Walnut oil oxidation degree analysis method based on olfaction analysis sensor array
Through olfactory analysis sensor array and HS-SPME-GC-MS technology, combined with Pearson correlation analysis and PLSR model, the complexity and inaccuracy of walnut oil oxidation degree detection are solved, and a fast and accurate judgment of oxidation degree is achieved. , improves the accuracy and environmental protection of testing.
Patent Information
- Application Number
- CN202411009490.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-26
- Publication Date
- 2024-11-15
AI Technical Summary
In the prior art, the detection method for oxidation of walnut oil is complicated to operate, requires a large number of chemical reagents, is not environmentally friendly, and the measurement accuracy of the rapid detection test strips is poor and the repeatability is poor, so it is impossible to accurately judge the oxidation degree of walnut oil.
Using the olfactory analysis sensor array combined with HS-SPME-GC-MS technology, characteristic oxidative markers were screened through Pearson correlation analysis, and PLSR prediction model was constructed to quickly judge the oxidation degree of walnut oil, and to predict peroxidation using the olfactory analysis sensor response value. Value and acid value.
It realizes rapid and accurate judgment of the degree of oxidation of walnut oil, improves the accuracy and repeatability of detection, reduces the requirements for operators, and reduces the use of chemical reagents, and improves the environmental protection of detection.
Smart Images

Figure CN118961916A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the field of walnut oil oxidation degree analysis, and in particular to a walnut oil oxidation degree analysis method using an olfactory analysis sensor array. Background Art
[0002] Walnut oil is rich in unsaturated fatty acids, which react with oxygen in the air in a certain environment to produce an oxidation reaction, which is often called "rancid". When the oxidation of walnut oil reaches a certain level, it will produce a "rancid smell". This unpleasant smell is caused by the volatile components produced by the oxidation of fatty acids.
[0003] In order to ensure the quality and food safety of oils and fats, my country's edible vegetable oil standards stipulate the limits of acid value and peroxide value of oils and fats. At present, the detection methods for peroxide value and acid value in my country's national standards include titration and potentiometric titration, but both methods have prominent problems such as complex operation, requiring a large amount of chemical reagents, high operating experience requirements for operators, high testing costs, and environmental pollution. Secondly, there are also rapid test strips on the market, but the accuracy of their measurements cannot be guaranteed and their repeatability is poor.
[0004] The prior art generally uses the electronic nose response value to construct a linear fitting curve of the acid value and the peroxide value for the oxidation degree detection of volatile oils. However, the above method simply linearly fits the detection response value of the electronic nose with the peroxide value and the acid value. The fitting equation is also relatively simple, and the result of judging the oxidation degree of the oil is not accurate enough.
[0005] Therefore, it is necessary to provide a method for accurately and quickly judging the degree of oxidation of walnut oil. Summary of the invention
[0006] In order to solve the problems in the prior art of peroxide value and acid value detection, such as complicated operation, large amount of chemical reagents required, environmental pollution, failure to ensure the measurement accuracy of rapid test strips, and poor repeatability, the present invention provides a walnut oil oxidation degree analysis method based on an olfactory analyzer sensor array.
[0007] To achieve the above purpose, the specific technical solution of the present invention is as follows.
[0008] A method for quickly determining the oxidation degree of walnut oil comprises the following steps:
[0009] The walnut oil to be tested is tested by the olfactory analysis characteristic reaction sensor, and whether the walnut oil is oxidized is determined according to the |r| value of the correlation curve;
[0010] If |r| of the correlation curve is less than 0.6, the judgment result is no, and the acid value and peroxide value of the walnut oil are not predicted;
[0011] If |r| of the correlation curve is greater than 0.6, the judgment result is yes, and the walnut oil to be tested is subjected to olfactory analysis to obtain the olfactory analysis sensor response value of the walnut oil to be tested, and the response value of the walnut oil oxidation degree is brought into the PLSR prediction model of the walnut oil oxidation degree to obtain the peroxide value prediction value and acid value prediction value of the walnut oil to be tested, which are used to judge the oxidation degree of the walnut oil to be tested.
[0012] In a preferred embodiment, the specific process of obtaining the olfactory analysis characteristic reaction sensor is as follows:
[0013] Walnut oil with different oxidation degrees was subjected to HS-SPME-GC-MS analysis to obtain the volatile substances of walnut oil with different oxidation degrees. The volatile substances with VIP>1 were retained and subjected to Pearson correlation analysis with the response values of the olfactory analysis sensor. The volatile substances with p<0.05 and |r|>0.8 were retained to obtain characteristic oxidation markers, as well as olfactory analysis characteristic reaction sensors with characteristic reactions to the characteristic oxidation markers.
[0014] In a preferred embodiment, the specific construction process of the PLSR prediction model for the degree of oxidation of walnut oil is as follows:
[0015] Walnut oil with different oxidation degrees was subjected to olfactory analysis and the olfactory analysis sensor response values of walnut oil with different oxidation degrees were obtained; the olfactory analysis sensor response values of walnut oil with different oxidation degrees were respectively combined with the acid value and peroxide value of walnut oil with different oxidation degrees to construct a partial least squares regression equation to obtain the PLSR prediction model of walnut oil oxidation degree.
[0016] In a preferred embodiment, the PLSR prediction model for the degree of oxidation of walnut oil is as follows:
[0017]
[0018] Where: b i is the regression coefficient corresponding to the i-th eigenvalue collected by the olfactory analysis sensor; i = 1, 2, 3...n; b i It is calculated by the least squares formula after the data matrix is standardized;
[0019] X i is the i-th eigenvalue data collected by the olfactory analysis sensor;
[0020] Yes i and X i The sum of the products of ;
[0021] C is the predicted value when the feature values collected by the olfactory analysis sensor are all 0. C is calculated by minimizing the squared error between the dependent variable and the model prediction value;
[0022] Y is the predicted value of the oxidation degree of walnut oil.
