Materials and Design 76 (2015) 7785

Contents lists available at ScienceDirect

Materials and Design

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

Bio-based polyurethane foams toward applications beyond thermal

insulation
Nuno V. Gama a, Belinda Soares a, Carmen S.R. Freire a, Rui Silva b, Carlos P. Neto a, Ana Barros-Timmons a,,
Artur Ferreira a,c
a
CICECO, Aveiro Institute of Materials and Department of Chemistry, University of Aveiro, Campus Santiago, 3810-193 Aveiro, Portugal
b
Sapec-Qumica SA, Zona Industrial de Ovar Lote 18, 3880 Ovar, Portugal
c
Escola Superior de Tecnologia e Gesto de gueda, Rua Comandante Pinho e Freitas, n 28, 3750-127 gueda, Portugal

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

Article history: In this work the preparation of viscoelastic bio-based polyurethane foams (PUFs) using polyols obtained
Received 12 November 2014 via acid liquefaction of coffee grounds wastes has been optimized. In a rst stage, the effect of different
Revised 8 March 2015 ratios of isocyanate content to hydroxyl number (0.6, 0.7 and 0.8) and of three distinct percentages of
Accepted 19 March 2015
catalyst (3%, 5% and 7%) on the extent of the polymerization reaction was studied by infrared spec-
Available online 19 March 2015
troscopy. Next, different percentages of surfactant (14%, 16% and 18%) and blowing agent (12%, 14%
and 16%) were used to assess their effect on the density, thermal conductivity and mechanical properties
Keywords:
of the foams, including their recovery time. The mechanical properties of the ensuing foams proved to be
Polyurethane foams
Viscoelastic behavior
very interesting due to their viscoelastic behavior. PUFs were also characterized by scanning electron
Liqueed coffee grounds microscopy (SEM) revealing a typical cellular structure and by thermogravimetric analysis (TGA) which
proved that these materials are thermally stable up to 190 C. These results suggest other potential appli-
cations for these materials beyond heat insulation in areas where damping properties can be an added
value.
2015 Elsevier Ltd. All rights reserved.

1. Introduction insulation consisting in a signicant contribution to energy man-

Polyurethane foams (PUFs) are a class of lightweight porous
materials with enormous interest because of their specic proper-
ties and potential application in several elds [1]. PUFs can be clas-
sied into two major categories: exible and rigid foams. In some
classications PUFs are subdivided in semi-rigid and semi-exible
foams [1]. Moreover, some exible foams can also be classied as
viscoelastic when they show a delayed recovery. In fact, other
expressions like, visco-hyperelastic, memory, slow recovery, con-
trolled recovery or low resilience foams are also used [24].
Besides applications in thermal insulation, the range of applica-
tions of viscoelastic PUFs includes acoustic absorbing materials
for noise and vibration control due to their great potential for the
damping of mechanical vibrations [5]. Therefore, they are often
used to reduce vibrations and harshness and consequently increase
comfort [6]. For these reasons, viscoelastic PUFs are commonly
used in seats in automobile and aircraft industries. Furthermore,
depending on their viscoelastic properties PUFs may also be used
as panels in buildings for heat [7], sound [8] and vibration [9]
Corresponding author.
E-mail address: anabarros@ua.pt (A. Barros-Timmons).
agement, as well as helping to reduce noise pollution. Yet, the
use of these porous materials to hamper vibrations requires speci-
c viscoelastic characteristics when they are bound onto the
vibrating structure [10].
Currently, the polyurethane (PU) industry is still heavily petro-
leum-dependent because its two major feedstocks, polyols and iso-
cyanates derive entirely from it. However, due to the uncertainty
about the cost of petroleum in the future, as well as the need to
move toward more environmentally friendly feedstocks, many
recent efforts have been focused on replacing all or a portion of
the conventional petroleum-based polyols by counterparts
obtained from renewable resources. In fact, many non-petroleum
resources, as well as different processes of production of non-pet-
roleum derived polyols have been used in PUs production.
Processes such as oxypropylation [11] or acid liquefaction [12] of
several biomass residues, as well as the modication of vegetable
oils following different strategies [13] have been used to produce
renewable polyols for the PUs industry. The acid liquefaction of
biomass resources, like starch [14], soybean [15], alginic acid
[16], palm [17], sugar-cane bagasse [18], lignin [19], cork [20] or
coffee grounds [21], to obtain products which can be used as the
polyol component in the production of PUFs is a particular

