Keywords

Introduction

In accordance with GOST 22391-2015 [1], the oil-and-fat industry identifies the seed oil content as the main quality indicator of raw materials during their gathering and processing. The oil content comprises the content of raw fat and accompanying lipoid substances, which, together with fat, pass into the ether extract from the seeds. In accordance with GOST 10857-64 [2], the determination of this indicator is based on the method of Soxhlet exhaustive extraction. A drawback of this method consists in its duration, application of toxic chemical solvents, low productivity, and high requirements to the personnel qualification.

At present, due to the wide development of technical equipment and information technologies, instrumental methods, such as IR spectroscopy and pulsed NMR methods, which can be used to obtain information about the seed oil content, are quite common. The determination of oil content using the method of IR spectroscopy is typically performed in accordance with GOST 32749-2014 [3]. In the case of sunflower seeds, this method is unpopular due to significant errors (up to 2 abs%) and the dependence of measurement results on the seed appearance and their fineness degree. In addition, this method requires the calibration of IR analyzers using a large number of seeds with a certain value of determined indicators obtained using referee methods [4,5,6,7,8].

In the contemporary Russian oil-and-fat industry, the most widespread instrumental method for determining the oil content of seeds and products of their processing involves pulsed nuclear magnetic resonance (NMR). This is explained by its analytical simplicity, lack of a complex sample preparation, high accuracy (error of not more than 0.6 abs%, which is comparable in accuracy with a referee chemical method), promptness (the analysis of one sample takes < 30 s), lack of the effect caused by the operator’s subjectiveness on the analytical results, as well as ease of implementation. Currently, AMV-1006M (V.S. Pustovoit All-Russian Research Institute of Oil Crops, VNIIMK, Russia) is the most common NMR analyzer used at Russian enterprises. The determination of oil content using this instrument is based on the substantiated scientific and methodological approach with the rank of 0.04 [9].

The calibration of operating NMR analyzers for the determination of oil content in sunflower seeds is carried out using GSO 3107-84Footnote 1 state reference materials [10, 11], which were used for the verification of AMV-1006M NMR analyzers until the expiration of their Type Approval Certificate in 2016.

These reference materials (RM) are made on the basis of chemically inert substances-imitators, thermally resistant, resistant to thermal oxidation and ultraviolet effects, as well as having the dielectric properties (of organosilicon liquids) [12,13,14] of oil NM relaxation characteristics.

Imitated values of indicated GSO 3107-84 were assigned using an NMR analyzer, whose calibration and testing were performed using sunflower seeds with the oil content determined by the Soxhlet method of exhaustive extraction. The drawbacks of this calibration method include its duration, high requirements to the operator’s qualification, as well as the application of specially prepared traceable seeds. It should be noted that even specially prepared seeds have a scattering in the measured oil content values in individual samples, isolated from one specimen using the method of exhaustive extraction to 1 abs%. This is explained by the fact that seeds represent a natural heterogeneous object [15, 16].

Let us note that one of the key features of the NMR method comprises a functional dependence between the oil content in the analyzed seed sample and the amplitude of the NMR signal obtained from the protons of triacylglycerins (TAG) contained in the oil.

At present, the NMR method is widely used for determining the oil content of seeds and oil-containing raw materials [17,18,19,20,21,22,23,24],Footnote 2 which is explained by its application simplicity, high metrological characteristics, sample preparation simplicity, ecological safety, as well as the non-destructive nature of analysis. More than 350 enterprises of the Russian oil-and-fat industry currently use AMV-1006M NMR analyzers to quickly obtain information on the oil content at all stages of gathering, storage, and processing of oil seeds.

At the same time, the calibration of quantitative NMR analyzers represents a complex problem.

Currently, several main methods of calibrating quantitative analyzers, used for the quality assessment of lipid-containing raw materials, can be found in literature:

  • using natural samplesFootnote 3,Footnote 4;

  • using imitator RMs produced from chemically inert substances [13, 14, 24];

  • using samples, obtained on the basis of natural components, e.g., oil cakes and oil (see footnote 3) [25].

In this study, we aim to scientifically and experimentally substantiate the application of sunflower oil samples in the calibration of NMR analyzers.

Materials and Methods

Experiments were carried out at the Central Experimental Base of V.S. Pustovoit All-Russian Research Institute of Oil Crops (Krasnodar) in 2021–2022. Traceable sunflower seed samples of VNIIMKFootnote 5 breeding and two samples of commercially distributed oil were prepared, including the refined and deodorized oil of the “BLAGO” trademark (Russia) and non-refined oil of the “STAVROPOL’E” trademark (Russia), acquired in a retail network.

