Introduction

Beer is a fermented beverage mainly composed by water, ethanol, non-fermentable carbohydrates and a small quantity of proteins [1–5]. The proportion of these molecules determines the flavor and textural attributes defined as the mouthfeel of the beer [5, 6]. Malted barley which has an appropriate level of protein, low level of lipids and generates high levels of amylolytic enzyme activity during germination, is the most used source of carbohydrates (mainly soluble starch, arabinoxylan and β-glucan) for brewing processes [1, 5, 7]. Starch is a complex semi-crystalline structure composed basically of two homopolymers of d-glucose: amylose and amylopectin. Amylose is a linear, α-(1, 4)-linked polymer, whereas amylopectin corresponds to a highly branched polymer formed by chains of α-(1, 4)-linked glucose with α-(1, 6)-branching links [8]. The access of amylases to the granule which is essential for starch degradation [8] is determined by the grade of dissolution of the cellular wall of the starch granule [8] which depends on the activity of some proteases and β-glucanases also activated during the malting process [3, 9, 10]. Chemical degradation of the malt can be measured using the Kolbach Index which has been defined as the amount of soluble nitrogen (or protein) in a grain related to the total amount of nitrogen (or protein) [11]. Thus, malt with a Kolbach Index between 36 and 42 % is considered as highly modified and suitable for mashing [3, 9].

During mashing, the malt amylolytic enzymes break down the starch into fermentable sugars [9, 10]. Both α- and β-amylase can hydrolyze α-(1, 4) bonds. α-amylase hydrolyzes intact starch granule and degrades amylose to oligosaccharides, maltose and glucose [8]. However, amylopectin gives rise to small branched dextrins which cannot be further broken down during mashing [8, 10]. Thus, carbohydrates in beer include small residual fermentable sugars and a variable amount of higher dextrins (20–30 g/L in a pilsner beer), most of which contain two or more branches [1–5]. Dextrins are branched polysaccharides including poly-disperse species [12]. The most known are maltodextrins [13, 14] which are soluble dextrins with a higher reducing power than starch. They are usually characterized according to their dextrose equivalent (DE) which has been defined as the amount of reduced sugar expressed as glucose as a percentage of the dry substance weight. For example, maltose has a DE of 50. Maltodextrins are known to have a dextrose equivalent of less than 20 [14]. According to their DE value, maltodextrins have different physicochemical properties like solubility, freezing temperature and viscosity among others [13, 14]. However, maltodextrins with similar DE value may have very different properties depending on the hydrolysis procedure, botanical source of starch and amylose/amylopectin ratio [14].

The mouthfeel of the beer comprises three key attributes which are carbonation, palate fullness and aftertaste [15–17]. Carbonation is often the first attribute perceived in the mouth. It is felt as a particular sting or tingle that is linked to the amount of carbon dioxide in the beer [15–17]. Palate fullness refers to the weight and flow resistance of a beer while it is being consumed [15]. The physical properties associated with the palate fullness are density and viscosity [15]. The final attribute of mouthfeel is the aftertaste which is associated with the lasting sensations recognized in the mouth such as stickiness, astringency, dryness and bitterness [15–17]. The aftertaste in beer is mainly determined by the bitter that comes from the hops [17]. Beer components influencing the perception of mouthfeel, mainly dextrins [18, 19], and small amount of proteins [15] can be considered to originate from its raw materials and the variation of the mashing process [20]. The characterization of the structural properties of these components would play an important role in the development of new technologies for the optimization of brewing processes in the industry. In this work, we evaluated the influence of the molecular weight distribution of beer components on the intensity of the palate fullness of the beer.

