Optimisation of Not-from-Concentrate Goji Juice Processing Using Fuzzy Mathematics and Response Surface Methodology and Its Quality Assessment

Abstract

Not-from-concentrate (NFC) juice effectively retains the original characteristics of the fruit. Despite the various health benefits of Goji berries, reports on the processing technology and quality changes of NFC goji juice are lacking. We optimised the processing technology of NFC goji juice. Employing fuzzy mathematics evaluation combined with response surface methodology based on single-factor experiments, the physicochemical, nutritional, and microbiological indicators of the juice were determined. Gas chromatography-ion mobility spectroscopy was used to analyse changes in volatile compounds before and after processing. The optimal processing parameters were: times for homogenisation, ultrasonication, and sterilisation of 2 min, 3 min, and 85 s, respectively, and sterilisation temperature of 102 °C. The resulting product had a sensory evaluation score of 85.5 and a rich, pleasant taste. The ratio of total soluble solids to titratable acidity and turbidity were 12.8 and 1420 NTU, respectively. Post-processing, the contents of β-carotene, polysaccharides, and betaine increased by 57.3%, 26.7%, and 31.5%, respectively. Microbiological indicators met Chinese national limits for food pollutants and microorganisms. The total relative content of volatile substances in NFC goji juice decreased by 19.86% after processing. This study provides a theoretical reference for the intensive processing and development of high-value goji berries.

1. Introduction

Goji berries, the fruit of the medicinal and edible plant Lycium barbarum, are rich in active ingredients [1] such as polysaccharides, flavonoids, alkaloids, and carotenoids. Goji berries offer various health benefits, including reducing cholesterol [2], alleviating atherosclerosis [3], lowering blood sugar [4], and boosting immunity [5]. Currently, the Chinese market offers a wide range of goji-processed products, including goji puree, juice, yogurt, fruit paste, and goji vinegar [6]. Among these, goji juice is particularly favoured by consumers for its unique flavour. The main types of goji juice products include fermented goji beverages [7], clear goji juice [8], and goji compound juice [9]. However, excessive processing often results in significant loss of nutrition and flavour, thereby affecting product quality. To address this issue, developing minimally processed goji juice products that retain both nutritional value and flavour will be a new trend. Not-from-concentrate (NFC) juice is 100% pure fresh juice obtained by directly squeezing fresh fruits. NFC juice retains the original colour, nutrition, and flavour of the fruit better than from-concentrate juice. NFC juice represents a new type of processed product with enhanced health, functional, and nutritional qualities, and holds broad market prospects [10]. Current research on NFC juice primarily focuses on the quality and flavour analysis of apple [11], orange [12], and pear juices [13] using various processing methods. However, there are no reports on the processing technology and quality changes of NFC goji juice.

Fuzzy-mathematics-based comprehensive sensory evaluation is a method that applies fuzzy mathematics theory to product sensory evaluation [14], suitable for assessing subjective and hard-to-quantify characteristics. It has been used to evaluate the sensory characteristics of products such as food, beverages, and cosmetics [15,16,17]. Response surface optimisation is a statistical technique aimed at determining the optimal parameter combination in multivariable systems to maximise or minimise the target function (response value) [18]. While widely used in food research, it is prone to subjective intervention by evaluators. There have been no reports on using fuzzy mathematics sensory evaluation combined with response surface methodology to study the processing technology of NFC goji juice.

In this study, fresh Xinjiang goji berries were used as experimental materials. The ratio of total soluble solids to titratable acidity (TSS/TA) and sensory evaluation scores were used as evaluation criteria. The processing parameters of goji juice extraction, homogenisation, ultrasonic treatment, and sterilisation were optimised using fuzzy mathematics sensory evaluation combined with response surface methodology. The multidimensional effects of processing technology on the quality of NFC goji juice products, including physicochemical, microbiological, nutritional, and volatile substances, were evaluated. This study aims to provide technical guidance and a theoretical reference for the diversified development and production of processed goji products.

2. Materials and Methods

2.1. Experimental Materials and Instruments

Fresh goji berries of the variety “Ningqi No. 7” were harvested from Jinghe County, Xinjiang, China, with a soluble solids content (SSC) of 19.9% and a total acidity of 0.64 g/L. After harvesting, the fruits were transported to the laboratory of the Xinjiang Academy of Agricultural Sciences and stored at 0–5 °C for later use.

• Turbidity standard solution: Analytical grade, supplied by Howei Pharmaceutical Technology Co., Ltd. (Guangzhou, China).
• Organic solvents: 2-pentanone, 2-hexanone, 2-heptanone, 2-octanone, and 2-nonanone (Sinopharm Chemical Reagent Beijing Co., Ltd., Beijing, China), all analytical grade, were used as external references to calculate the retention index (RI) of volatile compounds.
• PAL-BX/ACID2 digital sugar acidity meter (Atago Co., Ltd., Tokyo, Japan) was used to measure TSS; WZS-188 turbidity meter (INESA Lighting Ltd., Shanghai, China) was measured the turbidity of juice; FT74XTS high/ultra-high temperature instantaneous sterilisation machine (Armfield Ltd., Ringwood, UK) was used for high-temperature sterilisation; JY92-IIN ultrasonic cell disruptor (Ningbo Xinzhi Biotechnology Co., Ltd., Ningbo, China) was used for ultrasonic processing of fruit juice; T18 homogeniser (IKA GmbH, Staufen, Germany) was used for juice averages.

2.2. Experimental Methods

2.2.1. NFC Goji Juice Processing Technology Flow

The processing process of NFC wolfberry juice is shown in Figure 1. Fresh wolfberry was washed, after juicing, filtration, homogenisation, ultrasound and sterilisation.

2.2.2. Single-Factor Experimental Design of NFC Goji Juice Processing Technology

The homogenisation time, ultrasonication time, sterilisation temperature, and sterilisation time were selected as independent variables. The processing technology was optimised by measuring the TSS/TA ratio, turbidity, and sensory evaluation scores of the NFC goji juice.

