You are on page 1of 14

Journal of Thermal Analysis and Calorimetry (2021) 146:1049–1062

https://doi.org/10.1007/s10973-020-10096-0

Effect of quantitative characteristic structure of resole phenolic


prepolymer resin on thermal stability, pyrolysis behaviors,
and ablation properties
Honglin Hu1 · Ying Zhang1 · Liang Liu1 · Yunhua Yang1 · Ruilian Yu1 · Jinming Wang1

Received: 28 December 2019 / Accepted: 17 July 2020 / Published online: 4 August 2020
© Akadémiai Kiadó, Budapest, Hungary 2020

Abstract
The chemical structure of resole phenolic prepolymer resin is difficult to be accurately described because it is a mixture
composed of different polymerization degree structures. Limited by the knowledge of prepolymer resin structure, the cross-
linking degree (CLD) of cured resin and the effects of structure on the ablation properties remain unclear. The objective of
this work is to understand the relationship between prepolymer resin structure, CLD of cured resin, and performance. The
quantitative characteristic structure of resole prepolymer and CLD of curing resin was proposed in this work based on the
full understanding of the chemical structure. A fast and accurate identification method was provided for the quality conform-
ity identification of mass production of resole prepolymer phenolic resin. A novel CLD characterization method for high
cross-linking resole phenolic resin was proposed. The results showed that the ablation properties of resole phenolic resin are
optimum when formaldehyde/phenol value of prepolymer is in the range of 1.20–1.46. This study may provide meaningful
information to the understanding of the ablation mechanism of resole phenolic resin, which is a benefit for developing novel
anti-ablation resin.

Keywords Chemical structure · Ablation properties · Cross-link degree · Pyrolysis behavior · Thermal stability

Introduction being quantified. As is known that resole phenolic prepoly-


mer resin is a mixture composed of different polymeriza-
Resole phenolic resin-based composites are widely applied tion degree structures. It is difficult to describe the chemical
in aerospace [1–3], military industry [4], and so on [5, 6], structure exactly. Compared with novolac phenolic resin,
due to their excellent heat resistance, anti-ablation, mechani- the cross-linking reaction of resole phenolic resin occurs to
cal performance, reliability, and low cost. With the rapid form three-dimensional network structures without adding
advancement of high-technologies, requirement of resin any hardener. Lots of excellent work had been reported in
matrix used in rigorous environments including high speed, the literature [7]. King and his co-workers [8] had focused
high temperature, high pressure, high corrosion, is ever on the analysis of the chemical structure of resole resin by
increasing. However, the lack of a full understanding of the the method of nuclear magnetic resonance (NMR) and gel
relationship between structure and ablation for resole phe- permeation chromatography (GPC). Infrared (IR) spectros-
nolic resin may be one of the biggest obstacles for develop- copy was used to quantitatively characterize the structure
ing novel anti-ablation composites. of phenol-formaldehyde resole resin [9]. Nuclear magnetic
For the purpose of comprehension of the relationship, the resonance (13C-NMR) characterization was employed to
chemical structure of prepolymer resin should be prior to identify the characteristic structure of resole prepolymer
phenolic resin [10]. All these studies pointed to the fact
* Yunhua Yang that important influencing factors on the properties of cured
nanofibre@sina.com resin determined quantitatively for resole prepolymer res-
ins were the formaldehyde/phenol (F/P) ratio after synthesis
1
Science and Technology on Advanced Functional reaction and the average molecular weight. Therefore, the
Composites Laboratory, Aerospace Research Institute
of Materials & Processing Technology, Beijing 100076, F/P and average molecular weight can be as the quantitative
China characteristic structure of resole phenolic prepolymer resin.

13
Vol.:(0123456789)
1050 H. Hu et al.

Specifically, the F/P has a significant effect on the perfor- that the apparent activation energy of the pyrolysis of phe-
mance of cured resin including thermal stability, pyrolysis nolic resin was in the range of 150–300 kJ mol−1 [14–16].
behaviors, and ablation. Average molecular weight has an The relationship between the structure and thermal prop-
influence on the performance of the composite molding erties of phenolic resin had been investigated [17–20]. Chen
process such as the resin transfer molded (RTM) process- et al. [21] studied the influence of feed F/P on pyrolysis
ing. However, these analytical ways can not provide a fast temperature and char yield of phenolic resin. A maximum
and accurate identification method. For example, in order to in the benzene ring substitutions, methylene bridge, and the
obtain a high-resolution 13C-NMR spectrogram, it is neces- result of the thermal properties showed that the resin with
sary to increase the number of scanning, which will inevi- F/P molar ratio 1.2 has the highest cross-link density and
tably increase the test time and reduce test efficiency. It is highest char yield, over 73%. Hu et al. [22] reported that the
unfavorable for the quality conformity identification of mass effects of feed F/P ratios on physicochemical characteristics
production of resole phenolic resin. Another example, the and foaming properties of the resulting resins were analyzed
replacement of a chromatographic column with the old one based on viscosity, solids content, hydroxymethyl index,
would bring great results fluctuation of average molecular residual monomer content, and molecular structure. Wang
weight by GPC, which would influence the quality evalua- et al. [23] focused on the influence of different cross-link
tion of resin. density on the heat resistance of novolac resin by adjusting
Subsequently, the quantitative characteristic structure the amount of hardener hexamethylenetetramine (HMTA).
for cured resole phenolic resin should be defined so as to The results reveal that the cross-link of novolac phenolic
build the relationship of prepolymer before and after curing resin can improve the thermal stability and char yield of
reaction. Cross-link density is a wide consensus for describ- cured PR simultaneously. However, the relationship between
ing the cured resin structure [11]. But the measurement of the structure of resole prepolymer resin, cross-link density
cross-link density is still a challenge. Several methods are of cured resin, and performance including thermal stability
available for testing the cross-link density of resole phenolic and ablation remains unclear.
resin, including the dynamical mechanical analysis (DMA), To address the problems mentioned previously, quantita-
swelling, and glass transition temperature (Tg) measure- tive characteristic structure of resole prepolymer and cross-
ments. DMA method is widely recognized as a more accu- link degree (CLD) of curing resin was proposed in this work
rate one for evaluating cross-link density. The elastic modu- based on the full understanding of the chemical structure.
lus of resin in the rubbery regime (i.e., T > Tg) is measured A fast and accurate identification method of the chemical
by using a thermal-mechanical testing instrument [12]. The structure of prepolymer was provided for the quality con-
rubber elasticity theory is applied to calculate the cross-link formity identification of mass production of resole prepoly-
density. However, it is difficult to prepare DMA samples mer phenolic resin. A novel CLD characterization method
without inner bubbles. Because the curing process of resole for high cross-linking resole phenolic resin was proposed.
phenolic resin will always release small molecular gas such The relationship between the structure of resole prepolymer
as water. Besides, rubber elasticity theory becomes invalid resin, CLD of cured resin, and performance was studied. The
due to the failure of the Gaussian chain assumption when investigation focused on the effect of prepolymer structure
the cross-link density is too high. In most cases, the cross- on the CLD and properties that may provide meaningful
link density of the fully cured resole phenolic resin used in information to the understanding of the ablation mechanism
ablative materials is high. Therefore, the DMA method is of resole phenolic resin. It may be a benefit for developing
not entirely applicable in that case. novel anti-ablation resin.
Pyrolysis mechanism of the phenolic resin has been
reported. Typically, Trick [13] proposed a mechanism of
pyrolysis consisting of three stages. First, the functional Experimental
groups of the cured phenolic resin react with each other to
form additional cross-links. Then, the cross-link is destroyed, Raw materials
leading to the formation of methane, hydrogen, and carbon
monoxide. Finally, the hydrogen atom is stripped from the Phenol (P) and formaldehyde (F) were purchased from Sin-
phenol ring structure and hydrogen is released. Phenolic opharm Chemical Reagent Co. Ltd. Ba(OH)2·8H2O used as
resin is converted into amorphous carbon, accompanied by a catalyst was purchased from Sinopharm Chemical Reagent
the release of a large number of gas products. Each of the Co., Ltd. Phosphoric acid was obtained from Tianjin Chemi-
three stages does not represent an elementary reaction but cal Plant, China.
rather a relatively complicated set of reactions. On the base
of this, Pyrolysis kinetics of phenolic resin was investigated
by non-isothermal thermogravimetry. These results showed

