Precise Quantification of Tungsten in Tool Steel Samples by Simultaneous Measurement of Vanadium Internal Standard in Continuum-light-source Flame Atomic Absorption Spectrometry

Abstract

This article describes a procedure using flame atomic absorption spectrometry for specimen preparation to quantify tungsten in alloyed tool steel and high-speed steel. The sample solution was prepared using a microwave-assisted digestion with a mixed acid of hydrofluoric, nitric, and phosphoric acids. Vanadium was added as an internal standard element to the sample solution to measure tungsten and vanadium absorption lines simultaneously using a spectrometer equipped with a multi-wavelength system, which comprised a xenon lamp as a continuum light source and an echelle grating. The internal standard method, with its tungsten-to-vanadium absorbance ratio, contributed to accurate and precise quantification of high-content tungsten in tool steel samples.

1. Introduction

High-speed tool steel, a kind of high-alloyed steel, contains large amounts of tungsten (e.g., up to 19 mass%¹⁾). The alloyed element of tungsten instills excellent abrasion and heating resistance characteristics to the steel, enabling it to withstand severe working conditions. The content of tungsten must be strictly controlled: it is much more expensive than iron. X-ray fluorescence spectrometry and spark discharge optical emission spectrometry are generally used for on-site analyses, but more reliable quantification is sometimes necessary for precise and accurate analytical results. Conventional gravimetric analysis, which requires a skillful analyst, is tedious and time-consuming. Although inductively coupled plasma atomic emission spectrometry (ICP-AES) is frequently used,²⁾ the ICP-AES certainty is not necessarily sufficient for quantifying a large content of tungsten. Therefore, an alternative analytical procedure with better precision must be developed. As one example described in previous work, absorption-free X-ray fluorescence quantification of high-content tungsten in high-speed steel has been suggested.³⁾

Atomic absorption spectrometry (AAS) has been used to analyze various metallic elements.^4,5,6) Tungsten has often been used for various modifications of AAS, such as tungsten metal furnace,⁷⁾ tungsten coil,⁸⁾ and tungsten-coated graphite furnace.⁹⁾ However, quantifying tungsten content in a material is difficult because of its lower atomization ability. This difficulty of the tungsten AAS quantification was largely overcome by internal standard method using a continuum-light-source spectrometer.^10,11) The analytical precision would be improved with the internal standard method using analyte-to-standard absorbance ratio. However, selecting internal standard element was restricted because it must provide an effective absorption peak in the vicinity of an analyte analytical absorption within 0.1 nm–0.15 nm wavelength. For the present work, vanadium internal standard method was adopted in the tungsten quantification using simultaneous measurement of tungsten I 255.135 nm and vanadium I 255.265 nm as an internal standard. Several applications of steel analysis on the basis of similar internal standard methods have been suggested: vanadium quantification with nickel added as the internal standard element (analyte vanadium I 318.398 nm and standard nickel I 318.437 nm);¹²⁾ nickel quantification with phosphoric acid added for the iron internal standard method, corrected with a molecular absorption line of phosphorus oxide (analyte nickel I 232.003 nm and standard iron I 232.036 nm corrected by phosphorus oxide 232.06 nm);¹³⁾ and a multi-elemental application for vanadium, chromium, nickel, and copper with cobalt, nickel, and phosphoric acid added as internal standard agents (e.g., analyte chromium I 359.349 nm and standard cobalt I 359.486 nm; analyte nickel I 232.003 nm and standard cobalt I 232.091 nm; and analyte copper I 324.754 nm and standard nickel I 324.846 nm).¹⁴⁾ These earlier works indicated that analytical precision and accuracy were improved using the internal standard method in flame AAS.

An appropriate AAS procedure for tungsten quantification was examined for this study. A sample of alloyed tool steel or high-speed steel was digested using a microwave oven. This microwave-assisted digestion was conducted with an acidic mixture of hydrofluoric acid, nitric acid, and phosphoric acid with a 1:1:1 volume ratio.¹⁵⁾ A specimen solution was prepared for the flame AAS by adding vanadium as an internal standard element. The most sensitive atomic absorption of tungsten I 255.135 nm was used for the present quantification. Subsequently, another atomic absorption of vanadium I 255.265 nm was measured simultaneously. The tungsten-to-vanadium absorbance ratio was used for the quantification calculations, contributing to more reliable and more precise analysis of tungsten in tool steel samples of several kinds.

