Escalating the technical bounds for the production of cellulose-aided peach leathers: From the benchtop to the pilot plant

Carbohydrate Polymers 245 (2020) 116437

Contents lists available at ScienceDirect

Carbohydrate Polymers
journal homepage: www.elsevier.com/locate/carbpol

Escalating the technical bounds for the production of cellulose-aided peach
leathers: From the benchtop to the pilot plant

T

Giuliana T. Francoa,b,1, Caio G. Otonia,c,1,*, Beatriz D. Lodia, Marcos V. Lorevicea,
Márcia R. de Mourad, Luiz H.C. Mattosoa,*
a

Nanotechnology National Laboratory for Agriculture (LNNA), Embrapa Instrumentation, Rua XV de Novembro, 1452, São Carlos, SP 13560-979, Brazil
PPGQ, Department of Chemistry, Federal University of São Carlos, Rod. Washington Luís, km 235, São Carlos, SP 13565-905, Brazil
c
Department of Materials Engineering, Federal University of São Carlos, Rod. Washington Luís, km 235, São Carlos, SP 13565-905, Brazil
d
Department of Physics and Chemistry, FEIS, São Paulo State University, Av. Brasil, 56, Ilha Solteira, SP 15385-000, Brazil
b

ARTICLE INFO

ABSTRACT

Keywords:
Prunus persica
Continuous casting

Edible film
Ternary mixture design
Response surface methodology
Food packaging

This contribution falls within the context of sustainable functional materials. We report on the production of fruit
leathers based chiefly on peach pulp, but combined with hydroxypropyl methylcellulose (HPMC) as binding
agent and cellulose micro/nanofibrils (CMNF) as fillers. Increased permeability to moisture (from 0.9 to 5.6 g
mm kPa−1 h−1m−2) and extensibility (from 10 to 17%) but reduced mechanical resistance (67–2 MPa) and
stiffness (1.8 GPa–18 MPa) evidenced the plasticizing effect of peach pulp in HPMC matrix, which was reinforced
by CMNF. A ternary mixture design allowed building response surfaces and optimizing leather composition. The
laboratory-scale leather production via bench casting was extended to a pilot-scale through continuous casting.
The effect of scaling up on the nutritional and sensory features of the peach leather was also depicted. The herein
established composition-processing-property correlations are useful to support the large-scale production of
peach leather towards applications both as packaging materials and as nutritional leathers.

1. Introduction
Strategies towards increased shelf life have become increasingly
demanded due to the complex food supply chain faced by today’s society, involving long distances and periods of transportation and storage. Packaging systems have gained prominence due to their role as
physical hurdles against food dehydration and spoilage. Ongoing is also
the trend of replacing petrochemical building blocks of non-biodegradable materials by rapidly renewable raw materials for biodegradable packaging (Ahmadzadeh & Khaneghah, 2019). This drives one’s
attention not only to what comes from Nature, but also to what goes
back to the environment. Edible packaging – i.e., that comprising exclusively food-grade components, regardless of processing (Cerqueira,
2019) – stands out in this context, particularly for single-use applications, as waste is either not generated or readily biodegradable. Should

such materials be also intended to protect foodstuffs, these must perform suitably from the physical-mechanical standpoint, as extensively
demonstrated for edible coating and self-standing films based on
naturally occurring polysaccharides and polypeptides (Atarés & Chiralt,
2016; Dehghani, Hosseini, & Regenstein, 2018).
Fruits and vegetables have been increasingly exploited as edible

film-forming matrices, in combination with natural polymer or by
themselves thanks to their typically high loads of carbohydrates and/or
proteins (Otoni et al., 2017). Depending on the formulation and processing conditions, self-standing edible layers having fruit pulp or puree
as main ingredient can behave rheomechanically like thermoplastics –
bioplastics – or present a leathery consistency – fruit leathers (Otoni
et al., 2017). Both approaches may not only contribute to solving the
aforementioned sustainability related issues, but also to human health
because of their nutritional load, as well as to consumer perception due

Abbreviations: ANOVA, analysis of variance; CMNF, cellulose micro/nanofibrils; DPPH, 2,2-diphenyl-1-picrylhydrazyl; HPMC, hydroxypropyl methylcellulose; LFF,
leather-forming formulations; MW, molecular weight; RDI, recommended daily intakes; RH, relative humidity; Trolox, 6-hydroxy-2,5,7,8-tetramethylchroman-2carboxylic acid; WVP, water vapor permeability
⁎
Corresponding authors at: Nanotechnology National Laboratory for Agriculture (LNNA), Embrapa Instrumentation, Rua XV de Novembro, 1452, São Carlos, SP
13560-970, Brazil.
E-mail addresses: (G.T. Franco), (C.G. Otoni), (B.D. Lodi),
(M.V. Lorevice), (M.R.d. Moura), (L.H.C. Mattoso).
1
These authors contributed equally.
/>Received 5 February 2020; Received in revised form 9 April 2020; Accepted 9 May 2020
Available online 24 May 2020
0144-8617/ © 2020 Elsevier Ltd. All rights reserved.


