Method for predicting raw material function in final product

Abstract

A method is proposed for predicting how the addition of unknown raw materials will affect the predictable and realizable properties of the final product.

Description

Field of the Invention

[0001] The present invention belongs to the field of food technology and relates to a method for predicting raw material functions in final products - in particular dairy products, mainly milk-based products, including cheese, their plant-based milk substitutes and corresponding mixed products - by collecting defined analytical and raw material data and performing statistical evaluations. Technical Background

[0002] The vegan diet trend has arrived in Germany. Around 1.3 million people follow a vegan diet. In contrast, around 8 million people eat a vegetarian diet. But this number is increasing every day. This has also increased interest in plant-based raw materials that are suitable alternatives to animal products. In the dairy sector, people are looking for alternatives for the important milk proteins. For example, candidates are plant proteins, such as those obtained from peas, beans or oats. However, the following aspect is problematic: Milk proteins cannot be replaced with any other type of protein in a 1:1 ratio, because on the one hand, the formula uses milk as a protein source, and on the other hand, proteins can have very different structures and functions. The same applies to the replacement of animal and / or plant macronutrients, such as fats and carbohydrates. Another aspect that can affect the functionality of the raw materials in the final product is that the raw material production process varies from supplier to supplier, resulting in different raw material qualities and batches. Therefore, when exchanging the same weight of raw materials, the resulting product may suddenly have a floury, rubbery or dull texture instead of a structured texture.

[0003] To address this problem, it has hitherto been necessary to first incorporate potentially suitable candidates into various end products and examine the resulting desired properties. Given the fact that the market now offers hundreds of different raw materials and additives to influence dairy and plant-based end product composition, structure and stability, the necessary pre-selection is very time-consuming. Therefore, a procedure is needed that allows for rapid prediction of the expected properties of the end product. Prior art

[0004] Evaluation methods are known in the prior art and can be used, for example, to predict the shelf life of food in packaging. For example, EP1626275B1 (WILD) proposes a method for evaluating the shelf life of a product in a packaging material, the method comprising the following steps: a) placing a sealed packaging material (4) filled with a filling material into a pressurized chamber (1); b) storing the packaging material (4) in a test gas atmosphere at an overpressure p1 and a specific temperature T1 in a pressurized chamber (1) for a period of time t1, wherein a defined amount Q of the test gas penetrates into the packaging material; and c) storing the packaging material (4) under the influence of heat and / or light for a specific time t2, and d) Analytical and / or sensory testing of filling materials.

[0005] A disadvantage of this method is the high equipment costs. In addition, the functional principle - simulation of the amount of oxygen introduced into the filling material and the acceleration of the reaction by increasing the pressure - ultimately allows only rough estimates, which are generally not suitable for qualitative statements, especially since it is known that for many polymer / gas systems the solubility and diffusion coefficient of the penetrating substance in the polymer (e.g. plastic bottles) depends on the concentration and pressure if the tests are carried out below the glass transition temperature Tg of the polymer and / or below the critical temperature of the gas [Müller, K. "Sauerstoff- von Kunststoffflaschen und Verschlüssen"Dissertation TU München (2003)].

[0006] Completely different approaches to visually indicate the shelf life of packaged food products to consumers are the subject of the patents EP 2378354 B1, EP 2581785 B1 and WO 2009 056591 A1 (UNI MüNSTER). These involve sensor devices, in particular electrochemical processors, which are short-circuited when the package is opened. An aluminum oxide layer is formed, the progression of which is proportional to the shelf life of the product. For example, an increase in temperature accelerates the formation of the oxide, which naturally also reflects a decrease in the freshness of the product. However, this solution requires a separate sensor for each individual package, which is also technically complex.

[0007] EP 3875953 B1 (DMK) describes an accelerated storage test for determining the quality and shelf life of food products, wherein a first sample is stored at a temperature T1 and a similar second sample is stored at a significantly higher temperature T2, whereby the samples are subjected to a sensory evaluation and compared with one another via a calibration curve. Since the sensory evaluation of products always contains a subjective component, a method based entirely on objective parameters would be more advantageous.

