JP2011209145A - Inclusion analysis method by emission spectrometric analysis - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a high-reliability inclusion analysis method when analyzing an inclusion in the metal by the spark discharge emission spectrometric analysis, by increasing the appearance frequency of emission pulses with the information on the inclusion, and by obtaining emission pulses influenced less by a main metal component.SOLUTION: The inclusion in the metal is analyzed by attaching substance particles to the surface of a sample, then by measuring emission pulses by performing spark discharge, and by selecting/analyzing the emission pulses with the information on the inclusion.

Description

  The present invention relates to a technique for rapidly analyzing inclusions in a metal using spark discharge emission spectrometry.

  In the spark discharge optical emission spectrometry, intermittent spark discharge is caused between a metal sample and a counter electrode in an inert gas atmosphere, and light emission generated by evaporation / excitation of the sample is measured. This analysis method is used for process control analysis because it can rapidly analyze many elements of a metal sample simultaneously.

  On the other hand, when an inclusion such as an oxide (eg, alumina) is present in the metal sample, a discharge phenomenon occurs between the discharge electrode and the inclusion of the metal sample (hereinafter referred to as “the inclusion is discharged. The emission intensity of inclusion constituent elements (eg, Al) is detected higher than the emission intensity of the base metal (solid solution). In recent years, this phenomenon has been used to detect the spectral line intensity of each element for each discharge pulse, and by selecting and analyzing the information with inclusion information from this light emission pulse, the particle size of inclusions in the metal Methods have been proposed for quickly determining distribution, abundance, composition, and the like.

  For example, in Patent Document 1, the emission intensity of inclusion constituent elements is set based on the actual measurement value obtained by microscopic measurement for the emission intensity of the inclusion constituent element in a predetermined intensity range from 0 to several hundred pulses at the initial stage of discharge. In addition, a rapid evaluation method for inclusions in a metal by an emission spectroscopic analysis method for calculating the number of existing inclusions, diameter, content, and average diameter using a predetermined formula is disclosed.

  Moreover, in patent document 2, the intensity value of the emission spectrum line obtained for every discharge is divided | segmented into the solid solution part and inclusion part of each element by the preset discharge time, and the intensity value of an inclusion part is calculated | required. The composition (type) of the inclusion is known from the detected element. In addition, for emission pulses with a value greater than the value determined by the oxygen emission intensity, the intensity value of inclusions is converted into a particle diameter based on the relationship between the intensity value obtained in advance with a standard sample and the particle diameter, and this operation is performed many times. A composition of inclusions in metal and a particle size distribution measuring method are disclosed in which the accumulation is repeated every discharge and the particle size distribution of inclusions is obtained.

  Furthermore, in Patent Document 3, the concentration of the element is obtained by using a calibration curve set in advance from the intensity value of the emission spectrum line of the inclusion constituent element, for an emission pulse having a value equal to or greater than the value determined by the oxygen emission intensity. The amount of inclusions is calculated by multiplying this by the amount of evaporation of the sample per discharge, and by dividing this by the density of inclusions, the volume of one spherical inclusion is converted into a particle size. A method for measuring the particle size distribution of inclusions in a metal to obtain a particle size distribution by repeating the operation is disclosed.

JP-A-4-238250 JP-A-9-43150 JP-A-9-43151

  However, in the above prior art, the frequency of appearance of light emission pulses having information on inclusions is not so high, and the light emission pulses having information on inclusions are also affected by the metal. For this reason, a highly reliable analysis result could not be given as a result.

  The present invention has been made to provide a highly reliable inclusion analysis method by increasing the frequency of appearance of light emission pulses having information on inclusions and obtaining light emission pulses that are less affected by metal.

As a result of studies to solve the above problems, the present inventors have obtained the following knowledge.
(1) If substance particles are attached to the surface of the metal sample, the substance particles protrude from the metal surface. Therefore, in the initial stage of discharge (specifically, 1 to 500 pulses), the substance particles are preferentially used. The discharge falls. At this time, the light emission pulse derived from the inclusion on the surface of the metal sample is generated in preference to the light emission pulse derived from the surface of the metal sample.

