GC: Gas Chromatography; MS: Mass spectrometry; O: Olfactometry; AEDA: Aroma Extract Dilution Analysis; EO: Essential Oil; F.I.D./FID: Flame Ionization Detector; FD: Flavor Dilution (factor); VCF: Volatile Compounds in Food; FEMA: Flavor and Extract Manufacturers Association (ingredients generally recognized as safe under conditions of intended use as flavors from the FEMA expert panel are listed); RI: Retention Index.

Cinnamomum cassia (Nees and T. Nees) J. Presl, Lauracee family, is one of the oldest known spices. The volatile oil from leaf and bark and the oleoresin from bark are used in soaps, perfumes, spice essences, food and beverages. In traditional medicine it is widely used to treat dyspepsia, gastritis, blood circulation disturbance and inflammatory diseases [1]. The essential oil (EO) obtained from cassia is widely used in cosmetics and foods especially for its antioxidant, antifungal and antibacterial properties [2,3]. Beside its use in fragrances, cassia is used as a flavoring agent in foods for its spicy, sweet and warm aromatic notes both in savory (ham and meat) and sweet food (e.g., beverages such as cola), although with some limits provided by law (Regulation EC N°1334/2008).

The need of analytical methods for quality assessment is directly related to the economic importance of cassia and his derivatives as raw materials in the food industry. This point is crucial if cassia end-use is to make flavors.

In this study a commercial-grade cassia essential oil was analyzed with GC-MS and GC-O was used to investigate the olfactory profile [4]. In literature, several sample of different origin and grade of Cassia EO, from laboratory scale to commercial, have been screened with significant variability on reported results [5]. Moreover, quantitative analyses of constituents in cassia oil are not always available as GC percentages are more often reported.

Literature papers tend agree on cassia oil major component being cinnamic aldehyde, ranging from 50% to 93%, o-methoxy cinnamaldehyde (0.1-25.4%), cinnamyl acetate (0.3-7.6%), benzaldehyde (0.3-2.9%) [5,6]. In literature, the number of compounds identified with GC-olfactometric analysis in food product usually varies with the nature of the raw material.

Literature data [7] for spices such as black pepper shows that important odorants are about 14, mainly terpenes and oxygenated compounds. For some spices, the odor is well represented by that of the main component of the volatile fraction. An example is cinnamon (Cinnamomum zeylanicum, C. aromaticum, C. burmanii) whose odor is mainly characterized by the high content in cinnamic aldehyde [7].

Generally the concentration of cinnamic aldehyde determines the flavor quality of cassia too, low levels being known to represent material of low quality [8,9] for these reasons, other aromatic compounds in the oil have not been examined in depth so there is a lack of odor profile characterization of cassia oil in literature.

Despite the considerable amount of cinnamic aldehyde, cassia oil have a more complex odor profile with spicy, warm and sweet notes with a woody and earthy background.

Gas chromatography-olfactometry (GC-O) can help achieve this goal since GC-O allows to determinate the contribution of single constituents to the overall flavor of a product. GC-O enables the assessment of odor-active components in complex mixtures, through the specific correlation with the chromatographic peaks of interest; this is possible because two detectors, one of them being the human olfactory system, perceive the eluted substances simultaneously.

Different GC-O methods are available such as dilution, time intensity, detection frequency and posterior intensity [10,11]. Dilutions methods are the most used methods and they are based on successive dilutions of an aroma extract. AEDA-Aroma Extraction Dilution Analysis-is one of the most used GC-O methods because it permits to identify the most important aroma compounds [12,13].

In AEDA the odorants are separated by gas chromatography on a capillary column (Figure 1). To determine the retention times of the aroma substances, the carrier gas stream, after leaving the capillary column, goes to sniffing port for detection by panelist (GC-O). The sensory assessment of a single GC run is not very meaningful because the perception of aroma substances in the carrier gas stream depends on limited quantities, e.g., the degree of concentration of the volatile fraction, and the amount of sample separated by gas chromatography [7]. These limitations are overcame by the stepwise dilution of the volatile fraction with solvent (Figure 2), followed by the gas chromatographic/olfactometric analysis of each dilution. The dilution process is repeated until no more aroma substance can be detected by the panelist.

