Rheological and physical properties of Inulin-containing milk chocolate prepared at different process conditions

Abstract

In the present study, the effects of inulin as a prebiotic substance at various levels (PSL) (0, 60, 90 and 120 g/kg) in milk chocolate, as well as the use of varying conching time (CT) (3.50, 4.0 and 4.50 h) and refining conditions (20.0, 25.0 and 28.0 µm as the mean particle size, D_[4,3]) in the sample preparation process were examined with respect to changes in the physical (colour, hardness, water activity) and rheological properties of the samples. Specific surface area (SSA) and largest particle size (D₉₀) values increased with increasing values of PSL; however, intended D_[4,3] values were reached for all PSL values. CT, PSL and PSD (particle size distribution) had no significant effect on the brightness and chroma in the inulin-containing milk chocolate; however, it has been found that hardness, water activity, yield stress and viscosity produced significant changes (P < 0.01).

En el presente estudio, los efectos de la inulina como una sustancia prebiótica a varios niveles (PSL) (0, 60, 90 y 120 g/kg) en el chocolate con leche, así como el uso de variaciones en el tiempo de conchado (CT) (3,50, 4,0 y 4,50 h) y las condiciones de refinado (20,0, 25,0 y 28,0 µm como tamaño medio de las partículas, D_[4,3]) en el proceso de preparación de la muestra se examinaron con respecto a los cambios en las propiedades físicas (color, dureza, actividad de agua) y reológicas de las muestras. Los valores de área de superficie específica (SSA) y mayor tamaño de partícula (D₉₀) aumentaron con los valores crecientes de PSL; sin embargo,los valores D_[4,3] buscados se alcanzaron con todos los valores de PSL. CT, PSL y PSD (distribución de tamaño de partícula) no tuvieron ningún efecto significativo sobre el brillo y el croma en el chocolate con leche que contiene inulina; sin embargo, se ha encontrado que la dureza, la actividad del agua, el límite elástico y la viscosidad tuvieron cambios significativos (P < 0,01).

Keywords:

• chocolate
• inulin
• conching
• particle size distribution
• prebiotic

Palabras clave:

• chocolate
• inulina
• conchado
• distribución del tamaño de partícula
• prebiótico

Introduction

Modern healthier foods require modifications in ingredients and recipe formulation that impact product rheology and structure (Afoakwa, Paterson, Fowler, & Ryan, 2009). In developed economies, a key trend at the moment is confectionary products that deliver functional benefits for health and well-being, such as sugarless sweets and functional chocolate (Belscak-Cvitanovic et al., 2012).

Recently, dietary fibres have become popular as functional food additives. These fibres are used because they have a low digestive energy content, a high capacity to absorb water and are said to support the digestive system (Bolenz, Amtsberg, & Schape, 2006). A prebiotic can be defined as a non-digestible food ingredient that beneficially affects the human body by selectively stimulating the growth and/or activity of one or a limited group of colon bacteria (Aragon-Alegro, Alegro, Cardarelli, Chiu, & Saad, 2007; Volpini-Rapina, Sokei, & Conti-Silva, 2012). Prebiotics can be applied to a variety of foods (Morris & Morris, 2012; Volpini-Rapina, 2012). Inulin is a plant-derived carbohydrate and a polydisperse β (2→1) fructan which is bifidogenic, prebiotic and fermentable dietary fibre (Farzanmehr & Abbasi, 2009). It presents 10% of the sweetness power of sucrose (Franck, 2002), allowing it to partially replace sucrose in certain formulations (De Castro, Cunha, Barreto, Amboni, & Prudencio, 2009; Villegas, Tarrega, Carbonell, & Costell, 2010; Wang, 2009). Whether a prebiotic can be used in chocolate production depends on its functional properties as a sugar replacement and a fibre (Wang, 2009). Inulin has also found applications in chocolate as a low-calorie bulking agent that can be added without adding sugar (Roberfroid, 1999).

Because milk chocolates are the most preferred chocolates among children, as well as a high percentage of adults (Belscak-Cvitanovic et al., 2012), in this study, inulin was used as a fibre and as a prebiotic substance to formulate a functional milk chocolate. In the previous studies, chocolates with 50 g/kg inulin, added either as a fibre or as a prebiotic substance, did not show considerable differences from the standard samples (Morris & Morris, 2012). Therefore, we added inulin to samples in mass ratios greater than 50 g/kg.

Chocolate is, in essence, cocoa mass and sugar suspended in a cocoa butter matrix (Andarea-Nightingale, Lee, & Engeseth, 2009). Chocolate production principally consists of five stages: mixing the ingredients, refining, conching, tempering and, finally, crystallisation (Schumacher et al., 2009). The production process greatly influences the quality of the final product (Cidell & Alberts, 2006). The refining process must result in the correct particle size (Bolenz, Thiessenhusen, & Schape, 2003). Processing technology is just as important as using the right ingredients for achieving the desired quality. Refining chocolate using a three- or five-roll refiner leads to a reduction of particle size, which is an important step towards obtaining a smooth texture (Lucisano, Casiragh, & Mariotti, 2006; Torres-Moreno, Tarrega, Costell, & Blanch, 2012). Final particle size critically influences the rheological and sensory properties of milk chocolate (Afoakwa, Paterson, & Fowler, 2007; Torres-Moreno et al., 2012). Conching is normally carried out by agitating chocolate at high temperatures (>40°C) and is regarded as an essential process for obtaining an adequate texture of chocolate mass. Variation in conching time (CT) and temperature can affect the viscosity, final texture and flavour of the chocolate (Bolenz et al., 2003; Bolenz, Amtsberg, & Lipp, 2005; Bordin et al., 2009; Torres-Moreno et al., 2012). The function of the conch was initially attributed to be reducing the particle size and guaranteeing the fluidity of the mass (Schumacher et al., 2009). The flow properties of molten chocolate are affected by the processing steps (refining, conching and tempering), as well as by the composition of the chocolate (amount of fat, amount and type of emulsifiers, particle size distribution) (Fernandes, Müller & Sandoval, 2012).

