Application of Palladium- and Rhodium- plating of the Graphite Furnace in Electrothermal Atomic Absorption Spectrometry EWA BULSKA AND WOJCIECH JEDRAL Department of Chemistry University of Warsaw 02-093 Warsaw Poland Palladium- and Rh-plating of the graphite furnace has been evaluated as a method of introducing the metallic form of Pd and Rh for chemical modification in electrothermal atomic absorption spectrometry. It is shown that by electroplating Pd and Rh onto the inner surface of the tube the pretreated graphite surface may resemble the behaviour of the corresponding modifier. The resulting metallic layer is very effective in inhibiting the loss of volatile elements (e.g. As and Se) as well as reducing the influence of oxide and carbide formation (e.g. Si). The advantage of the proposed procedure of introduction of a solid chemical modifier is that the pretreated tubes exhibit an extended analytical lifetime for the determination of As and Se up to 80 and 160 firings in the presence of Pd and Rh respectively.In the case of Si the Rh-plated graphite tube could last for about 100 firings. Keywords Electrothermal atomic absorption spectrometry; modijiers; palladium and rhodium-plated graphite tubes; arsenic selenium and silicon Chemical modification has been incorporated as an integral part of the stabilized temperature platform furnace concept by Slavin et a2.l The aim of chemical modification is to obtain optimal analytical conditions by changing the thermochemical behaviour of both the analyte and matrix components.' Since the introduction of Pd as a chemical modifier for electrothermal atomic absorption spectrometry3 it has been established as a modifier with universal appli~ability.~.~ Palladium combined with magnesium nitrate5-' or has become the most widely used modifier under routine analytical conditions.The high efficiency and universality of Pd-based modifiers can be explained by the unique catalytic properties of this metal. It has been reported9-12 that Pd must be in a particular chemical form such as the metallic form in order to act as an effective modifier. Palladium is known to form inter-metallic compounds and solid solutions with the elements to be determined in the graphite furnace.13 Besides the thermal stabilization of volatile elements another aspect of modifier action could be related to impregnation of the tube surface enhancing the direct contact of analyte with graphite.This is of special importance when carbide forming elements are determined. Ortner and Kantuscher14 improved the detection of Si by impregnation of the electrographite tubes with tungsten solution. Also W Ta Zr Mo V and Ti were used for impregnation of the porous graphite s~rface.'~ In the presence of Pd a 5-fold increase in the sensitivity was observed.16 In this case Pd must be introduced with each sample injection. Michaelis et ~ 1 . ' ~ reported life-times of more than 400 atomization cycles for TaC-coated platforms. Long- time performance of a modifier was first introduced as a concept of 'permanent modifier' by Shuttler et a1.l' The purpose of the present study was to investigate the performance of Pd and Rh after electroplating onto the inner Journal of Analytical Atomic Spectrometry surface of the graphite tube as a permanent chemical modifier with respect to a possible increase in the maximum pyrolysis temperature for volatile elements (e.g.As and Se) and a decrease in the formation of oxide and carbide for carbide forming elements (e.g. Si). Electroplating results in a continu- ous metallic film on the graphite surface. The film may significantly enhance losses of the elements determined. The quality of the deposited metallic film is not expected to be critical because the furnace operating temperature is usually higher than the melting-point of the metal. Therefore the structure of the metallic film should change during the first heating to atomization temperature.The thickness of the film however must be very small. A thick film might decrease the overall electric resistance of the furnace. This can change the rate of furnace temperature increase and finally can cause the real tube temperature to deviate from the