Reduction of Background Absorption in the Measurement of Cadmium, Lead and Selenium in Whole Blood Using Iridiumsputtered Graphite Tubes in Electrothermal Atomic Absorption Spectrometry CORNELIUS J. RADEMEYERa, BERNARD RADZIUKb, NATALYA ROMANOVAc , YNGVAR THOMASSEN*c AND PAOLO TITTARELLId aDepartment of Chemistry, University of Pretoria, Pretoria, South Africa bBodenseewerk Perkin-Elmer GmbH, P.O. Box 101761, D-88662 U� berlingen, Germany cNational Institute of Occupational Health, P.O.Box 8149 DEP, N-0033 Oslo, Norway dStazione Sperimentale per i Combustibili, V iale A. De Gasperi 3, I-20097 San Donato Milanese, Italy The thermal behaviour during pyrolysis and of the vapour stant for more than 700 atomization cycles. It was also shown that the pyrolysis temperature of Cd, Pb and Se in acidified phase during atomization for Cd, Pb and Se in acid-digested whole blood using Ir-sputtered tubes is described. The aqueous solutions is increased to that obtainable with an ordinary pyrolytic graphite-coated graphite tube in conjuction performance of Ir as a permanent modifier was affected unfavourably by the complex matrix compared with with modifiers added in solution form.In this work, the suitability of such a permanent modifier conventional modifiers. Background absorption was measured using an atomic absorption spectrometer in addition to a for the measurement of Cd, Pb and Se in acid-digested whole blood was assessed. The effects on thermal stabilization and diode-array spectrometer and compared with the background obtained in pyrolytic graphite-coated graphite tubes.Both background absorbance were investigated for the three elements. The contribution of the background was assessed by methods of measurement indicated that the background was much reduced in the Ir-sputtered tubes. The decrease in monitoring the background absorption detected by a Zeemaneffect atomic absorption spectrometer and also by monitoring background absorption improves conditions for the measurement of these elements. Background molecular molecular absorption in the UV range using a diode-array spectrometer.The latter approach provided insight into the absorption was also measured as a function of time. Molecular species such as NO were detected in the vapour phase using evolution of the molecular species deriving from the whole blood sample decomposition. pyrolytic graphite-coated tubes, whereas CS and CO were detected using Ir-sputtered tubes.Keywords: Electrothermal atomic absorption spectrometry; iridium permanent modifier; background reduction; cadmium; EXPERIMENTAL lead; selenium; whole blood Instrumentation A Perkin-Elmer (Norwalk, CT, USA) Zeeman 5100 atomic The concept of chemical modification has become essential in the measurement of elements at trace and ultratrace concen- absorption spectrometer equipped with an HGA 600 atomizer and an AS-60 autosampler was used for the thermal stabiliz- trations using ETAAS.1 It has been demonstrated that a variety of reagents, including metals such as palladium, can thermally ation and background absorption investigations.Standard tubes and iridium sputtered pyrolytic graphite coated tubes stabilize many analyte elements, making possible the selective removal of major portions of the sample matrix. Chemical (Perkin-Elmer Part No. B009–1504) were used to assess the pyrolysis characteristics of the sputtered surface for Cd and modification by Pd may be accomplished by heating a Pd-containing solution in the graphite tube to 1200 °C prior Pb in whole blood and integrated platform tubes with and without sputtered Ir (Perkin-Elmer Part.No. B300–1260) were to addition of the sample. The resulting metallic form of Pd has proved to be a reliable modifier for the measurement of also used for Se and background measurements. A Jasco (Tokyo, Japan) KS-100M diode-array spectrometer Cd, Pb and Se in whole blood.This has the additional advantage that any Cd and Pb contained in the modifier is equipped with a Perkin-Elmer HGA-400 atomizer and an AS-1 autosampler was used for the recording of the vapour- volatilized, eliminating the risk of contamination. However, such procedures can increase the analysis time by a factor of phase spectra collected during the atomization step of the elements. This spectrometer has been used previously for the almost two.It would therefore be advantageous if the modifier were ‘permanent’, i.e., added only once to each tube rather study of the thermal behaviour of slurries.4 The diode array of the spectrometer is a Reticon 512 S (512 elements 25 mm wide). than before every sample and used throughout