Selected ion flow tube mass spectrometry (SIFT-MS) is now recognized as the most versatile analytical technique for the identification and quantification of trace gases down to the parts-per-trillion by volume, pptv, range. This statement is supported by the wide reach of its applications, from real-time analysis, obviating sample collection of very humid exhaled breath, to its adoption in industrial scenarios for air quality monitoring. This review touches on the recent extensions to the underpinning ion chemistry kinetics library and the alternative challenge of using nitrogen carrier gas instead of helium. The addition of reagent anions in the Voice200 series of SIFT-MS instruments has enhanced the analytical capability, thus allowing analyses of volatile trace compounds in humid air that cannot be analyzed using reagent cations alone, as clarified by outlining the anion chemistry involved. Case studies are reviewed of breath analysis and bacterial culture volatile organic compound (VOC), emissions, environmental applications such as air, water, and soil analysis, workplace safety such as transport container fumigants, airborne contamination in semiconductor fabrication, food flavor and spoilage, drugs contamination and VOC emissions from packaging to demonstrate the stated qualities and uniqueness of the new generation SIFT-MS instrumentation. Finally, some advancements that can be made to improve the analytical capability and reach of SIFT-MS are mentioned.

There is interest in the development of exhaled breath tests for the detection of lower airway infection in children with cystic fibrosis. The first stage of this process is the identification of volatile organic compounds (VOCs) released into the gas phase by CF pathogens that can be used as breath test biomarkers. Selected ion flow tube mass spectrometry (SIFT-MS) is ideally suited to these in vitro studies as it allows simultaneous quantification of multiple VOCs in real time. We review a decade of in vitro experiments using SIFT-MS to analyse the VOCs released by respiratory pathogens. This includes identification and quantification of VOCs and the investigation of the in vitro factors that affect their production. We also report on how our culture methodology has been refined over the years to better account for variations in bacterial mass. Finally, we discuss how these in vitro findings have been translated into clinical trials and assess possible future applications.

Introduction Pseudomonas aeruginosa (PA) is one of the most important pathogens in cystic fibrosis (CF). Although there is a wealth of data about the prevalence of chronic PA infection, there is a paucity of evidence about the incidence of new PA infection. Methods The SPACE (Sensitivity and specificity of PA detection using the hydrogen Cyanide concentration of Exhaled breath) study investigated if exhaled breath hydrogen cyanide is an early marker of PA infection in children with CF. Breath samples, clinical data and microbiology samples were collected at each out-patient appointment from a large cohort of children with CF who had not isolated PA for >12 months. This abstract reports the PA acquisition data. Results 233 children were followed for a median of 2.0 (1.7–2.3) years. The median (IQR) age was 8.0 (5.0–12.2) years. 71 children isolated PA during the study period. The incidence rate (95% CI) of new PA infections was 0.15 (0.10–0.22) cases per patient year for those that had never previously isolated PA and 0.19 (0.13–0.27) cases per patient year for those that had been free from PA for >12 months. This rate varied between 0.08 (0.04–0.18) and 0.28 (0.14–0.49) cases per patient year at the 8 recruiting centres. 42% of children were asymptomatic at the time of PA acquisition. The median (IQR) number of antibiotic courses per patient year varied between the centres: 0.6 (0.2–1.3) to 3.6 (3.1–4.2) for oral and 0.0 (0–0) to 0.4 (0–1.2) for intravenous. Conclusions This is the first prospective study to report the incidence of new PA infection in a large cohort of children with CF, considered to be free of PA airway infection. Incidence rate was higher in children who had isolated PA previously. The variation between centres is not easily explained and needs further investigation. Acknowledgments We would like to thank the Principal Investigators, research nurses and co-ordinators at each of the recruiting centres as well as the children and their families.

