aiwiki.page
English
Science / Chemistry / high-performance-liquid-chromatography

High-Performance Liquid Chromatography

High-performance liquid chromatography separates dissolved substances using a pressurized mobile phase and a selective stationary phase for identification, quantification, and purification.

30 keywords9 linked from8 not yet writtenWritten by AI
ChromatographyAnalytical Chemi…Nobel Prize in C…Mass Spectrometr…AdsorptionTemperatureDiffusionViscosityHigh-Perfo…

High-performance liquid chromatography (HPLC) is a form of chromatography in which a liquid mobile phase carries a sample through a column containing a stationary phase. Components travel through the column at different rates because they interact differently with the two phases. Pumps drive the liquid through a tightly packed bed or another porous separation medium, allowing efficient separations that would be impractically slow under gravity alone.

HPLC is a central technique in analytical chemistry, pharmaceutical testing, environmental analysis, food science, and biochemical research. It can separate many substances that are insufficiently volatile or too thermally unstable for gas chromatography, including drugs, metabolites, peptides, and large biological molecules. Depending on the detector and method, it provides quantitative measurements, evidence of chemical identity, or purified fractions for subsequent investigation.

The term describes an instrumental platform rather than a single separation mechanism. Reversed-phase, normal-phase, ion-exchange, size-exclusion, and affinity separations can all be performed with high-performance liquid-chromatographic equipment. Their effectiveness depends on the coordinated selection of column chemistry, mobile-phase composition, flow, temperature, sample preparation, and detection.

Historical development and scope

Liquid chromatography originated before the development of high-pressure instruments. In the early twentieth century, Mikhail Tswett separated plant pigments on adsorbent columns, establishing the principle that substances could be resolved by their different affinities for a solid phase and a moving liquid. Early column methods generally used relatively coarse particles and gravity-driven flow. They were useful for isolation but often required long analysis times and substantial solvent volumes.

A major conceptual advance came in 1941, when Archer Martin and Richard Synge described partition chromatography. Their work demonstrated separation through distribution between phases rather than exclusively through surface adsorption. Martin and Synge received the 1952 Nobel Prize in Chemistry for this development. Subsequent advances in chromatographic theory clarified how smaller particles, controlled flow, and reduced band spreading could increase separation efficiency.

Modern HPLC emerged during the 1960s as reliable pumps, sensitive detectors, and improved column packings became available. Josef Huber and other researchers investigated rapid liquid separations, while J. J. Kirkland developed influential superficially porous packing materials. Csaba Horváth and colleagues described high-pressure liquid-chromatographic separations using pellicular particles in 1967. Commercial instrumentation subsequently made tightly controlled liquid separations accessible to routine laboratories.

The abbreviation HPLC initially commonly expanded to “high-pressure liquid chromatography.” “High-performance liquid chromatography” became the prevailing name because pressure is an enabling condition rather than the fundamental basis of separation. A high-pressure system does not necessarily produce high resolution: column selectivity, efficiency, sample loading, and instrument dispersion remain decisive.

From the 1970s onward, bonded stationary phases and increasingly uniform micrometre-scale particles helped establish HPLC as a routine quantitative technique. Automated injection, programmable solvent gradients, computerized data processing, and improved detector electronics increased reproducibility and throughput. Later developments included robust coupling to mass spectrometry and commercial ultrahigh-pressure instruments using sub-2-micrometre particles.

The boundaries between conventional HPLC and newer implementations are not defined by one universally applicable pressure threshold. Conventional instruments often have maximum operating pressures around 400 bar, whereas many ultrahigh-performance systems operate at 1,000 bar or more. Actual operating pressure depends on column dimensions, particle size, solvent viscosity, temperature, and flow. The underlying retention mechanisms remain continuous across these platforms.

