Saffron adulteration detection by high-resolution mass spectrometry
Author : Hamed Biglari | 2024 Jan 13

Saffron Adulteration Detection By High-Resolution Mass Spectrometry

Saffron, a spice obtained by drying the scars of saffron flowers (Crocus sativus L.), is by far the most valuable spice in the world, and, not surprisingly, it often becomes an object of adulteration. In this study, non-target screening of volatile compounds in 38 authentic saffron samples and 25 samples of plant materials, potential saffron adulterants (safflower, calendula, capsicum, and turmeric) was performed using headspace solid-phase microextraction for sample extraction followed by gas chromatography coupled to high-resolution mass spectrometry (HS-SPME-GC-HRMS) for non-target volatiles screening. The chemometric analysis of generated data by principal component analysis (PCA) and partial least squares discriminative analysis (PLS-DA) showed good separation of authentic saffron from potential plant adulterants. Significant compounds (‘markers’) for each group of these plants, as well as for saffron, were tentatively identified. The target screening of selected ‘markers’ in model admixtures enabled simple and reliable detection levels as low as 2% w/w of safflower, calendula, capsicum, or turmeric present in saffron.

A preprint edition of Filatova M et al. is shared.

Introduction

Saffron, the most valuable spice in the world, is obtained by drying the stigma of saffron flowers (Crocus sativus L.) [1]. The high price of saffron is due to the traditional collection method, where the stigma must be collected by hand, and the collection yield is minimal [2, 3]. Considering the high economic value and the limited raw material sources, saffron is vulnerable to fraud. One of the most frequent fraudulent practices is dilution by low-priced plant materials such as safflower petals (Carthamus tinctorius), calendula petals (Calendula officinalis), turmeric powder (Curcuma longa), beet fibers (Beta vulgaris), pomegranate fibers (Punica granatum), arnica flowers (Arnica Montana), and achiote seeds (Bixa Orellana) [4–6] with similar color and morphology to saffron [3, 5].

Like other spices, saffron's chemical composition is complex, containing many secondary metabolites responsible for characteristic bioactivity, ties, and overall quality. The significant phytochemicals distinct from saffron are crocetin glycosidic esters associated with a typical golden-yellow tinge of flavored food [1, 5], safra- nal, the critical component of volatiles fraction, and picrocrocin, which is responsible for the characteristic bitter taste of this spice. Picrocrocin, the unique secondary metabolite, is a known molecular marker of saffron [1, 3]. Regarding saffron aroma, isophorone-related compounds (structurally similar to safranal), C13-nor isoprenoids derived from lipophilic carotenoids, and saturated hydrocarbons such as octadecane, docosane, tricosane, and pentacosane are typically represented [5, 7, 8]. Quality and classification of saffron is defined by ISO 3632–2 2010 (microscopic analysis for detection of extraneous material in saffron) and ISO 362–1 2011 (UV–Vis), which is based on measurement of the absorption maxim values of aqueous saffron extracts at 257 nm (detection of picrocrocin), 330 nm (detection of safranal) and 440 nm (detection of crocins)[1, 6]. However, the UV mentioned above–Vis spectrophotometric method has been shown as insufficient when saffron is adulterated by plant material with a similar color and morphology. The study was performed by Sabatino et al. [9] and Demonstrated that the UV–Vis method did not enable the detection of the addition of impurities (safflower, calendula, and turmeric) at a level lower than 20% w/w. Besides the molecular biology-based approach employing polymerase chain reaction (PCR) followed by sequencing [10, 11], several instrumental strategies such as nuclear magnetic resonance (NMR) [2, 12, 13], liquid chromatography coupled with mass spectrometry (LC–MS) [3, 14, 15], infrared spectroscopy (IR) [16–18] were used for saffron authentication. In this study, we aimed to investigate in depth the application potential of headspace sampling of saffron volatiles by solid-phase microextraction extraction followed by gas chromatography coupled to high-resolution mass spectrometry (HS-SPME-GC-HRMS) for authentication purposes. There was only one study employing SPME-GC-HRMS for the detection of plant-derived pollutants [6], and we were wondering whether our approach using chemometric data handling would be able to identify similar markers and which lowest concentration of plant adulterants (safflower, calendula, capsicum, and turmeric) in the admixture could be recognized.

