A combined magnetic sector-time-of-flight instrument has been constructed and used with both continuous (liquid secondary ion mass spectrometry and electrospray ionization) and pulsed (matrix-assisted laser desorptionhonization) ion sources. The design includes provision for both low-and high-energy gas collision induced decompositions and orthogonal acceleration of the product ions into the time-of-flight mass analyser. Tandem mass spectra may be acquired for singly and multiply charged precursor ions and reliable mass measurements are easily obtained. Product-ion mass spectra are quite reproducible regardless of any instability in the precursor ion beam intensity.Instrumentation for molecular structure determination studies by tandem mass spectrometry (MSlMS) should ideally be able to provide unit mass selection of the precursor ion with high transmission, and to acquire full product-ion mass spectra within chromatography time scales at the highest achievable sensitivity. The instrumentation should yield reliable mass assignments of product ions, and in addition should ideally operate with both continuous and pulsed ion sources. To date, gas-phase ion/molecule collisions have been the most commonly used and generally most successful method for generating structurally informative decomposition products. Low and high energy collision-induced decompositions (CID) can provide different and complementary information and MSlMS instrumentation should ideally be able to accommodate both types. In particular, high energy CID of peptides generally implies the capability to induce the side-chain cleavages (i.e. 'd', 'v' and 'w' sequence ions) which may be used ?o distinguish isobaric amino acid residues such as leucine and isoleucine, or lysine and glutamine.In 1992 we suggested that the use of a magnetic sector instrument as the first mass analyser (MS-1) of an MUMS instrument offers unit mass precursor ion selection with high transmission, whilst the use of a time-of-flight (TOF) instrument as MS-2 offers the potential for high sensitivity and fast acquisition of full product-ion spectra.' Such an instrument could be smaller and more convenient to operate than foursector instruments equipped with array detectors and yet be comparable in performance. This paper describes a combined magnetic sector-TOF instrument which has been designed for structural determination studies by MSlMS, and which includes provision for both low-and high-energy gas collision-induced decompositions. COLLISION-INDUCED DECOMPOSITIONS (CID)Gas phase ion/molecule collisions have, so far, been the most widely studied method for generating structurally informative decomposition products. To generate * Author for correspondence.side-chain cleavage ions, such as the 'd' fragment ions in peptides, it has been shown that a minimum value of 'centre-of-mass' collision energy (E,,,) of 20 -25 eV is Such collision processes are generally termed 'high energy' collisions. The 'centre-of-mass' collision energy is defined in Eqn ( l ) , where Elab is the laborat...

Hitherto, exact mass measurement experiments have usually been performed using high-resolution mass spectrometry. However, under the right circumstances, measurements with comparable accuracy may be made at low resolution. Here we demonstrate the use of a low-resolution single-quadrupole mass spectrometer to accurately mass measure organic samples analyzed by electrospray ionization, using a variety of glycol polymers for internal calibration. Results are presented from 11 samples which yield molecular signals in the m/z range 190-750, including data for positive, negative, and multiply-charged sample and reference ions. Replicate determinations of the masses of 12 ions gave values within 4.5 ppm (1.1 millimass units, mmu) of their calculated values, with standard deviations no larger than 3 ppm (1.7 mmu). From a total of 88 individual 1 min measurements, 83 were within 5 ppm, and 87 within 2 mmu of the theoretical mass. The accuracy, precision, and sensitivity shown here are comparable to those achievable using high-resolution mass spectrometers, with the added benefits of simpler instrumentation and analytical technique afforded by the quadrupole mass analyzer.

SummaryThe use of HPLC-NMR-MS for the detection and identification of the metabolites of ibuprofen present in a solid phase extract of human urine is described. Gradient reversed-phase HPLC was used to separate the components present in the extract, which were then characterised by a combination of stopped-flow 1H NMR and on line electrospray-MS. This approach led to the rapid identification of the known phase 1 human metabolites of ibuprofen, including hydroxy-and carboxy-metabolites, together with their respective glucuronide conjugates. In addition a probable artefact resulting from the dehydration of one of the side chainhydroxylated glucuronides was also identified.