Materials and methods

Molecular weight distribution of maltodextrins and beer samples

The range of the molecular weight distribution (MWD) of maltodextrin in a spiking trial (beer + maltodextrin) as well as of different commercial pilsner beer and experimental beers produced in the laboratory was determined. To corroborate the accuracy of the methodology by asymmetrical flow field-flow fractionation (AF4) (Wyatt Technology, Germany) for the determination of the range of the MWD of different commercial maltodextrins, the range of the MWD of the maltodextrins alone (dissolved in water) was first determined. The procedure was carried out using 5 different types of commercial maltodextrins (Lehmann & Voss & Co, Hamburg, Germany), identified as M1, M2, M3, M4 and M5. For sample preparation, the corresponding amount of maltodextrin (10 g) was weighted and dissolved in 500 mL of water to reach a concentration of 20 g/L. Carbon dioxide was removed through a filtration process. For beer-maltodextrin samples, the same procedure was carried out using the same maltodextrins (M1, M2, M3, M4 and M5), and a concentration of 20 g/L was used for analysis. Commercial beer samples as well as experimental beers were filtered before further analysis to determine the range of MWD. All the samples (maltodextrin + water, maltodextrin + beer, commercial beers and experimental beers) were analyzed using AF4. The samples were measured three times. Weight average molar mass (Mw) and the range of molecular weight distribution (MWD) were determined by multiangle laser light scattering (MALLS) (Dawn, Heleos II, Wyatt Technology, Germany) and by refractive index (RI) as quantitative detector (Agilent Series 1200 G1362A, Agilent Technologies, Germany) [8, 21]. 100 μL of each sample was injected to the channel. For the separation process, the inserted spacer had a height of 490 μm, a width of 21.5 mm at the widest position approximately where the sample was first focused (500 seg) in the long channel (240 mm) before eluted to the detectors. Measurement was carried out at 25 °C. Carrier eluent was composed of 50 mM NaNO3 and 200 ppm NaN3 dissolved in Millipore water. Separation was conducted with a 1 kDa membrane (Polyethersulfon, Wyatt Technology, Germany). Elution flow was 1 mL min−1 and cross flow was decreased from 4 to 0.1 mL min−1 during 30 min. Then cross flow was constant for 20 min at 0.1 mL min−1 and finally reduced to zero. The intensity of scattered light was measured simultaneously at 18 different scattering angles ranging from 10° to 160°. A dn/dc of 0.146 mL g−1 was used to determine the concentration of polysaccharides in aqueous solvents by the RI detector. Baselines of RI detector were subtracted by baselines from blank runs since RI baselines are influenced by different salt and pressure conditions during a cross flow gradient in AF4. Before analysis of the samples, the accuracy of the calibration of the analytics (AF4/MALLS/RI) was verified. Pullulan standards (PSS GmbH, Mainz, Germany) of known molar mass (6–708 kDa; 1 g/L) as well as bovine serum albumin (BSA) (Sigma Aldrich), (66.5 kDa; 1 g/L) were measured. The results were compared with those provided by the manufactures. Table 1 shows that the values of MWD obtained by the AF4/MALLS/RI confirmed those reported by the manufacturer. The MWD expressed as the cumulative weight fraction as a function of the molar mass was obtained using the ASTRA V 6.0.3.16 software (Wyatt Technology, Germany).

Table 1 Molecular weight of standard solutions
Full size table

Sensory analysis

Sensory analysis was carried out by the team of the chair of Brewing and Beverage Technology, Weihenstephan. Fifteen individuals certified as beer tasters by the Deutsche Landwirtschafts-Gesellschaft, German Agriculture Society (DLG) participated in this sensory evaluation. Previous to this work, they have been trained to acquire the ability to evaluate the mouthfeel of several types of beers [22]. The intensity of the palate fullness for each maltodextrin-beer sample, as well as the threshold concentration of the samples was determined. Also, the intensity of the palate fullness of different commercial and experimental pilsner beers was evaluated. The trials were conducted as follows:

Spiking trial

Preliminary trials

To establish the adequate conditions for the spiking trial, a preliminary tasting was carried out. A concentration of 40 g/L of 5 different types of commercial maltodextrins identified as M1, M2, M3, M4 and M5 was added to different samples of the same pilsner beer. A total of 6 samples were presented to the tasting panel: A control commercial pilsner beer (without addition of maltodextrin) and 5 mixed stock model solutions (maltodextrin + beer). The maltodextrins were selected according their range of molecular weight distribution. For sample preparation, the corresponding amount of maltodextrin was weighted (40 g) and then dissolved in 500 ml of beer. With the aim of establishing adequate conditions for sensory analysis such as temperature and beer sparkling (CO2 content), approximately 20 min before the tasting was held, 500 ml mixtures of maltodextrins-beer were mixed with 500 ml of fresh beer at 2 °C. The resulting samples were presented among the tasting panel (100 mL for each taste). All tests were performed at around 15 °C. Samples were evaluated according to their intensity of palate fullness. A scale from 0 (no sense) to 7.0 (extremely strong) was used according to sensory analytical procedures (DIN 10952) described by Gastl et al. [23].

Intensity of the palate fullness according to the concentration of beer-maltodextrins samples

Different concentrations (2.5, 5.0, 10.0, 20.0, 40.0 and 80.0 g/L) of distinct maltodextrins (M1, M2, M3, M4 and M5) were independently added to samples of the same pilsner beer. All samples were prepared as explained above. Thus, six samples corresponding to five different stocks of maltodextrin-beer, at the same concentration and also a standard control pilsner beer (without maltodextrin), were randomly presented to the panel. For each maltodextrin concentration, three different tasting sessions were carried out. Samples were evaluated according to their intensity of palate fullness using the procedures described above.

Threshold taste of maltodextrins in the spiking trial

The DIN ISO 4120 sensory analysis triangle test [23] was used to evaluate the threshold taste with respect to the standard pilsner beer, corresponding to each maltodextrin-beer sample. For this tasting, 3 coded samples for each concentration (1.25, 2.5, 5.0, 10.0, 20.0, 40.0 g/L) of maltodextrin were submitted, two of which were identical (i.e., 6 sets of 3 samples each), and these were arranged in ascending concentration. For this procedure, three samples were presented to the panel, and the tasters identified which of the three samples was different (odd sample). The maltodextrin taste threshold was defined as the arithmetic mean of the group threshold values at which the maltodextrins were detected unambiguously.

Palate fullness of different commercial beers

For determining the influence of the range of the MWD of pilsner beer on the intensity of the palate fullness, nine different commercial pilsner beers, similar in alcohol and extract content (Table 2), were chosen from the market. The samples from each trademark belonged to the same production batch. Sensory analysis was carried out using a ranking test according to DIN 10963/ISO 8587 [24]. Beer samples were tasted at 15 oC. The beers were each time randomly presented for tasting, and the panel had to arrange and label the samples in an ascending order according to the intensity of the palate fullness. The procedure was carried out ten times.

Table 2 Alcohol and Wort extract content of the different commercial pilsner beers
Full size table

Modification of technological parameters of the brewing process and evaluation of the MWD and the palate fullness of experimental beers

An amount of experimental beers were produced in the laboratory at a scale of 10 L. Different initial temperatures of the mashing process were tested in a range of 45, 55 and 63 °C, respectively. Figure 1 shows the evolution of the mashing procedures. All further phases of the mashing process (boiling wort, wort filtration, fermentation, beer filtration and beer filling storage) were performed identical, for all trials, according to standard procedures [25–27]. Briefly, boiling wort was performed during 80 min (98–99 °C). Primary fermentation was carried out at a temperature of 12 °C during 6 days. Secondary fermentation was carried out at a temperature of 15 °C at a pressure of 1 bar during 3 days. Lagering was performed during 3 weeks at a temperature of 0 °C. The obtained beers were stored at 2 °C. In order to obtain a good reproducibility of the results, four experiments were carried out for each temperature. The final wort concentration for each beer was of 11, 5 % W/V. For this purpose, 1.5 kg of malt for 10 L of wort was required. Common malt was made at the chair of Brewing and Beverage Technology, Weihenstephan, according to standard MEBAK procedures [25]. Malts with different Kolbach Index: 41 and 36 % (Mathe, Saaten union, Germany, produced with different steeping degree) were used [25]. All the experiments were carried out using infusion mashing process. Bitterness was similar for all trials (same quantity of hops during wort bowling) resulting in 20 IBU for each type of final beer. The range of the MWD for each resulting beer was determined using AF4/MALLS/RI as explained above. Sensory analysis was performed using a scale from 0 (no sense) to 7.0 (extremely strong) according to DIN 10952, described by Gastl et al. [23]. In this analysis, 10 tasting sessions were carried out. In each session, six beers were tasted (two for each temperature); therefore, the intensity of the palate fullness for each beer was evaluated five times.

