Journal of Food Engineering xxx (2015) 1e9

Contents lists available at ScienceDirect

Journal of Food Engineering

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

Puff pastry with low saturated fat contents: The role of fat and dough
physical interactions in the development of a layered structure
Stefano Renzetti a, b, *, Roderik de Harder b, Albert Jurgens a, b
a
TI Food and Nutrition, Wageningen, The Netherlands
b
TNO Functional Ingredients Expertise Group, Zeist, The Netherlands

a r t i c l e i n f o a b s t r a c t

Article history: In puff pastry, fat and dough rheological behavior during sheeting control pastry dough development by
Received 16 March 2015 formation of the layered structure which is essential for product quality. The aim of this work was to
Received in revised form unravel the influence of fat and dough physical interactions during sheeting, as affected by variations in
20 June 2015
fat composition, to the rheological behavior, the microstructure and baking performance of pastry dough.
Accepted 10 September 2015
Available online xxx
Rheological data at small and large deformations were collected on bulk dough and pastry dough as
function of sheeting steps and fat composition. The resistance of fat to work related deformation, i.e. fat
consistency, and softening behavior during sheeting were also studied by means of a multi extrusion cell.
Keywords:
Saturated fatty acids
The physical interaction of bulk dough and bulk fat during sheeting promoted progressive structuring of
Puff pastry the pastry dough in a multi-layer system of solid layers alternating with soft gel layers. Structuring
Dough development resulted in increasing elastic-like behavior and consistency of the pastry dough. Pastry dough elastic-like
Fat rheology behavior and consistency increased with increasing fat consistency, until a maximum was reached. Such
maximum corresponded to an ideal fat-layered structure, as indicated by microscopic analysis, and
provided puff pastry with the best baking and textural quality. A further increase in fat consistency had
detrimental effect on pastry dough structure and baking performance.
© 2015 Elsevier Ltd. All rights reserved.

1. Introduction well as their position in the triglyceride molecule, the crystal

During the past decades, international dietary recommenda-
tions have focused on the reduction of fat intake in relation to
correlated diseases such as coronary heart disease. In response, the
food industry has been prompted to develop and launch commer-
cial products labeled as low-fat or fat-free (Hu et al., 2001). How-
ever, simply lowering the percentage of energy from fat intake is
considered unlikely to reduce incidence of coronary heart diseases.
In fact evidence from metabolic studies, epidemiologic in-
vestigations and clinical trials have indicated a link between the
type of dietary fats, i.e. saturated fatty acids (SAFA), and incidence
of coronary heart diseases (Hu and Willett, 2002). In view of these
findings, most recent nutritional guidelines have advised to keep
the daily intake of saturated fats below 10% energy (EFSA, 2010).
Puff pastry usually contains 30% or more fat (Baldwin et al.,
1972), of which about 50% in the form of SAFA (Simovic et al.,
2009; Yella Reddy and Jeyarani, 2001). Fatty acid composition as
* Corresponding author. TNO Functional Ingredients Expertise Group, Zeist, The
Netherlands.
E-mail address: stefano.renzetti@tno.nl (S. Renzetti).
structure, polymorphism, solid fat content (SFC), hardness and the
processing conditions determine the physical properties of the fat
(Garcia-Marcias et al., 2011; Marangoni and Narine, 2002; Campos
et al., 2002). In order to obtain puff pastry with good baking quality,
the key characteristics of the fat (i.e. margarine or shortening)
include: plasticity, consistency and melting properties (Brekke,
1980; Skogerson et al., 2006; Keogh and Morrissey, 1990). Puff
pastry dough is characterized by a layered structure of several
alternating layers of fat and dough resulting from a repeated
folding and sheeting process. The fat layered structure is essential
for the puff and flaky texture of the baked product (Baldwin et al.,
1972); it keeps the dough layers separate and traps the water vapor
during baking, leading to the final expanded structure. To achieve
an optimal layered structure, plasticity and consistency of the fat
are required to retain a continuous, unbroken layer during sheeting
(Yella Reddy and Jeyarani, 2001; Chrysam, 1985). It has been sug-
gested that fat for pastry application should have the same con-
sistency as the bulk dough across the temperature range which
includes the processing temperatures (Stauffer, 1994; Johnson,
1999). As the fat deviates from this consistency, sheeting quality
decreases (Stauffer, 1994).