set temperature or may decrease the life time of the graphite tube. EXPERIMENTAL Apparatus An atomic absorption spectrometer Model Video 22 (Thermo Jarrel Ash) equipped with a Perkin-Elmer HGA-400 graphite furnace atomizer was used throughout this work. Since the system was not equipped with an autosampler solutions were injected manually using a 20 pl micro-pipette. Hollow cathode lamps for As Se and Si were used at wavelengths of 197.3 196.0 n and 251.6 nm respectively. Different types of graphite tubes were used for the investi- gation uncoated electrographite (EG) tubes pyrolytic graphite coated electrographite (PC) tubes and electroplated (EPC) tubes.All tubes (EG and PC) were commercially available Perkin-Elmer products. The EPC tubes were prepared from EG tubes by electrodeposition of the modifier (Pd or Rh) onto the surface. The time-temperature programme used is given in Table 1. Reagents Standard stock solution containing lo00 mg 1-1 of the elements being investigated were prepared from Titrisol concentrate (Titrisol Merck). Working standard solutions were prepared daily by appropriate dilution with 0.1 % HNO,. Palladium nitrate (1 mg ml-l) and rhodium chloride (1 mg ml-') were used as chemical modifiers.The composition of the electrolytic bath for Pd- and Rh-plating is shown in Table 2. All sample containers glassware and autosampler cups were soaked in 10% v/v nitric acid for 24 h and then rinsed thoroughly with doubly distilled water. All reagents and solu- tions were regularly checked for possible contamination with the elements investigated. Journal of Analyticd Atomic Spectrometry January 1995 Vol. 10 49Table 1 Graphite furnace temperature programme Programme I used for the determination of investigated elements- Step 1 2 3 4 5 Temperat ure/"C 90 120 -* -t 2700 Parameter Ramp time/s 5 10 5 0 1 Hold time/s 10 20 10 4 2 Read - ON - - - Programme 2 used for thermal pre-treatment of the modifier$ in graphite furnace- Parameter Step 1 2 3 TemperaturePC 90 120 1200 Ramp time/s 5 5 5 Hold time/s 10 20 10 *Variable.7 Atomization temperature As 2200 "C; Se 2200 "C; Si 2500 "C. $20 p1 contained 2 pg of Pd or 3 p of Rh. Procedures The modifiers were used as follows. (i) The modifier solution was injected first dried and pyrolysed (Table 1 Programme 2) to obtain Pd or Rh in its reduced form then the sample was injected. (ii) The modifier was deposited by electroplating onto the inner surface of the graphite tube. Preparation of the metal-plating tubes EPC A simple set-up consisting of a constant current source a Pt anode and a holder for the graphite furnace (acting as a cathode) was used to electrodeposit the metal (Fig. 1). An additional voltmeter was used for measuring the voltage between the electrodes to assure that no short connection occurred between the electrodes during the electroplating.The EC graphite tube was cleaned before use by heating at 2500 "C for 5 s in the atomizer while purging argon gas through it. After cooling the tube was wrapped with a Teflon band to protect the outside surface from metal deposition then fixed into the metal holder. Platinum wire was carefully introduced inside the tube and both electrodes (graphite tube and Pt-wire connected as shown in Fig. 1) were immersed into a cylindrical glass cell (id. 0.9 cm) containing 2ml of electrolytic bath solution. Experimental details for the metal-plating procedure are shown in Table 3. After several minutes of electroplating a significant amount of the metal has been moved from the bath to the tube surface.Decreasing the concentration of the metal compound in the bath influences the current efficiency of the electroplating process. To obtain good repeatability of the process for each electroplating a fresh electrolytic bath solution was used. After the electroplating cycle was finished the tube was washed in a stream of doubly distilled water and dried at room temperature. Next the tube was fixed in the HGA unit slowly dried at a temperature of about 80°C and then heated Table 2 Electrolytic bath composition 1 H- Fig. 1 Unit used for electroplating Pt platinum anode; GF graphite tube; H Holder; S glass separator; CS constant current