[0008] None of these existing testing methods address the early, targeted prediction of raw material functionality in the final product. Purpose of the Invention

[0009] The object of the present invention is therefore to provide a test method with which, by correlative observation and statistical evaluation of different raw material properties, a reliable prediction of the raw material functionality in end products such as dairy products, predominantly milk-based products, including cheese, their plant-based milk substitutes and corresponding mixed products can be made in a reproducible manner at an early stage. Summary of the invention

[0010] The present invention relates to a method for predicting the functionality of raw materials in a final product - a dairy product, a product mainly based on milk, including cheese, their plant-based milk substitutes and corresponding mixed products - comprising or consisting of the following steps: (a) providing a set of raw materials of different qualities; (b) establishing a set of test parameters; (c) applying the set of test parameters to each individual sample from step (a) to generate a data set; (d) incorporating the raw materials of different qualities from step (a) into final products, such as dairy products, mainly milk-based products, including cheese, their plant-based milk substitutes and corresponding blended products; (e) record and evaluate product characteristics; (f) Use principal component analysis (PCA) to check the correlation of the data in the data set, obtain a coordinate system (score map), and determine the largest variance of the data as the first component, and the second largest variance of the data as the second component; (g) assigning the product characteristics obtained in step (e) to the data points in the coordinate system of step (f); and (h) Identify clusters in the score plot to identify raw materials that, when used in the final product, will produce similar or identical desired product characteristics.

[0011] Surprisingly, it was found that with the aid of the method according to the invention it is possible to quickly, reliably and reproducibly assess in advance the different properties and functionalities of new raw materials when incorporated into different dairy products, predominantly milk-based products, including cheese and their plant-based milk alternatives, and corresponding mixed products, based on the physicochemical properties that were primarily recorded during the initial testing of the raw materials, but not used for the described predictions.

[0012] The method according to the invention is divided into four parts: -Choose appropriate analysis parameters -Generate data - PCA was used to verify the correlation and - Identify clusters in the score plot to identify raw materials that produce similar or identical desired product properties when used in a product. Dairy products, primarily and theoretically milk-based products, including cheese or their plant-based milk alternatives and corresponding blended products

[0013] The choice of product to which the method according to the invention can be applied is not important per se. Typically, the product is selected from the group consisting of whole milk, skimmed milk, UHT milk, milk powder, cream, curd, cheese and yogurt, but may also be a product mainly based on milk, such as a variety of desserts. Also included are their plant-based milk substitutes, such as plant-based beverages, spreads, cream substitutes, yogurt substitutes and cheese substitutes. The process can also be applied to mixed products, in which only part of the animal milk component is replaced by a plant-based substitute, or vice versa (substituting a plant-based component with an animal-based raw material or ingredient). This list is not exhaustive. raw materials

[0014] The raw materials whose properties and functions are to be predicted in the final product can be selected from the group consisting of raw materials such as proteins, fats, carbohydrates, and semi-finished products such as acid whey, milk permeate and milk retentate, their mixtures and their corresponding plant substitutes or raw materials. They can be in solid, powdered form, or in liquid and concentrated form.

[0015] Typically, a raw material set of 5 to 50, preferably 7 to 15 raw materials is needed for the desired prediction of the raw material functionality in the final product, which preferably results in at least three different final product properties. Test Parameters

[0016] An essential feature of the method according to the invention is the selection of the test parameters based on which the sample is evaluated. For this purpose, the following factors are preferably taken into account: - pH value; - solubility; - Thermal behavior of powders (DSC); - Particle size distribution (D10, D50, D90, specific surface area); - Fat absorption capacity (rapeseed oil); - Water absorption capacity; - dissolved oxygen content; - Lumisizer; - Lightness color value white / black (L*); - Color coordinates red / green (a*); - Color coordinates yellow / blue (b*); - Free amino nitrogen (PAN); - ammonia content; - urea content; - Calcium content; -acidity; - Rheological behavior; - Robustness; and -Degree of syneresis.

[0017] DSC data are the results obtained from differential thermal analysis. This is a thermal analysis method that measures the amount of heat released or absorbed by a sample during heating, cooling or isothermal processes. A packaged container containing the sample (5-40 mg) and a second identical container without contents (reference) or reference sample are exposed to the same temperature change program in a temperature chamber. Due to the heat capacity of the sample and any exothermic or endothermic processes or phase changes (such as melting or evaporation), there is a temperature difference between the sample and the reference, because heat energy flows into or out of the sample during the study. DSC is mainly used to determine the degree of crystallinity, melting or decomposition points, and the degree of protein denaturation. Among them, the temperature at the beginning (start) and end (end) of the heat-induced reaction, the temperature at which the reaction reaches a maximum value (peak), and the height of the peak are measured. The reaction enthalpy is also determined.

[0018] The specific surface area and particle size distribution of the powders were determined using static laser scattering (Horiba LA-960). For the purposes of the method according to the invention, the specific surface area of ​​the powder is a volume-related quantity Sv in the units of m2 / m3. The mass-related specific surface area SM is usually determined, since this is easier to perform experimentally. The two values ​​can be converted using the factor p SV ＝ ρSM

[0019] The particle size distribution includes the specification of the D10, D50 and D90 values, ie the definition of the particle size range that contains 10%, 50% or 90% by weight of all the powder particles.