the lifetime of the tube.2 The UV range (190–350 nm) is focused on the diode array by a concave holographic grating (500 linesmm-1) blazed at Recently, we showed that the inside surface of a graphite tube may be coated with a homogeneous Ir layer by means of 230 nm.The calculated resolution is about 0.3 nm per diode. The data processor allows the collection of spectra with a sputtering process in a low-pressure argon discharge.3 When sufficient Ir was deposited, this metallic modifier surface collection times varying from 0.05 to 1 s and with interval times between the start of consecutive spectra varying from changed only slightly during the entire lifetime of the graphite tube and the integrated absorbance remained essentially con- 0.05 to 25 s.Journal of Analytical Atomic Spectrometry, January 1997, Vol. 12 (81–84) 81Acid Homogenization of Human Whole Blood A 2.5 ml volume of sub-distilled 65% nitric acid (Chemscan, Elverum, Norway) was added to 2 ml of reconstituted lyophilized human whole blood (Seronorm 404108, NycoMed, Oslo, Norway) in a polypropylene digestion tube. After degasification at room temperature overnight, the tube was heated at 95°C for 1 h in a laboratory oven.The sample was allowed to cool to room temperature before final dilution to a previously calibrated volume (13.7 ml). De-ionized water (18 MV cm) obtained from a Milli-Q water-purification system (Millipore, Watford, UK), was used throughout. Recording of Absorption Spectra UV absorption spectra were continuously collected during the atomization step using the diode-array spectrometer. The collection time of each spectrum and the interval between the starts of consecutive spectra were set to 0.2 s.Hence the spectra were recorded continuously without delay between the end of Fig. 2 Pyrolysis curves for Se in aqueous solution (15 ml injection) a spectrum and the start of the next one. Particular attention (%) and whole blood (15 ml injection) (2) using an Ir-sputtered was paid to the molecular absorption occurring at the 196.0 nm intergrated platform graphite tube under normal internal gas flow analytical wavelength of Se.It was not possible to employ (R) and stopped flow (2---2) conditions. The background collection times shorter than 0.2 s owing to the combined effect absorbance was also measured under normal (&) and stopped flow conditions (Q). of poor efficiency of the diodes and low emission intensity of the deuterium lamp at this wavelength. A faster measurement rate may have shed light on transient phenomena. stop conditions, volatile Se species released from the platform above 400 °C are likely to be collected and thermally stabilized by the Ir-modified inner surface of the graphite tube rather RESULTS AND DISCUSSION than being deposited at the cooler ends or flushed out of Figs. 1 and 2 show the pyrolysis curves for Cd, Pb and Se in the tube. aqueous solutions and whole blood using Ir-sputtered tubes. It can be expected for a complex matrix such as acid The maximum pyrolysis temperature for each of these elements homogenized whole blood, which contains undigested organic in aqueous standard solutions is increased to a temperature compounds, that the surface modified by iridium may not be similar to that reported for modifiers such as Pd added to the fully available for reaction with the analyte.The reason for analyte in solution form.5 The data for whole blood displayed this is that the permanent modifier is not dissolved in the in Fig. 1 show that the maximum pyrolysis tperature is solution.It was demonstrated previously that, e.g., Sb was approximately 500 °C for Cd, 750 °C for Pb and only 400°C stabilized to a comparable extent by Ir added to aqueous for Se. solutions and to whole blood and urine.6 With pre-reduced A unique feature of a graphite tube fully coated by Ir noble metal modifiers, the modifier is distributed on the surface sputtering can be seen in Fig. 2. Selenium is retained in the of the tube in finely dispered form. On addition of the sample tube at much higher pre-treatment temperatures under solution on the surface, it will mix intimately with the modifier stopped-flow conditions compared with a flow rate of 300 ml and the latter may even be partly redissolved in the solution.min-1, which is normally used during pyrolysis. Under gas- While the permanent Ir modifier is not as effective in the thermal stabilization of the analyte in this matrix, it is very effective in modification of the matrix itself. Fig. 1 also shows the background absorption measured at the analytical wavelengths of Cd and Pb.In both cases background