Regular Article
Determination of Alloying and Impurity Elements from Matrix and Inclusions from a Process Sample of a Double Stabilized Stainless Steel
2016 Volume 56 Issue 8 Pages 1445-1451
Details
2016 Volume 56 Issue 8 Pages 1445-1451
The determination of alloying and impurity elements was performed from a stainless steel matrix and inclusions in process samples. An electrolytic extraction method was applied for the separation of inclusions using two different but commonly used electrolytes, 10% HCl and 10% acetylacetone in methanol. The elemental analyses were performed using atomic absorption spectrometry. The elements of interest were aluminum, arsenic, copper, vanadium, titanium and chromium. The aluminum containing inclusions were imaged using a field emission scanning electron microscope. The results for copper and chromium in both electrolytes, vanadium in 10% HCl electrolyte and arsenic in 10% acetylacetone electrolyte were in good agreement with industrial data. Titanium and aluminum were measured from the dissolved steel matrix but titanium was also detected in the inclusions. It was concluded that the analytical results for titanium and aluminum measured using an optical emission spectrometer is affected by the inclusions within the stainless steel.
As stainless steels are getting cleaner, the analysis of its microstructure, inclusions and impurities is getting more attention. The trend is to develop methods that are suitable for the simultaneous study of steel matrix, inclusions and impurity elements, preferably as an online method during the steel manufacturing process. The current industrial analysis technique of steel chemical composition does not offer information concerning inclusions (spark excitation optical emission spectrometry, spark-OES). An optical emission spectrometer – pulse discrimination analysis (OES-PDA) method has been suggested as a possible method for fast online analysis of steel matrix and inclusions.1) Laser induced breakdown spectrometry (LIBS) is another proposed method for online use in the steel industry.2) Online methods for inclusions analysis are still under development and more studies are required. Before the online techniques can be utilized to an online use, it should be studied if inclusions affect the elemental analysis of a solid steel piece.
An electrolytic extraction method is a commonly used technique in analysis of inclusions. The study of extracted inclusions offers information about their origin, modification processes and even information concerning the mechanical properties of the final steel products. The current steel studies are focused on the influence of inclusions on steel properties and chemical analysis of soluble alloying elements.3,4,5) The elemental analysis of inclusions can offer additional information concerning their origin and how much of the alloying and impurity elements they contain. In addition to alloying materials, inclusions can contain impurity elements and refractory materials. As an example, lead as an impurity in austenitic steel can have an effect on the workability properties even in ppm levels.6) Therefore it is essential to study impurities in stainless steel as well as inclusions. In this study, arsenic and copper are considered impurity elements. Arsenic and copper can originate from several sources, such as recycled raw material.7) Copper can be present in steel as sulphide inclusions and sulphur is known to cause hot-shortness in steels.8) Sulphide inclusions can act as a starting point in pitting and crevice corrosion.9) On the other hand, copper also improves corrosion resistance and is therefore an important alloying element for further studies.10) The influence of copper on corrosion behavior in ferritic and ferritic-austenic stainless steel has been studied by Banas et al.11) Aluminum is a common deoxidizing agent and can therefore have an influence on titanium content in steel and in inclusions.12) Titanium is involved in the studies due to its inclusion formation being similar to that of aluminum. The stability of titanium containing inclusions in non-aqueous electrolytes has been studied before and it has been concluded that 10% acetylacetone electrolyte is suitable in extraction of Ti2O3 and TiAl2O5 inclusion particles.13) Vanadium is a carbide forming element as are titanium and niobium. Vanadium carbides have an effect on strength, creep resistance and impact resistance and were therefore chosen to be one of the studied elements in this study.14)
In the previous study, microalloying elements titanium, niobium, nickel, manganese, and chromium were determined from a stainless steel matrix and inclusions using the method of electrolytic extraction and elemental analysis.15) In this study, alloying elements and impurities with even smaller contents were determined from steel matrix and inclusions from a stainless steel process sample. An electrolytic extraction method is used to separate stainless steel matrix and inclusions in two common electrolytes in inclusions studies. A chemical analysis was performed on dissolved steel matrices and dissolved inclusions. The elements of interest were aluminum, arsenic, copper, vanadium, titanium and chromium. The arsenic, vanadium, aluminum, and titanium contents were measured using a graphite furnace atomic absorption spectrometry (GFAAS). Chromium and copper contents were measured using a flame atomic absorption spectrometry (FAAS). A field emission scanning electron microscope (FESEM) was used in the examination of inclusions. The study was executed to determine the amount of alloying elements and impurities in inclusions and steel matrix. The used method offers new, more detailed information concerning stainless steels composition.
