Multicomponent Internal Standard Methods for Determination of Vanadium, Chromium, Nickel, and Copper in Tool Steel Samples in Continuum-light-source Flame Atomic Absorption Spectrometry

Abstract

This paper suggests an improved method for the sample preparation to quantify vanadium, chromium, nickel and copper in tool steel and high-speed steel by using flame atomic absorption spectrometry. The suggested method prepared two specimen solutions containing different internal standard elements: one included nickel for the quantifications of vanadium or copper and cobalt for the quantification of chromium and another included cobalt and phosphoric acid for the quantification of nickel, after the sample was digested with an acid mixture of hydrofluoric acid, nitric acid and phosphoric acid by using a microwave oven. A multi-wavelength high-resolution spectrometer system was employed to be simultaneously measured absorption lines for several pairs of an analyte and an internal standard element, such as vanadium and nickel, chromium and cobalt, nickel and iron, cobalt and phosphorus oxide, and copper and nickel. The internal standards contributed to more accurate and precise quantification of vanadium, chromium, nickel, and copper in tool steel and high-speed steel samples.

1. Introduction

Tool steel is classified into a kind of high-alloyed steel containing relatively large amounts of alloyed elements, such as chromium, nickel, cobalt, molybdenum, and tungsten, which contribute to the excellent properties of this alloy type. Because these alloyed elements are much more expensive than iron, their contents have to be strictly controlled in the production process. While spark discharge optical emission spectrometry is now employed for the on-site analysis, more accurate determination is sometimes needed to check the precision and accuracy of the analytical values. Such a referee analysis has generally relied on inductively coupled plasma atomic emission spectrometry (ICP-AES); however, the precision of ICP-AES is not necessarily sufficient for determining the allowed contents of the alloyed elements along with the product specifications completely. Therefore, an alternative analytical method, in which better precision can be obtained rather than the detection sensitivity, should be developed.

Atomic absorption spectrometry (AAS) has also been used for steel and iron analysis, such as alloyed element analysis of a bulk steel sample digested with an appropriate acid mixture,¹⁾ impurity analysis of pure iron treated by selective precipitation,²⁾ inclusion analysis of stainless steel prepared with electrolytic extraction,³⁾ and trace analysis of pure iron based on a direct atomization technique.⁴⁾ An internal standard method is effectively applied to accurate and precise AAS measurements; however, sequential measurements^5,6,7) of an analyte and an internal standard element have been insufficient to improve the precision further. On the other hand, a continuum-light-source system can be applied for the precise AAS measurement based on the internal standard method.^8,9,10) This is because this system combined with an echelle spectrometer can obtain an atomic absorption spectrum in a certain wavelength range,^9,11,12,13) and then, the absorption line of an internal standard element can be simultaneously measured with that of an analyte. For example, iron as the internal standard (the absorption line of 232.036 nm) can be used to quantify nickel (that of 232.003 nm) in steel samples.^11,12) Another application by using nickel as the internal standard (the absorption line of 318.437 nm) can be used to quantify vanadium (that of 318.398 nm) in a high-speed tool steel sample.¹³⁾ These previous studies reported that the analytical precision was improved by using the internal standard method in AAS.

It is a critical issue whether the sample solution can be prepared appropriately for the multi-element AAS, because tool steels contain several alloyed elements having different chemical properties. For instance, an appropriate sample solution for above-mentioned steel samples containing tungsten, which is readily precipitated in the prepared sample solution, was difficult to be prepared. Our previous paper¹⁴⁾ has described a microwave digestion for various kinds of steel samples including tool steel to prepare a specimen solution in ICP-AES. The microwave digestion has been applied to decompose metal carbides and oxides in a steel sample.¹⁵⁾ A tertiary acid mixture of hydrofluoric acid, nitric acid, and phosphoric acid with 1:1:1 volume ratio¹⁴⁾ was suitable to digest a tool steel sample without a residue. As referring to these previous reports on the sample preparation for tool steel, this paper suggests an improved method for the decomposition and dissolution of tool steel, in order that vanadium, chromium, nickel and copper can be precisely determined with an internal standard method in the multi-element AAS.

