Technical Paper
Classification of various types of sugar products by mineral composition
2021 年 27 巻 1 号 p. 49-56
詳細
2021 年 27 巻 1 号 p. 49-56
Various types of sugar products were classified by measuring their mineral compositions. As a result, the sugar products could be classified according to the results of principal component analysis (PCA) and cluster analysis of the mineral compositions. It also became possible to distinguish between kokuto (black sugar) from China and kokuto from Okinawa, which were classified into the same group in the PCA using a taste sensing system in the authors' previous report. These results agree with the previous report showing that the synthetic variables obtained by PCA of taste information values by the three sensors of a taste sensing system were correlated with the kokuto taste indexes, which indicate the taste intensity of kokuto, and that sugar products can be broadly classified into three groups. Consequently, the taste of sugar products can be now predicted from their mineral compositions.
As reported previously, non-centrifugal sugar, such as brown sugar and black sugar, has come to be used as raw material for processed food products around the world (Fujii et al., 2020). Non-centrifugal sugar is made by concentrating the juice of a plant rich in sucrose, crystallizing sucrose and then cooling into a solid form. It contains a small percentage of plant-derived components, which gives it a specific raw material plant-derived or production process-derived taste and flavor. Non-centrifugal sugars are popular among natural or organic food-conscious consumers. Especially, sugarcane-derived non-centrifugal sugar is widely used in Japanese style sweets, reflecting the Japanese food boom. Also, sugar beet-derived non-centrifugal sugar is recommended by those who practice macrobiotic diets. On the other hand, refined sugar is a sugar product with high sucrose purity made by crystalizing sucrose, and is used in various kinds of food because it has a mild flavor. Examples of these foods are described below. According to Mintel Global New Products Database (GNPD) search results, the number of products using brown or black sugar has increased since 2013 when Japanese food (Washoku) was recorded on the list of the World Heritage (Fujii et al., 2020, Appendix Figure 1), and many of these products are desserts or baked confectioneries.
Digital microphotography pictures of sugars.
Color of the cells under the sample name shows color sample culculated from L*a*b* value.
Ref1: refined sugar and Raw3: raw sugar are centrifugal sugars. Palm6: palm sugar is a non-centrifugal sugar.
BKA4: black sugar from Kagoshima, BOK2: black sugar from Okinawa, and BCN3: black sugar from China are non-centrifugal sugars. Microscopic observations were performed at 100x using Hirox Digital Microscope KH-7700 (Hirox Co., Ltd., Tokyo, Japan).
When sugar is used as an ingredient for processed food products, it is important that its taste quality can be evaluated objectively. Objective evaluation means quantifying the degree of taste felt by the five senses using instrumental analysis.
On the other hand, it is important to identify components that influence how the taste of sugar is felt, and if components with a large influence can be identified and standardized, ingredients for processed food products can be selected effectively (Ikezaki, 2007; Toida, 2012; Enobi et al., 2017).
The authors have so far evaluated various sugar products and measured their components using taste sensing systems and other analytical equipment, and studied the characteristics of those products (Fujii et al., 2019a, 2019b, 2020).
First, we reported that the synthetic variables obtained by principal component analysis (PCA) of taste information values by the three sensors of a taste sensing system are correlated with the kokuto taste indexes, which indicate the taste intensity of kokuto (Fujii et al., 2019a, 2019b). Next, we reported that different types of sugar made from different plants, such as sugarcane sugar, palm sugar, and maple sugar, or sugar products produced in different production areas, such as Okinawa, Kagoshima, China, and South Korea, can also be roughly classified into three groups by PCA of the taste estimation values using a taste sensing system (Fujii et al., 2020).
We also made a report about components that influence taste, stating that the polyphenol contents and color values are highly positively correlated with the kokuto taste indexes obtained from the results of evaluation using a taste sensing system.
Minerals such as salt are the basis of taste and are indispensable ingredients as seasoning for cooking (Endo and Ishikawa, 2015). Compared to refined sugars, non-centrifugal sugars have a high mineral content of several percent, and some sensors of the taste sensing system respond to minerals (Fujii et al., 2019b). Regarding sugar products, we thought that the characteristic ingredients affecting the taste were minerals, so we focused on the mineral compositions.
In this report, the mineral compositions of various types of sugar products were measured and then classified according to the results. Then, the relationship between the mineral compositions and how sugar taste is felt was examined.
