- Email: [email protected]
Direct determination of boron in a cobalt-based alloy by graphite furnace-atomic absorption spectrometry
Talanta ELSEVIER
Talanta 42 (1995) 1419-1423
Direct determination of boron in a cobalt-based alloy by graphite furnace-atomic absorption spectrometry Benling Gong *, Yongming Liu, Yuli Xu, Zhuanhe Li, Tiezheng Lin Dalian Institute of Chemistry Physics, Academia Sinica, 161 Zhongshan Road, Dalian, People's Republic of China
Received 10 November 1994; revised 28 February 1995; accepted 10 March 1995
Abstract
A matrix modifier composed of nickel and zirconium, and a graphite tube treated with zirconium solution were proposed for the determination of boron in cobalt-based alloys by graphite furnace-atomic absorption spectrometry. The effects of this matrix modifier and the treated graphite tube were studied, and the combination of 60 lag of nickel and 20 lag of zirconium as matrix modifier, and a graphite tube soaked with 10 g 1-~ of zirconium solution were found to give the highest analytical sensitivity. The interference effects of major components (cobalt) and eight minor components (chromium, nickel, tungsten, iron, tantalum, molybdenum, titanium, aluminium and manganese) were studied. Boron in four cobalt-based alloys was determined by graphite furnace-atomic absorption spectrometry employing the proposed matrix modifier and the treated graphite tube, without the preseparation of matrix. The relative standard deviation was 3.3% for 0.048% of boron. A characteristic mass was 500 pg.
1. Introduction
The addition of appropriate amounts of some elements improves the performance of steels and alloys appreciably. A microamount of boron increases the degree of quenching and the strength and toughness of steels and alloys. However, excess amounts of boron may cause their performance to deteriorate. Therefore the exact determination of boron in steels and alloys is required for metallurgical applications. Cobalt-based alloys contain many components (chromium, nickel, tungsten, aluminium, tantalum, iron, manganese, silicon, carbon, etc.), as well as its main component (cobalt) and boron. Such a complex composition causes severe spectral and chemical interferences in the determination of boron by atomic spectroscopy. Although separation of the matrix from the sample solution before the measurements are performed and overcome these inter-
* Corresponding author. Fax: (86)411-363-2426.
ferences, this requires a tedious treatment, easily resulting in either contamination or loss. Therefore it is preferable to determine boron in such alloys by direct measurements without the preseparation o f the matrix from the sample solution. Inductively coupled plasma-atomic emission spectrometry has been widely used for the determination of microamounts o f boron [1 - 11], but this method is unsuitable for direct determinations because it necessitates preseparating the matrix from the sample solution in order to remove spectral interferences. Graphite furnace-atomic absorption spectrometry, another sensitive method, also suffers severely from matrix interferences. Moreover, boron combines very easily with the carbon of the graphite tube, resulting in low analytical results and severe memory effects. Nevertheless, the selective application of a suitable matrix modifier and the pretreatment of the graphite tube could resolve the above problems. T o date, some matrix modifiers such as
0039-9140/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved SSDi 0039-9140(95)01585-X
1420
B. Gong et al. / Talanta 42 (1995) 1419-1423
Table 1 Instrumental parameters and operating conditions Parameter/conditions
Value
Wavelength (nm) Slit width (nm) Lamp current (mA) Integration time (s) Mode Injected sample volume (~tl)
249.7 0.7 15 6 Peak height 20
Heating programme Parameter/conditions
Temperature (*C) Ramp time (s) Hold time (s) Read time (s) Internal gas flow rate (ml min -~)
Step I
2
3
120 10 20 300
1200 10 20 300
2700
20
2650
1
1
1
I
3
3
3
3
50
300
300
300
calcium [12,13], strontium [14], magnesium [13,14], nickel [15] and some pretreatment methods for the graphite tube have already been proposed. In this work, a mixture of nickel and zirconium as a matrix modifier, and the coating of the graphite tube with zirconium solution are proposed, because zirconium was often used for coating the graphite tube to protect against the formation of carbide during the analyses of elements that easily form carbides, and nickel was one of the commonly used matrix modifiers. It was found that this combination of a mixed matrix modifier and a zirconiumcoated graphite tube resulted in excellent performance, increasing the analytical sensitivity and suppressing the memory effect for the determination of boron. With the application of this mixed matrix modifier and a zirconiumcoated graphite tube, satisfactory results for the determination of boron in cobalt-based alloys by graphite furnace-atomic absorption spectrometry without the preseparation of the matrix were obtained.