Effect of Homogenisation Time on Quality of NFC Goji Juice

Homogenisation times were set to 1, 2, 3, 4, and 5 min, with a fixed ultrasonication time of 2 min, sterilisation temperature of 90 °C, and sterilisation time of 60 s. The TSS/TA ratio and turbidity of NFC goji juice samples were measured, and sensory evaluation was performed using fuzzy mathematical models to analyse the impact of homogenisation time on juice quality.

Effect of Ultrasonication Time on Quality of NFC Goji Juice

Ultrasonication times were set to 1, 2, 3, 4, and 5 min, with a fixed homogenisation time of 2 min, sterilisation temperature of 90 °C, and sterilisation time of 60 s. The TSS/TA ratio and turbidity of the processed goji juice samples were measured, and sensory evaluation was performed using fuzzy mathematical models to analyse the effect of ultrasonication time on juice quality.

Effect of Sterilisation Temperature on Quality of NFC Goji Juice

Sterilisation temperatures were set to 80 °C, 90 °C, 100 °C, 110 °C, and 120 °C, with a fixed homogenisation time of 2 min, ultrasonication time of 2 min, and sterilisation time of 90 s. The TSS/TA ratio and turbidity of NFC goji juice samples were measured, and sensory evaluation was performed using fuzzy mathematical models to analyse the impact of sterilisation temperature on juice quality.

Effect of Sterilisation Time on Quality of NFC Goji Juice

Sterilisation times were set to 30, 60, 90, 120, and 150 s, with a fixed homogenisation time of 2 min, ultrasonication time of 2 min, and sterilisation temperature of 90 °C. The TSS/TA ratio and turbidity of NFC goji juice samples were measured, and sensory evaluation was performed using fuzzy mathematical models to analyse the impact of sterilisation time on juice quality.

2.3. Design of Response Surface Experiments

Based on the results of the single-factor experiments, response surface methodology was used to analyse the effects of homogenisation time (A), ultrasonication time (B), sterilisation temperature (C), and sterilisation time (D) on the TSS/TA ratio and sensory evaluation scores of the processed goji juice. Models were established to optimise the processing technology.

Table 1. Levels of factors in response surface experiments.

			Factor
Level	A. Homogenisation Time (min)	B. Ultrasonication Time (min)	C. Sterilisation Temperature (°C)	D. Sterilisation Time (s)
−1	2	1	90	60
0	3	2	100	90
1	4	3	110	120

lists the levels of the factors used in the response surface experiments.

2.4. Method for Index Determination

2.4.1. Total Soluble Solids (TSS)

The PAL-BX/ACID2 sugar acidity meter was used to measure TSS in 20 μL samples. Each sample was tested in triplicate, and the average value was recorded.

2.4.2. Titratable Acidity (TA)

A 20 μL sample was extracted, diluted to 1000 μL, and measured using the PAL-BX/ACID2 sugar acidity meter (Atago Co., Ltd., Tokyo, Japan). Each sample was tested in triplicate, and the average value was recorded.

2.4.3. TSS/TA Ratio

The TSS/TA ratio was calculated as the ratio of total soluble solids to TA.

2.4.4. Turbidity

Turbidity was measured using a WZS-188 turbidimeter (Wuhan Huaxing, Wuhan, China). Each sample was tested in triplicate, and the average value was recorded.

2.4.5. Microbiological Indicator Detection

The numbers of E. coli were determined according to Prisacaru. A.’s method [19]. From the dilutions made, 1 mL of sample was taken and inoculated on Petrifilm (3M, St. Paul, MN, USA), and then incubated at 37 °C for 24 h.

The total number of germs was determined according to Prisacaru. A. and Wei, W.’s method [19,20], respectively. Thus, 1 mL sample was taken from each sample and 7 dilutions were made. From the last dilution, 1 mL was taken and inoculated on a Petri plate, using nutrient agar as a nutrient medium. The plates were inoculated at 35 °C for 24 h, and the colonies were counted.

2.4.6. Colour

Colour measurements were performed using a YS6060 desktop colourimeter (Yashica Co., Ltd., Tokyo, Japan), with values expressed as L, a, and b, according to the Commission Internationale de l‘Eclairage (CIE) system. The total colour difference (ΔE) was calculated using the formula: ΔE = (ΔL*² + Δa*² + Δb*²)^1/2.

2.4.7. Nutritional Quality Determination

Betaine content was determined according to Subeen. D’s method [21]. The determination of β-carotene, polysaccharides, and total phenol content was based on Singleton V L [22]. and other relevant literature.

2.4.8. Determination of Volatile Components

Volatile substances in both processed and unprocessed goji juice were analysed using gas chromatography-ion mobility spectrometry (GC-IMS). A 3.00 g sample of goji juice was placed in a 20 mL (2 cm × 4 cm) headspace vial. Each experiment was repeated three times. The sample was incubated at 40 °C for 10 min and then injected with 0.5 mL, with the injection needle at 45 °C and an incubation speed of 500 rpm. High-purity nitrogen (≥99.999%) was used as the carrier gas. The chromatographic column was FS-SE-54-CB-0.5 15 m ID: 0.53 mm, with a column temperature of 45 °C. The analysis time was 20 min, with a program flow rate set to 2.00 mL/min for 10 min, linearly increasing to 20.00 mL/min over 5 min, then to 100.00 mL/min over 15 min, and maintained at 100.00 mL/min for 5 min.

Standards of 2-pentanone, 2-hexanone, 2-heptanone, 2-octanone, and 2-nonanone were used to establish standard curves and determine the retention times of flavour substances in the samples.

2.5. Establishment of Fuzzy Mathematics Model

2.5.1. Sensory Evaluation

Referring to the method of Ma et al. [23], sensory evaluation used taste, colour, odour, and texture as indicators (

Table 2. Sensory evaluation criteria.

