Preparation of reconstituted milk

Reconstituted milk with 15% (w/w) total solids was prepared by dispersing skim milk powder in distilled water. After dispersion, the milk was kept at refrigerator temperature for approximately 12 h to ensure complete hydration of the solids. Finally, it was filtered through a double-layered cheesecloth with a pore diameter of approximately 1 mm just before use.

Lyophilization of soapwort extract

Liquid soapwort extract was frozen at −80°C overnight, lyophilized (ScanVac CoolSafe Pro95-15, LaboGene ApS, Lynge, Denmark), ground with a coffee grinder (Premier Prg 259-170, Nasmina Inc., Istanbul, Türkiye) and stored at −20°C until use. Prior to yogurt production, a 1.0% (w/w) dispersion of the dried extract was prepared in reconstituted milk (15% total solids) within an airtight glass jar and used within 2 h.

Production of yogurt with soapwort extract

Four groups of nonfat yogurt (<0.15% fat) containing 15% (w/w) total solids and varying proportions of lyophilized soapwort extract (Control: 0%; SE25: 0.025%; SE50: 0.050% and SE100: 0.100% w/w) were produced according to the formulations in Table 1 and the process outlined in Fig. 1. These concentrations were selected based on preliminary experiments testing a range of 0.01–0.50%. The yogurt milk bases were heat-treated in a water bath at 85°C for 20 min. Following heat treatment, the bases were rapidly cooled to approximately 45°C and inoculated with 3% (w/w) yogurt culture. The samples were then transferred to 100 ml polypropylene containers and incubated at 42°C in an incubator (IB 15G, Jeio Tech. Ltd., Seoul, South Korea) until a final pH of 4.60 was reached. During incubation, pH was monitored by keeping a pH electrode immersed in a yogurt container from each of the four groups. Upon reaching the target pH, the samples were refrigerated at 4°C. Physicochemical analyses were performed on days 1 (24 h after production), 15 and 30 of storage; however, total solids and protein contents were evaluated only on days 1 and 30, while sensory evaluation was carried out solely on day 15. The entire production process and all subsequent analyses were performed in three independent experimental replicates conducted over a 6-month period. Each replicate was monitored throughout a 30-day storage period.

Table 1.

Formulation of yogurt milk bases supplemented with dried soapwort extract

Table 1 Long description

Yogurt milk base formulations are compared across four treatments with increasing dried soapwort extract concentration: control, SE25, SE50, and SE100. Each formulation totals 100.0 g. Soapwort extract dispersion rises stepwise from 0 g in the control to 2.5 g, 5.0 g, and 10.0 g in SE25, SE50, and SE100. Reconstituted milk decreases in the same pattern from 100.0 g to 97.5 g, 95.0 g, and 90.0 g. The data indicate a direct substitution of milk by soapwort dispersion to keep total mass constant. Values describe ingredient amounts per batch and do not provide outcomes such as texture, flavor, or fermentation performance.

^a Percentages in the column headers indicate the final concentration of dried soapwort extract in the yogurt milk base.

Figure 1.

Flowchart for the manufacturing process of set-type yogurt supplemented with dried soapwort extract. Refer to Table 1 for specific ingredient quantities.

Analyses of skim milk powder and lyophilized soapwort extract

The protein content of the skim milk powder and dried soapwort extract was determined using the Bradford method, as modified by Zor and Selinger (Reference Zor and Selinger1996). The dry matter content of both materials was measured by evaporating the water in an airflow oven (G11600S model, Thermal Laboratory Appliances Industry and Trade Co., Istanbul, Türkiye) at 102 ± 2°C (IDF, 1993a). Additionally, the saponin content of the dried soapwort extract was estimated according to the method described by Lalitha et al. (Reference Lalitha, Seshadri and Venkataraman1987).

Physicochemical analysis

The pH values of the yogurts were measured using a digital pH meter (pHenomenal 1000L, VWR International Ltd., Dublin, Ireland). The titratable acidity (expressed as % lactic acid), total solids and fat content were quantified using AOAC International protocols 947.05, 990.20 and 2000.18, respectively, with slight modifications for yogurt matrices (AOAC, Reference Horwitz and Latimer2005). The protein content was analysed using the Kjeldahl method (IDF, 1993b).

