served. The repeating pattern of 16 mass units for C,DZ,+I ions, as well as the abundant high mass peaks available a t higher temperatures, makes this material useful as a general purpose internal standard. Isobutane reagent gas produced a spectrum from the mixture of perdeuterated hydrocarbons which was similar t o that shown in Figure 2, with the exception that peaks corresponding t o C,DZ, or C,D2n-lH2 were of increasing intensity in the lower mass region, and were greater than CnDzn+lbelow C I ~ . The high fractional mass of deuterium (2.01410) results in very high fractional mass values of deuterioalkane ions (upper abscissa, Figure 2), which increase a t the rate of approximately 1.78 millimass units per amu of alkane mass. The large, favorable mass difference between marker and sample ions is illustrated in Figure 3, which shows a small portion of the high resolution spectrum of the nucleoside N6,Nf-dimethyladenosine.The ion of mje 178.1092 (5.3 relative intensity) is an adduct representing the base moiety plus a rearranged hydrogen from ribose and a methyl group from the reagent gas (5). The substantial mass difference between CBHI2NS and CllDZ3, even in this low mass region of the spectrum, assures that for virtually all organic ions there is no danger of overlap with deuterioalkane ions. Also, the lower groups of ions, such as CllHD22, CllD22-CllH2D21 (difference = 0.0015 mass unit), and CllD21, are potentially useful as additional standard mass peaks. (Lists of standard masses and for the minor ions C,D,,H and for the series CnD2n+l, C,D2, - to 4, are available from the authors upon request.) The use of perdeuterioalkanes as reference standards is ~
~
~
_
__-
( 5 ) M. S. Wilson, I. Dzidic, and J. A. McCloskey, Biocliim. Binplrys. Acta, 240, 623 (1971). _ _ _ _ _ _ _ ~ ~ ~
clearly not required in many instances, and unlabeled hydrocarbons may be equally satisfactory. However, since the mass difference between CnHln+land most organic ions is less than in the case of CnDln+l,resolution requirements will in general be higher. In addition, if CnH2n+lions are produced by the sample under investigation, they cannot be resolved from reference ions of the same composition arising from the hydrocarbon standard. As shown in the upper abscissa in Figure 2, the fractional mass of C,D2,+1ions reaches 1.0 amu a t approximately m/e 560. Therefore the region mje 650-730, which corresponds t o the fractional mass range 0.15 t o 0.30, becomes opaque with respect t o possible overlap between sample and reference peaks for many organic ions. The hydrogen content of the molecule and the resolving power employed then become important considerations, should perdeuterioalkanes be used in that region of the spectrum. ACKNOWLEDGMENT
The authors thank F. E. Montgomery for his assistance in the conversion of the instrument for chemical ionization; Mrs. N. R . Earle far calculation of standard mass values; and Drs. J. H. Futrell, M. S. B. Munson, a n d F . H. Field for their comments and discussions regarding instrument modifications for high pressure work. We thank Dr. Wojcik for a copy of his manuscript prior to publication.
