Graphite Braid Atomizer for Atomic Absorption and Atomic Fluorescence Spectrometry Akbar Montaser, S. R. Goode,' and S. R. Crouch' Department of Chemistry, Michigan State University. East Lansing, Mich. 48824
Non-flame atomizers for atomic absorption (AA) and/or atomic fluorescence (AF) spectrometry have become increasingly useful in recent years. A wide variety of nonflame atomizers have been proposed. The most common of these are "thermal" atomizers, which are heated by dissipating electrical energy in the atomization element. Thermal atomizers include furnace types (1-3), rods ( 4 , 5 ) . and metal strips or filaments (6-9). The atomizers made from carbon or graphite appear to be highly applicable because of the high operating temperatures (25003000 "C) which can be achieved. The real and potential advantages of non-flame atomizers have been summarized recently by Kirkbright (10). Despite the many attractive features of non-flame atomization, some of the previously described atomizers have several practical limitations. Graphite or carbon tubes and furnaces, for example, may require several kilowatts of power to achieve atomization temperatures in the range of 2500 to 3000 "C (2, 5 ) . Even smaller carbon rods and filaments may require 1 kilowatt of power to achieve similar temperatures ( 5 ) . Since the transfer of electrical energy into thermal energy is not perfectly efficient, a fraction of the electrical energy will be dissipated as electromagnetic radiation. Extensive shielding and/or careful circuit design must be employed to minimize these electromagnetic interferences. The sheer bulk of a system which requires kilowatts of power also prevents its application as a portable field analyzer. In this communication. a new filament-type non-flame atomizer. the graphite braid atomizer (GBA), is described and characterized. The electrically heated GBA is capable of providing temperatures similar to other carbon or graphite atomizers, but with lower applied powers. The graphite braid has been used in the AA and AF determination of several elements and shows good sensitivity and linearity. A procedure for the direct analysis of copper in a serum matrix is presented to illustrate the application of the GBA to biological samples. Only a dilution of the serum sample is required prior to the analysis step. EXPERIMENTAL Graphite Braid Atomizer. T h e graphite braid (Union Carbide Corp.. Parma. Ohio) contains high strength. flexible. continuous Present address. Department of Chemistry, University of South Carolina, Columbia. S.C. 29208. Author t o whom reprint requests should be addressed. (1) H . Massmann, Spectrochim Acta.. 2 3 6 , 215 (1968). (2) R. Woodriff, R. W. Stone, and A. M. Held, Appi. Spectrosc., 22, 408 (1968). (3) M. S. Black, T. H. Glenn, M . P. Bratzel. and J. D. Winefordner, Anal. Chem., 43, 1769 (1971). ( 4 ) T. S. West and X. K. Williams, Ana/. Chim. Acta., 45, 27 (1969). (5) M . D. Amos, P. A . Bennett, K. G. Brodie, P. W. Y . Lung, and J. P. Matousek, Anai. Chem.. 43, 211 (1971). (6) M. P. Bratzel, R . M. Dagnall, and J. D. Winefordner, Ana/. Chim. Acta.. 48, 197 (1969). (7) M. P. Bratzel, R . M. Dagnall, and J. D. Winefordner, Appi. Spectrosc.. 24, 518 (1970). ( 8 ) S. R. Goode, Akbar Montaser, and S. R. Crouch, Appi. Spectrosc.. 27, 355 (1973). (9) H. M. Donega and T. E. Burgess, Anal. Chem., 42, 1521 (1970) (10) G. F . Kirkbright, Anaiyst ( L o n d o n ) ,96, 609 (1971).
