Cigarette smoke component difference identification and difference quantification method

Abstract

The invention discloses a cigarette smoke component differential component identification and difference quantification method. The method comprises the following steps: S1, carrying out quantitative detection on cigarette smoke components; s2, aiming at a quantitative detection result, performing difference analysis on the data by adopting a statistical method to obtain a visual score chart with identification difference; s3, based on the visual score plot, drawing an ellipse covering all data of the group through grouping, determining the mass center of the ellipse, and calculating the distance between the mass centers of the ellipses of different groups, namely the difference quantization between the smoke component groups; and S4, extracting components having great contribution to the difference to formulate a VIP compound list. Compared with the prior art, the method has the advantages that the integrated breakthrough of'overall difference identification-difference evaluation-key contribution component positioning 'of the cigarette smoke components is realized, and the method for identifying the difference components of the cigarette smoke components and quantifying the difference is provided from knowing the difference to knowing the components with great contribution to the difference.

Description

Technical Field

[0001] This invention relates to the field of cigarette smoke analysis technology, specifically to a method for identifying and quantifying differences in cigarette smoke components. Background Technology

[0002] Cigarette smoke, a complex byproduct of tobacco combustion, directly determines the product's quality, sensory experience, and safety characteristics. It contains thousands of chemical substances, including tar, nicotine, volatile aroma compounds, and harmful trace components (such as nitrosamines and polycyclic aromatic hydrocarbons). In tobacco industry product development, quality control, and market competition, accurately identifying the overall differences in smoke composition between different cigarette products (e.g., different brands) and under different smoking environments in different consumer markets, and further pinpointing the key components that play a dominant and significant role in these differences, is crucial for achieving precise product optimization, stable quality control, and replication of core styles. For example, when benchmarking against standard products, it's necessary to identify which component differences cause discrepancies in aroma, strength, etc., between the target product and the benchmark. In batch quality fluctuation analysis, it's necessary to quickly identify which abnormal changes in components have triggered overall quality deviations. However, existing technologies still have significant shortcomings in meeting this core requirement.

[0003] Currently, researchers typically use sophisticated instruments such as gas chromatography-mass spectrometry (GC-MS), high-performance liquid chromatography (HPLC), and ultra-high-performance combined phase chromatography (UPC²) to quantitatively analyze specific components in cigarette smoke. For example, according to industry standards, GC-MS is used to determine the content of 16 polycyclic aromatic hydrocarbons in cigarette smoke, HPLC is used to detect polyphenolic aroma precursors, and ion chromatography (IC) is used to analyze organic acids in cigarette smoke. Simultaneously, rapid detection techniques such as near-infrared spectroscopy (NIRS) and Raman spectroscopy can quickly screen conventional indicators such as tar and nicotine in batches of samples, shortening the detection cycle. These techniques can obtain content data of various components in different samples, providing basic data support for subsequent analysis. However, currently, based on this cigarette smoke composition data, it is only possible to preliminarily determine that "there are component differences" between different samples; it is not possible to evaluate the differences or discrepancies in cigarette smoke composition.

[0004] Since differences in smoke composition are ultimately reflected in sensory experience (such as aroma intensity, off-flavor strength, and aftertaste comfort), existing technologies often combine professional sensory evaluation systems. This involves evaluators scoring cigarette sensory indicators to indirectly infer the direction of component differences. For example, using industry-standardized sensory evaluation scales, samples are scored based on dimensions such as aroma type, aroma quantity, irritation, and aftertaste. If the target sample and benchmark sample differ significantly in the "aroma quantity" indicator, the content of volatile aroma components in the two types of samples is further compared to infer the possible influencing components. However, this process is only a preliminary correlation between "sensory differences and component inference," lacking precision and certainty.

[0005] Therefore, the core limitation of existing technology is that it can only determine whether there are "component differences" or "overall characteristic differences" between different samples, but it cannot quantify the contribution of each component to the overall difference, let alone locate the key components that dominate the difference.

[0006] To address the above problems, this invention is proposed. Summary of the Invention

[0007] The present invention aims to overcome the shortcomings of the prior art and provide a method for identifying and quantifying the differences in cigarette smoke components.

[0008] The technical solution adopted in this invention is as follows:

[0009] This invention provides a method for identifying and quantifying differences in cigarette smoke components. It includes the following steps:

[0010] S1, quantitative detection of cigarette smoke components;

[0011] S2, based on the quantitative detection results, uses statistical methods to perform differential analysis on the data and obtain a visual score chart that can identify differences;

[0012] S3, and based on the visualized score map, by grouping, draw an ellipse covering all the data in the group, determine the centroid of the ellipse, and calculate the distance between the centroids of the ellipses of different groups, which is the quantification of the difference between the flue gas component groups;

[0013] S4, extract the components that contribute significantly to the difference and compile a VIP compound list.

