0
20
40
hr'ie sec
0
60
Figure 2. Effect of mixing on the IgG/anti-lgG reaction (0 - 0 . reaction mixed for 1 sec, 2 sec prior to each reading: 0 - 0, no mixing during the reaction. Antiserum dilution 1:20 in PEG:lgG concentration in the reaction mixture. 0.66 mg/dl)
proteins of serum, urine, or cerebrospinal fluid. Using the laser-modified parallel fast analyzer system, kinetic data were collected on the IgG/anti-IgG reaction. Regular mixing of the solutions was necessary t o prevent settling of large antigen-antibody complexes, and the effect of mixing on the IgG/anti-IgG reaction is shown in Figure 2. A significant difference in both the apparent rate and the shape of the curves is evident. Investigations carried out with the IgG/anti-IgG system using other instrumentation provided data which were consistent with the rate and shape of the curve observed when the solutions were mixed prior to reading (8). The effect of mixing on the within-run precision of the rate measurement for 56 replicate samples representing 4 separate runs yielded a RSD of 3.8%. These data demonstrate that the reproducibility of kinetic measurements of immunochemical reactions carried out with this instrument are not adversely affected by the slight settling of the antigen-antibody complexes. Changes in light scattering with time for different concentrations of IgG are shown in Figure 3. From this, it can be seen that measurable changes in intensity are observed for IgG concentrations as low as 0.13 mg/dl in the reaction mixture as early as 55 sec after initiation of the reaction.
20
40
tlve sec
60
Figure 3. Change in light scattering with time for different coneentiations of IgG (4) 0.66 mg/dl. (3) 0.33 mg/dl, (2) 0.22 rng/dl, and (1) 0.13 mg/dl; antiserum
dilution 1:20 in PEG
CONCLUSION This report describes the modification of a commercially available fast analyzer system to allow the performance of light scattering measurements and has been applied to the kinetic measurement of the IgG/anti-IgG reaction. This type of kinetic measurement of immunochemical reactions has considerable potential for the measurement of many specific proteins in biological fluids, since results could be obtained rapidly with the consumption of only small quantities of antiserum. The incorporation of an ultraviolet laser into the system would also make possible rapid and precise fluorescence measurements. This could represent a means by which the time required for quantitation of enzymatic activity could be significantly shortened, thereby providing for increased throughput in a clinical laboratory situation. ACKNOWLEDGMENT The authors thank Richard Jarnagin and Steve Peterson (Chemistry Department, The University of North Carolina a t Chapel Hill) for the use of their laser during the initial stages of the experimentation.
RECEIVED for review April 23, 1974. Accepted July 10, 1974.
(8) J. Savory. G. J. Buffone, and R. Reich, Clin. Chern.(in press).
Temperature-Dependent Response of the Glass Electrode in Deuterium Oxide R. Ken Force and James D. Carr Department of Chemistry, University of Nebraska, Lincoln, Neb.
68508
Several investigators have examined the response of the glass electrode in deuterium oxide in an attempt to establish a pD scale equivalent to the p H scale in protium oxide. I t has been observed that a correction factor must be added to the pH meter reading to Obtain a that is equivalent to the activity of the deuterium ion in deuterium oxide. Co-
vington, Paabo, Robinson, and Bates ( I ) of the National Bureau of Standards have determined the correction factor to be 0.41 (molar scale) or 0.45 (molal scale) a t 25 "C for 2 (1) A. K. Covington, M. Paabo, R. A. Robinson, and R G. Bates, Anal. Chern., 40, 700 (1968).
