Sample preparation in analytical laboratories and how it affects reliable results

Why sample preparation determines analytical reliability

Sample preparation is the controlled sequence used to turn collected material into a test portion that an instrument or analytical method can measure reliably. Depending on the work, it may include homogenizing, grinding, drying, filtering, dissolving, digesting, extracting, diluting, concentrating, or preserving a sample before analysis. The aim is not just to make the sample easier to test. The prepared portion has to remain representative, stable, uncontaminated, and compatible with the chosen method.

For that reason, sample preparation deserves the same scrutiny as the instrument method itself. A well-calibrated chromatograph, spectrometer, microscope, or balance cannot compensate for a sample that changed during storage, lost analyte during extraction, absorbed contamination from a container, or was subsampled from a non-homogeneous material. Published guidance from organizations such as IUPAC, NIST, EPA, FDA, and ISO points to the same principle: reliable analysis begins before the measurement step.

What sample preparation must achieve

The right workflow depends on the matrix, analyte, technique, and the decision that will be made from the result. A food sample prepared for pesticide residue analysis, a water sample preserved for metals testing, a polymer specimen prepared for microscopy, and a plasma sample extracted for LC-MS/MS do not use the same procedure. They do, however, share several technical goals.

• Representativeness: the test portion should reflect the material or population being evaluated.
• Stability: the analyte and matrix should not change materially before measurement.
• Compatibility: the prepared sample must suit the instrument, detector, column, reagent system, or imaging technique.
• Recovery: the procedure should not lose a meaningful amount of the target analyte unless the loss is measured and corrected.
• Cleanliness: the procedure should avoid introducing contamination, carryover, or interfering substances.
• Traceability: every handling, transfer, dilution, and storage step should be documented well enough to reconstruct the result.

IUPAC terminology distinguishes sampling error from analytical error and links sampling error to non-homogeneity in the parent material. The distinction is important because many laboratories invest heavily in instrument precision while underestimating the variability introduced when a test portion is taken from a larger, uneven sample. For solids, powders, biological tissues, soils, composites, and heterogeneous liquids, subsampling can contribute more uncertainty than the final instrumental measurement.

Common preparation steps and the risks they control

Most preparation steps are intended to solve a specific analytical problem. When a step is selected by habit rather than by method need, it can introduce a new source of error. The table below summarizes common laboratory steps, why they are used, and what has to be controlled.

Preparation step	Typical purpose	Main risk to control
Mixing or homogenization	Reduce variability before subsampling	Incomplete mixing, heat generation, or segregation after mixing
Grinding or milling	Reduce particle size and improve extraction or digestion	Analyte loss, cross-contamination, moisture change, or metal wear debris
Drying	Remove water or standardize moisture basis	Loss of volatile compounds, oxidation, or thermal degradation
Filtration or centrifugation	Remove particulates that interfere with analysis	Adsorption of analyte to filters, incomplete separation, or contamination from consumables
Dilution	Bring concentration into the working range	Pipetting error, wrong dilution factor, or matrix mismatch
Digestion or dissolution	Transfer analyte into solution for analysis	Incomplete digestion, reagent contamination, or analyte volatilization
Extraction	Separate analyte from matrix components	Poor recovery, emulsion formation, solvent impurities, or variable phase separation
Derivatization	Improve detectability, volatility, or chromatographic behavior	Incomplete reaction, unstable derivatives, or reagent batch effects

NIST sample preparation materials for engineered nanomaterials and related measurement work illustrate a broader point: each measurement may require the sample to be in a specific physical form. A specimen suitable for electron microscopy may not be suitable for bulk chemical analysis, and a solution prepared for ICP analysis may not preserve the structural information needed for particle characterization. The preparation route should therefore be chosen by working backward from the measurement objective, not by routine alone.

Where avoidable error usually enters the workflow

Preparation errors are often practical rather than theoretical. They occur when routine handling steps are not treated as measurement-critical operations. A label is handwritten unclearly, a sample warms on the bench while a batch is assembled, a homogenized powder segregates during transfer, a pipette is used outside its reliable range, or a blank is omitted because the run is already behind schedule. Each issue may look minor in isolation, but each can change the reported result.

Non-homogeneous samples

Non-homogeneity is one of the most persistent challenges in sample preparation. Powders may segregate by particle size or density. Environmental samples may contain localized contamination. Biological samples may vary within a tissue or fluid fraction. If the laboratory takes a small portion before adequate mixing, milling, splitting, or compositing, the final result may describe that portion rather than the original material. Replicate preparation, duplicate subsampling, and documented homogenization criteria help reveal and reduce this risk.

Contamination and carryover

Contamination can come from glassware, plasticware, filters, solvents, reagents, milling tools, digestion vessels, the work area, or previous samples. Carryover is especially important when high-concentration samples are prepared near trace-level samples. FDA bioanalytical guidance highlights selectivity, matrix effect, carryover, and stability because these factors can change measured concentrations even when the instrument appears to be functioning normally. Preparation blanks, reagent blanks, equipment rinses, and clean batch sequencing are practical safeguards.

Stability and holding time

Some samples change quickly after collection. Microbial populations may increase or decline, volatile compounds may evaporate, metals may adsorb to container walls, and redox-sensitive species may transform. EPA SW-846 guidance discusses preservation, storage containers, and holding times, while noting that specific methods can provide more targeted requirements than general tables. In practice, holding time is not an administrative detail. It is part of the measurement conditions and should be planned before samples arrive.

