Thermal degradation kinetics of semi-interpenetrating polymer network based on polyurethane and siloxane

Article abstract of DOI:10.1016/j.tca.2013.03.015

A series of semi-interpenetrating polymer networks based on linear polyurethane and 1,3-bis(glycidyloxypropyl)-1,1,3,3-tetramethyldisiloxane, which was cross-linked with diethylenetriamine, were obtained. The samples have been submitted to thermal stability investigations at non-isothermal conditions in nitrogen. The kinetic parameters of the degradation process were determined both by isoconversional methods of Friedman and Ozawa-Flynn-Wall as well as by model fitting multivariate non-linear regression method. It was observed that the thermal stability of the semi-interpenetrating polymer networks depends on siloxane content. The best fit of the f(α) function with the experimental data was found for three-step degradation mechanism. Successive stages of thermal decomposition occurred by diffusion, phase-boundary and autocatalysis models, respectively.

Full text of DOI:10.1016/j.tca.2013.03.015

Accepted Manuscript
Title: Thermal degradation kinetics of semi-interpenetrating
polymer network based on polyurethane and siloxane
Author: <ce:author id="aut0005"> Łukasz
Byczy n´ ski<ce:author id="aut0010"> Michał
Dutkiewicz<ce:author id="aut0015"> Hieronim Maciejewski
PII:
S0040-6031(13)00149-4
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.tca.2013.03.015
TCA 76395
To appear in:
Thermochimica Acta
Received date:
Revised date:
Accepted date:
30-1-2013
4-3-2013
12-3-2013
Please cite this article as: Ł. Byczy n´ ski, M. Dutkiewicz, H. Maciejewski, Thermal
degradation kinetics of semi-interpenetrating polymer network based on polyurethane
and siloxane, Thermochimica Acta (2013), http://dx.doi.org/10.1016/j.tca.2013.03.015
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

1
Thermal degradation kinetics of semi-interpenetrating polymer network
based on polyurethane and siloxane
2
3
4
a,*
b,c
b,c
Łukasz Byczyński , Michał Dutkiewicz , Hieronim Maciejewski
5
6
7
8
9
0
1
2
3
4
5
a
Faculty of Chemistry, Rzeszow University of Technology, Al. Powstańców Warszawy 6,
35-959 Rzeszow, Poland
b
Faculty of Chemistry, Adam Mickiewicz University, Grunwaldzka 6, 60-780 Poznan,
Poland
c
1
1
1
1
1
1
Poznań Science and Technology Park, A. Mickiewicz University Foundation, Rubież 46, 61-
612 Poznan, Poland
*
Corresponding author. (Ł. Byczyński) E-mail address: lbyczynski@prz.edu.pl, Tel.: +48 17
8651440.
16
17
18
19
20
21
22
23
24
25
26
27
Abstract
A series of semi-interpenetrating polymer networks based on linear polyurethane and
1,3-bis(glycidyloxypropyl)-1,1,3,3-tetramethyldisiloxane, which was cross-linked with
diethylenetriamine, were obtained. The samples have been submitted to thermal stability
investigations at non-isothermal conditions in nitrogen. The kinetic parameters of the
degradation process were determined both by isoconversional methods of Friedman and
Ozawa-Flynn-Wall as well as by model fitting multivariate non-linear regression method. It
was observed that the thermal stability of the semi-interpenetrating polymer networks
depends on siloxane content. The best fit of the f(α) function with the experimental data was
found for three-step degradation mechanism. Successive stages of thermal decomposition
occurred by diffusion, phase-boundary and autocatalysis models, respectively.
2
2
3
8
9
0
Key words
Polyurethane; Siloxane; Thermal degradation; Kinetics; Semi-interpenetrating polymer
network.
31
1. Introduction
3
3
3
3
2
3
4
5
Polyurethanes (PUs) are segmented polymers consisting of alternating hard and soft
segments. Owing to their structure, PUs offer very good elasticity with reasonably high
mechanical strength and abrasion resistance at the same time, and also controllable hardness
[1]. These polymers are used in wide variety of application, such as foams, elastomers,
1
Page 1 of 25

36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
adhesives, and coatings [1]. Polyurethanes are commonly known to be thermally stable up to
180 C, and those based on polyesters are more stable then those based on polyethers.
Thermal decomposition is initiated in hard segments where the thermally weakest urethane
linkages are predominant, and it terminates in elastic segments, which comprise ester or ether
bonds. [2]. To improve the thermal resistance of PUs different modifications were employed
e.g. by introduction of heterocyclic structures such as isocyanurate, oxazolidone, imide,
triazine, pyridine or phosphazene [2,3]. One of the known method for improving the thermal
stability of PUs is their modification by siloxane resulting in inorganic-organic hybrid
materials [4-7]. Polysiloxanes offer high thermal stability up to 300 °C. That resistance results
from the presence of Si-O bonds, for which the dissociation energy is higher (ca. 460.5 kJ
-1
-1
-1
-1
mol ) in relation to C-O (358 kJ mol ), C-C (304 kJ mol ) and C-NH (98 kJ mol ) bonds
[2,8,9]. Moreover, those polymers show low free surface energy and glass temperature, but
they have poor tensile strength and abrasion characteristic [10].
The chemical modification of polyurethanes by siloxane is mostly achieved by
introduction of usually linear polydimethylsiloxane into PU backbone as soft segments or as a
part of them [4-6]. However, the investigations on polyurethanes with other polysiloxane
structures including polyhedral oligomeric silsequioxane (POSS) have been recently
developed [7,11-14]. The thermal degradation studies were also carried out for poly(urethane-
siloxane) networks obtained by sol-gel process [15], by cross-linking with hyperbranched
polyester [16], or by moisture curing [17].
So far, no papers can be found for detailed studies on kinetics of the thermal stability
of polyurethane–siloxane semi-interpenetrating polymer networks (semi-IPNs). In this kind of
hybrid material linear polyuretahane is entangled at molecular scale in the network of
crosslinked siloxane. Due to the lack of chemical linkage between siloxane and urethane unit,
this thermosetting material is expected to provide different thermal characteristic than
crosslinked poly(urethane-siloxane) copolymer. Hence, the aim of this work is to investigate
the effect of the composition of semi-IPNs composed of polyurethane and crosslinked 1,3-
bis(glycidyloxypropyl)-1,1,3,3-tetramethyldisiloxane on thermal stability, including detailed
kinetic studies. The results of the kinetic investigations may be useful to determine the
lifetime of semi-IPNs exposed to different temperatures.
2
Page 2 of 25