Regarding the fat requirements for obtaining desirable flow properties, Afoakwa, Paterson, and Fowler (2008) and Beckett (1999) concluded that (1) the largest particle size and specific area of solids are two key parameters; (2) the particle diameter has an impact on the coarseness and (3) the surface area of the particles is important. As particle size increases, particles become more spherical, leading to a broadening of the PSD (particle size distribution) with a consequent reduction in solid loading as the fat content increases. Reduction in specific surface area (SSA) with increasing particle size, as observed in the PSDs of individual components, has been reported in previous studies (Afoakwa, Paterson, Fowler, & Vieira, 2008a; Beckett, 1999; Sokmen & Gunes, 2006; Ziegler & Hogg, 1999). Smaller particles improve sensory properties of chocolate (Ziegler, Mongia, & Hollender, 2001). PSD influences rheological and textural properties of molten and tempered dark chocolates, affecting the microstructure, product spread, tempering and pre-crystallisation behaviour, hardness and sensorial qualities (Afoakwa, Paterson, Fowler, & Vieira, 2008b).

Briefly, to determine the effect of process conditions and composition in producing a prebiotic milk chocolate containing inulin at various levels (60, 90 and 120 g/kg), we investigated the effects of refining (targeted mean particle size) and conching (processing time) on the rheological properties (such as yield stress, viscosity and rate index) and physical properties (such as colour [brightness, hue angle and chroma], water activity and hardness) of milk chocolate samples.

Materials and methods

Materials

For the preparation of milk chocolate samples containing prebiotic substances, the following compounds were used: cocoa butter, cocoa mass (Altinmarka, Istanbul, Turkey), fine sugar (SMS Kopuz, Istanbul, Turkey), whole milk powder (Besel, Konya, Turkey), soy lecithin (Bremtag Chemistry, Istanbul Turkey), polyglycerol polyricinoleate (PGPR) (Palsgaard, Zierikzee, Netherlands), vanillin (Ekin Chemistry, Istanbul, Turkey) and inulin (Beneo Orafti, Oreye, Belgium). All materials were obtained from Tayas Food Company (Gebze, Kocaeli, Turkey).

Apparatus

A pilot-scale chocolate line and temper machine (Aasted, Farum, Denmark), pilot-scale refiner (Lehman, Asslen, Germany) and pilot-scale conch (BSA Schneider Anlagentechnik, Aachen, Germany) were used for preparing chocolate samples. In addition to general laboratory equipment, a TA-TXPLUS Texture Analyser (Microstable Systems, UK), LA-300 Laser-Scattering Particle Size Distribution Analyser (Horiba Scientific, USA), CR400 colorimeter (Konica Minolta, Japan), Master aw Water Activity Measurement device (Novasina, Switzerland) and a rheometer (Brookfield R/S Plus, USA) were used for the analysis.

Sample preparation

Each sample group was prepared in a 10-kg batch using the formulations presented in Table 1. For this purpose, the melted fat components (20% of the total cocoa butter) and dry powders (fine sugar, milk powder and cocoa mass with or without inulin) were mixed until homogeneous and simultaneously warmed to 40°C. At the end of the mixing and warming, the chocolate mass (nearly 5.98% fat) was first pre-refined on a pilot scale, 3-roll refiner (Lehmann, Aaelen, Germany), subsequently mixed again and warmed to 50°C. To achieve a mean particle size of 20.0, 25.0 and 28.0 µm, the gap size/pressure between rollers of the three-roll refiner was adjusted and the PSD was measured using a laser-diffraction particle size analyser (Horiba, USA). After measuring the particle size, dry conching was performed for 45 min, and the remaining cocoa butter (80% of the total), vanillin, soy lecithin and PGPR were then added (up to 24.0% fat). The total CT was 210, 240 and 270 min at 55°C.

Table 1. Formulation used for the chocolate samples (g/kg).

Table 1. Formulaciones utilizadas para las muestras de chocolate.

		Sample group
Ingredient	Control (K)	A	B	C
Fine sugar	390.0	350.0	340.0	320.0
Cocoa butter	240.0	240.0	240.0	240.0
Cocoa mass	140.0	130.0	124.5	124.5
Whole powdered milk	224.5	214.5	200.0	190.0
Soy lecithin	3.0	3.0	3.0	3.0
PGPR	2.20	02.20	2.20	2.20
Vanilla	0.30	00.30	0.30	00.30
Inulin	0.0	60.0	90.0	120.0

Afterwards, a three-stage tempering process (33–35, 24–25 and 25–26°C) was implemented (temper index value, as measured by temper metre [Chocometer, Aasted Farum, Denmark]: 5.50–6.0). Subsequently, the moulding and vibration process (Aasted Farum, Denmark) was conducted at 27–30°C. After 20 min of cooling (Aasted Farum, Denmark) at 5°C, the process was completed, with the samples output at a temperature between 13 and 15°C. The samples were stored away from light and heat prior to analysis.

In the preparing of the sample formulation, the quantity of the components that have the characteristics of dry powder are decreased in accordance with the amount of inulin to be added. Moreover, the quantities of both cocoa mass and butter in the samples from groups B and C (Table 1) are kept constant to monitor the effects of the change in the amount of inulin.

Experimental design

To determine the physical and rheological properties of milk chocolate containing various levels of the prebiotic substance inulin, three experimental variables were used: the CT, prebiotic substance level (PSL), and PSD (Table 2). Other variables, including conching temperature, were held constant. In sample formulations, depending on the percentage of inulin used, the quantities of ingredients in dry powdered forms were decreased for each sample group according to Table 1.

Table 2. Experimental design of the study.

Table 2. Diseño experimental del estudio.