  (2) The light emission pulse emitted at the beginning of the discharge based on the discharge that has fallen on the surface of the metal sample under the condition that the substance particles are attached to the surface is derived from the metal in the metal sample due to the intensity of the emission intensity of Fe. Can be clearly distinguished from those derived from inclusions.

  (3) For this reason, from the light emission pulse containing light emission of an element contained in both the metal and the inclusion such as Al, the one derived from the inclusion can be selected regardless of the intensity of the emission intensity.

  (4) Therefore, compared to the prior art in which only those having a high emission intensity are determined as emission pulses derived from inclusions, the distribution of inclusions containing such elements on the surface of the metal sample (for example, the particle size) Distribution) can be accurately measured.

The following present invention has been completed based on the above findings.
Regarding the method of analyzing inclusions in a metal sample by spark discharge emission spectrometry, after depositing substance particles on the sample surface, a spark discharge is performed to measure the emission pulse, and information on the inclusion is obtained. A method for analyzing inclusions by emission spectroscopic analysis, which comprises analyzing and analyzing inclusions in a metal by selecting and analyzing emitted light pulses.

  Here, the substance particles have no compositional restriction unless the composition overlaps with the inclusions to be measured. In addition, as long as the particle size has a minimum size required to realize preferential discharge for the material particles, the viewpoint of workability such as ease of attaching the material particles is taken into consideration. Is set as appropriate. The amount of substance particles adhering to the surface also takes into account the above-mentioned effects and the disadvantages of adhering the substance particles (for example, light emission from a metal sample cannot be obtained if excessively adhering). Set as appropriate.

  The selection of the light emission pulse may be determined to be caused by inclusions when the Fe emission intensity in the light emission pulse is equal to or less than the threshold value. The threshold value can be set based on the emission intensity distribution of Fe at the stage where inclusions substantially disappear from the surface of the metal sample as in the late stage of discharge.

  The analysis method of the light emission pulse selected as being derived from the inclusion is appropriately set according to the information to be obtained.

  According to the present invention, inclusions can be analyzed by highly reliable spark discharge optical emission spectrometry.

It is a graph which shows each luminescence intensity of Fe, Si, Al, and Mg at the time series at the time of apply | coating SiC to the surface of a steel sample. It is a graph which shows the correlation with the light emission intensity of Fe, and each light emission intensity of Si, Al, and Mg in the discharge early stage (a) and the discharge late stage (b) at the time of apply | coating SiC to the surface of a steel sample. It is a graph which shows the correlation with the Al light emission intensity in the initial stage of discharge, and Mg light emission intensity at the time of apply | coating SiC to the surface of a steel sample. It is a graph which shows each luminescence intensity of Fe, Al, and Mg at the time series when measured by the measuring method which concerns on a prior art. It is a graph which shows the correlation with Al light emission intensity in the initial stage of discharge when measured by the measuring method which concerns on a prior art, and Mg light emission intensity. It is a graph which shows the relationship between the amount of substance particles adhering to the surface of a steel sample, and the number of light emission pulses with the information of inclusions. It is a histogram which shows the particle size distribution measurement result of an inclusion, (a) is a result measured by the measuring method concerning the present invention, and (b) is a result measured by extraction-SEM image processing. Observation image (a-1) of the sample surface after the discharge has dropped 10 times on the sample coated with SiC particles on the metal surface, and observation image (a-) of the sample surface after the discharge has dropped 500 times on the same sample 2), and an observation image (b-1) of the sample surface after the discharge has dropped 10 times on the sample on which the SiC particles are not coated on the metal surface, and the sample surface after the discharge has dropped 500 times on the sample It is an observation image (b-2).

  The present invention has been made on the basis of a phenomenon found when various kinds of substance particles are applied to a metal surface, and a spark discharge emission spectroscopic analysis is performed to investigate the behavior of various kinds of substance particles. That is, this is a phenomenon in which the discharge falls on the inclusions existing in the vicinity of the metal surface as well as the discharge falls on the material particles applied to the metal surface. By using this, many light emission pulses having information on inclusions with little influence from the metal can be obtained.

