- Email: [email protected]
Experimental parameters of vanadium determination by atomic-absorption spectroscopy with graphite-furnace atomization
Talanta. Vol. 27, pp. 214 to 216 O Pergamon Press Lid 1980. Printed in Great Britain
0039-9140/80/0201-0214102.00/0
EXPERIMENTAL PARAMETERS OF V A N A D I U M DETERMINATION BY ATOMIC-ABSORPTION SPECTROSCOPY WITH GRAPHITE-FURNACE ATOMIZATION ADAM HULANICKI, REGINA KARWOWSKAand JOANNA STANCZAK Institute of Fundamental Problems in Chemistry, University of Warsaw, Warsaw, Poland (Received 6 April 1979. Accepted 17 Auoust 1979) Summary--It was found that impregnation of a graphite cuvette (HGA-72) with salts of elements which form stable carbides (Ta, Si, Nb, Zr, W, La) decreases the absorbance signal for vanadium. The slope of the atomization curves indicates that formation of vanadium atoms is inhibited, probably by formation of a ternary compound between the impregnating element, vanadium and graphite. On the contrary bigger signals and better repeatability of results may be achieved when the cuvette is coated with pyrolytic graphite and methane is added to the sheath gas. The presence of methane increases the atomization efficiency and compensates for the disadvantageous influence of any air present in the sheath gas.
The advantage of the electrothermal atomization of technique of atomic-absorption spectroscopy in determination of metals in low concentrations and small samples has stimulated investigations on determination of vanadium by this technique in various materials such as sea-water, t air," biological samples) steep and most commonly in petrochemicals. s-s The authors of these papers have pointed out numerous effects which influence the absorbance of vanadium. It is mentioned that formation of vanadium carbides in graphite cuvettes 9 makes the detection limit higher and that this harmful effect can be diminished by covering the graphite surface with pyrolytic c a r b o n ) ° It is also suggested that a similar effect may be achieved by impregnating the graphite atomizer with other carbide-forming elements.t t Such a procedure has been used by Ortner and Kantuscher ~2 for determination of silicon and by Runnels et al) 3 for beryllium, manganese, chromium and aluminium. Nevertheless there are lacking more systematic studies of the atomization parameters in vanadium determination. In this work the effect of modification of the graphite cuvette surface and composition of the gas atmosphere on vanadium determination have been investigated.
metavanadate in dilute ammonia, acidifying with conc. nitric acid and diluting to volume. Working solutions were obtained by dilution with 0.01M hydrochloric acid. Sodium tungstate, sodium silicate, zirconium nitrate and lanthanum chloride solutions were prepared, containing 1% of the element of interest. Niobium chloride solution (1% Nb) was prepared as a colloidal suspension. Tantalum powder (1 g) was dissolved in 5 mt of hydrofluoric acid, 10 ml of conc. hydrochloric acid and 10 mi of 30?/0 hydrogen peroxide, then 5 ml of conc. sulphuric acid were added, and the solution was evaporated to a small volume and diluted to 100 ml. Measurements Standard measurements were made on 50-~1 samples containing 1 ppm of vanadium, dried at 110° for 75 sec, ashed for 20 sec at 1700°, and atomized during 30 sec at 2660~. Atomization curves for 100 ng of vanadium were recorded over the temperature range from 1750° to 2660 ° (rate of heating 2400°/min°). The absorbance was measured at 318.4 nm, with a 0.7 nm band-pass, at a lamp current of 30 mA. The rate of argon flow was 1.5 l./min and other gases were mixed with argon by means of a peristaltic pump. In the background correction the absorbance of a blank was taken into account. Cuvettes were impregnated by injection of solutions which contained 1-5 mg of the element, drying at 100° and then heating to 2660 ° at 1200°/min. In the case of tantalum the procedure followed that of Ortner and Kantuscher, 12 i.e., the cuvette was soaked in tantalum solution for 24 hr, dried for 12 hr at 120° and finally heated to 2660 °.