2. Experimental 2.1. Apparatus A Perkin-Elmer model 5000 atomic absorption spectrometer equipped with a model HGA 500 graphite furnace, a Data Station 10 and a PR I00 printer was used. An Eppendorf mi-
4
5
6 20
cropipette was used for injecting the sample solution into the graphite furnace. A pyrolytic graphite tube was soaked in a 10 g 1-i solution of zirconium for 2 days, then dried at 120 °C for 2 h and stored in a desiccator. Such a tube can be used for 200-300 firings. Alternatively, the pyrolytic graphite tube was placed in the graphite furnace, 20 lal of 10gl -~ zirconium solution were injected and one heat cycle of the heating programme was performed. This operation was repeated five times. The life of the tube treated using the latter method was shorter than that of the tube treated using the soaking method. Instrumental parameters and operating conditions were optimized by means of experiments, and are shown in Table 1.
2.2. Reagents All chemicals used were of analytical grade, and distilled, deionized water was used throughout. A stock standard solution of boron (1000 mg l - t ) w a s prepared by dissolving boric acid in water, and was stored in a polyethylene bottle. The working standard solution (10mgl -t) was prepared by dilution of the stock standard solution with 0.2% nitric acid solution. A 20 g 1- t solution of nickel was prepared by dissolving 9.90 g of Ni(NO3)2 • 2H20 in 100 ml of water.
B. Gong et al. / Talanta 42 (1995) 1419-1423
1421
sample completely until some carbon particles only remained in the sample solution. After cooling, the sample solution was transferred to a 20 ml graduated test tube with the minimum volume o f 0.2% nitric acid solution. A yellow precipitate of tungstic acid may stick on the beaker bottom, but this did not affect the analytical results. The sample solution was neutralized to pH 5 - 7 carefully with concentrated ammonia solution, and was diluted to the mark with water. After shaking, a 5 ml aliquot of this solution was pipetted into a l0 ml graduated test tube containing 1.5 ml of a 20 g l-~ nickel solution, and the solution was diluted to the mark with water. After shaking well, a 20 lal aliquot o f this final sample solution and 101al o f a 2 0 0 0 m g i - ' solution of zirconium were injected into the graphite furnace, and the peak height was measured.
A 20 g 1-I solution o f zirconium was prepared by dissolving 7.06 g o f ZrOCl 2 • 8 H 2 0 in 100 ml o f water. A 20 g l-~ solution o f iron was prepared by dissolving 9.68 g of FeC13 • 6H2 O and diluting to 100 ml with 0.2% nitric acid. A 10 g 1-~ solution o f chromium was prepared by dissolving 1.00g of metallic chromium in a minimum volume of concentrated hydrochioric acid and diluting to 100 ml with water. A l0 g l - ' solution o f cobalt was prepared by dissolving 4.94 g o f Co(NO3)2" 6 H 2 0 and diluting to 100 ml with 0.2% nitric acid, A l0 g l - t solution o f tungsten was prepared by dissolving 17.95g o f N a 2 W O 4 - 2 H 2 0 in about 200 ml o f water, 100 ml of 10% N a O H solution were added, and the solution was diluted to 1 l with water and stored in a polyethylene bottle. A 5 g 1-~ solution of aluminium was prepared by dissolving 0.50 g o f aluminium foil in 10ml o f hydrochloric acid, adding several drops o f nitric acid, and diluting to 100 ml with water. A 5 g 1-i solution o f titanium was prepared by dissolving 1.61 g of TiCI3 in 20ml o f hydrochloric acid and diluting to 100 ml with water. A 2 g 1-~ solution o f manganese was prepared by dissolving 2.00 g o f manganese metal in a minimum volume o f (1:1) HNO3 and diluting to 1 1 with 0.2% HNO3. A 2 g 1- ~ solution of tantalum was prepared by dissolving 3.96 g of TaCI5 in 1 1 o f water. A 1 g 1- ~ solution o f molybdenum was prepared by dissolving 1.00 g of metallic molybdenum in 25 ml of nitric acid and diluting to 1 1 with water. To construct the calibration graphs, 0, 0.5, 1, 1.5 and 2 ml o f a 5 mg 1- J working solution of boron were placed in 10 ml volumetric flasks, 1.5ml o f a 2 0 g l -~ solution of nickel were added to each flask, and the solutions were diluted to the mark with water. The pH values of the final solutions were 5-7.
