LWT - Food Science and Technology 116 (2019) 108456

Contents lists available at ScienceDirect

LWT - Food Science and Technology

journal homepage: www.elsevier.com/locate/lwt

Effect of apple fibre addition and temperature on the rheological properties T

of apple juice and compensation study
D.E. Salinas, A. Garvin∗, R. Ibarz, A. Ibarz
Food Engineering Unit, Food Technology Department, University of Lleida, Av. Alcalde Rovira Roure, 191 25198, Lleida, Catalonia, Spain

A R T I C LE I N FO A B S T R A C T

Keywords: The addition of dietary fibre has become a trend in the food sector and, as it changes some physical properties
Rheology like flow parameters, knowing these changes is essential for food process design. In this work, the influence of
Thixotropy both the temperature (5-60 °C) and the fibre content (0–16%, w/w) on the flow behaviour of apple juice were
Viscosity studied. The apple juice presented thixotropic behaviour above a fibre content of 10% for all the range of
Fibre
temperatures tested. The thixotropy followed the kinetic model of Figoni and Shoemaker. Flow behaviour was
Fruit juice
found to vary depending on the fibre content: the Newtonian model for 0–8%, the power law model (with
pseudoplastic behaviour) for 10–12% and the Herschel-Bulkley model for 14–16%. The temperature effect was
found to agree with the Arrhenius - equation. The iso-consistency temperatures were 79.6 and −135.9 °C for the
Newtonian and non-Newtonian models, respectively. The compensation for the Newtonian juices was found to
be no more significant than the propagation of the experimental errors. Nonetheless, the compensation for the
non-Newtonian juices was found to be real, the control being entropic for all the working temperatures.

1. Introduction In the design of flow processes, including pumping, mixing and heat
transfer (Ibarz & Barbosa-Cánovas, 2003) and in process and storage
Fruit dietary fibre is the dehydrated residue obtained from the po- control, like quality and stability control (Falguera, Vélez-Ruiz, Alins, &
mace obtained after squeezing the fruit to extract juice. It is known to Ibarz, 2010), it is important to know the rheological properties of fluid
be an outstanding ingredient for enriched foods, because it provokes an and semi-solid foods.
increase on the nutritional properties of the original food, such as an Some fruit fluids show thixotropy, which is defined by a continuous
increase in the insoluble/soluble fibre ratio and a reduction in its en- reduction in the viscosity with the shear time and a recovery once the
ergetic value. It also causes some functional changes, like the formation shear rate has stopped (Lozano & Ibarz, 1994). The thixotropic response
of gel, an increase in water and oil binding, a decrease in syneresis, an of a fluid can be determined by measuring the shear stress as a function
increase in viscosity, modification of texture and formation and stabi- of time at a given steady shear rate. Figoni-Shoemaker (1981) modelled
lization of emulsions and foams (Augusto, Falguera, Cristianini, & Ibarz, thixotropy using the following equation:
2011; Ibarz & Barbosa-Cánovas, 2003). Thus, adding fruit fibre to foods
σ = σe + (σi − σe ) exp (−αt ) (1)
involves three benefits, an increase in the healthiness of the foods, a
change in their functionality and the use of an industrial by-product where σ is the shear stress (Pa), σe is the steady shear stress at the
that would otherwise need resources to be disposed of (Grigelmo- equilibrium reached after sufficiently long time of shearing (Pa); σi is
Miguel, Ibarz, & Belloso, 1999b; Sabanis, Lebesia, & Tzia, 2009; the initial shear stress (Pa); α is the kinetic constant of structure de-
Sanchez Zapata et al., 2010; Sendra et al., 2010; Staffolo, Bertola, gradation (s-l) and t is time (s). Thus, the difference (σi – σe) represents
Martino, & Bevilacqua, 2004). the breakdown structure.
The apple is one of the most popular fruit consumed worldwide There are quite a few rheological models that describe the flow
(Kaushal & Joshi, 1995). Sixty three% is sold as fresh apples, and the behaviour (Canet, Alvarez, Fernández, & Luna, 2005; Kokini, 1992;
remaining 37% is processed: 18% as juices or cider, 13% as canned Rao, Cooley, & Vitali, 1984; Rha, 1978), but the most usual for food
fruit, 3% as dehydrated fruit, 2% as frozen fruit and 1% as other mis- fluids are the following:
cellaneous products such as vinegar, wine and jellies (Root, 1996). Newton: σ = ηγ˙ (2)

∗
Corresponding author.
E-mail addresses: [email protected] (D.E. Salinas), [email protected] (A. Garvin), [email protected] (R. Ibarz), [email protected] (A. Ibarz).

https://doi.org/10.1016/j.lwt.2019.108456
Received 11 April 2019; Received in revised form 27 July 2019; Accepted 29 July 2019
Available online 30 July 2019
0023-6438/ © 2019 Elsevier Ltd. All rights reserved.
D.E. Salinas, et al. LWT - Food Science and Technology 116 (2019) 108456

Power law: σ = kγ˙ n (3) obtained from the slope b:

Herschel − Bulkley: σ = σ0 + kγ˙ n (4) 1

Tiso =
bR (7)
where σ is the shear stress (Pa), ẏ the shear rate (s−1), η the Newtonian
viscosity (Pa·s), k the consistency index (Pa·sn), n the flow behaviour lnk 0 = ln (kTiso) −
1
Ea
index (dimensionless) and σ0 the yield stress (Pa). RTiso (8)
The Herschel-Bulkley model mathematically includes the other two
Only a few studies have been published regarding the rheological
simpler equations. When the yield stress (σ0) is equal to zero, the model
behaviour of juices containing fibre: orange and peach fibre dispersion
is converted to the Newton one when the flow behaviour index (n) is
in water (Grigelmo-Miguel et al., 1999a and 1999b), orange fibre ad-
equal to unity and otherwise it is the power law one. Thus, the
dition to yogurt (Sendra et al., 2010), peach fibre addition to peach
Herschel-Bulkley equation can be used to evaluate the three models,
juice (Augusto et al., 2011) and orange fibre addition to orange juice
defining the consistency as the resistance to flow, being either the
(Siche et al., 2012, 2015). In some cases, The studies were carried on at
viscosity for the Newtonian model, or the consistency index for both the
only one specific fibre content and/or at only one specific temperature
power-law and Herschel-Bulkley models.
value. Besides, in none prior study the compensation was studied.
The food fluids that follow the power-law model usually show flow
Guillon and Champ (2000) highlighted the need for more research to
behaviour indexes (n) lower than 1, behaving as shear thinning
understand the change in a material by the addition of fibre. Moreover,
(pseudoplastic). They can also follow the Herschel-Bulkley model with
the rheological behaviour of any fluid needs to be known prior to de-
n values lower than 1 (Bingham pseudoplastic), yield stress (σ0) being
signing any industrial equipment. Thus, the two main aims of this work
the minimum value of shear stress needed to start the flow, being re-
were to explore the changes in the rheological behaviour of apple juice
lated to the internal structure that resists the flow.
by adding apple fibre for both wide ranges: temperature (0–60 °C) and
The Arrhenius equation is usually used to relate the dependence of
fibre content (0–16 g/100 g); and to study whether the phenomenon
the viscosity (η) or the consistency index (k) to the temperature.
called compensation took place.
E
Arrhenius equation: k = k 0 exp ⎛ a ⎞
⎝ RT ⎠ (5) 2. Material and methods
where k is the consistency, either the viscosity (Pa·s) or the consistency
2.1. Samples
index (Pa·sn), k0 is the predicted parameter k at extremely high tem-
peratures (usually called frequency factor in the case of kinetic con-
The apple juice contained 12.5 °Brix of total soluble solids, similar
stants) (Pa·s or Pa·sn), Ea is the activation energy (J·mol−1), R is the
to the commercial ones, and was obtained by diluting with distilled
ideal gas constant (8.314 J mol−1·K−1) and T is the absolute tempera-
water a clarified and depectinised 66.5-°Brix concentrated juice. Both
ture (K). In this case, unlike for the kinetic constant, the sign inside the
the apple juice and the fibre were provided by the Indulleida company,
exponential is positive because both the viscosity or consistency index
which obtained them from a mixture of 95% Golden delicious and 5%
decrease when the temperature increases (Garvin, Ibarz, & Ibarz, 2017).
Granny Smith varieties. Indulleida is one of the most important com-
The concept called kinetic compensation for chemical reactions takes
panies involved in the industrial production of fruit juices in both
place when the logarithm of the frequency factor (lnk0) happens to
Catalonia and Spain. The fibre was added and then homogeneized using
depend linearly on the activation energy (Ea) for the different values of
an Ultra-Turrax T25 (IKA, Staufen, Germany) for 3 min at 15,000 rpm.
the environmental variable at which the Arrhenius equation was pre-
Both the rheological behaviour and thixotropy were studied at fibre
viously fitted (Garvin et al., 2017). It is not a scientific law, but this
contents of 0, 2, 4, 6, 8, 10, 12, 14 and 16 g/100 g.
behaviour has been reported in quite a few studies regarding chemical,
physical and biological processes ((Liu & Guo, 2001)) but also in food
processes (Garvin et al., 2017). When this phenomenon takes place, the 2.2. Analysis of the industrial apple fibre used
isokinetic temperature is a mathematical consequence, being the tem-
perature at which the reaction rates have the same value regardless the The fibre profile was obtained following the AOAC method
environmental variable. In these cases, the control is enthalpic for (Grigelmo-Miguel, Ibarz, & Belloso, 1999c; Prosky, Asp, Schweizer,
working temperatures lower than the isokinetic temperature, otherwise DeVries, & Furda, 1988). The method is based on the enzymatic re-
the control is entropic (Lefler, 1955; Garvin et al., 2017). Other con- moval of starch and protein and consequent separation into the soluble
sequences are that the kinetic constant depends significantly on the and insoluble fibre fractions by centrifugation.
environmental variable and that the reaction mechanism remains the Protein, ash, fat and moisture determination were carried out using
same within the studied range of the environmental variable. For the standard AOAC method numbers 978.04, 930.05, 920.85 and
physical changes that follow Arrhenius equation, the change in the 930.04, respectively (AOAC, 1997).
physical parameter can be thought of as a chemical reaction. Ewell
(1938) described viscosity as a movement of molecules that can be 2.3. Soluble solids analysis
activated to escape from a structure formed by their closest neigh-
bouring molecules prior to sliding over them. The total soluble solids were determined using an ATAGO RX-100
In the case of the flow, instead of kinetic compensation, the phe- refractometer (Atago Bussan Co., Tokyo, Japan) at a temperature of
nomenon can be called consistency compensation and the temperature 20 °C.
at which the studied flow parameter remains constant can be called the
iso-consistency temperature. In this case, the parameter studied can be 2.4. Rheological and thixotropy measurements
either the viscosity for Newtonian models or the consistency index for
non-Newtonian models. The straight-line equation that causes con- A rheometer Haake RS 600 (Haake, Karlsruhe, Germany) was used,
sistency compensation is the following (Ozilgen & Bayindirli, 1992; using a Couette-geometry sensor (concentric cylinder, Haake Z40-DIN)
Garvin et al., 2017): with cup and bob radius ratio of 1.0847 (bob radius = 20.000 ±
0.004 mm), except in the case of 16 g/100 g of fibre content, for which,
lnk 0 = a − bEa (6)
due to its high consistency, the grooved plate-plate geometry (60 mm
in which a is the value of the logarithm of either the viscosity or the diameter, Haake PP 60Ti) was used.
consistency index at the iso-consistency temperature, which can be Thixotropy was previously evaluated by shearing each sample at a

