A method for designing a leaf group formula of a Chinese-style roasted tobacco type cigarette with ultra-low tar release

Abstract

The present application relates to a kind of Chinese-style smoked tobacco type cigarette ultra-low tar release product leaf group formula design method, comprising the following steps: establishing sample library, according to the function of tobacco raw material is classified, respectively obtain the initial sample group of style class module, quality enhancement fragrance class module and harmonize smoke module is highlighted;According to screening index, the tobacco in the initial sample group of three modules is sorted, and the selected sample group of three modules is obtained, and through single tobacco sensory quality evaluation, unqualified tobacco sample is removed, and the preferred sample group of three modules is obtained;Part or all of the preferred sample group of three modules is entered into formula personnel leaf group design, combined with product compliance sensory quality evaluation and tar content detection, and the ultra-low tar release leaf group formula is obtained.The design method provided by the present application can realize the ultra-low tar release of product while meeting the feeling, and it has important significance for the continuous development of low-tar cigarette product "reducing harm".

Description

Technical Field

[0001] This invention relates to the field of tobacco technology, and in particular to a leaf blend formulation design method for a Chinese-style flue-cured cigarette with ultra-low tar release. Background Technology

[0002] As research into the health effects of smoking deepens, the development of low-tar tobacco products has become an inevitable direction for the global tobacco industry. my country, with its large proportion of flue-cured cigarettes and high tar content, faces an even more urgent task in developing low-tar tobacco products.

[0003] Reducing tar and harm is an undeniable responsibility of the tobacco industry. The industry's medium- and long-term science and technology innovation plan has continuously laid out plans for tar reduction and harm reduction, and has set higher requirements. Because cigarette tar contains a large number of aroma components, reducing the total amount of tar will result in a bland smoke that fails to meet consumer needs. Therefore, to reduce tar while ensuring cigarettes still offer a certain level of satisfaction and aroma, higher demands are placed on cigarette blending technology. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a method for designing leaf blend formulations for ultra-low tar release products in Chinese-style flue-cured cigarettes. This invention introduces screening indicators from three modules—character-enhancing tobacco leaves, aroma-enhancing and quality-improving tobacco leaves, and smoke-harmonizing tobacco leaves—for formulation design. This further improves the core technical level of ultra-low tar release leaf blend formulation design, achieving ultra-low tar release while maintaining a satisfying experience. This is of great significance for the sustainable development of low-tar cigarette products that "reduce tar and harm."

[0005] To achieve this objective, the present invention adopts the following technical solution:

[0006] This invention provides a leaf blend formulation design method for ultra-low tar release products of Chinese flue-cured cigarettes, the leaf blend formulation design method comprising the following steps:

[0007] (1) Establish a sample library and classify the tobacco raw materials according to their functions in the product formula to obtain initial sample groups for the style highlighting module, the quality preservation and aroma enhancement module and the blending smoke module respectively.

[0008] (2) The tobacco samples in the initial sample groups of the style-highlighting module, the quality preservation and aroma-enhancing module and the blending smoke module are sorted according to the screening criteria to obtain the candidate sample groups of the three modules; then the sensory quality evaluation of single tobacco in each module is carried out, and unqualified tobacco samples are removed to obtain the preferred sample groups of the three modules.

[0009] The first screening criterion for the "Style Highlighting" module is the aroma intensity per unit of tar, and the second screening criterion is the nicotine content per unit of tar smoke; the first screening criterion for the "Preservation and Flavor Enhancement" module is the amount of tar aroma per unit, and the second screening criterion is the nicotine content per unit of tar smoke; the first screening criterion for the "Blending Smoke" module is the concentration of tar smoke per unit, and the second screening criterion is the nicotine content per unit of tar smoke.

[0010] (3) Introduce some or all of the preferred sample groups of the style-highlighting module, the quality preservation and aroma-enhancing module and the blending smoke module into the leaf group design of the formulation personnel to obtain the primary leaf group formula. Combine the sensory quality evaluation of product compliance and the tar content test to obtain the ultra-low tar release leaf group formula that takes into account both the aroma and the satisfaction.

[0011] In this invention, by analyzing the role of different grades of tobacco leaves from different production areas in the leaf blend formulation, the tobacco leaves are divided into three modules: tobacco leaves that highlight the style, tobacco leaves that preserve quality and enhance aroma, and tobacco leaves that blend the smoke. The raw material selection indicators for each module are determined according to its function.

[0012] The functions of the style-enhancing module, the quality-preserving and aroma-enhancing module, and the smoke-blending module in the ultra-low tar cigarette formulation are as follows:

[0013] The "Style Highlighting" module highlights the main style of the product and serves as its framework. It emphasizes a sweet, elegant, and graceful feel, and generally consists of tobacco leaves with high aroma and high taste.

[0014] The "Preservation and Aroma Enhancement" module enhances the richness of the natural aroma of tobacco, improves the fineness of the smoke, and enhances the purity of the taste. This type of tobacco generally has a sufficient amount of aroma, less off-flavors, and a pure taste.

[0015] Blending Smoke Module: Balances smoke, increases smoothness and comfort, and improves product taste characteristics. This type of tobacco leaf has a wide selection range and can be selected according to the main leaf blend formula requirements of the product.

[0016] This invention introduces unit tar aroma quality, unit tar smoke nicotine, unit tar aroma quantity, and unit tar smoke concentration as screening indicators for three modules: style-enhancing tobacco leaves, aroma-enhancing and quality-improving tobacco leaves, and blended smoke tobacco leaves. This allows for a greater breakthrough in the tar release of current low-tar and low-harm tobacco leaf blends, which is of great significance for the sustainable development of low-tar cigarettes.

[0017] In this invention, the unit tar aroma quality refers to the ratio of aroma quality to tar content, and the definitions of other screening indicators follow the same principle.

[0018] Preferably, the establishment of the sample library in step (1) includes removing samples based on any one or at least two of the following indicators: tobacco leaf sample inventory, production cycle, or aging cycle, to obtain the sample library.

[0019] Preferably, the sorting in step (2) includes sorting the tobacco leaf samples from largest to smallest using a first screening index, selecting the top-ranked tobacco leaf samples to obtain a first screening sample group, and sorting the first screening sample group from largest to smallest using a second screening index to obtain a candidate sample group.