2. Experimental

2.1. Apparatus

Flame atomic absorption measurement was conducted using a spectrometer (contrAA 700; Analytik Jena AG, Jena, Germany). The analytical conditions are presented in Table 1.

Table 1. Analytical condition of the flame atomic absorption spectrometer.

Analytical line	Tungsten I 255.135 nm
Flame type	Acetylene–nitrous oxide
Flow rate of acetylene	230 dm3 min−1
Observation height	5 mm
Accumulation time	2.0 s
Repeated measurement	5 times
Internal standard line	Vanadium I 255.265 nm

Tool steel samples were digested using a microwave oven (TOPwave; Analytik Jena AG, Jena, Germany). The heating operation of the oven, which increased the temperature to 433 K in 15 min and maintained the temperature for 30 min, had been optimized in earlier work.¹⁵⁾

2.2. Reagents and Reference Materials

The following acid reagents were used for steel digestion: 70 mass% of nitric acid and 85 mass% of phosphoric acid (for electronics industry; Fujifilm Wako Pure Chemical Corp., Osaka, Japan), and 50 mass% of hydrofluoric acid (electronic grade; Morita Chemical Inds. Co., Ltd., Osaka, Japan). A tungsten standard solution was obtained by dissolution of sodium tungstate dihydrate (>99.0 mass%; Fujifilm Wako Pure Chemical Corp., Osaka, Japan) with distilled water. An iron solution was prepared with pure iron (99.99 mass%, MAIRON SHP; Toho Zinc Co., Ltd., Tokyo, Japan) and nitric acid. A vanadium solution was prepared using pure vanadium (99.9 mass%; Furuuchi Chemical Corp., Tokyo, Japan) and nitric acid. An aluminum solution was obtained by dissolution of aluminum (III) chloride hexahydrate (guaranteed reagent; Fujifilm Wako Pure Chemical Corp., Osaka, Japan) using distilled water.

The following certified reference materials (CRMs) were used for validation analysis of the suggested AAS: alloyed tool steel of JSS 602-10, and high-speed steels of JSS 609-11, JSS 610-11, and JSS 606-9 (Japanese Iron and Steel Federation, Tokyo, Japan).

2.3. Sample Preparation

A steel sample (0.050 g) was placed in a 100 cm³ of polytetrafluoroethylene digestion vessel. An acid mixture (5.0 cm³) of hydrofluoric acid, nitric acid, and phosphoric acid with 1:1:1 volume ratio was added to the vessel. Digestion was conducted using the microwave oven. After the vessel had cooled, the resulting solution was placed in a 50 cm³ of polypropylene volumetric flask. The following agents were added to the flask: 60 mg of aluminum as a matrix modifier and 80 mg of vanadium as an internal standard element. The final volume of a specimen solution was set to 50 cm³ with distilled water.

3. Results and Discussion

3.1. Effects of Additional Agents

The prepared specimen solutions were measured using flame AAS to obtain atomic absorption spectra as presented in Fig. 1. Appropriate specimen solutions were prepared by adding aluminum for the improved atomization of vanadium.^12,14) Furthermore, tungsten atomization might be influenced by co-existing phosphoric acid, which was added as a digestion agent to the specimen solutions.

Fig. 1.

Atomic absorbance spectra (254.99 nm–255.28 nm) of 50 cm³ specimen solutions prepared with 0.050 g of high-speed steel samples of JSS 602-10 (broken line; certified value of tungsten, 3.62 mass%) and JSS 606-9 (solid line; certified value of tungsten, 17.17 mass%) containing 1.6 mg cm⁻³ of vanadium as the internal standard. ●, vanadium I 254.997 nm; ○, tungsten I 225.038 nm; ▲, tungsten I 255.135 nm; and △, vanadium I 255.265 nm.

The tungsten absorbance was enhanced about 1.2 times as significant with 95% confidence level by adding aluminum in a specimen solution as presented in Fig. 2. The 50 cm³ specimen solution containing 60 mg of aluminum was used for the present AAS. However, the added vanadium only slightly influenced the absorbance of tungsten, as depicted in Fig. 3. Solutions containing 80 mg of vanadium were commonly prepared.

Fig. 2.

Variation in the absorbance of tungsten I 255.135 nm (●) and vanadium I 255.265 nm (○) as the internal standard line with the amount of aluminum in the 50 cm³ synthetic solutions containing 0.10 mg cm⁻³ of tungsten, 0.90 mg cm⁻³ of iron as matrix constituent, and 1.6 mg cm⁻³ of vanadium.

Fig. 3.