2
D.E. Salinas, et al. LWT - Food Science and Technology 116 (2019) 108456

constant shear rate of 100 s−1 for 8 min and recording the shear values Table 2
as a function of time. A new sample without mechanical history was Values of the thixotropic parameters of the Figoni-Shoemaker model as a
used after standing for 10 min still inside the rheometer, in order to function of the percentage of fibre and temperature at a shear rate of 100 s−1
achieve the desired temperature and to assure that the sample structure Fibre (g/ T (°C) α (s−1) σi (Pa) σe (Pa) R2
was the same as before handling it. 100 g)
Each flow rheological evaluation was carried out with the same
16 5 0.0090 ± 0.0002 45.45 ± 0.06 27.96 ± 0.03 0.9440
samples used for checking the thixotropy and after eliminating it by
16 15 0.0091 ± 0.0005 44.60 ± 0.02 28.00 ± 0.06 0.9733
shearing for 5 min at the maximum shear rate admitted by each sample. 16 30 0.0073 ± 0.0007 16.98 ± 0.07 9.35 ± 0.06 0.9520
For each measure of shear stress, the shear rate was varied from 0.01 to 16 45 0.0069 ± 0.0009 25.40 ± 0.24 18.34 ± 0.14 0.9820
300 s−1, with the exception of the samples with 0 and 2 g/100 g fibre 16 60 0.0137 ± 0.0003 10.15 ± 0.02 5.98 ± 0.03 0.9720
content, for which the maximum shear rate were 80 s−1 and 150 s−1, 14 5 0.0071 ± 0.0005 38.66 ± 0.05 18.62 ± 0.02 0.9725
14 15 0.0084 ± 0.0002 32.76 ± 0.05 16.71 ± 0.03 0.9689
respectively. Each sample was checked in triplicate, first increasing the 14 30 0.0089 ± 0.0003 23.95 ± 0.09 12.82 ± 0.02 0.9880
shear rate, then decreasing it and finally increasing it yet again. As 14 45 0.0122 ± 0.0005 26.93 ± 0.07 15.37 ± 0.02 0.9901
thixotropy was previously eliminated, the three ramps coincided and 14 60 0.0140 ± 0.0005 20.22 ± 0.02 9.31 ± 0.03 0.9968
only the data from the third ramp was used in order to evaluate the 12 5 0.0061 ± 0.0002 9.22 ± 2.15 5.78 ± 1.88 0.9322
12 15 0.0057 ± 0.0005 14.55 ± 0.01 6.56 ± 0.03 0.9900
corresponding model (Genovese & Rao, 2005; Sato & Cunha, 2009;
12 30 0.0079 ± 0.0005 11.52 ± 1.13 5.24 ± 0.09 0.9629
Silva et al., 2010; Tonon, Alexandre, Hubinger, & Cunha, 2009). 12 45 0.0140 ± 0.0009 12.09 ± 0.06 4.75 ± 0.03 0.9872
12 60 0.0113 ± 0.0009 10.85 ± 0.02 4.92 ± 0.01 0.9770
10 5 0.0064 ± 0.0002 4.13 ± 0.05 2.93 ± 0.03 0.9814
10 15 0.0055 ± 0.0003 3.80 ± 0.62 2.74 ± 0.06 0.9734
2.5. Temperature
10 30 0.0061 ± 0.0005 4.65 ± 0.82 2.57 ± 0.07 0.9620
10 45 0.0070 ± 0.0003 4.48 ± 0.07 2.32 ± 0.02 0.9203
The temperature was kept constant ( ± 0.3 °C), using a water bath 10 60 0.0068 ± 0.0002 2.88 ± 0.07 1.54 ± 0.02 0.9410
Phoenix Thermo Haake C25P (Haake, Karlsruhe, Germany). The tem-
peratures checked were 5, 15, 30, 45 and 60 °C for both the rheological p-values < 0.0001 for all cases
and thixotropy measurements.
3.2. Thixotropic characterization of apple juice

2.6. Statistical analysis The apple juices with fibre content up to 8 g/100 g did not show any
thixotropy. Table 2 shows the estimates of the Figoni-Shoemaker model
Each test was carried out with three replicates, using the average for each temperature and fibre content above 10 g/100 g. Since all the
values to fit the model equations. The parameters of each model were determination coefficients were higher than 0.92 and all the p-values
obtained by linear and non-linear regression with the STATGRAPHICS were clearly lower than 0.001, thixotropy was concluded to be properly
Centurion XVI (Statistical Graphics Corp., Warrenton, VA, USA) statis- represented by the Figoni-Shoemaker model.
tical programme using a significant probability level of 95%. From Table 2 it can be seen that the kinetic constant of the de-
The data was analysed using one-way analysis of variance. When a gradation of the internal structure (α) shows no clear tendency but
significant difference was found among the sample means, Tukey's test roughly increases with temperature and fibre content. Although letters
was used to determine which means were different. Data were reported were omitted in order to simplify the reading, Tukey test concluded that
as mean ± standard deviation (SD), and significance was set at both temperature and fibre content have a significant effect on the in-
P ≤ 0.05 using JMP Statistical Software Version 5 (JMP Statistical ternal structure, not only the initial value (σi) but also the final one (σe).
Discovery Software, Cary, NC). The lower the temperature and the higher the fibre content, the higher
the internal structure level for both the initial and the final one. Since
the initial internal structure (σi) is more sensitive to temperature and
3. Results and discussion the fibre content than the final one (σe), the breakdown structure (σi –
σe) also was significantly concluded to increase when the temperature
3.1. Apple fibre profile decreases and the fibre content increases.

The total fibre content was found to be 60.1 g/100 g, consisting of 3.3. Flow rheological behaviour
13.8 of soluble fibre and 46.3 of insoluble fibre. Table 1 shows the
content analysis of the industrial dietary fibre used. Fig. 1 shows the rheograms for all the apple juice-apple fibre mix-
tures at 30 °C. For the rest of the temperatures, the rheograms obtained
were similar, showing a significant variation for fibre contents higher
Table 1
than 14 g/100 g, especially for 16 g/100 g fibre content, while a slight
Composition of the dietary fibre used.
variation for fibre contents between 0 and 12 g/100 g. The dependence
Dietary fibre composition (% w/w) changes from straight lines (Newtonian model) for low fibre contents to
curves (non-Newtonian models) for high fibre contents.
Water 8.6
Sugars 27.2 Fig. 2 shows the rheograms at all the temperatures for the 10 g/
Fat 2.5 100 g fibre content enriched juice. For the rest of the fibre contents, the
Ash 1.6 rheograms obtained were similar, showing that the shape of the curves
Soluble proteins 4
do not change by changing the temperatures.
Soluble fibre 13.8
Insoluble fibre 46.3 Table 3 shows the Herschel-Bulkley estimates for all the tempera-
Insoluble polysaccharides 30.9 tures and fibre contents. Since this equation mathematically includes
Klason lignin 12.5 the other two simpler equations, these same estimates allow the model
Resistant proteins 1.1 followed by the fluid to be known. Applying the Tukey test for yield
Tannins 1.8
stress, the fibre content was concluded to be significant for fibre con-
Subtotal fibre 60.1
Total 100 tents higher than 12 g/100 g and the temperature was not found sig-
nificant unless the sample reaches the 16 g/100 g of fibre content. For

3
D.E. Salinas, et al. LWT - Food Science and Technology 116 (2019) 108456

Fig. 1. Effect of the fibre content on the flow curves for apple juice at 30 °C. Each flow curve consists of the mean of three replicates of shear stress (σ) in Pa versus
shear rate (γ̇ ) in s−1.