Colour

The CIELAB parameters (L*, a* and b*) of the yogurt samples were determined from three different locations under C light using a CR-400 Minolta colorimeter (Konica Minolta Optics Inc., Tokyo, Japan). In this system, the a* value ranges from green (−) to red (+), the b* value ranges from blue (−) to yellow (+) and the L* value is an estimation of lightness varying from 0 (black) to 100 (white). The colorimeter was calibrated using a standard white plate (with a standard C light source, Y = 87.1, x = 0.3157 and y = 0.3234). In addition, the total colour difference (ΔE*) of the trial yogurt samples with respect to the control yogurt was calculated using the following equation:

\begin{equation*}\Delta {E^*} = \sqrt {{{({L^*} - L_0^*)}^2} + {{({a^*} - a_0^*)}^2} + {{({b^*} - b_0^*)}^2}} \end{equation*}

where L*, a* and b* are the colour values of the trial yogurts, and L ₀*, a ₀* and b ₀* are the corresponding colour values of the control yogurt.

Gel hardness

The gel hardness of the yogurts was determined using a texture analyser (TA-XTplus, Stable Micro Systems Ltd., Godalming, England) equipped with a 5 kg load cell, according to the YOG1/BEC method recommended by Stable Micro Systems Ltd. Briefly, the measurement was conducted at 10°C on samples contained within plastic cups (diameter: 5.5 cm; depth: 6.8 cm). A probe (diameter: 35 mm; thickness: 5 mm) was inserted to a depth of 30 mm at a constant speed of 1 mm/s. The maximum force registered in the resulting force–time graph was recorded as the measure of hardness.

Gel viscosity

The viscosity of the experimental yogurts was determined using a viscometer (DV2Textra, Brookfield AMETEK, Middleborough, MA, USA). Prior to analysis, the yogurt samples (in plastic containers of 5.5 cm diameter and 6.8 cm depth) were stirred for 1 min with a hand mixer and allowed to rest for 5 min to equilibrate. Viscosity measurements were then conducted at 10°C using a No. 64 spindle with an LVK protector at 40 rpm for 3 min. The value recorded at the first minute was used to represent sample viscosity (mPa · s).

Serum separation (syneresis)

The serum separation analysis was performed according to the method of Gauche et al. (Reference Gauche, Tomazi, Barreto, Ogliari and Bordignon-luiz2009), with minor modifications. After mixing with a hand mixer for 1 min, a 30 g yogurt sample was transferred to a centrifuge tube and centrifuged (NF 800R, Nüve, Ankara, Türkiye) at 350 × g for 10 min at 10°C. The supernatant was subsequently discarded, and the remaining residue (pellet) was weighed. The serum separation values (%) were calculated using the following formula:

\begin{align*}& Serum\;separation\;(\%)\\ & =\frac{Weight\;of\;initial\;sample\;(g)-Weight\;of\;residue\;(g)}{Weight\;of\;initial\;sample\;(g)}\times100\\\end{align*}

Scanning electron microscopy

Approximately 100 µL of the yogurt sample was spread as a thin layer onto aluminium foil, frozen and subsequently lyophilized using a freeze dryer (ScanVac CoolSafe Pro95-15, LaboGene ApS, Lynge, Denmark). The dried samples were sputter-coated with gold to a thickness of 15 nm and examined using a scanning electron microscope (SEM) (Zeiss EVO 40, Carl Zeiss AG, Oberkochen, Germany). Micrographs were captured at an accelerating voltage of 10 kV and a magnification of 5000×. To ensure accuracy, at least five distinct fields of view were scanned per sample, from which representative images were selected.

pH values

The fermentation process involves the acidification of milk to pH 4.60 via the metabolic activity of L. delbrueckii subsp. bulgaricus and S. thermophilus. This acidification induces the formation of the yogurt gel, a three-dimensional network comprised of casein clusters and chains (Lee and Lucey, Reference Lee and Lucey2010). In the present study, no significant difference was observed in the incubation time required for the yogurt milk bases containing varying concentrations of dried soapwort extract to reach the target pH of 4.60, which ranged uniformly between 210 ± 15 min across all treatments (P > 0.05).

