The effect of grinding time on particle size, colour and moisture adsorption properties of carob powder
Carob powder was prepared by vibratory grinder using processing times of 30 to 180 s in this study. Physical parameters are shown in Table 1. The grinding time had a significant effect on
Table 1
Particle size (µm), water activity (aw), and colour (CIEL*a*b*) of carob powder after vibratory grinding
| Vibratory grinding time/s |
| 30 | 60 | 90 | 120 | 180 | |
D10 | 17.8 ± 1.1b | 15.2 ± 0.5c | 16.3 ± 2.1bc | 21.4 ± 0.8a | 19.4 ± 0.3a | |
D50 | 141.3 ± 2.5a | 87.9 ± 1.7d | 110.1 ± 3.5c | 129.7 ± 2.9b | 135.1 ± 1.8b | *** |
D90 | 427.7 ± 12.5d | 572.7 ± 8.1b | 494.6 ± 7.9c | 810.6 ± 21.3a | 429.7 ± 5.6d | |
Span | 2.9 | 6.3 | 4.3 | 6.1 | 3.0 | |
aw | 0.380 ± 0.005d | 0.374 ± 0.002d | 0.403 ± 0.001c | 0.417 ± 0.001b | 0.433 ± 0.003a | *** |
L* | 47.1 ± 0.5b | 51.1 ± 0.4a | 49.4 ± 0.2c | 48.0 ± 0.3d | 47.0 ± 0.3b | *** |
a* | 6.3 ± 0.4b | 5.8 ± 0.6b | 6.3 ± 0.3bc | 6.8 ± 0.4ac | 6.9 ± 0.2a | *** |
b* | 19.0 ± 0.5b | 17.9 ± 0.4c | 18.7 ± 0.3b | 20.1 ± 0.4a | 19.8 ± 0.3a | *** |
Results are expressed as mean ± standard deviation (particle size N = 2; aw N = 3; colour N = 5). Different letters in superscript indicate significant differences in row according to Duncan multiple pairwise test (p < 0.05). Significant effect of grinding time using ANOVA was marked as ***=p < 0.001. D10, D90, and D50 = lower decil, upper decil, and a median of cumulative weight
the particle size distribution. The D50 values decreased significantly from 141.3 ± 2.5 to 87.9 ± 1.7 µm (p < 0.001) when ground for 30 and 60 s, respectively. A further increase in grinding time resulted in a gradual increase in D50 values up to 135.1 ± 1.8 µm for CP180. A similar pattern was observed for the D10 values. However, the particle size in the upper percentile (D90) showed a different behaviour; high D90 values of 527.7 ± 8.1 and 810.6 ± 21.3 µm were observed for CP60 and CP120, respectively. This finding also corresponded with high span values showing the lack of uniformity of the distribution (spans 6.3 and 6.1, respectively.). It was previously published that grinding time significantly influenced the particle size distribution of various food powders [10–13]. Reduction in particle size was observed with increasing ball milling time for onion peel [10], ginger rhizome [11], horseradish [12] or soybean protein isolate [13] powder. The long grinding time resulted in an increase in the value of D90 caused by the sticking of the particles together and the formation of agglomerates for horseradish powder [12]. The agglomeration process occurred during the entire grinding of the dried mushrooms and was effectively disturbed by ultrasound treatment [9].
The grinding time of the carob pods significantly influenced all the values that describe the colour of the carob powder (p < 0.01). Lightness increased after 60 s of griding (L*= 51.1 ± 0.4), then gradually decreased with the increase of processing time (Table 1). Carob powder ground for 120 s and 180 s had higher values of redness (a*=6.8 ± 0.4 and 6.9 ± 0.2, respectively) and yellowness (b*=20.1 ± 0.4 and 19.8 ± 0.3, respectively).
