2 Physical Methods and Techniques Part (ii) Mass Spectrometry By M. A. BALDWIN Department of Pharmaceutical Chemistry School of Pharmacy University of London 29/39 Brunswick Square London WClN 1AX 1 Introduction This report will begin by updating some of the analytical techniques dealt with in the preceding review,’ in particular those relating to the rapid growth of new ionization methods. It will deal with developments in mass analysers with particular emphasis on ion trapping techniques. Finally the ability of mass spectrometry to produce and study novel chemical species will be discussed. It is not intended to update the remarks on the general mass spectrometry literature as the sources listed previously are still recommended.’ The proceedings of the 10th International Mass Spectrometry Conference held in Swansea in September 1985 have been published.2 The two-volume set contains 31 invited review papers and 528 two-page abstracts of submitted contributions.This provides an excellent snap- shot of the state of mass spectrometry at that time. A new volume has appeared in the series Specialist Periodical Reports in Mass Spe~trometry,~ and a further very comprehensive Analytical Chemistry review of the field has been published by Burlingame et aL4 2 Fast Atom Bombardrnent/Secondary Ion Mass Spectrometry Grouping these two techniques together is an acknowledgement that fast atom bombardment mass spectrometry (FABMS) and secondary ion mass spectrometry (SIMS) are essentially indistinguishable.SIMS was already long-established for inorganic and surface analysis when Barber et al. introduced FAB as a new tech- niq~e.~ At that time there was an implicit assumption that the use of inert gas atoms to bombard a sample to produce sputtered ions was significantly different to bom- bardment by ions. However the crucial innovation was in fact the use of a liquid matrix as a solvent and support for the sample in FAB allowing the successful analysis of a range of involatile and thermally labile biomolecules.6 More recently ’ M. A. Baldwin Annu. Rep. Frog. Chem. Sect. B 1986 82 15. ’‘Advances in Mass Spectrometry 1985’ ed. J. F. J. Todd Wiley Chichester 1986. ‘Mass Spectrometry’ Vol. 9 ed. M. E. Rose A Specialist Periodical Report The Royal Society of Chemistry London 1987.A. L. Burlingame T. A. Baillie and P. J. Derrick Anal. Chem. 1986 58 165R. M. Barber R. S. Bordoli R. D. Sedgewick and A. N. Tyler J. Chem. SOC. Chem. Commun. 1981 325. M. Barber R. S. Bordoli G. J. Elliot R. D. Sedgewick and A. N. Tyler Anal. Chem. 1982 54 645A. 15 16 M. A. Baldwin it has been appreciated that there are advantages in dealing with ion rather than with neutral beams i.e. higher kinetic energies can be achieved the beams are readily focused and the residual pressure of the inert gas is avoided. This has encouraged the use of caesium ion guns for the bombardment of liquid samples,' which with 35 keV primary ions has extended the molecular mass range for FAB/SIMS of proteins to about 23 OOO.* Some authors persist in describing this as FAB which in fact is a FIB i.e.fast ion bombardment or an untruth! A grammatically unacceptable alternative in current use is liquid SIMS (LSIMS) ie.SIMS of liquid rather than solid samples. This topic of nomenclature has been discussed elsewhere without any very satisfactory conclusions being reached." The role of the matrix has been identified as providing a fresh sample surface for bombardment with sample molecules diffusing to replace those damaged by the high energy impact. (10 keV particles have kinetic energies equivalent to lo6kJ mol-I.) Surface-active samples are advantageous and the choice of matrix may be critical. It has been pointed out that enhanced sensitivity may be attained by derivatizing hydrophilic samples with more hydrophobic groups to encourage surface activity e.g.esterification of carboxylic acid groups in peptides.' A number of groups have experimented with cooled sample probes that allow more volatile liquids to be analysed without added matrix. The mass spectrum of propanol was obtained at temperatures close to its melting point (-lOS0C) and at higher sub- ambient temperatures a range of terpenes and related compounds such as geraniol and terpinyl acetate gave good mass spectra." Quantitative studies in solution using FAB analysis are still problematical as suppression of certain ions is commonplace. The