Materials and methods

Samples

In total, 38 authentic saffron (Crocus sativus L.) samples from different origins (Iran, Spain, Greece, and Italy) were obtained for this study in collaboration with the Czech Agriculture and Food Inspection Authority. Also, 25 samples of spices and plant material, which could be used for an adulteration of saffron, were collected for this study. They were Turmeric (Curcuma longa L., n = 6), Calendula (Calendula officinalis L., n = 9), Capsicum (Capsicum annum L., n = 6) and Safflower (Carthamus tinctorius L., n = 4). Some of the samples were supplied as ground material (powder). An acceptable homogenization was required in some samples (mostly saffron obtained as dried stigma). To avoid the loss of volatiles, the homogenization was manually performed in the ceramic grinding bowl. All samples were stored in gas-tight containers in the freezer at -18 °C.

Chemicals and materials

SPME fibers coated with divinylbenzene/carboxy/poly- methylsiloxane (DVB/CAR/PDMS, 50/30 μm, 1 cm) were purchased from Supelco (Bellefonte, USA). Sodium chloride was purchased from Penta (Prague, Czech Republic). Water purified by a Milli-Q® Integral system supplied by Merck was used throughout the study. C7–C30 saturated alkanes mixture (certified reference material, 1000 μg/mL each component in hexane), used for unknown compounds retention index (RI) based identification, was obtained from Sigma-Aldrich (Steinheim, Germany).

Sample preparation

Non‑target screening of volatile profiles of saffron and adulterants

Finely ground samples (50 mg) were placed in SPME crimp cap vials (10 mL), and 2 mL of a saturated solution of NaCl (26.4% w/w) was added. The vials were immediately crimped with a gas-tight SPME cap. Each sample was prepared in one repetition. In addition, samples for background measurement (blanks) were prepared (2 mL of a saturated solution of NaCl), as well as quality control samples (QC). For the QC sample, 50 mg of the pooled samples were used, and the QC sample was prepared in six repetitions. It should be noted that during the development of the SPME-GC- HRMS method, the different amounts of the sample (25 mg, 50 mg, and 100 mg) were considered, as well as the addition of a saturated solution of NaCl. Finally, the combination of 50 mg of the sample with the 2 mL of saturated solution of NaCl was chosen.

Artificially adulterated saffron samples

In this part of the experiment, saffron was artificially adulated and treated by mixing it with selected plant materials in different proportions (percentage by mass – % w/w). Specifically, for the mixtures of saffron with turmeric, paprika, safflower, and calendula, the amounts of saffron/adulterant were as follows: 80/20% w/w, 90/10% w/w, 96/4% w/w, and 98/2% w/w. The artificially adulterated, authentic saffron and pure adulterant samples were prepared as described in Sect. "Non-target screening of volatile profiles of saffron and adulterants." In two repetitions.