CT (h)	0 g/kg Inulin	60g/kg Inulin	90 g/kg Inulin	120 g/kg Inulin	D[4,3] (μm)
3.50	K1.1	A1.1	B1.1	C1.1
4.0	K1.2	A1.2	B1.2	C1.2	20.0
4.50	K1.3	A1.3	B1.3	C1.3
3.50	K2.1	A2.1	B2.1	C2.1
4.0	K2.2	A2.2	B2.2	C2.2	25.0
4.50	K2.3	A2.3	B2.3	C2.3
3.50	K3.1	A3.1	B3.1	C3.1
4.0	K3.2	A3.2	B3.2	C3.2	28.0
4.50	K3.3	A3.3	B3.3	C3.3

CT, conching time; D_[4,3], mean particle size.

A 3 × 3 × 3 factorial design was used that included the following:

1. CT; 3.50, 4.0 and 4.50 h

2. PSL; 60, 90 and 120 (g/kg)

3. PSD; 20.0, 25.0 and 28.0 µm (mean)

Particle size distribution

A method for determining the PSD of samples was followed according to the method used by Afaokwa, Paterson, and Fowler (2008). A LA-300 Laser-Scattering Particle Size Distribution Analyser (Horiba, USA) was used. Approximately, 0.20 g of each chocolate sample was dispersed in vegetable oil (refractive index, RI: 1.45) at ambient temperature (20 ± 2°C) until an obscuration of 0.20 was obtained. Ultrasonic dispersion was maintained by stirring for 2 min to ensure the particles were freely dispersed. The size distribution was quantified as the relative volume of particles in size bands presented as size distribution curves. Data analysis was based on Mie theory used by Do, Hargreveas, Wolf, Hort, and Mitchell (2007). The obtained PSD parameters included the following: D_[4,3] (μm), mean; D_[3,2] (μm), median; D₉₀ (μm), diameter for which 90% of particles were smaller in size; SSA (cm²/cm³); D₁₀ (μm), diameter for which 10% of particles were smaller in size; [(D₉₀ − D₁₀)/D₅₀] (μm), span; and standard deviation (St Dev).

Hardness measurements

The mechanical properties of the chocolates, such as the hardness, were measured using a TA-TXPlus Texture Analyser (Stable Micro Systems, UK). A cylindrical, flat-ended, stainless steel probe with a diameter of 2 mm was used to penetrate each sample to a depth of 5 mm at a speed of 1 mm/s at 20 ± 2°C. The trigger force was set to 0.05 N. The results for the hardness (N) are expressed as the mean value of five replicates conducted on different samples of the same lot of each chocolate.

Determination of water activity

A 10.0-g quantity of each milk chocolate sample was homogenised, and 2.0 g of the homogenised sample was used to determine the water activity (aw) at 25°C using a Lab-Master aw (Novasina, Switzerland); aw values of each sample were measured in triplicate after a follow-up day of sample preparation.

Colour measurement

Instrumental analyses were performed according to the instruction manual of a portable colorimeter (Chroma Meter CR-400, Konica Minolta, Japan), obtaining L*, the luminance, ranging from 0 (black) to 100 (white); as well as a* (green to red) and b* (blue to yellow), with values from −120 to +120. All of these values were obtained at 25°C using the CIELAB System. The colour parameters in the study were brightness (L*) and chroma (c* = [(a*²) + (b*²)]^1/2). All data were expressed as the mean value of five replicates conducted on different samples of the same lot of each chocolate.

Rheological measurement

Rheological properties of the milk chocolate samples were carried out in a rheometer, (Brookfield R/S Plus, USA) using bob and cup geometry, according to a modification of the method used by Sokmen and Gunes (2006). Each chocolate sample was incubated at 50°C for 75 min, melted, transferred to the rheometer and sheared at a rate of 5.0 s⁻¹ for 10 min at 40°C in the rheometer before the measurement cycles were started. The shear rate was ramped up from 5.00 to 60.0 s⁻¹ over a period of 120 s and was ramped back down from 60.0 to 5.0 s⁻¹. During each ramp, 50 measurements were taken. This measurement cycle was repeated 30 times consecutively until thixotropy was eliminated from the samples. The data from the 30th measurement were analysed according to the Casson model, and related rheological parameters, such as yield stress and viscosity, were determined.

Statistical analysis

The results were analysed using a factorial design with analysis of variance (ANOVA). Duncan’s multiple comparison test was applied to determine if there were significant differences. Statistical analyses were performed using the MINITAB-15 (Minitab Inc., State College, PA) and MSTAT statistical packages (Michigan State University, East Lansing, MI). Values with P < 0.01 were considered significant.

Results and discussion

Particle size distribution

Particle sizes in chocolate range between approximately 1 and 50 µm (Do et al., 2007). Traditionally, continental European chocolate has been described as having a fineness of 15–22 µm particle diameter, while North American chocolate has a fineness of 20–30 µm (Afoakwa et al., 2007). Particle size above 35 µm often results a coarse mouthfeel, although the acceptance interval for this quality parameter changes for each region (Schumacher et al., 2009; Sokmen & Gunes, 2006). Normally, values approximately 20 µm are used as standards (Schumacher et al., 2009). Thus, we have used 20.0, 25.0 and 28.0 µm as mean particle sizes (D_[4,3]) to investigate influence of varying PSD on the physical and rheological properties of milk chocolate containing inulin. The effect of varying inulin levels on achieving the targeted PSD was also a parameter.

The largest particle size (D₉₀), smallest particle size (D₁₀), mean (D_[4,3]), median (D_[3,2]), span and SSA of each sample group are presented in Table 3. Also, representative PSD curves of milk chocolate samples are given in Figure 1. After refining, at all inulin levels, the targeted D_[4,3] could be reached with tolerable deviations. The deviations did not vary with varying inulin level. However, reduction in specific area and incremental changes in D₉₀, D₁₀, span and D_[3,2] with increasing D_[4,3] were identified. Beckett (1999) concluded that D₉₀ and SSA were two key parameters in manufacturing chocolate. SSA was inversely correlated with the different component of PSD in previous studies (Afoakwa et al., 2009; Beckett, 1999; Sokmen & Gunes, 2006; Ziegler & Hogg, 2001) and the present study confirmed the same results.

