University of California at Berkeley. Registry No. Br2.nH20, 16053-02-4.