Carbohydrate Polymers 245 (2020) 116437

G.T. Franco, et al.

to their unique sensory features. The possibilities of using overripe
produce (Aguirre-Joya et al., 2018) or even side streams from its processing (Otoni, Lodi et al., 2018) make such a novel class of materials
further appealing by valorizing typically underutilized resources, which

in turn is expected to culminate in diminished food waste. It is noteworthy that, owing to their low water activity levels, fruit leathers may
be taken as dehydrated foods and ought to be stable as far as microbial
spoilage as long as they are protected from moisture (Otoni et al.,
2017). From the shelf life standpoint, while fruit leathers are ideal to
wrap dry foodstuffs within also dry environments, these can be suitably
used in humid atmospheres provided they are further enclosed within a
high-barrier secondary packaging, but they are less likely to serve as
primary packaging for moist foods.
In line, the exploitation of underutilized resources is also expected
to reduce production costs and to add value to otherwise discarded raw
materials. Also related with the production costs of bioplastics and fruit
leathers is the material-forming method itself, which is traditionally
carried out in batch-mode solvent casting, process that involves laboratory scale, long drying times, and low yields – herein referred to as
bench casting (de Moraes, Scheibe, Sereno, & Laurindo, 2013; Otoni
et al., 2017). The continuous analogue of bench casting has been demonstrated to remarkably increase the yield of edible layers made up of
fruits and vegetables (Munhoz et al., 2018; Otoni, Lodi et al., 2018),
even though maintaining the properties remains challenging because of
the relatively low thermal stability of some of their constituents. From
the mechanical standpoint, specifically, should fruit-only bioplastics
perform poorly, these can be combined to other food-grade matrices
and fillers, such as lignocellulosics – e.g., hydroxypropyl methylcellulose (HPMC), a nonionic cellulose ether that is widely known for its
film-forming ability (Hay et al., 2018) and for being generally recognized as safe (FDA; GRAS Notice No. GRN 000213, 2007) and approved as a food additive (US FDA, 21 CFR 172.874, 2011; and European Union, EPCD No. 95/2/EC, 1995). Among these biorenewables,
cellulose micro/nanofibrils (CMNF) denote efficient fillers for mechanically reinforcing edible bioplastics (Valencia, Nomena, Mathew, &
Velikov, 2019; Viana, Sá, Barros, Borges, & Azeredo, 2018).
In the context presented above, this contribution is devoted to fruit
leathers comprising peach pulp as main component and HPMC and
CMNF, herein exploited as binding and reinforcing agents, respectively.
The role played by each of these components on the physical-mechanical properties of the resulting materials was fully depicted.
Finally, this study also set out to scale up the production of the peach
leathers from bench casting to its continuous equivalent, as well as to

either confirm or refute the hypothesis that the leather-forming protocol influences the key properties of such materials, including nutritional and sensory aspects.

USA) was used in all experiments.
2.2. Leather production via bench casting
The components listed above were mechanically stirred at 1500 rpm
for 30 min under vacuum (500 mmHg) into leather-forming formulations (LFF) that were allowed to rest under vacuum for another 30 min
to remove bubbles before being spread with uniform thickness onto a
poly(ethylene terephthalate) sheet, where they were dried at room
temperature and 50 ± 10% relative humidity (RH) for 24 h. Dried
leathers were peeled from the casting surface and equilibrated at room
temperature and 50% RH for at least 48 h in a desiccator containing
saturated magnesium nitrate solution before used for testing.
2.3. Leather production via continuous casting
Leathers were also produced in a continuous fashion on a KTF-B
lamination system (Werner Mathis AG, Switzerland) that comprised
four main steps, namely: 1, feeding: the LFF was poured on a Mylar
(DuPont Teijin Films U.S. Ltd., USA) conveyor belt; 2, lamination: the
LFF was forced through a gap between the polyester substrate and a
knife into a 1.50-mm-thick, 26-cm-wide wet layer; 3, drying: the wet
LFF layer was conveyed through an infrared pre-drying stage (at ca. 45
°C and 0.10 m⋅min−1 for 30 cm) and two convective drying stages (at
120 °C and 0.10 m min−1 for 92 cm each); and 4, winding. The feedingto-winding distance and time were 3 m and 30 min, respectively. Prior
to any testing, dried leathers were stored as described previously.
2.4. Experimental design, optimizations, and scale up
Peach pulp, HPMC, and CMNF were combined at different proportions according to a ternary mixture design (Table S1). The weight ratio
(dry basis) of each component ranged from 0 to 1, their weighted
contributions together adding up to 1. Bench casting (Section 2.2) was
used at this stage. Tensile strength ( T ), Young’s modulus (E), and
elongation at break ( B ) were determined by tensile assay on a DMA
Q800 dynamic-mechanical analyzer (TA Instruments, Inc., USA), calculated by Eq. 1 to Eq. 3, and taken as dependent variables for optimization purposes.

