Review State-of-the-art in liquid chromatography–mass ...€¦ · Impressive progress has been made in the technology and application of combined liquid chromatography–mass spectrometry

Journal of Chromatography A, 856 (1999) 179–197www.elsevier.com/ locate /chroma

Review

State-of-the-art in liquid chromatography–mass spectrometryW.M.A. Niessen

Hyphen MassSpec Consultancy, De Wetstraat 8, 2332 XT Leiden, The Netherlands

Abstract

Impressive progress has been made in the technology and application of combined liquid chromatography–massspectrometry (LC–MS) in the past decennium. From a technique, that could only be used by a specialist, it has developedinto a routinely applicable technique. LC–MS has become the method-of-choice of analytical support in many stages of drugdevelopment within pharmaceutical industries and has found its way into environmental, biochemical and other laboratories.This paper provides a perspective on the current technology, principles and applications of LC–MS. 1999 ElsevierScience B.V. All rights reserved.

Keywords: Reviews; Mass spectrometry; Liquid chromatography–mass spectrometry; Ionization mechanisms; Atmosphericpressure ionization methods

Contents

1. Introduction ............................................................................................................................................................................ 1792. General view on API instrumentation ....................................................................................................................................... 1813. Recent advances in interface technology ................................................................................................................................... 1824. Ionization mechanisms ............................................................................................................................................................ 184

4.1. Liquid-phase processes in analyte ionization ..................................................................................................................... 1864.2. Gas-phase process in analyte ionization ............................................................................................................................ 187

5. Advances in MS instrumentation .............................................................................................................................................. 1886. LC–MS applications in perspective .......................................................................................................................................... 189

6.1. Pharmaceutical applications ............................................................................................................................................. 1896.2. Quantitative bioanalysis in pharmaceutical applications ..................................................................................................... 1916.3. Environmental applications.............................................................................................................................................. 1916.4. Biochemical and miscellaneous applications ..................................................................................................................... 192

7. Discussion and conclusions ...................................................................................................................................................... 193References .................................................................................................................................................................................. 194

1. Introduction development and its acceptance and high spread,especially considering its price tag, is astonishing.

Combined liquid chromatography–mass spec- LC–MS has become the method-of-choice for ana-trometry (LC–MS, [1]) can be considered as being lytical support in many stages of drug developmentone of the most important techniques of the last within the pharmaceutical industry. In addition, itdecade of twentieth century. The extreme speed of plays an important role in environmental analysis,

0021-9673/99/$ – see front matter 1999 Elsevier Science B.V. All rights reserved.PI I : S0021-9673( 99 )00480-X

180 W.M.A. Niessen / J. Chromatogr. A 856 (1999) 179 –197

while LC–MS instrumentation is also heavily used in a special emphasis on the development of interfacesbiochemical and biotechnological applications, as for atmospheric-pressure ionization (API) [3]. Thewell as in many other fields of application. current paper provides an overview of current tech-

In their recent book on LC–MS, Willoughby and nology with an emphasis to commercially availableco-workers [2] analyzed the development of LC–MS technology. At present, most of the instrumentalas well as its acceptance as an analytical technique. developments are taking place within the instrumen-As pictured in Fig. 1, at present we are still in the tal manufacturers laboratories and workshops, al-early stage of acceptance and use of LC–MS in though to a high extent such developments arereal-world applications. While new developments in steered and initiated by the demands from users,instrumentation are slowing down, significant pro- especially the pharmaceutical industries. Their ef-gress is made in the implementation and application forts and research and economical interests in theof LC–MS in many different fields. This stage puts implementation of LC–MS in the various stage ofdifferent demands on the development of LC–MS as drug development demands more advanced instru-an analytical technique, as is for instance indicated mentation as well as software support.by the currently growing interest in software de- Next to reviewing the state-of-the-art in instru-velopment, especially for more efficient data pro- mentation, the current status of LC–MS as ancessing after LC–MS analysis. analytical tool is evaluated. What are the possibilities

LC–MS interfacing and its applications have and limitations of current instrumentation and tech-frequently been reviewed. Recently, the advances in nology? What are topics which still require consider-the instrumentation in LC–MS were reviewed, with able research interest and/or development? What to

Fig. 1. Model for the distribution of people applying LC–MS (based on ref. [3]).

W.M.A. Niessen / J. Chromatogr. A 856 (1999) 179 –197 181

expect from LC–MS in the first decade of the next [14]). In addition, ultrasonically assisted electrospraycentury? In addition, attention is paid to some of the nebulization has also been described [15]. From themost important applications of LC–MS, perhaps in aerosol, gas-phase ions are generated by mechanismsorder to allow the reader to answer questions like: Is discussed in more detail below. These ions, togetherthe current huge interest in LC–MS a transitory with solvent vapor and nitrogen bath gas, are sam-hype? Why so many laboratories are fascinated by pled by an ion sampling device into a first pumpingthe potential of LC–MS in their developing applica- stage. The mixture of gas, solvent vapour and ions istions? supersonically expanded into this low-pressure re-

gion (10–100 Pa). The core of the expansion,containing the ions and other (neutral) material of

2. General view on API instrumentation higher molecular mass, is sampled by a skimmer intoa second pumping stage (pressure 0.1–1 Pa), con-

An API interface /source consists of five parts (cf. taining an ion focussing and transfer device toFig. 2): (1) the liquid introduction device or spray optimally transport and focus the ions in a suitable

23probe, (2) the actual atmospheric-pressure ion source manner to the mass analyzer region (pressure ,10region, where the ions are generated by means of Pa). In most systems, the ion transfer device consistselectrospray ionization (ESI), atmospheric-pressure of an RF-only quadrupole, hexapole or octapole.chemical ionization (APCI), or by other means, (3) From the vacuum point-of-view, it is not importantan ion sampling aperture, (4) an atmospheric-pres- whether a high flow-rate or a low flow-rate of liquidsure to vacuum interface, and (5) an ion optical is nebulized, because the sampling orifice actuallysystem, where the ions are subsequently transported acts as a fixed restriction between the atmospheric-into the mass analyzer. API interfaces are available pressure region and the first pumping stage. From thefrom all major mass spectrometer manufacturers [4– ionization point-of-view, it is also not important how13]. the ions are generated, i.e., by ESI or APCI, although

The operational principle of an API interface and (slightly) different tuning of voltages in the ionion source for LC–MS is as follows. The column optics might be needed due to some differences ineffluent from the LC is nebulized into an atmos- the ion kinetic energies. In addition, ESI-generatedpheric-pressure ion source region. Nebulization is ions generally contain less internal energy than theeither performed pneumatically, i.e., in heated-nebul- ions generated via APCI.izer APCI, by means of the action of a strong Two major mechanisms contribute to the ioniza-electrical field, i.e., in ESI, or by a combination of tion of the analytes: gas-phase ion–molecule re-both, i.e., in pneumatically assisted ESI (‘ionspray’ actions and a process generally called ion evapora-

Fig. 2. General scheme of an atmospheric-pressure interface and ion source.


tion. Ionization mechanisms are dealt with in more more-or-less distinct approaches. In most LC–MSdetail below. applications, one aims at introducing the highest

possible flow-rate to the interface. While early ESIinterfaces showed best performance at 5–10 ml /min,

3. Recent advances in interface technology present pneumatically-assisted ESI interfaces areoptimized for flow-rates between 50 and 200 ml /

The historical development of API interfaces for min. Although instrument manufacturers indicateLC–MS was reviewed in detail elsewhere [3]. A that their ESI interfaces can be applied with flow-variety of API interfaces is now commercially rates up to 1 ml /min (or even higher), such a highavailable from a number of instrument manufactur- flow-rate is hardly used in practice and also noters. The interfaces are not discussed in detail here. recommended. In practice, the response is not ad-Important design characteristics of the most widely versely influenced by using a solvent splitter betweenapplied interfaces are summarized in Table 1. Other column outlet and ESI probe to reduce the flow-ratemanufacturers apply slightly different concepts, i.e., of 1 ml /min from the column to ca. 100 ml /min feddifferent combinations of the devices described in into the interface probe.the table. All API interfaces operate in the way It is generally believed that under API conditions,described above. However, in evaluating the avail- the mass spectrometer acts as a concentration-sensi-able interfaces, a number of interesting trends can be tive detector, i.e., the response of the detector shouldobserved. be independent of the flow-rate [16–19]. It is in fact

First, ESI interfacing has developed into two difficult to prove experimentally, whether this is

Table 1Design characteristics of the most widely applied LC–MS interfaces

Interface type Design characteristics Ref.