Variation in the absorbance of tungsten I 255.135 nm (●) and vanadium I 255.265 nm (○) as the internal standard line with the amount of vanadium in the 50 cm³ synthetic solutions containing 0.10 mg cm⁻³ of tungsten and 0.90 mg cm⁻³ of iron as matrix constituent.

The specimen solution contained about 0.5 mol dm⁻³ phosphoric acid of the digesting acid mixture. The effects of phosphoric acid were tested as portrayed in Fig. 4. A large amount of phosphoric acid (3.2 mol dm⁻³), which has a high viscous property, reduced the absorbance of tungsten because of a marked decrease of the solution introduced to the flame. However, the absorbance found through this analysis would be little affected by co-existing phosphoric acid.

Fig. 4.

Variation in the absorbance of tungsten I 255.135 nm (●) and vanadium I 255.265 nm (○) as the internal standard line with phosphoric acid concentration in the synthetic solutions containing 0.10 mg cm⁻³ of tungsten, 0.90 mg cm⁻³ of iron as matrix constituent, and 1.6 mg cm⁻³ of vanadium.

3.2. Flame Condition

An acetylene–nitrous oxide flame was operated by optimizing the conditions of (1) flow rate of acetylene and (2) observation height. An appropriate acetylene flow rate was tested as presented in Fig. 5(a). The flow rate of 230 dm³ min⁻¹ was preferred to the present AAS. The absorbance tended to decrease with observation height (Fig. 5(b)). Then, 5-mm of observation height was used.

Fig. 5.

Variation in the absorbance of tungsten I 255.135 nm (●) and vanadium I 255.265 nm (○) as the internal standard line with the acetylene flow rate (a) and observation height (b) for the measurements of the synthetic solution containing 0.10 mg cm⁻³ of tungsten, 0.90 mg cm⁻³ of iron as matrix constituent, and 1.6 mg cm⁻³ of vanadium.

3.3. Internal Standard Method

Analytical precision was tested by continuous measurement of a synthetic solution, which was an imitated solution of the steel sample comprising 90 mass% of iron and 10 mass% of tungsten. Variations of absorbance were confirmed by calculating the normalized absorbance, as obtained from {(each absorbance) – (mean value)}/(standard deviation),^13,14) as depicted in Fig. 6. The variations well agreed with one another. The relative standard deviations (RSDs) were 5.5% for absorbance of tungsten I 225.135 nm, 8.4% for that of vanadium I 254.997 nm, 6.2% for that of vanadium I 255.265 nm, 3.2% for the absorbance ratio of tungsten I 225.135 nm and vanadium I 254.997 nm, and 1.2% for the ratio of tungsten I 225.135 nm and vanadium I 255.265 nm. The RSDs were improved considerably in terms of their absorbance ratios with a 95% confidence level. Vanadium I 225.265 nm was used as internal standard line for this AAS quantification.

Fig. 6.

Variation in the normalized absorbance of tungsten I 255.135 nm (●), vanadium I 254.997 nm (○), and vanadium I 255.265 nm (△) as the internal standard line with time in absorption measurements of the synthetic solution containing 0.10 mg cm⁻³ of tungsten, 0.90 mg cm⁻³ of iron as matrix constituent, and 1.6 mg cm⁻³ of vanadium.

3.4. Calibration and Validation

Calibration curves were obtained by measuring calibration standard solutions, which contained 0, 18.3, 36.6, 73.3, 110, 147, 183, 275, and 366 μg cm⁻³ of tungsten together with about 900 μg cm⁻³ of iron as the matrix constituent. Analytical values were obtained using the following calibrations: (1) common linear calibration with the absorbance of tungsten (Fig. 7(a)), (2) linear calibration using a tungsten-to-vanadium absorbance ratio (broken line as presented in Fig. 7(b); A = 4.87 × 10⁻² + (3.74 × 10⁻³)C, A stands for absorbance ratio and C signifies concentration in μg cm⁻³), (3) nonlinear calibration with a quadratic equation using the absorbance ratio (solid line as presented in Fig. 7(b); A = 7.19 × 10⁻³ + (4.67 × 10⁻³)C – (2.62 × 10⁻⁶)C²), and (4) weighted linear calibration using the absorbance ratio (bold line as presented in Fig. 7(b); A = 2.61 × 10⁻² + (3.75 × 10⁻³)C). The correlation coefficients were improved slightly by using the internal standard method with tungsten-to-vanadium absorbance ratio: 0.995 for the calibration (1) and 0.997 for the calibration (2). Analytical results based on calibration (4) agreed well with the certified values (Table 1). The calibration performed for (2) yielded a large excess of the intercept of the calibration curve as presented in Fig. 7(b). Consequently, the following appropriate weights were found for calibration (4): 10 for the plots of 0, 18.3, 275, and 366 μg cm⁻³, and 1 for the other plots. This was because improved calibration (4) required smaller intercept than calibration (2) and similar slope.