http://dx.doi.org/10.1016/j.jfoodeng.2015.09.009
0260-8774/© 2015 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Renzetti, S., et al., Puff pastry with low saturated fat contents: The role of fat and dough physical interactions in
the development of a layered structure, Journal of Food Engineering (2015), http://dx.doi.org/10.1016/j.jfoodeng.2015.09.009
2 S. Renzetti et al. / Journal of Food Engineering xxx (2015) 1e9
Margarine and shortenings can be generally described as par-
ticle gels, consisting of a network of weakly bonded solid particles
associated with triacylglycerol fluid (Ghotra et al., 2002; Narine and
Marangoni, 1999a,b). Such structural organization is directly
responsible for the mechanical behavior (i.e. elasticity) of the fat
(Narine and Marangoni, 1999a,b; Pernetti et al., 2007). The appli-
cation of large deformations, i.e. sheeting, induces structural
changes to the particle network of fat, resulting in decreased
elasticity of the fat (Gonzalez-Gutierrez and Scanlon, 2013; Kloek
et al., 2005). Therefore, the product consistency during sheeting
changes from brittle to spreadable. Given the relation between the
particle crystal network and elasticity, fats for puff pastry applica-
tion are generally characterized in terms of SFC as a function of
temperature (Doerry, 1996; Stauffer, 1996). Despite important, lipid
composition cannot adequately predict the rheological properties
of fats (Narine and Marangoni, 1999a,b), which addresses the need
for understanding their behavior for specific processing
applications.
During rolling and sheeting, wheat flour dough is subjected to a
combination of uniaxial extension and shear at relatively high
deformation rates (MacRitchie, 1986; Morgenstern et al., 1999)
which determine the rheological properties of the dough at each
step in the process. Puff pastry quality (e.g. volume) is strongly
related to the rheological properties of the dough before baking. In
fact, the mechanical response of the dough at large deformations is
associated to the strain-hardening behavior of the gluten network
(van Vliet, 2008) and it is suggested to control the formation of thin
dough layers as required in puff pastry (Sliwinski et al., 2004).
With the increasing desire to reduce SAFA contents in bakery
products, such as puff pastry, the fundamental understanding of
the effect of variations in the characteristics of the fat on dough
development and final product quality becomes important. To the
best of our knowledge, there are no studies which address changes
in the rheological behavior of pastry dough during sheeting in
relation to variations in the rheological behavior of fat. Therefore,
the aim of this work was to relate the rheological behavior of fat
and pastry dough during processing to puff pastry quality, as
affected by variations in fat composition. In order to do so, small
and large deformation rheological tests were performed on fat and
pastry dough at different steps during the sheeting procedure. Bulk
dough, i.e. pastry dough without the layered fat, was also studied at
similar sheeting steps in order to gain insight of the contribution of
fat to the rheological behavior of the pastry dough. Light micro-
scopy of selected pastry dough samples was performed in order to
gain insights on the effects of fat variation on dough microstruc-
ture. Finally, baking tests were performed for each of the pastry
doughs varying in fat composition.
2. Materials and methods
2.1. Materials
Commercial wheat flour for pastry application (Meneba Meel
BV, Rotterdam, The Netherlands) was used in conjunction with
table salt and tap water. The flour had a moisture, protein and ash
content of 15.5%, 11.6% and 0.6%, respectively.
Palm stearin, palm oil and rapeseed oil were used for the
preparation of six different fat blends varying in SAFA content
(Table 1). All fats were purchased from Remia BV (Den Dolder, The
Netherlands).
2.2. Pilot production of shortenings
Fats were first melted and mixed in a jacketed tank at 70
C. Fat
blends were left to cool to 55
C and then transferred to a scraped
Please cite this article in press as: Renzetti, S., et al., Puff pastry with low saturated fat contents: The role of fat and dough physical interactions in
the development of a layered structure, Journal of Food Engineering (2015), http://dx.doi.org/10.1016/j.jfoodeng.2015.09.009
surface heat exchanger (VK 60/400-F, Gerstenberg Schro €der, Ger-
many) at 100 rpm. The pump was set at 12.8 bar and the flow rate
was 79.5 L/h. The temperature at the outlet was 20
C. Once pro-
cessed, fats were stored for 1 month at 20
C.
2.3. Fat pre-working
Prior to their use, all fats were conditioned at 18
C and then pre-
worked in a Kemper spiral mixer (Kemper SP15, Shelton, USA). Pre-
working was performed by mixing 2500 g of fat for 2 min at speed 2
of the mixer. The pre-worked fat was stored overnight at 18
C.
2.4. Bulk dough preparation
Bulk doughs were prepared and used in combination with the
different fat blends. The water amount to be added to the flour was
calculated taking into account flour moisture content and the far-
inograph water absorption. To 100 g of flour 10 g of fat were added
together with 1 g of salt. The ingredients were conditioned over-
night at 18
C. After addition of water (10
C), the ingredients were
mixed in a Kemper mixer (Kemper Spiral Mixer, Shelton, USA) for
1 min at low speed (speed I of the mixer). The speed was then
changed to a higher rate (speed II of the mixer) until the temper-
ature of the dough reached 18
C as measured by thermocouples
which were placed in the mixer. The dough was weighed (4500 g),
rounded and cooled at 5
C for 5 min. Afterwards, the bulk dough
was sheeted with a reversible Rondo sheeter (Rondostar Eco,
Rondo, Burgdorf, Switzerland) into a square shape of
600  600 mm and 8 mm thickness.
2.5. Pastry dough preparation
In order to prepare the pastry dough, the pre-worked bulk fat
from the different fat blends (2000 g) was sheeted with a reversible
Rondo sheeter (Rondostar Eco, Rondo, Burgdorf, Switzerland) into a
square shape of 400  400 mm and 13 mm thickness. The fat was
then conditioned overnight at 18
C. The next day the bulk dough
was prepared and the bulk fat was closed inside the bulk dough like
an envelope, and the edges sealed together (see supplementary
material for description of sheeting steps). The pastry dough was
sheeted to 10 mm thickness (pre-sheeting step). Then the pastry
dough was folded according to the French method (three folds),
turned 90
, sheeted to 10 mm and folding by the Dutch method
(four folds), which was followed by 30 min rest at 5
C (first
sheeting step). After resting, the dough was sheeted to 10 mm,
folded by the French method, then sheeted again to 10 mm and
folded by the Dutch method, which was followed by 30 min rest at
5
C (second sheeting step). Finally the pastry dough was sheeted to
12 mm, folded in two and sheeted to 6 mm (final sheeting step).
The puff pastry was then cut using a custom-made device resulting
in squares of 100 mm length and with an inner hole of 50  50 mm
length. The pastry dough was then rested for 60 min at 18
C.
2.6. Puff pastry baking
Puff pastry samples were baked in a Rothoterm oven (Werner &
Pfleiderer, Dinkelsbühl, Germany) in two steps: 21 min at 200
C
and 8 min at 160
C. After baking, the puff pastries were cooled for
60 min at 28
C and then stored overnight at 18
C in sealed plastic
bags prior to analyses. Baking was performed in duplicates for each
of the six fat variations, producing 12 puff pastries for each inde-
pendent duplicate.
 S. Renzetti et al. / Journal of Food Engineering xxx (2015) 1e9 3

Table 1
Fat blend composition and fat blend consistency (wrk1) and softening behavior (n1/2) at different stages in the process.