Initial design On-axis ESI probe with three concentric tubes:sample, sheath and gas. Countercurrent drying gas.Ion sampling via glass capillary with metallized ends.Lens stack.

Heated capillary On-axis pneumatically-assisted ESI probe with three [12]concentric tubes: sample, sheath and gas.No drying gas.Ion sampling via heated stainless-steel capillary.RF-only quadrupole or octapole.

Turboionspray Pneumatically-assisted ESI probe at 458 and heated gas at 908. [10]Countercurrent curtain gas.Ion sampling via orifice.High-pressure RF-only quadrupole.

Z-spray Orthogonal pneumatically-assisted ESI probe. [9,13]Concurrent desolvation gas.Ion sampling via conical orifice and orthogonal skimmer.RF-only hexapole.

Orthogonal Orthogonal pneumatically-assisted ESI probe. [4–6]Countercurrent drying gas.Ion sampling via glass capillary with metallized ends.RF-only octapole.

aQa Orthogonal pneumatically-assisted ESI probe. [12]No drying gas.Ion sampling via conical orifice (washed with solvent)and orthogonal skimmer. Flow focussing.RF-only hexapole.


actually true. A proper experimental design for such nano-ESI devices are applied in combination withan experiment is difficult, because it is difficult to capillary electrophoresis, nano-LC, microcapillaryachieve a constant mass-flow in the analyte peak LC, and capillary electrochromatography. Typicalwhile changing the solvent flow-rate. The behaviour flow-rates are in the range of 30–1000 nl /min. Thereof the detection system as a concentration-sensitive is considerable development in the design and op-device can be considered to result from the splitting timization of dynamic nano-ESI devices [25–28].at the sampling orifice, where only a fixed part of the Nano-ESI interfaces and supplies are commerciallysource volume can be sampled (cf. Fig. 3). In available from major instrument manufacturers [4–addition, the concentration of ions in the spray plume 13], but also from two specialized companiesgenerated in the electrospray nebulization is limited [29,30].by space charging [20]. Attempts to improve the ion In the high flow-rate ESI interfacing, a clear trendsampling by electrostatic means are bound not to is the application of heat in the source, in order tosucceed as they adversely influence the potential stimulate the evaporation of the mostly aqueousrequired to produce an efficient electrospray nebuli- droplets. Heated nitrogen gas is applied to enhancezation. droplet desolvation, where the gas is introduced

In sample limited cases, nano-ESI interfaces are either in countercurrent flow, in concurrent flow (inapplied which can efficiently be operated at sub-ml / Z-spray interfaces [9,13]), or perpendicular to themin flow-rates [21,22]. Two types of nano-ESI spray (in turboionspray interfaces [10]). Relativelyinterfaces should be distinguished: ‘static’ and ‘dy- high gas temperatures must be applied to achievenamic’ interfaces. Static nano-ESI devices are espe- sufficient heat transfer to the evaporating droplets.cially applied in the field of protein characterization Another trend in interface development is related[21,22]: a narrowbore nano-ESI needle is filled with to the need to avoid source contamination during thethe protein solution and positioned on a probe. The routine analysis of large series of biological samples.probe is positioned in the API source and the protein These generally contain significant amounts of non-solution is sprayed at liquid flow-rate in the range of volatile material (salts, proteins, etc.), which will10–100 nl /min. In this way, about one hour of mass contaminate the ion source and especially the ion-spectrometric experiments can be performed with sampling cone. Off-axis electrospray nebulizationonly about 1 ml of sample. The optimum design of instead of on-axis with the sampling orifice appearsthe nano-ESI needle is studied [23,24]. Dynamic to be an important progress in that respect. Ionspray

Fig. 3. Splitting at ion-sampling orifice: the sampling orifices acts as a splitter which enables the introduction of only a limited part of theions generated in atmospheric-pressure ionization.


Fig. 4. Schematic diagram of a turboionspray source (based on information from the manufacturer [10]).

and turboionspray interfaces [10] are generally used skimmer. No tuning of voltages of ion sourceoff-axis, positioned at 30–458 relative to the axis (see devices is needed in the aQa source.Fig. 4). In addition, the use of a curtain gas flowing Obviously, all manufacturers continuously pursuearound the sampling cone greatly reduces the con- improved performance of their interfaces by optimiz-tamination of the cone and orifice. ing the design, e.g., with respect to the shape of the

In most interface designs (c.f. Table 1), an ortho- ion focusing parts or with respect to pumping at thegonal spray probe is applied [4–6,9,12,13]. An ion source.example of an orthogonal ESI source is shown inFig. 5. Some manufacturers use orthogonal sprayprobes in combination with additional devices toavoid ion source contamination, i.e., in the Z-spray 4. Ionization mechanisms[9,13] and in the aQa source [12]. In the Z-spraysource (see Fig. 6, [9]), the electrospray nebulization In the discussions on the ionization mechanismsis performed orthogonally to the sampling cone. Ions active in atmospheric-pressure ionization (API), gen-are extracted orthogonally from the spray into the erally a clear distinction is made between ESI andsampling cone, while large droplets and nonvolatile APCI. The ionization in an ESI interface is consid-material are collected onto a baffle plate. Sub- ered primarily a liquid-phase ionization technique:sequently, the ions are extracted orthogonally from preformed ions in solution are desorbed or evapo-the expansion behind the sampling cone into the high rated to the gas phase and can subsequently be massvacuum of the mass spectrometer. The aQa source analyzed. The ionization in APCI is considered to be(see Fig. 7, [12]) is described as a self-cleaning primarily based on gas-phase ion–molecule reactionssource: a constant low flow-rate of solvent is de- between analyte molecules and a solvent-basedlivered to the edge of the ion sampling orifice to reagent gas, generated by a series of ion–moleculeprevent the build-up of nonvolatile material. The LC reactions initiated by electrons from the coronaeffluent is sprayed orthogonally to the sampling discharge needle. Whether this distinction is actuallycone. After passing through the entrance cone, the valid, may be questioned, especially in the analysisgas flow is disrupted and ions are send to the of small compounds.


Fig. 5. Schematic diagram of an orthogonal electrospray source (based on information from the manufacturer [5,6]).

Fig. 6. Schematic diagram of a Z-spray ion source (based on information from the manufacturer [9,13]).


Fig. 7. Schematic diagram of an aQa ion source (based on information from the manufacturer [12]).

4.1. Liquid-phase processes in analyte ionization serving desolvation of the droplets. This ESI mecha-nism is currently known as the charge-residue model.

Nebulization of the column effluent is a common Subsequently, both in the discussion on the mech-feature of the current API interfaces. It was demon- anism of the thermospray buffer ionization by Vestalstrated by Dedieu et al. [31] in 1982, using a direct et al. [34–36] and in the discussion on the ESIliquid introduction interface, that the nebulization phenomena observed in the early experiments ofplays a major role in the soft transfer of analytes Yamashita and Fenn [37–39], the ion-evaporationfrom the liquid phase to the gas phase prior to concept introduced by Iribarne and Thomson [40–chemical ionization. Highly labile molecules like the 43] was considered to be the most adequate model.vitamin B could be transferred from the liquid Both the ion-evaporation and the charge residue12

phase to the gas phase in this way. This process, model [44,45] start from the nebulization of theindicated as soft desolvation, is active in all inter- liquid stream into small droplets. Due to statisticalfaces based on liquid nebulization, and is certainly random sampling of the positive and negative (buf-important in heated-nebulizer APCI systems. In fact, fer) ions present in solution [46], positively andthe underlying concept resembles the ionization negatively charged droplets are generated. In thermo-mechanism proposed by Dole et al. [32,33] for ESI: spray and heated nebulizer interfaces, both positivelyby spraying a solution of preformed protein ions, and negatively charged droplets are generated. Insmall droplets will be generated, from which gas- electrospray nebulization, droplets of only one po-phase protein ions will emerge due to charge-pre- larity are generated, because the droplets of opposite


polarity are discharged prior to their formation at the droplets will actually contain only one preformedneedle surface, as a result of electrochemical nature analyte molecule, which is transferred to the gasof the electrospray process [44,47]. phase by subsequent solvent evaporation from the

