The optical spectrum rapid nondestructive detection method of fruit and vegetable vitamin C content
Technical field
The present invention relates to the lossless detection method field, relate in particular to the method that a kind of application of spectral analytical technology quick nondestructive detects.
Background technology
In recent years, along with the continuous increase of fruit, yield of vegetables and improving constantly of people's living standard, consumer's requirement to the inside quality index of fruits and vegetables when choosing fruits and vegetables is more and more higher, and vitamin C is as a kind of important nutritional labeling, the main reference index the when consumer selects fruits and vegetables beyond doubt.And at present to the mensuration of fruit and vegetable vitamin C often based on destructive chemical analysis or use the expensive experimental instrumental analysis, the method for common detection fruit and vegetable vitamin C mainly contains:
2,6-dichloroindophenol titrimetry: by 2 of blueness, 6-dichloroindophenol solution carries out redox titration to containing ascorbic fruits and vegetables sample acid solution, during beginning 2,6-dichloroindophenol blue solution is reduced to colourless, when reaching titration end-point, unnecessary 2,6-dichloroindophenol blue solution presents light red in acid medium, by 2, and the ascorbic content of reduced form in the 6-dichloroindophenol solution usage calculation sample.2, though 6-dichloroindophenol titrimetry is the national standard method (GB/T6195-1986) that China's fruit, vegetables Vit C contents are measured, but measuring process need be carried out processing such as fruit making beating, constant volume, filtration, titration, demarcation, it is bigger that measurement result is influenced by human factor, detection speed is slow, detection efficiency is low, and needs to buy expensive chemical reagent, detects the cost height.
2, the 4-dinitro phenyl hydrazine method: total ascorbic acid comprises the ancient happy saccharic acid of reduced form, dehydrogenation type and diketone, by activated charcoal reduced form ascorbic acid (vitamin C) in the fruit sample is oxidized to hydroascorbic acid, again with 2, the effect of 4-dinitrophenylhydrazine generates red osazone, be directly proportional with total ascorbic acid content according to the content of osazone in sulfuric acid solution, carry out colorimetric assay.2, though the 4-dinitro phenyl hydrazine method is the national standard method (GB/T12392-1990) that total ascorbic acid is measured in China vegetables, fruit and the goods thereof, but measuring process need be carried out processing such as fruit making beating, constant volume, filtration, water-bath, colorimetric, curve plotting, it is bigger that measurement result is influenced by human factor, detection speed is slow, detection efficiency is low, and needs to buy expensive chemical reagent, detects the cost height.
High performance liquid chromatography: by high performance liquid chromatograph the fruits and vegetables samples of juice is analyzed, the vitamin C in the fruits and vegetables sample after oxalic acid solution extracts rapidly, separation determination on reverse-phase chromatographic column.Though high performance liquid chromatography be multiple component content is measured in China's food, the feed national standard method (for example: GB/T 19861-2005: the detection method of tonyred dyestuff in the food; GB/T 17817-1999: the assay method of vitamin A in the feed), but measuring process needs processing such as making beating, constant volume, filtration, stratographic analysis, and must be equipped with equipment such as high performance liquid chromatograph, UV-detector, integraph, so the method complicated operation, the instrument costliness of using, detection cost height, sense cycle is long.
Summary of the invention
In order to overcome the existing various defectives of above-mentioned detection method, the invention provides a kind of method that detects fruit, vegetables Vit C contents with the spectral analysis technique quick nondestructive.The method can not only realize that quick nondestructive detects the content of fruit and vegetable vitamin C, and does not need sample is carried out Chemical Pretreatment, does not need to buy expensive chemical reagent.From testing result, the present invention can carry out fast fruit and vegetable vitamin C content, harmless, effective and easy detection.
The optical spectrum rapid nondestructive detection method of fruit and vegetable vitamin C content may further comprise the steps:
(1) sample collecting: collect representational sample in the fruits and vegetables, comprise from now on desired sample chemical character in the actual analysis as far as possible;
(2) the modeling sample collection is selected: select the sample sets of the wider sample of concentration range as mathematical modeling from representational sample;
(3) modeling sample spectra collection: fruit and vegetable surfaces is after simple process, and each sample position is under the line gathered a spectrum every 120 degree, gets three spectrum and asks average;
(4) measure with reference to chemical score: use the national standard assay method carries out the inside quality Vit C contents to the fruits and vegetables sample of having gathered mensuration;
(5) mathematical model is set up: use the multivariate calibration methods in the Chemical Measurement, set up the mathematical model between the Vit C contents and absorption spectrum in the tested fruits and vegetables sample;
(6) calibration model checking: get the sample of known fruit and vegetable vitamin C content, under the same conditions, measure its absorption spectrum, the Vit C contents according to the concentrated fruits and vegetables sample of having set up of correction calculated with mathematical model checking requires the checking error less than 10%;
(7) unknown sample prediction: behind the scanning unknown sample spectrum, fruits and vegetables sample Vit C contents is predicted comparison with building mathematical model and spectrum thereof.
