SIFT-MS: A Tool for Objective Instrumental Odor Analysis?

The human olfactory system’s response to odor is extremely complex and has proved difficult to reproduce instrumentally, and so human sensory panels remain the gold standard approach. Sensory panels are, however, expensive and can only assess a limited number of samples per day before panelists become fatigued, meaning they are not practically scalable for applications ranging from the assessment of recycled packaging to environmental odor complaints. Gas chromatography–mass spectrometry (GC–MS) and GC-olfactometry (GC-O) are powerful instrumental techniques, but throughput and coverage of all odorants in a single analysis are significant impediments to wider use. Selected ion flow tube mass spectrometry (SIFT-MS) provides an alternative approach using direct sample analysis with soft chemical ionization that sensitively and specifically quantifies a wide range of odorants in real time. This article reviews SIFT-MS odor analysis in food-flavor, packaging, and environmental applications, suggesting potential utilization as an objective screening tool.

Aroma or odor assessment is important across various industries because it contributes to consumer acceptance (for foods, beverages, and other consumer products). In the environmental sector, unpleasant odors contribute to poorer quality of life (1). Since it is a human perception, expert sensory panels currently serve as the gold standard for assessing odor —from food through to environmental nuisance. Panel size typically ranges from four to 12 individuals, and thorough training is necessary to make the process as objective as possible.

Clearly, there is significant set-up cost for a sensory panel, and it is expensive for each sample. In addition, sensory fatigue means that most panels can only assess a small number of samples per day. These challenges have prompted the evaluation of numerous technologies for sensory analysis in recent decades. While it is outside the scope of this article to review these in detail, several widely used technologies are summarized in Table I (2), together with the emerging technology that is the subject of this article, selected ion flow tube mass spectrometry (SIFT-MS).

When applying analytical technology to odor evaluation, it is very important to recognize that the human sense of smell is extremely complex, varies significantly from one person to another, and is still only partially understood (1). Furthermore, odors are often complex mixtures (and they frequently include volatiles with significantly lower odor impact), and odorants are of diverse chemical functionality. Hence comprehensive instrumental analysis of odor that mimics the olfactory system response is extremely challenging.

This article discusses the potential application of SIFT-MS to objective instrumental odor analysis. Published case studies are summarized for several food, packaging, and environmental applications, while a broader view of literature on each topic is provided in recent general reviews (3,4). SIFT-MS is not proposed as a universally applicable solution, but it is hoped that readers may be assisted in identifying candidate systems via the summary of relevant SIFT-MS attributes coupled with varied examples.

SIFT-MS: A Brief Introduction in the Context of Odor Analysis

SIFT-MS and its applications have been described in detail elsewhere (3,5,6,7).Briefly, SIFT-MS uses multiple, rapidly switchable chemical ionization agents (so-called reagent ions) that detect diverse volatile compounds in the gas phase with high specificity (4,7). The SIFT-MS reagent ions are generated by a microwave plasma through moist air, then selected individually using a quadrupole mass filter. The mass-selected reagent ions first encounter carrier gas, which cools the ions and provides consistent ionization and quantitation from reagent ion—analyte reactions that occur when sample is introduced. After a few milliseconds of reaction as they traverse the flow tube, product ions and unreacted reagent ions are mass-filtered and detected, and analyte concentrations are calculated in real time. In the studies reviewed below, commercial SIFT-MS instruments were utilized. Where analysis was automated, “xyz” robotic syringe-injection autosamplers coupled with the SIFT-MS instrument were used (8).

This section summarizes the attributes relevant to its applications in instrument-based sensory analysis: (i) breadth of analysis, (ii) specificity (including the ability to identify specific odorants), (iii) quantitation (including sensitivity and dynamic and linear ranges), (iv) rapid analysis (real-time and high throughput), and (v) practicality (from simple reporting of results to suitability for mobile laboratories).