[0014] Preferably, in step S1, the components of cigarette smoke are quantitatively detected using gas chromatography-mass spectrometry (GC-MS) or gas chromatography-triple quadrupole mass spectrometry (GC-MS / MS).

[0015] Preferably, in step S1, ① quantitative detection of cigarette smoke components under different smoking environment conditions, including temperature, humidity, and altitude; ② quantitative detection of cigarette smoke components from different brands, formula iterations, and production batches.

[0016] Preferably, in step S2, principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS-DA) are used to perform differential analysis on the data, and the resulting score plots are PCA and OPLS-DA score plots.

[0017] Preferably, in step S2, the following steps are performed: ① a difference analysis is conducted on the cigarette smoke component detection data under different smoking environment conditions, including temperature, humidity, and altitude; ② a difference analysis is conducted on the cigarette smoke component detection data of different brands.

[0018] Preferably, in step S2, PCA is a data-driven unsupervised method, mainly suitable for exploratory analysis; OPLS-DA is a goal-oriented supervised method, mainly suitable for classification and variable selection.

[0019] PCA and OPLS-DA Analysis: For smoke component data, PCA and OPLS-DA statistical methods were used to conduct differential analysis on smoke component detection data under different smoking environmental conditions (temperature, humidity, altitude), obtaining visualized PCA and OPLS-DA score plots that can identify differences. When conducting differential analysis of smoke component, it is important to note that: PCA is a data-driven unsupervised method, mainly suitable for exploratory analysis; OPLS-DA is a goal-oriented supervised method, mainly suitable for classification and variable selection.

[0020] Preferably, in step S3, the grouping is based on ① different smoking environment conditions, including temperature, humidity, and altitude; and ② different brands, formula iterations, and production batches.

[0021] Scoring ellipse drawing and centroid distance calculation: Based on the visualized PCA scoring chart, ellipses covering all data in each group are drawn based on ① different smoking environment conditions (temperature, humidity, altitude) and ② different brands, formula iteration versions, and different production batches. The centroids of the ellipses are determined, and the distance between the centroids of the ellipses in different groups is calculated, which is the quantification of the differences between the smoke component groups.

[0022] Preferably, in step S3, in the OPLS-DA data analysis model, VIP>1 is used as the initial screening threshold for important indicators to measure the importance of each "feature / variable" to the model's "predictive ability".

[0023] VIP compound screening: This is an indicator used in OPLS-DA to evaluate the importance of each feature or variable in the data analysis model to the model's predictions. This document uses VIP>1 as the initial screening threshold for importance indicators. VIP>1 indicates that the variable's contribution to the model's classification / prediction is higher than the average level of all variables.

[0024] Compared with the prior art, the present invention has the following beneficial effects:

[0025] 1. Compared with the prior art, the present invention has achieved an integrated breakthrough in the "overall difference identification - gap evaluation - key contributing component location" of cigarette smoke components, and further from knowing the differences to knowing the components that contribute the most to the differences, providing a method for identifying the difference components and quantifying the gaps in cigarette smoke components.

[0026] This invention first uses PCA and OPLS-DA statistical methods to perform differential analysis on the data, and obtains visualized PCA and OPLS-DA score maps that can identify differences. These score maps can be used to identify the overall differences in cigarette smoke components.

[0027] Furthermore, based on the visualized score map, by grouping, an ellipse covering all the data in the group is drawn, the centroid of the ellipse is determined, and the distance between the centroids of different groups of ellipses is calculated to quantify the differences between the flue gas component groups.

[0028] Furthermore, this invention is used to evaluate the importance of each feature or variable in the OPLS-DA data analysis model to the model prediction. It uses VIP>1 as the initial screening threshold for importance indicators and extracts components that contribute significantly to the difference to formulate a VIP compound list.

[0029] 2. This invention provides a foundation for cigarette improvement. By linking the overall difference identification results, gap quantification data, and key contributing component location results, this invention constructs a complete closed loop of "difference phenomenon - core cause - improvement solution." The method of this invention can extract components that contribute significantly to the difference to create a VIP compound list, providing direction for the next step of cigarette improvement, significantly reducing R&D trial-and-error costs, and improving product optimization efficiency. This closed-loop capability of the invention allows the analysis results to go beyond the data level and be directly transformed into implementable technical solutions, greatly enhancing the application value of the technological achievements. Attached Figure Description

[0030] Figure 1 This is a schematic diagram of the process for identifying and quantifying the differences in cigarette smoke components according to the present invention.