2. Experimental

2.1. Apparatus

A spectrometer of contrAA 700^® (Analytik Jena AG, Jena, Germany) was used in the flame AAS for the determination of vanadium, chromium, nickel, and copper. The analytical conditions were listed in Table 1. The spectrometer comprises a xenon lamp as the continuum light source and an echelle grating system, which has an ability to measure an absorption spectrum in a wavelength range of 0.2 nm to 0.3 nm. A microwave oven of TOPwave^® (Analytik Jena AG, Jena, Germany) was used for the digestion of steel samples. The oven was operated by increasing temperature to 433 K in 15 min and keeping the temperature for 30 min.

Table 1. Analytical conditions for flame atomic absorption spectrometric analysis.

Analyte	Vanadium	Chromium		Nickel	Copper
Wavelength/nm	318.398	359.349		232.003	324.754
Flame type	C2H2–N2O	C2H2-air	C2H2–N2O	C2H2-air	C2H2-air
Flow rate of C2H2/dm3 min−1	230	80	210	50	50
Observation height/mm	4	10	5	5	6
Internal standard	Ni 318.437 nm	Co 359.486 nm		Fe 232.036 nm	Ni 324.846 nm
				PO* 232.06 nm
				Co 232.091 nm

*   Molecular line of phosphorus oxide.

2.2. Reagents and Standards

Acid reagents for the digestion of tool steel samples were 50-mass% hydrofluoric acid, 70-mass% nitric acid, and 85-mass% phosphoric acid. An acid mixture of these acids with 1:1:1 volume ratio was prepared. This acid mixture was suited for the steel samples containing tungsten such as tool steel and high-speed steel, which has been confirmed in the previous work.¹⁴⁾

A nickel sulfate (hexahydrate) solution (50 mg cm⁻³) and a cobalt chloride solution (50 mg cm⁻³) were prepared to be added as the internal standards. An aluminum chloride (hexahydrate) solution¹³⁾ (20 mg cm⁻³) was prepared to modify the atomization of vanadium and chromium (Fig. 1).

Fig. 1.

Flow chart of procedure of sample preparation.

The calibration standard solutions were prepared by using pure metals of vanadium, chromium, nickel, copper, and iron as matrix constituent. Two sets of six calibration solutions contained different amounts of iron (the matrix element) as shown in Table 2. The specimen solution would have different concentrations of iron when tool and high-speed steel samples were digested. Therefore, different amounts of iron were added to calibration solutions to emulate the concentration of iron in the specimen solution. Then, the absorbance of iron was corrected by the actual iron concentration and was applied for the internal standard method in the AAS quantification of nickel.¹²⁾

Table 2. Concentrations (in μg cm⁻³) of analytes and iron as matrix constituent in prepared calibration standard solutions A and B.

	Calibration standard solutions A				Calibration standard solutions B
Constituents for internal standards	5.0 mg cm−3 of nickel 2.0 mg cm−3 of cobalt				10.0 mg cm−3 of cobalt 1.5 mol dm−3 of phosphoric acid
	Vanadium	Chromium	Copper	Iron	Chromium	Nickel	Iron
Solution 1	0	0	0	360	0	0	360
2	1.86	2.41	0.415	300	2.41	0.434	300
3	4.64	6.03	0.934	400	6.03	0.759	400
4	8.36	10.9	1.56	340	10.9	1.19	340
5	13.0	16.9	2.28	380	16.9	1.73	380
6	18.6	24.1	3.11	320	24.1	2.38	320

2.3. Reference Materials

Validation analysis was conducted with several certified reference materials (CRMs). The following CRMs of tool steel and high-speed steel were used: JSS 602-11 of tool steel and JSS 604-9 of high-speed steel (the Japanese Iron and Steel Federation, Tokyo, Japan), and ECRM 290-1 of high-speed steel (Bundesanstalt für Materialforschung und-prüfung, Berlin, Germany).

3. Results and Discussion

3.1. Internal Standards

A sensitive atomic absorption line of V I 318.398 nm was measured for the determination of vanadium in tool steels and an atomic absorption line of Ni I 318.437 nm was also appeared in the spectrum by the continuum-light-source AAS, as shown in Fig. 2(a), which was employed as the internal standard. A specimen solution A (Fig. 1) containing an appropriate amount of nickel was prepared for the internal standard method.