Materials The samples used are shown in Table 1. The following seven products were used as refined sugars (abbreviated as Ref): granulated sugar (standard sugar sample), crystal brown sugar, two san-onto products (soft brown sugar), wasanbon sugar, white sugar produced in South Korea, and zarame (coarse sugar) produced in Kagoshima. Three products produced in Kagoshima were used as raw sugars (abbreviated as Raw). The following products were used to represent kokuto sugars: 19 products from Kagoshima (abbreviated as BKA), 21 products from Okinawa (abbreviated as BOK), five products from China (abbreviated as BCN), three products from South Korea (abbreviated as BKR), one product from Taiwan (abbreviated as BTW), and six products with unknown production areas (abbreviated as BUN). Products labeled as “black sugar,” “dark brown sugar,” or “red sugar” were referred to as “black sugar.” A total of 15 products were used as brown sugar; these were produced in Okinawa, Kagoshima, Thailand, the Philippines, Australia, South Korea, and unknown production areas. The two products used as maple sugars (abbreviated as Maple) were both produced in Quebec, Canada. A total of seven products were palm sugars (abbreviated as Palm); these were produced in Cambodia, Thailand, Indonesia, and the Philippines. Four products produced in Hokkaido were used as beet sugars (abbreviated as Beet).
| Original plant | Centrifugal sugars | Non-centrifugal sugars | Processed sugars | |
|---|---|---|---|---|
| Refined sugars | Unrefined sugars | |||
| Sugar cane (Saccharum officinarum) |
White sugar -Granulated sugar (Ref1) -Soft white sugar “Jyo-hakuto” -Crystal white sugar “Shirozarato” -Rock sugar Brown sugar -Soft brown sugar “San-onto” (Ref2,3,4) -Crystal brown sugar “Cyu-zarato” (Ref5,6) |
Raw sugar (brown sugar) (Raw1–3) |
Black sugar “Kokuto” (BKA1–19 Kagoshima) (BOK1–21 Okinawa) Red sugar (BCN1–5 China) Dark brown sugar (BKR1–3 Korea) (BTW Taiwan) (BUN1–6) |
Brown sugar made by blending black sugar, raw sugar, refined sugar and/or molasses (Brown1–15) |
| Japanese sugar cane (Saccharum sineuse) |
“Wasanbon”(in Japan)(Ref7) | |||
| Maple | Maple sugar (Maple1–2) | |||
| Sugar palm or Coconut palm | Palm sugar including Coconut sugar (Palm1–7) |
|||
| Sugar beet | White sugar -Soft white sugar -Granulated sugar |
Raw sugar | Beet sugar (Beet1–4) | |
Measurement of mineral content Cations (sodium, potassium, magnesium, calcium, and ammonium ions) and anions (chlorine, phosphate, and sulfate ions) were measured using an Integrion AS-AP system (Thermo Fisher Scientific K.K., Tokyo, Japan). Cation measurement used a Dionex IonPac CS12A column (Thermo Fisher Scientific K.K.), multication standard solution III (Fujifilm Wako Pure Chemical Corporation, Osaka, Japan), and 20 mM methanesulfonic acid as an eluent. Anion measurement used a Dionex IonPac AS14A column (Thermo Fisher Scientific K.K.), cation mixed standard solution IV (Kanto Chemical Co., Inc., Tokyo, Japan), and 8.0 mM Na2CO3/1.0 mM NaHCO3 as an eluent.
Measurement of color difference For each sample, such as the refined sugars (Ref1), raw sugars (Raw3), palm sugars (Palm6), Kagoshima kokuto (BKA4), Okinawan kokuto (BOK2), and Chinese kokuto (BCN3), 10% (w/w) aqueous solution was prepared, which was then filtered with No. 2 filter paper (Advantec Co., Ltd., Tokyo, Japan) to remove the effects of cloudiness. After filtration, the filter paper appeared white, and no colored materials were noted. Each filtrate was measured using an SE6000 spectrophotometer (Nippon Denshoku Industries Co., Ltd., Tokyo, Japan) based on the CIE L*a*b* system for transmittance with transmission cells, C/2º. To inspect color samples, L*a*b* was converted to RGB to create a color sample (Daichi et al., 2007; Yamashita, 2016).