EXPERIMENTAL RESULTS AND DISCUSSION
Apparatus • Perkin-Elmer model 300 atomic-absorption spectrophotometer with HGA 72 graphite atomizer and Hitachi Perkin-Elmer model 159 recorder. Perkin-Elmer hollowcathode lamp. Reaoents and solutions Vanadium standard stock solution (1 mg/ml) was prepared by dissolving the appropriate amount of ammonium
Impregnated cuvettes Following the investigations of other authors 12'13 the effect of Si, Zr, Ta, W, N-b and La on determination of vanadium was investigated. As shown in Table 1, the absorbance of vanadium was seriously depressed by the presence of lanthanum or tungsten, and this may be attributed to mechanical losses on 214
215
SHORTCOMMUNICATIONS Table 1. Effect of various elements on absorbance of 50 ng of vanadium in 0.02M HCI --
Absorbance with I mg of element added to the sample Absorbance with cuvette impregnated with element indicated
the particles of tungsten and lanthanum salts. Suchl an effect was not observed when the cuvette was impregnated with these (and the other salts investigated) before the sample was injected. Nevertheless, in the case of some cuvettes the absorbance was considerably lowered by impregnation of the atomizer with tungsten whereas impregnation with lanthanum or silicon caused little lowering of the signal (Table 1). It seems more probable that such behaviour is due to formation of three-component systems containing carbon and two other elements (one of them vanadium) than to modification of the graphite structure.It The formation of V-Ti-C, V-Nb-C, V-Ta-C and V-Hf-C systems has been mentioned in the literature; 14 change in the surface of the cuvette would be expected to increase the absorbance rather than decrease it. Similar conclusions may be reached from study of the atomization curves (Fig. 1). Comparison of the atomizatiofi rate in untreated and impregnated cuvettes indicates decrease of the peak value and of area under the curve, and delay of the appearance maximum when the impregnated cuvettes are used. For untreated cuvettes the maximum appears in the range 2600-2620°, whereas impregnation shifts it to 2660 ° i.e., to the highest temperature obtained with our instrument.
Si
La
Ta
Zr
Nb
0.40 0.12 0.045 0.15 0.38 0.38 0 . 2 8 0.31 0.32
formed in the furnace extends the life-time of the cuvettes and compensates for the presence of oxidants. This is important in the case of vanadium, for which a rather high atomization temperature is needed. The absorbance of vanadium in the presence of methane (Figs. 2 and 3) added at a flow-rate between 1.2 and 5.4 ml/min is larger than that with pure argon. Methane influences the atomization eft]-
0.70
/
/
0.60
/'(
0.40
~
I
/
/
/
/
0.50
.,*'°
i! °,'°
..'1
...'"
..2
I
030 Jg
020
The effect of pyrolytic graphite
0.10
Decomposition of methane at ca. 2200° produces pyrolytic graphite on the atomizer surface, and this advantageously influences the peak height and area without, however, changing the temperature of occurrence of the peak. The effect of the presence of methane in the argon sheath-gas was observed by Kantor et al. is who noted that reaction of methane with water vapour
W
I.!"
,b 2b
4b
io
Amount of V, ng
Fig. 2. Calibration curves for vanadium, l--Untreated cuvette; 2--cuvette covered with pyrolytic graphite; 3--covered with pyrolytic graphite, with methane flow (5.4 ml/min) during atomization. 3
O. 50
I
4
2500
O.4C 0.3C
// - ' " "
020
r
O.lO 0
L _ 5
...
" .
® 0.60
2300 u
2100
04(3
2100 ~.
1900
0.2C
1900
2
~
I0 15 20 25 30 35 40 45
Fig. 1, Atomization curves (temperature increase 2400°/rain) for 100 rig of V. I--Untreated cuvette; 2--impregnated with 1 mg of W; 3--impregnated with 1 mg of Ta; 4---temperature change during atomization. 27/2--J
2500
2300
Time, se c
TAL
0.8(
0
5
I0 15 20 ;=5 30 35 40 45 Time, sec
Fig. 3. Atomization curves for 100 ng of V in cuvettes covered with pyrolytic graphite. 1--Pure argon flow; 2--argon with addition of methane (5.4 ml/min); 3 - - t e m perature change during atomization.