  In FIG. 1, when SiC particles are applied to the surface of a steel sample containing alumina inclusions containing a small amount of MgO and subjected to spark discharge optical emission spectrometry, it is obtained at the beginning of discharge (1-500 pulses). The emission intensity of Fe, Si, Al, and Mg is shown in time series.

  In the early stage of discharge, the emission of Fe as a sample matrix is suppressed, and the emission of Si caused by SiC particles as substance particles adhering to the sample surface is detected. At the same time, inclusion constituent elements Al and Mg are detected. It can be seen that luminescence is detected.

FIG. 2A shows the correlation between the emission intensity of Fe and the emission intensity of Si, Al, and Mg using the same data as the data shown in FIG.
From the graph showing the correlation between the light emission intensity of Si and the light emission intensity of Fe (the upper part of FIG. 2A), it is understood that there are two types of light emission pulses having different tendencies. One is a light emission pulse in which the light emission intensity of Si increases as the light emission intensity of Fe increases, that is, has a positive correlation. The other is a light emission pulse having a low correlation with the emission intensity of Fe and a relatively high emission intensity of Si.

  The latter light emission pulse is understood to be obtained on the basis of the fact that the discharge dropped on the substance particles adhering to the sample surface because of the high light emission intensity of Si. Further, since the latter emission pulse has a low emission intensity of Fe (in the example of FIG. 1, the emission intensity of Fe <1500), the emission pulse generated on the basis of the discharge that has fallen on the substance particles adhering to the sample surface is not grounded. It is understood that there is not much information derived from gold.

  Regarding the graph showing the correlation between the emission intensity of Al and the emission intensity of Fe (FIG. 2 (a) middle stage) and the graph showing the correlation between the emission intensity of Mg and the emission intensity of Fe (lower stage of FIG. 2 (a)), Similar to the graph showing the correlation between the emission intensity of Si and the emission intensity of Fe, emission pulses having a low correlation with the emission intensity of Fe and high emission intensity of Al and Mg are concentrated in a region where the emission intensity of Fe is low. There is a tendency to be. Since the correlation with the emission intensity of Fe is low and the emission intensity of these elements is high, it is understood that the emission pulse in this region has information mainly derived from inclusions.

  From the above, the point group plotted in the region where the emission intensity of Fe is low (in the example of the figure, the emission intensity of Fe <1500), the spark discharge is generated in the inclusions simultaneously with the SiC particles adhering to the sample surface. It is understood that it represents a fallen emission pulse.

Here, the result of observing the surface after discharging 10 times or 500 times for each of the sample coated with SiC particles on the metal surface and the sample not coated is shown in FIG.
The images shown in FIG. 8 are an observation image (a-1) of the sample surface after the discharge has dropped 10 times on the sample in which the SiC particles are coated on the metal surface, and the discharge has dropped 500 times on the same sample. Later observation image (a-2) of the sample surface, observation image (b-1) of the sample surface after the discharge has dropped 10 times on the sample on which no SiC particles are applied on the metal surface, and 500 for the same sample It is an observation image (b-2) of the sample surface after a circular discharge fell.

  As shown in FIG. 8, the sample coated with SiC particles has a less rough surface due to the discharge than the sample not coated at any number of discharges. It is also confirmed that the discharge has dropped in preference to the substance particles. In particular, in the case of the discharge of about 10 times, the polishing mark of the bare metal is clearly confirmed, and it is considered that the discharge is not substantially dropped on the bare metal at this number of discharges.

  From the above examination, the light emission pulse in the region where the emission intensity of Fe is low is an emission pulse having information on inclusions. By selecting the emission pulse with low emission intensity of Fe and analyzing the selected emission pulse In other words, the information on the inclusions can be selectively extracted.

  FIG. 2 (b) shows a late discharge period (1501-2000 pulses) as a comparison. It is considered that no emission pulse having Si emission intensity having a low correlation with the emission intensity of Fe is generated, and the substance particles are not substantially attached to the analysis region of the sample surface. In addition, there are almost no emission pulses plotted in a region where the emission intensity of Fe is low (Fe <1500). Al and Mg are similar to Si and have a substantially single tendency. From these results, it is assumed that the discharge almost fell to the metal in the late stage of discharge.

