Low-field nuclear magnetic resonance is differentiated the method for edible oil quality
Technical field
The invention belongs to low-field nuclear magnetic resonance oils attributional analysis field, be specifically related to the method that a kind of low-field nuclear magnetic resonance is differentiated the edible oil quality.
Background technology
The food that a certain amount of edible oil of feed that people every day is all direct or indirect or edible oil were handled.The quality of edible oil is directly related with health with human diseases.In recent years, all parts of the country the lawless person occurs and collect food and beverage waste oil and grease or poor oil (waste oil), and processing is sold eating house, grease workshop with knockout price after handling, and mixes in the edible oil and reaps staggering profits.The waste oil complicated component contains multiple poisonous and harmful element.
Though the multinomial physical and chemical index of waste oil all is different from normal edible oil and fat,, be difficult to distinguish with oil with common edible from outward appearance and sense organ through " waste oil " of refining.The waste oil detection method has much at present, but has the defective (separate sources " waste oil " characteristics difference) that can not detect all sources " waste oil " with a kind of detection method.And it is big to detect error, and the possibility of erroneous judgement is arranged.If " waste oil " mixes content more after a little while, then can't detect.Therefore, in order to ensure edible oil quality safety, press for a kind of waste oil of can differentiating fast and mix the reliable detection method of pseudo-edible oil.
Nowadays, catering industry consumes a large amount of edible oils every day, but the quality of oil never has better approach and means are tested.Also emerged in large numbers various low grade oilses on the market, wherein waste oil just belongs to wherein a kind of.
The low-field nuclear magnetic resonance analytical technology can be distinguished the difference of certified products edible oil and waste oil fast and accurately by relaxation time spectrum data analysis.Survey pseudo-measuring method with traditional grease and compare, nuclear magnetic resonance method can keep the integrality of sample, is a kind of nondestructive detection means; Method of operating is simply quick, measures accurately, and repeatability is high; Sample need not to add solvent, and quantitative measurement need not standard specimen; Measurement result is not subjected to the influence of sample of material size and appearance luster, and operator's technology and decision influence are less.Therefore, nuclear magnetic resonance technique is that a kind of very potential grease is surveyed pseudo-new technology fast.
The present invention is fully recognized that importance and the necessity that the edible oil quality is differentiated, and recognized the huge advantage that NMR (Nuclear Magnetic Resonance) relaxation time series analysis method is differentiated in the oils quality, adopt the relaxation time spectrum data to differentiate that oil quality can not only improve detection speed and accuracy greatly, also making on-the-spot oil quality identify simultaneously becomes possibility.
Summary of the invention
The present invention seeks to: provide a kind of low-field nuclear magnetic resonance to differentiate the method for edible oil quality, this method not only can be suitable for the quality measurement of various edible oils, and test speed is fast, accuracy is high, good stability, be a kind of can really applying in the middle of the actual edible oils quality discriminating work, effective waste oil is mixed pseudo-edible oil quality and is differentiated detection method.
Technical scheme of the present invention is: a kind of low-field nuclear magnetic resonance is differentiated the method for edible oil quality, it is characterized in that at first under identical parameters, carrying out the NMR signal collection at the oils sample of different qualities, utilize T2 spectrum and T2 and coefficient of diffusion correlativity spectrum that known standard specimen is analyzed, find out the nuclear magnetic resonance test data difference of different quality oils samples, and same quality oils sample is carried out the reperformance test checking; According to the result who analyzes test data and oils quality are carried out correspondingly then, determine the correlativity of oils quality and nuclear magnetic resonance data; Then, set up oils attribute test database, and form the analytical test standard according to the test data of different quality oils; At last, by the test data in the database unknown oils sample being carried out quality identifies.