Following production, pH levels generally continue to decline (post-acidification) due to the ongoing metabolic activity of the starter culture; thus, pH is a critical determinant of both sensory properties and shelf life. The pH values of the yogurt samples assessed on days 1, 15 and 30 of storage at 4°C are presented in Table 2. As shown, the pH values ranged from 4.20 to 4.30. Statistical analysis revealed that neither the addition of soapwort extract nor the storage period had a significant effect on the pH of the yogurts (P > 0.05). This lack of influence can be attributed to the chemical composition of the soapwort extract; it does not contain significant levels of fermentable carbohydrates (Slobodianiuk et al., Reference Slobodianiuk, Budniak, Marchyshyn, Kostyshyn and Zakharchuk2021) that would accelerate acid production, nor does it possess antimicrobial properties strong enough to inhibit the specific lactic acid bacteria (L. bulgaricus and S. thermophilus) at the concentrations used. Consequently, the fermentation kinetics remained consistent across all treatments, indicating that the extract is chemically inert regarding the acidification process. Throughout the storage period, pH values exhibited only minor, insignificant fluctuations. The highest average value (4.30 ± 0.05) was recorded in the control sample on day 1, while the lowest (4.20 ± 0.05) was observed in the SE100 sample on day 15. This pH stability over the 30-day storage period is attributed to the high protein content (>5%, see Table 2) of the samples, which provides a significant buffering capacity. Mechanistically, milk proteins (caseins and whey proteins) act as amphoteric polyelectrolytes capable of resisting changes in pH. The basic amino acid side chains (such as histidine) and colloidal calcium phosphate within the casein micelles bind the excess hydrogen ions (H⁺) generated by the starter culture during post-acidification (Lee and Lucey, Reference Lee and Lucey2010). By absorbing these protons, the proteins prevent the concentration of free H⁺ ions from rising sharply, thereby stabilizing the pH even as organic acids accumulate. This limited post-acidification likely had a positive impact on the overall quality of the samples, particularly their sensory attributes, by preventing the development of excessive sourness. Similarly, Bruzantin et al. (Reference Bruzantin, Daniel, da Silva and Spoto2016) reported higher pH values (4.32) in yogurts produced from milk fortified with skim milk powder, attributing this to the enhanced buffering capacity provided by the increased protein concentration.

Table 2.

Changes in pH, titratable acidity (expressed as % lactic acid w/w), total solids (% w/w) and protein content (% w/w) of yogurt samples during storage

Table 2 Long description

Yogurt samples with four soapwort extract levels (control, 0.025 percent, 0.050 percent, 0.100 percent) are compared across storage days 1, 15, and 30 for pH, titratable acidity, total solids, and protein. pH remains stable for all types, ranging from about 4.20 to 4.30; the lowest value occurs for the 0.100 percent sample on day 15 (about 4.20), while most other values cluster around 4.23 to 4.30. Titratable acidity changes only slightly, staying roughly between 1.33 and 1.39 percent across all types and days, with a small rise by day 30 in the control (about 1.39). Total solids are reported for days 1 and 30 only (day 15 not determined), staying around 14.6 to 14.9 percent; the 0.100 percent sample is highest on day 1 (about 14.86) and the 0.050 percent sample is highest on day 30 (about 14.83). Protein is also reported for days 1 and 30 only (day 15 not determined), remaining near 5.1 to 5.5 percent; the 0.100 percent sample is highest on day 30 (about 5.54). Mean rows indicate very small overall shifts over time, suggesting storage and extract level have limited impact within the reported variability. Interpretation should consider that some measures are missing at day 15 and values are presented with variability estimates.

C, control (0% dried soapwort extract); SE25, 0.025% dried soapwort extract; SE50, 0.050% dried soapwort extract; SE100, 0.100% dried soapwort extract containing yogurt sample; ND, not determined.

^a Mean ± standard deviation.

^b Mean ± standard error.

Titratable acidity

The titratable acidity of yogurt is influenced by numerous factors, including the specific starter culture strains, incubation and storage conditions, as well as the total solids and protein content. The changes in titratable acidity (expressed as % lactic acid) of the yogurt samples on days 1, 15 and 30 of storage are presented in Table 2. Statistical analysis indicated that the addition of dried soapwort extract did not have a significant effect on titratable acidity (P > 0.05). This suggests that the extract had no adverse impact on the growth of the starter bacteria or the buffering capacity of the yogurts. Throughout the storage period, titratable acidity showed a slight but statistically insignificant increase (P > 0.05), mirroring the stability observed in the pH values. The titratable acidity values, obtained in the present study, are consistent with those reported by Paseephol et al. (Reference Paseephol, Small and Sherkat2008) for yogurts supplemented with inulin, Delikanli and Ozcan (Reference Delikanli and Ozcan2014) for yogurts supplemented with whey protein derivatives, Bruzantin et al. (Reference Bruzantin, Daniel, da Silva and Spoto2016) for yogurts fortified with skim milk powder and Çoban et al. (Reference Çoban, Demir Özer and Yildirim2025) for yogurts fortified with dried kefir grains. The minor differences in titratable acidity observed across these studies can be primarily attributed to variations in the metabolic activity of the starter culture strains and the initial buffering capacity of the milk base. Furthermore, the incubation temperature and the rate of cooling post-fermentation are significant factors; variations in these parameters can lead to different levels of residual lactose fermentation and, consequently, varying concentrations of accumulated lactic acid.