Fig. 2.

Atomic absorption spectra of specimen solutions containing tool steel samples. (a) V 318.398 nm (●) and Ni 318.437 nm (○) in specimen solution A of JSS 602-11; (b) Cr 359.349 nm (■) and Co 359.486 nm (□) in specimen solution A of JSS 604-9; (c) phosphorous oxide (PO) 231.89 nm (▲), Co 231.916 nm (△), Ni 232.003 nm (◆), Fe 232.036 nm (◇), PO 232.06 nm (×), and Co 232.091 nm (+) in specimen solution B of JSS 602-11; and (d) Cu 324.754 nm (▼) and Ni 324.846 nm (▽) in specimen solution A of JSS 602-11.

An atomic absorption line of Cr I 359.349 nm was measured, together with an atomic absorption line of Co I 359.486 nm in the neighborhood of the Cr I analytical line (Fig. 2(b)). A specimen solution A containing 2.0 mg cm⁻³ of cobalt as the internal standard was prepared.

A sensitive atomic absorption line of Ni I 232.003 nm was measured as the analytical line. Atomic absorption lines of Fe I 232.036 nm and Co I 232.091 nm, and a molecular absorption of phosphorus oxide (PO) 232.06 nm were observed simultaneously (Fig. 2(c)). These absorption lines were used for several internal standard methods: (1) nickel-to-iron absorbance ratio, (2) nickel-to-(iron corrected by PO) absorbance ratio, (3) nickel-to-PO absorbance ratio, and (4) nickel-to-cobalt absorbance ratio. The correction of iron¹²⁾ was conducted through the follow procedure. A calibration curve of the iron concentration (from 300 μg cm⁻³ to 400 μg cm⁻³) in a specimen solution was obtained by using the iron-to-PO absorbance ratio. A corrected absorbance of iron was then estimated, when the iron concentration was 400 μg cm⁻³. The nickel-to-(the resulting corrected iron) absorbance ratio was finally calculated. A specimen solution B (Fig. 1), which contained 10.0 mg cm⁻³ of cobalt and 1.5 mol dm⁻³ of phosphorus acid, was prepared.

An atomic absorption line of Cu I 324.754 nm was measured, together with the neighboring atomic absorption line of Ni I 324.846 nm (Fig. 2(d)). The specimen solution containing nickel was able to be used for internal standard methods of vanadium-to-nickel absorbance ratio and copper-to-nickel absorbance ratio. Then, a specimen solution A containing 5.0 mg cm⁻³ of nickel was prepared for internal standard methods for vanadium and copper.

3.2. Effect of Phosphoric Acid

The absorbance of each analyte was reduced by the addition of phosphoric acid as shown in Fig. 3. On AAS measurements by using the acetylene-air flame, even a small amount of phosphoric acid (about 0.10 mol dm⁻³) decreased the absorbances of chromium and nickel. This was because heat-stable compounds of these elements might be formed in the flame. On the other hand, a large amount of phosphoric acid generally reduced the absorbance of all the analytical lines because of a significant decrease of solution introduced to the flame.

Fig. 3.

Variations in the absorbance-to-concentration ratio of vanadium for acetylene-nitrous oxide flame (●), chromium for acetylene-nitrous oxide flame (▲), chromium for acetylene-air flame (△), nickel for acetylene-air flame (□), and copper for acetylene-air flame (◇) with concentration of phosphoric acid.