6
6
6
7
2. Experimental
2.1. Materials
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
Isophorone diisocyanate (IPDI), diethylenetriamine (DETA) and dibutyltin dilaurate
(DBTDL) from Aldrich were used without further purification. Poly(oxytetramethylene)diol
(PTMO, M = 2000) was purchased from Aldrich and was dried in a vacuum oven at 105 °C
n
before use. 1,4-butanediol (BD, Aldrich) used as a chain extender was dried over 4 Å
molecular sieves. 2-butanone (MEK, POCh, Poland) were distilled and dried over 4 Å
molecular sieves as well. 1,3-Bis(glycidyloxypropyl)-1,1,3,3-tetramethyldisiloxane (GPSil)
was prepared according to the following procedure: 1,1,3,3-Tetramethyldisiloxane and 1.1
equiv of allyl glycidyl ether (calculated to each Si–H group) were placed into the reaction
vessel and heated up to 90 °C. Then the appropriate amount of [{Rh(OSiMe )cod} ] (in the
3
2
-5
ratio 10 mol per mol Si–H) was added as a catalyst, and the reaction was continued for the
next 2 h to give a crude product, which was purified by vacuum distillation, and the fraction
was collected at 184 °C/2 mm Hg. The formation of the desired product was verified by NMR
1
analysis: H NMR: (C
6
D
6
, 298 K, 300 MHz) d(ppm): 0.05 (12H, SiCH ); 0.53 (4H, SiCH
3
2
–);
1.60 (4H, SiCH
2
CH
2
–); 2.20, 2.21 (t, 2H, CH –CHO); 3.12 (m, 1H, CH –CHO); 3.26 (m, 2H,
2
2
13
OCH CH ); 3.38 (m, 2H, CHCH O); C NMR: (C D , 298 K, 79 MHz) d(ppm): 0.02
2
2
2
6
6
(SiCH
3
); 14.6 (SiCH
2
–); 24.0 (–CH
2
–); 43.6 (CH
2
2
–CHO); 50.7 (CH –CHO); 71.8
29
(OCH CH ); 74.3 (CHCH O); Si NMR: (C D , 298 K, 59.61 MHz) d(ppm): 7.8
2
2
2
6
6
3 2
(OSi(CH ) ).
86
2.1.1. Synthesis of polyurethane (PUR)
87
88
89
90
91
92
93
94
The polyurethane (PUR) was obtained in one step solvent synthesis in MEK. First
PTMO 2000 (22.00 g) and BD (4.96 g) were diluted in MEK (97.13 g) in 250 ml three-
necked flask equipped with a mechanical stirrer, thermometer, reflux condenser and nitrogen
inlet. Then IPDI (14.67 g) was added drop by drop to the flask and DBTDL as catalyst (0.2
o
wt% with reference to PTMO) was added. The reaction was allowed to proceed at 70 C till
the unreacted –NCO content reached 0 %. The conversion of –NCO at each step was
determined by a standard dibutylamine back titration method every 30 min [18]. The molar
ratio of IPDI:PTMO:BD was 6:1:5.
3
Page 3 of 25

95
2.1.2. Synthesis of siloxane network (GS)
96
97
98
99
1,3-Bis(glycidyloxypropyl)-1,1,3,3-tetramethyldisiloxane (GPSil) was mixed with
diethylenetriamine (DETA) at equimolar ratio of epoxy groups to reactive hydrogen atoms
from amine (–CHCH O : –NH = 1:1). The mixture was homogenized with the use of
2
mechanical stirrer at ambient temp. for 10 min, poured on polytetrafluoroethylene plate and
100 cured at 110 °C for 6 h in an circulating oven.
1
01 2.1.3. Synthesis of semi-interpenatrating polymer networks (PGS) with PUR and GS
1
1
1
1
1
1
1
1
1
1
1
1
02
The semi-IPNs were obtained with the use of polyurethane and 1,3-
03 bis(glycidyloxypropyl)-1,1,3,3-tetramethyldisiloxane (GPSil) that was cured with
04 diethylenetriamine (DETA). Appropriate amount of PUR was completely diluted in MEK
05 followed by the addition of GPSil. The mixture was homogenized with the use of mechanical
06 stirrer at ambient temp. for 10 min. Then DETA at equimolar ratio of epoxy groups to
07 reactive hydrogen atoms from amine (–CHCH O : –NH = 1:1) was added. The mixture was
2
08 once again homogenized for the next 10 min, poured on polytetrafluoroethylene plate and
09 cured at 110 °C for 6 h in an circulating oven. Using above described procedures 3 samples of
10 PGS were obtained. The detailed composition of semi-interpenatrating polymer networks is
11 listed in Table 1 and the structures of the compounds are presented in Scheme 1.
12
13
Scheme 1
Table 1
1
1
14 2.2. Methods
15 2.2.1. NMR spectroscopy
1
1
1
1
16
H NMR spectrum of the polyurethane was recorded with the use of the spectrometer
II
17 FT NMR Bruker Avance 500 . The sample of PUR was dissolved in CDCl
3
and the solution
-3
18 with the concentration of about 0.2 g dm was prepared. TMS was used as a standard.
1
19 2.2.2. IR spectroscopy
1
1
1
1
20
IR spectra were recorded with the spectrophotometer Bruker Tensor 27, within the
–
1
21 range of 4000 – 650 cm , with the use of ATR technique. FT-IR spectra were recorded on a
22 Bruker Tensor 27 Fourier transform spectrophotometer equipped with a SPECAC Golden
-1
23 Gate diamond ATR unit. In all cases 16 scans at a resolution of 2 cm were collected to
4
Page 4 of 25