All the glassware was properly cleaned in an acid bath, rinsed with ultra pure (Milli-Q) water and the final rinse was carried out using methanol. Ultra pure water was used in the diluting of samples and standards. A methanolic HCl electrolyte was prepared using a 32% HCl acid (EMSURE), tartaric acid (EMSURE) and methanol (EMPARTA) all provided by Merck. The waterless acetylacetone electrolyte was prepared using methanol, acetylacetone (EMSURE) provided by Merck and tetramethylammoniumchloride (purity +98%) provided by Acros Organic.
The matrix modifiers in determination performed using a GFAAS were all provided by Merck except for the reagent that was used in preparation of the tungsten modifier, which was provided by Acros organic. The LaCl3 and NH4Cl reagents that were used in FAAS determinations were provided by Merck.
2.2. InstrumentalAn ultrasonic bath was used to clean the sample from grinding paper traces using an Elmasonic P 30 H ultrasonic cleaning unit (frequency 37 kHz, room temperature). The same ultrasonic unit was used in order to remove adhered inclusions from the stainless steel sample after the electrolytic extraction.
A BioLogic SP-150 potentiostat was used to control the electrolytic extraction of the stainless steel sample. The filtering was assisted with a BUSCHI vacuum pump. The decomposition of inclusions was performed in a SYSTEC autoclave in PTFE containers.
2.2.1. GFAASA PerkinElmer AAnalyst600 graphite furnace atomic absorption spectrometer equipped with an AS-800 autosampler and PerkinElmer THGAGraphite tubes (standard platform B0504033) was used in the determination of aluminum, arsenic, titanium and vanadium. Hollow cathode lamps (HCL) and an electrodeless discharge lamp (EDL) were used as radiation sources and they were provided by PerkinElmer, except for the one used with vanadium, which was provided by Halstead.
The instrumental parameters and the furnace programs used in the determinations are presented in Tables 1 and 2. The internal gas flow was set to 250 mlmin−1 during steps 1–3 and 5. During the atomization step (step 4) the gas flow was 0 mlmin−1.
| V | Al | As | Ti | |
|---|---|---|---|---|
| λ (nm) | 318.4 | 309.3 | 193.7 | 364.3 |
| Lamp | HCL | HCL | EDL | HCL |
| Current (mA) | 15 | 25 | 390 | 42 |
| Slit (nm) | 0.7 | 0.7 | 0.7 | 0.2 |
| Sample Volume (μl) | 10 | 20 | 20 | 20 |
| Matrix modifier | Cr 0.05 μg | Pd 0.5 μg* | Ni 0.4 μg | No/Pd*1 |
| W 54 μg | ||||
| Linear calibration | 0–400 | 0–50 | 0–50 | 0–100 |
| range*2 μgL−1 |
| Furnace programs | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| V | Al | As | Ti | |||||||||
| Step | Temp K (°C) | Ramp (s) | Hold (s) | Temp (°C) | Ramp (s) | Hold (s) | Temp (°C) | Ramp (s) | Hold (s) | Temp (°C) | Ramp (s) | Hold (s) |
| 1 | 383+ (110) | 10 | 30 | 393 (120) | 5 | 30 | 383 (110) | 5 | 30 | 383 (110) | 1 | 30 |
| 2 | 413+ (140) | 10 | 30 | 433 (160) | 5 | 25 | 423 (150) | 10 | 30 | 403 (130) | 15 | 30 |
| 3 | 1573 (1300) | 10 | 20 | 2473 (1200) | 10 | 30 | 1473 (1200) | 10 | 30 | 1773 (1500) | 10 | 20 |
| 4 | 2673 (2400) | 0 | 6 | 2623 (2350) | 0 | 3 | 2273 (2000) | 0 | 4 | 2823 (2550) | 0 | 6 |
| 5 | 2773 (2500) | 1 | 5 | 2723 (2450) | 1 | 3 | 2723 (2450) | 1 | 3 | 2773 (2500) | 1 | 3 |
A Perkin Elmer AAnalyst 400 flame atomic absorption spectrometer equipped with an S10 autosampler and a 10 cm N040-0102 burner head was used in the determination of chromium and copper. The linear calibration graph had a correlation coefficient R2 value at least 0.999 for both elements. The instrumental parameters in determination of chromium and copper are presented in Table 3.