The effect o f coating the pyrolytic graphite tube was studied for the determination of 500 ~tg l - t o f boron with 60 lag of nickel and 20 ~tg of zirconium as matrix modifier, using a zirconium-coated and an uncoated pyrolytic graphite tube. The results, shown in Table 2, indicated that coating with zirconium enhanced the sensitivity approximately two fold.
2.3. Procedure
Table 2 Effect of coating the graphite tube on the sensitivity~
About 30 mg o f finely granulated sample was weighed exactly in a quartz beaker, 2.5-3.0 ml o f aqua regia were added, and the solution was allowed to stand for at least 1 h at room temperature. The beaker was then heated in a water bath at about 70 *C to decompose the
3. Results and discussion
3.1. Neutralization o f the sample solution to be measured After the decomposition, the sample solution must be carefully neutralized with ammonia solution to pH 5 - 7 to match the pH of the standard solution used for preparing the calibration graph. At this time, a pink precipitate o f cobalt hydroxide may appear. The acidity of the injected sample solution affects the analytical sensitivity considerably, i.e. the sensitivity decreases with increase of acidity. 3.2. Effect o f coating the graphite tube
Type of tube
Absorbance, A
Coated Uncoated
0.090 0.048
" For 500 lag I- ~boron, 60 lag of nickel/20 lag of zirconium modifier.
B. Gong et al. / Talanta 42 (1995) 1419-1423
1422
Table 3 Effects of nickel, zirconium and nickel-zirconium as matrix modifiers on the sensitivity ~
Table 5 Effect of different amounts of zirconium with a constant amount (60 lag) of nickel on sensitivity~
Amount of Ni
Amount of Zr
(lag)
(~g)
0 0 60 60
0 20 0 20
Absorbance. A
Factor
Amount of Zr (lag)
Absorbance, A
0.015 0.024 0.036 0.090
I 1.6 2.4 6
0 20 40
0.036 0.090 0.000
For 500 lag 1- ~ of boron. For 500 lag I- ' of boron.
3.3. Effect o f matrix modifier The bond strength of Zr-C (134 + 6 kcal mol-~) is larger than that of B - C (107 ___7 kcal m o l - ~) [16], so when boron and zirconium are atomized simultaneously, zirconium may combine predominantly with carbon in the graphite tube. This results in the retarding of the combination of boron with carbon. Hence it makes the sensitivity increase, and at the same time causes the memory effect to be suppressed. Nickel was known to be useful as a matrix modifier for many elements, and was found by us to be able to increase the ashing temperature for the analysis of steels and alloys up to 1200 *C. This is undoubtedly useful for decreasing the matrix effect. The effect was studied of a matrix modifier composed o f nickel and zirconium on the determination of 500 lag 1- ~ of boron with a zirconium-coated pyrolytic graphite tube. The results obtained are shown in Tables 3, 4 and 5. In Table 3 it is worth noticing that although both nickel and zirconium have some enhancing effects on the analytical sensitivity, the combination o f these two elements shows a multiple enhancing effect, i.e. a matrix modifier
Table 4
Effect of different amounts of nickel with a constant amount (20 lag) of zirconium on the sensitivity ~ Amount of Ni
Absorbance, A
{lag)
20 60 100 200 a For 500 lag I- ~ of boron.
0,075 0.090 0.042 0.043
composed of nickel and zirconium increases the analytical sensitivity six times, which is higher than the sum (4 times) of the enhancing factor of nickel (2.4 times) and that of zirconium (1.6 times). In Tables 4 and 5 it is seen that nickel and zirconium both have an optimal quantity for achieving the best sensitivity, i.e. a larger or smaller quantity than the optimal one causes the sensitivity to decrease. Combination of these elements leads to the same situation, i.e. the combination of 60 lag of nickel and 20 lag of zirconium gives the highest analytical sensibility. Moreover, the above experiments also show that the m e m o r y effect decreases rapidly with increasing amount of zirconium.