Different colours can be visually observed only for CP60 in our study. Correlation analysis revealed a strong association between the median particle size (D50) and the colour in terms of L* (r=-0,961, p < 0.001), a* (r = 0.634, p < 0.01), and b* (r = 0.764, p < 0.001). Interestingly, all of the colour values exhibited an association with the water activity of the carob powder. Whereas L* value showed negative association with aw (r=-0.548, p < 0.05), the increase in aw corresponded to increase in values of a* (r = 0.781, p < 0.01) and b* (r = 0.795, p < 0.001). The water activity of carob powder was similar for CP30 and CP60, then its increase was observed with increased grinding time. It might seem that the smaller the particles, the lower the water activity, as can be seen in Table 1. However, the correlation coefficient indicates a weak and negligible association (r = 0.464, p > 0.05).
Colour changes have been observed during the grinding of flours [6], rose-myrtle powder [7], or black kidney bean powder [8]. Drakos et al. [6] found that the colour changes during grinding were product-specific, i.e., different changes in the values of a* and b* were determined for jet milled rye and barley flours. Various fractions of particle size of kidney bean powder exhibited a weak association with colour stimuli, for example, the values of b* decreased with increasing particle size from 125 to 250 µm but increased significantly with a further reduction in particle diameter [8]. It should be noted that colour changes can also be affected by increasing the temperature during milling. The colour of the carob powder was influenced by temperature treatment, including roasting [27] or spray-drying [28]. The temperature of the metal equipment after 30 and 180 s of grinding was 22–24°C and 37–39°C, respectively, in this study.
According to the IUPAC classification, the shape of the moisture adsorption isotherms of all carob powder samples is type III [29], that is, an increase of EMC to ~ 25 mg/g in the range of 0–40% RH followed by a steep increase to ~ 259–268 mg/g in the range of 40 to 80% of RH. This is an obvious behavior for high-sugar food products [30]. The differences in the adsorption isotherm plots for the carob powder samples are barely visible; therefore, we compared the
EMC for two levels of RH. As can be seen in Fig. 1, EMC was significantly higher (p < 0.05) for the CP60 sample at both 10% (7.1 mg/g) and 60% (79.3 mg/g) of RH compared to other grinding times. EMC decreased with increasing grinding time from 60 to 180 s. These findings corresponded to the particle size distribution, where a strong association was found between D50 and EMC at 10% (r=-0.883, p < 0.05) and 60% (r=-0.884, p < 0.05) of RH. It was well documented that reducing the particle size distribution results in improved water adsorption measured by water holding capacity, i.e., the smaller the particles, the higher the surface area available for moisture uptake [5, 14].
Grinding Time As Affected Antioxidant Properties And Phenolic Content In Carob Powder
The effect of grinding time on the phenolic content and antioxidant properties is presented in Table 2. ANOVA revealed that the grinding time had a significant effect on all variables (p < 0.001) excluding vanillic acid. Cinnamic acid was the most abundant phenolic constituent with a maximum value of 4.41 ± 0.08 µg/g in the CP180 sample. The carob powder ground for 180 s had a significantly higher content of all phenolics determined in this study compared to the CP120 sample. The TPC values ranged from 4.73 ± 0.21 to 6.07 ± 0.15 GAE/(mg/g), with a maximum value for CP180. The TFC value was significantly higher for CP60 (0.33 ± 0.13 QUE/(mg/g)) compared to CP30 (p < 0.05). The increase in grinding time from 90 to 120 s led to the same TFC values (p > 0.05). Although the CC values appeared to increase with increased grinding time, the differences were not statistically significant (p > 0.05). DPPH values increased after 60 s of grinding to 11.91 ± 0.51 TEAC/(mg/g) (p < 0.05), then decreased significantly for CP90 (p < 0.05) followed by a further increase reaching its maximum value of 11.93 ± 0.74 TEAC/(mg/g) for CP180. FRAP values followed the same pattern with the highest antioxidant capacity for the CP120 sample (16.95 ± 1.08 TEAC/(mg/g)). As was expected, positive associations were found between antioxidant properties in terms of DPPH and FRAP assays, and TPC, TFC, and catechin content (r = 0.656–0.836, p < 0.05). The particle size distribution was associated with the release of phenolic compounds during the extraction procedure [15]. In contrast, some experiments did not confirm the correlation between particle size and phenolic content [16] or antioxidant capacity [6], probably due to different materials, grinding techniques, or particle size ranges. In our study, weak negative association (p > 0.05) between D50 and TPC (r=-0.421), ferulic acid (r=-0.439), and naringenin (r=-0.460) was observed. The higher ability to scavenge DPPH radical was significantly associated with lower D50 values (r=-0.606, p < 0.05). It suggests that carob powder with small particle size can release more phenolics during extraction; however, other factors should also be considered.