ions observed from gfycine/Cu'' equilibria in glycerol did not include any copper-containing species even though it was known that chelated structures were present in solution." By contrast using stable isotope dilution techniques several very successful analyses of trace metals have been carried out with dry rather than liquid samples.Calcium has been analysed in plasma with a coefficient of variation of 0.2'/0,'~ and zinc was assayed in faeces with a CV of 1% at an enrichment of 1'/o .I3 Isotope dilution minimizes the sup- pression problem as both the isotopes will be similarly affected and the use of stable rather than radioisotopes allows their use in human studies. The use of mass spectrometry for trace analysis of metals has been reviewed.I4 The experience with the copper-glycine complex is perhaps atypical in that the use of FABMS in inorganic analysis is widespread. The choice of matrix is probably more critical with inorganic samples and the nature of the material from which the probe has been constructed may also be important.The complex (PC3H7.Hg12)2 was analysed successfully from an aluminium surface but with a copper probe there was complete exchange of the mercury by ~opper.'~ Solubility in the matrix is ' G. Elliot and J. S. Cottrell Proceedings of the 34th Annual Conference on Mass Spectrometry and Allied Topics Cincinatti Ohio 1986. a M. Barber and B. N. Green Rapid Commun. Mass Spectrom. 1987 1 80. S. A. Naylor A. F. F. Findeis B. W. Gibson and D. H. Williams J. Am. Chem. Soc. 1986 108 6359. 10 K. Heckles R. A. W. Johnstone and A. H. Willby Tetrahedron Lett. 1987 28 103. I' M. J. Connolly and R. G. Orth Anal. Chem.1987 59 903. 12 X. Jiang and D. L. Smith Anal. Chem. 1987 59 2570. l3 P. L. Peirce K. M. Hambidge C. H. Goss L. V. Miller and P. V. Fennessey Anal. Chem. 1987,59,2034. 14 D. E. Pratt J. Eagles and R. Self in ref. 3 p. 407. '' P. R. Ashton and M. E. Rose Org. Mass Spectrom. 1986 21 388. Physical Methods and Techniques -Part (ii) Mass Spectrometry 17 essential for effective ionization hence the need for a DMSO-thioglycerol mixture for the analysis of cisplatin analogues which gave useful spectra in both positive and negative ion modes.16 Continuous Flow FAB.-A very interesting development is the introduction of con- tinuous flow methods. 17*18 Hitherto sample introduction has usually been effected with the sample dissolved in a relatively viscous and involatile liquid such as glycerol a small drop of the solution being inserted into the mass spectrometer via a vacuum lock.This is then bombarded by the atom beam for a period of several minutes the maximum lifetime of the experiment being determined by the time taken for the liquid matrix to evaporate. The range of satisfactory matrix materials is limited but continuous flow methods allow a much wider choice of solvents and allow the experiment to continue almost indefinitely. The solvent generally contains a small proportion of glycerol to improve the flow characteristics and to prevent too rapid evaporation but the clustering of glycerol -producing interfering ions of quite high mass -is greatly reduced. This technique provides improved signal-to-noise for trace analysis and has given spectra of peptides ranging from substance P (RMM 1347) up to insulin," and a number of 'difficult' samples have been analysed in this way.'' This technique also offers new opportunities in interfacing high performance liquid chromatography with mass spectrometry.The optimum flow rate for con- tinuous flow FAB is about 5 p1 min-' which is ideal for microbore HPLC.20 This has been used for direct analysis of peptide digests,21,22 peptides and proteins,21 and antibiotics.22 HPLC with standard packed columns at a flow rate of 1ml min-' has been used. After the column the eluent (methanol-water-trifluoroacetic acid) was mixed with a little glycerol from a second pump. It then flowed through a standard UV detector before being split to deliver approximately 1% to the con- tinuous flow FAB probe via a fused silica tube.Peptides of different masses that were not well-resolved in the UV chromatogram were readily separated by the mass spectrometer using selected ion monitoring. Alternatively the mass spectrometer could produce a mass spectrum for each component of interest in the UV Caprioli has investigated continuous flow FAB as an alternative to 'static' FAB for continuous monitoring of reactions such as tryptic digestion of peptides. One advantage is the ability to use predominantly aqueous solvent for the reactions rather than glycerol making the results more consistent with other techniques. Water provides another benefit in reduced suppression of certain peptides.There are nine peptides of masses greater than 700 produced by tryptic digestion of myoglobin and they have various hydrophilicity/ hydrophobicity indices. The five hydrophobic peptides are surface active in glycerol and they were detected strongly in standard 16 M. M. Seigel P. Bitha R. G. Child J. J. Hlavka Y.