Anal. Chem. 1980, 52, 2353-2355
Table 111. Average Recoveries of Plutonium
and Americium av recoveries, %
sample
no.
soils air dusts
22 23 5 8
sediments waters
x + Fx
quantity log
compositea < l g
400mL
Pu 87 86
94 93 90
i
i i i ?
Am 4 6
4 3
4
89 89 93 88
5 5 6 2 9oi 2 i k i i
Composite of 1 3 47-mm membrane filters per sample. Samples collected from a retention pond.
a
cause degradation of the a spectra. Therefore, the DBHQ is purified by recrystallization. Iron and other contaminants are removed from the HDEHP by preparation of its copper salt (11)and by washing with an alkaline citrate solution after conversion of the copper salt to the acid. Additional washes with perchloric and nitric acids and water eliminate the formation of a yellow organic phase during extraction of the lanthanides and actinides and a white residue which would otherwise appear upon evaporation of the americium and plutonium strip solutions. A summary of recoveries obtained by using the current procedure on soils, air dusts, and waters is shown in Table 111. All the results were obtained by using a barium sulfate precipitation for the initial separation. Water samples were evaporated to a few milliliters and decomposed by the recommended procedure for 1-g soil samples (12). Sediment samples were decomposed by the procedure for 1 g of soil. After dry ashing, the air dust filters were decomposed by the procedure described earlier (2). With the exception of the water samples, the americium results were obtained after separation from the lanthanides by column chromatography (13). Final determinations on both nuclides were obtained by a spectrometry. The extraction procedure provides a rapid method for the separation of americium and plutonium after their initial
2353
separation on barium sulfate in environmental samples. Beginning with the dissolution of the barium sulfate in perchloric acid, the HDEHP extractions and separations on eight samples require about 2 h to obtain final fractions ready for electrodeposition or separation of americium from the lanthanides. Adaptation of the procedure to other sample types such as vegetation ash, fecal ash, and urine, to the determination of thorium and protactinium, and to the use of cerium fluoride instead of barium sulfate for the initial separation on samples of 1 g or less prior to the HDEHP separations is currently being investigated.
LITERATURE CITED Wessman, R. A.; Leventhal, L. I n "Transuranics I n Natural Environments"; White, M. G.. Dunaway. P. 6.. Eds.; USERDA Report NVO-176; NTIS: Springfield, VA., 1978; pp 545-574. Sill, C. W.; Puphal, K. W.; Hindman, F. D. Anal. Chem. 1974, 46, 1725-1 737. Mason, G. W.; McCarty Lewey, S.; Peppard, D. F. J. I m g . Nucl. Chem. 1978, 4 0 , 1427-1430. Qureshi, I. H.; McClendon, L. F.; LaFleur, P. D. Radlochim. Acta 1989, 72,107-111. USAEC Report ORNL-4145; Oak Ridge National Laboratory: Oak Ridge, TN, 1967; pp 132-133. USAEC Report ORNL-4272; Oak Ridge National Laboratory: Oak Ridge, TN, Sept 1968, 95-105. Lawrence, F. 0.; Hoffman, D. C. USAEC Report LA-1721, 3rd ed.; Los Alamos Scientific Laboratory: Los Alamos, NM, 1967; p 94. Chilton, J. M.; Fardy, J. J. J. Inorg. Nucl. Chem. 1989, 31, 3871-3874. Peppard, D. F.; Mason, G. W.; McCarty, S. J. Incrg. Nucl. Chem. 1980, 13, 138-150. Cleveland, J. M. I n "Plutonium Handbook"; Wick, 0. J., Ed.; Gordon and Breach: New York, 1967; Vol. I, Chapter 13. Partridge, J. A,; Jensen, R. C. J. Inorg. Nucl. Chem. 1989, 31, 2587-2589. Sill, C. W. Anal. Chem. 1977. 4 9 , 618-621. Martin, D. 6.; Pope, D. G., "Improved Separation of Tervalent Lanthanides from Actinides by Extraction Chromatography"; Radiologicaland Environmental Sciences Laboratory, Department of Energy: Idaho Falls. ID, unpublished work. Puphal, K. W.; Olson, D. R. Anal. Chem. 1972, 4 4 , 284-269.
RECEIVED for review March 27, 1980. Accepted September 2,1980. Use of commercial product names is for accuracy in technical reporting and does not constitute endorsement of the product by the United States Government.
Electron Spectroscopy for Chemical Analysis and Other Studies of the Anomalous Behavior of the Copper Ion Selective Electrode in Acetonitrile J.