| Cr | Cu | |
|---|---|---|
| λ (nm) | 357.9 | 324.8 |
| Slit width/height (mm) | 2.7/0.8 | 2.7/0.8 |
| Lamp | HCL | HCL |
| Current (mA) | 25 | 25 |
| Gas flows (lmin−1) | ||
| Air | 10.0 | 10.0 |
| Acetylene | 3.3 | 2.5 |
| Linear calibration range mgL−1 | 0–2.0 | 0–4.0 |
A control sample was measured containing 1 mgL−1 chromium in a 1% (V/V) methanol matrix. The control sample resulted in 1.01 mgL−1 chromium concentration which shows that methanol did not affect the results of the samples of soluble chromium with 1% (V/V) methanol content.
The influence of 50% (V/V) methanol in samples in the determination of copper was also examined. A control sample with a known copper content of 500 μgL−1 was measured. The methanol content decreased the absorbance of the control sample by approximately 30%. The instrument was recalibrated using a water-methanol solution in the standards to match the matrix in the samples of soluble copper with methanol content of 50% (V/V).
2.2.3. FESEMA Zeiss Ultra Plus field emission scanning electron microscope (FESEM) coupled with an Energy-Dispersive X-ray Spectroscopy (EDS) was used in the examination of extracted inclusions on a film filter. An EDS analysis was performed on selected inclusions.
2.3. Pretreatment of Steel SamplesProcess samples was chosen, because inclusion studies are usually conducted using laboratory scale samples or pilot scale samples, although some inclusion studies have been conducted with industrial samples.16) It is known that the solidification process considerably affects the inclusions sizes and distribution of inclusions.17) Lollipop samples are used in sample analysis from steel melt and a proper piece can be cut from a lollipop for electrolytic extraction. Sometimes the lollipop samples contain air bubbles or pieces of slag in them so it is not possible to obtain a suitable piece for electrolytic extraction. A slab is therefore considered as a good source for industrial samples. Figure 1 contains an illustration of the position on the slabs from where the samples were obtained.
An illustration of the sample piece of this study obtained from the stainless steel slab and the pretreatment of the steel sample (not in scale).
Two pieces of stainless steel for electrolytic extraction was cut from a corner piece of two different slabs using a Struers Secotom-10 cutting machine with an Al2O3 cutting wheel. The steel samples were ground and polished using a Struers Labopol-6 grinding machine with a SiC grinding paper. The polished and cleaned steel sample was used in electrolytic extraction experiments. The final size of the sample before the electrolytic extraction was approximately 15×10×5 mm. The same sample piece was used in the extraction in two different electrolytes. The chemical composition of the studied samples is presented in Table 4. The chemical compositions of the studied samples were determined by a spark-OES commonly used in steel industry (Table 4).
| Element (w-%) | C | Al | As | Ti | Cu | V | Cr | Fe |
|---|---|---|---|---|---|---|---|---|
| Sample 1 | 0.03 | 0.012 | 0.012 | 0.151 | 0.14 | 0.05 | 18 | 80 |
| Sample 2 | 0.02 | 0.022 | 0.007 | 0.168 | 0.12 | 0.05 | 18 | 80 |
The stainless steel grade was the same for both samples. The largest difference was in the amount of alloyed aluminum.
2.4. Electrolytic Extraction and FiltrationAn electrolytic extraction was performed in two different electrolytes. The electrolytes in this study were 10% HCl (10 V/V% HCl, 1 w/V% tartaric acid, and methanol) and 10% acetylacetone (10 V/V% acetylacetone, 1 w/V% tetramethylammoniumchloride, and methanol).