Figure 1: AEDA aromagram of Cinnamomum cassia EO. All the molecules perceived during AEDA session versus LogFD. Higher LogFD value is due to an odorant perception until high dilution level.

Figure 2: Olfactogram of first dilution of Cinnamon cassia obtained by Panelist1: retention time (X Axis) versus signal from GC-O joystick. When the panelist push the joystick, the instrument records a 100 signal that lasts until the panelist release the joystick.

Our GC-O equipment splits the flux into sniffing port (so human nose as detector) and analytical detector F.I.D. In order to get a precise identification of the peaks, a GC-MS analysis of the EO was run as well. GC-MS has been proven to be a powerful and suitable tool for the determination of volatile compounds because of its high separation efficiency and sensitive detection [14]. In order to avoid errors due to peaks overlapping, an equipment with double column was used.

Essential oil and reagents

A commercial grade cassia (Cinnamomum cassia (Nees and T.Nees) J. Presl.) essential oil obtained by steam distillation from cassia bark, leaves and twigs was used for all tests. Ethyl alcohol (96°, food grade) was used in the experiments.

Analysis of volatile composition

The identification of volatile aroma compound was performed on a gas-chromatograph Thermo Trace coupled with a Thermo ISQ massspectrometer. The GC-MS system was equipped with a DB1 (30 m, 0.25 mm i.d., 0.25 μm film thickness) capillary column (Agilent JandW). The starting temperature of the column was 50°C, which was held for 3 min, then increased 5°C/min to 280°C, where was held for 10 min. The constant column flow was 1,5 ml/min, using helium as carrier gas, the injector was in split mode at 280°C. Mass spectrometer parameters were as follow: ionization mode EI at 70 eV, source temperature 250°C, scan range m/z 33 to 350, scan mass 1 s. The components were identified by comparing their spectra with those present in the Wiley and NIST spectra collection and in an authentic chemicals spectra library.

The mass spectra identifications were confirmed using Kovats Indices (n-alkanes) on DB1 and DB1701 columns (Figure 3). To calculate the Kovats Indices a GC analysis was performed using an Agilent 5890 gas-chromatograph equipped with an auto sampler. The two capillary columns (Agilent JandW) DB1 and DB1701 (30 m, 0.25 mm i.d., 0.25 μm film thickness) were assembled on the same injector. Helium (flow 1.5 ml/min) was used as carrier gas, the injector was in split mode (ratio 1:50) and two F.I.D. detector were used. The injector and detectors temperatures were 280°C and 285°C respectively. The column temperature was initially maintained at 50°C for 3 min before increasing to 280°C at a rate of 5°C/min and held for 10 min.

Figure 3: Chromatogram (FID) Cinnamon cassia essential oil (FID obtained by elution from column DB1).

The Kovats Indices were compared to those of authentic chemicals elute in the same conditions. For quantitative analysis, the response factor relative to the internal standard (benzyl benzoate) was previously determined and stored in an inland library for each molecule identified. The choice of benzyl benzoate as internal standard has been made after a preliminary evaluation confirmed negligible quantities of this component in our sample. The percentage of every component was then calculated using the response factors.

Response factor (K) is calculated (for each molecule, on both capillary column) using the following:

K=W(x) × A(IS)/W(IS) × A(x)

Where: W(x)=weight (g) of molecule “x”; W(IS)=weight (g) of internal standard (Benzyl benzoate); A (IS)=Peak area of internal standard; A (x)=peak area of molecule “x”.

GC-O analysis

Cassia essential oil was analyzed with GC-O using an Agilent 5890 gas chromatograph equipped with DB-1 (30 m, 0.25 mm i.d., 0.25 μm film thickness) capillary column (Agilent J and W).