http://dx.doi.org/10.1016/j.matdes.2015.03.032
0261-3069/ 2015 Elsevier Ltd. All rights reserved.
78 N.V. Gama et al. / Materials and Design 76 (2015) 7785
interesting strategy. This is due to the fact that it does not involve
the use of a large amount of harmful solvents or reagents nor pres-
sure thus, it presents a relatively low environment impact.
However, the majority of these studies are generally associated
with the production of rigid PUFs as a result of the functionality
of the ensuing polyols.
Due to their unique taste and avor, coffee brews are among the
most consumed beverages in the world and have grown steadily in
commercial importance during the last 150 years. With an annual
worldwide production of approximately 120 million tons, large
quantities of residues are generated [22]. The spent coffee grounds,
the solid residues obtained from the treatment of coffee powder
with hot water to prepare instant coffee, are the main industrial
residues with a worldwide annual generation of 6 million tons
[22]. Considering the huge amount of coffee residues produced
all over the world, its reutilization and valorization is of major rele-
vance [23]. The composition of coffee grounds may vary from spe-
cies to species, but they are typically rich in polysaccharides (34
53%, w/w) [22,24], which make them suitable for liquefaction into
bio-based polyols [21] to be used in PUFs formulations.
In the present study liqueed coffee grounds were used as the
polyol component for the production of viscoelastic bio-based
PUFs. To the best of our knowledge this is the rst report on the
production of PUFs with a distinct viscoelastic behavior from that
of typical rigid foams produced using bio-based polyols. The
formulation used in the production of PUFs was rst optimized
and the resulting foams were characterized by Fourier transform
infrared spectroscopy (FTIR), scanning electron microscopy
(SEM), density and thermal conductivity measurements, compres-
sive tests, dynamic mechanical analysis (DMA) and thermogravi-
metric analysis (TGA).
2. Materials and methods
2.1. Materials
Spent coffee grounds, typically consisting of hemicelluloses
(37 wt%), lignin (25 wt%), extractives (15 wt%), cellulose (9 wt%)
and ashes (12 wt%), used in this study were kindly provided by
NESTL S.A. (Portugal). The material with initial moisture content
of 80 wt% was air dried (moisture content achieved 10 wt%), milled
in a Retsch cross beater mill SK1 (Haan, Germany), sieved and the
1860 mesh fraction was selected for the liquefaction experiments.
This fraction was oven dried at 105 2 C for 24 h (moisture con-
tent around 12 wt%) before the liquefaction step. Polyethylene
glycol (PEG, 99% purity and average molecular weight of 400), sup-
plied by Merck, and glycerol (99.5% purity), purchased from
SigmaAldrich, were used as solvents in the liquefaction process.
Sulfuric acid (H2SO4, 9597% purity) supplied by Fluka was used
as catalyst. The resulting polyol presented an hydroxyl number
(OH number) of 287.2 mgKOH/g, an average molecular weight (Mn)
of 575.0 g/mol, an acid value (AC) of 9.8 mgKOH/g and a water con-
tent of 2.0% (w/w), and was used to produce PUFs by the reaction
with a polymeric isocyanate in the presence of a catalyst, a surfac-
tant and a blowing agent. The polymeric isocyanate Voranate
M229 MDI (methylene diphenyl diisocyanate) with an average
molecular weight of 340380 g/mol, a NCO content of 31.1%
(weight percent of free isocyanate groups), a functionality of 2.7,
a viscosity of 190 mPa s (at 25 C) and an isocyanate equivalent
of 135 was kindly supplied by Dow Chemicals. A mixture of
Tegostab B8404 (70% w/w) and Dabco DC3043 (30% w/w) was used
as surfactant. Tegostab B8404, supplied by Evonik, is a polyether-
modied polysiloxane with a density of 1.0451.065 g/cm3 (at
25 C), while Dabco DC3043, supplied by Air Products, and is a sili-
cone glycol copolymer with a density of 0.98 g/cm3 (at 21 C).
Dichloromethane was supplied by Fluka and used as blowing
agent, while Polycat 34, supplied by Air Products, is a tertiary
amine with a density of 0.84 g/cm3 (at 25 C) and was used as cata-
lyst. All the solvents used were of analytical grade and were used
as received.
2.2. Liquefaction of coffee grounds
Coffee grounds were used to produce the polyol via acid
liquefaction reaction in a 300 cm3 stainless steel vessel (PARR,
model 4566), equipped with an independent controller (PARR,
model 4842), stirring (200 rpm), heating resistance (1000 W),
internal water cooling system and temperature sensor, following
the procedure reported by Soares et al. [21]. The dried coffee
grounds powder, solvent mixture (PEG/glycerol, 90/10 wt/wt)
and sulfuric acid (4 wt%) were reacted at atmospheric pressure
and moderate temperature (160 C) during 80 min. The liquefac-
tion solvents/coffee grounds powder ratio used was 3/1 (wt/wt).
2.3. Characterization of coffee grounds derived polyol
2.3.1. Acid Value and hydroxyl number
The AC and OH number of the polyol were determined according
to the procedures described in our previous report [21].
2.3.2. Average molecular weight
The number average molecular weight (Mn) of polyol was mea-
sured using a K-7000 Vapor Pressure Osmometer (Knauer) cali-
brated using four standard solutions of benzil in acetone, at
40 C, in the concentrations range of 4.5883 g/L up to 20.6790 g/L.
2.3.3. Water content
The determination of the water content was carried out using a
KF 756 Coulometer for Karl Fisher titration, according ASTM
D6304-07. The samples of polyols were analyzed using Hydranal
(Hydranal Coulomat AG, Sigma) as reagent. The analyses were per-
formed in triplicate and the results averaged.
2.4. First step optimization of isocyanate/polyol ratio and catalyst
content
For the optimization of isocyanate/polyol molar ratio (RNCO/OH)
and percentage of catalyst (%cat), the coffee grounds derived polyol
was mixed with the isocyanate in the presence of the catalyst.
Polyol and different amounts of isocyanate and catalyst were
placed in polypropylene cups and the mixtures were homogenized
using an IKA Ost Basic mixer with rotating blades, during ca. 10 s at
700 rpm. The resulting products were analyzed by FTIR and the
area of the isocyanate peak (2250 cm1) measured to monitor
the extent of reaction.
According to Kapps and Buschkamp [25] the RNCO/OH used in the
PUFs production was determined using Eq. (1).
%NCO
miso  M
RNCO=OH NCO
1
mpolyol  OHnumber AC
mH2O  EqH2O
M KOH
where RNCO/OH is dened as the number of moles of NCO groups of
the isocyanate per OH mole of the polyol, miso is the mass (g) of iso-
cyanate, %NCO is the quantity of NCO groups in the isocyanate
(31.1%), MNCO is the molecular weight of NCO group (0.042 g/mmol),
mpolyol is the mass (g) of the polyol, OHnumber and AC are the hydroxyl
number and the acid value of polyol, respectively (mgKOH/g) and
MKOH is the molecular weight of KOH (56.1 mg/mmol). Finally,
mH2O is the mass of water present in the polyol while EqH2O is the
equivalent of OH groups present in the water (111 mmol/g).
 N.V. Gama et al. / Materials and Design 76 (2015) 7785
The catalyst percentage was determined using Eq. (2).
mcat
%cat
mpolyol misoc
where mcat is the mass (g) of catalyst.
2.5. Second step optimization of surfactant and blowing agent
content
After selecting the suitable RNCO/OH and %cat, the second step of
the optimization of PUFs formulation consisted in the selection of
the best surfactant and blowing agent percentages. The coffee
grounds derived polyol, the corresponding amount of isocyanate
and catalyst (selected in the previous optimization step) and differ-
ent amounts of surfactant and blowing agent were placed in
polypropylene cups. The mixtures were homogenized using an
IKA Ost Basic mixer with rotating blades, during ca. 10 s at
700 rpm and the foams were obtained by free expansion in the
mold at room temperature. The resulting foams were characterized
by SEM, density measurements, thermal conductivity, compressive
tests, TGA and DMA.
A mixture of two surfactants was used in order to produce
foams with good expansion and dimensional stability. PUFs pro-
duced using only Tegostab B8404 surfactant presented poor
dimensional stability (shrunk) while the foams prepared using
Dabco DC3043 surfactant on its own showed reduced expansion.
Different proportions of Tegostab B8404/Dabco DC3043 were then
tested and the best dimensional stability and foam expansion were
achieved for 70/30 (w/w), respectively. The surfactant percentage
was determined using Eq. (3).
msurf
%surf
mpolyol
where msurf is the mass (g) of the surfactant.
Dichloromethane, used as blowing agent, vaporizes due to the
exothermic reaction between the polyol and isocyanate and its
percentage was determined using Eq. (4).
mBA
%BA
mpolyol
where mBA is the mass (g) of the blowing agent.
2.6. Characterization of PUFs
2.6.1. Fourier transform infrared spectroscopy
The FTIR spectra of PUFs were collected on a Perkin Elmer FTIR
System Spectrum BX Spectrometer equipped with a single horizon-
tal Golden Gate ATR cell. All data were recorded at room tempera-
ture, in the 4000600 cm1 range by accumulating 32 scans with a
resolution of 4 cm1.
2.6.2. Scanning electron microscopy
SEM analyses were performed in a SU-70 (Hitachi) scanning
electron microscope after vacuum-coating with gold to avoid elec-
trostatic charging during examination and at accelerating voltage