absorption decreases to an almost constant level at pyrolysis temperatures higher than 300 °C. The background absorption recorded during atomization at the analytical wavelengths of Cd, Pb and Se is shown in Fig. 3. The solid curves (Ir-sputtered) were obtained using the same type of tubes for the respective elements which had been sputtered by Ir and with no further addition of modifier in any case.The background is considerably smaller for all three elements when Ir-sputtered tubes are employed. The effect is particularly remarkable for Se. As long-term tests of Ir-sputtered tubes have shown that the modifier is little affected by age, it may be presumed to be chemically inert. Therefore, a possible explanation for the decrease in background involves a catalytic role of the Ir-sputtered surface of the tube while sample components are reduced during the thermal decomposition of the organic part of the blood matrix.This modification of the matrix by the Ir-sputtered surface is also indicated in Fig. 1 Pyrolysis curves for Cd and Pb in aqueous solution (15 ml the shift of appearance time of the background absorption injection) (%) and whole blood (15 ml injection) (2) using Ir-sputtered in Fig. 3. graphite tubes and wall atomization. The background (&) from whole blood is measured for both elements. The molecular absorption spectra that were obtained using 82 Journal of Analytical Atomic Spectrometry, January 1997, Vol. 12Fig. 4 Absorption spectra obtained at different times for 15 ml of nitric acid-digested whole blood after commencement of the atomiz- Fig. 3 Background absorbance–time profiles obtained for Cd, Pb ation cycle in ordinary pyrolytic graphite-coated graphite tubes as (wall atomization) and Se (platform atomization) for 15 ml of nitric measured with a diode-array spectrometer between 190 and 340 nm.acid-digested whole blood for ordinary pyrolytic graphite-coated Collection and interval time, 0.2 s. Scale marks on the time scale also graphite tubes (broken lines) compared with those obtained in represent zero absorbance for the corresponding spectrum. Ir-sputtered graphite tubes (solid lines). Pyrolysis temperatures: Cd and Pb, 100 °C; Se, 200 °C. Atomization temperatures: Cd and Pb, 2000 °C (ramp 1); Se, 2300°C (maximum power). the diode-array spectrometer under the same experimental conditions as employed for the recording of background profiles are shown in Figs. 4 and 5. Fig. 4 shows the absorption of the vapour phase when an ordinary pyrolytic graphitecoated integrated platform tube is employed, and Fig. 5 shows the absorption when the tube is sputtered with Ir. As the period at the beginning of the atomization step is particularly affected by the evolution of molecular species, these spectra were recorded from the onset of this step for a period of 1.8 s.In the case of the pyrolytic graphite-coated tubes (Fig. 4), the background absorption is very strong below 220 nm with a maximum around 200 nm. The behaviour of the molecular absorption as a function of time at the analytical wavelengths of Cd, Pb and Se, recorded with the diode-array spectrometer, is similar to that shown in Fig. 3 and recorded using the atomic absorption spectrometer. The collection time of the diode-array spectrometer (0.2 s) does not allow the characterization and distinction of very fast phenomena such as those evidenced at 196 nm.The absorbance values measured were also lower than those recorded with the atomic absorption spectrometer. This difference can be attributed to the different spectral and time resolution between these spectrometers. The Se line is located between various overlapping molecular bands whose features are changing very rapidly, as can be seen Fig. 5 Absorption spectra obtained at different times for 15 ml of nitric acid-digested whole blood after commencement of the atomiz- in the first three spectra in Fig. 4. For the vapour-phase spectra ation cycle in Ir-sputtered graphite tubes as measured with a diode- obtained with the Ir sputtered tube (Fig. 5), it is clearly array spectrometer between 190 and 340 nm. Collection and interval noticeable that the molecular absorption is greatly reduced. time, 0.2 s. Scale marks on the time scale also represent zero absorbance Further, the molecular absorption decreases to a minimum for the corresponding spectrum.value at a much earlier stage during the atomization step than in the case of the ordinary pyrolytic graphite-coated tube. Fig. 6 shows the molecular absorption spectrum collected 214.3, 225.9 and 235.3 nm belong to the d-system of NO.7 The appearance of NO species at