Breadth of Analysis: Although the SIFT-MS technique is blind to the bulk components of air due to its soft ionization, it has remarkable breadth of analysis for volatile organic compounds (VOCs )and trace inorganic gases. This is due to the use of multiple reagent ions that ionize compounds via a wide range of ion-molecule reaction mechanisms (Figure 1) (3,7). SIFT-MS readily detects and quantifies common, pungent odorants including:

• Hydrogen sulfide and organosulfur compounds,
• Ammonia, amines, and other nitrogen-containing species,
• Aldehydes and ketones, and
• Volatile fatty acids and esters.
  

Furthermore, SIFT-MS can analyze most compounds in a single method on a single instrument configuration, without derivatization, preconcentration, or drying. The data presented below demonstrate this capability.

Specificity: SIFT-MS achieves specific analysis in real time by combining highly controlled, ultra-soft chemical ionization with mass spectrometry (4,7). Moreover, the ability to rapidly switch reagent ions enables the most appropriate reagent ion-product ion pair(s) to be utilized to achieve specific and sensitive analysis in a single run. Figure 1 summarizes the elements that contribute to specific analysis in SIFT-MS. Note, however, that the specificity of SIFT-MS is not as high as gas chromatography–mass spectrometry (GC–MS), since the latter temporally separates analytes.

Quantitation: SIFT-MS is inherently quantitative, reporting gas-phase concentrations (in part-per-billion by volume; ppbV) in real time based on first-principles calculations (9), with dynamic and linear ranges of five orders of magnitude. Conventional calibration approaches are readily applied to SIFT-MS data (8). Modern instruments have quantitation limits lower than most human odor recognition thresholds (ORTs), typically in the part-per-trillion by volume range (pptV).

Rapid Analysis: Since SIFT-MS is a chromatography-free, direct MS technique, it analyzes samples continuously in real time (3) or with high throughput when coupled with an autosampler (8).

Practicality: Ultra-soft chemical ionization ensures the stability of the SIFT-MS technique, enabling it to be highly automated, and facilitating a high degree of software automation. SIFT-MS instruments are also robust, even acquiring data in real time in moving laboratories (10,11).

Food

Human sensory panels play important roles in both development of new food products and in routine quality assurance. Correlation of sensory panel data with SIFT-MS measurements of odorants have consistently demonstrated high correlation. Here, two examples are discussed: parmesan cheese and beef.

Parmesan cheese studies using SIFT-MS have targeted the most impactful odorants identified in earlier published studies (12,13). Concentration data for these odorants were converted into odor-activity values (OAVs) using published ORTs, then processed using multivariate statistical analysis (soft independent modeling by class analogy; SIMCA) to evaluate the ability of SIFT-MS to classify the sample types (see reference 14 for a description of how this statistical technique is used with SIFT-MS). In the first study, using manual headspace analysis, Langford and colleagues (15) differentiated genuine Italian and imitation New Zealand Parmesan cheeses according to both factory and country of origin. A follow-up study by Perkins et al. (16) utilized automated headspace analysis to differentiate six genuine Parmesan products from three Italian manufacturers by targeting the odor-active compounds. Figure 2 summarizes the results obtained when the odor-active compound concentrations determined by SIFT-MS were used to classify six products. As measured by interclass distances (all three or greater [14]), all products were readily distinguished using this approach. Products P1, P4, P5, and P6 (bottom left in Figure 2) are all products from the same manufacturer. When these were aggregated into the same class and SIMCA analysis applied, an interclass distance of 2.6 indicated that they were nearly distinguished from P2, while both groups were well separated from P3. In summary, targeting the dominant odor-active compounds using quantitative SIFT-MS analysis, coupled with multivariate statistical analysis, is a potentially powerful tool for product and manufacturer identification.

SIFT-MS combined with multivariate statistical analysis can also be used to more directly correlate with human sensory evaluation, as demonstrated by a study that classified prime and defective beef samples (17). Beef aroma is an important characteristic for acceptance by consumers, and preference is often culturally dependent. Certain VOCs impart favorable or unfavorable characteristics to the aroma, but grading—if conducted at all—has traditionally been achieved using limited sensory testing because robust, high-throughput analytical technologies were not available.