-I. Lin and T. T. Chang Biomed. Enuiron. Muss Spectrom. 1986 13 25. Y. Ito T. Takeuchi D. Ishi and M. J. Goto J. Chromutogr. 1985 346 161. '* R. M. Caprioli T. Fan and J. S. Cottrell Anal. Chem. 1986 58 2949. 19 M. Barber L. W. Tetler D. Bell A. E. Ashcroft R. S. Brown and C. Moore Org. Muss Spectrom. 1987 22 647. 2o A. E. Ashcroft J. R. Chapman and J. S. Cottrell J. Chromatogr. 1987 394 15. 21 D. W. Hutchinson A. R. Woolfitt and A.E. Ashcroft Org. Muss Spectrom. 1987 22 304. 22 A. E. Ashcroft Org. Mass Spectrom. 1987 22 754. 23 D. E. Games S. Pleasance E. D. Ramsey and M. A. McDowell Biomed. Enuiron. Mass Specworn. 1988 15 179. 18 M. A. Baldwin FAB whereas the four hydrophilic peptides gave no signals in the presence of the hydrophobics. This suppression problem was much reduced with continuous introduction of aqueous solutions only one hydrophilic peptide showing no signal at Ionization Mechanisms in Sputtering Techniques.-The development of FAB/SIMS and other desorption methods has been an outstanding success in broadening the scope of mass spectrometry in organic and biochemical analysis. Many examples of new areas of application are described in two recent books one on general applications of mass spectrometry to bio~hemistry,~~ and the other on the analysis of large molecules.26 This latter topic is not restricted to biochemistry and also embraces the analysis of polymers.The production and characterization of cluster ions is also a new area which has been reviewed independently by Derri~k,~’ Mark,28 and Sta~e.~~ All three authors address the fascinating subject of magic numbers i.e. a particular number of individual atoms or molecules that provide the cluster with enhanced stability. Examples include the caesium ion clusters CS1& which is a 3 x 3 x 3 cube and Cs231r2 which is 3 x 3 x 5. Amongst protonated water clusters H(H20)11,which forms a pentagonal dodecahedron is favoured; for carbon C60 Buckminsterfullerene shows enhanced stability both as a cation and as an anion.It has been pointed out by many authors that the spectra from FAB/SIMS are extremely similar to those from other sputtering/desorption mass spectrometry techniques including field desorption (FDMS) laser desorption (LDMS) and plasma desorption (PDMS) induced by fission fragment bombardment (252Cf fission) or particles from accelerators. For example the LDMS spectrum of bovine insulin has recently been p~blished,~’ and is very similar to that from FABMS. Desorption and ionization induced by MeV energy particles is essentially the same as that from 10 keV FAB although the higher energy particles are better at ionizing high mass compounds and they are more efficient in terms of secondary ion yield per initial bombardment which is consistent with statistical RRK-based kinetic theory.31 In practical terms the latter advantage of PDMS may be cancelled out by the more intense primary beams attainable in FAB/ SIMS.Nevertheless the ionization mechanisms are still not clearly understood. Derrick and Sundqvist have edited a special volume of International Journal of Mass Spectrometry and Ion Processes entitled ‘Gaseous ions from involatile ther- mally-sensitive materials energetics and mechanisms’. In a preface they state that ionization precedes volatilization -far more neutrals than ions are sputtered into the gas phase and the ionization process is inefficient and uncontrolled and they suggest that for more effective techniques it will be necessary to achieve volatilization before ionization.32 This is consistent with a report from Williams et al.that neutral species in solution are generally sputtered as neutrals giving weak FAB spectra 24 R. M. Caprioli W. T. Moore and T. Fan Rapid Commun. Mass Spectrom. 1987 1 15. 25 ‘Mass Spectrometry in Biomedical Research’ ed. S. J. Gaskell Wiley Chichester 1986. 