Standard Sample Classification for Optical Emission Spectrometer Calibration

Optical emission spectrometer is a relative analytical technique, not an absolute measurement technique. This means that its parameters, including elemental concentrations or quantities, depend on the testing light intensity. The ratio of the measured parameter value to the light intensity is determined through calibration using certified reference materials. Without calibration, the spectrometer can only measure light intensity and cannot perform any quantitative chemical analysis.

Optical emission spectrometer is a fast and accurate analytical method, but it is also highly sensitive. Some necessary validation, maintenance, and calibration work can help ensure that the analyzer is always in optimal performance.

I. Factors Affecting Optical Emission Spectrometer Stability

1. Aging components affect sensitivity.

2. Contamination of the spectroscopic system lenses.

3. Energy decay of the optical system.

4. Dependence on fluctuations in external factors, such as humidity and temperature, fluctuations in power supply voltage, or fluctuations in the composition of the supplied gas (argon with a minimum purity of 99.995%).

5. Shocks caused by instrument location and its stability.

II. Optical Emission Spectrometer Calibration Steps

1. Tracing

This involves calibrating the optical system. Due to thermal expansion and contraction of materials, slight shifts in the instrument's exit slit may occur. We adjust the impact of this shift on the analytical results by moving the entry slit position.

2. Spectrum Calibration

Automatic optical path calibration. The optical system automatically scans spectral lines to ensure correct reception, eliminating the need for tedious peak scanning. The instrument automatically identifies specific spectral lines and compares them with the original stored lines to determine the drift position and locate the current pixel position of the analytical line for measurement.

3. Standardization

Using a standard sample with known accurate concentration as a reference, the instrument acquires the characteristic spectral signal of the standard sample (such as atomic emission line intensity) and establishes a calibration curve with the standard concentration of the standard sample. By comparing the deviation between the current signal and the initial calibration curve, correction coefficients (such as intensity correction and wavelength correction) are automatically calculated to update the instrument's detection parameters, offset system errors, and ensure the accuracy of subsequent sample measurements.

4. Type Standardization

If the channel analysis range of an analytical program is relatively large or there are interferences (such as superposition interference, matrix effects, etc.), and standardization still results in deviations in the control sample analysis, then type standardization is required. Type standardization can effectively calibrate the impact of various interferences on analytical results. It is particularly suitable for multi-sample analysis of metals and their alloys with basically the same content or grade. Type standardization is a method of locally correcting the working curve of the elemental analysis channel.

III: Classification of Calibration Standard Sample

1. High and Low Standard Sample

The "high and low standards" of spectrometer standardization is a two-point calibration method. The core is to use "low-concentration standards" and "high-concentration standards" to define the concentration range of the calibration curve, correcting the linear error of the instrument across the entire concentration range. This is more accurate than single-point calibration (especially suitable for wide concentration range detection, such as alloy element analysis of automotive parts).

(1). Core Logic and Operation

- Low Standard: Certified standards close to the detection limit of the analyte (e.g., 0.1% low standard for Cr in steel);

- High Standard: Certified standards close to the maximum permissible concentration of the analyte (e.g., 10% high standard for Cr in steel);

- Key: The high and low standards must be consistent with the matrix of the sample to be tested (e.g., aluminum alloy high and low standards for aluminum alloy samples) to avoid matrix interference.

(2). Advantages and Applicable Scenarios

- Advantages: Covers low and high concentration endpoints, solving the deviation problem of single-point calibration at concentration extremes, meeting the high-precision testing needs of automotive parts, etc. (e.g., alloy element concentration deviation ≤ ±1%);

- Scenarios: Applicable to samples with large fluctuations in element concentration (e.g., different batches of high-strength steel, aluminum alloy castings), or calibration of linear offset after long-term instrument use.

2. Control Sample

The "control sample" in a spectrometer is a quality control sample. Its core is a standard sample of known concentration that is "consistent with the matrix and similar in composition" to the sample to be tested. It is used to verify the accuracy of instrument detection in real time and to promptly detect measurement deviations (e.g., calibration drift, environmental interference). It is a key link in industrial quality inspection (e.g., alloy analysis of automotive parts).

(1). Essence

Not only a calibration standard sample, but also a "verification tool for test results." Its composition must be infinitely close to the sample to be tested (e.g., a retained sample of the same batch of aluminum alloy castings, a standard sample of the same grade of steel).

(2). Key Usage Points

- The control sample must be completely matched with the matrix of the sample to be tested (e.g., when testing 304 stainless steel parts, use 304 stainless steel as the control sample), otherwise the verification will fail due to the matrix effect;

- The concentration of the control sample must be within the commonly used range of the element to be tested (e.g., when testing Ni in steel at 5%-8%, the Ni content of the control sample should be around 6%).