Entropy analysis and grey cluster analysis of multiple indexes of 5 kinds of genuine medicinal materials

5 kinds of genuine medicinal materials, including Diding (Latin name: Corydalis bungeana Turcz), Purslane (Latin name: Portulaca oleracea L.), straw sandal board (Latin name: Hoya carnosa (L.f.) R. Br), June snow (Latin name: Serissa japonica (Thunb.) Thunb.), pine vine rattan (Latin name: Lycopodiastrum casuarinoides (Spring) Holub. [Lycopodium casuarinoides Spring]), were selected as the research objects. The combustion heat, thermo gravimetric parameters, and fat content, calcium content, trace element content, ash content of 5 kinds of genuine medicinal materials were measured. The combustion heat, differential thermal gravimetric analysis, fat content, calcium content, trace elements content, and ash content of 5 kinds of genuine medicinal materials were used to build a systematic multi-index evaluation system by gray pattern recognition and grey correlation coefficient cluster analysis, which can make up for the gaps in this area and provide scientific basis and research significance for the study of genuine medicinal materials quality. The results showed that the order of combustion heat of 5 kinds of genuine medicinal materials, including Diding, Purslane, straw sandal board, June snow, pine vine rattan, was Diding > June snow > straw sandal board > Purslane > pine vine rattan, the order of fat content (%) of 5 kinds of genuine medicinal materials was straw sandal board > Diding > pine vine rattan > June snow > Purslane, the order of calcium content (%) was pine vine rattan > June snow > Purslane > straw sandal board > Diding, the order of ash content was June snow > Purslane > straw sandal board > pine vine rattan > Diding. From the analysis of thermogravimetric analysis results and thermogravimetric combustion stability, the order of combustion stability of 5 kinds of genuine medicinal materials was June snow > pine Vine rattan > straw sandal board > Diding > Portulaca oleracea. The order of the content of 12 trace elements in 5 kinds of genuine medicinal materials, in terms of trace element content, June snow contains the highest trace elements in all samples. According to combustion heat, combustibility (combustion stability of genuine medicinal materials), fat, calcium, ash, trace element content, the comprehensive evaluation results of multi-index analysis constructed by gray correlation degree, gray correlation coefficient factor analysis, and gray hierarchical cluster analysis showed that the comprehensive evaluation multi-index order of 5 genuine medicinal materials, including Diding, Purslane, straw sandal board, June snow and pine vine rattan, was June snow > straw sandal board > Diding > Purslane > pine vine rattan. Therefore, the comprehensive evaluation results of the quality of genuine medicinal materials selected in this study were June snow the best, followed by straw sandal board. This research has important theoretical and practical significance for the multi-index measurement and comprehensive evaluation of genuine medicinal materials, and can provide scientific basis and research significance for the research of multi-index quality control of genuine medicinal material.


Methods
[Lycopodium casuarinoides Spring]), were purchased from Laibin traditional Chinese medicine market in June 2020, in Guangxi, China, and were selected as the research objects and the analysis samples, all samples were sieved through 40 mesh pharmacopoeia. Repeat the test multiple times for each sample to reduce errors. The Diding plant was identified as dry whole grass of Papaveraceae (Fig. S1), and Portulaca oleracea plant was identified as dry whole herb of Portulaca oleracea L. (Fig. S2), and straw sandal board was the dry leaf of dicotyledonous plant medicine bulbus of Asclepiadaceae (Fig. S3), and June snow was a dry whole herb of Serissa serissoides (DC.) Druce of Rubiaceae (Fig. S4), and pine vine rattan was the dry stem of a plant of the Convolvulaceae fangchiaceae (Fig. S5). The above materials GKS20190600010, GKS20190600011, GKS20190600012, GKS20190600013 and GKS20190600014 were respectively identified from the source, eye observation, hand touch, nose smell, mouth taste, comparison, etc. by Prof. Caiyun  www.nature.com/scientificreports/ Determination method of genuine medicinal material combustion heat. In this experiment, bomb calorimeter was used to determine the heat values under constant volume. The data processing formula is: In the formula: m, Qv, ∆T, W cal , Q ignition wire , ∆m respectively are the quality of the sample to be tested, the constant volume combustion heat, the temperature change before and after combustion, the calorimeter water equivalent, and the combustion heat of the ignition wire (Q ignition wire = 1400.8 J/g), the actual mass of the ignition wire participating in the combustion reaction 27 . Instruments and reagents. BH series combustion heat measurement experimental device, oxygen cylinder, oxygen meter, grinder, sheet press, ignition wire (nickel-chromium wire, Changsha Changxing Higher Education Instrument Equipment Co., Ltd.); electronic balance(model FA2004, Shanghai Shunyu Hengping Scientific Instrument Co., Ltd.), benzoic acid (AR, Tianjin KERMEL Chemical Reagent Co., Ltd.), medicinal capsules.
