CN114460121B - Method for detecting moisture and fat content of poultry meat by using low-field nuclear magnetic resonance technology - Google Patents

Abstract

The invention relates to a method for detecting the moisture and fat content of livestock meat by using a low-field nuclear magnetic resonance technology, which comprises the following steps of (1) sample treatment: pretreating livestock meat by using a manganese chloride solution; (2) detection of a sample; (3) establishing a calibration curve; and (4) calculation of moisture and fat content. The method applies the pretreatment process of the manganese chloride solution to the field of detecting livestock meat by using a low-field nuclear magnetic resonance technology, verifies the feasibility of the pretreatment process for the application in the field of foods, determines the specific use condition of the manganese chloride solution through condition optimization, and ensures the accurate detection of the moisture and fat content of the livestock meat.

Description

Method for detecting moisture and fat content of poultry and livestock meat using low-field nuclear magnetic resonance technology

Technical Field

The invention belongs to the field of food detection, and in particular relates to a method for detecting the moisture and fat content of poultry and livestock meat.

Background Art

With the development of science and technology and the improvement of people's living standards, consumers have put forward higher requirements for food quality and safety, and rapid food safety detection technology has therefore been more widely developed. At present, the rapid detection technologies commonly used in the food field mainly include chemical colorimetry, enzyme-linked immunosorbent assay and near-infrared spectroscopy. Compared with other detection technologies, low-field nuclear magnetic resonance (LF-NMR) technology has the advantages of being economical and environmentally friendly, accurate data, fast and non-destructive, and is therefore widely used in the food field.

Since 1945, the first discovery of the nuclear magnetic resonance phenomenon by American physicists Bloch and Purcell has made the application research of nuclear magnetic resonance technology more and more extensive. It is mainly that specific atomic nuclei in the sample material undergo nuclear transitions under certain effects and generate nuclear magnetic resonance signals to reflect some inherent properties of the sample material. Generally, nuclear magnetic resonance instruments can be divided into two types according to the magnetic field strength: high-field nuclear magnetic resonance and low-field nuclear magnetic resonance. Low-field nuclear magnetic resonance refers to nuclear magnetic resonance with a magnetic field strength below 0.5T, which is mainly used to analyze the physical properties of samples. The main measurement index of LF-NMR is relaxation time, and relaxation refers to the process in which ^1H nuclei change from high energy state to low energy state in a non-radiative way. LF-NMR mainly reflects the motion properties of specific protons ( ^1H ) in the sample by measuring the longitudinal relaxation time _T1 , transverse relaxation time _T2 and diffusion coefficient D. In order to obtain information such as the physical and chemical environment and water flow distribution inside the system to analyze the target components in the sample, the relaxation time T ₁ and T ₂ of the sample can be analyzed, that is, the interaction between the spin, environment and spin. This technology can observe the internal structure of the sample without damage, and measure various indicators such as oil, water and protein in the sample. It is generally not affected by the size and shape of the sample, and the detection effect and accuracy are good.

At present, the standard method for measuring moisture and fat content is mainly the 105℃ drying constant weight method. These traditional water removal methods use physical means to evaporate or sublimate the water in the meat. Although they can achieve the effect of removing moisture, the traditional dehydration method requires a long pre-processing time.

Prior art has applied low-field nuclear magnetic resonance technology to detect moisture and fat content in food. For example, CN105606637A discloses a method for detecting moisture and fat content in abalone using low-field nuclear magnetic resonance technology, which can simultaneously detect moisture and fat content in abalone, is not affected by the surface properties of abalone, and the measurement process does not damage abalone itself. However, the method quantitatively uses the method of establishing a prediction model, and the water-oil signal is not clearly distinguished in the pre-treatment. It is a big data learning performed using statistical means. If it is desired to obtain a more accurate quantitative value, it is necessary to select as many representative samples as possible when establishing the model.

Currently, there is no method in the prior art to accurately detect the water and fat content in poultry and livestock meat simultaneously through low-field nuclear magnetic resonance technology.

Summary of the invention

In order to solve the above technical problems, the first aspect of the present invention provides a method for detecting the moisture and fat content of poultry and livestock meat using low-field nuclear magnetic resonance technology, comprising the following steps:

(1) Sample preparation: Poultry meat is minced and evenly divided into two portions. One portion of the sample with a mass of M _f is weighed, and MnCl ₂ ·4H ₂ O solution is added. The mixture is vortexed and mixed. The mixture is allowed to stand and kept warm to obtain the sample 1 to be tested. The other portion of the sample with a mass of M is weighed, and the MnCl ₂ ·4H ₂ O solution is not added. The mixture is kept warm to obtain the sample 2 to be tested.