[0023] In a preferred embodiment, the characteristic oxidation markers are as follows:
[0024] Hexanal, acetonitrile, trimethylsilanol, (E,E)-2,4-heptadienal, hexanoic acid, (E)-2-octenal, (E,E)-2,4-decadienal;
[0025] The characteristic reaction sensors of the olfactory analysis are: W8, W13, W11, W12;
[0026] The model number of W8 is TGS2618;
[0027] The model number of W11 is TGS2602;
[0028] The model number of W12 is WSP2110;
[0029] The model of W13 is MQ137.
[0030] In a preferred embodiment, the conditions for the olfactory analysis detection are:
[0031] The injection flow rate is 350-400 mL / min; the detection time is 60-90 s, and the injection time is 30-50 s.
[0032] In a preferred embodiment, the chromatographic conditions of the HS-SPME-GC-MS analysis are as follows:
[0033] Chromatographic column: DB-WAX (30.0m×250μm, 0.25μm); initial temperature was 40℃ for 4min, then increased to 110℃ at a rate of 4℃ / min, and then increased to 230℃ at a rate of 10℃ / min for 10min; vaporization chamber temperature was 250℃; transfer line temperature was 230℃; carrier gas was He; carrier gas flow rate was 1.0mL / min; no split;
[0034] The mass spectrometry conditions for the HS-SPME-GC-MS analysis are as follows:
[0035] EI source; electron energy 70eV; ion source temperature 230℃; quadrupole 150℃; scanning mode is Scan; scanning mass range is 20~500u.
[0036] In a preferred embodiment, the specific preparation method of the walnut oil with different oxidation degrees is as follows:
[0037] In an oven at 100°C, oxygen was introduced at a rate of 3L / min to conduct a 48h accelerated oxidation test of walnut oil. A 5mL sample was taken every 4h and placed in a 20mL headspace chromatography injection bottle. After balancing the odor in a 60°C constant temperature water bath for 30min, samples at different oxidation stages were obtained. Because the accelerated oxidation test method was used, the oxidation rate of walnut oil was about 20 times the normal oxidation rate, and the oxidation of walnut oil was a limited reaction. In addition, since the fresh walnut oil sample did not have a large number of oxidation initiators in the early stage, the continuous introduction of oxygen would accelerate the oxidation of walnut oil, thereby inducing the oxidative decomposition of walnut oil. The short-chain fatty acids obtained by oxidative decomposition would further induce the oxidative decomposition of fresh walnut oil, forming a chain-proliferation-like oxidative decomposition, so sampling was selected every 4h.
[0038] In a preferred embodiment, the walnut oils with different oxidation degrees are subjected to gas balance under water bath conditions before being subjected to olfactory analysis.
[0039] The temperature of the water bath is 40-60°C;
[0040] The gas balancing time is 20 to 30 minutes.
[0041] In a preferred embodiment, the acid value of walnut oil with different oxidation degrees is measured using the national food safety standard GB 5009.229-2016;
[0042] The peroxide value of walnut oil with different oxidation degrees was measured using the national food safety standard GB 5009.227-2023.
[0043] Compared with the prior art, the present invention has the following beneficial effects:
[0044] (1) The present invention separates the olfactory analyzer gas sensor array, performs Pearson correlation analysis on the oxidation characteristic markers of walnut oil in each oxidation stage and the response value of the olfactory analysis sensor, and then screens out the olfactory analysis characteristic reaction sensor, so as to quickly and accurately determine whether the walnut oil is oxidized. The walnut oil oxidation degree PLSR prediction model is used to determine the oxidation degree of the walnut oil. The method of the present invention can not only quickly and accurately determine whether the walnut oil is oxidized, but also determine the oxidation degree of the walnut oil, effectively improving the accuracy of determining the oxidation degree of the walnut oil.
[0045] (2) The present invention combines sensor technology with information processing technology to convert gas volume fraction into corresponding electrical signals, uses walnut oil raw materials with different oxidation degrees, and establishes a one-to-one mathematical relationship with the acid value and peroxide value of vegetable oil after processing through multi-channel olfactory analysis detection data, analyzes the quantitative relationship between the measured value and the olfactory analyzer data, and constructs a rapid discrimination model and forms a method for rapid discrimination of the oxidation degree of walnut oil after repeated training and verification. In this way, the peroxide value and acid value of the corresponding model feedback sample after olfactory analysis detection of the oxidation degree of walnut oil can be realized, the sample pre-treatment is simple, the entire detection process is very fast, and the requirements for operators are very low. BRIEF DESCRIPTION OF THE DRAWINGS
[0046] Figure 1 It is the technical roadmap of the present invention.
[0047] Figure 2 The electric signal radar diagram and PCA principal component analysis of the olfactory analyzer of the present invention are shown in Figure 1. A is the electric signal radar diagram of the olfactory analyzer of the present invention, and B is the PCA principal component analysis of the olfactory analyzer of the present invention.
[0048] Figure 3 GC-MS volatile substance type and quantity diagram of the present invention, wherein A is the type of GC-MS volatile substance of the present invention, and B is the quantity of volatile substance of the present invention.
[0049] Figure 4 The OPLS-DA analysis diagram and cross-validation diagram of GC-MS volatile substances of the present invention are shown in Figure 1. A is the multivariate statistical analysis diagram of OPLS-DA of GC-MS volatile substances, and B is the cross-validation diagram.
[0050] Figure 5 This is a clustering heat map of volatiles of the GC-MS characteristic markers of the present invention.
[0051] Figure 6 It is a Pearson correlation analysis diagram of the olfactory analyzer of the present invention and volatile representatives.
[0052] Figure 7 The figure is a modeling result diagram of the peroxide value and acid value of walnut oil of the present invention. Among them, A is a modeling result diagram of the peroxide value of walnut oil of the present invention, and B is a modeling result diagram of the acid value of walnut oil of the present invention. DETAILED DESCRIPTION
[0053] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.
[0054] Unless otherwise specified, the experimental methods described in the following examples are all conventional methods; the reagents and materials described are all commercially available unless otherwise specified.
[0055] With the continuous development and progress of sensor technology, the application of gas sensors has become very extensive and is used in many industries such as medical, security, and food industry.