26 ‘Mass Spectrometry in the Analysis of Large Molecules’ ed. C. J. McNeaI Wiley Chichester 1986. 27 P. J. Derrick in ref. 2 p. 85. 28 T. D. Mark Int. J. Mass Spectrom. Ion Processes 1987 79 1. 29 A. J. Stace in ref. 3 p. 96. 30 J. Grotemeyer and E. W. Schlag Org. Mass Spectrom. 1987 22 758. 31 I. S. T. Tsong J. W. Christiansen S. H. Lin and B.V. King in ref. 26 p.67. 32 P. J. Demck and B. Sundqvist Int. J. Mass Spectrom. Ion Processes 1987 78 ix. Physical Methods and Techniques -Part (ii) Mass Spectrometry 19 whereas preionized species such as peptides containing basic amino acids are sputtered as ions and neutrals to a comparable extent.33 It has also been shown that addition of organic bases to the matrix promotes (M-H)- formation for nucleic acids in negative ion FAB.34 However this view of the ionization mechanism is not universally held Sunner et al. have developed a phase explosion model of desorption in which most of the ionization occurs in the relatively high pressure region immedi- ately above the liquid. Here the densities are 10%-30% of those in the liquid phase and kinetic modelling can be achieved with a gas collision Ionization occurs by processes analogous to chemical ionization hence the importance of gas phase basicities in determining the protonated species observed in FAB.36 Undoubtedly the tendancy is for many more neutrals than ions to be sputtered into the vapour phase.For LDMS this has been exploited by using two lasers operating at different wavelengths one for sample evaporation and the other for ionization thereby achieving temporal and spatial separation of these two processes. In one example a C02 infra-red laser was used for evaporation and a ND:YAG UV laser at 266nm was pulsed approximately 100 ps later to cause resonance enhanced multiphoton ionization (REMPI). This instrument gave good detection of PTH-amino acids at the 1 pmole level and gave a linear response up to 1 nm01e.~’ However it should be emphasized that the PTH-amino acids are by no means as demanding to vaporize and ionize as the proteins that are now being observed in FAB/SIMS and PDMS experiments.The insulin molecule has 2358 normal modes of vibration about 450 of which must be excited for the intact molecule to be desorbed with an average energy deposition of about 800 ev/m~lecule.~’ Multiphoton ionization (MUPI) LDMS from the solid phase with an excimer laser operating at different wavelengths in the UV shows that spectral characteristics change with the photon energy and this has been related to competitive photoioniz- ation and energy transfer processes in the solid.38 Ionization mechanisms in MUPI have been reviewed by Neusser and he also discusses the production of state-selected and energy-selected ions and their role in studying unimolecular ion reaction kinetics.39 Mass Analysers for Sputtering Techniques.-The majority of mass spectrometers in widespread use employ either magnetic sectors (and perhaps electric sectors) or quadrupole analysers for the separation of ions.Both of these are ‘continuous mode’ techniques in that a continuous beam of ions is produced in the ion source and passed through the instrument. The flux of ions is usually great enough for this to be treated as an electrical current which is amplified as an analogue signal. Weak signals will result in statistical ‘noise’ but commercial mass spectrometers are rarely equipped for the ion counting that very weak signals might require.This approach is compatible with FABISIMS as continuous high intensity beams of primary atoms or ions are readily produced in turn giving secondary ion beams of reasonable 33 D. H. Williams A. F. F. Findeis S. Naylor and B. W. Gibson J. Am. Chem. SOC.,1987 109 1980. 34 A. Sandstroem and J. Chattopadhyaya J. Chem. SOC. Chem. Commun. 1987 862. 3s J. Sunner A. Morales and P. Kebarle Anal. Chem. 1988 60,98. 36 J. Sunner A. Morales and P. Kebarle Anal. Chem. 1987 59 1378. 31 F. Engelke J. H. Hahn W. Henke and R. N. Zare Anal. Chem. 1987 59 909. 38 B. Spengler M. Karas U. Bahr and F. Hillenkamp J. Phys. Chem. 1987 91 6502. H. J. Neusser Inf. 1.Mass Spectrom.Ion Processes 1987 79 141. 