Retention, efficiency, and resolution

Chromatographic separation results from repeated exchange of analyte molecules between the mobile phase and the stationary phase. A component that spends a greater fraction of its passage associated with the stationary phase moves more slowly. This behavior can involve adsorption, partitioning, electrostatic interactions, molecular recognition, or restricted access to pores. Retention therefore reflects both analyte chemistry and the physical and chemical properties of the separation system.

For an isocratic separation, in which mobile-phase composition remains constant, retention is commonly expressed by

k=tR−tMtM,k=\frac{t_R-t_M}{t_M},

where tRt_R is the analyte retention time and tMt_M is the hold-up time of an effectively unretained substance. The retention factor kk is dimensionless. For two substances with k2>k1k_2>k_1, their selectivity is expressed as α=k2/k1\alpha=k_2/k_1. Changes in solvent composition, stationary-phase chemistry, or temperature can alter selectivity and sometimes reverse elution order.

Efficiency describes the extent to which an initially narrow sample band spreads during passage through the column. For an approximately Gaussian peak, the number of theoretical plates can be estimated from its baseline width wbw_b:

N=16(tRwb)2.N=16\left(\frac{t_R}{w_b}\right)^2.

The plate height is H=L/NH=L/N, where LL is column length. These quantities are operational measures, not counts of physical layers inside the column. Their interpretation becomes less straightforward for strongly asymmetric peaks, gradient separations, and measurements substantially affected by extra-column broadening.

The van Deemter equation provides a simplified description of how plate height varies with mobile-phase linear velocity uu:

H=A+Bu+Cu.H=A+\frac{B}{u}+Cu.

The terms represent multiple flow paths, longitudinal diffusion, and resistance to mass transfer. Liquid-phase diffusion is relatively slow, making mass-transfer effects especially important at high velocities. Smaller particles shorten diffusion distances and reduce some sources of dispersion, but they also increase hydraulic resistance.

For two adjacent peaks, resolution is commonly calculated as

Rs=2(tR,2−tR,1)wb,1+wb,2.R_s=\frac{2(t_{R,2}-t_{R,1})}{w_{b,1}+w_{b,2}}.

A value near 1.5 is often associated with baseline separation of approximately Gaussian peaks of similar size, but it does not guarantee accurate measurement when peaks differ greatly in concentration or shape. Under idealized conditions, resolution depends on efficiency, selectivity, and retention through

Rs≈N4(α−1α)(k21+k2).R_s\approx\frac{\sqrt{N}}{4} \left(\frac{\alpha-1}{\alpha}\right) \left(\frac{k_2}{1+k_2}\right).

This relationship explains why changing selectivity can be more effective than simply lengthening a column. It also assumes linear, approximately equilibrium-controlled behavior; overloaded or strongly tailing peaks require more cautious interpretation.

Instrument architecture and fluid handling

An HPLC system usually contains solvent reservoirs, a degassing unit, one or more pumps, a mixing arrangement, an injector or autosampler, a column compartment, a detector, and data-acquisition software. Liquid leaving the detector either enters a waste container or is collected in fractions. Although the column performs the principal separation, every component can influence retention reproducibility, peak shape, and quantitative response.

Reciprocating-piston pumps are widely used because they deliver controlled flow against substantial backpressure. Multiple pistons, check valves, and electronic compensation reduce flow pulsation. Solvent compressibility and changes in viscosity affect delivery, especially during gradients. In low-pressure gradient systems, solvents are proportioned before entering the pump; in high-pressure mixing systems, separately pumped streams combine downstream. The arrangements differ in mixing behavior, complexity, and gradient delay.

Dissolved gases are commonly removed using membrane degassers. Bubbles can interfere with pump operation or generate detector disturbances. Mobile phases also require appropriate cleanliness because particles may obstruct inlet filters, narrow tubing, or column frits. Solvent compatibility matters during changes between methods: mixing an aqueous salt solution directly with a strong organic solvent can precipitate salts inside the fluid path.

