JP5748897B1 - Determination of free fatty acids in vegetable oils and their raw materials using near infrared spectroscopy - Google Patents

Abstract

[PROBLEMS] To provide a method for quickly, simply and accurately quantifying free fatty acids in a sample containing vegetable oil and / or a plant body as a raw material thereof. A regression equation for quantifying free fatty acids in a sample containing vegetable oil and / or plant as a raw material thereof is obtained by a method including the following steps (1) to (3). And a method of quantifying free fatty acids in a sample using near-infrared light spectrum data of a test sample in the wavelength region used to create the regression equation. (1) A step of measuring a near infrared light spectrum of a sample, a solvent extract of the sample, or a lipid concentrate obtained by removing the solvent from the solvent extract (2) Free fatty acids in the sample subjected to step (1) Step of specific quantification (3) Analyze spectral data obtained in the whole or part of the wavelength range where the near-infrared light spectrum was measured and the quantitative value of free fatty acid by multivariate analysis method. Process for determining factors related to fatty acids [selection figure] None

Description

  The present invention relates to a method for quantifying free fatty acids in vegetable oils or their raw materials using near infrared spectroscopy.

Rice oil obtained from rice bran is attracting attention as a high-value-added edible oil that is rich in nutrition and physiological functions, and has excellent taste and texture, and its use is expanding. On the other hand, rice bran, a raw material for rice oil, is limited in supply and easily deteriorated, so it cannot be stored, and the rice crude oil after extraction has many impurities and the refining process is complicated. Rice oil has the disadvantage of high production costs and difficulty in increasing production compared to other vegetable oils. In addition, rice oil contains abundant functional components, but the content varies depending on the raw materials. Therefore, quality control of raw rice bran is very important in the production of rice oil.
Unlike other vegetable oils and fats, rice bran is characterized by being rich in lipolytic enzyme lipase. This lipase is inactive in the state of brown rice, but it is activated and rapidly activated when rice bran is separated by milled rice. It is known that oil (triacylglycerol, phospholipid, etc.) is decomposed to produce free fatty acid (FFA) (Formula 1) (Non-Patent Document 1). The generated FFA, particularly an FFA having a double bond, is susceptible to oxidation, generates oxidation products such as lipid peroxide harmful to the human body, and promotes degradation of lipid components. Rice crude oil extracted from degraded rice bran contains a large amount of oxidation products and the content of triacylglycerol (TAG) is also reduced, leading to an increase in refining load and a decrease in yield. Therefore, in order to efficiently obtain good quality rice crude from rice bran, it is necessary to evaluate not only the raw rice bran but also the degree of deterioration of rice crude, and evaluate and manage its quality.

In recent years, rice varieties (oil-fed rice) that are excellent in utilization as edible oil and fat materials have been developed, and in order to prevent rapid degradation of rice bran, reduction of enzymes that produce FFA such as lipase and phospholipase. One of the themes is breeding development of varieties that have been improved. Therefore, if it becomes possible to quickly evaluate the degree of deterioration of rice bran, it is possible to select rice bran that does not easily deteriorate by evaluating the degree of deterioration over time. Efficiency can be improved dramatically.
In view of the above, in order to maintain and improve the production efficiency and quality of rice oil, a rapid and accurate evaluation method for the degree of deterioration of rice bran and rice crude oil is required.

As described above, since deterioration of rice bran and rice crude oil occurs with an increase in FFA, it is effective to monitor the amount of FFA as an index for evaluating the degree of deterioration of rice bran and rice crude oil. FFA in fats and oils is generally evaluated by an acid value (amount of potassium hydroxide necessary to neutralize an acid contained in 1 g of fats and oils) measured using an alkali titration method. However, since there are acidic substances that react with potassium hydroxide in addition to FFA in lipids, it cannot be said that the acid value represents the exact amount of FFA. Further, the alkali titration method has a drawback that the measured value fluctuates due to the influence of impurities.
As a method for accurately measuring FFA, a gas chromatography (GC) method (Patent Document 1), various high performance liquid chromatography (HPLC) methods (Non-Patent Document 2), and the like have been reported. However, since the GC method and the HPLC method quantify fatty acids for each molecular species and add the quantified values, there are problems in complexity and measurement accuracy. On the other hand, in addition to the above-mentioned titration method, a method using an enzymatic method (colorimetric determination method) (Patent Document 2) or Fourier transform infrared spectroscopy (FT-IR) as a method capable of measuring the total amount of fatty acid content at one time ( Non-patent document 3) has been reported. Since these do not require calculation for each molecular species, they can be analyzed and evaluated more easily than the GC method or the HPLC method. However, these methods are also difficult to operate as a rapid analysis method because an oil and fat extraction step is required in advance when analyzing FFA in rice bran.

On the other hand, in order to evaluate the deterioration degree of fats and oils and fat components in food (such as rice), an analytical method for quantifying acid value using a near infrared spectroscopy (NIR) method has been reported (Patent Document 3, 4) Actually, a quantitative model has been put into practical use. The NIR method is a technique for obtaining information on the physical or chemical properties of a sample from the spectrum of scattered light or transmitted light obtained by irradiating the sample with near infrared light. Can be predicted in a very short time without destroying. However, since the NIR method creates a calibration model from the correlation between the chemical analysis value and the NIR spectrum, it is necessary to establish a chemical analysis method with high analytical accuracy in order to obtain a highly accurate calibration model.
The inventors have established a method for predicting the TAG content in brown rice and rice bran using the NIR method in the past (Patent Document 5). The NIR method is a very promising tool for determining the degree of degradation by quantifying the amount of FFA in rice bran and rice crude oil due to its rapidity, but the amount of FFA contained in rice bran and rice crude oil is applicable to the NIR method. It is necessary to establish a chemical analysis method that can accurately measure A calibration model of FFA by NIR has been prepared so far, but its chemical analysis value is based on the alkali titration method (acid value), and this method is a method that can specifically quantify FFA as described above. Absent. Therefore, the selection of chemical analysis methods has become an issue for the construction of an NIR calibration model that can predict the exact FFA content of rice bran and rice crude oil.

  When evaluating the quality of rice bran and rice crude oil from the viewpoint of raw materials for rice oil products, not only the deterioration degree of fat and oil components but also rice oil such as γ-oryzanol (OZ), sterol (ST), sterol ester (SE), etc. The content of characteristic functional components is also an important indicator. Recently, glycidol fatty acid esters in fats and oils are carcinogenic, which is a problem. This substance is thought to be produced from diacylglycerol in rice crude oil in the refining process, and the pattern of acylglycerol (tri, di, mono) is also an evaluation index. Although there are many reports on the analysis methods of these components, all of them are very complicated methods because they are evaluation methods for each molecular species. In terms of quality control / evaluation, it is not always necessary to evaluate each molecular species, and an analytical method that can analyze and evaluate quantitative patterns (crude oil profile) of these components simply and quickly is required. The inventor has developed a rapid analysis method using NIR for γ-oryzanol, which is a major functional component of rice oil, but no analysis method has been reported that can quickly obtain a crude oil profile. .

  A recently developed method for analyzing fat and oil components is a HPLC method (Non-Patent Document 4) that combines two separation modes of molecular weight fractionation (GPC) and polar interaction. By this method, TAG, diacylglycerol (DAG), monoacylglycerol (MAG) and fatty acid methyl ester, which are main components in biodiesel synthesized from vegetable oils and fats, can be separated as one peak. This method has so far been used for biodiesel synthesized from palm oil (Non-patent Document 4) and composition analysis of wax derived from rice bran (Non-patent Document 5). However, there is no report using this method for the analysis of rice crude oil, and there is no report on the analysis of functional components characteristic of rice oil. Also, no analysis examples of fats and oils containing OZ and ST have been reported.

JP 2009-204596 A Japanese Patent Laid-Open No. 2-184759 JP-A-4-77648 JP-A-6-91181 JP2013-72726A

Journal of the Japan Food Industry Association, 36 (6), 519 (1989) Journal of Chromatographic Science, 48, 663 (2010) Analytical chemistry, 54 (5), 381 (2005) Eur. J. Lipid Sci. Technol, 110, 422 (2008) Journal of the American Oil Chemists' Society, 88 (10), 1497 (2011)

  An object of the present invention is to provide a method for quickly, simply and accurately quantifying free fatty acids in a sample containing vegetable oil and / or a plant body as a raw material thereof. Another object of the present invention is to provide a method for quickly and appropriately evaluating the degree of deterioration and the ease of deterioration (stability) of an oil and fat component in a plant body that is a raw material for vegetable oil and fat using the same method. Furthermore, an object of the present invention is to provide a method for obtaining a quantitative pattern (crude oil profile) of γ-oryzanol, sterol, sterol ester and the like contained in order to evaluate the quality of vegetable fats and oils.