20 M. A. Baldwin intensity. However LDMS and PDMS are both pulsed techniques neither of which is ideally suited to the production of continuous signals. For LDMS this is determined by the duty cycle of the laser whereas for 252Cf fission fragment PDMS the flux of high energy particles is low and the resulting ion current is intermittent. The advent of LDMS and PDMS has led to the renascence of a pulsed mass analysis technique namely time-of-flight (TOF). For LDMS the TOF analysis can be readily synchron- ized with the firing of the laser and for fission-fragment PDMS the detection of a second fission fragment is used to gate the mass analyser. There is no theoretical upper mass limit for TOFMS which is obviously advantageous for high mass analysis.However TOF instruments have tended to have low resolving power due to a spread of ion velocities. This disadvantage can be remedied by an ion optical device known as a reflectron and a REMPI TOFMS instrument has been described having a mass resolution of 10OO0.40 An alternative is to form the ions in a burst and then to trap them allowing the build-up of sufficient ions in the trap before analysing them. This topic will be covered in the following section. 3 Ion Trapping Ion traps have recently been re~iewed.~' A very simple ion trap has been marketed in recent years as a GC detector.42 This is based on the two-dimensional quadrupole trap consisting of a circular electrode with two end caps but this has not been employed in studies of high mass or involatile species.However the Fourier transform ion cyclotron resonance instrument (FTICR or FTMS) has been used in this area. Very low kinetic energy ions can be stored in the cell and built up over a period of time. These are then excited into circular orbits by a pulse of radiation and their cyclotron frequencies specify their masses with high precision. Unlike a scanning sector instrument this technique allows all masses to be monitored simul- taneously and thus enjoys the Fellget advantage. (Early sector instruments used photographic detection at a focal plane and thereby allowed simultaneous detection of a range of masses. This has been up-graded by one manufacturer of double focusing instruments with a multichannel electro-optical detector allowing all masses within a 4% spread to be monitored.Recording a 200 amu wide spectrum in region of the molecular ion of bovine insulin with this detector gave an eightfold improve- ment in signal-to-noise compared with conventional scanning.43) LD is the most frequently used ionization method for high mass FTMS. Shomo et al. obtained comparable or better sensitivity with LDFTMS compared with FD and FABMS using sector instruments for a number of drugs of low volatility in the mass range 404-819 and they obtained mass measurements accurate to *0.003 am^.^^ Tabet et al. obtained the spectrum of Leu-enkephalin by 252Cf PDFTMS,45 and McLafferty et al. have experimented with the use of both caesiurn 4" K. Waiter U. Boesl and E.W. Schlagg Int. J. Mass Specrrom. fon Processes 1986 71 309. 41 J. Allison and R.M. Stepnowski Anal. Chem. 1987 59 1072A. 42 Finnigan MAT San Jose California. 43 J. S. Cottrell and S. Evans Anal. Chem. 1987 59 1990. 44 R. E. Shomo A. G. Marshall and R. P. Lattimer fnt. J. Mass Spectrom. fon Processes 1986 72 209. 45 J. C.Tabet J. Rapin M. Poreti and T. Gaumann Chimia 1986 40 169. Phvsical Methods and Techniques -Part (ii) Mass Spectrometry 21 ion SIMS,4h and 252Cf PDMS with FTMS.47 It is desirable to use dry samples with FTMS as the vapour pressure of the liquid matrix gives an unacceptably high pressure in the cell. With Cs' SIMS from dry samples they showed that addition of the tripeptide glutathione enhanced the spectra.36 With 52Cf PDMS they were able to produce spectra for compounds of masses up to 2000 by desorption from mylar or nitrocellulose but the trapping efficiency was very low and was estimated to be only 0.1% giving poor sensitivity compared with PDMS using TOF analysis.47 The use of glutathione or nitrocellulose was already established in PDTOFMS and it has been shown for peptides up to RMM 14 000 that ions desorbed from nitrocel- lulose have lower internal energies and they are less susceptible to unimolecular decay.48 Hunt et al.have shown that with FTMS it is possible to detect protonated molecular ions of cytochrome c (RMM 12384) formed externally to the analyser cell by FAB i~nization.~~ LDFTMS has been compared favourably with FAB/SIMS for the analysis of polymers more regular polymer distributions being obtained with less fragmentation and less mass dis~rimination.