13
Effect of quantitative characteristic structure of resole phenolic prepolymer resin on thermal… 1051

Synthesis of phenolic prepolymer resin the area of the resonance signal peaks is equivalent to the
simplest integer ratio of the number of hydrogen atoms. The
1
F and P were mixed in a 3000 mL three-neck round-bot- H-NMR spectrum has a very high resolution and accuracy
tomed flask with mechanical stirring at 40 °C. The mole for describing the characteristic functional groups of phe-
ratio between F and P was 1.0:1, 1.2:1, 1.4:1, 1.6:1, 1.8:1. nolic resin. Therefore, the characteristic structure of phe-
Then, Ba(OH)2·8H2O was added in the mixture under the nolic resin is described by the calculated methylene index
agitation. Subsequently, the system temperature was raised (MI), hydroxymethyl index (HI), and F/P value.
to 60 °C and kept for 2 h, to 90 °C kept for 1 h. 10 mass% MI can be calculated according to formula (1) as follows:
phosphoric acid solution was added into the mixture for
5S1
neutralizing the Ba(OH)2·8H2O. The phenolic resin was MI = (1)
S4
obtained after vacuum dewatering and filter. S1 + S3 + 0.5S2 + 3

Samples preparation HI can be calculated according to formula (2) as follows:


5S2
20 g resin was put in a closed steel mold. The curing process HI = (
S4
)
(2)
was as follows: rising temperature to 80 °C kept for 1 h, 2 S1 + S3 + 0.5S2 + 3
100 °C kept for 2 h, 120 °C kept for 2 h, 140 °C kept for 2 h,
160 °C kept for 2 h, 180 °C kept for 2 h. F/P value can be calculated according to formula (3) as
The cured samples were ground to ensure its uniform follows:
particle size. Before the thermogravimetric analysis, the ( )
5 S1 + S2
samples were put into an oven at 110 °C for 1 h and then F∕P = ( (3)
put into a dryer.
)
S
2 S1 + S3 + 0.5S2 + 34
T300 carbon fiber-reinforced phenolic resin composites
were fabricated via RTM, the curing process was the same Pyrolysis gas-chromatography–mass-spectrometry (PY-
with resin. GC/MS) system was employed to separate and identify the
pyrolysis volatiles. A PY-2020S (Frontier, Japan) pyro-
Characterization lyzer joining with GCMS-QP2010 (Shimadzu Co., Japan)
mass spectrometer was used. The pyrolysis temperatures
Characteristic structures of resole phenolic prepolymer resin of 400 °C and 550 °C were used. The temperature of the
were characterized by Agilent DD2 600 MHz nuclear mag- pyrolysis chamber before heating and during the separation
netic resonance (NMR) instrument. The phenolic resin was of the volatiles was maintained at 200 °C. The separation
dissolved in 99.8% DMSO of the CIL company, and the con- of the volatile products was performed in a 30 m capillary
centration was 20 mg mL−1. The number of scans was used quartz column. Before the chromatography separation, the
as 16. Tetramethylsilane was used as an internal standard. temperature of the chromatographic column was progres-
The resonance signal peak at the chemical displacement of sively increased as follows: (a) 40 °C for 3 min; (b) from 40
3.5–4.0 ppm corresponds to the methylene proton signal, to 260 °C at a rate of 10 °C min−1; (c) the capillary column
and the integral area is defined as ­S1; the resonance signal was maintained at 260 °C for 10 min. Helium gas was used
peak at 4.2–4.75 ppm is hydroxymethyl proton signal, and as the carrier gas. The mass range used for the mass selective
the integral area is defined as S
­ 2; the resonance signal peak detector was 19–500 m/z. The decomposition products were
at 6.5–7.5 ppm corresponds to the proton signal of phenol identified using the comparison between the experimental
ring, and the integral area is defined as S ­ 3; the resonance mass spectrum and the mass spectrum library attached to
signal peak at 2.0–2.4 ppm corresponds to the proton signal the PY-GC/MS apparatus. The identification of each vola-
of methyl, and the integral area is defined as ­S4. Because the tile product can be confirmed if the qualification percentage
natural abundance of the hydrogen atom is nearly 100%, the reaches 85% and even higher. In this paper, a quantitative
number of NMR signal peaks is equivalent to the number description of cross-link density is firstly suggested in for-
of hydrogen atom chemical environment, and the ratio of mula (4), i.e., cross-link degree (CLD), which was obtained
from the result of Py-GC/MS at 550 °C.

sum of dimethylphenol and trimethylphenol


CLD = × 100% (4)
sum of benzene and its homologues, phenol and its homologues

13
1052 H. Hu et al.