this stage is due to the decompo- between 0.4 and 0.6 s from the beginning of the atomization step using a pyrolytic graphite-coated tube (third spectrum in sition of the nitric acid employed in the dissolution of the blood sample resulting in the transformation of the bulk of Fig. 4). The well defined bands with maxima located at 204.2, Journal of Analytical Atomic Spectrometry, January 1997, Vol. 12 83lines of Pd, Mg and Na can be observed in the lower spectrum in Fig. 7. After 2.8 s only the lines of Pd can be identified. While Na and Mg are major components of the matrix, no Pd was contained or added to the samples in Ir-sputtered tubes.The absorption is therefore attributed to the volatilization of Pd from the tube ends contaminated by the graphite cones which had been used in previous experiments in which Pd was added. CONCLUSIONS It has been found that Cd, Pb and Se, when contained in acidified aqueous solutions, are thermally stabilized to at least the same extent in Ir-sputtered tubes as by conventional modifiers. However, the performance of the permanent modifier Fig. 6 Absorption spectrum obtained between 0.4 and 0.6 s using a was affected unfavourably by the complex matrix contained in pyrolytic graphite-coated graphite tube (third spectrum in Fig. 4). nitric acid-digested whole blood samples. Maximum pyrolysis temperatures were lowered, compared with those for acidified solutions, from 800 to 500°C for Cd, from 1200 to 750°C for Pb and from 1400 to 400 °C for Se. The background absorption that was measured during the atomization phase was dramatically lowered in comparison with that measured when reduced palladium was used as the modifier, in spite of the much lower permissable pyrolysis temperatures.This decrease in background absorption may be explained in terms of a catalytic effect of the Ir-treated surface on the dissociation of volatile components originating from proteins contained in the whole blood matrix. The decrease in background absorption improves conditions for the measurement of these elements in blood, in particular for Se, which is measured at an analytical wavelength situated among molecular bands that are rapidly changing in intensity and spectral position during the atomization step.Fig. 7 Absorption spectra obtained after 1.6 (lower spectrum, ninth spectrum in Fig. 5) and 2.8 s (upper spectrum) using an Ir-sputtered graphite tube. REFERENCES 1 Tsalev, D. L., and Slaveykova, V. I., J. Anal. At. Spectrom., 1991, the inorganic residue to nitrate form during drying.The 7, 147. appearance of NO bands is also observed during the heating 2 Tsalev, D. L., D’Ulivo, A., Lampugnani, L., Di Marco, M., and Zamboni, R., J. Anal. At. Spectrom., 1995, 10, 1003. or atomization of samples treated with nitric acid as in the 3 Rademeyer, C. J., Radziuk, B., Romanova, N., Skaugset, N. P., case of sea water.8 The presence of NO in the vapour phase is Skogstad, A., and Thomassen, Y., J. Anal. At. Spectrom., 1995, observed to a much smaller extent when Ir-sputtered tubes are 10, 739.used (Fig. 5), indicating that nitrates are reduced at the sputt- 4 Tittarelli, P., and Biffi, C., J. Anal. At. Spectrom., 1992, 7, 409. ered surface during atomization. This behaviour is in agreement 5 Tsalev, D. L., Slaveykova, V. I., and Mandjukov, P. B., Spectrochim. with a general decrease in the appearance of vapours of Acta Rev., 1990, 13, 225. 6 Dahl, K., Thomassen, Y., Martinsen, I., Radziuk, B., and Salbu, monoxides during the atomization step when a noble metal B., J. Anal. At. Spectrom., 1994, 9,1. such as Pd is added to slurries.4 7 Pearse, R. W. B., and Gaydon, A. G., T he Identification of The appearance of CS molecules is enhanced in the case of Molecular Spectra, Chapman and Hall, London, 3rd edn., 1963. Ir-sputtered tubes after 1.6 s and CO after 2.8 s (Fig. 7). This 8 Katskov, D. A., Marais, P. J. J. G., and Tittarelli P., Spectrochim. may be a result of reduction of sulfur compounds of the matrix. Acta, Part B, 1996, 51, 1169. Both CS and CO are observed at high temperatures when 9 Tittarelli, P., Biffi, C., and Kmetov, V., J. Anal. At. Spectrom., 1994, 9, 443. mixed oxides and sulfates are atomized, as a result of the decomposition of the matrix on the graphite surface.9 The Paper 6/04783A appearance of CS and CO occurs at the end of the atomization Received July 8, 1996 cycle and does not affect the background absorption measured Accepted October 4, 1996 during the atomization of Cd, Pb and Se. Apart from the molecular bands of CS, the atomic absorption 84 Journal of Analytical Atomic Spectrometry, January 1997, Vol. 12