Total solids content

Total solids content is a critical determinant of yogurt quality, directly influencing the physical and sensory properties of the gel. In commercial production, the solids content typically ranges from 14 to 15% (Tamime and Robinson, Reference Tamime and Robinson2007). In this study, the total solids content of all experimental yogurt milk bases was standardized to approximately 15% to minimize variability in the resulting gel properties.

Since the total solids content is not expected to change significantly during storage, analysis was conducted only on the 1st and 30th days. The total solids data obtained during storage are presented in Table 2. Statistical analysis indicated that neither the addition of dried soapwort extract nor the storage period had a significant effect on the total solids content (P > 0.05). Throughout the storage period, slight fluctuations in total solids content were detected across all yogurt formulations. This stability was expected, as the samples were stored in tightly sealed containers to prevent water loss via evaporation, and no post-processing treatments were applied. The total solids content measured in the present study is consistent with those reported by Güven and Karaca (Reference Güven and Karaca2003) for yogurts fortified by different methods, Paseephol et al. (Reference Paseephol, Small and Sherkat2008) for yogurts supplemented with inulin and Çoban et al. (Reference Çoban, Demir Özer and Yildirim2025) for yogurts fortified with dried kefir grains, though higher than the values reported by Bruzantin et al. (Reference Bruzantin, Daniel, da Silva and Spoto2016) for yogurts fortified with skim milk powder. These variations in total solids content across the literature are primarily attributed to differences in the standardization protocols and the initial composition of the raw milk employed.

Fat content

Since the experimental yogurt samples were produced from reconstituted milk prepared from skim milk powder, no fat content was detected in any sample when analysed by the Gerber method.

Protein content

In addition to its nutritional value, the protein in yogurt is crucial because it significantly influences both textural and sensory properties. Since protein levels in yogurt are not expected to change significantly during storage, analysis was conducted only on the 1st and 30th days. The protein content determined during these periods is presented in Table 2.

The protein levels of the experimental samples ranged between 5.11 and 5.54%. Statistical analysis indicated that neither the addition of dried soapwort extract (P > 0.05) nor the storage period (P > 0.05) had a significant effect on protein content. Given that the protein content of the dried soapwort extract itself was low (3.42 ± 0.22%), its addition was not expected to significantly alter the protein levels of the yogurt. Furthermore, since no biological or chemical processes that significantly alter crude protein ratios occur during storage, the stability of the protein content over time was expected. While minor fluctuations were observed in the data (Table 2), these were statistically insignificant. The protein contents obtained in this study were higher than the values reported by Isleten and Karagul-Yuceer (Reference Isleten and Karagul-Yuceer2006) (4.10–5.05%) for yogurts fortified by dried dairy ingredients and Güven and Karaca (Reference Güven and Karaca2003) (3.76–4.61%) for yogurts fortified by different methods. In contrast, they were similar to the values (4.81–5.12%) reported by Çoban et al. (Reference Çoban, Demir Özer and Yildirim2025) for yogurts supplemented with dried kefir grains. These variations are primarily attributable to the different standardization methods applied to the yogurt milk. The final protein concentration is directly dependent on the initial protein content of the raw milk and the specific fortification strategy used.

Colour values

Colour is a critical sensory attribute as it is the first quality perceived by consumers. The characteristic white appearance of skim milk, and consequently nonfat yogurt, results from the reflection and scattering of light by casein micelles (Walstra et al., Reference Walstra, Wouters and Geurts2005). The L* values (lightness) determined for the yogurt samples using the CIELAB colour system are presented in Table 3. Throughout the storage period, L* values ranged between 85.97 and 87.78. Although the values exhibited minor fluctuations during storage, statistical analysis revealed that neither the addition of dried soapwort extract (P > 0.05) nor the storage time (P > 0.05) had a significant effect on the samples’ lightness. The lightness values measured in the present study are lower than the values reported by Delikanli and Ozcan (Reference Delikanli and Ozcan2014) (90.23–95.10) and Çoban et al. (Reference Çoban, Demir Özer and Yildirim2025) (90.83–93.44). These variations in lightness are likely attributed to differences in the intensity of the thermal treatment. Mechanistically, more intense or prolonged heat treatment promotes the Maillard reaction (non-enzymatic browning) between lactose and milk proteins (Walstra et al., Reference Walstra, Wouters and Geurts2005), which can slightly darken the matrix and reduce the L* value.

Table 3.