216
SHORTCOMMUNICATIONS I
4
0.901
PSOO
0.70
i
g~ i .
Jl
"
o.6o
23oo 3
,("" - ,
0.4 5
-,.
2100
19oo ff
0.1
o.15
0 0 .g
I,'8
3'.0 4'. 2 ,5.4 Flow rote, mi/rain
,.5
I0
15 20
25 30 35 40 45
Time,sec
616
Fig. 5. Atomization curves for 100 ng of V in 0.01M HCI
Fig. 4. The effect of air added to argon, on the absorbance of 50 ng of V in 0.01M HCI.
in cuvettes covered with pyrolytic graphite. 1--argon gas; 2--argon + air (2.7 ml/min); 3--argon + air (2.7 ml/ rain) + methane (5.4 ml/min).
ciency, and the peak occurs earlier, at 2600 °. On the basis of the mechanism postulated by Sturgeon and Chakrabarti 1~ it may be supposed that dissociation of gaseous vanadium oxide is shifted towards formation of free metal atoms by the reaction promoted between oxygen and methane:
atomizer surface. The most probable explanation seems to be the increase of the atomization rate through the additional exothermai process between methane and oxygen, which raises the vapour temperature in the cuvette. This is similar to the effect of hydrogen, which increases the temperature by about 100° at 1700°.
VO---,V + O O + CH4 ~ 2H2 + CO
(l) (2)
The effect t~'air in the sheath gas Small quantities of air may occur in the argon as an impurity or because of diffusion from the surroundings. Therefore the effect of the presence of air was tested by mixing small amounts with the argon with the aid of a peristaltic pump. Increasing the amount of air decreases the absorbance of vanadium (Fig. 4). When air is present in the argon the atomization curve shows a decrease in both peak height and area (Fig. 5). Also, maximal absorbance occurs at higher temperature, which may be explained by the unfavourable shift of the equilibrium of reaction (l). To compensate for this, methane was also added to the argon. Under these conditions the maximum of the absorbance appears at 2420 °, which is about 200 ° lower than without methane added. The height of the peak and its area increase, favourably influencing the results. This indicates that atomization in these conditions is faster and more efficient. Such an effect can be explained neither by a change of atomization mechanism nor by a change in graphite structure on the
REFERENCES
!. D, A. Segar and J. G. Gonzales, Anal. Chim. Acta, 1977, 58, 714. 2. A. Kamiya, Eisei Kagaku, 1975, 21,267. 3. P, Schramel, Anal. Chim, Acta, 1973, 67, 69. 4. F. Shaw and J. Ottaway, Analyst, 1975, 100, 217. 5. S. H. Omang, Anal. Chim. Acta, 1971, 56, 470. 6. K. G. Brodie and J. P. Matousek, Anal. Chem., 1971, 43, 1557. 7. C. L. Chakrabarti and G. Hall, Spectrosc. Lett., 1973, 6, 385. 8. L. A. May and B. J. Presley, ibid., 1975, 8, 20l. 9. B. Welz, Atom-Absorptians-Spektroskopie, Verlag Chemie, Weinheim, 1975. 10. C. W. Fuller, Analyst, 1976, 101,798. 1!. R. E. Sturgeon and C. L. Chakrabarti, Anal." Chem., 1977, 49, 90. 12. H. M. Ortner and E. Kantuscher, Talanta, 1975, 22, 581. 13. J. H. Runnels, R. Merryfield and H. B. Fischer, Anal, Chem., 1975, 47, 1258. 14. T. A. Nikolskaya and R. G. Awarbe, Vyssokotemperaturnye Karbidy (High-temperature carbides), p. 58, Ukrainian Academy of Sciences, Institute of Material Science, Naukova Dumka, Kiev, 1975. 15. T. Kantor, S. A. Clyburn and C. Veillon, Anal. Chem., 1974, 46, 2205.