F. Coetzee,"
W. K. Istone, and M. Carvalho
Department of Chemistry, University of Pittsburgh, Pittsburgh, Pennsylvania
I n the Ion-selective electrode for copper( 11) Ion, which has the composition Ag,,,,Cu,,,S, copper Is present as copper( I). It responds satisfactorily in a wide range of solvents but not In acetonHrlle. I t was shown by ESCA and other studies that, In solutions of certain copper( 11) salts in acetonltrile, the electrode fails because copper( 11) replaces copper( I)and sllver(1) Ion Is leached from the electrode.
T h e behavior of ion-selective and other electrodes in nonaqueous solvents is of both fundamental and applied interest. Considerable information is available on the response of electrodes of the first and second kinds. For example, the 0003-2700/80/0352-2353$01.OO/O
15260
Ag+/Ag electrode behaves reversibly in a wide range of solvents, while the response of electrodes of the type Ag,AgX/Xin aprotic solvents is complicated by formation of stable complexes such as AgX,-. As far as the behavior of ion-selective membrane electrodes in nonaqueous solvents is concerned, most of what is known is limited to the glass electrode for hydrogen ion. It responds reversibly to hydrogen ion activity in a wide range of solvents and shows an impressive overall dynamic range of some lo4"in activity in solutions having adequate buffer capacity ( 1 , 2 ) . As far as electrodes for other ions are concerned, only scattered information of limited scope is available, with few exceptions. Nakamura ( 3 ) has compared the response of the "monovalent cation" glass electrode with that of amalgam electrodes to lithium, 0 1980 American Chemical Society
2354
ANALYTICAL CHEMISTRY, VOL. 52, NO. 14, DECEMBER 1980
Table I. ESCA Binding Energies of Jalpaite (Referred to C (1s) at 285.0 eV) sample powder, original pellet, original pellet, soaked in acetonitrile pellet, soaked in 1M Cu( C104)z in water pellet, soaked in 1M
S(2P)
other
Table 11. Results of ESCA and W e t Chemical Analysis of Pressed Polycrystalline Membranes of Jalpaite and of Soaking Solutions
933.5
368.6
162.3
933.3
368.4
162.2
soaking solution
934.2
368.9
162.5
933.3
368.3
162.1
water acetonitrile Cu(ClO,), in water Cu(CIO,), in acetonitrile Cu(NO,), in acetonitrile Cu(Ph,B), in acetonitrile
933.6 943.6a
368.3
~
Cl(2p)at 199.6
pellet,
~
Cu+,Ag’, S 2 CUI, Ag’, S 2 Cu+,Ag’, Si2Cu2+,Ag’ Cuz+,Ag’, S z Cu+, Ag’, S 2 - , SO,’., Ph,B-
Ag’, SO,’Ag’
a Only new ions, not present in the original soaking solution, are listed; Cu’ could not be tested for in presence of the large excess of Cu2+. Pellet turned color from original silvery gray to dull black.
cu(c104)Z
in acetonitrile pellet, soaked in 1 M Cu(NO3)* in acetonitrile
major ions identified after soaking membrane sblutiona
RESULTS AND DISCUSSION 935.5 943.9a
367.9
934.2
368.5
soaked in 0.1 M Cu(Ph4B), in acetonitrile a “Shake-up” - (. 7),. ion (8).
162.3
N (1s)
not seen
162.6 168.gC
Reduced in intensity.
B (1s) at 192.5
Sulfate
sodium, cesium, thallium(I), and silver(1) ions in several protic and aprotic solvents. We w ill report elsewhere on the response of the ion-selective electrode for fluoride ion ( 4 ) and have reported before on the response of the electrode for copper(I1) ion in a variety of solvents and solvent mixtures (5). References to other studies of ion-selective electrodes in nonaqueous solvents are given in ref 3-5. The copper electrode responds satisfactorily in a range of solvents having widely different solvating abilities toward copper(I1) ion. The extremes of the range tested are represented by propylene carbonate and dimethyl sulfoxide; at constant concentration, the activity of copper(I1) ion is loz2higher in the former solvent than in the latter. Acetonitrile was the only solvent tested in which the electrode did not respond. This observation in itself would merit further investigation; in addition, acetonitrile-water mixtures are important eluants in liquid-solid chromatography, and the electrode may be thought to be a likely detector. We have therefore further investigated the behavior of the electrode in acetonitrile by ESCA and wet chemical tests and report the results here.