The implementation step that the inventive method is concrete is as follows:
1) sample test analysis: be listed as by cpmg sequence certified products oil, poor oil are carried out the transverse relaxation analysis, and adopt the anti-Laplce's algorithm of one dimension to compose inversion algorithm as T2, draw the relaxation time spectrum of certified products oil and poor oil, if the relaxation time spectrum of the two is not exclusively overlapping, then the miscella to them carries out the transverse relaxation analysis, and draws relaxation time spectrum; If certified products oil still exists in miscella with the poor oil non-overlapping portions, and proportional with poor oil content in the miscella, then select this part as oils quality judging standard; If certified products oil exists in miscella with the poor oil non-overlapping portions but is disproportionate with miscella, then certified products oil, poor oil and their miscella three are carried out T2 and coefficient of diffusion correlation analysis by the PGSE-CPMG sequence, and adopt two-dimentional anti-Laplce's algorithm as T2 and coefficient of diffusion correlativity spectrum inversion algorithm, draw the T2 of certified products oil, liquid inferior and miscella thereof and coefficient of diffusion correlativity spectrum, find in this spectrum fully independently certified products oil or poor oil composition as oils quality judging standard;
2) standard specimen reperformance test: above as the described test analysis of step 1) by the certified products oil of same quality or poor oil sample being carried out twice, with twice or the error of test data repeatedly before and after understanding, thereby understand the probability of miscarriage of justice of low-field nuclear magnetic resonance method in the edible oil quality is identified;
3) the sample data storehouse is set up: after oils quality judging standard is established, carry out corresponding with the oils quality test data according to analysis result, determine the correlativity of oils quality and nuclear magnetic resonance data, then, set up oils attribute test database according to the nuclear magnetic resonance test data of different quality oils samples, with the formation of convenient edible oil quality measurement and industry testing standard;
4) sample measurement: utilize the test data of the database the inside of step 3) foundation that unknown oils sample is carried out the quality evaluation.
Can carry out temperature variation testing to certified products oil, poor oil and miscella thereof simultaneously in the described step 1) of the inventive method, described temperature variation testing is meant the oils sample is heated up or cooling test, to draw the influence of temperature to different oils qualities, and find out correlativity, thereby determine the correlativity of oils quality and temperature.
The described step 3) of the inventive method specifically is according to determined oils quality judging standard, the ratio that accounts for the bulk sample signal with this part signal in the miscella is Y, certified products oil in the miscella or poor oil content are X, carry out the once linear match and obtain fitting a straight line y=kX+b, then y is the measured value of Y; According to the mode analytic sample of standard specimen signal measurement, find the sample signal y of standard specimen signal area correspondence; Utilize certified products oil or poor oil content X in the y value calculation sample that obtains in calibration line and the sample analysis, thereby find out the difference of different quality oils samples and set up oils attribute test database with this.
The cpmg sequence row that relate in the inventive method are to be used for the conventional nuclear-magnetism pulse train that transverse relaxation is analyzed; The PGSE-CPMG sequence then is to be used for transverse relaxation and the conventional nuclear-magnetism pulse train that spreads correlation analysis; The anti-Laplce's algorithm of one dimension that relates among the present invention simultaneously is the conventional algorithm that is used for the relaxation time spectrum inverting; And two-dimentional anti-Laplce's algorithm is the conventional algorithm that is used for T2 and the inverting of coefficient of diffusion correlativity spectrum.
Advantage of the present invention is:
1. the inventive method can guarantee fast and accurately edible oil sample is carried out the nuclear magnetic resonance spectroscopy test, so that understand the quality of edible oil sample, thereby differentiates low grade oils fast and accurately, for people's diet provides quality assurance.Compare with traditional method, the inventive method has not only improved detection speed, and accuracy that detects and also raising greatly of stability.
2, the inventive method is fast for the detection speed of edible oil sample quality, weak point consuming time, and whole test can be finished with interior at 4 minutes substantially.
