Orthogonal Design And Sensory Evaluation Of A Functional Beverage Containing Cistanche, Honey, And Lycium Extracts
Nov 21, 2025
2 Results and Discussion
2.1 Optimization of the Formulation and Process for Cistanche Blend Liquor
Table 3. Sensory Evaluation of the Voting Results of Each Sample
| Number | Aroma - Excellent | Aroma - Satisfactory | Aroma - Bad | Taste - Excellent | Taste - Satisfactory | Taste - Bad | Color - Excellent | Color - Satisfactory | Color - Bad | Appearance - Excellent | Appearance - Satisfactory | Appearance - Bad | Sensory Scores |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 3 | 5 | 2 | 3 | 4 | 3 | 5 | 4 | 1 | 4 | 3 | 3 | 5.741 |
| 2 | 5 | 2 | 1 | 3 | 6 | 2 | 4 | 7 | 0 | 7 | 3 | 0 | 6.398 |
| 3 | 4 | 4 | 1 | 4 | 5 | 1 | 6 | 3 | 1 | 6 | 4 | 0 | 6.104 |
| 4 | 5 | 1 | 0 | 5 | 4 | 1 | 6 | 2 | 2 | 7 | 2 | 1 | 6.530 |
| 5 | 7 | 2 | 0 | 8 | 1 | 0 | 7 | 2 | 0 | 7 | 3 | 0 | 7.076 |
| 6 | 6 | 3 | 0 | 7 | 2 | 0 | 6 | 3 | 0 | 6 | 4 | 0 | 6.416 |
| 7 | 4 | 4 | 1 | 6 | 3 | 0 | 5 | 4 | 0 | 4 | 5 | 0 | 5.876 |
| 8 | 5 | 3 | 1 | 6 | 2 | 1 | 6 | 3 | 0 | 5 | 4 | 0 | 6.512 |
| 9 | 2 | 4 | 3 | 4 | 4 | 3 | 4 | 5 | 2 | 4 | 3 | 4 | 5.837 |
Table 4. Orthogonal Test Design and Results
| Number | A: Cistanche deserticola extract (Level) | B: Honey (Level) | C: Lycium extract (Level) | Sensory Scores |
|---|---|---|---|---|
| 1 | 1 | 1 | 1 | 5.741 |
| 2 | 1 | 2 | 2 | 6.398 |
| 3 | 1 | 3 | 3 | 6.104 |
| 4 | 2 | 1 | 2 | 6.530 |
| 5 | 2 | 2 | 3 | 7.076 |
| 6 | 2 | 3 | 1 | 6.416 |
| 7 | 3 | 1 | 3 | 5.876 |
| 8 | 3 | 2 | 1 | 6.512 |
| 9 | 3 | 3 | 2 | 5.837 |
| A: Cistanche deserticola extract | B: Honey | C: Lycium extract | |
|---|---|---|---|
| K₁ | 18.243 | 18.147 | 17.994 |
| K₂ | 20.022 | 19.986 | 19.440 |
| K₃ | 18.225 | 18.357 | 19.056 |
| R² | 1.797 | 1.839 | 1.446 |
Weights: Aroma = 0.230, Taste = 0.360, Color = 0.210, Appearance = 0.200
Based on the optimal Cistanche extract conditions obtained from preliminary single-factor tests and response surface methodology, an orthogonal experiment with three factors at three levels was conducted. The selected factors were the addition amounts of Cistanche extract (A), honey (B), and goji berry extract (C), with the fuzzy mathematics sensory evaluation score as the response. A 3-factor, 3-level orthogonal design was arranged according to the L9(3^4) orthogonal table.
Cistanche Deserticola Liqour
Supportive Service Of Wecistanche-The largest cistanche exporter in China:
Email:wallence.suen@wecistanche.com
2.2 Physicochemical properties of wine samples treated with different aging methods
Fig. 1 Heat map of physicochemical indexes of wine samples with different accelerating aging treatments
As shown in Tables 3 and 4, the order of influence of raw material ratios on sensory evaluation was B (honey addition) > A (Cistanche extract addition) > C (goji berry extract addition). The optimal combination was A2B2C3, corresponding to 100 mL Cistanche extract, 10 mL honey, and 14 mL goji berry extract, i.e., 80.6% Cistanche extract, 8.1% honey, and 11.3% goji berry extract. Five validation trials were performed under the optimal formulation, yielding a final sensory score of 7.114, higher than each item in the table, indicating good accuracy of the optimized formula.