Table 3. PSD values of chocolate samples.

Table 3. VAlores PSD de muestras de chocolate.

Sample	SSA (cm^3/cm^2)	D90 (μm)	D10 (μm)	D[4,3] (μm)	Span	D[3,2] (μm)	SD
K1.1, K1.2, K1.3	5,268.5	39.80	5.29	20.06	2.1034	16.41	14.49
K2.1, K2.2, K2.3	4,879.0	54.23	5.37	25.37	2.5825	18.92	21.49
K3.1, K3.2, K3.3	4,415.9	61.5	5.5	28.11	2.8071	20.88	23.90
A1.1, A1.2, A1.3	5,313.1	42.33	5.12	20.91	2.2178	16.78	15.85
A2.1, A2.2, A2.3	5,188.7	55.03	5.16	25.01	3.0103	16.8	24.01
A3.1, A3.2, A3.3	4,707.6	62.01	5.52	28.03	2.8499	19.82	24.93
B1.1, B1.2, B1.3	5,679.2	44.33	5.12	20.72	2.2178	15.10	17.73
B2.1, B2.2, B2.3	4,978.3	55.54	5.65	25.18	2.5419	18.33	22.32
B3.1, B3.2, B3.3	4,746.6	62.25	5.83	28.2	2.8576	19.74	25.37
C1.1, C1.2, C1.3	5,747.1	44.49	4.81	19.94	2.5598	14.72	16.49
C2.1, C2.2, C2.3	4,768.4	55.71	5.38	25.13	2.9183	17.79	21.99
C3.1, C3.2, C3.3	4,678.0	62.61	5.45	27.91	2.9735	19.22	25.37

SSA; specific surface area, D₉₀; largest particle size, D₁₀; smallest particle size, D_[4,3]; mean, D_[3,2]; median, SD; standard deviation.

Figure 1. Representative PSD curves of milk chocolate samples.

Figure 1. Curvas PSD representante de muestras de chocolate con leche.

As the amount of inulin in the preparation of the samples increased, the SSA values also increased. This result is also valid for D₉₀ values. For samples containing amounts of inulin between 0% – 120 g/kg and processed to produce a mean particle size of 20.0, 25.0 and 28.0 µm, the D₉₀ values increased and showed statistical significance (P < 0.01), being in the ranges of 39.80–44.49, 54.23–55.71 to 61.50–62.61 cm³/cm², respectively. However, while SSA values also showed significant differences, these changes occurred in the opposite direction. It has been concluded that using inulin at various levels in milk chocolate preparation might affect the sensory, textural and rheological properties of the chocolate via changes in the PSD.

Smaller particle sizes in chocolates are known to improve sensory properties (Ziegler et al., 2001). An unpleasant texture has been found to occur when too many coarse particles are present in the chocolate (Bolenz et al., 2006). These particles (size above 30 µm) lead to a gritty or sandy product. Grittiness is a textural defect in chocolate, but it is one of many possible defects – difficulty swallowing, poor melting and hardness. In some cases, grittiness represents a minor issue (Do et al., 2007). Considering that using inulin in milk chocolate might cause sensory and textural deteriorations, as reflected in changes in the PSD, these other deteriorations could possibly increase as the amount of inulin is increased. However, considering other possible interactions, this potential for deterioration can be tolerated.

Another remarkable result of this study is that the intended D_[4,3] values under the same processing conditions have been achieved in both the controls samples and the samples containing inulin levels between 60 and 120 g/kg. In addition, the difference (P < 0.01) between each sample group was not determined; however, variation has been observed within narrow ranges, such as 19.94–20.91 µm, for 20.0 µm; 25.01–25.37 µm, for 25.0 µm; and 27.91–28.20 µm, for 28.0 µm. The relationship between this variation and variation in the amount of inulin is not clear. Optimisation of PSD in chocolate requires consideration of palate sensitivity (Afoakwa et al., 2007). This optimisation also has a significant impact on the rheological and instrumental textural properties of model chocolates. Improvements in rheological and textural characteristics via PSD optimisation are based on the reduction of interparticle contact and particle aggregation strength (Do et al., 2007).

It has been considered that the PSD can be used to observe changes in the rheological properties of chocolate because PSD is a key determinant of the flow (rheological) properties of chocolate. In previous studies, Beckett (2008) stated PSD influences chocolate rheology, with SSA and mean particle size (D_[4,3]) influencing yield stress.) A chocolate with particle sized according to the infinite model distribution may give lowest plastic viscosity (Afoakwa et al., 2007; Bouzas & Brown, 1995). In this study, although there may be insignificant differences in the D_[4,3] values of the sample groups, significant differences have been observed in the yield stress values (Table 4). The reason for this has been thought to result from the availability and quantity of inulin on the PSD, as well as the impact of these values on the product matrix. Yield value usually correlates with measured SSA. This, however, is not the case with Casson plastic viscosity, where at large particle sizes, the plastic viscosity can increase due to increasing amounts of bound molecules arranging themselves together (Belscak-Cvitanovic et al., 2012; Beckett, 2008). However, in the findings obtained from this study, despite the increase in the values of SSA and D_[4,3], a decrease in the values of yield stress and viscosity has been observed in the control samples, from both the sample groups that do contain inulin and those that do not contain inulin.

Table 4. Rheological measurements of chocolate samples.

Table 4. Color, actividad de agua y dureza de muestras de chocolate.