Level	Taste	Colour	Odor	Texture
V1	Pronounced fruity flavour, appropriately sweet and sour, rich taste, moderate astringency	Colour ranges from orange-red to brown-red, rich and appealing colour, good gloss	Noticeable fruity aroma, pure aroma without off-notes	Juice is uniform, no pulp precipitation
V2	Moderate fruity flavour, slightly sour or sweet, mediocre taste, slight astringency	Colour deviates slightly, good gloss	Aroma is somewhat faint, slight off-notes, mediocre aroma	Layering present but not significant, minimal sediment at the bottom
V3	Faint fruity flavour, inappropriate sweetness or acidity, unbalanced taste, slightly more astringent	Colour is uneven, slight gloss	Fruity aroma lacking, off-notes in aroma, poor aroma	Significant layering, noticeable sediment at the bottom
V4	No fruity flavour, overly sour or sweet, poor taste, heavy astringency	Dull colour, poor gloss	No fruity aroma, pronounced off-notes, unbalanced aroma	Serious layering, flocculation present, significant sediment at the bottom, uneven colour

). A panel of 10 food science graduate students (five males and five females) aged 20–30 years and in good health was organised. Smoking, drinking, and eating were prohibited for 2 h before evaluation. During evaluation, no discussion was allowed, and each sample was evaluated at 15-min intervals, with rinsing with water in between. The panel objectively evaluated and scored the juices.

2.5.2. Determination of Sensory Factor Set and Comment Set

Fuzzy-mathematics-based comprehensive sensory evaluation is a method that applies fuzzy mathematics theory to product sensory evaluation [14], which is suitable for assessing subjective and hard-to-quantify characteristics.

The sensory factor set U = {taste, colour, odour, texture} = {U₁, U₂, U₃, U₄} was established using the respective states of the goji juice as evaluation factors. The sensory comment set V = {V₁, V₂, V₃, V₄} was established where V₁ = “excellent (90–100)”, V₂ = “good (80–91)”, V₃ = “fair (60–79)”, and V₄ = “poor (0–59)”, as shown in Table 2. The fuzzification method of boundary clarification was applied to obtain the scores, as listed in

Table 3. Correspondence between comment set and scoring range.

Comment Set	Excellent	Good	Fair	Poor
Scoring range	90–100	80–89	60–79	0–59
Boundary clarification	95	85	70	30

.

2.5.3. Determination of Evaluation Weight Set

A questionnaire survey method was employed [24], involving 100 respondents who were food-related teachers or students in good health and without sensory impairments. The taste, colour, odour, and texture of the goji juice were analysed and evaluated to determine the weight set (

Table 4. Weight value in non-concentrated reduced (NFC) goji berry juice.

Factor	Taste	Colour	Odor	Texture
Weight value	30/100	20/100	25/100	25/100

).

2.5.4. Comprehensive Evaluation Set of Fuzzy Relationships

The completed score sheets were analysed to calculate the ratio of votes for each factor at each level relative to the total number of respondents. This data was used to create the fuzzy relationship matrix.

2.6. Data Statistics and Analysis

Excel 2016 and Origin 2021 were used for the analysis and graphing of results. Design-Expert 8.0.6 was utilised for response surface analysis, optimisation, and modelling. Laboratory Analytical Viewer (LAV) 2.2.1 software included with the GC-IMS equipment was used to qualitatively determine the characteristic volatile substances. Information on each compound was obtained using the built-in 2014 NIST database in the GC-IMS Library1.0.3 Search software. The Reporter and Gallery plugins in the LAV data analysis software were applied to construct fingerprint diagrams of the volatile organic compounds.

3. Results and Discussion

3.1. Results Analysis of Single-Factor Experiments

3.1.1. Effect of Homogenisation Time on NFC Goji Juice Quality

Homogenisation breaks down and refines suspended particles in fruit juice, reducing average particle size, enhancing dispersion, and improving sensory properties [25,26]. As shown in Figure 2, a homogenisation time of 3 min resulted in the highest turbidity and sensory scores for NFC goji berry juice, measured at 1582 NTU and 79 points, respectively. This is because an optimal homogenisation time impacts large particles, such as fats, fibres, and proteins, through impact and shearing, resulting in smaller, uniformly dispersed particles that enhance juice stability.

With longer homogenisation times, the TSS/TA ratio showed no significant change, consistent with Paola Maresca’s [27] findings on the TSS/TA ratio of NFC apple juice under different processing stages. At a homogenisation time of 5 min, the sensory score decreased to 73 points, turbidity dropped to 1547 NTU, and the TSS/TA ratio decreased to 10.5. This decrease may be attributed to the increased temperature in the juice system with prolonged homogenisation, which raises interparticle energy, reduces juice viscosity, and ultimately decreases juice stability and quality [28,29].

3.1.2. Effect of Ultrasonication Time on NFC Goji Juice Quality

Ultrasonic treatment can enhance fruit juice quality and delay its deterioration [30]. As shown in Figure 3, the NFC goji juice exhibited the highest TSS/TA ratio (11.15) and sensory score (79%) at an ultrasonication time of 2 min, along with increased turbidity. This improvement is attributed to ultrasonic cavitation effects, which break down large molecules such as pectin in the juice, thereby reducing viscosity and enhancing juice quality [31]. This finding aligns with the results of Tiwari [32], who reported improved stability and quality of blackberry and tangerine juices through ultrasonic treatment.

However, with prolonged ultrasonication, both the TSS/TA ratio and sensory score of the juice tended to decrease. At an ultrasonication time of 5 min, the TSS/TA ratio dropped to 10.6, the sensory score fell to 60 points, and turbidity increased to 1589 NTU. The decline in the TSS/TA ratio may be related to the mild disruption of cell structures by ultrasonic treatment [33], leading to the release of intracellular substances such as sugars (glucose, fructose, and sucrose) [34]. This result is consistent with previous findings on the impact of ultrasonic treatment on strawberry juice quality [35].

3.1.3. Effect of Sterilisation Temperature on NFC Goji Juice Quality

The effect of sterilisation temperature on juice quality is illustrated in Figure 4. At a sterilisation temperature of 100 °C, the juice achieved the highest TSS/TA ratio and sensory score, at 18 and 75 points, respectively, with a turbidity of 1294 NTU. However, as the temperature increased to 110 °C, juice quality declined, with a 15.56% decrease in the TSS/TA ratio and a 6.67% decrease in the sensory score compared to those at 100 °C. This decline is attributed to higher sterilisation temperatures causing flocculation in the juice, where the precipitate adsorbs some of the total soluble solids (TSS), resulting in decreased TSS content [36]. Yang’s [36] study on the impact of heat treatment on noni juice quality indicated that low-temperature heat treatment (85 °C for 5 min) significantly increased juice TSS (p < 0.05), whereas medium and high-temperature treatments (100 °C and 115 °C for 5 min each) decreased TSS content. This suggests that low-temperature heat treatment can better maintain juice quality. However, in the current study, 100 °C provided the best juice quality, likely due to the short duration of high-temperature instantaneous sterilisation, which did not adversely affect the juice quality.