(2) Sample determination: After collecting signals from the first sample to be tested, two main relaxation peaks are obtained using a low-field nuclear magnetic resonance analyzer. The second relaxation peak is the fat peak, and the peak area is recorded as A _f ; after collecting signals from the second sample to be tested, a T ₂ spectrum is obtained. After inversion calculation, the total peak area is obtained and recorded as A;

(3) Establishment of calibration curve: MnCl ₂ ·4H ₂ O solution and livestock oil of different masses were selected as standard samples for water and fat quantification, respectively. The NMR signals were collected using the same parameters as those for sample detection to obtain T ₂ spectra. Linear fitting was performed with the abscissa being the water mass or livestock oil mass and the ordinate being the peak area. The obtained linear equation for fat mass was recorded as Y=a _f X+b _f , and the linear equation for water mass was recorded as Y=a _w X+b _w ;

(4) Calculation of water and fat content: The fat content was calculated according to the following formula: F = ( _Af - _bf )/( _af · _Mf ) × 100%, where F is the fat content, _Af is the fat peak area of the poultry and livestock meat sample to be tested, and _Mf is the mass of the sample pre-treated with manganese chloride solution; the water content was calculated according to the following formula: W = [(A/M)-(Af _{/ Mf} ) _-bw ]/ _aw × 100%, where W is the water content, A is the total peak area of fresh meat, and M is the mass of the fresh meat sample.

Preferably, the poultry and livestock meat in step (1) of the method is pork.

Preferably, the mass concentration of the MnCl ₂ ·4H ₂ O solution added in step (1) of the method is 5-50%, more preferably, the mass concentration of the MnCl ₂ ·4H ₂ O solution added is 10-30%, further preferably, the mass concentration of the MnCl ₂ ·4H ₂ O solution added is 20-30%, and most preferably, the mass concentration of the MnCl ₂ ·4H ₂ O solution added is 20%.

Preferably, the volume of the MnCl ₂ ·4H ₂ O solution added in step (1) of the method is 1-3 ml, and more preferably, the volume of the MnCl ₂ ·4H ₂ O solution added is 1.5 ml.

Preferably, the vortex mixing time in step (1) of the method is 5-20 minutes, and more preferably, the vortex mixing time is 10 minutes.

Preferably, in step (1) of the method, the mixture is allowed to stand for more than 30 minutes after vortex mixing, and more preferably, the mixture is allowed to stand for more than 60 minutes after vortex mixing.

Preferably, in step (1) of the method, the insulation temperature of the test sample 1 and the test sample 2 is 50-70°C, and more preferably, the insulation temperature is 50°C.

Preferably, the detection parameters in step (2) of the method are set as follows: the repeated sampling waiting time TW is 2000 ms, the echo time TE is 0.3 ms, and the repeated sampling number NS is 32.

Preferably, the detection parameters in step (2) of the method are set as follows: the receiver bandwidth SW is 200, the analog gain RG1 is 20, the digital gain DRG1 is 3, the preamplifier gain PRG is 2, and the parameter control RFD of the start sampling time is 0.08ms.

Preferably, the livestock oil in step (3) of the method is lard.

Preferably, the mass concentration of the MnCl ₂ ·4H ₂ O solution in step (3) of the method is 0.85%.

Preferably, in step (3) of the method, the different masses of MnCl ₂ ·4H ₂ O solutions are 0.4958 g, 0.9916 g, 1.4884 g, 1.9808 g, 2.4823 g and 2.9867 g, respectively, and the different masses of poultry and livestock meat are 0.0248 g, 0.0398 g, 0.0809 g, 0.1637 g, 0.2492 g, 0.3282 g, 0.4868 g and 0.6413 g, respectively.

Preferably, the linear equation of fat mass obtained in step (3) of the method is Y=34319X+82.457, and the linear equation of water mass is Y=34468X+98.647.

The second aspect of the present invention provides the use of MnCl ₂ ·4H ₂ O solution in a method for detecting the moisture and fat content of poultry and livestock meat using low-field nuclear magnetic resonance technology.