T

(1)

= F / A0

B=[(LB

L0 )/ L0]·100

E = lim / L
L

L0

(2)
(3)

Where F is the maximum load, L is the extension (at a given point:
no index; at the beginning of the assay: index 0; at break: index B), σ is
the stress at a given point, and A0 is the initial cross-sectional area, the
latter considering sample thickness, which was measured with a digital
micrometer (Mitutoyo Corp., Japan) by averaging five random measurements. The load cell was 18 N and the stretching rate was 0.1%
min−1. The response variables were used to build response surfaces.
To investigate the isolated effect of peach pulp on the mechanical
and barrier properties of the leathers, 2.5, 5.0, or 7.5% (w/v) of peach
pulp was mixed with 2% (w/v) of HPMC in water. The LFF were converted into leathers by bench casting (Section 2.2) and characterized as
described in Sections 2.5 and 2.6. Finally, to assess the effect of scaling
up leather production on its key properties, the optimum LFF (XPP =
0.85, XHPMC = 0.15; further discussion below) was also processed

through continuous casting (Section 2.3), and the nutritional and sensory properties of the continuously and bench-cast leathers were compared.

2. Materials and methods
2.1. Materials
Pasteurized peach pulp (De Marchi, Brazil), HPMC Methocel® E4M
(CAS no. 9004-65-3; The Dow Chemical Company, Brazil) – comprehensively characterized elsewhere [degree of substitution: 1.9; Mw: ca.
350,000 g mol−1 (Otoni, Lorevice, Moura, & Mattoso, 2018)] – and 2,2diphenyl-1-picrylhydrazyl (DPPH) and 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid, or Trolox (Sigma-Aldrich Co. LLC,
USA) were used as received. Microcrystalline cellulose Sigmacell® Type
50 (CAS no. 9004-34-6; Sigma-Aldrich Brasil Ltd, Brazil) – ζ potential:
-1.0 ± 0.4 mV; apparent particle size: 1.6 ± 0.9 μm; crystallinity index:
62 ± 2% (Otoni, Carvalho et al., 2018) – was dispersed in water at 1%
(w/v) without any pretreatment or purification and high-pressure microfluidized (Microfluidizer® M-110 P; Microfluidics Corp., USA) for
seven cycles at 138 MPa into CMNF – ζ potential: -30 to -24 mV; apparent particle size: 160−250 nm; crystallinity index: 70–75% (Otoni,
Carvalho et al., 2018). Ultrapure water (Barnstead Nanopure Diamond,

2.5. Mechanical properties
The leathers were submitted to uniaxial tensile assay on a DL3000
2


Carbohydrate Polymers 245 (2020) 116437

G.T. Franco, et al.

Fig. 1. Surface response plots for the mechanical properties of biocomposites comprising different weight ratios of peach pulp (PP), hydroxypropyl methylcellulose
(HPMC), and cellulose micro/nanofibrils (CMNF).

universal testing machine (EMIC Equipamentos e Sistemas de Ensaio
Ltda., Brazil) equipped with a 10-kgf load cell. At least six leather
specimens per treatment, shaped in accordance with ASTM D882−18,

were stretched at 10 mm min−1 from an initial length (L0 ) of 100 mm
until rupture. The mechanical attributes T , B , and E were calculated
using Eq. (1), Eq. (2), and Eq. (3), respectively.

Williams, Cuvelier, and Berset (1995). Briefly, 0.1 mL of different dilutions (in methanol) of the supernatant was mixed with 3.9 mL of a
0.0024% (w/v) solution of DPPH, also in methanol. The mixtures were
kept in the dark at room temperature for 30 min before having their
absorbances at 515 nm measured on a UV-1601PC spectrophotometer
(Shimadzu Co., Japan). Leather-free methanol and DPPH solutions were
used as blank and control, respectively. An analytical curve was adjusted by varying the concentrations of Trolox – a standard antioxidant
analogous to vitamin E – from 80 to 800 mg L−1 and used to quantitatively express the antioxidant capacity of the leathers, in μg of Trolox
equivalent per g of sample (μg g−1).

2.6. Barrier to water vapor
The water vapor permeability (WVP) of the leathers was measured
in accordance with the modification of ASTM E96−80 proposed by
McHugh, Avena-Bustillos, and Krochta (1993). In brief, at least four
leather specimens per treatment were fixed onto the edges of poly
(methyl methacrylate) capsules with circular openings of 5.1 cm in
diameter, serving as semi-permeable barriers between an inner high-RH
environment and a chamber at 30 ± 1 °C and 30 ± 2% RH. The capsules were weighed periodically to determine WVP.

