Polymer Degradation and Stability 105 (2014) 140e149

Contents lists available at ScienceDirect

Polymer Degradation and Stability

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

Relationship between structure and properties of reprocessed glass

fiber reinforced flame retardant poly(butylene terephthalate)
Zhang Shuidong, Tan Lingcao, Liang Jizhao, Huang Hanxiong, Jiang Guo*
Lab for Micro Molding and Polymer Rheology, College of Mechanical and Automotive Engineering, South China University of Technology,
Guangzhou 510640, China

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

Article history: Considering environment protection and economic requirements, the reuse of glass fiber reinforced
Received 13 August 2013 flame retardant PBT (GFFRPBT) is of significant importance. In this paper, the properties of reprocessed
Received in revised form GFFRPBT (RGFFRPBT) including rheological, mechanical, thermal and flame retardant properties were
20 March 2014
evaluated. To explore the relationship between structure and properties of RGFFRPBT, the carboxyl
Accepted 11 April 2014
Available online 24 April 2014
content and intrinsic viscosity of PBT, average length of glass fiber and its distribution, and the micro-
scopic appearance of RGFFRPBT were investigated. The results demonstrated that when GFFRPBT was
injection molded for five cycles, the intrinsic viscosity of PBT, average length of the glass fiber and the
Keywords:
Glass fiber reinforced flame retardant PBT
percentage of glass fiber length, which was higher than critical length decreased, whereas, the carboxyl
Reprocessed content of PBT increased significantly. The rheological and mechanical properties became sensitive to
Intrinsic viscosity these variations: the shear viscosity, impact strength, tensile and flexural strength of RGFFRPBT
Average length of glass fiber decreased. Furthermore, the influence of degradation on the PBT molecular weight and the glass fiber
Properties length after the reprocessing cycles were negative on the thermal properties of RGFFRPBT. However, the
flame retardant properties of RGFFRPBT was independent on the degradation; RGFFRPBT with five in-
jection molding cycles could achieve a UL-94 V-0 classification 32.0 LOI, and pass the GWIT (775
C) test.
The results showed that, when GFFRPBT was reprocessed twice, its properties matched the requirements
of electronic and electric fields well.
Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction reinforced PBT becomes important [4]. Moreover, as the glass fiber
has a major effect on the flaming behavior of PBT, for example, it is
Polyesters, especially poly(butylene terephthalate) (PBT), is one harder to remove burning PBT by melt flow or dripping during the
of the most important engineering polymers with a wide range of burn. Thus many studies focused on the fabrication and improve-
applications as it possesses good dimensional stability, low mois- ment of the flame-retarded glass fiber reinforced PBT. Up to now,
ture absorption, high heat distortion temperature, good chemical the most commonly used flame retardant systems for glass fiber
resistance, excellent electrical properties, creep resistance with reinforced PBT are composed of additives. Typical flame retardant
high strength and rigidity [1,2]. Thus, it is one of the most additives for glass fiber reinforced PBT includes: halogenated sys-
frequently used engineering thermoplastic materials [3]. Typical tems, organic, or inorganic compounds containing phosphorus,
applications include electrical connectors, switches, automotive inorganic metal hydroxides, boron, and nitrogen compounds and
under hood components, electrical appliance components and combinations [5e10].
outdoor telecommunications enclosures. Fiber reinforced com- Injection molding is a very common processing method for glass
posite materials have been adapted to improve the mechanical and fiber reinforced flame retardant PBT (GFFRPBT), because it provides
thermal properties of PBT. Today, glass fiber reinforced PBT is used products with dimensions steadiness and precision, a low manu-
in various insulating parts for electrical engineering and elec- facture cycle and low cost [1,9]. Denault et al. [11] investigated the
tronics, such as electronic relay, communication and automobile mechanical performances of a new class of injection molded long
applications. Therefore, the investigation of the processing of fiber fiber composites based on PP and PBT matrix. Effects of material
parameters such as glass fiber concentration, orientation, breakage,
and PBT matrix composition were analyzed. It was concluded that
* Corresponding author. Fax: þ86 20 87110029.
higher mechanical performances of the long fiber reinforced ther-
E-mail address: jiangguo@scut.edu.cn (J. Guo). moplastics will be attained by the injection molding process to