of 15.0 kV. SEM images were used to determine the average cells
size of the foams. Cell diameter was obtained by inspecting visually
at least 15 porous cells on the surface of the each foam.
2.6.3. Density
PUFs specimens (10  10  10 mm3) were cut and weighed to
determine the density. Densities were obtained, dividing the
weight of the specimens by the calculated volume. The values pre-
sented correspond to the average density determined for 10 speci-
mens of each foam.
79
2.6.4. Thermal conductivity
A KD2 Pro (Decagon Devices) was used to measure the thermal
2 conductivity of the PUFs. Samples were analyzed by introducing
the thermal conductivity sensor inside the foams. An 8 min gap
between analyses was used to ensure the stabilization of the
sensor.
2.6.5. Mechanical properties
An Instron 5966 universal mechanical test analyzer was used to
measure the compressive stress of the foams, according to the
ASTM D 695. Before analysis, PUFs specimens
(10  10  10 mm3) were conditioned at 20 C and 41% relative
humidity, in a chamber with humidity control, for 24 h. Samples
were then placed between the two parallel plates and compressed
at 10 mm/min (100%/min) up to 70% of compression. The Youngs
modulus was calculated by the slope of the tangent of the linear
portion up to 5% strain (where elastic deformation occurs) of the
stressstrain curve. The compressive stress is the stress at which
the plastic deformation begins to occur. Toughness was calculated
by measuring the area underneath the stressstrain curve (up to
30% of strain). The Instron 5966 universal mechanical test analyzer
was also used to determine the recovery time of the PUFs. The
recovery time of samples was measured from the moment right
after attaining 70% of compression until their initial geometry
was regained. All presented values correspond to the average of
10 samples.
2.6.6. Thermogravimetric analysis
TGA of the PUFs was performed using a SETSYS Evolution 1750
thermogravimetric analyzer (Setaram) from room temperature up
to 800 C, at a heating rate of 10 C/min and under oxygen ux
3
(200 mL/min).
2.6.7. Dynamic mechanical analyses
DMA were carried out using a Tritec 2000 equipment (Triton
Technologies). Samples with dimensions of 10  9  6 mm3 were
compressed from 55 C up to 150 C at a constant heating rate
of 2 C/min and at a frequency of 1 Hz. The annealing of BA-16-S-
4
14 samples was carried out at 80 C in a ventilated oven for 4 h.
The annealed samples were cooled to room temperature in a des-
iccator for 4 h and 24 h (respectively) prior to being analyzed.
3. Results and discussion
Pursuing our goal to contribute to convert biomass residues into
high added value products, this paper consists in the use of a bio-
based polyol, obtained by acid liquefaction of coffee grounds [21],
for the production of PUFs. Specically, different values of NCO
index (RNCO/OH) and percentages of catalyst (%cat) have been tested
to select the appropriate RNCO/OH and %cat, upon which different
amounts of surfactant (%surf) and blowing agent (%BA) were tested.
3.1. Optimization of isocyanate and catalyst content
For the optimization of the RNCO/OH and %cat, different quantities
of isocyanate and catalyst were tested keeping the amount of
polyol constant, and FTIR spectroscopy was used to monitor the
extent of the reaction between the isocyanate and the polyol.
The determination of RNCO/OH includes the necessary correction of
the OH number taking into account the water content of the polyol.
In order to minimize the amount of isocyanate used, aiming at pro-
ducing PUFs with improved performance, the formulations dis-
cussed here only consider samples prepared using RNCO/OH values
between 0.6 and 0.8.
80 N.V. Gama et al. / Materials and Design 76 (2015) 7785
The FTIR spectra proles of the products obtained using the dif-
ferent formulations are very similar as illustrated in Fig. 1. The
band in the 32003450 cm1 region is attributed to the non-
bonded stretching vibrations of OAH groups and symmetric and
asymmetric stretching vibrations of the NAH of the urethane and
of urea groups. The latter results from the reaction between water
and some of the isocyanate groups and may even form domains if
the water quantity is signicant. The bands between 2950 and
2850 cm1 correspond to the asymmetric and symmetric stretch-
ing vibrations of CAH respectively. Near 2250 cm1 the peak
associated with the stretching vibration of the NCO group can be
detected when this region is expanded. The peak at 1710 cm1 is
due to the stretching vibrations of the C@O while the nearly over-
lapped bands between 1540 and 1517 cm1 are attributed to the
stretching and bending vibrations of the CAN and NAH of the
urethane moieties, respectively. The bands at 1220 and
1100 cm1 are associated with the C-O stretching vibrations.
To evaluate the effect of the RNCO/OH and %cat the area of the NCO
peak (Aisoc  2250 cm1) and the total area (Atotal) of the normal-
ized spectrum of each product were measured and their ratio
determined for each RNCO/OH and %cat studied. Fig. 2 shows the trend
of the Aisoc/Atotal ratio as a function of %cat for the three values of
RNCO/OH studied.
As expected, these data show that the Aisoc/Atotal ratio is higher
for higher RNCO/OH, while as the %cat increases the Aisoc/Atotal ratio
decreases. As the catalyst promotes the reaction between the
NCO and OH groups, for higher %Cat the isocyanate is consumed
in a higher extent. Theoretically, for RNCO/OH values below unity,
all the NCO groups of the isocyanate react completely with OH
groups of polyol [26]. Yet, in the FTIR spectra of the products
(including the one prepared using the lowest RNCO/OH) the peak
attributed to NCO groups is still detectable. This is thought be
associated with the fact that during the foam expansion some of
the reactive functional groups may become hindered within the
forming network.
Varying the RNCO/OH is a general strategy to control the mechani-
cal properties of PUFs [27] as higher RNCO/OH values yield foams
with a higher content of hard segments and thus improved proper-
ties [28], namely dimensional stability and mechanical properties.
For that reason, the RNCO/OH = 0.6 was discarded for the next step of
optimization. On the other hand, as discussed before, the extent of
reaction of isocyanate groups with OH groups can be limited
Fig. 1. Normalized FTIR spectra of the products prepared using different RNCO/OH
values with the 25002100 cm1 range highlighted.
Fig. 2. Aisoc/Atotal ratio for three different RNCO/OH and %cat used.
during foam expansion; therefore the use of the RNCO/OH = 0.8 was
discarded and the RNCO/OH = 0.7 was selected for the optimization
of the %surf and the %BA. These values were chosen for the optimiza-
tion of PUFs discussed next as they suggest a good balance
between the reactivity and the mechanical properties of the foams.
As regards the percentage of catalyst, since PUFs production
relies on two competing reactions isocyanate/polyol and iso-
cyanate/water (present in the polyol), a balance between them is
required to yield foams with good cellular structures and desired
physical properties. While these reactions may proceed in the
absence of catalysts, the reaction rates are too slow.
Nevertheless, a correct balance is required between the rate of
the blowing reaction and that of gelation. Indeed, if the blowing
reaction proceeds relatively fast, the foam may collapse. On the
other hand, if the gelation reaction overtakes the blowing reaction,
it might lead to foam shrinkage [29]. From the results obtained, the
%Cat = 7 was selected for the optimization of PUFs discussed next, as
it corresponds to a reasonable rate of consumption of NCO groups
that ensures that the foams produced do not collapse nor shrink.
3.2. Optimization of blowing agent and surfactant content
Having selected the values for RNCO/OH (0.7) and %cat (7%), differ-
ent amounts of blowing agent (between 12% and 16%) and surfac-
tant (between 14% and 18%) were tested. Based on our experience
on the production of PUFs using bio-based polyols [30], a 70/30 (w/
w) mixture of Tegostab B8404/Dabco DC3043 surfactants was used
in the present study since it affords the best foams dimensional
stability and expansion. The density, thermal conductivity and
thermomechanical properties of the resulting foams were evalu-
ated to assess the effect of the content of these reagents on the per-
formance of the ensuing PUFs. Table 1 summarizes the %surf and %BA
used and the properties of the resulting foams.
From the analysis of Table 1 it is clear that the density increases
with the increase of %surf. Although the role of the surfactant is to
control the process of foaming, higher concentrations of surfactant
limit the cell opening process and for that reason the foam density
increases [3]. On the other hand, the density of the foams decreases
with the increase of %BA. This correlation is related to the foaming
process: during mixing, air bubbles are usually introduced in the
reaction mixture acting as nucleation sites for the blowing gas.
The growing bubbles result in a closely packed network of bubbles
responsible for the typical cellular structure of PUFs [14]. In this
study dichloromethane was used as blowing agent which vapor-
izes as a result of the heat released by the exothermic reaction
between the polyol and isocyanate. Thus, when the amount of
 N.V. Gama et al. / Materials and Design 76 (2015) 7785 81