Sample	Inulin content (%, w/w)	Mean particule size (µm)	Conching time (h)	Yield stress (Pa)	Viscosity (Pas)	Standard error
K1.1	0.00	20.0549	3.50	18.16^a, a, a	3.96^a, a, a	2.86
K1.2	0.00	20.0549	4.00	17.22^a, a, b	3.46^a, a, b	3.67
K1.3	0.00	20.0549	4.50	15.45^a, a, c	3.07^a, a, c	5.72
K2.1	0.00	25.3709	3.50	16.45^a, b, a	3.7^a, b, a	3.91
K2.2	0.00	25.3709	4.00	15.37^a, b, b	3.29^a, b, b	5.79
K2.3	0.00	25.3709	4.50	14.38^a, b, c	3.06^a, a, c	4.01
K3.1	0.00	28.1107	3.50	13.57^a, c, a	3.3^a, c, a	4.63
K3.2	0.00	28.1107	4.00	12.59^a, c, b	3.03^a, c, c	2.71
K3.3	0.00	28.1107	4.50	11.09^a, c, c	3.11^a, a, b	4.32
A1.1	6.00	20.9062	3.50	13.67^b, a, a	3.26^b, a, b	3.21
A1.2	6.00	20.9062	4.00	13.30^b, a, a	3.32^a, a, a	3.14
A1.3	6.00	20.9062	4.50	12.71^b, a, b	3.16^a, a, c	4.12
A2.1	6.00	25.0065	3.50	10.46^b, b, a	2.86^b, b, a	2.18
A2.2	6.00	25.0065	4.00	9.399^b, b, b	2.81^b, b, a	3.02
A2.3	6.00	25.0065	4.50	8.704^b, b, c	2.64^b, b, b	2.14
A3.1	6.00	28.0340	3.50	9.842^b, c, a	3.27^a, a, a	3.18
A3.2	6.00	28.0340	4.00	8.892^b, c, b	2.39^b, c, c	11.03
A3.3	6.00	28.0340	4.50	8.518^b, b, b	2.51^b, c, b	4.12
B1.1	9.00	20.7175	3.50	9.370^c, a, a	2.81^c, b, a	2.45
B1.2	9.00	20.7175	4.00	8.724^c, a, b	2.65^b, a, b	3.91
B1.3	9.00	20.7175	4.50	6.657^c, b, c	2.51^b, a, c	5.79
B2.1	9.00	25.1805	3.50	8.352^c, b, b	3.03^a, a, a	7.79
B2.2	9.00	25.1805	4.00	8.792^b, a, a	2.47^c, b, b	6.02
B2.3	9.00	25.1805	4.50	8.704^b, a, a	2.42^c, b, b	8.32
B3.1	9.00	28.1999	3.50	6.798^c, c, a	2.35^b, c, a	2.37
B3.2	9.00	28.1999	4.00	5.012^c, b, b	2.25^b, c, b	3.92
B3.3	9.00	28.1999	4.50	5.179^c, c, b	2.16^c, c, c	2.8
C1.1	12.0	19.9372	3.50	8.703^d, a, a	2.46^d, a, a	5.53
C1.2	12.0	19.9372	4.00	7.158^d, a, a	2.35^b, a, b	2.82
C1.3	12.0	19.9372	4.50	5.245^c, a, c	2.28^c, a, c	2.12
C2.1	12.0	25.1289	3.50	6.096^d, b, a	2.32^c, b, a	2.94
C2.2	12.0	25.1289	4.00	5.824^c, b, b	2.3^d, a, b	11.78
C2.3	12.0	25.1289	4.50	5.321^c, a, c	2.28^d, a, b	4.74
C3.1	12.0	27.9069	3.50	4.179^d, c, a	1.82^c, c, a	3.26
C3.2	12.0	27.9069	4.00	3.406^d, c, b	1.79^c, b, b	2.26
C3.3	12.0	27.9069	4.50	3.354^d, b, c	1.72^d, b, c	2.51

For each parameter, means followed by the same letter are not significantly different (P > 0.01). First letters are used for varying PSL at the same PSD and CT, second letters are used for varying PSD at the same PSL and CT, third letters are used for varying CT at the same PSD and PSL.

Colour

One of the most important parameters influencing sensory perception of chocolate is colour (Viaene & Januszewska, 1999). Visual information characteristics of objects, including colour, brightness and translucency, are summarised into appearance attributes. Relevant information on colour can be acquired from modern technologies, such as computer vision and calibrated colour imaging analysis via such models as HunterLab and CIELab (Afoakwa, Paterson, Fowler, & Vieira, 2008a; Briones & Aguilera, 2005; Hatcher, Symons, & Manivannan, 2004; Jahns, Nielsen, & Paul, 2001; Lawless & Heymann, 1998). In this study, we used the CIELab model to determine brightness (L*) and chroma (C*) for each chocolate sample.

In previous studies, differences in L* (38.25–43.49) and C* (11.04–14.36) were found to depend on the PSD, packaging conditions, and storage (Afoakwa, Paterson, & Fowler, 2008; Aguilera, Michel, & Mayor, 2004; Mexis, Badeka, Riganakos, & Kontominas, 2010). Colour changes are likely to be specific to each chocolate sample (Aguilera et al., 2004). In our study, it was determined that variations in the L* values depend on PSD (mean values of 37.29, 37.76 and 36.78 for targeted D_[4,3] values of 20.0, 25.0 and 28.0 µm, respectively), CT (mean values of 37.09, 36.92 and 36.81 for 3.50, 4.0 and 4.50 h CT, respectively) and PSL (mean values of 36.16, 37.36, 37.19 and 37.06 for 0, 60, 90 and 120 g/kg of PSL, respectively). The results were found to be in accordance with other studies (Afoakwa, Paterson, & Fowler, 2008; Mexis et al., 2010). In addition to the L* value, it was also determined that the C* value varies depending on PSD (mean values of 16.28, 16.13 and 16.27 for targeted D_[4,3] values of 20.0, 25.0, 28.0 µm, respectively), CT (mean values of 16.14, 16.25 and 16.29 for 3.50, 4.0 and 4.50 h CT, respectively) and PSL (mean values of 15.37, 16.35, 16.55 and 16.64 for 0, 60, 90 and 120 g/kg of PSL, respectively).