Description of drawings
Fig. 1 is the relaxation time spectrum of pure soy bean sample;
Fig. 2 is the soybean oil sample relaxation time spectrum behind the adding 2% frying oil sample;
Fig. 3 is the soybean oil sample relaxation time spectrum behind the adding 5% frying oil sample;
Fig. 4 is the soybean oil sample relaxation time spectrum behind the adding 10% frying oil sample;
Fig. 5 is the soybean oil sample relaxation time spectrum behind the adding 20% frying oil sample;
Fig. 6 is the soybean oil sample relaxation time spectrum behind the adding 50% frying oil sample;
Fig. 7 is the soybean oil sample relaxation time spectrum behind the adding 100% frying oil sample;
Fig. 8 is old oil content of frying and small peak ratio correlation figure.
Embodiment
Embodiment: the invention discloses the method that a kind of low-field nuclear magnetic resonance is differentiated the edible oil quality, its concrete implementation step is as follows:
1) sample test analysis: be listed as by cpmg sequence certified products oil, poor oil are carried out the transverse relaxation analysis, and adopt the anti-Laplce's algorithm of one dimension to compose inversion algorithm as T2, draw the relaxation time spectrum of certified products oil and poor oil, if the relaxation time spectrum of the two is not exclusively overlapping, then the miscella to them carries out the transverse relaxation analysis, and draws relaxation time spectrum; If certified products oil still exists in miscella with the poor oil non-overlapping portions, and proportional with poor oil content in the miscella, then select this part as oils quality judging standard; If certified products oil exists in miscella with the poor oil non-overlapping portions but is disproportionate with miscella, then certified products oil, poor oil and their miscella three are carried out T2 and coefficient of diffusion correlation analysis by the PGSE-CPMG sequence, and adopt two-dimentional anti-Laplce's algorithm as T2 and coefficient of diffusion correlativity spectrum inversion algorithm, draw the T2 of certified products oil, liquid inferior and miscella thereof and coefficient of diffusion correlativity spectrum, find in this spectrum fully independently certified products oil or poor oil composition as oils quality judging standard;
2) standard specimen reperformance test: above as the described test analysis of step 1) by the certified products oil of same quality or poor oil sample being carried out twice, with twice or the error of test data repeatedly before and after understanding, thereby understand the probability of miscarriage of justice of low-field nuclear magnetic resonance method in the edible oil quality is identified;
3) the sample data storehouse is set up: after oils quality judging standard is established, carry out corresponding with the oils quality test data according to analysis result, determine the correlativity of oils quality and nuclear magnetic resonance data, then, set up oils attribute test database according to the nuclear magnetic resonance test data of different quality oils samples, with the formation of convenient edible oil quality measurement and industry testing standard;
4) sample measurement: utilize the test data of the database the inside of step 3) foundation that unknown oils sample is carried out the quality evaluation.
Can carry out temperature variation testing to certified products oil, poor oil and miscella thereof simultaneously in the described step 1), described temperature variation testing is that finger is heated up the oils sample or cooling test, to draw the influence of temperature to different oils qualities, and find out correlativity, thereby determine the correlativity of oils quality and temperature.
Described step 3) specifically is according to determined oils quality judging standard, the ratio that accounts for the bulk sample signal with this part signal in the miscella is Y, certified products oil in the miscella or poor oil content are X, carry out the once linear match and obtain fitting a straight line y=kX+b, and then y is the measured value of Y; According to the mode analytic sample of standard specimen signal measurement, find the sample signal y of standard specimen signal area correspondence; Utilize certified products oil or poor oil content X in the y value calculation sample that obtains in calibration line and the sample analysis, thereby find out the difference of different quality oils samples and set up oils attribute test database with this.