Table 1
Effect of %surf and %BA used on the properties of the resulting foams.

Sample code %BA (%) %surf (%) Density Thermal conductivity Youngs modulus (kPa) Toughness (J/m3) Compressive stress (kPa) Recovery time (s)
(kg/m3) (W/m K)
BA-12-S-14 12 14 100.5 5.1 0.043 85.8 7.1 1590.33 257.49 2.5 0.5 10.9 1.6
BA-12-S-16 12 16 106.4 3.6 0.042 86.9 7.7 1724.20 224.05 4.3 0.3 12.3 2.7
BA-12-S-18 12 18 119.5 9.8 0.041 88.6 6.6 1840.83 149.41 4.4 0.5 13.4 0.9
BA-14-S-14 14 14 86.0 3.4 0.040 75.9 3.4 1340.14 315.06 3.7 0.6 10.6 3.9
BA-14-S-16 14 16 97.5 2.1 0.039 81.5 4.0 1455.13 246.41 5.1 0.3 11.6 1.7
BA-14-S-18 14 18 105.0 6.8 0.040 86.3 6.3 1696.20 152.13 6.3 0.7 12.5 2.0
BA-16-S-14 16 14 80.1 5.4 0.040 73.2 5.8 1230.20 262.89 4.2 0.6 9.9 2.4
BA-16-S-16 16 16 92.8 5.2 0.039 79.7 6.7 1391.38 271.10 6.0 0.5 10.9 1.6
BA-16-S-18 16 18 95.6 7.6 0.040 84.7 5.6 1622.20 204.50 7.3 0.7 11.1 1.0