[0031] Figure 2 PCA analysis score chart (using chempattern software) of the detection data of fruit aroma components in the mainstream smoke of five brands of cigarettes in Example 1 (difference identification).

[0032] Figure 3 The OPLS-DA analysis score chart (using SIMCA software) shows the detection data of fruit aroma components in the mainstream smoke of five brands of cigarettes in Example 1 (difference identification).

[0033] Figure 4The VIP scores of the OPLS-DA (using SIMCA software) for detecting fruit aroma components in the mainstream smoke of five brands of cigarettes in Example 1 (VIP compound list).

[0034] Figure 5 PCA score plots (using SIMCA software) of the target component detection data in the mainstream cigarette smoke of different smoking environments (cities) in Example 2 (difference identification).

[0035] Figure 6 OPLS-DA score map (using SIMCA software) of the target component detection data in the mainstream cigarette smoke of different smoking environments (city) in Example 2 (difference identification).

[0036] Figure 7 The VIP score map of OPLS-DA (using SIMCA software) for the detection data of targeted components in the mainstream cigarette smoke of different smoking environments (city) in Example 2 (VIP compound list). Detailed Implementation

[0037] The present invention will be further described below through embodiments, but is not limited to these embodiments. Experimental methods not specifically described in the embodiments generally use conventional conditions and conditions described in manuals, or conditions recommended by the manufacturer. The general equipment, materials, reagents, etc., used are all commercially available unless otherwise specified.

[0038] Example 1: Different cigarette brands (five cigarette products)

[0039] This embodiment provides a method for identifying and quantifying the differences in cigarette smoke components, the specific steps of which are as follows:

[0040] S1, targeted quantitative detection of cigarette smoke components:

[0041] The TD-GC-MS / MS method was used to determine the fruit aroma components in single-puff cigarette smoke. Fifty-eight fruit aroma components in the mainstream smoke of five brands of cigarettes (A, B, C, D, and E) were determined. The quantitative results were obtained by calculating the average value based on six measurements. The results are shown in Table 1 below.

[0042] Table 1. Determination results of 58 fruit aroma components in the mainstream smoke of five cigarette products.

