Mechanism of Action of Different d-Spacings Clays on the

Intumescent fire Retardance of Polymers

Simone Pereira da Silva Ribeiro,1 Luciana Rocha de Moura Esteva
~ o,2
Celeste Margarida Correia Pereira, Regina Sandra Veiga Nascimento1
3
1
Instituto de Quımica, DQO, UFRJ, CT Bloco A, 6 andar, Cidade Universita ria, Rio de Janeiro, RJ, CEP: 21941-590, Brazil
2
Age^ ncia Nacional do Petro  s Natural e Biocombustıveis-ANP, SCM, Av. Rio Branco 65, 17 andar, Centro, Rio de Janeiro,
 leo, Ga
RJ, CEP: 20090-004, Brazil
3
INEGI, Instituto de Engenharia Meca
^ nica e Gesta
~ o Industrial, Rua Doutor Roberto Frias 378, 4200-465 Porto, Portugal
Correspondence to: S. P. da Silva Ribeiro (E - mail: [email protected])

ABSTRACT: A series of sodic and organophilic clays with different d-spacings was added to a polymeric matrix of poly (ethylene-co-
butyl acrylate) EBA-30, containing an intumescent formulation of ammonium polyphosphate (APP) and pentaerythritol (PER), in
order to investigate the influence of the d-spacings of the clays on their synergistic effects with the intumescent formulation. A series
of samples was evaluated through cone calorimetry, SEM, TG-FTIR, FTIR, and XRD of burned residues. The results revealed that the
addition of clays with smaller d-spacings led to a synergistic interaction with the intumescent formulation, and consequently to an
improvement in the flame retardance of the materials. This effect was not observed with the addition of clays with d-spacings larger
than 30 Å. For these materials, the formation of a less homogeneous and structured intumescent layer, and a delay in the formation
of the phosphocarbonaceous species which act as char precursors was observed. This delay could be responsible for the loss of
synergy and also for the type of char morphology formed when larger d-spacings clays were used. XRD analyses results indicated
that the presence of clays in the materials promoted changes in the crystalline phase of the char when the samples were submitted
to higher temperatures. Those changes probably allowed for the maintenance of the structures at high temperatures. V C 2013 Wiley

Periodicals, Inc. J. Appl. Polym. Sci. 130: 1759–1771, 2013

KEYWORDS: flame retardance; nanostructured polymers; clay; degradation

Received 18 December 2012; accepted 19 March 2013; Published online 6 May 2013
DOI: 10.1002/app.39349

INTRODUCTION agent. These additives, under heating, form a tumid carbona-

ceous surface layer, known as “char,” which prevents the trans-
The development of polymeric materials with effective flame re-
fer of heat, fuel, and oxygen, hence ceasing the flame.3–6 In this
tardant properties is becoming more important with time.1 It
study, ammonium polyphosphate (APP) was used as the acid
has been recognized that the addition of only clay nanoparticles
source and blowing agent, and pentaerythritol (PER) was added
to polymeric matrices does not avoid the complete degradation
as the carbonific source.
of the polymer, despite the increase in thermal stability and
decrease in the heat release rate registered through the relevant The sole use of an intumescent formulation rarely leads to suita-
analyses. Furthermore, the exclusive use of clays does not inter- ble LOI results. Therefore, several authors have been searching for
fere significantly or positively with other important flammability additives that could act as synergistic agents. Zeolites, montmoril-
tests results, as for instance LOI (Limit of Oxygen Index) and lonite, sepiolite, and even spent catalysts from fluid-bed catalytic
UL-94, where other factors such as the time for ignition are cracking units (FCC) in oil refining processes have been reported
more determinant.2 Therefore, the concomitant use of addi- as synergy agents in fire retardant formulations.7–12 Our group
tional fire retardant additives is an important practice, especially has observed that both sodic and organophilic clays are able to
if the additive plays a role as a synergic element. play this role in a poly[ethylene(30%)–butyl acrylate] copolymer
Intumescent formulations are composed of a set of additives matrix containing an intumescent formulation of APP and
which increase the flame retardant properties of polymers. An PER.13–15 In addition, it has been observed that the d-spacings of
intumescent formulation is commonly composed of three ingre- clays influence the synergistic effect as well, so that LOI values
dients: an acid source, a carbonific compound, and a blowing and UL-94 classification decrease when the d-spacings increase.16

C 2013 Wiley Periodicals, Inc.