[0020] Preferably, the number of samples in the candidate sample group in step (2) accounts for 40-55% of the initial sample group of the corresponding module, for example, it can be 40%, 42%, 45%, 50%, 55%, etc.

[0021] In this invention, screening the above-mentioned number of samples can ensure the quality of the samples while ensuring that there are still enough samples to form a combination formula after some samples are removed in the later stage.

[0022] Preferably, the evaluation indicators for the sensory quality evaluation of the single-material tobacco in step (2) include aroma type, style prominence, aroma quality, aroma quantity, off-flavors, irritation and aftertaste.

[0023] Preferably, the unqualified sample is characterized by incorrect aroma, style prominence <6, aroma quality <4.6, aroma quantity <4.6, off-odors <6, irritation <4.5, and aftertaste <4.5.

[0024] In this invention, a first sensory quality evaluation is performed after screening and sorting to eliminate unqualified samples, which helps to improve the sample quality of the later formulation.

[0025] Preferably, in step (3), the leaf blend formulation with ultra-low tar release includes 35-60% (e.g., 35%, 40%, 45%, 50%, 55%, 60%) of a style-enhancing module, 25-40% (e.g., 25%, 30%, 35%, 40%) of a quality-preserving and aroma-enhancing module, and 15-30% (e.g., 15%, 20%, 25%, 30%) of a smoke-blending module, based on mass percentage.

[0026] In this invention, since reducing tar will reduce the strength, concentration, and aroma of the smoke, the second screening index for all three modules is nicotine per unit of tar. The proportion of modules that highlight the style and modules that preserve the quality and enhance the aroma is appropriately increased. This can compensate for the loss of strength, concentration, and aroma of the smoke in the leaf blend formula to the greatest extent possible, and ensure that the physiological strength does not decrease significantly while reducing tar.

[0027] Preferably, step (3) involves combining the top 20-100% (e.g., 30%, 40%, 50%, 60%, 70%, 80%, 90%, etc.) of the preferred sample group to obtain the formulation.

[0028] Preferably, in step (3), the formulation is first evaluated for its conformity to product design style, and unqualified tobacco leaf samples are removed from the formulation. Then, the tar content is tested, and formulations with unqualified tar content are removed to obtain a leaf group formulation with ultra-low tar release.

[0029] After removing substandard tobacco samples, this invention can optionally combine sensory quality evaluation based on product design style conformity with the selection of tobacco grades with similar screening indicators in different modules to gradually replace the samples and conduct sensory quality evaluation based on product design style conformity again until a leaf blend formula that meets the product's sensory quality requirements is formulated. Further, tar release is tested in conjunction with ultra-low tar product design auxiliary materials. If the tar release is lower than a set value, a product leaf blend formula that meets the design requirements is obtained. If the tar release is higher than the set value, the three modules can be used to select the best tobacco leaf groups, choosing tobacco grades with higher preferred indicators and tar ratios and similar style characteristics for gradual, small-scale replacement. Sensory evaluation based on product design style is conducted again. If the requirements are met, tar release is further tested until the leaf blend formula conforms to the product design style and the tar release meets the design value, thus obtaining a leaf blend formula with ultra-low tar release and a satisfying sensory experience.

[0030] Preferably, the evaluation indicators for the sensory quality assessment of the product design style conformity include the degree of style prominence, strength, concentration, aroma quality, aroma quantity, penetration, harmony, sweetness, off-flavors, irritation, and aftertaste.

[0031] Preferably, the total sensory quality evaluation score of the product design style of the substandard tobacco leaf samples in the formula is <85.

[0032] Preferably, the tar release of the unqualified leaf formulation exceeds 3.05 mg / tube.

[0033] In this invention, Nanjing (3mg Yuhua stone) is selected as an auxiliary material to prepare cigarettes.

[0034] Preferably, the analytical method for obtaining the screening indicators in the style-enhancing module, the quality-preserving and aroma-enhancing module, and the smoke-blending module in step (1) includes:

[0035] (1) Establish a database of tar release, aroma characteristics, smoke characteristics, taste characteristics and style characteristics of tobacco leaf samples;

[0036] (2) Cluster analysis was performed on the tar release of tobacco leaf samples to obtain tobacco leaf groups with low tar release.

[0037] (3) Correlation analysis was conducted on the aroma characteristics, smoke characteristics, taste characteristics and style characteristics of samples in the low tar release tobacco leaf group to obtain the screening indicators in the style highlighting module, quality preservation and aroma enhancement module and blended smoke module.

[0038] Preferably, the evaluation indicators of aroma characteristics in step (1) include aroma quality, aroma quantity, permeability, and off-odors.

[0039] Preferably, the evaluation indicators of the taste characteristics in step (1) include irritation, dryness, and aftertaste.

[0040] Preferably, the evaluation indicators of the flue gas characteristics in step (1) include the number of puff ports, total particulate matter, carbon monoxide, nicotine, moisture and tar.

[0041] Preferably, the evaluation indicators of the style features in step (1) include the degree of prominence, concentration, and intensity of the style features.

[0042] Preferably, the correlation analysis in step (3) includes simple correlation analysis, multivariate correlation analysis and canonical correlation analysis.

[0043] Compared with the prior art, the present invention has at least the following beneficial effects:

[0044] This invention introduces unit tar aroma quality, unit tar aroma quantity, unit tar smoke concentration, and unit tar smoke nicotine as screening indicators for three modules: style-enhancing tobacco leaves, aroma-enhancing and quality-improving tobacco leaves, and smoke-blending tobacco leaves. This allows for a greater breakthrough in the tar release of current tar-reducing and harm-reducing leaf blend formulations, which is of great significance for the sustainable development of tar reduction and harm reduction in Chinese flue-cured cigarettes. Detailed Implementation

[0045] To facilitate understanding of the present invention, the following embodiments are provided. Those skilled in the art should understand that these embodiments are merely illustrative and should not be construed as limiting the scope of the invention.

[0046] Experimental Example 1

[0047] This experimental example provides an analytical method for obtaining screening indicators in modules that highlight style, preserve quality and enhance aroma, and blend smoke. The analytical method includes the following steps:

[0048] (1) Establish a database of the tar release, aroma characteristics, smoke characteristics, taste characteristics and style characteristics of tobacco leaves from different production areas and parts of the brand raw materials.

[0049] (2) Based on the tar content of tobacco leaves, cluster analysis was conducted to classify the raw materials in the core production area into three grades: low tar, medium tar, and high tar, according to their parts, so as to understand the regional distribution of tar content and the differences in parts.