[0043] Serial Number Compound Name Cigarette A content (ug / cigarette) Cigarette B content (ug / cigarette) Cigarette C content (ug / cigarette) Drug content in cigarettes (ug / cigarette) Cigarette E content (ug / cigarette) 1 Ethyl propionate 0.012 0.035 0.017 0.000 0.066 2 n-Propyl acetate 0.022 0.032 0.012 0.006 0.006 3 2-Hexanone 0.104 0.304 0.092 0.146 0.158 4 Ethyl butyrate 0.025 0.076 0.036 0.040 0.039 5 Butyl acetate 0.000 1.441 0.004 0.498 0.000 6 Ethyl isovalerate 0.000 0.002 0.000 0.001 0.001 7 Isoamyl acetate 0.007 0.020 0.000 0.010 0.010 8 2-Hepagonal 0.018 0.941 0.021 0.012 0.026 9 5-Methyl-3-hexen-2-one 0.007 0.017 0.000 0.007 0.007 10 Ethyl valerate 0.000 0.992 0.000 0.000 0.000 11 Amyl acetate 0.134 0.284 0.000 0.074 0.246 12 γ-valerol 0.000 0.073 0.004 0.014 0.000 13 3-Methyl-2-(5H)-furanone 0.235 0.265 0.000 0.271 0.000 14 6-Methyl-5-hepten-2-one 0.026 0.021 0.019 0.016 0.019 15 2-Octanone 0.003 0.005 0.004 0.003 0.003 16 Butyl butyrate 0.000 0.001 0.000 0.007 0.000 17 Ethyl hexanoate 0.000 0.014 0.001 0.001 0.000 18 Isoamyl isobutyrate 0.255 0.134 0.000 0.131 0.000 19 4-Isopropyltoluene 0.034 0.077 0.047 0.040 0.049 20 Phyllanthrene 0.564 1.146 0.965 0.640 0.768 21 Dimethyl succinate 0.034 0.055 0.028 0.028 0.042 22 α-terpinene 1.644 4.491 2.332 1.762 2.257 23 2,5-Dimethyl-4-hydroxy-3(2H)-furanone 0.000 0.000 0.000 0.000 0.000 24 (±)-Limonene 0.000 0.000 0.000 0.000 0.000 25 Ethyl levulinate 0.181 0.247 0.000 0.118 0.293 26 4-Isopropenyltoluene 0.016 0.046 0.019 0.016 0.025 27 Amyl butyrate 0.000 0.000 0.000 0.000 0.000 28 Allyl hexanoate 0.000 0.000 0.000 0.000 0.000 29 2-Methylbutyric acid-2-methylbutyl ester 0.000 0.011 0.000 0.027 0.020 30 Nononal 0.078 0.023 0.088 0.011 0.119 31 Benzaldehyde dimethyl acetal 0.000 0.009 0.000 0.001 0.000 32 2-Methylbutyric acid-3-methylbutyl ester 0.000 0.105 0.000 0.006 0.073 33 Isoamyl isovalerate 0.000 2.056 0.000 95.668 0.000 34 benzyl acetate 0.000 0.000 0.000 103.250 0.000 35 Diethyl succinate 0.000 0.000 0.000 4.586 0.000 36 Ethyl octanoate 0.000 0.020 0.092 4.808 0.000 37 decanal 0.149 1.605 0.000 0.000 0.000 38 γ-Octinolone 0.000 0.000 0.000 0.670 0.000 39 DL-diethyl malate 0.000 0.005 0.000 0.000 0.000 40 2-Undecane 0.000 0.000 0.012 2.714 0.090 41 benzyl butyrate 0.000 0.000 0.000 0.000 0.000 42 α-Damaconone 0.000 0.000 0.285 0.051 0.000 43 nerol acetate 0.023 0.035 0.026 0.007 0.026 44 Damasne 0.018 0.019 0.011 0.040 0.000 45 Methyl cinnamate 0.000 0.000 0.000 0.000 0.022 46 2-Phenylethyl isobutyrate 0.000 0.003 0.001 0.001 0.002 47 Ethyl decanoate 0.000 0.167 0.000 0.000 0.000 48 β-Damaconone 0.000 0.000 0.000 0.003 0.000 49 Anisyl acetate 0.000 0.000 0.000 0.000 0.000 50 Isoamyl benzoate 0.000 0.000 0.000 0.000 0.000 51 Cinnamyl acetate 0.002 0.005 0.001 0.003 0.001 52 Ethyl cinnamate 0.000 0.002 0.000 0.000 0.000 53 Isoamyl phenylacetate 0.000 0.001 0.000 0.001 0.000 54 Raspberry ketone 0.000 0.000 0.000 0.000 0.014 55 Neroli tertrol 0.011 0.027 0.009 0.012 0.014 56 γ-Undecanoic acid lactone 0.000 0.000 0.000 0.000 0.000 57 benzyl benzoate 0.000 0.000 0.007 0.000 0.003 58 Benzyl cinnamate 0.000 0.000 0.000 0.000 0.000

[0044] S2. For the quantitative detection results, statistical methods were used to perform differential analysis on the data, obtaining a visual score chart that identifies these differences. The PCA analysis score chart of the fruit aroma components in the mainstream smoke of five cigarette brands (using Chempattern software) is shown below. Figure 2 The OPLS-DA analysis score chart of the fruit aroma components in the mainstream smoke of five cigarette brands (using SIMCA software) is shown below. Figure 3 ;

[0045] S3, and based on the visualized score map, by grouping, draw an ellipse covering all the data in the group, determine the centroid of the ellipse, and calculate the distance between the centroids of the ellipses of different groups, which is to quantify the difference between the flue gas component groups. The distance results between the centroids of the ellipses of different groups are shown in Table 2.

[0046] Table 2. Results of distances between the centroids of different groups of ellipses

[0047] PCA centroid distance between groups Cigarette A Cigarette B Cigarette C Cigarette D Cigarette E Cigarette A 0 / / / / Cigarette B 7.10 0 / / / Cigarette C 3.18 7.95 0 / / Cigarette D 0.71 7.11 2.48 0 / Cigarette E 2.67 6.02 5.65 3.27 0

[0048] S4, extract the components that contribute significantly to the difference and compile a VIP compound list.

[0049] The VIP scores of OPLS-DA (using SIMCA software) for detecting fruit aroma components in the mainstream smoke of five cigarette brands are shown below. Figure 4 The list of compounds that contributed significantly to the differences (VIP>1) is shown in Table 3.