3.1.4. Effect of Sterilisation Time on NFC Goji Juice Quality

The impact of sterilisation time on the quality of NFC goji juice is depicted in Figure 5. As sterilisation time increased, the turbidity of the juice also increased. When the sterilisation time was 90 s, the juice achieved a higher TSS/TA ratio and sensory score, measuring 12.48 and 79%, respectively. At 120 s, the TSS/TA ratio and sensory evaluation score decreased by 7.6% and 3.2%, respectively, compared to the 90-s mark. Significant differences (p < 0.05) were observed in TSS/TA ratio, turbidity, and sensory evaluation values across different sterilisation times. When the sterilisation time was extended to 150 s, the TSS/TA ratio decreased to 12.1. This is attributed to longer medium-to-high-temperature heat treatments, which alter the content of TSS and organic acids in the juice system, thereby affecting the taste and overall quality of the juice [33].

3.2. Comprehensive Results of Fuzzy-Mathematics-Based Sensory Evaluation for NFC Goji Juice Processing Technology

Ten sensory evaluators assessed 29 different combinations of processing technologies for goji juice across four criteria. The distribution of votes for each factor and level is shown in

Table 5. Statistics of sensory evaluation indicators of NFC goji juice.

Sample	Taste				Colour				Odor				Texture
	V1	V2	V3	V4	V1	V2	V3	V4	V1	V2	V3	V4	V1	V2	V3	V4
1	0.6	0.3	0	0.1	0.1	0.5	0.1	0.3	0.6	0	0	0.4	0.3	0.5	0.2	0
2	0.6	0.3	0.1	0	0.1	0.5	0.3	0.1	0.6	0.4	0	0	0.4	0.6	0	0
3	0.6	0.2	0	0.2	0.7	0.3	0	0	0.6	0.4	0	0	0.6	0.3	0.1	0
4	0.6	0.3	0.1	0	0.1	0.5	0.3	0.1	0.6	0.3	0.1	0	0.4	0.6	0	0
5	0	0.1	0.3	0.6	0.2	0.3	0.5	0	0	0.4	0.3	0.3	0.2	0.3	0.2	0.3
6	0.1	0.2	0.3	0.4	0.1	0	0.5	0.4	0.3	0	0.3	0.4	0	0.4	0.6	0
7	0	0.1	0.4	0.5	0	0.1	0.4	0.5	0.1	0	0.4	0.5	0	0.1	0.4	0.5
8	0	0.2	0.2	0.6	0.1	0	0.3	0.6	0.6	0.3	0	0.1	0.3	0.4	0	0.3
9	0.6	0	0.1	0.3	0.1	0.3	0.1	0.5	0.6	0.4	0	0	0.6	0.3	0.1	0
10	0.6	0.3	0	0.1	0.1	0.5	0	0.4	0.6	0	0	0.4	0.3	0.5	0.2	0
11	0.4	0.5	0	0.1	0	0.5	0.2	0.3	0.1	0	0.4	0.5	0	0.1	0.4	0.5
12	0.3	0.2	0.3	0.2	0.1	0	0.5	0.4	0.6	0	0	0.4	0	0.4	0.6	0
13	0.6	0.3	0	0.1	0.1	0	0.5	0.4	0.6	0	0	0.4	0.3	0.5	0.2	0
14	0.4	0.5	0	0.1	0	0.5	0.2	0.3	0.1	0	0.4	0.5	0	0.1	0.4	0.5
15	0.1	0.2	0.6	0.1	0.2	0.3	0.4	0.1	0	0.4	0.3	0.3	0.2	0.3	0.2	0.3
16	0.2	0.3	0.1	0.4	0.1	0.3	0.2	0.4	0.6	0.4	0	0	0.3	0.5	0.2	0
17	0	0.2	0.2	0.6	0.1	0	0.3	0.6	0.6	0.3	0	0.1	0.3	0.4	0.2	0.1
18	0.4	0.5	0	0.1	0	0.5	0.2	0.3	0.1	0.4	0.2	0.3	0	0.1	0.4	0.5
19	0.6	0	0.2	0.2	0.1	0.3	0.2	0.4	0.6	0.3	0	0.1	0.6	0.1	0	0.3
20	0.6	0.2	0.2	0	0	0.3	0.6	0.1	0.6	0	0.4	0	0	0.3	0.1	0.6
21	0.4	0.3	0.2	0.1	0.1	0.5	0.3	0.1	0.6	0.3	0.1	0	0.1	0.3	0.4	0.2
22	0.6	0	0.2	0.2	0.1	0.3	0.1	0.5	0.6	0.4	0	0	0.6	0.3	0.1	0
23	0.3	0.2	0.3	0.2	0.1	0	0.5	0.4	0.3	0	0.3	0.4	0	0.4	0.6	0
24	0.6	0.2	0	0.2	0.1	0	0.5	0.4	0.6	0	0	0.4	0.3	0.5	0.2	0
25	0.2	0.4	0.2	0.2	0.1	0.5	0.3	0.1	0.6	0.3	0.1	0	0.1	0.3	0.4	0.2
26	0.2	0.4	0.2	0.2	0.1	0.5	0.3	0.1	0.6	0.3	0.1	0	0.3	0.4	0.2	0.1
27	0.6	0.2	0.2	0	0.1	0	0.3	0.6	0.6	0.3	0	0.1	0.6	0.1	0	0.3
28	0.6	0.2	0.2	0	0.1	0.3	0.2	0.4	0.6	0.3	0	0.1	0.6	0.1	0	0.3
29	0.1	0.1	0.6	0.2	0.2	0.4	0.4	0	0.3	0.4	0.3	0	0.4	0.4	0.2	0

.