Fig. 1
figure 1

Time evolution of the mashing process according to different initial temperatures

Full size image

Statistical analysis

Statistical analysis was carried out using GraphPad InStat version 3.05, (Graph Pad software, San Diego California, USA). Non-parametric ANOVA (Kruskall–Wallis test with Dunn’s multiple comparison test) was used to compare the mean values of the intensity of the palate fullness among the different maltodextrin-beer samples. The threshold concentration values were compared using one-way analysis of variance (ANOVA) with Bonferroni’s multiple comparison test. Friedman’stest (non-parametric repeated measures ANOVA with Dunn’s multiple comparison test) was used to compare the values of the ranks of the intensity of the palate fullness between the different commercial beers. Non-parametric ANOVA (Kruskall–Wallis Test with Dunn’s multiple comparison test) was used to compare the values of the intensities of the palate fullness of the experimental beers according to the variation of the initial temperature of the mashing process. Kruskal–Wallis test (non-parametric ANOVA) was performed to compare the molar weight distributions of the different commercial pilsner beers. Spearman’s rank correlation was used to evaluate the associations between the molar weight distributions with the intensity of the palate fullness of the beers. Repeated measures analysis of variances (Repeated ANOVA) with Tukey–Kramer multiple comparisons post-test was used to compare the differences between the molar weight distributions according to the distinct mashing procedures.

Results

MWD from different commercial pilsner beer

An amount of nine Pilsner beers with similar content of alcohol and wort extract were selected for this trial (Table 2), and the determination of the MWD was carried out by AF4/MALLS/RI as described previously. The results are shown in Fig. 2. It was found that the MWD of the beers ranged between 3 and 13 kDa. Regardless of their similar extract content, the range of the MWD differed significantly (Kruskal–Wallis Statistic: KW = 19.169; ρ = 0.0140) within the distinct selected beers.

Fig. 2
figure 2

Molecular weight distribution of different commercial pilsner beers

Full size image

MWD of standard commercial maltodextrins

In order to evaluate at which range of MWD of maltodextrins corresponded those observed previously for the commercial beers, five different types of commercial maltodextrins varying in their range of MWD were selected.

MWD of the water-maltodextrin samples

The results of this study showed that standard maltodextrins (dissolved in water) varied in a wide range of MWD (1.8–105 kDa). It can be observed in Fig. 3 that the MWD differed among the different types of maltodextrins. A good reproducibility of the values was also confirmed (standard deviation of the three samples for each type of maltodextrin). Moreover, Table 3 showed that larger elution volumes obtained by AF4 analysis corresponded to increasing ranges of MWD, indicating that the application of AF4 technique [8] was suitable for the separation of maltodextrins at this range of MWD.