RECEIVED for review June 1, 1971. Accepted July 19, 1971. This work was supported by the Robert A. Welch Foundation _ (4-125), the National Institutes of Health (GM-13901, N I H 69-2161, GM-02055) and computer facilities through NIH grant F R 259. ~
~
Pulse Polarography of Halide Ions in Molten Nitrates William O’Deen and R. A. Osteryoung Department of Chemistry, Colorado Stute Unicersity, Fort Collins, Colo. 80521
DEPOLARIZATION OF MERCURY by halide ions has been studied in both aqueous and fused salt solvents. Kolthoff and Miller found a two-electron reversible formation of mercurous halide if halide concentrations were kept below millimolar level in aqueous solutions ( I ) . They employed conventional dc polarography. Biegler employed both ac and dc polarography on this system and found evidence for an initial oneelectron faradaic step (2). A “halidomercury” product was also postulated as an intermediate occurring in calomel formation by Hills and Ives (.?). Bewick, Fleischmann, and Thirsk have concluded that calomel formation is a crystal growth process most likely involving a one-electron oxidation of mercury ( 4 ) . Similar studies have been carried out in molten nitrates. Swofford and Holifield employed conventional polarography
and found the halide depolarization of mercury t o be an irreversible two-electron process (5). However, by using linear sweep voltammetry, Francini, Martini, and Monfrini found the process t o occur via a one-electron reversible step (6). We have extended these studies in nitrate melts using integral (normal) pulse polarography in an attempt t o clarify these conflicting results (7). In integral (or normal) pulse polarography (I.P.P.), pulses of successively increasing amplitude are applied t o the electrode from a fixed base potential. With a dropping mercury electrode (DME), the fixed initial potential may be maintained for the life of the drop prior t o the time of the pulse application. In the case of a DME, the pulse is applied once during the life of a drop and is synchronized so that each pulse is applied t o a drop of the same area. Current is
( I ) I. M. Kolthoff and C . S . Miller, J. Amer. Clirrn. Sor., 63, 1405 ( 194I). (2) T. Biegler, J. Elecirouiiol. Cliem., 6 , 357, 365, 373 (1963). (3) G. J. Hills and D. J. G. Ives, J. Clirm. Soc., (Ln/rdo/i)1951, 311. ~. (4) A. Bewick, M. Fleischmann, and H. R. Soc., 58, 2200 (1962).
( 5 ) H. S . Swofford, Jr., and Charles L. Holifield, ANAL. CHEM., 37, 1513 (1965). (6) M. Francini. S. Martini, and C. Monfrini. EIectrochin~.Metal. 2 (3), 325 (1967). (7) E. P. Parry and R . A. Osteryoung, ANAL. CHEM.,37, 1634 (1965) (and references therein).
ANALYTICAL CHEMISTRY, VOL. 43, NO. 13, NOVEMBER 1971
1879
I
MELABS I.i?f? OF B r 10
p' II
I
-I CA)
T= 255OC
/
C = 0.50 mMOLAL T=240'C WIDTH = 100 mSEC
20
30
- 0.4
- 0.2
CONCENTRAT I ON (mMOLAL)
- 0.6
E (VI
Figure 1. Anodic pulse polarogram of bromide at a dropping mercury electrode in fused NaN03-KN03 at 240 "C. Melabs instrument
Figure 3. Diffusion current us. concentration of bromide at a mercury electrode in the NaN03-KN03melt
2.5
. ... /,.......... ....
COMPUTER 1.P. F! OF Br-
2 .o
T.240 ' C
1+Logt(i,-iP/i7
1.5 .*
3001-
500
.
C = 4 . 0 0 mMOLAL T=242'C PULSE WIDTH = 10 mSEC.
0.5
1. - 0.2
- 0.4 E (V)
- 0.6 -0.30
Figure 2. Anodic pulse polarogram of bromide at a hanging mercury drop electrode in fused NaN03-KN03 at 242 "C. Computer system measured just prior t o pulse application and again at some fixed time ( 5 t o 100 milliseconds) after pulse application. The difference in the current is read out. Appropriate modifications can be made if stationary electrodes are used (8). The technique is effectively sampled chronoamperometry. EXPERIMENTAL
Apparatus. Integral pulse polarography was performed using two different systems. A Melabs Pulse Polarographic Analyzer employing a dropping mercury electrode with a fixed pulse width of 100 milliseconds was used in some instances. A computer controlled system, employing a hanging mercury drop, was used in those instances where a variation of pulse width was advantageous. The computer hardware and software have been described elsewhere (8). The potentiostat, through which the computer-generated functions were fed to the cell, was a solid state version of a multipurpose instrument similar in design to one described by Lauer and Osteryoung (9). A Moseley 7004A X-Y
..__ _ _ _ _ ~ ___. (8) H. Keller and R . A. Osteryoung, ANALCHEM.. 43, 342 (1971). (9) G. Lauer, H. Schlein, and R . A. Osteryoung, ibid.. 35, 1789 (1963).