~~ Ion traps are also used with the more traditional ionization techniques of EI and CI and the quadrupole ion trap has shown itself to be a very versatile device for studies in ion chemistry as well as for conventional mass spectrometry.It is possible to select and trap ions of a specific mass. Trapping the ions for long periods of time can allow low energy collisions with background gases to bring about collisional dissociation with high efficiency. Approximately 2 ev is deposited in the ions on collision and sequential reactions can be studied." In a similar way it is possible to carry out low pressure chemical ionization extended trapping times leading to the elimination of peaks due to electron impact.The use of a range of reagent gases with selective ion trapping allows versatile ion/molecule studies.52 Another advantage of ion trapping may prpve to be important in tandem mass spectrometry of high mass ions. Analysis of high mass compounds by techniques such as FAB provide protonated molecular ions that show little tendency to fragment. Small groups may be lost from these ions to give some peaks in the upper part of the spectrum but there are usually very few structurally diagnostic peaks. This is a particular problem for compounds of RMM greater than 2000 but even below 1000 the fragmentation of these even-electron ions is unpredictable. Tandem mass spectrometry allows ions of a particular mass to be selected the ions to be energized by collisions with an inert gas and the resulting fragments to be analysed.Biemann et al. have used this technique for the analysis of several unknown peptides up to RMM 3000 and have reported that the fragmentation obtained in this way is far more predictable and comprehensive often allowing unambiguous structure determi- nation.53 They have demonstrated superior performance for FAB/tandem mass 46 I. J. Amster J. A. Loo J. J. P. Furlong and F. W. McLafferty Anal. Chem. 1987 59 313. 47 J. A. Loo E. R. Williams 1. J. Amster J. J. P. Furlong B. H. Wang F. W. McLaffertl B. T. Chait and F. H. Field And. Chem. 1987 59 1880. 4n B. T. Chait Int. J. MASSSpectrom. Ion Processes 1987 78 237.49 D. F. Hunt J. Shabanowitz J. R. Yates 111 N.-Z. Zhu D. H. Russell and M. E. Castro Proc. Narl. Acad. Sci.,1987 84 620. 50 L. M. Nuwaysir and C. L. Wilkins Anal. Chem. 1988 60 279. J. N. Louris R. G. Cooks J. E. P. Syka. P. E. Kelley G. C. Stafford Jun. and J. F. J. Todd AnalChem.. 1987 59 1677 52 J. S. Brodbelt J. N Louris and R. G. Cooks Anal. Chem. 1987 59 1278. 53 S. A. Martin and K. Biemann Inr. .I. Mass Spectrom. Ion Processes 1987 78 213. 22 M. A. Baldwin spectrometry over acid hydrolysis/FABMS for the analysis of substance P,54 and they have described a computer aided graphics-based interactive interpretation system that rapidly calculates all possible amino acid sequences corresponding to observed mass differences in the collision ~pectra.~' However as the mass of ions increases there are more and more vibrational modes available.Consequently there is increasing difficulty in introducing sufficient energy in a collision to induce fragmentation within the lifetime of the experiment so as to obtain coherent structural information. Ion trapping has been recommended as a means of extending the time available for the fragmentation of large molecules thereby increasing the chance of obtaining structural inf~rmation.'~ 4 Chromatography/Mass Spectrometry The combination of chromatography and mass spectrometry continues to be an important growth area. Despite the increasing importance of HPLC/MS GC/MS is still widely used; in reviewing GC/MS over a two-year period Evershed has listed nearly 500 papers that contain novel development^.'^ GC/MS is generally more sensitive than LC/MS e.g.for the determination of cortisol in blood serum at physiological levels GC/ MS gave better precision and better sensitivity. However after an initial extraction the GC assay involved HPLC fractionation and then derivatization to form the bis(methy1)-oxime tris(trimethylsily1)-ether derivative whereas thermospray HPLC was carried out on the crude extract." This is a common situation in that GC/MS gives greater accuracy and precision whereas LCMS gives a quicker analysis. The obvious target for improvement in LC/MS is sensitivity. GC/MS has been compared favourably with radioimmunoassay in terms of sensitiv- ity and linearity of response for the assay of leukotriene