The char residue of pure PF resins was measured by using m0 − mi


TGA (Netzsch STA409PC), which was carried out under
𝛼=
m0 − mf (9)
the protection of nitrogen gas in a range of 120–900 °C at
a constant heating velocity of 5 °C min−1, 10 °C min−1, where m0, mf and ­mi are the mass at initial, final, and
15 °C min−1, 20 °C min−1. A scanning electron microscope moments, respectively.
(SEM) with the type of APOLLO 300 was employed to The equation of activation energy E calculated by Ozawa
investigate the surface morphologies of the specimens after [25] method is as follows:
ablation. The average molecule weight was measured by Gel AE E
Permeation Chromatography(GPC) (DAWN 8+, WYATT). lg𝛽 = lg ( ) − 2.315 − 0.4567 (10)
Rg 𝛼T RT
The temperature of the environment fluctuation is less than
2 °C. The temperature inside and outside of the detector is At the same conversion rate αT, lgβ is linearly correlated
35 °C. The flow rate is 1 mL min−1 with 5–8 mg mL−1 for with T−1. The activation energy E is calculated according to
sample concentration and the injection volume is 20 L. The the slope of the linear regression equation.
sample solution is injected into the chromatographic col-
umn through the injection system. Then, the spectrogram
is recorded. The ablation performance of the composite is Results and discussion
investigated by the arc heating ablation. The sample with a
size of 8.5 cm × 5.0 cm is inlaid on the sample table. During Characteristic structure of resole phenolic
the test, the high-temperature airflow heated by the lami- prepolymer resin
nated heater flows through the supersonic nozzle, forming
the required flow field in the test section, and conducting Figure 1 shows the NMR spectrum of phenolic prepolymer
the ablation test on the sample. The surface of the sample is resin synthesized under different F/P ratios and fixed other
heated for t = 10 s. Mass ablation rate and linear ablation rate synthesis conditions. According to Fig. 1, the characteris-
can be calculated according to formula (4) and (5) as follows: tic structure including MI, HI, and F/P value was listed in
m0 − mf Table 2. It indicates that the feed ratio of F/P is close to
Mass ablation rate = (5) the F/P value of 1H-NMR. This proves that describing the
8.5 × 5 × t
characteristic structure of phenolic resin by 1H-NMR is fea-
where m0 and mf are the sample mass at initial and final, sible. In addition, it should be noted that the F/P value of
respectively. 1
H-NMR is slightly higher than the feed ratio of F/P, which
h0 − hf may be caused by the volatilization of free phenol during
Linear ablation rate = (6) the vacuum dehydration stage, resulting in F/P value of 1H-
t
NMR slightly higher than the feed ratio of F/P.
where h0 and hf are the sample thickness at initial and final, The MI value increases with the increase in the content of
respectively. –CH2-according to the formula (1), which means that the MI
has a directly proportional relation with molecular weight. The
Theoretical method bigger the MI value is, the higher the molecular weight is. It is
worth noted that the MI value of sample No. 3 is lower than
The equation of activation energy E calculated by Kissinger that of sample No. 2 in Table 1, which infers that the molecular
[24] method is as follows : weight of sample No. 2 should be higher than that of sample No.
( ) 3. In order to further prove the relationship between MI value
dln 𝛽TP−2 E and molecular weight, GPC analysis was carried out to evaluate
( −1 ) = − (7)
d TP R molecular weight in Table 2. It can be obvious observed that
the weight average molecular weight (Mw) of sample No. 2 is
where β and Tp are the heating rate and peak temperature, higher than that of sample No. 3. As a result, 1H-NMR has a
respectively. good resolution for the evaluation of the characteristic structure
The reaction frequency factor A calculated by Kissinger of resole phenolic prepolymer resin. MI value mainly influences
method is as follows: on weight average molecular weight of the resin.
Much of the works had been focused on the effect of
𝛽Ee(E∕RTP ) feed ratio F/P on the properties of resin [20, 21]. F/P value
A≈ (8)
RTP2 obtained by 1H-NMR finally decides the ratio of alkyl bridge
to phenol ring of resin after curing processing, which is
The kinetic analysis from the TGA was carried out using important for heat resistance, pyrolysis behaviors, and
the conversion rate α according to the formula:

13
Effect of quantitative characteristic structure of resole phenolic prepolymer resin on thermal… 1053

(a)
1.88
(b)

0.55

9.61

0.80

0.90

4.26

1.00
0.52

0.48

6.49

0.61

0.60
0.48
1.27

0.65

3.41
0.20

1.00
9.5 9.0 8.5 8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0
10 9 8 7 6 5 4 3 2
f1/ppm
f1/ppm

(c) (d)

4.59

0.80

1.00
0.94
0.80

0.60

0.55

4.13
1.41

0.72

7.83

1.07

1.02

5.86

1.00
0.61

9.5 9.0 8.5 8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 10 9 8 7 6 5 4 3 2
f1/ppm f1/ppm

(e)
0.66

0.19

0.55

0.54
0.66

0.87

4.42

1.00
4.22

10 9 8 7 6 5 4 3 2
f1/ppm

Fig. 1  1H-NMR spectrum of different feed ratios of F/P with a 1.0:1; b 1.2:1; c 1.4:1; d 1.6:1; e 1.8:1

ablation properties. Therefore, the influence of F/P value on kinetic via TG/DTG analysis. Figure 2 shows the TG curves
the performance is systematically investigated as follows. and DTG curves of phenolic resin at the different heating
rates with different F/P values. The peak temperatures in
Influence of F/P value on thermal stability every DTG curves are listed in Table(3.
According to formula (7), the dln 𝛽TP−2 is used to plot
)

The thermal stability of resins with different F/P values TP−1. The linear regression equations of phenolic resin with
was investigated by the method of non-isothermal pyrolysis the different structures are obtained. According to the slope

13
1054 H. Hu et al.