There is increasing interest in using the cyanogenic properties of Pseudomonas aeruginosa to develop a nonmicrobiological method for its detection. Prior to this, the variation in cyanide production between different P. aeruginosa strains needs to be investigated.Hydrogen cyanide (HCN) released into the gas phase by 96 genotyped P. aeruginosa samples was measured using selected ion flow tube-mass spectrometry after 24, 48, 72 and 96 h of incubation. The HCN produced by a range of non-P. aeruginosa cultures and incubated blank plates was also measured.All P. aeruginosa strains produced more HCN than the control samples, which generated extremely low levels. Analysis across all time-points demonstrated that nonmucoid samples produced more HCN than the mucoid samples (p=0.003), but this relationship varied according to strain. There were clear differences in the headspace HCN concentration for different strains. Multivariate analysis of headspace HCN for the commonest strains (Liverpool, Midlands_1 and Stoke-on-Trent, UK) revealed a significant effect of strain (p < 0.001) and a borderline interaction of strain and phenotype (p=0.051).This evidence confirms that all P. aeruginosa strains produce HCN but to varying degrees and generates interest in the possible future clinical applications of the cyanogenic properties of P. aeruginosa.

Aims In cystic fibrosis, Pseudomonas aeruginosa (PA) infection is associated with significant morbidity and mortality. Unfortunately, inpatients who cannot expectorate sputum current diagnostic methods are unreliable, unpleasant or invasive. This has lead to increased interest in using the cyanogenic properties of PA to develop a non-microbiological method for its detection. Prior to this development it needs to be determined if cyanide production varies according to PA strain. Methods The hydrogen cyanide (HCN) released into the gas phase by 96 genotyped PA samples was measured using Selected Ion Flow Tube Mass Spectrometry after 24, 48, 72 and 96 h of incubation. The HCN produced by a range of non-PA cultures and incubated blank agar plates was also measured. Results The 96 samples included 26 different strains; four were previously described epidemic strains (Liverpool, Midlands1, Midlands2 and Stoke). Extremely low levels of HCN (<10 parts per billion) were produced by the control samples. All the PA strains produced more HCN than the controls. Across all four time points the non-mucoid samples produced more HCN than the mucoid samples (p=0.003), they also produced more HCN after 24 h when the time points were analysed separately (p=0.008). When samples were separated according to strain and phenotype, some strains (Liverpool and Stoke) continued to show higher HCN production by non-mucoid samples but others (Midlands1) showed higher production by mucoid samples. Clear differences were apparent in the HCN production between the different strains. Multivariant analysis of the three commonest strains (Liverpool, Midlands1 and Stoke) revealed a significant effect of strain (p<0.001) and a borderline interaction of strain and phenotype (p=0.051). Discussion This supports previous studies showing PA is one of a limited number of organisms to produce cyanide and SIFT-MS is a sensitive and reproducible way of analysing this. It is the first study to demonstrate that cyanide production is dependent on the PA strain and that the effect of phenotype may vary between strains. As all the PA strains produced more HCN than the controls, it creates further interest in using the cyanogenic properties of PA to develop a diagnostic test.

Using selected ion flow tube mass spectrometry (SIFT-MS), measurements have been made of the levels of several metabolites in the exhaled breath of 200 healthy school children. Thus, concentration distributions of each metabolite have been obtained for the first time in the paediatric age range. The median values (in parentheses) of the concentrations in parts per billion, ppb, were ammonia (628), acetone (297), methanol (193), ethanol (187), isoprene (37), propanol (16), acetaldehyde (23) and pentanol (15). Hydrogen cyanide was not present in the breath above the detection limit of 2 ppb in the majority of subjects. The water vapour level (humidity) of the breath samples was routinely measured as a check on the sample integrity. Such data are essential if SIFT-MS breath analyses are to be used as a clinical tool to aid diagnosis and/or as a monitor of disease in children. The levels of metabolites usually followed a log-normal distribution and the levels of some compounds were similar to those obtained previously in adults. Lower values were found in the levels of acetone, ammonia, methanol and isoprene. There were no major variations in relation to gender. Some metabolites showed significant variation in relation to age and body mass index. To our knowledge, these are the first measurements of exhaled mouth breath pentanol levels. The median ammonia levels in mouth-exhaled breath of these children decreased with age, whereas in older adults, ammonia has been shown to increase with age. Breath acetone levels were significantly increased for those who had not eaten for more than 6 h prior to providing the breath sample, although dietary control was not a mandatory aspect of the protocol.