3.4. Interferences of coexisting elements The investigated cobalt-based alloys contain 5 3 - 5 7 % of cobalt, 25% of chromium, 10-14% of nickel, 7 - 7 . 5 % of tungsten, 0.8% of aluminium, 0.15-0.35% of zirconium, 0.25% of tantalum, 0.16% of iron, 0.15% of titanium, 0.07% of silicon, 0.03% of manganese and 0.48-0.8% of carbon, as well as 0.027-0.05% of boron. The interference effects from the above elements except nickel, zirconium, silicon and carbon were studied with use of a matrix modifier composed of 60 lag o f nickel and 20 lag of zirconium, and a zirconium-coated tube. The results are shown in Table 6. It is seen that among these coexisting elements, chromium shows the most severe interference effects relatively, but the use of a matrix modifier composed of nickel and zirconium and a zirconium-coated graphite tube could avoid the interferences from all coexisting elements including chromium, provided the amount of sample does not exceed 35 rag.
1423
B. Gong et al. / Talanta 42 (1995) 1419-1423
Table 6 Permissible amounts of coexisting elements ~ Element
Maximum content in sample (%)
Maximum amount b calculated (mg l - ')
Permissible amount obtained (mg I - ')
Co Cr W Al Ta Fe Ti Mn
57 25 7.5 0.8 0.25 0.16 0.15 0.03
427 187.5 56.25 6 ! .875 1.2 !.i25 0.225
1375 225 375 125 Il0 3000 200 250
" For 500 lag I - ' of boron. b The amount contained in I0 ml of final sample solution prepared from a 30 mg sample. Table 7 Determination of boron in cobalt-based alloys by use of nickel-zirconium as the matrix modifier and a zirconiumcoated graphite tube Sample number
Certified value (%)
Obtained value (%)
Difference (%)
Permissible difference" (%)
Co-68 Co-35
0.027 0.048
0.0269b 0.0484~
0.0001 0.0004
0.005 0.005
Co-70 Co-21
0.049 0.050
0.0490 b 0.0507 b
0.0000 0.0007
0.005 0.005
"Based on the Chinese national standard for the determination of boron in steels and alloys, GB223.6-81, the permissible difference is 0.005% for the content of 0.0260.050% of boron. b Mean of four determinations. c Mean of ten determinations.
3.5. Determination of boron in cobalt-based alloys The proposed method was applied to the determination of boron in four cobalt-based alloys. The results are shown in Table 7. It is seen that the differences between the values obtained and the certified values are all smaller than the permissible differences. The relative standard deviation was 3.3% for 0.0480% boron (n = 10; the values were 0.0475,
0.0472, 0.0463, 0.0493, 0.0507, 0.0487, 0.0508, 0.0465, 0.0480 and 0.0489; with a mean of
0.0480). A characteristic mass was 500 pg.
References [I] G. Mezger, E. Grallath, U. Stix and G. Tolg, Fresenius' Z. Anal. Chem., 317 (1984) 765. [2] V.K. Pin, Anal. Chim. Acta, 159 (1984) 387. [3] J.N. Walsh, Analyst, ! 10 (1985) 959. [4] K. Fujimoto, T. Okano, Y. Matumura and S. Harima, Bunseki Kagaku, 35 (1986) 651. [5] Y. Takahashi, Bunseki Kagaku, 36 (1987) 693. [6] G. Bauer, I. Rehana, W. Wegscheider and H.M. Ortnet, Spectroehim. Acta, part B, 43 (1988) 971. [7] G. Zeibig, Mikroehim. Acta, 1989111, 389. [8] I. Hlavacek, Mikroehim. Acta, 1989111, 309. [9] Hu Bin, Fresenius' Z. Anal. Chem., 340 (1991) 435. [10] J. Ciba and B. Smolec, Fresenius' Z. Anal. Chem., 348 (1994) 215. [!i] T. Miyatani, H. Suzuki and O. Yosimoto, Bunseki Kagaku, 43 (1994) 475. [12] Operational Manual, Perkin-Eimer model HGA-500. [13] Yongqing Jiang, Jinyu Yao and Benli Huang, Fenxi Huaxue, 17 (1989) 456. [14] Yongqing Jiang, Jinyu Yao and Benli Huang, Fenxi Shiyanshi, 7 (1988) 21. [15] M. Luguera, Y. Madrid and C. Camera, J. Anal. At. Spectrom., 6 (1991) 669. [16] R.C. Weast (Ed.), CRC Handbook of Chemistry and Physics, 70th edn., CRC Press inc., Boca Raton, FL, 1990, FI97-198.