SPME‑GC‑HRMS

For the extraction of head-spaced volatiles, the HS-SPME technique was used. The SPME parameters were first optimized to achieve the most efficient extraction of volatiles to the SPME fiber; simultaneously, it was crucial to test different desorption times to ensure the complete desorption of volatiles from the fiber to avoid cross-contamination. For the extraction temperature, 40 °C, 45 °C, and 50 °C were tested. Furthermore, the extraction time was tested at 10 minutes, 15 minutes, and 20 minutes. As for desorption time, 1 min and 3 min were considered. The final parameters of the SPME method are described below. The temperature of incubation and extraction was 45 °C, sample incubation time was 20 min, extraction time was 10 min, and sample desorption to the GC system was 3 min (desorption temperature was 240 °C). The Agilent 7200B system consists of an Agilent 7890B gas chromatograph equipped with a multimode inlet, PAL RSI 85 for automated headspace–solid-phase microextraction (HS–SPME) and direct injection, and quadrupole – time of flight mass spectrometer (Q-TOF) (Agilent Technologies, Palo Alto, California, USA) was employed. Mass-Hunter GC/MS Acquisition (Agilent Technologies, Palo Alto, California, USA) software (B.10.0.384.1) was used for instrument control and data acquisition. Sample components were separated on a 30 m HP–5MS capillary column (0.25 mm id, film thickness: 0.25 µm; Agilent Technologies, Palo Alto, California, USA). Samples were injected using split mode (5:1), the injector temperature was 240 °C, and the oven temperature program was as follows:

40 °C (2 min), 2 °C/min to 64.5 °C, 0.5 °C

min to 70 °C, 2 °C/min to 80 °C, 5 °C/min to 100 °C, 0.5 °C

min to 114 °C, 2 °C/min to 120 °C, 5 °C/min to 184 °C

20 °C/min to 240 °C (5 min)

The mass spectrometric detector was operated in the electron ionization (EI) mode. The temperature of the transfer- line was 280 °C, and the ion source temperature was 230 °C. The mass range setting was 40–550 m/z, the acquisition speed was five specter/s, and the resolution of the mass analyzer was set > 12,500 (FWHM).

GC–MS data processing and chemometric analysis

First, data obtained by SPME-GC-HRMS was transferred to the SureMass format. Then, our SureMass data was subjected to peak detection, deconvolution, and library search (NIST17.L). The data was then converted to CEF files (.CEF). All these data processing steps were performed in MassHunter Unknowns Analysis (Version B.10.1). Based on the converted QC CEF file in MassHunter Quantitative Analysis (Version.B.10.1) was created a data processing method and performed a recursive analysis. All software was from Agilent Technologies, Palo Alto, California, USA. After completing a recursive analysis, data were exported to the MetaboAnalyst 5.0 software (https://www.metab oanalyst.ca). Employing this software, the following data treatment operations were performed: sum-normalization (relative intensities of each signal obtained by division of the sum of all signals), log transformation, and Pareto scaling. These data processing operations ensure low abundant compounds' normal distribution and higher significance. In addition, in MetaboAnalyst, principal component analysis (PCA) was performed to oversee exported data. Analysis of Variance (ANOVA) or t-test (False Discovery Rate (FDR), p-value < 0.05) was employed to filter out the statistically insignificant compounds. The filtered data were normalized by a sum in MS Excel and loaded to SIMCA (Sartorius, Göttingen, Germany), where unsupervised principal component analysis (PCA) and supervised partial least squares discriminant analysis (PLS-DA) were performed. R2 (cumulative) and Q2 (cumulative) parameters were obtained for each model by fourfold internal cross-validation and were used to determine the validity of the models. R2 (cum) shows the variation described by all components in the model, and Q2 (cum) is a criterion of how accurately the model can predict class membership. The prediction and recognition abilities of the models were also obtained by fourfold internal cross-validation. The significant volatile compounds ('markers') were selected based on their Variable Importance in Projection (VIP) score in PLS-DA models. This study considered the most significant compounds with a VIP score of more than 1.0. The Pattern Search function of MetaboAnalyst software, which allows the specification and search for specific patterns of interest in the data, was also used to detect characteristic 'markers.' This function uses a template matching method, and the results are expressed as a ranked list of variables with the Spearman correlation coefficient and p-value. This study contemplated compounds with the Spearman correlation coefficient ≥ 0.80. Identification of selected 'markers' was performed using mass spectra NIST 17 library; for tentative identification, only compounds with match factor ≥ 750 and relevant Kovats retention indexes (RI) relative to n-alkanes (C7–C30) were considered. The compliance of the exact mass of detected ions (mass error < 5 ppm) and isotopic pattern were used to confirm the identification.