Sample code (BA-x-S-y) x %BA; y %surf.

dichloromethane is increased, more blowing gas is generated, so
the bubbles begin to expand and the foam begins to rise affording
lighter PUFs [31,32].
The results discussed above were actually conrmed by SEM
analysis. Fig. 3 shows the SEM micrographs of the cross-section
surface of PUFs and the typical cellular structure of PUFs is clearly
observed in all cases. Furthermore, it is also noticeable that the
cells are mainly closed. The increase of the amount of surfactant
reduces the cells size and increases their number [33], which can
be seen by comparing Fig. 3(a) and (b). In fact, when higher
amounts of surfactant were used the ensuing foams have more
cells with smaller sizes. The BA-12-S-14 image presents around
38 cells with an average cell size of 589.7 lm whereas that of
BA-12-S-18 presents around 68 cells with an average cell size of
466.5 lm. As regards the density, Fig. 3(b) and (c) prove that the
use of higher amounts of blowing agent leads to the formation of
bigger bubbles and thus less dense foams [18]. While BA-12-S-14
presents an average cell size of 589.7 lm, BA-16-S-14 presents
an average cell size of 738.8 lm.
Besides the effect on density, ne tuning of surfactant and
blowing agent is also important to adjust pore size and dis-
tribution. In fact, recent studies have highlighted the importance
of pore size distribution and structure in promoting acoustic per-
formances such as absorption, acoustic attenuation and vibration
damping properties. Zhang et al. have investigated the inuence
of the design parameters (pore cell size, open porosity) on the
sound absorption performance of PU porous materials and
reported that PUFs with cells diameter up to 1050 lm showed
good sound absorption performance over the range of 125
2000 Hz [6]. Coffee grounds derived PUFs presented cells diameter
between 466.5 and 738.8 lm, so according to Zhang et al. these
materials may have good sound absorbing properties.
The density of foams is also relevant for their thermal perfor-
mance, namely in terms of thermal conductivity (k). From Table 1
it is clearly observed that the variation of k values is not signi-
cant which is in agreement with the small differences between
the %surf and %BA used. The low thermal conductivity values
obtained for the coffee grounds derived PUFs make them good
candidates as thermal insulation materials with potential applica-
tions in buildings, refrigerator trucks, upholstery, etc. In fact,
these foams present a lower thermal conductivity when com-
pared to some PUFs obtained from other renewable resources
whose k values generally vary between 0.0233 and 0.0505 W/
m C [26,34].
Compressive tests were also performed to study the effect of the
%surf and %BA on the mechanical properties of foams. The Youngs
modulus, toughness, compressive stress and recovery time were
determined from compressive tests. Fig. 4(a) displays a representa-
tive compressive test up to 70% of deformation, while Fig. 4(b)
highlights the compressive test up to 30% of deformation. The
toughness values of all the foams, i.e. the work to failure
determined from the area under each curve, are summarized in
Table 1, together with those of Youngs modulus, compressive
stress and recovery time.
The compressive properties are related, among other, to the
foam density as well as foam structure and analyzing Table 1 it
is clear that the Youngs modulus and toughness increase with
the increase of surfactant and with the decrease of blowing agent
contents. In other words, the mechanical properties improved with
foam density which is in agreement with the literature [29].
For both PUFs presented in Fig. 4, the stressstrain curves exhi-
bit the typical three regions observed for elastomeric foams: (i)
approximately linear behavior for strains below 5%, due to the
bending of cell edges, (ii) a plateau where strain increases at nearly
constant stress up to a strain of approximately 50%, due to elastic
buckling of the cell edges and (iii) a densication region where the
stressstrain curve increases exponentially as collapsed cell edges
come into contact with each other, causing the material behavior
to approach that of the bulk polymer [35]. As regards the recovery
times registered in Table 1, even though the values are not very
distinct, a trend correlating higher recovery times for PUFs pre-
pared using higher amounts of surfactant and lower amounts of
blowing agent can be identied. This is in agreement with the fact
that the differences between the %surf and %BA are not very signi-
cant. The recovery time is a critical characteristic of viscoelastic
foams as it is related to the pneumatic effect. When compressed,
air needs time to leave and reenter the foam delaying the compres-
sion as well as re-expansion of the foam [3]. This effect is related to
the foam structure hence, foams with higher density may take a
longer time to recover the original geometry after compression
[31]. Fig. 5 shows the recovery behavior of BA-16-S-14 after 70%
of creep compression (a), 4 s after creep compression (b) and after
regaining the original geometry (c).
The compressive tests results, together with the fact that the
specimens regained their original geometry after 70% of creep
compression highlighting their amazing elastic behavior, show
that the coffee grounds derived PUFs present a similar mechanical
performance to those used as acoustic absorbing materials.
Therefore, this type of materials may nd higher added value
applications, for example in acoustic insulation and aeronautics
[4,28].
The understanding of porous materials mechanical properties,
in particular their viscoelastic ones, is a prior step to their elastic
and damping characterization [10]. Hence, complementary to com-
pressive tests, DMA was also used to study the thermo-mechanical
properties of the polymer phase. Fig. 6 displays the temperature
dependence of the storage modulus and tan (d) of BA-12-S-14
and BA-16-S-14. Based on our experience in this type of materials,
only the PUFs produced using the higher and the lower %BA were
analyzed as the behavior of the other PUFs was expected to be
intermediate. As the temperature goes through the glass transition
temperature (Tg) (measured at the top of the tan (d)), the increase
82 N.V. Gama et al. / Materials and Design 76 (2015) 7785