V

WWW.MATERIALSVIEWS.COM WILEYONLINELIBRARY.COM/APP J. APPL. POLYM. SCI. 2013, DOI: 10.1002/APP.39349 1759

 ARTICLE

Therefore, this study aimed to assess, by cone calorimetry analy- isotherm is important to ensure that all chemical reactions
sis, the influence of the d-spacings of clays on the fire retardant liable to occur at a given temperature took place for char for-
properties of materials containing an intumescent formulation. mation. These analyses aimed at identifying possible changes in
An attempt to elucidate this phenomenon was made through the crystalline arrangement of the char with the increase of tem-
the analysis of Scanning Electron Microscopy (SEM) images of perature and with the addition of clays with different basal
the intumescent layer formed, TG-FTIR, FTIR, and XRD of the spacings. The analyses were performed in a X-ray diffraction
burned residues. Rigaku Miniflex instrument with a copper anode. Scanning was
carried out in a 0.05 /s from 2 to 30 (2h), using 30 kV and
EXPERIMENTAL 15 mA voltage and filament current, respectively.
The polymer matrix used was a poly[ethylene(30%)–butyl acry- Scanning Electron Microscopy (SEM)
late] copolymer supplied by Elf-Atochem under the trade name SEM analyses were performed on the intumescent layers formed
Lotryl 30BA02, hereafter referred to as EBA-30. The intumes- in the samples, after being analyzed according to UL-94 tests.
cent formulation used was composed of ammonium polyphos- These analyses were carried out to identify changes in char mor-
phate (APP), supplied by Clariant under the trade name of phology when clays with different basal spacings were added.
Exolit 422, and pentaerythritol (PER) from Sigma-Aldrich. In The analyses were conducted in a Jeol JSM-6460LV microscope
the polymeric samples produced, the intumescent formulation coupled to a Thermo-Noran, Six 200 EDS (Energy Dispersive
represented 30 wt %, with a 3 : 1 ratio of APP : PER where, Spectroscopy) system. A 15 kV operating voltage in a tungsten
according to the literature, maximum flame retardancy proper- filament was used, and the images were observed through
ties are observed for polyethylene materials.5 The sodic clay and secondary electrons mode.
two organophilic clays studied were supplied by Southern Clays
under the trade names Cloisite Na, Cloisite 30B, and Cloisite TG-FTIR Analysis
15A, respectively. The clays were added to the polymeric mix- Thermogravimetric Analysis (TG) coupled to Fourier Transform
ture in order to account for 3 wt % of the total mass involved. Infrared Spectroscopy (FTIR) was carried out to verify whether
In a previous article, the chemical composition of the sodic there were different reactions products formed in the gas phase
clay, the chemical structure of the organic modifiers used in the for the various systems studied. The samples were previously
clay organophilization procedures and the concentrations of milled in a cryogenic mill to ensure the homogeneity of the
these organophilic agents were presented.16 material. The analyses were performed in a NETZSCH 209
TG instrument, coupled with the Bruker Vector 22 FTIR
The polymers and additives were mixed in a Haake Rheocord
spectrometer.
9000 rheometer, equipped with a rheomix chamber 600 and
roller blades rotor, at 160 C under 50 rpm for 15 min. The FTIR Residue Analysis
mixture was then pressed at 150 C in a Carver press applying Since the reactions for char formation occur in the condensed
9000 kgf load on a 10 3 10 cm area to obtain sheets of 3 mm phase,17 a series of FTIR analyses of the residues obtained from
thick, from which all test specimens were obtained. heating the samples at different temperatures was carried out.
By identifying the reactions taking place in the condensed phase
Cone Calorimetry
and associating them with the intumescent layer formation, a
Cone Calorimetry was performed using a FTT Thermal Science
better understanding of the influence of d-spacing of the clays
Stanton Redcroft equipment following the ASTM 1354-04 A
on the synergistic effect could be envisaged.
standard. The duration of the tests followed the standard rec-
ommendation on data collection until 2 min after any flaming After processing, the samples were heated in a furnace under air
or combustion signs cease. Samples measuring 100 3 100 3 3 flow, with a heating rate of 10 C/min up to 280, 350, 430, and
mm were exposed horizontally to a 50 kW heat flux under irra- 560 C, with a final isotherm of 12 h. The analyses were per-
diation by a conical-shaped radiant electric heater. Each sample formed using a Nicolet 740 Fourier transform infrared spec-
was tested in three replicates. trometer and KBr pellets.
The heat release rate (HRR) is the key parameter obtained by RESULTS AND DISCUSSION
the equipment, although it performs simultaneous acquisition
d-Spacing Measurements
of other complementary data.2 The other basic data registered
X-ray Analysis. XRD analyses were performed in order to com-
by the cone calorimeter along all the burning tests was the sam-
pare the d-spacings of clays before and after processing with the
ple mass decay, smoke release, and carbon dioxide (CO2) and
matrix. When the d-spacings of clays are kept unchanged after
carbon monoxide (CO) emissions.
processing with the polymer, it can be considered that a com-
X-Ray Diffraction Analysis (XRD) posite was formed. On the other hand, if the d-spacings
X-ray diffraction analysis was carried out in order to measure increase, as shown by the peak associated with montmorillonite
the d-spacings of the clays, before and after processing with the structure in a diffractogram, an intercalated nanocomposite is
polymer and intumescent additives. The samples were pulver- liable to have been produced,2 although a complete characteri-
ized in a cryogenic mill in order to obtain higher homogeniza- zation of the nanomorphology of the samples requires the com-
tion. Moreover, samples were burned at 280, 350, 430, and bined analysis of XRD patterns and other techniques, especially
560 C in a furnace under a synthetic air flow and a heating rate those involving image generation, such as transmission electron
of 10 C/min with isotherm of 12 h and analyzed by XRD. This microscopy (TEM) and atomic force microscopy (AFM). Figure

1760 J. APPL. POLYM. SCI. 2013, DOI: 10.1002/APP.39349 WILEYONLINELIBRARY.COM/APP

 ARTICLE

Figure 3. HRR versus Time of the studied materials.

Figure 1. XRD analyses of Cloisites Na, 30B, and 15A.

1 presents the diffractograms of clays before processing and the In this study, the inflammability of the produced materials is
calculated d-spacings. evaluated by cone calorimetry. Figure 3 shows a decrease in the
HRR with the addition of the intumescent formulation. The
From the results, shown in Figure 2, it is possible to note that
results were also influenced by the d-spacings of the added
there is a small displacement of the Cloisite Na peak after proc-
clays. Two peaks are clearly identifiable in the curves of all sam-
essing, reaching the calculated d-spacing around 13.5 Å. How-
ples containing the intumescent formulation. The first peak is
ever, Cloisites 30B and 15A, respectively presented substantial
probably related to the formation of the intumescent layer,
changes in the displacements towards smaller angles. Cloisite
whereas the second could be attributed to its degradation.18 The
30B presented an increase in d-spacing from 18.2 Å to 21.8 Å
addition of Cloisites Na and 30B, with d-spacings smaller than
after processing. For the Cloisite 15A, the increase was from
24Å, decreased the HRR peaks and delayed the occurrence of
32.1 Å to 35.3 Å. These results suggest that the processing of
the second peak, and hence, char degradation. On the other
these clays led to the formation of intercalated nanocomposite
hand, the addition of Cloisite 15A—with a d-spacing larger
materials, where clays interacted with the polymeric matrix due
than 35.3 Å—leads to a slight increase in the HRR, compared
to their organophilization. In the case of the sample containing
to the results obtained for the sample containing only the intu-
Cloisite Na, the discrete displacement occurred due to the polar
mescent formulation. Therefore, according to this technique,
characteristic of polymer matrix.
the addition of this clay, with a d-spacing greater than 30 Å,
Inflammability Evaluation did not lead to a synergistic effect with the intumescent
Cone Calorimetry. Previous studies16 have demonstrated that formulation.
sodic clays with d-spacings of up to 24 Å can act as synergistic
Figure 4 presents the dependence of mass loss of materials with
agents in polymeric materials containing an intumescent formu-
cone calorimetry analysis time. It can be noticed that the
lation when evaluated through LOI. However, when clays with
addition of the intumescent formulation, despite the clay’s
higher d-spacings are added, the synergistic effect virtually dis-
d-spacing, induces mass loss early on. This phenomenon is related
appears and the materials cease to classify according to the UL-
to the formation of the intumescent layer, which is directly
94 standard.16 These results have suggested that the clay struc-
linked to the chemical reactions involving the degradation of
ture needed to promote synergy disappears with the increase in
d-spacings, hence reducing the synergistic effect.16