[0056] The prior art usually uses an electronic nose to construct a linear fitting curve of acid value and peroxide value for the detection of the oxidation degree of oils. However, the above method simply linearly fits the detection response value of the electronic nose with the peroxide value and the acid value, and the fitting equation is relatively simple, and the judgment result of the corruption degree of ethyl ester fish oil is not accurate enough. In order to improve the accuracy, the partial least squares method is also used to comprehensively analyze the collected physical and chemical index data and the electronic nose test data and construct the acid value and peroxide value prediction models respectively. However, this method only uses the electronic nose electrical signal response value to make a radar chart and PCA principal component analysis, which is not accurate enough for judging the corruption degree during the storage process of oils.
[0057] The present invention provides a walnut oil oxidation degree analysis method based on an olfactory analysis sensor array, which can quickly and accurately determine the oxidation degree of walnut oil. The olfactory analyzer gas sensor array is split, and each olfactory analyzer electrical signal value is normalized and six characteristic values are extracted, and the detected headspace solid phase microextraction-gas chromatography mass spectrometry signal is integrated, and the oxidation characteristic markers of the walnut oil in each oxidation stage and the response value of the olfactory analysis sensor are sorted out for Pearson correlation analysis, and then the most suitable olfactory analyzer sensor array, that is, the olfactory analysis characteristic reaction sensor, is screened out, which is used to quickly and accurately determine whether the walnut oil is oxidized, and then the peroxide value and acid value detection values of the walnut oil with different oxidation degrees obtained after accelerated oxidation are used to establish a PLSR partial square regression equation with the response value obtained by the olfactory analyzer sensor array, so that the walnut oil oxidation degree PLSR prediction model is more scientific and rigorous, and the predicted value has a higher degree of familiarity with the true value.
[0058] The following is a detailed description of a method for analyzing the degree of oxidation of walnut oil based on an olfactory analysis sensor array.
[0059] 1. Material preparation:
[0060] 1.1 Main materials and instrument parameters
[0061] 1.1.1 Main materials
[0062] Walnut oil, the raw materials of walnut variety 'Xin 2' for oil pressing are purchased from Wensu County, Aksu, Xinjiang. After being transported back, they are immediately peeled, washed and drained of surface moisture, and placed in a 45℃ AGHD-15ELC air-energy heat pump drying oven to dry until the moisture content is lower than ≤5% specified in GB / T22327-2019 "National Standard of the People's Republic of China for Walnut Nut Quality Grades". The dried walnut kernels are shelled and walnut oil is produced using hydraulic cold pressing technology. After filtering, they are stored in a cold storage at 4-6℃ for later use.
[0063] 1.1.2 Instrument parameters
[0064] 1) Chromatographic conditions:
[0065] Chromatographic column: DB-WAX, 30.0m×250μm, 0.25μm;
[0066] Initial temperature: 40℃ for 4 min, then increase to 110℃ at a rate of 4℃ / min, then increase to 230℃ at a rate of 10℃ / min and then maintain for 10 min;
[0067] Vaporization chamber temperature 250°C; transfer line temperature 230°C; carrier gas He; carrier gas flow rate 1.0 mL / min; no split.
[0068] Mass spectrometry conditions: EI source; electron energy 70 eV; ion source temperature 230°C; quadrupole 150°C; scanning mode Scan; scanning mass range 20-500u.
[0069] 2. Analysis method of walnut oil oxidation degree based on olfactory analyzer
[0070] 2.1 Walnut oil accelerates oxidation
[0071] The accelerated oxidation test was carried out for 48 hours in a 100℃ oven and 3L / min oxygen flow rate. A 5mL sample was taken every 4 hours and placed in a 20mL headspace chromatography injection bottle. The sample was taken out for olfactory analysis after the odor was balanced in a 60℃ constant temperature water bath for 30 minutes. The sampling every 4 hours was marked as Tx, and x represented the sampling time.
[0072] 2.2 Obtaining peroxide value and acidity
[0073] The acidity and peroxide value of samples at different oxidation stages were obtained according to the national standards GB 5009.227-2016 "National Food Safety Standard Determination of Peroxide Value in Food" and GB 5009.229-2016 "National Food Safety Standard Determination of Acid Value in Food". Each sample was tested three times independently and the average value was calculated.
[0074] 2.3 Olfaction analyzer and HS-SPME-GCMS detection
[0075] 2.3.1 Olfactory analyzer detection
[0076] The samples at different oxidation stages were balanced in a water bath at 40-60°C for 20-30 minutes. Before sampling, indoor air filtered with standard activated carbon was introduced to clean the channel to eliminate the influence of the external odor environment. Then, the olfactory analyzer was tested. The olfactory analyzer test conditions were as follows: the gas flow rate was 350-400 mL / min; the detection time was 60-90 s, and the injection time was 30-50 s. The olfactory analyzer data of the samples at different oxidation stages were obtained, and each sample was tested three times independently. The olfactory analyzer contains 14 metal oxide sensor arrays. The names and performance descriptions of each sensor are shown in Table 1.
[0077] Table 1 Name and performance of each sensor in the olfactory analyzer
[0078]
[0079]
[0080] 2.3.2 HS-SPME-GC-MS detection
[0081] Weigh 2g of sample and transfer it into a 15mL extraction bottle, then add a stirring magnet and seal it quickly. Aging the SPME extraction fiber head at 250℃ in the GC-MS injection port until there is no impurity peak. Place the sample bottle on the solid phase microextraction device, set the temperature to 60℃, and the stirring speed to 500rpm; preheat the sample bottle in the extraction device for 15min; insert the SPME extraction head into the headspace of the sample through the bottle cap, push out the fiber head, the extraction head is about 1.0cm above the upper surface of the sample, and the headspace extraction is 40min; withdraw the fiber head and pull the extraction head out of the sample bottle; then insert the extraction head into the GC-MS injection port, push out the fiber head, analyze at 250℃ for 3min, inject and analyze, and obtain the mass spectrometry data of walnut oil samples with different oxidation degrees.