3.3. Improvement in Precision by Internal Standard Method

The effect of internal standards was checked by calculating a normalized absorbance, which was obtained from {(each absorbance value) – (mean value of the absorbances)}/(standard deviation of the absorbance).^12,13) Continuous measurements of each analyte were conducted by using the specimen solutions of A and B containing JSS 602-11, which contained about 0.9 μg cm⁻³ of vanadium, about 1.5 μg cm⁻³ of chromium, and about 0.6 μg cm⁻³ of nickel and copper. On a continuous measurement of copper as shown in Fig. 4, the absorbance of Cu I 324.754 nm was similarly varied along with the absorbance of Ni I 324.846 nm as the internal standard line. The relative standard deviations (RSD) were 5.8% for the raw absorbance of copper, 7.4% for that of nickel, and 2.8% for the copper-to-nickel absorbance ratio. The RSD was significantly improved by using the absorbance ratio with 95% confidence level. Similarly the effects of internal standards for vanadium, chromium, and nickel were also confirmed: (1) 8.1% for the vanadium absorbance, 7.2% for the nickel absorbance, and 4.2% for the vanadium-to-nickel absorbance ratio; (2) 4.5% for the chromium, 4.9% for the cobalt, and 1.6% for the chromium-to-cobalt ratio; (3) 6.6% for the nickel, 11.4% for the iron, 5.7% for phosphorous oxide (PO), 12.9% for the cobalt, 13.5% for the nickel-to-iron ratio, 4.2% for the nickel-to-PO ratio, and 13.2% for the nickel-to-cobalt ratio. The RSD of the absorbance ratios was significantly improved except for nickel. As shown in Fig. 2(c), the background level for the Ni I 232.003 nm was not corrected completely due to some overlapping with the neighboring absorption lines, thus making the internal standard correction insufficiently. An appropriate evaluation of the background requires a certain area of no absorption line. In spite of the large RSD of nickel for a short measuring time, an improvement in accuracy was achieved on the quantification of nickel by using the absorbance ratios as shown in section 3.5. and Table 3. This result indicated that long-period fluctuation in the absorbance for the quantification of nickel could be compensated by the internal standard methods.

Fig. 4.

Variations in the normalized absorbance of copper and nickel absorption lines with the number of continuous measurement for 12 s. Cu 324.754 nm (●) and Ni 324.846 nm (○).

Table 3. Analytical results (in mass%) with standard deviation (n = 3) of alloyed elements in certified reference materials of tool steel.

						JSS602-11, tool steel		JSS604-9, tool steel		ECRM290-1, high-speed tool steel
	Flame type	Measured solution*	Internal standard method	Detection limit	Effective quantification range	Analytical value	Certified value	Analytical value	Certified value	Analytical value	Certified value
Vanadium	C2H2–N2O	A	Without	0.011	0.56 to 4.6	0.199 ± 0.007	0.218	0.464 ± 0.007	0.512	1.86 ± 0.04	1.91
		A	With V/Ni ratio	0.008	0.43 to 4.6	0.214 ± 0.004		0.508 ± 0.012		1.96 ± 0.07
Chromium	C2H2-air	A	Without	0.018	1.5 to 6.0	0.436 ± 0.019	0.366	4.39 ± 0.29	4.90	4.02 ± 0.27	4.18
		A	With Cr/Co ratio	0.015 (0.012**)	1.6 to 6.0 (0.41 to 2.7**)	0.346 ± 0.004**		4.94 ± 0.08		4.08 ± 0.09
	C2H2–N2O	A	Without	0.008	2.2 to 6.0	0.435 ± 0.043		3.59 ± 0.06		2.82 ± 0.07
		A	With Cr/Co ratio	0.008 (0.005**)	1.5 to 6.0 (0.17 to 1.5**)	0.359 ± 0.010**		4.98 ± 0.14		4.21 ± 0.04
	C2H2–N2O	B	Without	0.005	1.5 to 6.0	0.515 ± 0.020		4.88 ± 0.25		4.38 ± 0.09
		B	With Cr/Co ratio	0.007 (0.005**)	2.3 to 6.0 (0.22 to 1.5**)	0.345 ± 0.008**		4.58 ± 0.14		4.15 ± 0.05
Nickel	C2H2-air	B	Without	0.017	0.044 to 0.60	0.146 ± 0.017	0.1547	0.182 ± 0.018	0.177	0.378 ± 0.073	0.329
		B	With Ni/Fe ratio	0.015	0.12 to 0.60	0.143 ± 0.008		0.168 ± 0.011		0.370 ± 0.022
		B	With Ni/(corrected Fe by PO***) ratio	0.016	0.025 to 0.60	0.136 ± 0.006		0.175 ± 0.013		0.326 ± 0.020
		B	With Ni/PO ratio	0.016	0.017 to 0.60	0.136 ± 0.007		0.174 ± 0.013		0.327 ± 0.015
		B	With Ni/Co ratio	0.014	0.051 to 0.60	0.155 ± 0.010		0.172 ± 0.014		0.316 ± 0.006
Copper	C2H2-air	A	Without	0.006	0.11 to 0.57	0.158 ± 0.005	0.1513	0.061 ± 0.002	0.0630	0.080 ± 0.002	0.081
		A	With Cu/Ni ratio	0.005	0.031 to 0.57	0.150 ± 0.002		0.060 ± 0.002		0.081 ± 0.002