By comparing the physicochemical indices of the experimental liquors (Figure 1), it was found that, relative to the control (CK), all three aged samples (WB, CS, and HXT) showed reduced alcohol content. The smallest decrease occurred in CS, about 0.5% vol, followed by WB at about 1.29% vol. This may be due to ethanol participating in ester synthesis or undergoing thermal decomposition [28,29]. Because activated carbon aging can readily cause ethanol loss [18], the HXT group showed the greatest decrease, up to 3.01% vol. Organic acids are important aroma contributors in Cistanche liquor and largely indicate liquor quality [30]. The data showed that, except for CS, total acidity in the other two treated groups decreased significantly (P < 0.05). Meanwhile, since activated carbon lacks selectivity in adsorption, the total acidity in HXT was significantly lower than in other treatments (P < 0.05), leading to the highest pH in HXT, while pH in the other treatment groups did not change significantly. These findings are consistent with Gong Dexuan et al. [31]. Compared with CK, soluble solids content decreased in all three aging treatments, with CS having the highest soluble solids among the treated groups, possibly because ultrasonic cavitation and thermal effects accelerate biochemical reactions and enhance amylase activity, promoting starch breakdown to reducing sugars [32]. Antioxidant capacity analysis showed that both CS and WB significantly enhanced total antioxidant capacity (P < 0.05), with CS increasing T‑AOC by 56.02% over CK. In summary, different aging methods affected the physicochemical properties of Cistanche blend liquor. Among them, CS yielded relatively superior total antioxidant capacity, total acidity, and soluble solids, which is beneficial for quality improvement.
Cistanche Deserticola Liqour
2.3 Color Parameters of Samples with Different Aging Treatments
Color metrics for the differently aged samples are shown in Figure 2. The L* values of all three aged samples were higher than CK, with WB exhibiting the highest L*, 45.97% greater than CK. The a* values of HXT and CK were the lowest and not significantly different (P > 0.05), whereas CS had the highest a* with significant differences from the other groups (P < 0.05). The b* value was highest in HXT, followed by WB, indicating more yellowish hues in HXT and WB, consistent with the visual results in Figure 3. This may be because WB increases collisions among carotenoid molecules, causing π-orbital electron disruption [33], which facilitates the release and dissolution of carotenoids from organelles and cell walls, promoting carotenoid release from the matrix. The Cab value incorporates contributions from a and b*; a higher Cab indicates more concentrated color and higher saturation [34]. All three aging treatments increased Cab, with CS being the highest, indicating the greatest color saturation. The hue angle (Hab) reflects color tone; higher values indicate more pronounced color characteristics. As shown in Figure 2, WB had the most distinct color features, followed by CS; CK and HXT shared the same Hab. Chroma increases with darker body color; CS and CK had the highest chroma, and Figure 3 shows a clear difference between CS and the other treatments (P < 0.05). Since lower hue implies a redder color and HXT had the highest hue while CK and CS had the lowest and identical hue, HXT skewed yellow, whereas CK and CS skewed red. Overall, the three aging treatments produced distinct color differences; CS yielded a bright, vivid, soft reddish-brown hue.
Fig. 2 Colour index of wine samples with different accelerating aging treatments
Fig. 3 Color characterization of wine samples with different accelerating aging treatments
2.4 Aroma Indices of Samples under Different Aging Treatments
Volatile compounds are critical indicators of flavor quality in blend liquors. A total of 58 volatile compounds were detected, including 7 higher alcohols, 5 fatty acids, 28 esters, 9 benzene derivatives, and 9 other compounds (Figure 4A). The CS group had the highest total volatiles at 20,884.69 μg/L, 10.10% higher than CK (18,968.72 μg/L).