The injector introduces a defined sample volume without interrupting flow. Autosamplers commonly use fixed-loop, variable-loop, or flow-through-needle arrangements. Injection volumes must be compatible with column volume and analyte retention. A large plug of strong solvent can distort peaks by preventing analytes from concentrating near the column inlet. Needle washing and valve design influence carryover, particularly for strongly adsorbing substances or measurements spanning large concentration ranges.

Column ovens stabilize temperature, which affects solvent viscosity, analyte diffusion, and retention selectivity. Guard columns or replaceable inlet cartridges can retain contaminants before they reach the analytical bed. Their use adds volume and potential dispersion, so they must be compatible with the analytical column and separation requirements.

Two instrument volumes are particularly important. Dwell volume is the volume between the point where a gradient is formed and the column inlet; it determines when the programmed composition change reaches the column. Extra-column volume includes components outside the separation bed that broaden analyte bands. Large connecting tubes, poorly fitted connections, injector passages, and detector cells can erase much of the efficiency gained from a short, narrow-bore column. Consequently, transferring a method between instruments requires attention to fluid-path design, not merely matching the displayed flow and gradient program.

Columns, stationary phases, and pressure

Most analytical HPLC columns are tubes packed with porous particles, commonly based on silica. Representative conventional dimensions include lengths of 50–250 millimetres and internal diameters of 2.1 or 4.6 millimetres, although capillary, microbore, and preparative formats extend well beyond these ranges. Particle diameters commonly span approximately 1.7–5 micrometres in modern high-efficiency analytical separations. Column dimensions and particle properties jointly determine sample capacity, solvent consumption, efficiency, and pressure requirements.

Silica provides a rigid porous support with substantial internal surface area. Its surface silanol groups can be chemically modified to attach organic ligands. Octadecyl, or C18, phases are widely used, alongside C8, phenyl, cyano, and other chemistries. Bonding density, ligand architecture, endcapping, and residual silanol activity affect selectivity. Columns carrying the same broad designation, such as C18, are therefore not necessarily interchangeable.

Pore dimensions influence access to internal surface area. Small molecules can enter pores that may substantially exclude large proteins. Columns intended for proteins and other macromolecules therefore often use wider pores than columns optimized for small-molecule analysis. A larger pore does not automatically improve every separation: surface area, loading capacity, and mass-transfer behavior also change.

Fully porous particles contain an interconnected pore network throughout their volume. Core–shell particles, also called superficially porous particles, contain a solid core surrounded by a porous layer. Their shorter intraparticle diffusion paths and favorable packing characteristics can provide high efficiency without requiring the smallest particle diameters. Monolithic columns instead contain a continuous porous structure, allowing permeability and mass-transfer behavior to be engineered through interconnected flow channels and smaller pores.

For comparable packed beds, the pressure drop follows the approximate scaling

ΔP∝ηLudp2,\Delta P\propto\frac{\eta L u}{d_p^2},

where η\eta is mobile-phase viscosity and dpd_p is particle diameter. The proportionality also depends on bed porosity and particle structure. Halving particle diameter therefore increases pressure requirements approximately fourfold at the same linear velocity and length. Raising temperature often lowers viscosity, but it can also alter selectivity or accelerate degradation of sensitive analytes and stationary phases.

Column durability depends on chemical as well as mechanical conditions. Conventional silica-based bonded phases have restricted pH operating ranges, with limits varying by product and conditions. Polymeric and hybrid organic–inorganic supports extend the range available for some applications. Exposure to unsuitable solvents, precipitated salts, contaminated samples, or excessive loading can damage a column or change its surface chemistry even when the pressure remains within its rated limit.

Separation modes and their chemical selectivity

Reversed-phase chromatography uses a relatively nonpolar stationary phase and a more polar mobile phase, commonly a mixture of water with acetonitrile or methanol. Increasing the organic-solvent fraction usually reduces retention for neutral hydrophobic analytes. The mechanism involves differential solvation and interactions with the bonded phase, rather than a simple universal ranking by molecular polarity. Aromatic interactions, molecular shape, and residual surface groups can also influence selectivity.