0039-9140/80/0201-0214102.00/0
EXPERIMENTAL PARAMETERS OF V A N A D I U M DETERMINATION BY ATOMIC-ABSORPTION SPECTROSCOPY WITH GRAPHITE-FURNACE ATOMIZATION ADAM HULANICKI, REGINA KARWOWSKAand JOANNA STANCZAK Institute of Fundamental Problems in Chemistry, University of Warsaw, Warsaw, Poland (Received 6 April 1979. Accepted 17 Auoust 1979) Summary--It was found that impregnation of a graphite cuvette (HGA-72) with salts of elements which form stable carbides (Ta, Si, Nb, Zr, W, La) decreases the absorbance signal for vanadium. The slope of the atomization curves indicates that formation of vanadium atoms is inhibited, probably by formation of a ternary compound between the impregnating element, vanadium and graphite. On the contrary bigger signals and better repeatability of results may be achieved when the cuvette is coated with pyrolytic graphite and methane is added to the sheath gas. The presence of methane increases the atomization efficiency and compensates for the disadvantageous influence of any air present in the sheath gas.
The advantage of the electrothermal atomization of technique of atomic-absorption spectroscopy in determination of metals in low concentrations and small samples has stimulated investigations on determination of vanadium by this technique in various materials such as sea-water, t air," biological samples) steep and most commonly in petrochemicals. s-s The authors of these papers have pointed out numerous effects which influence the absorbance of vanadium. It is mentioned that formation of vanadium carbides in graphite cuvettes 9 makes the detection limit higher and that this harmful effect can be diminished by covering the graphite surface with pyrolytic c a r b o n ) ° It is also suggested that a similar effect may be achieved by impregnating the graphite atomizer with other carbide-forming elements.t t Such a procedure has been used by Ortner and Kantuscher ~2 for determination of silicon and by Runnels et al) 3 for beryllium, manganese, chromium and aluminium. Nevertheless there are lacking more systematic studies of the atomization parameters in vanadium determination. In this work the effect of modification of the graphite cuvette surface and composition of the gas atmosphere on vanadium determination have been investigated.
metavanadate in dilute ammonia, acidifying with conc. nitric acid and diluting to volume. Working solutions were obtained by dilution with 0.01M hydrochloric acid. Sodium tungstate, sodium silicate, zirconium nitrate and lanthanum chloride solutions were prepared, containing 1% of the element of interest. Niobium chloride solution (1% Nb) was prepared as a colloidal suspension. Tantalum powder (1 g) was dissolved in 5 mt of hydrofluoric acid, 10 ml of conc. hydrochloric acid and 10 mi of 30?/0 hydrogen peroxide, then 5 ml of conc. sulphuric acid were added, and the solution was evaporated to a small volume and diluted to 100 ml. Measurements Standard measurements were made on 50-~1 samples containing 1 ppm of vanadium, dried at 110° for 75 sec, ashed for 20 sec at 1700°, and atomized during 30 sec at 2660~. Atomization curves for 100 ng of vanadium were recorded over the temperature range from 1750° to 2660 ° (rate of heating 2400°/min°). The absorbance was measured at 318.4 nm, with a 0.7 nm band-pass, at a lamp current of 30 mA. The rate of argon flow was 1.5 l./min and other gases were mixed with argon by means of a peristaltic pump. In the background correction the absorbance of a blank was taken into account. Cuvettes were impregnated by injection of solutions which contained 1-5 mg of the element, drying at 100° and then heating to 2660 ° at 1200°/min. In the case of tantalum the procedure followed that of Ortner and Kantuscher, 12 i.e., the cuvette was soaked in tantalum solution for 24 hr, dried for 12 hr at 120° and finally heated to 2660 °.