[0050] The cluster analysis results of tar content in the upper leaves and the middle leaves are shown in Table 1 and Table 2, respectively:

[0051] Table 1

[0052]

[0053]

[0054] Table 2

[0055] Tar content Number of samples Minimum value Maximum value mean Standard deviation variance Coefficient of variation (%) Skewness coefficient Kurtosis coefficient Low 27 9.21 10.99 10.41 0.51 0.03 4.90 -0.97 -0.06 middle 18 11.21 12.41 11.82 0.36 0.13 3.05 -0.35 -1.02 high 8 12.62 13.84 13.29 0.34 0.12 2.56 -0.68 2.53

[0056] The cluster analysis of tar content in B2F grade tobacco leaves from 2020 to 2021 is shown in Table 1. The results show that among the 50 samples submitted for testing, there were 18 samples with low tar content, with a mean of 11.42 mg / cig, 31 samples with medium tar content, with a mean of 11.52 mg / cig, and 1 sample with high tar content, with a mean of 16.81 mg / cig.

[0057] The cluster analysis of tar content in C3F grade tobacco leaves from 2020 to 2021 is shown in Table 2. The results show that among the 53 samples submitted for testing, 27 samples had low tar content with a mean of 10.41 mg / cig, 18 samples had medium tar content with a mean of 11.82 mg / cig, and 8 samples had high tar content with a mean of 13.29 mg / cig.

[0058] (3) Study the covariance relationship between aroma characteristics, taste characteristics and smoke characteristics of tobacco raw materials, analyze the main factors affecting the quality characteristics of tobacco leaves, further clarify the tobacco raw material groups with excellent aroma and taste characteristics and low tar release, and provide technical support for the selection of tobacco leaf grades in the three modules.

[0059] ① A simple correlation analysis between the aroma characteristics and taste characteristics of tobacco leaves

[0060] Table 3 shows a simple correlation analysis between the aroma and taste characteristics of flue-cured tobacco. It can be seen that there is a highly significant positive correlation between the five aroma characteristics and the four taste characteristics (P < 0.01).

[0061] Table 3

[0062]

[0063]

[0064] Note: ** indicates that the correlation reached a significant level of 0.01.

[0065] ②Multivariate correlation analysis of single aroma characteristics of tobacco leaves on taste characteristics

[0066] Table 4 shows the typical correlation coefficients between aroma and taste characteristics of flue-cured tobacco, and Table 5 shows the multivariate correlation analysis between aroma and taste characteristics.

[0067] Table 4

[0068] Canonical correlation coefficient Eigenvalues Wilke Statistics F Molecular degrees of freedom Denominator degrees of freedom p-value 0.905 4.520 0.181 73.820 3.000 49.000 0.000

[0069] Table 5

[0070] Taste characteristics index variables Variable coefficients Irritating 0.718 dryness -0.186 Aftertaste 0.465

[0071] A multivariate correlation analysis was performed on the aroma and taste characteristics, and the canonical correlation coefficient was 0.905, which was highly significant (P < 0.01). As shown in the table above, the irritation was most strongly correlated with the aroma, followed by the aftertaste.

[0072] Table 6 shows the typical correlation coefficients between aroma content and taste characteristics of flue-cured tobacco, and Table 7 shows the multivariate correlation analysis between aroma content and taste characteristics.

[0073] Table 6

[0074] Canonical correlation coefficient Eigenvalues Wilke Statistics F Molecular degrees of freedom Denominator degrees of freedom Significance 0.712 1.028 .493 16.797 3.000 49.000 0.000

[0075] Table 7

[0076] Taste characteristics index variables Variable coefficients Irritating 0.999 dryness 0.058 Aftertaste -0.055

[0077] Multivariate correlation analysis was performed on the aroma content and taste characteristics of the samples. The canonical correlation coefficient was relatively large, at 0.712, which was highly significant (P < 0.01). As shown in the table above, the aroma content of tobacco leaves was most strongly correlated with irritation, followed by dryness.

[0078] Table 8 shows the typical correlation coefficients between the permeability and taste characteristics of flue-cured tobacco, and Table 9 shows the multivariate correlation analysis between permeability and taste characteristics.

[0079] Table 8

[0080] Canonical correlation coefficient Eigenvalues Wilke Statistics F Molecular degrees of freedom Denominator degrees of freedom Significance 0.586 0.523 0.657 8.543 3.000 49.000 0.000

[0081] Table 9

[0082] Taste characteristics index variables Variable coefficients Irritating 0.806 dryness 0.192 Aftertaste 0.016

[0083] Multivariate correlation analysis was performed on the translucency and taste characteristics of the samples, and the canonical correlation coefficient was 0.586, which was highly significant (P < 0.01). As shown in the table above, translucency was most strongly correlated with irritation, followed by dryness.

[0084] Table 10 shows the typical correlation coefficients between off-flavors and taste characteristics of flue-cured tobacco, and Table 11 shows the multivariate correlation analysis between off-flavors and taste characteristics.

[0085] Table 10

[0086] Canonical correlation coefficient Eigenvalues Wilke Statistics F Molecular degrees of freedom Denominator degrees of freedom Significance 0.883 3.542 0.220 57.856 3.000 49.000 0.000

[0087] Table 11

[0088] Taste characteristics index variables Variable coefficients Irritating 0.373 dryness -0.134 Aftertaste 0.762

[0089] Multivariate correlation analysis was performed on the off-flavors and taste characteristics of the samples. The canonical correlation coefficient was relatively large, at 0.883, and reached a highly significant level (P < 0.01). As shown in the table above, the aftertaste of the samples was most strongly correlated with off-flavors, with a corresponding coefficient of 0.762, followed by irritation, with a corresponding coefficient of 0.373.

[0090] Table 12 shows the canonical correlation coefficients between the total aroma characteristics score and the taste characteristics of flue-cured tobacco, and Table 13 shows the multivariate correlation analysis between the total aroma characteristics score and the taste characteristics.

