High-Performance Liquid Chromatography, or HPLC, is one of the most widely used techniques in analytical chemistry. Ever since Mikhail Tswett discovered that a mixture of pigments in solution could be separated by allowing the liquid to drain through a glass column packed with calcium carbonate, scientists have been finding ways to improve the efficiency of separations and apply the technique to a mind-boggling array of substances from simple ions to complex polymer mixtures. The principles used to separate mixtures by HPLC are the same as for all types of chromatography, but technology lends itself to this method in ways that allow analyses that could never be attempted using open columns, thin layer chromatography, flash chromatography or gas chromatography. HPLC is basically the same technique Tswett used, except the solvent is forced through the column using a pump, and electronic instruments are used to detect separated components rather than the human eye. HPLC can still be used for simple qualitative analysis like Tswett performed, but it can also be used for very accurate quantitative analysis, and can be used for purifying compounds (preparative, semi-prep chromatography). HPLC is especially useful for detecting and measuring impurities and degradates of pharmaceutical compounds.
A basic HPLC system has a pump, sample injector, column, and detector. The pump is the key to allowing HPLC to be "high-performance". In order to provide efficient separations, modern analytical chromatography columns are typically packed at 10,000 PSI with solid particles ten microns or smaller in diameter, and vary from about ten to thirty centimeters in length. Allowing the mobile phase to drain through the packing using nothing but the force of gravity as Tswett did simply won't work. A typical HPLC assay requires a pressure of about 1500 PSI to pump mobile phase at 1 milliliter per minute through a 25 centimeter column. This pressure is usually supplied by a dual-piston reciprocating pump, capable of supplying pressures up to 5000 PSI at flow rates from 0.1 to 5.0 milliliters per minute. To ensure accurate and reproducible results, the pump must also provide a consistent flow rate with a bare minimum of pressure fluctuation, and it must be built out of materials resistant to a wide variety of solvents and buffers. The pistons are typically made of sapphire, the piston seals are usually ultrahigh-molecular weight polyethylene or polytetrafluoroethylene, check valves often contain a small ruby ball, and most other parts including the pump head are stainless steel. Tubing connecting the various components in the flow path is usually stainless steel or PEEK (polyetheretherketone) and can have an internal diameter as small as 0.005".
A single pump with only one solvent reservoir is a isocratic pump. A gradient pump can reduce the time necessary to separate complex mixtures by changing the solvent composition during an analysis. For instance, a mixture containing both very polar and nonpolar compounds could be separated using a column packed with nonpolar particles and a mobile phase consisting of an aqueous buffer containing a small percentage of an organic solvent. The polar compound would travel through the column very quickly, but the nonpolar compound would elute very slowly. The nonpolar compound can be eluted more quickly by increasing the percentage of organic solvent in the mobile phase after the polar compound has passed through the column. This is accomplished in one of two ways. A low-pressure gradient pump uses a single pump head at a constant flow rate, but a proportioning valve upstream from the pump allows solvent to be drawn from multiple reservoirs. The relative proportions of solvents can be adjusted during an analysis as necessary. A high-pressure gradient pump accomplishes this task by using multiple pump heads with a mixer downstream. Each pump head supplies only one solvent, with proportioning accomplished by varying the flow rate the individual pump heads are operating at. High-pressure gradients usually provide more accurate and sensitive separations than low-pressure, but this advantage is not always significant enough to justify the higher cost of hardware and maintenance for multiple pump heads vs. the cost of a single pump for low-pressure separations.
Because of the high pressures involved, an HPLC system is a closed system, making the introduction of a sample into the solvent stream problematic. A sample injector typically consists of a rotary valve, a sample loop, and either a syringe or a metered pump. The valve has more than one flow path, but solvent is always passing through it from the pump to the column. The sample loop is a small piece of stainless steel tubing connected to the valve. The following is a feeble attempt at an ASCII diagram of a six-port injection valve. The valve connects three pairs of ports at a time, and because two ports are bridged by the sample loop, this results in two distinct flow paths. In the first position the flow paths are 1,6,3,2 where the sample is loaded and 4,5 where mobile phase passes through to the column. When the valve rotates, the sample loop becomes part of the path 4,3,6,5 such that mobile phase flows through the sample loop flushing the sample into the column. The 1,2 path is idle in this position.