  From the examination based on the results of FIG. 2 above, when substance particles are attached to the surface according to the present invention, only the light emission pulses from the inclusions are selected by selecting the light emission pulses having a low light emission intensity of Fe. It is guided that we can do it.

This point will be further described with reference to FIG.
FIG. 3 shows the correlation between the Al emission intensity and the Mg emission intensity at the initial stage of discharge, that is, 1-500 pulses. Here, in FIG. 3, when plotting so that the emission intensity of Fe in the emission pulse is 1500 or more and less than 1500, it is understood that both have independent correlations. .

  More specifically, the former (Fe emission intensity of 1500 or more) forms a point group having an inclination reflecting the Al / Mg concentration ratio of the base metal (solid solution). On the other hand, the latter (light emission pulse having information of inclusions) is a point group having a slope reflecting the concentration ratio of Al and Mg in the inclusions. Therefore, according to the measuring method according to the present invention, by attaching substance particles to the sample surface, the light emission pulse having the information of inclusions and the light emission pulse having the information of the bare metal are included in the pulse. It can be clearly separated by the intensity of the emission intensity of Fe. Thus, by determining that the light emission pulse is derived from inclusions when the light emission intensity of Fe is relatively weak, information about the inclusions can be obtained from the light emission pulse having a low light emission intensity of Al. In this regard, since the emission intensity of Al has a positive correlation with the particle size of inclusions, according to the present invention, the particle size distribution of inclusions can be measured more accurately than in the prior art.

  On the other hand, using the same sample, sample pretreatment is performed by cutting or belt-polishing, and the emission intensity of each of Fe, Al, and Mg in the emission pulse measured by a conventional technique in which material particles do not adhere to the sample surface. This is shown in FIG. 4 in a time series (1-500 pulses).

  Although the emission intensity of Al is high (in the example of the figure, the emission intensity of Al> 1500), the emission pulse can be identified as an emission pulse having information on inclusions, but the emission intensity of Fe detected at the same time is high Therefore, information obtained from the light emission intensity of Al is affected by the metal.

  In addition, the number of light emission pulses having information on inclusions is about 80 in the example of the present invention, but is about 20 in the example of the prior art. From this result, it is understood that the generation of light emission pulses from inclusions is concentrated in the early stage of discharge by attaching substance particles to the sample surface. Needless to say, this contributes to shortening the measurement time and improving the measurement accuracy.

  Further, FIG. 5 shows the correlation between the emission intensities of Al and Mg. A plurality of point groups having a correlation that can be clearly distinguished from each other as shown in FIG. 3 of the present invention is not recognized. For this reason, in the measurement according to the prior art, it must be determined that a light emission pulse having a high Al light emission intensity is derived from inclusions. However, as described above, since the emission intensity of Al has a positive correlation with the particle size of inclusions, if it is determined that the emission pulse with low emission intensity of Al does not originate from inclusions, It is not possible to obtain any information about inclusions with small.

  As described above, according to the present invention, many emission pulses having inclusion information that is not affected by the metal can appear at the beginning of the discharge, and more information on the metal can be obtained by using the emission intensity of Fe. It becomes possible to easily distinguish it from the light emission pulse it has.

  The number of emission pulses having inclusion information detected in the present invention depends on the amount of substance particles adhering to the sample surface. The amount of the adhered substance particles can be estimated by the sum of the emission intensity of the elements constituting the substance. For example, when the adhered particles are SiC, the sum of the Si emission intensities of the emission pulses having an Fe emission intensity of less than 1500 in FIG. FIG. 6 shows the correlation between the amount of adhering substance particles and the number of emission pulses having information on detected inclusions. Note that the number of light emission pulses on the vertical axis in FIG. 6 is the number of light emission pulses with Fe emission intensity of less than 1500. A good correlation is observed. Note that the slope of this correlation reflects the number of inclusions in the sample.

As the material particles to be adhered to the sample surface, any inorganic material (SiC, BN, TiO ₂ , CaCO ₃ etc.) or organic material (sucrose, flour etc.) can be used as long as the sample is not corroded. The size of the particles is preferably fine particles of the order of μm or less, which easily adhere to the sample surface. If the amount to be adhered is too large, it becomes impossible to measure a light emission pulse having information on the surface of the sample. As a method for adhesion, it is possible to directly apply substance particles, disperse / dissolve them in a solvent, and then apply and dry, or a type of polishing in which polishing powder remains during sample polishing.