The beneficial effects of the present invention are as follows: the method for detecting the moisture and fat content in food by using low-field nuclear magnetic resonance technology reported in the prior art lacks a pre-treatment process, and cannot effectively distinguish the moisture and fat detection signals, resulting in inaccurate detection results. The present invention applies the pre-treatment process of manganese chloride solution to the field of detecting poultry and livestock meat by using low-field nuclear magnetic resonance technology for the first time, verifies the feasibility of the pre-treatment process for application in the food field, and at the same time determines the specific use conditions of the manganese chloride solution through condition optimization, thereby ensuring the accurate detection of the moisture and fat content in poultry and livestock meat.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 shows the effect of manganese chloride solution on the multi-component relaxation spectrum of lard;

FIG2 shows the effect of manganese chloride solution on the multi-component relaxation spectrum of pork;

FIG3 shows the effect of repeated sampling waiting time on signal stability;

FIG4 shows the effect of echo time on relaxation spectra;

FIG5 shows the effect of echo time on the degree of signal attenuation;

FIG6 shows the effect of the number of repeated scans on the signal-to-noise ratio;

FIG7 shows the effect of the number of repeated scans on the relaxation spectrum;

FIG8 shows the correlation between the number of repeated scans and A ₂₂ ;

FIG9 shows the effect of the mass concentration of manganese chloride solution on the relaxation spectrum;

FIG10 shows the effect of solution volume on _T2 spectrum;

FIG11 shows the effect of mixing time on the _T2 spectrum;

FIG12 shows the effect of rest time on T2 spectrum;

FIG13 shows T2 spectra of lard of different qualities;

Figure 14 shows the standard curves of lard of different qualities;

FIG15 shows the calibration curves of manganese chloride solutions with different mass concentrations;

FIG16 shows T2 spectra of manganese chloride solutions of different masses (mass concentration is 0.85%);

FIG17 shows the standard curves of different mass manganese chloride solutions (mass concentration is 0.85%);

FIG. 18 shows a comparison of fat content between the low-field NMR method and the Soxhlet extraction method.

DETAILED DESCRIPTION

Test Example 1: Detecting the moisture and fat content of poultry and livestock meat using low-field nuclear magnetic resonance technology

1. Test methods

1.1 Determination of moisture and fat content

According to GB 5009.6-2016 Determination of fat in food and GB 5009.3-2016 Determination of water in food.

1.2. Low-field NMR signal acquisition

1.2.1. Sample processing method

Take the evenly stirred minced pork, weigh 2.0g ( _Mf ) and put it in the NMR sample bottle, add a certain mass concentration and a certain volume of _MnCl2 · _4H2O solution, mix it and let it stand at a constant temperature before signal collection. The same sample cannot be collected continuously. After each signal collection, put it back into the dry thermostat to keep the temperature constant before the next signal collection.

Take the evenly stirred minced pork, weigh 2.0 g (M) and put it into the NMR sample bottle, without adding MnCl ₂ ·4H ₂ O solution, and directly perform signal acquisition after constant temperature.

1.2.2 Signal Acquisition Method

According to the instrument operating instructions, the instrument was first calibrated using the free induction decay (FID) sequence. Certain detection parameters were set under the CPMG sequence to collect the transverse (T ₂ ) relaxation signal of the sample. Three replicates were taken for each sample and each sample was scanned twice. The relaxation data were expressed as mean ± standard deviation.

1.3. Determination of instrument parameters

The parameters of instrument detection are mainly composed of two parts, namely system parameters and sampling parameters. The system parameters are determined by the instrument characteristics, environment and corresponding pulse sequence. The instrument will automatically adjust the parameters under the FID sequence. Including center frequency, 90° pulse and 180° pulse. The sampling parameters are determined by the instrument characteristics and research purposes, mainly including: number of repeated scans (NS), repeated sampling waiting time (TR), echo time (TE) and echo count (NECH). TR has no significant effect, while NECH needs to ensure that the sample signal is completely attenuated. Therefore, this experiment mainly examines NS and TE.

1.3.1 Waiting time (TW)

The waiting time for repeated sampling is the time from the end of the previous sampling to the beginning of the next sampling. If TW is set too short, the signal amplitude will decrease during the next sampling. Generally, TW can be set to be greater than 5 times T1 to ensure that the signal of the sample is restored by at least 98%. This test sets the waiting time to 1000ms, 1500ms, 2000ms, 2500ms, 3000ms, 3500ms and 4000ms, and compares the maximum value of the signal modulus and makes it relatively stable as the best condition.