1880
1.0
ANALYTICAL CHEMISTRY, VOL. 43,
NO. 13,
-0.35
-0.40
E (VI Figure 4. Plots of E L;S. log (current function) for bromide depolarization of mercury in fused NaN03-KN03 at 240 "C recorder was used. The cell was a 4-cm diameter borosilicate glass tube sealed at one end. Fusion was carried out in a pot furnace of conventional design. The melt temperature was measured with a mercury in glass thermometer and controlled t o *2 "C by manual adjustment of a variable transformer through which line power was applied t o the heating coils. Electrodes. A Melabs D M E Electrode Stand was used with the Melabs Analyzer and a Kemula-type hanging mercury drop electrode with the computer system. The area of the hanging drop employed was 0.0357 cm2. The counter electrode in all cases was a platinum wire immersed directly in the nitrate melt. The reference electrode was Ag/AgNOa (0.06M) N a N 0 3 - K N 0 3contained in 7-mm glass tubing and separated from the bulk nitrate melt by a fine porosity sintered glass frit. Chemicals. The alkali metal halides were reagent grade material dried in vacuum and stored in a desiccator until used. The nitrates were reagent grade chemicals (Mallinckrodt) which gave clean polarographic background when used without purification. Technique. Sixty grams of a n equimolar mixture of N a N 0 3 and K N 0 3 was placed in the cell and fused. The melt was bubbled with pre-purified nitrogen t o dry it, although traces of water did not appear t o have any marked
NOVEMBER 1971
-
Table I. Half-Wave Potentials for Bromide Depolarization of Mercury Concentration of bromide (millimolal) Eli2 (volts)
..., 10 -
-0.323 -0,326 -0.326 -0.326 -0,324 -0.323 -0.325
0.5 1.o 2.0 3.0 4.0 5.0 6.0
-I -
GA)
-0,298
C = 1.50 mMOLAL
30 -
Table 11. Pulse Polarographic Half-Wave Potentials and Diffusion Coefficients for Halide Depolarization of Mercury in Fused NaN03-KN03 at 245 “C El y (volts) 1’s. Ag/AgNOa (1 M ) Diffusion coefficientsa Halide ( 1 Molar) cm*-sec-l x lo6 1-0.709 6.76 i 0.21 Br-0.453 6.99 f 0.21 Cl-
CONVENTIONAL POLAROGRAM
- 0.5 - 1.0 E (VI Figure 5. Conventional anodic polarogram for iodide at DME in NaN03-KN03melt 0.0
6.91 & 0.36
D values based on eight measurements on each of three different solutions, except for 1-. Deviations are at the 95% coiifidence limits. (1
effect o n the halide polarograms. The alkali metal halide was introduced directly as a solid. Stirring with a stream of nitrogen aided the dissolution of the halide and ensured homogeneity. Analysis began once the temperature reached a steady value. RESULTS
Figures 1 and 2 show typical integral pulse polarograms of bromide obtained with the Melabs instrument and the computer systems, respectively. Figure 3 shows the linearity of the limiting current with concentration. Similar straight line plots were observed for chloride and iodide. The log plots in Figure 4 show clearly that the wave represents a reversible one-electron step with the formation of a soluble product. The 0-value calculated from the slope of the straight line is 0.97 rt 0.03 t o the 9 5 z confidence interval. The slope is theoretical for a one-electron process. In contrast, the twoelectron soluble or insoluble log plots show marked curvature. Similar results are observed for both chloride and iodide. Table I shows half wave potentials for the bromide depolarization of mercury obtained a t various concentrations of bromide. The constancy of Eli? with increasing bromide concentration is again consistent with a one-electron reversible production of a soluble product. A cathodic shift of El,?of 50 millivolts per decade increase in concentration would be expected for the two-electron production of soluble mercurous or mercuric halide. A cathodic shift of E112of 100 millivolts per decade increase in concentration would be expected for the two-electron production of insoluble mercurous o r mercuric halide. Similar results were obtained for chloride and iodide. El/?for bromide was also unaffected by changing pulse width, again indicating reversibility. Diffusion coefficients and Eli2values at 245 “C are given in Table 11. The Eli? values are with respect t o the reference Ag/AgNOa ( 1 M ) . The diffusion coefficients for chloride and bromide were calculated using the Cottrell equation and substituting in the computer pulse width, the concentration of halide, the dift‘usion current, and the area o f t h e hanging drop. The diffusion coefficients represent a n average obtained for