Higher alcohols are important aroma components in sauce-aroma Baijiu; they impart elegant alcoholic notes, enhance body complexity, and increase sweetness perception [35]. Compared with CK, both WB and CS had more species of higher alcohols, and the total higher alcohol content in CS was significantly higher than CK, notably including 1-pentanol, 1-hexanol, and 1-nonanol, which together contributed green, citrus, and sweet notes to CS. Fatty acids can bring unpleasant rancid, cheesy, and fatty odors; total fatty acids in WB and CS decreased by 44.78% and 28.30% versus CK, respectively. During aging, fatty acids esterify with higher alcohols to form esters, thereby improving aroma quality [36].
In Cistanche blend liquor, esters account for over 90% of the total volatiles. Studies have identified ethyl acetate, ethyl hexanoate, and ethyl butyrate as key esters in blend liquors [37]. In this study, only CS showed a significant increase in total esters versus CK (P < 0.05), with total ethyl esters 11.57% higher than CK. CS possessed a richer volatile profile, including ethyl acetate, isoamyl acetate, ethyl decanoate, and isoamyl butyrate, as well as δ-decalactone unique to CS, collectively forming a rich bouquet. Since CK lacked refined aroma from untreated raw materials, aging promoted the formation of floral and fruity esters, improving aroma quality [28]. Moreover, per JIA [33] and others, increasing esters is a hallmark of aging, indicating that ultrasound and microwave treatments both exert aging effects on Cistanche blend liquor [30].
Benzene derivatives and other compounds also contribute significantly. Ethyl benzoate, ethyl phenylacetate, furfural, and 2-nonanone were abundant in CS, imparting pleasant rosy, floral, and caramel notes, positively enhancing aroma complexity.
A heatmap was generated to visualize differences in volatile contents among treatments (Figure 4B). CS clearly exhibited higher levels of many volatiles, including 1-nonanol, ethyl acetate, isoamyl acetate, isoamyl butyrate, δ-decalactone, ethyl phenylacetate, and furfural. WB had relatively higher contents of geranyl acetone and lauryl alcohol. HXT generally showed lower contents across classes, likely because activated carbon adsorbed and removed some aroma compounds while removing off-flavors [38].
Fig. 4 Analysis of volatile compounds in liquor samples with different accelerating aging treatment contents
2.5 Differential Aroma Compounds among Aging Treatments
Using the 58 detected volatile compounds as dependent variables and the different treatments as independent variables, orthogonal partial least squares-discriminant analysis (OPLS-DA) effectively distinguished the four groups (Figure 5A). The model's explanatory power for X (R^2X) was 0.999, for Y (R^2Y) was 0.997, and predictive ability (Q^2) was 0.962; with both R^2 and Q^2 > 0.5, the model fit was acceptable [39]. After 200 permutation tests (Figure 5B), the intercept of the Q^2 regression line with the y-axis was less than 0, indicating no overfitting and valid model verification; thus, the results are suitable for differential analysis.
Marker compounds with VIP > 1 in the OPLS-DA model (seven esters, one higher alcohol, and one other compound) were analyzed. The combined heatmap (Figure 6) shows that most marker aroma compounds, especially esters, were significantly higher in CS than in other samples (P < 0.05). These compounds enhanced floral and fruity notes, further demonstrating that ultrasonic aging positively improves the aroma quality of Cistanche blend liquor.
2.6 Sensory Evaluation of Samples under Different Aging Treatments
Indicator weights were determined by user survey and paired comparison methods, and results were normalized to obtain the weight set. The fuzzy comprehensive evaluation method quantifies the relationships among multiple sensory factors and idealizes their essential attributes and dynamic processes, thereby reducing subjectivity between evaluation indices and sensory factors and improving the scientific rigor and accuracy of the results [40]. The CS sample achieved the highest sensory score, 7.691. This sample exhibited a vivid reddish-brown color and the highest total antioxidant capacity (2.59), showing a close association (Figure 1).