source; and V voltmeter Table 3 Conditions used for Ph- and Rh-plating of the graphite tubes Electrolytic bath Current density/mA cm-2 Current efficiency (YO) TemperaturePC Time/min Pd 0.2 50 80 60 Rh 0.6 20 40 45 to 2000°C.After this procedure the modified EPC tube was ready to be used in further investigations. RESULTS AND DISCUSSION As many elements are subjected to losses during the pyrolysis step chemical modification techniques are recommended to decrease the volatility of the analyte.2 Arsenic and Se were studied as an example of volatile elements for which Pd has been reported to be a suitable chemical modifier.6 On the other hand Si was chosen to investigate the application of a modifier to the determination of the refractory elements. Although Si is an element of low volatility the formation of oxide and carbide is the main reason for decreased atomization efficien~y.'~~~~ Therefore the application of Pd and Rh for those elements was investigated.The application of a modifier can be realized by pre-mixing with a sample or by separate injection into the furnace. The effectiveness of thermal pre- treatment of Pd has been reported in many application^.^^^ In these studies the modifier solution was injected first then dried and pyrolysed in order to obtain Pd in its reduced forrn. Therefore the appropriate aliquot containing Pd must be introduced before each sample injection. In order to simplify the procedure efforts have been made to stabilize the reduced- Pd modifier. Shuttler et used Ir to stabilize Pd for permanent trapping of the hydride. In this study electroplating pre-treatment of the graphite tube was used in order to obtain a permanent modification effect for the determination of As Se and Si.Pd-plating Rh-plating Compound Amount/g per 100 ml Compound Amount/g per 100 ml PdCI 1.66 RhZ ( s04)3 0.25 Na2HP04 * 12 H,O 20 Sulfuric acid (d = 1.8 g cm-3) 30 (NH4)2HP04 * 12 H2O 5 Benzoic acid 0.2 50 Journal of Analytical Atomic Spectrometry January 1995 Vol. 10Table 4 Maximum pre-treatment temperatures ("C) in the presence of Pd and Rh modifiers Pd Rh Thermal Electro- Thermal Electro- Element No modifier pre-treatment* plating pre- treatment * plating As 900 1300 1300 1450 1450 Se 300 1200 1200 1400 1400 Si 1200 1200 1200 1500 1600 *Pd or Rh solution (20 p1; 1 mg m1-l) was injected dried and pyrolysed before sample injection. Thermal stabilization studies The pyrolysis curves for As and Se under different conditions e.g.in the absence and in the presence of Pd and Rh are shown in Fig. 2,(a) and (b). The maximum pre-treatment temperature applied under the different conditions are presented in Table 4. The results indicate that Pd modifier used in both modes e.g. thermal pre-treatment or electroplating thermally stabil- izes As up to 1300°C and Se up to 1200°C. Comparison of the results obtained with Pd and Rh modifiers shows that the latter cause better thermal stabilization for both elements up to 1450°C for As and 1400°C for Se. The data are in good agreement with the observations of Tsalev and Slaveykova.21 This is to be expected when comparing the melting- and boiling-points for Pd (m.p. = 1550 "C; b.p. = 2900 "C) and Rh (m.p.=1970°C; b.p.=370OoC).These data lead to the con- clusion that the substantial improvement in the maximum pre- treatment temperature observed experimentally in the presence of Rh is related to its higher melting-point. Different behaviour was observed in the case of Si [Fig. 2(c)]. Silicon is a refractory element with a high melting-point (m.p. = 1420°C). Therefore in the presence of Pd the pyrolysis curve for Si exhibits similar behaviour as compared with the case 0.4 I 1 1 0.2 0.1 0 0 500 1000 1500 2000 Temperatu re/"C Fig.2 Thermal pre-treated curves for (a) 1 ng of Se (b) 1.5 ng of As and (c) 2 ng of Si. Graphite tubes were used 0 without modification; 0 Pd-plated; and . Rh-plated where no Pd was added. However the significant increase in the integrated absorbance value indicates that Pd acts as a modifier enhancing the efficiency of the atomization. This could be explained by the formation of inter-metallic (Si-Pd) compounds decreasing the possibility of forming silicon oxide or silicon