For ionizable compounds, pH can change retention substantially by altering the fraction present in charged and neutral forms. Weak acids generally become more negatively charged as pH rises, whereas weak bases become less positively charged. Neutral forms are often more strongly retained in reversed-phase systems, although surface interactions and ion-pairing effects complicate this pattern. Buffer composition and concentration influence both analyte ionization and the chemical environment of the stationary phase.

Normal-phase chromatography employs a polar stationary phase, such as bare silica, with a comparatively nonpolar mobile phase. A more polar modifier usually increases elution strength by competing for adsorption sites. It is useful for separations sensitive to polar functional groups and certain structural differences. Water content can strongly affect surface activity, making solvent preparation and equilibration important.

Hydrophilic-interaction chromatography, or HILIC, also uses a polar stationary phase but typically operates with an acetonitrile-rich mobile phase containing water. Retention is often described partly by partitioning into a water-enriched interfacial layer, with adsorption and electrostatic interactions contributing according to the phase and analyte. Increasing the aqueous fraction generally increases elution strength. HILIC is useful for highly polar compounds that show insufficient retention on conventional reversed-phase columns.

Ion-exchange chromatography separates substances through interactions with fixed charged groups. Anion exchangers retain negatively charged species; cation exchangers retain positively charged species. Elution commonly uses increasing salt concentration or changing pH. Size-exclusion chromatography instead separates by access to pores: larger species explore less internal pore volume and usually elute earlier. Ideally, adsorption is minimized, so elution reflects hydrodynamic size rather than chemical affinity or molecular mass alone.

Affinity chromatography uses selective binding between an immobilized ligand and a target, as in antibody capture with protein A. Chiral chromatography employs a chiral environment to discriminate between enantiomers, often through several simultaneous noncovalent interactions. Mixed-mode phases combine mechanisms, such as reversed-phase retention and ion exchange. These examples demonstrate why “HPLC” cannot be treated as synonymous with reversed-phase separation, even though reversed-phase methods are especially widespread.

Detection and structural identification

An HPLC detector converts changes in column effluent into a recorded signal. Detector selection depends on analyte properties, required sensitivity, mobile-phase composition, and whether the sample must remain recoverable. No detector gives equal response to all compounds. A chromatogram is therefore a detector-specific representation of a separation, not a complete inventory of everything injected.

Ultraviolet and visible absorbance detectors are widely used for compounds containing suitable chromophores. Their response is governed approximately by the Beer–Lambert relationship,

A=εbc,A=\varepsilon bc,

where AA is absorbance, ε\varepsilon is molar absorptivity, bb is optical path length, and cc is concentration. Mobile-phase absorbance, detector noise, and analyte absorptivity influence sensitivity. Diode-array detectors record absorbance across multiple wavelengths, enabling spectral comparison within and between peaks. Such spectroscopic information can support identification, but similar spectra do not prove that two substances are identical.

Fluorescence detectors can provide high sensitivity and selectivity for naturally fluorescent compounds or analytes converted into fluorescent derivatives. Electrochemical detectors measure oxidation or reduction at an electrode and are useful for selected electroactive compounds. Their performance depends on electrode material, applied potential, mobile-phase composition, and surface condition.

Differential refractive-index detectors respond to differences in refractive index between the mobile phase and column effluent. They are useful for some carbohydrates and polymers lacking strong ultraviolet absorption, but their temperature sensitivity and response to mobile-phase composition make them poorly suited to ordinary solvent gradients. Evaporative light-scattering and charged-aerosol detectors extend detection to many nonvolatile analytes by removing volatile mobile phase and measuring the remaining particles. Their responses are often nonlinear and depend on operating conditions.

Liquid chromatography–mass spectrometry adds mass-selective detection. Electrospray ionization is particularly important for polar and ionic compounds, whereas atmospheric-pressure chemical ionization serves many less polar, thermally suitable analytes. In tandem mass spectrometry, selected precursor ions are fragmented and product ions measured, providing additional selectivity. Accurate-mass measurements and fragmentation patterns can constrain structural assignments but do not necessarily distinguish every isomer.