[0091] Table 12

[0092] Canonical correlation coefficient Eigenvalues Wilke Statistics F Molecular degrees of freedom Denominator degrees of freedom Significance 0.846 2.527 0.284 41.272 3.000 49.000 0.000

[0093] Table 13

[0094] Taste characteristics index variables Variable coefficients Irritating 0.701 dryness -0.049 Aftertaste 0.356

[0095] A multivariate correlation analysis was performed on the total score of aroma characteristics and taste characteristics of the samples. The canonical correlation coefficient was relatively large, at 0.846, and reached a highly significant level (P < 0.01). As shown in the table above, the irritation was most strongly correlated with the total score of aroma characteristics, with a coefficient of variation of 0.701; followed by the aftertaste, with a coefficient of variation of 0.356.

[0096] ③ Multivariate correlation analysis of single taste characteristics of tobacco leaves on aroma characteristics

[0097] The multivariate correlation coefficients between irritation and aroma characteristics are shown in Table 14, and the multivariate correlation analysis between irritation and aroma characteristics is shown in Table 15.

[0098] Table 14

[0099] Correlation Eigenvalues Wilke Statistics F Molecular degrees of freedom Denominator degrees of freedom p-value 0.901 4.319 0.188 51.827 4.000 48.000 0.000

[0100] Table 15

[0101] Aroma characteristic index variables Variable coefficients Fragrance 0.832 Aroma -0.150 translucency 0.072 Mixed gases 0.255

[0102] The multivariate correlation coefficient results show that the multivariate correlation coefficient between irritation and aroma characteristics is relatively large, at 0.901, and reaches a highly significant level (P < 0.01). As shown in the table above, irritation has the strongest correlation with aroma characteristics, followed by off-odors.

[0103] The multivariate correlation coefficients between dryness and aroma characteristics are shown in Table 16, and the multivariate correlation analysis between dryness and aroma characteristics is shown in Table 17.

[0104] Table 16

[0105] Correlation Eigenvalues Wilke Statistics F Molecular degrees of freedom Denominator degrees of freedom p-value 0.822 2.085 0.324 25.016 4.000 48.000 0.000

[0106] Table 17

[0107] Aroma characteristic index variables Variable coefficients Fragrance 0.682 Aroma -0.166 translucency 0.106 Mixed gases 0.400

[0108] The multivariate correlation coefficient results show that dryness is highly correlated with aroma characteristics (0.822), which is extremely significant (P < 0.01). The table above shows that dryness is most strongly correlated with aroma characteristics, followed by off-odors.

[0109] The multivariate correlation coefficients between aftertaste and aroma characteristics are shown in Table 18, and the multivariate correlation analysis between aftertaste and aroma characteristics is shown in Table 19.

[0110] Table 18

[0111] Correlation Eigenvalues Wilke Statistics F Molecular degrees of freedom Denominator degrees of freedom p-value 0.907 4.655 0.177 55.865 4.000 48.000 0.000

[0112] Table 19

[0113] Aroma characteristic index variables Variable coefficients Fragrance 0.646 Aroma -0.305 translucency 0.067 Mixed gases 0.572

[0114] The multivariate correlation coefficient results show that the aftertaste and aroma characteristics are highly correlated (0.907), reaching a highly significant level (P < 0.01). As shown in the table above, the aftertaste has the strongest correlation with aroma characteristics, followed by off-flavors.

[0115] The multivariate correlation coefficients between the total taste characteristics and aroma characteristics are shown in Table 20, and the multivariate correlation analysis between the total taste characteristics and aroma characteristics is shown in Table 21.

[0116] Table 20

[0117] Correlation Eigenvalues Wilke Statistics F Molecular degrees of freedom Denominator degrees of freedom p-value 0.900 4.285 0.189 51.418 4.000 48.000 0.000

[0118] Table 21

[0119] Aroma characteristic index variables Variable coefficients Fragrance 0.723 Aroma -0.211 translucency 0.080 Mixed gases 0.413

[0120] A multivariate correlation analysis was performed on the total taste characteristics and aroma characteristics, and the multivariate correlation coefficients are shown in Table 3.7. The results show that the total taste characteristics and aroma characteristics are highly correlated (0.900), reaching a highly significant level (P < 0.01). As shown in the table, the total taste characteristics are most strongly correlated with aroma characteristics, followed by off-odors.

[0121] ④ Typical correlation analysis between aroma characteristics and taste characteristics

[0122] Table 22 shows the typical correlations between aroma characteristics and mouthfeel characteristics; Table 23 shows the standardized typical correlation coefficients of aroma characteristics; and Table 24 shows the standardized typical correlation coefficients of mouthfeel characteristics.

[0123] Table 22

[0124] Serial Number Correlation Eigenvalues Wilke Statistics F Molecular degrees of freedom Denominator degrees of freedom Significance 1 0.916 5.238 0.139 11.287 12 121.996 0.000 2 0.353 0.142 0.865 1.18 6 94 0.324 3 0.111 0.012 0.988

[0125] Table 23

[0126] variable 1 2 3 Fragrance -0.736 1.479 1.925 Aroma 0.250 1.308 -0.509 translucency -0.063 0.068 -0.696 Mixed gases -0.442 -2.589 -1.117

[0127] Table 24

[0128] variable 1 2 3 Irritating -0.518 3.038 1.734 dryness 0.196 0.338 -2.621 Aftertaste -0.674 -3.317 0.652

[0129] From the table above, we can see that the correlation coefficient of the first pair of canonical variables is 0.916 (P < 0.01); the correlation coefficient of the second pair of canonical variables is 0.353 (P > 0.01); and the correlation coefficient of the third pair of canonical variables is 0.111 (P > 0.01). The first pair of canonical correlation coefficients is highly significant. The standardized canonical correlation coefficients of X and Y for the first pair of canonical variables are shown in the table.

[0130] Based on the table above, the standardized expressions for the first pair of canonical correlation variables U1 and V1 can be derived:

[0131] U1 = -0.736 * Aroma Quality + 0.250 * Aroma Amount - 0.063 * Transparency - 0.442 * Off-odors;

[0132] V1 = -0.518 * Irritation + 0.196 * Dryness - 0.674 * Aftertaste;

[0133] As can be seen from the standardized expressions of the first pair of canonical related variables, since the absolute values ​​of aroma quality, aroma quantity, and off-odor coefficients are relatively large, the canonical variables reflecting aroma characteristics are mainly determined by aroma quality, aroma quantity, and off-odor; the absolute values ​​of irritation and aftertaste are relatively large, so the canonical variables reflecting taste characteristics are mainly determined by irritation and aftertaste.