3 Conclusions
A Cistanche blend liquor was developed using Cistanche, honey, and goji berry as raw materials. Through single-factor tests and response surface optimization, and using fuzzy mathematical sensory evaluation as the criterion, an orthogonal design determined the optimal formulation as 80.6% Cistanche extract, 8.1% honey, and 11.3% goji berry extract, under which the highest sensory score was 7.114. Subsequently, three aging methods-ultrasound, microwave, and activated carbon-were applied to evaluate their impacts on liquor quality. All three influenced aroma and physicochemical properties. Notably, ultrasound significantly increased total antioxidant capacity (P < 0.05) while maintaining relatively small changes in basic physicochemical indices; the treated liquor appeared reddish-brown; the types and contents of volatile compounds increased significantly (P < 0.05); and sensory evaluation (7.691 points) indicated optimal aroma and taste. In summary, ultrasonic treatment shows strong application potential in blend liquors for improving the quality of Cistanche blend liquor and provides data support for production.
References
[1] Munire·Mutalifu, Aini·Ersman, Ye Yijie, et al. Diversity of endophytic bacteria in Cistanche and isolation and screening of functional strains [J]. Microbiology China, 2024: 1–19.
[2] Wang F, Li R, Tu P, et al. Total glycosides of Cistanche deserticola promote neurological function recovery by inducing neurovascular regeneration via Nrf-2/Keap-1 pathway in MCAO/R rats [J]. Frontiers in Pharmacology, 2020, 11: 236.
[3] Li Yingman. Study on functional factors and mechanisms of Cistanche in delaying aging [D]. Tianjin: Tianjin University of Traditional Chinese Medicine, 2023, 9–10.
[4] Chen A, Wan F, Ma G, et al. Radio frequency vacuum drying study on the drying characteristics and quality of Cistanche slices and analysis of heating uniformity [J]. Foods, 2024, 13(17): 2672–2672.
[5] Ministry of Health of the People's Republic of China. National Food Safety Standard Distilled Spirits and Their Blended Spirits: GB 2757-2012 [S]. Beijing: China Standards Press, 2013: 1.
[6] Zhang Jun, Zhang Dongxue, Zhou Guowang, et al. Development of a composite blend liquor with sea buckthorn and Hibiscus manihot [J]. Liquor-Making Science & Technology, 2017, (2): 33–36.
[7] Zhang Lipan, Guan Bingfeng, Hu Guifang, et al. Development of a peony–cornus–goji wine and its antioxidant activity [J]. Food Science and Technology, 2021, 46(4): 70–74.
[8] Raquel R S, Laura V A, Manuel S J, et al. Optimization of the process of aromatic and medicinal plant maceration in grape marc distillates to obtain herbal liqueurs and spirits [J]. Journal of the Science of Food and Agriculture, 2016, 96(14): 4760–4771.
[9] Chen Jiangping. Processing technology for a blend liquor with okra, goji, and chrysanthemum [J]. Modern Food, 2018(8): 4.
[10] He Famei. Study and functional evaluation of sauce-aroma honey blend liquor [D]. Guizhou University, 2020.
[11] Ball, David W. The chemical composition of honey [J]. Journal of Chemical Education, 2007, 84(10): 1647–1650.
[12] Zhang Min. Effects of brewing processes on polyphenolic antioxidant components in mead [D]. Guangzhou: South China Agricultural University, 2016: 27–28.
[13] Zhang Miaomiao, Cao Guozhen, Miao Jianshun, et al. Advances in physical methods for accelerating the aging of brewed liquors [J]. Science and Technology of Food Industry, 2015, 36(12): 395–399.
[14] Hu Li, Lu Zhen, Li Wei. Maturation acceleration of grain Baijiu by activated carbon for alcoholic beverage use [J]. Brewing, 2017, 44(02): 99–101.
[15] Wu Zhifei, Liu Zheng, Lu Yanxiang, et al. Effects of different aging acceleration methods on sauce-aroma Baijiu [J]. Food Research and Development, 2023, 44(16): 143–149.
[16] Zhu Meiling, Ni Shugan, Luo Gaojian, et al. Application of novel activated carbon for alcoholic beverages in blend liquors [J]. Liquor-Making Science & Technology, 2022(4): 86–89.
[17] Wu Z, Lv S, Xiao P, et al. Comprehensive analysis of risk factors (methanol, acetaldehyde and higher alcohols) in alcoholic beverages and their reduction strategies: GC–MS analysis and modified activated carbon adsorption and characterization [J]. Food Chemistry, 2024, 460: 140461.