1.3.2 Echo Time (TE)

After the RF pulse, the time interval from the initial generation of transverse magnetization intensity to the reception of the signal is called the echo time. The half echo time is generally required to be greater than 6 times the 90° pulse width setting value. In this experiment, TE was set to 0.1, 0.15, 0.2, 0.25, 0.3 and 0.35ms respectively, and the attenuation degree and relaxation information of the curve were compared. The final attenuation signal amplitude was the average of the last ten signal amplitudes.

1.3.3. Number of repeated scans (NS)

The number of repeated scans is the number of times the instrument performs repeated sampling. The number of NS settings will affect the signal strength and sampling time of the sample, so it should be set according to the actual signal strength of the sample. In this experiment, it was set to 16, 32, 64 and 128 times respectively. After sampling, the relaxation characteristics and signal-to-noise ratio of the _T2 spectrum were compared and analyzed. The signal-to-noise ratio is the average of the first point signal/the last point 10 signals.

1.4. Determination of sample pretreatment conditions

This experiment mainly investigated the effects of mass concentration of manganese chloride solution, added volume, mixing time, standing time and sample temperature on the fat peak area (A ₂₂ ).

1.4.1. Mass concentration of manganese chloride solution

1 ml of manganese chloride solution with different mass concentrations (5%, 10%, 15%, 20%, 30%, 40%, 50% and 60%) was added to the samples respectively. After the samples were allowed to stand at 50°C, signals were collected in the NMR sample tube to investigate the changes in signals of different samples.

1.4.2. Volume of manganese chloride solution

After adding different volumes (0.5 ml, 1.0 ml, 1.5 ml, 2.0 ml and 3.0 ml) of manganese chloride solution with a mass concentration of 20%, the signals were collected to observe the changes in the signals of different samples.

1.4.3 Mixing time

Add 1 ml of 20% manganese chloride solution to 2 g of minced meat, vortex mix for different times (3 min, 5 min, 10 min, 15 min and 20 min), and then let it stand at 50°C for 30 min. After collecting the signals, observe the changes in the signals of different samples.

1.4.4. Standing time

After weighing 2g of minced meat, add 1ml of 20% manganese chloride solution, vortex mix for 10min, and then stand at 50℃ for different time periods (10min, 30min, 60min, 120min, 180min and 240min). After collecting the signals, the changes in the signals of different samples were observed.

1.4.5. Sample temperature

1 ml of 20% manganese chloride solution was added to the sample, and the sample was left to stand at different temperatures (32°C, 40°C, 50°C, 60°C and 70°C) for 30 min. After collecting the signals, the changes in the signals of different samples were observed.

1.5. Determination of the concentration of water quantitative standard samples

Weigh 1g of manganese chloride solution of different concentrations (0.5%, 1.0%, 3.0%, 5.0%, 8.0% and 10.0%), establish a calibration curve of water peak area and solution concentration, and obtain equation E1. According to the change degree of peak area when the sample reduces 1g of water, substitute it into E1 to calculate the solution concentration with consistent hydrogen proton density.

1.6. Drawing of standard curve

Drawing of fat quantitative standard curve: Weigh different masses of lard, and establish a linear equation by least square regression with mass as the abscissa and fat peak area (A ₂₂ ) as the ordinate, as follows:

Y＝a _f X+b _f (1).

Drawing of the moisture quantitative standard curve: Weigh different masses of 0.85% manganese chloride solution, and establish a linear equation by least squares regression with the mass of water in the solution as the horizontal and vertical coordinates and the water peak area (A ₂₁ ) as the vertical coordinate, as follows:

Y＝a _w X+b _w (2)

1.7 Calculation of fat and water content

The fat content is calculated according to the following formula: F = (A _f - b _f )/(a _f ·M _f )×100%

Where F is the fat content, _Af is the fat peak area of the pork sample to be tested, and _Mf is the mass of the sample pre-treated with manganese chloride solution.

The moisture content is calculated according to the following formula: W = [(A/M) - (A _f /M _f ) - b _w ] / a _w × 100%

Where W is the moisture content, A is the total peak area of fresh meat, and M is the mass of fresh meat sample.

1.8 Data Analysis

SPSS software was used to perform significant variance analysis on the relaxation data under different treatment conditions, and origin software was used to draw the graph. The _T2 spectrum was obtained by multi-component inversion using the analysis software provided by the low-field nuclear magnetic resonance instrument, and the data were expressed as mean ± standard deviation.