[0020] The fat absorption capacity (here the absorption capacity of rapeseed oil) and the water absorption capacity are related to the swelling capacity of the powder and are specified in grams of fat or grams of water per gram of powder.

[0021] The lightness color value L* and the two color coordinates a* and b* belong to the so-called L*a*b* color system, also known as the CIELAB system. It is the most commonly used color measurement system today and is widely used in almost all fields of application. It was defined by the CIE in 1976 as one of the equidistant color spaces in order to solve the main problem of the original Yxy system: equal geometrical difference distances in the x,y color triangle do not lead to equal color differences in terms of sensitivity. The color space of the L*a*b* system is characterized by the lightness L* and the color coordinates a* and b*. The sign indicates the color direction: +a* represents the red component, while -a* points to green. Thus, +b* represents yellow and -b* represents blue. The origin of the coordinates (the intersection of the axes) is a neutral gray without any chromaticity. As the a*b* value increases, i.e. the color position is further from the center, the chromaticity increases. The CR 400 or 410 device has proven itself to be able to determine color values ​​in the context of the method according to the invention. More information can be found in the consumer information "Exact colorcommunications" from Konica Minolta.

[0022] Preferably, the set of test parameters should contain at least 5, preferably at least 7 parameters. Principal Component Analysis (PCA)

[0023] The data pool obtained in this way is evaluated in a further step by means of a so-called “principal component analysis” (PCA).

[0024] This is an established statistical method for analyzing large data sets with a large number of parameters per observation, for improving the interpretability of the data while retaining a maximum amount of information, and for visualizing multidimensional data. Formally speaking, PCA is a statistical technique for reducing the dimensionality of a data set. This is done by linearly transforming the data into a new coordinate system in which (most of) the variation in the data can be described with fewer dimensions than the original data. In the method according to the invention, only the first two principal components are used to represent the two-dimensional data and to visually identify clusters of closely related data points.

[0025] In PCA data analysis, the first principal component of a set of P variables, assuming a joint normal distribution, becomes the derived variable, which is formed as a linear combination of the original variables and explains the maximum variance. The second principal component explains the maximum variance remaining when the effect of the first component is eliminated, and we can continue with P iterations until all variance is explained. PCA is most often used when many variables are highly correlated with each other and their number needs to be reduced to independent sets. Therefore, the first principal component or the second principal component can be defined as the direction that maximizes the variance of the projected data. PCA is the simplest truly eigenvector-based multivariate analysis and is closely related to factor analysis. A detailed description of the method can be found, for example, at https: / / en.wikipedia.org / wiki / Principal_component_analysis.

[0026] In short, PCA provides a coordinate system in which the component with the largest variance or deviation for each data point (the "first component") is determined relative to the component with the second largest variance (the "second component"). In the final step, each data point in the coordinate system is now assigned a corresponding previously determined product characteristic.

[0027] Surprisingly, it was found that this led to the formation of clusters that could be clearly associated with the properties and functions to be studied. Properties and functions that can be studied or predicted in this way include texture, creaminess, synergistic effects and combinations of two, three or more of these properties.

[0028] This means that only the test parameters mentioned at the beginning of the unknown raw material sample need to be analyzed in order to be able to assign coordinates to it with the help of PCA. Based on the position of the coordinates, the expected raw material function in the final formulation can be deduced quickly, reliably and, if necessary, automatically. Therefore, faster simulations based on initial values ​​and future predictions as well as selection criteria for raw material selection at a faster speed by AI (artificial intelligence) are obviously conceivable and implementable as a second stage. In addition, early and rapid selection of suitable alternative raw materials is possible in any industrial implementation. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 The third step of the method according to the invention is shown.

[0030] Figure 2 A score plot of a cluster of substances with desired properties is shown. Example 1A / B Parameter selection and data generation

[0031] Hereinafter, the method according to the invention is explained based on the evaluation of various plant-based powders. For this purpose, the protein fraction was first extracted from the fruits of various plants (peas, beans, chickpeas, oats) and then made into a powder state. The following parameters were then selected from the available parameters: -pH -Solubility -DSC data (start value, final value, peak, peak height, reaction enthalpy) -Particle size distribution (D10, D50, D90, specific surface area) - Fat absorption capacity (rapeseed oil) -Water absorption capacity - Color coordinates (L*, a*, b*) The corresponding measured values ​​were determined; these values ​​are summarized in Table 1. Various samples of the plant-based powders were then blended into a standard recipe of vegan plant-based quark and the texture of the resulting product was determined. The results can also be found in the last column of Table 1. Example 1C Use PCA = Principal Component Analysis to check for correlation in the data

[0032] Figure 1 The third step of the method according to the invention is shown. The previously obtained data set is processed with the aid of principal component analysis. PCA provides a coordinate system in which the scores of the first two principal components are plotted for each data point. Figure 1 In the example above, principal component 1 explains 40.8% of the variance. The second principal component explains 26.1%. Overall, PCA explains 66.9% of the variance in the original data. Knowing which parameters characterize the two principal components is irrelevant to this procedure, as this step only involves adjusting the variance to a scalable size.