EXPERIMENTAL SECTION Experimental conditions were generally the same as those employed before (5). Samples of Agl,5,Cuo,45SCjalpaite) were prepared by the method of Van der Linden (6). Electrode pelleta were made by pressing the material in a steel die under vacuum at lo4 kg cm-2. The pellets were mounted into the end of a glass tube with epoxy cement, and silver wire connectors were attached to the pellets by means of silver epoxy cement. In dimethyl sulfoxide and particularly dimethylformamide, slow degradation of epoxy occurs, so that such electrodes should not be left long in those solvents. ESCA spectra were recorded with an AEI Model ES200A spectrometer and an AEI Model DSlOO data system. The aluminum anode (A1 Koc = 1486.6 eV) was operated at 12 kV and 25 mA. Spectra were recorded at a pressure of torr.
Samples of jalpaite (powder or pressed pellets) were soaked in various solutions for 24 h and were then thoroughly washed with the corresponding pure solvent, dried in vacuo a t 25 “C for 24 h, and finally analyzed by ESCA. Results are shown in Tables I and 11. Assignments of ESCA binding energies were based on the tabulations of Carlson (8). The ESCA data confirm Van der Linden’s (9) conclusion derived from X-ray diffraction that copper is present in jalpaite as copper(I), not copper(I1). When the pellet is exposed t o either copper(I1) perchlorate or nitrate in acetonitrile, it turns black and its ESCA spectrum shows the copper doublet (“shake-up”) which indicates that copper(1) has been converted into copper(I1). No such conversion is observed with copper(I1) tetraphenylborate in acetonitrile or with the perchlorate in water. In addition, with the tetraphenylborate in acetonitrile, this ion and also sulfate ion are found on the surface. Wet chemical analysis of the soaking solutions shows (Table 11) that silver(1) ion is leached from the pellet by either copper(I1) perchlorate or nitrate in acetonitrile. Extraction of pellets with carbon disulfide showed only trace amounts of elemental sulfur. These observations are consistent with the occurrence of the following reaction with copper(I1) nitrate and perchlorate in acetonitrile, writing for simplicity Ag2S + CuzS for Ag1.55CU0.45S
Ag,S(s)
-
+ Cu2S(s)+ 2Cu2+
2CuS(s)
+ 2Cu++ 2Ag+ (1)
I t is possible that in copper(I1) perchlorate an additional reaction occurs since chloride, but no sulfide, is found on the surface and sulfate ion appears in solution.
S2-(s)
+ C104-
-
S042- + Cl-(s)
In view of the weak oxidizing power of perchlorate ion in acetonitrile at room temperature, the occurrence of reaction 2 would imply that the activation energy for the process is much lower on the surface of the pellet than in solution. Finally, when the pellet is exposed t o copper(I1) tetraphenylborate, tetraphenylborate appears on the surface, apparently by adsorption [pKsp of AgPhlB in acetonitrile = 7.5 (IO)]. Some sulfate ion also appears on the surface, probably by atmospheric oxidation which is known to occur with other electrodes (e.g., the lead electrode) containing sulfides. It appears that formation of the surface film protects the jalpaite from attack by copper(I1) ion. The fact that degradation of jalpaite occurs in solutions of certain copper(I1) salts in acetonitrile but not in other solvents tested (5)is undoubtedly caused by the stabilization of silver(1) and copper(1) ions by acetonitrile (and other nitriles) derived from P back-bonding and also by the high reactivity of cop-
Anal. Chem. 1980, 52, 2355-2357
per(I1) ion in acetonitrile, as shown by the following free energies of transfer of these ions from water to acetonitrile: Cu2+,+14.2; Cu+,-11.4; Ag+,-4.2 kcal mol-' at 298 K (5). The result is that reaction 1 is favored much more in acetonitrile than in water or other solvents; this will probably remain the case in acetonitrile-water mixtures containing up to 80 mol 70,or even more, of water ( 5 ) .
ACKNOWLEDGMENT We thank D. M. Hercules for the use of his ESCA facilities. LITERATURE CITED (1) Ritchie, C. D.; Uschokl, R. E. J. Am. Chem. SOC. 1967, 89,1721-1725, 2752-2753.