Statistical processing The measured values of the granulated sugar were analyzed statistically to clarify the differences among their parameters. All statistical analyses of the measured values were performed using the statistical software package BellCurve for Excel (ver. 3.00; Social Survey Research Information, Japan). The PCA was carried out with a variance-covariance matrix. In the cluster analysis, conditions other than the Ward method were determined, and are shown in the results. The correlation coefficient of each pair of evaluation items was calculated using Excel's functions.
Measurement of mineral content Table 2 shows the measured content (mg/kg) of cations (sodium, potassium, magnesium, calcium, and ammonium ions) and anions (chlorine, phosphate, and sulfate ions) in the samples. A covariance analysis of principal components was performed for these values, and principal component scores were calculated with no constant term set. Figure 1 shows the result of mapping based on the PC1 scores and PC2 scores. PC1's contribution rate was 86.6%, while PC2's contribution rate was 6.4%, and the cumulative contribution rate was 93.0%, which indicates that these two components account for most of the data. According to the eigenvectors, the PC1 scores are positive for all the eight items and reflect how large the total mineral amount is, while the PC2 scores are positive for some items and negative for other items, and reflect the mineral balance (or mineral compositions). Based on the plotted data, the samples were classified into four groups: kokuto from China with high PC1 and PC2 scores, kokuto from Okinawa with high PC1 scores and low PC2 scores, kokuto from Kagoshima with intermediate PC1 scores, and refined sugar, raw sugar, and beet sugar with low PC1 scores. The correlation coefficient between the PC1 scores and the kokuto taste indexes reported previously was as high as 0.82 (Fujii et al., 2020).
 |
Principal component scores of each sugar product using PCA of minerals.
Figure 2 shows a dendrogram of the results of a cluster analysis of the mineral contents of the samples (the number of clusters was six, and the Euclidean distance and the distance after combining were calculated using Ward's method). Clusters 1 and 2 are groups of centrifugal sugar comprised mainly of refined sugar, raw sugar, and beet sugar, and kokuto from Okinawa and kokuto from China are concentrated in Clusters 3 to 6. The two maple sugar products are close to each other and contained in Cluster 5, but are distant from other products.
Dendrogram of the result of cluster analysis of minerals.
In the PCA using a taste sensing system in the previous report, sugar products were classified into three groups: kokuto from China and from Okinawa with high PC1 scores, kokuto from Kagoshima with intermediate PC1 scores, and refined sugar, raw sugar, and beet sugar with low PC1 scores. In the PCA of mineral contents, on the other hand, it was found that sugar products can be classified by the combination of PC1 and PC2. In the PCA of mineral contents performed in this report, the samples were classified into four groups: kokuto from China with high PC1 and PC2 scores, kokuto from Okinawa with high PC1 scores and low PC2 scores, kokuto from Kagoshima with intermediate PC1 scores, and refined sugar, raw sugar, and beet sugar with low PC1 scores. These results agree with the classification using a taste sensing system, and the sugar groups were further divided (Fujii et al., 2020). As described in the previous report, each sensor of the taste sensing system reacts to various minerals and umami substances. In the saltiness, which is the pre-taste of the CT0 sensor of taste sensing system, anions such as chloride and sulfate respond to the sensor output, and in the mineral bitterness, which is the aftertaste of the AN0 sensor, cations such as calcium and magnesium ions respond to sensor output (Fujii et al., 2019b). A high correlation was observed between the PC1 scores obtained from the PCA of the contents of cations and anions in the samples and the kokuto taste indexes reported previously, indicating that the contents of cations and anions can also provide a measure of the taste intensity of kokuto (Fujii et al., 2020). In other words, it was suggested that the contents of component minerals influence the taste intensity.
It became possible to classify sugar products by PCA of mineral contents and to distinguish between kokuto from China and kokuto from Okinawa, which were classified into the same group in the PCA using a taste sensing system in the previous report, using the mineral balance. Kokuto from China and from Okinawa has the highest content of minerals, which is followed by kokuto from Kagoshima and the group of refined sugar, raw sugar, and beet sugar. This agrees with the kokuto taste indexes reported previously, indicating that the taste of sugar products can also now be predicted from the mineral contents (Fujii et al., 2020).
Principal component scores of each sugar product using PCA of minerals. This figure is an enlargement of the box in Fig. 1 of the main text.
