Figure 1: Overlay of the TIC of triplicates analysis of headspace vapors of coffee sample (variety #1).

Coffee is widely consumed as a beverage because of the stimulating effect it produces in humans. The aroma of coffee contributes to the flavor and taste of the beverage and has consequently led to extensive research on its benefits. Hundreds of volatile/semi- volatile aroma and flavor compounds have been identified in coffee using traditional laboratory based headspace GCMS systems. This is a study for rapid fingerprinting of coffee volatile/semi-volatile compounds using solid phase micro-extraction (SPME) coupled to a portable GCMS system for separation and detection.

The on-site analysis of coffee using portable technology can be used for quick quality control check of raw and finished products, comparison of competitor products, analysis of storage conditions or for process development.

Coffee (1 gm) was placed in headspace vials (20 mL vial volume), capped and placed at room temperature for at least an hour to allow for saturation of the aroma volatiles in the headspace. The SPME fiber was directly exposed to the headspace vapors in the vial for 15 seconds prior to analysis. Below are the GC and  MS conditions:
•    Sampling: SPME
•    SPME phase: DVB/PDMS, 65 um
•    GC injector temp: 270 C
•    GC column: MTX-5, 5 m x 0.1 mm, 0.4 u df
•    GC carrier gas: Helium, 0.2 mL/min
•    GC column temp: 50- 270 C at 2 C/sec, end hold time for 60 sec
•    Transfer line: 250 C
•    Injector split ratio: 10 to 1
•    Mass analyzer: Toroidal ion trap
•    Mass range: 42-500 Da
•    Detector: Electron multiplier

Figure 2: Overlay of the TIC of triplicates analysis of headspace vapors of decaffeinated coffee sample (variety #1).

Results and discussion

Triplicate analysis of coffee (variety #1) is shown in Figure 1. The overlay of the total ion current (TIC) suggests the analysis is very reproducible between injections. 
The overlay of TIC of caffeinated and decaffeinated samples of coffee (variety #1) showed peaks with similar retention time but in many cases with varying intensities, suggesting similar compounds are present in the two samples but at different concentrations (Figure 3). An extra peak was observed in the caffeinated sample that was not observed in the decaffeinated coffee.

The extraneous peak observed in the caffeinated coffee resembled the spectra of toluene, and comparisons confirmed its identification. The presence of toluene in the caffeinated sample is not surprising as literature suggests toluene can be produced in roasted coffee.

Figure 3: Overlay of the TIC of caffeinated and decaffeinated coffee variety #1 samples.

The decaffeinated form of variety #1 coffee was similarly analyzed in triplicates (Figure 2) and also showed excellent reproducibility of analysis. 

Two other varieties of coffee (variety #2 and #3) were analyzed, and the TIC overlay of the three coffee types (Figure 4) showed some distinct profiles between the samples. For instance, a peak at ~30 secs was common to #1 and #2, but not observed in #3. The spectra of this peak was matched against the NIST library and the probability of the match suggested it was likely methyl pyrrole.

Literature studies suggest that methyl pyrrole degradation is accelerated in coffee with high moisture content, which may be the reason for variety #3 having little or no methyl pyrrole.

A brief study of coffee fingerprinting using the Torion T-9 coupled to SPME sampling resulted in reproducible profiles for triplicate analysis. Separation of the volatile/semi-volatile compounds by GC and the specificity of detection provided by MS along with spectral matching to the NIST library helped identify differences in coffee varieties. Therefore, portable GCMS provides a rapid and reliable screening technology to study coffee aroma profiles within a couple of minutes, making it an ideal tool for on-site analysis. 

Figure 4: Overlay of the TICs of three different coffee varieties.