-1
1
24 record the spectra in a range of 4000 – 650 cm . The obtained spectra were presented as the
–1
1
25 relation of absorbance versus wave number [cm ].
1
26 2.2.3. Gel permeation chromatography (GPC)
1
1
1
1
1
1
1
1
27
GPC chromatograms were performed using a Waters Alliance 2695 GPC system
28 equipped with a Waters 2414 RI detector and a set of three serially connected 7.8 × 300 mm
29 columns (Waters Styragel HR1, HR2 and HR4). The pore sizes of columns are as follows: 10,
30 100 and 1000 Å. Molecular weights and polydispersity indices were calculated on the basis of
3
31 point to point calibration curve of polystyrene Shodex standards in the range from 1.31×10 to
6
-1
32 3.64×10 Da. Tetrahydrofuran (THF) was used as an eluent in a 0.6 mL min isocratic flow.
33 Columns were maintained at 35 °C. All samples were prepared as ca. 10 wt% solutions in
34 THF.
1
35 2.2.4. Thermogravimetric analysis (TGA)
1
1
1
1
36
Thermogravimetric analysis was performed using a Mettler Toledo TGA/DSC1. The
37 TGA experiments have been carried out in nitrogen from 25 to 500 °C with varying heating
-1
38 rates 5; 10; 20 and 40 °C min . The measurement conditions were as follows: sample weight
-1
39 ~5 mg, gas flow 50 mL min , 150 μl open alumina pan.
1
40 2.2.5. Kinetic analysis of the decomposition process
1
1
1
41
In order to study thermal degradation of PGS semi-IPNs in detail, kinetic analysis of
42 the observed thermal decomposition processes of those polymers was performed.
43
44
The general expression for the kinetic description of degradation of solids is [19]:
d
dt
1
 k(T) f ()
(1)
1
1
45 where α is the conversion degree defined as the ratio of the actual weight loss to the total
46 weight loss, k(T) is the reaction rate constant, and f(α) is the kinetic model function. In non-
dT
147 isothermal condition at constant heating rate  
when the reaction rate constant is
dt
1
1
48 described by the Arrhenius equation, the conversion degree can be analyzed as a function of
49 temperature with Eq. (2).
d
dT
A

E 
 RT 
a
150

exp
 f ()
(2)

5
Page 5 of 25

1
1
1
1
1
51 where:
52 A – frequency factor (s ),
53 – activation energy (kJ mol ),
-1
-1
E
a
-3
-1 -1
54 R – universal gas constant (8.314·10 kJ mol K ),
55 T – absolute temperature (K).
1
1
1
1
1
1
56
Two isoconversional methods by Friedman [20] and Ozawa-Flynn-Wall (OFW)
57 [21,22] were employed to evaluate of kinetic parameters in thermal decomposition of
58 polymers under dynamic conditions in nitrogen. These methods may be used to determine and
59 to monitor changes in the activation energy during the degradation process, without
60 assumption of reaction model.
61
62
The Friedman isoconversional method [20] is based the following equation:
d
dt
Ea
1
ln
 ln A  ln f () 
(3)
RT
1
1
63 In order to find the apparent activation energy value (E
a
) for a given degree of conversion (α),
64 one should take a series of measurements at different heating rates β. Then, for a fixed degree


d 
  f (1/T) for which
dt 
1
65
66
of conversion (α = const), straight lines are obtained in the plot ln
Ea
1
the slope is defined as n  
.
R
1
1
67
The Ozawa-Flynn-Wall method is based on the Doyle’s approximation [23] and it
68 resolves itself to the use of the following equation:


A E 
Ea
a
169
ln   ln
  ln g()  5.3305 1.052
R 
(4)
RT
1
1
1
a
70 In order to find the apparent activation energy value (E ) for a given degree of conversion (α),
71 one should take a series of measurements at different heating rates β. Then, for a fixed degree
72 of conversion (α = const), straight lines are obtained in the plot lnβ = f(1/T) for which the
Ea
173
slope is defined as m 1.052
.
R
1
1
1
1
74
After the values of the activation energy have been calculated with the use of the
75 isoconversional methods, the determination of reaction model f(α) was performed using
76 model fitting multivariate non-linear regression method. The iterative procedures were used to
77 estimate kinetic triplet. Table 2 presents the specification of models, which were used in our
6
Page 6 of 25