Talanta 42 (1995) 1419-1423
Direct determination of boron in a cobalt-based alloy by graphite furnace-atomic absorption spectrometry Benling Gong *, Yongming Liu, Yuli Xu, Zhuanhe Li, Tiezheng Lin Dalian Institute of Chemistry Physics, Academia Sinica, 161 Zhongshan Road, Dalian, People's Republic of China
Received 10 November 1994; revised 28 February 1995; accepted 10 March 1995
Abstract
A matrix modifier composed of nickel and zirconium, and a graphite tube treated with zirconium solution were proposed for the determination of boron in cobalt-based alloys by graphite furnace-atomic absorption spectrometry. The effects of this matrix modifier and the treated graphite tube were studied, and the combination of 60 lag of nickel and 20 lag of zirconium as matrix modifier, and a graphite tube soaked with 10 g 1-~ of zirconium solution were found to give the highest analytical sensitivity. The interference effects of major components (cobalt) and eight minor components (chromium, nickel, tungsten, iron, tantalum, molybdenum, titanium, aluminium and manganese) were studied. Boron in four cobalt-based alloys was determined by graphite furnace-atomic absorption spectrometry employing the proposed matrix modifier and the treated graphite tube, without the preseparation of matrix. The relative standard deviation was 3.3% for 0.048% of boron. A characteristic mass was 500 pg.
1. Introduction
The addition of appropriate amounts of some elements improves the performance of steels and alloys appreciably. A microamount of boron increases the degree of quenching and the strength and toughness of steels and alloys. However, excess amounts of boron may cause their performance to deteriorate. Therefore the exact determination of boron in steels and alloys is required for metallurgical applications. Cobalt-based alloys contain many components (chromium, nickel, tungsten, aluminium, tantalum, iron, manganese, silicon, carbon, etc.), as well as its main component (cobalt) and boron. Such a complex composition causes severe spectral and chemical interferences in the determination of boron by atomic spectroscopy. Although separation of the matrix from the sample solution before the measurements are performed and overcome these inter-
* Corresponding author. Fax: (86)411-363-2426.
ferences, this requires a tedious treatment, easily resulting in either contamination or loss. Therefore it is preferable to determine boron in such alloys by direct measurements without the preseparation o f the matrix from the sample solution. Inductively coupled plasma-atomic emission spectrometry has been widely used for the determination of microamounts o f boron [1 - 11], but this method is unsuitable for direct determinations because it necessitates preseparating the matrix from the sample solution in order to remove spectral interferences. Graphite furnace-atomic absorption spectrometry, another sensitive method, also suffers severely from matrix interferences. Moreover, boron combines very easily with the carbon of the graphite tube, resulting in low analytical results and severe memory effects. Nevertheless, the selective application of a suitable matrix modifier and the pretreatment of the graphite tube could resolve the above problems. T o date, some matrix modifiers such as
0039-9140/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved SSDi 0039-9140(95)01585-X
1420
B. Gong et al. / Talanta 42 (1995) 1419-1423
Table 1 Instrumental parameters and operating conditions Parameter/conditions
Value
Wavelength (nm) Slit width (nm) Lamp current (mA) Integration time (s) Mode Injected sample volume (~tl)
249.7 0.7 15 6 Peak height 20
Heating programme Parameter/conditions
Temperature (*C) Ramp time (s) Hold time (s) Read time (s) Internal gas flow rate (ml min -~)
Step I
2
3
120 10 20 300
1200 10 20 300
2700
20
2650
1
1
1
I
3
3
3
3
50
300
300
300
calcium [12,13], strontium [14], magnesium [13,14], nickel [15] and some pretreatment methods for the graphite tube have already been proposed. In this work, a mixture of nickel and zirconium as a matrix modifier, and the coating of the graphite tube with zirconium solution are proposed, because zirconium was often used for coating the graphite tube to protect against the formation of carbide during the analyses of elements that easily form carbides, and nickel was one of the commonly used matrix modifiers. It was found that this combination of a mixed matrix modifier and a zirconiumcoated graphite tube resulted in excellent performance, increasing the analytical sensitivity and suppressing the memory effect for the determination of boron. With the application of this mixed matrix modifier and a zirconiumcoated graphite tube, satisfactory results for the determination of boron in cobalt-based alloys by graphite furnace-atomic absorption spectrometry without the preseparation of the matrix were obtained.