[0050] Table 3 List of compounds (VIP>1)

[0051] Serial Number Compound Name VIP value Serial Number Compound Name VIP value 1 α-Damaconone 1.41424 14 Isoamyl acetate 1.12409 2 benzyl benzoate 1.40548 15 methyl 2,2-dimethoxypropionate 1.09987 3 Raspberry ketone 1.34352 16 Propyl acetoacetate 1.0966 4 n-Propyl acetate 1.24777 17 Ethyl butyrate 1.09495 5 2-Hydroxypropylacetate 1.24771 18 4-Isopropyltoluene (p-Cymene) 1.08223 6 Isoamyl isovalerate 1.22282 19 α-terpinene 1.07792 7 Ethyl levulinate 1.22154 20 Diethyl succinate 1.06324 8 Methyl nonyl ketone (2-undecane ketone) 1.21972 21 β-Damaconone 1.05386 9 Amyl acetate 1.20629 22 2-Hexanone 1.0486 10 Isoamyl isobutyrate 1.20389 23 Isoamyl phenylacetate 1.03049 11 Phyllanthrene 1.20173 24 4-Isopropenyltoluene (α,p-Styrane) 1.0181 12 Ethyl propionate 1.15224 25 5-Methyl-3-hexen-2-one 1.01055 13 2-Methylbutyric acid 2-methylbutyl ester 1.13815

[0052] Example 2: Different smoking environments (same cigarette product in different smoking environments)

[0053] Example 2

[0054] This embodiment provides a method for identifying and quantifying the differences in cigarette smoke components. The specific steps are the same as in Embodiment 1, except that the cigarettes and the smoking environment (city) are different.

[0055] S1, quantitative detection of cigarette smoke components:

[0056] Cigarette Z was targeted for analysis using multiple reaction monitoring (MRM) mode of TD-GC-MS / MS, with six parallel collections of cigarette smoke samples from the same city. A total of 123 target compounds were detected in cigarette smoke samples from three cities (Kunming, Chengdu, and Hangzhou). Quantitative analysis of these chemical components was performed, and the results are shown in Table 4.

[0057] Table 4. Quantitative analysis results of the chemical composition of cigarette smoke samples from three cities (Kunming, Chengdu, and Hangzhou).