1
1
1
78 calculations to describe thermal degradation of polymers. The degradation kinetics parameters
79 were evaluated with the use of the Netzsch Thermokinetic 3.1 software.
80
Table 2
1
1
81 3. Results and discussion
82 3.1. Chemical structure and molecular mass distribution of PUR
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
83
The H NMR and GPC studies were performed only for polyurethane. Owing to linear
84 structure PUR could be dissolved in solvents, which are commonly used to those analyses.
1
85
The H NMR analyses confirmed the predicted structure of polyuretahane. The
86 signals, which were present at δ = 0.93-1.07 ppm and 1.68 ppm, were assigned to protons
87 from -CH and -CH - groups from IPDI structure, respectively. The build-in of PTMO was
3
2
88 confirmed by the presence of the chemical shift at 3.41 ppm, which is assigned to protons
89 from methylene group in -O-CH -. The signal at 1.62 ppm is related to protons in -CH
2
-
2
90 group from PTMO and BD. Moreover the shift at δ = 4.08 ppm was assigned to the protons
91 from methylene group at urethane link -CH -OCONH-. which are presented both in PTMO
2
92 and BD. Those protons derive only from PTMO and BD compounds, which reacted with
93 IPDI.
94
The GPC results revealed that PUR is characterized by narrow polydispersity (M
w n
/M )
95 amounts to 2.15. The number average molecular mass (M
n
) and weight average molecular
96 mass (M ) are 20,339 and 43,811, respectively. This may suggest that PUR contains ca. 7 – 8
w
97 structural units of PTMO.
1
98 3.2. Structural analysis of polyurethane-siloxane semi-IPNs
1
2
2
2
2
2
2
2
2
2
99
The chemical structures of all synthesized samples were verified on the basis of
00 infrared spectroscopy analysis. The IR spectra of PUR, GS and semi-IPNs were presented in
01 Fig. 1.
02
03
Fig. 1.
-1
The absence of NCO peak at 2270 cm in PUR spectrum indicates that the isocyanate
04 conversion was complete. There are characteristic absorption peaks for polyurethanes, both in
-1
05 PUR and semi-IPNs spectra, at 1656 – 1699 cm (–C=O stretching, first amide band) and at
-1
06 1531 – 1534 cm (–NH deformation, second amide band). The intensity of both bands
07 decreases with decreasing amount of polyurethane in crosslinked siloxane (GS), which
08 confirms the presence of PUR in semi-IPNs.
7
Page 7 of 25

2
2
2
2
2
2
2
2
2
2
2
2
09
The occurrence of siloxane structures in PGS was confirmed by the presence of bands
-1
10 at about 1250 and 795 cm that derive from deformation vibrations in Si–CH
3
group.
11 Moreover the intensity of those bands increases with the increase of the crosslinked siloxane
12 in the semi-IPN structure.
-1
13
14 results from stretching vibrations of C-H bond in alkyl groups (-CH-,-CH
15 characteristic IR bands present both in PUR and semi-IPNs can be found in the range 3324 -
In the spectra of all the samples there is the band in the range 2796 - 2952 cm , which
2
-, CH -). Other
3
-1
-1
16 3391 cm (–NH and –OH stretching) and around 1040 - 1110 cm (-Si-O-Si- and/or -C-O-C-
17 bending).
18
In PUR sample the third amide band related to C-N stretching can be found at 1237
-
1
19 cm . In the other samples this band is overlapped by the more intense band of approx. 1250
-1
3
20 cm originating from the deformation vibration of C-H bond in Si-CH group.
2
21 3.3 Thermal degradation studies
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
22
The samples of semi-IPNs (PGS25, PGS50, PGS75), polyurethane as well as
23 crosslinked 1,3-bis(glycidyloxypropyl)-1,1,3,3-tetramethyldisiloxane were investigated by
24 thermogravimetric analysis to determine their thermal stability. TG and DTG curves recorded
-1
25 at 10 °C min in nitrogen were presented in Figure 2 whereas Table 3 provides interpretation
26 of both profiles.
27
28
29
Table 3
Fig. 2.
On the TG and DTG curves two basic degradation stages for all the samples can be
30 distinguished. This may suggest that the samples have at least two major stages of
31 degradation.
32
For the polyurethane, temperature of maximum weight loss at the first stage of
33 degradation (Tmax1) amounts to 310 °C and is related to the decomposition within the hard
34 segments, derived from IPDI and BD. The second stage of the PUR degradation is associated
35 with thermolysis of PTMO flexible segments and occurs at temperature of maximum weight
36 loss (Tmax2) equal to 391 °C. The mass loss (Δm) of both stages is similar and amounts to ca.
37 50 %, which corresponds to the amounts of hard and soft segments in the PUR composition.
38
39 380 °C than PUR. The mass loss at the first and second step of degradation for GS sample is
40 Δm = 9 % and Δm = 88.5 %, respectively. Similar two-stage decomposition was obtained
However, siloxane sample (GS) is characterized by lower Tmax1= 227 °C and Tmax2 =
1
2
8
Page 8 of 25

2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
41 for the siloxane network, which was prepared from 1,3-bis(glycidyloxypropyl)-1,1,3,3-
42 tetramethyldisiloxane, but cross-linked with 1,2-diaminoethane at equimolar ratio [25].
43
The thermal stability of the semi-IPNs depends on siloxane content. It was observed
44 that the increasing amount of GS resulted in the increase of both temperature of 5 % mass loss
45 (T ) from 276 °C for PUR to 299 °C for PGS75, and Tmax1. The formation of siloxane
5%
46 network greatly improved the thermal stability of neat PUR. The DTG curves of PGS display
47 not only maximum peaks (Tmax1 and Tmax2), but also a shoulder peak at the beginning of the
48 degradation. The shoulder peak corresponds to ca. 3 % of mass loss at 170 – 270 °C and
49 indicates a complex mechanism of degradation. Moreover, for PGS75 the peak related to
50 Tmax1 is shifted to higher temperature and overlapped with the peak of Tmax2, which hinders
51 proper determination of Tmax1. The second stage of degradation appears at 330 – 450 °C and
52 may correspond to the decomposition of PUR soft segment and GS network. At this stage the
53 highest Tmax2 = 427 °C showed the PGS25 sample, contained 25 wt% of GS. Further
54 increasing of the siloxane content resulted in a reduction of this temperature to 380 °C, which
55 was obtained for the comparative GS sample. The complexity of the mechanism was also
56 noticed at the second stage of degradation for PGS50 sample, which was manifested by the
57 presence of split peak on DTG curve. The mass loss of the first step of degradation (Δm
1
)
58 decreased whereas at the second stage Δm increased with the decrease of PUR. This may
2
59 suggest that at the first stage the degradation polyurethane predominates. The amount of solid
60 residue after degradation increased with the increase of siloxane. Those findings seem to
61 make the evidence for the formation of complex silicon-based structures in the thermolysis
62 process. The structures are formed on the surface and probably create the insulating layer,
63 which slows down further decomposition of the polymer, as it was observed for polyurethane-
64 siloxane copolymers [26].
2
2
65 3.4 Thermal degradation kinetics
66 3.4.1. Isoconversional methods
2
2
2
2
2
2
67
68 isoconversional as well as model fitting methods were employed. Figure 3 illustrates the
69 relationship between the activation energy (E ) of the synthesized samples and the degree of
70 conversion by Friedman and Ozawa-Flynn-Wall, at α values ranging between 0.05 and 0.95.
71 On Fig. 3a the increase in E with the increase of α was observed, which may result from
To study the thermal degradation in more detail, the kinetic analyses based on
a
a
72 complex mechanism of degradation. The presence of inflection point, for almost each sample
9
Page 9 of 25