Figure 2. XRD patterns of polymeric samples. Figure 4. Mass loss over cone calorimetry analysis time.

WWW.MATERIALSVIEWS.COM WILEYONLINELIBRARY.COM/APP J. APPL. POLYM. SCI. 2013, DOI: 10.1002/APP.39349 1761

 ARTICLE

Table I. Parameters Analyzed by Cone Calorimetry

THR SEA SEA time SEA Average CO time CO average CO2 time CO2 average
Sample (kJ) peak to peak (s) (m2/kg) to peak (s) (kg/kg) to peak (s) (kg/kg)
EBA-30 899 1316 177 462 220 0.0180 175 1.77
EBA-30 1 APP/PER 709 1741 258 796 280 0.0641 260 1.58
EBA-30 1 APP/PER 1 683 1346 292 781 335 0.0591 335 1.69
Cloisite Na
EBA-30 1 APP/PER 1 709 1246 382 787 420 0.0522 380 1.75
Cloisite 30B
EBA-30 1 APP/PER 1 741 1993 277 868 280 0.0648 245 1.73
Cloisite 15A

pentaerythritol, ammonium polyphosphate, and even of the The production of CO and CO2 is also an important parameter
polymeric matrix. In Figure 4, it is possible to observe that the to be measured because these are two gases commonly pro-
degradation profile of the sample containing only the intumes- duced in fires and can cause fatalities. The formation of carbon
cent formulation (EBA-30 1 APP/PER) was not affected by the monoxide is linked to an incomplete combustion, while the car-
addition of Cloisite 15A. Also, assigning a defined mass loss, as bon dioxide, produced in greater quantity, is related to the
for instance 60%, the degradation time was virtually the same complete combustion. Regarding the production of CO, it is
for both samples. On the other hand, the delay in the degrada- observed that the addition of APP–PER increases considerably
tion for the sample containing Cloisites Na and 30B is evident, the formation of this gas. The addition of the Cloisite 30B
indicating the formation of a more effective intumescent layer. slightly decreases the concentration of CO, while Cloisite 15A
increased the production of CO. Furthermore, the time for the
Table I also presents another relevant parameter measured
maximum CO peak to be reached is significantly higher than
through the cone calorimeter. Besides the aforementioned effect
for EBA 1 APP–PER when Cloisites Na and 30B are added.
on HRR, the addition of the intumescent formulation decreases
However, this parameter is unchanged with the addition of
the total heat released (THR) of the matrix. The addition of
Cloisite 15A.
Cloisite Na clay further decreased the THR, when compared
to the results obtained for the sample EBA-30 1 APP/PER. The amount of CO2 produced stayed practically unchanged
However, the addition of Cloisite 15A increased it slightly. when the intumescent formulation and clays were added.
However, the maximum concentration of carbon dioxide was
The addition of the intumescent formulation increased the
reached after a longer period of time when APP–PER was
amount of smoke produced (SEA, Specific Extinction Area)
added.
during burning, as previously observed in other studies.11 Thus,
although the APP/PER system increases the fire resistance of the The addition of Cloisites Na and 30B led to yet longer times.
polymer, once the burning process begins, the mixture contain- However, the presence of the Cloisite 15A decreased the time
ing the formulation releases more smoke than the pure polymer. for the CO2 peak to be reached, when compared with the
This is a common disadvantage in the use of intumescent flame sample containing APP/PER only.
retardants. However, the addition of Cloisites Na and 30B
Thus, the presence of Cloisite Na and 30B, with d-spacings of
decreased the maximum peak of obscuration of the pure poly-
13.5 and 21.8 Å respectively, led to an increase in flame retard-
mer, while the addition of Cloisite15A led to a considerable
ancy, as measured by cone calorimetry, while the sample con-
increase of SEA, with a greater peak than that of the sample
taining the Cloisite 15A, with greater d-spacing (35.3 Å),
containing only the intumescent formulation. The same is
showed no improvement over the samples containing only the
observed for the SEA average values. Furthermore, the time nec-
intumescent formulation. Thus, cone calorimetry also showed
essary for the SEA peak to be reached is greater with the intu-
that the increase in d-spacings leads to a decrease in flame
mescent formulation, and is further increased when Cloisites Na
retardance, as previously observed by UL-94 and LOI testing.16
and 30B are present. On the other hand, Cloisite 15A decreases
this time. These results indicate that the addition of clays with Morphology of Char Formed
smaller d-spacings decreases the obscuration due to smoke pro- Scanning Electron Microscopy. Considering the results
duction as well as delaying its formation. However, the addition obtained by cone calorimetry, which show a correlation between
of clay with larger d-spacing did lead to worse results in com- the basal spacings of clays and flame retardance, Scanning Elec-
parison to the EBA-30 1 APP/PER sample. tron Microscopy (SEM) was employed to observe the morphol-
ogy of the char formed with the different clays used. Figure 5
It is noteworthy that it is very important to obtain synergistic
shows SEM images of the char formed upon firing of the sam-
agents that reduce the smoke generated and increase its time of
ples according to the UL-94 test.
appearance in systems containing an intumescent formulation,
since obscuring generates panic, causing disorientation of the vic- The images show that the pure polymer did not form any type
tims who will have more difficulty to access the emergency exits. of char, showing a completely smooth surface. Chemical analysis