Fat blend Fat composition wrk1 (Nmm) n1/2

Palm stearine (%) Palm oil (%) Rapeseed oil (%) SAFAa (%) Unworked Pre-sheeting Final sheeting Unworked Pre-sheeting Final sheeting

F1 66.3 19.4 14.30 54.2 38108a,A 26275a,B 23081a,B 3.6a,B 6.0bc,A 5.3c,A
F2 62.9 21.6 15.5 53.1 32080b,A 23401b,B 19656a,B 3.49a,B 5.2b,A 5.3c,A
F3 58.9 23.3 17.8 51.5 25186c,A 19583c,B 19438a,B 3.55a,B 6.3bc,A 6.5b,A
F4 49.1 25.1 25.8 46.4 16155d,A 11786d,B 10140b,C 4.71a,C 8.2a,B 8.9a,A
F5 43.33 25.16 31.51 43.0 13736e,A 8084e,B 7365b,C 4.00a,B 7.5ac,A 8.1a,A
F6 36.9 24.7 38.4 38.9 9613f,A 5214f,B 5180b,B 4.23a,B 8.5a,A 9.0a,A

Lower case letter indicate significant differences among rows (p < 0.05).
Upper case letter indicate significant differences among columns (p < 0.05).
a
Estimated based on fat composition.

2.7. Rheology of pastry dough, bulk dough and bulk fat
2.7.1. Sheeting of bulk dough and bulk fat for rheological tests
In order to gain insight in the rheological behavior of the main
pastry dough components, i.e. the bulk dough and the bulk fat, at
the different steps in the sheeting process, bulk dough and bulk fat
were separately subjected to the same sheeting procedure as
applied to the pastry dough. In the final sheeting step, the final
thickness was kept at 10 mm in order to keep the same gap settings
for all rheological tests. All rheological tests were performed on
dough and fat samples taken at the end of the different steps
described in the sheeting procedure, i.e. pre-sheeting, first
sheeting, second sheeting and final sheeting.
2.7.2. Frequency sweep tests of pastry dough and bulk dough
Frequency sweeps were performed on the bulk and pastry
dough samples taken at the different stages in the sheeting pro-
cedure. The tests were performed on a stress controlled rheometer
(Anton Paar MCR 301, Ostfildern, Germany), using a cross-hatched
parallel plate geometry (25 mm diameter) with a gap of 10 mm. A
piece of sheeted dough was cut out with a diameter of 22 mm. Next,
the sample was gently compressed between the plates and the
open edge of the sample was coated with a thin layer of paraffin oil
using a brush in order to avoid evaporation. After loading, the
dough was rested for 5 min to allow relaxation of residual stresses.
Frequency sweeps were performed from 0.1 Hz until 10 Hz with a
strain of 105 which was well within the linear viscoelastic region
of the samples. Measurements of pastry and bulk dough samples
were performed in duplicates by preparing new independent
samples for each of the sheeting step.
2.7.3. Stress-relaxation tests of pastry dough and bulk dough
Stress-relaxation tests were performed on bulk and pastry
dough samples taken at the different stages in the sheeting pro-
cedure. The sample preparation follows the same procedure as
indicated in the frequency sweep tests. After loading, the dough
was first rested for 5 min without mechanical stress. A shear rate of
0.02 s1 was then applied for 50 s, resulting in a total strain of 1.
Afterwards, the strain was stopped and the stress was monitored
for 5 min. Temperature was kept constant at 18
C throughout the
tests. Measurements of pastry and bulk dough samples were per-
The cell had an inner length of 100 mm and an inner diameter of
25 mm. For extrusion, a probe was used consisting of a 6 holes disk
of 3 mm depth and with inner holes of 6 mm diameter. The MEC
was filled with 40 g fat previously equilibrated at 18
C and placed
in a water bath of 18
C. The plunger was moved up and down 50
times (i.e. 100 extrusions) at 10 mm/s. Time, force and distance
were recorded. For each extrusion the work per extrusion step
(Nmm/step) was calculated from the curve of force against dis-
tance. The obtained values were plotted as a function of the number
of extrusions, forming the breakdown profile of the bulk fat. Mea-
surements of bulk fat samples were performed in duplicates by
preparing new independent samples for each of the sheeting step.
2.8. Puff pastry baking quality evaluation
Standard baking tests were performed on each of the puff pastry
samples obtained from the six fat variations. Puff pastry volume
(mL) was measured with rapeseed displacement (n ¼ 8). Puff pastry
weight was recorded and used for the calculation of specific volume
(SV).
The height of the puff pastries was derived from the distance at
trigger force during textural analysis of pastries (n ¼ 144). The
thickness (mm) of the corresponding pastry dough was previously
measured with a caliper (n ¼ 4) in order to calculate puff pastry lift.
2.9. Texture analyses puff pastry
A puncture test was performed with a Texture Analyser XT2i
(Stable Micro Systems, Cardiff, United Kingdom), with a cylindrical
probe of 3 mm and a load cell of 50 N. The penetration speed during
the test was 5 mm/s with a maximum penetration depth of 90%
(trigger force 0.1 N). 6 puff pastry samples were measured for each
fat variation in duplicate, with 12 penetrations for each sample
(n ¼ 144). From the penetration profile the force peaks for pene-
tration through the puff pastry layered structure are calculated
along with the distance between peaks.
The force number distribution was obtained using a log normal
distribution function:
!2 


lnðFavg ÞlnðFmax Þ

vNðFÞ Fsig

formed in duplicates by preparing new independent samples for ¼ NFmax e (1)

each of the sheeting step.
2.7.4. Multi-extrusion cell (MEC) tests on fat
MEC tests were performed on the bulk fats at the different steps
in the sheeting procedure in order to measure changes in their
consistency profile induced by mechanical “work”. The multi
extrusion cell was connected to a Texture Analyser HD (Stable
Micro Systems, Cardiff, United Kingdom) with a load cell of 250 N.
vðlnðFÞÞ
where NFmax is the maximum of the distribution (Force amplitude),
Fmax is the force with maximum counts and Fsig is the width of the
force distribution. The force number distribution was calculated as
the number of force peaks (nF) as a function of the force F, assuming
vNðFÞi =vlnðFi ÞzvNðFÞ=vlnðFÞ where Fi stands for the individual
force peaks from the puncture test. From the force number distri-
bution the average force (Favg) needed to fracture the puff pastry

