Trace-Level Volatile Quantitation by Direct Analysis in Real Time Mass Spectrometry following Headspace Extraction: Optimization and Validation in Grapes

• Published in: Journal of agricultural and food chemistry Vol. 65; no. 42; pp. 9353 - 9359
• Main Authors: Jastrzembski, Jillian A, Bee, Madeleine Y, Sacks, Gavin L
• Format: Journal Article
• Language: English
• Published: United States American Chemical Society 25.10.2017
• Subjects: Flavoring Agents - chemistry; Flavoring Agents - isolation & purification; Mass Spectrometry - methods; Solid Phase Microextraction; Vitis - chemistry; Volatile Organic Compounds - chemistry; Volatile Organic Compounds - isolation & purification; DART; low-level odorants; volatile analysis; wine grapes; ambient ionization
• ISSN: 0021-8561; 1520-5118
• DOI: 10.1021/acs.jafc.7b03638

Ambient ionization mass spectrometric (AI-MS) techniques like direct analysis in real time (DART) offer the potential for rapid quantitative analyses of trace volatiles in food matrices, but performance is generally limited by the lack of preconcentration and extraction steps. The sensitivity and selectivity of AI-MS approaches can be improved through solid-phase microextraction (SPME) with appropriate thin-film geometries, for example, solid-phase mesh-enhanced sorption from headspace (SPMESH). This work improves the SPMESH-DART-MS approach for use in food analyses and validates the approach for trace volatile analysis for two compounds in real samples (grape macerates). SPMESH units prepared with different sorbent coatings were evaluated for their ability to extract a range of odor-active volatiles, with poly­(dimethylsiloxane)/divinylbenzene giving the most satisfactory results. In combination with high-resolution mass spectrometry (HRMS), detection limits for SPMESH-DART-MS under 4 ng/L in less than 30 s acquisition times could be achieved for some volatiles [3-isobutyl-2-methoxypyrazine (IBMP) and β-damascenone]. A comparison of SPMESH-DART-MS and SPME-GC-MS quantitation of linalool and IBMP demonstrates excellent agreement between the two methods for real grape samples (r 2 ≥ 0.90), although linalool measurements appeared to also include isobaric interference.