1762 J. APPL. POLYM. SCI. 2013, DOI: 10.1002/APP.39349 WILEYONLINELIBRARY.COM/APP

 ARTICLE

Figure 5. SEM of the char surfaces formed: (A) EBA-30, (B) EBA-30 1 APP/PER, (C) EBA-30 1 APP/PER 1 Cloisite Na, (D) EBA-30 1 APP/PER 1
Cloisite 30B, and (E) EBA-30 1 APP/PER 1 Cloisite 15A.[Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

of this surface revealed the presence of only carbon and oxygen. smooth regions, indicating that the char was not fully formed
Sample EBA-30 1 APP/PER presented both spongy islands, across the materials surface. Chemical analysis showed the pres-
demonstrating the formation of the intumescent layer, and ence of phosphorus in the spongy region, which is characteristic

WWW.MATERIALSVIEWS.COM WILEYONLINELIBRARY.COM/APP J. APPL. POLYM. SCI. 2013, DOI: 10.1002/APP.39349 1763

 ARTICLE

structure of the char.19 Between 280 C and 350 C occurs the

swelling of the char through the release of ammonia generated
from APP degradation.19 From 500 C to 550 C weight loss
decreases and the sample mass is stabilized, indicating the for-
mation of a protective layer which prevents the degradation of
the material. However, APP/PER alone does not avoid the com-
plete degradation of the material since at higher temperatures
the amount of residue approaches zero. However, the addition
of all clays, regardless of their basal spacings, led to an increase
in the thermal stability of the char formed, maintaining a resi-
due of about 8% at 900 C. A similar result was obtained in a
previous study.16
Montmorillonite clays can act as catalysts for various chemical
reactions, including Diels–Alder reactions and esterification.20,21
Thus, sodium and organophilizated clays could favor the reac-
Figure 6. TG curves obtained by TG-FTIR.
tions that give rise to the polyaromatic structure of the char.
The slight decrease of the decomposition temperature with the
of the intumescent layer. However, the addition of Cloisites Na
addition of clays could be an indication of this catalysis. Fur-
and 30B led to the formation of a more uniform and well-struc-
thermore, it has been observed that the interlayer spacing of
tured intumescent layer which possibly accounts for the
clays have an important role in the catalysis of many chemical
enhanced flame retardant properties observed. The chemical
reactions. Alkenes, for example, are protonated in the interlayer
analysis results also show the presence of the elements Si and
of clays, being generated carbocation intermediates.22 In addi-
Al, which indicates clay participation in the intumescent layer.
Finally, Cloisite 15A which has the largest d-spacing, led to the tion, dimerization of unsaturated fatty acids, occurs preferably
formation of a char with similar morphology to that of the in the montmorillonite interlayer.23,24 According to Weiss,23 the
EBA-30 1 APP/PER sample, which might explain the low flame dimerization reaction of oleic acid only occurs because the p
retardance properties of the samples containing this clay. Once bond is in direct contact with the silicate layer of montmoril-
again the char formed is neither homogeneous nor well struc- lonite interlayer space, which is probably activated by the empty
tured, alternating smooth and spongy regions similar to those 3d orbital of silicon. Furthermore, compounds such as alcohols
of the sample EBA-30 1 APP/PER. The chemical analysis and nitriles containing long unsaturated chains, which are
results show the presence of phosphorus, silicon, and aluminum arranged in the interlayer space perpendicular or tilted to a
only in the foamed parts. In the smooth regions, only carbon large angle, do not dimerize. According to the same author, in
and oxygen were detected as in the pure matrix. these cases, the p connection is far from the silicate layer and
cannot be activated. Thus, the increase of the interlayer spacings
Thus, the addition of clays with various basal spacings induced
could hamper the activation of alkenes, leading to a decrease in
changes in the morphology of the char formed that influenced the
the reaction rate of formation of polyaromatic hydrocarbons,
flame retardant properties measured through cone calorimetry.
precursors of char.
Understanding the Action of Clays in Synergy From the TG curves, shown in Figure 6, it is not possible to
Analysis of the gas phase—TG-FTIR. In order to elucidate the detect differences in the decomposition temperature of the sam-
action of clays with different basal spacings on the flame retard- ples containing clays with different basal spacings. However, the
ance properties of the materials containing the APP/PER formu- addition of all clays, regardless of their basal spacings, led to an
lation TG-FTIR analyses were also performed. This technique increase in the thermal stability of the char formed, maintaining
was employed in order to investigate whether the clays have any a residue of about 8% at 900 C. A similar result was obtained
influence on the gas phase reactions. Figure 6 shows the curves in a previous study.15 No significant changes in the decomposi-
obtained through TG. tion temperatures were detected when using clays with different
d-spacings.
The thermograms show that the addition of the intumescent
formulation increases the degradation temperature compared to In order to identify the volatile substances released during the
that of the pure polymer. Between 220 and 500 C mass losses process of thermal decomposition of the materials studied, TG-
occur in samples containing APP/PER which are related to char FTIR analyses were performed. Figure 7 shows the absorbance
formation reactions. At temperatures below 280 C, a mixture of spectra of the 3D FTIR overheating. The most evident band for
esters are formed by the reaction of acid species, such as phos- all samples appears only at temperatures above 400 C at
phoric acid and orthophosphates generated by the degradation approximately 2360 cm21 and can be attributed CO2 mole-
of APP, and the carbonific agent PER. From 280 C on, the car- cules.25 For the samples containing Cloisites 15A and 30B, at
bonization process occurs through the formation of double 400 C, there is also the appearance of band at 2940 cm21 that
bonds, generated by the thermal degradation of the esters is more intense than in the other samples and that is related to
formed in the previous step, followed by Diels–Alder reactions the CH2 groups present in aliphatic hydrocarbons.25 Although
and free radical processes, which will increase the polyaromatic there is some variation in peak intensity, it appears that all