2. Test results

2.1 Effect of manganese chloride solution pretreatment on the multi-component relaxation spectra of lard and pork

The present invention firstly uses a low-field nuclear magnetic resonance instrument to investigate the possible influence of using a manganese chloride solution for pre-treatment of poultry and livestock meat.

Figure 1 shows the changes in _T2 spectra before and after adding 1ml of 40% manganese chloride solution to lard of different masses. From the figure, two completely separated relaxation peaks can be observed, and the distribution ranges of _T21 and _T22 are 0.01-0.977ms and 16.832-622.257ms respectively. The _T21 peak in the sample with manganese chloride solution shows the hydrogen protons in the solution, while the T21 peak in the pure oil sample may be a "false positive" caused by the system error of the instrument due to the small volume. _{The T22 peak} is a relaxation peak generated by the fat in the sample, so only the changes in the fat main peak _T22 are compared in the subsequent studies. It is obvious that the distribution curves of lard of different masses before and after adding manganese chloride solution are almost completely consistent, and there is no significant difference in _A22 (peak area), indicating that manganese chloride solution does not affect the relaxation characteristics of lard. Manganese ions are paramagnetic ions, soluble in water but not in oil, so they only affect the hydrogen ions in water, and do not affect the signals of hydrogen ions in fat or oil.

As can be observed from FIG. 2 , the T ₂ spectrum of pork is also composed of two peaks. The T ₂₁ peak with a shorter relaxation time in the figure represents bound water, and the T ₂₂ peak is produced by the water and fat in the meat, and it is difficult to distinguish the two from the curve. Therefore, manganese ions are added to pork, and they are infiltrated into the water of the sample through ion migration. After adding manganese chloride solution, the present invention obviously observes that the T ₂₂ peak area is greatly reduced, while the T ₂₁ peak becomes larger. On the one hand, this is because the water in the sample is affected by manganese ions, and the relaxation time moves to the left. On the other hand, it is the hydrogen signal brought by the added manganese chloride solution. There is no significant difference between the T ₂₂ of the processed pork sample and lard. At this time, it is believed that the T ₂₂ peak is only produced by hydrogen protons in fat. Many studies have also shown that the relaxation time of lipids is generally about 100ms. This is because the hydrogen in lipids has a more complex structure than water, so the fluidity is limited.

2.2. Test results of instrument detection parameter optimization

2.2.1 Waiting time

As shown in Figure 3, the maximum value of the modulus of the sample at each waiting time remains stable, and the variation range of the two maximum values of the modulus is less than 1%, indicating that the samples can return to equilibrium within the selected waiting time. The waiting time of different samples is different, which is related to the speed at which they release energy. Considering that the fat content of different samples is different, their _T2 is also different (see Table 1), so setting TW to 2000ms can meet the testing needs of all livestock and poultry meat samples.

Table 1 Effect of fat content on relaxation time

Fat content (%) Relaxation time T ₂ 0.818±0.007 7.42±2.22 1.365±0.066 46.36±8.9 3.428±0.1102 94.17±0.96 5.936±0.245 128.33±1.9 6.450±0.121 126.25±2.93 23.052±0.334 171.3±3.33

2.2.2 Echo time

Figure 4 shows the _T2 spectra at different echo times. It can be clearly observed that the echo time causes the _T21 peak to move to the right and the peak area to decrease significantly. This is because the self-diffusion of water is stronger than that of oil, and the change in echo time will affect the effect of molecular self-diffusion and molecular exchange on relaxation characteristics. The longer the echo time, the stronger the molecular self-diffusion, which leads to the extension of the transverse relaxation time _T2 . Within 0.35ms, the change in echo time does not have a significant effect on the fat peak area and relaxation time. At the same time, the echo time will affect the attenuation degree of the CPMG curve and the distribution range of the relaxation time.

Figure 5 shows the attenuation degree of different echo times. When the echo time is 0.1ms, the signal amplitude of the sample decays to 257.42. As TE increases, the _T2 relaxation interval gradually increases. At the same time, the final signal amplitude of the attenuation curve gradually decreases to 57.31. After the echo time is 0.3ms, the final signal attenuation amplitude no longer changes significantly. It is believed that the signal has decayed completely at this time, and the stability of the peak area is also good, with RSD less than 5%. Therefore, the echo time is selected to be 0.3ms.