2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
73 indicates, that degradation take place in two main steps. The activation energy of the first step
-1
74 of degradation calculated by Friedman method is in the range 150 – 175 kJ mol and at the
-1
75 second step exceeds 200 kJ mol . The inflection point shifts clearly to lower values of degree
76 of conversion with the decrease of PUR in PGS composition, from 0.45 to 0.35. However,
77 only for PGS75 sample the absence of the inflection point and the increase of the activation
-1
78 energy with the increase of α from ca. 150 to ca. 250 kJ mol is observed. This results from
79 the complex degradation mechanism, which was also seen on DTG curve as overlapped
80 peaks, related to the maximum weight loss of the first and second stage of degradation. With
a
81 the increase of polyurethane in semi-IPNs composition the decrease of E is observed,
82 resulting from the reduction of Si-O bonds with the large dissociation energy. Similar
83 conclusions can be driven from analysis of the results obtained by Ozawa-Flynn-Wall method
84 (Fig. 3b). However, the activation energy values in this case are lower than those calculated
85 by Friedman method. The differences between kinetic parameters calculated using Friedman
86 and Ozawa – Flynn – Wall methods were explained in literature [27].
87
Fig. 3.
2
88 3.4.2. Model-fitting methods
2
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
89
Based on the analysis of the results of the activation energy from isonversional
90 methods as well as the shape of profiles and isoconversional lines from Friedman plots, the
91 multivariate non-linear regression studies for finding the most probable kinetic model of
92 thermal degradation of the samples was employed. The degree of the best fitting was
93 determined with the use the regression coefficient. Table 4 summarizes the results of those
94 calculations for the three most probable degradation kinetics mechanisms. Moreover, Fig. 4
95 shows the comparison between experimental points and calculated curves obtained from
96 multivariate non-linear regression for the exemplary PGS50 sample.
97
98
99
Fig. 4.
Table 4
Both for the linear polyurethane (PUR) and crosslinked siloxane (GS) the best match
00 between the experimental data and the models considered was obtained for the two-stage
01 degradation mechanism. The activation energy of PU of the first and second step of
-1
-1
02 degradation amounts to ca. 158 kJ mol and 238 kJ mol , respectively. In case of GS sample
03 the first step of degradation occurs with high probability by diffusion mechanism. The E of
04 this stage is ca. 73 kJ mol . However, the second main stage may be described by n-th order
a
-1
1
0
Page 10 of 25

-
1
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
a
05 reaction (Fn) with definitely larger E ca. 270 kJ mol . The initial decomposition process may
06 be associated with the degradation that occurs due to the chain end scission of the polymer
07 network. At the second stage, chain ends are consumed and subsequently, the degradation
08 occurs toward the random chain scission, which shows higher values of activation energy
09 [28].
10
For semi-IPNs the best fit between the experimental data and the models was obtained
11 for the three-stage degradation mechanism, which was confirmed by the high correlation
12 coefficient above 0.9998. The first stage of thermal decomposition occurs with the relatively
-1
13 low activation energy in the range 50 - 65 kJ mol and reveals diffusion character (D2, D3,
14 D4). This stage may be related to the first step of GS degradation with similar E value. The
a
15 occurrence of this additional step is reflected by presence of shoulder peak at the beginning of
16 degradation only on DTG curve (Fig. 2b), which is associated with slight 3 % of mass loss.
17 The kinetic analyzes indicate that the main second step of degradation occurs by the Prout-
18 Tompkins autocatalytic reaction (Bna) and the third - by the reaction of n-th order with
19 autocatalysis (Cn). The increase of GS content in semi-IPN improved their thermal stability,
20 which was evidenced by the increase of the activation energy of the second stage of
-1
-1
21 degradation (Ea2) from ca. 163 kJ mol to ca. 175 kJ mol . These values are within the range
22 of the results determined by two employed isoconversional methods. Additionally, the results
23 obtained by Friedman indicated that the first essential stage of the degradation might occur by
24 the phase-boundary reaction. This is consistent with results of non-linear regression. The Bna
25 model is employed to describe the kinetics of the solid phase processes that occur at phase
26 boundary [28]. The equation has been used to study of polymer systems such as
27 nanocomposites [29] or modified epoxy resin network [30]. Since this stage of degradation is
28 related to the decomposition of urethane hard segments, it is possible that the thermal
29 degradation of PGS occurs at the phase boundary between PUR and GS. For the third stage of
30 decomposition of semi-IPNs the most probable kinetic model is the n-th order reaction with
31 autocatalysis (Cn). Processes related to the decomposition of soft polyurethane segments and
32 of the siloxane network occur simultaneously at this stage. This is evidenced by the consistent
a
33 results of activation energy of the third stage of semi-IPNs degradation (Ea3) and the E of the
-1
34 last stage of degradation of PUR and GS samples, exceeding the value of 230 kJ mol .
1
1
Page 11 of 25