Fig. 3
figure 3

Molecular weight distributions of the different commercial maltodextrins

Full size image
Table 3 Molecular weight distribution and elution volumes of different commercial maltodextrins
Full size table

MWD of the beer-maltodextrin samples

The range of MWD of different maltodextrins dissolved in samples of the same pilsner beer (20 g/L) was determined. The results are shown in Fig. 4. It can be observed that the range of MWD among the maltodextrin + beer samples varied noticeably according to the type of maltodextrin added to the sample of beer. Also, and similar to the results obtained when the maltodextrins were dissolved in water, the elution volumes increased according to the range of MWD (Table 3). Thus, the lower MWD (2.7–8.9 kDa) corresponded to the smaller elution time (14.1–17.6 min), whereas the higher MWD (26.5–108 kDa) corresponded to the largest elution time (27.1–36.2 min) (Table 4). The results showed that the MWD of the same standard maltodextrin, dissolved in beer was slightly greater than the observed when the corresponding maltodextrin, at a similar concentration, was dissolved in water. This effect may be due to the presence of other components of the pilsner beer which may elute together with each type of maltodextrin. It is important to point out that MWD of the commercial beers (3–13KDa) were similar to those found for maltodextrins with intermediated ranges of MWD (3.4–22.3 KDa).

Fig. 4
figure 4

Molecular weight distribution of different commercial maltodextrins in beer

Full size image
Table 4 Molecular weight distribution and elution volumes of maltodextrin–beer samples
Full size table

Palate fullness of the spiking trial

The intensity of the palate fullness according to the range of MWD and concentration of the maltodextrin-beer samples were determined (Table 5). As explained above, a preliminary test (at adequate conditions of temperature and beer sparkling) was carried out. In this trial, a comparison of the palate fullness between the different samples of beer-maltodextrin was performed when a high concentration of maltodextrin (40 g/L) was added to a pilsner beer sample. At this concentration, strong palate fullness was perceived by the panel for all types of maltodextrins. Moreover, a slight increase (p > 0.005) in the intensity of the palate fullness was perceived in those samples corresponding to the higher range of MWD compared to those with the lower ranges of MWD. Regardless of the intensity of the palate fullness, the panel stated that the characteristic flavor of beer was maintained in all the samples tasted. Under these conditions, we proceeded to study the possible differences in the palate fullness according to the type of maltodextrin at different concentrations of maltodextrin-beer sample using the same approach.

Table 5 Intensity of the palate fullness of the beer samples according to the concentration and the range of molecular weight distribution of maltodextrins
Full size table

Differences on the intensity of the palate fullness among distinct maltodextrin-beer samples

Table 5 shows the mean ± standard deviation of the values of the intensity of the palate fullness of different maltodextrin-beer samples, according to the maltodextrin’s range of MWD and maltodextrin concentrations of the samples. It was found that at a concentration of 2.5 g/L there were slight significant differences on the intensity of the palate fullness within the samples according to the range of MWD (Kruskall Wallis statistic: 29.34, p < 0.05). Noteworthy, a significant difference (p < 0.005) on the palate fullness was reported only for maltodextrins with the higher ranges of MWD (M4 and M5) with respect to that of the lower range of MWD (M1). At a concentration of 5 g/L, also an important difference on the intensity of the palate fullness within the samples was reported (Kruskall Wallis statistic: 39.59, p < 0.005). At this concentration, greater differences (p < 0.0001) were appreciated among the samples with lower ranges of MWD (M1 and M2) compared with those with the higher ranges of MWD (M4 and M5), and also a slight significant difference (p < 0.005) was found among samples with low ranges of MWD (M1 and M2) compared with that of intermediated range of MWD (M3). At a concentration of 10 g/L, the differences in the palate fullness within the beer samples were more noticeable (Kruskall Wallis statistic: 56.89, p < 0.0001). In addition, a great increase in the intensity of the palate fullness was perceived in those samples corresponding to the higher MWD (M4 and M5) with respect to those samples with the lower MWD (M1 and M2). Also, it was perceived an increase (p < 0.005) in the sample with intermediate palate fullness (M3) with respect to those with the lower palate fullness (M1 and M2). The same pattern was reported at a concentration of 20 g/L (Kruskall–Wallis statistic: 69.63, p < 0.0001). As mentioned above, at 40 g/L, the palate fullness was perceived as strong in all the samples tasted. Even though a significant difference within the samples was still reported (Kruskall–Wallis statistic: 34.61, p < 0.005), only those samples corresponding to the higher range of MWD (M4 and M5) showed a significant (p < 0.005) increase in the palate fullness. At a concentration of 80 g/L, an extremely strong palate fullness was perceived for all the samples such that differences with respect to the range of MWD were not appreciated by the panel.