carbide molecules. After electroplating of the graphite tube with Rh the increase of sensitivity is not as pronounced compared with Pd [see Fig.2(c)]. However the maximum pyrolysis temperature was increased up to 1600°C for the determination of Si. Analytical figure of merit The sensitivity [characteristic masses (mo) defined as the mass of analyte in pg which provides an integrated absorbance of 0.00441 for the elements investigated under various conditions are shown in Table 5. From these data it is clear that in the case of As and Se the Pd or Rh modifiers have no influence on sensitivity.Note that when no modifier was used the data obtained relate to conditions when the pyrolysis temperature did not exceed 900 "C for As and 300 "C for Se. It should be noted from Table 5 that the use of Pd or Rh results in a lower characteristic mass for Si. This is probably due to the fact that the modifier protects the analyte from direct contact with the graphite surface and/or with oxygen present in the gas phase. Therefore the formation of silicon oxide and carbide is less probable hence the sensitivity for the determination of Si is enhanced. It is interesting to note that the sensitivity is always better when electroplating pre-treatment of the modifier was used for the determination of Si.Thermal pre-treatment of the modifier improves the sensitivity by a factor of 3.6 for Pd and 2.3 for Rh compared with when no modifier was used. The better results (by a factor of 1.6) were obtained when the graphite surface was covered with a metallic layer of the modifier. It is believed that in case of Pd- or Rh-plating the whole surface of the tube is coated while the thermal pre-treatment allows only a limited area (e.g. the site of the droplet) to be covered. However it is difficult to conclude whether the improvements obtained when using electroplating are caused only by a larger Table 5 Characteristic masses (m,) of As Se and Si for different types of graphite tubes surface (in pg). Note that the maximum pre-treatment temperatures for the determination of m were as described in Table 4.EG uncoated electrographite; PC pyrolytic graphite-coated electro- graphite; TP thermal pre-treatment of modifier solution (Table 1 Program 11); EPC electroplating of the modifier onto the graphite surface EG - tubes PC-tubes No No Element modifier Pd-TP Pd-EPC modifier Rh-TP Rh-EPC As 22 18 18 21 19 16 Se 28 30 26 32 30 28 Si 340 94 58 210 90 52 Journal of Analytical Atomic Spectrometry January 1995 Vol. 10 51area being covered by the modifier and/or by the different more even and dense coverage of the graphite surface. Analytical lifetime of the modified tubes In view of the improved performance of the electroplated graphite tubes relatively long useful lifetimes were achieved. In the case of Rh-plated tubes the integrated absorbance for the determination of Si remained the same [relative standard deviation (RSD) = 3% for 10 determinations] for about 100 firings (Table 6).However after 100 atomization cycles the integrated absorbance decreased to the level obtained with uncoated graphite tubes. Microscopic studies of the graphite tubes with and without plating Fig. 3 shows three types of electrographite tubes at 200x magnification an unused EC tube; an EPC tube after Pd-plating cycle; and an EPC tube after 60 firings. Microscopic images of the surface at 200 x magnification are not as detailed as scanning electron micrographs which can be found in a number of publication^.'^.^^ However the important infor- mation relating to the history of the surface can be clearly evaluated from the images presented.Two images [Fig. 3(a) and (b)] show the difference between the surface of an unused EC tube before and after Pd-plating. The surface showed on Fig. 3(b) is much smoother compared with that of the new tube [Fig. 3(a)] and shows the deposition of a Pd metallic layer onto the tube after the plating process (as described under Experimental). Fig. 3(c) which shows an image of the Pd-plated surface after 60 firings demonstrates that the coating decreases substantially exposing the initial structure of the electrographite surface. This may explain the experimental