The sensory scores of the processed juice samples were calculated using matrix multiplication to ensure accuracy and avoid algorithmic errors [37]. Given the weight set X = {0.30, 0.20, 0.25, 0.25} for NFC goji juice, the sensory evaluation results for each juice sample were obtained by applying the fuzzy principle Y₁ = X × T₁:

$${\mathrm{Y}}_1=\mathrm{X}\times {\mathrm{T}}_1=|0.30,0.20,0.25,0.25|\times \left|\begin{array}{cccc}0.6& 0.3& 0& 0.1\\ 0.1& 0.5& 0.1& 0.3\\ 0.6& 0& 0& 0.4\\ 0.3& 0.5& 0.2& 0\end{array}\right|=\{0.425,0.315,0.07,0.19\}$$

(1)

Similarly, Y₂–Y₂₉ were calculated. The fuzzy comprehensive evaluation score was determined by W = Y × K with the evaluation set K = {95, 85, 75, 30} and Y₁ = {0.425, 0.315, 0.07, 0.19}. The comprehensive score for the first combination was calculated as:

$${\mathrm{W}}_1={\mathrm{Y}}_1\times \mathrm{K}=\{0.425,0.315,0.07,0.19\}\times \left|\begin{array}{c}95\\ 85\\ 75\\ 30\end{array}\right|=78.1$$

(2)

Similarly, the results for W₂–W₂₉ were computed and are listed in

Table 6. Design and results of response surface experiment.

Experimental No.	Homogenisation Time (min)	Ultrasonication Time (min)	Sterilisation Temperature (°C)	Sterilisation Time (s)	TSS/TA Ratio (%)	Sensory Evaluation (Points)
1	2	1	100	90	13.63	78.1
2	4	1	100	90	14.1	87.5
3	2	3	100	90	14.3	87.65
4	4	3	100	90	14.3	87.25
5	3	2	90	60	9.98	64.6
6	3	2	110	60	10.5	65.6
7	3	2	90	120	9.8	53.75
8	3	2	110	120	10	64.25
9	2	2	100	60	11.8	78.8
10	4	2	100	60	12.4	77.2
11	2	2	100	120	11.85	65.35
12	4	2	100	120	11.7	71
13	3	1	90	90	11.85	76.2
14	3	3	90	90	11.8	65.35
15	3	1	110	90	11.77	71.35
16	3	2	110	90	12.5	75.85
17	2	2	90	90	10.8	66.5
18	4	2	90	90	11.4	68.6
19	2	2	110	90	12.3	75.8
20	4	2	110	90	11.2	75.9
21	3	1	100	60	12.3	80.2
22	3	3	100	60	12.8	80.15
23	3	1	100	120	12.3	69.5
24	3	3	100	120	12.1	74.55
25	3	2	100	90	14.2	77.95
26	3	2	100	90	14.03	79.2
27	3	2	100	90	13.8	79.1
28	3	2	100	90	13.66	76.7
29	3	2	100	90	13.7	80.3

.

3.3. Experimental Design and Results

Based on the single-factor experiments, the processing parameters of NFC goji juice were optimised using the Box–Behnken design. The factors selected for the four-factor, three-level response surface analyses were homogenisation time (A), ultrasonication time (B), sterilisation temperature (C), and sterilisation time (D). The response values measured were the TSS/TA ratio (Y1) and the sensory evaluation score based on fuzzy mathematics (Y2). The experimental design and results are presented in Table 6.

3.4. Model Establishment and Significance Analysis

As shown in

Table 7. Table of Variance of Regression Model.

Y1: TSS/TA Ratio (%)					Y2: Sensory Evaluation (Points)
Source	Quadratic Sum	Mean Square	F	p	Quadratic Sum	Mean Square	F	p
Model	51.16	3.65	61.74	<0.0001 **	1701.2	121.51	50.99	<0.0001 **
A-Homogenisation time	0.015	0.01	0.25	0.6260	19.38	19.38	8.13	0.0128*
B-Ultrasonication time	0.33	0.33	5.65	0.0323 *	25.83	25.83	10.84	0.0053 **
C-Sterilisation temperature	0.90	0.9	15.14	0.0016 **	169.95	169.95	71.32	<0.0001 **
D-Sterilisation time	0.34	0.34	5.8	0.0304 *	193.2	193.2	81.08	<0.0001 **
AB	0.06	0.06	0.93	0.3505	24.01	24.01	10.08	0.0068 **
AC	0.72	0.72	12.21	0.0036 **	1	1	0.42	0.5276
AD	0.14	0.14	2.38	0.1455	13.14	13.14	5.51	0.0341 *
BC	0.20	0.2	3.35	0.0887	112.76	112.76	47.32	<0.0001 **
BD	0.12	0.12	2.07	0.1722	6.5	6.5	2.73	0.1208
CD	0.03	0.03	0.43	0.5214	22.56	22.56	9.47	0.0082 **
A2	0.37	0.37	6.32	0.0247 *	18.05	18.05	7.57	0.0156 *
B2	0.53	0.53	8.91	0.0098 **	136.62	136.62	57.33	<0.0001 **
C2	27.47	27.47	464.15	<0.0001 **	458.32	458.32	192.34	<0.0001 **
D2	20.55	20.55	347.21	<0.0001 **	386.72	386.72	162.29	<0.0001 **
Residual error	0.83	0.06			33.36	2.38
Lack-of-fit	0.62	0.06	1.16	0.4801	25.84	2.58	1.37	0.4069
Error	0.21	0.05			7.52	1.88
Total Sum of Squares	51.99				1734.56

Note: * indicates a significant difference (p < 0.05); ** indicates an highly significant difference (p < 0.01); R² = 0.9861; R²_Adj = 0.9721.