[0082] The mass spectrometry data of walnut oil samples with different oxidation degrees were qualitatively analyzed using the MS database NIST11 and retention time; the column bleed peak was deducted from the database screening results to obtain representative volatiles of walnut oil samples with different oxidation degrees.
[0083] Quantitative analysis: The quantitative analysis by area normalization method is to use the percentage of the peak area of the identified component to the sum of the areas of all identified components as the quantitative result. The formula is as follows:
[0084]
[0085] Wherein: Ci——content of a certain identified component, %; Ai——peak area corresponding to a certain identified component; n——represents the total number of identified components.
[0086] 2.4 Peroxide value PLSR formula and acid value PLSR formula
[0087] A p correlation analysis was performed on the representative volatiles obtained in 2.3 and the olfactory analyzer response values corresponding to the olfactory analyzer sensor array, and finally 14 sensor arrays were obtained that were positively correlated with the volatile representatives of each oxidation degree; the six eigenvalues extracted by the 14 sensor arrays were arranged horizontally in sequence, and the detected peroxide value and acid value were arranged vertically to form an equation with the olfactory analyzer eigenvalue as X and the peroxide value as Y.
[0088] The above arrangement data is input into PLSR modeling analysis, and the specific process is as follows:
[0089] Let (x, y) be a pair of observations, and x = [x1, x2, ..., x n ]T, y = R satisfies the theoretical function: y = f(x,w), where w = [w1,w2,…,w n ]T is the unknown parameter. Calculate the optimal estimate of the parameter w in the function f(x,w). Given m sets of observation data (xi,yi), solve the objective function:
[0090]
[0091] Assume p dependent variables y1, y2, ..., y P with m independent variables x1, x2, …, x m ; The n-time observation data matrices of the independent variable group and the dependent variable group are An×m and Bn×P respectively (one sample per row):
[0092]
[0093] Extract the first pair of components u1 and v1 of the two variable groups respectively, and make their correlation maximum. u1 is the independent variable set X = [x1, x2, ..., x m ] Linear combination of T:
[0094] u1=α 11 x1+…α 1m x m =ρ (1)T X
[0095] v1 is a linear combination of the dependent variable set Y = [y1,…,yP]T:
[0096] ν1=β 11 y1+…α 1p y p =γ(1)T Y
[0097] u1 and v1 each extract as much variation information of their variable groups as possible; the correlation between u1 and v1 reaches the maximum.
[0098] From the standardized observation data matrices A and B of the two sets of variables, the score vector of the first pair of components can be calculated, denoted as and
[0099]
[0100] At this point, the above two requirements are transformed into mathematical conditional extreme value problems:
[0101]
[0102] Using the Lagrange multiplier method, the problem is transformed into calculating the matrix M = A T BB T The eigenvalues and eigenvectors of A, where the unit eigenvector corresponding to the largest eigenvalue is ρ(1), and
[0103]
[0104] Obtain and Available express
[0105] Establish y1, yP pair The regression of x1, xm The regression model is assumed to be:
[0106]
[0107] Where σ(1)=[σ11,…,σ1m]T, τ(1)=[τ11,…,τ1P]T are the parameter vectors in the regression model, and A1 and B1 are the residual matrices.
[0108] The least squares estimate of the regression coefficient vector σ(1), τ(1) is:
[0109]
[0110] σ(1), τ(1) are the model effect loadings.
[0111] Repeat the above steps using the residual matrices A1 and B1 instead of A and B.
[0112]
[0113] Assume that the rank of the n×m independent variable data matrix A is r≦min(n-1,m), then there are r components u1, u2, …, u r So that:
[0114]
[0115] put u k =α k1 x1+…+α km x m (k=1,2,…,r), substitute into Y=u1τ(1)+…+urτ(r), and we get the partial least squares regression equation of P dependent variables:
[0116] y j =C j1 x1+…+C jm x m , j=1,2,…,P.
[0117] The partial least squares regression equation obtained by the above method can not only perform regression modeling under the condition that there are serious multiple correlations in the independent variables, but also allow regression modeling under the condition that the number of samples is less than the number of independent variables; the partial least squares regression method obtained in the present invention has the advantages of simple calculation, high prediction accuracy and easy explanation compared with other modeling methods. The peroxide value PLSR formula and the acid value PLSR formula are shown in Table 2.
[0118] Table 2 Peroxide value PLSR formula and acid value PLSR formula
[0119]
[0120] The final partial least squares (PLSR) regression equation is:
[0121]
[0122] Where: b i is the regression coefficient corresponding to the i-th eigenvalue collected by the olfactory analysis sensor. i=1,2,3……n; b i It is obtained by standardizing the data matrix and calculating it using the least squares formula.
[0123] X i is the i-th eigenvalue data collected by the olfactory analysis sensor.
[0124] C is the predicted value when the eigenvalues collected by the olfactory analysis sensor are all 0. C is calculated by minimizing the square error between the dependent variable and the model prediction value.
[0125] Y is the predicted value of the oxidation degree of walnut oil.