B4 in human neutrophil samples.59 Several interface techniques are available for LC/ MS including the continuous flow FAB described earlier.Thermospray which was described in the last review in this series,' has become established as the most widely used method. This relies on desolvation of preionized species and favours polar materials and reversed-phase chromatography although additional ionization can be achieved in the gas phase by corona discharge or electron bombardment from a heated filament. An alternative ionization method is atmospheric pressure ionization (API) with an ion spray interface which reportedly can give high sensitivity. Widely varying flow rates can be accommodated and the limits of detection have been reported as 10 ng for full scans and 1Opg with single ion monitoring.60 This technique has been compared with thermospray/API.6' Chromatography/mass spectrometry is widely used in biochemistry phar- macology drug metabolism etc.and it is commonplace to use heavy isotope-labelled 54 H. A. Scoble S. A. Martin and K. Biemann Biochem. J. 1987 245 621. 55 H. A. Scoble J. E. Biller and K. Biemann Fresenius' Z. Anal. Chem. 1987 327 239. 56 P. Demirev J. K. Olthoff C. Fenselau and R. J. Cotter Anal. Chem.. 1987 59 1951. 57 R. P. Evershed in ref. 3 p. 196. 58 S. J. Gaskell K. Rollins R. W. Smith and C. E. Parker Biomed. Enuiron. Mass Spectrom. 1987 14,717. 59 W. R. Mathews G. L. Bundy M. A. Wynalda D. M. Guido W. P. Schneider and F. A. Fitzpatnck Anal.Chem. 1988 60,349. 60 A. P. Bruins T. R. Covey and J. D. Henion Anal. Chem. 1987 59 2642. 61 T. R. Covey A. P. Bruins and J. D. Henion Org. Mass Spectrom. 1988 23 178. Physical Methods and Techniques -Part (ii) Mass Spectrometry 23 standards as these are readily identified by mass spectrometry e.g. isotope dilution can be used to monitor the kinetics of metabolism of endogenous compounds and thermospray LC/MS has been used for quantitative analysis in a number of such studies.62 The binding of caffeine to human serum albumin showed a significant isotope effect when deuterated standards were employed. This points to ways in which the biochemical mechanisms can be detected but it also provides a warning against the ubiquitous use of labelled compounds as standards which are assumed to differ from the compounds of interest only in terms of mass.63 The powerful combination of GC/FTIR/MS has been and has been described using an ion trap as the mass analy~er.~~ The much greater sensitivity of mass spectrometry over Fourier transform infra-red spectroscopy means that a sample split of the GC eluent of perhaps 50:l is required in favour of the FTIR.Chromatographic resolution is retained in the mass spectrum by splitting rather than having a flow-through IR cell preceding mass spectrometry. The techniques of IR and MS provide highly complementary information and the greatest advantage is obtained through searching separate IR and MS databases.64 ITMS has also been interfaced with chromatography including supercritical fluid chromatography (SFC).As explained above it is desirable to maintain high vacuum in the FTMS analyser cell and this was achieved through the use of a differentially pumped dual-cell instrument which gave a one hundred-fold pressure differential in the ionization and analysis regions.66 The analysis of caffeine has been described using SFC/FTMS with ‘self CI’ i.e. chemical ionization in which the primary ions collide with un-ionized sample molecules to produce new ionic species.67 SFC/MS is a technique of great promise that has yet to become firmly established. A recent review lists more than 100 literature references although the authors concede that early hopes that LC/ MS would be replaced by SFC/ MS were ‘naive’.68 5 Novel Neutral Species Mass spectrometry is well established as a means of studying the unimolecular and bimolecular chemistry of ions free of the solvent effects associated with ~olutions.~~ Tandem mass spectrometry allows oxidation and reduction reactions to be studied through collisional charge exchange (CE) reactions ‘charge permutation reac-tion~’,’~ although it is normally necessary for both the reactants and the products to be ions.More recently it has become possible to study neutral species which have been produced from an ion beam in a mass spectrometer in one of a number of ways (i) by unimolecular fragmentation of an ion giving a neutral fragment (ii) by a similar collision-induced reaction with an atom such as He or (iii) by collisional CE with an atom such as Hg or Xe which neutralizes an ion.The neutrals thus 62 A. L. Yergey N. V. Esteban and D. J. Liberato Biomed. Environ. Mass Spectrom. 