Premium beef samples were derived from eight prime beef cattle (labelled “Prime”), while defective beef samples were classified by the expert sensory panel as “Bull 1,” “Bull 2,” “Cow 1,” “Cow 2,” “High pH 1,” “High pH 2,” “Norm pH,” and “Over-aged” (17). Figure 3 shows headspace-SIFT-MS concentration data (averaged across multiple replicates) for the premium and defective beef samples, illustrating broad-spectrum odorant detection in a single analysis. Although the volatile profiles for the various classifications are visually different in Figure 3, for rapid screening applications (such as the testing laboratory or on the process line), classification using multivariate statistical methods is the preferred approach, providing an immediate, objective answer. As for Parmesan above, the SIFT-MS data were correlated with sensory classifications using the SIMCA multivariate statistics approach (Figure 4). Each colored point represents a replicate measurement. All sensory classifications are well separated in the SIFT-MS analysis. This suggests that SIFT-MS has potential as a rapid, economical grading tool for beef, facilitating grading on a much wider scale than has been possible using the conventional sensory panel approach.

Packaging

Packaging materials can transmit odorous and/or harmful volatile compounds to the consumer products that they contain, so quality assurance testing of these materials is important prior to use. Increased recycling of packaging materials amplifies testing requirements, since risks of contamination—and hence recalls and brand damage—from prior usage increase significantly compared to virgin materials. SIFT-MS analysis coupled with multivariate statistical analysis has potential to provide high-throughput screening of materials and in production quality control.

Paper is experiencing a renaissance as consumer goods and pharmaceutical industries move toward more sustainable forms of packaging. However, by its very nature, paper is rich in volatiles—both with and without significant olfactory impact. Human sensory analysis (using, for example, the DIN 1230-1 method [18]) has long been the benchmark for determining whether packaging is acceptable to consumers, but it is expensive to deploy for routine product analysis. In parallel with human sensory testing, SIFT-MS has been used to analyze recycled, virgin, and mixed paperboard samples destined for packaging applications (19). This study investigated the correlation with odor intensity rating and odor descriptor (also termed character or note), in addition to distinguishing paper composition and mill of origin.

Paper samples from four mills (24 products in total) were analyzed, together with one pulp sample.The key discriminating volatiles varied depending on the classification model, but typically included hydrogen sulfide, ethanol, acetone, hexanal, dimethyl sulfide, and formaldehyde. Headspace-SIFT-MS analysis coupled with multivariate statistical data analysis (SIMCA) exhibited near-perfect classification of the fiber type and the mill of origin. Classification of products according to the odor intensity rating determined by the sensory panel proved effective overall. An exception was observed for odor intensity ratings in the range from 2.75 to 3.5, that is, quite odorous samples (“moderate to strong” odor intensity). Samples in this range are likely to be poorly suited to the packaging of food, beverage, cosmetic, and drug products, so the ability to rapidly identify such odorous samples across a range of intensities and odor descriptors is sufficient for this purpose.In addition, samples were assigned one of 14 combinations of odor descriptors by the sensory panel. The combined SIFT-MS and SIMCA approach achieved effective separation for all combinations except for the following: “paperlike/sourly/musty/horse” and “paperlike/horse/cheesy/woody” were not differentiated. The results obtained in this study demonstrate the potential application of SIFT-MS as an objective, rapid instrument-based classification tool for odor intensity ratings and odor descriptors.

Environment

Odors in the environmental context can arise from natural and industrial sources; in the case of natural sources, these can be amplified through anthropogenic interventions, such as intense farming operations. In addition to the challenge of using instrumentation to objectively quantify odor, the often-transient nature of environmental odor events can make identifying odor sources challenging. In addition to the analytical benefits discussed above for laboratory or manufacturing applications, the demonstrated robustness of SIFT-MS that enables monitoring in mobile laboratories (11) means that sensory analysis can be conducted in the field and in real time, if required.