3
35 4. Conclusions
3
3
3
3
3
3
3
3
3
3
3
36
In this work a series of semi-interpenetrating polymer networks based on linear
37 polyurethane and cross-linked 1,3-bis(glycidyloxypropyl)-1,1,3,3-tetramethyldisiloxane at
38 different weight ratios, were obtained. The formation of siloxane network greatly improved
39 the thermal stability of PUR, which was reflected in the increase of the temperature of 5 %
40 mass loss and the activation energy with the increasing amount of cross-linked siloxane.
41 Thermogravimetric studies carried out at four different heating rates showed that the thermal
42 degradation of semi-IPNs is basically a two-stage process. However, the kinetic analysis
43 based on model fitting multivariate non-linear regression method revealed three-stage
44 degradation mechanism. The first-, second- and third-stage of the thermal decomposition of
45 polyurethane-siloxane semi-IPNs may be generally assumed to follow by diffusion, phase-
46 boundary and autocatalysis model, respectively.
3
47 5. References
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
48 [1] P. Król, Synthesis methods, chemical structures and phase structures of linear
49 polyurethanes. Properties and applications of linear polyurethanes in polyurethane elastomers,
50 copolymers and ionomers, Prog. Mater. Sci. 52 (2007) 915–1015.
51 [2] D.K. Chattopadhyay, D.C. Webster, Thermal stability and flame retardancy of
52 polyurethanes, Prog. Polym. Sci. 34 (2009) 1068–1133.
53 [3] S. Chen, H. Zhuo, S. Chen, Z. Ge, H. Yuan, J. Luo, Studies on the thermal stability of
54 polyurethanes containing pyridine: Thermogravimetric analysis, Thermochim. Acta 543
55 (2012) 281– 287.
56 [4] P. Król, K. Pielichowska, Ł. Byczyński, Thermal Degradation of Polyurethane-Siloxane
57 Anionomers, Thermochim. Acta 507-508 (2010) 91–98.
58 [5] J.T. Yeh, Y.C. Shu, Characteristics of the Degradation and Improvement of the Thermal
59 Stability of Poly(siloxane urethane) Copolymers, J. Appl. Polym. Sci. 115 (2010) 2616–2628.
60 [6] F.S. Chuang, H.Y. Tsi, J.D. Chow, W.C. Tsen, Y.C. Shu, S.C. Jang, Thermal degradation
61 of poly(siloxane-urethane) copolymers, Polym. Degrad. Stab. 93 (2008) 1753–1761.
62 [7] Ł. Byczyński, Effect of different polyethers on surface and thermal properties of
63 poly(urethane-siloxane) copolymers modified with side-chain siloxane, J. Therm. Anal.
64 Calorim. DOI 10.1007/s10973–012–2903–4.
65 [8] G. Camino, S.M. Lomakin, M. Lazzari, Polydimethylsiloxane thermal degradation. Part 1.
66 Kinetic aspects, Polymer 42 (2001) 2395–2402.
1
2
Page 12 of 25

3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
4
67 [9] G. Camino, S.M. Lomakin, M. Lazzari, Polydimethylsiloxane thermal degradation. Part 2.
68 The dergradation mechanisms, Polymer 43 (2002) 2011–2015.
69 [10] P. Rościszewski, M. Zielecka, Silikony właściwości i zastosowanie, WNT, Warszawa
70 2002.
71 [11] Ł. Byczyński, P. Król, Badania nad syntezą oraz właściwościami termicznymi i
72 aplikacyjnymi anionomerów poli(uretanowo-dimetylosiloksanowych) Część I – Badania
73 strukturalne, (Synthesis, and performance properties of poly(urethane-dimethylsiloxane)
74 anionomers. Part I. Structural studies), Polimery 58 (2013) 24–30.
75 [12] Ł. Byczyński, P. Król, Badania nad syntezą oraz właściwościami termicznymi i
76 aplikacyjnymi anionomerów poli(uretanowo-dimetylosiloksanowych) Część II – Właściwości
77 termiczne i aplikacyjne, (Synthesis, and performance properties of poly(urethane-
78 dimethylsiloxane) anionomers. Part II. Thermal and performance properties), Polimery 58
79 (2013) in press.
80 [13] H.J. Naghash, I. Mohammadidehcheshmeh, M. Mehrnia, Synthesis and characterization
81 of a novel hydroxy terminated polydimethylsiloxane and its application in the waterborne
82 polysiloxane–urethane dispersion for potential marine coatings, Polym. Adv. Technol. (2012)
83 DOI: 10.1002/pat.3084,
84 [14] J.P. Lewicki, K. Pielichowski, P. Tremblot De La Croix, B. Janowski, D. Todd, J.J.
85 Liggat, Thermal degradation studies of polyurethane/POSS nanohybrid elastomers, Polym.
86 Deg. Stab. 95 (2010) 1099–1105.
87 [15] M. Alexandru, M. Cazacu, M. Cristea, A. Nistor, C. Grigoras, B.C. Simionescu,
88 Poly(siloxane-urethane) Crosslinked Structures Obtained by Sol-Gel Technique, J. Polym.
89 Sci. Part A: Polym. Chem. 49 (2011) 1708–1718.
90 [16] J.V. Džunuzović, M.V. Pergal, R. Poręba, V, V. Vodnik, B.R. Simonović, M. Špírková,
91 S. Jovanović, Analysis of dynamic mechanical, thermal and surface properties of
92 poly(urethane-ester-siloxane) networks based on hyperbranched polyester. J. Non-Crystalline
93 Solids 358 (2012) 3161–3169.
94 [17] Z. Shi, reparation and characterization of crosslinked polyurethane-
95 blockpoly(trifluoropropylmethyl)siloxane elastomers with potential biomedical applications,
96 Polym. Int. (2013) DOI 10.1002/pi.4428.
97 [18] H.E. Stagg, A method for the determination of isocyanates, The Analyst 71 (1946) 557–
98 559.
99 [19] P. Šimon, Isoconversional methods, fundamentals, meaning and application, J. Therm.
00 Anal. Calorim. 76 (2004) 123–132.
1
3
Page 13 of 25