filaments of graphite which are woven to form a braid of 1.5- t o 2-mm diameter. T h e total impurity content of braids is reported to be less t h a n 150 ppm (11). Braids used in this work were outgassed by resistive heating t o about 2500 "C for a period of 2-3 sec. T h e cleaning procedure was repeated until t h e AA or AF base line was constant. This usually required two or three repetitions of t h e cleaning cycle. Apparatus. Graphite braids 3 cm in length were mounted below t h e optical path of an AA/AF spectrometer by an arrangement similar to t h a t described previously (8). The spectrometer consisted of a grating monochromator (EU-500, Heath Co.. Benton Harbor. Mich). a photomultiplier module (EU-701-30. Heath Co.). radiation sources to be described below, a current-to-voltage converter (Model 427, Keithley Instruments, Inc., Cleveland, Ohio) and a P D P Lab 8 / e computer (Digital Equipment Corp.. Maynard, Mass.). Details of t h e computer-controlled non-flame spectrometer have been described (f2).Computer integration of the peak-shaped signals was used in all cases. In addition. background correction, conversion of transmittance to absorbance. and other calculations were performed by the computer, and the analytical results were printed on a teletype. Cadmium and zinc metal discharge lamps (George W . Gates & Co., Inc.. Franklin Square, K.Y.) were used as AF excitation sources. while a lead hollow cathode lamp (WL 22925, Westinghouse Electric Corp., Electronic T u b e Division, Elmira. K.Y.) and a Cu-Zn-Pb-Ag multielement hollow cathode l a m p (JA 45448. Fisher Scientific, Waltham. Mass.) were used as AA sources. Samples were dispensed onto t h e graphite braid atomizer by either a pneumatically operated automatic sampler (8).or a 5-pl syringe (Unimetrics Universal Corp.. Anaheim, Calif.). T h e GBA current was controlled by a programmable. currentregulated power supply ( 1 2 ) . A three-stage current program for desolvation. ashing. and atomization was used for biological samples. For aqueous solutions, only t h e desolvation and atomization steps were utilized. T h e GBA was shielded by two concentric sheaths of argon gas. Lnless otherwise stated. the parameters given in Table I were used throughout this stud of the atomizer temperature were made with an optical pyrometer (Catalog Yo. 8622-C, Leeds & Northrup Co., Philadelphia. P a . ) . All temperature measurements had an uncertainty of approximately k50"C. Reagents. All stock solutions were prepared from the pure metal dissolved in a minimum amount of hydrochloric or nitric acid. Aqueous standards were prepared by dilution of the stock solution with distilled water which had been purified by an Osmonics Inc. "Ultrapure Water System" (Osmonics, Inc.. Hopkins. Minn.). For t h e blood serum analysis. t h e serum matrix (Monitrol I, Scientific Products. Inc.. Detroit. Mich.) was diluted tenfold and then spiked with standard solutions of copper. Procedure. When aqueous solutions were analyzed. 2-111 s a m ples were placed on t h e graphite braid, and the integrated absorption or fluorescence was measured under the conditions shown in Table I. T h e GBA was allowed to cool for 10 seconds between analyses. For t h e analysis of copper in serum, the sample was dried for 10 seconds at 100 "C. ashed for 30 seconds a t 500 "C. and atomized for 3 seconds at 1900 "C.
RESULTS AND DISCUSSION To evaluate the graphite braid atomizer for AA and AF spectrometry, several tests were performed. First, the D. J. Page, Union Carbide Corp., Parma, Ohio, personal correspondence, 1973 (12) Akbar Montaser and S. R. Crouch, The Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, Cleveland, Ohio, March 1973, Paper 119. (11)
~~
~~~~~~~~
Table I. Experimental Parametersa Selected in the S t u d y of Graphite Braid Atomizer
Method
Element
AF
Analysis wavelength, nm
Monochromator slit, width, mm
Sheath gas flow rate, 1. 'min
Atomization power, W
Atomization temperature, " C
228.8 213.86 283 ,31 324.75
1.5 1.5 0.2 0.2
1 1 2 2
95 95 165 165
1485 1485 1900 1900
Cadmium Zinc Lead
A.4
Copper
The following parameters were the same for all elements: P M T supply voltage, 800 V; drying time, 11 sec; drying power, 3 W; atomization time, 3 sec; and sample size. 2 pl.
t 0
I20
60
I80
240
360
300
TIME
Figure 1. Fluorescence
peaks for 150-pg Cd samples
Vertical scale, l o - ' A/div; horizontal scale, 60 sec/div; relative standard deviation, 4%
X
-+!