A major wastewater treatment plant (WWTP) study (20) compared dilution olfactometry and SIFT-MS analysis for 16 odor sources through a multi-step wastewater treatment process. Sampling was conducted from late spring to early fall/autumn and comprised 107 samples. For consistency across the sensory and instrumental analyses at geographically separated laboratories, sampling was conducted using Nalophanodor bags prepared in-house by the sensory laboratory. The sensory laboratory conducted analyses according to Australia/New Zealand Standard 4323.3:2001 (21) using an olfactometer with four to six panelists. SIFT-MS analysis targeted 66 analytes derived from the literature, comprising sulfur-, nitrogen- and chorine-containing compounds, alcohols, aldehydes, ketones, volatile fatty acids, hydrocarbons, and monoterpenes and their derivatives. SIFT-MS selectively detected and quantified many, but not all, key odorants. Overall, the SIFT-MS data correlated well with the olfactometry data (Figure 5). The outlier at upper left in Figure 5 corresponds to a sample determined informally by the odor panel to have an “earthy/musty” odor. Typically, this odor note arises from geosmin and/or 2-methylisoborneol, which have very low human odor recognition thresholds in air (7 pptV), and currently cannot be selectively quantified in the complex matrices throughout the WWTP process using SIFT-MS. In contrast, the characteristic “sulfur” (or “sewage”) odor was sensitively detected, even in samples where informal sensory analysis missed it. The SIFT-MS results also demonstrated that hydrogen sulfide is not a reliable sole marker for WWTP odor, as is commonly assumed by industry in its widespread use of hydrogen sulfide sensors.

In summary, SIFT-MS has the potential to provide instrument-based sensory analysis for environmental applications, based on this substantial comparative evaluation alongside gold-standard dilution olfactometry (which is not without limitations [22]).

Conclusion

The SIFT-MS technique has found broad application in the detection and quantitation of diverse odorous volatile compounds since it was introduced in the mid-1990s. This article has used case studies in food, packaging, and environmental applications to demonstrate various approaches that have leveraged quantitative odorant analysis for correlation with human sensory analysis. Overall, the correlation between SIFT-MS and sensory results is very good, and hence SIFT-MS—both in the stationary and mobile/field laboratory—offers potential for high-throughput and real-time objective instrumental odor analysis. SIFT-MS may therefore offer the opportunity for wider-scale sensory screening by substantially reducing the cost per sample compared to sensory panels.

References

(1) Springer Handbook of Odor; Buettner, A., Ed.; Springer, 2017.

(2) Reineccius, G. Flavor Chemistry and Technology; CRC Press, 2006.

(3) Smith, D.; Španěl, P.; Demarais, N.; Langford, V. S.; McEwan, M. J. Recent Developments and Applications of [SIFT-MS]. Mass Spec. Rev. 2025, 44, 101–34. DOI: 10.1002/mas.21835

(4) Langford, V. S. SIFT-MS: Quantifying the Volatiles you Smell… And the Toxics You Don’t. Chemosensors 2023, 11, 111. DOI: 10.3390/chemosensors11020111

(5) Smith, D.; Španěl, P. Selected Ion Flow Tube Mass Spectrometry (SIFT-MS) for On-line Trace Gas Analysis. Mass Spec. Rev. 2005, 24, 661–700. DOI: 10.1002/mas.20033

(6) Španěl, P.; Smith, D. Progress in SIFT-MS: Breath Analysis and Other Applications. Mass Spectrom. Rev. 2011, 30, 236–267. DOI: 10.1002/mas.20303

(7) Langford, V. S.; Perkins, M. J. SIFT-MS: From Method Concept to Routine Analysis. Royal Society of Chemistry, 2025. DOI: 10.1039/9781837677917

(8) Langford, V. S.; Perkins, M. J. Improved Volatiles Analysis Workflows Using Automated [SIFT-MS]. Anal. Methods 2024, 16, 8119–8138. DOI: 10.1039/D4AY01707B 