The charged droplets enter into a series of pro- droplet. Whatever mechanism is the correct one –cesses. Charge-preserving solvent evaporation results and it has even been argued that different mecha-in a decreasing droplet size, which is accompanied nisms are valid for small and large molecules [49] –,by an increasing electric field at the droplet surface. both mechanisms indicate the importance of pre-At a certain stage, the coulomb repulsion between formed ions in solution, e.g., by (de)protonation inthe surface charges exceeds the cohesive forces due the liquid phase. The pH of the solution is thusto surface tension and the droplet will explode. This considered an important parameter. However, the pHCoulomb explosion was photographed using a in the liquid will change during the droplet evapora-shadowgraph technique by Gomez and Tang [48]. In tion as a result of the differences in volatility of thethis way, it was demonstrated that the droplet fission compounds applied to set the pH, as discussed for

ˇtake place from a surface deformation; a droplet-jet instance by Appfel et al. [50] and Gatlin and Turecekis formed at the surface, as schematically drawn in [51].Fig. 8. Microdroplets carrying a relatively highnumber of charges are emitted. The processes of 4.2. Gas-phase process in analyte ionizationcharge-preserving solvent evaporation and elec-trohydrodynamic droplet disintegration may proceed Not only analyte ions will ion evaporate from theuntil sufficiently small droplets are produced for the droplets, but the buffer ions will do so as well.final step. Therefore, next to the protonated analyte ions, buffer

It is in this final step where the ion-evaporation ions like the ammonium ion will be present in themodel and the charge-residue model differ. Iribarne gas phase. In the discussions of the ESI mechanism,and Thomson [32–35] argued that ions may be like previously in the discussion of the thermospraydirectly emitted (‘evaporated’) into the gas phase ionization mechanism [52], the importance of gas-when the droplets are sufficiently small and charged. phase ion–molecule reactions is often neglected orThe charge-residue model assumes that these small ignored. Protonated analyte molecules may also be

formed by gas-phase reactions between the proton-ated ammonia from the buffer and a neutral analyteion, transferred to the gas phase by the soft desolva-tion process described above. In addition, the ionsgenerated in the gas phase undergo series of colli-sions to neutral solvent and nitrogen molecules in theatmospheric-pressure ion source. In combinationwith acceleration of the ions, collisionally induceddissociation (CID) of the ions may occur. But ingeneral, the multiple collisions may alter the com-position of the ionic distribution initially generatedby the ESI process. The potential difference betweennozzle and skimmer may, for instance, alter thecharge state distribution of multiple-charged proteinions [53,54].

The gas-phase ion–molecule reactions describedabove are conventional chemical ionization (CI)reactions [55]. CI is the prime ionization mechanismin APCI. The process is initiated by electrons fromthe corona discharge [56]. The electrons ionize theFig. 8. Representation of the Coulomb explosion, leading to the

production of small and highly-charged microdroplets. nitrogen in the source and the nitrogen molecular ion


reacts with the solvent molecules. Protonated solvent analyte response as a result of competitive effects inspecies are formed. The most abundant species in the ion evaporation. In real matrices, often a significantreagent gas is due to the solvent constituent with the reduction of the analyte response is observed withouthighest proton affinity (in positive-ion mode) or the a clear indication on what compound is responsiblelowest gas-phase acidity (in negative-ion mode). for the ion suppression. In general, no response isWhen the proton affinity of the analyte exceeds that observed from the competing species. This type ofof the reagent gas, the protonated analyte will be matrix effects often requires considerable attention ingenerated. Similarly, the deprotonated analyte will be the development of quantitative bioanalysis. A nicegenerated when the gas-phase acidity of the reagent and well documented example is the development ofgas exceeds that of the analyte. In addition, adduct a bioanalytical method for the drug finasteride, asions are generated when the proton affinities or described by Matuszewski et al. [60]. On-columngas-phase acidities of analyte and reagent gas are injections of standard solutions indicate a five-foldsimilar. better signal-to-noise ratio in ESI than in APCI.

These gas-phase ion–molecule reactions are effec- However, due to a severe and not-reproducibletive in the gas–vapor mixture generated in the ESI matrix effect in ESI, the R.S.D. of the peak area ofion source as well. Reactions between ion-evapo- the internal standard is unacceptably high, both forrated buffer ions, e.g., protonated ammonia, and repetitive injections of one particular plasma sampleanalytes result in protonated analyte molecules. From and between plasma samples from different patients.the mass spectrum, it is impossible to determine In order to avoid these matrix effects, the methodwhether the protonated molecule was formed in the was developed for APCI.liquid phase or in the gas phase. For many smallmolecules, the ESI response can be predicted fromCI rules. 5. Advances in MS instrumentation

Similarly, ion-evaporation-like processes may beimportant in heated nebulizer APCI, i.e., for some Almost equally important to optimization of APIcompounds good response can be achieved in the interfaces is the progress in MS instrumentation. Thecorona discharge-off mode in APCI. In method computer-controlled operation of modern instrumen-development for quantitative analysis, it is worth- tation helps in achieving optimum performance. Inwhile to investigate the response in APCI discharge- addition, new developments in the instrumentationoff mode as well. help to improve the overall performance of LC–MS.

In conclusion, both liquid-phase and gas-phase The majority of LC–MS applications are still runionization processes should be taken into account in on single and triple quadrupole instruments. In thedescribing the ionization in both ESI and APCI. The past years, the performance of triple quadrupoleactual ion yield is the resultant of mixed mecha- (QqQ) instruments for MS–MS is improved bynisms, especially in the analysis of small molecules. replacing the RF-only quadrupole collision cell byIn the analysis of large molecules like proteins, the RF-only hexapole [9] and octapole [12] collisionmajor effect of gas-phase reaction is often gas-phase cells. The latter provide an improved product ionreneutralization. Gas-phase reactions of electro- collection and transmission. This is also pursuedsprayed protein ions were studied by Ogorzalek-Loo with the modified quadrupole collision cell describedet al. [57,58]. by Mansoori et al. [61] and the linear accelerating

Another important issue, especially in relation to high-pressure collision cell (LINAC, [10]). TheESI, is the influence of the other solvent constituents LINAC also allows shorter dwell times in selectiveand the matrix. In the Coulomb explosion, pref- reaction monitoring (SRM), thereby allowing toerentially the ions at the droplet surface are trans- speed up the analysis or to monitor more SRMferred to the microdroplets. This explains why the transitions during one (fast) chromatographic run.analyte response may be greatly reduced by the API on an ion-trap mass spectrometers appears topresence of surfactants in the mobile phase [59]. be a very successful combination, which is availableHigh concentrations of other ionic species reduce the from two manufacturers [5,12]. While in low-level


quantitative analysis SRM on a QqQ instrument has MS–MS prior to TOF mass analysis is explored asto be preferred [62,63], the API ion-trap instrument well [71].especially proves its power in qualitative analysis, Finally, multiply-charging of proteins by ESI haswhere the multiple stages of MS–MS can be applied also stimulated the use of Fourier-transform ion-to achieve structure elucidation of the unknown. cyclotron resonance mass spectrometers (FT–ICR–Data-dependent acquisition further enhances the MS) [72,73]. Again, the enhanced resolution is anperformance in structure elucidation of minor com- important feature. The high-resolution operation inponents in a mixture [64,65]. combination with various dissociation techniques,

Orthogonal-acceleration reflectron time-of-flight such as CID, sustained off-resonance irradiation,(oa-TOF) instruments, available from several manu- infrared multiphoton dissociation, and surface in-facturers [4,9,10], combine the ability to perform duced dissociation, enables the use of ESI FT–ICR–accurate mass determination with an excellent full- MS for advanced structure elucidation of proteins. Inscan sensitivity. Mass accuracies of better than 20 addition, reaction chemistry and gas-phase confirma-ppm without and better than 5 ppm with an internal tion studies can be performed [72,73].lock mass can be achieved routinely. For compounds Among these new and exciting developments, thewith an M below 1000 Da, this accuracy provides an role of ESI on magnetic sector instruments hasr

excellent confirmation of identity based on calculated diminished. The main reason for the use magneticelemental compositions. An oa-TOF instrument is an sector instrument appears to be the use of an array-integrating rather than a scanning system. In practice, type of detector to enhance sensitivity. However, forthe ‘all-ion-detection’ capability of the oa-TOF many applications a single TOF instrument is asystem provides a 20–100-fold improvement in viable alternative, being easier to operate and with asensitivity, compared to a scanning QqQ system. better price /performance characteristic. The need forObviously, the QqQ instrument operated in SRM high-energy CID in peptide sequencing appears to bemode will provide better sensitivity, but only at the overcome to a large extent by the use of proteinexpense of the information content. databases [74,75].