1764 J. APPL. POLYM. SCI. 2013, DOI: 10.1002/APP.39349 WILEYONLINELIBRARY.COM/APP

 ARTICLE

Figure 7. FTIR 3D spectra over temperature: (A) EBA-30 1 APP–PER;(B) EBA-30 1 APP/PER 1 Cloisite Na; (C) EBA-30 1 APP/PER 1 Cloisite 30B;
and (D) EBA-30 1 APP/PER 1 Cloisite 15A.

samples produced the same bands, indicating no significant dif- correlated with C–H in C5C groups,26 indicating the formation
ferences between the products in the gas phase. Therefore, in of unsaturated carbon compounds. Considering that this band
principle, the variation of the basal spacings in clays does not is move evident at higher temperatures, it is more likely to
influence the formation of volatile products during thermal derive from the decomposition of the polycondensed aromatic
decomposition, nor gives rise to new reactions at the gas phase. species of the carbonaceous layer than from the formation of
However, for a more detailed analysis, it is necessary to analyze
unsaturated char precursors.
the 2D FTIR spectra obtained at certain temperatures, chosen
according to the TG curves in Figure 6. Figure 8 shows the The bands at 2940, 2865, and 1460 cm21 correspond to the
curves of the thermal degradation products at 200, 400, 600, stretching of C–H in CH2 groups of aliphatic hydrocarbons.26–
28
and 800 C, including the temperature ranges where the char is These compounds can be generated by the degradation of the
formed (200–400 C), stabilized (600 C), and degraded (800 C). polymer matrix at 400 C, as shown in Figure 6. The bands at
2360 and 670 cm21 correspond to CO2 molecules, and the
In Figure 8, there are no significant differences between the
bands at 2180 and 2105 cm21 correspond to CO. For all sam-
spectra of the EBA-30 1 APP–PER and clay containing samples,
ples, CO bands have the greatest intensity at 600 C, temperature
where all samples exhibit very similar bands. The trend
at which the char is already formed but has not yet degraded,
observed in 3D graphs (Figure 7) is maintained when analyzed
as observed in Figure 6. The presence of CO gas is an indication
through 2D spectra (Figure 8). Thus, the addition of clay does
of incomplete combustion due to the presence of the intumes-
not seem to lead to significant changes in the formation of vol-
cent layer. The band at 1740 cm21 in the samples containing
atile products during thermal decomposition, and there is no
clay corresponds to C5O stretching and indicates the volatiliza-
evidence of new gas phase reactions. Moreover, greater inten-
tion of products which contain the carbonyl group of carboxylic
sities of absorbance are obtained at 400 C, where the char is
acids.26,27 It is worthwhile to note that the phosphocarbona-
fully formed, and at 600 C, where its degradation begins, as
ceous structures, suggested by Bourbigot et al.18,29 presented
shown in Figure 6.
carboxylic acid carbonyl groups attached to polyaromatic struc-
Looking in more detail at the spectra obtained (Figure 8), all tures. The band at 1520 cm21 can be attributed to the angular
samples showed bands in the 3700–3900 cm21 region which deformation of the N–H group. Finally, the bands at 965 and
can be attributed to hydroxyl groups from the water generated 920 cm21 correspond to ammonia molecules generated by
by the thermal degradation of [0]the sample, for example.26,27 the degradation of ammonium polyphosphate and that function
A band at 3240 cm21 is evident for all samples and can be as the expansion gas for the intumescent layer. In samples

WWW.MATERIALSVIEWS.COM WILEYONLINELIBRARY.COM/APP J. APPL. POLYM. SCI. 2013, DOI: 10.1002/APP.39349 1765

 ARTICLE

Figure 8. FTIR curves of the thermal degradation products at temperatures 200, 400, 600, and 800 C, obtained by TG-FTIR (A) EBA 1 APP/PER; (B)
EBA 1 APP/PER 1 Cloisite Na; (C) EBA 1 APP/PER 1 Cloisite 30B; and (D) EBA 1 APP/PER 1 Cloisite 15A.

containing organophilic clays, the intensity of these bands is According to Bourbigot et al.30 the infrared frequency range of
greater, probably due to the quaternary ammonium salts gener- 850–1350 cm21 corresponds to different types of P–O bonds,
ally used as modifying agents that also release ammonia on and allows the characterization of phosphoric species, confirm-
thermal degradation. ing the presence of a phosphocarbonaceous structure. Another
region of interest includes the 2500–3700 cm21 range which
From Figure 8, it is also possible to verify that there is no
corresponds to the absorption of aliphatic groups and allows
involvement of phosphorous groups in the gas phase, since no
the observation of the participation of the polymer matrix in
bands related to the P5O (1268 cm21) and P–O (1031 cm21)
the formation of the intumescent layer. Thus, for a better of vis-
groups were identified.26 Thus, the reactions involving the phos-
ualization of the spectra, Figure 9 shows these two bands in the
phorous groups for the formation of the intumescent layer sys-
wavelength ranges mentioned above.
tems must have all occurred in the condensed phase.
Analyzing the first range of 2500–3700 cm21, two bands are
Analysis of the Condensed Phase—FTIR. In order to observe clearly identified at 2850 and 2920 cm21 in all samples at 30,
the action of clays with different basal spacings in the con- 280, and 350 C. These bands correspond to the symmetrical
densed phase, mixtures containing these clays and APP/PER in axial deformation and asymmetrical axial deformation of C–H
the polymer matrix were analyzed by FTIR before (30 C) and bonds in polyethylenic chain, respectively.30–32 The presence of
after heating the samples at 280, 350, 430, and 560 C. The spec- this group provide greater flexibility to the intumescent layer,
tra obtained are shown in Figure 9. connecting the polyaromatic structures which are being