EXPERIMENTAL RESULTS AND DISCUSSION
Apparatus • Perkin-Elmer model 300 atomic-absorption spectrophotometer with HGA 72 graphite atomizer and Hitachi Perkin-Elmer model 159 recorder. Perkin-Elmer hollowcathode lamp. Reaoents and solutions Vanadium standard stock solution (1 mg/ml) was prepared by dissolving the appropriate amount of ammonium
Impregnated cuvettes Following the investigations of other authors 12'13 the effect of Si, Zr, Ta, W, N-b and La on determination of vanadium was investigated. As shown in Table 1, the absorbance of vanadium was seriously depressed by the presence of lanthanum or tungsten, and this may be attributed to mechanical losses on 214
215
SHORTCOMMUNICATIONS Table 1. Effect of various elements on absorbance of 50 ng of vanadium in 0.02M HCI --
Absorbance with I mg of element added to the sample Absorbance with cuvette impregnated with element indicated
the particles of tungsten and lanthanum salts. Suchl an effect was not observed when the cuvette was impregnated with these (and the other salts investigated) before the sample was injected. Nevertheless, in the case of some cuvettes the absorbance was considerably lowered by impregnation of the atomizer with tungsten whereas impregnation with lanthanum or silicon caused little lowering of the signal (Table 1). It seems more probable that such behaviour is due to formation of three-component systems containing carbon and two other elements (one of them vanadium) than to modification of the graphite structure.It The formation of V-Ti-C, V-Nb-C, V-Ta-C and V-Hf-C systems has been mentioned in the literature; 14 change in the surface of the cuvette would be expected to increase the absorbance rather than decrease it. Similar conclusions may be reached from study of the atomization curves (Fig. 1). Comparison of the atomizatiofi rate in untreated and impregnated cuvettes indicates decrease of the peak value and of area under the curve, and delay of the appearance maximum when the impregnated cuvettes are used. For untreated cuvettes the maximum appears in the range 2600-2620°, whereas impregnation shifts it to 2660 ° i.e., to the highest temperature obtained with our instrument.
Si
La
Ta
Zr
Nb
0.40 0.12 0.045 0.15 0.38 0.38 0 . 2 8 0.31 0.32
formed in the furnace extends the life-time of the cuvettes and compensates for the presence of oxidants. This is important in the case of vanadium, for which a rather high atomization temperature is needed. The absorbance of vanadium in the presence of methane (Figs. 2 and 3) added at a flow-rate between 1.2 and 5.4 ml/min is larger than that with pure argon. Methane influences the atomization eft]-
0.70
/
/
0.60
/'(
0.40
~
I
/
/
/
/
0.50
.,*'°
i! °,'°
..'1
...'"
..2
I
030 Jg
020
The effect of pyrolytic graphite
0.10
Decomposition of methane at ca. 2200° produces pyrolytic graphite on the atomizer surface, and this advantageously influences the peak height and area without, however, changing the temperature of occurrence of the peak. The effect of the presence of methane in the argon sheath-gas was observed by Kantor et al. is who noted that reaction of methane with water vapour
W
I.!"
,b 2b
4b
io
Amount of V, ng
Fig. 2. Calibration curves for vanadium, l--Untreated cuvette; 2--cuvette covered with pyrolytic graphite; 3--covered with pyrolytic graphite, with methane flow (5.4 ml/min) during atomization. 3
O. 50
I
4
2500
O.4C 0.3C
// - ' " "
020
r
O.lO 0
L _ 5
...
" .
® 0.60
2300 u
2100
04(3
2100 ~.
1900
0.2C
1900
2
~
I0 15 20 25 30 35 40 45
Fig. 1, Atomization curves (temperature increase 2400°/rain) for 100 rig of V. I--Untreated cuvette; 2--impregnated with 1 mg of W; 3--impregnated with 1 mg of Ta; 4---temperature change during atomization. 27/2--J
2500
2300
Time, se c
TAL
0.8(
0
5
I0 15 20 ;=5 30 35 40 45 Time, sec
Fig. 3. Atomization curves for 100 ng of V in cuvettes covered with pyrolytic graphite. 1--Pure argon flow; 2--argon with addition of methane (5.4 ml/min); 3 - - t e m perature change during atomization.