, a regression fitting analysis of the experimental results was conducted using Design-Expert 8.0.6 software. The quadratic polynomial regression models obtained are as follows:Y₁ = 13.92 + 0.035A + 0.18B + 0.28C − 0.17D − 0.12AB − 0.43AC − 0.19AD + 0.28BC − 0.18BD − 0.08CD − 0.24A² + 0.3B² − 2.13C² − 1.77D²; Y₂ = 78.86 + 0.81A + 1.10B + 3.88C − 4.01D − 1.07AB − 0.5AC + 1.81AD + 6.54BC + 1.27BD + 2.38CD + 2.12A² + 5.22B² − 8.95C² − 7.91D².

Analysis of Y₁ (TSS/TA Ratio): The analysis of the quadratic regression model for Y₁ indicated significance after optimisation, with a non-significant lack-of-fit (p > 0.05), R² = 0.9841, and R²_Adj = 0.9681, indicating good predictive capability. According to the p-values, the linear term C, interaction term AC, and quadratic terms B², C², and D² have significant effects on the TSS/TA ratio of the juice (p < 0.01). Additionally, the linear terms B and D and the quadratic term A² significantly affected the TSS/TA ratio (p < 0.05). The influence of each factor on the TSS/TA ratio was in the order: C (sterilisation temperature) > D (sterilisation time) > B (ultrasonication time) > A (homogenisation time). Through optimisation, the parameters yielding the maximum TSS/TA ratio are homogenisation time of 2.73 min, ultrasonication time of 2.94 min, sterilisation temperature of 102.06 °C, and sterilisation time of 85.69 s, resulting in a predicted TSS/TA ratio of 14.4105 and an actual value of 14.3.

Analysis of Y₂ (Sensory Evaluation Score): The analysis of the quadratic regression model Y₂ indicated that the optimised model F was 61.74, which was highly significant with p < 0.0001, and the lack-of-fit term was not significant (p > 0.05). The correlation coefficients R² = 0.9861 and R²_Adj = 0.9721 suggest that the predicted values of the model closely reflect the actual values. Based on the p-values, the linear terms C and D, interaction terms BC and CD, and quadratic terms A², B², C², and D² had significant effects on the sensory evaluation of the juice (p < 0.01). The linear term B and interaction term AD had a significant effect on the sensory evaluation (p < 0.05). The influence of each factor on the sensory evaluation score was in the order: D (sterilisation time) > C (sterilisation temperature) > B (ultrasonication time) > A (homogenisation time). Through optimisation, the parameters yielding the maximum sensory score are homogenisation time of 2.83 min, ultrasonication time of 2.95 min, sterilisation temperature of 105.46 °C, and sterilisation time of 89.95 s, resulting in a predicted sensory score of 87.6536 points and an actual value of 87.65 points.

3.5. Analysis of Interaction of Various Factors

To visually represent the interaction effects of factors A (homogenisation time), B (ultrasonication time), C (sterilisation temperature), and D (sterilisation time) on the response variable Y, three-dimensional surface plots and contour plots were generated using Design Expert 8.0.6 software. As shown in Figure 6, a steeper response surface with a larger inclination indicates a greater impact of the interaction between the factors on the response value. Conversely, a flatter surface suggests a lesser impact. If the contour line is oval or saddle-shaped, it signifies a highly significant interaction between the two factors, whereas a circular contour line indicates a non-significant interaction [38]. According to the plots, the interactions between A and C, and the interactions between A and B, the difference is striking (p < 0.01). The interactions between C and B and between C and D were significant (p < 0.05). These interactions significantly affected the response value, which is consistent with the results presented in

Table 8. Quality indices of fresh-pressed goji juice and NFC goji juice.

Measuring Item	Fresh-Pressed Goji Juice	NFC Goji Juice
TSS/TA ratio	11.91 ± 0.2 b	12.80 ± 0.1 a
TSS (%)	19.9 ± 0.15 b	20.10 ± 0.17 a
TA (g/L)	1.67 ± 0.2 a	1.57 ± 0.15 a
Turbidity/NTU	1231 ± 3.5 b	1420 ± 2.8 a
L	11.58 ± 0.8 a	10.22 ± 0.2 b
a	34.68 ± 1.8 a	30.82 ± 1.3 b
b	19.82 ± 0.3 a	17.46 ± 0.2 b
ΔE	41.59 ± 0.5 a	36.87 ± 0.3 b
Total phenols (mg/mL)	3.40 ± 0.01 a	2.60 ± 0.02 b
β-carotene (mg/mL)	3.02 ± 0.1 b	4.75 ± 0.2 a
Polysaccharides (mg/mL)	7.50 ± 0.2 b	9.50 ± 0.2 a
Betaine (g/100 g)	0.276 ± 0.01 b	0.403 ± 0.01 a
E. coli (MPN/mL)	Not detected	Not detected
Total colony count (CFU/mL)	<10	<10

Note: Different lowercase letters in the same line indicate significant differences (p < 0.05).

.

3.6. Determination of Optimal Processing Conditions and Validation of Regression Model

Optimal processing conditions for NFC goji juice were determined using the TSS/TA ratio and sensory evaluation scores as response indicators. The ideal parameters were: homogenisation time of 2.05 min, ultrasonication time of 2.94 min, sterilisation temperature of 102.38 °C, and sterilisation time of 85 s. For practical production, these parameters were adjusted to: homogenisation time of 2 min, ultrasonication time of 3 min, sterilisation temperature of 102 °C, and sterilisation time of 85 s. Validation experiments yielded a final TSS/TA ratio of 14.03 and a sensory evaluation score of 85.5, differing from predicted values by only 1.9% and 0.5 points, respectively. The close match between validation results and model predictions indicates the applicability of the model for optimizing NFC goji juice processing. Under these conditions, the NFC goji juice exhibited a rich red colour, glossy appearance, distinctive goji aroma, stable and uniform juice consistency, and a smooth, balanced taste of sweetness and acidity.

3.7. Quality Analysis of NFC Goji Juice

Physical, chemical, and nutritional quality analyses were conducted on freshly pressed goji juice and NFC goji juice produced under optimised processing conditions. The results are detailed in Table 8.