2.2.3. Repeat scan times

The number of repeated scans has a direct impact on the signal-to-noise ratio of the sample. The higher the signal-to-noise ratio, the more repeated scans are required. In Figure 6, when the number of repeated scans increases, the signal-to-noise ratio also increases, from the initial 184.82 to 290.44. Of course, the time spent on acquiring the signal also doubles. There is no significant difference in the signal-to-noise ratio when scanning 32 times and 64 times. Although scanning 128 times significantly improves the signal-to-noise ratio, it takes longer time, thereby reducing the efficiency of detection. As can be seen from Figure 7, the T ₂ spectra obtained by different scanning times all show two peaks. From Figure 8, it can be found that NS and A ₂₂ have a good linear relationship (Y = 64.41X-5.0905), and R ² reaches 0.9999. Studies have shown that the increase in the number of repeated scans may cause the magnet to heat up and cause a slight change in the sample temperature, which can be reflected from the right shift of the T ₂₂ peak. Therefore, it is considered that it is sufficient to repeat the scan 32 times, at which point the relaxation characteristics of the samples show good stability (RSD is less than 5%).

2.3. Optimization test of pretreatment conditions of manganese chloride solution

2.3.1. Mass concentration of manganese chloride solution

Figure 9 is the _T2 spectrogram of pork after adding different mass concentrations of manganese chloride solution. As the concentration increases, the _T21 peak caused by the water content in the sample moves to the left. When the concentration reaches about 10%, the relaxation time is basically stable, and only the peak area gradually decreases. When the manganese ion content increases, the peak area of water will decrease, because manganese ions accelerate the relaxation of hydrogen in water, so most of the hydrogen proton signals are lost. This may be because when the manganese ion concentration is too low, there is no manganese ion that can dissolve in the water in the meat. At the same time, the _T22 peak also moves significantly to the left, until the mass concentration is 20%, the _T22 peak shape is similar to lard, and _A22 no longer changes significantly. As the solution concentration increases, the RSD also gradually increases. When the concentration is 50%, the RSD is as high as 14.52%, and when the concentration is 20%, the RSD is only 2.38% (see Table 2), which can meet the requirements. Based on the above analysis, it is most appropriate to set the concentration of manganese chloride solution at 20%.

Table 2 Effect of mass concentration of manganese chloride solution on fat peak area (A ₂₂ )

Mass concentration A ₂₂ RSD 1% 292.84 ^c ±78.52 26.81% 5% 437.03 ^b ±11.32 2.59% 10% 433.51 ^b ±3.83 0.88% 20% 472.38 ^a ±11.24 2.38% 30% 488.24 ^a ±14.67 3.01% 40% 480.95 ^a ±30.67 6.38% 50% 465.41 ^ab ±67.59 14.52%

Note: The same letters (a, b, c) in the upper right corner of the _A22 data indicate that the peak area difference is not significant (p>0.05), while different letters indicate that the peak area difference is significant (p<0.05). The meaning of the letters in the upper right corner of the data in Tables 3-6 below is the same as in Table 2.

2.3.2. Volume of manganese chloride solution added

Table 3 shows the change of relaxation characteristics of fat in the sample when different volumes of manganese chloride solution are added. When the added volume is 1.5-3.0 ml, the fat peak area is significantly higher than that of the treatment group with 0.5-1.0 ml, and as the volume of the solution increases, the relative standard deviation of the sample gradually decreases. When 0.5 ml is added, the sample stability is the worst, with an RSD of 5.05%. It can also be observed from _T2 spectrum 10 that the distribution of relaxation peaks is basically stable when the added volume is greater than 1.5 ml, and the smaller the _T21 peak is when the added volume is larger. Adding a larger volume of solution means that more manganese ions are added to the sample, thus affecting more water hydrogen protons. When the solution volume is too small, the _T21 peak may account for a large proportion, resulting in a greater deviation in the result of _T22 peak inversion. Therefore, the added volume is selected to be 1.5 ml.