2. Experimental 2.1. Apparatus A Perkin-Elmer model 5000 atomic absorption spectrometer equipped with a model HGA 500 graphite furnace, a Data Station 10 and a PR I00 printer was used. An Eppendorf mi-
4
5
6 20
cropipette was used for injecting the sample solution into the graphite furnace. A pyrolytic graphite tube was soaked in a 10 g 1-i solution of zirconium for 2 days, then dried at 120 °C for 2 h and stored in a desiccator. Such a tube can be used for 200-300 firings. Alternatively, the pyrolytic graphite tube was placed in the graphite furnace, 20 lal of 10gl -~ zirconium solution were injected and one heat cycle of the heating programme was performed. This operation was repeated five times. The life of the tube treated using the latter method was shorter than that of the tube treated using the soaking method. Instrumental parameters and operating conditions were optimized by means of experiments, and are shown in Table 1.
2.2. Reagents All chemicals used were of analytical grade, and distilled, deionized water was used throughout. A stock standard solution of boron (1000 mg l - t ) w a s prepared by dissolving boric acid in water, and was stored in a polyethylene bottle. The working standard solution (10mgl -t) was prepared by dilution of the stock standard solution with 0.2% nitric acid solution. A 20 g 1- t solution of nickel was prepared by dissolving 9.90 g of Ni(NO3)2 • 2H20 in 100 ml of water.
B. Gong et al. / Talanta 42 (1995) 1419-1423
1421
sample completely until some carbon particles only remained in the sample solution. After cooling, the sample solution was transferred to a 20 ml graduated test tube with the minimum volume o f 0.2% nitric acid solution. A yellow precipitate of tungstic acid may stick on the beaker bottom, but this did not affect the analytical results. The sample solution was neutralized to pH 5 - 7 carefully with concentrated ammonia solution, and was diluted to the mark with water. After shaking, a 5 ml aliquot of this solution was pipetted into a l0 ml graduated test tube containing 1.5 ml of a 20 g l-~ nickel solution, and the solution was diluted to the mark with water. After shaking well, a 20 lal aliquot o f this final sample solution and 101al o f a 2 0 0 0 m g i - ' solution of zirconium were injected into the graphite furnace, and the peak height was measured.
A 20 g 1-I solution o f zirconium was prepared by dissolving 7.06 g o f ZrOCl 2 • 8 H 2 0 in 100 ml o f water. A 20 g l-~ solution o f iron was prepared by dissolving 9.68 g of FeC13 • 6H2 O and diluting to 100 ml with 0.2% nitric acid. A 10 g 1-~ solution o f chromium was prepared by dissolving 1.00g of metallic chromium in a minimum volume of concentrated hydrochioric acid and diluting to 100 ml with water. A l0 g l - ' solution o f cobalt was prepared by dissolving 4.94 g o f Co(NO3)2" 6 H 2 0 and diluting to 100 ml with 0.2% nitric acid, A l0 g l - t solution o f tungsten was prepared by dissolving 17.95g o f N a 2 W O 4 - 2 H 2 0 in about 200 ml o f water, 100 ml of 10% N a O H solution were added, and the solution was diluted to 1 l with water and stored in a polyethylene bottle. A 5 g 1-~ solution of aluminium was prepared by dissolving 0.50 g o f aluminium foil in 10ml o f hydrochloric acid, adding several drops o f nitric acid, and diluting to 100 ml with water. A 5 g 1-i solution o f titanium was prepared by dissolving 1.61 g of TiCI3 in 20ml o f hydrochloric acid and diluting to 100 ml with water. A 2 g 1-~ solution o f manganese was prepared by dissolving 2.00 g o f manganese metal in a minimum volume o f (1:1) HNO3 and diluting to 1 1 with 0.2% HNO3. A 2 g 1- ~ solution of tantalum was prepared by dissolving 3.96 g of TaCI5 in 1 1 o f water. A 1 g 1- ~ solution o f molybdenum was prepared by dissolving 1.00 g of metallic molybdenum in 25 ml of nitric acid and diluting to 1 1 with water. To construct the calibration graphs, 0, 0.5, 1, 1.5 and 2 ml o f a 5 mg 1- J working solution of boron were placed in 10 ml volumetric flasks, 1.5ml o f a 2 0 g l -~ solution of nickel were added to each flask, and the solutions were diluted to the mark with water. The pH values of the final solutions were 5-7.