Valve in load position:
| sample |
| v |
| 1 waste| <--sample loop
| / ^ |
---6 2 |
column<- 5 3___|
mobile phase from pump
Valve in inject position:
| sample |
| v |
| 1 waste| <--sample loop
| \ ^ |
---6 2 |
column<- 5 3___|
mobile phase from pump
A manual injector would have a septum on port one where a syringe would be used to inject the sample into the loop. An autosampler is a device which has a tray of vials containing the samples to be analyzed and is programmed to perform a sequence of sample injections over a period of time. In an autosampler, port one would be connected to a sampling needle and port two to an automated syringe or metered pump. The syringe or pump pulls the sample into the loop. The sequence of valve switches and some of the plumbing varies in autosamplers depending on how the syringe/pump is returned to its starting position. Autosamplers come in many different designs, but the principle is the same: each sample vial must be placed under the sampling needle at some point. Common designs include sample trays that rotate such that each vial position is eventually moved under the needle, or a robotic arm removes the vials from a stationary tray and places them under the needle. Some autosamplers have temperature-controlled sample trays so that temperature sensitive samples will not degrade while waiting to be injected.
Columns have come a long way from glass tubes packed with calcium carbonate. Most are made from stainless steel tubes, but plastics are used for some applications. The stationary phase varies widely, and is chosen depending on the types of substances to be separated. Most commonly an organic compound is chemically bonded to silanol groups on the surface of very fine silica particles. Common organic groups are hydrocarbon chains of various lengths, phenyl groups, and cyano groups. These column packings are named after the length of the hydrocarbon chains or the name of other organic groups attached to the silica. The most common packings are C8, C18, and phenyl. These packings are used for "reversed-phase" HPLC, meaning that the stationary phase is nonpolar and the mobile phase is polar. HPLC was originally performed using polar packings such as silica gel, and a nonpolar solvent such as hexane. This came to be known as "normal phase". Other types of column packings include ion-exchange resins and polymers/resins with carefully controlled pore sizes designed to seperate molecules based on their relative sizes.
After the sample has been injected and separated, a detector is needed to measure it. The detector is a device which measures either a bulk property (such as refractive index) of the liquid flowing through it, or a solute property (such as ultraviolet absorbance at a specific wavelength). The method of detection is usually optical or electrical in nature. A UV detector works by passing the the eluent from the column through a quartz flow cell with a source of ultraviolet light shining through it. As different compounds pass through the cell, UV light is absorbed. The detector measures the change in UV light passing through the cell and sends out a continuous analog signal that can either go to a chart recorder or a computerized data system. Chromatography data is typically analyzed as a graph of signal response vs. time, this is 2-dimensional chromatographic data. More complex analyses can be performed using a diode-array detector. This type of detector is similar to a UV detector, but can collect data from more than one wavelength at a time. This results in 3-dimensional data (elapsed time, signal response, wavelength). This type of detection can be useful for identification of compounds, since different substances will usually have distinct UV absorbance spectra. A more accurate method for identifying compounds is to attach a mass spectrometer to the HPLC system after the bulk/solute property detector.
There are many sub-categories of HPLC, some named by the mode of separation, some by the method of detection, some a combination of both. The following is a brief description of the more common types of HPLC:
Reversed-phase: Uses a nonpolar stationary phase (usually bonded silica) and a polar mobile phase to separate nonpolar to moderately polar compounds. UV detection is most common for this type of HPLC. Compounds that do not strongly absorb UV light can be chemically modified so that they do. Refractive index, fluorescence, and electrochemical detectors can also be used.
Normal-phase: Uses polar stationary phase and nonpolar mobile phase. Detection methods are the same as reversed-phase.
Ion-chromatography: Uses an ionized resin as stationary phase and an aqueous mobile phase modified with acid, base, and/or buffers to give pH optimized for the seperation and detection the ion or ions of interest. Analytes are measured using a conductivity detector. To increase the sensitivity of the detector, a suppressor is sometimes connected after the column to remove mobile phase ions.
Ion pair chromatography: This is a technique used to measure ionizable analytes. Instead of adjusting the mobile phase pH as in ion chromatography, an ionic compound with an alkyl chain (such as 1-octanesulfonate, sodium salt) is added to the mobile phase. This compound then forms an ion pair with the ionizable analyte. The separation is typically performed using reversed-phase conditions with UV detection.
Size exclusion/gel permeation chromatography: This type of chromatography is typically used in the analysis of polymers. The column packing is typically a stable polymer or gel with carefully controlled pore sizes, and a refractive index detector is typically used. Unlike other types of chromatography where different compounds are separated based on chemical or electrical properties, size exclusion is intended to separate different molecular weights of the same polymer or separate large biological molecules based on size. When characterizing polymers it is often necessary to connect several columns in series, each one designed to separate based on excluding a different range of sizes.
Preparative/semi-prep chromatography: Any application of HPLC used as a purifying step in chemical synthesis. Just about any type of HPLC can be used, but the columns are much larger and flow rates are higher. The HPLC system must be equipped with a fraction collector after the detector.
Affinity chromatography: Refer to this node.
Sources: This was noded entirely from memory (I'm not kidding) based on years of experience as an analytical chemist performing chromatographic separations. I've had a lot of training and done some reading along the way, but too much time has passed for me to remember any specific sources. Please /msg me if you have suggestions for better organizing this information, or if you notice anything that needs to be corrected.