Sunflower samples were cleaned off damaged seeds and waste admixtures. The oil and moisture content of seeds was preliminarily determined using an AMV-1006M NMR analyzer in accordance with GOST 8.597-2010 [26]. The seed samples presented in the study belong to contemporary high-productive varieties. The acid number of the seed oil was determined according to GOST 31933-2012 [24, 27] using the titrimetric method with visual indication.

The determination of the oil content in seeds was carried out by the method of exhaustive extraction in four repetitions for each sample in order to reduce the measurement error, in accordance with the procedure developed based on GOST 10857-64. The duration of extraction was 24 h, at a temperature providing a number of siphonings from 7 to 10 per hour. The completeness of extraction was checked using a watch glass sample. The average value of four measurements was taken as the final result.

The oil from sunflower seeds was obtained using a Laboratoroff PR-L laboratory hand press (LLC Eltemix, Russia) with a force of 12 tons. The obtained oil was further filtered to remove seed particles, trapped in the oil during its pressing, using the FS laboratory filtering paper.

In order to construct the oil content calibration curves, from each oil sample, both obtained by pressing and acquired at a retail network, five weighments evenly distributed in the range from 2000 to 7000 g with the accuracy of 0.001 g were taken using an AND HK-50AG laboratory weights (AND, Japan).

Prior to measuring the oil proton NMR signals, the prepared samples were thermostated at a temperature of 23 ± 0.5 °C for 2 h in a TVL-K(50)B thermostat (CJSC INSOVT, Russia). The thermostating of samples comprises an important stage, since the temperature has the essential effect on the NM relaxation characteristic of oil samples [21].

Measurements were conducted using typical AMV-1006M NMR analyzers of the oil and moisture content (VNIIMK, Russia). In accordance with the analytical procedure, when performing measurements by an AMV-1006M NMR analyzer, the volume of each analyzed seed weighment is equal to 25 cm3.

For the control of NMR analyzers and primary processing of NMR signals, obtained using the studied oil samples, the own software was used [28]. The obtained data were statistically analyzed using Statistica and Excel software.

Results and Discussion

The main characteristics of sunflower seeds used in the study are presented in Table 1.

Table 1 Characteristics of the studied sunflower seeds
Full size table

It is known [2] that, for oilseeds with a moisture content of < 8%, the intensity of NM relaxation characteristics is determined exclusively by the protons of the oil and accompanying substances. This is due to the high degree of connection between water molecules and the protein part of the seeds. The data in Table 1 show that the moisture content of the samples under study is much lower. Therefore, the resulting analytical parameter will further characterize only TAG protons.

The seed samples prepared for the analysis characterize the range of the oil content from 32.1 to 56.3%. The seeds are healthy; their acid number corresponds to the first-class sunflower and does not exceed 0.8 mg KOH/g for all samples.

At the first stage of the study, we obtained the NM relaxation characteristics of the prepared oilseeds and determined their oil content using the Soxhlet method of exhaustive extraction. The obtained calibration curve, describing the dependence of the oil mass on the amplitude of the NMR signal, is shown in Fig. 1.

Fig. 1
A line graph of P oil, g versus A 0 oil, a. u.. It plots an upward tangent line along with dots. Y = 0.0071 x negative 0.008 and R squared = 0.9999 are labeled at the bottom.

Dependence of the mass of the sunflower oil, obtained by the Soxhlet extraction, on the amplitude of the NMR signal

Full size image

The resulting curve is characterized by a high correlation coefficient of 0.9999. The main drawback of the described calibration technique is the complexity of implementation, duration, and high requirements for the qualification of personnel.

Table 2 shows the calculated values of the oil mass in sunflower seeds and the actual values obtained by the extraction method.

Table 2 Calculated and actual (obtained by the extraction method) values of the oil mass in sunflower seeds
Full size table

Table 2 shows that the maximum error in measuring the oil mass in the analyzed sample by the NMR method does not exceed 28 mg, which, in terms of the oil content, taking into account the seed moisture content and the mass of the sample with a volume of 25 cm3, is < 0.29%.

At the next stage, the possibility of using oil, obtained by pressing sunflower seeds, for the calibration of an NMR analyzer was investigated. Figure 2 illustrates the dependence of the oil mass on the amplitude of the NMR signal obtained using the oil pressed from sunflower seeds of three contemporary high-oil varieties (Dzhin, Imidzh, SPK).