Approximately 200 ml of electrolyte was used in an extraction. The initial weight of the stainless steel sample 1 was 13.2 g and sample 2 was 6.5 g. The voltage of the working electrode (the sample) was set to 0.150 V vs. standard calomel electrode (SCE). The sample was suspended in the electrolyte in a platinum basket and a platinum ring was used as a counter electrode. The electric current was approximately 22 mA in the 10% acetylacetone electrolyte and 300 mA in the 10% HCl electrolyte. The detailed experimental design is presented in the previous article.15) The sample was weighed before and after both extractions to obtain the dissolved mass of the sample. The dissolved sample masses for sample 1 and sample 2 in the 10% acetylacetone electrolyte were 0.15 g and 0.16 g. The dissolved sample masses in 10% HCl electrolyte were 0.83 g for sample 1 and 0.44 g for sample 2. The total amounts of dissolved elements were obtained by multiplying the relative amount of the element (w-%) by the amount of dissolved masses of the sample. Each extraction was performed once.
After the electrolytic extraction the electrolyte was filtered through a 0.05 μm pore size polycarbonate filter. The steel sample was exposed to ultrasound in methanol in order to collect the all of the extracted particles. The methanol was filtered with the electrolyte. The steel sample was dried and weighed. The filtered electrolyte containing soluble elements was diluted to volume with water in a volumetric flask. The filtered electrolyte is from now on referred to as the sample for soluble elements. A blank sample was prepared by filtering a proper amount of unused electrolyte through a PC-filter.
For the sample preparation from the filter for elemental analysis of inclusions see section 2.4.1. It should be noted that this method is limited to studying inclusions larger than 0.05 μm in diameter, although it is clear that inclusions and precipitates of even smaller size are present in this stainless steel type.18)
An additional extraction was performed in a 10% acetylacetone electrolyte for a closer examination of inclusions. Approximately 0.2 g of sample was dissolved during the extraction. The electrolyte was filtered through a 0.1 μm pore size PC-filter. The filter was dried in a desiccator overnight before sample preparation for FESEM. For the sample preparation from the filter see section 2.4.2.
2.4.1. Sample Preparation from Extracted Inclusions for Elemental AnalysisThe polycarbonate filter was dried with the extracted inclusions in a desiccator overnight. The dried filter was put in a gently closed PTFE container with 5 ml concentrated HNO3 and 2 ml HF. The decomposition program in an autoclave was 30 minutes in 120°C (393.15 K). The containers were cooled to room temperature before removing the filter and diluting the sample to a volume with water. The filter sample is from now on referred to as sample for dissolved inclusions. A blank sample was prepared using a filter through which an unused electrolyte was filtered and then was put through the same decomposition program.
2.4.2. Sample Preparation for FESEMA square shaped piece was cut from the film filter with inclusions (15×15 mm). The filter piece was placed on a carbon tab and coated with carbon to ensure the conductivity of the sample.
The amount of alloying elements and impurity elements was determined from stainless steel matrix and inclusions from two stainless steel slab samples. The soluble elements were measured from the filtrated 10% acetylacetone and 10% HCl after the extraction. The results for soluble and inclusion bound elements as well as impurity elements are presented in Table 5. The same elements were measured from extracted inclusions after the HNO3-HF dissolution. It can be stated that, when the dissolved mass of soluble alloying and impurity elements equals to the dissolved total mass, the element is among the steel matrix. If the soluble mass is notably less than the total dissolved mass, the element is mostly bound in the inclusions. The dissolved total mass of the elements was calculated using the data presented in Table 4 and the dissolved sample mass in an extraction.