http://dx.doi.org/10.1016/j.polymdegradstab.2014.04.009
0141-3910/Ó 2014 Elsevier Ltd. All rights reserved.
 Z. Shuidong et al. / Polymer Degradation and Stability 105 (2014) 140e149 141

further reduce fiber breakage. During the injection molding, the length. In this work, the inner structure of RGFFRPBT, such as the
glass fiber would be damaged by screw shear and degrade the glass carboxyl content and intrinsic viscosity of PBT, average glass fiber
fiber reinforced composites. Joshi et al. [12] prepared the blends of length and distribution and the morphology were studied. Simul-
PBT with high density polyethylene (80:20 wt %) reinforced with taneously, the properties of RGFFRPBT including rheological, me-
short glass fibers to enhance their mechanical properties. Since it chanical, thermal, flame retardant properties and content of
has been observed that, considerable glass fiber damage occurs bromine were characterized by capillary rheometer, tensile, flexural
during injection molding process, reduction of the average fiber and notched impact test, TG, LOI value, UL-94 test, Glow Wire
length and decreasing the reinforcement efficiency. The fiber Ignition Temperature (GWIT) and X Ray Fluorescence (XRF).
orientation and length distribution, interfacial bond strength, as
well as the glass fiber concentration was responsible for the vari- 2. Experimental
ation in mechanical properties of the composite. Fung et al. [1,13e
15] found that some parameters of the injection molding process, 2.1. Materials
greatly affects the quality of products. The Taguchi method and
principal component analysis was suitably applied to find out the PBT-3316 incorporated with 30% glass fiber and flame retardant
most influential injection molding parameter for single and mul- (brominated polystyrenics (BPS) and Sb2O3) was purchased from
tiple response problems, optimizing the injection molding process Polyplastic Co., Ltd. (Japan) and used as received. All the analytical
for friction properties of glass fiber reinforced PBT. Michael et al. reagents, containing phenol, tetrachloroethane, benzyl alcohol,
[16] investigated the glass fiber reinforced PBT by injection molding chloroform and sodium hydroxide were all and purchased from
in a ribbed plaque. Flat strips, tensile bars, and rib sections were cut Third Chemical Reagent Factory of Tianjin (China) and used as
from the plaque in the flow and cross flow directions. The result received.
illuminated fiber alignment and the loading mode strongly influ-
enced the resultant SeN (stress versus number of cycles to fail) 2.2. Sample preparation
curve. Khamsehnezhad et al. [17] studied the mechanical proper-
ties of single-gated (SG) and double-gated (DG) injection molded PBT-3316 was dried at 120
C, for 4 h before usage. After that,
blend of polybutylene terephthalate (PBT) and polycarbonate (PC), mechanical, thermal and flammability tests specimens were in-
the composites contained different contents of (øf) short glass fi- jection molded from the dried PBT-3316 using an injection molding
bers (SGF) such as 15, 20 and 30 wt%. Results showed that the machine (Germany, ARBUR 420M, ø ¼ 25 mm, L/D ¼ 28) with 2-
weldline (WIF) factor for the mechanical properties of PBT/PC/SGF plate mold at 265
C, 270
C, 270
C, 275
C, 280
C, respectively.
decreased linearly with increasing SGF content. Hashemi et al. [18] The mold temperature was kept at 55
C. The nozzle diameter of the
studied how the fiber concentration, strain rate, temperature and machine was 2.5 mm and the sprue diameter of the 2-plate mold
the type of loading, affects the strength of single and double-gate was 2.8 mm. The rotational speed of screw was 25 m/min and
injection molded PBT reinforced by short glass fiber with 0, 10, 20 three-stage injection rate was 80, 120 and 100 mm/s, respectively.
and 30% by weight. It was found that, the tensile strength of the The injection pressure and reserve pressure was 1850  100 bar and