Given these features of an oa-TOF analyzer, alogical combination is the hybrid of a quadrupolefront-end and an oa-TOF back-end for MS–MS. This 6. LC–MS applications in perspectiveso-called Q–TOF, with unsurpassed performance interms of specificity and sensitivity, was first intro- At present, the three major application areas ofduced and built by one manufacturer [9,66,67], while LC–MS technology are in the pharmaceutical, en-recently a similar Qq–TOF instrument was intro- vironmental and biochemical fields. The applicationsduced by another [10,68]. The most interesting in the field of drug development and testing can befeature of the Q–TOF hybrid is its ability to perform considered as an important driving force in theaccurate mass determination at excellent sensitivity current development of LC–MS technology.after conventional low-energy CID in a hexapolecollision cell. This greatly facilitates identification of 6.1. Pharmaceutical applicationsunknowns, not only in the field of protein chemistry,where the instrument was originally built for, but LC–MS technology is applied in virtually everyalso in studies related to impurities, degradation stage of drug development. The three main issues inproducts and metabolites of drugs. this field are sensitivity, selectivity and speed. Com-

Another powerful hybrid instrument is the ion-trap pared to UV and UV-photodiode array (PDA) de-storage – reflectron TOF instrument (IT–reTOF), as tection, the LC–MS combination provides enhanceddeveloped by the group of Lubman [69,70]. Initially, confirmation of identity and often enhanced selectivi-the ion-trap part of the instrument was primarily ty. In addition, the ease-of-operation and theapplied to achieve a pulsed ion introduction to the achieved level of automation make LC–MS anreTOF with an improved duty cycle. Currently, the attractive tool in drug development.potential of the ion trap in (multiple stages of) The role of LC–MS already starts in the drug


discovery stage, irrespective whether this is per- reduced data can be viewed by means of theformed by conventional ‘intelligent’ synthesis or by browser: in a graphical representation of the 96-wellcombinatorial chemistry. Three powerful software plate the confirmed samples are colored green, whiletools have been developed for LC–MS to support in the others are colored red. More advanced colordrug discovery. These software tools are available schemes have been described as well [84]. In somefrom all major instrument manufacturers. applications, the screening for biological activity and

The first and most generally applicable tool is the mass spectral characterization is combined intoOpen Access [76,77]. This tool transforms the API- one system [85], e.g., by the application of on-lineMS instrument into a walk-up ‘black-box’ for syn- bioaffinity columns [86] or integrated biodetectionthetic chemist in need for a rapid confirmation of the systems based on antigen–antibody or ligand–re-good progress of their synthesis by molecular mass ceptor interaction [87]. The latter systems are com-determination of their product. A remote computer mercially available as tailor-made solutions to aserves as a log-in to the system. After entering the particular application [88].sample identification code and the type of LC–MS The third tool combines the rapid screening ofexperiments to be performed, the computer indicates combinatorial libraries or of series of extracts fromthe position in the autosampler rack to be used. The natural products to preparative scale purification ofsample is run automatically, e.g., in both positive-ion biologically active compounds. The fractionation isand negative-ion mode and at both a high and a low controlled by the response of the compound ofin-source CID potential. The resulting spectra are interest in the LC–MS. Initially, these systems wereplaced onto the LIMS network or sent to the chemist developed using conventional LC columns, but pre-by electronic mail. Recently, open-access LC–MS parative-scale LC columns are applied now as wellsystems are extended by the implementation of on- [89–91].line LC–NMR as well [78]. While the structure confirmation in open-access

To a growing extent, drug discovery is based on and combinatorial-chemistry strategies is primarilysynthesis by combinatorial chemistry procedures. based on molecular mass of the intact compound,The combinatorial libraries need to be screened for more elaborate structure elucidation is required inbiological activity, while a rapid characterization of subsequent stages of drug development, i.e., theidentity is also required. LC–MS technology is stages related to impurity screening, identification offrequently applied for the latter step [79]. The drug metabolites, and the search for degradationanalytical system consists of an x–y autosampler, products in drug substances and drug formulations.enabling sample introduction from a 96-well plate, Screening strategies based on precursor-ion andconnected to a column-bypass API-MS [80] or fast neutral-loss scans in MS–MS are frequently appliedLC–MS system [81]. Single quadrupole systems are to search for structurally-related compounds. Basedoften used for this purpose, although the use of on the product-ion mass spectrum of the parentoa-TOF instruments is of growing importance; the compound and educated guesses of possible prod-accurate mass determination allows a better con- ucts, selective neutral-losses and/or possible com-firmation of identity. The software allows the rapid mon product-ions are selected. An example of theanalysis of large series of samples at a high sample use of these complementary strategies in the impuritythroughput, i.e., up to 60 samples per hour in profiling of butorphanol tartrate is described by Volkcolumn-bypass mode [82]. On-line UV–PDA as well et al. [92]. Similar MS–MS strategies can be appliedas other LC detectors may be used to establish the in drug metabolism studies as well. Common neutralcompound purity [83]. The data-processing software losses in the identification of Phase-II metabolites areprovides a user interface, often called a data browser, the losses of 80 and 176 Da, for aryl-O-sulphate andwhich allows rapid answers on whether or not the O-glucuronic acid conjugates, respectively. An ex-expected products are present. In reaching this ample is described by Brownshill et al. [93]. Thedecision, data from both positive-ion and negative- advantages of accurate mass determination in theseion acquisition are used, taking into account the studies is also clearly recognized. FT–ICR–MS [94]various adduct ions that might be generated. The and the Q–TOF hybrid [95] are applied for this


purpose. Another interesting approach, circumvent- and/or the chromatography. The rationale of this ising the need for on-line LC separation, is the use of an increase in the sample throughput. However, thisnano-ESI: a small volume of the sample solution is often puts serious demands on the sample pretreat-introduced via nano-ESI. The mass spectrum is ment methods. As an example, Knebel et al. [101]searched for minor impurities and product-ion MS– described the fast bioanalysis of saquinavir, a selec-MS spectra are subsequently generated. The applica- tive inhibitor of HIV proteinases, using an off-linetions of multiply MS–MS stages, as available on solid-phase extraction (SPE) using a Gilson ASPECion-trap MS systems, can be of great help in followed by fast LC–APCI-MS–MS on 3034.6 mmstructure elucidation, as demonstrated by Tiller et al. I.D. columns with a run time of only 1.5 min. In the[96] in a glyburide metabolism study. Another overall sample throughput, the off-line sample pre-emerging trend is the on-line combination of LC– treatment appeared to be the rate-limiting step. In theNMR and LC–MS in metabolite studies, using either case of saquinavir, the sample throughput was |100(triple) quadrupole [97–99] or ion-trap [100] MS samples per analyst per day.systems. This indicates the clear need for speeding up the

sample pretreatment. One of most widely applied6.2. Quantitative bioanalysis in pharmaceutical approaches is the use of parallel SPE procedures onapplications short SPE columns or Empore disks, mounted in a

96-well plate format [102]. Allanson et al. [103], forQuantitative bioanalysis is the most important instance, achieved a 4–7-fold improvement in sam-

application area of LC–MS, in terms of number of ple throughput in the sample pretreatment by replac-instruments applied and the number of analyses ing SPE on individual samples by a robotic SPEperformed. Quantitative bioanalysis is required to sample pretreatment in 96-well plate format. Similarsupport preclinical and clinical drug testing and to approaches were described by others [104,105].provide pharmacokinetic and pharmacodynamic data. While off-line sample pretreatment seems to beAutomated unattended operation of the LC–MS preferred by most researchers, on-line strategies, e.g.,instrumentation is required. Fast and routine LC–MS based on Prospekt SPE instrumentation [106], areanalysis also demands fast and automated sample described as well. An example is the rapid de-pretreatment strategies and advanced data-processing termination of pranlukast and metabolites in humansoftware. plasma [107].