Table 2
Phenolic content and antioxidant properties of carob powder prepared by vibratory grinding at various processing times
| Vibratory grinding time/s |
| 30 | 60 | 90 | 120 | 180 | ANOVA |
w/(µg/g) | | | | | | |
Vanillic acid | 3.95 ± 0.40ab | 4.17 ± 0.40ab | 3.58 ± 0.43ab | 4.08 ± 0.26b | 4.41 ± 0.08a | |
Ferulic acid | 8.64 ± 0.04d | 10.77 ± 0.22c | 10.85 ± 0.43c | 10.09 ± 0.51b | 11.28 ± 0.09ac | *** |
Cinnamic acid | 45.20 ± 1.20c | 50.30 ± 1.13b | 47.80 ± 1.64bc | 50.00 ± 2.15b | 54.28 ± 1.42a | *** |
Luteolin | 13.62 ± 1.82b | 16.11 ± 1.44b | 15.17 ± 1.74b | 16.97 ± 1.00b | 20.52 ± 0.57a | ** |
Naringenin | 2.76 ± 0.06c | 6.77 ± 0.25ab | 6.67 ± 0.44b | 7.28 ± 0.30a | 7.28 ± 0.13a | *** |
Apigenin | 1.29 ± 0.19d | 1.94 ± 0.05c | 2.17 ± 0.06ab | 1.97 ± 0.11bc | 2.27 ± 0.04a | *** |
w/(mg/g) | | | | | | |
TPC as GAE | 4.73 ± 0.21d | 5.84 ± 3.98ae | 5.41 ± 2.19c | 5.72 ± 0.20be | 6.07 ± 0.15a | *** |
TFC as QUE | 0.19 ± 0.15b | 0.33 ± 0.13ac | 0.27 ± 0.14bc | 0.32 ± 0.07bc | 0.39 ± 0.05ac | * |
CC as CAT | 0.33 ± 0.13a | 0.36 ± 0.03a | 0.35 ± 0.12a | 0.44 ± 0.06a | 0.46 ± 0.19a | |
DPPH as TEAC | 9.29 ± 3.16b | 11.91 ± 0.51a | 9.80 ± 0.53b | 9.59 ± 1.62b | 11.93 ± 0.74a | * |
FRAP as TEAC | 13.78 ± 1.33b | 16.11 ± 1.80c | 13.12 ± 1.28b | 16.95 ± 1.08ac | 16.52 ± 0.65a | *** |
All results were on a dry mass basis and expressed as mean ± standard deviation (N = 3). Different superscript letters indicate significant differences in row according to Duncan multiple pairwise test (p < 0.05). Significant effect of grinding time using ANOVA was marked as *=p < 0.05, **=p < 0.01, ***=p < 0.001. TPC = total phenolic content, GAE = gallic acid equivalent, TFC = total flavonoid content, QUE = quercetin equivalent, CC = catechin content, CAT = catechin equivalent, DPPH = 1,1-diphenyl-2-picrylhydrazyl radical assay, FRAP = ferric reducing antioxidant capacity, TEAC = Trolox equivalent antioxidant capacity
The Effect Of Grinding Time On The Bioaccessibility Of Antioxidant Properties And Phenolic Content
The bioaccessibility of phenolic compounds from carob products was recently studied [18–20]. Chait et al. [18] observed an increase in free phenolic compounds, but the degradation of bound and conjugated compounds was determined after complete in vitro digestion. Pure phenolic acids and flavonoids were degraded during the oral, gastric, and intestinal digestion process, but the authors found that bioaccessibility depended on the type of the product [19]. For example, TPC retention was similar for carob powder, syrup, fibre, and extract. However, higher retention of DPPH was observed for carob powder than for carob extract.