Table 3 Effect of manganese chloride solution volume on fat peak area

Solution volume A ₂₂ RSD 0.5ml 404.3 ^a ±20.41 5.05% 1ml 413.15 ^a ±14.68 3.55% 1.5ml 442.93 ^b ±13.27 3.00% 2ml 439.93 ^b ±10.09 2.29% 3ml 440.26 ^b ±9.58 2.18%

2.3.3 Mixing time

In order to ensure that the added manganese chloride solution can fully contact the sample, different mixing times are compared. The results are shown in Table 4. The A ₂₂ of the sample mixed for 3 minutes is significantly lower than that of other mixing times. After adding manganese chloride, it needs to fully combine with the water in the meat to achieve the purpose of water-oil signal separation. When the mixing time is too short, the manganese chloride solution may not be fully mixed with the minced meat, thereby affecting the magnetic field of the hydrogen ions in the lipids. When the mixing time is higher than 5 minutes, no significant changes will occur. At the same time, observing the T ₂ spectrum (Figure 11), it can be found that there is no significant difference in T ₂₂ with different mixing times, and the T ₂₂ peak distribution between the treatment groups is almost consistent, and the T ₂₂ peak area RSD is less than 5%. However, in order to ensure that the sample can be fully mixed, it is considered that the optimal mixing time is 10 minutes.

Table 4 Effect of mixing time on fat peak area

Mixing time A ₂₂ RSD 3min 415.77 ^a ±10.54 2.54% 5min 425.83 ^b ±12.83 3.01% 10min 427.78 ^b ±12.13 2.84% 15min 432.08 ^b ±15.56 3.60% 20min 435.97 ^b ±8.40 1.93%

2.3.4. Standing time

The sample needs to be left standing for a period of time to ensure complete penetration of manganese ions and to allow the sample to reach a constant temperature. By observing the _T2 spectrum (Figure 12), and comparing different standing times, it is found that there is almost no effect on the _T22 peak shape and relaxation time, and the fat peak area will not change significantly with the change of standing time. The RSD at different times is less than 5%. Therefore, in order to ensure that the sample is constant and the temperature is consistent, it is recommended to ensure that the sample is left standing for at least 30 minutes to allow the sample to reach a constant temperature.

Table 5 Effect of standing time on fat peak area

Standing time A ₂₂ RSD 10min 462.29 ^a ±14.89 3.22% 30min 468.51 ^a ±10.51 2.24% 60min 469.47 ^a ±7.5 1.60% 120min 473.24 ^a ±18.06 3.82% 180min 474.72 ^a ±14.62 3.08% 240min 473.98 ^a ±6.98 1.47%

2.3.5. Sample temperature

The differences in relaxation time (T ₂₂ ) and peak area (A ₂₂ ) under different temperature conditions were analyzed. It can be found from Table 6 that the relaxation time increases with the increase of temperature. This is because the binding force of hydrogen protons is weakened, which increases the fluidity of oil molecules. The peak area reaches a peak value at a temperature of 40°C, and the peak area is significantly higher than that of other samples, but the peak area ratio (P ₂₂ ) shows a certain downward trend. When the temperature is higher than 50°C, the peak area no longer changes significantly. Since the melting point of animal fat is 23-48°C, when the temperature is lower than 50°C, the existence state of lipids may have both solid fat crystals and liquid components. Therefore, in order to ensure that all the fat in the sample is liquid, the sample temperature is determined to be 50°C. At this time, the relative standard deviation of the sample fat peak area is 3.52%, indicating that the stability of the sample under this condition is good.

Table 6 Effect of sample temperature on relaxation characteristics

Sample temperature T ₂₂ A ₂₂ P ₂₂ 32℃ 69.13±12.1 474.46 ^a ±18.16 9.95±0.95 40℃ 88.49±13.39 534.33 ^b ±31.73 9.59±0.61 50℃ 106.12±14.58 483.39 ^a ±17.03 8.54±1.15 60℃ 137.45±12.49 485.23 ^a ±25.83 6.8±0.55 70℃ 142.33±40.64 489.72 ^a ±37.09 5.69±0.33

2.4. Drawing of standard curve

Figure 13 is the result of signal acquisition of standard samples prepared from lard of different qualities, where the T ₂₂ peak is the fat peak. As the quality of lard increases, the T ₂₂ peak also increases. It can be seen intuitively from Figure 14 that within the range of lard weight of 0.0248-0.6413g, A ₂₂ increases linearly with the increase of weight. The fitted regression equation is Y=34319X+82.457, and the correlation coefficient R ² is 1.000, indicating that there is a very significant correlation between the two. According to the peak area reading, the corresponding fat weight in the sample can be calculated, and thus the fat content can be calculated.