The TSS/TA ratio significantly affects the sensory characteristics and consumer acceptance of juice [33]. As shown in Table 8, under optimised processing conditions, the TSS/TA ratio (12.80 ± 0.1a) of NFC goji juice is significantly higher than that of fresh-pressed goji juice (11.91 ± 0.2b) (p < 0.05). This is because high-temperature sterilisation during processing causes the evaporation of organic acids, reducing the TA content [37] and increasing the TSS/TA ratio in NFC goji juice compared to fresh-pressed juice. Turbidity is an important indicator of system stability; the higher the turbidity, the more stable the juice [34]. In this study, the turbidity of goji juice increased by 15% after processing compared to fresh-pressed juice. This increase is attributed to homogenisation and ultrasonic treatment, which reduce the size of large particles such as fats, fibres, and proteins, ensuring uniform dispersion and enhancing juice stability [35]. Colour is a key factor influencing consumer acceptance of juices. The L, a, and b values of NFC goji juice are lower compared to fresh-pressed juice, indicating a darker juice colour. The total colour difference ΔE > 5 suggests significant changes due to homogenisation, ultrasonication, and sterilisation, likely caused by polyphenol oxidation and the Maillard reaction involving amino acids and reducing sugars during processing [34,36]. The total phenolic content in processed NFC goji juice decreased by 23.5% compared to fresh-pressed goji juice, likely due to the degradation of phenolic substances under high-temperature treatment [36]. Conversely, the contents of β-carotene, polysaccharides, and betaine in NFC goji juice increased by 1.73 mg/mL, 2 mg/mL, and 0.127 g/100 g, respectively, compared to fresh-pressed goji juice. This increase is due to processing steps like homogenisation, ultrasonication, and sterilisation, which disrupt the cellular matrix and facilitate nutrient release [36,38]. Both freshly pressed goji juice and NFC goji juice had a total colony count of <10 CFU/mL, and no E. coli was detected, meeting national standards. Therefore, the quality of processed NFC goji juice was superior to that of freshly pressed juice.

3.8. Analysis of Volatile Components in NFC Goji Juice

The ion peaks of all volatile compounds in freshly pressed and processed NFC goji juice were summarised to construct a fingerprint of volatile organic compounds using the Gallery plugin in GC-IMS LAV. This analysis identified differences in volatile compounds between freshly pressed and processed NFC goji juice. As shown in Figure 7, 49 volatile substances were detected by GC-IMS, including ten alcohols (four monomers and dimers), 17 aldehydes (six monomers and dimers), nine ketones (three monomers and dimers), five esters, two furans, one olefin, one ether, and four alkanes (one monomer and dimer).

The results indicated significant differences in the volatile substances between freshly pressed and processed goji juice, highlighted in red and green boxes. Volatile compounds in the red box were present in both juices before and after processing (

Table 9. Qualitative analysis information of volatile substances in non-processed and processed NFC goji juice.

No.	Compound	CAS	Molecular Formula	Retention Index	Retention Time (s)	Drift Time (s)	Relative Content (%)
							Fresh-Processed Juice	Processed Juice
1	Ethanol-M	C64175	C2H6O	46.1	489.9	1.0418	0.86	0.61
2	Pentanal-M	C110623	C5H10O	86.1	700.1	1.1887	0.42	0.57
3	Nonanal	C124196	C9H18O	142.2	1104.3	1.4808	0.24	0.18
4	2-Butanone	C78933	C4H8O	72.1	580.5	1.0587	1.32	1.4
5	Ethylmethylpropanoate	C97621	C6H12O2	116.2	769.7	1.1909	0.34	0.33
6	1-Butanol	C71363	C4H10O	74.1	649.9	1.176	1.07	1.16
7	2-Hexenal-M	C505577	C6H10O	98.1	846.2	1.1822	1.56	1.63
8	Propyl acetate	C109604	C5H10O2	102.1	717.6	1.172	0.46	0.49
9	Methyl acetate	C79209	C3H6O2	74.1	510.2	1.1989	0.65	0.87
10	Ethanol-D	C64175	C2H6O	46.1	451.8	1.1168	4.02	5.06
11	3-Pentanone-M	C96220	C5H10O	86.1	684.3	1.1148	1.04	1.08
12	6-Methyl-5-hepten-2-one	C110930	C8H14O	126.2	984.4	1.18	0.4	0.36
13	Hexanal-M	C66251	C6H12O	100.2	792.7	1.2638	1.32	1.4
14	Hexanenitrile	C628739	C6H11N	97.2	880.6	1.2658	0.73	1.09
15	Hexanal-D	C66251	C6H12O	100.2	782.9	1.5636	4.2	3.13
16	2-Butanone-D	C78933	C4H8O	72.1	579.3	1.2475	3.71	3.91
17	2-Hexenal-D	C505577	C6H10O	98.1	847.9	1.521	3.77	3.23
18	Heptanal-M	C111717	C7H14O	114.2	890.6	1.3318	1.32	0.83
19	2-Methylbutanal	C96173	C5H10O	86.1	680.2	1.3947	0.42	0.6
20	1,2-Dimethoxyethane	C110714	C4H10O2	90.1	634.6	1.3151	1.27	1.13
21	1-Hexanol-M	C111273	C6H14O	102.2	865.5	1.3296	1.1	0.79
22	(E)-2-pentenal-M	C1576870	C5H8O	84.1	742.7	1.1065	1.42	0.83
23	3-Hydroxy-2-butanone-M	C513860	C4H8O2	88.1	721.5	1.2476	0.76	0.48
24	3-Pentanone-D	C96220	C5H10O	86.1	685.3	1.3474	4.12	1.97
25	2-Pentylfuran	C3777693	C9H14O	138.2	980.5	1.257	0.89	0.36
26	3-Octanol	C589980	C8H18O	130.2	994.9	1.4088	0.32	0.18
27	Ethylformate	C109944	C3H6O2	74.1	518	1.2181	0.56	0.62
28	1-Pentanol-M	C71410	C5H12O	88.1	759.8	1.2556	0.88	0.44
29	1-Penten-3-one-M	C1629589	C5H8O	84.1	666.2	1.0786	0.87	0.38
30	1-Hexanol-D	C111273	C6H14O	102.2	862.4	1.6442	0.53	0.25
31	2,3-Butanedione	C431038	C4H6O2	86.1	608.4	1.17	0.4	0.25
32	3-Hydroxy-2-butanone-D	C513860	C4H8O2	88.1	722.1	1.4986	0.33	0.17
33	2-Ethylfuran	C3208160	C6H8O	96.1	741.2	1.3078	0.42	0.18
34	(Z)-4-Heptenal	C6728310	C7H12O	112.2	889.2	1.1492	1.25	0.31
35	3-Heptanol	C589822	C7H16O	116.2	887.8	1.6596	0.76	0.38
36	1-Propanol	C71238	C3H8O	60.1	537.2	1.253	0.43	0.21
37	(E)-2-octenal-M	C2548870	C8H14O	126.2	1055.9	1.3364	0.97	0.1
38	(E)-2-octenal-D	C2548870	C8H14O	126.2	1055.9	1.8278	0.32	0.14
39	Dimethyl trisulfide	C3658808	C2H6S3	126.3	980.5	1.3098	0.45	0.11
40	(E)-2-heptenal-M	C18829555	C7H12O	112.2	949.2	1.2592	1	0.12
41	(E)-2-heptenal-D	C18829555	C7H12O	112.2	946.9	1.6728	0.41	0.12
42	Heptanal-D	C111717	C7H14O	114.2	890.6	1.6948	1.88	0.29
43	1-Pentanol-D	C71410	C5H12O	88.1	755.3	1.5145	0.43	0.13
44	(E)-2-pentenal-D	C1576870	C5H8O	84.1	741.6	1.3621	3.12	0.37
45	1-Penten-3-one-D	C1629589	C5H8O	84.1	667.1	1.3112	4.32	0.21
46	Butyl lactate	C138227	C7H14O3	146.2	1016.5	1.2675	1.35	0.28
47	alpha-Phellandrene	C99832	C10H16	136.2	1017.3	1.692	0.64	0.18
48	Phenylacetaldehyde	C122781	C8H8O	120.2	1021	1.2457	0.67	0.88
49	2-Heptanone	C110430	C7H14O	114.2	881.2	1.6332	0.5	0.57