With the addition of apple fibre, not only did the consistency in-
crease, but also the flow behaviour changed. The higher the fibre
content, the more complex the model (Table 3). The Newton model was
followed up to a fibre content of 8 g/100 g. It then changed to the
power law model, that was followed up to a 12 g/100 g fibre content. At
a fibre content of 12 g/100 g, for temperatures higher than 15 °C the
fluid behaved following the power law model, but for a temperature of
5 °C, the fluid behaved according to the Herschel-Bulkley model. For
fibre contents of 14 and 16 g/100 g, the behaviour followed the Her-
schel-Bulkley model for all the temperatures. Since all the values of R2
were higher than 0.9 and all the p-values clearly lower than 0.0001, the
models were concluded to represent properly the fluid flow.
As expected, the increase in the fibre content resulted in a pro-
gressive increase in the fluid consistency. The deviation of the
Fig. 2. Effect of the temperature on the flow curves for apple juice with fibre Newtonian flow also increased with the addition of fibre, not only re-
content of 10 g/100 g. Each flow curve consists of the mean of three replicates presented by the change of model but also by the decrease in the index
of shear stress (σ) in Pa versus shear rate (γ̇ ) in s−1 at the specific temperature of flow behaviour (n) and the increase in the yield stress (σ0) once it
(■5 °C, ▲15 °C, ●30 °C, ▼45 °C, ՜ 60 °C). appeared. All this was due to the fact that the fibre content interacts
with the particles of the juice to produce new bonds that create new
the consistency index, a significant effect was also observed for the internal structure hindering the flow. The level of internal structure
16 g/100 g fibre content sample while the effect of the temperature against the flow (σ0) increased clearly, from a value over 2 Pa to a value
increases as the fibre content also increases. For flow behaviour index, over 40 Pa when the fibre content was increased from 14 to 16 g/100 g,
the effect of the fibre content is significant for fibre contents higher than respectively. The initial structure that hinders the flow increased by
10 g/100 g, coinciding with the thixotropic behaviour. In order to more than 20 times after adding only 2 g/100 g of fibre content. Siche
clarify the reading of Table 3, the letters of the Tukey test were omitted. et al. (2012 and 2015) and Augusto et al. (2011) found similar results in
Finally, as expected, the clarified and depectinised, ready to drink the rheological behaviour of orange and peach juices, respectively.
apple juice with 12.5 Brix and 0 g/100 g fibre showed a Newtonian flow Data in Table 3 were fitted to Equations (9)–(11), excluding data for
behaviour, with the viscosity decreasing from 2.2 mPa s at 5 °C to zero fibre content because the behaviour was clearly different. Table 4
0.6 mPa s at 60 °C. shows the estimates of the parameters of Equation (9), showing how
both parameters decreased while the temperature increased. Table 5
3.3.1. Fibre content effect shows the estimates of the parameters of Equation (11), showing better
The dependence of the flow parameters on any kind of concentra- R2 and p values, but showing no tendency for the three parameters with
tion has been found to fit the following most frequently-used mathe- the temperature. Equation (10) was found not to fit the data properly.
matical expressions (Vitali & Rao, 1982; Rao et al., 1984; Ibarz, Vicente, Similar equations were found for other fruit juices, e.g. orange (Siche,
& Graell, 1987, 1989, 1992): 2012, Vitali & Rao, 1984a,b)), peach (Augusto et al., 2011), malus
floribunda (Cepeda, Villarán, & Ibarz, 1999), sloe (Ibarz, Garvín, &
Exponential function: k = a1 exp (b1 C ) (9) Costa, 1996) and kiwi (Ibarz, Giner, Pagan, Gimeno, & Garza, 1995).
When increasing the fibre content, the yield stress (σ0) showed a sig-
Potential function: k = a2 C b2 (10) nificant increase and the flow behaviour index (n) a significant de-
crease, similar to other cases, such as tomato products (Sharma,
Exponential − Polynomial function: k = a3 exp (b3 C + d3 C 2) (11) Lemaguer, Liptay, & Poysa, 1996; Yoo & Rao, 1994(Yoo and Rao,
where k is the consitency, C is the concentration and a1, b1, a2, b2, a3, b3 1996)).
and d3 are constants (units in Tables 4 and 5).
Fig. 1 shows that in order to achieve the same shear rate, the shear 3.3.2. The effect of the temperature and consistency compensation
stress needed increases clearly when the fibre content is raised, the Fig. 2 shows that in order to achieve the same shear rate, the shear
variation being especially high for fibre contents upper than 14 g/100 g. stress needed decreased clearly (and thus, the consistency) while the

4
D.E. Salinas, et al. LWT - Food Science and Technology 116 (2019) 108456

Table 3
Values for the parameters of the Herschel-Bulkley model (0.1 s−1 < γ̇ < 300 s−1, with an average of three replicates ± standard deviation). Herschel-Bulkley model
mathematically includes the power Law (σ0 = 0) and Newton models (σ0 = 0 and n = 1).
Fibre (g/100 g) T (°C) σ0 (Pa) k (Pa.sn) n R2 Rheological Behaviour