Figure 15 is the calibration curve of manganese chloride solution with different mass concentrations. It can be found that the solution concentration is negatively correlated with the peak area, and the two are exponentially correlated. The correlation coefficient ^R2 is 0.9505, and the fitting curve is y=43.496x ^-0.82 . According to the previous experiment, the change degree of the peak area when the sample reduces 1g of water, and the peak area is consistent when the solution mass concentration is 0.85%. Therefore, a standard curve is established with different masses of manganese chloride solution as the x-axis and _A21 as the y-axis, as shown in Figure 15. Figure 16 is the _T2 spectrum of manganese chloride solution with different masses. It is obvious that with the increase of solution mass, the _T21 peak increases accordingly, and similar to fat, the water mass in the solution has a very high correlation with _A21 . The regression equation obtained by fitting is Y=34468X+98.647, and ^R2 is 1.000 (Figure 17). It shows that the water content in the sample can be calculated according to this standard curve.

2.5. Quantitative detection test of water and fat in pork

The method of the present invention is compared with the water content and fat determination method in the national standard (GB 5009.6-2016 Determination of fat in food and GB 5009.3-2016 Determination of water content in food), wherein FIG18 is the linear relationship of fat content determined by different methods. Both methods have good linear correlation, and the correlation coefficient is 0.99, indicating that the fat content can be quantified by using lard as a standard sample to establish a calibration curve. Table 7 compares the water content measured by the traditional drying method and the low-field nuclear magnetic resonance method. The results show that the relative error between the two can reach less than 5%, indicating that the method can achieve the effect of simultaneously determining water and oil, and compared with the traditional determination method, the low-field nuclear magnetic resonance method has certain advantages, and the pre-treatment operation is simple and fast, which is very suitable as a technical means for rapid detection.

Table 7 Comparison of moisture content between low field NMR method and direct drying method

Claims (7)

1. a method utilizing low-field nuclear magnetic resonance technology to detect poultry meat moisture and fat content, is characterized in that, comprises the following steps: (1) Sample preparation: Stir poultry and livestock meat into minced meat, divide it evenly into two parts, first weigh one part of the sample with a mass of _Mf , add MnCl ₂ 4H ₂ O solution, vortex mix, stand still, and keep warm , to obtain sample 1 to be tested; then weigh another sample with a mass of M, without adding MnCl ₂ 4H ₂ O solution, and directly keep it warm to obtain sample 2 to be tested; (2) Determination of samples: Use low-field nuclear magnetic resonance analysis instrument to obtain two main relaxation peaks after collecting the signal of the sample to be tested, the second relaxation peak is the fat peak, and the peak area is recorded as A _f ; the sample to be tested 2. Obtain the _T2 spectrum after collecting the signal, and record the total peak area as A after inversion calculation; (3) Establishment of the calibration curve: Select different masses of MnCl ₂ 4H ₂ O solution and poultry oil as standard samples for water and fat quantification respectively, use the same parameters as the sample detection to collect NMR signals, and obtain the T ₂ spectrum ; Take the abscissa as the water quality or poultry oil quality, and the ordinate as the peak area to carry out linear fitting, the obtained fat linear equation is marked as Y=a _f X +b _f , and the water linear equation is marked as Y=a _w X + b _w ; (4) Calculation of water and fat content: the fat content is calculated according to the following formula: F=(A _f -b _f )/(a _f M _f )×100%, where F is the fat content and A _f is the poultry to be tested Fat peak area of meat samples, M _f is the mass of samples pretreated by manganese chloride solution; water content is calculated according to the following formula: W=[(A/M)-(A _f /M _f )-b _w ]/a _w × 100%, where W is the water content, A is the total peak area of fresh meat, and M is the mass of the fresh meat sample; The mass concentration of the MnCl ₂ ·4H ₂ O solution added in the step (1) is 10-30%; The detection parameters in the step (2) are set as follows: the resampling waiting time TW is 2000ms, the echo time TE is 0.3ms, and the resampling times NS is 32. 2. The method according to claim 1, characterized in that the poultry meat in step (1) is pork. 3. The method according to claim 1, characterized in that the mass concentration of the MnCl ₂ ·4H ₂ O solution added in step (1) is 20-30%. 4. The method according to claim 1, characterized in that the volume of the MnCl ₂ ·4H ₂ O solution added in step (1) is 1-3 ml. 5. The method according to claim 1, characterized in that the vortex mixing time in step (1) is 5-20 minutes. 6. The method according to claim 1, characterized in that in step (1), vortex mixing and then standing for more than 30 minutes. 7. The method according to claim 1, characterized in that, in step (1), the holding temperature of the first sample to be tested and the second sample to be tested is 50-70°C.

Images