ANALYTICAL CHEMISTRY, VOL. 46, NO. 13, NOVEMBER 1974
2049
Table I. Response of the Glass Electrode at Different Temperatures for Basic Solutions in H 2 0 and D20. p = 0.100 PKH20 = 13*997 Base
T = 25.0 'C
Acidity, m
pKDZO = 14.955
Meter reading
A ~ D molal , scale,
concentration, m x 103
-log
c
-log H+
c
D+
A calcd
in H 2 0
in D 2 0
A exptl
11.530 11.542 11.517 11.847 11.856 11.833 12.023 12.042 12.045
12.020 12.025 12.000 12.330 12.344 12.327 12.517 12.521 12.530
0.490 0.483 0.483 0.483 0.488 0.494 0.485 0.479 0.485
0.469 0.476 0.476 0.476 0.471 0.465 0.474 0.480 0.474
0.465 0.490 0.475 0.535 0.500 0.505 0.510 0.505 0.500
0.477 0.452 0.467 0.407 0.442 0.437 0.432 0.437 0.442
0.480 0.495 0.490 0.500 0.505 0.485
0.431 0.431 0.436 0.426 0.421 0.441
4.581
11.657
12.616
0.959
9.160
11.958
12.917
0.959
12.134
13.093
0.959
13.74
pKHZO = 13.680
4.581
11.341
12.283
0.942
9.160
11.642
12.584
0.942
11.818
12.760
0.942
13.74
~
9.160
13.74
K = H 13.396 ~
11.358
12.284
0.926
11.534
12.460
0.926
T = 35.0 'C
< pD < 9. A number of earlier workers (2-5) have obtained values within experimental error of the value obtained by the workers at the National Bureau of Standards. Fife and Bruice (6) have examined the temperature-dependent response of the glass electrode in acid solution of DzO by using DC1 solutions. However, a literature search indicates that no one has previously examined the temperature response of the glass electrode at other pD values. In this work, the temperature-dependent response of the glass electrode in high pD solutions is examined and correlated with the work of Fife and Bruice. EXPERIMENTAL Preliminary experiments showed that with very dilute solutions of deuteroxide or hydroxide ion, absorption of carbon dioxide from the air was a serious problem and measurements were nonreproducible. Therefore, all further work which led to the data reported herein was carried out in a glove box under a nitrogen atmosphere. Deuterium oxide (99.8% D20, Stohler Isotope Chemicals) was made free of dissolved carbon dioxide by refluxing under a stream of nitrogen for two hours and transferring t o a storage bottle while hot and stored in the glove box. Deionized, distilled protium oxide was boiled vigorously for two hours, transferred to a polyethylene storage bottle while hot, and kept in the glove box under nitrogen. (2)P. K. Glasoe and F. A. Long, J. Phys. Chem., 64, 188 (1960). (3)R. Lumry, E. L. Smith, and R. R. Glantz, J. Amen Chem. Soc., 73, 4330 (1951). (4)E. Mikkelsen and S. 0. Mielsen. J. Phys. Chem., 64, 632 (1960). (5) P. Saiomaa, L. Schaleger, and F. A. Long, J. Amer. Chem. SOC.,86, 1 (1964). (6)T. H. Fife and T. C. Bruice, J. Phys. Chem., 65, 1079 (1961).
