RAPID ODOUR SCREENING OF PAPERBOARD USING STATIC HEADSPACE-SIFT-MS

Significance of the Topic

Paperboard products often carry unwanted odours from volatile organic compounds (VOCs) generated during raw material processing and handling. Traditional sensory panels are costly, time consuming and subject to variability. An instrumental method for rapid odour screening can improve quality control, reduce reliance on human assessors and provide consistent, real-time data on odour-active compounds in paperboard.

Objectives and Study Overview

The study aimed to evaluate static headspace selected ion flow tube mass spectrometry (SH-SIFT-MS) combined with multivariate statistical analysis as a tool for rapid odour profiling of paperboard. Eleven paperboard samples with known sensory odour ratings were analyzed to determine:

• Whether SH-SIFT-MS can discriminate among different samples.
• How instrumental measurements correlate with trained panel odour ratings.
• Key VOC markers responsible for odour differences.

Methodology and Instrumentation

Paperboard coupons (21 cm × 4 cm, ~1.3 g) were sealed in 20 mL headspace vials and incubated at 60 °C for 15 minutes. A 2.5 mL sample of the headspace was drawn using a heated syringe (150 °C) at 50 µL/s and introduced into the SIFT-MS inlet with make-up gas for a total flow of ~420 µL/s. The instrument employs soft chemical ionization without chromatography, enabling real-time quantification of diverse VOCs. Multivariate analysis used the SIMCA algorithm (Soft Independent Modelling by Class Analogy) in Pirouette software, applying principal component analysis (PCA) to model classes, calculate interclass distances and determine discriminating power of individual compounds.

Main Results and Discussion

• Class Discrimination: Eleven unique sample classes achieved good separation in PCA space, with most interclass distances exceeding the threshold of 3.
• Correlations with Sensory Ratings: Samples with extreme odour ratings (highest and lowest) were clearly distinguished. Mid-range ratings showed some overlap but remained broadly separable.
• Key Markers: Methanol and short-chain aldehydes (e.g., hexanal, heptanal, octanal), along with volatile fatty acids and hydrogen sulfide, contributed most to sample differentiation.

Benefits and Practical Applications

• Throughput: Analysis of at least 12 samples per hour enables rapid QC screening.
• Cost-effectiveness: Eliminates lengthy sensory panels and extensive sample preparation.
• Broad Coverage: Captures a wide range of odour-active VOCs in a single measurement.
• Objectivity: Provides consistent, quantitative odour ratings aligned with sensory panel assessments.

Future Trends and Potential Applications

• Model Refinement: Expanding sample libraries and odor descriptors to improve prediction accuracy.
• Automation: Integration into inline monitoring systems for real-time odour control in production lines.
• Advanced Data Analytics: Application of machine learning and pattern recognition for faster classification and odour fingerprinting.
• Industry Expansion: Adapting the method to other packaging materials and odor-sensitive products.

Conclusion

This preliminary study demonstrates that SH-SIFT-MS combined with multivariate statistical analysis can effectively replicate sensory panel odour ratings for paperboard samples. The method offers a rapid, economical and objective approach for odour screening and quality control. Further validation with larger sample sets and refined models will enhance its robustness and predictive power.

Reference

• Spanel P., Smith D. Selected ion flow tube: a technique for quantitative trace gas analysis of air and breath. Med. Biol. Eng. Comput. 1996, 24, 409.
• Smith D., Spanel P. Selected ion flow tube mass spectrometry (SIFT-MS) for on-line trace gas analysis. Mass Spec. Rev. 2005, 24, 661.
• Prince B.J., Milligan D.B., McEwan M.J. Application of SIFT-MS to real-time atmospheric monitoring. Rapid Commun. Mass Spectrom. 2010, 24, 1763.
• Shen W., Wang C. Multiple headspace extraction for quantitative determination of residual monomer and solvents in polystyrene pellets using the Agilent 7697A Headspace Sampler. Agilent Technologies Application Note 2012.
• Kvalheim O.M., Karstang T.V. SIMCA – Classification by mean of disjoint cross validated principal components models. In Multivariate Pattern Recognition in Chemometrics, Elsevier: Amsterdam, 1992, 237.