*A and B are specimen solutions shown in Fig. 1; **Alternative calibration curve was used as described in section 3.5.; ***Molecular absorption line of phosphorus oxide.

3.4. Calibration

Two sets of calibration standard solutions were prepared as shown in Table 2. One is for the measurements of V, Cr, and Cu in the specimen solution A, which contained these analytes, different amounts of iron, 5.0 mg cm⁻³ of Ni as an internal standard element for the V and Cu quantifications, and 2.0 mg cm⁻³ of Co as another internal standard element for the Cr quantification. The other is for the measurements of Ni and Cr in the specimen solution B, which contained these analytes, different amounts of iron, 10.0 mg cm⁻³ of Co as an internal standard element for the Ni and Cr quantifications, and 1.5 mol dm⁻³ of phosphoric acid to obtain the molecular absorption of phosphorus oxide as another internal standard constituent for the Ni quantification.

The effective quantification ranges, which were evaluated as concentration values in a steel sample, for the prepared calibration curves were listed in Table 3. Their lower values were estimated from the lower use-limit of the calibration curve, which has been suggested in previous works.^16,17) The following calculated result was used for the present estimation: (3/b) [{Σ(A_i – A_i′)²}/(m – 2)]^1/2, where b is the slope of a calibration curve (in cm³ μg⁻¹), A_i is the measured absorbance (or absorbance ratio) of ith calibration standard solution, A_i′ is the calculated absorbance (or absorbance ratio) of ith calibration standard solution, and m is the number of calibration standard solutions. Their upper values were calculated by the highest concentration value of the used calibration standard solution on each calibration calculation. Common detection limits (3s/b), which were calculated from standard deviation (s) of the blank measurements, were also listed in Table 3.

3.5. Validation

Analytical results of CRMs were listed in Table 3. Analytical values were generally improved by using the internal standard method with several exceptions: (1) extreme lower content values of chromium like in JSS 602-11 in the internal standard method, (2) analytical values of chromium in the chromium-to-cobalt absorbance ratio using the specimen solution B, and (3) analytical values of nickel in the nickel-to-iron absorbance ratio. In the case of (1), the obtained calibration curve, which covered to 24.1 μg cm⁻³ of chromium in a specimen solution, had a low linearity and high intercept leading to the analytical values (about 0.01 mass%) with a poor precision; therefore, for the trace level of chromium, an alternative calibration was needed on the basis of another calibration curve in the content range to 6.0 μg cm⁻³ of chromium. In the case of (2), the large amount of phosphoric acid (1.5 mol dm⁻³) in the specimen solution B would prevented the atomization of chromium (on the other hand, the specimen solution A contained 0.1 mol dm⁻³ of phosphoric acid). In the case of (3), an appropriate calibration curve could not be obtained because the iron concentrations in the calibration standard solutions widely varied from 300 μg cm⁻³ to 400 μg cm⁻³ as shown in Table 2. Correlation coefficients of the nickel calibration curves were 0.9982 for the nickel absorbance, 0.9867 for the nickel-to-iron absorbance ratio, 0.9994 for the nickel-to-(corrected iron by PO) absorbance ratio, 0.9997 for the nickel-to-PO absorbance ratio, and 0.9976 for the nickel-to-cobalt absorbance ratio.