1987 14 623. 63 Y. Cherrah J. B. Falconnet M. Dessage J. L. Brazier R. Zini and J. P. Tillement Biomed. Environ. Mass Spectrom. 1987 14 653. 64 C. L. Wilkins Anal. Chem. 1987 59 571A. 65 E. S. Olson and J. W. Diehl Anal. Chem. 1987 59 443. 66 D. A. Laude Jun. S. L. Pentony Jun. P. R. Griffiths and C. L. Wilkins Anal. Chem. 1987 59 2283. 67 E. D. Lee J. D. Henion R. B. Cody and J. A. Kinsinger Anal. Chem. 1987 59 1309. 68 R. D. Smith H. T. Kalinoski and H. R. Udseth Mass Spectrom. Rev. 1987 6 445. 69 M. A. Baldwin in ref. 3 p. 59. 70 J. H. Beynon R. K. Boyd and A.G. Brenton in ref. 2 p. 437. 24 M. A. Baldwin produced are then ionized by CE with a molecule such as O2and analysed by mass spectrometry. If the neutrals are formed by process (iii) and then reionized this can properly be termed neutralization/reionization mass spectrometry (NRMS). However if He is used for the ionization step this brings about fragmentation as well as ionization and this may be called collision-induced dissociative ionization (CIDI). Yet another term that has been used to describe such experiments is neutralized ion beam spectroscopy (NIBS). Thus it can be seen there is anarchy amongst the acronyms and the suggestion that all such experiments should be described by the terms NRMS71 is perhaps as welcome as it is inaccurate.There have been two recent reviews of this which highlight the very interesting chemistry that can be studied. A number of compounds not obtainable by conventional chemical means can be formed in the gas phase by neutralization of the corresponding radical cations. Vinyl alcohol CH2=CHOH hydroxymethylcarbene CH,COH and methoxycarbene HCOCH3 which are isomers of acetaldehyde can be formed in this way e.g. the elimination of ethene from ionized cyclobutanol gives the radical cation of vinyl alcohol which is then subjected to CE. There are high energy barriers to the unimolecular isomerization of the various neutrals and these have been identified and characterized by NRMS7’ Many ions formed by indirect processes such as rearrangement do not conform to normal valency rules and when neutralized they can provide hypervalent species.The interpretation of the observed data is not without controversy for a number of these species. The CH; radical has been reported to exist with a lifetime > 3.9 x s although this conflicts with earlier results.73 Hi can exist in metastable and electronically excited dissociative states.74 Isotope effects can play a significant role in increasing the lifetimes of such species e.g. D,O’ is substantially more stable than H30’.75 Reference was made in the preceding review’ to the species CH2ClH. That this exists as a stable radical cation is undisputed and new MS/MS/MS experiments from Wesdemiotis et al. reinforce the claim that neutralization with Hg gives a neutral of the same structure that is stable for at least 1 ps.76This stability is disputed by Hop et al.who have neutralized the ions with Xe and claim isomerization occurs to give CH3C1.77 A6 initio molecular orbital calculations predict that the ground-state neutral exists only in a shallow well and therefore can readily is~merize.’~ If neutral species such as this do exist they may be regarded as ylides i.e.-CH,-ClH+. NRMS spectra have been obtained for species that could possibly exist as betaines -CH2CH20Hl and -CH2CH2NHl or as dipole-alkene c~mplexes.’~ 71 C. Wesdemiotis and F. W. McLafferty Chem. Rev. 1987 87 485. 72 J. K. Terlouw and H. Schwarz Angew. Chem. Int. Ed. Engl. 1987 26 805. 73 W. J. Griffiths F. M. Harris A. G. Brenton and J. H. Beynon Inr.J. Mass Spectrorn. Ion Processes 1986 74 317. 74 G. I. Gellene and R. F. Porter Int. J. Mass Spectrom. Ion Processes 1985 64 55. 75 G. I. Gellene and R. F. Porter J. Chem. Phys. 1984 81 5570. 76 C. Wesdemiotis R. Feng M. A. Baldwin and F. W. McLafferty Org. Mass Spectrom. 1988 23 166. 77 C. E. C. A. Hop J. Bordas-Nagy J. L. Holmes. and J. K. Terlouw Org. Mass Spectrom. 1988 23 155. 78 B. F. Yates W. J. Bouma and L. Radom J. Am. Chem. Soc. 1987 109 2250. 79 C.Wesdemiotis P. 0. Danis R. Feng J. Tso and F. W. McLafferty J. Am. Chem. SOC.,1985 107 8059.