Results and discussion

Optimization of SPME‑GC‑HRMS

For this study, HS-SPME was the technique of choice. Sample preparation, extraction temperature, and desorption conditions in the GC injector are described in detail in the sections. "Chemicals and Materials "Sample Preparation" "SPME-GC-HRMS" Regarregard different patterns of volatiles in the analyzed plant species, dendrimeric and ric, were used for method optimization. To control the repeatability of measurements, the QC sample (described in Sect. "Non-target screening of volatile profiles of saffron and adulterants") containing all the tested plants was analyzed throughout the study (the retention time tolerance was 0.2 min, and the repeatability of the peaks areas 5%).

Non‑target screening of saffron and plant adulterants

In the first phase of the experiments, 38 authentic safety samples and 25 samples of different plant materials representing four plant species were analyzed. The chromatograms in Fig. 1 demonstrate the difference in the main volatile profiles between the saffron samples and potential contaminants. While volatiles with retention times not exceeding 45 min under used GC conditions on HP–5MS capillary column dominated in saffron (A), capsicum (D), and safflower (E), volatiles with relatively high intensities with retention times up to 60 min were present in turmeric (B) and calendula (C). When using the m/z value range 40–500, a high number of compounds ranging on average from 150 (capsicum) to 550 (turmeric) was detected after applying the deconvolution function.

Results of chemometric data evaluation for all samples

To investigate the differences and get an in-depth insight into the volatile profiles of saffron and its potential contaminants, a chemometric analysis starting with PCA and followed by PLS-DA was performed. The score plot in Fig. 2 shows the distinct separation of saffron samples from other plant matrices. The clustering was relatively tight despite the diverse geographic origin of measured saffron samples, their harvest time, and types (both powders and stigmas were contained in the experimental set). Regarding matrices of other botanical origin, samples of turmeric were well separated from all other plant adulterants, which overlapped to some extent; no improvement was obtained by supervised PLS-DA analysis. The tight cluster of QC samples documents a good reproducibility of the measurements.

 

Fig 1 GC–HRMS total ion chromatograms (40–550 m/z) of saffron (A), turmeric (B), calendula (C), capsicum (D), and safflower (E) samplesFig. 1 GC–HRMS total ion chromatograms (40–550 m/z) of saffron (A), turmeric (B), calendula (C), capsicum (D), and safflower (E) samples

 

Fig. 2 PCA score plot based on the data obtained by non-target screening of headspace volatiles of saffron, turmeric, calendula, capsicum, and safflower

Chemometric analysis of saffron vs. single plant species (binary models)

Although PCA would allow the identification of 'pure' authentic saffron, total replacement by other plant species is improbable under actual life conditions; dilution is more common. On these terms, knowing the adulteration's specific 'markers' is essential for fraud detection. To simplify their selection, binary PLS-DA classification models for saffron and potential adulterants were constructed (see Fig. 3), and then, internal fourfold cross-validation was performed.Table 1. shows the models' excellent (100%) prediction and recognition abilities, as well as high values of Q2 and R2 parameters (the lowest but still very high as in the case of the 'Saffron and Safflower' binary model).

Selection and identification of characteristic 'markers'

In the next phase of experiments, deconvoluted peaks characterized by particular m/z value and retention time with VIP-scores higher than 1.0 were considered candidate significant' markers.' The other (complementary) tool that was used for the potential 'markers' selection was the Pattern search function of MetaboAnalyst software; in that case, the criterion was the value of the Pearson correlation coefficient (≥ 0.8 required). The identification of compounds that met at least one of these selection criteria was based on the match of (i) measured spectra with the records on the NIST library,(ii) accurate mass of detected m/z (mass error < 5 ppm) corresponding to the expected elemental composition, and (iii)

The compliance with Kovats indexes is on the respective capillary column. It should be noted that we identified several unique volatile components [19–21] characteristic of saffron aroma (see Table 2); nevertheless, as the aim of this study was to find out 'markers' of impurities, more emphasis is placed on the identification of volatiles occurring in potential plant adulterants. The results of their tentative identifications are summarized in Table 3. Some of them, such as α-phellandrene, β-myrcene, α-terpinene, p-cymene, D-limonene, and eucalyptol, were reported to be contained also in saffron [19, 20]. Nevertheless, their concentration in samples available in this study did not exceed the detection threshold.