Table 1  Analysis of characteristic structure of phenolic resin by 1H- cases of F/P value with 1.68 and 1.82, the cross-linking den-
NMR sity of resin may be higher than the others due to the excess
No. Feed ratio of F/P MI HI F/P value of formaldehyde to phenol. However, excessive formaldehyde
1
H-NMR forms more alkyl chains or bridges between phenols. Exces-
sive alkyl structure may lead to the decrease in activation
1 1.0:1 0.392 0.836 1.03
energy of thermal decomposition in the range of 723–873 K.
2 1.2:1 0.544 0.927 1.20
Table 5 shows the char yield of phenolic resin with dif-
3 1.4:1 0.425 1.246 1.46
ferent F/P at 1073 K. It indicates that the char yield of resin
4 1.6:1 0.653 1.349 1.68
at 1073 K from high to low in turn is F/P = 1.20, 1.46, 1.68,
5 1.8:1 0.673 1.487 1.82
1.82. The results further prove the fact that the thermal sta-
bility of resole phenolic resin may be optimal when the F/P
value is in the range of 1.20–1.46.
of the equation, the apparent activation energy of the thermal The apparent activation energy of phenolic resin at dif-
decomposition of resin is calculated. The pre-exponential ferent F/P values was also calculated by the Ozawa method.
factor A is calculated according to formula (8). The results Table 6 shows the temperature corresponding to the conver-
are listed in Table 4. The correlation coefficient R2 is above sion ratio αT at different heating rates. According to formula
0.95, which shows that the linear relationship is significant. (10), for the same conversion αT, lgβ is linearly correlated
As the increase in F/P value, the apparent activation with T−1. The linear regression curves of phenolic resin at
energy of thermal decomposition in the range of 723–873 K different F/P values are shown in Fig. 3. According to the
tends to decrease from Table 5. It can be seen that the appar- slope of the equation, the apparent activation energy of the
ent activation energy in this temperature range is the highest thermal decomposition of resin is calculated and showed in
when the F/P value is 1.03, but it does not mean that the Fig. 4. The value of the average activation energy obtained
thermal stability of resin is the best. It is worth noting that by the Ozawa method is slightly smaller than the activation
there is an obvious pyrolysis rate peak at 673 K from the energy obtained by the Kissinger method, which is con-
DTG curve (Fig. 3a) when the F/P value is 1.03, correspond- sistent with other research [26]. It infers that the results of
ing apparent activation energy of 174.12 kJ mol−1. For the apparent activation energy are reliable.
reason, the cross-linking density of resin with F/P = 1.03 As a result, thermal stabilization of resin may be optimum
may be lower than the others, which results from methyl- when F/P value in the range of 1.20–1.46. In order to further
ene formed by formaldehyde is not enough to make bridges prove the results discussed above, the Py-GC/MS experi-
between the phenols. Partial end group phenols as pyroly- ments and were carried out, and CLD of resin at different
sis product may release, which leads to the mass loss in F/P values were evaluated as follows.
this temperature range of 600–700 K (DTG curves of resin
with F/P = 1.03 in Fig. 2a). Residual resin carbon becomes Influence of F/P value on pyrolysis behaviors
compact under the slowly heating rate of 5–20 K min−1, and CLD
which may result in the highest apparent activation energy
of thermal decomposition in the range of 723–873 K. Com- The ideal cross-linked structure of resole phenolic resin is
pared with other resins (Fig. 2b–e), only the resin with that: all the three activated points of phenol ring are involved
F/P = 1.03 has a large weight loss in the temperature range in the formation of body structure. However, because of the
of 600–700 K. When the F/P value is 1.20 and 1.46, the synthesis, curing, and other reasons, the actual curing struc-
corresponding apparent activation energy of thermal decom- ture contains different proportions of terminal groups of sin-
position is 239.37 kJ mol−1, 213.60 kJ mol−1, respectively. gle bridge biphenyl, a certain proportion of double bridge
It infers that the thermal stability of the resin with the F/P biphenyl (Fig. 5), which results in the decrease in cross-
value in the range of 1.20–1.46 may be optimum. For the link density of the cured resin. Methylene bridges between

Table 2  The molecular weight No. Feed ratio of F/P Number average molecular Weight average molecular Mw/Mn
of the different characteristic weight/Mn weight/Mw
structure
1 1.0:1 326.2 571.1 1.75
2 1.2:1 357.5 701.2 1.96
3 1.4:1 384.1 637.5 1.66
4 1.6:1 478.8 883.3 1.84
5 1.8:1 512.6 940.8 1.84

13
Effect of quantitative characteristic structure of resole phenolic prepolymer resin on thermal… 1055

0.05
(a) (b)
100 20 K min–1 100
0.00 20 K min–1 0.00
15 K min–1 15 K min–1
10 K min–1
10 K min–1 90 5 K min–1 – 0.05

DTG/% min–1
5 K min–1

DTG/% min–1
20 K min–1 – 0.05
Mass/%

Mass/%
15 K min–1
80 80 20 K min–1 – 0.10
10 K min–1 – 0.10 15 K min–1
5 K min–1 10 K min–1
5 K min–1 – 0.15
70
– 0.15
– 0.20
60 60
– 0.20
– 0.25
600 800 1000 600 800 1000
Temperature/K Temperature/K

0.05 110 0.05


(c) (d)
100
20 K min–1 0.00 100
0.00
15 K min–1 20 K min–1
10 K min–1 15 K min–1

DTG/% min–1
DTG/% min–1
90 90
5 K min–1 – 0.05 10 K min–1
– 0.05

Mass/%
5 K min–1
Mass/%

– 0.10 80
80 20 K min–1
15 K min–1 – 0.10
10 K min–1 20 K min–1
– 0.15 70 15 K min–1
70 5 K min–1
10 K min–1
5 K min–1 – 0.15
– 0.20 60
60
– 0.25 50 – 0.20
600 800 1000
600 800 1000
Temperature/K
Temperature/K
(e)