The present invention includes the following inventions in order to solve the above problems.
[1] A method for obtaining a regression equation for quantifying free fatty acids in a sample containing vegetable oil and / or plant as a raw material thereof, comprising the following steps (1) to (3) To obtain a regression equation
(1) A step of measuring a near infrared light spectrum of a sample, a solvent extract of the sample, or a lipid concentrate obtained by removing the solvent from the solvent extract (2) Free fatty acids in the sample subjected to step (1) Step of specific quantification (3) Analyze spectral data obtained in the whole or part of the wavelength range where the near-infrared light spectrum was measured and the quantitative value of free fatty acid by multivariate analysis method. A step of determining a factor related to fatty acid [2] The method of obtaining a regression equation according to [1], wherein the sample contains rice bran and / or rice oil.
[3] The method for obtaining the regression equation according to [1] or [2], wherein free fatty acids are specifically quantified by HPLC analysis using a non-aqueous GPC column in step (2).
[4] The toluene according to [3], wherein the composition of the mobile phase used in the HPLC analysis method using a non-aqueous GPC column is toluene containing 0 to 0.03% (v / v) acetic acid. How to get a regression equation.
[5] The above [3], wherein the sample solvent used in the HPLC analysis method using a non-aqueous GPC column is a mixed solution of toluene and alcohol, and the alcohol concentration is 20% (v / v) or more. The method of obtaining the regression equation as described in [4].
[6] The step or (1) is characterized in that the sample or lipid concentrate is dissolved in a solvent selected from chloroform, acetonitrile, t-butanol and dimethyl sulfoxide, and the near-infrared spectrum of the resulting solution is measured. The method of obtaining the regression equation in any one of said [1]-[5].
[7] The wavelength region in step (3) includes at least one region selected from 1435 ± 10 nm, 1662 ± 10 nm, 1863 ± 10 nm, 2162 ± 10 nm, and 2245 ± 10 nm [1] A method for obtaining the regression equation according to any one of to [6].
[8] The method for obtaining a regression equation according to any one of [1] to [7], wherein the near-infrared light spectrum is a spectrum of diffusely reflected light or transmitted light.
[9] A method for quantifying free fatty acids in a sample containing vegetable oil and / or plant as a raw material thereof, obtained by the method for obtaining a regression equation according to any one of [1] to [8]. A method for quantifying free fatty acid, comprising using a regression equation and near-infrared light spectrum data of a test sample in the wavelength region used to create the regression equation.
[10] At least selected from the group consisting of free fatty acids, triacylglycerols, diacylglycerols, monoacylglycerols, sterol esters, sterols and γ-oryzanols in a sample containing vegetable oils and / or plants as raw materials thereof An analysis method for simultaneously analyzing two or more components, wherein an HPLC analysis method using a non-aqueous GPC column is used.
[11] The analysis method according to [10], wherein the sample contains rice bran and / or rice oil.
[12] The above [10] or [11], wherein the mobile phase used in the HPLC analysis method using a non-aqueous GPC column is toluene containing 0 to 0.03% (v / v) acetic acid. ] The analysis method of description.
[13] The above-mentioned [10], wherein the sample solvent used in the HPLC analysis method using a non-aqueous GPC column is a mixed solution of toluene and alcohol, and the alcohol concentration is 20% (v / v) or more. [12] The analysis method according to any one of [12].
[14] A method for evaluating the degree of deterioration of a sample containing vegetable oil and / or a plant body as a raw material thereof, wherein the method for quantifying free fatty acids according to [9] is used.
[15] A composition analysis method for a lipid component of a sample containing vegetable oil and / or a plant body as a raw material thereof, wherein the analysis method according to any one of [10] to [13] is used.

  ADVANTAGE OF THE INVENTION According to this invention, the method of quantifying the free fatty acid in the sample containing the vegetable oil and / or the plant body used as the raw material rapidly, simply and with high precision can be provided. By using the quantification method, it is possible to quickly and appropriately evaluate the deterioration degree and stability of a sample containing vegetable oil and / or a plant body as a raw material thereof. In addition, according to the present invention, the amount of free fatty acid in a sample containing vegetable oil and / or plant as a raw material thereof can be analyzed more easily and accurately than conventional chemical analysis methods, and vegetable oil and fat and / or its raw material can be analyzed. It is possible to provide an analysis method capable of simultaneously analyzing other components contained in a sample containing a plant body.

It is a figure which shows the result of having examined the influence of the acetic acid concentration in the mobile phase (acetic acid containing toluene) in a HPLC-GPC method, (a) shows the result in case an acetic acid concentration is 0.25% (v / v). , (B) shows the results when the acetic acid concentration is 0.03% (v / v), (c) shows the results when the acetic acid concentration is 0.01% (v / v), (d) Indicates the result when the acetic acid concentration is 0.003% (v / v). TAG is triacylglycerol, DAG is diacylglycerol, MAG is monoacylglycerol, OZ is γ-oryzanol, SE is a sterol ester, ST is a sterol, and FFA is a free fatty acid. It is a figure which shows the result of having examined the influence of 2-propanol concentration in the sample solvent (toluene: 2-propanol) in HPLC-GPC method, (a) is the case where 2-propanol concentration is 0% (v / v) (B) shows the result when the 2-propanol concentration is 20% (v / v), and (c) shows the result when the 2-propanol concentration is 50% (v / v). (D) shows the results when the 2-propanol concentration is 80% (v / v), and (e) shows the results when the concentration is 2-propanol 100% (v / v). Peak of each component (FFA, TAG, DAG, ST, OZ) in the RI chromatogram when the 2-propanol concentration in the sample solvent (toluene: 2-propanol) in the HPLC-GPC method is 50% (v / v) It is a figure which shows the isolation | separation state. It is a figure which shows the correlation of the predicted value by NIR method, and each analysis value, (a) shows a result in case a chemical analysis value is an analysis value by HPLC-GPC method, (b) shows alkali analysis titration of a chemical analysis value. It is a figure which shows a result in case it is the analytical value by a method. It is a figure which shows the result of having examined the influence of (gamma)-oryzanol in the analytical value with a HPLC-GPC method and an alkali titration method. It is a figure which shows the correlation of the near-infrared-light-spectrum estimated value (quantitative value) and the HPLC-GPC method analytical value (actual value) about the free fatty acid content in rice bran, (a) is the free fatty acid content per lipid concentrate (B) is a result about the free fatty acid content per rice bran. It is a figure which shows the correlation of the near-infrared-light-spectrum estimated value (quantitative value) and the HPLC-GPC method analysis value (actual measurement value) about the free fatty acid content (per lipid concentrate) in the lipid concentrate of rice bran. It is a figure which shows the relationship between the storage time of rice bran, and a free fatty acid content (per lipid concentrate). It is a figure which shows the relationship between the storage time of rice bran, the chemical analysis value of free fatty acid content, and the predicted value by NIR method, (a) is the result about the free fatty acid content per lipid concentrate, (b) is the free per rice bran It is a result about a fatty acid content.

In the present specification, “vegetable oil and fat” refers to a lipid mainly composed of TAG separated from a plant body as a raw material, and its preparation method is not particularly limited, and includes both unpurified and purified ones. “Raw material” refers to a plant that is the main raw material for vegetable oils and fats, and does not include processing agents. “Sample solvent extract” refers to a solution obtained by extracting a sample using a solvent suitable for lipid extraction, and “lipid concentrate” refers to an unpurified product obtained by distilling off the solvent from the solvent extract of the sample. Refers to lipid. When simply described as “lipid concentrate”, the raw material and composition thereof are not limited.
In this specification, “rice koji” refers to koji (red koji, middle koji, white koji) obtained as a by-product of white rice when brown rice is milled, and the pericarp, seed coat, germ, aleurone layer, suballeurone layer of brown rice surface layer, Powdered endosperm mixed in the processing process is included. “Crude oil” refers to unrefined fats and oils separated from natural raw materials, and “rice crude oil” refers to crude oil made from rice bran. In this specification, “rice oil” refers to any of rice crude oil using rice bran as a raw material, rice oil being refined, and refined rice oil as a final product, and the production method is not particularly limited. This is not limited to rice oil, but is the same for all the ingredients described as “raw material name + oil”.

  In this specification, a free fatty acid (FFA) will not be specifically limited if it is contained in a vegetable oil, However, A C8-C24 single-chain monocarboxylic acid is preferable. Preferred FFAs include octanoic acid, decanoic acid, lauric acid, myristic acid, palmitic acid, palmitoleic acid, stearic acid, oleic acid, linoleic acid, linolenic acid, arachidic acid, eicosenoic acid, behenic acid, erucic acid, lignoceric acid, etc. Can be mentioned.

  In this specification, triacylglycerol (TAG), diacylglycerol (DAG), and monoacylglycerol (MAG) are not particularly limited, but are preferably ester forms of the above-mentioned FFA and glycerin. In the present specification, sterol (ST) includes triterpene alcohols and sterol analogs in which a part of the structure of sterols and triterpene alcohols is changed (substitution, isomerization, etc.) in addition to general sterols. ST is not particularly limited as long as it is contained in vegetable oil. Preferred STs include cholestanol, campestanol, ergostar, stigmasterol cholesterol, desmosterol, campesterol, dihydrobrass castrol clinosterol, brass casterol, 24-methylene cholesterol, kodisterol, β-sitosterol, stigmasterol, fuco 4-desmethyl sterols such as sterol, isofucostol, clerolosterol, 22-dehydrocrerolosterol, latosterol, fungisterol, episterol, 22-dihydrospinasterol, spinasterol, chondrasterol, abenasterol , Gramisterol, citrostadienol, 29-norlanosterol, obtusifoliol, cycloeucarenol, 29- 4α-methyl sterols such as lucyclolaudenol, 24-dihydrolanosterol, parqueol, cycloartanol, cycloartenol, cyclosador, 24-methylenecycloartanol, cyclobranol, cyclolaudenol, butyrospermol, Examples include tetracyclic triterpene alcohols such as Δ7-tilcanol and 10α-cucurbitadienol, and sterol analogs such as brassinosteroids. In the present specification, γ-oryzanol (OZ) is preferably an ester of ST and ferulic acid. In this specification, the sterol ester (SE) is preferably an ester of the above ST and FFA.