Fig. 7.

Calibration curves of tungsten without internal standard method (a) using linear calibration and with vanadium internal standard method (b) using linear calibration (broken line), nonlinear calibration (solid line), and weighted linear calibration (bold line). (Online version in color.)

Table 1. Analytical results (in mass%) with standard deviation (n = 3) of tungsten in certified reference materials of alloyed tool steel and high-speed steel.

Sample name	Steel type	Without internal standard method	With internal standard method			Certified value
		Linear calibration	Linear calibration	Nonlinear calibration	Weighted linear calibration
JSS 602-10	Alloyed tool steel	3.02 ± 0.35	3.08 ± 0.11	3.28 ± 0.08	3.66 ± 0.11	3.62
JSS 609-11	High-speed steel	5.64 ± 0.14	5.50 ± 0.04	5.09 ± 0.02	6.08 ± 0.04	6.15
JSS 610-11	High-speed steel	8.87 ± 0.37	8.52 ± 0.18	7.35 ± 0.12	9.09 ± 0.18	9.22
JSS 606-9	High-speed steel	16.94 ± 0.42	16.63 ± 0.12	13.39 ± 0.08	17.17 ± 0.11	17.17

4. Conclusion

Reliable tungsten quantification was accomplished using a continuum-light-source AAS along with the internal standard method. Specimen solutions containing added vanadium as an internal standard element were prepared with tool steel and high-speed steel samples. The tungsten-to-vanadium absorbance ratio was applied to quantify high-content of tungsten in several CRMs, yielding good results. Reliable analytical results are obtainable when the line of regression for AAS calibration is selected appropriately.

References

• 1)  JIS G 4403: 2015, High speed tool steels.
• 2)  JIS G 1258-8: 2017, Iron and steel—ICP atomic emission spectrometric method—Part 8: Determination of tungsten content—Decomposition with phosphoric and sulfuric acids.
• 3)   K.  Nakayama and  K.  Wagatsuma: X-Ray Spectrom., 49 (2020), 332. https://doi.org/10.1002/xrs.3134
• 4)   T.  Taniai,  A.  Sakuragawa and  A.  Uzawa: ISIJ Int., 44 (2004), 1852. https://doi.org/10.2355/isijinternational.44.1852
• 5)   K.  Oguma,  T.  Izumi,  H.  Sakurai and  J.  Noro: Tetsu-to-Hagané, 97 (2011), 48 (in Japanese). https://doi.org/10.2355/tetsutohagane.97.48
• 6)   T.  Sipola,  T.  Alatarvas,  T.  Fabritius and  P.  Perämäki: ISIJ Int., 56 (2016), 1445. https://doi.org/10.2355/isijinternational.ISIJINT-2016-071
• 7)   A.  Uzawa,  H.  Minamisawa and  T.  Okutani: Anal. Sci., 16 (2000), 1085. https://doi.org/10.2116/analsci.16.1085
• 8)   G. L.  Donati,  K. E.  Pharr,  C. P.  Calloway, Jr.,  J. A.  Nóbrega and  B. T.  Jones: Talanta, 76 (2008), 1252. https://doi.org/10.1016/j.talanta.2008.05.003
• 9)   Y.  Nakamoto,  T.  Ishimaru,  N.  Endo and  K.  Matsusaki: Anal. Sci., 20 (2004), 739. https://doi.org/10.2116/analsci.20.739
• 10)   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
• 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: ISIJ Int., 60 (2020), 812. https//doi.org/10.2355/isijinternational.ISIJINT-2019-601
• 13)   K.  Nakayama and  K.  Wagatsuma: Bunseki Kagaku, 69 (2020), 219 (in Japanese). https://doi.org/10.2116/bunsekikagaku.69.219
• 14)   K.  Nakayama and  K.  Wagatsuma: ISIJ Int., 61 (2021), 2122. https//doi.org/10.2355/isijinternational.ISIJINT-2020-732
• 15)   K.  Nakayama and  K.  Wagatsuma: ISIJ Int., 60 (2020), 193. https//doi.org/10.2355/isijinternational.ISIJINT-2019-393