[0058] Serial Number Compound Name Chengdu (ug / piece) Nanjing (ug / piece) Kunming (ug / piece) 1 2-Pentanone 0.308222 0.195392 0.386532 2 Ethyl propionate 0.02788 0.013654 0.020222 3 n-Propyl acetate 0.006258 0.006668 0.012322 4 Pyridine 0.004894 0.01525 0.019022 5 Pyrrole 0.094724 0.06814 0.049408 6 2-Hexanone 0.01726 0.006504 0.014926 7 Ethyl butyrate 0.009482 0.00695 0.011814 8 Butyl acetate 0.000598 0.003912 0.000742 9 trans-2-hexenal 0.034656 0.017434 0.024326 10 Ethyl isovalerate 0.001034 0.00031 0.000186 11 Furfuryl alcohol [2-furanethanol] 0.812236 0.452518 0.173006 12 α-Angelactone 1.168394 0.627066 0.420314 13 Isoamyl acetate 0 0.000494 0.00125 14 2-Hydroxypropylacetate 0.308126 0.155766 0.11667 15 4-Cyclopenten-1,3-dione 1.188036 0.448308 0.30567 16 2-Hepagonal 0.005852 0.004126 0.003306 17 5-Methyl-3-hexen-2-one 0.000796 0.000488 0.000548 18 Naphthalene 0 0.080996 0 19 2-Hydroxy-2-cyclopenten-1-one 0.002357 0.001338 0.001386 20 methyl 2,2-dimethoxypropionate 0.115422 0.040306 0.030212 21 (+)-α-pinene 0.366508 0.341486 0 22 γ-valerol 0.007898 0.176874 0.1357 23 3-Ethylpyridine 0.005756 0.00364 0.00863 24 5-Methylfurfural 0.378074 0.284088 0.157212 25 3-Ethyltoluene 0.068944 0.055162 0.039208 26 Trimethylbenzene 0.003442 0.00271 0.003042 27 phenol 0.012348 0.053104 0.032854 28 6-Methyl-5-hepten-2-one 0.03649 0.025506 0.03296 29 2-Octanone 0.001756 0.002026 0.001326 30 Benzofuran 0.002882 0.00312 0.002234 31 3-Methylstyrene 0.01424 0.011274 0.010766 32 Butyl butyrate 0.67676 0.444782 0.642718 33 Ethyl hexanoate 0.002872 0 0.0016 34 sapinene 0.317624 0.21821 0.259068 35 (-)-β-pinene 0.010294 0.008908 0.006736 36 Isoamyl isobutyrate 0.014658 0.020964 0.024148 37 3-Methyl-1,2-cyclopentanedione 0 0.00023 0.000308 38 4-Isopropyltoluene (p-Cymene) 0.012096 0.009892 0.009276 39 Phyllanthrene 0.687494 0.453316 0.381116 40 2-Ethylhexanol (Isooctanol) 0.01007 0.01037 0.00735 41 benzyl alcohol 0.329134 0.222544 0.186192 42 Propyl acetoacetate 0.000312 0.000197 5.98E-05 43 Indene 0.009892 0.00748 0.005446 44 Dimethyl succinate 0.039278 0.018948 0.023188 45 3-carene 0.516394 0.407096 0.257136 46 phenylacetaldehyde 0.217564 0.163546 0.104596 47 α-terpinene 0.77667 0.630302 0.551572 48 γ-caprolactone 0.000863 0.000788 0.000417 49 2,5-Dimethyl-4-hydroxy-3(2H)-furanone 0 1.02E-05 0 50 Acetophenone 0.039998 0.03455 0.020686 51 Ethyl levulinate 0.000296 0.000205 0.000134 52 Linalool Oxide: Mixture of Isomers 0.00203 0.001678 0.001104 53 furanyl hydroxymethyl ketone 1.2377 0.852814 0.33062 54 2,3,5,6-Tetramethylpyrazine 0.000101 7.69E-05 4.32E-05 55 4-Isopropenyltoluene (α,p-Styrane) 0.014394 0.01045 0.008402 56 Linalool 0.002646 0.001694 0.0015 57 Nononal 0.156262 0.076866 0.029844 58 maltol 0.03468 0.023414 0.211644 59 Benzaldehyde dimethyl acetal 0.004392 0.003984 0 60 Isophorone 0.000399 0.212753 0.00011 61 2-Methylbutyric acid-3-methylbutyl ester 0 0.045736 0 62 Phenylacetyl alcohol 0.081126 0.041124 0.01821 63 Methylcyclopentenolone 0.913182 0.917978 0.478404 64 Isoamyl isovalerate 0.001346 48.47199 2.96E-05 65 (-)-Menthone 0.000548 0.000402 0 66 benzyl acetate 0.000264 8.53E-05 9.57E-05 67 (±)-Menthol 0.031992 0.018514 0.016422 68 4'-Methylacetophenone 0.019024 0.01424 0.007338 69 Diethyl succinate 0 0 4.8E-07 70 Artemisia argyi [Artemisia argyi, Anethole, Piperol Methyl Ether, 4-Allyl Anisole] 0.000702 0.00057 0.000198 71 α-Terpineol 0 0.001306 0.001808 72 Methyl salicylate 3.57E-07 4.82E-07 1.16E-07 73 decanal 0.07495 0.237828 0.646796 74 5-Hydroxymethylfurfural 0.171232 0.011222 0.017324 75 D-Citronellol 0.012892 0.003944 0.000368 76 (S)-(β)-citronellol 0.000132 2.8E-05 1.39E-06 77 Menthol [(+)-Pulleone] 0 1.161782 1.152766 78 Ethyl phenylacetate 0 0.041466 0.037874 79 p-Anisaldehyde [4-methoxybenzaldehyde] 0.000094 0.00012 0.000184 80 γ-Octinolone 0 0.300632 1.242192 81 Benzaldehyde propylene glycol acetal 0.023852 0.010246 0.008192 82 Citral [neraldehyde, geranialdehyde]: a mixture of cis and trans isomers 0.573044 0.59665 0.180282 83 Cinnamaldehyde 0 0.000636 0.003842 84 Indole 0.014192 0.010504 0.00346 85 Methyl nonyl ketone (2-undecane ketone) 0.017146 0.000614 4.862928 86 α-Methylcinnamaldehyde 0 0.001158 0.00136 87 Eugenol 0 0 0.01643 88 p-Methoxyacetophenone 0 0.00709 0.006056 89 α-Damaconone 0.673066 0.982544 0.986452 90 Neroli acetate (neroli acetate) 0.559074 0.392148 0.311384 91 Anisaldehyde dimethyl acetal 0.000608 0.000791 0.000684 92 Damasne 0 0 0.188236 93 Dihydrocoumarin 0 0 0.008354 94 Vanillin 0.17226 0.160748 0.090826 95 1-Tetradecene 0.042248 0.02897 0.024612 96 2-Acetamidothiazole 0 0.01354 0.047708 97 Ethyl decanoate 0 0.000204 0 98 β-Damaconone 0.04045 0.007216 0.034138 99 Anisyl acetate (4-methoxybenzyl acetate) 0 0 0.014636 100 β-Caryophyllene 0 0 0.03963 101 Coumarin 0.048056 0 0.009976 102 Cinnamyl acetate 0.015854 0.007928 0.024514 103 Hexyl butyrate 0.002335 0.000596 0 104 γ-decanolide 1.2E-06 5.37E-07 3.87E-07 105 Isoeugenol 0.590884 0.497582 0.250652 106 Geraniol 0.315658 0.215472 0.18358 107 Geraniol Acetone 0.020602 0.012613 0.011692 108 Ethyl lactate 1.08E-06 7.97E-06 1.39E-06 109 Ethyl vanillin 0 0 0.031662 110 Perilla stalk 0.085844 0.0547 0.031276 111 Ethyl cinnamate 0.002484 0.002472 0 112 Methyl vanillin 0.001272 0.002102 0 113 Isoamyl phenylacetate 0.00425 0 0 114 (+) Thujone (Kopalene) 0 0 0.001058 115 Myristyl ether 0.014534 0.018292 0.013048 116 Nerolidol: a mixture of cis and trans isomers 0.03586 0.027182 0.023092 117 trans, trans-farnesol 0.040886 0.028744 0.021844 118 Cedarwood 0.028464 0.026302 0.03451 119 Jasmine aldehyde propylene glycol acetal 0.082818 0.079128 0.049102 120 Methyl 2,4-dihydroxy-3,6-dimethylbenzoate 0.007534 0.00451 0.004514 121 benzyl benzoate 0.048158 0.036596 0.0252 122 methyl 1,4-methylpentadecanoate 0.00242 0.001602 0.001604 123 Benzyl cinnamate 0 0 0.006622