Digestion liquid was used to examine the effect of various griding times on the release of phenolic substances from carob powder after three-stage in vitro digestion process in our study. We assume that phenolics in digestion fluid are readily absorbed in the small intestine.
As can be seen in Table 3, luteolin and apigenin were the flavonoids most affected by the digestion process of carob powder samples in this study. Luteolin content decreased by 12.5–19.5 µg/g, which is equivalent to 5–8% of its initial values. Apigenin content was reduced to 94–95%. The losses of apigenin were similar to those observed in deffated lupin seed during two-stage in vitro digestion [31]. Ferulic and cinnamic acid were the most stable during in vitro digestion showing 42–51% and 33–39% of bioaccessibility. The grinding time strongly affected the content of vanillic acid (p < 0.001), cinnamic acid (p < 0.001), luteolin (p < 0.01) and apigenin (p < 0.001) after three-stage digestion process, but no trend can be pointed out. Whereas vanillic acid content was the highest in CP30, the maximal content values for cinnamic acid, luteolin and apigenin were observed in different CP samples. Particle size did not influence the phenolic content after in vitro digestion. By comparing the bioavailability of substances in carob powder samples, we get a clearer overview of their fate during digestion. The bioaccessibility was remarkably different in CP30 and CP60 samples. For example, 23% and 14% of vanillic acid
Table 3
Phenolic content and antioxidant properties of carob powder after in vitro digestion, and its bioaccessibility
| Vibratory grinding time/s |
| 30 | 60 | 90 | 120 | 180 |
w/(µg/g) | | | | | | |
Vanillic acid | 0.91 ± 0.01a (23) † | 0.57 ± 0.03b (14) | 0.56 ± 0.00b (15) | 0.55 ± 0.01b (13) | 0.67 ± 0.23b (12) | *** |
Ferulic acid | 4.45 ± 2.12ab (51) | 4.28 ± 0.17b (40) | 4.68 ± 0.28ab (43) | 4.65 ± 0.24ab (46) | 4.71 ± 0.18a (42) | |
Cinnamic acid | 17.69 ± 0.37b (39) | 18.34 ± 0.38a (36) | 17.35 ± 0.35b (36) | 16.61 ± 0.28c (33) | 18.02 ± 0.14a (33) | *** |
Luteolin | 1.15 ± 0.03a (8) | 1.16 ± 0.04a (7) | 1.16 ± 0.03a (8) | 1.05 ± 0.04b (6) | 1.05 ± 0.07b (5) | ** |
Naringenin | 0.80 ± 0.02a (29) | 0.81 ± 0.02a (12) | 0.83 ± 0.01a (12) | 0.83 ± 0.03a (11) | 0.80 ± 0.03a (11) | |
Apigenin | 0.08 ± 0.00c (6) | 0.09 ± 0.01c (5) | 0.12 ± 0.00a (6) | 0.10 ± 0.01b (5) | 0.10 ± 0.01b (5) | *** |
w/(mg/g) | | | | | | |
TPC as GAE | 5.10 ± 0.10c (109) | 5.51 ± 0.27a (94) | 5.09 ± 0.12c (94) | 5.66 ± 0.33a (99) | 5.32 ± 0.29b (88) | ** |
TFC as QUE | 0.24 ± 0.02ab (126) | 0.28 ± 0.05a (85) | 0.20 ± 0.01b (74) | 0.23 ± 0.03b (72) | 0.19 ± 0.02b (49) | * |
CC as CAT | 0.59 ± 0.23a (148) | 0.52 ± 0.18a (141) | 0.50 ± 0.13a (143) | 0.57 ± 0.18a (129) | 0.53 ± 0.16a (116) | |
DPPH as TEAC | 10.24 ± 0.47b (110) | 12.17 ± 1.23a (102) | 11.21 ± 0.64ab (114) | 11.12 ± 1.16ab (116) | 11.09 ± 1.26ab (93) | * |
FRAP as TEAC | 13.97 ± 0.26 (101)d | 16.12 ± 1.43a (100) | 15.07 ± 0.47bc (115) | 16.05 ± 0.69a (95) | 14.70 ± 0.31cd (89) | *** |