The keys to the success of LC–MS in quantitative Serious matrix problems may be experienced inbioanalysis are: (a) typical detection limits in the pg quantitative bioanalysis, especially in ESI. Signaland in favorable cases even sub-pg range, (b) suppression due to unknown matrix interferences isexcellent selectivity against possibly interfering com- often observed. Changes in the sample pretreatmentpounds in the biological matrix, especially when procedures may be successful in solving the prob-operated in SRM mode, (c) enhanced confidence of lem, but in some cases changing-over to APCI, whenidentity of the compound(s) analyzed, and (d) the applicable, appears to be the only feasible solution.ability to use the ideal internal standards: isotopically A well-documented example of signal suppressionlabelled compounds. LC–MS–MS is often as easy to was described by Matuszewski et al. [60].operate as LC–UV–PDA, but provides better selec-tivity. As a result, LC–MS–MS has become the 6.3. Environmental applicationsmethod-of-choice in quantitative bioanalysis withinpharmaceutical industries. For proprietary reasons, Environmental analysis is another important appli-unfortunately, there are not many reports on success- cation area of LC–MS [108]. The strategies requiredful LC–MS applications in quantitative bioanalysis are often different from those in pharmaceuticalavailable in the public literature. applications due to different aims of the analysis. In

The higher selectivity achievable due to the use of pharmaceutical applications, the analysis is directedSRM procedures is often immediately given away by to particularly one or only a few target compounds.decreasing the quality of the sample pretreatment SRM provides additional confirmation of identity


and selectivity. In environmental analysis, often a system, applied in the field, must be identified bymultiresidue screening is required. The analysis is MS, i.e., often LC–MS. In this step, a more target-directed at many different compounds from various compound oriented approach can be applied becausecompound classes. Restrictions to a more limited often the compound class of the unknown could benumber of compounds or compound classes are determined by UV–PDA. SAMOS-like systems canmade for practical reasons. For regulatory purposes, be on-line coupled to an LC–MS interface, as forgenerally a broad screening is required. instance demonstrated by the research groups of

´A clear example of the dilemma in developing Brinkman [111,112] and Barcelo [113,114]. In recentanalytical strategies in environmental analysis is the years, the volume to be applied in sample pretreat-screening of surface and ground waters prior to their ment could be reduced, e.g., from 100 down to 10use in the production of drinking water. According to ml, as a result of the improved performance of theEuropean regulations, the determination of any in- LC–MS instrumentation. Another approach is to usedividual pesticide at the level of 0.1 mg/ l is de- single short columns, applied for both preconcen-manded. Both quantitation and identification is re- trating sample pretreatment and minimum separationquired at this level. Detection of pesticides at these of the target compounds, as demonstrated by Hogen-levels can only be achieved by the use of pre- boom et al. [115]. The use of the so-called RFDconcentrating sample pretreatment, e.g., off-line or mode, which enables CID of all ions produced in theon-line liquid–liquid extraction or SPE, in combina- ion source above a certain cut-off mass, on a TSQ-70tion with selective ion monitoring or SRM. Such a QqQ instrument [12] has been described by Kienhuisprocedure is obviously not a general screening et al. [116]. Off-line SPE strategies combined withprocedure. Increasing the sample volume often re- flow-injection analysis MS–MS have been describedsults in the preconcentration of not only the analytes by Geerdink et al. [117].of interest, but also of interferences. The latter Recently, the advantages of the enhanced full-scanhamper full-scan analysis. In addition, the various sensitivity of an oa–TOF instrument were demon-compound classes show quite different responses in strated in the quantitation and identification ofESI or APCI, e.g., some compounds should be pesticides in environmental surface water [118]. On-analyzed in positive-ion mode, while others only line SPE–LC–MS on the oa–TOF was performed toprovide sufficient response in negative-ion mode. For determine accurate masses of the pesticides. Calcu-optimum SRM performance, optimization of the lated elemental compositions were searched against atuning parameters of ion source and collision cell is pesticide database for identification.required, basically for each individual compound or Next to strategies directed at the analysis ofcompound class. The problems increase even further pesticides and related compounds, the environmentalwhen not only pesticides but other environmental analysis of various other contaminants has attractedmicrocontaminants like surfactants, (azo) dyes, and significant attention as well, e.g., surfactants [119],perhaps even drug residues [109], have to be taken organotin compounds [120], and dyes [121].into account.

A typical approach in environmental analysis is anintegrated system called SAMOS [110], which en- 6.4. Biochemical and miscellaneous applicationsables the automated, unattended analysis of filtered100 ml surface water samples by means of on-line ESI is frequently applied in the various stages ofSPE on a short cartridge column, and subsequent the characterization of peptides and proteins: molec-gradient-HPLC analysis with UV–PDA detection. ular mass determination, amino acid sequencing,Compounds with sufficient UV activity can be determination of nature and position of chemical andquantified. In addition, using a UV spectral library, post-translational modifications of proteins, inves-provisional identification of compound or compound tigation in protein tertiary and quaternary conforma-class is possible, at least when the UV spectrum tion, and the study of noncovalent associates. In mostacquired is not too much distorted by the presence of cases, no on-line separation is applied, but thehumic acids. Any compound detected by a SAMOS sample solution is introduced directly via the ESI or


nano-ESI interface. Impressive results have been Unlike in GC–MS, the analyte peaks have to beachieved in this area [122–124]. searched for against a relatively high background of

On-line LC–MS for peptide and protein charac- solvent-related ion current in LC–MS. While a total-terization has also been described, especially in ion chromatogram (TIC) in GC–MS often revealsrelation to peptide sequencing and characterization of the presence of a number of compounds, even at lowsecondary protein structure, e.g., post-translational injected amounts, this is not necessarily true in LC–modifications. The determination of the N-glycosyla- MS. As such, low-level multiresidue screening and/tion sites in recombinant Human Factor VIII protein or the search for unknowns at low concentrations is aby reversed-phase LC–ESI–MS [125] may serve as difficult topic in LC–MS. Due to the high back-an example in this area. A microcapillary column- ground ion current, hardly any peaks show up in theswitching system to be applied in combination with TIC. By generating reconstructed mass chromato-LC–MS has been described by van der Heeft et al. grams, these hidden peaks can be detected, but[126] for the direct identification of peptides present searching for peaks over a wide mass range might bein major histocompatibility complex class I mole- a tedious and time-consuming procedure. The use ofcules. base-peak chromatograms can be helpful in this

Similarly, the application of LC–MS for the respect, but the implementation of the base-peakcharacterization of oligosaccharides [127,128] and chromatograms in most commercial MS softwareoligonucleotides [129,130] has also been described. packages is rather poor: in most cases the m /z range

LC–MS is applied in many other fields as well, to be searched for base peaks cannot be specified.e.g., in the study of natural products [131–133], Even more powerful chemometric approaches mightendogenous compounds like acylcarnitines and be available for peak recognition in a TIC obtainedarachidonic acid metabolites [134], and DNA ad- from LC–MS.ducts [135,136]. LC–MS has become a routinely Conventionally, MS is considered as an importantapplicable technique. It is rapidly entering the chro- tool in the identification of (totally) unknown ana-matography laboratories to act as an LC detector in a lytes. For GC–MS, clear successes in this respectvariety of analyses. As such, LC–MS is appreciated have been achieved. The excessive fragmentation infor its sensitivity and selectivity, its specificity and electron ionization, eventually in combination withthe information obtained, e.g., on molecular mass of high-resolution MS, can be applied to identify anthe analyte. The operation of an LC–MS system is unknown. The additional use of computer libraryno longer reserved to a MS specialist. searching helps in identification because the acquired

mass spectrum shows sufficient agreement with oneof the library entries, or the entries found pinpoint