Changes in the L* [lightness (100) − darkness (0)], a* [red (+) − green (−)], b* [yellow (+) − blue (−)] and total colour difference (ΔE*) values of the yogurt samples

Table 3 Long description

Color measurements for yogurt with four soapwort extract levels (including a control) are reported at days 1, 15, and 30, with an overall mean for each yogurt type and storage day. Lightness values were consistently high and tightly clustered, roughly 86 to 88 across all treatments and days, indicating minimal lightness change during storage. Red to green values stayed negative for all samples, generally near minus 3.6 on day 1 and around minus 3.0 on day 15, with the largest variability at day 15 for the highest extract level. Yellow to blue values decreased over time on average, from about 8.9 at day 1 to about 8.0 at days 15 and 30, and the day 1 average was statistically higher than days 15 and 30. Total color difference was zero for the control at all days, while extract-added yogurts showed small differences, with overall means around 0.79 to 1.14 depending on extract level. Across storage days, the average total color difference peaked at day 15 (about 1.04) and was lower at days 1 and 30 (about 0.59 to 0.60). Statistical letter markers indicate some significant differences by day and by treatment, but most changes are small in magnitude and should be interpreted alongside the reported variability.

C, control (0% dried soapwort extract); SE25, 0.025% dried soapwort extract; SE50, 0.050% dried soapwort extract; SE100, 0.100% dried soapwort extract containing yogurt sample. a, b: Values with different letters in the same row are significantly different (P ≤ 0.05). A, B: Values with different letters in the same column are significantly different (P ≤ 0.05).

^c Mean ± standard deviation.

^d Mean ± standard error.

During storage, the red/green colour (a*) values of the yogurts ranged between −3.67 and −2.14 (Table 3). Although the values fluctuated slightly, statistical analysis indicated that neither the addition of dried soapwort extract (P > 0.05) nor the storage period (P > 0.05) had a significant effect on a* coordinate values. Throughout the storage period, the yogurt samples exhibited negative a* values, indicating a greenish hue. Previous studies attribute this greenness to the presence of riboflavin (vitamin B₂) in the serum; consequently, as serum separation occurs, the a* value typically decreases (i.e., the green hue intensifies) (Shirai et al., Reference Shirai, Pedraza, Gutierrez-duran, Marshall, Revah-moiseev and Garcia-garibay1992). In comparison to our findings, Delikanli and Ozcan (Reference Delikanli and Ozcan2014) reported higher Hunter a values (−1.40 to −0.43) for yogurts supplemented with whey protein derivatives, indicating a less intense green hue. Conversely, Çoban et al. (Reference Çoban, Demir Özer and Yildirim2025) reported lower a* values (−4.98 to −3.75) for yogurts containing kefir grain powder, indicating a more intense green hue than observed in our study. These variations in a* values across the literature are primarily governed by the opacity of the yogurt gel and the extent to which the protein matrix masks the serum pigments. The higher values (less green) reported by Delikanli and Ozcan (Reference Delikanli and Ozcan2014) are likely due to the increased light-scattering effect of the added whey protein derivatives, which effectively masks the green hue of the serum riboflavin.

Throughout the storage period, the b* values (yellow/blue colour) of the experimental yogurts ranged between 7.32 and 9.52 (Table 3). A general decreasing trend in b* values was observed during storage, indicating a reduction in the yellow hue of the experimental samples over time. A similar decreasing trend was also observed by Çoban et al. (Reference Çoban, Demir Özer and Yildirim2025) in yogurts supplemented with dried kefir grain. Statistical analysis confirmed this finding: the storage period had a significant effect on the b* value (P < 0.05), while the addition of dried soapwort extract had no significant effect (P > 0.05). Since the storage period was a significant factor, the mean b* values across storage days were analysed using the Duncan test. The results showed that the mean b* value on day 1 (8.91) was significantly higher than the values recorded on days 15 and 30 (P ≤ 0.05), consistent with the highest yellow hue being present initially. This reduction in b* values during storage can be attributed to the degradation of riboflavin and pigments from Maillard reaction. Over the 30-day storage period, even under refrigeration, these pigments can undergo slow oxidative changes or be affected by the metabolic activity of the starter cultures, leading to a loss of the initial yellow intensity. The Hunter b values reported by Delikanli and Ozcan (Reference Delikanli and Ozcan2014) for yogurts produced with the addition of whey protein concentrate, isolate or hydrolysate were higher than those in our study, ranging between 12.32 and 13.79 during a 14-day storage period. These disparities in b* values are primarily due to the composition of the milk base and the type of fortification used.