2.8. Content of minerals
The content of minerals in the fruit leathers was determined by
atomic absorption spectrometry. Continuously and bench-cast leathers
(1−2 g, measured accurately) were digested in a mixture of nitric acid
P.A. (10 mL), hydrochloric acid P.A. (5 mL), and 30 vol. hydrogen
peroxide (3 mL) on a Kjeldahl digester (model SL-25/40, Solab
Equipamentos para Laboratório Ltda., Brazil) at 140 °C for 24 h and
then filtered through filter paper into 50-mL volumetric flasks that were

further completed with ultrapure water. The resulting solutions were
analyzed on a PinAAcle 900 T spectrometer (PerkinElmer, Inc., USA),
with flame and radiation sources at specific wavelengths to determine
the following minerals: Fe, 248.33 nm; K, 766.49 nm; Mg, 285.21 nm;

2.7. Antioxidant capacity
Continuously and bench-cast leathers (1−2 g, measured accurately)
were dipped in 20 mL 99.8% pure methanol at 4 °C for 24 h. The
methanolic extracts were centrifuged (Rotina 380 R, Andreas Hettich
GmbH & Co. KG, Germany) for 10 min at 15 °C and 10,000 rpm and had
their antioxidant capacities determined as described by Brand3


Carbohydrate Polymers 245 (2020) 116437

G.T. Franco, et al.

Ca, 422.67 nm; Mn, 279.48 nm; and Na, 589.00 nm. Analytical curves
were previously built with standard solutions of each of these minerals
and used for the quantitative interpretation of the results. The runs
were carried out in triplicates.

tensile behavior of the peach leathers. The surface response plots resulting from the ternary mixture design are presented in Fig. 1. The
statistical procedure allowed establishing models that were efficient in
fitting the acquired mechanical data. The regression coefficients of such
models are presented in Table S2.
The response surface plots elucidate the role played by each of the
components: HPMC serves as binding agent and leather-forming matrix,
higher HPMC contents being accompanied by also higher tensile
strengths; CMNF act as fillers for mechanical reinforcement, thus

leading to stronger leathers; and peach pulp, introduced as the major
functional ingredient, but from the mechanical standpoint behaving as
a plasticizer by increasing leather extensibility and decreasing its resistance and stiffness. This mapping strategy allows one to engineer the
mechanical properties of materials made up of such a system without
the need for further testing. Importantly, we demonstrate the possibility
of producing biocomposite leathers with varying mechanical performances by simply altering the composition of the LFF, ranging from
rubbery self-standing layers – suitable for applications as edible leathers, for instance – all the way to stiff sheets – suitable for e.g. food
packaging. Because we targeted at leather that are functional – and as
preliminary tests showed that neither HPMC nor CMNF themselves
have antioxidant activity – but still provide sufficient mechanical
strength to be used as self-supporting layers, the LFF chosen for the
further steps comprised 85 wt.% peach pulp and 15 wt.% HPMC. This
approach is illustrated in Fig. 2. It is worth stressing out that, should
superior mechanical properties be required for a given application, the
established relationships can be used to alter the LFF and produce
materials with tailored mechanical performance.

2.9. Colorimetric parameters
The color of the fruit leathers was determined on a CR-400 digital
colorimeter (Konica Minolta Sensing, Inc., Japan). The brightness (L*)
and the chromaticity parameters a* (red-to-green) and b* (yellow-toblue) were determined in at least three random positions along the
leather surfaces. The LFF served as reference for calculating the total
color difference (ΔE) by Eq. 4, wherein the subscript indexes L and LFF
refer to leather and leather-forming formulations, respectively. Yellowness index (YI) and whiteness index (WI) were calculated through
Eq. 5 and Eq. 6, respectively.

E = [(LL*

*
LLFF

)2 + (aL*

YI = 142.86 bL* LL*

WI = 100

[(100

*
aLFF
) 2 + (bL*

*
bLFF
)2]0.5

(4)
(5)

1

LL* ) 2 + aL*2 + bL*2]0.5

(6)

2.10. Statistical treatment of data
Quantitative data were submitted to analysis of variance (ANOVA)
at 5% of significance followed by regression analysis or Tukey test at
the same level of significance, as suitable. Mechanical data were fitted
to quadratic regression models according to the mixture design. The

importance of model components was examined by ANOVA (also at
5%), and nonsignificant effects were disregarded from the models that
were used to plot response surfaces.

3.2. On the role of peach pulp on HPMC matrix
Because peach pulp and HPMC were selected as the constituents of
the peach leathers, the effect of mixing them at different proportions
was further investigated as far as mechanical and water barrier properties (Fig. 3).
All of the evaluated properties were affected (P < 0.05) by composition. Peach pulp itself formed a continuous layer that was easily
detachable from the casting surface, although with extremely low
stiffness and resistance while with high extensibility. As expected,
higher peach pulp contents led to leathers featuring lower resistance
and stiffness, but higher extensibility and permeability to moisture. This

3. Results and discussion
3.1. On the optimization of leather formulation
Physical-mechanical cohesion and integrity were herein taken as the
primary technical requirements for fruit leathers to be suitably used as
self-standing layers, application that would be prevented otherwise.
This was the rationale for relying the optimization procedures upon the

Fig. 2. Left: formulations of peach leathers using hydroxypropyl methylcellulose (HPMC) as binder and of biocomposites comprising cellulose micro/nanofibrils
(CMNF) as fillers; right: leather-forming formulation before (top) and after (bottom) drying either on lab- (bench casting) or pilot-scale (continuous casting). The
insets show dried leather specimens shaped for tensile assays.
4


Carbohydrate Polymers 245 (2020) 116437

G.T. Franco, et al.