(8) mathematical model after establishment is proofreaied and correct.
(9) testing sample is carried out spectroscopic assay.
(10) spectroscopic data that records is imported the mathematical model of determining.
(11) obtain ascorbic content.
Has fibre-optical probe in the pick-up unit according to such scheme foundation, in fibre-optical probe, has light source, fibre-optical probe is contained in movably on the erecting frame, handle is housed on fibre-optical probe, handle and light source irradiation are 45 degree to the depression angle of fruits and vegetables glazed thread, be equipped with in the handle and accept catoptrical optical fiber transmission line, optical fiber transmission line is connected with spectrometer, spectrometer is connected with the computing machine with spectral manipulation software by the data cable line, and described light source, spectrometer and computing machine are connected with external power supply through supporting power lead.
The acquisition range of spectrum is 350-1800nm, belongs to visible light and near infrared spectrum district, and light source adopts the 50W halogen tungsten lamp.
Near infrared spectrum (Near Infrared Spectroscopy, NIRS) being electromagnetic wave between visible light (VIS) and mid-infrared light (MIR), is 780-2526nm (12820-3959cm according to the wavelength coverage in the near infrared spectrum district of American Society for Tests and Materials (ASTM) definition
-1).Near infrared spectrum mainly is because the anharmonicity of molecular vibration makes molecular vibration produce when the high level transition from ground state, what write down mainly is the frequency multiplication and the sum of fundamental frequencies absorption of hydrogeneous radicals X-H (X=C, N, O) vibration, different groups (as methyl, methylene, phenyl ring etc.) or near infrared absorption wavelength and the intensity of same group in the different chemical environment all have significant difference, near infrared spectrum has abundant structures and composition information, and the composition and the character that are suitable for very much carbon-hydrogen organic are measured.The spectral signature of visible-range can reflect visual signatures such as the color and luster, shape, texture of fruit sample.The essential characteristic of visible light and near-infrared diffuse reflection spectrum is that quick nondestructive detection fruit and vegetable vitamin C content has been established theoretical foundation.The fruits and vegetables sample of known vitamin C chemical measurement value and its visible light and near-infrared diffuse reflection spectrum are carried out the correlativity mathematical modeling, after model adaptability and reliability demonstration, as long as record the visible light and the near-infrared diffuse reflection spectrum of unknown fruits and vegetables sample, spectrum is directed into the mathematical model of building, just can realizes that quick nondestructive detects the testing goal of fruit sample Vit C contents.
The present invention solves the testing scheme that its technical matters adopts:
A kind of fruit of mutual relationship between fruit, vegetables vitamin C and the spectrum, fast non-destructive detection method of vegetables Vit C contents of being based upon; Spectrometer links to each other with computing machine by two-way parallel interface cable data line, links to each other with external power supply through supporting power lead simultaneously; Visible light and near-infrared diffuse reflection fibre are connected on spectrometer and the light source handle, make the light that comes out through the inner diffuse reflection of fruits and vegetables be transmitted through the fiber to detecting device in the spectrometer; Light source is fixed in the fibre-optical probe, and the light source depression angle is 45 degree; Fix the relative position of fruits and vegetables, light source, optical fiber and spectrometer, light source irradiation is at the fruit and vegetable surfaces with the optical fiber probe support, through fruits and vegetables internal reflection and scattering, by the detecting device of Optical Fiber Transmission to the spectrometer, collect the spectrum picture of fruits and vegetables sample, convert the spectrum master data to by spectral manipulation software again, by multivariate calibration methods in the Chemical Measurement software ascorbic chemical measurement value of fruit and spectrum master data are carried out mathematical then, set up model, through overcorrect, the mathematical model after determining at last to proofread and correct.The sample that needs are measured carries out spectroscopic assay, and the spectrum that records is imported the mathematical model of determining, obtains the Vit C contents of sample at last.
The fruit that is fit to the method for spectrum quick nondestructive detection fruit and vegetable vitamin C content can be tangerine orange, navel orange, apple, pears, strawberry etc., and vegetables can be cucumber, eggplant, tomato, taro, potato, radish, sponge gourd etc.