Threshold concentration of maltodextrins

The mean ± standard deviation of the threshold concentration of the maltodextrin-beer samples are shown in Fig. 5. It was found that for the samples with the lower ranges of MWD: 2.7–8.9 kDa, the threshold concentration by which the palate fullness was perceived as different with respect to the control standard pilsner was significantly higher than the threshold concentration corresponding to the higher ranges of MWD: 12.3–51 kDa (Bonferroni’s multiple test: t = 3.609, p < 0.0001) and 26.5–108 kDa (Bonferroni’s multiple test: t = 3.601, p < 0.0001). A slightly significant difference was also observed between the threshold concentration of the sample with the lowest range of MWD: 2.7–8.9 with that of intermediated MWD: 4.5–22.3 (Bonferroni’s multiple test: t = 3.06, p < 0.05).

Fig. 5
figure 5

Threshold concentration values of maltodextrins in the spiking trial. Mean ± standard deviation of the threshold concentration values for each maltodextrin, defined as the concentration value by which the palate fullness was perceived as different with respect to the control pilsner beer (without maltodextrin)

Full size image

Palate fullness of the selected commercial pilsner beers

Table 6 shows the individual ranks of each beer arranged and labelled in an ascending order according to the intensity of the palate fullness by the tasting panel and the total rank obtained from the sum of the individual ranks for each beer. It was found that the ranks of the intensity of the palate fullness within pilsner beers differed significantly (Friedman statistic, Fr: 68.397, p < 0.0001). Applying Dunn’s multiple test for individual comparisons, it was found that beers A and B exhibited significant lower ranks of palate fullness compared to beers C, D and E (p < 0.005), being this difference extremely different between beers A and B compared to beers F, G, H and I (p < 0.0001). Also there was a slight significant difference between beer E with beers H and I (p < 0.005). There were no differences between beers A and B as well as among beers F, G, H and I. Figure 6 shows the range of MWD of the different beers arranged according to the intensity of the palate fullness by the panel (1-9). The range of MWD was determined from different points (1, 50, 70, 85, 100 %) of the cumulative weight fraction calculated for each beer sample (Fig. 2). Significant correlations were found between the increase in the range of the molecular weight distribution and the rank of the intensity of the palate fullness (r = 0.8667, Spearman ρ = 0.0045 at a cumulative weight fraction of 50 %; r = 0.8443, Spearman ρ = 0.0052 at a cumulative weight fraction of 70 %, r = 0.8234, Spearman ρ = 0.0007 at a cumulative weight fraction of 85 % and r = 0.8356, Spearman ρ = 0.002 at a cumulative weight fraction of 100 %). Also it can be seen in the figure that those molecular weight distributions corresponding to 50, 70 and 100 % from the cumulative weight fraction were more associated with the increase in the palate fullness of the beers.

Table 6 Intensity rank of the palate fullness of different commercial pilsner beer according to ten different tastings
Full size table
Fig. 6
figure 6

Association of the molecular weight distribution with the intensity rank of the palate fullness of different commercial pilsner beers. Range of MWD of the different beers arranged according to the intensity of the palate fullness. The range of MWD was calculated from 5 different points of the cumulative weight fraction determined for each beer sample

Full size image

Effect of technological aspects of brewing on the MWD and the palate fullness of experimental beers