1766 J. APPL. POLYM. SCI. 2013, DOI: 10.1002/APP.39349 WILEYONLINELIBRARY.COM/APP

 ARTICLE

Figure 9. FTIR analyses of the samples produced (30 C) and of their residues of at 280, 350, 430, and 560 C: (A) EBA 1 APP/PER; (B) EBA 1 APP/
PER 1 Cloisite Na; (C) EBA 1 APP/PER 1 Cloisite 30B; and (D) EBA 1 APP/PER 1 Cloisite 15A.

formed.18,19 A flexible intumescent layer is able to withstand the Analyzing the bands with wavenumbers between 850–1350
tensions generated by the release of the gases formed by the cm21, the 1100–1300 cm21 range shows a band at 1160 cm21
burning material, causing a better encapsulation of these gases. that corresponds to the P–O–C stretching mode in phosphate
However, at 350 C the intensity of these bands is greater for carbon complexes.30,32–34 The EBA-30 1 APP/PER sample
samples containing Cloisites Na and 30B considering the ratio presents this band at 280 and 350 C, but at 560 C, it is no lon-
between the bands. The sample containing Cloisite 15A, with ger observed. Samples containing Cloisites Na and 30B not only
higher basal spacing, showed intensities very close to those of presented these bands at 280 to 350 C, but also maintained
the EBA-30 1 APP/PER sample. Hence, the presence of greater them at temperatures above 430 C. However, a different behav-
amounts of these fragments at 350 C suggests that Cloisites Na ior was observed in the sample containing Cloisite 15A. In this
and 30B may give rise to protective layers with better shielding mixture, the 1160 cm21 band was not present at 280 C, and
properties and that are able to retard the degradation of these only became clearly visible at 350 C. Nevertheless, once formed,
fragments at higher temperatures. this band remained identifiable up to 560 C.

WWW.MATERIALSVIEWS.COM WILEYONLINELIBRARY.COM/APP J. APPL. POLYM. SCI. 2013, DOI: 10.1002/APP.39349 1767

 ARTICLE

Figure 10. XRD curves of the samples that did not undergo heat treatment (30 C).

FTIR bands in the 980–1020 cm21 range can be observed for all materials containing Cloisite 15A. Intumescence is a complex
the samples heated from 280 to 560 C, and correspond to the process in which all reactions must occur at a specific tempera-
symmetrical axial deformation of PO2 and PO3 in complex car- ture and time to ensure the formation of a protective shield
bon phosphates.30 with optimal properties.19,37,38 In the process, ester formation
and APP decomposition must follow a certain rate in order to
Analyzing the aforementioned results, it is worth noting that
produce ammonia at the moment where the carbonaceous
the EBA-30 1 APP/PER sample presents no P–O–C bonds at
structure has adequate properties to withstand swelling.19,38
560 C. This suggests that the layer becomes more rigid by the
scission of P–O–C bonds and condensation of the polyaromatic Analysis of the Condensed Phase—XRD. The intumescent
structure. The rigid layer is more brittle and creates cracks on layer is a very heterogeneous material, composed of gaseous
its surface, exposing the material to the flame and permitting its products stored in the phosphocarbonaceous structure of the
complete degradation. This suggestion is consistent with the TG condensed phase. This phase is a mixture of liquid (tar) and
results shown in Figure 6. solid phases, and has dynamic properties that allow the storage
of gaseous and liquid products from the degradation of the
The addition of clays, regardless of their basal spacings, led to
polymer.18
the stabilization of the P–O–C bonds up to 560 C. Therefore,
the addition of clays prevents the complete scission of the P–O– The carbonaceous fraction of the condensed phase consists
C bonds, leading to a more flexible intumescent layer. The TG of polyaromatic species arranged in stacks and the phospho-
curve shown in Figure 6 indicates that the clays did lead to the carbonaceous material consists of macromolecules and crys-
formation of a more thermally stable intumescent layer. The talline cells of polyaromatic chemical structures connected by
possible formation of aluminophosphoric complexes could assist polymeric chain fragments and phosphate groups. Besides
in stabilizing the phosphocarbonaceous structure of the char, this crystalline phase, the phosphocarbonaceous material
therefore increasing its thermal stability.30,35,36 has an amorphous phase surrounding the crystalline
domains.18,38
However, the addition of Cloisite 15A, which has the largest ba-
sal spacing among the clays studied, despite stabilizing the P– This amorphous phase consists of small polyaromatic molecules,
O–C bonds to higher temperatures, delayed their formation to alkyl chains formed by the degradation of the additives and
350 C. For all other samples, including the EBA-30 1 APP/ polymer chain fragments. This phase must be large enough to
PER, the 1160 cm21 band was already present at 280 C. This completely coat the crystalline domains. Finally, the intumescent
result suggests that increasing the basal spacings of the clays layer must have an appropriate balance between stiffness and
hinders the formation of the phosphocarbonaceous structure viscosity that will give the dynamic properties of interest and
that produces char. This hypothesis is consistent with the afore- that will prevent dripping, and also accommodate the tensions
mentioned assumption that the greater distance between the sil- generated by solid particles and from the pressure of the gases
icate layers in the clays could prevent the activation of the p produced by the burning process18,38,39
bond for the aromatization reactions that lead to char
In order to observe changes in the crystalline phase of the intu-
formation.19,28,37
mescent layer with the addition of clays with various basal spac-
The delay in the formation of the intumescent layer, compared ings, XRD analysis was performed on the samples before and
to the other samples, could explain the poor performance of the after heating up to 280, 350, 430, and 560 C.

1768 J. APPL. POLYM. SCI. 2013, DOI: 10.1002/APP.39349 WILEYONLINELIBRARY.COM/APP

 ARTICLE

Figure 11. XRD curves of the samples that underwent heat treatment at 280 C.