[Method for Obtaining Regression Equation for Quantifying FFA in Sample Containing Plant Oil and / or Plant as Raw Material]
The present invention provides a method for obtaining a regression equation for quantifying FFA in a sample containing vegetable oil and / or a plant body as a raw material thereof (hereinafter referred to as “method for obtaining the regression equation of the present invention”). The method for obtaining the regression equation of the present invention only needs to include the following steps (1) to (3).
(1) A step of measuring a near infrared light spectrum of a sample, a solvent extract of the sample, or a lipid concentrate obtained by removing the solvent from the solvent extract (2) Free fatty acids in the sample subjected to step (1) Step of specifically quantifying (3) Factors related to FFA by analyzing spectral data obtained in all or part of the wavelength range in which the NIR spectrum was measured and the quantitative value of FFA by a multivariate analysis method The method of obtaining the regression formula of this invention may include processes other than (1)-(3), and the content is not ask | required.

  In step (1), the NIR spectrum of a sample containing vegetable oil and / or plant as a raw material, a solvent extract of the sample, or a lipid concentrate obtained by removing the solvent from the solvent extract is measured. The measuring method of the NIR spectrum is not particularly limited, but the diffuse reflection method or the transmission method is preferable, and it is more preferable to use these properly depending on the sample. If the sample is solid, it is preferable to measure mainly using the diffuse reflection method, and if the sample is liquid, it is preferable to measure mainly using the transmission method. However, in the case of a liquid, the sample is highly colored and difficult to transmit light. Preferably measures diffusely reflected light. When the sample is a solution, it is preferable to measure the transmitted light by appropriately adjusting the concentration. The measurement cell is not particularly limited, but it is preferable to use properly depending on the measurement sample and the measurement method. NIR spectra are measured not only in off-line analysis but also in at-line, on-line, in-line, non-invasive analysis, etc. Also good. The NIR spectrum can be measured by taking a sample and measuring with a near-infrared analyzer in the analysis room, or by bringing a near-infrared analyzer into a production site or warehouse and measuring it on the spot. It is also possible to continuously measure the NIR spectrum of a sample by incorporating a near-infrared analyzer in the production line.

  For measurement of the NIR spectrum, a commercially available near-infrared spectrometer can be used, but those capable of measuring transmitted light and diffuse reflected light are preferred. Moreover, what was equipped with the solid measurement module, the liquid measurement module, the optical fiber module, and the solid permeation | transmission measurement module according to a sample can be utilized. Moreover, what provided the temperature control function of the measurement cell is more preferable. Specifically, MPA (Bruker Optics), TANGO series (Bruker Optics), NIRFlex N-500 (Buch), NIRFlex N-400 (Buch), NIRLab N-200 (Buch), NIRMaster (Buch), NIRS6500 ( Nireco), XDS / NIR series (Nireco), SpectraStar Series (BLTEC), Light Quality Checker HOS-F Series (SAIKA), Light Quality Checker Portable HQC-F30 (SAIKA), IRAffinity-1 (Shimadzu Corporation), LAMBDA 750/950 / 1050 and Frontier NIR (Perkin Elmer), FT-IR4000 series and FT-IR6000 series (JASCO), FT-NIR MB 600 (ABB Bomem) and the like, MPA and TANGO series (Bruker Optics) is particularly preferred. The obtained spectrum is preferably a Fourier transform spectrum.

In the step (1), either a sample containing vegetable oil and / or plant as a raw material thereof, a solvent extract of the sample, or a lipid concentrate obtained by removing the solvent from the solvent extract is used for NIR spectrum measurement. What is necessary is just to be able to do, and it is not limited about the plant body used as the raw material of vegetable oil and fat, vegetable oil and fat.
Plants that can be used as raw materials for vegetable oils include rice, corn, sesame, safflower, soybeans, rapeseed, sunflower, olives, barley, grapes, palm, palm, cotton, flax, castor seed And algae. Among these, it is preferable that it is a part of rice seed, more preferably a part of brown rice, and still more preferably rice bran. Vegetable oils include rice oil, corn oil, sesame oil, safflower oil, soybean oil, rapeseed oil, sunflower oil, olive oil, barley oil, grape seed oil, coconut oil, palm oil, palm kernel oil, cottonseed oil, coconut oil, linseed oil , Castor oil, algal oil and the like, and this domestic rice oil is preferred. The vegetable oil may be after purification, before purification, or during purification.

  Preferable rice bran includes, for example, rice bran obtained by milling brown rice, processed rice bran, defatted rice bran, rice bran oil residue, rice bran during degreasing, and the like. Among these, those not subjected to degreasing treatment are preferable, and those not processed after milling are more preferable. The rice varieties used as the raw material for the rice bran as a sample are not particularly limited. The rice bran is preferably mainly composed of red rice bran, medium rice bran, or white rice bran obtained by polishing brown rice, and particularly preferably red rice bran or medium rice bran. Moreover, although white rice may be mixed in these rice bran, the mixing rate is preferably 20% by weight or less, and more preferably 10% by weight or less.

  The purity of the sample fat used for the measurement of the NIR spectrum is not particularly limited. When using vegetable fats and oils as a sample, the manufacturing method is not limited, but those extracted with a solvent excellent in oil solubility and those extracted by pressing are preferable, and vegetable oils and fats in the middle of purification or purified vegetable oils and fats are used. Also in some cases, those refined from such crude oil are preferred. Preferred extraction solvents include hexane, diethyl ether, ethanol, 2-propanol, acetone, chloroform, or a mixture thereof. When using refined vegetable oils and fats, those before use or after use (waste oil) may be used. In addition, by-products such as deodorized scum, soap stock, FFA, wax (wax), gum and the like produced as by-products in the essential oil process can be suitably used as a sample.

In step (1), a mixture of different vegetable oils and fats may be subjected to measurement of the NIR spectrum. In addition, the mixture of different vegetable oils and fats may contain animal oils such as beef tallow, lard, chicken oil (chii oil), fish oil, etc., and processed for food, feed, cosmetics, pharmaceuticals, etc. Also good.
The lipid content in the sample is preferably 5% by weight or more, and more preferably 10% by weight or more. When the lipid content in the sample is less than 10% by weight, it is preferable to use a solvent extract obtained by performing lipid extraction or a lipid concentrate obtained by removing the solvent from the solvent extract for NIR spectrum measurement. The lipid extraction method will be described later.

  When the sample is a processed product of the raw material for vegetable oils and fats, the sample may be any of food, feed, cosmetics, and pharmaceuticals using a plant body that is the raw material for vegetable oils and fats. Examples of food include confectionery (rice cake, steamed confectionery, rock cake, crepe, rice cake, hot cake, muffin, steamed bun, cupcake, baked confectionery, grilled rice cake, dumpling, cookies, castella, madeleine, waffle, etc.) , Breads (bread, scones, naan, etc.), noodles (noodles, ramen, buckwheat, pasta, macaroni, etc.), kneaded products (kamaboko, chikuwa, hanpen, sweet potato fried, sausage, etc.), dishes in general (white rice, rice balls, rice porridge, tofu) , Cereal, gnocchi soup, okonomiyaki, curry, stew, gratin, tempura, pasta soup, dango soup, pizza, hamburger, fried rice, soup base, etc., flour (rice flour, brown rice flour, corn starch, bread crumbs, potato starch, tempura Powder), drinks (tea, coffee, cocoa, soft drinks, fruit drinks, Gender drinks, alcoholic beverages, such as energy drinks), seasoning (soy sauce, source, cooking wine, mirin, dashi, such as miso), tablet candies, and the like supplements such. Examples of the feed include feed for livestock such as cattle, horses and pigs, feed for poultry such as chicken, pet feed such as dogs and cats, feed for farmed fish, and the like. Examples of cosmetics include creams, lotions, gels, mists, masks, packs, shampoos, rinses, bathing agents, and the like, which are blended with a plant that is a raw material for vegetable oils and fats. Examples of pharmaceuticals containing such raw materials and materials include tablets, powders, capsules and the like using a plant body as a raw material for vegetable oils and fats as excipients.

  When the sample is a processed product of vegetable oils and fats, there are no particular limitations on the plant body, composition, production method, and the like as the raw material. Processed products include, for example, tempura, tempura, deep-fried, fried, margarine, butter, cheese, fried potato, deep-fried, thick fried, fresh cream, coffee milk (powder), whipped cream, dressing, mayonnaise, tartar sauce, Foods such as oyster sauce, white sauce, potato chips, chocolate, biscuits, corn snacks, popcorn, crackers, fried rice crackers, rice crackers, short cakes, eclairs, cream puffs, wafers, bavaroa, donuts, karinto, and ice cream. Moreover, for example, cosmetics mainly composed of fats and oils such as soaps, detergents, foundations, cleansing oils, aroma oils, lip balms, and nail polishes can be used.