The possible influence of differences in the mashing process and/or the quality of the raw material on the variation of the range of MWD of beer was studied. Also it was evaluated whether these variations were perceived as differences on the palate fullness by the tasting panel. Table 7 shows the associations of the ranges of MWD with the palate fullness of the beers. The resulting beers were obtained using malted barley with a Kolbach Index of 41 % and mashing procedures initiated at different temperatures (45, 55, 63 °C). It can be observed that significant higher ranges of MWD were obtained when the mashing process was initiated at 63 °C compare to that obtained when the mashing process was initiated at 45 °C (Tukey–Kramer test: q = 9.516, p < 0.001) or 55 °C (Tukey–Kramer test: q = 7.018, p < 0.001). Also, there were significant differences on the intensity of the palate fullness according to the variation of the initial temperature of the mashing process (Kruskal–Wallis statistic KW = 252.41). Therefore, beers produced at an initial temperature of 63 °C exhibited significantly stronger palate fullness compared to those produced at initial temperatures of 45 °C (p < 0.0001) and at 55 °C (p < 0.0001) (Table 7). There were no statistical differences on the MWD and on the intensity of the palate fullness using initial temperatures of 45 °C compared to those obtained using 55 °C. Table 8 shows the associations of the ranges of MWD with the palate fullness of the resulting beers according to the trials carried out using barley malted with a Kolbach Index of 36 % and at different initial temperatures (45, 55, 63 °C). Higher ranges of MWD were obtained when the mashing process was initiated at 55 and 63 °C compared to those obtained using a temperature of 45 °C (Tukey–Kramer test: q = 13.523, p < 0.0001 between 45 and 55 °C; q = 12.730, p < 0.0001 between 45 and 63 °C). There were no statistical differences among the range of MWD using initial temperatures of 55 °C compared with those observed using initial temperatures of 63 °C. It was found that those beers elaborated at initial temperatures of 55 and 63 °C showed a tendency to be perceived with greater palate fullness than those elaborated with initial temperatures of 45 °C, although a significant difference on the palate fullness was not appreciated.

Table 7 Range of molecular weight distribution and intensity of the palate fullness of experimental beers (Kolbach Index: 41 %) according to different initial temperatures of the mashing process
Full size table
Table 8 Range of molecular weight distribution and intensity of the palate fullness of experimental beers (Kolbach index: 36 %) according to different initial temperatures of the mashing process
Full size table

Discussion

The structural properties of beer components play an important role in the determination of the palate fullness. However, the quantitative associations of structural properties of dextrins as well as from other beer components with the palate fullness have not yet been well elucidated. Early studies carried out by Vermeylen [28] have shown a relationship between wort dextrin concentration and beer palate fullness. On the other hand, studies carried out by Ragot et al. [29] have shown that a concentration of 50 g/L of maltodextrin must be added to a light beer to produce a detectable increase in viscosity and thus to enhance significantly its palate fullness. Because the natural content of dextrins in a beer ranges from 10 to 50 g/L, the authors concluded that dextrins alone do not account for the viscosity or thickness of beer. In contrast, in this work, the results of the spiking trial indicated that the effect of the maltodextrin concentration on the intensity of the palate fullness was influenced by the range of MWD of the maltodextrins. Nevertheless, this effect could be appreciated only when intermediated concentrations of maltodextrins (5–20 g/L) were added to the beer samples. When lower concentrations (2.5 g/L) of maltodextrins were added, only small differences could be perceived on the intensity of the palate fullness (according to the range of MWD), whereas by the addition of higher concentrations (40–80 g/L), the intensity of the palate fullness was too strong so that the influence of the range of MWD could not be clearly appreciated. Moreover, the threshold concentration by which the palate fullness was perceived as different with respect to a control pilsner beer was also found to be highly dependent on the range of MWD of the maltodextrins. These results suggest that the range of MWD plays an important role in the production of a beer with a defined palate fullness.