How to choose a fit-for-purpose preparation method

A strong sample preparation method starts with the analytical question. The laboratory should define what must be measured, in which matrix, at what concentration range, with what uncertainty, and for what decision. From there, the preparation design can be evaluated against the method and the sample.

1. Define the measurand. Decide whether the result should represent total content, extractable content, dissolved fraction, bioavailable fraction, surface contamination, or another operationally defined quantity.
2. Understand the matrix. Consider moisture, fat, salt, protein, organic matter, particle size, pH, viscosity, and likely interferences.
3. Match the instrument requirement. Instruments may need clear solutions, specific solvents, limited dissolved solids, particle-free extracts, defined pH, or particular specimen dimensions.
4. Assess analyte stability. Check whether light, temperature, oxygen, microbes, enzymes, or container materials can alter the analyte.
5. Control recovery and bias. Use spikes, certified reference materials where available, surrogate standards, or matrix-matched controls to evaluate losses and interferences.
6. Document the limits. State when the procedure is not appropriate, such as unsuitable matrices, high particulate loads, unstable analytes, or concentrations outside the validated range.

In regulated or quality-managed environments, laboratories should not treat sample preparation changes as minor unless the method allows them. A different extraction solvent, digestion temperature, filtration material, dilution scheme, or storage condition can alter recovery, matrix effect, or stability. If the change can affect the result, verification or validation evidence is needed before routine use.

Quality controls that make preparation defensible

Quality control should cover the preparation stage, not only the final analysis. A calibration curve confirms instrument response, but it does not prove that the sample was extracted completely or that the container was clean. The controls below help connect the reported result to the actual material received. See also: buying guides.

• Method blanks: detect contamination from reagents, vessels, tools, and the preparation environment.
• Preparation duplicates: show variability introduced by subsampling and preparation, not only instrument repeatability.
• Matrix spikes: evaluate recovery in the actual or similar matrix.
• Surrogate or internal standards: help track losses and instrument variability when appropriate for the method.
• Certified reference materials: provide an external check on bias when a suitable material exists.
• Control charts: show whether recovery, blanks, or duplicate precision are drifting over time.
• Chain-of-custody and condition checks: document receipt temperature, container condition, preservation status, sample mass or volume, and any deviations.

ISO/IEC 17025 emphasizes procedures for transportation, receipt, handling, protection, storage, retention, and disposal of test or calibration items. Even outside formal accreditation, the same logic applies: if the laboratory cannot show how the sample was protected, it becomes harder to defend the result when it is questioned.

A practical checklist before testing begins

Before a prepared sample reaches the instrument, the analyst or reviewer should be able to answer a short set of questions. This checklist is not a substitute for a validated method or standard operating procedure, but it helps identify gaps that commonly lead to avoidable rework.

• Was the sample identity confirmed against the submission record, label, and batch worksheet?
• Were receipt conditions documented, including temperature, preservation, container condition, and holding time where relevant?
• Was the sample mixed, split, or homogenized using a procedure suitable for its matrix?
• Were preparation tools, vessels, filters, and solvents appropriate for the target analyte and detection limits?
• Were blanks, duplicates, spikes, or reference materials included at the preparation stage?
• Were all masses, volumes, dilution factors, extraction times, temperatures, and deviations recorded contemporaneously?
• Was the prepared sample stored under conditions that preserve the measurand until analysis?
• Does the final preparation match the validated or verified method conditions?

The strongest workflows make these checks routine. They also separate facts from assumptions. For example, the fact may be that a sample was received above a required temperature; the analytical judgment may be that the result is still usable for a stable inorganic analyte; the report note may need to disclose the deviation. Keeping those distinctions clear improves both scientific quality and client communication.

Frequently asked questions

What is the difference between sampling and sample preparation?

Sampling is the act of selecting material from a larger population, lot, environment, process, or batch. Sample preparation happens after collection and turns that material into a form suitable for testing. The two are connected because poor sampling usually cannot be repaired by excellent preparation.

Why is sample preparation often a major source of error?

It involves many manual or semi-manual steps: weighing, transferring, mixing, extracting, filtering, diluting, heating, cooling, and storing. Each step can introduce variability, contamination, loss, or transformation. Unlike instrument noise, preparation error may be difficult to detect unless the workflow includes blanks, duplicates, spikes, and reference materials.

Can one sample preparation method work for all instruments?

No. A method must match the matrix, analyte, measurement technique, and reporting objective. For example, a strong acid digestion may be suitable for total metals analysis but inappropriate if the goal is to measure a specific chemical species or particle structure.

When should a laboratory revalidate or verify a preparation change?

Verification or validation should be considered when a change can affect recovery, selectivity, stability, contamination, matrix effect, detection limit, or uncertainty. Examples include changing solvent, extraction time, filtration material, digestion temperature, sample mass, dilution ratio, or storage condition.

What is the simplest way to improve preparation quality?

The most effective first step is to document the workflow as a measurement process rather than a set of informal handling habits. Clear acceptance criteria, trained analysts, controlled consumables, preparation-stage quality controls, and deviation records usually improve reliability faster than changing instruments.

Conclusion

Sample preparation is not a preliminary chore. It is a decisive part of analytical measurement. The prepared test portion must remain representative of the original material, stable enough for the intended analysis, and compatible with the method. Laboratories that control preparation through documented procedures, appropriate quality controls, and clear acceptance criteria reduce avoidable uncertainty before the first injection, scan, titration, or reading occurs. That discipline is what turns a measured signal into a result that can be trusted.