0 60 - 0.0008 ± 0.00022 1 0.9373 Newton

45 - 0.0012 ± 0.00018 1 0.9029 Newton
30 - 0.0014 ± 0.00007 1 0.9285 Newton
15 - 0.0015 ± 0.00016 1 0.9654 Newton
5 - 0.0022 ± 0.00023 1 0.9834 Newton
2 60 - 0.0021 ± 0.00005 1 0.9790 Newton
45 - 0.0024 ± 0.00006 1 0.9808 Newton
30 - 0.0024 ± 0.00016 1 0.9914 Newton
15 - 0.0034 ± 0.00006 1 0.9945 Newton
5 - 0.0036 ± 0.00024 1 0.9941 Newton
4 60 - 0.0033 ± 0.00018 1 0.9902 Newton
45 - 0.0033 ± 0.00004 1 0.9891 Newton
30 - 0.0039 ± 0.00068 1 0.9784 Newton
15 - 0.0042 ± 0.00055 1 0.9872 Newton
5 - 0.0045 ± 0.00035 1 0.9894 Newton
6 60 - 0.0035 ± 0.00006 1 0.9877 Newton
45 - 0.0036 ± 0.00025 1 0.9925 Newton
30 - 0.0042 ± 0.00021 1 0.9956 Newton
15 - 0.0054 ± 0.00043 1 0.9956 Newton
5 - 0.0087 ± 0.00046 1 0.9990 Newton
8 60 - 0.0031 ± 0.00104 1 0.9957 Newton
45 - 0.0058 ± 0.00014 1 0.9915 Newton
30 - 0.0065 ± 0.00114 1 0.9993 Newton
15 - 0.0091 ± 0.00373 1 0.9962 Newton
5 - 0.0122 ± 0.00512 1 0.9998 Newton
10 60 - 0.0162 ± 0.00887 1.002 ± 0.070 0.9991 Power law
45 - 0.0235 ± 0.01676 0.970 ± 0.073 0.9995 Power law
30 - 0.0288 ± 0.02465 0.955 ± 0.081 0.9997 Power law
15 - 0.0823 ± 0.01714 0.901 ± 0.064 0.9998 Power law
5 - 0.0913 ± 0.03167 0.899 ± 0.070 0.9998 Power law
12 60 - 0.0742 ± 0.01132 0.778 ± 0.241 0.9978 Power law
45 - 0.1010 ± 0.00000 0.750 ± 0.268 0.9995 Power law
30 - 0.2513 ± 0.01215 0.753 ± 0.205 0.9988 Power law
15 - 0.2721 ± 0.02853 0.714 ± 0.218 0.9996 Power law
5 1.594 ± 0.423 0.2821 ± 0.01118 0.719 ± 0.153 0.9995 Herschel-Bulkley
14 60 2.564 ± 0.122 0.5248 ± 0.02381 0.681 ± 0.033 0.9993 Herschel-Bulkley
45 2.443 ± 0.860 0.6496 ± 0.03519 0.646 ± 0.027 0.9996 Herschel-Bulkley
30 2.345 ± 0.792 0.8938 ± 0.02349 0.620 ± 0.017 0.9998 Herschel-Bulkley
15 2.149 ± 0.815 1.4094 ± 0.06956 0.603 ± 0.007 0.9987 Herschel-Bulkley
5 2.272 ± 0.190 1.6482 ± 0.07929 0.592 ± 0.018 0.9994 Herschel-Bulkley
16 60 46.060 ± 0.707 9.4981 ± 0.0700 0.464 ± 0.005 0.9995 Herschel-Bulkley
45 53.285 ± 1.414 13.776 ± 0.0707 0.424 ± 0.002 0.9994 Herschel-Bulkley
30 55.385 ± 4.642 16.680 ± 0.0591 0.412 ± 0.004 0.9996 Herschel-Bulkley
15 55.431 ± 3.944 20.885 ± 0.6852 0.399 ± 0.007 0.9992 Herschel-Bulkley
5 61.495 ± 6.215 28.043 ± 1.8360 0.377 ± 0.001 0.9999 Herschel-Bulkley

p-values < 0.0001 for all cases

Table 4 temperature increased. This fact can also be seen in Table 3.

Influence of the fibre content on the consistency (either viscosity or consistency In order to quantify the effect that the temperature exerts on the
index) at each temperature. Equation (9). k = a1 exp (b1 C ) consistency of each fluid, the consistency estimates for each model were
T (°C) a1 (mPa·sn) b1 (g/100 g)−1 R2 p-value fitted to the Arrhenius equation, obtaining a pair of estimates for lnk0
and Ea for each fibre content (Table 6). Since all the coefficients of
5 0.3028 0.6226 0.9125 < 0.0001 determination were always higher than 0.85 and all the p-values clearly
15 0.2570 0.6199 0.9070 < 0.0001
lower than 0.05, the Arrhenius equation (Equation (5)) was concluded
30 0.1948 0.6118 0.8866 0.0001
45 0.1897 0.5859 0.8649 0.0003
60 0.1743 0.5666 0.8386 0.0005 Table 6
Parameters of the Arrhenius equation for the dependence of consistency on
temperature.
Table 5
Influence of the fibre content on the consistency (either viscosity or consistency Fibre (g/100 g) Ea (kJ/mol) k0 (mPas.sn) R2 p-value
index) at each temperature. Equation (11). k = a3 exp (b3 C + d3 C 2)
0 12.17 0.0107 0.9160 0.0106
T (°C) a3 (mPa·sn) b3 (g/100 g)−1 d3 (g/100 g)−2 R2 p-value 2 7.87 0.1191 0.9330 0.0075
4 4.74 0.5794 0.9466 0.0053
5 4.8959 −0.2123 0.0464 0.9936 < 0.0001 6 12.12 0.03874 0.8612 0.0229
15 1.7200 2.8284 −0.1303 0.9508 < 0.0001 8 17.42 0.00657 0.9455 0.0055
30 2.8298 3.0148 −0.1555 0.9456 < 0.0001 10 26.04 0.00123 0.9413 0.0061
45 2.5914 3.3497 −0.1893 0.9494 < 0.0001 12 20.17 0.05631 0.8532 0.0250
60 5.7684 3.2394 −0.1874 0.9099 < 0.0001 14 16.87 1.1437 0.9898 0.0004
16 14.09 61.72 0.9804 0.0012