single-gate and double-gate moldings SGFPBT increased linearly 2000  100 bar, respectively. The cycle time of the process was 16 s,
with the natural logarithm of strain rate and decreased linearly including injection time (1.2 s), reserve pressure time (2.1 s) and
with increasing temperature. cooling time (12 s). Ten specimens were collected to measure the
Up to now, waste electrical and electronic equipment (WEEE) properties. The other specimens and recycled materials were
has become a serious problem in environmental protection and smashed into about 5  3  2 mm size piece by low speed crusher
human health. It was estimated that 1.8 Mt of waste plastics con- (Sweden, Rapid Co ltd, PMD-2L), named as RPBT-3316-1st. The
taining at least 159 000 tons of brominated flame retardants from RPBT-3316-1st collected by injection molding was reprocessed and
monitors every year. The degradation of flame-retarded polyester marked as RPBT-3316-2nd. When RPBT-3316-2nd was reprocessed,
from e-waste became a point of interest due to the increasing the recycled material was marked as RPBT-3316-3rd, the PBT-3316
awareness of the need to recycle plastic waste. PBT, polyamide 6 reprocessed for four and five cycles were prepared by the same
(PA6) and polyamide 6, 6 (PA66), polycarbonates (PC), thermo- method and marked as RPBT-3316-4th and RPBT-3316-5th,
plastic polyurethanes (PUR), as well as their blends, are thermally- respectively. RPBT-3316-1st, RPBT-3316-2nd, RPBT-3316-3rd,
processed at very high temperatures with the rate of these degra- RPBT-3316-4th, RPBT-3316-5th (were called by a joint name as
dation reactions extremely rapid [19]. These degradation reactions RPBT-3316) were crashed by high-speed knapper and power sam-
result in severe loss of molecular weight, mechanical and thermal ples with a total of 0.2 g immersed in 500 ml 1,2-dichloroethane at
properties, reducing the possibility for reprocessing large amounts 40
C for 12 h by ultrasonic to dissolve the brominated poly-
of post-industrial regrind or postconsumer recycle into demanding styrenics (BPS). Then the solution was cooled to room temperature
engineering applications. When GFFRPBT is applied in various and filtered through a disposable membrane filter made from
insulating parts for electrical engineering and electronics, during Teflon. The obtained solid was washed three times with 100 ml
the process of injection molding, it would produce recycled ethanol and dried at 100
C for 12 h. The dried solid was maintained
GFFRPBT from scraps and unqualified injection products. In general, in 1000 ml mixture of phenol/1,1, 2,2-tetrachloroethane (50/50, w/
the proration part of recycled GFFRPBT was higher than 45 wt%. It w) at 80
C for 24 h to achieve complete PBT dissolution. Then the
was important to reuse the recycled GFFRPBT due to economical solution was cooled to room temperature and filtered. The solution
and environment protection. was reduced using pressure distillation to remove the phenol/
In this paper, the reprocessed GFFRPBT acting as the main part of 1,1,2,2-tetrachloroethane and the solid PBT was collected to mea-
the recycled GFFRPBT is collected from GFFRPBT at different injec- sure the carboxyl end-group content and intrinsic viscosity.
tion molding cycles. The aim of this investigation was to establish
the relationship between structure and properties of reprocessed 2.3. Characterization
GFFRPBT (RGFFRPBT) and evaluate how the rheological, mechanical,
thermal and flame retardant properties of GFFRPBT varied during 2.3.1. Carboxyl end-group content measurement
injection molding. In theory, the properties of RGFFRPBT are Carboxyl end-group content (CC) of the PBT was determined
significantly affected by the degradation of PBT and glass fiber according to Pohl’s method [20] by titrating solution of the resin in
142 Z. Shuidong et al. / Polymer Degradation and Stability 105 (2014) 140e149