Fig. 3. Mechanical properties and water vapor permeability (WVP) of peach leathers made up of the combination, at different weight ratios (dry basis) of peach pulp
and hydroxypropyl methylcellulose. Dotted lines indicate fittings to actual data points, whose equations are presented as Supplementary Material (Eq. S1 to Eq. S4).

drying the LFF indeed boosted water removal: whereas the dried leather
was detached from the bench casting substrate 24 h after the deposition
of the LFF, the feeding-to-winding time in the continuous casting was
only 30 min, i.e., a significant 48-time reduction. This was allowed by
the infrared pre-drying stage and the convective drying at higher
temperatures. As reported by Otoni, Lodi et al. (2018), who produced
carrot-based biocomposites using the same apparatus and conditions,
this aspect is important because ca. 75 m2 would be needed to produce
such an area of leather through bench casting, number that would be
reduced to ca. 5 m2 for continuous casting. This discrepancy gets increasingly enlarged as the production volume increases, corroborating
the relevance of scaling up the production of fruit leathers in the direction of large-scale industrial operations.
As important as the yield is the maintenance of the key features of
fruit leathers, which are to be preserved even after harsher processing
conditions. In this context, the antioxidant capacity, color, and content
of minerals were compared in the continuously and bench-cast leathers.
The antioxidant capacity of both leathers as well as their contents of the
minerals K, Na, Mg, Ca, Fe, and Mn are presented in Table 1.
The processing method had either low or no influence on the mineral content, which was somehow expected due to the high thermal
stability of these compounds. Considering the recommended daily intakes (RDI) of these minerals in adult diets, as defined by the Brazilian
Health Regulatory Agency (ANVISA) through Resolution RDC No. 269,
the produced leathers, regardless of the processing method, can be
classified as sources – 100 g of leather provide more than 15% of the
RDI – of the macronutrient potassium and the micronutrient manganese. Additionally, these leathers are classified as having high contents
– 100 g of leather provides more than 30% of the RDI – of the micronutrient iron. Also importantly, it is classified as having low sodium
content – 100 g of leather comprises less than 120 mg of sodium.
As for the antioxidant capacity, the processing at higher temperatures led to a reduction, but both continuously and bench-cast leathers

were highly antioxidant. This is also expected as fruits and vegetables
are overall known to have several antioxidant compounds, including
phenolics, flavonoids, terpenes, sterols, saponins, glycosinolates, and
carotenoids, being often associated with reduced risk of cardiovascular
disease, cancer, arteriosclerosis, and other diseases related to the aging
process and induced by the formation of free radicals (Du, AvenaBustillos, Breksa, & McHugh, 2014). Through different mechanisms,

Table 1
Antioxidant capacity (AC), contents of minerals, and their recommended daily
intakes (RDI) in peach leathers produced on laboratory and pilot scales.
Mineral

Bench casting
/mg 100 g−1

Continuous casting
/mg 100 g−1

RDI
/mg 100 g−1

K
Na
Mg
Ca
Fe
Mn
AC (μg Trolox g−1)

995 ± 53a

48.5 ± 0.2a
32.5 ± 0.1a
17 ± 1a
4.9 ± 0.1a
0.66 ± 0.02a
2920 ± 21b

920 ± 34a
53.4 ± 0.2b
35.7 ± 0.1a
17.7 ± 0.1a
4.8 ± 0.1a
0.64 ± 0.01a
2555 ± 58a

4700
–
260
1000
14
2.3
–

ab

Within a row, different mean ± standard deviation values (P < 0.05) are
followed by different superscript letters.

effect, which is typical of plasticizers, had already been reported after
the incorporation other fruit pulps and purees (e.g., guava and papaya)

into HPMC films (Lorevice, Moura, de, Aouada, & Mattoso, 2012;
Lorevice, Moura, de, & Mattoso, 2014). This effect can be attributed to
low-molecular weight (MW) compounds that are naturally found in
fruits, such as mono, di, and oligosaccharides (Espitia et al., 2014). As a
matter of fact, 100 g fresh peach has been shown to contain 5.2–8.8 g
sucrose – MW: 342.3 g⋅mol−1 – ca. 1.2 g of glucose – MW: 180.2 g mol−1
– and 1.1–1.4 g of fructose – MW: 180.2 g⋅mol−1 (Cascales, Costell, &
Romojaro, 2005). These low-MW components can easily accommodate
themselves between HPMC chains – MW ca. 330,000 g mol−1 for the
E4M grade –, separating them apart. In the context of mechanical
performance during uniaxial tensile assay, this plasticizing behavior
lessens the level of intermolecular interaction and allows adjacent
HPMC chains to flow more freely one over another. Regarding barrier
to moisture, separating HPMC chains increases the intermolecular volume and reduces the tortuosity of the path followed by the permeant,
in this case water vapor. Water molecules can therefore diffuse more
easily from the high- to the low-RH environments, culminating in
higher WVP values.
3.3. On the effect of scaling up leather production
Peach leathers were successfully produced through both bench and
continuous casting procedures (Fig. 2). The pilot-scale approach of
5