results obtained for Si an element which is particularly sensitive to the quality of the tube surface. The Si sensitivity test indicates a gradual decrease of integrated absorbance after about 55-60 atomization cycles.CONCLUSION The practical advantages of the use of electroplated tubes were illustrated through thermal stabilization studies characteristic mass values and analytical life-time of the plating coating. An empirical investigation of the thermal stability of the metals investigated showed that both Pd and Rh permitted signifi- cantly higher pyrolysis temperatures for Se and As. In the case of Si both modifiers improved the sensitivity by a factor of 4.5 for Pd and a factor of 3 for Rh. It can be assumed that in the pre-injection mode the reducing properties of hot graphite would reduce Pd and Rh to the elemental form. Once reduced both metals become very effective modifiers. However this investigation showed that pre-injection of the modifier solution into the graphite tube and heating to 1000°C was not nearly as effective as electro- reduction in an electrolytic-bath.The graphite surface pre- treated by plating with Pd or Rh may resemble the behaviour of the corresponding modifiers and therefore act as a solid chemical modifier. Microscopic images show that in the case of Pd the metallic layer is stable up to about 60 atomization cycles. In view of the improved performance of the electroplated Table 6 Analytical lifetime (firings) of improved tubes performance Element Pd-thermal Pd-plating Rh-thermal Rh-plating As 8 80 20 160 Se 10 80 20 160 Si 1 60 10 100 (a ) Fig. 3 Surfaces of the tubes used in HGA-400 atomizer made from electrographite (a) surface of an unused EC tube; (b) EPC tube surface after Pd-plating cycle followed by heating to 2000°C; (c) EPC tube surface after 60 firings under condition for silicon determination (Table 1 Programme I) coating of the graphite tube low characteristic masses were obtained for both modifiers and a relatively long useful lifetime was achieved in the case of Rh-plating.Pyrolysis temperatures of 1400 "C-1600 "C were possible for the elements investigated in this study. Obviously before a chemical modifier can be widely used it must be carefully evaluated in a variety of sample matrices; detailed investigations are in progress. This research was carried out as part of the BST 472/2/94 project. 52 Journal of Analytical Atomic Spectrometry January 1995 Vol. 10REFERENCES 1 Slavin W. Manning D. C. and Carnrick G.R. At. Spectrosc. 1981 2 135. 2 Tsalev D. L. Slaveykova V. I. and Mandjukov P. B. Spectrochim. Acta Rev. 1990 13 225. 3 Shan X.-Q. and Ni Z-M Acta Chim. Sin. 1979,37 261. 4 Welz B. Schlemmer G. and Mudakavi J. R. J. Anal. At. Spectrom. 1987 2 45. 5 Schlemmer G. and Welz B. Spectrochim. Acta. Part B 1986 41 1157. 6 Welz B. Schlemmer G. and Mudakavi J. R. J. Anal. At. Spectrorn. 1988 3 93. 7 Welz B. Schlemmer G. and Mudakavi J. R. J. Anal. At. Spectrom. 1988 3 695. 8 Welz B. Schlemmer G. and Mudakavi J. R. Anal. Chem. 1988 60 2567. 9 Voth-Beach L. M. and Shrader D. E. J. J. Anal. At. Spectrom. 1987 2 45. 10 Bulska E. Grobenski Z. and Schlemmer G. J. Anal. At. Spectrom. 1990 5 203. 11 Knowles M. B. and Brodie K. G. J. Anal. At. Spectrom. 1988 3 511. 12 13 14 15 16 17 18 19 20 Knowles M. B. and Brodie K. G. J. Anal. At. Spectrom. 1989 4 305. Volynsky A. Tikhomirov S. and Elagin A. Analyst 1991 116 145. Ortner H. M. and Kantuscher E. Talanta 1975 22 581. Fritzche H. Wegscheider W. Knapp G. and Ortner H. M. Talanta 1979 26 219. Fuchs-Pohl G. R. Solinska K. and Feig H. Fresenius J. Anal. Chem. 1992,343 711. Michaelis M. R. A. Wegscheider W. and Ortner H. M. J. Anal. At. Spectrom. 1988 3 503. Shuttler I. L. Feuerstein M. and Schlemmer G. J. Anal. At. Spectrom. 1992 7 1299. Frech W. and Cedergren A. Anal. Chim. Acta 1980 113 227. Rademeyer C. J. and Vermaak I. J. Anal. At. Spectrom. 1992 7 347. 21 Tsalev D. and Slaveykova I. Spectrosc. Lett. 1992 25 221. 22 Ortner H. M. Schlemmer G. Welz B. and Wegscheider W. Spectrochim. Acta Part B 1985 40 959. Paper 4/02 755 H Received May 10 1994 Accepted August 31 1994 Journal of Analytical Atomic Spectrometry January 1995 Vol. 10 53