materials may well require further formulation adjustments to

obtain materials with higher Tg values. However, this is beyond
the scope of the present study. In Fig. 6(a), it is also important to
refer that the Tg is followed by a small hump (63.1 C and 60.8 C
for BA-12-S-14 and BA-16-S-14 respectively). This behavior,
though less intense, has already been observed for cork derived
PUFs [30] and was attributed to the fact that these foams were pro-
duced using a mixture of surfactants which as the temperature
increases seems to allow localized rearrangements of polymer seg-
ments which are somehow segregated. To test that, two other sam-
ples of BA-16-S-14 have been annealed for 4 h at 80 C, and then
cooled to room temperature in a desiccator for 4 h and 24 h respec-
tively. Afterward, the samples were submitted to DMA using the
same conditions as those used for the original (not annealed) BA-
16-S14. As it can be observed in Fig. 6(b), upon annealing the small
hump after the Tg disappeared, becoming a single peak centered at
48.1 C and 36.5 C depending on the time elapsed after annealing.
This suggests that annealing promotes mixing of localized domains
which in principle consists in a reversible process. Indeed, although
the reversibility of this process has not been fully proven, the fact
that the Tg value of the foam cooled to room temperature for 24 h

Fig. 3. SEM micrographs of BA-12-S-14 (a), BA-12-S-18 (b) and BA-16-S-14 (c).

of molecular motion causes the storage modulus to drop abruptly.

Below 0 C the foams behave as hard solids showing relatively high
storage modulus values (above 1.7  104 and 4.0  103 kPa for BA-
12-S-14 and BA-16-S-14, respectively). As regards the storage
modulus (E0 ) it is clear that the use of a higher %BA leads to a
decrease of E0 which is associated with the BA-16-S-14 lower den-
sity. Fig. 6(a) indicates that the increase of %BA induces a minor
effect on the Tg (19.4 C and 19.0 C for BA-12-S-14 and BA-16-S-
14, respectively). Moreover, as these values are around room tem-
perature, they may simply reect oscillations in atmospheric con- Fig. 4. Compressive behavior of BA-12-S-14 and BA-16-S-14 up to 70% of strain (a)
ditions and for that reason, potential applications of this type of and highlighted region up to 30% of strain (b).
 N.V. Gama et al. / Materials and Design 76 (2015) 7785 83

Fig. 6. Temperature dependence of storage modulus and tan (d) of BA-12-S-14 and
BA-16-S-14 (a) and temperature dependence of storage modulus and tan (d) of BA-
16-S-14 and BA-16-S-14 annealed for 4 h and analyzed 4 h and 24 h later (b).

Fig. 5. Recovery behavior of BA-16-S-14 after 70% of creep compression (a), 4 s after
creep compression (b) and after regaining the original geometry after 10 s (c).