4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
01 [20] H.L. Friedman, Kinetics of thermal degradation of char-forming plastics from
02 thermogravimetry. Application to a phenolic plastic, J. Polym. Sci. C: Polym. Symp. 6 (1964)
03 183–195.
04 [21] T. Ozawa, A new method of analyzing thermogravimetric data. Bull. Chem. Soc. Jap. 8
05 (1965) 1881–1886.
06 [22] J.H. Flynn, L.A. Wall, A quick, direct method for the determination of activation energy
07 from thermogravimetric data, J. Polym. Sci. B:Polym. Lett. 4 (1966) 323–328.
08 [23] C.D. Doyle, Estimating isothermal life from thermogravimetric data, J. Appl. Polym. Sci.
09 6 (1962) 639–642.
10 [24] J. Opfermann, Kinetic analysis using multivariate non-linear regression. I. Basic
11 concepts, J. Therm. Anal. Calorim. 60 (2000) 641–658.
12 [25] H. Maciejewski, I. Dąbek, R. Fiedorow, M. Dutkiewicz, M. Majchrzak, Thermal stability
13 of hybrid materials based on epoxy functional (poly)siloxanes, J. Therm. Anal. Calorim. 110
14 (2012) 1415–1424.
15 [26] L.F. Wang, Q. Ji, T.E. Glass, T.C. Ward, J.E. McGrath, M. Muggli, G. Burns, U.
16 Sorathia, Synthesis and characterization of organosiloxane modified segmented polyether
17 polyurethanes, Polymer 41 (2000) 5083–5093.
18 [27] S. Vyazovkin, A.K. Burnham, J.M. Criado, L.A. Pérez-Maqueda, C. Popescu, N.
19 Sbirrazzuoli, ICTAC Kinetics Committee recommendations for performing kinetic
20 computations on thermal analysis data, Thermochim. Acta 520 (2011) 1–19.
21 [28] A. Goswami, G. Srivastava, A.M. Umarji, G. Madras, Thermochim. Acta 547 (2012) 53–
22 61.
23 [28] K. Chrissafis, Kinetics of thermal degradation of polymers. Complementary use of
24 isoconversional and model-fitting methods, J. Therm. Anal. Calorim. 95 (2009) 273–283.
25 [39] S.M. Lomakin, L.A. Novokshonova, P.N. Brevnov, A.N. Shchegolikhin, Thermal
26 properties of polyethylene/montmorillonite nanocomposites prepared by intercalative
27 polymerization, J. Mater. Sci. 43 (2008) 1340–1353.
28 [30] D. Rosu, L. Rosu, M. Brebu, Thermal stability of silver sulfathiazole‒epoxy resin
4
4
29 network, J. Anal. Appl. Pyrol. 92 (2011) 10‒18.
30
1
4
Page 14 of 25

4
4
4
4
4
4
30 Table Captions
31 Table 1 Composition of synthesized semi-interpenetrating polymer network
32 Table 2 Kinetic models for thermal degradation of polymers [24]
-1
33 Table 3 Interpretation of TG and DTG curves recorded at 5,10, 20, 40 ºC min in nitrogen
34 Table 4 The results of the kinetic analysis obtained using multivariate non-linear regression
35
4
36 Captions for figures
4
4
4
4
4
4
4
37 Scheme 1. Compounds used in semi-IPN compositions
38 Fig. 1. FTIR spectra of the synthesized samples.
-1
39 Fig. 2. TG thermograms (a) and DTG curves (b) recorded at 10 °C min in nitrogen.
40 Fig. 3. Dependence of apparent activation energy (E ) on the degree of conversion (α)
a
41 determined by (a) Friedman method and (b) Ozawa-Flynn-Wall method in nitrogen.
42 Fig. 4. Best fit of kinetic models of the thermal decomposition for PGS50 semi-IPN.
43
1
5
Page 15 of 25

443 Table 1 Composition of synthesized semi-interpenetrating polymer network
Sample
GS wt.%
PUR wt.%
PUR
0
25
50
75
100
100
75
50
25
0
PGS25
PGS50
PGS75
GS
4
4
44
45
1
6
Page 16 of 25

445 Table 2 Kinetic models for thermal degradation of polymers [24]
Model
Symbol
f(α)
n
Reaction of n-th order
Fn
(1 - α)
1
One-dimensional diffusion
Two-dimensional diffusion
D1
2

1
D2
D3
D4
[
ln(1)]
2
/ 3
3
(1)
[1 (1)
Three-dimensional diffusion (Jander equation)
1
/ 3]
2
3
Three-dimensional diffusion (Ginstling-Brounshtein)
Prout-Tompkins equation

1/ 3
2
[(1)
1]
n
a
Bna
(1- α) α
Reaction of n-th order with autocatalysis by X
X – a component in the complex model
n
CnX
(1 - α) (1 + Kcat α)
4
4
46
47
1
7
Page 17 of 25