2050
pKDZ0 = 14.622
11.170 11.175 11.165 11.495 11.500 11.510 11.695 11.700 11.690 T~ = 45.0 "C
A calcd - ~ e x p t l
PKD 2
11.230 11.215 11.230 11.400 11.405 11.410
11.635 11.665 11.640 12.030 12.000 12.015 12.205 12.205 12.190 = 14.322
11.710 11.710 11.720 11.900 11.910 11.895
Reagent grade potassium chloride (Mallinckrodt Chemical Company) was recrystallized twice from water, dried for two hours a t 100 OC and stored in a desiccator over P205. Standard solutions of NaOD and NaOH were prepared by dissolving reagent grade NaOH pellets (minimum NaOH 98%) in C02free deuterium oxide and protium oxide, respectively, and standardized by titration with primary standard potassium hydrogen phthalate. The use of NaOH in D20 leads to an atom fraction of 2.77 X of 'H in D20. All emf measurements were made with a Corning Model 1 2 expanded-scale pH meter, which has sensitivity to 0.2 mV. Comparison with a Dial-A-Volt precision voltage source showed that this electrometer was reliable to 0.2 mV (including measurements involving polarity change). Two different glass electrodes (both Corning No. 476022) were used. One electrode had been in use for several months in this laboratory; the other was new. Both electrodes gave essentially identical values in duplicate experiments. A Corning saturated calomel electrode (No. 476002) was used as the reference electrode. Before the determination of pD values, both the glass and reference electrodes were pre-equilibrated for 1 2 hours a t the temperature to be employed. The pH meter was standardized a t each temperature with standard phosphate and borax buffers in protium oxide. Working solutions in protium oxide were prepared in 25.00-ml volumetric flasks; deuterium oxide solutions were prepared in 5.00-ml volumetric flasks. The concentration of the different solutions of sodium hydroxide and sodium deuteroxide used in the measurements are given in Table I and were equal to each other in each set of measurements.
RESULTS AND DISCUSSION Meter readings (as apparent pH) for three pairs of equimolal solutions in protium and deuterium oxide at temperatures of 25, 35, and 45 "C at an ionic strength of 0.100, controlled with potassium chloride, are given in Table I.
ANALYTICAL CHEMISTRY, VOL. 46, NO. 13, NOVEMBER 1974
Table 11. Response of the Glass Electrode at Different Temperature for Basic Solutions in H 2 0 a n d D 2 0 at High Ionic Strength ( p = 2.00) T = 25.P 'c'
Base
Acidit), m
A pD, mol31 scale,
kletrr readiii!l
concentTation,
- log c
x 103
13.74
:I-.
-log c
A calcd
in li,@
in D20
13.093
0.959
12.020 12.015 12.015
12.520 12.515 12.505
0.500 0.500 0.490
0.459 0.459 0.469
12.205 12.190 12.220
0.505 0.505 0.515
0.437 0.437 0.427
11.940 11.930 11.940
0.313 0.525 0.520
0.413 0.4C1 0.406
D-
12.134
A
exptl
2 colcd - A e x y t l
T = 3 5 . 0 'C
13.74
11.818
12.760
0.942
11.700 11.685 11.705 T = 45.0
13.74
__
11.534
12.460
0.926
C
11.427 11.405 11.420
______
-
-
Table 111. Average ApD at Different Ionic S t r e n g t h and Temperatures u = 0.100
Temp,
~v A ~ D , "molal s i d l e
C
* *
25.0 35.0 45.0
0.47 0.01 0.44 0.01 0.43 i 0.01 'I
= 2.00
25.0 35.0 45.0 a
0.46 i 0.01 0.43 i 0.01 0 . 4 1 i 0.01
Figure 1. Plot of ApD vs. lo3/ T. 1.1 = 0.100 (KCI) ( A ) measurements from Ref. 6. (0) measured in this work.
Uncertainty quoted is one standard deviation.