0.05
100
(e)
20 K min–1 0.00
90 15 K min–1
10 K min–1
DTG/% min–1

5 K min–1 – 0.05
80
Mass/%

20 K min–1
15 K min–1
10 K min–1 – 0.10
70 5 K min–1

60 – 0.15

50 – 0.20
600 800 1000
Temperature/K

Fig. 2  The TG and DTG curves of phenolic resin at the different heating rates with F/P value a 1.03; b 1.20; c 1.46; d 1.68; e 1.82

phenol rings are well-known formed by polycondensation of compared to cement. The proportion of cement and brick
formaldehyde [27]. Under the condition of the resin cured should directly determine the performance of a brick wall.
in an airtight container such as RTM processing, the ratio of Therefore, cross-link density can be defined as the ratio of
methylene bridge to phenol ring should equal to F/P value alkyl bridge to the phenol ring.
obtained by 1H-NMR. It means that CLD value increases Figure 6 shows the total ion chromatograms of phenolic
with the enhance of F/P value. The phenolic resin could be resin pyrolysis at 400 °C with different F/P value, the peak
compared to the brick wall. The phenol ring is compared to identification of pyrolysis are listed in Table 7. It indicates
brick. The methylenes and little other bridging structures are that mass loss of F/P value with 1.03 at 400 °C is mainly

13
1056 H. Hu et al.

Table 3  Peak temperature in DTG curves under different F/P ratios Table 5  Char yield of phenolic resin with different F/P at 1073 K
F/P value/ Peaks in DTG curves, Tp/K F/P value Char yield/%
temperature −1 −1 −1 −1
range/K 5/K min 10/K min 15/K min 20/K min 5/K min−1 10/K min−1 15/K min−1 20/K min−1

1.03/653–703 661.50 673.75 684.67 688.81 1.03 60.906 63.895 62.170 60.455
1.03/723–873 793.07 810.16 816.15 822.46 1.20 63.486 65.550 65.031 64.598
1.20/723–873 785.73 802.52 809.93 814.27 1.46 62.309 62.776 63.325 62.503
1.46/723–873 782.46 798.47 807.16 815.22 1.68 57.140 59.781 57.724 58.398
1.68/723–873 761.59 776.31 791.48 801.11 1.82 59.139 58.475 55.541 57.625
1.82/723–873 750.42 762.96 776.68 787.45

composite is investigated by the arc heating ablation under


the same ablation state. The results are shown in Table 9 and
owing to releasing end group phenol. As the increase in F/P Fig. 9. It can be seen that the linear ablation rate and mass
value, the content of pyrolysis product phenol decreases ablation rate of the composite from low to high in turn is
obviously. Therefore, the content of single bridge biphenyl F/P = 1.20, 1.46, 1.68, 1.82, 1.03. It infers that the thermal
increases with the decrease in F/P, leading to a decrease in stability of the resin is optimum when the F/P value is in the
thermal stability of the resin. range of 1.20–1.46.
Pyrolysis gases of resole phenolic mainly include CO, Figure 10 shows the surface morphology of compos-
­CH4, ­CO2, ­H2O, benzene and its homologs, phenol and its ites after ablation. When the F/P value is 1.03, due to the
homologs, which has been reported in the literature [20, low CLD of the cured resin (10.9%), the mass loss of resin
28, 29]. Figure 7 shows the total ion chromatograms of the increases during the ablation process, which leads to less
pyrolysis gas of phenolic resin with different characteristic residual resin carbon between the carbon fibers (Fig. 10a).
structure at 550 °C. The peak identifications of pyrolysis Heat penetration of composites quickly occurs due to the
gas are listed in Table 8. It should be noted that the sum lack of resin carbon connection between the fibers, which
of the ion intensity of small molecules increases with the results in the increase in the ablation rate of composites.
increase in F/P value. The sum of the ion intensity of small When the F/P value increases to 1.20, the CLD of the
molecules was calculated by the ion intensity of CO, C ­ H4, resin increases to 15.7%. The compact resin carbon can be
­CO2, ­H2O. The CLD of resin also raises with the increase in observed between the fibers (Fig. 10b). When the F/P value
F/P value (Fig. 8). Methanol is the polymerization inhibitor increases to 1.46, the CLD of the resin increases to 22.3%.
contained in formalin, which is unnecessary to count in. It The resin carbon can be obviously observed between the fib-
infers that the excess formaldehyde may form more alkyl ers. It is worth noting that the pore morphology of the resin
structures, which leads to an increase in CLD and release carbon area appears (Fig. 10c), which should be caused by
more small-molecule pyrolysis products such as ­CH4, CO, the release of small molecular pyrolysis gases such as C ­ H4,
­H2O, and ­CO2. CO, ­H2O, and C ­ O2. With the increase in F/P value to 1.68,
the CLD of the cured resin increases to 35.4%. The ratio of
Influence of F/P value on ablation properties small molecular pyrolysis gases increases. The inner pres-
sure caused by small molecular gases leads to micro-erosion
In order to verify that the thermal stability of the resin is of composite (Fig. 10d), which results in an increase in the
optimal when the F/P value is in the range of 1.20–1.46, ablation rate. When the F/P value increases to 1.82, the CLD
carbon fiber-reinforced composites with different F/P values of the cured resin increases to 40.1%. A large number of
of resin were fabricated. The ablation performance of the small molecular pyrolysis gases were released, resulting

Table 4  Activation energy F/P value/tempera- Slope Ea/KJ mol−1 Average of A value Correlation coefficient
and A value obtained by the ture range/K of linear regression R2
Kissinger method of phenolic
resin with different F/P ratios 1.03/653–703 − 20,943 174.12 1.37817 × 1013 0.98624
1.03/723–873 − 29,190 242.69 5.99079 × 1010 0.98624
1.20/723–873 − 28,790 239.37 1.84695 × 1015 0.98071
1.46/723–873 − 25,691 213.60 3.82531 × 1013 0.99807
1.68/723–873 − 19,412 161.40 2.10130 × 1010 0.97261
1.82/723–873 − 20,181 167.79 9.42428 × 1010 0.95279

13
Effect of quantitative characteristic structure of resole phenolic prepolymer resin on thermal… 1057

1.4 (a) 1.4 (b)


1.3 1.3

1.2 1.2

1.1 1.1
Igβ

Igβ
1.0 1.0

0.9 0.9

0.8 0.8

0.7 0.7

0.6
α :90% α :5% 0.6 α :90% α :5%

0.0010 0.0015 0.0020 0.0010 0.0012 0.0014 0.0016 0.0018 0.0020


T – 1/K–1 T – 1/K–1

1.4 (c) 1.4 (d)


1.3 1.3

1.2 1.2

1.1 1.1

1.0 1.0
Igβ

Igβ

0.9 0.9

0.8 0.8

0.7 0.7

0.6 α :90% α :5% 0.6 α :90% α :5%

0.0010 0.0015 0.0020 0.0010 0.0015 0.0020

T – 1/K–1 T – 1/K–1

1.4 (e)
1.3

1.2

1.1

1.0
Igβ

0.9

0.8

0.7
α :90% α :5%
0.6

0.0010 0.0012 0.0014 0.0016 0.0018 0.0020 0.0022


T – 1/K–1

Fig. 3  Curves of lgβ versus T−1 with F/P value a 1.03; b 1.20; c 1.46; d 1.68; e 1.82

13
1058 H. Hu et al.