7. Discussion and conclusions towards specific structural characteristics in theunknown.

The previous sections indicate the huge progress Identification by means of LC–MS technology ofthat has been made in LC–MS. Fascinating applica- unknowns is more difficult. LC–MS is an excellenttions are indicated, which clearly prove that LC–MS technique to confirm the identity of a target com-has developed into a mature technique. While it is pound, even when the target compound is a memberoften not difficult to achieve an appreciation of the of a relatively large group of (related) compounds.benefits and advantages of LC–MS, the limitations Either in-source CID or preferentially product-ionof the current technology appear to be somewhat MS–MS spectra can be applied for this purpose.neglected. Clearly, current LC–MS instrumentation Accurate mass determination by means of an oa-can solve problems that could not be solved (as TOF–MS can be useful as well. However, theefficiently) a couple of years ago. Nevertheless, LC– fragmentation of protonated molecules in low-energyMS requires a different appreciation of MS as an CID often leads to only a limited number of frag-analytical tool than does GC–MS. This is evaluated ments. This will often not allow unambiguous identi-below in relation to peak finding and analyte identifi- fication of the unknown, because the information incation. the MS–MS spectrum is insufficient. In addition, the


interpretation of the MS–MS spectrum is often tion as well as the data processing. Excellent inte-hampered by the lack of insight in the fragmentation grated and user-friendly software tools are availablerules. While fragmentation of the odd-electron radi- for the evaluation of quantitative analyse by LC–cal cation generated in electron ionization is well MS. Dedicated software packages are available tounderstood, this is not true for the fragmentation of operate LC–MS in specific applications, e.g., open-the even-electron protonated molecule. The available access, screening of combinatorial libraries, peptideknowledge is also not available in a systematic way. sequencing via (Internet) database searching, neonat-

Because of the influence of the experimental al screening of blood of newborn children forconditions on the appearance of the MS–MS or metabolic disorders, and MS-controlled fractionationin-source CID spectra, it is not possible to produce in preparative LC. These tools assist in establishingMS–MS libraries, which are applicable for a variety LC–MS as a useful tool in the respective applicationof instruments from different manufacturers. This areas, as they effectively give access to the parame-statement is based on the assumption that such a ters important to a specific application and guidelibrary should be searched by similar algorithms as through the complete analytical process.used for the conventional electron ionization li- These type of applications will dominate the usebraries, i.e., by taking into account both the m /z of LC–MS in the first decade on the next century.value and the relative intensity. Building a library Powerful applications, e.g., quantitative bioanalysis,where the m /z values of the fragment peaks is screening and confirmation of target compounds, andconsidered of prime important appears to be a to a limited extent identification of partially un-potential alternative [137]. knowns, will continue to give an important impetus

The softer and step-wise fragmentation achieved to the application and development of LC–MS.in CID in an ion trap often is a very helpful tool in Although there are some indications of emergingstructure elucidation. The high number of ion-trap new interface technology, e.g., sonic spraysystems currently installed perhaps enables the build- [138,139], laser spray [140], and continuous-flowing of spectral libraries. However, such a library will and aerosol matrix-assisted laser desorption / ioniza-only be applicable with ion traps, because MS–MS tion [141,142], ESI and APCI will continue to be thespectra from QqQ instrument in most cases are most important interfacing and ionization ap-significantly different. The stepwise fragmentation proaches. Improvements in the instrumentation andmight be applied to study fragmentation mechanisms especially in the software are to be expected, in orderof protonated molecules. However, the general ap- to accommodate LC–MS to perform specific tasksplicability of this knowledge may be questionable within particular applications. And for the time tobecause the apparent differences in the CID process. come, applications within the pharmaceutical indus-This is exemplified by the ability to fragment try continue to be the most important applicationsodiated molecules in an ion trap, while this is area for LC–MS.generally not possible in a QqQ instrument.

The current interest in the use of oa-TOF, Q–TOFand FT–ICR–MS instruments in structure elucida-

Referencestion is mainly based on the assumption, that accuratemass determination of the (product) ions facilitates

[1] W.M.A. Niessen, Liquid Chromatography–Mass Spec-the interpretation of the fragments.trometry, 2nd ed, Chromatogr. Sci. Series, Vol. 79, Marcel

As indicated above, the identification of a total Dekker Inc, New York, NY, 1999.unknown is not required in most of the current [2] R. Willoughby, E. Sheehan, S. Mitrovich, A Global View of

LC?MS, Global View Publishing, Pittsburgh, PA, 1998.applications of LC–MS. Generally, target-compound[3] W.M.A. Niessen, J. Chromatogr. A 794 (1998) 407.like strategies are applied. In these areas, the per-[4] Analytica of Branford, Branford, CT, http: / /www.aob.com.formance of current LC–MS instrumentation is[5] Bruker Daltonics, Bremen, Germany, http: / /www.bruker-

generally satisfactory. Improvements in software .com/daltonics.facilitates the rapid optimization of experimental [6] Hewlett-Packard, Palo Alto, CA, http: / /chem.exter-parameters, the automated unattended data acquisi- nal.hp.com/cag.


[7] Hitachi, Tokyo, Japan, http: / /www.hii.hitachi.com. [42] B.A. Thomson, J.V. Iribarne, P.J. Dziedzic, Anal. Chem. 54[8] Jeol, Tokyo, Japan, http: / /www.jeol.com/ms. (1982) 2219.[9] Micromass, Manchester, UK, http: / /www.micromass.co.uk. [43] J.V. Iribarne, P.J. Dziedzic, B.A. Thomson, Int. J. Mass

[10] Perkin-Elmer Sciex, Concord, Ontario, Canada, http: / Spectrom. Ion Phys. 50 (1983) 331./www.pesciex.com. [44] P. Kebarle, Y. Ho, in: R.B. Cole (Ed.), Electrospray

[11] Shimadzu, Kyoto, Japan, http: / /www.shimadzu.com. Ionization Mass Spectrometry, Wiley & Sons, New York,´[12] Thermoquest, San Jose, CA, http: / /www.finnigan.com. NY, 1997, p. 3.

[13] Waters, Milford, MA, http: / /www.waters.com. [45] J.B. Fenn, J. Am. Soc. Mass Spectrom. 4 (1993) 524.[14] A.P. Bruins, T.R. Covey, J.D. Henion, Anal. Chem. 59 [46] E.E. Dodd, J. Appl. Phys. 24 (1953) 73.

(1987) 2642. [47] A.T. Blades, M.G. Ikonomou, P. Kebarle, Anal. Chem. 63[15] J.F. Banks, J.P. Quinn, C.M. Whitehouse, Anal. Chem. 66 (1991) 2109.

(1994) 3688. [48] A. Gomez, K. Tang, Phys. Fluids 6 (1994) 404.[16] L. Tang, P. Kebarle, Anal. Chem. 65 (1993) 3654. [49] R.D. Smith, K.J. Light-Wahl, Biol. Mass Spectrom. 22[17] G. Hopfgartner, K. Bean, J.D. Henion, R.A. Henry, J. (1993) 493.

Chromatogr. 647 (1993) 51. [50] A. Appfel, S. Fischer, G. Goldberg, P.C. Goodley, F.E.[18] T.R. Covey, The Realities and Misconceptions of Electro- Kuhlmann, J. Chromatogr. A 712 (1995) 177.

ˇspray Ionization and HPLC Flow-rates, Thornhill, Ontario, [51] C.L. Gatlin, F. Turecek, Anal. Chem. 66 (1994) 712.1993, Perkin-Elmer Sciex Ionspray Application Note. [52] W.M.A. Niessen, U.R. Tjaden, J. van der Greef, J. Chroma-

[19] A.P. Bruins, J. Chromatogr. A 794 (1998) 345. togr. 554 (1991) 3.[20] M. Busman, J. Sunner, C.R. Vogel, J. Am. Soc. Mass [53] J.A. Loo, H.R. Udseth, R.D. Smith, Rapid Commun. Mass

Spectrom. 2 (1991) 1. Spectrom. 2 (1988) 207.[21] M.S. Wilm, M. Mann, Int. J. Mass Spectrom. Ion Processes [54] D.S. Ashton, C.R. Beddell, D.J. Cooper, B.N. Green,

136 (1994) 167. R.W.A. Oliver, Org. Mass Spectrom. 28 (1993) 721.[22] M.S. Wilm, M. Mann, Anal. Chem. 68 (1996) 1. [55] A.G. Harrison, Chemical Ionization Mass Spectrometry,[23] G.A. Valaskovic, F.W. McLafferty, J. Am. Soc. Mass 2nd ed, CRC Press, Boca Raton, FL, 1992.