216
SHORTCOMMUNICATIONS I
4
0.901
PSOO
0.70
i
g~ i .
Jl
"
o.6o
23oo 3
,("" - ,
0.4 5
-,.
2100
19oo ff
0.1
o.15
0 0 .g
I,'8
3'.0 4'. 2 ,5.4 Flow rote, mi/rain
,.5
I0
15 20
25 30 35 40 45
Time,sec
616
Fig. 5. Atomization curves for 100 ng of V in 0.01M HCI
Fig. 4. The effect of air added to argon, on the absorbance of 50 ng of V in 0.01M HCI.
in cuvettes covered with pyrolytic graphite. 1--argon gas; 2--argon + air (2.7 ml/min); 3--argon + air (2.7 ml/ rain) + methane (5.4 ml/min).
ciency, and the peak occurs earlier, at 2600 °. On the basis of the mechanism postulated by Sturgeon and Chakrabarti 1~ it may be supposed that dissociation of gaseous vanadium oxide is shifted towards formation of free metal atoms by the reaction promoted between oxygen and methane:
atomizer surface. The most probable explanation seems to be the increase of the atomization rate through the additional exothermai process between methane and oxygen, which raises the vapour temperature in the cuvette. This is similar to the effect of hydrogen, which increases the temperature by about 100° at 1700°.
VO---,V + O O + CH4 ~ 2H2 + CO
(l) (2)
The effect t~'air in the sheath gas Small quantities of air may occur in the argon as an impurity or because of diffusion from the surroundings. Therefore the effect of the presence of air was tested by mixing small amounts with the argon with the aid of a peristaltic pump. Increasing the amount of air decreases the absorbance of vanadium (Fig. 4). When air is present in the argon the atomization curve shows a decrease in both peak height and area (Fig. 5). Also, maximal absorbance occurs at higher temperature, which may be explained by the unfavourable shift of the equilibrium of reaction (l). To compensate for this, methane was also added to the argon. Under these conditions the maximum of the absorbance appears at 2420 °, which is about 200 ° lower than without methane added. The height of the peak and its area increase, favourably influencing the results. This indicates that atomization in these conditions is faster and more efficient. Such an effect can be explained neither by a change of atomization mechanism nor by a change in graphite structure on the
REFERENCES
!. D, A. Segar and J. G. Gonzales, Anal. Chim. Acta, 1977, 58, 714. 2. A. Kamiya, Eisei Kagaku, 1975, 21,267. 3. P, Schramel, Anal. Chim, Acta, 1973, 67, 69. 4. F. Shaw and J. Ottaway, Analyst, 1975, 100, 217. 5. S. H. Omang, Anal. Chim. Acta, 1971, 56, 470. 6. K. G. Brodie and J. P. Matousek, Anal. Chem., 1971, 43, 1557. 7. C. L. Chakrabarti and G. Hall, Spectrosc. Lett., 1973, 6, 385. 8. L. A. May and B. J. Presley, ibid., 1975, 8, 20l. 9. B. Welz, Atom-Absorptians-Spektroskopie, Verlag Chemie, Weinheim, 1975. 10. C. W. Fuller, Analyst, 1976, 101,798. 1!. R. E. Sturgeon and C. L. Chakrabarti, Anal." Chem., 1977, 49, 90. 12. H. M. Ortner and E. Kantuscher, Talanta, 1975, 22, 581. 13. J. H. Runnels, R. Merryfield and H. B. Fischer, Anal, Chem., 1975, 47, 1258. 14. T. A. Nikolskaya and R. G. Awarbe, Vyssokotemperaturnye Karbidy (High-temperature carbides), p. 58, Ukrainian Academy of Sciences, Institute of Material Science, Naukova Dumka, Kiev, 1975. 15. T. Kantor, S. A. Clyburn and C. Veillon, Anal. Chem., 1974, 46, 2205.