Please cite this article in press as: Renzetti, S., et al., Puff pastry with low saturated fat contents: The role of fat and dough physical interactions in
the development of a layered structure, Journal of Food Engineering (2015), http://dx.doi.org/10.1016/j.jfoodeng.2015.09.009
4 S. Renzetti et al. / Journal of Food Engineering xxx (2015) 1e9
layers was calculated.
2.10. Light microscopy of pastry dough
Pastry dough samples were cut to slices of 10 mm using a
cryostat microtome (Microm HM 550, Adamas, The Netherlands)
set at a temperature of 35
C. In order to stain fat a stock solution
of Oil Red O (0.5%) in isopropanol was prepared. A staining solution
was then prepared by diluting 30 mL of the stock solution with
20 mL distilled water, allowing to stand for 10 min and then
filtering. Samples were then stained for 10 min and then rinsed
with 30% ethanol solution. Pictures of microstructure were taken
with a DP 71 camera attached to an Olympus BX-61 microscope
(Olympus, Germany).
2.11. Statistical analysis
Statistical analyses were performed with SPSS Statistics (IBM,
version 22, US) on all data using one-way ANOVA. Fisher's least
significant differences test was used to describe means at 5% sig-
nificance level.
3. Results and discussion
3.1. Physical properties of fat blends: MEC tests at various stages
during processing
Fat blends were produced with different ratios of palm stearin,
palm oil and rapeseed oil in order to obtain variation in SAFA
content (Table 1). As a result of these variations, differences in the
consistency of the fat blends were expected. The rheological
behavior of the fat blends prior to their use in puff pastry applica-
tion was assessed by MEC tests. The tests provide a profile of the fat
consistency, i.e. work, per extrusion step (Fig. 1A). Mainly, two
Fig. 1. Characteristic softening profile of the six unworked fat blends used in this study
(A). Softening profile of fat blend F1 after pre-working and at different sheeting steps (B).
Please cite this article in press as: Renzetti, S., et al., Puff pastry with low saturated fat contents: The role of fat and dough physical interactions in
the development of a layered structure, Journal of Food Engineering (2015), http://dx.doi.org/10.1016/j.jfoodeng.2015.09.009
parameters were derived from the tests: the initial consistency of
the fat, i.e. the work at step 1 of the extrusion (wrk1), and the
number of extrusion steps required to reach half of the residual
consistency (n1/2). n1/2 provides information regarding the soft-
ening profile of the fat during processing. As expected, the initial
consistency of the fat blends (unworked wrk1) significantly
decreased with decreasing SAFA content (Table 1). The initial con-
sistency of the unworked fat (wrk1) was a power law function of
the SAFA content (R2 ¼ 0.99), as previously observed in a number of
studies (Narine and Marangoni, 1999a,b; Vreeker et al., 1992). All fat
blends showed a softening behavior with increasing number of
extrusion steps, with a rather sharp decrease in consistency during
the initial extrusion steps while reaching a similar plateau with
higher number of extrusions (Fig. 1A). The softening behavior was
similar for all the unworked fat blends, as indicated by the un-
worked n1/2 (Table 1).
All fat blends required pre-working in order to be properly
sheeted in sheets of 400  400 mm with 13 mm thickness and be
used in the pastry dough preparation. During the sheeting process,
the consistency of the fat blends changed compared to the un-
worked fat blend. A typical profile of the softening behavior of the
fat blends measured at the different stages during the sheeting
procedure is shown in Fig. 1B (fat blend F1). After pre-sheeting, the
initial consistency of fat blend F1 (pre-sheeting wrk1) was signifi-
cantly lower than the unworked fat (Table 1). For all fats a decrease
in wrk 1 was observed going from pre-sheeting to final sheeting
(Table 1), which was significant for fat blends F4 and F5. Sheeting
induces breakages in the inter-particle linkages of the fat crystal
network (Tang and Marangoni, 2007), thus resulting in a decrease
consistency of the fat with progressive sheeting (Heertje, 1993).
When comparing the different fat blends at similar sheeting
steps, a significant decrease in wrk 1 was observed with decreasing
SAFA content after pre-sheeting (Table 1). At final sheeting, a
similar dependency of wrk 1 with SAFA content was still observed.
Wrk 1 of fat blends F1, F2 and F3 was significantly higher than those
of F4, F5 and F6 (Table 1).
For all fats, the softening profile (n1/2) at pre-sheeting and final
sheeting was significantly smoother than the unworked fats
(Table 1). On the contrary, no changes in softening profile were
observed between pre-sheeting and final sheeting, with the
exception of fat F4 (Table 1). At both pre-sheeting and final
sheeting, significant differences in softening profile n1/2 were
observed among the fat blends (Table 1). In general, the softening
behavior became sharper with increasing SAFA content.
Consistency and work softening behavior are important rheo-
logical properties for pastry dough applications, as the plasticity of
the fat largely contributes to the formation of an homogeneous
layered structure (Ronholt et al., 2014). The variations obtained
within the fat blends of this study are hence expected to result in
considerable differences in the microstructure and baking quality
of the pastry dough.
3.2. Frequency sweep tests of bulk and pastry dough
Independent of the fat blend used, all pastry dough showed that
the elastic modulus (G0 ) was higher than the viscous modulus (G00 ),
indicating that the pastry dough behaved as an elastic solid. All
pastry dough showed a progressive increase in complex modulus
(G*) during sheeting, regardless of the frequency, indicating an
increased resistance to deformation (Fig. 2A). Additionally, the tan
(d) showed to decrease during sheeting, indicating increased
elastic, solid-like behavior (Fig. 2B). On the contrary, no changes in
G* and in tan (d) were observed with progressive sheeting of the
bulk dough (data not shown). The frequency sweep curves for the
G0 could be all fitted with a power law equation (R2 ¼ 0.99):
 S. Renzetti et al. / Journal of Food Engineering xxx (2015) 1e9 5

Fig. 2. G* of pastry dough produced with fat blend F1 at different steps during sheeting (A). Tan (d) of pastry dough produced with fat blend F1 (B). Power law index of pastry dough
produced with the different fat blends and of bulk dough with fat blend F3 derived from the frequency sweep test at different points in the sheeting process (C); different letters
within the same sample indicate significant differences (p < 0.05). G0 (at 1 Hz) of pastry dough from different fat blends obtained from the frequency sweep test as a function of fat
consistency at different points in the sheeting process (D); the G0 (at 1 Hz) at final sheeting of the bulk dough with fat blend F3 is also plotted; different letters indicate significant
differences for the last sheeting step (p < 0.05).