Table 6  Temperature (K) αT/% 5 10 15 20 25 30 35 40 45


corresponding to the conversion
ratio αT at different heating rates F/P = 1.03
5/K min−1 513.43 577.08 622.14 648.01 668.78 689.88 710.48 729.25 746.33
10/K min−1 524.57 588.30 628.23 658.67 682.43 699.10 718.92 737.75 756.70
15/K min−1 527.49 593.42 635.40 668.32 693.79 707.56 726.73 746.25 765.81
20/K min−1 533.22 601.65 643.05 676.99 704.50 716.47 734.80 754.76 774.22
αT/% 50 55 60 65 70 75 80 85 90
5/K min−1 762.10 776.62 790.10 803.20 816.74 832.03 851.27 877.96 913.81
10/K min−1 773.75 786.14 802.41 815.54 828.91 845.44 863.59 890.09 926.66
15/K min−1 784.42 795.56 813.88 826.71 839.92 857.08 875.15 901.67 938.70
20/K min−1 793.59 805.03 824.59 835.82 850.70 867.42 885.76 913.78 950.27
F/P = 1.20
αT/% 5 10 15 20 25 30 35 40 45
5/K min−1 524.22 633.36 669.66 694.86 715.99 732.89 747.35 760.53 773.01
10/K min−1 530.43 641.48 676.77 702.74 724.36 741.85 756.23 770.75 782.88
15/K min−1 536.70 649.45 684.68 711.31 733.58 751.51 764.89 779.91 792.54
20/K min−1 542.92 657.22 692.56 720.27 741.64 759.72 774.11 788.57 801.42
αT/% 50 55 60 65 70 75 80 85 90
5/K min−1 784.86 796.56 808.37 821.06 835.54 853.17 875.50 902.89 935.05
10/K min−1 794.22 806.22 820.05 832.84 849.22 865.50 887.16 916.12 950.73
15/K min−1 804.90 816.90 829.71 842.50 860.88 876.86 899.82 928.24 964.46
20/K min−1 814.76 825.50 841.39 853.73 869.99 887.08 909.51 941.45 976.44
F/P = 1.46
5/K min−1 503.45 570.23 650.78 681.32 704.90 725.05 740.30 752.96 764.50
10/K min−1 510.12 578.85 659.64 692.28 715.56 735.71 749.57 763.64 773.22
15/K min−1 517.38 585.04 666.62 702.22 724.21 743.84 760.83 773.13 784.45
20/K min−1 522.26 592.81 676.48 705.51 733.76 752.95 771.05 784.49 794.16
αT/% 50 55 60 65 70 75 80 85 90
5/K min−1 775.43 786.28 797.20 808.78 821.87 837.59 858.30 885.97 921.22
10/K min−1 784.40 794.97 808.56 818.44 832.23 850.25 870.65 897.31 930.10
15/K min−1 795.63 807.33 819.42 830.04 839.09 858.81 880.31 913.85 944.46
20/K min−1 805.74 818.90 831.75 842.61 854.46 874.17 894.85 924.19 961.01
F/P = 1.68
αT/% 5 10 15 20 25 30 35 40 45
5/K min−1 503.31 546.85 615.43 663.99 692.50 714.76 730.72 743.50 755.49
10/K min−1 515.11 559.08 623.09 676.66 708.38 728.33 744.94 754.75 768.71
15/K min−1 521.23 567.63 633.41 686.40 717.07 741.00 755.29 768.32 781.26
20/K min−1 527.48 576.78 641.63 694.91 726.94 749.70 767.84 781.73 795.47
αT/% 50 55 60 65 70 75 80 85 90
5/K min−1 767.57 779.68 792.11 805.20 819.92 837.56 860.33 883.88 915.36
10/K min−1 780.69 794.89 806.33 820.87 834.06 853.33 878.01 901.10 931.52
15/K min−1 794.35 806.36 820.58 835.08 846.91 869.67 893.23 916.64 949.50
20/K min−1 806.94 820.59 835.03 848.32 863.22 882.65 909.60 933.87 966.05
F/P = 1.82
αT/% 5 10 15 20 25 30 35 40 45
5/K min−1 494.64 531.17 582.83 651.70 685.09 709.79 727.97 741.25 753.22
10/K min−1 502.99 539.76 591.06 662.58 697.85 721.44 741.08 755.57 766.77
15/K min−1 511.65 547.31 600.60 671.89 708.46 733.68 752.35 766.80 779.54
20/K min−1 519.54 554.52 609.71 679.44 718.13 742.67 762.82 779.39 790.87
αT/% 50 55 60 65 70 75 80 85 90
5/K min−1 765.17 777.22 789.35 802.05 815.86 831.63 851.69 878.88 914.34
10/K min−1 777.52 793.51 805.03 816.74 831.43 848.30 866.92 895.61 931.90

13
Effect of quantitative characteristic structure of resole phenolic prepolymer resin on thermal… 1059

Table 6  (continued) αT/% 5 10 15 20 25 30 35 40 45


−1
15/K min 791.17 805.68 816.62 830.11 845.55 862.62 881.01 913.42 948.51
20/K min−1 802.08 819.13 829.63 843.66 859.28 877.84 897.74 928.55 964.05