| 10% acetylacetone | 10% HCl | |||||
|---|---|---|---|---|---|---|
| Soluble mass (μg) | Inclusions (μg) | Dissolved total mass in extraction* (μg) | Soluble mass (μg) | Inclusions (μg) | Dissolved total mass in extraction* (μg) | |
| Element | Sample 1 | |||||
| Al | 69 | < 1 | 18 | 123 | < 1 | 100 |
| As | 10 | < 1 | 18 | 32*1 | < 1 | 100 |
| Ti | 102 | 220 | 233 | 479 | 216 | 1260 |
| V | 76 | < 1 | 77 | 290 | < 1 | 420 |
| Cr (mg) | 32 | 0.05 | 28 | 160 | 0.01 | 150 |
| Cu | 210 | < 1 | 220 | 1130 | < 1 | 1170 |
| Element | Sample 2 | |||||
| Al | 59 | < 1 | 35 | 71 | – | 97 |
| As | 12 | < 1 | 11 | 11*1 | < 1 | 31 |
| Ti | 231 | 95 | 267 | 552 | 271 | 741 |
| V | 60 | < 1 | 80 | 110 | < 1 | 220 |
| Cr (mg) | 27 | 0.03 | 29 | 82 | 0.04 | 79 |
| Cu | 190 | < 1 | 190 | 520 | < 1 | 530 |
In addition to the soluble mass, titanium was the most common element by mass and the only element that was clearly detected in the inclusions dissolved in HNO3-HF. For sample 1, 220 μg of titanium was detected in the sample for dissolved inclusions that were extracted from the stainless steel sample in the 10% acetylacetone electrolyte and 216 μg of titanium was detected in the dissolved inclusions that were extracted using the 10% HCl electrolyte. The results for titanium mass in the extracted inclusions for sample 2 were 95 μg for 10% acetylacetone electrolyte and 271 μg for the 10% HCl electrolyte. It is clear, that the total amount of titanium is divided between the soluble titanium in the steel matrix and the titanium containing inclusions.
A small amount of vanadium was also detected in inclusions in addition to the soluble mass. Vanadium is a carbide forming element so it is plausible that they are present as inclusions in this stainless steel grade.
The results showed no clear difference in the amounts of aluminum present in the sample and a blank sample for inclusions extracted with 10% acetylacetone. It is possible that the amount of aluminum containing inclusions is too small to be detected when less than one gram of sample is dissolved during an electrolytic extraction.
The results for soluble copper and chromium in both electrolytes, arsenic in a 10% acetylacetone electrolyte and aluminum in a 10% HCl electrolyte were considered to be adequate. According to the results copper, chromium and arsenic are in the soluble form in the steel matrix –not bound in inclusions.
3.2. Examination of InclusionsThe inclusions in the process sample were studied from the steel sample after the electrolytic extraction using a FESEM. An inclusion size distribution and chemical composition has been determined in the previous article15) on the stainless steel sample 1 and it was concluded that a clear majority of the inclusions were Ti and Nb containing inclusions of 1–3 μm in size.
Aluminum was one of the elements of interest in this study for its low content in steel. Although the majority of the inclusions were similar, titanium containing inclusions as in sample 1, aluminum containing inclusions were also detected. Aluminum containing inclusions from sample 2 were imaged and an EDS analysis was performed. The images of the inclusions and their representative EDS analyses can be seen in Figs. 2(a)–2(d).
Figures of aluminum containing inclusions in stainless steel process sample 2 and their corresponding EDS analyses.
Aluminum containing inclusions were presumed to dissolve during extraction using 10% HCl. A 10% acetylacetone electrolyte is usually used in the examination of Al-containing inclusions.19) Even though Al2O3 is only very slightly soluble in acid, in this stainless steel a number of aluminum containing inclusions are present as a form of calcium aluminates20) which are soluble in HCl. Calcium was detected in aluminum containing inclusions in FESEM EDS analyses, thus supporting the assumption of the presence of calciumaluminates.
Determination of alloying and impurity elements from a stainless steel sample using an electrolytic extraction method and elemental analyses was performed. The elements investigated were aluminum, vanadium, titanium and chromium as alloying elements and arsenic and copper as impurity elements. The electrolytic extraction method and elemental analysis of soluble and insoluble alloying elements as well as impurities give additional information regarding stainless steel. The results can be compared to those obtained using the conventional spark-OES.
The purpose of this study was to determine whether the electrolytic extraction method is suitable for the determination of alloying elements and impurities in various content levels in a stainless steel sample and inclusions. The results for the determined elements in both electrolytes is presented in Table 6. The amount of titanium includes both soluble and inclusion containing titanium. Other elements that were measured were not detected in the dissolved inclusions, but only as a soluble form in the steel matrix.