The present invention's beneficial effect that has of comparing with background technology is:
(1) analysis speed is fast, and the spectroscopic assay process can be finished in 1 second;
(2) the fruits and vegetables sample does not need pre-service, reaches nondestructive measurement;
(3) have higher degree of accuracy, reduce personal error;
(4) analysis result accuracy is approached standard method;
(5) do not need chemical reagent, reduce and detect cost, free from environmental pollution;
(6) can measure multiple composition and index simultaneously.
Description of drawings
Fig. 1 is the spectrum quick nondestructive pick-up unit figure of fruit and vegetable vitamin C content of the present invention;
Fig. 2 is a process flow diagram of the present invention;
Fig. 3 is the original averaged spectrum curve map of the typical modeling sample of three kinds of different cultivars of Nanfeng orange;
Fig. 4 is that 90 Nanfeng orange modeling samples are proofreaied and correct and inner validation-cross model scatter diagram fully;
Fig. 5 is the outside prediction of 20 a Nanfeng oranges scatter diagram.
Embodiment
With Nanfeng orange fruit sample is the method that example explanation spectrum quick nondestructive detects Vit C contents.
Three kinds of different cultivars of Nanfeng orange (Jiangxi provinces and cities mountain town fruitlet system, the big fruit in Jiangxi provinces and cities mountain town system, Jiangxi Province's POLO town fruitlet system) are used to set up calibration model, and detection procedure as shown in Figure 2.
(1) sample collecting: totally 110 and guarantee that these fruit sample surfaces and inside do not have open defect of three kinds of Nanfeng orange of plucking the maturity stage from Jiangxi provinces and cities mountain town, POLO town;
(2) the modeling sample collection is selected: select the sample sets of wide and 90 samples being evenly distributed of Vit C contents scope as mathematical modeling from 110 representational fruit samples;
(3) modeling sample spectra collection: fruit surface is after simple process, utilize the Indico V4.0 of U.S. ASD company or the Unsorambler V9.5 spectra collection software of Norway CAMO company that each sample is spent spectrum of collection in interval, position 120 under the line, get three spectrum and ask average.The typical mean absorption spectrum of three kinds of different cultivars of Nanfeng orange as shown in Figure 3;
(4) measure with reference to chemical score: use 2,6-dichloroindophenol titrimetry (GB/T 6195-1986) is carried out the mensuration of inside quality Vit C contents to 90 Nanfeng orange samples having gathered;
(5) mathematical model is set up: utilize the PLS1 method in the Unscrambler V9.5 multivariate statistics software of Norway CAMO company, set up the correction mathematical model between the vitamin C chemical assay value and visible light and near-infrared diffuse reflectance absorption spectrum in the sample, the results are shown in Figure 4;
(6) calibration model checking: in Unscrambler V9.5 multivariate statistics software, adopt inner cross verification (leave-one-out proof method) fully when setting up the correction mathematical model, model to be verified, the results are shown in Figure 4, its checking square error is 1.6284mg/100g;
(7) unknown sample prediction: behind the Nanfeng orange unknown sample scanning visible light and near-infrared diffuse reflection spectrum of 20 different cultivars, using positive mathematical model for school building and spectrum thereof predicts the sample Vit C contents, the results are shown in Figure 5, the mathematical model after determining to proofread and correct;
(8) sample to be measured is carried out spectroscopic assay, the spectroscopic data that records is imported the mathematical model of determining, obtain ascorbic content at last.
Fig. 1 has shown a kind of pick-up unit, tangerine orange 1 is placed on the
rubber washer 2 on the fibre-
optical probe 3, in fibre-optical probe, has
light source 4, be the 50W halogen tungsten lamp, fibre-optical probe is contained in movably on the erecting
frame 5, is equipped with to accept catoptrical optical
fiber transmission line 6 in the fibre-optical probe handle, adopts the standard reflection probe of U.S. ASD company, optical fiber transmission line is connected with
spectrometer 7, and spectrometer adopts the Quality Spec of U.S. ASD company
The pro spectrometer, spectrometer is connected with the
computing machine 9 with spectral manipulation software by
data cable line 8, data processing software is the Indico V4.0 of U.S. ASD company, detecting device is silicon detecting device (350-1000nm) and indium gallium arsenic detecting device (1000-1800cm), and optical fiber adopts the standard reflection fibre-optical probe of U.S. ASD company.Multivariate calibration methods is partial least square method (PLS) in the mathematical modeling.20 tangerine orange are listed in the table 1 through the ascorbic value that chemical measurement and spectral measurement obtain.
Show the contrast of 1:20 unknown sample predicted value and chemical reference value numerical value
Table 1 has shown spectrum predicted value and the chemical reference value that each sample records, and the checking square error that difference has between the two met mathematical model is 1.6248mg/100g.