three different solution preparations for each halide and the average of eight separate runs for each solution. Solutions containing iodide gave less consistent results than those containing bromide and chloride. A calculation of the diffusion coefficient of iodide by plotting the diffusion limited current, Id, cs. the inverse of the square root of the pulse width gave a least squares slope which yielded a diffusion coefficient cm2/sec. This is the value which appears in of 6.76 X Table IT. The intercept of the I d L’S. l/t1’2least squares line intercepts the Id axis at f 4 FA, whereas in theory the intercept should be zero. The dependence reported above on the inverse square root of time indicates a diffusion controlled process. DISCUSSION
The results obtained in this work appear t o be more in agreement with those of Francini, Martini, and Monfrini than with Swofford and Holifield-that is, we find a reversible oneelectron formation of a soluble product. Problems associated with employing conventional polarography, the technique of Swofford and Holifield, may be seen in Figure 5 which shows a conventional polarogram for the mercury depolarization by iodide obtained at a DME. The polarogram shows a marked maximum and the current is significantly larger on the maximum than would be expected for a simple one- or two-electron wave. An arrow on the current axis indicates the current (6 FA) expected for a one-electron wave using the previously cited diffusion coefficient for iodide. A pre-wave appears t o exist a t -0.7 volt which is close to the half-wave potential of the integral pulse polarographic one-electron wave as indicated by the arrow o n the potential axis in Figure 5. It is possible that this “pre-wave” i s the start of the reversible one-electron wave which, however, becomes obscured by the much larger current maximum. However, any attempt to use the current maximum a s the limiting current of the faradaic step and t o obtain information from log plots would lead t o potentially meaningless results. Attempts t o measure a “half-wave” potential would also be fruitless in that the erratic nature of the wave would clearly result in marked shifts with concentration. This would explain the unreasonable positive shifts in potential found by Swoford and Holifield for increasing halide concentration (5). The ability to obtain well defined waves of halides under these conditions with integral pulse polarography is probably
ANALYTICAL CHEMISTRY, VOL. 43, NO. 13, NOVEMBER 1971
1881
due to the ljmited charge passed in this technique. The charge passed during the life of a 6-second drop held on the diffusion plateau of a wave in normal polarography is almost an order of magnitude greater than that for the same 6second drop subjected t o a 50-millisecond pulse t o the diffusion plateau a t the end of drop life. Presumably calomel formation takes place when too large a concentration of product is formed, resulting in deposition and maximum formation. The fixed 100-millisecond pulse width of the Melabs system required the use of concentrations below 1 millimolal in order t o prevent maximum formation. The variable pulse width of the computer system allowed the use of halide concentrations up t o 7 millimolal, since shorter pulse times mean less charge is passed per pulse. The integral pulse polarograms were slightly more well defined at larger concentrations and, therefore, the computer system was used t o obtain the half-wave potential and diffusion coefficient data for the halide electrooxidations. More difficulty was encountered in obtaining well defined iodide polarograms, as
indicated above, probably because of the lower solubility of mercurous iodide. The pulse polarographic method applied to the study of filmed electrodes may well be a useful general technique. Activity is under way in our laboratory to study systems, both aqueous and nonaqueous, which give erratic and ill-defined waves where conventional polarography is employed. This technique also would appear to be better than linear sweep voltammetry for similar studies in that, again, the amount of charge passed would be smaller than that in conventional sweep experiments unless one employed very high sweep rates. This would, of course, necessitate large corrections for charging current, whereas most charging current problems are eliminated by the pulse polarographic technique.