As an example of the present invention, inclusions in steel samples a to d (0.2 mass% C-0.2 mass% Si-0.8 mass% Mn steel) containing alumina inclusions containing a small amount of MgO. An analysis example of particle size distribution, abundance and composition will be described. As a method for attaching the substance particles to the sample surface, a type of polishing in which polishing powder remains at the time of sample polishing was adopted. Since this polishing contained KBF ₄ in addition to the SiC of the abrasive grains, the substitute value of the amount of material particles adhering to the sample surface is the sum of the emission intensity of B which is an element not included in the steel sample. (ΣB) was used.

  After the sample was polished as described above, it was set in a spark discharge optical emission spectrometer. Spark discharge was performed 2000 times, and the emission intensity of Fe, B, Al, and Mg was measured and stored for each discharge pulse. The same measurement was repeated several times with different discharge locations. A light emission pulse having an Fe light emission intensity smaller than a preset Fe light emission intensity (1500 in the present embodiment) in the 1-500 pulse region was selected as a light emission pulse having inclusion information.

Since the Al emission intensity of the emission pulse having the information of the inclusion is proportional to the mass (volume) of the inclusion (Al ₂ O ₃ ) where the discharge has dropped, the 1/3 power of the Al emission intensity is the inclusion (Al ₂ O). _It is proportional to the particle size of ₃ ). Therefore, obtain the 1/3 power of the Al emission intensity of the emission pulse having the information of the inclusions detected in the sample, and convert the numerical value obtained by the conversion factor that is determined in advance using the standard sample into the particle size. The emission pulses were integrated to determine the particle size distribution of inclusions (Al ₂ O ₃ ).

Since the Al emission intensity and Mg emission intensity of the emission pulse having the information of the inclusion are proportional to the Al ₂ O ₃ mass and MgO mass of the inclusion where the discharge has dropped, the sum of the Al emission intensity (ΣAl), the Mg emission intensity The sum (ΣMg) is proportional to the Al ₂ O ₃ abundance and MgO abundance. However, in the case of the present invention, the number of emission pulses having information of inclusions is proportional to the amount of substance particles adhering to the sample surface (monitor: ΣB), so ΣAl / ΣB and ΣMg / ΣB are present in the amount of Al ₂ O ₃ . , Proportional to the amount of MgO present.

Therefore, ΣAl / ΣB and ΣMg / ΣB of the sample are calculated using a calibration curve [Al ₂ O ₃ abundance vs.. ΣAl / ΣB, MgO abundance vs. The sample was converted into Al ₂ O ₃ abundance and MgO abundance by ΣMg / ΣB], and averaged over the measurement times to calculate the Al ₂ O ₃ abundance and MgO abundance of the sample. Furthermore, MgO (mass%) composition was calculated | required from both.

FIG. 7 (a) shows the particle size distribution of inclusions determined by the measuring method based on the present invention, and FIG. 7 (b) shows the particle size distribution of inclusions determined by SEM / EDS-image processing after chemical extraction and separation. Indicates. Comparing the two, in the present invention, some of the inclusions were scraped when polishing the sample, so it was not completely consistent with the extraction method, but a particle size distribution reflecting the difference in the inclusion particle size distribution of the sample was obtained . In addition, the results of Al ₂ O ₃ abundance, MgO abundance and MgO (mass%) composition obtained by (a) a measurement method based on the present invention and (b) a chemical analysis method after chemical extraction separation can be compared. Table 1 shows. A fairly good agreement was observed.

  As described above, according to the present invention, the particle size distribution, abundance, and composition of inclusions can be favorably obtained.

Claims (1)

  Regarding the method of analyzing inclusions in a metal sample by spark discharge emission spectrometry, after depositing substance particles on the sample surface, a spark discharge is performed to measure the emission pulse, and information on the inclusion is obtained. A method for analyzing inclusions by emission spectroscopic analysis, which comprises analyzing and analyzing inclusions in a metal by selecting and analyzing emitted light pulses.