Thermogravimetric analysis method [28][29][30] . Thermogravimetric analysis is a technology to measure the relationship between the mass change of samples and temperature or time under a programmed temperature and a certain atmosphere. The temperature values on the curves are often used to compare the thermal stability of samples, which is the data basis for evaluating the combustibility of combustibles.
An appropriate amount of 2-10 mg samples were placed into an alumina crucible with a heating rate of 10 °C min −1 , a reference compound of ɑ-Al 2 O 3 , N 2 atmosphere (flow rate of 100 mL min −1 ), and a temperature range of 30-600 °C. TG, DTG and DTA analysis were performed at the same time. In order to reduce the experimental error, the experiment was repeated for each sample 31 . Instruments and reagents. Thermogravimetric analyzer (Germany NETZSCH STA 2500), crucible.
Fat determination method. Soxhlet extraction, the fat content was determined by gravimetric method, that is, after the solvent of the extraction sample was extracted, the fat of the measured substance was extracted from the sample, dried and weighed, and calculated [32][33][34][35] . W fat = m 1 /m 2 × 100%, m 1 is the fat mass (g), m 2 is the sample mass (g).
Determination method of calcium content. Calcium and aminocarboxylate can form metal complexes quantitatively, and its stability is stronger than that of calcium and indicator [36][37][38][39][40][41][42][43] . In the appropriate pH range, EDTA was titrated with ammonia-carboxylate complexing agent. When the measurement was reached, EDTA captured calcium ions in the indicator complex, making the solution present the color of free indicator (end point). According to the amount of EDTA complexing agent, calcium content can be calculated.
Ash determination method. The residual inorganic matter in genuine medicinal materials after burning is called ash 44-48 . Instruments and reagents. Muffle furnace, electric furnace, crucible, electronic balance.
Determination method of trace elements in 5 kinds of genuine medicinal materials. Determination method of trace elements content. In this experiment, 5 kinds of genuine medicinal materials, including Diding, Purslane, straw sandal board, June snow, pine vine rattan, in Guangxi, China, are all medicinal materials produced [49][50][51][52][53][54][55] . After sampling the uniform sample has been crushed with a crushing grinder, accurately weigh the sample 0.3 g (accurate to 0.0001 g) to 50 mL digestion container. Add aqua regia 8 mL and hydrogen peroxide 3 mL to the container. After standing overnight, put it into microwave digestion until the sample is completely digested, and the acid is removed. After cooling, it is filtered and the volume is fixed to a 50 mL volumetric flask, and the instrument is tested. The measurement of 12 kinds of trace elements is in accordance with GB/T 30903-2014. At the same time, the reagent blank experiment was performed, and the number of collection points and the number of repetitions were both 6 times.

Calculation of combustion heat of five kinds of genuine medicinal materials.