G* ¼ Kðf Þn
where G* is the complex moduli, K is a constant, f is the frequency
and n is the power law index. The power law index provides in-
formation on the frequency dependence of the visco-elastic
behavior (Leroy et al., 2010). In general, a decrease in power law
index was observed during progressive sheeting. For all pastry
dough, a significant reduction was observed at final sheeting as
compared to the pre-sheeting step (Fig. 2C). On the contrary, no
significant changes could be observed on the bulk dough (Fig. 2C). A
decrease in the power law index indicates that the viscoelastic
behavior of the material is becoming less dependent on the applied
frequency and it is generally related to the formation of an
extended network, i.e. structuring, within the system (Letang et al.,
1999). This finding is in agreement with the observation that tan(d)
decreases during sheeting.
The G0 of the different pastry dough was plotted as a function of
the initial fat consistency (wrk1) at each of the different steps
during sheeting (Fig. 2D). For each of the plotted G0 values, the work
softening related fat consistency (wrk1) is taken into account by
taking its value according to the corresponding sheeting step (data
partially shown in Table 1). A progressive increase in G0 is observed
with increasing fat consistency for the pre-sheeting and first
sheeting steps and, with the exception of F1, also for the second
sheeting step (Fig. 2D). It should be noted that the G0 of the bulk
dough, which was sheeted without the bulk fat, is considerably
lower (by four fold) than that of the pastry dough (Fig. 2D). In view
of these results, it can be suggested that (i) fat has a relevant
contribution to the elasticity of the pastry dough; and that (ii) the
specific G0 of the puff pastry dough depends on the consistency and
work softening behavior of the fat used.
At final sheeting, G0 becomes independent of the type of fat used
for fat consistencies larger than 10,000 Nmm (p < 0.05), i.e. F1, F2,
F3 and F4, despite the significant differences that still exist in the
consistency, i.e. wrk1, of the bulk fats at this last sheeting step
(Table 1). During sheeting and repeated folding, the characteristic
layered structure of alternating fat and dough layers is obtained in
the pastry dough. At the beginning of the sheeting process (i.e. pre-
sheeting, first sheeting and second sheeting), the low number of fat
layers present in the pastry dough (i.e. 1, 3 and 36, respectively) can
explain the distinct contribution of the fat type to the pastry dough
rheology, also in view of the low G0 of the bulk dough. At the final
sheeting step, a high number of fat layers is achieved, i.e. 288,
which result in a higher dispersion of fat in the pastry dough. In this
multi-layered structure it can be suggested that the fat behaves as a
solid compared to the bulk dough phase which can be instead
described as a soft gel (Leroy et al., 2010). The layered fat provides a
spanning structure which imbues the pastry dough with its solid-
like behavior, thus resulting in similar pastry dough rheology
independently of the fat used. The structuring role of the layered fat
is confirmed by the progressive increase in the elastic-like behavior
of the pastry dough during sheeting, irrespective of the type of fat
used.
3.3. Rheological behavior of pastry dough and bulk dough at large
deformations
Stress-relaxation tests were performed on pastry dough with
different fat blends in order to assess changes in dough strength
and elasticity for large deformations at the different steps in the
sheeting process. Stress-relaxation tests are considered a viable
mean to evaluate differences in the molecular structuring of a
polymeric matrix (Rao et al., 2000) and hence could provide in-
formation on structure development during sheeting.
Dough consistency at the targeted strain of 1, i.e. tmax, increased
for all pastry dough during the sheeting process (Fig. 3A). At the end
of the final sheeting step, the highest consistency was observed for
the pastry dough with fat blend F3. Above and below the