  When the sample contains a plant body that is a raw material for vegetable oil and is solid, it may be used as it is for NIR analysis in step (1), but it is preferable to perform pulverization for homogenization, and further to perform drying. It is preferable to remove moisture. The drying method in this case is not particularly limited, and a heat drying method, a freeze drying method, a vacuum drying method, or the like can be used, but a method that can avoid denaturation of the sample is preferable. The pulverization method is not particularly limited. For example, when the sample is solid and the sample is too large to be crushed as it is, it is preferable to cut and break it to an appropriate size before applying the pulverization treatment. In this case, it is preferable to first perform freeze-drying treatment or vacuum drying treatment before crushing. In the case where the sample is liquid, the oil and fat as the main component is handled in the same manner as the oil and fat sample described later. For oils and fats that do not contain oil and fat as the main component, dry treatment is performed to remove moisture, and the resulting residue is handled in the same manner as a solid sample, or solvent extraction is performed from the residue, and the resulting extract is used as a solvent extract of the sample described below. Treat as well.

  When the sample is an oil or fat containing vegetable oil or fat, it is preferable that foreign matters have been removed and the sample is in a homogeneous state. When the sample is a liquid, it may be used for the NIR spectrum measurement in the step (1) as it is by the transmission method. In the case of solid oils and fats, the NIR spectrum may be measured by the diffuse reflection method as it is, but if it can be dissolved by heating, it is heated to a liquid state and the NIR spectrum is measured by the transmission method. Is preferred. In the case where the sample is liquid oil and fat and there is a precipitate (insoluble matter), when the precipitate is dissolved by heating, it is preferable to measure by the permeation method after heating the measurement cell and dissolving the precipitate. If the precipitate does not dissolve even after heating, homogenize the sample and measure the NIR spectrum by the diffuse reflection method, or use the solvent extract obtained by solvent extraction or dissolution as a sample. It is preferable. When the viscosity of the sample is high, it is preferable to lower the sample viscosity by heating the measurement cell containing the sample. When the sample is opaque and can be clarified by heating, it is preferable to measure in a state where the measurement cell is heated and clarified.

  When the sample is a processed product of vegetable oils and fats, the sample is handled in the same manner as the oil and fat samples when oils and fats are the main component. If fats and oils are not the main component, the NIR spectrum may be measured after homogenization if the sample can be homogenized or easily homogenized, but if the sample is difficult to homogenize, the measurement accuracy will be improved. In order to improve, it is preferable to use a solvent extract obtained by performing a solvent extraction process or a dissolution process described later as a sample. The method for measuring the NIR spectrum of the homogenized sample is handled in the same manner as the sample containing the fat and oil raw material, and when analyzing the solvent extract obtained by performing solvent extraction, it is handled in the same manner as the solvent extract of the sample described later.

  When the sample contains protein regardless of whether it is solid or liquid, it is preferable to perform protein removal treatment. The method of protein removal treatment is not particularly limited, and examples thereof include denaturation precipitation methods such as addition of acids and organic solvents and heat treatment, ultrafiltration methods, dialysis methods, and column methods. When the sample is solid, the sample is preferably broken and pulverized and then suspended in a suitable solvent for protein removal treatment.

  When the sample is a solid or a liquid that does not contain oils and fats as a main component, the FFA concentration (per lipid concentrate) in the sample is preferably 1.5% by weight or more, and more preferably 3.0% by weight or more. In the case where the sample is a liquid sample mainly containing fats and oils, an oily fat sample or an oil processed sample, the FFA concentration (per lipid concentrate) in the sample is preferably 0.5% by weight or more, more preferably 1.0% by weight or more. .

  In the step (1), the sample solvent extract or the lipid concentrate obtained by removing the solvent from the extract can be suitably used for the measurement of the NIR spectrum. The solvent used for extraction is preferably a solvent capable of dissolving FFA. Examples of the solvent capable of dissolving FFA include ethers such as diethyl ether, dimethyl ether, methyl ethyl ether, diisopropyl ether, 2-methoxyethanol, and tetrahydrofuran (THF), methanol, ethanol, propanol, isopropanol (2-propanol), Alcohols having 3 to 11 carbon atoms such as butanol, t-butanol, glycerin and ethylene glycol, alkanes having 5 to 17 carbon atoms such as pentane and hexane, alkenes having 5 to 20 carbon atoms such as pentene and hexene, cyclo Cycloalkanes such as pentane and cyclohexane, acetone, acetonitrile, chloroform, ethyl acetate, dioxane, trifluoroacetic acid (TFA), benzene, dimethyl sulfoxide (DMSO), petroleum ether , Toluene, dimethylformamide, dichloromethane, carbon tetrachloride, etc. t- butyl acetoacetate are exemplified. These solvents may be used alone, or may be used in an arbitrary ratio as long as they can be mixed uniformly. The extraction method is not particularly limited, and a known lipid extraction method can be used. For example, a Soxhlet extraction method, a chloroform-methanol extraction method, an acid decomposition method and the like can be mentioned. Among them, the Soxhlet extraction method is preferable in terms of extraction efficiency. When a solvent extract is used, if the solvent is a preferable solvent for NIR measurement described later, the NIR spectrum may be measured as it is. If the solvent is not a preferred solvent for NIR measurement, it is preferable to redissolve the lipid concentrate obtained by removing the solvent once in a preferred solvent for NIR measurement or to directly analyze the lipid concentrate. Solvents preferable for NIR measurement are not particularly limited, and examples thereof include chloroform, acetonitrile, t-butanol, dimethyl sulfoxide, etc. Among these, chloroform and acetonitrile are more preferable. If the sample is a solvent extract or a solution obtained by re-dissolving a lipid concentrate in a solvent preferred for NIR measurement, it can be directly subjected to NIR analysis by the permeation method. . When the sample is a solvent extract, the FFA concentration (per solvent extract) is preferably 0.03% (w / v) or more, and more preferably 0.06% (w / v) or more.

  The regression equation obtained by the present invention can also be used for the quantification of FFA in a sample derived from a living body for the purpose of clinical trials. Although the sample derived from a living body is not particularly limited, for example, a biological component of an animal is preferable. Examples of animal-derived samples include blood, serum, plasma, tissue fluid, lymph fluid, cerebrospinal fluid, breast milk, pus, mucus, runny nose, sputum, urine, feces, ascites, semen from humans, laboratory animals, livestock, farmed fish, etc. And body fluids such as skin, mucous membrane, various organs, tissues such as bone, hair, body hair and the like.

When the sample is a sample derived from a living body and is intended for nondestructive analysis, the sample may be directly used for the NIR spectrum measurement in the step (1), but the sample is prepared in order to improve the measurement accuracy. It is preferable. When the sample is a solid, it is preferable to remove foreign substances, and when the sample is a liquid, it is preferable to perform homogenization by removing the foreign substances and precipitation.
When non-destructive analysis is not aimed, it is preferable to perform homogenization or protein removal treatment by crushing or stirring. However, since the FFA contained in the sample is predicted to have a very low concentration, there is a possibility that sufficient measurement sensitivity cannot be obtained by direct analysis. Therefore, it is a condition that a stable recovery rate is obtained, but it is more preferable to prepare a lipid concentrate extracted using a solvent as a measurement sample, rather than directly analyzing the sample. The extraction method is not particularly limited, but it is preferable to perform the protein removal treatment first in order to improve extraction efficiency. The method of protein removal treatment is not particularly limited, and examples thereof include precipitation method by adding acid or organic solvent, ultrafiltration method, dialysis method, column method and the like. When the sample is directly analyzed, the FFA concentration (per lipid concentrate) in the measurement sample is preferably 1.5% by weight or more, and more preferably 3.0% by weight or more.

  It is preferable to use a plurality of samples having different FFA contents as the sample to be used in the step (1). For example, when using rice bran as a sample, store a large amount of fresh rice bran immediately after milling and the rice bran to a certain extent, and sample and measure the amount necessary for measurement over time according to the progress of deterioration, suitable for measurement Can be sampled from the beginning and stored, and the same sample can be measured over time. For example, when using rice oil as a sample, a plurality of rice oils having different FFA contents can be prepared and used by mixing rice oil having a low FFA content and rice oil having a high FFA content. Prepare multiple samples to be subjected to step (1) so that the FFA content in the test sample to be quantified using the regression equation obtained is included in the range of the minimum and maximum values of FFA included in the sample. Is preferred. The larger the number of samples used in the step (1), the better. In order to obtain sufficient prediction accuracy, 30 samples or more are preferable, and 50 samples or more are more preferable.

  In step (2), FFA in the sample subjected to step (1) is specifically quantified. The method for specifically quantifying FFA in the step (2) is not particularly limited. For example, an enzymatic method, an LC / MS method, a HPLC analysis method using a non-aqueous gel permeation chromatography (GPC) column (hereinafter “HPLC”). Among these, the HPLC-GPC method is preferable. In addition, since the alkali titration method cannot specifically quantify FFA in a sample, it is not included in the method of specifically quantifying FFA.