[0134] ⑤ Typical correlation analysis between aroma characteristics and style features

[0135] Table 25 shows the typical correlation between aroma characteristics and style features, Table 26 shows the standardized typical correlation coefficients of aroma characteristics, and Table 27 shows the standardized typical correlation coefficients of style features.

[0136] Table 25

[0137] Serial Number Correlation Eigenvalues Wilke Statistics F Molecular degrees of freedom Denominator degrees of freedom Significance 1 0.904 4.496 0.104 13.773 12 121.996 0.000 2 0.649 0.728 0.57 5.082 6 94 0.000 3 0.122 0.015 0.985

[0138] Table 26

[0139] variable 1 2 3 Fragrance -0.061 1.088 -1.09 Aroma -0.098 -1.132 -1.737 translucency -0.418 -0.529 1.139 Mixed gases -0.518 0.573 1.634

[0140] Table 27

[0141]

[0142]

[0143] From the table above, we can see that the correlation coefficient for the first pair of canonical variables is 0.904 (P < 0.01); the correlation coefficient for the second pair of canonical variables is 0.649 (P < 0.01); and the correlation coefficient for the third pair of canonical variables is 0.122 (P > 0.01). The standardized canonical correlation coefficients of x and y are shown in the table.

[0144] Based on the table above, the standardized expressions for the first pair of canonical correlation variables U1 and V1 can be derived:

[0145] U1 = -0.061 * Aroma Quality - 0.098 * Aroma Amount - 0.418 * Transparency - 0.518 * Off-odors;

[0146] V1 = -0.876 * Degree of Style Characteristics Emphasis - 0.287 * Concentration + 0.171 * Vigor;

[0147] The standardized expressions for the second pair of canonical correlation variables U2 and V2 are as follows:

[0148] U2 = +1.088 * Aroma Quality - 1.132 * Aroma Amount - 0.529 * Transparency + 0.573 * Off-odors;

[0149] V2 = +0.611 * Degree of Style Feature Emphasis - 0.429 * Concentration - 0.803 * Vigor;

[0150] The standardized expressions for the first pair of canonical correlation variables show that, due to the relatively large absolute values ​​of the transparency and off-odor coefficients, the canonical variables reflecting aroma characteristics are mainly determined by transparency and off-odor; similarly, the relatively large absolute value of the degree of style feature prominence indicates that the canonical variables reflecting style features are mainly determined by the degree of style feature prominence. The standardized expressions for the second pair of canonical correlation variables show that the canonical variables reflecting aroma characteristics are determined by aroma intensity and aroma quantity, while the canonical variables reflecting style features are mainly determined by strength.

[0151] ⑥ Typical correlation analysis between aroma characteristics and smoke characteristics

[0152] Table 28 shows the typical correlation between aroma characteristics and smoke characteristics; Table 29 shows the standardized typical correlation coefficients of aroma characteristics; and Table 30 shows the standardized typical correlation coefficients of smoke characteristics.

[0153] Table 28

[0154] Serial Number Correlation Eigenvalues Wilke Statistics F Molecular degrees of freedom Denominator degrees of freedom Significance 1 0.578 0.503 0.53 1.26 24 151.219 0.202 2 0.370 0.159 0.796 0.701 15 121.866 0.779 3 0.219 0.05 0.922 0.465 8 90 0.878 4 0.177 0.032 0.969 0.497 3 46 0.686

[0155] Table 29

[0156] variable 1 2 3 4 Fragrance -0.255 -2.217 -0.839 -1.509 Aroma -0.521 -0.243 -1.015 1.922 translucency 0.204 -0.638 1.504 -0.405 Mixed gases -0.432 2.862 0.754 -0.071

[0157] Table 30

[0158] variable 1 2 3 4 Number of suction ports (ports / piece) -0.526 0.265 -0.11 -1.398 Total particulate matter (mg / cig) 90.443 56.278 5.694 -8.205 Carbon monoxide (mg / cig) -0.406 -0.634 0.949 -0.434 Nicotine (mg / cig) -32.258 -20.298 -1.462 4.006 Moisture (mg / cig) -22.415 -14.884 -1.14 1.44 Tar (mg / cig) -67.37 -41.593 -4.779 6.387

[0159] From the table above, we can see that the correlation coefficient for the first pair of canonical variables is 0.578 (P > 0.01); the correlation coefficient for the second pair is 0.370 (P > 0.01); the correlation coefficient for the third pair is 0.219 (P > 0.01); and the correlation coefficient for the fourth pair is 0.177 (P > 0.01). None of the canonical correlation coefficients are significant.

[0160] ⑦ Typical correlation analysis between taste characteristics and style features

[0161] Table 31 shows the typical correlation between taste characteristics and style characteristics; Table 32 shows the standardized typical correlation coefficients of taste characteristics; and Table 33 shows the standardized typical correlation coefficients of style characteristics.

[0162] Table 31

[0163] Serial Number Correlation Eigenvalues Wilke Statistics F Molecular degrees of freedom Denominator degrees of freedom Significance 1 0.909 4.786 0.146 15.314 9 114.536 0.000 2 0.386 0.176 0.846 2.092 4 96 0.088 3 0.073 0.005 0.995 0.264 1 49 0.609

[0164] Table 32

[0165] variable 1 2 3 Irritating -0.08 -1.872 -2.999 dryness -0.295 2.601 -0.409 Aftertaste -0.646 -0.69 3.315

[0166] Table 33

[0167] variable 1 2 3 Degree of emphasis on style characteristics -0.934 -0.843 -0.188 concentration -0.315 1.668 0.104 energy 0.712 -0.795 -0.991

[0168] From the table above, we can see that the correlation coefficient for the first pair of canonical variables is 0.909 (P < 0.01); the correlation coefficient for the second pair of canonical variables is 0.386 (P > 0.01); and the correlation coefficient for the third pair of canonical variables is 0.073 (P > 0.01). The first pair of canonical correlation coefficients is significant. The standardized canonical correlation coefficients for x and y for the first pair of canonical variables are shown in the table above.