Please cite this article in press as: Renzetti, S., et al., Puff pastry with low saturated fat contents: The role of fat and dough physical interactions in
the development of a layered structure, Journal of Food Engineering (2015), http://dx.doi.org/10.1016/j.jfoodeng.2015.09.009
6 S. Renzetti et al. / Journal of Food Engineering xxx (2015) 1e9
Fig. 3. Pasrty dough consistency tmax with different fat blends plotted as a function of
fat consistency at different points in the sheeting process (A); different letters indicate
significant differences for the last sheeting step (p < 0.05). Relation between tmax and
fat rheology (i.e. wrk1∙n1/2) as measured for the last sheeting step (B).
consistency of fat blend F3, pastry dough consistency gradually
decreased, thus suggesting an optimum. On the contrary, the bulk
dough obtained with fat blend F3 and which was subjected to the
same sheeting procedure as the pastry dough (see section 2.7.1), did
not show any increase in consistency (data not shown). Similar
results were obtained for the other bulk dough, independently
from the type of fat used (data not shown).
During sheeting, energy (expressed as unit of work per mass or
volume) is transferred into the dough, affecting its quality and
baking performance (Levine et al., 2002). Pastry dough is composed
of two bulk phases, i.e. the bulk dough and the bulk fat, which are
structured in a progressive number of alternating layers during
sheeting. When mechanical deformations are applied to a fat
layered system, the local stresses in the layered bulk dough phase
increase with increasing layers of the bulk fat as well as with
increasing consistency of the bulk fat phase. Compared to bread
dough, the bulk dough for pastry applications is less elastic and
ruptures at smaller deformations (Dobraszczyk and Morgenstern,
2013). At the beginning of the sheeting operations, the dough is
undeveloped and is sensitive to piercing of the fat through the
dough layer, with detrimental effects on the formation of a layered
structure. The sheeting operations induce dough development (i.e.
the observed increase in tmax), with strain-hardening properties
being enhanced by the orientation of the gluten with progressive
sheeting (van Vliet, 2008; Dobraszczyk and Morgenstern, 2013).
The maximum development is achieved at final sheeting (Fig. 3A).
On the contrary, sheeting of the bulk dough in absence of the bulk
fat does not provide enough deformation for proper dough devel-
opment. Based on these considerations, it could be suggested that
the optimum observed for intermediate fat consistency, corre-
sponds to an ideally layered fat structure. Such structure is the
result of an optimum in the interplay between strain-hardening
behavior with progressive sheeting and fat consistency and work
softening behavior. Above a certain fat consistency, detrimental
effects on the layered structure may occur, due to excessive
deformation and shear destruction of the bulk dough.
Please cite this article in press as: Renzetti, S., et al., Puff pastry with low saturated fat contents: The role of fat and dough physical interactions in
the development of a layered structure, Journal of Food Engineering (2015), http://dx.doi.org/10.1016/j.jfoodeng.2015.09.009
Variation in pastry dough rheological properties was obtained
exclusively by variations in the type of fat used. According to the
methodology used in the current study, the overall rheological
behavior of the fat is provided by its initial consistency at each
sheeting step (wrk1) and by the softening behavior during pro-
gressive working (n1/2). For this reason, pastry dough consistency
tmax at final sheeting was plotted as function of the interaction
between wrk1 and n1/2 at final sheeting (Fig. 3B), showing a strong
correlation among these parameters. Similar strong correlations
were also found for dough consistency tmax and fat rheology (i.e.
wrk1∙n1/2) at pre-sheeting, first and second sheeting. These results
strongly suggest that, within the conditions of this study, final
dough consistency tmax is modulated by the interplay of fat con-
sistency (wrk1) and fat softening behavior (n1/2) during sheeting.
3.4. Microscopy of pastry dough
Microscopic analysis of pastry dough with fat staining was
performed in order to gain insight in the layered structure of the
dough at the end of the sheeting process. The microstructure of
pastry dough with fat blends F1, F3, F4 and F6 is displayed in Fig. 4.
For fat blend F6, a large variation in the thickness of the fat layers
could be observed. Several discontinuities in the layered structure
were also present due to breakages in the fat layers and/or merging
between different layers. Furthermore, the thickness of the bulk
dough layers seems quite large when compared to that of the fat
layers. Despite some variation in the thickness of the fat layers,
pastry dough with fat blend F4 showed a rather continuous layered
structure with a much clearer alternation of dough and fat layers
compared to fat blend F6. The thickness of fat and bulk dough layers
appeared to be considerably more homogeneous than those
observed with fat blend F6. Similarly, a continuous layered struc-
ture was obtained also with fat F3, despite the presence of some fat
globules which were not fully sheeted. Further increasing fat con-
sistency, i.e. F1, resulted in a deterioration of the continuous layered
structured observed for fat blend F4 and F3. Although the conti-
nuity of the layered structure appeared to be better retained with
fat blend F1 than with fat blend F6, considerable variation in
thickness of the fat layers could be observed. In particular, the
presence of large fat globules at several points along one fat layer
was noticeable, thus suggesting that the fat was too hard to obtain
homogeneous fat layers. Furthermore, the bulk dough layers
appeared often considerably thin, thus suggesting excessive
deformation and shear destruction of the dough as a result of
increased fat consistency.
Taken together, the rheological and microcopy results indicate
that the optimum observed in pastry dough development (i.e. tmax)
for intermediate fat consistency, is related to an ideal layered-fat
structure at final sheeting. A further increase in fat consistency
does not result in further increase in elastic-like behavior of the
pastry dough and is associated with a loss of homogeneity in the
layering of both fat and bulk dough. Based on these considerations,
it can be expected that the baking and textural quality of puff pastry
is optimal in the ranges of fat consistency corresponding to F3 and
F4.
3.5. Baking and textural quality of puff pastries
Baking quality of puff pastries was assessed by measuring
maximum average height, lift, volume and specific volume (SV).
Puff pastry SV was a function of the initial fat consistency, i.e. wrk1
at pre-sheeting, with a maximum SV observed between 8000 and
20,000 Nmm (F5 and F3 fat blend, respectively) (Fig. 5A). Outside
this range a significant decrease in SV was observed, going to either
higher (>20,000 Nmm) or lower (<8000 Nmm) fat consistencies.
 S. Renzetti et al. / Journal of Food Engineering xxx (2015) 1e9 7

Fig. 4. Microstructure of pastry dough with fat blend F6 (A,E), F4 (B,F), F3 (C, G) and F1 (D,H); cross-section of puff pastry obtained with fat blends F6 (I), F4 (j), F3 (k) and F1 (l).

Similar trends could be observed also for the other baking pa-
rameters (data not shown). Within the condition of this study, the
observed trends in baking quality as function of fat consistency
suggests an optimum for wrk1 at pre-sheeting around
15,000 Nmm. That is intermediate to the consistencies of fat blends
F4 and F3. Given the previously observed trends in pastry dough
microstructure and pastry dough tmax as function of fat consis-
tency, these results confirm a relation between baking performance
and optimal structuring of the fat in homogeneous layers. On the
other hand, the relationship between tmax and any of the baking
quality parameters was not as good as that observed with fat
consistency (data not shown). It should be taken into account that
strain-hardening behavior in large deformations such as bubble
expansion is of key importance for baking performance (van Vliet
et al., 1992; Dobraszczyk et al., 2003; van Vliet, 2008). Strain-
hardening allows the expanding gas cell walls to resist failure by
locally increasing resistance to extension as the bubble walls
become thinner, resulting in better gas retention. Unfortunately,
evaluation of strain-hardening by biaxial deformation in
compression tests could not be successfully performed by the au-
thors, likely due to the layered structure in the samples. It could be
suggested that a combination of strain-hardening behavior and
homogeneity of the multi-layered pastry dough structure, may
explain the increase in baking performance with increasing fat
consistency until an optimum is reached with fat blends F3 and F4.
With higher fat consistencies, i.e. F1 and F2, baking quality

Fig. 5. SV of puff pastry obtained from different fat blends as a function of fat consistency at pre-sheeting (A). Number of force peaks (nF) obtained from puncture tests on puff
pastry as a function of fat consistency at pre-sheeting (B). Average force (Favg) for puncturing of puff pastry as a function of fat consistency at pre-sheeting (C). For all graphs the type
of fat used is indicated by the labels. Different letters indicate significant differences between samples (p < 0.05).