  When the method for specifically quantifying FFA is the HPLC-GPC method, the separation column used in the HPLC-GPC method is not particularly limited, but a column capable of using a low-polar solvent such as toluene as the mobile phase is preferable. . The column shape is not particularly limited as long as necessary separation can be obtained. The material of the column filler is not particularly limited, but is preferably a polymer base, more preferably a polystyrene base, and further preferably a styrene divinyl copolymer base. Moreover, the exclusion limit molecular weight of a GPC column needs to be 300 or more, 1000 or more are preferable and 1000-5000 are more preferable. The penetration limit molecular weight needs to be 200 or less, preferably 100 or less, and more preferably 50 or less. The column temperature is not particularly limited as long as it does not cause any problem in column performance. Specific examples of GPC columns suitable for the present invention include Phenogel series (Phenomenex), GPC UT-800, GPC HT-800, GPC AT-800 series (Showa Denko), HSPgel series (Waters), TSKgel series (Tosoh) , PLgel series (agent technology), SDV series (PSS), etc., among which Phenogel series (Phenomenex) is preferable. Although it is preferable to obtain the necessary separation with a single GPC column, a plurality of the same or different types of GPC columns may be connected as necessary. Further, a GPC column and other HPLC columns may be connected and used. Examples of HPLC columns other than GPC columns include normal phase columns, reverse phase columns, amino columns, phenyl columns, and silver columns.

  The flow rate of the mobile phase is not particularly limited as long as it does not cause any problem in column performance. The detector is not particularly limited as long as it can detect FFA, but a detector that can detect components such as TAG, DAG, MAG, sterol ester (SE), ST, and OZ simultaneously with FFA is preferable. Specifically, a fluorescence (FL) detector, an ultraviolet-visible spectroscopy (UV-Vis) detector, a photodiode array (PDA) detector, an evaporative light scattering (ELSD) detector, a mass (MS) detector, an electrochemical Examples include (EC) detectors, differential refractive index (RI) detectors, electrical conductivity (CD) detectors, and corona CAD detectors. Among these, it is preferable to use an RI detector, an MS detector, an ELSD detector, a UV-Vis detector, a PDA detector, and a corona CAD detector, and an RI detector, an ELSD detector, and a corona CAD detector are more preferable. If necessary, a plurality of these different detectors may be connected and used.

  The mobile phase used in the HPLC-GPC method is preferably a solvent that can dissolve FFA and is suitable for GPC, and more preferably a solvent that can dissolve components such as TAG, DAG, MAG, SE, ST, and OZ. The mobile phase is preferably a solvent that can be used in a GPC column. Specifically, THF, chloroform, N, N-dimethylformamide (DMF), hexafluoroisopropyl alcohol (HFIP), dichloromethane (methylene chloride), toluene, and these Examples thereof include mixed solvents. Of these, a low-polar solvent is preferably the main component, and a small amount of acid is preferably added. Low polar solvents include aromatic hydrocarbons such as toluene, benzene and xylene, alkanes having 5 to 17 carbon atoms such as pentane and hexane, cycloalkanes such as cyclopentane and cyclohexane, carbons such as pentene and hexene. Although several to 20 alkenes, petroleum ethers and the like can be used alone or mixed (when homogeneously mixed), toluene is preferably used. As the acid to be added, for example, acetic acid, phosphoric acid, citric acid, tartaric acid, boric acid, formic acid, TFA and the like can be preferably used, and among them, acetic acid is preferable. For example, when Phenogel is used as a separation column, a mixed solution of toluene and an acid is preferable. When a mixed liquid of acetic acid and toluene is used for the mobile phase, it is preferable to contain 0.003-0.03% (v / v) of acetic acid, more preferably 0.005-0.03% (v / v). Preferably, 0.01 to 0.03% (v / v) is further included.

  If homogenous mixing is possible, the mobile phase is THF, chloroform, DMF, HFIP, diethyl ether, trichlorobenzene, pyridine, o-dichlorobenzene, trichlorobenzene, dichloromethane, quinoline, m-cresol, alcohols having 3 to 11 carbon atoms. , Ethers, dioxane, acetonitrile, dimethyl sulfoxide, ethyl acetate, acetone, methyl ethyl ketone, dimethylformamide, dimethylacetamide, dichloromethane, carbon tetrachloride, t-butylacetoacetate and the like may be added as necessary. Among these, THF, DMF, HFIP, and dichloromethane are preferable.

  In step (2), when FFA in a sample is quantified by HPLC-GPC method, a lipid concentrate is extracted from the sample by a known lipid extraction method (eg, diethyl ether extraction method), and this is used as a sample solvent. It is preferable to melt | dissolve and use for HPLC-GPC method. When quantifying FFA in a sample by the HPLC-GPC method, the sample may be directly dissolved in a sample solvent and subjected to the HPLC-GPC method. Similarly to the sample, a lipid concentrate is extracted by a known lipid extraction method. Then, it may be dissolved in a sample solvent and subjected to the HPLC-GPC method. The sample concentration used for the HPLC-GPC method is not particularly limited as long as the lipid component is completely dissolved, and the FFA concentration in the sample solution is not particularly limited as long as it can be quantitatively measured with a detector.

The lipid concentrate extracted from the sample may be dissolved in the sample solvent as it is and subjected to the HPLC method. However, if necessary, pretreatment for reducing components other than FFA may be performed. Examples of pretreatment methods include precipitation / insoluble matter removal by centrifugation, dehydration treatment, degumming treatment, wax removal treatment by wintering, liquid-liquid extraction treatment using multiple solvents, solid-phase extraction treatment, etc. It is done.
There are two types of solid-phase extraction processes, a column method using a column preliminarily filled with a stationary phase and a batch method in which a stationary phase is added to a sample, and any of these may be used. Suitable columns used in the column method include normal phase columns, reverse phase columns, amino columns, phenyl columns, ion exchange columns, and the like, and these may be used alone or in combination. Examples of columns suitable for solid-phase extraction include Sep-pak series, Oasis series (Waters), Inertsep series (GL Science), BondElute series (Agilent Technology), and Presep series (Wako Pure Chemical Industries). On the other hand, in the batch method, the same stationary phase packed in the column used in the column method can be used. The stationary phase in the batch method may be used alone, but a plurality of stationary phases may be combined, or a dehydrating agent may be combined. An example of a treatment method combining a stationary phase and a dehydrating agent is the QuEChERS method and its improved method.
In performing these pretreatments, an antioxidant may be added to avoid oxidation of the sample. Examples of the antioxidant to be added include pyrogallol (pyrogallolic acid), dibutylhydroxytoluene (BHT), butylhydroxyanisole (BHA), t-butylhydroquinone (TBHQ), and ethoxyquin.

The sample solvent should just be a solvent which can melt | dissolve FFA, Furthermore, the solvent which can melt | dissolve components, such as TAG, DAG, MAG, SE, ST, and OZ, is preferable. Examples of the sample solvent used in the step (2) include toluene and benzene, alkanes having 5 to 17 carbon atoms such as pentane and hexane, alkenes having 5 to 20 carbon atoms such as pentene and hexene, and petroleum ethers and the like. A polar solvent is used alone or in combination, or a solution obtained by adding and mixing a polar solvent such as THF, chloroform, DMF, HFIP, t-butylacetoacetate, etc., as necessary is preferable. When an RI detector is used as the detector, the main component of the sample solvent is preferably the same as that used for the mobile phase.
As an example, when Phenogel (manufactured by Phenomenex) is used, a mixed solution of toluene and alcohol is preferable. The alcohol is not particularly limited, but a lower alcohol having 4 or less carbon atoms is preferable. Specific examples include ethanol, methanol, 1-propanol, 2-propanol, n-butanol, t-butanol, etc. Among them, 2-propanol is preferable. preferable. When a mixed solution of toluene and alcohol is used as a sample solvent, the alcohol concentration is preferably 20 to 100% (v / v), more preferably 40 to 100% (v / v), and 40 to 80%. More preferably, it is (v / v).

The HPLC-GPC method is superior in that FFA can be quantified with high accuracy as compared with a conventional chemical analysis method such as an alkali titration method. The HPLC-GPC method does not require pretreatment other than dissolving the analysis sample in a solvent, and can separate the FFA peak containing various molecular species as a single peak. Therefore, the fatty acid is quantified for each molecular species, Compared to the conventional analysis method that adds the quantitative values, the quantitative value is very easy and the reproducibility of the measured values is high.
The above-mentioned HPLC-GPC method realizes separation of MAG and FFA with similar molecular weights and reveals quantitative patterns of components such as TAG, DAG, MAG, SE, ST, OZ as well as FFA analysis. Is possible. The analysis conditions in the step (2) are preferably set so that the peaks of these components can be separated and detected.

  The HPLC-GPC method of step (2) can be used alone as an analysis method for FFA. The analysis sample in this case is not particularly limited, but preferably contains an oil and fat raw material and / or an oil and fat. In this case, the oil and fat raw material is preferably a plant body, more preferably rice bran. Other conditions are the same as in the HPLC-GPC method.

  In step (3), the spectrum data obtained in all or a part of the wavelength range in which the NIR spectrum is measured and the quantitative value of FFA are analyzed by a multivariate analysis method to determine factors related to FFA. The part of the wavelength range in which the NIR spectrum is measured is an arbitrary partial region included in the wavelength range in which the NIR spectrum is measured. Although a part of wavelength range which measured the NIR spectrum is not specifically limited, It is preferable that it is 800-2500 nm or its part, and it is more preferable that it is 1063-2355 nm or its part. In addition, it is more preferable that the region includes one or more regions, more preferably two or more regions, of a group related to FFA (1435 ± 10, 1662 ± 10, 1763 ± 10, 2162 ± 10 and 2245 ± 10 nm). Further, this one wavelength region is preferably a continuous region where the difference between the maximum wavelength and the minimum wavelength is 20 nm or more, more preferably a continuous region of 100 nm or more, and even more preferably 150 nm or more. Only one wavelength region or a plurality of wavelength regions may be set.