Carbohydrate Polymers 245 (2020) 116437

G.T. Franco, et al.

these compounds are able to stop oxidative reactions in their early
stages by the accepting free radicals, donating hydrogen atoms to serve
as free radical acceptors, or chelating metals that catalyze oxidative

reactions (Eỗa, Machado, Hubinger, & Menegalli, 2015; Reis et al.,
2015). In peach, specifically, high levels of carotenoids, flavonoids,
anthocyanins, and hydroxycinnamate have been reported (Gil, TomásBarberán, Hess-Pierce, & Kader, 2002; Zhao et al., 2015). The RDI of
antioxidant compounds for benefiting human health is from 0.75 to
0.90 g of Trolox, and fruit and vegetable intake accounts for 0.3−0.4 g
Trolox d−1 (Prior & Cao, 2000), corroborating the relevance of having
leathers as an extra form of fruit intake. In addition to the benefits to
human health, the antioxidant capacity of bioplastics, and therefore of
fruit leathers, can also be advantageous in the case of food preservation,
since oxidation reactions of organic molecules represent one of the
main mechanisms of food spoilage: besides causing nutritional – by the
loss of vitamins and essential fatty acids – and sensory depreciation – by
the occurrence of oxidative rancidity –, toxic compounds can be produced (Reis et al., 2015).
Color was taken as an indicator of the sensory quality of the peach
leathers, as it is indeed one of the most important attributes affecting
food appearance and having a pronounced influence on consumers'
perception. The colorimetric parameters of the leathers are presented in
Table 2.
Simply drying the LFF into leathers, regardless of the method, increased the values of the three coordinates of the CIELab scale, behavior
which can be attributed to the concentration of chromophore compounds. The similar total color difference values – which takes the three
colorimetric coordinates into account – between the continuously and
bench-cast leathers, in relation to the precursor LFF, suggest that the
overall color variation was not affected by the processing protocol.
When the coordinates are analyzed as isolated variables, however, the
scaled up leathers were darker than those produced by bench casting, as
indicated by the higher L* values and whiteness index of bench cast
leathers. This is attributed to non-enzymatic browning reactions, like
the Maillard reaction, which involves the interaction between reducing
carbohydrates (mainly fructose and glucose in the case of peach) and
amino acids and/or proteins upon heating, culminating in the production of dark-colored compounds such as melanoidins. Peach darkening

kinetics has been shown to be temperature-dependent, being faster at
higher temperatures, as well as to involve carotenoid degradation
during heat treatment, leading to the depreciation of the yellowish
coloration and the enhancement of the reddish hue (Ávila & Silva,
1999). Indeed, the herein produced peach leathers attained a stronger
red color when processed at 120 °C – indicated by the higher a* value –
but yellowing did not follow the same trend. It is important mentioning
that bench casting peach leather could also cause enzymatic browning
reactions, but the previous pasteurization of the peach pulp is expected
to inactivate the oxidative enzymes (e.g., peroxidase and polyphenoloxidase) responsible for such processes. Finally, although high
mechanical and barrier performances were not targeted for the herein
produced peach leathers, it is noteworthy that the WVP of bench

(7.7 ± 0.9 g mm kPa−1 h−1 m−2) and continuously (6.8 ± 0.2 g mm
kPa−1 h−1 m−2) cast leathers was not affected (P > 0.05) by the leather-forming method. The tensile resistance (3.5 ± 0.1 to 2.6 ± 0.1
MPa), stiffness (44 ± 1 to 36 ± 2 MPa), and extensibility (14.4 ± 0.3 to
12.6 ± 0.6%), conversely, were slightly reduced when the casting
process was scaled up.

Table 2
Colorimetric parameters of peach leather-forming formulations (LFF) and leathers produced on laboratory (bench casting) and pilot (continuous casting)
scales.

Declaration of Competing Interest

Parameter
L*
a*
b*
ΔE

Yellowness index
Whiteness index

LFF

Bench casting
a

35.1 ± 0.3
0.1 ± 0.0a
13.7 ± 0.3a
–
55.7
33.7

c

55 ± 2
14 ± 1b
37 ± 1b
33.7
97.6
39.7

4. Conclusions
The pilot-scale production of peach leathers was herein demonstrated to be feasible in terms of physical-mechanical, nutritional, and
sensory properties. The interdependency among leather composition,
leather-forming parameters, and leather properties was also elucidated,
confirming the roles played by peach pulp as the main component,
HPMC as binding agent, and CMNF as filler. In particular, peach pulp