5
D.E. Salinas, et al. LWT - Food Science and Technology 116 (2019) 108456

Fig. 3. Viscosity compensation. Linear tendency between lnk0 (k0 in mPa·sn) Fig. 4. Tendency of the viscosity to the iso-consistency point for the fibre
and Ea (kJ/mol) previously obtained by fitting Arrhenius equation. contents related to the Newton model (< 8 g/100 g). Extrapolation of lnη (η in
mPa·s) versus temperature (°C) for each fibre content (■ 2 g/100 g, Δ 4 g/100 g,
՜ 6 g/100 g, ᆫ 8 g/100 g, ● Iso-consistency point).
to describe properly the effect of the temperature on the juice con-
sistency, regardless the rheological model. The values obtained for the
activation energy are similar to those obtained in other systems (Garvin
et al., 2017). Fig. 3 shows the dependence of the lnk0 on the activation
energy (Ea) when the fibre content varied, showing two different
straight lines, one for the parameters obtained from the Newton model
for fibre content from 2 to 8 g/100 g, and another one for the para-
meters obtained from both the power law and the Herschel-Bulkley
models from 10 to 16 g/100 g of fibre content. Since two straight lines
were obtained, the existence of consistency compensation and a dif-
ferent fluid flow mechanism for each straight line were concluded. The
existence of each straight line implies that the environmental variable,
in this case the fibre content, does not change the flow mechanism for
the whole range of values of the fibre content for which the relation is a
straight line. Thus, when the fibre content increases, at a value of 10 g/
100 g not only did the rheological model change from Newtonian to
non-Newtonian model, but the flow mechanism also changed. This Fig. 5. Tendency of the consistency index to the iso-consistency point for the
fibre contents related to the Non-Newton models (10–16 g/100 g).
specific fibre content value of 10 g/100 g does not follow the straight
Extrapolation of ln k (k in mPa·sn) versus temperature (°C) for each fibre content
line corresponding to the Newton model, but it is clearly close to it. So,
(• 10 g/100 g, ■ 12 g/100 g, Δ 14 g/100 g, ՜ 16 g/100 g, ● Iso-consistency
comparing both straight lines, the flow mechanism apparently changes point).
for a fibre content between 8 and 10 g/100 g.
The data in Table 6 were fitted to Equation (8). Table 7 shows the
estimates and confidence intervals for both parameters and each model each apple juice containing fibre contents from 10 to 16 g/100 g. These
with high values of the coefficient of determination and low p-values. temperatures at which either the viscosity or the consistency coefficient
Figs. 4 and 5 show the fact that, according to the compensation values are the same regardless of the fibre content are called the iso-
theory, the value for the consistency would be the same at the iso- consistency temperatures. Obviously, the iso-consistency temperature is
consistency temperature for each model. These figures show the ex- only a concept and in some cases it is not possible to check it experi-
trapolated values as a function of the temperature, considering that the mentally, e.g., at the iso-consistency temperature (−135.9 °C) found for
behaviour would be the same as the one found for the range of tem- the non-Newtonian models (10–16 g/100 g) the juice would be frozen
peratures studied. Fig. 4 shows how the Neperian logarithm of the and would not flow.
viscosity expressed in mPa·s approximately approached the same value In general, some authors have claimed that any kind of compensa-
(0.858) at the same temperature (78.9 °C) for each apple juice with tion is not real, but only Krug, Hunter, and Grieger (1976) demon-
fibre contents up to 8 g/100 g. Fig. 5 shows how the Neperian logarithm strated that in some cases this phenomenon is caused by the propaga-
of the consistency index expressed in mPa·sn would approximately ap- tion of experimental errors. When the harmonic mean temperature
proach the same value (15.6) at the same temperature (−135.9 °C) for (Thm, Equation (12)) is located within the confidence interval of the iso-
consistency temperature, the phenomenon is concluded to be statisti-
cally caused by the propagation of the experimental errors, but in the
Table 7 other cases, nothing artificial has been shown to cause this phenom-
Consistency compensation parameters. enon (Garvin et al., 2017).
Fibre Content (g/100 g) Newtonian Non-Newtonian n
Thm = n 1
∑i = 1 Ti (12)
0–8 10–16

Tiso (°C) Value 78.9 −135.9 According to the criterion of Krug et al. (1976), Thm was 29.6 °C.
Confidence interval −5.1 to 239.7 −185.6 to −43.4 Since this temperature is inside the confidence interval for the iso-
ln(k(Tiso)) Value 0.858 15.6 consistency temperature for fibre content up to 8 g/100 g (Newtonian
Confidence interval −0.39 to 2.10 5.75 to 25.4
model), the compensation cannot be concluded to be real because it
R2 0.9858 0.9665
p-value 0.0071 0.0169 could be exclusively due to the propagation of the experimental errors.
With no real compensation and without an iso-consistency temperature,