Spectrom. 7 (1996) 1270. [56] A.P. Bruins, Trends Anal. Chem. 13 (1994) 37.[24] J.C. Hannis, D.C. Muddiman, Rapid Commun. Mass Spec- [57] R.R. Ogorzalek Loo, J.A. Loo, H.R. Udseth, J.L. Fulton,

trom. 12 (1998) 443. R.D. Smith, Rapid Commun. Mass Spectrom. 6 (1992) 159.[25] K.P. Bateman, R.L. White, P. Thibault, Rapid Commun. [58] R.R. Ogorzalek Loo, R.D. Smith, J. Am. Soc. Mass

Mass Spectrom. 11 (1997) 307. Spectrom. 5 (1994) 207.[26] J.N. Alexander IV, G.A. Schultz, J.B. Poli, Rapid Commun. [59] K.L. Rundlett, D.W. Armstrong, Anal. Chem. 68 (1996)

Mass Spectrom. 12 (1998) 1187. 3493.[27] S. Geromanos, J. Philip, G. Freckleton, P. Tempst, Rapid [60] B.K. Matuszewski, M.L. Constanzer, C.M. Chavez-Eng,

Commun. Mass Spectrom. 12 (1998) 551. Anal. Chem. 70 (1998) 882.[28] K.W.Y. Fond, T.-W.D. Chan, J. Am. Soc. Mass Spectrom. 10 [61] B.A. Mansoori, E.W. Dyer, C.M. Lock, K. Bateman, R.K.

(1999) 72. Boyd, B.A. Thomson, J. Am. Soc. Mass Spectrom. 9[29] Protana, Odense, Denmark, http: / /www.protana.com. (1998) 775.[30] New Objective Inc., Cambridge, MA, http: / /www.newob- [62] M. Wakefield, L. Lopez, D. Buhrman, J. Newton, B. Folk,

jective.com. R. Burton, Thermoquest Finnigan LCQ/TSQ Application´[31] M. Dedieu, C. Juin, P.J. Arpino, G. Guiochon, Anal. Chem. Report [258, Finnigan Corporation, San Jose, CA., 1996.

54 (1982) 2372. [63] A.C. Hogenboom, W.M.A. Niessen, U.A.Th. Brinkman, J.[32] M. Dole, R.L. Hines, L.L. Mack, R.C. Mobley, L.D. Chromatogr. A 794 (1998) 201.

Ferguson, M.B. Alice, J. Chem. Phys. 49 (1968) 2240. [64] P.R. Tiller, Z. El Fallah, V. Wilson, J. Huysman, D. Patel,[33] J. Gieniec, L.L. Mack, K. Nakamae, C. Gupta, V. Kumar, Rapid Commun. Mass Spectrom. 11 (1997) 1570.

M. Dole, Biomed. Mass Spectrom. 11 (1984) 259. [65] D.M. Drexler, P.R. Tiller, S.M. Wilbert, F.Q. Bramble, J.C.[34] M.L. Vestal, Mass Spectrom. Rev. 2 (1983) 447. Schwartz, Rapid Commun. Mass Spectrom. 12 (1998)[35] C.R. Blakley, J.J. Carmody, M.L.Vestal, J. Am. Chem. Soc. 1501.

102 (1980) 5931. [66] H.R. Morris, T. Paxton, A. Dell, J. Langhorne, M. Berg,[36] V. Katta, A.L. Rockwood, M.L. Vestal, Int. J. Mass Spec- R.S. Bordoli, J. Hoyes, R.H. Bateman, Rapid Commun.

trom. Ion Proc. 103 (1991) 129. Mass Spectrom. 10 (1996) 889.[37] M. Yamashita, J.B. Fenn, J. Phys. Chem. 88 (1984) 4451. [67] H.R. Morris, T. Paxton, M. Panico, M. McDowell, A. Dell,[38] M. Yamashita, J.B. Fenn, J. Phys. Chem. 88 (1984) 4671. J. Protein Chem. 16 (1997) 469.[39] J.B. Fenn, M. Mann, C.K. Meng, S.F. Wong, C.M. [68] A. Shevchenko, I. Chernushevich, W. Ens, K.G. Standing,

Whitehouse, Mass Spectrom. Rev. 9 (1990) 37. B.A. Thomson, M.S. Wilm, M. Mann, Rapid Commun.[40] J.V. Iribarne, B.A. Thomson, J. Chem. Phys. 64 (1976) Mass Spectrom. 11 (1997) 1015.

2287. [69] M.G. Qian, D.M. Lubman, Anal. Chem. 67 (1995) 234A.[41] B.A. Thomson, J.V. Iribarne, J. Chem. Phys. 71 (1979) [70] J.-T. Wu, M.G. Qian, M.X. Li, K. Zheng, P. Huang, D.M.

4451. Lubman, J. Chromatogr. A 794 (1998) 377.


[71] J.-T. Wu, L. He, M.X. Li, S. Parus, D.M. Lubman, J. Am. [98] E. Clayton, S. Taylor, B. Wright, I.D. Wilson, Chromato-Soc. Mass Spectrom. 8 (1997) 1237. graphia 47 (1998) 264.

[72] D.A. Laude, E. Stevenson, J.M. Robinson, in: R.B. Cole [99] G.J. Dear, J. Ayrton, R. Plumb, B.C. Sweatman, I.M.(Ed.), Electrospray Ionization Mass Spectrometry, Wiley & Ismail, I.J. Fraser, P.J. Mutch, Rapid Commun. MassSons, New York, NY, 1997, p. 291. Spectrom. 12 (1998) 2023.

[73] E.R. Williams, Anal. Chem. 70 (1998) 179A. [100] J.P. Shockcor, S.E. Unger, I.D. Wilson, P.J.D. Foxall, J.K.Nicholson, J.C. Lindon, Anal. Chem. 68 (1996) 4431.[74] M. Mann, P. Hojrup, P. Roepstorff, Biol. Mass Spectrom.

22 (1993) 338. [101] N.R. Knebel, S.R. Sharp, M.J. Madigan, J. Mass Spectrom.30 (1995) 1149.[75] A.L. McCormack, D.M. Shieltz, B. Goode, S. Yang, G.

Barnes, D. Drubin, J.R. Yates III, Anal. Chem. 69 (1997) [102] B. Kaye, W.J. Heron, P.V. McRae, S. Robinson, D.A.767. Stopher, R.F. Venn, W. Wild, Anal. Chem. 68 (1996) 1658.

[76] F.S. Pullen, G.L. Perkins, K.I. Burton, R.S. Ware, M.S. [103] J.P. Allanson, R.A. Biddlecombe, A.E. Jones, S. Pleasance,Taegue, J.P. Kiplinger, J. Am. Soc. Mass Spectrom. 6 Rapid Commun. Mass Spectrom. 10 (1996) 811.(1995) 394. [104] J. Janiszewski, R.P. Schneider, K. Hoffmaster, M. Swyden,

[77] R.C. Spreen, L.M. Schaffter, Anal. Chem. 68 (1996) 414A. D. Wells, H. Fouda, Rapid Commun. Mass Spectrom. 11(1997) 1033.[78] R.M. Holt, M.J. Newman, F.S. Pullen, D.S. Richards, A.G.

Swanson, J. Mass Spectrom. 32 (1997) 64. [105] H. Simpson, A. Berthemy, D. Buhrman, R. Burton, J.Newton, M. Kealy, D. Wells, D. Wu, Rapid Commun. Mass[79] J.N. Kyranos, J.C. Hogan Jr., Anal. Chem. 70 (1998) 389A.Spectrom. 12 (1998) 75.¨[80] G. Hegy, E. Gorlach, R. Richmond, F. Bitsch, Rapid

[106] Spark Holland, Emmen, the Netherlands, http: /Commun. Mass Spectrom 10 (1996) 1894./www.spark.nl.[81] D.B. Kassel, M.D. Green, R. Wehbie, R. Swanstrom, J.