Detection of plant adulterants in saffron samples

As mentioned in the Introduction, saffron is a high-value commodity susceptible to fraud, typically by dilution by a cheaper plant material. Several admixtures were prepared to verify the function of 'markers' selected in the first phase of experiments to detect this type of fraud; the authentic saffron samples were artificially adulterated by turmeric, calendula, capsicum, and safflower. The sample preparation procedure is described in Sect. "Artificially adulterated saffron samples." Using markers summarized in Tables 2 and 3 for PCA and PLS-DA model construction, additions of potential plant contaminants as low as 2% w/w could be reliably identified. The paragraphs below provide more detailed information.

The PLS-DA score plot (see Fig. 4A) obtained for safe- from-turmeric admixtures shows good separation of the sample groups. The classification model was built based on 27 volatile compounds (16 'markers' of turmeric and 11 'markers' of saffron). The specific 'markers' of tur- meric were α-turmerone, ar-turmerone, cis-γ-atlantone, β-turmerone, 6S, 7R-bisabolene, and trans-α-atlantone, an example of specificity of β-turmerone characteristic 'marker' ion m/z = 105.0695 ± 5 ppm is illustrated in Fig. 4B.

 

Fig 3 Binary models for saffron and potential contaminants, PLS-DA score plots: A saffron and turmeric samples; B saffron and calendula samples; C saffron and capsicum samples; D saffron and safflower samplesFig. 3 Binary models for saffron and potential contaminants, PLS-DA score plots: A saffron and turmeric samples; B saffron and calendula samples; C saffron and capsicum samples; D saffron and safflower samples

 

Table 1 Parameters of the PLS-DA binary classification models for saffron and adulterants

Binarymodel                                          No.ofsamplesR2

Q2                    RecognitionabilityPrediction

 

 

 

 

(%)

ability (%)

'SaffronandTurmeric'

44

0.996

0.991

100

100

'SaffronandCalendula'

47

0.996

0.988

100

100

'SaffronandCapsicum'

44

0.996

0.985

100

100

'SaffronandSafflower'

42

0.989

0.973

100

100

Another plant material used for artificial adulteration was calendula. The statistical evaluation was performed based on 19 volatiles (8 'markers' of calendula and 11 'markers' of saffron). In the case of this classification model, the statistical difference was observed at all % w/w levels of the addition of calendula, which could be observed from the PLS-DA score plot (see Fig. 5A). The specific 'markers' of calendula were α-calcarine and thymoquinone. Interestingly, the latter compound's presence has not yet been reported in Calendula officinalis. Figure 5B, showing the extracted ion chromatograph (EIC) of thymoquinone char- characteristic 'marker' ion m/z = 164.0827 ± 5 ppm, documents its specificity.

Table 2. Identified volatile 'markers' of saffron with VIP value > 1.0, based on PLS-DA binary models and the Pearson correlation coefficient ≥ 0.8

Saffron' markers'

 

Retentiontime (min)

TentativeidentificationbyNIST

Matchfactor*

CAS

Kovatsindex exp

References NIST

 

2.15

2-Methylfuran

94.4

534–22-5

606

[5, 19–21]

8.32

2(5H)-Furanone

89.2

497–23-4

915

915

[5, 19–21]

10.01

6-Methyl-2-heptanone

75.7

928–68-7

954

957

[5, 19–21]