Bolenz et al. (2006) reported a chocolate sample containing sample to be darker and different in colour than standard. Also, in another study, CIELAB parameters, as well as L* and C* values, increased as the inulin content of the chocolate decreased (Shourideh, Taslimi, Azizi, & Mohammmadifar, 2012). They used inulin as a sugar replacer at various levels in their study. However, in this study, samples containing 60 g/kg inulin were identified to be brighter than control samples, although these samples contained smaller quantities of cocoa particles for balancing the dry powder content during sample preparation. In the present study, sample groups of B and C contained identical quantities of cocoa mass, and brightness was not found to vary as a function of inulin content. Both of these findings were valid for C* values as well.

In the control sample group (n = 9) containing 0 g/kg inulin, the change in brightness has been statistically confirmed (P < 0.01) and the samples whose D_[4,3] value is intended to be 20.0 µm have a higher degree of brightness than those samples for which this value is 25.0 and 28.0 µm. However, similar data have not yet been obtained to confirm this relationship in samples containing inulin. Different CTs for samples that have the same levels of PSD and inulin produced no significant change statistically in the values of both brightness and chroma.

As was the case for the colour parameters, as a consequence of the variance analysis conducted for C* and L* (Table 5), the interaction of PSD × PSL × CT and the differences between the averages of the PSD, PSL and CT factors were found to be statistically significant (P < 0.01). Giving special consideration to the data from the control samples, there might be a decrease in the brightness along with an increase in the particle size of milk chocolate. It has also been hypothesised that CT might not have a determining effect on the colour properties; however, the connection between the change in the level of inulin and the colour properties has not been established.

Table 5. Colour, water activity and hardness of chocolate samples.

Table 5. Color, actividad de agua y dureza de muestras de chocolate.

Sample	Brightness (L*)	Chroma (C*)	Water activity (aw)	Hardness (N)
K1.1	37.61^a, a, a ± 0.887	15.89^a, a, a ± 0.148	0.34^a, a, a ± 0.027	11.82^a, a, a ± 0.312
K1.2	37.38^b, a, a ± 0.564	16.636^b, a, a ± 0.110	0.34^a, a, b ± 0.004	12.32^b, a, a ± 0.227
K1.3	37.17^bc, a, a ± 0.906	16.11^bc, a, a ± 0.263	0.337^a, a, b ± 0.041	11.76^bcd, b, a ± 0.474
K2.1	35.88^b, b, a ± 1.015	14.806^b, b, a ± 0.142	0.34 ^a, a, a ± 0.031	12.51^ab, a, a ± 0.186
K2.2	35.65^b, b, a ± 1.273	14.53^b, b, a ± 0.229	0.32 ^a, b, b ± 0.015	11.42^bc, b, a ± 0.495
K2.3	35.62^b, b, a ± 0.526	15.49^b, b, a ± 0.120	0.30 ^a, b, b ± 0.002	12.74^b, a, a ± 0.627
K3.1	35.65^c, b, a ± 1.537	14.23^c, b, a ± 0.141	0.34 ^a, b, a ± 0.015	10.37^b, b, a ± 0.058
K3.2	35.52^b, b, a ± 0.481	15.17^b, b, a ± 0.327	0.34^a, b, b ± 0.023	10.90^bc, b, a ± 0.364
K3.3	34.94^c, b, a ± 1.410	15.53^c, b, a ± 0.200	0.34 ^a, b, b ± 0.021	10.86^c, c, a ± 0.636
A1.1	37.42^a, a, a ± 0.855	15.64^a, a, a ± 0.159	0.33 ^b, a, a ± 0.007	11.00^bc, c, b ± 0.642
A1.2	37.93^a, a, a ± 0.599	16.60^a, a, a ± 0.231	0.31 ^b, a, b ± 0.013	11.61^bc, b, ab ± 0.555
A1.3	37.75^a, a, a ± 0.489	17.05^a, a, a ± 0.185	0.30 ^b, a, b ± 0.003	12.01^bc, a, b ± 0.324
A2.1	37.58^a, a, a ± 0.577	16.83^a, a, a ± 0.263	0.32 ^a, a, a ± 0.014	12.78^a, a, a ± 0.136
A2.2	37.37^a, a, a ± 0.425	16.32^a, a, a ± 0.128	0.33 ^a, a, b ± 0.017	12.50^a, a, a ± 0.282
A2.3	36.54^a, b, b ± 1.284	16.00^a, b, b ± 0.270	0.32 ^a, a, b ± 0.013	12.19^bc, a, a ± 0.430
A3.1	37.25^b, a, a ± 0.628	16.19^b, a, a ± 0.226	0.33 ^ab, a, a ± 0.032	11.96^a, b, a ± 0.221
A3.2	37.20^a, a, a ± 0.575	16.19^a, a, a ± 0.168	0.32 ^ab, a, b ± 0.016	12.08^a, ab, a ± 0.065
A3.3	37.23^ab, ab, a ± 0.541	16.29^ab, ab,a ± 0.146	0.329 ^ab, a, b ± 0.024	12.67^a, a, a ± 0.101
B1.1	37.49^a, a, a ± 0.440	16.45^a, a, a ± 0.431	0.347 ^a, a, a ± 0.033	10.40^b, b, b ± 0.362
B1.2	37.52^a, a, a ± 0.985	16.31^a, a, a ± 0.509	0.34 ^a, a, b ± 0.021	13.18^a, a, a ± 0.204
B1.3	37.20^b, a, a ± 0.123	15.72^b, a, a ± 0.622	0.34 ^a, a, b ± 0.014	12.85^a, b, a ± 0.206
B2.1	37.26^a, a, a ± 0.793	17.10^a, a, a ± 0.319	0.32 ^ab, b, a ± 0.011	12.0996^bc, a, b ± 0.096
B2.2	37.28^a, a, a ± 0.738	17.11^a, a a ± 0.449	0.31 ^ab, b, b ± 0.006	11.02^c, c, c ± 0.029
B2.3	36.79^a, a, a ± 0.395	15.887^a, a, a ± 0.546	0.31 ^ab, b, b ± 0.009	13.96^a, a, a ± 0.535
B3.1	36.95^b, a, a ± 0.511	17.05^b, a, a ± 0.871	0.33 ^b, b, a ± 0.017	11.92^a, a, a ± 0.701
B3.2	37.18^a, a, a ± 0.290	16.47^a, a, a ± 0.415	0.32 ^b, b, b ± 0.037	12.10^a, b, a ± 0.048
B3.3	37.01^b, a, a ± 0.375	16.89^b, a, a ± 0.517	0.32 ^b, b, b ± 0.017	11.43^bc, c, a ± 0.573
C1.1	36.60^b, b, a ± 0.300	16.15^b, b, a ± 0.343	0.32 ^b, a, a ± 0.012	11.01^bc, ab, b ± 0.439
C1.2	36.63^c, ab, a ± 0.331	16.21^c, ab, a ± 0.328	0.32 ^b, a, b ± 0.019	10.73^d, c, b ± 0.401
C1.3	36.78^bc, b, a ± 1.278	16.66^bc, b, a ± 0.238	0.328 ^b, a, b ± 0.027	12.35^ab, a, a ± 0.564
C2.1	37.10^a, b, a ± 0.087	16.17^a, b, a ± 0.301	0.31 ^bc, b, a ± 0.007	11.66^cd, a, a ± 0.776
C2.2	37.18^a, a, a ± 0.547	16.83^a, a, a ± 0.240	0.29 ^bc, b, b ± 0.008	11.83^ab, b, a ± 0.688
C2.3	36.85^a, b, a ± 0.177	16.50^a, b, a ± 0.166	0.30 ^bc, b, b ± 0.010	11.49^c, b, a ± 0.210
C3.1	38.34^a, a, a ± 0.398	17.21^a, a, a ± 0.194	0.33 ^b, b, a ± 0.011	10.74^b, b, b ±0.266
C3.2	36.24^b, b, b ± 0.219	16.68^b, b, b ± 0.146	0.31 ^b, b, b ± 0.068	12.66^a, a, a ± 0.656
C3.3	37.84^c, a, a ± 0.142	17.36^c, a, a ± 0.222	0.31 ^b, b, b ± 0.016	11.08^c, b, b ± 0.537

