Sampling and Sample Preparation

Introduction

The first chemical operation in an analysis may happen before the instrument or laboratory ever sees the material: selecting a sample. A highly precise method can give a misleading answer if the sample is not representative. Preparation then has to expose the analyte to measurement without losing it or changing the chemical form being sought. The sampling and preparation plan is part of the measurement, not administrative paperwork.

Core explanation

Define the population before taking a portion. “Nitrate in a lake” might mean a particular depth and date, a daily average at one outlet, or a spatial average across many sites. These are different targets and need different sampling designs. For a heterogeneous solid, collect multiple increments across the material rather than taking only a convenient surface scoop. Combine and reduce the bulk sample using a documented procedure that avoids systematic selection of large or fine particles.

Homogenisation can reduce variation inside a collected sample. Grinding a dry solid, mixing a liquid or blending suitable material makes subsequent portions more similar. However, homogenisation cannot repair a collection plan that omitted an important part of the population. It can also change the sample: grinding may heat a material, release volatile components or expose it to air. Therefore the preparation must match the analyte's stability.

Containers and storage matter. Trace-metal work may require carefully cleaned containers and appropriate preservation, while species-sensitive measurements may require rapid separation or analysis to prevent oxidation and biological change. A volatile organic compound can escape from a partly filled container. A dissolved-metal measurand may require filtration at a specified stage; a total-metal measurand may instead require complete digestion of suspended solids. These choices define what the eventual number means.

Sample preparation often changes physical form: dry to determine mass basis, dissolve to create a homogeneous solution, digest a resistant matrix, extract a target into another phase, or dilute a concentrated extract. Each step has potential loss, contamination or incomplete recovery. A procedural blank can reveal reagent contamination; a matrix spike, in which a known amount of analyte is added before processing, can test recovery in that sample type. A spike result does not prove every native analyte form behaves identically, but it is useful evidence.

An aliquot is a known portion of a larger mixture. If 10.00 mL is removed from a 100.0 mL solution that is homogeneous, the aliquot contains one tenth of the dissolved analyte. If the solution has particles or has not been mixed, that factor can fail. Record initial mass, final volumes, transfer volumes and any losses so that the measured concentration can be converted back to the original material.

Step-by-step reasoning

1. Define the material, place, time and chemical form represented by the result. 2. Choose independent collection points or increments that cover relevant heterogeneity. 3. Select containers, preservation and transport suited to analyte stability. 4. Homogenise and prepare the laboratory portion while avoiding loss or contamination. 5. Track every mass, aliquot and final volume; check blanks and, where useful, recovery.

Visual explanation

Picture a soil pile divided into a grid. Mark increments taken from different depths and positions, then show them mixed into a bulk sample. A narrowing sequence takes a laboratory portion, dissolves it and transfers a known aliquot to a flask. Beside each stage label one hazard: spatial bias, poor mixing, incomplete dissolution, transfer loss or contamination.

Real-world analogy

To judge a pot of soup, tasting only a spoonful of oil from the surface tells little about the whole pot. Stirring before tasting improves the portion's representativeness, but if the pot contains chunks at the bottom, one spoonful may still miss them. Chemical samples have analogous gradients and particles, and some stirring methods may themselves change the material.

Real-world example

A laboratory analysing metal content in a bag of crushed ore takes increments from several locations, mixes them and reduces the composite to a manageable mass. It records the dry mass basis, digests a weighed portion and dilutes to a known volume. If a resistant mineral remains undissolved, the measured solution may understate total metal despite perfect instrument calibration.

Why?

Why can a small sampling error dominate an otherwise precise instrument? The instrument measures only the submitted portion. If that portion contains a different analyte fraction from the material about which the decision is made, every repeat reading inherits the same mismatch. More repeated readings of the wrong portion reduce instrumental scatter but do not fix representativeness.

Common misconception

“Three readings are three independent samples” is wrong if all readings come from one prepared solution. Replicate instrument injections test measurement repeatability; independent field samples test spatial or temporal variability. Another error is assuming that every acid digestion measures total element content; resistant phases may remain unless the method is validated for that matrix.

Worked example

Suppose a 2.000 g solid sample is fully dissolved and made to 250.0 mL. A 25.00 mL aliquot is analysed and found to contain 3.00 mg analyte. The aliquot is one tenth of the homogeneous solution, so the original 2.000 g portion contains 30.0 mg analyte. Mass fraction = 30.0 mg/2000 mg = 0.0150, or 1.50%. If dissolution was only 80% complete, the calculation would be systematically low and should not be reported as total content without correction and validation.

Quick check

1. Does analysing the same prepared solution five times establish that the field sample represented an entire river? Answer: No. It tests repeatability for that solution. Representative river sampling requires a plan covering relevant locations, depths and times, depending on the defined measurand.

Exam focus

Distinguish a field sample, laboratory portion, aliquot and replicate reading. State whether the target is dissolved, total or a particular chemical form. Track masses and volumes through each preparation step. In a question about reliability, identify sampling bias and incomplete recovery before blaming instrument noise.

Advanced insight

Sampling uncertainty can be estimated by collecting independent replicate field samples and processing them separately. Comparing their variation with repeated measurements of one prepared sample helps separate between-sample heterogeneity from analytical repeatability. The two contributions can be large at different scales, and the correct study design depends on the population to which the result will be applied.

Summary

Representative sampling connects the laboratory result to the material of interest. Preparation must preserve the measurand while making the analyte accessible. Homogenisation, dissolution, extraction, aliquoting and preservation can each introduce bias. Independent samples, blanks, recovery checks and complete records help establish what a reported concentration actually means.

Practice questions

1. Why is a surface scoop from a heterogeneous ore pile risky? Answer: Particle size and mineral composition may vary with position and depth. A surface scoop may systematically over- or underrepresent the metal content of the whole pile.

2. What does a procedural blank reveal? Answer: It follows the preparation without intentional analyte and can reveal contamination from reagents, vessels, transfer steps or the laboratory environment.

3. Why might immediate filtration change the reported result? Answer: Filtration removes particle-bound analyte. The result then refers to the operationally defined dissolved fraction rather than total analyte in the original unfiltered sample.