The total colour difference (ΔE*) is an important composite parameter that takes into account all differences in the L*, a* and b* colour parameters. Therefore, the ΔE* value is crucial for determining the overall colour difference between the control sample (C) and the treated samples containing dried soapwort extract (SE25, SE50 and SE100). The ΔE* values, determined relative to the control, are presented in Table 3. The ΔE* values of the samples fluctuated irregularly – showing increases and decreases – throughout storage. Statistical analysis revealed that the effect of storage time on ΔE* values was not significant (P > 0.05), whereas the addition of soapwort extract had a significant effect (P < 0.05). Due to the significant effect of the extract, the mean ΔE* values of the different treatments were subjected to the Duncan test. The Duncan test confirmed that the soapwort extract-supplemented samples exhibited significantly greater colour differences compared to the control (P ≤ 0.05). This indicates that the addition of dried soapwort extract did indeed alter the samples’ colour. However, no significant difference (P > 0.05) was observed among the three soapwort extract-supplemented samples (SE25, SE50 and SE100). Despite the statistical significance, the measured ΔE* values were lower than 2. This is the widely accepted threshold below which colour differences are generally not easily detectable by the human eye. Thus, the change caused by the soapwort extract, while statistically real, is not likely to be perceived by consumers.

Gel hardness

Gel hardness is a critical parameter reflecting the structure and consistency of yogurt. A high-quality yogurt requires a fine, smooth texture and a firm body to retain its shape when spooned. Defined as the measure of the gel’s resistance to penetration, hardness is directly related to total dry matter content and, specifically, the protein concentration. A higher protein content facilitates greater cross-linking within the gel network, resulting in a denser and firmer structure (Tamime and Robinson, Reference Tamime and Robinson2007).

The hardness values of the experimental yogurts throughout storage are presented in Table 4. Values ranged between 244.1 and 375.8 g. Notably, on days 1 and 30 of storage, the control yogurt exhibited the highest hardness values (310.1 and 375.8 g, respectively). Statistical analysis revealed that both the addition of lyophilized soapwort extract (P < 0.01) and the storage period (P < 0.01) had significant effects on hardness, whereas their interaction was not significant (P > 0.05). According to the Duncan test applied to the mean values, the control sample (343.8 g) and sample SE25 belonged to the same statistical group (P > 0.05). However, the control sample had significantly higher hardness values than samples SE50 and SE100 (P ≤ 0.05). A clear trend was observed: as the concentration of soapwort extract increased, the hardness of the yogurt decreased. Consequently, sample SE25 (325.2 g) was significantly harder than sample SE100 (P ≤ 0.05). This significant reduction in hardness is likely attributable to the components of the soapwort extract – particularly saponin, the primary active compound (15.47%) – which may interfere with casein micelles and weaken the gel matrix. Saponins are surface-active agents that may associate with hydrophobic regions of milk proteins, potentially acting as ‘spacers’ that prevent the formation of a dense, highly cross-linked casein network during acid-induced gelation. A similar decreasing trend was observed by Paseephol et al. (Reference Paseephol, Small and Sherkat2008) in yogurts supplemented with inulin (65.2–106.0 g/0.64–1.04 N) and by Çoban et al. (Reference Çoban, Demir Özer and Yildirim2025) in yogurts fortified with dried kefir grains (340.5–456.6 g). Conversely, Delikanli and Ozcan (Reference Delikanli and Ozcan2014) reported much lower hardness values (26.50–38.74 g/0.26–0.38 N) for yogurts supplemented with whey protein derivatives compared to our findings. The pronounced differences in gel hardness across the literature are primarily driven by variations in total solids and protein concentrations, as well as the specific processing conditions and instrumental settings employed.

Table 4.

Changes in gel hardness (g), viscosity (mPa · s) and serum separation (%) of yogurt samples during storage

Table 4 Long description

Measurements compare yogurt texture and whey-off for four formulations (control and three increasing dried soapwort extract levels) at 1, 15, and 30 days of storage, with overall means also reported. Hardness increases over time on average, from about 279 grams at day 1 to about 329 grams at days 15 and 30; within formulations, the control rises from 310 to 376 grams, while the highest extract level rises from 244 to 292 grams. Across formulations, higher extract corresponds to lower mean hardness: about 344 grams for control versus about 280 grams for the highest extract level. Viscosity also increases with storage on average, from about 2124 millipascal seconds at day 1 to about 2600 at day 15 and about 2710 at day 30; formulation means are similar, ranging roughly from 2391 to 2585 millipascal seconds. Serum separation is low at day 1 (about 5.7 percent on average) but increases sharply by day 15 (about 31.1 percent) and remains high at day 30 (about 29.5 percent), with only small differences among formulations. Overall, storage time drives the largest changes, while soapwort extract mainly reduces hardness and has minimal impact on viscosity and serum separation within the reported variability.