[0126] Therefore, the PLSR equation for the peroxide value of walnut oil is:
[0127] Y=0.0030X1+0.0015X2-0.012X3-0.0714X4-0.0006X5+0X6+0.0054X7+0.0133
[0128] X8-0.0041X9-0.2428X 10 -0.0006X 11 -0.0002X 12 +0.0201X 13 +0.0280X 14 -0.0007X 15 -0.0406X 16 +0.0009X 17 +1X 18 +0.0269X 19 +0.0369X 20 -0.0026X 21 -0.1380X 22 +0.0013X 23 +0.0001X 24 +0.0122X 25 +0.0082X 26 +0.0021X 27 +0.1206X 28 -0.0003X 29 +0X 30 +0.0252X 31 +0.0388X 32 -0.0015X 33 -0.0787X 34 +0.0043X 35 +0.0001X 36 +0.0072X 37 +0.0177X 38 -0.0003X 39 -0.0127X 40- 0.0023X 41 -0.0001X 42 +0.0126X 43 +0.0145X 44 -0.0006X 45 -0.0421X 46 -0.0004X 47 +0X 48 +0.0343X 49 +0.0449X50 +0.0028X 51 +0.1655X 52 +0.0042X 53 +0.0002X 54 +0.0241X 55 +0.0297X 56 +0.0047X 57 +0.2834X 58 +0.0036X 59 +0.0001X 60 +0.0213X 61 +0.0280X 62 -0.0021X 63 -0.1118X 64 +0.0027X 65 +0.0001X 66 +0.0169X 67 +0.0195X 68 +0.0034X 69 +0.1986X 70 +0.0021X 71 +1X 72 +0.0175X 73 +0.0206X 74 +0.0025X 75 +0.1432X 76 +0.0017X 77 +0.0001X 78 +0.0184X 79 +0.0245X 80 -0.0006X 81 -0.0325X 82 +0.0027X 83 +0.0001X 84 +0.0168X 85 +0.0223X 86 -0.0018X 87 -0.1164X 88 +0.0023X 89 +0.003X 90 +0.1200;
[0129] The PLSR equation for the acid value of walnut oil is:
[0130] Y=0.002X1+0.003X2+1X3+0.0019X4+1X5+1X6+0.002X7+0.02X8+1X9+0.009
[0131] X 10 +1X 11 +1X 12 +0.002X 13 +0.03X 14 +1X 15 +0.004X 16 +1X 17 +1X 18 +0.002X 19 +0.003X 20 +1X 21 -0.011X 22 +0X 23 +0X 24 +0.002X 25 +0.004X 26 +0X 27 +0.004X 28 +0.001X 29 +0X 30 +0.002X 31 +0.004132+0X 33 -0.011X 34 +0.001X 35 +0X 36 +0.003X 37 +0.006X 38 +0X 39 +0.009X 40 +0X 41 +0X 42 +0.02X 43 +0.004X 44 +0X 45 -0.013X 46 +0X 47 +0X 48 +0.002X 49 +0.003X 50 +0X 51 +-0.011X 52 +0.001X 53 +0X 54 +0.002X 55 +0.003X 56 +0X 57 +0.010X 58 +0.0016X 59 +0X 60 +0.003X 61 +0.006X 62 +0X 63 -0.016X 64+0.0017X 65 +0X 66 +0.002X 67 +0.003X 68 +0X 69 +0.011X 70 +0.001X 71 +0X 72 +0.003X 73 +0.003X 74 +0X 75 +0.027X 76 +0.001X 77 +0X 78 +0.03X 79 +0.005X 80 +0X 81 -0.07X 82 +0.003X 83 ++0X 84 +0.003X 85 +0.005X 86 +0X 87 +0.0028X 88 +0.001X 89 +0X 90 +0.337424;
[0132] 2.5 Walnut oil oxidation degree detection
[0133] 2.5.1 Index measurement standards
[0134] The acid value and peroxide value are respectively based on the national standards GB 5009.227-2016 “National Food Safety Standard - Determination of Peroxide Value in Foods” and GB 5009.229-2016 “National Food Safety Standard - Determination of Acid Value in Foods”.
[0135] 2.5.2 Data processing
[0136] Origin 2021 software was used for data processing and plotting of acid value and peroxide value. Multivariate statistical analysis was performed on the volatile odor information collected by the olfactory analyzer, including principal component analysis (PCA) and bar radar chart analysis. Partial least squares (OPLS-DA), cluster analysis, and Pearson correlation analysis were performed on the GC-MS data.
[0137] 2.5.3 Experimental Results
[0138] 1) Analysis of walnut oil olfactory analyzer radar chart at different oxidation degrees
[0139] The olfactory analyzer is a high-tech product that simulates the olfactory organs of animals or humans. It can respond quickly and detect gas components with good repeatability. The olfactory analyzer was used to detect the volatile substances in walnut oil at different storage times. The response values of the volatile substances in walnut oil on 14 sensors varied. Figure 2 A shows that the response values of each metal odor sensor show a gradual upward trend and reach the maximum value at T48, indicating that the volatile odor of walnut oil becomes more intense after storage. The changes in volatile odors during different storage periods are mainly reflected in the W8, W13, W11, W12, W5, W7, and W2 sensors, which are sensitive to odors such as methane, natural gas, biogas, nitrogen oxides, ammonia, sulfides, benzene steam fluorine, plastic packaging and metal packaging, ozone, carbon monoxide, and ammonia. These results show that the volatile substances in walnut oil can be effectively distinguished by the olfactory analyzer, and the degree of oxidation of walnut oil can also be judged by the size of the response value of the olfactory analyzer, further indicating that the changes in volatile substances change with the changes in the degree of oxidation of walnut oil.
[0140] In order to further clarify the differences in volatile compounds in walnut oils with different oxidation degrees, PCA analysis was performed on the response values of the olfactory analyzer sensor. Figure 2 B shows that the contribution rate of PC1 is 93.7%, the contribution rate of PC2 is 3.4%, and the total contribution rate is 97.1%. The repeatability between each group of samples is good. The PCA score chart clearly distinguishes the maximum response values of the olfactory analyzer of walnut oil with different oxidation degrees, indicating that PCA can well reflect the information between sample groups. According to the PC1 axis, the control group CK is compared with the others, and the other groups show greater differences with the extension of storage time, indicating that with the extension of storage time, the difference in volatile substances between samples gradually increases; but it is worth noting that there is gradually some overlap between T20 and T24, indicating that the difference in volatile compounds between the two is not large; during T28-T36, there is a large difference, which is speculated to be due to the high-speed oxidation stage of walnut oil; and there is also a little overlap between T40-T48 samples, indicating that the difference in volatile compounds between samples at this stage is not large or the change trend is not obvious.
[0141] 2) Analysis of relative content and types of volatile compounds in walnut oil with different oxidation degrees
[0142] The HS-SPME-GC-MS method was used to analyze the volatile components and relative contents of walnut oils with different oxidation degrees, and to clarify the differences in characteristic volatile organic compounds in walnut oils with different oxidation degrees. A total of 298 compounds were identified, including 31 aldehydes, 20 acids, 43 esters, 26 ketones, 13 benzenes, 3 nitriles, 39 olefins, 39 alcohols, 44 alkanes, 14 amines, and 26 other heterocyclic substances. Figure 3 A shows that a total of 12 types of volatile organic compounds were detected in the accelerated oxidation process of walnut oil. It is worth noting that during the accelerated oxidation process, the types and quantities of volatile compounds underwent obvious dynamic changes.