† bioaccessibility in % of initial content (in brackets); different superscript letters indicate significant differences in row according to Duncan multiple pairwise test (p < 0.05); significant effect of grinding time using ANOVA was marked as *=p < 0.05, **=p < 0.01, ***=p < 0.001.TPC = total phenolic content, GAE = gallic acid equivalent, TFC = total flavonoid content, QUE = quercetin equivalent, CC = catechin content, CAT = catechin equivalent, DPPH = 1,1-diphenyl-2-picrylhydrazyl radical assay, FRAP = ferric reducing antioxidant capacity, TEAC = Trolox equivalent antioxidant capacity
was observed after digestion of CP30 and CP60 samples, respectively. Further increase of grinding time resulted in similar retention as for CP60 sample. The same pattern was observed for ferulic acid (decrease from 51–40%) and naringenin (decrease from 29 to 12%). It should be noted that the level of bioaccessibility is dependent on the experimental procedure applied for initial content determination. The in vitro digestion process takes place in an aqueous environment, so the extraction of phenolics into water solution is a better option [32, 33]. However, water/organic solution mixtures were also used for the determination of the initial content of phenolic compounds [34, 35], which were then used for the bioaccessibility calculation.
Total phenolic (p < 0.01) and flavonoid (p < 0.05) contents of carob powder after in vitro digestion were affected by the grinding time. While CP60 and CP120 have the highest TPC values (5.51 ± 0.27 and 5.66 ± 0.33 mg GAE/g, respectively), the maximum TFC values were observed in CP30 and CP60 samples. Catechin content was similar for all the carob samples. Whereas bioaccessibility of phenolic individuals decreased after digestion of carob powder ground for 30 s, meanwhile TPC, TFC and CC increased to 109, 126, and 148%, respectively. This discrepancy can be explained by the subsequent release of other phenolic compounds after in vitro digestion process. An increase of total phenolic content and at the same time the decrease of some hydroxybenzoic acids after gastrointestinal digestion of soursop was reported [36]. Although the reduction of vanillic acid, caffeic acid, and catechin content in Martricaria recutita flower after duodenal digestion was observed, TPC value increased [37]. This was evidenced by the increase of rutin, quercitrin or quercetin content in their research. Further increase of grinding time resulted in the decrease of both TPC and TFC bioaccessibility reaching their minimum level of 88% and 49% for CP180 sample, respectively. Bioaccessibility of catechins also exhibited a gradual decrease with the increase of grinding time, but it was still improved in carob powder ground for the longest time. Antioxidant properties of digestive liquid ranged from 10.24 ± 0.47 to 12.17 ± 1.23 mg Trolox/g using DPPH assay, and from 13.97 ± 0.26 to 16.12 ± 1.43 mg Trolox/g using FRAP assay. The bioaccessible phenolic, flavonoid and catechin contents were not affected by the particle size of carob powder. It is likely that the bioaccessibility of phenolics can be influenced by particles smaller than those used in our study. Li et al. [38] observed that bioaccessible phenolic content and antiradical activity in wheat bran were enhanced in the powder fraction smaller than 19.16 µm. Our findings suggest that carob powder ground for 30 s represents a good source of bioactive substances.