-1
4
47 Table 3 Interpretation of TG and DTG curves recorded at 5,10, 20, 40 ºC min in nitrogen
Sample
T
5%
T
[ C]
15%
T
[ C]
50%
T
[ C]
max1
Δm
[%]- [ C]
1
T
max2
Δm
[%]
2
Remaining
mass [%]
o
o
o
o
o
[
C]
1
Heating rate at 5 ºC min
PUR
265
270
282
286
221
288
302
312
320
325
346
365
370
371
369
307
313
317
-
48.7
44.3
33.6
-
383
397
378
374
51.1
54.8
65.5
98
0.2
0.9
0.9
2.0
3.0
PGS25
PGS50
PGS75
GS
205
9.7 - 375
87.3
1
Heating rate at 10 ºC min
PUR
276
289
296
299
235
294
315
323
333
339
351
377
379
381
379
310
325
328
-
49.8
43.8
35.2
-
391
427
391
384
50.0
55.6
63.8
98.6
88.5
0.2
0.6
1.0
1.4
2.5
PGS25
PGS50
PGS75
GS
227
9.0 - 380
1
Heating rate at 20 ºC min
PUR
289
292
307
308
255
310
323
334
341
355
363
382
390
388
386
325
337
342
-
48.8
45.1
33.8
-
401
415
403
392
50.9
54.5
65.3
98.5
90.6
0.3
0.4
0.9
1.5
2.0
PGS25
PGS50
PGS75
GS
225
7.4 - 386
1
Heating rate at 40 ºC min
PUR
294
306
316
321
262
312
329
340
350
359
365
387
391
392
388
326
342
349
-
49.9
46.7
35
-
7.5
400
420
400
399
390
49.9
53.1
64.2
98.9
90.8
0.2
0.2
0.8
1.1
1.7
PGS25
PGS50
PGS75
GS
253
4
4
48
49
1
8
Page 18 of 25

4
49 Table 4 The results of the kinetic analysis obtained using multivariate non-linear regression.
E
a1
,
log
E
a2
,
log
Parameter
a
a3
E ,
n
Model
n
Model
log A
3
kJ/mol
A
1
kJ/mol
A
2
kJ/mol
Sample
PUR
Stage I
Stage II
St
158.07
2.2
58.01
2.2
58.13
12.00
± 0.2
11.99
± 0.2
12.00
± 0.2
3.55
1.178
± 0.04
1.176
± 0.06
1.178
± 0.06
238.14
± 4.8
237.58
± 4.8
237.84
± 4.8
163.52
± 1.4
16.66
± 0.4
16.62
± 0.4
16.64
± 0.4
12.30
± 0.1
12.26
± 0.1
12.25
± 0.1
12.84
± 0.3
12.84
± 0.3
12.66
± 0.2
12.82
± 0.3
12.94
± 0.3
13.00
± 0.3
1.25
± 0.1
1.25
± 0.1
1.26
± 0.1
1.74
± 0.1
1.70
± 0.1
1.69
± 0.1
1.36
± 0.1
1.36
± 0.1
1.34
± 0.1
1.09
± 0.1
1.09
± 0.1
1.14
± 0.1
Fn
Fn
-
Fn
Bna
Bna
Bna
Bna
Bna
Bna
Bna
Bna
Bna
Cn C
Cn C
Fn
-
-
-
-
-
-
±
1
±
1
-5
6.89∙10
-6
± 7∙10
-5
4.43∙10
Cn B
D2
D2
D4
D3
D3
D4
D2
D4
D4
D3
D4
D4
-6
±
2.2
60.80
2.5
1.46
2,7
1.05
2.8
65.56
4.3
5.56
4.3
8.98
3.9
52.62
2.6
5.24
2.2
5.01
2.2
72.81
1.6
3.14
1,5
5.10
1,6
± 7∙10
PGS25
PGS50
PGS75
GS
0.270
± 0.03
0.266
± 0.02
0.249
± 0.02
0.179
± 0.04
0.179
± 0.04
0.194
± 0.04
234.92
± 2.5
15.99
± 0.2
15.92
± 0.2
15.99
± 0.2
15.95
± 0.3
15.95
± 0.3
15.92
± 0.3
18.20
± 0.2
18.15
± 0.2
18.19
± 0.3
1.
±
1.
±
1.
±
1.
±
1.
±
1.
±
2.
±
2.
±
-
-
-
-
-
-
-
-
-
-
-
-
±
6
±
6
± 0.3
3.62
± 0.3
3.05
± 0.3
3.47
± 0.5
3.47
± 0.5
2.66
± 0.4
2.40
± 0.3
2.08
± 0.2
2.06
± 0.2
3.46
± 0.1
2.58
± 0.1
163.14
± 1.4
163.22
± 1.5
172.18
± 3.5
172.17
± 3.5
170.12
± 2.7
175.16
± 3.3
176.45
± 3.8
177.27
± 3.7
234.06
± 2.5
234.90
± 2.5
231.38
± 3.4
231.38
± 3.4
230.97
± 3.4
256.71
± 3.2
256.07
± 3.1
256.63
± 3.4
±
±
6
±
5
±
-
4
1.25∙10
± 4∙10
-5
±
5
±
5
-
2.
±
-
-
-
±
270.18
± 6.0
238.92
± 2.9
241.31
± 3.1
19.63 2.210
± 0.5 ± 0.03
16.98 2.265
± 0.2 ± 0.06
17.37 2.118
± 0.2 ± 0.04
-
-
-
-
-
-
±
6
±
6
Cn C
Bna
2.72
± 0.1
0.133
±
± 0.02
4
4
50
51
1
9
Page 19 of 25

4
4
4
4
4
51
52
53
54
55
Highlights
1. The thermal decomposition of the polyurethane-siloxane semi-IPNs was investigated.
2. The thermal degradation kinetic parameters were determined.
3. The possible degradation model f(α) was proposed.
4
4
56
57
2
0
Page 20 of 25

Scheme 1
Page 21 of 25

Figure 1
Page 22 of 25

Figure 2
Page 23 of 25

Figure 3
Page 24 of 25

Figure 4
Page 25 of 25