[0143] Specifically, aldehydes showed the greatest change, with the relative content increasing from 24.58% to 62.9%; the number increased from 5 to 12. This may be due to the fact that under high temperature and oxygen conditions, the unsaturated fatty acids in walnut oil underwent lipid peroxidation and amino acid degradation.
[0144] Ester compounds mainly give walnut oil fruity and floral aromas, mainly from the esterification reaction between alcohols and acids. Its relative content is 2.11% at the initial stage, reaching a maximum of 22.25% after 4 hours, and then gradually decreases to 5.55% as the oxidation time increases. Studies have shown that lipid substances will gradually decrease with the extension of storage time at high temperatures, which is similar to our research.
[0145] The source of alcohol compounds is mainly the decomposition of fatty acids. The relative content reaches the highest at 12h of oxidation, increasing from 4.8% to 17.81%, which is 4.1 times the initial value, indicating that alcohol compounds are important products in the oxidation process of walnut oil. Studies have shown that alcohol is one of the key substances in the aroma component characteristics of walnut oil, which is similar to the conclusion of this study. Which changes in alcohols cause the flavor changes of walnut oil needs further analysis.
[0146] The initial content of amides was 22.48%, which could not be detected after 32 hours of oxidation. The loss of this type of substance reached the maximum during the oxidation process of walnut oil, and it has not been reported in walnut oil. Studies have shown that amines are an important indicator for evaluating the freshness of fish. Hydrocarbons, alkanes, and alkenes come from the oxidative degradation of unsaturated fatty acids in walnut oil, which usually give walnut oil a sweet taste. The relative content in the CK group was 23.75%, but it decreased to 5.96% with the increase of oxidation time. Studies have shown that they are important intermediates in the formation of heterocyclic compounds, and alkenes will form aldehydes and ketones under certain conditions, which is considered to be an important factor in the composition of flavor.
[0147] In addition, walnuts also produce some complex heterocyclic substances during the oxidation process, such as furan, pyridine and other heterocyclic compounds, which give walnut oil a roasted nut, coffee and caramel flavor. Studies have shown that furan is the result of free radical cyclization of linoleic acid and oxygen, and is an important aroma feature in walnut oil. In this experiment, 2,3-dihydrofuran was detected at the 16th hour of the oxidation process. In addition, 2-pentylfuran was also detected. Studies have shown that this substance is an undesirable aromatic compound.
[0148] In summary, the oxidation process of walnut oil is involved in many complex oxidation, degradation and polymerization reactions, accompanied by the dynamic changes of many aldehydes, acids, lipids and heterocyclic compounds. Therefore, the monitoring and identification of these volatile compounds plays a vital role.
[0149] 3) OPLS-DA analysis of volatile substances in walnut oil at different oxidation degrees
[0150] To further evaluate the characteristic changes of volatile substances in walnut oil at different oxidation stages, supervised OPLS-DA was used to perform multivariate statistical analysis on 298 substances detected by HS-SPME-GC-MS. Figure 4 As shown in Figure A, this model effectively distinguishes the differences between samples with different oxidation degrees in walnut oil and accurately reflects the contribution of the model to the changes in the volatile components of the samples. The model parameter is R 2 X=0.581,R 2 Y=0.966,Q 2 =0.945, R 2 X and R 2 The closer the value of Y is to 1, the higher the degree of explanation of the sample components by the model, and the more reliable the predictability and accuracy of the model. In addition, the model was cross-validated for 200 iterations, as shown in Figure 4 As shown in B, the low intercept R 2 =0.22, Q 2 = -0.497 indicates that there is no overfitting. This is similar to the PCA analysis results of the electronic nose. Based on the identified volatile compounds, the score plot shows that there is a clear separation between the walnut oil samples, indicating that as the degree of oxidation increases, there are differences between the volatile substances of walnut oils with different oxidation degrees.
[0151] The role of VIP value is to screen out the variables that contribute most to the model classification. Specifically, VIP value is used to evaluate the importance of variables in the model. Variables with values greater than 1 are considered to be variables that significantly affect the model classification results, and are therefore often used as important indicators for screening differential metabolites. Variable importance projection can be used to characterize the specific markers of volatiles between different oxidation degrees in walnut oil. The results are shown in Appendix 1. 108 volatile compounds with VIP values greater than 1 were screened, including 19 aldehydes, 15 acids, 12 ketones, 10 esters, 10 alkenes, 9 alcohols, 8 benzenes, 8 alkanes, 8 heterocyclics, 6 amines, 3 nitriles, 1 furan, and 1 pyrazine.
[0152] 4) Thermal clustering analysis of volatile substances in walnut oil at different oxidation degrees
[0153] The above 108 characteristic volatiles with VIP values > 1 were subjected to thermal clustering analysis, and the samples were standardized to obtain Figure 5 As shown in the figure, the size of the average value of each component is represented by different shades of color. The darker the color, the more content, and the lighter the color, the less content. It vividly shows the distribution of each volatile characteristic substance in the sample. The darkest substance can be considered as the characteristic volatile substance with different oxidation degrees in walnut oil. According to the VIP value screening and quantitative results, in the oxidation process of walnut oil, due to the high temperature and high oxygen content environment, complex biochemical reactions such as oxidative decomposition of unsaturated fatty acids and degradation of amino acids occur, which causes dynamic changes in the oxidation process of volatile compounds.