11.61

6-Methyl-5-hepten-2-one

90.0

110–93-0

993

988

[5, 19–21]

14.71

Dihydroisophorone

83.9

873–94-9

1037

1080

[19]

14.98

Isophorone

75.2

78–59-1

1045

1086

[20, 21]

23.50

4-Oxoisophorone

77.9

1125–21-9

1152

1145

[20, 21]

23.68

2-Hydroxy-4-oxo isophorone

80.4

4883–60-7

1230

1237

[22]

27.14

Safranal

75.8

116–26-7

1199

1207

[19]

33.83

Mint furanone

75.3

13,341–72-5

1302

1314

[23]

40.95

4-Hydroxy-2,6,6-trimethyl-3-oxy- clohexa-1,4-dienecarbaldehyde

81.4

35,692–95-6

1382

1396

[20]

*Match factor is a comparison of the unknown's mass spectrum's peaks the peaks in the NIST library

In Fig. 6A, the PLS-DA score plot demonstrates the capsicum separation of the artificially adulterated saffron sample. For constructing the classification model in this case, 21 compounds were used (10 'markers' of capsicum and 11 'markers' of saffron). Among the capsicum-specific 'markers' were 3-carene, D-carvone, and 2,2,6-tri- methylcyclohexanone. The EIC of typical 'marker' ion m/z = 108.0929 ± 5 ppm for D-carvone is shown in Fig. 6B. The last plant material tested for artificial saffron adulteration was safflower. From the PLS-DA score plot (see Fig. 7A), all % w/w levels were separated. For the construction of this classification model, 20 volatiles (9' safflower markers' and 11 'markers' of saffron) volatiles were used. Specific and characteristic 'markers' of safflower were 3-methyl butanal, 3-methyl-1-butanol, and γ-terpinene. Figure 7B shows the EIC of specific terpinenen 'marker' ion m/z = 136.1070 ± 5 ppm.

The critical comparison of the results with the studies previously performed

In the final phase of this study, we compared the results we obtained with similar studies concerned with the use of GC for saffron authentication. For instance, Morozzi et al. [38] used flash HS-GC-FID to analyze 28 saffron samples and admix- tures with turmeric and calendula (20, 15, 10, and 5% w/w). Using the created chemometric models, it was possible to separate the mixed samples from the authentic ones in the case of both pollutants, even at the lowest level of contamination. However, no concrete' markers' were identified concerning using FID for volatile fingerprinting. Di Donato et al. [6]. also studied the adulteration of authentic saffron by dilution with three plant materials (calendula, safflower, and turmeric). They used SPME–GC–MS (ion trap mass analyzer) to screen volatile substances. Like in our study, the presence of impurities at low levels, 2–3% w/w, could be detected; nevertheless, contrary to our results, only one (trans-anethole for safflower and ar-turmerone for turmeric), or in maximum two (γ-cadinene and δ-cadinene for calendula) 'markers' were identified for respective plants. These substances were also detected in our samples of adult tenants; nevertheless, their VIP scores, except for ar-turmerone, were relatively low compared to those found in our study. The discrepancy in identified 'markers' might be due to reasonably different chromatographic conditions while starting the temperature of GC separation in Di Donato et al. [6]. The study was 120 °C; we jumped at 40 °C, enabling our detection and identification of more volatile 'marker' compounds with high VIP scores. It is worth mentioning that a complementary authentication study employing ultra-high performance reverse-phase liquid chromatography coupled to tandem high-resolution mass spectrometry (UHPLC- HRMS/MS) was performed in our lab on the same unique samples set by Ryparova Kvirencova et al.[39]. Comparable results in Q2 and R2 parameters and the ability to detect the addition of contaminants at the level of 2% w/w were achieved.