[0169] Based on the table above, the standardized expressions for the first pair of canonical correlation variables U1 and V1 can be derived:

[0170] U1 = -0.08 * Irritation -0.295 * Dryness -0.646 * Aftertaste;

[0171] V1 = -0.934 * Degree of Style Characteristics Emphasis - 0.315 * Concentration + 0.712 * Vigor;

[0172] As can be seen from the standardized expressions of the first pair of canonical related variables, since the absolute value of the aftertaste coefficient is relatively large, the canonical variables reflecting taste characteristics are mainly determined by the aftertaste; the absolute values ​​of the degree of prominence of style characteristics and the intensity are relatively large, so the canonical variables reflecting style characteristics are mainly determined by the degree of prominence of style characteristics and the intensity.

[0173] ⑧ Typical correlation analysis between taste characteristics and smoke characteristics

[0174] Table 34 shows the typical correlation between taste characteristics and smoke characteristics; Table 35 shows the standardized typical correlation coefficients of taste characteristics; and Table 36 shows the standardized typical correlation coefficients of smoke characteristics.

[0175] Table 34

[0176] Serial Number Correlation Eigenvalues Wilke Statistics F Molecular degrees of freedom Denominator degrees of freedom Significance 1 0.657 0.758 0.463 2.173 18 124.936 0.007 2 0.407 0.198 0.814 0.977 10 90 0.469 3 0.158 0.026 0.975 0.295 4 46 0.88

[0177] Table 35

[0178] variable 1 2 3 Irritating -0.589 -0.859 -3.379 dryness -1.295 2.19 0.74 Aftertaste 0.959 -1.738 2.819

[0179] Table 36

[0180] variable 1 2 3 Number of suction ports (ports / piece) -0.067 -1.213 -0.349 Total particulate matter (mg / cig) 83.528 0.792 70.019 Carbon monoxide (mg / cig) -0.731 0.742 -0.283 Nicotine (mg / cig) -29.604 -0.145 -25.179 Moisture (mg / cig) -20.387 -0.255 -18.531 Tar (mg / cig) -62.647 0.053 -52.013

[0181] From the table above, we can see that the correlation coefficient for the first pair of canonical variables is 0.657 (P < 0.01); the correlation coefficient for the second pair of canonical variables is 0.407 (P > 0.01); and the correlation coefficient for the third pair of canonical variables is 0.158 (P > 0.01). The first pair of canonical correlation coefficients is significant. The standardized canonical correlation coefficients for x and y for the first pair of canonical variables are shown in the table.

[0182] Based on the table above, the standardized expressions for the first pair of canonical correlation variables U1 and V1 can be derived:

[0183] U1 = -0.589 * Irritation - 1.295 * Dryness + 0.959 * Aftertaste;

[0184] V1 = -0.067 * number of puffs (puffs / cigarette) + 83.528 * total particulate matter (mg / cig) - 0.731 * carbon monoxide (mg / cig) - 29.604 * nicotine (mg / cig) - 20.387 * moisture (mg / cig) - 62.647 * tar (mg / cig);

[0185] As can be seen from the standardized expressions of the first pair of canonical related variables, since the absolute values ​​of the dryness and aftertaste coefficients are relatively large, the canonical variables reflecting taste characteristics are mainly determined by dryness and aftertaste; the absolute values ​​of total particulate matter and tar are relatively large, so the canonical variables reflecting smoke characteristics are mainly determined by total particulate matter and tar.

[0186] Based on the above results, the first screening criterion for the "Style Highlighting" module is the aroma intensity per unit of tar, and the second screening criterion is the nicotine content per unit of tar smoke; the first screening criterion for the "Preservation and Flavor Enhancement" module is the amount of aroma per unit of tar, and the second screening criterion is the nicotine content per unit of tar smoke; and the first screening criterion for the "Blended Smoke" module is the concentration of tar smoke per unit of tar, and the second screening criterion is the nicotine content per unit of tar smoke.

[0187] Example 1

[0188] This embodiment provides a leaf blend formulation design method for an ultra-low tar release product of Chinese flue-cured tobacco, including the following steps:

[0189] (1) Based on the functional positioning of tobacco leaf samples in the product design leaf group formula, tobacco leaves are divided into three modules: highlighting style, preserving quality and enhancing aroma, and blending smoke. Samples are removed based on inventory, production cycle, and aging cycle to obtain three initial sample groups (each module has a sample size of 40).

[0190] (2) The initial sample group of the style-highlighting module was sorted by the unit tar aroma quality as the evaluation index. The tobacco samples were sorted from largest to smallest. The top 50% of the tobacco samples were sorted by the unit tar smoke nicotine as the index, and the candidate sample group of the style-highlighting module was obtained (sample size 20). The unqualified tobacco samples were eliminated through sensory quality evaluation of single tobacco, and the preferred sample group of the style-highlighting module was obtained (sample size 15).

[0191] (3) The initial sample group of the quality preservation and aroma enhancement module was sorted by the unit tar aroma content as the evaluation index. The tobacco samples were sorted from large to small. The top 50% of the tobacco samples were sorted by the unit tar nicotine content as the index, and the candidate sample group of the quality preservation and aroma enhancement module was obtained (sample size 20). The unqualified tobacco samples were eliminated through sensory quality evaluation of single tobacco, and the preferred sample group of the quality preservation and aroma enhancement module was obtained (sample size 16).

[0192] (4) The initial sample group of the blended smoke module is sorted by unit tar smoke concentration as the index, and the tobacco samples are sorted from largest to smallest. The top 50% of the tobacco samples are sorted by unit tar smoke nicotine as the index, and the candidate sample group of the blended smoke module is obtained (sample size is 20). Through sensory quality evaluation of single tobacco, unqualified tobacco samples are eliminated to obtain the preferred sample group of the blended smoke module (sample size is 16).

[0193] (5) Select the top 8 samples from the preferred sample groups of the style highlighting module, the quality preservation and aroma enhancement module and the smoke blending module respectively. Mix the 8 samples in each module in equal proportions. Formulate the mixed sample of the three modules according to the mass fraction of 35% for style highlighting module, 35% for quality preservation and aroma enhancement module and 30% for smoke blending module to obtain formula 1.

[0194] Adjust the first 8 samples in the above operation to the first 9 samples, and repeat the above operation to obtain formula 2;

[0195] Adjust the first 8 samples in the above operation to the first 10 samples, and repeat the above operation to obtain formula 3;

[0196] Adjust the first 8 samples in the above operation to the first 11 samples, and repeat the above operation to obtain formula 4;

[0197] Adjust the first 8 samples in the above operation to the first 12 samples, and repeat the above operation to obtain formula 5;

[0198] (6) Conduct sensory quality evaluation on the product design style conformity of formulas 1-5. When fine-tuning, remove unqualified tobacco leaf grades in the formula and replace them with the increment of other grades used in the same module of the formula. Keep the quality scores of the three modules unchanged to form new formulas 1-5. Conduct tar content testing and remove leaf group formulas with tar release exceeding 3mg / cigarette to obtain leaf group formulas with ultra-low tar release.