5
300
a 12 3
4 e
b
c
200 d
e
E/KJ mol–1

Ion intensity
100
Average value of E:
a: 210.7 KJ mol–1 c
0
b: 226.7 KJ mol–1
c: 207.5 KJ mol–1
– 100 b
d: 164.8 KJ mol–1
e: 166.9 KJ mol–1 6
– 200 a
0 20 40 60 80 100
α T/% 0 10 20 30 40 50 60 70 80
Time/min
Fig. 4  Activation energy obtained by the Ozawa method of phenolic
resin with F/P value a 1.03; b 1.20; c 1.46; d 1.68; e 1.82 Fig. 6  Total ion chromatograms of phenolic resin pyrolysis at 400 °C
with F/P value a 1.03; b 1.20; c 1.456; d 1.68; e 1.82

OH H2C in the higher inner pressure and a large number of micro-


H2 H2 erosion on the composite’s surface (Fig. 10e). The ablation
C C CH2 C OH
H2 rate further increases. Therefore, the decrease in residual
HO OH CH2 CH2 resin carbon will lead to losing the connection between fib-
CH2 H2C H2C
CH2 OH ers when the F/P value is low, which results in the decrease
in the ablative performance of the composite. The increase
OH
H2C
CH2
OH CH2
in inner pressure caused by pyrolysis gases will result in
OH CH2
micro-erosion when the F/P value is high, which leads to
the decrease in ablative properties of the composite. As a
OH consequence, the optimum F/P value of resin using in the
ablation composite is in the range of 1.20–1.46.
Fig. 5  Cross-linking structure of resole-typed resin

Table 7  Pyrolysis products of Peak no. Pyrolysis gases Peak area of ion intensity under different F/P values (× 10−8/mg)
Py-GC/MS at 400°C
1.03 1.20 1.46 1.68 1.82

1 CO 0.031 0.030 0.324 0.603 0.239


2 CO2 0.074 0.097 0.420 0.681 0.566
3 H2O 0.814 1.181 1.531 2.144 2.248
4 HCHO 0 0 0 0 0.205
5 Methanol 0.037 0.085 0.094 2.692 2.907
6 Phenol 1.249 0.074 0.032 0.037 0.034

13
1060 H. Hu et al.

11
10 40
5 121314
123 4 6 15 e
35

d 30
Ion intensity

CLD value/%
25
c
20

b
15

7 89 a 10
0 20 40 60
1.0 1.2 1.4 1.6 1.8
Time/min
F/P value of 1H-NMR
Fig. 7  Total ion chromatograms of phenolic resin pyrolysis at 550 °C
with F/P value a 1.03; b 1.20; c 1.456; d 1.68; e 1.82 Fig. 8  CLD of phenolic resin pyrolysis at 550 °C with different F/P
values

Conclusions Table 9  Ablation performance of carbon fiber-reinforced phenolic


resin composites
F/P and MI value obtained by 1H-NMR can be used for F/P value of 1H-NMR Mass ablation Linear ablation
quantitative characteristic structure of resole phenolic pre- rate/g cm−2 s−1 rate/mm s−1
polymer resin. The CLD is defined as the percentage of the
1.03 0.08871 0.44250
sum of dimethylphenol and trimethylphenol to sum of ben-
1.20 0.07026 0.28667
zene and its homologs, phenol and its homologs, which was
1.46 0.07763 0.35000
obtained from the result of Py-GC/MS at 550 °C. F/P value
1.68 0.08072 0.40833
has a directly proportional relation with CLD. The phenolic
1.82 0.08936 0.41667
skeleton structure such as phenol ring is easy to break and

Table 8  Analysis of pyrolysis Peak no. Pyrolysis gases Peak area of ion intensity under different F/P
products of Py-GC/MS at values (× 10−8/mg)
550 °C
1.03 1.20 1.46 1.68 1.82

1 CO 0.282 0.365 0.405 0.415 0.414


2 CH4 0.171 0.276 0.548 0.616 0.605
3 CO2 1.004 1.354 1.539 1.674 1.662
4 H2O 2.931 3.160 3.228 3.247 3.418
5 Methanol 0.050 0.306 0.288 2.810 2.680
6 pentadiene 0.054 0.119 0.068 0.113 0.246
7 benzene 0.249 0.100 0.083 0 0
8 toluene 0.532 0.198 0.271 0.089 0.126
9 xylene 0.335 0.087 0.193 0.074 0.113
10 Phenol 14.849 5.465 4.637 3.707 4.336
11 O-cresol 12.575 4.965 7.552 8.079 9.451
12 P-cresol 8.011 2.303 3.584 2.893 3.922
13 2,6-dimethylphenol 2.180 1.224 2.022 3.390 4.532
14 2,4-dimethylphenol 2.295 1.126 2.417 3.895 6.064
15 2,4,6-trimethylphenol 0 0.086 0.235 0.842 1.407
Sum of ion intensity of small molecules (× 10−8/mg) 4.388 5.155 5.72 5.952 6.099
CLD/% 10.9 15.7 22.3 35.4 40.1

13
Effect of quantitative characteristic structure of resole phenolic prepolymer resin on thermal… 1061

0.50 0.12 micro-erosion when the CLD is high, leading to a decrease


in ablative properties. As a consequence, heat resistance and

Mass ablation rate/g cm–2 s–1


Linear ablation rate/mm s–1

0.45
a ablation properties of resin are optimum when the F/P value
0.40
0.10 of resin is in the range of 1.20–1.46.
0.35
b Acknowledgements The authors would like to acknowledge Science
0.30 and Technology on Advanced Functional Composites Laboratory
0.08 for funding the research under the Grant No. HTKJ2019KL703002.
0.25
Thanks to Prof. Zhongping Li for his useful suggestions on the research
0.20 of the paper.