C, control (0% dried soapwort extract); SE25, 0.025% dried soapwort extract; SE50, 0.050% dried soapwort extract; SE100, 0.100% dried soapwort extract containing yogurt sample. Values with superscript “a, b, c” letters in the same row are significantly different (P ≤ 0.05). Values with superscript “A, B, C” letters in the same column are significantly different (P ≤ 0.05).

^d Mean ± standard deviation.

^e Mean ± standard error.

Regarding the storage effect, the lowest hardness value (278.5 g) was observed on day 1. The Duncan test indicated that the mean hardness on day 1 was significantly lower than on days 15 and 30 (P ≤ 0.05). However, no statistical difference was found between day 15 (328.9 g) and day 30 (329.3 g) (P > 0.05). These findings align with Güven and Karaca (Reference Güven and Karaca2003), who reported that yogurt firmness increased by the end of a 12-day storage period, attributing this to the progressive stabilization of the casein matrix. In contrast, other studies have reported statistically insignificant changes (Paseephol et al., Reference Paseephol, Small and Sherkat2008) or irregular fluctuations (Izadi et al., Reference Izadi, Nasirpour, Garoosi and Tamjidi2015) in hardness during storage.

Gel viscosity

Viscosity is a critical parameter influencing yogurt quality, determined largely by the number and strength of bonds between casein micelles/molecules, as well as their three-dimensional structural distribution.

The viscosity values of the experimental yogurts during storage are presented in Table 4, ranging between 1959 and 2801 mPa · s. On day 1, sample SE50 exhibited the highest viscosity (2291 mPa · s), while sample SE100 showed the lowest (1959 mPa · s). By day 30, the control sample recorded the highest value (2801 mPa · s). Statistical analysis indicated that the addition of lyophilized soapwort extract did not have a significant effect on viscosity (P > 0.05). However, the storage period showed a significant effect (P < 0.001). Consequently, a Duncan test was applied to the mean values across storage days. The viscosity on day 1 was significantly lower than on days 15 and 30 (P ≤ 0.05). No significant difference was found between day 15 and day 30 (P > 0.05), with the highest mean value recorded on day 30 (2710 mPa · s). Overall, while the soapwort extract did not significantly alter viscosity, the values increased with storage time. This increase is attributed to the rearrangement of the gel structure and the strengthening of casein–casein contact points over time. Comparisons with the literature reveal that our viscosity values are lower than those reported by Paseephol et al. (Reference Paseephol, Small and Sherkat2008) (2280–4390 mPa · s) and Isleten and Karagul-Yuceer (Reference Isleten and Karagul-Yuceer2006) (5243–11,069 mPa · s). These discrepancies are expected, as viscosity is heavily influenced by factors such as starter culture type, dry matter and fat content, heat treatment intensity and homogenization. Furthermore, measurement parameters – specifically the instrument type, spindle geometry, rotational speed and temperature – play a definitive role in the resulting values.

Serum separation

Serum separation (syneresis) is an undesirable defect in yogurt that negatively affects consumer acceptance. It typically results from the rearrangement of the casein matrix during storage or from physical factors such as vibration and shear, which reduce the ability of the serum phase to remain entrapped within the three-dimensional gel network (Lucey, Reference Lucey2001). Yogurts with a densely cross-linked, branched structure and few pores exhibit less serum separation. The addition of milk-derived components, adjustments to starter culture concentration, the use of stabilizers such as xanthan gum and pectin and modifications to processing conditions are among the strategies that can be employed to reduce serum separation (Amatayakul et al., Reference Amatayakul, Halmos, Sherkat and Shah2006; Nguyen et al., Reference Nguyen, Ong, Kentish and Gras2015). Although the centrifugation method is not ideal for assessing spontaneous serum separation, it is widely used in research due to its simplicity and high reproducibility. For this reason, the centrifugation method was employed in the present study.

The serum separation values of the experimental yogurts during storage are presented in Table 4. Values ranged from 5.42 to 31.49%. At the beginning of storage (day 1), the lowest value (5.42%) was observed in the SE100 sample, while the control sample exhibited the highest (5.91%). By the end of the storage period, the control yogurt again exhibited the highest value (30.14%). Statistical analysis revealed that the addition of dried soapwort extract had no significant effect on serum separation (P > 0.05). However, the storage duration had a highly significant effect (P < 0.001). The Duncan test indicated that all storage periods differed statistically from one another (P ≤ 0.05). The lowest mean serum separation (5.68%) was recorded on day 1, while the highest mean value (31.12%) was observed on day 15. The significant increase in serum separation during storage (up to 15 days) is primarily attributed to the spontaneous rearrangement and contraction of the three-dimensional casein network. As yogurt is stored under refrigeration, casein particles continue to interact, forming additional and stronger protein–protein bonds (Lucey, Reference Lucey2001). This structural reorganization results in contraction of the gel matrix, promoting serum expulsion. However, after prolonged cold storage (up to 30 days), serum separation decreased slightly but significantly. This reduction may be explained by the partial re-entrapment of previously expelled serum within the reorganized protein network as the gel approaches a more thermodynamically stable, lower-energy structure. Serum separation values reported in the literature vary widely, largely due to differences in analytical methods and experimental conditions. Reported ranges include 18.12–33.00% (Güven and Karaca, Reference Güven and Karaca2003) and 8–24% (Isleten and Karagul-Yuceer, Reference Isleten and Karagul-Yuceer2006).