6
D.E. Salinas, et al. LWT - Food Science and Technology 116 (2019) 108456

Table 8 E
k = aexp ⎛ a + bC + dC 2 ⎞
Combined effect of the temperature and the fibre content on the consistency ⎝ RT ⎠ (14)
(either the viscosity or the consistency effect).
where the parameters are the same as the previous equations and a, b
Equation 13 Equation 14 and d are constants (units in Table 8).
k = aexp ( Ea
RT
+ bC ) k = aexp ( Ea
RT
+ bC + dC2 ) The combined effect of temperature and fibre content was studied
by fitting the data in Table 3 to Equations (13) and (14). Table 8 shows
n −5
a (mPa·s ) 2.197·10 0.01 the parameters fitted to both equations. As expected (Ibarz, Pagán,
Ea (kJ/mol) 13.12 13.84
Gutiérrez, & Vicente, 1989), the activation energy found this way was
b (g/100 g)−1 0.952 −0.321
d (g/100 g)−2 0.054 approximately the average mean for the activation energy found by the
R2 0.9883 0.9977 Arrhenius equation for each fibre content. The parameters related to the
p-value < 0.0001 < 0.0001 concentration were also approximately the average mean of the same
parameters in equations (9) and (11). These activation energy values
are similar to the values found for other similar systems (Garvin et al.,
nothing could be stated about the control mechanism of the flow pro- 2017).
cess.
In the case of the non-Newtonian models (10–16 g/100 g), Thm is 4. Conclusions
clearly out of the confidence interval for the iso-consistency tempera-
ture and thus real compensation was concluded. In this case, the con- The apple juice at 12.5°Brix showed thixotropy for fibre contents
sistency of the juices was concluded to depend on the fibre content and above 10 g/100 g, following the Figoni-Shoemaker model. The decrease
the mechanism of flow was also concluded to be the same for the whole in temperature and increase in fibre content caused an increase in all
range of fibre contents. Since the whole confidence interval for the iso- the parameters related to the internal structure, while the kinetic con-
consistency temperature was found to be lower than the whole range of stant of degradation of the internal structure shows no clear tendency
experimental temperatures, the control of the flow process was con- but it roughly increases with the temperature and fibre content.
cluded to be entropic rather than enthalpic for all the working tem- The flow behaviour followed the Newton, power law and Herschel-
peratures. Bulkley models for fibre contents up to 8 g/100 g, from 10 to 12 g/
Although many studies have included the concept of compensation 100 g, and between 14 and 16 g/100 g, respectively. For all cases, either
in physical changes, as far as we know only two of them are related to the viscosity or the consistency coefficient increased while the tem-
the flow behaviour (Garvin et al., 2017) and neither apply all the perature decreased and the fibre content increased. The deviation from
concepts related. Ozilgen and Bayindirli (1992) studied the viscosity of the Newtonian flow increased with the addition of fibre, not only re-
some fruit juices using the soluble solid content as the environmental presented by the change of model but also by the decrease in the index
variable. The working temperature was between −15 and 80 °C. Sour of flow behaviour (n) and the increase in the yield stress (σ0) once it
cherry, clarified apple, apple and grape juices were studied at soluble appeared. The dependence on the concentration for each temperature
solid contents between 12 and 73.5°Brix and the consistency compen- was found to follow an exponential and a combination of exponential
sation was concluded for all cases, finding iso-consistency temperatures and polynomial functions.
of 187, 206, 104 and 114 °C, respectively. Since the harmonic mean The effect of the temperature was found to follow the Arrhenius
temperature was always located out of the confidence interval of the equation, showing two compensation lines, one for the juices with a
iso-consistency temperature, the compensation could be claimed as fibre content up to 8 g/100 g that followed the Newton model, and
real. As the iso-consistency temperature was always significantly higher another for the juices containing more than 10 g/100 g of fibre that
than the experimental range of temperatures, the control could have followed the non-Newtonian models. The iso-consistency temperature
been concluded as enthalpic. The existence of consistency compensa- was found to be 78.9 and −135.9 °C for the Newtonian and non-
tion showed that the soluble solid content influenced the viscosity Newtonian models, respectively. The compensation for the Newtonian
without changing the physical mechanism. Dik and Özilgen (1994) juices was found to be less significant than the propagation of the ex-
studied the rheological behaviour of bentonite-apple juice dispersions perimental errors while the compensation for the non-Newtonian juices
using the bentonite content as the environmental variable at a range of was found to be real, for which the control was entropic at the working
temperatures from 10 to 50 °C, concluding that the power law model temperatures.
fitted properly. They concluded consistency compensation and an iso- Two general equations were also found to describe the combined
consistency temperature of −268 °C. Since the iso-consistency tem- dependence of the temperature and the fibre content on either the
perature was lower than the whole range of temperatures, the com- viscosity or the consistency index.
pensation was concluded to be real and the control entropic. Thus, the
bentonite content influenced the consistency index without changing Acknowledgement
the physical mechanism.
For some chemical reactions, an equilibrium reaction can be asso- The authors would like to thank the companies of Indulleida, for
ciated to the reaction mechanism, the equilibrium showing a different providing the apple fibre and the apple juice.
but related compensation (Garvin et al., 2017). In this case, since none
equilibrium reaction is related to the flow, it did not make any sense to References
study the thermodynamic compensation.
Association of official analytical chemists (1997). Official methods of analysis. Plants (16th
ed.). Washington, DC: AOAC.
3.3.3. Combined effect of the temperature and the fibre content Augusto, P. E., Falguera, V., Cristianini, M., & Ibarz, A. (2011). Influence of fibre addition
on the rheological properties of peach juice. International Journal of Food Science and
There are some equations that can be used to evaluate the combined
Technology, 46, 1086–1092.
effect of the concentration and temperature (Ibarz et al., 1995, 1996, Canet, W., Alvarez, M. D., Fernández, C., & Luna, P. (2005). Comparison of methods for
1989; Khalil, Ramakrishna, Nanjundaswamy, & Patwardhan, 1989): measuring of yield stress in potato puree: Effect of temperature and freezing. Journal
of Food Engineering, 68, 143–153.
Cepeda, E., Villarán, M. C., & Ibarz, A. (1999). Rheological properties of cloudy and
E
k = aexp ⎛ a + bC ⎞ clarified juice of Malus floribunda as a function of concentration and temperature.
⎝ RT ⎠ (13) Journal of Texture Studies, 30, 481–491.
Dik, T., & Özilgen, M. (1994). Rheological behaviour of bentonite-apple juice dispersions.