  The multivariate analysis method used in the step (3) includes PLS (partial least squares) regression analysis method, multiple linear regression analysis (MLR) method, principal component regression analysis (PCR, principal component regression) method, There are CLS (classical least squares) regression analysis methods, among which the PLS regression analysis method is preferably used. PLS regression analysis can be performed using commercially available software. Examples of the software to be used include The Unscrambler (Camo Software Japan Co., Ltd.), OPUS (Bruker Optics), NIRCal (Nippon Büch), and Pirouette (GL Science). Using these software, the NIR diffuse reflection spectrum is analyzed, and the result is used in the present invention. The spectrum data or the Fourier transform spectrum data to be used may be original spectrum data, but it is preferable to use data obtained by processing the original spectrum data. Examples of data processing methods include, for example, first-order differentiation, second-order differentiation, third-order differentiation and the like (Derivative), smoothing (Smoothing), spectrum subtraction (Normalization), normalization (Normalization), MSC correction (Multiplicative). Scatter correction), SNV correction (standard normal variant correction), and the like. These processing methods may be used alone or in combination. Among them, preferable processing methods include SNV alone, secondary differentiation alone, a combination of primary differentiation and SNV correction, and a combination of primary differentiation and MSC correction.

  By using the spectral data of the wavelength region used for the analysis and the quantified FFA content (actual value of FFA) for multivariate analysis, a factor related to the FFA content can be determined and a regression equation can be obtained. The factor related to the FFA content refers to a virtual variable having a high correlation with the change in the FFA content inherent in the spectral data. A regression equation for predicting the FFA content from the spectrum data is created using the regression coefficient of the factor. In near-infrared spectroscopy, after creating a regression equation, the measurement accuracy of the created regression equation is evaluated using a sample other than the sample for obtaining the regression equation (evaluation sample). The measurement accuracy of the regression equation is evaluated by the standard deviation or average value of the residual (measurement error) between the actual measurement value of the evaluation sample and the analysis value based on the predicted value calculated by the created regression equation. 6 and 7 are scatter diagrams of the predicted value (quantitative value) and the actually measured value. The calibration curve regression equation is a regression equation derived so that the correlation between the actual measurement value and the predicted value (quantitative value) is maximized.

  In the present invention, it is assumed that the test sample contains a large amount of water, and in that case, it may be quantified per dry matter amount. In such a case, it is necessary to obtain and quantify NIR spectrum data after performing moisture removal treatment on the measurement sample, or to quantify the moisture content by some method and correct the predicted value (quantitative value). Examples of the method for determining the water content include a method in which a part of a sample is collected and measured by a loss on drying method or a Karl Fischer method. For the purpose of nondestructive analysis, a method for predicting the water content using NIR spectrum data can be suitably used.

[Quantification method of FFA using NIR spectroscopy]
The present invention is a method for quantifying FFA using the calibration curve regression equation obtained by the above-described method for obtaining the regression equation of the present invention and the NIR spectrum data of the test sample in the wavelength region used for the creation of the regression equation (hereinafter “ The FFA quantification method of the present invention ”) is provided. Specifically, first, the NIR spectrum of the test sample is measured. The form of the test sample is preferably the same form as the sample subjected to NIR spectrum measurement in order to obtain a calibration curve regression equation used for quantification of FFA. In other words, in the FFA quantification method of the present invention, it is preferable to use a calibration curve regression equation obtained based on NIR spectrum measurement values of a sample having the same form as that of the test sample. The wavelength range for measuring the NIR spectrum is not particularly limited as long as it includes the wavelength region used for preparing the calibration curve regression equation, and is preferably 800 to 2500 nm.

  The NIR spectrum data of the test sample can be acquired in the same manner as when acquiring the NIR spectrum data in the method for obtaining the calibration curve regression equation of the present invention. That is, a NIR spectrum is measured using a commercially available near-infrared spectrometer and spectrum data, preferably Fourier transform spectrum data, is acquired. The spectrum data or the Fourier transform spectrum data may be the original spectrum data, but it is preferable to use the original spectrum data. Examples of data processing methods include, for example, first-order differentiation, second-order differentiation, third-order differentiation and the like (Derivative), smoothing (Smoothing), spectrum subtraction (Normalization), normalization (Normalization), MSC correction (Multiplicative). Scatter correction), SNV correction (standard normal variant correction), and the like. These processing methods may be used alone or in combination. Among them, preferable processing methods include SNV alone, secondary differentiation alone, a combination of primary differentiation and SNV correction, and a combination of primary differentiation and MSC correction.

  The predicted value of the FFA content is calculated by analyzing the NIR spectrum data of the test sample (preferably those processed in the previous section) and applying the previously obtained calibration curve regression equation. The analysis of the spectrum data here and the calculation of the predicted value (quantitative value) can be performed using commercially available software. Examples of the software to be used include The Unscrambler (Camo Software Japan Co., Ltd.), OPUS (Bruker Optics), NIRCal (Nippon Büch), and Pirouette (GL Science).

[Analyzing Method of Various Component Contents in Samples Containing Plant Oil and / or Plant as Raw Material Using HPLC-GPC Method]
In the process of obtaining the calibration curve regression equation of the present invention and the NIR quantification method of the present invention, the present inventors have made various components (FFA, A method for simultaneously analyzing the content of TAG, DAG, MAG, SE, ST, etc. (including rice bran and rice oil including OZ) using the HPLC-GPC method was completed. That is, the present invention simultaneously analyzes at least two components selected from the group consisting of FFA, TAG, DAG, MAG, SE, ST, and OZ of a sample containing vegetable oil and / or a plant body that is a raw material thereof. Provided is an analysis method characterized by using an HPLC analysis method using a GPC column (hereinafter referred to as “simultaneous analysis method of the present invention”).

By using the simultaneous analysis method of the present invention, FFA, TAG, DAG, MAG, SE, ST, and OZ, which are plant oils and / or plant components that are raw materials thereof, can be detected as separate peaks. Therefore, it becomes possible to simultaneously analyze two or more of these components, which was difficult with the conventional analysis method.
The simultaneous analysis method of the present invention can be carried out by the HPLC-GPC method described in the step (2) of the method for obtaining the regression equation of the present invention. In the simultaneous analysis method of the present invention, as in the step (2) of the method for obtaining the regression equation of the present invention, the sample is directly dissolved in the sample solvent, or the lipid concentrate is extracted by a known lipid extraction method, You may melt | dissolve in a sample solvent and use for HPLC-GPC method. The concentration of the component to be measured is not particularly limited as long as it can be quantitatively measured by the detector.

When components to be analyzed are limited in the simultaneous analysis method of the present invention, pretreatment for reducing components that are not to be analyzed may be performed as necessary. Pretreatment methods include precipitation / insoluble matter removal by centrifugation, dehydration treatment, degumming treatment, wax removal treatment by wintering, liquid-liquid extraction treatment using multiple solvents, saponification treatment, alkali treatment, distillation Treatment, decolorization treatment, solid-phase extraction treatment, and the like.
The solid phase extraction treatment and the use of the antioxidant are the same as in the case of using the HPLC-GPC method for the analysis of FFA in the step (2) of the method for obtaining the regression equation of the present invention.

The preparation used in the simultaneous analysis method of the present invention is not particularly limited, but among the components to be measured, those containing at least one molecular species contained as a main component in the sample are preferable. The preparation may be a commercially available preparation having a single composition, or a mixture of a plurality of molecular species may be used. Moreover, you may use as a standard what refine | purified only the component for the purpose of determination from the sample.
For example, when analyzing FFA using rice oil as a sample, oleic acid, linoleic acid, and palmitic acid, which are the main components of FFA contained in rice oil, may be used alone or as a mixture. Alternatively, FFA (molecular species mixture) separately purified from rice oil may be used as a sample. Similarly, when quantifying TAG of rice oil, a standard composition such as triolein may be used alone, mixed or used, or purified rice oil is used as a TAG standard. May be.

[Evaluation method of deterioration degree of sample containing vegetable oil and / or plant body as raw material]
The present invention provides a method for evaluating the degree of deterioration of a sample containing a vegetable oil and / or plant as a raw material using the FFA quantitative method of the present invention. The criteria for evaluation are not particularly limited. For example, in the case of rice bran, the degree of deterioration can be evaluated by the degree of increase in the FFA concentration in the rice bran after storage, compared to the FFA concentration in the rice bran immediately after milling. For example, in the case of rice oil, the degree of deterioration can be evaluated by the degree of increase in the FFA concentration in the rice oil after storage, compared to the FFA concentration in the rice oil immediately after purification. The present invention is very excellent in that FFA in rice bran can be quantified quickly, simply and non-destructively, and the degree of deterioration of the rice bran can be quickly evaluated.