I
[dl 01
02
I
I
I
1
I
1
I
I
03
04
05
06
07
08
09
IO
CONCENTRATION OF COPPER IN 1-10 D I L U T E D S E R U M , pg/rnl
Standard addition analytical curve for copper added to tenfold diluted serum Figure 2.
temperature us. applied power relationship was established. Studies were made to determine typical braid lifetimes under a variety of operating conditions. The general characteristics of AA and AF signals with graphite braid atomization were studied, and the atomizer was used for the analysis of copper in serum to illustrate its potential application for biological samples. Braid Temperature. The temperature of the atomizer was found to be variable up to 2600 "C. The temperature dependence on power for a GBA 3 cm long, was approximately linear over the range 1000 to 2500 "C and had a slope of = 5 'C/W in this range, with only 350 W needed to heat the atomizer to 2500 "C. When the length of the braid is shortened. the required power needed to reach a 600
A N A L Y T I C A L CHEMISTRY, VOL. 46,
NO. 4 ,
A P R I L 1974
given temperature is considerably less than the power needed for the longer braid. In contrast to carbon rods or graphite furnaces, no cooling system for the atomizer holders was required with the GBA. Braid Lifetime. The lifetime of any non-flame atomizer depends on a variety of parameters such as sample size, matrix, gas flow, and applied power. Although no extensive study was made of the effects of all these parameters, the results of four different sets of AF experiments on aqueous samples containing 100 pg of Cd, performed under the selected conditions shown in Table I, gave a n average of 228 determinations before the braid deteriorated. This decreased to about 60 determinations at 2000 "C before the braid was weakened significantly. Deterioration of the braid can be seen as a hot spot develops a t a point which is usually located between the inner and outer gas flow. The braid separates if heating is continued after a hot spot has developed. AA a n d A F Signal Characteristics. During the atomization step. the atomic vapor above the atomizer gives rise to an AA or AF signal with a time duration which depends on a variety of factors such as the sample size, matrix, and atomization temperature. Typical signals were approximately Gaussian in shape with half-widths on the order of 0.5 sec. Integration of the absorption and fluorescence peaks was used to improve the accuracy and to extend the linearity of the working curves. The photometric signal was integrated for the duration of the atomization step. After a n aqueous sample is placed on top of the braid, it soaks down between the graphite fibers. Heating the fibers produces a furnace-like environment. The heating processes are very efficient because a high surface area is kept in contact with the sample. The possibility of sample explosion is minimized, again due to the furnace-type environment around the sample. An explosion and a concurrent loss of sample could be observed if the sample were too viscous to soak into the braid and the desolvation temperature were set well above the boiling point of the solvent. The effects of varying the length of time which the sample remained on the atomizer prior to desolvation were studied. For soaking times of up to 20 seconds, the mean AF signal for a 1 pg/ml aqueous zinc solution remained constant, within a 95% confidence interval. The standard deviation of replicate trials was also unaffected by the soaking time, probably because the entire length of the braid is heated to a uniform temperature. Soaking of the sample into the braid may be a problem with different sample matrices, however. Analytical Results. Analytical data were obtained for four different elements in aqueous solutions. Analyses were performed for cadmium and zinc by atomic fluorescence, and the experimental detection limits (S/N = 2) were approximately 10-11 gram for both elements, which corresponds to concentrations of about 5 ng/ml in both
Only a rough optimization for the ashing step was performed to obtain these data. Figure 2 demonstrates the excellent linearity in a complex matrix without any difficult sample preparation steps. The relative standard deviations ranged from 4 to 16% with a sample population of 3. Since sample volumes are often limited in clinical situations, the automated sampler, with a dead volume of 150 pl, could not be used. A syringe was used to transport the sample, and the high relative standard deviation of individual points can be attributed to the difficulty of placing samples on the atomizer in a reproducible manner.
cases. Copper and lead were analyzed by atomic absorption, and detection limits of approximately 2 X gram were obtained. All of the analytical curves showed good linearity over two to three orders of magnitude in concentration. The relative standard deviation was usually between 4-770 at concentrations one order of magnitude greater than the detection limit. In Figure 1, typical recorder tracings of AF signals from 10 consecutive samples of 150 pg of Cd are shown. The relative standard deviation of the peak areas of these signals is 4%. These data were obtained without optimization of instrumental parameters. The detection limits and the dynamic range can probably be improved substantially if the system is optimized (8). The ability of a non-flame atomizer to utilize samples in complex matrices is one of its greatest assets. Because of its clinical significance, copper ( 1 3 ) in a serum matrix was chosen as a good test of the GBA in a practical analysis. When copper is analyzed by flame spectrometry, the serum matrix affects the results. probably because of changes in the transport parameters. The transport is nearly 100% efficient in the non-flame atomizers, and matrix effects are minimized when appropriate temperatures are used to desolvate, ash, and atomize the sample. Different amounts of aqueous copper were added to the tenfold diluted serum samples, and the standard addition analytical curve shown in Figure 2 was obtained by AA.