1. The metabolic fate and urinary excretion of 2-bromo-4-trifluoromethylaniline has been studied in rat using 19F-NMR spectroscopic and directly coupled HPLC-NMR-MS methods. The compound was dosed to Sprague-Dawley rats (50 mg kg-1, i.p.) and urine collected over 0-8, 8-24 and 24-48 h post-dosing. 2. A total urinary recovery of 53.5 +/- 7.0% of the dose was achieved up to 48 h after dosing. The major metabolite in the urine was identified as 2-amino-3-bromo-5-trifluoromethylphenylsulphate accounting for a total of 35.7 +/- 6.2% of the dose. 3. Further metabolites detected were 2-bromo-4-trifluoromethylphenylhydroxylamine-1V-glucuronide (9.7 +/- 0.2% of the dose), 2-bromo-4-trifluoromethylaniline-N-glucuronide (3.0 +/- 0.3%) and 2-amino-3-bromo-5-trifluoromethylphenylglucuronide (2-St 0-4). Minor metabolites, including 2-bromo-4-trifluoromethylphenylhydroxylamine-O-glucuronide, 2-amino-3-bromo-5-trifluoromethylphenol and 2-bromo-4-trifluoromethylphenylsulphamate, in total accounted for 2.3 +/- 0.9% of the dose. 4. Directly coupled HPLC-NMR-MS and 19F-NMR spectroscopy proved to be efficient techniques for the unequivocal and rapid determination of the urinary metabolic fate and excretion balance of fluorinated xenobiotics without the need for radiolabelling.

Applications of liquid Chromatography/nuclear magnetic resonance/mass spectrometry have recently been described which have shown the benefits of this double hyphenation for identification and structure determination. However, the combination of HPLC/NMR and HPLC/MS is not without its difficulties and, as the number of applications of this technology has increased in our laboratory, we have gained considerable insight into the pitfalls that can be encountered when attempting to perform HPLC/NMR/MS. Guidelines are presented to facilitate the effective combination of HPLC with NMR and MS in tandem together with some suggestions as to how the technique may develop. # 1998 John Wiley & Sons, Ltd. Received 24 August 1998; Accepted 24 August 1998 HPLC/MS has been employed for many years but only since the advent of electrospray ionization has it become a truly robust and routine method for the analysis of mixtures. HPLC/NMR is, by comparison, a relatively recent introduction; however, technological advances in this area have resulted in HPLC/NMR being extensively employed in the pharmaceutical industry.1,2 Now that both of these spectroscopic techniques have been coupled to HPLC, it has for the first time become possible to acquire both NMR and MS data simultaneously from a single chromatographic analysis. A number of groups have begun to investigate the technique of HPLC/NMR/MS (e.g see Refs. 3-9) for the analysis of mixtures, especially those of pharmaceuticals and drug metabolites. However, because of the novelty of the approach, published work has tended to concentrate on the results of these studies without detailed discussion of the practical problems associated with this technique. Here we describe our own experiences of HPLC/NMR/MS and offer our suggestions as to the direction of future developments. EXPERIMENTAL HPLC conditionsChromatography was performed using a variety of C18 250 Â 4.6mm columns, attached to a standard Bruker LC system comprising Bruker LC22 pump, autosampler, UV detector and BPSU-12 collector (Bruker Spectrospin, Coventry, UK). For the majority of experiments the outlet of the UV detector was connected immediately to a splitter which directed the flow to the NMR and MS in the ratio 95:5. Both linear and gradient elution were used with 0.1% formic or trifluoroacetic (TFA) acids in 2 H 2 O (99.9 atom %, Fluorochem, Glossop, UK) and acetonitrile (Riedel de Haen, Seelze, Germany) at a flow rate of 1.0 mL/min. UV detection was performed at 254nm. The splitter was a simple Valco stainless steel T-piece and all transfer lines were in PEEK. A small length of silica capillary was used to connect the PEEK tubing to the mass spectrometer and to adjust the split ratio. The use of an all silica capillary transfer line to the mass spectrometer was not investigated as this has been known to act as a chromatographic stationary phase and alter the elution times and orders of the peaks. The samples used in this study were from various sources and are described, where appropriate, in the text. NMR spectrosc...