Five replicates of the hardness and colour (L* and C*) experiments were performed. Water activity experiments were performed in triplicate. All data reported as the mean ± SD. For each parameter, means followed by the same letter are not significantly different (P > 0.01). First letters are used for varying PSL at the same PSD and CT, second letters are used for varying PSD at the same PSL and CT, third letters are used for varying CT at the same PSD and PSL.

Hardness

Chocolate is a complex food product with sensory properties that are controlled mainly by the crystal structure and polymorphism of cocoa butter (Le Reverend, Smart, Fryer, & Bakalis, 2011). These properties of the microstructure determine macro-scale properties, such as hardness and melting profile in the mouth (Braipson-Danthine & Deroanne, 2004; Le Reverend et al., 2011). Hardness plays an important role in the sensory assessment of chocolate (Viaene & Januszewska, 1999).

Farzanmehr and Abbasi (2009) studied milk chocolates containing inulin. These researchers obtained different hardness values (10.1–15.0 N), depending on the composition of the samples. The findings obtained in this study have shown compatibility with those values mentioned above, and when all the samples are taken into consideration, the values of hardness vary between approximately 10.4 and 13.960 N.

The hardness of chocolates was correlated with the type of fat and its content, the type of sugar, PSD, tempering conditions, conching temperature (Afoakwa, Paterson, Fowler, & Vieira, 2008a, 2008b; De Clerq et al., 2012; Jovanovic & Pajin, 2004; Kieran Keogh, Murray, & O’Kennedy, 2003; Konar, 2013; Shourideh et al., 2012) and also inulin content (Farzanmehr & Abbasi, 2009). Hardness showed an inverse relationship with particle size, attributed mainly to the relative strengths of the particle-to-particle interactions within the different particulate structures of products exhibiting different PSDs (Do et al., 2007; Campos, Narine, & Marangoni, 2002; Afoakwa et al., 2009). In our study, the interaction of PSD × PSL × CT were found to be statistically significant (P < 0.01). Moreover, a decrease has been observed with the increase of the values of PSD and hardness in the samples. This situation most likely results from the increase in the contact level between particles and from the increase in PSD parameters.

As a result of the statistical evaluation of the samples prepared using the same parameters, it has been confirmed that CT has no significant effect on hardness (P < 0.01). However, when the milk chocolates in the B and C sample groups, which are prepared using the same levels of cocoa butter and cocoa mass, as well as the same CT and PSD parameters, are compared, it is observed that this situation does not hold for PSL. A statistically significant decrease in the values of hardness can be observed when the PSL parameter is increased (P < 0.01). Inulin is a food component that has a wide area of usage in many food products as both a prebiotic fibre and a fat replacer. Do et al. (2007) stated that fat content had a significant effect on hardness and that fat content had a higher impact than the PSD. In this related study, multiple comparison tests indicated that the hardness of the samples varied significantly for different fat contents and that increasing fat content decreased the hardness.

Water activity

Several factors, such as the raw materials used, the surface area of the materials, and the temperature and the humidity of refining and conching, can influence the water activity in chocolate (Biquet & Labuza, 1988; Rossini, Norena, & Brandella, 2011; Vercet, 2003). Due to its composition, white chocolate is a product with low water activity (Vercet, 2003); Rossini et al. state a water activity value of 0.40 aw for it (Rossini et al., 2011). In this study, water activity of all samples was determined to be below 0.34 aw (Table 5). This difference might be the result of the implementation differences in sample preparation procedures, as well as the distinctiveness of the properties of the inulin included in the samples, as discussed below. Additionally, for milk chocolates produced with inulin (10.45 g/100 g), Farzanmehr and Abbasi (2009) determined a water activity of 0.34.

As a result of the variance analysis carried out in relation to the findings, the interaction of PSD × PSL and the difference between the level averages in CT have been found to be statistically significant in terms of water activity (P < 0.01). In all the sample groups, the water activity values of the samples that are obtained with a 3.50 h CT are higher than those that are produced under the same conditions and formulation but with a 4.0 and 4.50 h CT. However, no significant difference has been observed between 4.0 and 4.50 h CTs in terms of water activity (P < 0.01).