Figures 10–14 show the XRD patterns obtained for the previ- basal spacing, showed the same profile as the EBA-30 1 APP/
ously heated samples at 280, 350, 430, and 560 C and that of PER sample at the respective temperatures. Thus, in this tem-
the non-heated sample (30 C). It can be observed that at 280 C perature range, the addition of clays with various basal spacing
(Figure 11) and 350 C (Figure 12), there is a change in the does not influence the crystalline structure of the char.
crystalline structure, since the curve’s profiles are different from
At 430 C (Figure 13), there is a significant decrease in the
that of the non-heated sample (Figure 10). This change in the
intensities of the diffraction patterns for all samples. In addi-
crystalline organization is related to the formation of the intu-
tion, at this temperature, all clay containing samples showed the
mescent layer. More specifically, it corresponds to crystalline
same profile. However, this profile differs from that of the EBA-
phase changes in the intumescent layer which, as mentioned
30 1 APP/PER sample. These results suggest that temperature
before, is composed of both an amorphous and a crystalline
increase causes a decrease in the crystalline organization of the
phase. In this complex medium, crystalline cells of polyaromatic
structure. Furthermore, in EBA-30 1 APP/PER, it appears that
structures are linked together by fragments of polymer chain
the crystalline structure of the phosphocarbonaceous material is
and by phosphate groups.18
destroyed, coinciding with the P–O–C bonds disappearance at
Although the diffractogram profiles altered from 280 C to this temperature as showed in FTIR analysis [Figure 9(a)].
350 C, indicating changes in the crystalline structure due to From these results, it is possible to infer that clay addition pre-
temperature, all clay containing samples, regardless of their serves the crystalline structure and the P–O–C bonds at this

Figure 12. XRD curves of the samples that underwent heat treatment at 350 C.

WWW.MATERIALSVIEWS.COM WILEYONLINELIBRARY.COM/APP J. APPL. POLYM. SCI. 2013, DOI: 10.1002/APP.39349 1769

 ARTICLE

Figure 13. XRD curves of the samples that underwent heat treatment at 430 C.

temperature [Figure 9(B–D)]. No significant differences were in relation to the basal spacings of the clays used. Bourbigot
observed regarding the basal spacing of the clays studied. et al.30 in studying an ethylenic terpolymer containing APP/PER
Figure 14 shows a significant increase in peak intensities at and zeolite 4A by Nuclear Magnetic Resonance (NMR) of Al27,
560 C for the samples containing clays, whereas the EBA-30 1 Si29, and P31, observed an increase in the thermal stability of
APP/PER peaks did not undergo this increase. Once again, the the char formed with the zeolite, and attributed this effect to
diffraction patterns observed for all clay containing samples the formation of organic aluminosilicophosphate complex dur-
were very similar, regardless of the basal spacing of the clays, ing burning. The production of these species would reduce the
but differed from that of the sample without the clay. These scission of P–O–C bonds and therefore the increase in the size
results indicate that the presence of the clays lead to a more of polyaromatic stacks. This fact would justify the increase of
organized crystalline phase, but the presence of P–O–C bonds, the organization of the crystalline structure in samples with
still detected at 560 C [Figure 9(B–D)] confers flexibility to the clays and submitted to 560 C. The aluminophosphate species
protective layer and stabilizes it at high temperatures. [0]Yet could also increases the efficiency of the intumescent char
again, no differences were observed in the crystalline structure shield.40

Figure 14. XRD curves of the samples that underwent heat treatment at 560 C.