We have carried out a selected ion flow tube mass spectrometry (SIFT-MS) study of the concentrations of the sulfur-containing compounds H(2)S (using H(3)O(+) precursor ions), CH(3)SH (H(3)O(+)), (CH(3))(2)S (O(2)(+)), (CH(3))(2)S(2) (NO(+)) and CS(2) (O(2)(+)) in single exhalations of mouth-exhaled breath and nose-exhaled breath and in the static gas in the oral cavity for two healthy volunteers. The primary purpose of the study was to show how compounds present in breath at levels as low as a part per billion (ppb) can be identified and quantified if the overlap of 'impurity' isobaric ions with the analytical product ions for each trace compound is identified and accounted for. The H(2)S measurements are straightforward using H(3)O(+) precursor ions, since no overlapping ions are recognized and its breath concentration is relatively high at typically 20-70 ppb. Thus, its concentration distribution for two healthy volunteers has been obtained over a period of a few weeks. The situation is very similar for CH(3)SH, but to analyse this compound we had to study the kinetics of its reactions with the SIFT-MS reagent ions H(3)O(+), NO(+) and O(2)(+) in order to provide the required kinetics library data for this compound. It is seen that CH(3)SH, (CH(3))(2)S and (CH(3))(2)S(2) are present in the mouth breath/cavity at lower levels of <10 ppb. The measurements of the levels of H(2)S and these compounds in the nose-exhaled breath and the closed mouth indicate that they are largely produced in the oral cavity, although there is some indication that (CH(3))(2)S is partially systemic in these two volunteers. It was not possible to quantify CS(2) in the breath because of serious interference (overlapping ions) due to the presence of carbon dioxide and acetone that inevitably occur in exhaled breath. This study paves the way for the accurate analysis of these sulfur compounds in halitosis and potentially for probing the diseased state, especially liver disease, by breath analysis. To demonstrate the simplicity of measuring these compounds when they are present at levels of about 100 ppb and greater, data are presented on the emissions of these sulfur-containing compounds from Pseudomonas bacterial cultures in vitro.

The ingestion of relatively large doses of the vitamin niacin by healthy volunteers results in a reddening of the skin, a skin 'flush'. Thus, we have carried out a study of the breath metabolites of two healthy volunteers following (i) the ingestion of 200 mg of immediate-release niacin, (ii) as (i) but preceded by the ingestion of 325 mg of aspirin that diminishes the skin 'flush', (iii) ingestion of 500 mg of slow-release niacin. On-line breath analysis was carried out using selected ion flow tube mass spectrometry, SIFT-MS. The interesting new observation is that the breath ammonia levels of both volunteers clearly increased following (i) and (ii), and an obvious skin flush did occur following (i) but not following (ii). The slow-release niacin (iii) did not result in a flush and the breath ammonia levels increased more slowly and did not reach the higher levels produced by (i) and (ii). The results of these experiments demonstrate that breath ammonia levels are dependent on the blood/plasma levels of niacin, but are not directly related to the flushing phenomenon, and that the observed increases in blood/breath ammonia levels are consistent with current knowledge of the metabolic pathways of niacin. The parallel measurements of breath isoprene are presented, which demonstrate the quality of breath analyses that can be achieved using SIFT-MS.

Selected-ion flow-tube mass spectrometry (SIFT-MS) has been used to monitor the volatile compounds in the exhaled breath of 30 volunteers (19 male, 11 female) over a 6-month period. Volunteers provided breath samples each week between 8:45 and 13:00 (before lunch), and the concentrations of several trace compounds were obtained. In this paper the focus is on ethanol and acetaldehyde, which were simultaneously quantified by SIFT-MS using H3O+ precursor ions. The mean ethanol level for all samples was 196 parts-per-billion (ppb) with a standard deviation of 244 ppb, and the range of values for breath samples analysed is 0 to 1663 ppb. The mean acetaldehyde level for all samples was 24 ppb with a standard deviation of 17 ppb, and the range of values for breath samples analysed is 0 to 104 ppb. Background (ambient air) levels of ethanol were around 50 ppb, whereas any background acetaldehyde was usually undetectable. Increased ethanol levels were observed if sweet drink/food had been consumed within the 2 h prior to providing the breath samples, but no increase was apparent when alcohol had been consumed the previous evening. The measured endogenous breath ethanol and acetaldehyde levels were not correlated. These data relating to healthy individuals are a prelude to using breath analysis for clinical diagnosis, for example, the recognition of bacterial overload in the gut (ethanol) or the possibly of detecting tumours in the body (acetaldehyde).