[107] A. Marchese, C. McHugh, J. Kehler, H. Bi, J. MassBerman, Anal. Biochem. 228 (1995) 259.Spectrom. 33 (1998) 1071.[82] T. Wang, L. Zeng, T. Strader, L. Burton, D.B. Kassel, Rapid

´[108] D. Barcelo (Ed.), Applications of LC?MS in EnvironmentalCommun. Mass Spectrom. 12 (1998) 1123.Chemistry, Elsevier Science, Amsterdam, 1996.[83] E.W. Taylor, M.G. Qian, G.D. Dollinger, Anal. Chem. 70

[109] R. Hirsch, T.A. Ternes, K. Haberer, A. Mehlich, F.(1998) 3339.Ballwanz, K.-L. Kratz, J. Chromatogr. A 815 (1998) 213.¨[84] E. Gorlach, R. Richmond, I. Lewis, Anal. Chem. 70 (1998)

´[110] U.A.Th. Brinkman, J. Slobodnık, J.J. Vreuls, Trends Anal.3227.Chem. 13 (1994) 373.[85] J.R. Krone, R.W. Nelson, D. Dogruel, P. Williams, R.

´ ´[111] H. Bagheri, J. Slobodnık, R.M. Marce, R.T. Ghijsen, U.A.Granzow, Anal. Biochem. 244 (1997) 124.Th. Brinkman, Chromatographia 37 (1993) 159.[86] M.L. Nedved, S. Habibi-Goudarzi, B. Ganem, J.D. Henion,

´[112] J. Slobodnık, A.C. Hogenboom, J.J. Vreuls, J.A. Rontree,Anal. Chem. 68 (1996) 4228.B.L.M. van Baar, W.M.A. Niessen, U.A. Th. Brinkman, J.[87] T. Schenk, H. Irth, A.J. Oosterkamp, E.S.M. Lutz, U.Chromatogr. A 741 (1996) 59.Schobel, Anal. Chem., 71 (1998) in press.

´[113] S. Lacorte, D. Barcelo, Anal. Chem. 68 (1996) 2464.[88] ScreenTec, Leiden, The Netherlands, http: / /www.screen-´ ´[114] C. Aguilar, I. Ferrer, F. Borrull, R.M. Marce, D. Barcelo, J.tec.nl.

Chromatogr. A 794 (1998) 147.[89] L. Zeng, L. Burton, K. Yung, B. Shushan, D.B. Kassel, J.[115] A.C. Hogenboom, P. Speksnijder, R.J. Vreeken, W.M.A.Chromatogr. A 794 (1998) 3.

Niessen, U.A.Th. Brinkman, J. Chromatogr. 777 (1997) 81.[90] L. Zeng, D.B. Kassel, Anal. Chem. 70 (1998) 4380.[116] P.G.M. Kienhuis, R.B. Geerdink, A. Sijpersma, Anal.[91] D.M. Drexler, P.R. Tiller, Rapid Commun. Mass Spectrom.

Methods Instrum. 2 (1995) 236.12 (1998) 895.´[117] R.B. Geerdink, in: D. Barcelo (Ed.), Applications of LC?[92] K.J. Volk, S.E. Klohr, R.A. Rourick, E.H. Kerns, M.S. Lee,

MS in Environmental Chemistry, Elsevier Science, Am-J. Pharm. Biomed. Anal. 14 (1996) 1663.sterdam, 1996, p. 135.[93] R. Brownshill, J.-P. Combal, A. Taylor, M. Bertrand, W.

[118] A.C. Hogenboom, W.M.A. Niessen, D. Little, U.A.Th.Luijten, B. Walther, Rapid Commun. Mass Spectrom. 8Brinkman, Rapid Commun. Mass Spectrom. 13 (1999) 125.(1994) 361.

[119] A. DiCorcia, J. Chromatogr. A 794 (1998) 165.[94] N.J. Haskins, C. Eckers, A.J. Organ, M.F. Dunk, B.E.[120] S. White, T. Catterick, B. Fairman, K. Webb, J. Chromatogr.Winder, Rapid Commun. Mass Spectrom. 9 (1995) 1027.

A 794 (1998) 211.´[95] C. Eckers, N.J. Haskins, M.J. Lee, S. Monte, J. Sanderson,´[121] J. Yinon, L.D. Betowski, R.D. Voyksner, in: D. Barceloin: Presented at MS in Proteomics and Drug Discovery,

(Ed.), Applications of LC?MS in Environmental Chemistry,Shrigley Hall Hotel, UK, 18–20 October, 1998, organizedElsevier Science, Amsterdam, 1996, p. 187.by Micromass Ltd.

[96] P.R. Tiller, A.P. Land, I. Jardine, D.M. Murphy, R. Sozio, [122] S. Akashi, M. Shirouzu, T. Terada, Y. Ito, S. Yokoyama, K.A. Ayrton, W.H. Schaefer, J. Chromatogr. 794 (1998) 15. Takio, Anal. Biochem. 248 (1997) 15.

[97] K.I. Burton, J.R. Everett, M.J. Newman, F.S. Pullen, D.S. [123] H. Deissler, M.S. Wilm, B. Genc, B. Schmitz, T. Ternes, F.Richards, A.G. Swanson, J. Pharm. Biomed. Anal. 15 Naumann, M. Mann, W. Doerfler, J. Biol. Chem. 272(1997) 1903. (1997) 16761.


[124] G. Neubauer, A. Gottschalk, P. Fabrizio, B. Seraphin, R. [133] G. Lin, K.-Y. Zhou, X.-G. Zhao, Z.-T. Wang, P. Pui HayLuhrmann, M. Mann, Proc Nat Acad Sci USA 94 (1997) But, Rapid Commun. Mass Spectrom. 12 (1998) 1445.385. [134] C.S. Newby, A.I. Mallet, Rapid Commun. Mass Spectrom.

[125] K.F. Medzihradszky, M.J. Besman, A.L. Burlingame, Anal. 11 (1997) 1723.Chem. 69 (1997) 3986. [135] W. Appruzzese, P. Vouros, J. Chromatogr. A 794 (1998) 97.

`[126] E. van der Heeft, G.J. ten Hove, C.A. Herberts, H.D. [136] E.L. Esmans, D. Broes, I. Hoes, F. Lemiere, K. Vanhoutte,Meiring, C.A.C.M. van Els, A.P.J.M. de Jong, Anal. Chem. J. Chromatogr. A 794 (1998) 109.70 (1998) 3742. [137] P. Kienhuis, R. Geerdink, I. Freriks, in: Presented at the

[127] R.A.M. van der Hoeven, A.J.P. Hofte, U.R. Tjaden, J. van 15th Montreux Symposium of LC–MS, Montreux, Switzer-der Greef, N. Torto, L. Gorton, G. Marko-Varga, C. land, November 11–13, 1998.Bruggink, Rapid Commun. Mass Spectrom. 12 (1998) 69. [138] A. Hirabayashi, M. Sakairi, H. Koizumi, Anal. Chem. 67

[128] Y. Mechref, M.V. Novotny, Anal. Chem. 70 (1998) 455. (1995) 2878.[129] A. Apffel, J.A. Chakel, S. Fischer, K. Lichtenwalter, W.S. [139] Y. Hirabayashi, A. Hirabayashi, Y. Takada, M. Sakairi, H.

Hancock, Anal. Chem. 69 (1997) 1320. Koizumi, Anal. Chem. 70 (1998) 1882.[130] D.L.D. Deforce, J. Raymackers, L. Meheus, F. van Wijnen- [140] K. Hiraoka, S. Saito, J. Katsuragawa, I. Kudaka, Rapid

daele, A. de Leenheer, E.G. van den Eeckhout, Anal. Chem. Commun. Mass Spectrom. 12 (1998) 1170.70 (1998) 3060. [141] R.M. Whittal, L.M. Russon, L. Li, J. Chromatogr. A 794

[131] J.-L. Wolfender, S. Rodriguez, K. Hostettmann, J. Chroma- (1998) 367.togr. A 794 (1998) 299. [142] X. Fei, G. Wei, K.K. Murray, Anal. Chem. 68 (1996) 1143.

¨[132] M. Stadler, T. Henkel, H. Muller, K. Weber, H. Schreckler,J. Chromatogr. A 818 (1998) 187.

Documents

Review State-of-the-art in liquid chromatography–mass ...€¦ · Impressive progress has been made in the technology and application of combined liquid chromatography–mass spectrometry