CONCLUSIONS The graphite braid non-flame atomizer has been shown to be a medium power alternative to the more commonly used graphite rods and furnaces. In addition to the high operating temperature and other advantages of non-flame atomizers which are preserved, the graphite braid provides a furnace-type environment with uniform temperature throughout the GBA. The atomizer requires relatively low power and no cooling system is necessary. The good detection limits and precision suggest that the GBA should have widespread application in AA and AF spectrometry. Further investigations of the GBA and its applications are being carried out in these laboratories. Received for review August 13. 1973. Accepted November 16,1973.
(13) Norbert W. Tietz. "Fundamentals of Clinical Chemistry," W. B. Saunders Go.. Philadelphia, Pa., 1970, p 663.
Double Modulation Atomic Fluorescence Flame Spectrometry W. K. Fowler, D. 0. Knapp, and J. D. Winefordnerl Department of Chemistry, University of Florida, Gainesville, Fla. 32607
A continuum source and double modulation-i. e., modulation of the source radiation and modulation of the spectral image over a small wavelength range-has been used previously in atomic absorption flame spectrometry by Elser and Winefordner ( I ) . Some of the advantages presented for such a system are applicable to atomic fluorescence flame spectrometry (AFFS). For example, emission radiation from the flame cell and incident radiation scattering can be minimized in AFFS by such a system. Modulation of the source radiation compensates for thermal emission from the flame cell, and wavelength modulation, which results in a derivative of the signal with respect to the wavelength, compensates for scattering from particles in the flame. A continuum source offers several advantages distinct from line sources among which are a savings in analysis time and cost of many sources and the convenience of having one source for all elements. Continuum sources unfortunately have rather low spectral radiance over an absorption line compared to the spectral radiance of intense line sources-e.g., thermostated electrodeless discharge lamps (EDL) (2). 1
Author to whom reprint requests should be sent.
( 1 ) R. C . Elser and J. D. Winefordner, A n a / . Chem , 44, 698 (1972). (2) R . F . Browner, B. M. Patei, T. H. Glenn, M. E. Rietta, and J. D. Winefordner. Spectrosc. Lett., 5 , 311 (1971).
In this work, the limits of detection of nine elements in an air-acetylene flame and three elements in a nitrous oxide-acetylene flame using a 900-watt xenon arc are measured by double modulated atomic fluorescence flame spectrometry (DMAFFS). Specifically, the spectral radiation from a 900-W xenon arc (XBO, 900 W/P, Osram, Berlin, Germany) enclosed in a suitable housing (LH 151 N, Schoeffel Instrument Co., Westwood, N.J.) and powered by a regulated dc supply (Sola Electric Co., Elk Grove Village, Ill.) was modulated a t 666 Hz by a mechanical chopper (Model 125, Princeton Applied Research, Princeton, N.J.) and focused on the center of the flame cell. The burner system, which utilized capillaries, has been described elsewhere ( 3 ) . Gas flow rates of 14.3 and 2.6 1. min-I and 12.0 and 6.4 1. min-l were used with air/CzHz and NzO/CzHz flames, respectively. All measurements were made approximately 2 cm above the burner top. The monochromator, refractor plate for wavelength modulation, and associated electrical components have been described previously ( 1 ) except that in the present case a low noise preamplifier (PAR, Model CR-4) and lock-in amplifier (PAR, Model JB4) were utilized. All aqueous solutions were prepared from reagent-grade chemicals . ( 3 ) L. M . Fraser and J. D. Winefordner, A n a / . Chem.. 43, 1693 (1971). A N A L Y T I C A L C H E M I S T R Y , V O L . 46, N O . 4 , A P R I L 1974
601