The interface imposes chemical constraints. Volatile additives such as formic acid or ammonium acetate are commonly favored, whereas nonvolatile phosphate buffers can contaminate the source and suppress signals. Coeluting matrix components can reduce or enhance ionization, causing a response change without a corresponding concentration change. Consequently, a clean-looking chromatogram or a selective mass transition does not alone establish freedom from analytical interference.

Method development and sample preparation

Method development begins by defining the analytical question: which substances must be measured, in what material, over what concentration range, and with what required discrimination. A purity test for a pharmaceutical ingredient differs from a trace contaminant measurement or a preparative isolation. The relevant constraints include analyte stability, available sample volume, detection chemistry, expected interferences, and the consequences of unresolved components.

Knowledge of molecular structure helps narrow the initial conditions. Ionizable groups, hydrophobicity, solubility, and approximate size guide selection of separation mode and pH. Reversed-phase screening often starts with a gradient broad enough to reveal the retention range of a mixture. Comparisons among methanol and acetonitrile, different bonded phases, and suitable pH conditions can uncover useful selectivity changes. Increasing organic content does not reproduce the effects of changing solvent identity, because solvents differ in their interactions with both analytes and the stationary phase.

In gradient elution, mobile-phase composition changes during the run. A common reversed-phase gradient progressively increases organic-solvent content, allowing early separation of less retained components while eluting strongly retained compounds within a practical time. Gradient duration, initial composition, flow, and column volume jointly influence separation. The column must subsequently return to its initial composition and re-equilibrate; insufficient equilibration produces drifting retention times and inconsistent selectivity.

Sample preparation removes material that could interfere with separation, detection, or instrument operation. Depending on the sample, it can involve dilution, centrifugation, protein precipitation, liquid–liquid extraction, or solid-phase extraction. Biological samples often require removal of proteins and reduction of matrix components, while environmental waters may require concentration of trace analytes. Each operation can introduce selective losses, contamination, or chemical transformation.

Filtration is not chemically neutral. Analytes may adsorb to membranes, and extractable substances from filters or containers can appear as chromatographic peaks. Sample diluent composition is also important: a solution that dissolves an analyte well may be too strong for the initial chromatographic conditions. Incompatible pH or solvent composition can produce precipitation or poor peak focusing after injection.

Development therefore considers separation and sample handling together. A method with excellent standards-only performance may fail in real samples because of extraction losses or matrix-dependent response. Deliberate variation of temperature, flow, buffer concentration, and composition reveals sensitivity to practical changes. Statistical experimental design can distinguish important interactions among variables, particularly when several critical peak pairs must be resolved simultaneously.

Quantification, validation, and interpretation

Quantitative HPLC usually relates integrated peak area, and sometimes peak height, to analyte amount or concentration. With external-standard calibration, standards of known concentration establish the response relationship and samples are measured under corresponding conditions. The calibration model must reflect the detector response: linearity cannot be assumed for all detectors, and a high correlation coefficient alone does not demonstrate that a model is suitable.

Internal-standard methods compare analyte response with that of a reference substance added in a controlled amount. An effective internal standard behaves similarly during preparation and measurement while remaining distinguishable from the analyte. In mass-spectrometric assays, a stable-isotope-labelled analogue can compensate for extraction variability and some matrix effects. Compensation is not automatic, however, especially if the labelled compound elutes differently or is added after losses have already occurred.

Peak identification and peak integration are separate problems. Retention-time agreement with a reference standard supports identification but is rarely definitive in a complex mixture. Co-injection, spectral comparison, alternative stationary phases, and mass-spectrometric evidence can strengthen an assignment. Integration establishes how signal is divided into peaks and baseline. Unresolved shoulders, drifting baselines, detector saturation, and inconsistent integration rules can introduce bias.