[0199] The sensory quality evaluation method for single-material tobacco in the above steps is shown in the table below. Unqualified samples are those with incorrect aroma type, style prominence <6, aroma quality <4.6, aroma quantity <4.6, off-flavors <6, irritation <4.5, and aftertaste <4.5.

[0200]

[0201] The method for evaluating the sensory quality of product design is shown in the table below. The full score is 99, and the total sensory evaluation score of the product design style of the unqualified sample is <85.

[0202]

[0203]

[0204] Example 2

[0205] This embodiment provides a leaf blend formulation design method for an ultra-low tar release product of Chinese flue-cured tobacco, which differs from Embodiment 1 only in that:

[0206] In steps (2)-(4), the selection amount of tobacco leaf samples is replaced by 40% instead of 50%. After the first sensory quality evaluation, the sample quantity of the preferred sample group for the style-highlighting module is 13, the sample quantity of the preferred sample group for the quality preservation and aroma enhancement module is 15, and the sample quantity of the preferred sample group for the blended smoke module is 14.

[0207] In step (5), the formula is prepared by mass fraction according to 60% of the style-enhancing module, 25% of the quality-preserving and aroma-enhancing module, and 15% of the smoke-blending module;

[0208] Other examples are shown in Example 1.

[0209] Example 3

[0210] This embodiment provides a leaf blend formulation design method for an ultra-low tar release product of Chinese flue-cured tobacco, which differs from Embodiment 1 only in that:

[0211] In steps (2)-(4), the selection amount of tobacco leaf samples is replaced by 55% instead of 50%. After the first sensory quality evaluation, the sample quantity of the preferred sample group for the style-highlighting module is 18, the sample quantity of the preferred sample group for the quality preservation and aroma enhancement module is 18, and the sample quantity of the preferred sample group for the blended smoke module is 16.

[0212] In step (5), the formula is prepared by mass fraction according to the following proportions: 40% for highlighting style, 40% for preserving quality and enhancing aroma, and 20% for blending smoke.

[0213] Other examples are shown in Example 1.

[0214] Example 4

[0215] This embodiment provides a leaf blend formulation design method for an ultra-low tar release product of Chinese flue-cured tobacco, which differs from Embodiment 1 only in that:

[0216] In step (5), the formula is prepared by mass fraction according to the following proportions: 10% for highlighting style, 55% for preserving quality and enhancing aroma, and 35% for blending smoke.

[0217] Other examples are shown in Example 1.

[0218] Example 5

[0219] This embodiment provides a leaf blend formulation design method for an ultra-low tar release product of Chinese flue-cured tobacco, which differs from Embodiment 1 only in that:

[0220] In step (5), the formula is prepared by mass fraction according to the following proportions: 25% for highlighting style, 50% for preserving quality and enhancing aroma, and 25% for blending smoke.

[0221] Other examples are shown in Example 1.

[0222] Comparative Example 1

[0223] This comparative example provides a leaf blend formulation design method for an ultra-low tar release product of Chinese flue-cured tobacco, which differs from Example 1 only in that:

[0224] Step (2) Replace the unit tar aroma quality screening index with aroma quality, and replace the unit tar smoke nicotine screening index with smoke nicotine; through the sensory quality evaluation of single tobacco, remove unqualified tobacco leaf samples to obtain the preferred sample group (sample size is 16) of the style-highlighting module.

[0225] Step (3) Replace the unit tar aroma quantity screening index with aroma quantity, and replace the unit tar smoke nicotine screening index with smoke nicotine; through the sensory quality evaluation of single tobacco, remove unqualified tobacco leaf samples to obtain the preferred sample group (sample size is 17) of the quality preservation and aroma enhancement module.

[0226] Step (4) Replace the unit tar smoke concentration screening index with smoke nicotine, and replace the unit tar smoke nicotine with smoke nicotine; through the sensory quality evaluation of single tobacco, remove unqualified tobacco leaf samples to obtain the preferred sample group (sample size is 15) of the blended smoke module.

[0227] Other examples are shown in Example 1.

[0228] Comparative Example 2

[0229] This comparative example provides a leaf blend formulation design method for an ultra-low tar release product of Chinese flue-cured tobacco, which differs from Example 1 only in that:

[0230] Step (2) Replace the unit tar aroma quality screening index with tar, and through the sensory quality evaluation of single tobacco, remove unqualified tobacco leaf samples to obtain the preferred sample group (sample size is 14) of the style-highlighting module.

[0231] Step (3) Replace the unit tar aroma content screening index with tar through single-material tobacco sensory quality evaluation, remove unqualified tobacco leaf samples, and obtain the preferred sample group (sample size is 17) of the quality preservation and aroma enhancement module.

[0232] Step (4) Replace the unit tar smoke concentration screening index with tar, and through the sensory quality evaluation of single tobacco, remove unqualified tobacco leaf samples to obtain the preferred sample group (sample size is 16) of the blended smoke module.

[0233] Other examples are shown in Example 1.

[0234] Test Example 1

[0235] The tar content test results and compliance rates of formulations 1-5 obtained from Examples 1-5 and Comparative Examples 1-2 were summarized, and sensory quality evaluation was conducted. The evaluation method was sensory quality evaluation of product design style conformity.

[0236] The results are summarized in Table 1.

[0237] Table 1

[0238]

[0239]

[0240]

[0241] * indicates a qualified formulation after comprehensive evaluation of tar release and sensory quality.

[0242] The test results show that:

[0243] (1) As can be seen from Examples 1 to 5, the present invention divides tobacco leaves into modules that highlight style, modules that preserve quality and enhance aroma, and modules that blend smoke, and determines the screening indicators for each of the three modules. During the screening and formulation process, unqualified samples are removed through sensory evaluation. The resulting formulation has a low overall tar content, a high compliance rate, and high sensory quality. It obtains a high-quality leaf blend formulation on the basis of meeting the requirements of ultra-low tar release (below 3.05 mg / cigarette).