1770 J. APPL. POLYM. SCI. 2013, DOI: 10.1002/APP.39349 WILEYONLINELIBRARY.COM/APP

 ARTICLE

CONCLUSIONS 12. Huang, N. H.; Chen, Z. J.; Wang, J. Q.; Wei, P. Express.
Polym. Lett. 2010, 4, 743.
Cone calorimetry revealed that the addition of clays with basal
spacings greater than 30 Å for materials containing an EBA-30 13. Ribeiro, S. P. S.; Estev~ao, L. R. M.; Nascimento, R. S. V. J.
matrix and an APP/PER intumescent formulation lead to a loss Therm. Anal. Calorim. 2007, 87, 661.
of the synergistic action obtained with clays with smaller basal 14. Ribeiro, S. P. S.; Estev~ao, L. R. M.; Pereira, C.; Rodrigues, J.;
spacings, such as 13 and 22 Å. Nascimento, R. S. V. Polym. Degrad. Stab. 2008, 94, 421.
The SEM analysis results showed that the addition of Cloisite 15. Ribeiro, S. P. S.; Estev~ao, L. R. M.; Csaba, N.; Nascimento,
15A with a basal spacing greater than 30 Å, led to the formation R. S. V. J. Therm. Anal. Calorim. 2011, 106, 535.
of an intumescent layer with a morphology very similar to the 16. Ribeiro, S. P. S.; Estev~ao, L. R. M.; Nascimento, R. S. V. Sci. Tech-
one of the sample containing only APP–PER. These layers are nol. Adv. Mater. 2008, 9, doi:10.1088/1468-6996/9/2/024408.
less homogeneous and structured than those containing clays 17. Lewin M. In Fire Retardancy of Polymers: The Use of Intu-
with basal spacings of 13 or 22 Å. mescence; Le Bras, M., Camino, G., Bourbigot, S., Delobel,
R., Eds.; The Royal Society of Chemistry: London, 1998;
The results of the TG-FTIR analysis suggest that clays act exclu-
Chapter 1, pp 3–32.
sively in reactions taking place in the condensed phase. The
FTIR analyses of the residues obtained from the heating treat- 18. Bourbigot, S.; Duquesne, S. In Flame Retardant Polymer
ments employed indicate a delay in the formation of the phos- Nanocomposites Intumescence and Nanocomposites; Mor-
gan, A. R., Wilkie, C. A., Eds.; Wiley: New Jersey, 2007;
phocarbonaceous species when clays with larger basal spacings
Chapter 6, pp 131–162.
are added. This effect can justify the loss of synergy and the dif-
ferences in morphology of the char formed. 19. Bourbigot, S.; Duquesne, S. J. Mater. Chem. 2007, 17, 2283.
20. Soma, Y.; Soma, M. Environ. Health Perspect. 1989, 83, 205.
Finally, XRD results indicate that the presence of clays in the materi-
als promoted changes in the crystalline phase of the char when the 21. Adams, J. M.; Dyer, S.; Martin, K.; Matear, A.; McCabe, R.
samples were submitted to temperature increase. Those changes W. J. Chem. Soc. Perkin Trans. 1994, 1, 761.
allowed for the maintenance of the structures at higher temperatures. 22. Laszlo, P. Science 1987, 235, 1473.
23. Weiss, A. Angew. Chem. Int. Ed. Engl. 1981, 20, 850.
24. Koster, R. M.; Borget, M.; de Leeuw, B.; Poels, E. K.; Bliek,
REFERENCES A. J. Mol. Catal. A: Chem. 1998,134, 159.
1. Laoutid, F.; Bonnaud, L.; Alexandre, M.; Lopez-Cuesta, J. 25. Miller, T. W. J. Therm. Anal. Calorim. 2011, 106, 249.
M.; Dubois, Ph. Mater. Sci. Eng. R. Rep. 2009, 63, 100. 26. Yi, L.; Jiang Song, Y.; Xufu, C. J. Therm. Anal. Calorim.
2. Kiliares, P.; Papaspyrides, C. D. Prog. Polym. Sci. 2010, 35, 902. 2012, 107, 1191.
3. Jimenez, M.; Duquesne, S.; Bourbigot, S. Surf. Coat. Technol. 27. Wang, K.; Lv, P.; Hu, Y.; Hu, K. J. Anal. Appl. Pyrol. 2009,
2006, 201, 979. 86, 207.
4. Le Bras, M.; Bourbigot, S. In Fire Retardancy of Polymers: 28. Herrera, M.; Mattuschek, G.; Kettrup, A. J. Therm. Anal.
The Use of Intumescence; Le Bras, M., Camino, G., Bourbi- Calorim. 2000, 59, 385.
got, S., Delobel, R., Eds.; The Royal Society of Chemistry: 29. Bourbigot, S.; Le Bras, M.; Delobel, R. Carbon 1993, 31, 1219.
London, 1998; Chapter 3, pp 64–75 30. Bourbigot, S.; Le Bras, M.; Delobel, R. J. Chem. Soc. Faraday
5. Bourbigot, S.; Le Bras, M.; Breant, P.; Tremillon, J. M.; Trans. 1996, 92, 3435.
Delobel, R. Fire Mater. 1996, 20, 145. 31. Liu, W.; Chen, D.-Q.; Wang, Y.-Z.; Wang, D.-Y.; Ming-Hai,
6. Bourbigot, S.; Le Bras, M.; In Fire Retardancy of Polymers: Q. Polym. Degrad. Stab. 2007, 92, 1046.
The Use of Intumescence; Le Bras, M., Camino, G., Bourbi- 32. Le Bras, M.; Bourbigot, S.; Revel, B. J. Mater. Sci. 1999, 34, 5777.
got, S., Delobel, R., Eds.; The Royal Society of Chemistry:
33. Thomas, L. C.; Chittenden, R. A. Spectrochim. Acta 1964,
London, 1998; Chapter 13, pp 222–235.
20, 489.
7. Shumao, L.; Hua, Y.; Tao, Y.; Weizhong, Y.; Jie, R. Polym.
34. Yi, J.; Liu, Y.; Pan, D.; Xufu, C.. J. Appl. Polym. Sci. 2012,
Adv. Techol. 2009, 20, 1114.
127, 1061.
8. Yun, L.; Jun-Sheng, W.; Cheng-Liang, D.; De-Yi, W.; Yan-
35. Tang, Y.; Hu, Y.; Li, B.; Liu, L.; Wang, Z.; Chen, Z.; J. Polym.
Peng, S.; Yu-Zhong, W. Polym. Adv. Techol. 2010, 21, 789.
Sci. Part A: Polym. Chem. 2004, 42, 6163.
9. Estev~ao, L. R. M.; Nascimento, R. S. V.; Le Bras, M.; Delo-
36. Tang, Y.; Hu, Y.; Wang, S.; Gui, Z.; Chen, Z. Polym. Int.
bel, R. In Fire Retardancy of Polymers New Application on
2003, 5, 1396.
Mineral Fillers; LeBras, M., Wilkie, C. A., Bourbigot, S.,
Duquesne, S., Jama, C., Eds.; The Royal Society of Chemis- 37. Antonov, A. V. Russ. Chem. Rev. 1999, 68, 605.
try: Cambridge, 2005; Chapter 23, pp 313–326. 38. Wei, P.; Hao, J.; Du, J.; Han, Z.; Wang, J. J. Fire Sci. 2003, 21, 17.
10. Estev~ao, L. R. M.; Nascimento, R. S. V. Polym. Degrad. Stab. 39. Shartel, B.; Wei, A.; Mohr, F.; Kleemeier, M.; Hartwig, A.;
2002,75, 517. Braun, U. J. Appl. Polym. Sci. 2010, 118, 1134.
11. Estev~ao, L. R. M.; Le Bras, M.; Delobel, R.; Nascimento, R. 40. Chen, Y.; Fang, Z.; Yang, C.; Wang, Y.; Guo, Z.; Zhang, Y. J.
S. V. Polym. Degrad. Stab. 2005, 88, 444. Appl. Polym. Sci. 2010, 115, 777.

WWW.MATERIALSVIEWS.COM WILEYONLINELIBRARY.COM/APP J. APPL. POLYM. SCI. 2013, DOI: 10.1002/APP.39349 1771