The range of MWD varied significantly among the commercial beers despite their similar alcohol and wort extract content, suggesting a role of technological parameters of the brewing process in the variation of the range of the MWD of beer components as discussed below. It is important to point out that the values of the intensity of palate fullness obtained by the addition of maltodextrins of intermediated ranges of MWD in the spiking trial corresponded to those observed for the commercial beer samples, suggesting that pilsner beers contain dextrins of intermediated MWD (3.4–22.3 kDa). It is well known that dextrins are the major components (75–80 %) of the residual unfermented solids present in the beer [4, 5, 18]; however, nitrogenous compounds in a minor proportion (6–9 %) and β-glucans in very low quantities [4, 15, 29] may also contribute to the MWD of the final beverage. The variations observed in the range of MWD of the beers strongly reflected the differences in the intensity of the palate fullness perceived by the panel. In addition, it was found that variation of technological parameters of the brewing process also influenced the range of MWD of experimental beers. Thereby, using a barley malt with a Kolbach index of 41 %, significant higher ranges of MWD were obtained when the mashing process was initiated at 63 °C compared to those obtained when the mashing process was initiated at 45 or 55 °C. As expected, these beers also exhibited stronger palate fullness. It is possible that at these initial temperatures (45 or 55 °C) occurs an intense mashing process, favoring a better degradation of proteins and β-glucans. Also partial degradation of starch by amylases may occur [8]. On the other hand, gelatinization of the starch granules of this malted barley was found to occur at 62.4 °C. At this temperature, the crystalline structure of starch is broken rendering glycosidic bonds accessible to the amylases [8]. However, at 65 °C the activity of the β- amylase decreases rapidly whereas the α-amylase activity is favored, leading to the production of dextrins with high molar masses [30, 31] which may contribute to a higher range of molecular weight distribution of the total carbohydrates of the final beer. Also it must be considered that during malting temperatures can reach up to 80 °C [6, 32, 33]. This also causes partial breaking of the crystalline structure of the starch granule [8, 30, 31], allowing further a slow degradation of starch at temperatures lower than gelatinization temperature, during the mashing process. In contrast, using barley malted with a Kolbach index of 36 % and at different initial temperatures (45, 55, 63 °C) of mashing process, it was shown that higher ranges of MWD were obtained when the mashing process was initiated at 55 and 63 °C compared to those obtained using a temperature of 45°, while there were no differences using initial temperatures of 55 °C compared with those observed using initial temperatures of 63 °C. Because these experiments were performed using malted barley with a low degree of modification, the cellular wall of the starch granule would not be degraded completely during the process of malting [8, 30, 31], therefore hindering the access of the enzymes to the granule, at this range of temperature. In contrast, at 45 °C, the activity of proteases and glucanases intensively degrades the cellular wall of the granules during the mashing process [30, 31], allowing a better action of amylases, thus leading to the presence of carbohydrates of lower ranges of MWD in the final beer. According to the sensory analysis, beers produced with malted barley with a Kolbach index of 36 % exhibited a stronger palate fullness compared with those elaborated with a Kolbach index of 41 %. Among those produced with malted barley with a Kolbach index of 36 %, those elaborated at initial temperatures of 55 and 63 °C showed a tendency to be perceived with greater palate fullness than those elaborated with initial temperatures of 45 °C, although a significant difference on the palate fullness was not appreciated. In contrast, a great difference on the intensity of the palate fullness of the final beer was perceived according to the initial temperature of the mashing process using barley with a Kolbach index of 41 %. These results demonstrate that variations in the initial temperature during the mashing process influenced the range of MWD of the beer, through the modulation of the access of the amylases to the glycosidic bonds, thus leading to the production of dextrins of different MWD which in turn would determine variations in the MWD of the total carbohydrate content. Although other components such as proteins of low molecular weight contribute to the MWD [27], it must be considered that the carbohydrate content (30 g/L) of the beer is around sixfold the content of proteins (5 g/L). Therefore, our results suggested that dextrins would be an important contributor to the range of MWD of the beer, thus influencing the intensity of the palate fullness of the beer.