[Method of evaluating rice bran stability]
By storing rice bran, evaluating the degree of deterioration using the above-described evaluation method, and recording the change over time, the deterioration rate can be obtained. This rate of deterioration is an indicator of rice bran stability. By selecting a variety with a lower rate of rice bran degradation, it is possible to select rice varieties that are resistant to oxidation and excellent in processability to refined rice oil.
Since the increase rate of FFA is affected by the TAG concentration, the degree of deterioration of the rice bran is low, and while the TAG concentration is sufficiently high, it progresses at a constant rate (initial speed). As the rate decreases, the rate of increase in FFA gradually decreases. The deterioration rate of rice bran is preferably evaluated based on the initial increase rate of FFA in a stage where the degree of deterioration of rice bran is low, and it is preferable that the rice bran at the start of the storage test has not deteriorated. The FFA concentration (per lipid concentrate) at the start of the storage test is preferably 10% by weight or less, and more preferably 5% by weight or less. Since the moisture content and oxygen concentration also affect the deterioration rate of rice bran, it is preferable to prepare samples so that the moisture content and oxygen concentration are comparable.
Although the storage conditions of rice bran are not particularly limited, it is preferably stored in a dark place, and is preferably stored in an environment where the humidity and temperature are constant using an incubator or the like.

[Measurement of quantitative patterns of lipid components in samples containing vegetable oils and / or plants as raw materials]
The present invention provides a composition analysis method for a sample containing vegetable oil and / or a plant body as a raw material thereof using the simultaneous analysis method of the present invention. According to the present invention, by one HPLC-GPC analysis, various components (FFA, TAG, DAG, MAG, SE, ST, OZ, etc.) in a sample containing vegetable oil and / or a plant body as a raw material thereof are analyzed. A quantitative pattern can be easily obtained.

  EXAMPLES Hereinafter, although an Example demonstrates this invention in detail, this invention is not limited to these.

[Example 1: Examination of acetic acid concentration in mobile phase of HPLC-GPC method]
First, the conditions of the acetic acid concentration in the mobile phase of the HPLC-GPC method for analyzing the main lipid components contained in rice crude oil, FFA, TAG, DAG, MAG, ST, SE, and OZ were examined. The following samples were used as samples and analyzed by the HPLC-GPC method described later.
<List of standard products>
Use oleic acid as FFA, triolein as TAG, glyceryl dioleate as DAG, glyceryl monooleate as MAG, β-sitosterol as ST, β-sitosteryl oleate as SE, and γ-oryzanol molecular species mixture as OZ It was.
<HPLC-GPC method>
(1) The sample was added to the sample solvent (toluene) and dissolved to 25 mg / mL.
(2) A differential refractive index detector (RI detector) was used as a detector for the HPLC-GPC method, and “Phenogel” (7.8 × 300 mm, particle size 5 μm) manufactured by Phenomenex was used as a column. The sample of (1) was analyzed individually with a column temperature of 60 ° C., a flow rate of 1.0 mL / min, and toluene having an acetic acid concentration of 0.25% (v / v) as a mobile phase.
(3) The same analysis was performed by changing the acetic acid concentration of the mobile phase to 0.003, 0.01, 0.03% (v / v).
The obtained RI chromatogram (displayed by superposing the chromatograms of the respective samples) is shown in FIGS. (A) is the result when the acetic acid concentration is 0.25% (v / v), (b) is the result when the acetic acid concentration is 0.03% (v / v), and (c) is the acetic acid concentration is 0. The result in the case of 0.01% (v / v), (d) is the result in the case where the acetic acid concentration is 0.003% (v / v). Note that TAG, DAG, MAG, and FFA appear as negative peaks under these conditions, but the chromatogram is shown with the positive and negative reversed for convenience.
As shown in FIGS. 1A to 1D, the FFA and ST peaks cannot be separated at an acetic acid concentration of 0.25% (v / v), but at an acetic acid concentration of 0.03% (v / v) or less, It was found that good separation of ST and ST was obtained. However, when the acetic acid concentration is less than 0.003% (v / v), the elution of FFA becomes very slow and overlaps with the peak of the sample solvent. Therefore, the appropriate concentration of acetic acid is 0.003 to 0.03% (v / V). When a detector that does not detect a solvent peak, such as an evaporative light scattering (ELSD) detector, is used as the detector, the peak broadens even if the acetic acid concentration is 0.003% (v / v) or less. It was possible to analyze.

[Example 2: Investigation of solvent used as sample solvent in HPLC-GPC method]
In Example 1, the sample was rice crude oil, and the sample solvent was a mixture of toluene and 2-propanol each having a concentration (0, 20, 50, 80, 100% (v / v)) of 2-propanol, and acetic acid in the mobile phase. The HPLC-GPC method was performed under the same conditions as in Example 1 except that the concentration was 0.01% (v / v). In order to identify the FFA, TAG, DAG, ST, SE, and OZ peaks in the crude oil from the obtained RI chromatogram, the sample used in Example 1 was analyzed under the same conditions.
The RI chromatograms of rice crude oil having various concentrations of 2-propanol are shown in FIGS. (A) is the result when the 2-propanol concentration is 0% (v / v), (b) is the result when the 2-propanol concentration is 20% (v / v), and (c) is the 2-propanol concentration. (D) is the result when the 2-propanol concentration is 80% (v / v), (e) is the concentration of 2-propanol 100% (v / v) Is the result of. For the sake of convenience, the RI chromatogram is shown with its polarity reversed in order to make the FFA peak, which is the main analysis target, positive.
As shown in FIGS. 2 (a) to 2 (e), when the 2-propanol concentration is 0% (v / v), the FFA peak cannot be separated from the unidentified peak and cannot be accurately quantified. At 100% (v / v), FFA and unidentified peak were separated and obtained as a single peak. In addition, since MAG and SE are hardly contained in US crude oil, they were not detected as peaks.
As an example, FIG. 3 shows a peak separation state of each component (FFA, TAG, DAG, ST, OZ) in an RI chromatogram when the 2-propanol concentration is 50% (v / v). As is apparent from FIG. 3, good separation was obtained for the peaks of the five components contained in the rice crude oil.

[Example 3: Composition analysis of rice crude oil by HPLC-GPC method]
The contents of FFA, TAG, DAG, ST, and OZ in rice crude oil were calculated from the data in FIG. It shows in Table 1 about the quantitative value (weight%) of each component. Hereinafter, a toluene / 2-propanol mixed solution having a 2-propanol concentration of 50% (v / v) as a sample solvent in the HPLC-GPC conditions in which the data of FIG. 3 was obtained, that is, the HPLC-GPC method of Example 1, The condition where the concentration of acetic acid in the mobile phase is 0.01% (v / v) is referred to as “HPLC-GPC method of Example 3”.

  As is clear from Table 1, by optimizing the conditions of the HPLC-GPC method, the content of TAG and its decomposed products (DAG, FFA) and typical functional components of rice crude oil are quantified in a single analysis. Became possible.

[Example 4: Comparison of chemical analysis values by measurement method of free fatty acid]
For various rice crude oils whose FFA concentration was precisely analyzed by the LC / MS method described later, the FFA was quantified by the HPLC-GPC method of Example 3 and a known alkali titration method. Correlation was determined.
<FFA precision measurement by LC / MS method>
(1) Rice crude oil was dissolved in 2-propanol to prepare an analytical sample.
(2) The analysis was performed by the LC / MS method under the following conditions. An ODS column was used as a separation column, and methanol / acetic acid = 99/1 (v / v) was used as a mobile phase. The column temperature was 40 ° C. and the flow rate was 0.45 mL / min. Using an LC / MS detector, four types of oleic acid, linoleic acid, palmitic acid, and stearic acid were set and analyzed for each molecular species to obtain a chromatogram of the monitor ions.
(3) Each sample was analyzed under the same conditions as in (2) and quantified with monitor ions.
(4) The total value of the quantitative values of the four fatty acids was calculated as a precision analysis value of the fatty acid content.

  The results are shown in FIGS. 4 (a) and (b). (A) is the correlation between the analytical value by the HPLC-GPC method and the precise analytical value, and (b) is the correlation between the analytical value by the alkali titration method and the precise analytical value. As can be seen from FIGS. 4 (a) and 4 (b), the HPLC-GPC method shows a higher correlation with the precision analysis value than the alkali titration method, and the accuracy as a simple measurement method of FFA is excellent. Became clear.

[Example 5: Examination of influence of OZ concentration on chemical analysis value]
Deoxidized rice crude oil (rice crude oil subjected to distillation deacidification treatment and alkali deoxidation treatment according to a conventional method to make the FFA concentration 0.1% by weight or less) and rice with no OFA content of about 30% by weight Oil (“germ oil gamma 30” (trade name), manufactured by Tsukino Food Industry Co., Ltd.) is mixed so that the final concentration of OZ is 10 to 20% by weight, and contains a high concentration of OZ and a small amount of FFA. Samples were prepared. The FFA content of this model sample was measured by the HPLC-GPC method and alkali titration method of Example 3.
The relationship between the added amount of OZ and the quantitative value of the FFA content is shown in FIG. In the HPLC-GPC method, there was no change in the FFA quantitative value even when the OZ concentration in the sample increased, but in the alkali titration method, the titer increased as the OZ concentration in the sample increased, and the FFA quantitative value. Increased. From this result, when measuring the FFA content of the oil and fat sample containing OZ which is a characteristic component of rice oil, the HPLC-GPC method is more suitable than the alkali titration method which may erroneously detect FFA. It became clear.