Samples (210 in total) of broncholaveolar lavages (BALs), obtained from patients hospitalized with pneumonia in various departments of two hospitals, were analysed using the method of solid phase microextraction–gas chromatography (SPME-GC) with FID detection. Up to 20% (9% unequivocally, 11% probably) of these samples was found to contain volatile fatty acids (VFAs) in the series from acetic acid to heptanoic acid. Importantly, the presence of these acids indicates the presence of fermenting anaerobic bacteria, which were not detected by the conventional microbiological examination. Other compounds, namely the heptanol and cyclohexanone, were also detected by this method in some samples. Cyclohexanone occurred almost exclusively in samples from patients receiving intensive care with mechanical ventilation, and is suspected to originate from plastic parts of ventilators. Selected representative samples were also analysed using further methods, namely gas chromatography–mass spectrometry (GC–MS) of native and silylated samples, and selected ion flow tube mass spectrometry (SIFT-MS). These methods confirmed the identities of above mentioned compounds, and detected numerous other compounds tentatively identified as various alcohols, aldehydes, ketones, esters and hydrogen cyanide, HCN. Most of these compounds occurred in small amounts and their origin and diagnostic significance remains uncertain, except, that is, for the HCN, which indicates the presence of Pseudomonas aeruginosa.

The increase in ammonia and ethanol in the exhaled breath stream following mouthwashes by aqueous solutions of urea and sugar (sucrose), respectively, has been investigated by analysing exhaled breath in real time using selected ion flow tube mass spectrometry, SIFT-MS. It is shown that the measured levels of these compounds in the stream of exhaled breath can be much greater than the endogenous levels originating at the alveolar boundary. Thus, it is concluded that without careful preparation, mouth production of these compounds, and other compounds as yet unidentified, can seriously compromise the quantification of truly endogenous trace compounds present in blood and in the alveolar breath, as required for clinical diagnosis, and can probably introduce additional compounds into the breath stream that could seriously mislead breath analysis. The concentrations of both the urea and sucrose solutions used to enhance the ammonia and ethanol levels were larger than normally present in food and drinks and so in most situations such severe enhancements will not occur.

Selected ion flow tube mass spectrometry, SIFT-MS, has been used to monitor the volatile compounds in the exhaled breath of 30 volunteers (19 males, 11 females) over a 6 month period. Volunteers provided breath samples each week between 8:45 am and 1 pm (before lunch), and the concentrations of several trace compounds were obtained. In this paper the focus is on ammonia, acetone and propanol. It was found that the concentration distributions of these compounds in breath were close to log-normal. The median ammonia level estimated as a geometric mean for all samples was 833 parts per billion (ppb) with a multiplicative standard deviation of 1.62, the values ranging from 248 to 2935 ppb. Breath ammonia clearly increased with increasing age in this volunteer cohort. The geometric mean acetone level for all samples was 477 parts per billion (ppb) with a multiplicative standard deviation of 1.58, the values ranging from 148 to 2744 ppb. The median propanol level for all samples was 18 ppb, the values ranging from 0 to 135 ppb. A weak but significant correlation between breath propanol and acetone levels is apparent in the data. The findings indicate the potential value of SIFT-MS as a non-invasive breath analysis technique for investigating volatile compounds in human health and in the diseased state.