Validation examines whether a procedure is fit for its intended use. Common characteristics include selectivity, accuracy, precision, reportable range, detection capability, quantitation capability, and robustness. Repeatability concerns measurements under closely matched conditions, whereas intermediate precision includes changes such as different days, analysts, or instruments. Recovery experiments evaluate aspects of sample processing, but recovery alone does not establish trueness when the reference material or matrix is inadequately characterized.

Detection and quantitation limits are operational quantities that depend on the complete procedure. Signal-to-noise calculations and statistical estimates from calibration data are common approaches, but they require appropriate assumptions and confirmation. A concentration detectable in a clean standard may not be reliably measurable in a complex extract. Measurement uncertainty can include contributions from standard preparation, calibration, repeatability, extraction, sample heterogeneity, and matrix effects.

System-suitability tests check whether the instrument and column are functioning adequately for a particular method. Typical criteria include replicate-injection precision, retention behavior, peak asymmetry, and resolution of a critical pair. These tests do not substitute for validation or prove that every sample is interference-free. Likewise, chromatographic area percentage is not automatically a mass-purity percentage: different compounds may have different response factors, and some impurities may escape the selected detector entirely.

Applications, scale, and continuing development

Pharmaceutical analysis is a major application of HPLC. Methods measure active ingredients, related substances, degradation products, dissolution samples, and components of complex formulations. Stability-indicating procedures are designed to distinguish the substance of interest from relevant degradation products. For biological medicines, complementary modes provide different information: size exclusion can reveal aggregates, ion exchange can resolve charge variants, and reversed-phase peptide mapping can examine proteolytic fragments.

In proteomics, liquid chromatography commonly separates peptide mixtures before tandem mass spectrometry. Narrow-bore and nanoflow systems reduce solvent flow and can improve electrospray sensitivity under suitable conditions, although they impose demanding requirements for connections, sample cleanliness, and flow stability. Metabolite analysis often combines reversed-phase and HILIC methods because no single mode captures the full chemical diversity of cellular metabolism.

Environmental laboratories use liquid chromatography to investigate pesticides, pharmaceutical residues, and other contaminants in water, soil extracts, and biological material. Food analysis includes vitamins, preservatives, mycotoxins, sweeteners, and authenticity markers. Complex matrices make sample preparation and confirmation particularly important. A method that separates a target from a few standards may still encounter interferences in wastewater, plant extracts, or processed foods.

Preparative HPLC uses the same separation principles to recover material rather than merely measure it. Larger columns and higher flow rates increase throughput, while fraction collectors isolate selected portions of the effluent. Sample loading becomes a central design variable: high concentrations can produce nonlinear retention, broadened bands, and altered peak shapes. Process conditions therefore balance purity, recovery, productivity, and solvent consumption rather than maximizing analytical resolution alone.

Ultrahigh-performance liquid chromatography combines high-pressure-capable instruments with efficient columns and low-dispersion fluid paths. It can shorten analysis times, improve resolution, or reduce solvent use, depending on how a method is redesigned. Pressure itself does not create chemical selectivity, and very rapid separations may be limited by detector acquisition speed, injection-cycle time, or frictional heating inside the column.

Multidimensional liquid chromatography addresses mixtures too complex for one separation. Selected fractions, or the entire effluent in a comprehensive arrangement, undergo a second separation with different selectivity. The interface must manage sampling frequency, dilution, solvent compatibility, and the limited time available for the second dimension. Automated method screening and computational retention models also support development, but their usefulness depends on representative experimental data and reliable characterization of column behavior.

Resource consumption remains an important practical constraint. Smaller internal diameters, shorter methods, solvent recovery in suitable preparative settings, and alternative mobile phases can reduce waste, but changes must preserve adequate performance. Replacing a solvent alters viscosity, pressure, detector compatibility, and selectivity; transferring a method to a smaller column changes injection-volume and dispersion requirements. These physical and chemical trade-offs shape the design of modern HPLC systems.