The cpmg sequence row that relate in the inventive method are to be used for the conventional nuclear-magnetism pulse train that transverse relaxation is analyzed; The PGSE-CPMG sequence then is to be used for transverse relaxation and the conventional nuclear-magnetism pulse train that spreads correlation analysis; The anti-Laplce's algorithm of one dimension that relates among the present invention simultaneously is the conventional algorithm that is used for the relaxation time spectrum inverting; And two-dimentional anti-Laplce's algorithm is the conventional algorithm that is used for T2 and the inverting of coefficient of diffusion correlativity spectrum.
Below for adopting said method to carry out the concrete test example that the edible oil quality is differentiated:
One, the purpose of this test is:
1, the NMR signal difference of the different quality oil of test;
2, the poor oil that in certified products oil, adds different amounts, the detection signal difference, and find out the method for testing that adds poor oil content in the certified products oil;
3, set up the test model of nuclear-magnetism and oil quality;
Two, test material
The old oil of frying behind selection certified products soybean oil, the frying flour, nuclear-magnetism test tube, no nuclear magnetic signal bottle
Three, test apparatus
MicroMR magnetic resonance imaging analysis instrument
Four, the preparation of test specimen
1,7 oil samples altogether comprise the pure soy bean sample, and other six samples make for the old oil of frying that adds 2%, 5%, 10%, 20%, 50,100% mass parts in certified products soybean oil sample respectively; 2, the quality of all oil samples is all identical; 3, all oil samples are all put into same type and are not had the bottle of nuclear magnetic signal;
Five, process of the test
Adopt MicroMR magnetic resonance imaging analysis instrument that sample is carried out test analysis, specifically be to adopt the cpmg sequence row to measure the T2 of 7 samples after the relaxation time earlier, adopt the anti-Laplce's algorithm of one dimension to compose inversion algorithm again, draw the relaxation time spectrum of each sample, see Fig. 1~Fig. 7 as T2; From 7 width of cloth figure, can see, the main difference of the sample behind pure soy bean sample and the old oil of adding frying is to add a short relaxation can appear in the oily always sample of frying in test result small peak, all indicate the small peak ratio among Fig. 2 to Fig. 7 and accounted for the relation of resultant signal ratio, can see, when in the pure soy bean sample, adding the old oil of frying of different content, increase along with the old oily addition of frying, the area of small peak is also increasing gradually, and what following form disclosed is the pairing variation tendency of small peak:
Fry old oil content (%) small peak signal proportion (%)
0 0
2 0.25
5 0.36
10 0.68
20 0.9
50 2.38
100 5.18
From last table data as can be seen, increase along with the old oil content of frying, the small peak proportion is also increasing gradually, Fig. 8 is old oil content of frying and small peak ratio correlation figure, the ratio that accounts for the bulk sample signal with each small peak in the miscella is Y, the old oil content of frying in the miscella is X, carries out the once linear match and obtains fitting a straight line y=0.0502X+0.0506, and y is the measured value of Y; The content and the little peak area of the old oil of frying present extraordinary correlativity as can be seen.Analyze oils quality sample (non-test is prepared) at random according to the mode of standard specimen signal measurement, find the sample signal y of standard specimen signal area correspondence; Utilize the old oil content X of frying in the y value calculation sample that obtains in calibration line and the sample analysis, thereby find out the difference of different quality oils samples and set up oils attribute test database, with the formation of convenient edible oil quality measurement and industry testing standard with this; Thereby find out the difference of different quality oils samples and set up oils attribute test database with this.Utilize the test data of the database the inside of above-mentioned foundation that unknown oils sample is carried out the quality evaluation at last.
Above-mentioned experiment can prove low nuclear-magnetism method of testing can record in the certified products soybean oil, add 2% or the old oil of above frying after oil quality differentiate, and accurately in the specimen institute adds the oily always content of frying.The Blind Test checking is carried out in suggestion.From test data and test result, low-field nuclear magnetic resonance differentiates that the method for edible oil quality is a kind of method of testing very quickly and easily, and can obtain accurate test data, a kind of detection method of quick nondestructive is provided for oil quality analysis and Quality Detection.