Chromatographic conditions were 50°C held for 3 min, 5°C/min to 220°C, 15°C/min to 280°C held for 10 min. The first temperature ramp was the same used for the determination of Kovats indices. Helium was used as carrier gas (flux 1.5 ml/min). The injector was in splitless mode, column flow was divided (70:30) by a splitter between the sniffing port and a flame ionization detector (F.I.D.).

A tailor-made software (Simplicius^® from ABREG) was used to allow the recording of the odor intensity from the mechanical input of the panelists and to store the recorded signals that were further transformed into FD values or Log(FD).

GC-O analysis was performed using aroma extraction dilution analysis (AEDA) method. In order to find the most important odoractive compounds, which have the lowest odor thresholds, the amount of sample was reduced by ten dilution (w/v) of essential oil in ethyl alcohol (96°, food grade) factor dilution was 3. The sniffing started from the third dilution (1:27) due to peak overlapping (e.g., cinnamic aldehyde) and to overall too strong smell intensity of main eluting peak. According to AEDA method, samples were evaluated by the panelists in increasing dilution order and the impact of an odor-active component was measured as the last dilution value (FD).

The sniffing of cassia essential oil was carried-out by four trained assessors (1 male and 3 female, age 25-50 years), not smokers and without anosmia [15].

In total 72 components were identified through GC-MS analysis and confirmed by Kovats indices. Out of them 41 were quantified using internal standard and response factor while 31 were found in traces (<0.01%).

The majority of volatiles identified belongs to oxygenated compounds (e.g., aldehydes) while non oxygenated terpenes represent about 18% of the oil. In the analyzed cassia oil sample, cinnamic aldehyde is confirmed as the major component (about 78.4%), followed by o-methoxy cinnamaldehyde (7.40%), cinnamyl acetate (1.70%) and benzaldehyde (1.13%) (Table 1).

Table 1: Identifications and quantitative data of commercial sample of Cinnamon Cassia essential oil.

Different publications on volatiles components of cassia oil have also been reviewed by VCF [16]; 69 components have been listed, with only 25% have been associated with a content range. In our analyses, we are able to quantify as much as 55% of the overall molecules including an high numbers of components usually found in traces such as eucalyptol (0.01%), 3- phenylpropanol (0.01%), 2,4-decadienal (0.01%), cinnamyl formate (0.01%).

Due to safety concerns related to coumarine presence in flavor and fragrance, it was important to detect its content in cassia oil. In our sample, coumarine content reached 0.88%, consistent with the previously reported 0-12.2% [5]. Phenylacetaldehyde was not found in quantitative data in literature but in our sample it was present in small quantity (0.01%); this compound has a relevant odor impression since it was been perceived for many dilution in GC-O analysis. The odor active components determined in our cassia oil sample were 26. The odorants were identified by comparing their mass spectra, retention index, Kovats index and odor description with those of authentic standards.

The data in Table 2 show that most intense odor active compounds were aldehydes: 3- phenylpropanal and cinnamic aldehyde were perceived until last dilution. Some compounds present in smaller quantity had stronger olfactory impression (e.g., guaiacol) than other compounds present in higher quantities (e.g., o-methoxy cinnamaldehyde, benzaldehyde). The odor quality was obtained by olfactory evaluation and description perceived during GC-O session. Among perceived odorants there were some compounds that have a typical spicy connotation: cinnamic aldehyde, methoxy cinnamic aldehyde, cinnamyl acetate, eugenol and cinnamyl alcohol.

Table 2: AEDA results. Mean dilution indicates the last dilution until the molecule had been identified. Odor description indicates the odor perception of panelist.

The identification of a large number of odorants, some of which found in trace and others such as cinnamaldehyde and methoxy cinnamaldehyde that made up to approximately 90% of the oil, confirm AEDA as a powerful technique to analyze the olfactory profile of complex mixtures [17-19].

The author would thank the entire panel group that took part to the sniffing sessions. The author would thank Kerry Ingredients and Flavors that supported the study with materials and experts.