In conclusion, the implementation of the SPME-GC-HRMS technique for the detection of saffron dilution by the most common contaminants (turmeric, calendula, safflower, and capsicum) enabled the distinguishing authentic saffron samples from all other potential plant adulterants, all constructed classification models showed 100% prediction and recognition abilities and excellent Q2 and R2 parameters (> 0.9). Another outcome of the study was the identification of eleven unique 'markers' for saffron and twenty-seven volatile 'markers' for four potential plant adulterants. This enables fraud detection and recognition of the type of plant adulterant. When targeting identified 'markers' of plant pollutants, their presence in admixtures with saffron, even at the level of 2% w/w, could be recognized. The knowledge of respective 'markers' allows for simplified authentication procedures. For target analysis, standard triple quadrupole mass analyzers can be used.

Table 3 Identified volatile 'markers' of potential plant adulterants of saffron with VIP value > 1.0, based on PLS-DA binary models and the Pearson correlation coefficient ≥ 0.8

 

Retention

TentativeidentificationbyNIST

Matchfactor*

CAS

Kovats

index

Plant adulterants**

References

time (min)

 

 

 

exp

NIST

 

 

2.58

3-Methylbutanal

93.4

590–86-3

669

SF

[24]

3.71

3-Methyl-1-butanol

92.8

123–51-3

760

SF

[24]

6.44

(E)-2-Hexenal

89.9

6728–26-3

857

850

CL, SF

[24]

8.74

Methylhexanoate

77.1

106–70-7

925

925

CL, SF

[24]

9.02

α-Thujene

87.0

2867–05-2

932

929

TU,CL,CP, SF

[24–34]

10.93

Sabinene

86.6

3387–41-5

974

975

TU,CP

[25, 31, 32, 35]

11.83

β-Myrcene

89.5

123–35-3

991

984

TU,CP

[25, 31, 32, 35]

12.47

α-Phellandrene

93.3

99–83-2

1006

998

TU

[25]

12.8

3-Carene

89.8

13,466–78-9

1008

1003

CP

[31]

13.23

α-Terpinene

79.0

99–86-5

1015

1010

TU,SF

[24, 25]

13.75

p-Cymene

88.4

99–87-6

1022

1028

TU,CL,CP

[25, 30, 31]

14.01

D-Limonene

76.2

5989–27-5

1026

1023

TU,CL,CP, SF

[24–34]

14.14

Eucalyptol

87.2

470–82-6

1030

1022

TU,CP

[25, 31, 32, 35]

14.42

2,2,6-trimethylcyclohexanone

84.7

2408–37-9

1032

1026

CP

[32]

16.27

γ-Terpinene

80.1

99–85-4

1057

1056

SF

[36]

18.75

α-Terpinolene

79.8

586–62-9

1086

1089

TU,CP

[25, 31, 32, 35]

19.36

2-Nonanone

89.3

821–55-6

1092

1090

TU

[25]

25.65

Terpinen-4-ol

82.8

562–74-3

1180

1182

CL, SF

[30, 36]

29.48

D-Carvone

87.4

2244–16-8

1237

1238

CP

[31]

29.91

Thymoquinone

81.6

490–91-5

1244

1249

CL

58.69

α-Calacorene

78.1

21,391–99-1

1545

1546

CL

[30, 37]

62.7

α-Tumerone

75.2

180,315–67-7

1591

1632

TU

[25]

66.15

ar-Turmerone

75.4

532–65-0

1665

1672

TU

[25, 35]

66.28

cis-γ-Atlantone

87.4

108,549–48-0

1668

1679

TU

[35]

67.44

β-Turmerone

94.6

87,440–60-6

1696

1701

TU

[25]

68.88

6S,7R-Bisabolone

86.3

72,441–70-4

1746

1754

TU

[25]

69.72

trans-α-Atlantone

97.2

108,645–54-1

1774

1785

TU

[25, 35]

*Match factor is a comparison of the unknown's mass spectrum's peaks the peaks in the NIST library;

**Plant adulterants: TU turmeric, CL calendula, CP capsicum, SF safflower