Autoionization constants of both protium and deuterium oxide as a function of temperature were taken from Covington, Robinson, and Bates ( 7 ) and are given in Table I. Meter readings for one pair of equimolal solutions in protium and deuterium oxide a t temperature of 25, 3,5, and 45 "C at a n ionic strength of 2.0, controlled with potassium chloride, are given in Table 11. The analytical concentration of hydrogen ion and deuterium ion, CH+and CD+,respectively, were calculated from the autoionization constants for protium and deuterium oxide, respectively. and from t h e known hydroxide and deuteroxide conrentrat ion. The average correction factors as a function of temperature are given in Table 111. T h e correction factor (ApD) of 0.47 a t 25 "C is in good agreement with the value of 0.45 f 0.03 determined by Covington et al. ( I ) . Gary, Robinson, and Bates (8) found a correction of 0.467 for the phosphate buffer in D20 at 25 "C. A least-squares plot of ApD measured in this work LIS. 1/ T , shown in Figure 1,yields the following expression:
ApD =
2.02
x 102
T
- 0.204
where T is the absolute temperature. T h e values at high
temperature. indicated by 1 in Figare 1. are from Fife and Hruice (6).The measurements of' Fife and Rruice were reported as molarities, which can be converted to molalities by addition of log do t o the correction factor, where do is the density of DzO. The values shown in Figure 1 have been converted to molality. T h e high temperature measurements of Fife and Hruice were made with a Metrohm Type H electrc,de in acid solutions with DCl. Both the Corning No. 476022 and the Metrohm H, according t o the manufacturers' literature (the technical bulletin accompanying the Corning electrode, and Brinkmann Technical Catalog No. BEOA-827. BR%HI. are wide-pH-range and wide-temperature-range electrodes. Also. both are low-sodium-error electrodes. Excellent agreement is observed in Figure 1 for both acidic and basic solutions in D10, and over a wide temperature range. This agreement would seem t o indicate that if a wide pH and temperature-range, low-sodium-error electrode is employed, the necessary correction factor a t any temperature could be calculated from Equation 1. However, it appears that Equation 1 is not totally general in that measurements by Fife and Bruice with a Metrohm Type X electrode give a linear equation considerably different from Equation 1. The Metrohm X electrode is a lowresistance electrode designed for measurements at ambient temperature [lo0MR at 25 "C ( 9 ) ) .The Type X electrode
(7) A. K. Covington. R . A. Robinson, and R. G. Bates, J. Pbys. Cbem., 70,
3820 (1966). (8)R . Gary, R. A. Robinson, and R. G. Bates, J. Phys. Cbem., 68, 3806 (1964).
(9) R . G. Bates, "Determination of pH. Theory and Practice," 2nd ed.. Wiley-Interscience. New York, N.Y., 1973,p 387.
ANALYTICAL CHEMISTRY, VOL. 46, NO. 13, NOVEMBER 1974
2051
is also subject to a large sodium error at pH > 11. The pHsensitive glass in the Type X electrode consists of an unspecified mixture of LizO, CaO, and Si02 (IO). The Metrohm H electrode has a resistance of 1,400 MQ at 25 "C (9). and has a glass composition of Li20, BaO, and Si02 (10). Unfortunately, no comparison of glass composition and room temperature resistance can be made with the
Corning No. 476022 glass electrode, as this information is still regarded as proprietary by the Corning Glass Works. It is probably safe to conclude that Equation 1 is valid if pD measurements are made with an electrode that is widerange and has a low-sodium error. Equation 1 is not valid if the electrode employed does not meet these requirements.
(IO) W. Simon and D. Wegmann, Helv. Chim. Acta. 41, 2308 (1958).
RECEIVEDfor review May 6,1974. Accepted July 8,1974.
Interactions of Some Free Phosphorus(ll1) Compounds with Gold Vapor Detected by Means of X-Ray Photoelectron Spectroscopy Luis J. Matienzo' and Samuel 0. Grim Department of Chemistry, University of Maryland, College Park, Md. 20742
One of the major problems in any spectroscopic technique is the selection of a material suitable to serve as a standard of comparison for data obtained in different laboratories. In the particular case of X-ray photoelectron spectroscopy, charging effects on the surface of the sample, which occur when electrons are removed by incidental radiation, typify such technical difficulties. Generally speaking, if the sample is a good conductor, these charging effects do not take place-Le., electrical contact exists between the sample and the spectrometer. Conducting samples can be referenced with high reproducibility (f0.1eV) to the spectrometer's Fermi level, which can be determined by measuring the photoelectron levels of noble metals, such as gold (where the Fermi level can be detected quite readily). Because the position of the Fermi level for insulators is not known, it is necessary to use a standard reference to make accurate and meaningful binding energy determinations. The most common method of calibration so far has been the use of the C(1s) photoelectron signal arising from organic impurities present in the system (mostly pump oil vapor). Siegbahn et al. ( I ) , were able to obtain what they believed were sound correlations by assuming that pump oil vapor always gave the same C(1s) signal. Although significant variations of C(1s) signal sometimes occur (2) through the use of different kinds of pump oil and through the presence of other impurities, this system of calibration is still employed, probably because the C(1s) line arising from samples containing organic ligands dominates this spectral region. Other methods of calibration, based mainly on the internal mixing of the sample with the standard, have also been used-e.g., graphite ( 3 ) , carbon from the supporting tape ( 4 ) , Pb304 ( 2 ) , potassium chloride ( 5 ) , and molybdenum Present address, Pigments Department, E.