Thirty volunteers (19 males, 11 females) were recruited for a 6-month study of the volatile compounds in their exhaled breath using the selected ion flow tube mass spectrometry (SIFT-MS) analytical technique. Volunteers provided weekly breath samples between 8:45 am and 1 pm (before lunch), and the concentrations of several trace compounds were obtained. In this paper, we focus on the isoprene in alveolar breath, which was monitored by SIFT-MS using NO+ precursor ions. The mean isoprene level for all samples was 118 parts per billion (ppb) with a standard deviation of 68 ppb and the range of values for breath samples given is 0–474 ppb. Variability in isoprene levels was similar in most volunteers. Isoprene levels increased immediately after moderate exercise, but returned to normal within 2–3 min for those few volunteers that were investigated. Cholesterol levels analysed for only three of the subjects were not obviously correlated with isoprene concentration in breath. Differences in isoprene levels were not directly correlated to gender, age or body mass index.

Pseudomonas aeruginosa (PA) is associated with a distinctive smell produced by a combination of volatile compounds (VCs). Selected ion flow tube mass spectrometry (SIFT-MS) provides a novel and rapid methodology for rapid, accurate detection of trace quantities (parts per billion; ppb) of VCs in air. We studied the VCs produced by different isolates of PA cultures in vitro from patients with cystic fibrosis. Twenty-one patients with cystic fibrosis provided sputum and cough swab samples for culture. These were used to inoculate blood agar (BA) and Pseudomonas-selective media (PSM). These plates were incubated for 48 hr at 37 degrees C inside sealed plastic bags. The air surrounding the samples after 48 hr (headspace) was analyzed using SIFT-MS. PA growth was commonly associated with the production of significant quantities of VCs, notably hydrogen cyanide gas (HCN). This was detectable in the headspace of 15/22 of PA-positive samples. In contrast, it was only seen in the headspace of 1/13 control samples (6 sterile plates and 7 plates with only mixed upper respiratory tract flora). The concentration of HCN was significantly higher above PA-positive samples than above other bacterial growth (P < 0.01), and in our study, levels of HCN greater than 100 ppb were a sensitive (68%) and highly specific (100%) biomarker of PA. SIFT-MS can detect a range of VCs from PA in vitro. HCN may be a specific indicator of PA infection in vivo, and offers promise as a biomarker for noninvasive detection of PA infection by breath analysis. (c) 2005 Wiley-Liss, Inc.

A selected ion flow tube, SIFT, study has been carried out of the reactions of H3O+, NO+ and O2+ with some volatile organic compounds that are released by bacteria. The major intention is to prepare the way for an extensive study of the emissions from Pseudomonas bacteria in vitro using selected ion flow tube mass spectrometry, SIFT-MS, with a view to detecting the presence of these bacteria in vivo. This requires an extensive SIFT-MS database of the rate coefficients and product ion distributions for the reactions of the above precursor ions with those molecular species that are released by or implicated in the growth of bacteria. A partial list of these molecular species is given. The available SIFT-MS database already includes the kinetic data for the reactions of several of these compounds and the present study supplements this to include 2-methyl-1-butanol and 2-heptanol, 3-methyl-1-butyl acetate, 4-methyl-1,3-pentadiene, and dimethyl trisulphide and dimethyl tetrasulphide. The kinetic data obtained in the present study are compared with those obtained previously for classes of similar compounds.

We have carried out a selected ion flow tube (SIFT) study of the reactions of H3O+, NO+ and O2+• with the following 10 compounds: 2-hydroxyphenol, 2-, 3- and 4-methylphenol (o-, m- and p-cresol, respectively), 4-ethylphenol, 1-phenylmethanol (benzyl alcohol), 1- and 2-phenylethanol, 1,4-benzoquinone and cyclohexanone. The primary purpose of this work was to extend the kinetics database to allow these compounds (M), to be analysed in air by selected ion flow tube mass spectrometry (SIFT-MS). The initial step in all the H3O+ reactions is exothermic proton transfer to produce MH+ ions, which are observed as the only products for seven of the ten reactions, but for the three aromatic alcohols, H2O molecule elimination occurred from the nascent MH+ ions producing the corresponding hydrocarbon ion. This is an essential point to recognise when exploiting proton transfer to analyse these compounds using SIFT-MS and proton transfer reaction mass spectrometry, PTR-MS. NO+ reacts with six of the compounds via non-dissociative charge transfer producing M+ ions and this is a valuable route to their analysis by SIFT-MS. In the case of the NO+/quinone reaction, adduct formation occurs giving NO+M product ions, whilst for the remaining three reactions two or more ion products were formed. All the O2+• reactions proceeded via charge transfer with multiple ion products in most cases. A sample analysis is carried out to indicate the value of simultaneous use of both H3O+ and NO+ precursor ions to analyse a mixture containing some of these compounds.