The effect o f coating the pyrolytic graphite tube was studied for the determination of 500 ~tg l - t o f boron with 60 lag of nickel and 20 ~tg of zirconium as matrix modifier, using a zirconium-coated and an uncoated pyrolytic graphite tube. The results, shown in Table 2, indicated that coating with zirconium enhanced the sensitivity approximately two fold.
2.3. Procedure
Table 2 Effect of coating the graphite tube on the sensitivity~
About 30 mg o f finely granulated sample was weighed exactly in a quartz beaker, 2.5-3.0 ml o f aqua regia were added, and the solution was allowed to stand for at least 1 h at room temperature. The beaker was then heated in a water bath at about 70 *C to decompose the
3. Results and discussion
3.1. Neutralization o f the sample solution to be measured After the decomposition, the sample solution must be carefully neutralized with ammonia solution to pH 5 - 7 to match the pH of the standard solution used for preparing the calibration graph. At this time, a pink precipitate o f cobalt hydroxide may appear. The acidity of the injected sample solution affects the analytical sensitivity considerably, i.e. the sensitivity decreases with increase of acidity. 3.2. Effect o f coating the graphite tube
Type of tube
Absorbance, A
Coated Uncoated
0.090 0.048
" For 500 lag I- ~boron, 60 lag of nickel/20 lag of zirconium modifier.
B. Gong et al. / Talanta 42 (1995) 1419-1423
1422
Table 3 Effects of nickel, zirconium and nickel-zirconium as matrix modifiers on the sensitivity ~
Table 5 Effect of different amounts of zirconium with a constant amount (60 lag) of nickel on sensitivity~
Amount of Ni
Amount of Zr
(lag)
(~g)
0 0 60 60
0 20 0 20
Absorbance. A
Factor
Amount of Zr (lag)
Absorbance, A
0.015 0.024 0.036 0.090
I 1.6 2.4 6
0 20 40
0.036 0.090 0.000
For 500 lag 1- ~ of boron. For 500 lag I- ' of boron.
3.3. Effect o f matrix modifier The bond strength of Zr-C (134 + 6 kcal mol-~) is larger than that of B - C (107 ___7 kcal m o l - ~) [16], so when boron and zirconium are atomized simultaneously, zirconium may combine predominantly with carbon in the graphite tube. This results in the retarding of the combination of boron with carbon. Hence it makes the sensitivity increase, and at the same time causes the memory effect to be suppressed. Nickel was known to be useful as a matrix modifier for many elements, and was found by us to be able to increase the ashing temperature for the analysis of steels and alloys up to 1200 *C. This is undoubtedly useful for decreasing the matrix effect. The effect was studied of a matrix modifier composed o f nickel and zirconium on the determination of 500 lag 1- ~ of boron with a zirconium-coated pyrolytic graphite tube. The results obtained are shown in Tables 3, 4 and 5. In Table 3 it is worth noticing that although both nickel and zirconium have some enhancing effects on the analytical sensitivity, the combination o f these two elements shows a multiple enhancing effect, i.e. a matrix modifier
Table 4
Effect of different amounts of nickel with a constant amount (20 lag) of zirconium on the sensitivity ~ Amount of Ni
Absorbance, A
{lag)
20 60 100 200 a For 500 lag I- ~ of boron.
0,075 0.090 0.042 0.043
composed of nickel and zirconium increases the analytical sensitivity six times, which is higher than the sum (4 times) of the enhancing factor of nickel (2.4 times) and that of zirconium (1.6 times). In Tables 4 and 5 it is seen that nickel and zirconium both have an optimal quantity for achieving the best sensitivity, i.e. a larger or smaller quantity than the optimal one causes the sensitivity to decrease. Combination of these elements leads to the same situation, i.e. the combination of 60 lag of nickel and 20 lag of zirconium gives the highest analytical sensibility. Moreover, the above experiments also show that the m e m o r y effect decreases rapidly with increasing amount of zirconium.