Scanning electron microscopy

The microstructure of yogurt plays a pivotal role in determining the texture, consistency and visual appearance of the gel. Under an electron microscope, an ideal yogurt gel should exhibit a network structure characterized by small, uniform pores and a high degree of cross-linking and branching (Lee and Lucey, Reference Lee and Lucey2010). The SEM images of the yogurt samples on day 30 are presented in Fig. 2. The control sample exhibited a relatively homogeneous pore size and distribution. In contrast, the samples containing dried soapwort extract displayed a more irregular gel structure. While certain regions remained compact, the pore sizes in the extract-supplemented samples varied much more widely than in the control. In addition to small pores, significantly larger voids were observed, with pore size increasing progressively as the level of soapwort extract increased. Consequently, the ratio of pore volume to total volume was higher in the extract-containing samples. Theoretically, an increase in pore volume can lead to a weaker gel structure and greater susceptibility to serum separation (Tamime and Robinson, Reference Tamime and Robinson2007). Moreover, as the concentration of soapwort extract increased, the interaction among casein molecules appeared to diminish. This was evidenced by smaller casein aggregates and a more pronounced cotton-like, amorphous appearance. The significantly lower hardness values observed in the extract-supplemented samples (discussed in the previous section) can be directly attributed to these structural modifications. However, it must be emphasized that because the samples were not subjected to chemical fixation prior to lyophilization, some of the observed microscopic pores may represent artefacts induced by ice crystallization and sublimation during processing. Nonetheless, because all experimental treatments were processed under identical freeze-drying configurations, the clear relative trend of increasing structural porosity with higher extract concentration remains a valid indicator of the extract’s destabilizing effect on the native casein network.

Figure 2.

Scanning electron micrographs of yogurts on the 30th day of storage. Magnification: 5000× (C, control; SE25, 0.025%; SE50, 0.050%; and SE100, 0.100% dried soapwort extract containing yogurt sample).

Sensory analysis

The results of the sensory analysis conducted on day 15 of storage are presented in Fig. 3. Panellists first assessed the samples for body and texture (0–7 points) and then overall acceptability (0–7 points). In the body and texture assessment, the control yogurt received the highest score (5.8), while the SE50 yogurt received the lowest (5.1). The high score obtained by the control yogurt correlates well with its instrumental viscosity and hardness values. However, this consistency was not observed for the SE50 sample, where the lower sensory score did not fully align with the instrumental data. Statistical analysis revealed that the addition of dried soapwort extract had no significant effect on body and texture scores (P > 0.05). This finding aligns with the instrumental viscosity results, which also showed no significant difference. However, it contrasts with the instrumental hardness results, which were significantly influenced by the extract. This discrepancy suggests that while the texture analyser detected significant changes in gel firmness, these differences were not substantial enough to be perceived as statistically significant by the sensory panellists.

Figure 3.

Sensory evaluation of the experimental yogurt samples (C, control; SE25, 0.025%; SE50, 0.050%; and SE100, 0.100% dried soapwort extract containing yogurt sample). Values with different letters are significantly different (P ≤ 0.05).

Regarding overall acceptability, the control yogurt received the highest score (5.4), whereas the SE100 sample received the lowest (4.0). Overall acceptability scores decreased steadily as the level of soapwort extract increased. Statistical analysis confirmed that the addition of soapwort extract significantly reduced overall acceptability (P < 0.01). However, the SE25 sample exhibited an acceptability level comparable to that of the control (P > 0.05), whereas the higher concentrations (SE50 and SE100) received significantly lower scores (P ≤ 0.05).

Based on these results, the optimum inclusion level for soapwort extract was determined to be 0.025%. The reduction in acceptability at higher levels – particularly in SE100 – was attributed to a bitter taste detected by the panellists. This bitterness is likely due to the high saponin content of the extract, as saponins are characterized by a naturally bitter taste profile.