ours and Company, Wilmington, Del. 19898.
I. duPont de Nem-
(1) K . Siegbahn et a/., "ESCA: Atomic, Molecular and Solid State Structure Studied by Means of Electron Spectroscopy," Almqdst and Wiksells, Uppsala. Sweden, 1967. (2) W. J. Stec, W. E. Morgan, R. G. Albridge, and J. R. Van Wazer, lnorg. Chem., 11, 219 (1972). (3) R. Nordberg, H. Brecht, R. G. Albridge. A. Fahlman, and J. R. Van Wazer, horg. Chem., 9,2469 (1970). (4) C. K. Jorgensen, Chimia, 25, 213 (1971). (5) L. D. Hulett and T. A. Carlson. Appl. Spectrosc., 25,33 (1971).
2052
trioxide (6). Hnatowich et al. (7), have reported that vacuum deposition of vapor of metals (considered chemically stable) onto the surface of samples could be used as yet another method of compensation for charging effects. The method (which has been used in the present investigations) consists of applying very small amounts of gold or palladium onto the insulating samples. The metal upon condensation forms agglomerates which are not in electrical contact with each other and which will assume the insulator's electrical potential. The position of the metal photoelectron lines shows a shift from the metal photoelectron line corrected to the Fermi level of the spectrometer. This variation in binding energy is considered to be the correction for surface charging. Hnatowich et al. ( 7 ) were able to demonstrate the reproducibility of their measurements to be fO.l eV when applied to simple salts. Betteridge et al. ( 8 ) , however, pointed out some of the problems which might arise using this method of calibration, among them the possibility of reaction of the vacuum deposited gold with the sample. In the process of investigating the X-ray photoelectron spectra of molybdenum compounds containing phosphorus ligands (9), it was observed that some of the free ligands displayed broad P(2p) signals even after short periods of gold deposition. More detailed studies on this matter have made it possible to explain the origin of these signals and to establish some of the limitations of this method of calibration.
EXPERIMENTAL T h e compounds used in the present investigation were prepared by reported synthetic methods except that reagent grade sodium fluoride and o-phenanthroline were used as received (10, 11).All of the compounds, especially 1,2-bis(diphenylphosphino)ethaneand triphenylphosphine sulfide, were recrystallized several times. Furthermore, the purity of the compounds containing phosphorus was
(6)W. E. Swartz, Jr., P. H. Watts, Jr., E. R. Lippincott, J. C. Watts, and J. E. Huheey. horg. Chem., 11, 2632 (1972). (7) D. J. Hnatowich, J. Judis, M. L. Perlman. and R . C. Ragaini, J. Appl. Phys., 42, 4883 (1971). (8)D. Betteridge, J. C. Carver, and D. M. Hercules, J. Nectron Spectrosc. Re/. Phenomena, 2, 327 (1973). (9) S.0. Grim and L. J. Matienzo, to be published. (10) W. Hewertson and H. R. Watson, J. Chem. Soc.. 1490 (1962). (1 1) A. Michaelis and H. V. Scden, Ann. Chem., 229. 307 (1885).
ANALYTICAL CHEMISTRY, VOL. 46. NO. 13. NOVEMBER 1974