0.15 0.06 Author contributions Methodology: H.H., L.L. Formal analysis and
1.0 1.2 1.4 1.6 1.8
investigation: H.H., R.Y. Writing-original draft preparation: H.H.
F/P value of 1H-NMR
Writing-review and editing: H.H., Y.Z. Data collection and analysis:
H.H., Y.Z. Funding acquisition: J.W. Supervision: Y.Y.
Fig. 9  Analysis of ablation performance of carbon fiber-reinforced
phenolic resin composites. Curve a: Linear ablation rate; curve b:
Mass ablation rate
References
1. Naderi A, Mazinani S, Ahmadi SJ, Sohrabian M, Arasteh R. Mod-
ified thermo-physical properties of phenolic resin/carbon fiber
composite with nano zirconium dioxide. J Therm Anal Calorim.
2014;117:393–401.
2. Cheng H, Xue H, Hong C, Zhang X. Preparation, mechanical,
thermal and ablative properties of lightweight needled carbon
fibre felt/phenolic resin aerogel composite with a bird nest struc-
ture. Compos Sci Technol. 2017;140:63–72.
3. Zhao T, Ye H, Zhang L, Cai Q. Experimental investigation on the
specific heat of carbonized phenolic resin-based ablative materi-
als. Int J Thermophys. 2017;38:151.
4. Menapace C, Leonardi M, Secchi M, Bonfanti A, Gialanella S,
Straffelini G. Thermal behavior of a phenolic resin for brake pad
manufacturing. J Therm Anal Calorim. 2019;137:759–66.
5. Chen J, Zhang W, Liu J, Ge H, Tian M, Liu J, et al. Improved
thermal stability of phenolic resin by graphene-encapsulated
nano-SiO2 hybrids. J Therm Anal Calorim. 2019;135:2377–87.
6. Chen R, Xu X, Zhang Y, Lu S, Lo S. Characterization of igni-
tion and combustion characteristics of phenolic fiber-reinforced
plastic with different thicknesses. J Therm Anal Calorim.
2020;140:645–55.
7. Tonge LY, Hodgkin J, Blicblau AS, Collins PJ. Effects of initial
phenolformaldehyde (pf) reaction products on the curing proper-
ties of pf resin. J Therm Anal Calorim. 2001;64:721–30.
8. King PW, Mitchell RH, Westwood AR, Rue DL. Structural analy-
sis of phenolic resole resins. J Appl Polym Sci. 1974;18:1117–30.
9. Holopainen T, Alvila L, Rainio J, Pakkanen TT. IR spectroscopy
as a quantitative and predictive analysis method of phenol-formal-
Fig. 10  Surface morphology of carbon fiber-reinforced composites dehyde resol resins. J Appl Polym Sci. 1998;69:2175–85.
after ablation using different resins with F/P value a 1.03; b 1.20; c 10. Rego R, Adriaensens PJ, Carleer RA, Gelan JM. Fully quantitative
1.46; d 1.68; e 1.82 carbon-13 NMR characterization of resol phenol-formaldehyde
prepolymer resins. Polymer. 2004;45:33–8.
11. Bandyopadhyay A, Valavala PK, Clancy TC, Wise KE, Ode-
release during the ablation when the F/P value is too low. gard GM. Molecular modeling of crosslinked epoxy polymers:
When the F/P value is too high, a lot of small-molecule the effect of crosslink density on thermomechanical properties.
Polymer. 2011;52:2445–52.
pyrolysis gases produced by excessive alkyl structure will 12. Hill LW. Calculation of crosslink density in short chain networks.
release. Both of the situations will reduce the heat resistance Prog Org Coat. 1997;31:235–43.
and ablation properties of the resin. The decrease in resid- 13. Trick KA, Saliba TE. Mechanisms of the pyrolysis of phenolic
ual resin carbon will lead to losing the connection between resin in a carbon/phenolic composite. Carbon. 1995;33:1509–15.
14. Jiang H, Wang J, Wu S, Wang B, Wang Z. Pyrolysis kinetics of
fibers when the CLD is low, which results in a decrease phenol-formaldehyde resin by non-isothermal thermogravimetry.
in the ablative performance of composites. The increase Carbon. 2010;48:352–8.
in inner pressure caused by pyrolysis gases will result in

13
1062 H. Hu et al.

15. Trick KA, Saliba TE, Sandhu SS. A kinetic model of the pyroly- 23. Wang Y, Wang S, Bian C, Zhong Y, Jing X. Effect of chemical
sis of phenolic resin in a carbon/phenolic composite. Carbon. structure and cross-link density on the heat resistance of phenolic
1997;35:393–401. resin. Polym Degrad Stab. 2015;111:239–46.
16. Ma W, Wang S, Cui JP, Zhang ST, Fan BC, He YZ. Thermal 24. Kissinger HE. Reaction kineticsin differential thermal analysis.
decomposition kinetic model of phenolic resin. Acta Phys Chim Anal Chem. 1957;29:1702–6.
Sin. 2008;24:1090–4. 25. Ozawa T. A new method of analyzing thermogravimetric data.
17. Costa L, Di Montelera LR, Camino G, Weil ED, Pearce E. Struc- Bull Chem Soc Jpn. 1965;11:1881–6.
ture-charring relationship in phenol-formaldehyde type resins. 26. Wei H, Zhang Q, Li J, Li C, Li Z. Suggestion of the non-isother-
Polym Degrad Stab. 1997;56:23–35. mal decomposition kinetic mechanism function for phenolic resin.
18. Parker JA. The effects of molecular structure on the thermo- Polym Mater Sci Eng. 2014;30:100–5.
chemical properties of phenolics and related polymers. NASA 27. Jiang H, Wang J, Wu S, Yuan Z, Hu Z, Wu R, et al. The pyrolysis
TR R-276, Ames Research Center; 1967. mechanism of phenol formaldehyde resin. Polym Degrad Stab.
19. Fitzer E, Schäfer W. The effect of crosslinking on the forma- 2012;97:1527–33.
tion of glasslike carbons from thermosetting resins. Carbon. 28. Wang J, Jiang H, Jiang N. Study on the pyrolysis of phenol-for-
1970;8:353–64. maldehyde (pf) resin and modified pf resin. Thermochim Acta.
20. Hirano K, Asami M. Phenolic resins-100 years of progress and 2009;196:136–42.
their future. React Funct Polym. 2013;73:256–69. 29. Sobera M, Hetper J. Pyrolysis-gas chromatography-mass
21. Chen ZQ, Zeng WJ, Chen YF, Li WK, Liu AH. Influence of F/P spectrometry of cured phenolic resins. J Chromatogr A.
on structure and thermal property of phenolic resin. Key Eng 2003;993:131–5.
Mater. 2012;500:98–103.
22. Hu XM, Zhao YY, Cheng WM. Effect of formaldehyde/phenol Publisher’s Note Springer Nature remains neutral with regard to
ratio (f/p) on the properties of phenolic resins and foams synthe- jurisdictional claims in published maps and institutional affiliations.
sized at room temperature. Polym Compos. 2015;36:1531–40.

13

You might also like