(1) Sample name: the first group of experimental samples of Diding. According to the experimental data, ∆T Curve of Reynolds temperature of Diding is shown in Fig. 1. The experiment is repeated three times. Figure 1 shows ΔT Curve of Reynolds temperature. According to the calculation, ∆m ignition wire = 0.0095 g, according to, ∆m Diding Qv = W cal ∆T−Q ignition wire ∆m ignition wire −Q capsule m capsule , Q Diding = 58526.79 J/g. The average of Q Diding is 56915.503 J/g. (2) In the same way, determine the combustion heat of Purslane, straw sandal board, June snow, pine vine rattan, and repeat the test for three times.The heat of combustion of 5 kinds of genuine medicinal materials, is shown in Table 1. According to Table 1, the order of combustion heat of 5 kinds of genuine medicinal materials, including Diding, Purslane, straw sandal board, June snow, pine vine rattan, was Diding > June snow > straw sandal board > Purslane > pine vine rattan. The combustion heat of the 5 kinds of genuine medicinal material test samples has ranged from 49,779.54 to 56,915.503 J/g, CV% < 3.65%.The combustion heat of Diding is 56,915.503 J/g, and the energy is the highest. The combustion heat of pine vine rattan is 49,779.548 J/g, and the energy is relatively small. Combustion heat is regarded as an important physical data to measure genuine medicinal material energy.  Table 2. It can be seen from Fig. 2 that the sample begins to decompose at 34.8 °C, which may be due to the thermal desorption of the residual small molecular substances in the sample, resulting in a small amount of mass loss of the sample, with a loss rate of 7.07%; After heating for a period of time, the temperature reaches 184.1 °C and enters the second stage of decomposition. A large amount of mass loss begins to appear in the sample until 404.05 °C, and the loss rate is 37.64%; then, with the continuous increase of temperature, the sample is further decomposed, and the mass of the remaining sample is 43.83%. With the increase of temperature, the DTG curve shows two peak shapes, and the inflection points of the peak shapes are 76.2 °C and 316.8 °C respectively. In addition, the DTA curve of Diding has an exothermic peak, with a peak value of 119.6 °C, a temperature range of 91.5-153.1 °C, and a peak area of 32 J/g; there is a smaller endothermic peak, with a peak value of 282.9 °C, and the temperature range is 252.8-308.6 °C, the peak area is 15.58 J/g. Thermogravimetric analysis of Purslane. The thermal gravimetric data of Purslane are shown in Fig. 3 and Table 3. It can be seen from Fig. 3 that the sample begins to decompose at 31.3 °C, which may be due to the thermal desorption of the residual small molecular substances in the sample, resulting in a small amount of mass   www.nature.com/scientificreports/ loss of the sample, with a loss rate of 4.25%; After heating for a period of time, the temperature reaches 146.9 °C, entering the second stage of decomposition, the sample begins to have a large mass loss until 279.7 °C, and the loss rate is 23.55%; As the temperature continues to rise, it enters the third stage of decomposition until 414.9 °C, the loss rate is 27.48%, the temperature continues to rise, the sample continues to decompose, and finally the mass of the remaining sample is 34.69%. With the increase of temperature, the DTG curve shows three peaks, and the inflection points of the peaks are 94.1 °C, 243.5 °C and 312.7 °C respectively. In addition, the DTA curve of Purslane has three exothermic peaks, the first peak is 107.6 °C, the temperature range is 83.1-145.0 °C, and the peak area is 23.87 J/g; The second peak is 186.4 °C, the temperature range is 171.4-207.3 °C, and the peak area is 8.054 J/g; The third peak is 228.6 °C, the temperature range is 211.2-242.8 °C, and the peak area is 19.21 J/g. Thermogravimetric analysis of straw sandal board. The thermal gravimetric data of straw sandal board are shown in Fig. 4 and Table 4. It can be seen from Fig. 4 that the sample begins to decompose at 60.8 °C. This may be due to the thermal desorption of the remaining small molecules in the sample, causing a small amount of mass loss in the sample, with a loss rate of 4.77%; after a period of time, When the temperature is increased, the temperature reaches 192.8 °C and enters the second stage of decomposition. The sample begins to show a large mass loss until 281.4 °C, and the loss rate is 14.65%; as the temperature continues to rise, it enters the third stage of decomposition until 410.6 °C, the loss rate is 28.35%, the temperature continues to rise, the sample continues to decompose, and finally the remaining sample mass is 43.18%.
