Characterization is the systematic process of measuring and defining the properties of a material, substance, or biological sample so you know exactly what it is, how it behaves, and whether it meets a specific standard. The approach varies depending on what you’re working with, but the core workflow is consistent: prepare your sample, select the right analytical techniques, collect data, and validate your results against known references or acceptance criteria.
The General Workflow
Regardless of the field, characterization follows a predictable sequence. You start with a clear question: What do you need to know about this sample? That question determines which properties matter and which techniques to use. From there, the process moves through sample preparation, instrument selection, data collection, and interpretation. Each step feeds the next, and skipping one typically invalidates everything downstream.
In an industrial or pharmaceutical setting, the process often begins with a risk assessment. You identify which quality attributes are critical, then design studies around those attributes. For example, a drug manufacturer would first determine which properties of a nanoparticle (size, surface charge, stability) could affect how the drug performs in the body, then build a testing plan around those specific measurements. The final step is always validation: confirming that your measurements are accurate, reproducible, and sensitive enough to detect meaningful differences.
Chemical and Molecular Characterization
When you need to know what something is made of at the molecular level, you’re looking at two broad families of techniques: spectroscopy and chromatography. Spectroscopy works by measuring how a substance interacts with light or electromagnetic radiation. Different wavelengths reveal different things. Infrared spectroscopy identifies the types of chemical bonds in a molecule, which helps you determine its functional groups. Nuclear magnetic resonance spectroscopy maps the arrangement of atoms, giving you a detailed picture of molecular structure. Ultraviolet-visible spectroscopy tells you about electronic transitions, which is useful for identifying conjugated systems or metal complexes.
Chromatography separates a mixture into its individual components so you can identify and quantify each one. High-performance liquid chromatography is the workhorse for non-volatile compounds, pushing a liquid sample through a packed column that separates molecules based on their chemical properties. Gas chromatography does the same for volatile compounds using a gas carrier. The real power comes from combining these approaches. Liquid chromatography paired with mass spectrometry, for instance, first separates the components and then identifies each one by its molecular weight and fragmentation pattern. This combination is standard for characterizing biological samples like blood, urine, or tissue extracts.
Physical and Structural Characterization
Chemical identity is only part of the picture. Many materials behave differently depending on their physical structure, even when their chemistry is identical. A powder with particles averaging 50 nanometers will dissolve, scatter light, and interact with cells very differently than the same compound with particles averaging 5 micrometers.
Microscopy is the primary tool here. Scanning electron microscopy (SEM) is one of the most widely used techniques because it reveals both surface shape and topography at high magnification. It fires a focused beam of electrons at the sample and builds an image from the signals that bounce back. Transmission electron microscopy (TEM) sends electrons through the sample instead, producing flat two-dimensional images that show internal structure and fine detail, though it doesn’t capture surface topography the way SEM does. Atomic force microscopy (AFM) takes a different approach entirely, dragging a tiny physical probe across the surface to build a three-dimensional map. AFM is particularly good at measuring surface roughness and texture at the nanoscale.
Beyond microscopy, X-ray diffraction reveals whether a material is crystalline or amorphous and identifies its crystal structure. Thermal analysis techniques measure how a material responds to heat, telling you about melting points, decomposition temperatures, and phase transitions. Mechanical testing covers hardness, elasticity, and tensile strength.
Biological and Clinical Characterization
Characterizing biological samples adds layers of complexity because the material is inherently variable and often unstable. Blood, tissue, saliva, cerebrospinal fluid, and cell cultures all require careful preparation before any analysis can begin. Extraction techniques isolate the molecules of interest from the complex biological matrix surrounding them. Liquid-liquid extraction, solid-phase extraction, and solid-phase microextraction are all common preparation methods, each suited to different sample types and target molecules.
Once prepared, biological samples are typically analyzed using the same chromatography-mass spectrometry combinations used in chemistry, but with protocols tuned for biological complexity. Metabolomics studies, for instance, use mass spectrometry to profile hundreds or thousands of small molecules in a single urine or blood sample, building a chemical fingerprint of a patient’s metabolic state.
At the clinical level, characterization takes on a different meaning. Researchers characterize patient populations by defining phenotypes: observable traits like disease status, drug response, or physiological measurements. Electronic health records have become a major data source for this work, combining billing codes, lab results, vital signs, medication records, clinical notes, and family history into a rich dataset. The most accurate phenotype definitions use a combination of structured data (like diagnostic codes and lab values) and unstructured data (like physician notes). Billing codes alone are the most commonly used resource for identifying phenotypes, but validating whether a patient truly has a given condition generally requires reviewing the clinical notes themselves.
What Regulators Expect
If your characterization data will support a product application, regulatory agencies set specific expectations. The FDA’s guidance on nanomaterial-containing drug products, for example, requires you to measure and report five core attributes: chemical composition, average particle size, particle size distribution, general shape and morphology, and both physical and chemical stability. Depending on the product and how it’s administered, additional parameters may include surface charge, porosity, crystal form, coating properties, impurity levels, drug loading and release rates, and sterility.
On the testing side, international standards from organizations like ASTM International and ISO define how to prepare specimens and run tests so that results are comparable across laboratories. ASTM, established in 1898, maintains over 130 technical committees that develop consensus standards. ISO standards are adopted across the EU and most European countries, while ASTM standards see global use, particularly where more detailed testing procedures are needed. The two systems sometimes cover the same testing objectives but differ in technical procedures, which means data from one standard isn’t directly comparable to data from the other. Choosing the right standard before you begin testing matters.
Validating Your Results
Characterization data is only useful if it’s reliable. Validation confirms that your analytical method actually measures what you think it measures, with sufficient sensitivity and reproducibility. Two key thresholds define the lower limits of your method. The detection limit is the smallest amount of a substance your method can reliably distinguish from background noise, generally accepted at a signal-to-noise ratio of 3 to 1. The quantitation limit is the smallest amount you can measure with acceptable accuracy, set at a signal-to-noise ratio of at least 10 to 1.
Accuracy is reported as the mean percent recovery of a known amount of analyte spiked into your sample, along with a confidence interval. Precision is reported as the standard deviation or relative standard deviation (coefficient of variation) across repeated measurements, again with a confidence interval. These numbers let anyone reviewing your data judge whether your method is fit for purpose. A method with a coefficient of variation of 2% is far more trustworthy for quality control than one with a CV of 15%, even if both technically produce results.
Choosing the Right Technique
The most common mistake in characterization is selecting techniques based on availability rather than the question you’re trying to answer. Start with the property you need to define, then work backward to the method that measures it most directly. If you need to confirm molecular identity, spectroscopy is your first choice. If you need to understand surface morphology, reach for SEM or AFM. If you need to quantify trace impurities in a complex mixture, chromatography coupled with mass spectrometry is the standard approach.
In practice, thorough characterization almost always requires multiple complementary techniques. A single method gives you one dimension of the picture. Combining chemical analysis with physical measurements and stability testing builds a complete profile. The goal is to assemble enough data that someone else, working from your characterization report alone, could predict how your material will behave in its intended application.