Please cite this article in press as: Renzetti, S., et al., Puff pastry with low saturated fat contents: The role of fat and dough physical interactions in
the development of a layered structure, Journal of Food Engineering (2015), http://dx.doi.org/10.1016/j.jfoodeng.2015.09.009
8 S. Renzetti et al. / Journal of Food Engineering xxx (2015) 1e9
decreases to due excessive deformation of the bulk dough layers
resulting in (i) detrimental effects on the homogeneity of the multi-
layered structure and (ii) shear destruction of the gluten network
thus affecting the strain-hardening behavior. Additionally, it should
be considered that the physical properties of the fat during the
baking process, i.e. melting profile, were likely to contribute to the
final baking quality.
Two main parameters were derived from the puncture tests to
describe puff pastry texture: the number of force peaks >0.1 N (nF)
and the average puncturing force (Favg). nF significantly increased
with increasing fat consistency until reaching a maximum with fat
blend F3 (Fig. 5B). After that, a further increase in fat consistency
resulted in a significant decrease in nF (Fig. 5B). The relationship of
nF with fat consistency suggests an optimum around 15,000 Nmm
of fat consistency as observed for the baking quality parameters. A
high number of force peaks indicates a high number of dough
lamellae in the puff pastry, thus confirming proper expansion of the
layered structure during baking. Favg showed a minimum for fat
consistency between 8000 and 20,000 Nmm (F5 and F3 fat blend,
respectively), with a significant increase above and below those
ranges (Fig. 5C). Taken together, nF and Favg can describe the flaky
texture of the puff pastry. Overall, these results indicate that the
textural quality of puff pastry is optimal when using fat with an
initial consistency, i.e. wrk1 at pre-sheeting, between 10,000 and
20,000 Nmm, which is intermediate to the fat blends F4 and F3.
These are the fat blends for which pastry dough showed an ho-
mogeneous multi-layered structure (Fig. 4), thus confirming
the relationship between pastry dough structure and baking
quality.
4. Conclusion
The baking quality of the puff pastry dough is strongly associ-
ated with formation of a homogeneous structure of alternating fat
and dough layers. In fact, the highest pastry SV, lift, expansion of
the layered structure (nF) and flaky texture (high nF combined with
low Favg) were observed for those pastry dough in which the mi-
croscopy investigations showed the best layered structure.
Small and large deformation rheological tests were able to
monitor the development of the layered structure during sheeting,
showing an increase in elastic-like behavior and consistency (G0 ,
power law index and tmax) of the pastry dough. On the contrary, the
corresponding bulk dough did not show any change in rheological
behavior. These results clearly pointed to a structuring role of the
fat in the pastry dough: the progressive sheeting of the pastry
dough increases layering of the fat in the pastry dough, resulting in
a progressive increase of elastic-like behavior and pastry dough
consistency. Despite the softening behavior of the fat during
sheeting, the fat layers behave as a solid in comparison to the bulk
dough.
Within the conditions of this study, an optimum in pastry dough
development and baking quality was achieved for intermediate fat
consistency, i.e. 8000e20,000 Nmm (wrk 1). Above and below such
ranges, detrimental effects on pastry quality were observed. In
particular, excessive fat consistency induced excessive deformation
of the bulk dough, explaining the detrimental effects on pastry
dough development (tmax) and microstructure.
The observed variations in pastry dough development and
baking performance were modulated by the type of fat used, and
particularly by the interplay of initial fat consistency (wrk1) and fat
softening during working (n1/2). Therefore, baking quality of puff
pastry with low SAFA contents can be modulated by variations in
those two parameters.
Please cite this article in press as: Renzetti, S., et al., Puff pastry with low saturated fat contents: The role of fat and dough physical interactions in
the development of a layered structure, Journal of Food Engineering (2015), http://dx.doi.org/10.1016/j.jfoodeng.2015.09.009
Acknowledgments
The authors would like to thank their TNO colleagues Cees
Heddes and Therese Maarschalkerweerd for their technical assis-
tance with puff pastry production and dough microscopy,
respectively.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.jfoodeng.2015.09.009.
References
Baldwin, R.R., Baldry, R.P., Johansen, R.G., 1972. Fat systems for bakery products.
J. Am. Oil Chem. Soc. 49, 473e477.
Brekke, O.L., 1980. Soybean Oil Food Products: Their Preparation and Uses (In D.R.
Erickson).
Campos, R., Narine, S.S., Marangoni, A.G., 2002. Effect of cooling rate on the
structure and mechanical properties of milk fat and lard. Food Res. Int. 35,
971e981.
Chrysam, M.M., 1985. Table spreads and shortenings. In: Applewhite, T.H. (Ed.),
Bailey's Industrial Oil and Fat Products, vol. 3. John Wiley and Sons, New York,
pp. 41e125.
Dobraszczyk, B.J., Morgenstern, M.P., 2013. Rheology and the breadmaking process.
J. Cereal Sci. 38, 229e245.
Dobraszczyk, B.J., Smewing, J., Albertini, M., Maesmans, G., Schofield, J.D., 2003.
Extensional rheology and stability of gas cell walls in bread doughs at elevated
temperatures in relation to breadmaking performance. Cereal Chem. 80,
218e224.
Doerry, W.T., 1996. Laminated Dough Systems. American Baking Institute, Man-
hattan, Kansas, pp. 63e68.
EFSA Panel on Dietetic Products, Nutrition, Allergies (NDA), 2010. Scientific opinion
on dietary reference values for fats, including saturated fatty acids, poly-
unsaturated fatty acids, monounsaturated fatty acids, trans fatty acids, and
cholesterol. EFSA J. 8, 1461e1568.
Garcia-Marcias, P., Gordon, M.H., Frazier, R.A., Smith, K., Gambelli, L., 2011. Perfor-
mance of palm-based fat blends with a low saturated fat content in puff pastry.
Eur. J. Lipid Sci. Technol. 113, 1474e1480.
Ghotra, B.S., Dyal, S.D., Narine, S.S., 2002. Lipid shortenings: a review. Food Res. Int.
35, 1015e1048.
Gonzalez-Gutierrez, J., Scanlon, M.G., 2013. Strain dependence of the uniaxial
compression response of vegetable shortening. J. Am. Oil Chem. Soc. 90,
1319e1326.
Heertje, I., 1993. Microstructure studies in fat research. Food Microstruct. 12, 77e94.
Hu, F.B., Willett, W.C., 2002. Optimal diets for prevention of coronary heart disease.
Clin. Cardiol. 288, 2569e2578.
Hu, F.B., Manson, J.E., Willett, W.C., 2001. Types of dietary and risk of coronary heart
disease: a critical review. J. Am. Coll. Nutr. 2, 5e19.
Johnson, B., 1999. Bakery margarines. Cereal Foods World 9e13.
Keogh, M.K., Morrissey, A., 1990. Anhydrous milk fat. 6. Baked goods. Ir. J. Food Sci.
Technol. 14, 69e83.
Kloek, W., van Vliet, T., Walstra, P., 2005. Large deformation behaviour of fat crystal
networks. J. Texture Stud. 36, 516e543.
Leroy, V., Pitura, K.M., Scanlon, M.G., Page, J.H., 2010. The complex shear modulus of
dough over a wide frequency range. J. Non Newton. Fluid Mech. 165, 475e478.
Letang, C., Piau, M., Verdier, C., 1999. Characterization of wheat flour water doughs.
Part I: rheometry and microstructure. J. Food Eng. 41, 121e132.
Levine, L., Corvalan, C.M., Campanella, O.H., Okos, R.A., 2002. Model describing the
two-dimensional calendering of finite width sheets. Chem. Eng. Sci. 57,
643e650.
MacRitchie, F., 1986. Physicochemical processes in mixing. In: Blanshard, J.M.V.,
Frazier, P.J., Galliard, T. (Eds.), Chemistry and Physics of Baking: Materials,
Processes and Products. Royal Society of Chemistry, London.
Marangoni, A.G., Narine, S.S., 2002. Identifying key structural indicators of me-
chanical strength in networks of fat crystals. Food Res. Int. 35, 957e969.
Morgenstern, M.P., Zheng, H., Roos, M., Campanella, O.H., 1999. Rheological prop-
erties of sheeted wheat flour dough measured with large deformations. Int. J.
Food Prop. 2, 265e275.
Narine, S.S., Marangoni, A.G., 1999a. Fractal nature of fat crystal networks. Phys. Rev.
E 59, 1908e1920.
Narine, S.S., Marangoni, A.G., 1999b. Relating structure of fat crystal networksto
mechanical properties: a review. Food Res. Int. 32, 227e248.
€ ter, E., Bot, A., 2007. Structuring of edible oils by
Pernetti, M., van Malssen, K.F., Flo
alternatives to crystalline fat. Curr. Opin. Colloid Interface Sci. 12, 221e231.
Rao, V.K., Mulvaney, S.J., Dexter, J.E., 2000. Rheological characterization of long- and
short- mixing flours based on stresserelaxation. J. Cereal Sci. 31, 159e171.
Ronholt, S., Kirkensgaard, J.J.K., Hoyer, K.F., Mortensen, K., Knudsen, J.C., 2014. The
effect of capacity, rotational speed and storage on crystallization and rheolog-
ical properties of puff pastry butter. J. Am. Oil Chem. Soc. 91 (1), 29e38.
 S. Renzetti et al. / Journal of Food Engineering xxx (2015) 1e9 9