With the increase of temperature, the DTG curve of straw sandal board shows three peak shapes, and the inflection points of the peak shapes are 98.9 °C, 243.5 °C and 329.6 °C respectively. In addition, the DTA curve has an exothermic peak, with a peak value of 125.3 °C, a temperature range of 100.3-152.3 °C, and a peak area of 21.25 J/g; there is an endothermic peak with a peak value of 265.7 °C and a temperature range of 225.3-302.0 °C, the peak area is 30.28 J/g. Thermo gravimetric analysis of June snow. The thermal gravimetric data of June snow are shown in Fig. 5, and Table 5. It can be seen from Fig. 5 that the sample begins to decompose at 42.6 °C, which may be due to the thermal desorption of the residual small molecular substances in the sample, resulting in a small amount of mass     www.nature.com/scientificreports/ loss of the sample, with a loss rate of 5.21%; After heating for a period of time, the temperature reaches 170.5 °C and enters the second stage of decomposition until 278.7 °C, and the loss rate is 21.91%; As the temperature continues to rise and enters the third stage of decomposition, the sample begins to have a large mass loss until 387.5 °C, the loss rate is 33.23%, the temperature continues to rise, the sample continues to decompose, and finally the mass of the remaining sample is 26.25%. With the increase of temperature, the DTG curve of June snow presents three peaks, and the inflection points of the peaks are 105.0 °C, 243.9 °C and 335 °C respectively. In addition, the DTA curve has three exothermic peaks, the first peak is 122.3 °C, the temperature range is 92.2-167.2 °C, and the peak area is 54.72 J/g; The second peak value is 218.4 °C, the temperature range is 188.3-256.2 °C, and the peak area is 30.92 J/g; The third peak is 312.9 °C, the temperature range is 286.4-340.3 °C, and the peak area is 8.942 J/g. Thermo gravimetric analysis of pine vine rattan. The thermal gravimetric data of pine vine rattan are shown in Fig. 6 and Table 6. It can be seen from Fig. 6 that the sample begins to decompose at 46.9 °C. This may be due to the thermal desorption of the remaining small molecules in the sample, causing a small amount of mass loss in the sample, with a loss rate of 6.85%; after a period of heating up, the temperature reaches 180.3 °C, and enter the second stage of decomposition, the sample begins to show a large amount of mass loss, until 422.4 °C, the loss rate is 50.81%; the temperature continues to rise, the sample further decomposes, and the remaining sample mass is 30.59%.
With the increase of temperature, the DTG curve of pine vine rattan shows two peak shapes, and the inflection points of the peak shapes are 89.6 °C and 308.8 °C respectively. In addition, the DTA curve has an exothermic  Figure 6. Thermo gravimetric (TG %) curve, derivative thermo gravimetric (DTG) curve and differential thermal analysis (DTA) curve of pine vine rattan. www.nature.com/scientificreports/ peak, with a peak value of 116.8 °C, a temperature range of 90.9-161.7 °C, and a peak area of 31.43 J/g; there is a larger endothermic peak with the peak value of 302.6 °C and a temperature range of 259.8-364.5 °C, and the peak area is 82.84 J/g.

Combustion stability analysis of 5 kinds of genuine medicinal materials. The thermogravimetric
parameters, including index X 1 the first stage weight loss percentage, index X 2 the fastest temperature in the first stage of weight loss, index X 3 the weight loss percentage in the second stage, index X 4 the fastest temperature in the second stage of weight loss, index X 5 the percentage of weight loss in the third stage, index X 6 the remaining mass percentage, index X 7 the peak area of the first stage and index X 8 the second stage peak area, were used to build combustion stability of genuine medicinal materials through gray pattern recognition. The thermal gravimetric parameter data of 5 kinds of genuine medicinal materials were shown in Table S1. Table S1 is in Supplementary Materials. Based on the thermogravimetric parameter data of five genuine medicinal materials, a multi-index evaluation system for combustion stability was established. Thermo gravimetry is to study the combustion characteristic index of genuine medicinal particles at different heating rates by thermo gravimetry analyzer to judge the combustion stability of genuine medicinal materials. According to the method of grey pattern recognition 56 , this subject calculates the correlation coefficient between each scheme and the ideal scheme composed of the best indicators, obtains the correlation degree from the correlation coefficient, and then sorts and analyzes it according to the correlation degree to draw a conclusion. The greater the correlation degree Z is, the better the sample effect is. Finally, compare all the Z values to draw the evaluation conclusion.  Table 7. Table 7 showed, the order of fat content (%) of 5 kinds of genuine medicinal materials, including Diding, Purslane, straw sandal board, June snow, pine vine rattan, was straw sandal board > Diding > pine vine rattan > June snow > Purslane, the order of calcium content (%) was pine vine rattan > June snow > Purslane > straw sandal board > Diding, the order of ash content was June snow > Purslane > straw sandal board > pine vine rattan > Diding. The energy value of genuine medicinal materials can also be reflected by the combustion heat, fat content to a certain extent. The contents of ash, fat, and calcium are regarded as important physical data to measure the quality of genuine medicinal materials. The quality of genuine medicinal materials is evaluated from the aspect of energy, which provides a strong scientific basis for the classification of genuine medicinal materials.

