When a calculated result from inductively coupled plasma (ICP) analysis looks wrong, start with the input units and the stage of preparation that each quantity represents. The examples below use illustrative data and show how to check the values entered in the ICP sample concentration calculator.
Use the pattern of the discrepancy to choose what to check
↔ 表が見切れている場合は、左右にスクロールしてご覧ください。↔ Scroll the table horizontally to see any hidden columns.
| What you see | First records to check |
|---|---|
| About 1000 times or one thousandth of the expected value | mg/L versus µg/L, mL versus L, or mg versus g |
| A tenfold or fiftyfold discrepancy, close to a dilution factor | Duplicate dilution correction or a factor carried over from before remeasurement |
| One result differs greatly from the others | Sample mass, final volume and the match between the data row and the sample |
| Remeasurement of the same preparation gives a different result | Additional dilution, the selected instrument output column and cumulative dilution |
| The result turns negative after blank subtraction | Whether subtraction applies, whether both values are on the same basis and whether the software already corrected the result |
| Means agree but relative standard deviations differ | Individual results, replicate count, standard deviation formula and rounding |
These patterns point to records to check. Recover the inputs from the records before changing the calculation.
1. Mixing up mL and L, or mg/L and µg/L
50 mL is 0.050 L. Multiplying a concentration in µg/L by 50 instead of 0.050 gives an analyte mass 1000 times too high. The calculator expects the instrument concentration in µg/L, so convert any output reported in mg/L first.
For example, 0.010 mg/L is 10 µg/L. A 1 g sample made up to 100 mL and measured without further dilution gives 1.00 mg/kg. Typing 0.010 into the µg/L field gives 0.00100 mg/kg, one thousandth of the correct value. Likewise, enter 0.150 g for a 150 mg sample. Verify each number with its unit against the original record.
2. Applying the dilution factor twice
Use the concentration in the solution actually measured, before the software applies an additional dilution factor or calculates the original sample concentration. Check the instrument settings when the output column label is ambiguous.
For a 1 g sample brought to 100 mL, diluted tenfold and measured at 2 µg/L, the result is 2 mg/kg. If the software has already returned a dilution-corrected concentration of 20 µg/L and you enter that value with a factor of 10, the calculator gives 20 mg/kg. Follow the instrument output and preparation records to see which stage each concentration describes.
3. Entering the volume of acid as the final volume
If you used 5 mL of acid for digestion and brought the preparation to 50 mL, enter 50 mL. Record reagent additions, aliquot volumes and the final preparation volume separately. Entering 5 mL gives one tenth of the correct value with the other inputs unchanged. For a method without a make-up step, use the volume and equation specified by that method.
4. Subtracting a blank without checking the method
Even when both values are in µg/L, a blank and sample may represent different dilution stages. The blank field accepts a value on the same concentration basis as the measured result. Leave it empty when no subtraction applies or the software already made the same correction.
A method blank can be used to monitor contamination. For example, Method 6020B from the US Environmental Protection Agency (EPA) instructs laboratories to use a method blank for this purpose and not subtract it from associated sample results. Follow the analytical method when deciding whether the calculator's blank field applies. EPA Method 6020B, Section 9.7.1
- Confirm whether the method permits subtracting this blank.
- Check whether the same correction was already applied.
- Verify that the blank and sample are on the same measured-solution concentration basis.
Subtracting 0.10 µg/L from 0.05 µg/L gives −0.05 µg/L. Retain the measured value and check the analytical method's detection and reporting rules before assigning zero or reporting a non-detect.
5. Averaging instrument readings before converting independently prepared replicates
Convert each independently weighed and prepared replicate to mg/kg using its own sample mass, final volume and dilution factor. Average replicates that share the same original material, analyte and mass basis. The calculator's n counts independent preparations. Three instrument readings from one preparation yield information about repeat readings of that solution.
For example, suppose no blank subtraction or further dilution applies. Sample 1 is 1 g made up to 100 mL and measured at 10 µg/L. Sample 2 is 0.5 g made up to 100 mL and measured at 5 µg/L. Both yield 1 mg/kg. Averaging the instrument concentrations first gives 7.5 µg/L. Applying sample 1's preparation conditions to that mean gives 0.75 mg/kg instead of 1 mg/kg.
6. Judging a result using relative standard deviation alone
For the converted concentrations, the calculator uses the sample standard deviation (SD) with n − 1 in the denominator. When the mean is positive and n is at least 2, it displays the relative standard deviation (RSD) as 100 × SD ÷ mean. With n = 1, the sample SD and RSD are unavailable. The calculator suppresses RSD when the mean is zero or negative.
Near a mean of zero, small changes in the mean can produce large changes in RSD. Interpret it alongside the individual values, concentration level and the analytical conditions. US National Institute of Standards and Technology (NIST), Coefficient of Variation
↔ 表が見切れている場合は、左右にスクロールしてご覧ください。↔ Scroll the table horizontally to see any hidden columns.
| Three results (mg/kg) | Mean (mg/kg) | Sample SD (mg/kg) | RSD |
|---|---|---|---|
| 0.900, 1.00, 1.10 | 1.00 | 0.100 | 10.0% |
| 0, 0.100, 0.200 | 0.100 | 0.100 | 100% |
Both sets have the same absolute SD. The second mean is one tenth of the first, so the relative SD is ten times as high. When comparing with another spreadsheet, check whether it uses n − 1 or n. For 0.900, 1.00 and 1.10 mg/kg, the sample SD is 0.100 mg/kg and the SD with n in the denominator is about 0.0816 mg/kg. The standard error of the mean uses SD ÷ √n and answers a different question.
7. Using the displayed precision as the reporting precision
The calculator displays three significant figures. Apply the reporting rules of the method and the laboratory, including any requirements for measurement uncertainty. Retain the unrounded values during calculations.
For example, unrounded values of 1.001, 1.002 and 1.003 mg/kg have a mean of 1.002 mg/kg and a sample SD of 0.001 mg/kg. Each value displays as 1.00 at three significant figures. Calculating the SD from those displayed values alone gives zero.
8. Assuming that mg/kg always means a dry-mass result
The denominator is the mass you entered. The calculator applies no moisture adjustment. Check whether the reporting basis is dry mass or the material as weighed. Before comparing results, verify the mass basis and sample preparation as well as the unit.
Audit the calculation step by step
Trace the input values, any authorised blank subtraction, the additional dilution factor, the calculated analyte mass and the result in mg/kg. Start with a single sample and match its weighing record, preparation record and instrument output. Change one discrepant input at a time and recalculate.
If the input values and arithmetic agree, investigate the sample preparation or measurement. Keep these details for the review.
- Original instrument result, unit, output column and correction settings.
- Sample mass, final volume, dilution history and the measurement selected for reporting.
- Converted results for each replicate and the reporting mass basis.
- Inputs checked or corrected and how each change affected the result.
The calculator does not retain entries permanently. Save the calculation and its review in the laboratory record.