[0154] As the oxidation degree in walnut oil increases, the relative content of hexanal gradually decreases from CK 23.18% to 1.33% at T44, and decreases to 0% at T48; studies have shown that hexanal is a key volatile substance in the early stage of vegetable oil oxidation, which comes from lipid oxidation and amino acid degradation in walnut oil; trimethylsilanol decreases to 0% at T4, and the relative content of 4-ethylbenzamide decreases from CK 23.50% to 0% after the oxidation time reaches T12, and the relative content of acetonitrile gradually decreases from CK 7.33% to 0.26% at T48 during the oxidation process. The above-mentioned volatile organic compounds show a downward trend as the oxidation degree of walnut oil increases, indicating that they can be characteristic markers of walnut oil during the oxidation process. It is worth mentioning that acetonitrile and 4-ethylbenzamide were detected in walnut oil and dominated the initial sample of walnut oil, but they have not been reported in previous studies. The chemical reactions involved in the oxidation process of walnut oil are complex. It is speculated that the substances produced by unknown reactions in walnut oil may be due to high oxygen content and high temperature. The specific source needs further study. On the contrary, with the aggravation of the oxidation degree of walnut oil, the substances that show an increasing trend are (E,E)-2,4-nonadienal, (E,E)-2,4-decadienal, (E,E)-2,4-heptadienal, hexanoic acid, 1-(1-cyclohexen-1-yl)-ethanone, (E)-2-octenal, (E)-2-pentenal, 2-cyclohexene-1-one, 1-penten-3-one, 1H-pyrazole, (E,E)-3,5-octadien-2-one, 2,4-decadienal, 2,2-dimethyl-3-heptanone, (Z)-2-heptenal, (E)-5-(pentyloxy)2-pentene, dimethylhydrazoneformaldehyde, and hexanoic anhydride. It can be seen that the substances that show positively correlated dynamic changes during the oxidation process of walnut oil are mainly aldehydes, indicating that aldehydes are one of the main substances in walnut oil. Some previous studies believe that the generation and change of aldehyde compounds are mainly derived from the Maillard reaction and the Strecker degradation reaction. The second possibility is that unsaturated fatty acids (linoleic acid and linolenic acid) are exposed to high temperature and oxygen during heating, resulting in lipid peroxidation, which in turn produces a large number of aldehydes. Previous studies have shown that a small amount of aldehydes can make the oil present pleasant odors such as floral and fruity aromas, but as the degree of oxidation and rancidity of walnut oil deepens, it will make the walnut oil emit an unacceptable irritating odor. Hexanal, heptanal, octanal, (E)-2-octenal, (E)-2-nonenal, (E)-2-decenal, nonanal, decanal, (E)-2-heptenal, valeraldehyde and other substances have been confirmed to be the main cause of the unpleasant odor of nuts. At T48, walnut oil is viscous and accompanied by a pungent rancid smell, indicating that the increase of aldehydes is one of the reasons for the corruption and deterioration of walnut oil.It is worth noting that in this experiment, acids such as caproic acid, N-benzyloxycarbonyl-dl-norleucine, propionic acid, valeric acid, cyclohexanecarboxylic acid, and 3,4-didehydroproline were also detected. The content of these unsaturated fatty acids is relatively low, ranging from 0.25% to 6.02%, and changes dynamically with the increase in the oxidation time of walnut oil. Studies have shown that the acid substances in walnut oil are due to the oxidative decomposition reactions of numerous aldehydes, and caproic acid has the highest content of 6.02%, which is mainly responsible for the tacky and fishy smell of walnut oil. It is one of the three major products of lipid oxidation and has been confirmed to be one of the characteristic markers of sunflower seeds, which is similar to this experiment. In summary, caproic acid can be used as an oxidation characteristic marker of walnut oil.
[0155] 5) Correlation analysis between olfactometer and GC-MS of walnut oil volatile substances at different oxidation degrees
[0156] like Figure 6 As shown in the figure, a correlation network diagram was constructed based on 108 volatile substances with VIP>1 in the HS-SPME-GC-MS results and 14 sensor response values of the olfactory analyzer. The square blocks in the figure are the sensors of the electronic nose. The size and depth of the blocks represent the importance of the sensors. The size and depth of the dots represent the importance of the volatiles. The thickness of the curve represents the degree of correlation. A correlation network diagram was established based on the Pearson correlation coefficient p<0.05 and |r|>0.8. Figure 6 It is shown that the volatiles with strong correlation with the sensor response are: thiophene, N-dimethylaminomethylene methyl ester, 1-ethyl-3,5-dimethylbenzene, acetonitrile, 5-methyl-2-hexanone, trimethylsilanol, 1-penten-3-ol, trans-3-nonen-2-one, hexanoic acid, 1-butanol, methyl isobutyl ketone, hexanal, (E)-2-octenal, (E)-2-decenal, (E,E)-2,4-decadienal, (E,E)-2,4-heptadienal, (E,E)-1, 3,6-octatriene, propionic acid; combined with the cluster analysis results of walnut oil volatile substances, seven volatile organic compounds including hexanal, acetonitrile, trimethylsilanol, (E,E)-2,4-heptadienal, hexanoic acid, (E)-2-octenal, (E,E)-2,4-decadienal were identified as characteristic oxidation markers; the olfactory analyzer sensors with stronger correlation were: W8, W13, W11, W12, W5, W7, and W2. This result is consistent with the results of the radar chart in the olfactory analyzer sensor. Figure 2It shows a high degree of consistency; it is consistent with the result that the main volatile substances in the HS-SPME-GC-MS analysis are aldehydes. Therefore, the different oxidation degrees of walnut oil can be distinguished based on the response value of the electronic nose sensor to the volatile substances in the sample. In general, the integration of HS-SPME-GC-MS and olfactory analyzer not only produces the macroscopic results of the olfactory analyzer, but also provides a microscopic understanding of the volatile components, providing an effective means for characterizing the characteristic markers of walnut oil at different oxidation periods. The analysis of the network model further confirms the correspondence between the volatile compounds and the corresponding values of the olfactory analyzer, which provides a strong proof for the feasibility of using olfactory analyzer technology to quickly distinguish the oxidation degree of walnut oil.