Note: M: monomer; D: dimer.

: 1–36). Compared with fresh wolfberry juice, the relative contents of 2-hexenaldehyde-M, hexaldehyde-M, and 2-methylbutyraldehyde in NFC wolfberry juice increased by 0.07%, 0.08% and 0.18%, respectively. It may be due to the release of aldehydes promoted by processing. Esters such as ethyl propanoate and methyl acetate, which have a fruity aroma, increased by 0.03% and 0.22%, respectively, after processing. The relative contents of alcohols such as butanol and ethanol-D increased by 0.09% and 1.04%, respectively. Alcohols, important precursors of long-chain ester compounds, are mainly derived from the reduction of amino acids and oxidation of fats [39], serving as crucial aromatic substances. The relative content of 16 volatile compounds—1,2-dimethoxyethane, 1-hexanol-M, (E)-2-pentenal-M, 3-hydroxy-2-butanone-M, 3-pentanone-D, 2-pentylfuran, 3-octanol, 1-pentanol-M, 1-penten-3-one-M, 1-hexanol-D, 2,3-butanedione, 3-hydroxy-2-butanone-D, 2-ethylfuran, (Z)-4-heptenal, 3-heptanol, and propanol—in the red box decreased after processing. This decrease is possibly due to the degradation of some compounds caused by changes in conditions such as cell rupture, light, oxygen, and temperature during homogenisation, ultrasonication, and sterilisation. The compounds in the green box area—(E)-2-octenal-M, (E)-2-octenal-D, dimethyl trisulfide, 2-heptenal-M, 2-heptenal-D, heptanal-D, 1-pentanol-D, (E)-2-pentenal-D, 1-penten-3-one-D, butyl lactate, and alpha-phellandrene—had very low contents (0.1–0.37%) in the processed NFC goji juice. Among these 11 volatile substances, (E)-2-octenal (monomer, dimer), 2-heptenal (monomer, dimer), (E)-2-pentenal (dimer), and 1-pentene-3-one (dimer) are aldehydes and ketones containing unsaturated bonds, which degrade more quickly under high-temperature processing, reducing their relative content. Phenylacetaldehyde and 2-heptanone, which have fruity aromas, significantly increased in relative content after processing by 0.21% and 0.07%, respectively. This increase is likely due to high-temperature sterilisation promoting the degradation and oxidation of unsaturated fatty acids, enhancing the juice’s flavour. Overall, the total relative content of volatile flavour substances in NFC goji juice decreased by 19.86% after processing.

4. Conclusions

This study optimised the preparation of NFC goji juice using sensory evaluation scores and TSS/TA ratio as evaluation criteria, employing fuzzy mathematics combined with response surface methodology. The quality and changes in volatile compounds were systematically analysed and compared between freshly pressed goji juice and processed NFC goji juice. The optimal processing parameters were determined to be a homogenisation time of 2 min, ultrasonication time of 3 min, sterilisation temperature of 102 °C, and sterilisation time of 85 s. Under these conditions, the NFC goji juice exhibited a rich taste, moderate astringency, and appropriate sweetness and acidity, achieving a sensory score of 85.5 points. This optimised method is accurate and feasible, and it can be applied to the optimisation of other fruit juices. Compared to fresh-pressed goji juice, processed NFC goji juice not only showed a higher TSS/TA ratio and turbidity, but also contained higher levels of nutrients such as β-carotene, polysaccharides, and betaine. Additionally, some volatile compounds contributing to the fruity flavour (including 2-hexenal-M, hexanal-M, 2-methylbutanal, benzaldehyde, propyl acetate, methyl acetate, 1-butanol, ethanol-D, and 2-heptanone) exhibited significant increases in relative content. The microbiological indicators of the product met national food safety standards. Processed NFC goji juice products demonstrate good nutritional quality and align with the direction of sustainable food development, suggesting broad application prospects. This study provided a theoretical basis and technical support for the production and processing of NFC goji juice. Based on these findings, further exploration of the quality variation patterns and quality control mechanisms at different processing steps is recommended to accelerate the production and development of NFC goji juice.