presented a typical plasticizing behavior, decreasing both resistance
and stiffness while increasing extensibility and WVP of peach leathers.
CMNF, on the other hand, were efficient in providing mechanical reinforcement. The continuous casting approach was successful in removing the dispersant medium of the LFF in a significantly faster
fashion, boosting yield and productivity while maintaining the key
characteristics of the final materials, refuting the initial hypothesis. The
mineral content, for instance, was not impacted, while the continuously
cast leathers presented an even more reddish aspect than their benchcast analogues, increasing the association of the former with in natura
peach. Both leathers presented high antioxidant capacity, although
slightly reduced in the leather processed at higher temperatures.
Altogether, our results help pave the route for the large-scale production of fruit-based materials towards industrial applicability both as
packaging materials and as nutritional leathers.
Funding
This work was supported by the São Paulo Research Foundation
(FAPESP) [grant numbers 2013/14366-7, 2014/23098-9, and 2019/
06170-1], National Council for Scientific and Technological
Development (CNPq) [grant numbers 303796/2014-6, 312530/2018-8,
and 800629/2016-7], Coordination for the Improvement of Higher
Education Personnel (CAPES) [grant numbers 33001014005D‐6 and
88882.332747/2019-01], and Ministry of Science, Technology, and
Innovation (MCTI/SISNANO) [grant number 402287/2013-4].
CRediT authorship contribution statement
Giuliana T. Franco: Writing - original draft, Investigation. Caio G.
Otoni: Conceptualization, Investigation, Writing - original draft.
Beatriz D. Lodi: Investigation. Marcos V. Lorevice: Investigation,
Writing - original draft. Márcia R. de Moura: Supervision, Writing review & editing. Luiz H.C. Mattoso: Conceptualization, Funding acquisition, Project administration, Supervision, Writing - review &
editing.

None.

Continuous casting

Acknowledgements

48 ± 1b
20.8 ± 0.5c
36 ± 1b
33.2
106.8
33.6

The authors are thankful for the financial support of FAPESP (grants
no. 2013/14366-7, no. 2014/23098-9, and no 2019/06170-1), CNPq
(grants no. 303796/2014-6, no. 312530/2018-8, and no. 800629/
2016-7), SISNANO/MCTI (grant no. 402287/2013-4), CAPES (grants
no. 33001014005D‐6 and 88882.332747/2019-01), FINEP, and
Embrapa AgroNano research network. The gracious donation of HPMC
samples by The Dow Chemical Company is also acknowledged.

abc

Within a row, different mean ± standard deviation values (P < 0.05) are
followed by different superscript letters.
6


Carbohydrate Polymers 245 (2020) 116437

G.T. Franco, et al.

Appendix A. Supplementary data

G. W. (2018). Improved hydroxypropyl methylcellulose (HPMC) films through incorporation of amylose-sodium palmitate inclusion complexes. Carbohydrate
Polymers, 188, 76–84. />Lorevice, M. V., Moura, M., de, R., Aouada, F. A., & Mattoso, L. H. C. (2012).
Development of novel guava puree films containing chitosan nanoparticles. Journal of
Nanoscience and Nanotechnology, 12(3), 2711–2717. />2012.5716.
Lorevice, M. V., Moura, M., de, R., & Mattoso, L. H. C. (2014). Nanocomposite of papaya
puree and chitosan nanoparticles for application in packaging. Química Nova, 37(6),
931–936. />McHugh, T. H., Avena-Bustillos, R. J., & Krochta, J. M. (1993). Hydrophilic edible films:
Modified procedure for water vapor permeability and explanation of thickness effects. Journal of Food Science, 58(4), 899–903. />1993.tb09387.x.
Munhoz, D. R., Moreira, F. K. V., Bresolin, J. D., Bernardo, M. P., De Sousa, C. P., &
Mattoso, L. H. C. (2018). Sustainable production and in vitro biodegradability of
edible films from eellow passion fruit coproducts via continuous casting. ACS
Sustainable Chemistry & Engineering, 6(8), 9883–9892. />acssuschemeng.8b01101.
Otoni, C. G., Avena-Bustillos, R. J., Azeredo, H. M. C., Lorevice, M. V., Moura, M. R.,
Mattoso, L. H. C., & Mchugh, T. H. (2017). Recent advances on edible films based on
fruits and vegetables-A review. Comprehensive Reviews in Food Science and Food Safety,
16(5), 1151–1169. />Otoni, C. G., Carvalho, A. S., Cardoso, M. V. C., Bernardinelli, O. D., Lorevice, M. V.,
Colnago, L. A., ... Mattoso, L. H. C. (2018). High-pressure microfluidization as a green
tool for optimizing the mechanical performance of all-cellulose composites. ACS
Sustainable Chemistry & Engineering, 6(10), 12727–12735. />acssuschemeng.8b01855.
Otoni, C. G., Lodi, B. D., Lorevice, M. V., Leitão, R. C., Ferreira, M. D., Moura, M. R., &
Mattoso, L. H. C. (2018). Optimized and scaled-up production of cellulose-reinforced
biodegradable composite films made up of carrot processing waste. Industrial Crops
and Products, 121, 66–72. />Otoni, C. G., Lorevice, M. V., Moura, M. R., & Mattoso, L. H. C. (2018). On the effects of
hydroxyl substitution degree and molecular weight on mechanical and water barrier
properties of hydroxypropyl methylcellulose films. Carbohydrate Polymers, 185,
105–111. />Prior, R. L., & Cao, G. (2000). Analysis of botanicals and dietary supplements for antioxidant capacity: A review. Journal of AOAC International, 83(4), 950–956.
Reis, L. C. B., de Souza, C. O., da Silva, J. B. A., Martins, A. C., Nunes, I. L., & Druzian, J. I.
(2015). Active biocomposites of cassava starch: The effect of yerba mate extract and
mango pulp as antioxidant additives on the properties and the stability of a packaged