Simovic, D.S., Pajin, B., Seres, Z., Filipovic, N., 2009. Effect of low-trans margarine on
physicochemical and sensory properties of puff pastry. Int. J. Food Sci. Technol.
44, 1235e1244.
Skogerson, L., Boutte, J., Robertson, J., Yhang, F., 2006. Non-hydrogenated Vegetable
Oil Based Margarine for Puff-pastry Containing an Elevated Diglyceride Emul-
sifier. US Patent 20070148313.
Sliwinski, E.L., Kolster, P., van Vliet, T., 2004. On the relationship between large-
deformation properties of wheat flour dough and baking quality. J. Cereal Sci.
39, 231e245.
Stauffer, C.E., 1994. Fat & Oil Functions. Baking & Snack. January, 15e19.
Stauffer, C.E., 1996. Fats and Oils. In: Eagan Press Handbook Series (St. Paul,
Minnesota).
Tang, D., Marangoni, A.G., 2007. Modeling the rheological properties and structure
of colloidal fat crystal networks. Trends Food Sci. Technol. 18, 474e483.
van Vliet, T., 2008. Strain hardening as an indicator of bread-making performance: a
review with discussion. J. Cereal Sci. 48, 1e9.
van Vliet, T., Janssen, A.M., Bloksma, A.H., Walstra, P., 1992. Strain hardening of
dough as a requirement for gas retention. J. Texture Stud. 23, 439e460.
Vreeker, R., Hoekstra, L.L., den Boer, D.C., Agterof, W.G.M., 1992. The fractal nature of
fat crystal networks. Colloids Surfaces 65, 185e189.
Yella Reddy, S., Jeyarani, T., 2001. Trans-free bakery shortenings from mango kernel
and mahua fats by fractionation and blending. J. Am. Oil Chem. Soc. 78,
635e640.

Please cite this article in press as: Renzetti, S., et al., Puff pastry with low saturated fat contents: The role of fat and dough physical interactions in
the development of a layered structure, Journal of Food Engineering (2015), http://dx.doi.org/10.1016/j.jfoodeng.2015.09.009