[0059] S2. For the quantitative detection results, statistical methods are used to perform differential analysis on the data, obtaining a visual score chart that identifies differences. The PCA score charts for the detection of targeted components in mainstream cigarette smoke in different smoking environments (cities) (using the Maiwei Cloud Platform) are shown below. Figure 5 OPLS-DA score plots (using SIMCA software) of targeted component detection data in mainstream cigarette smoke from different smoking environments (cities) are shown below. Figure 6 .

[0060] S3, and based on the visualized score map, by grouping, draw an ellipse covering all the data in the group, determine the centroid of the ellipse, and calculate the distance between the centroids of the ellipses of different groups, which is to quantify the difference between the flue gas component groups. The distance results between the centroids of the ellipses of different groups are shown in Table 5.

[0061] Table 5. Results of distances between the centroids of different groups of ellipses

[0062] PCA centroid distance between groups Chengdu Nanjing Kunming Chengdu 0 / / Nanjing 8.231824 0 / Kunming 14.376870 9.146720 0

[0063] S4, extract the components that contribute significantly to the difference and compile a VIP compound list.

[0064] VIP score plot of OPLS-DA for the detection data of targeted components in mainstream cigarette smoke in different smoking environments (urban) (using SIMCA software) is shown below. Figure 7 The list of compounds that contributed significantly to the differences (VIP>1) is shown in Table 6.

[0065] Table 6 List of compounds (VIP > 1)