[Example 6: Preparation of FFA content calibration curve regression formula of rice bran]
(1) As a sample for obtaining a calibration curve regression equation, rice bran obtained by milling brown rice was stored at 30 ° C. and sampled over time to obtain rice bran having different degrees of deterioration.
(2) Using a FT-type near-infrared analyzer MPA (Bruker Optics) as the NIR spectrum analyzer, the NIR diffuse reflected light of rice bran of ( ¹ ) is in the range of 800 nm to 2500 nm (12500 to 4000 cm ⁻¹ ). It was measured.
(3) The lipid concentrate was extracted by Soxhlet extraction method using diethyl ether from the rice bran obtained from the spectrum, and the FFA content (per lipid concentrate or per rice bran) was determined using the HPLC-GPC method of Example 3. Quantified.
(4) Preprocessing (SNV correction) of the original spectrum obtained in (2), and using the wavelength data of the preprocessed spectrum (1333-1640, 2173-2355 nm) as explanatory variables, the measured FFA content (lipid concentrate) An FFA calibration model of rice bran was created by PLS regression analysis with the objective variable being hit or per rice bran. The correlation between the predicted NIR value (quantitative value) and the chemical analysis value (actual measurement value) is shown in FIGS. 6 (a) and 6 (b). (A) is a result about the FFA content per lipid concentrate, (b) is a result about the FFA content per rice bran. The NIR predicted value (% by weight, quantitative value) here is a calculated value derived by the measuring device from the calibration curve regression equation of the FFA content (per lipid concentrate or per rice bran) and NIR spectrum data. The quantitative value and the actual measurement value showed a high correlation (R ² = 0.955 and R ² = 0.962, respectively), and it became clear that this calibration curve regression equation has high prediction accuracy. From the PLS regression analysis results, five regions of 1435 nm ± 10 nm, 1662 ± 10 nm, 1762 ± 10 nm, 2162 ± 10 nm, and 2245 ± 10 nm are listed as important wavelength data related to the FFA content.

[Example 7: Preparation of FFA content calibration curve regression formula of lipid concentrate of rice bran]
(1) The rice bran lipid concentrate obtained in (3) of Example 6 was used as a measurement sample.
(2) NIR transmitted light of the sample was measured in the range of 800 nm to 2500 nm (12500 to 4000 cm ⁻¹ ) using an FT type near infrared analyzer MPA (Bruker Optics). In addition, in order to homogenize the fat and oil sample, the measurement cell was heated to 75 ° C. and the temperature stability was confirmed, and then the NIR spectrum was acquired.
(3) The FFA content was quantified using the HPLC-GPC method of Example 3 for the lipid concentrate from which the spectrum was obtained.
(4) Pre-processing (SNV correction) of the original spectrum obtained in (2), wavelength data of the pre-processed spectrum (1639 to 2175 nm) as explanatory variables, and PLS regression analysis with FFA content as an objective variable, A calibration model of the FFA content (per lipid concentrate) of the lipid concentrate was created. FIG. 7 shows the correlation between the predicted NIR value (quantitative value) and the chemical analysis value (actual measurement value). The predicted NIR value (weight%, quantitative value) here is a calculated value derived by the measuring device from the calibration curve regression equation of FFA content and NIR spectrum data. The quantitative value and the actually measured value showed a very high correlation (R ² = 0.99), and it became clear that this calibration curve regression equation has high prediction accuracy.

[Example 8: Examination of optimum wavelength range]
For the rice bran sample of Example 6, FFA calibration models were created by changing the wavelength range of the wavelength data of the preprocessed spectrum in various ways. The coefficient of determination R ² in all conditions of the obtained calibration models are shown in Table 2. The wavelength indicates the wavelength range used, and one section (Condition Nos. 4, 6-8, 10, 16-21) or two sections (Condition Nos. 1-3, 5) divided by the minimum wavelength and the maximum wavelength. , 9, 11-15). As can be seen from Table 2, condition no. In 1 to 16, a high correlation (R ² > 0.90) was observed between the NIR predicted value (quantitative value) and the chemical analysis value (actual measurement value). These highly correlated conditions include two or more regions of 1435 nm ± 10 nm, 1662 ± 10 nm, 1762 ± 10 nm, 2162 ± 10 nm, and 2245 ± 10 nm, which are wavelength regions related to the FFA shown in Example 6. There was something in common.

[Example 9: Correlation between chemical analysis value of FFA content of solvent extract of rice crude oil and predicted NIR value]
Rice crude oils having different FFA contents were extracted with various solvents (chloroform, acetonitrile, t-butanol, acetone, tetrahydrofuran, dimethyl sulfoxide), and NIR transmitted light spectra were measured as 1% (w / v) solutions. The correlation between the chemical analysis value of FFA content (per extract) by the HPLC-GPC method of No. 3 and the predicted value by NIR was examined. It shows the relationship of the correlation coefficient R ² and solvent shown in Table 3. There is a large difference in the correlation coefficient depending on the type of solvent used, and high correlation is obtained with chloroform and acetonitrile. Therefore, when diluting and measuring rice crude oil, chloroform and acetonitrile are suitable as solvents. It was.

[Example 10: Rice bran deterioration evaluation test]
Immediately after milling, rice bran having an FFA content (per lipid concentrate) of about 5% by weight was stored in an incubator at 30 ° C. Rice bran was sampled over time, and the FFA content (per lipid concentrate) was measured using the HPLC-GPC method of Example 3. FIG. 8 shows changes with time in the FFA content (per lipid concentrate) in rice bran. When the FFA content (per lipid concentrate) is about 5 to 10% by weight, it increases linearly with respect to the storage time, but thereafter increases in a gentle curve. When the degradation rate was evaluated by the increase in the FFA content from 0 to 96 hours with linearity, it was 0.056% by weight / h per lipid concentrate.

[Example 11: Comparison of chemical analysis value and NIR prediction value in degradation assessment test of rice bran]
Immediately after milling, rice bran having an FFA content (per lipid concentrate) of about 5% by weight was stored in an incubator at 40 ° C. The rice bran is sampled over time, and the NIR diffuse reflected light of the rice bran is in the range of 800 nm to 2500 nm (12500-4000 cm-1) using an FT type near infrared analyzer MPA (Bruker Optics) as an NIR spectrum analyzer. Measured with The measured NIR spectrum was applied to the calibration model of the FFA content (per lipid concentrate or per rice bran) of rice bran prepared in Example 6 to predict the FFA content. Further, the FFA content (per lipid concentrate or per rice bran) was measured using the HPLC-GPC method of Example 3. Changes over time in the FFA content of rice bran are shown in FIGS. 9 (a) and 9 (b). (A) is per lipid concentrate of rice bran, and (b) is the FFA content (% by weight) per rice bran. When the degradation rate is evaluated by the linear increase in FFA content from 0 to 96 hours, the NIR predicted value per lipid concentrate is 0.054% by weight / h, and the chemical analysis value is 0.055% by weight / h. For rice bran, the predicted NIR value was 0.011% by weight / h, and the chemical analysis value was 0.012% by weight / h, both of which showed no significant difference.

Claims (8)

A method for obtaining a regression equation for quantifying free fatty acids in a sample containing rice bran and / or rice oil , comprising the following steps (1) to (3).
(1) A step of measuring a near infrared light spectrum of a sample, a solvent extract of the sample, or a lipid concentrate obtained by removing the solvent from the solvent extract (2) Free fatty acids in the sample subjected to step (1) Step of specifically quantifying by HPLC analysis using non-aqueous GPC column (3) Spectral data obtained in all or part of the wavelength range in which the near-infrared light spectrum was measured, and free fatty acid The process of analyzing quantitative values by multivariate analysis method and determining factors related to free fatty acids The regression equation according to claim 1 , wherein the composition of the mobile phase used in the HPLC analysis method using a non-aqueous GPC column is toluene containing 0 to 0.03% (v / v) acetic acid. Method. Sample solvent used in the HPLC analysis method using a non-aqueous GPC column is a mixture of toluene and alcohol, according to claim 1 or 2, wherein the alcohol concentration is 20% (v / v) or higher How to get a regression equation. In step (1), the sample or lipid concentrate is dissolved in a solvent selected from chloroform, acetonitrile, t-butanol and dimethylsulfoxide, and the near-infrared spectrum of the resulting solution is measured. The method of obtaining the regression equation in any one of claim | item 1 -3 . Wavelength region in step (3) is any of 1435 ± 10nm, 1662 ± 10nm, 1763 ± 10nm, 2162 claims 1-4, characterized in that it comprises at least one region selected from ± 10 nm and 2245 ± 10 nm A method for obtaining a regression equation as described above. Near infrared light spectrum, a method of obtaining a regression equation according to any one of claims 1 to 5, characterized in that a spectrum of diffuse reflection light or transmitted light. A method for quantifying free fatty acids in a sample containing vegetable oil and / or plant as a raw material thereof, the regression equation obtained by the method for obtaining the regression equation according to any one of claims 1 to 6 , and A method for quantifying free fatty acid, comprising using near-infrared spectrum data of a test sample in a wavelength region used for preparing a regression equation. A method for evaluating the degree of deterioration of a sample containing vegetable oil and / or a plant body as a raw material thereof, wherein the method for quantifying free fatty acids according to claim 7 is used.