| Sample 1 | ||||||
| Spark-OES | Al | Cu | As | V | Ti | Cr |
| Weight-% | 0.012 | 0.14 | 0.012 | 0.05 | 0.15 | 18 |
| Electrolytic extraction + elemental analysis | ||||||
| 10% acetylacetone | 0.044 | 0.14 | 0.006 | 0.05 | 0.21 | 21 |
| 10% HCl | 0.015 | 0.14 | 0.004 | 0.03 | 0.08 | 19 |
| Sample 2 | ||||||
| Spark-OES | Al | Cu | As | V | Ti | Cr |
| Weight-% | 0.022 | 0.12 | 0.007 | 0.05 | 0.17 | 18 |
| Electrolytic extraction + elemental analysis | ||||||
| 10% acetylacetone | 0.037 | 0.12 | 0.007 | 0.04 | 0.21 | 21 |
| 10% HCl | 0.016 | 0.12 | 0.003 | 0.03 | 0.19 | 19 |
It is clear that the titanium content determined from the surface of a solid sample, is affected by the inclusions in this particular stainless steel grade. A remarkable amount of the total titanium amount was measured from the HNO3-HF dissolved inclusions from both samples and extractions. This means that the spark-OES data for titanium can be considered misleading. The distribution of inclusions in a slab might also affect the results.
The results showed an excess amount of aluminum in the sample dissolved in 10% acetylacetone. A 10% acetylacetone electrolyte is used in extracting of aluminates and some calciumaluminates in steel21) but only a small amount of aluminum was detected in the extracted inclusions as opposed to negative absorbance obtained for the sample for insoluble elements in 10% HCl. The soluble aluminum determinations might also be affected by inclusion distribution in a process sample.
There is some level of uncertainty in the results concerning vanadium. Vanadium is a challenging element to measure using GFAAS. A high atomization temperature is required since the fact that vanadium is a carbide forming element. It is necessary to measure sample blanks between the actual samples to reduce the memory effect. Similar difficulties have been reported elsewhere.22) In addition to difficulties in GFAAS measurements, vanadium is a carbide forming element in steel. It is therefore possible that the presence of carbides in inclusions require different decomposition reagents to the HNO3-HF-media that was used in this study.
The amount of copper measured from both electrolytes closely matched the calculated values. The proposed method could be useful when it is necessary to measure copper as an impurity in stainless steel and also for steel samples that would contain a remarkable amount of CuS inclusions.
The determination of arsenic from 10% acetylacetone electrolyte closely aligned with the expected value. The determination of arsenic in a 10% HCl electrolyte seemed to suffer from instrumental interferences. This might be due to the formation of arsenic chloride that is volatilized in the pyrolysis stage.23)
There are some challenges when it comes to the determination of elements below 0.01 w-% level using the electrolytic extraction method combined with elemental analysis using GFAAS and FAAS. The sample dissolution in the 10% acetylacetone electrolyte is slower than in the 10% HCl electrolyte. The longer time the dissolution takes, the more the electrolyte is prone to contamination. On the other hand, the amount of dissolved elements in the electrolyte must be in the working range of the analytical instrument. This means that the amount of the sample that is dissolved during the electrolytic extraction should have enough element concentrations to be detected with the chosen technique.
In conclusion, the applicability of the electrolytic extraction method in the separation of steel matrix and inclusions prior to elemental analysis has some drawbacks when it comes to elements of low content. The amount of alloying element that is analyzed from the surface of a solid sample is clearly affected by the inclusions as it was in this case, the amount of titanium. The proposed method can be applied for several low content elements, but for some elements such as aluminum the inclusion characteristics have to be taken into account. The determination of copper, chromium and titanium on the other hand can be performed fairly well. The results were considered successful in various concentration levels. Although an electrolytic extraction method can be considered time consuming, it is still the only method that provides detailed information of element distribution between the inclusions and the steel matrix. The determination of impurity elements is also possible. The method is therefore considered an excellent tool in the field of steel and inclusion research.
This research is a part of the System Integrated Metal Processing (SIMP) research program coordinated by the Finnish Metals and Engineering Competence Cluster (FIMECC). Outokumpu Stainless Oy and the Finnish Funding Agency for Technology and Innovation (TEKES) are acknowledged for funding this work. The author would also like to thank Elisa Wirkkala for her help and skilled work with the measurements and Outokumpu Stainless for providing the studied stainless steel process sample. A sign of recognition is also in place for the staff at the Center of Microscopy and Nanotechnology at the University of Oulu.