(9) Langford, V. S.; Dryahina, K.; Španěl, P. Robust Automated SIFT-MS Quantitation of Volatile Compounds in Air Using a Multicomponent Gas Standard. J. Am. Soc. Mass Spectrom. 2023, 34, 2630–2645. DOI: 10.1021/jasms.3c00312

(10) Wagner, R. L.;Farren, N. J.;Davison, J.;et al. Application of a Mobile Laboratory Using [SIFT-MS]for Characterisation of Volatile Organic Compounds and Atmospheric Trace Gases. Atmos. Meas. Tech. 2021, 14,6083–6100. DOI: 10.5194/amt-14-6083-2021

(11) Langford, V. S.; Cha, M. Y.; Milligan, D. B.; Lee, J. H. Adoption of SIFT-MS for VOC Pollution Monitoring in South Korea. Environments 2023, 10, 201. DOI: 10.3390/Environments10120201

(12) Qian, M.; Reineccius, G. A. Potent Aroma Compounds in Parmigiano Reggiano Cheese Studied Using a Dynamic Headspace (Purge-trap) Method. Flav. Frag. J. 2003, 18, 252–259. DOI: 10.1002/ffj.1194

(13) Qian, M.; Reineccius, G. A. Quantification of Arom Compounds in Parmigiano Reggiano Cheese by a Dynamic Headspace Gas Chromatography-Mass Spectrometry Technique and Calculation of Odor Activity Value. J. Dairy. Sci. 2003, 86, 770–776. DOI: 10.3168/jds.S0022-0302(03)73658-3

(14) Langford, V. S.; Perkins, M. J. Rapid Untargeted Screening of Food and Ingredient Aroma Using Direct-injection Mass Spectrometry. The Column 2023, 19 (6), 23–26. www.chromatographyonline.com/view/rapid-untargeted-screening-of-food-and-ingredient-aroma-using-direct-injection-mass-spectrometry

(15) Langford, V. S.; Reed, C. J.; Milligan, D. B.; et al. Headspace Analysis of Italian and New Zealand Parmesan Cheeses, Using SIFT-MS. J. Food Sci. 2012, 77, C719–C726. DOI: 10.1111/j.1750-3841.2012.02730.x

(16) Perkins, M. J.; Padayachee, D.; Langford, V. S. Rapid Parmesan Classification Using Automated Static Headspace-SIFT-MS Analysis, Syft Technologies application note. 2023. https://bit.ly/4d7zXAi (accessed 2025-07-25).

(17) Langford, V. S.; McEwan, M. J.; Cummings, T.; Simmons N.; Daly, C. Rapid Classification of Beef Aroma Quality Using SIFT‐MS. Adv. Food Bev. Anal. (Supplement to LCGC North Am.) 2018, 1, 8–15.

(18) German Institute for Standardization. DIN EN 1230-1:2010: Paper and Board Intended to Come into Contact with Foodstuffs – Sensory Analysis – Part 1: Odor. Berlin, 2010.

(19) Langford, V. S.; Du Bruyn, C.; Padayachee, D. An Evaluation of Selected Ion Flow Tube Mass Spectrometry for Rapid Instrumental Determination of Paper Type, Origin and Sensory Attributes. Packag. Technol. Sci. 2021, 34, 245–260. DOI:10.1002/pts.2555

(20) Langford, V. S.; Billiau, C.; McEwan, M. J. Evaluation of the Efficacy of SIFT-MS for Speciation of Wastewater Treatment Plant Odors in Parallel with Human Sensory Analysis. Environments 2020, 7, 90. DOI: 10.3390/environments7100090

(21) Standards Australia and Standards New Zealand. Determination of Odour Concentration by Dynamic Olfactometry (AS/NZS 4323.3:2001); Standards Australia International Ltd: Sydney, Australia; Standards New Zealand: Wellington, New Zealand, 2001.

(22) Braithwaite, S. K. Sensory Analysis and Health Risk Assessment of Environmental Odors. PhD Thesis. University of California-Los Angeles, 2019.