is closer to that of the original BA-16-S-14 sample (i.e. without

annealing) indicates that the polymer segments of the foam are
slowly undergoing rearrangements.
Finally, the thermal stability of PUFs is also a relevant aspect for
their potential applications hence, the thermal degradation of
these polymers was investigated by TGA and Fig. 7 displays the
corresponding TGA curves of BA-12-S-14 and BA-16-S-14. Again,
only PUFs produced using the higher and the lower %BA were ana-
lyzed because it was expected that the others PUFs show a very
similar behavior.
From Fig. 7, it is clear that the decomposition of PUFs is Fig. 7. Thermograms of BA-12-S-14 and BA-16-S-14 under O2 ux.
characterized by a very small weight loss at around 100 C attrib-
uted to the release of residual water, followed by two main decom- second, around 390 C related to the soft segments [36]. From this
positions steps: one around 220 C, related to the thermal study it is concluded that PUFs derived from liqueed coffee
decomposition of the hard segments (e.g. urethane groups), and a grounds are thermally stable up to 190 C which is comparable,
84 N.V. Gama et al. / Materials and Design 76 (2015) 7785
and in some cases better, than the performance of other PUFs
derived from other eco-friendly resources which have degradation
temperatures within the 125 and 215 C range [30,37,38].
4. Conclusion
Polyurethane foams with peculiar mechanical properties have
been obtained from liqueed coffee grounds upon optimization
of the formulation, as opposed to what is commonly achieved
when using other bio-based polyols which generally yield rigid
foams. The thermal stability and thermal conductivity of the
PUFs produced are within the values registered for other PUFs
derived from renewable resources making them suitable for ther-
mal insulation. Additionally, the mechanical behavior of coffee
grounds derived PUFs suggest they can nd potential application
in the production of viscoelastic foams whose properties can be
tuned by adjusting the percentages of blowing agent and surfac-
tant depending on the application in mind. Interestingly, while
the recovery time is sensitive to the proportion of the percentage
of surfactant versus blowing agent, little variations were registered
for the thermal conductivity values. Moreover, the fact that the
samples regained their original geometry after 70% of compression
highlights the amazing elastic behavior of these coffee grounds
derived PUFs. Therefore, higher added value applications for this
type of materials can be envisaged in acoustic insulation and aero-
nautics applications, i.e. beyond thermal insulation as it generally
happens for other bio-based PUFs.
Acknowledgements
The authors would like to acknowledge QREN/FEDER for fund-
ing the Ecopolyols Project (N 11435). Thermal Analysis
Laboratory was funded by FEDER Funds through Programa
Operacional Factores de Competitividade COMPETE and by
National Funds through FCT under the Project REEQ/515/CTM/
2005. The authors would also like to acknowledge Dow Chemical
for kindly supplying of the isocyanate. CICECO acknowledges FCT
for the FCOMP-01-0124-FEDER-037271 (PEst-C/CTM/LA0011/
2013). C. Freire acknowledges FCT/MCTES for a research contract
under the Program Investigador FCT 2012.
References
[1] K. Ashida, Polyurethane and Related Foams Chemistry and Technology, Taylor
& Francis Group, Florida, 2007.
[2] S. Kintrup, J.P. Treboux, H. Mlspreuve, Low resilience high performance
recent advances in viscoelastic exible slabstock foam, in: Polyurethane
conference 2000, Boston, 2000.
[3] R. Landers, R. Hubel, R. Borgogelli, The importance of cell structure for
viscoelastic foams, PU Mag. Int. (2008) 4048.
[4] L.M. Yang, V.P.W. Shim, A visco-hyperelastic constitutive description of
elastomeric foam, Int. J. Impact Eng 30 (2004) 10991110, http://dx.doi.org/
10.1016/j.ijimpeng.2004.03.011.
[5] N.N. Najib, Z.M. Ariff, A.A. Bakar, C.S. Sipaut, Correlation between the acoustic
and dynamic mechanical properties of natural rubber foam: effect of foaming
temperature, Mater. Des. 32 (2011) 505511, http://dx.doi.org/10.1016/
j.matdes.2010.08.030.
[6] C. Zhang, J. Li, Z. Hu, F. Zhu, Y. Huang, Correlation between the acoustic and
porous cell morphology of polyurethane foam: effect of interconnected
porosity, Mater. Des. 41 (2012) 319325, http://dx.doi.org/10.1016/
j.matdes.2012.04.031.
[7] N. Sarier, E. Onder, Thermal insulation capability of PEG-containing
polyurethane foams, Thermochim. Acta 475 (2008) 1521, http://dx.doi.org/
10.1016/j.tca.2008.06.006.
[8] A. Pereira, L. Godinho, D. Mateus, J. Ramis, F.G. Branco, Assessment of a
simplied experimental procedure to evaluate impact sound reduction of oor
coverings, Appl. Acoust. 79 (2014) 92103, http://dx.doi.org/10.1016/
j.apacoust.2013.12.014.
[9] Z. Li, M.J. Crocker, Effects of thickness and delamination on the damping in
honeycombfoam sandwich beams, J. Sound Vib. 294 (2006) 473485, http://
dx.doi.org/10.1016/j.jsv.2005.11.024.
[10] L. Jaouen, A. Renault, M. Deverge, Elastic and damping characterizations of
acoustical porous materials: available experimental methods and applications
to a melamine foam, Appl. Acoust. 69 (2008) 11291140, http://dx.doi.org/
10.1016/j.apacoust.2007.11.008.
[11] J.P.S. Aniceto, I. Portugal, C.M. Silva, Biomass-based polyols through
oxypropylation reaction, ChemSusChem 5 (2012) 13581368, http://
dx.doi.org/10.1002/cssc.201200032.
[12] M. Niu, G. Zhao, M.H. Alma, Polycondensation reaction and its mechanism
during lignocellulosic liquefaction by an acid catalyst: a review, For. Stud.
China 13 (2011) 7179, http://dx.doi.org/10.1007/s11632-011-0109-7.