RECEIVED for review May 24, 1971. Accepted July 19, 1971. This work was supported in part by a Grant-in-Aid of the Faculty Improvement Committee of Colorado State University,
Steric-Exclusion Chromatography at Pressures up to 3500 Kilograms per Square Centimeter B. A. Bidlingmeyerl and L. B. Rogers Department of Chemistry, Purdue University, Lafayette, Ind. 47907
THEPURPOSE of this study was t o demonstrate the feasibility of doing steric-exclusion chromatography at pressures very much higher than are conventionally used. This exploratory investigation utilized bovine plasma albumin on a column of porous glass beads. The native state of a protein is generally a compact, ordered form in water. The existence of this structure is believed to be due to the interaction of water with the hydrophobic side chains of the protein ( I , 2). Depending upon the magnitude of the pressure, protein denaturation may be retarded or accelerated (3,4). Pressure can also significantly influence the aggregation of macromolecules as has been demonstrated in ultracentrifugal studies ( 5 , 6 ) . Past investigations into pressure denaturations have generally relied upon classical methods to analyze the resulting solutions after pressure had been released (7-9). In addition t o the usual parameters such as duration of the pressure and Present address, Standard Oil Company (Tnd.), Standard Oil Research Center, P. 0. Box 400, Naperville, 111. 60540 (1) I. M. Klotz, Scieuce, 128, 815 (1958). (2) E. Wicke, Angew. Chem., Itit. Ed. E/ig/., 5, 106 (1966). (3) K . Suzuki and K. Kitamura, Reu. P h y ~ .Client. Jupati, 29, 81 (1960). (4) Ibid., p 86. (5) G . Kegeles, L. Rhodes, and J. L. Bethune, Proc. Nut. Acad. Sci. US.,58, 45 (1967). (6) L. F. TenEyck and W. Kauzmann, ibid., p 888. (7) K . Suzuki and Y . Miyosawa. J . Biochent. (Tokyo), 57, 116 (1965). (8) Y . Miyagawa, K. Sannoe, and K. Suzuki, Arch. Biochrm. Biopltys., 106, 467 (1964). (9) K. Aoki, K. Hirarnatsu. M. Tanaka, and S. Kaneshina, Biochim. Biophys. Acta., 160, 368 (1968). 1882
type of analysis, this approach introduced another experimental parameter, the time between pressure release and analysis. One report by Aoki et a/. ( 9 ) has described a boundary curve for the denaturation of bovine plasma albumin with respect to temperature and pressure. Their procedure was to expose the protein (1 solution) to a specified pressure and temperature in tris-boric acid buffer (pH = 8.9). After thirty minutes, the pressure was released and gel electrophoresis was performed t o determine if any portion of the protein had denatured. The pressure-temperature points where any portion of the protein had denatured defined the boundary curve for denaturation. The authors pointed out that the boundary was not absolute, but depended upon the sample of bovine plasma albumin used. The denatured form was of a higher molecular weight, and the authors suggested that it was mostly a dimer. The present communication describes the application of high-pressure steric-exclusion chromatography ( I O ) in a brief investigation of the modification of bovine plasma albumin under pressure. A qualitative comparison with the results of Aoki and coworkers is presented. EXPERIMENTAL
The Fraction V bovine plasma albumin was purchased from Armour Chemical Corp. (Chicago, Ill.) and was lot number D27309. The sample was defatted by the charcoal (10) B. A. Bidlingmeyer and L. B. Rogers, “High-pressure Liquid Chromatography.” paper 93, Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, Cleveland, Ohio, March 1971.
ANALYTICAL CHEMISTRY, VOL. 43, NO. 13, NOVEMBER 1971