[0157] 6) Rapid discrimination analysis of walnut oil oxidation degree based on partial least squares regression PLSR modeling
[0158] The partial least squares regression (PLSR) modeling method was used to establish a rapid detection model based on the response values of walnut oil with different oxidation degrees collected by the olfactory analyzer and the peroxide value and acid value of walnut oil. The modeling results are as follows: Figure 7 As shown in the figure, the Rc of the peroxide value is 0.9742, and the Rp is 0.9359; the Rc of the acid value is 0.9474, and the Rp is 0.9149, which intuitively shows the prediction results of the PLSR prediction model for the peroxide value and acid value of walnut oil with different oxidation degrees. The correlation coefficient is greater than 0.90, and the prediction root mean square error is low, indicating that the PLSR model established in this study has good prediction ability for the acid value and peroxide value content of walnut oil with different oxidation degrees.
[0159] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.
Claims
1. A method for analyzing the degree of oxidation of walnut oil based on an olfactory analysis sensor array, characterized in that: The following steps are involved: The walnut oil to be tested is tested by the olfactory analysis characteristic reaction sensor, and whether the walnut oil is oxidized is determined according to the |r| value of the correlation curve; If |r| of the correlation curve is less than 0.6, the judgment result is no, and the acid value and peroxide value of the walnut oil are not predicted; If |r| of the correlation curve is greater than 0.6, the judgment result is yes, and the walnut oil to be tested is subjected to olfactory analysis to obtain the olfactory analysis sensor response value of the walnut oil to be tested, and the response value of the walnut oil oxidation degree is brought into the PLSR prediction model of the walnut oil oxidation degree to obtain the peroxide value prediction value and acid value prediction value of the walnut oil to be tested, which are used to judge the oxidation degree of the walnut oil to be tested.
2. The method for analyzing the degree of oxidation of walnut oil based on an olfactory analysis sensor array according to claim 1, characterized in that: The specific process of obtaining the olfactory analysis characteristic reaction sensor is as follows: Walnut oil with different oxidation degrees was subjected to HS-SPME-GC-MS analysis to obtain the volatile substances of walnut oil with different oxidation degrees. The volatile substances with VIP>1 were retained and subjected to Pearson correlation analysis with the response values of the olfactory analysis sensor. The volatile substances with p<0.05 and |r|>0.8 were retained to obtain characteristic oxidation markers, as well as olfactory analysis characteristic reaction sensors with characteristic reactions to the characteristic oxidation markers.
3. The method for analyzing the degree of oxidation of walnut oil based on an olfactory analysis sensor array according to claim 1, characterized in that: The specific construction process of the walnut oil oxidation degree PLSR prediction model is as follows: Walnut oils with different oxidation degrees were subjected to olfactory analysis and the response values of the olfactory analysis sensors of walnut oils with different oxidation degrees were obtained. The partial least squares regression equation was constructed by combining the response values of the olfactory analysis sensors of walnut oils with different oxidation degrees with the acid value and peroxide value of walnut oils with different oxidation degrees to obtain the PLSR prediction model of walnut oil oxidation degree.
4. The method for analyzing the degree of oxidation of walnut oil based on an olfactory analysis sensor array according to claim 3, characterized in that: The PLSR prediction model for the degree of oxidation of walnut oil is as follows: Where: b i is the regression coefficient corresponding to the i-th eigenvalue collected by the olfactory analysis sensor; i = 1, 2, 3...n; X i is the i-th eigenvalue data collected by the olfactory analysis sensor; Yes i and X i The sum of the products of ; C is the predicted value when all the feature values collected by the olfactory analysis sensor are 0; Y is the predicted value of the oxidation degree of walnut oil.
5. The method for analyzing the degree of oxidation of walnut oil based on an olfactory analysis sensor array according to claim 2, wherein: The characteristic oxidation markers are as follows: Hexanal, acetonitrile, trimethylsilanol, (E,E)-2,4-heptadienal, hexanoic acid, (E)-2-octenal, (E,E)-2,4-decadienal; The characteristic reaction sensors of the olfactory analysis are: W8, W13, W11 and W12; The model of W8 is TGS2618; the model of W11 is TGS2602; the model of W12 is WSP2110; the model of W13 is MQ137.
6. A method for analyzing the degree of oxidation of walnut oil based on an olfactory analysis sensor array according to claim 1, characterized in that: The conditions for the olfactory analysis test are: The injection flow rate is 350-400 mL / min, the detection time is 60-90 s, and the injection time is 30-50 s.
7. A method for analyzing the degree of oxidation of walnut oil based on an olfactory analysis sensor array according to claim 2, characterized in that: The chromatographic conditions for the HS-SPME-GC-MS analysis are as follows: Chromatographic column: DB-WAX (30.0m×250μm, 0.25μm); initial temperature was 40℃ for 4min, then increased to 110℃ at a rate of 4℃ / min, and then increased to 230℃ at a rate of 10℃ / min and then maintained for 10min; vaporization chamber temperature was 250℃; transfer line temperature was 230℃; carrier gas was He; Carrier gas flow rate 1.0mL / min; no split; The mass spectrometry conditions for the HS-SPME-GC-MS analysis are as follows: EI source; electron energy 70eV; ion source temperature 230℃; quadrupole 150℃; scanning mode is Scan; scanning mass range is 20~500u.
8. A method for analyzing the degree of oxidation of walnut oil based on an olfactory analysis sensor array according to claim 2, characterized in that: The specific preparation method of the walnut oil with different oxidation degrees is as follows: In an oven at 100°C, oxygen was introduced at a rate of 3 L / min to accelerate the oxidation of walnut oil. Samples were taken every 4 hours to obtain walnut oils with different degrees of oxidation.
9. A method for analyzing the degree of oxidation of walnut oil based on an olfactory analysis sensor array according to claim 8, characterized in that: Before the walnut oils with different oxidation degrees are subjected to olfactory analysis, the method further comprises balancing the gas in a water bath; the temperature of the water bath is 40 to 60° C.; and the time for balancing the gas is 20 to 30 minutes.
10. The method for analyzing the oxidation degree of walnut oil based on an olfactory analysis sensor array according to claim 3, characterized in that: The acid value of the walnut oil with different oxidation degrees is measured using the national food safety standard GB 5009.229-2016; The peroxide values of the walnut oils with different oxidation degrees are measured using the national food safety standard GB 5009.227-2023.