7
D.E. Salinas, et al. LWT - Food Science and Technology 116 (2019) 108456

LWT-Food Science and Technology, 27, 55–58. Food Science and Technology, 27, 16–18.
Ewell, H. (1938). The reaction rate theory of viscosity and some of its applications. Ozilgen, M., & Bayindirli, L. (1992). Frecuency factor-activation energy compensation
Journal of Applied Physics, 9, 252–269. relations for viscosity of the fruit juices. Journal of Food Engineering, 17, 143–151.
Falguera, V., Vélez-Ruiz, J. F., Alins, V., & Ibarz, A. (2010). Rheological behaviour of Prosky, L., Asp, N. G., Schweizer, T. F., DeVries, J. W., & Furda, I. (1988). Determination
concentrated Mandarin juice at low temperatures. International Journal of Food of insoluble, soluble, and total dietary fiber in foods and food products:
Science and Technology, 45, 2194–2200. Interlaboratory study. Journal-Association of Official Analytical Chemists, 71(5),
Figoni, P. I., & Shoemaker, C. F. (1981). Characterisation of structure breakdown of foods 1017–1023.
from their flow properties. Journal of Texture Studies, 12, 287–305. Rao, M. A., Cooley, M. J., & Vitali, A. A. (1984). Flow properties of concentrated juices at
Garvin, A., Ibarz, R., & Ibarz, A. (2017). Kinetic and thermodynamic compensation. A low temperatures. Food Technology, 38, 113–119.
current and practical review for foods. Food Research International, 96, 132–153. Rha, C. (1978). Rheology of fluid foods. Food Technology, 7, 32–35.
Genovese, D. B., & Rao, M. A. (2005). Components of vane yield stress of structured food Root, W. H., Somogyi, L. P., Barret, D. M., & Hui, Y. H. (1996). Apples and apple pro-
dispersions. Journal of Food Science, 70, 498–504. cessing fruits. Science and Technology, 2, 1–35.
Grigelmo-Miguel, N., Ibarz, A., & Belloso, M. O. (1999a). Flow properties of orange Sabanis, D., Lebesia, D., & Tzia, C. (2009). Effect of dietary fibre enrichment on selected
dietary fibre suspensions. Journal of Texture Studies, 30, 245–257. properties of gluten-free bread. LWT – Food Science and Technology, 42, 1380–1389.
Grigelmo-Miguel, N., Ibarz, A., & Belloso, M. O. (1999b). Rheology of peach dietary fibre Sánchez-Zapata, E., Muñoz, C. M., Fuentes, E., Fernández-López, J., Sendra, E., Sayas, E.,
suspensions. Journal of Food Engineering, 39, 91–99. et al. (2010). Effect of tiger nut fibre on quality characteristics of pork burger. Meat
Grigelmo-Miguel, N., Ibarz, A., & Belloso, M. O. (1999c). Comparison of dietary fibre from Science, 85, 70–76.
by-products of processing fruits and greens and from cereals. LWT-Food Science and Sato, A. C. K., & Cunha, R. L. (2009). Effect of particle size on rheological properties of
Technology, 32(8), 503–508. jaboticaba pulp. Journal of Food Engineering, 91, 566–570.
Guillon, F., & Champ, M. (2000). Structural and physical properties of dietary fibres, and Sendra, E., Kuri, V., Fernández-López, J., Sayas-Barberá, E., Navarro, C., & Pérez-Alvarez
consequences of processing on human physiology. Food Research International, 33, (2010). Viscoelastic properties of orange fiber enriched yogurt as a function of fiber
233–245. dose, size and termal treatment. LWT-Food Science and Technology, 43, 708–714.
Ibarz, A., & Barbosa-Cánovas, G. V. (2003). Unit operations in food engineering. Boca Raton, Sharma, S. K., Lemaguer, M., Liptay, A., & Poysa, V. (1996). Effect of composition on the
FL(USA): CRC Press. rheological properties of tomato thin pulp. Food Research International, 29, 175–179.
Ibarz, A., Garvín, A., & Costa, J. (1996). Rheological behaviour of sloe (Prunus Spinosa) Siche, R., Falguera, V., & Ibarz, A. (2012). Influence of temperature and addition of fiber
fruit juices. Journal of Food Engineering, 27, 423–430. in the flow behavior of orange juice. Scientia Agropecuaria, 3, 303–308.
Ibarz, A., Giner, J., Pagan, J., Gimeno, V., & Garza, S. (1995). Rheological behaviour of Siche, R., Falguera, V., & Ibarz, A. (2015). Use of response surface methodology to de-
kiwi fruit juice concentrates. Journal of Texture Studies, 26, 134–145. scribe the combined effect of temperature and fiber on the rheological properties of
Ibarz, A., González, C., Esplugas, S., & Vicnete, M. (1992). Rheology of clarified fruit orange juice. Journal of Texture Studies, 46, 67–73.
juices. I: Peach juices. Journal of Food Engineering, 15, 49–61. Silva, V. M., Sato, A. C. K., Barbosa, G., Dacanal, G., Ciro-Velasquez, H. J., & Cunha, R. L.
Ibarz, A., Pagán, J., Gutiérrez, J. M., & Vicente, M. (1989). Rheological properties of (2010). The effect of homogenisation on the stability of pineapple pulp. International
clarified pear juice concentrates. Journal of Food Engineering, 10, 57–63. Journal of Food Science and Technology, 45, 2127–2133.
Ibarz, A., Vicente, M., & Graell, J. (1987). RheologicaI behaviour of apple juice and pear Staffolo, M. D., Bertola, N., Martino, M., & Bevilacqua, A. (2004). Influence of dietary
juice and their concentrates. Journal of Food Engineering, 6, 257–267. fibre addition on sensory and rheological properties of yogurt. International Dairy
Kaushal, N. K., & Joshi, V. K. (1995). Preparation and evaluation of apple pomace based Journal, 14, 263–268.
cookies. Indian Food Packer, 49, 17–24. Tonon, R. V., Alexandre, D., Hubinger, M. D., & Cunha, R. L. (2009). Steady and dynamic
Khalil, K. E., Ramakrishna, P., Nanjundaswamy, A. M., & Patwardhan, M. V. (1989). shear rheological properties of aç ai pulp (Euterpe oleraceae Mart.). Journal of Food
Rheological behaviour of clarified banana juice: Effect of temperature and con- Engineering, 92, 425–431.
centration. Journal of Food Engineering, 10, 231–240. Vitali, A. A., & Rao, M. A. (1982). Flow behaviour of guava as a function of temperature
Kokini, J. (1992). Rheological properties of foods. In D. R. Heldman, & D. B. Lund (Eds.). and concentration. Journal of Texture Studies, 13, 275–289.
Handbook of food engineering. New York (USA): Marcel Dekker. Vitali, A. A., & Rao, M. A. (1984a). Flow properties of low-pulp concentrated orange juice:
Krug, R. R., Hunter, W. G., & Grieger, R. A. (1976). Enthalphy-entropy compensation. 1. Serum viscosity and effect of pulp content. Journal of Food Science, 49, 876–881.
Some fundamental statistical problems associated with the analysis of Van't Hoff and Vitali, A. A., & Rao, M. A. (1984b). Flow properties of low-pulp concentrated orange juice:
Arhenius data. The Journal of Physical Chemistry, 80, 2235–2341. Effect of temperature and concentration. Journal of Food Science, 49, 882–888.
Leffler, J. E. (1955). The enthalpy-entropy relationship and its implications for organic Yoo, B., & Rao, M. A. (1994). Effect of unimodal particle size and pulp content on
chemistry. Journal of Organic Chemistry, 20, 1202–1231. rheological properties of tomato puree. Journal of Texture Studies, 25, 421–436.
Liu, L., & Guo, Q. (2001). Isokinetic relationship, isoequilibrium relationship, and en- Yoo, B., & Rao, M. A. (1996). Creep and dynamic rheological behaviour of tomato con-
thalpy-entropy compensation. Chemical Reviews, 201, 673–695. centrates: Effect of concentration and finisher screen size. Journal of Texture Studies,
Lozano, J. E., & Ibarz, A. (1994). Thixotropic behaviour of concentrated fruit pulps. LWT- 27, 451–459.