Fig. 2
A line graph of P oil, g versus A 0 oil, g. It plots an upward tangent line along with dots for the D Z h I N variety, triangles for the I M I D Z h variety, and diamonds for the S P K variety. Y = 0.0071 x negative 0.011 and R squared = 0.9999 are labeled at the top.

Dependence of the pressed oil mass on the amplitude of the proton NMR signal

Full size image

According to the presented curve, the variety special features of seeds cause no significant effect on the amplitude of the oil NMR signal and, consequently, on the obtained analytical dependence. The difference between the obtained curve and that shown in Fig. 1 can be explained by the introduction of different amounts of accompanying substances in the process of oil extraction and pressing.

Table 3 provides the calculated values of the oil mass in sunflower seeds according to the calibration of an NMR analyzer using pressed oil.

Table 3 Calculated values of the oil mass in sunflower seeds according to the calibration of an NMR analyzer using pressed oil
Full size table

The application of a calibration curve, obtained using pressed oil, leads to a maximum measurement error of not more than 32 mg or 0.33 abs% in terms of the oil content taking into account the seed moisture content and the mass of the sample with a volume 25 of cm3 in the entire studied range.

At the third stage of the study, two samples of sunflower oil were used: refined and deodorized of the “BLAGO” trademark and non-refined of the “STAVROPOL’E” trademark, acquired at a retail network.

Figure 3 plots the dependence of the oil mass on the amplitude of the NMR signal, obtained using oil acquired at a retail network.

Fig. 3
A line graph of P oil, g versus A 0 oil, g. It plots an upward tangent line along with dots for the B L A G O oil, and triangles for the S T A V R O P O L E oil. Y = 0.0071 x + 0.05 and R squared = 0.9999 are labeled at the top.

Dependence of the commercially distributed oil mass on the amplitude of the proton NMR signal

Full size image

As in the case of pressed oil, the obtained curve of the dependence between the mass of the analyzed oil sample and the amplitude of the oil NMR signal has a linear nature with the high correlation coefficient.

Table 4 provides the calculated values of the oil mass in sunflower seeds according to the calibration of the NMR analyzer using commercially distributed oil.

Table 4 Results of the oil mass measurement in sunflower seeds according to the calibration, obtained using commercially distributed oil
Full size table

The application of a calibration curve, obtained using commercially distributed oil, leads to a maximum measurement error of not more than 28 mg or 0.35 abs% in terms of the oil content taking into account the seed moisture content and mass of the sample with a volume 25 of cm3 in the entire studied range.

At the next stage, the oil samples were studied for the temporal stability of NM relaxation characteristics. Three weighments of both refined deodorized and pressed oil were selected for a storage at a temperature of 8 °C in tightly-closed glass cups. Once a week, the samples were taken out of the refrigerator and thermostated at a temperature of 23 ± 0.5 °C for 4 h followed by threefold measurements of the oil NMR signal amplitude. The mean of three measurements was taken as the measurement result. Further, using the obtained earlier calibration equations, the oil masses of the studied samples were calculated and the deviations from the basic value were determined (Table 5).

Table 5 Results of the oil mass measurement in the analyzed samples depending on the storage time
Full size table

According to Table 5, the NM relaxation characteristics of the refined oil and, consequently, the calculated values of the oil mass remain stable during the considered storage time. However, for the pressed oil samples, the measured amplitude of the NMR signal of TAG protons considerably altered and led to a decrease in the calculated oil mass value by 570 mg.

Thus, according to the results of the temporal stability study, pressed oil samples can preserve their proton NM relaxation characteristics during not more than 5–7 days, while the refined deodorized samples demonstrate the stability during not < 3 months at a temperature of 8 °C. This can be explained by the presence of a large quantity of free radicals, moisture, and other accompanying substances in pressed oil, facilitating the flow of oxidation processes, which lead to an increase in the quantity of free fatty acids and other products of oxidizing reactions.

Comparative characteristics of the studied methods for calibrating an AMV-1006M NMR analyzer are presented in Table 6.

Table 6 Comparative characteristics of NMR-analyzer calibration methods for determining the oil content of sunflower seeds
Full size table

Conclusion

The conducted research shows that the use of sunflower oil, obtained in different ways, for the calibration of NMR analyzers can significantly facilitate the calibration process, reduce the calibration time from 3–4 days to 3–4 h, as well as to eliminate the use of toxic solvents and additional expensive equipment from the process without significantly increasing the error in determining the oil content of sunflower seeds by the NMR method.