[0244] (2) By comparing Examples 1 and Examples 4-5, it can be seen that the present invention can achieve better technical effects of reducing tar and improving quality by further limiting the mass fraction of the style-enhancing module, the quality-preserving and aroma-enhancing module and the smoke-harmonizing module.

[0245] (3) A comparison between Example 1 and Comparative Example 1 shows that the present invention can improve the screening efficiency of low-tar leaf blend formulations by introducing screening indicators of three modules. However, after replacing the screening indicators, it cannot achieve the technical effect of reducing tar to the design value. A comparison between Example 1 and Comparative Example 2 shows that after replacing the screening indicators, the tar reduction effect is improved, but the total sensory quality score of the leaf blend formulation is far below the average, bland and tasteless, and the quality declines, which does not meet the product design requirements.

[0246] The applicant declares that the above description is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Those skilled in the art should understand that any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention fall within the protection and disclosure scope of the present invention.

Claims

1. A method for designing a leaf blend formulation for a Chinese-style flue-cured cigarette with ultra-low tar release, characterized in that, The leaf group formulation design method includes the following steps: (1) Establish a sample library and classify the tobacco raw materials according to their functions in the product formula to obtain initial sample groups for the style highlighting module, the quality preservation and aroma enhancement module and the blending smoke module respectively; (2) The tobacco samples in the initial sample groups of the style-highlighting module, the quality preservation and aroma-enhancing module and the blending smoke module are sorted according to the screening criteria to obtain the candidate sample groups of the three modules; then the sensory quality evaluation of single tobacco in each module is carried out, and unqualified tobacco samples are removed to obtain the preferred sample groups of the three modules. The first screening criterion for the "Style Highlighting" module is the aroma intensity per unit of tar, and the second screening criterion is the nicotine content per unit of tar smoke; the first screening criterion for the "Preservation and Flavor Enhancement" module is the amount of tar aroma per unit, and the second screening criterion is the nicotine content per unit of tar smoke; the first screening criterion for the "Blending Smoke" module is the concentration of tar smoke per unit, and the second screening criterion is the nicotine content per unit of tar smoke. The sorting process includes sorting the tobacco leaf samples from largest to smallest using a first screening index, selecting the top-ranked tobacco leaf samples to obtain a first screening sample group, and sorting the first screening sample group from largest to smallest using a second screening index to obtain a candidate sample group. (3) Some or all of the samples from the preferred sample groups of the style-enhancing module, the quality-preserving and aroma-enhancing module, and the blending smoke module are incorporated into the leaf group design by the formulation personnel to obtain the primary leaf group formulation. Combined with the product compliance sensory quality evaluation and tar content detection, an ultra-low tar release leaf group formulation that takes into account both tobacco aroma and satisfaction is obtained. The formulation is first subjected to product compliance sensory quality evaluation, and unqualified tobacco leaf samples are removed from the formulation. Then, the tar content is detected, and formulations with unqualified tar content are removed to obtain the ultra-low tar release leaf group formulation. The tar release of the unqualified leaf group formulation exceeds 3.05 mg / cigarette. By mass percentage, the ultra-low tar release leaf group formulation includes 35-60% style-enhancing module, 25-40% quality-preserving and aroma-enhancing module, and 15-30% blending smoke module.

2. The leaflet formulation design method according to claim 1, characterized in that, Step (1) involves establishing a sample library by eliminating samples based on any one or at least two of the following indicators: tobacco leaf sample inventory, production cycle, or aging cycle.

3. The leaflet formulation design method according to claim 1, characterized in that, The number of samples in the candidate sample group in step (2) accounts for 40-55% of the initial sample group of the corresponding module.

4. The leaflet formulation design method according to claim 1, characterized in that, The evaluation indicators for the sensory quality evaluation of single-material tobacco in step (2) include aroma type, style prominence, aroma quality, aroma quantity, off-flavors, irritation and aftertaste.

5. The leaflet formulation design method according to claim 1, characterized in that, The unqualified tobacco samples mentioned in step (2) are those with incorrect aroma, style prominence <6, aroma quality <4.6, aroma quantity <4.6, off-flavors <6, irritation <4.5, and aftertaste <4.

5.

6. The leaflet formulation design method according to claim 1, characterized in that, Step (3) Combine the top 20-100% of the tobacco leaf samples in the preferred sample group to obtain the formula.

7. The leaflet formulation design method according to claim 1, characterized in that, The evaluation indicators for the sensory quality assessment of the product's compliance include the degree of style prominence, strength, concentration, aroma quality, aroma quantity, penetration, harmony, sweetness, off-flavors, irritation, and aftertaste.

8. The leaflet formulation design method according to claim 1, characterized in that, The total score of the sensory quality evaluation of the non-compliant tobacco leaf samples in the formula described in step (3) is <85.

9. The leaflet formulation design method according to claim 1, characterized in that, The analytical methods for obtaining the screening indicators in the style-highlighting module, the quality preservation and aroma-enhancing module, and the smoke blending module mentioned in step (1) include: (1) Establish a database of tar release, aroma characteristics, smoke characteristics, taste characteristics, and style characteristics of tobacco leaf samples; (2) Cluster analysis was performed on the tar release of tobacco leaf samples to obtain tobacco leaf groups with low tar release; (3) Correlation analysis was conducted on the aroma characteristics, smoke characteristics, taste characteristics and style characteristics of samples in the low tar release tobacco leaf group to obtain the screening indicators in the style highlighting module, quality preservation and aroma enhancement module and blended smoke module.

10. The leaflet formulation design method according to claim 9, characterized in that, The evaluation indicators of aroma characteristics in step (1) include aroma quality, aroma quantity, permeability, and off-odors.

11. The leaflet formulation design method according to claim 9, characterized in that, The evaluation indicators of the taste characteristics mentioned in step (1) include irritation, dryness and aftertaste.

12. The leaflet formulation design method according to claim 9, characterized in that, The evaluation indicators of the flue gas characteristics in step (1) include the number of puff ports, total particulate matter, carbon monoxide, nicotine, moisture and tar.

13. The leaflet formulation design method according to claim 9, characterized in that, The evaluation indicators of style features in step (1) include the degree of prominence, concentration, and intensity of style features.

14. The leaflet formulation design method according to claim 9, characterized in that, The correlation analysis described in step (3) includes simple correlation analysis, multivariate correlation analysis and canonical correlation analysis.

Images