A study has been carried out of acetonitrile in exhaled breath and in the headspace of urine from several cigarette smokers and several non-smokers using selected ion flow tube mass spectrometry (SIFT-MS). The ion chemistry involved in the detection and quantification of acetonitrile in the gas phase by SIFT-MS using H3O+ ions is outlined. Further to this, experiments have been carried out to determine the Henry’s Law partition coefficients for dilute aqueous solutions of acetonitrile, which allow the determination of liquid phase urinary acetonitrile concentrations from headspace concentrations. The results of this study show that acetonitrile is readily detected by SIFT-MS in the breath and urinary headspace of smokers at levels dependent on the cigarette consumption, but is practically absent from the breath and urine headspace of non-smokers. Exploiting the measured values of the Henry’s Law partition coefficients, the urinary phase concentrations have been deduced. The results of this study show that typical breath concentrations of acetonitrile are within the range 17–124ppb (mean value 69ppb), which are in close agreement with previous studies. The urinary acetonitrile concentrations are within the range 0–150μg/L (mean value 57μg/L), which are close to the concentrations determined previously in blood. These collected data imply that the acetonitrile is equilibrated amongst the body fluids (blood, total body water and urine) and that excretion occurs via both exhaled breath and urine.

Selected ion flow tube mass spectrometry (SIFT-MS) has been used for a detailed study of the daily variations in the acetone and ammonia content of the headspace above urine from a healthy female subject over the course of three separate menstrual cycles. Midstream urine samples were taken every morning prior to any food intake and the headspace subsequently analysed for a number of metabolites. Concurrent with the time of ovulation, a 3-to-12-fold increase in the level of acetone in the urine headspace was observed. The successive peaks in acetone level and the subsequent return to baseline values were mirrored by similar increases in the ammonia levels, but these were a day out of phase. Interestingly, parallel breath analyses at ovulation showed no great increase in either acetone or ammonia above their normal morning levels, suggesting that these metabolites had been removed from the body during the night by the usual metabolic and physiological processes. The results of this study reveal what may be an important phenomenon at the time of ovulation and illustrate the potential and power of online SIFT-MS analysis in this area of research.

We have carried out a selected ion flow tube, SIFT, study of the reactions of H3O+, NO+ and O2+ ions with hydrogen peroxide, H2O2, in the presence of excess water vapour and peroxyacetic acid, CH3C(O)OOH, in the presence of comparable concentrations of acetic acid, CH3COOH. This study was initiated to investigate if these peroxides could be analysed in humid air using selected ion flow tube mass spectrometry, SIFT-MS, using the above precursor ions. Rate coefficients and product ions have been determined for the NO+ and O2+ reactions with H2O2 molecules (H3O+ ions do not react at a measurable rate with H2O2 molecules) and for the rapid reactions of H3O+, NO+ and O2+ with CH3C(O)OOH molecules. It turns out that both H3O+ and O2+ ions are unsuitable for SIFT-MS analyses of these peroxides, either because of low reactivity and/or the production of common product ions for their reactions with H2O and CH3COOH molecules. However, the results of this study show that NO+ precursor ions can be useful for the SIFT-MS analysis of both these peroxides, NO+H2O2 being the monitor ion for hydrogen peroxide analysis in moist air, the production of this ion actually being catalysed by the presence of H2O molecules, and NO2+ ion being suitable monitor ions for peroxyacetic acid analysis in the presence of acetic acid. The kinetic data for these peroxide reactions are presented and the likely mechanisms of the reactions are alluded to.