Determination of trace elements. Determination results of trace elements.
A method for the determination of 12 trace elements, including Mn, Mg, Fe, Co, Zn, Cu, Ni, Se, Sn, As, Li and Mo in 5 genuine medicinal materials, was established by inductively coupled plasma mass spectrometry (ICP-MS) based on microwave digestion [57][58][59][60][61] . Average value and standard deviation of 12 trace elements in 5 kinds of genuine medicinal materials, including Diding, Purslane, straw sandal board, June snow, pine vine rattan, were shown in Table 8.

Grey factor analysis of 12 trace elements in 5 kinds of genuine medicinal materials 62-66 .
Through grey factor analysis, the characteristic roots of grey factor correlation coefficient matrix and variance contribution rate of trace elements, including Mn, Mg, Fe, Co, Zn, Cu, Ni, Se, Sn, as, Li and Mo, are obtained, as shown in Table 9. According to Table 9, the cumulative contribution rate of the main factors of the first three grey factors reaches 93.403%, and the eigenvalues of the main factors of the first three grey correlation coefficient factors (λ > 1) are larger, that is, the main factors of the first three grey factors contribute the most to the explanatory variables. It is most appropriate to extract the main factors of the first three grey correlation coefficient factors, which represents 93.403% of the information of 12 trace elements in the five genuine medicinal materials in Guangxi, China. The gray correlation coefficient factor load matrix after rotation is shown in Table 10. It can be seen from Table 10 that the first main factor F 1 of the gray correlation coefficient mainly contains the original variables As, Co, Cu, Fe, Mn, Ni, Se trace element information necessary for the human body. The second main factor F 2 Table 7. Determination results of fat, calcium, ash content in 5 kinds of genuine medicinal materials (n = 3, CV% < 2.0%).

Sample
Fat content/% Calcium content/% Ash/% www.nature.com/scientificreports/ of the gray correlation coefficient mainly contains the information of the original variables Li, Mg and Mo. The third main factor F of the gray correlation coefficient mainly contains the information of the original variables Zn and Sn. The gray correlation coefficient factor score and the comprehensive gray correlation coefficient factor score are shown in Table 11. As can be seen from Table 11, the order of the content of 12 trace elements in 5 kinds of genuine medicinal materials, including Diding, Purslane, straw sandal board, June snow, pine vine rattan, is June snow > straw sandal board > Diding > Purslane > pine vine rattan. In terms of trace element content, June snow contains the highest trace elements in all samples.

Discussion
Construction of multi-index comprehensive evaluation system for 5 kinds of genuine medicinal materials. Construction of multi-index analysis and comprehensive evaluation system by entropy analysis. According to the combustion heat, thermogravimetric parameters, fat content, calcium content, trace Table 8. Average values ± standard deviation of 12 trace elements in 5 kinds of genuine medicinal materials (μg/g, n = 6).  Table 9. Grey factor coefficient characteristic root and variance contribution rate of 12 trace elements in 5 kinds of genuine medicinal materials.  www.nature.com/scientificreports/ element content and ash content, the multi-index comprehensive evaluation systems of five kinds of genuine medicinal materials including Diding, Purslane, straw sandal board, June snow, pine vine rattan were established by entropy method [67][68][69][70][71][72] .
According to the characteristics of entropy, this paper judges the randomness and disorder degree of an event by calculating the entropy value, and judges the dispersion degree of an index by using the entropy value [73][74][75][76] . The greater the dispersion degree of the index is, the greater the influence (weight) of the index on the comprehensive evaluation is, and the smaller the entropy value is. Using entropy method, 5 kinds of genuine medicinal materials were weighted to calculate the comprehensive score S.