2355
Coetzee, J. F.; Berlozzi, R. J. Anal. Chem. 1973, 45, 1064-1069. Nakamura, T. Bull. Chem. Soc. Jpn. 1975, 48, 2967-2968. Coetzee, J. F.; Martin, M. W., unpublished results. Coetzee, J. F.; Istone, W. K. Anal. Chem. 1980, 52, 53-59. Heijne, G.; Van der Linden, W.; Den Boef, 0.Anal. Chim. Acta 1977, 89,287-296. (7) Larsen, P. J. Electron Spectrosc. 1974, 4 , 213-218. (8) Plenum Press: . . Carlson, T. A. "Photoelectron and Auger - Spectroscopy"; . New York, 1975. (9) Heijne, G.; Van der Linden, W. Anal. Chim. Acta 1977, 93, 99-110. ( I o ) Alexander, R.; Parker, A. J.; Sharp, J. H.; Waghorne, W. E. J . Am. Chem. SOC. 1972, 9 4 , 1148-1158. (2) (3) (4) (5) (6)
RECEIVED for review June 16,1980. Accepted September 8, by the Science lg80. The research was Foundation under Grant No. CHE-7727699.
Time Delay Multiplexing of Optical Spectra with a Fiber Optic Array W. B. Whitten Analytical Chemistry Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37830
Fiber optic delay lines, a grating monochromator, a single photomultiplier, and timeeorrelated photon counting techniques are combined to make a 30-channel absorption spectrometer. The transit time delay in a series of optical fibers of different lengths permits the sequential arrival at the detector of the various wavelength components of a light pulse. Whlle of restricted practicality because of the required low countlng rate, the instrument suggests the feasibility of single pulse spectral measurements with a fiber optic multiplex spectrometer.
T h e transit time of a light pulse through an optical fiber depends on the wavelength of the light and on the length of the fiber. The dependence of transit time on wavelength can be used to measure optical spectra by time-of-flight techniques ( I , 2). The spectra are measured as a function of time so that a single detector can be used. In this article, a spectrometer is described that is based on the time delay properties of optical fibers. An array of fibers of different lengths is used to multiplex a large number of optical signals with subsequent reception by a single detector. The signals to be multiplexed can be an image or, as described in this investigation, timedependent monochromatic intensities of different wavelengths. T h e use of optical fibers as time delay elements goes back to the introduction of the first practical fibers. Pedinoff used an array of 10 fibers to time delay multiplex the optical signals from a polarization reference imaging system ( 3 ) . The first application of time delay multiplex spectroscopy may have been made by Colles and Walrafen ( 4 ) who used seven solid fused silica fibers with 2-ns incremental delay to study the spectral character of mode-locked laser pulses. Measurement of the spectrum of Thompson-scattered light to determine plasma temperatures by Schuss et al. ( 5 ) is another application. Tomlinson has developed a time multiplexed spectrometer in which the multiplexed spectrum is coupled into an optical fiber by the same diffraction grating which provides the initial dispersion (6). In the present investigation, an array of optical fibers of different lengths is used in conjunction with a grating 0003-2700/80/0352-2355$01 .OO/O
monochromator as a wavelength to time converter in a time-correlated photon counting absorption spectrometer. A pulse of polychromatic light, after passing through the sample or reference material, is spatially dispersed by a diffraction grating, and the wavelength components are focused onto an array of optical fibers of graduated lengths. The fibers delay the optical pulses and guide them to a photomultiplier where they are detected sequentially. Thus, a multiwavelength spectrum is obtained as a function of time. It is interesting to compare the multiple fiber time delay multiplexing system with the single fiber time-of-flight spectrometer described previously (I, 2). Both types of instruments yield spectra in the time domain and thus have a number of common qualities such as single detector operation and inherent time resolution. The multiple fiber system uses much shorter fibers so that fiber attenuation is not a serious problem, even in the near-ultraviolet region. The fibers can be arranged so that the shorter fibers transmit the more highly absorbed light. Also because of the shorter lengths, modal dispersion is small compared to the pulse widths normally employed so that large aperture step index fibers may be used for low cost and high light collecting efficiency. The incremental time delay of the fibers can be tailored to the time resolution of the source and detector; the spectral resolution of the system is determined primarily by the grating and the mechanical arrangement rather than by the fiber length and time resolution as in the single fiber spectrometer. The fibers can be arranged in the focal plane of the grating so that they are concentrated in regions of spectral interest. On the other hand, the multiple fiber system requires substantially more total fiber length for equal spectral coverage with equivalent time resolution except, perhaps, in the near-infrared region where fiber dispersion is very low. The additional mechanical complexity of the multiple fiber system and the cost of the grating and mechanical components must also be considered. The time resolution for both systems is source limited since the dispersion and detection are carried out in an expanded time domain. The total length of optical fiber required for a time delay multiplex spectrometer is determined by the number of channels desired and the time resolution capability of the 0 1980 American Chemical Society