[13] E. Sharmin, F. Zafar, S. Ahmad, Seed Oil Based Polyurethanes: An Insight,
Intech, India, 2012.
[14] O. Kwon, S. Yang, D. Kim, J. Park, Characterization of polyurethane foam
prepared by using starch as polyol, J. Appl. Polym. Sci. 103 (2006) 15441553.
[15] T.W. Pechar, S. Sohn, G.L. Wilkes, S. Ghosh, C.E. Frazier, A. Fornof, et al.,
Characterization and comparison of polyurethane networks prepared using
soybean-based polyols with varying hydroxyl content and their blends with
petroleum-based polyols, J. Appl. Polym. Sci. 101 (2006) 14321443, http://
dx.doi.org/10.1002/app.23625.
[16] S.-R. Yang, O.-J. Kwon, D.-H. Kim, J.-S. Park, Characterization of the
polyurethane foam using alginic acid as a polyol, Fibers Polym. 8 (2007)
257262, http://dx.doi.org/10.1007/BF02877267.
[17] R. Tanaka, S. Hirose, H. Hatakeyama, Preparation and characterization of
polyurethane foams using a palm oil-based polyol, Bioresour. Technol. 99
(2008) 38103816, http://dx.doi.org/10.1016/j.biortech.2007.07.007.
[18] A.A. Abdel Hakim, M. Nassar, A. Emam, M. Sultan, Preparation and
characterization of rigid polyurethane foam prepared from sugar-cane
bagasse polyol, Mater. Chem. Phys. 129 (2011) 301307. doi:10.1016/
j.matchemphys.2011.04.008..
[19] Y. Jin, X. Ruan, X. Cheng, Q. L, Liquefaction of lignin by polyethylene glycol
and glycerol, Bioresour. Technol. 102 (2011) 35813583, http://dx.doi.org/
10.1016/j.biortech.2010.10.050.
[20] B. Soares, N.V. Gama, C.S.R. Freire, A. Barros-Timmons, I. Brando, R. Silva,
et al., Ecopolyol production from industrial cork powder via acid liquefaction
using polyhydric alcohols, ACS Sustain. Chem. Eng. 2 (2014) 846854, http://
dx.doi.org/10.1021/sc400488c.
[21] B. Soares, N.V. Gama, C.S.R. Freire, A. Barros-Timmons, I. Brando, R. Silva,
et al., Spent coffee grounds as a renewable source for ecopolyols production, J.
Chem. Technol. Biotechnol. 64 (2014) 250275, http://dx.doi.org/10.1002/
jctb.4457.
[22] S.I. Mussatto, E.M.S. Machado, S. Martins, J.A. Teixeira, Production,
composition, and application of coffee and its industrial residues, Food
Bioprocess Technol. 4 (2011) 661672, http://dx.doi.org/10.1007/s11947-
011-0565-z.
[23] P.S. Murthy, M. Madhava Naidu, Sustainable management of coffee industry
by-products and value additiona review, Resour. Conserv. Recycl. 66 (2012)
4558. doi:10.1016/j.resconrec.2012.06.005..
[24] S.I. Mussatto, L.M. Carneiro, J.P.A. Silva, I.C. Roberto, J.A. Teixeira, A study on
chemical constituents and sugars extraction from spent coffee grounds,
Carbohydr. Polym. 83 (2011) 368374, http://dx.doi.org/10.1016/
j.carbpol.2010.07.063.
[25] M. Kapps, S. Buschkamp, The Production of Rigid Polyurethane Foam, Bayer
Material Science, 2004.
[26] S. Tan, T. Abraham, D. Ference, C.W. Macosko, Rigid polyurethane foams from a
soybean oil-based Polyol, Polymer 52 (2011) 28402846, http://dx.doi.org/
10.1016/j.polymer.2011.04.040.
[27] Y. Kurimoto, A. Koizumi, S. Doi, Y. Tamura, H. Ono, Wood species effects on the
characteristics of liqueed wood and the properties of polyurethane lms
prepared from the liqueed wood, Biomass Bioenergy 21 (2001) 381390.
[28] T. Wang, L. Zhang, D. Li, J. Yin, S. Wu, Z. Mao, Mechanical properties of
polyurethane foams prepared from liqueed corn stover with PAPI, Bioresour.
Technol. 99 (2008) 22652268, http://dx.doi.org/10.1016/
j.biortech.2007.05.003.
[29] M.L. Listemann, A.L. Wressell, K.R. Lassila, H.C. Klotz, G.L. Johnson, A.C. Savoca,
The inuence of tertiary amine structure on blow-to-gel selectivity, J. Cell.
Plast. 29 (1993) 473474, http://dx.doi.org/10.1177/0021955X9302900588.
[30] N.V. Gama, B. Soares, C.S.R. Freire, R. Silva, C.P. Neto, A. Ferreira, Rigid
polyurethane foams derived from cork liqueed at atmospheric pressure,
Polym. Int. 64 (2014) 250257, http://dx.doi.org/10.1002/pi.4783.
[31] N.J. Mills, R. Stmpi, F. Marone, P.A. Brhwiler, Finite element
micromechanics model of impact compression of closed-cell polymer foams,
Int. J. Solids Struct. 46 (2009) 677697, http://dx.doi.org/10.1016/
j.ijsolstr.2008.09.012.
[32] F. Saint-Michel, L. Chazeau, J.-Y. Cavaill, E. Chabert, Mechanical properties of
high density polyurethane foams: I. Effect of the density, Compos. Sci. Technol.
66 (2006) 27002708, http://dx.doi.org/10.1016/j.compscitech.2006.03.009.
[33] H. Lim, Effects of silicon surfactant in rigid polyurethane foams, Express
Polym. Lett. 2 (2008) 194200, http://dx.doi.org/10.3144/
expresspolymlett.2008.2.
[34] V. Ribeiro da Silva, M.A. Mosiewicki, M.I. Yoshida, M. Coelho da Silva, P.M.
Stefani, N.E. Marcovich, Polyurethane foams based on modied tung oil and
 N.V. Gama et al. / Materials and Design 76 (2015) 7785 85

reinforced with rice husk ash I: synthesis and physical chemical
characterization, Polym. Test. 32 (2013) 438445, http://dx.doi.org/10.1016/
j.polymertesting.2013.01.002.
[35] L.J. Gibson, Michael F. Ashby, Cellular Solids Structure and Properties,
Cambridge University Press, Cambridge, 1999.
[36] L. Shufen, J. Zhi, Y. Kaijun, Y. Shuqin, W.K. Chow, Studies on the thermal
behavior of polyurethanes, Polym. Plast. Technol. Eng. 45 (2006) 95108,
http://dx.doi.org/10.1080/03602550500373634.
[37] X. Luo, A. Mohanty, M. Misra, Lignin as a reactive reinforcing ller for water-
blown rigid biofoam composites from soy oil-based polyurethane, Ind. Crops
Prod. 47 (2013) 1319, http://dx.doi.org/10.1016/j.indcrop.2013.01.040.
[38] X. Luo, S. Hu, X. Zhang, Y. Li, Thermochemical conversion of crude glycerol to
biopolyols for the production of polyurethane foams, Bioresour. Technol. 139
(2013) 323329, http://dx.doi.org/10.1016/j.biortech.2013.04.011.