product. Food and Bioproducts Processing, 94, 382–391. />fbp.2014.05.004.
Valencia, L., Nomena, E. M., Mathew, A. P., & Velikov, K. P. (2019). Biobased cellulose
nanofibril–oil composite films for active edible barriers. ACS Applied Materials &
Interfaces, 11(17), 16040–16047. />Viana, R. M., Sá, N. M. S. M., Barros, M. O., Borges, M. F., & Azeredo, H. M. C. (2018).
Nanofibrillated bacterial cellulose and pectin edible films added with fruit purees.
Carbohydrate Polymers, 196, 27–32. />Zhao, X., Zhang, W., Yin, X., Su, M., Sun, C., Li, X., & Chen, K. (2015). Phenolic composition and antioxidant properties of different peach [Prunus persica (L.) Batsch]
cultivars in China. International Journal of Molecular Sciences, 16(12), 5762–5778.
/>
Supplementary material related to this article can be found, in the
online version, at doi: />References
Aguirre-Joya, J. A., De Leon-Zapata, M. A., Alvarez-Perez, O. B., Torres-León, C., NietoOropeza, D. E., Ventura-Sobrevilla, J. M., ... Aguilar, C. N. (2018). Basic and applied
concepts of edible packaging for foods. Food packaging and preservation1–61. https://doi.
org/10.1016/B978-0-12-811516-9.00001-4.
Ahmadzadeh, S., & Khaneghah, A. M. (2019). Role of green polymers in food packaging.
Reference module in materials science and materials engineering />1016/B978-0-12-803581-8.10576-4.
Atarés, L., & Chiralt, A. (2016). Essential oils as additives in biodegradable films and
coatings for active food packaging. Trends in Food Science & Technology, 48, 51–62.
/>Ávila, I. M. L. B., & Silva, C. L. M. (1999). Modelling kinetics of thermal degradation of
colour in peach puree. Journal of Food Engineering, 39(2), 161–166. />10.1016/S0260-8774(98)00157-5.
Brand-Williams, W., Cuvelier, M. E., & Berset, C. (1995). Use of a free radical method to
evaluate antioxidant activity. LWT - Food Science and Technology, 28(1), 25–30.
/>Cascales, A. I., Costell, E., & Romojaro, F. (2005). Effects of the degree of maturity on the
chemical composition, physical characteristics and sensory attributes of peach
(Prunus persica) cv. Caterin. Food Science and Technology International, 11(5),
345–352. />Cerqueira, M.Â. P. R. (2019). Edible packaging. Encyclopedia of food chemistry173–176.
/>de Moraes, J. O., Scheibe, A. S., Sereno, A., & Laurindo, J. B. (2013). Scale-up of the
production of cassava starch based films using tape-casting. Journal of Food
Engineering, 119(4), 800–808. />Dehghani, S., Hosseini, S. V., & Regenstein, J. M. (2018). Edible films and coatings in
seafood preservation: A review. Food Chemistry, 240, 505–513. />1016/j.foodchem.2017.07.034.
Du, W.-X., Avena-Bustillos, R. J., Breksa, A. P., & McHugh, T. H. (2014). UV-B light as a

factor affecting total soluble phenolic contents of various whole and fresh-cut specialty crops. Postharvest Biology and Technology, 93, 7282. />j.postharvbio.2014.02.004.
Eỗa, K. S., Machado, M. T. C., Hubinger, M. D., & Menegalli, F. C. (2015). Development of
active films from pectin and fruit extracts: Light protection, antioxidant capacity, and
compounds stability. Journal of Food Science, 80(11), C2389–C2396. />10.1111/1750-3841.13074.
Espitia, P. J. P., Du, W.-X., Avena-Bustillos, R., de, J., Soares, N., de, F. F., & McHugh, T.
H. (2014). Edible films from pectin: Physical-mechanical and antimicrobial properties - A review. Food Hydrocolloids, 35, 287–296. />2013.06.005.
Gil, M. I., Tomás-Barberán, F. A., Hess-Pierce, B., & Kader, A. A. (2002). Antioxidant
capacities, phenolic compounds, carotenoids, and vitamin C contents of nectarine,
peach, and plum cultivars from California. Journal of Agricultural and Food Chemistry,
50(17), 4976–4982. />Hay, W. T., Fanta, G. F., Peterson, S. C., Thomas, A. J., Utt, K. D., Walsh, K. A., ... Selling,

7