(3) Information entropy value e and information utility value d, information entropy value of item j is e j = − 1 ln m n i=1 P ij lnP ij Information utility value d j = 1 − e j . (4) Weight of evaluation indicators. The greater the information utility value is, indicating that the more important the indicators, the greater the importance of evaluation is. Finally, the weight of the j index is The weighted summation formula is used to calculate the evaluation value of the sample. The larger the comprehensive score S is, the better the sample effect is. Finally, compare all S values, that is, draw the evaluation conclusion. Using EXCEL calculation, the S values of Diding, Purslane, straw sandal board, June snow, pine vine rattan were 0.3764, 0.2777, 0.4876, 0.5744 and 0.2688 respectively.
According to combustion heat, combustibility (combustion stability of genuine medicinal materials), fat, calcium, ash, trace element content, the comprehensive evaluation results of multi-index analysis constructed by entropy analysis showed that the comprehensive evaluation multi-index order of 5 genuine medicinal materials, including Diding, Purslane, straw sandal board, June snow and pine vine rattan, was June snow > straw sandal board > Diding > Purslane > pine vine rattan. Therefore, the comprehensive evaluation results of the quality of genuine medicinal materials selected in this study were June snow the best, followed by straw sandal board.
Construction of multi-index analysis and comprehensive evaluation system by grey correlation coefficient cluster analysis. The gray correlation coefficient cluster analysis is based on the many properties of the sample, and the cluster analysis diagram is obtained from the correlation coefficient [77][78][79][80] . According to the literature 81,82 , the classification is carried out according to the degree of affinity of the nature of the sample. All cases are classified into different classes, making the same class Individuals in different classes have greater similarities, and individuals in different classes have greater differences. The multi-index comprehensive cluster analysis system of combustion heat, combustibility (combustion stability of genuine medicinal materials), fat content, calcium, ash and trace element content of 5 genuine medicinal materials in Guangxi was established. The gray correlation coefficient cluster analysis tree diagram was shown in Fig. 7. As can be seen from Fig. 7, 5 kinds of genuine medicinal materials from different producing areas, namely Diding, Purslane, straw sandal board, June snow and pine vine rattan, were divided into three categories according to the results of grey correlation coefficient cluster analysis. Straw sandal board, June snow were a class, Diding for one class, and Purslane and pine vine rattan for a class. Through the grey correlation coefficient cluster analysis, we found the similarity degree and genetic relationship between the properties of genuine medicinal materials from different origins, which can help better study the classification of genuine medicinal materials.

Conclusion
Thermogravimetric parameters were applied to the evaluation of the combustion stability of genuine medicinal materials through gray pattern recognition, which provided a strong scientific basis for the evaluation and research of the combustion stability of genuine medicinal materials by thermogravimetric analysis. www.nature.com/scientificreports/ A method for the determination of 12 trace elements, including Mn, Mg, Fe, Co, Zn, Cu, Ni, Se, Sn, As, Li and Mo in 5 genuine medicinal materials, was established by inductively coupled plasma mass spectrometry (ICP-MS) based on microwave digestion, and the data were comprehensively analyzed by the grey factor analysis method. From the content of trace elements, the trace elements contained in Junxue were the highest in all samples. These studies are for scientific research on trace elements and biological activities in medicinal materials, for the rational development of the quality control of Chinese medicinal materials in the standardized planting base of Chinese medicinal materials in Guangxi, and to provide reference for the medical and health care of Chinese medicinal materials.
In this paper, according to combustion heat, differential thermal-thermogravimetric analysis, fat content, calcium and ash content, trace element content data of genuine medicinal materials, systematic multi-index comprehensive evaluation systems were constructed through gray pattern recognition, gray factor analysis, entropy method and gray correlation coefficient cluster analysis. As long as the various parameters like combustion heat, thermogravimetric parameters, and fat content, calcium content, trace element content, ash content, etc. are measured, systematic multi-index comprehensive evaluation systems are constructed through gray pattern recognition, gray factor analysis, entropy method and gray correlation coefficient cluster analysis, and then can be utilized for the identification of genuine materials. This research has important theoretical and practical significance for the multi-index measurement and comprehensive evaluation of genuine medicinal materials, and can provide scientific basis and research significance for the research of multi-index quality control of genuine medicinal material.
The multi-index comprehensive evaluation system established in this study provides a new idea for the quantitative control of the quality of genuine medicinal materials, and provides a powerful way for the largescale development and classification research of genuine medicinal materials and provides basic support for the selection of raw materials of genuine medicinal materials and the application of quantitative control mode of multi-index ingredients to the quality control of genuine medicinal materials.