4. Conclusions

Specimen solutions containing several internal standard elements were prepared to quantify alloying metals in tool steel and high-speed steel samples when they were analyzed using a continuum-light-source AAS system. Internal standard methods were successfully applied to quantify the vanadium content by using the vanadium-to-nickel absorbance ratio, the chromium content by using the chromium-to-cobalt absorbance ratio, the nickel content by using the nickel-to-cobalt absorbance ratios, and the copper content by using the copper-to-nickel absorbance ratio in several CRMs of tool steel and high-speed steel, when the internal standard elements were selected and added appropriately.

References

• 1)  JIS G 1257-5: 2013, Iron and steel—Atomic absorption spectrometric method—Part 5: Determination of molybdenum—Flame atomization after decomposition with acids (in Japanese).
• 2)   H.  Matsumiya,  M.  Kuromiya and  M.  Hiraide: ISIJ Int., 53 (2013), 81. https://doi.org/10.2355/isijinternational.53.81
• 3)   T.  Sipola,  T.  Alatarvas,  T.  Fabritius and  P.  Perämäki: ISIJ Int., 56 (2016), 1445. https://doi.org/10.2355/isijinternational.ISIJINT-2016-071
• 4)   S.  Tsushima,  S.  Yamaoka,  J.  Bai,  H.  Minami,  N.  Takahashi,  S.  Inoue and  I.  Atsuya: Bunseki Kagaku, 58 (2009), 771 (in Japanese). https://doi.org/10.2116/bunsekikagaku.58.771
• 5)   M. A.  Bechlin,  F. M.  Fortunato,  E. C.  Ferreira,  J. A. G.  Neto,  J. A.  Nóbrega,  G. L.  Donati and  B. T.  Jones: Anal. Chim. Acta, 831 (2014), 24. https://doi.org/10.1016/j.aca.2014.05.002
• 6)   Z.  Kowalewska,  K.  Brzezińska,  K.  Cikorska,  J.  Pilarczyk and  Ł.  Gościniak: Energy Fuels, 33 (2019), 215. https://doi.org/10.1021/acs.energyfuels.8b03517
• 7)   L. M.  Roveda and  J. L.  Raposo, Jr.: Food Anal. Methods, 12 (2019), 1111. https://doi.org/10.1007/s12161-019-01439-1
• 8)   F. M.  Fortunato,  M. A.  Bechlin,  J. A. G.  Neto,  G. L.  Donati and  B. T.  Jones: Microchem. J., 122 (2015), 63. https://doi.org/10.1016/j.microc.2015.04.009
• 9)   D. V.  de Babos,  M. A.  Bechlin,  A. I.  Barros,  E. C.  Ferreira,  J. A. G.  Neto and  S. R.  de Oliveira: Talanta, 152 (2016), 457. https://doi.org/10.1016/j.aca.2014.05.002
• 10)   L.  Fernández-López,  B.  Gómez-Nieto,  M. J.  Gismera,  M. T.  Sevilla and  J. R.  Procopio: Spectrochim. Acta B, 147 (2018), 21. https://doi.org/10.1016/j.sab.2018.05.016
• 11)   T.  Itagaki,  T.  Ashino,  K.  Takada and  K.  Wagatsuma: Bunseki Kagaku, 64 (2015), 117 (in Japanese). https://doi.org/10.2116/bunsekikagaku.64.117
• 12)   K.  Nakayama and  K.  Wagatsuma: Bunseki Kagaku, 69 (2020), 219 (in Japanese). https://doi.org/10.2116/bunsekikagaku.69.219
• 13)   K.  Nakayama and  K.  Wagatsuma: ISIJ Int., 60 (2020), 812. https://doi.org/10.2355/isijinternational.ISIJINT-2019-601
• 14)   K.  Nakayama and  K.  Wagatsuma: ISIJ Int., 60 (2020), 193. https://doi.org/10.2355/isijinternational.ISIJINT-2019-393
• 15)   K.  Fujimoto,  M.  Shimura and  S.  Satoh: Tetsu-to-Hagané, 88 (2002), 285 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.88.6_285
• 16)   K.  Nakayama and  K.  Wagatsuma: Anal. Sci., 31 (2015), 851. https://doi.org/10.2116/analsci.31.851
• 17)   K.  Nakayama and  K.  Wagatsuma: Bunseki Kagaku, 67 (2018), 673 (in Japanese). https://doi.org/10.2116/bunsekikagaku.67.673