[0033] As can be seen in the figure, the data points are more or less evenly distributed over the coordinate system. However, after assigning the actual found textures to the data points, it becomes clear that the figure contains clusters of identical or similar structures. Therefore, the desired structured texture is only found in the x-coordinate range from -2.5 to 2.5 and in the y-coordinate range from +1 to +3. Samples that do not belong to this cluster are not structured; the texture here ranges from powdery, creamy to rubbery.

[0034] therefore Figure 1It can be understood as a raw material selection chart. In order to be able to predict the expected functional properties of the raw materials, it is only necessary to record the physicochemical data according to the selected test parameters - these parameters are usually either given by the specification or recorded during the incoming material inspection - and calculate the corresponding data points as described above. The position of the data points in the chart can then be used to draw conclusions about the expected raw material function in the end product. Example 2

[0035] Below, the method according to the invention is explained based on the evaluation of 17 different milk protein powders;

[0036] Select the following parameters from the available parameters: -pH -Solubility -DSC data (start value, final value, peak, peak height, reaction enthalpy) -Particle size distribution (D10, D50, D90, specific surface area) - Fat absorption capacity (rapeseed oil) -Water absorption capacity - Color coordinates (L*, a*, b*) The corresponding measured values ​​were determined; these values ​​are summarized in Table 2.

[0037] The various samples of milk protein powder were then blended into a standard recipe of milk powder with a high protein content and the texture of the resulting product was determined. The results can also be found in the last column of Table 2. PCA was then performed according to Example 1C. The score plot of the material cluster with the desired properties is shown in Figure 2 shown.

Claims

1. A method for predicting the functionality of raw materials in a final product, such as a dairy product, a product mainly based on milk, including cheese, their plant-based milk substitutes and corresponding blended products, comprising or consisting of the following steps: (i) providing a set of raw materials of different qualities; (j) establishing a set of test parameters; (k) applying the set of test parameters to each individual sample from step (a) to generate a data set; (l) incorporating the raw materials of different qualities from step (a) into final products, such as dairy products, mainly milk-based products, including cheese, their plant-based milk substitutes and corresponding blended products; (m) record and evaluate product characteristics; (n) Use principal component analysis (PCA) to check the correlation of the data in the data set, obtain a coordinate system (score map), and determine the largest variance of the data as the first component, and the second largest variance of the data as the second component; (o) assigning the product characteristics obtained in step (e) to the data points in the coordinate system of step (f); and (p) Identify clusters in the score plot to identify raw materials that, when used in a final product, produce similar or identical desired product characteristics.

2. The method according to claim 1, wherein the final product used is a dairy product, mainly milk-based products, including cheese and their plant-based milk substitutes and corresponding mixed products, which are selected from the group consisting of whole milk, skimmed milk, UHT milk, milk powder, cream, curd, cheese and yogurt, cheese and its plant-based substitutes and applications such as plant-based beverages, spreads, cream substitutes, yogurt substitutes and cheese substitutes.

3. The method according to claim 2, wherein dairy and / or vegetable based raw materials / ingredients are used, which are selected from the group consisting of proteins, fats, carbohydrates, acid whey, milk permeate and milk retentate, and mixtures and concentrates thereof.

4. The method according to claim 1, wherein a raw material sample set including 5 to 50 samples is used.

5. The method according to claim 1, wherein the sample set used contains at least three raw materials of different qualities.

6. The method according to claim 1, wherein a set of test parameters is used, which contains at least two of the following parameters: - pH value; - solubility; - Thermal behavior of powders (DSC); - Particle size distribution (D10, D50, D90, specific surface area); - Fat absorption capacity (rapeseed oil); - Water absorption capacity; - dissolved oxygen content; - Lumisizer; - Lightness color value white / black (L*); - Color coordinates red / green (a*); - Color coordinates yellow / blue (b*); - Free amino nitrogen (PAN); - ammonia content; - urea content; - Calcium content; -acidity; - Rheological behavior; - Robustness; and -Degree of syneresis.

7. The method according to claim 6, wherein a set of test parameters is used, which comprises at least 5, preferably at least 7 parameters.

8. The method according to claim 1, wherein the predicted property is selected from the group consisting of texture, creaminess, syneresis, mouthfeel, hardness, emulsion stability, synergistic effect and a combination of two, three or more of these properties and prediction of their application properties.

Images