[0066] Serial Number Compound Name VIP value Serial Number Compound Name VIP value 1 phenol 1.4643 30 Furfuryl alcohol [2-furanethanol] 1.10708 2 2-Hexanone 1.42276 31 α-Methylcinnamaldehyde 1.10637 3 β-Damaconone 1.4067 32 furanyl hydroxymethyl ketone 1.0911 4 Benzyl cinnamate 1.31257 33 2-Hydroxypropylacetate 1.08537 5 Ethyl vanillin 1.30647 34 methyl 2,2-dimethoxypropionate 1.08091 6 Damasne 1.2922 35 5-Hydroxymethylfurfural 1.08079 7 2,5-Dimethyl-4-hydroxy-3(2H)-furanone 1.26116 36 2-Acetamidothiazole 1.07543 8 Anisyl acetate (4-methoxybenzyl acetate) 1.25803 37 Methylcyclopentenolone 1.06076 9 β-Caryophyllene 1.25332 38 4'-Methylacetophenone 1.05836 10 Dihydrocoumarin 1.24453 39 Propyl acetoacetate 1.05441 11 Ethyl hexanoate 1.23276 40 5-Methyl-3-hexen-2-one 1.05351 12 maltol 1.23014 41 Cinnamaldehyde 1.03854 13 Butyl butyrate 1.21157 42 (S)-(β)-citronellol 1.03677 14 Cinnamyl acetate 1.20439 43 Phyllanthrene 1.03304 15 Benzaldehyde dimethyl acetal 1.19715 44 4-Isopropenyltoluene (α,p-Styrane) 1.03154 16 Isoamyl phenylacetate 1.18354 45 Isoamyl acetate 1.03102 17 3-Ethylpyridine 1.17358 46 α-Terpineol 1.02907 18 Ethyl propionate 1.16193 47 D-Citronellol 1.02835 19 Ethyl isovalerate 1.15886 48 Hexyl butyrate 1.02227 20 (+) Thujone (Kopalene) 1.14643 49 trans-2-hexenal 1.01715 21 (-)-Menthone 1.14502 50 phenylacetaldehyde 1.01687 22 2-Pentanone 1.14458 51 Methyl nonyl ketone (2-undecane ketone) 1.01232 23 4-Cyclopenten-1,3-dione 1.14007 52 5-Methylfurfural 1.01213 24 Methyl vanillin 1.13329 53 2,3,5,6-Tetramethylpyrazine 1.01101 25 α-Angelactone 1.13029 54 3-carene 1.00485 26 Isoamyl isovalerate 1.12403 55 (+)-α-pinene 1.00409 27 Coumarin 1.12348 56 Linalool 1.00327 28 Phenylacetyl alcohol 1.12101 57 Methyl 2,4-dihydroxy-3,6-dimethylbenzoate 1.00151 29 Ethyl butyrate 1.11696

[0067] The present invention has been described above by way of example. It should be noted that any simple modifications, alterations or other equivalent substitutions that can be made by those skilled in the art without creative effort without departing from the core of the present invention fall within the protection scope of the present invention.

Claims

1. A method for identifying and quantifying differences in cigarette smoke components, characterized in that, It includes the following steps: S1, quantitative detection of cigarette smoke components; S2, based on the quantitative detection results, uses statistical methods to perform differential analysis on the data and obtain a visual score chart that can identify differences; S3, and based on the visualized score map, by grouping, draw an ellipse covering all the data in the group, determine the centroid of the ellipse, and calculate the distance between the centroids of the ellipses of different groups, which is the quantification of the difference between the flue gas component groups; S4, extract the components that contribute significantly to the difference and compile a VIP compound list.

2. The method for identifying and quantifying differences in cigarette smoke components according to claim 1, characterized in that, In step S1, the components of cigarette smoke are quantitatively detected using gas chromatography-mass spectrometry (GC-MS) or gas chromatography-triple quadrupole mass spectrometry (GC-MS / MS).

3. The method for identifying and quantifying differences in cigarette smoke components according to claim 1, characterized in that, In step S1, ① quantitative detection of cigarette smoke components is performed under different smoking environment conditions, including temperature, humidity, and altitude; ② quantitative detection of cigarette smoke components is performed for different brands, formula iterations, and production batches.

4. The method for identifying and quantifying differences in cigarette smoke components according to claim 1, characterized in that, In step S2, principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS-DA) are used to perform differential analysis on the data. The resulting score plots are PCA and OPLS-DA score plots.

5. The method for identifying and quantifying differences in cigarette smoke components according to claim 1, characterized in that, In step S2, a difference analysis is performed on the cigarette smoke component detection data under different smoking environment conditions, including temperature, humidity, and altitude; and a difference analysis is performed on the cigarette smoke component detection data of different brands.

6. The method for identifying and quantifying differences in cigarette smoke components according to claim 1, characterized in that, In step S2, PCA is a data-driven unsupervised method, mainly suitable for exploratory analysis; OPLS-DA is a goal-oriented supervised method, mainly suitable for classification and variable selection.

7. The method for identifying and quantifying differences in cigarette smoke components according to claim 1, characterized in that, In step S3, the grouping is based on ① different smoking environment conditions, including temperature, humidity, and altitude; and ② different brands, formula iterations, and production batches.

8. The method for identifying and quantifying differences in cigarette smoke components according to claim 1, characterized in that, In step S3, in the OPLS-DA data analysis model, VIP>1 is used as the initial screening threshold for important indicators to measure the importance of each "feature / variable" to the model's "predictive ability".

Images