A similar situation has also been observed when PSD is regarded as a variance. In all samples, other than the sample group with a PSL value of 60 g/kg, a decrease has been observed in the water activity values when D_[4,3] value is changed from 20.0 µm to 25.0 µm, whereas no significant difference has been observed between the samples with D_[4,3] value of 25.0 µm and those with 28.0 µm (P < 0.01). As for the sample group with PSL value of 60 g/kg, no difference has been observed with regard to PSD.

When samples from the B and C groups are examined, an inverse relationship is observed between the amount of inulin and the value of water activity. The increase of PSL in the sample groups has brought about a decrease in the value of water activity. As a result of its chemical structure, inulin is a hydrophilic component. Thus, inulin can be hypothesised to bind water molecules and also cause a decrease in the water activity of the samples.

Rheology

Rheological measurements and rheological properties of chocolate are notably important in chocolate processing for obtaining high quality products with well-established texture (Beckett, 2008; Fernandes et al., 2012). Chocolate mass exhibits non-Newtonian rheology, defined by plastic flow and characterised by yield stress. In addition, chocolate mass shows thixotropic and rheopectic properties (Mezger, 2002; Pajin et al., 2013; Pieper, 1986; Solstad, 1983). The flow properties of chocolate are affected by processing (refining, conching and tempering), as well as composition (amount of fat, amount and type of emulsifiers, PSD) (Afoakwa et al., 2009; Schantz & Rohm, 2005; Vavreck 2004). As in this study, several investigations have focused on the viscosity of milk chocolates (Ziegleder, Amanitidis, & Hornik 2004).

With respect to techniques for characterising the rheological properties, the International Confectionary Association (ICA, previously IOCCC) suggests the use of the Casson equation (Goncalves & Lannes, 2010). The Casson method uses the shear rate and shear stress and calculates the Casson viscosity and Casson yield value. In previous studies, an acceptable correlation was found between the Casson viscosity, sensory perception and consumer preference (Bolenz et al., 2005). Therefore, we have examined the findings obtained in this study by using Casson modelling. When these findings are taken into consideration, it has been found that CT × PSD × PSL interaction has produced statistically significant changes in the parameters of yield stress and viscosity (P < 0.01). For all of the sample groups, the value of yield stress has been found to vary between 18.16 and 3.35 Pa, and viscosity has been found to vary between 3.96 and 1.72 Pas.

In both the control samples of milk chocolate and the samples that contain inulin, the yield stress and viscosity have shown changes in the opposite direction of changes in the PSD (P < 0.01). This result shows compatibility with the findings previously obtained from other chocolate samples, and it can be stated that the Casson yield stress and viscosity values for milk chocolate show changes in the opposite direction of changes in the PSD. Afoakwa, Paterson, and Fowler (2008) stated that an increase in average particle size results in a decrease of the Casson plastic viscosity, shear stress, yield stress and apparent viscosity. The milling process affects properties of chocolate mass, such as rheology, texture and sensory properties (Beckett, 2008; Pajin et al., 2013). PSD change and the differentiation on the humidity level of chocolate samples may be considered to be one of the reasons for this situation. In addition, it has been accepted that as a result of the change in the refining conditions, the interaction among particles has been affected and has produced a change in the rheological parameters. Servais, Ranc, and Roberts (2004) stated that any disturbance of the chocolate mass may affect the viscosity, as well as the amount of moisture. Also, the smaller the particle size, the higher the plastic viscosity because smaller particle size increases the contact surface with the cocoa butter (Schumacher et al., 2009; Sokmen & Gunes, 2006).

As is set for PSD, changing inulin level has been shown to produce changes in the yield stress and viscosity values in the opposite direction (P < 0.01). Also, a decrease in the yield value for an inulin-containing chocolate was observed by Bolenz et al. (2006). Therefore, the finding obtained from our study – that inulin has an effect on the chocolate rheology – exhibits compatibility with the other studies; additionally, the strength of this interaction is shown to depend on the inulin concentration (P < 0.01). Ziegleder et al. (2004) stated that, chocolate viscosity is increased by an increase in the moisture level. When the effects of inulin availability and inulin levels on the water activity value are considered in their study, one of the mechanisms found to be the source determined effect, as inulin water-binding property. However, results of the study conducted by Shourideh et al. (2012) indicated that when a high level of inulin was used, it produced an incremental effect in the plastic and apparent viscosity. A low level of inulin improves the flow properties but reduces the apparent viscosity. In the previous study, inulin was used as sugar substitute. Thus, these findings are not appropriate to compare with our results.

There have been no previous studies analysing the effect of CT on the rheological properties of inulin-containing milk chocolates. However, in a previous study, yield stress values decreased in line with increasing conching temperature for milk chocolate samples containing inulin (Konar, 2013). Also in this study, both yield stress and viscosity decreased significantly with increasing CT (P < 0.01). This situation is valid for samples having identical PSD and PSL features that are prepared using different CTs.

Conclusions

Although the use of inulin in milk chocolates has advantages for the functional properties of the product, the effects on the physical and rheological properties must be taken into consideration. It has been hypothesised that these effects might result from the fact that inulin is a hydrophilic component and has a polymer structure, as well as a fat-replacement quality. Taking into consideration the availability of inulin and the differences in the hardness, water activity and colour properties as a result of the change in the level of inulin, the effect of added inulin on these properties can be tolerated and even considered as having a positive effect because good chocolate should have a smooth, soft, velvety, texture while poor-quality chocolate feels hard, grainy or waxy (Aidoo, Sakyi-Dawson, Abbey, TanoDebrah, & Saalia, 2011). However, compared with the standard samples, significant deviations have been identified in the rheological properties. In light of these findings, it has been hypothesised that this effect cannot be inhibited through optimisation of CT. However, it is possible to reduce the magnitude of the change induced by a decrease in the value of mean particle size.