3.4. Interferences of coexisting elements The investigated cobalt-based alloys contain 5 3 - 5 7 % of cobalt, 25% of chromium, 10-14% of nickel, 7 - 7 . 5 % of tungsten, 0.8% of aluminium, 0.15-0.35% of zirconium, 0.25% of tantalum, 0.16% of iron, 0.15% of titanium, 0.07% of silicon, 0.03% of manganese and 0.48-0.8% of carbon, as well as 0.027-0.05% of boron. The interference effects from the above elements except nickel, zirconium, silicon and carbon were studied with use of a matrix modifier composed of 60 lag o f nickel and 20 lag of zirconium, and a zirconium-coated tube. The results are shown in Table 6. It is seen that among these coexisting elements, chromium shows the most severe interference effects relatively, but the use of a matrix modifier composed of nickel and zirconium and a zirconium-coated graphite tube could avoid the interferences from all coexisting elements including chromium, provided the amount of sample does not exceed 35 rag.
1423
B. Gong et al. / Talanta 42 (1995) 1419-1423
Table 6 Permissible amounts of coexisting elements ~ Element
Maximum content in sample (%)
Maximum amount b calculated (mg l - ')
Permissible amount obtained (mg I - ')
Co Cr W Al Ta Fe Ti Mn
57 25 7.5 0.8 0.25 0.16 0.15 0.03
427 187.5 56.25 6 ! .875 1.2 !.i25 0.225
1375 225 375 125 Il0 3000 200 250
" For 500 lag I - ' of boron. b The amount contained in I0 ml of final sample solution prepared from a 30 mg sample. Table 7 Determination of boron in cobalt-based alloys by use of nickel-zirconium as the matrix modifier and a zirconiumcoated graphite tube Sample number
Certified value (%)
Obtained value (%)
Difference (%)
Permissible difference" (%)
Co-68 Co-35
0.027 0.048
0.0269b 0.0484~
0.0001 0.0004
0.005 0.005
Co-70 Co-21
0.049 0.050
0.0490 b 0.0507 b
0.0000 0.0007
0.005 0.005
"Based on the Chinese national standard for the determination of boron in steels and alloys, GB223.6-81, the permissible difference is 0.005% for the content of 0.0260.050% of boron. b Mean of four determinations. c Mean of ten determinations.
3.5. Determination of boron in cobalt-based alloys The proposed method was applied to the determination of boron in four cobalt-based alloys. The results are shown in Table 7. It is seen that the differences between the values obtained and the certified values are all smaller than the permissible differences. The relative standard deviation was 3.3% for 0.0480% boron (n = 10; the values were 0.0475,
0.0472, 0.0463, 0.0493, 0.0507, 0.0487, 0.0508, 0.0465, 0.0480 and 0.0489; with a mean of
0.0480). A characteristic mass was 500 pg.
References [I] G. Mezger, E. Grallath, U. Stix and G. Tolg, Fresenius' Z. Anal. Chem., 317 (1984) 765. [2] V.K. Pin, Anal. Chim. Acta, 159 (1984) 387. [3] J.N. Walsh, Analyst, ! 10 (1985) 959. [4] K. Fujimoto, T. Okano, Y. Matumura and S. Harima, Bunseki Kagaku, 35 (1986) 651. [5] Y. Takahashi, Bunseki Kagaku, 36 (1987) 693. [6] G. Bauer, I. Rehana, W. Wegscheider and H.M. Ortnet, Spectroehim. Acta, part B, 43 (1988) 971. [7] G. Zeibig, Mikroehim. Acta, 1989111, 389. [8] I. Hlavacek, Mikroehim. Acta, 1989111, 309. [9] Hu Bin, Fresenius' Z. Anal. Chem., 340 (1991) 435. [10] J. Ciba and B. Smolec, Fresenius' Z. Anal. Chem., 348 (1994) 215. [!i] T. Miyatani, H. Suzuki and O. Yosimoto, Bunseki Kagaku, 43 (1994) 475. [12] Operational Manual, Perkin-Eimer model HGA-500. [13] Yongqing Jiang, Jinyu Yao and Benli Huang, Fenxi Huaxue, 17 (1989) 456. [14] Yongqing Jiang, Jinyu Yao and Benli Huang, Fenxi Shiyanshi, 7 (1988) 21. [15] M. Luguera, Y. Madrid and C. Camera, J. Anal. At. Spectrom., 6 (1991) 669. [16] R.C. Weast (Ed.), CRC Handbook of Chemistry and Physics, 70th edn., CRC Press inc., Boca Raton, FL, 1990, FI97-198.