In the early days of high performance liquid chromatography (HPLC), detection for qualitative or quantitative analyses often was carried out by collecting fractions and analyzing them off-line using gravimetric or wet chemical techniques. It wasn't until the 1940s and 1950s that the first online detectors for liquid chromatography, the refractive index (RI), and conductivity detectors appeared on the scene (3, 4), and while they were certainly an improvement over off-line approaches, neither detector was particularly sensitive. The search for more sensitive universal detectors for HPLC (much like flame ionization detection [FID] for gas chromatography [GC]) over the years led researchers to adapt GC detectors for use in HPLC (5–8), but the removal of the HPLC mobile phase through evaporation originally limited any real applicability. However, in the 1960s, the first ultraviolet (UV) detector for HPLC was introduced (9), and subsequent improvements in design led to better sensitivity (10) and improvements such as variable wavelength and diode array UV detectors. While a truly universal HPLC detector with the kind of sensitivity achieved in GC–FID is still elusive, many different types of detectors have been developed since the early UV, RI, and conductivity detectors that have been very successful for a wide variety of general or specific HPLC applications.
Detectors for HPLC are designed to take advantage of some physical or chemical attribute of either the solute or mobile phase in the chromatographic process in one of four ways (2):
- A bulk property or differential measurement
- Analyte specific properties
- Mobile phase modification
- Hyphenated techniques
Bulk property detectors are the most universal detectors for HPLC as they measure properties common to all analytes by measuring differences in the mobile phase with and without the sample. One of the most common bulk property detectors is the RI detector. Given the universal nature of bulk property detectors, they respond to all analytes, placing more emphasis on the selectivity of the chromatographic column. They are, however, inherently somewhat limited in sensitivity because they are the chromatographic equivalent of determining the weight of a sailor by weighing the battleship before and after the sailor departs for shore leave.
Analyte-specific property detectors respond to a characteristic that is unique to an analyte. The UV detector is the most common example of an analyte-specific property detector, responding to analytes that absorb UV light at a particular wavelength. UV detectors are usually thought of as somewhat specific, responding only to compounds with chromophores, but at low UV wavelengths (<210 nm), where just about every organic compounds absorb, UV detectors are actually somewhat universal. Other analyte specific detectors include fluorescence, conductivity, and electrochemical.
Table I: Desired detectorcharacteristics
Mobile phase modification detectors change the mobile phase postcolumn to induce a change in the properties of the analyte — for example, by creating particles suspended in a gas phase. Evaporative light scattering and corona discharge detectors fit into this category. Pre- or postcolumn derivatization of the analyte also is considered to fit into this category.
Table II: Example detector selection criteria
Hyphenated techniques refer to the coupling of a separate independent analytical technology to an HPLC system. The most common is mass spectrometry (LC–MS), and technologies such as infrared spectrometry (LC–IR) and nuclear magnetic resonance (LC–NMR) also have been used. While it can be argued that MS is seeing more and more use as a routine HPLC detector, good reviews of its use as well as other hyphenated techniques are available and can be consulted for more information (11–13).
Table III: Common HPLC detector properties
There are many characteristics to consider when choosing a detector, and Table I lists some of them. For additional details on many of the HPLC detectors addressed here, the reader is encouraged to consult the references in each individual section along with additional general reviews (14–16).
Table IV: Common HPLC detector attributes
Because no one detector has all of these characteristics, over time, a multitude of detectors have been designed, produced, and sold to answer one particular challenge or another. Ease of use, predictability, and reproducibility are all very important characteristics, however, recently, there has been an increased emphasis on the flow-cell contribution to band broadening and faster detector responses. This emphasis is due to new, low dispersion ultrahigh-pressure liquid chromatography (UHPLC) systems designed to take full advantage of sub-2-µm particle size column packings (17,18). A common rule of thumb says that for good peak integration, a minimum of 20 points should be collected across the peak. With the chromatographic efficiency of sub-2-µm particle size column packings often resulting in peaks that are less than 1-s wide, faster data rates are required to maintain good integration, sensitivity, and resolution. Narrower peaks also will require smaller volume detector flow cells to maintain peak concentration, and this lower dispersion shouldn't come at a loss of signal. Many detectors, including UV, photodiode-array (PDA), fluorescence, evaporative light scattering, and charged aerosol detectors are available commercially in both HPLC and UHPLC version depending upon the application. Some additional considerations to be made when selecting a detector are summarized in Table II. Most of the considerations in Table II are relatively straightforward; however, the use of complementary or orthogonal detectors used in combinations (series or parallel, depending upon back pressure limitations and whether or not they are destructive) should be noted as it is gaining in popularity, particularly in drug discovery or other screening-type applications. Table III lists some properties of the common HPLC detectors in use today, and a summary of some of their key attributes is presented in Table IV.