benzyl alcohol/chloroform with standard NaOH in benzyl alcohol in 3. Results and discussion
the presence of phenol red as indicator. Intrinsic viscosity of PBT
was performed using an Ubbelohde viscometer at 30
C in a 3.1. Carboxyl content and intrinsic viscosity of RPBT-3316
mixture of phenol/1,1,2,2-tetrachloroethane (50/50, w/w).
For PBT, the intrinsic viscosity decreased, while the carboxyl end-
groups increase during the thermal-process [21], due to suffering
2.3.2. Glass fiber length measurement the screw shear and thermal degradation. In this work, the carboxyl
A bum-out test (also called Ashing) was used to estimate the content (CC) and intrinsic viscosity of RPBT-3316 were investigated
average fiber length according to a standard procedure [12]. 10.0 g and summarized in Table 1. As shown in Table 1, the CC and intrinsic
sample of RPBT-3316-1st, RPBT-3316-2nd, RPBT-3316-3rd, RPBT- viscosity of PBT was 65 mg/kg and 1.03 for PBT-3316. For RPBT-3316-
3316-4th, RPBT-3316-5th was kept in a muffle furnace at 500
C for 1st, the CC increased to 106 mg/kg, while its intrinsic viscosity
3 h, respectively, to burn off the PBT matrix. The ash of fibrous decreased to 0.74. It means that PBT was degraded during the in-
material was subsequently spread on glass slides. The lengths of the jection molding process. When reprocessing cycle was extended
residual glass fibers were measured using scanning electron mi- from three to five, the CC of RPBT-3316 increased from 224 mg/kg to
croscope (SEM, S-3700, Hitachi, Japan) on a sampling of at least 600 262 mg/kg, while its intrinsic viscosity decreased from 0.54 to 0.41.
glass fibers were carried out. RPBT-3316-5th showed the highest CC and lowest intrinsic viscosity.
This showed that, the thermal degradation reaction of PBT-3316
occurs when the reprocessing cycle increases. During the thermal-
2.3.3. Scanning electron microscopy of RGFFRPBT
process with high pressure, both the shear by injection screw and
The features of the fracture surface of the test specimen, ob-
draw by runner could snip the PBT molecular chain, and the degree
tained after tensile failure at room temperature, were gold-
of damage increased with injection cycle increased. The scission of
sputtered and studied on the same scanning electron microscope.
PBT chain could also be induced by temperature (thermal degra-
dation). The result was related to degradation reactions causing the
2.3.4. Tensile testing, flexural testing and impact strength studies formation of short chain fragments with carboxylic and unsaturated
After PBT-3316 and RPBT-3316 were injection molded as stan- ester end-groups [22].
dard bars, tensile and flexural properties of samples were measured
on a Universal test machine (Reger, RGT-20A) according ASTM
D638-03 and ASTM D790-07, respectively. The notched impact 3.2. Average length glass fiber (L) and distribution
strength was performed by ASTM D256 using an impact tester
(Sans Tested, ZBC1400-2). In all cases, a minimum of five specimens The mechanical properties of fiber reinforced thermoplastic
were tested for determining an average value. composites are influenced by the length distribution of the fibers.
In this paper, the variation of glass fiber degradation and length
distribution of glass fiber during injection molding was investi-
2.3.5. Rheological measurement gated. The L of PBT-3316 and RPBT-3316 were listed in Table 1(list
The shear viscosities of all the samples were measured using a 3). It demonstrated that when PBT-3316 was processed by injection
capillary rheometer (Rheograph 25, GÖTTFERT). The capillary die molding [23,24], the fiber length was reduced. The L of PBT-3316
used has a length-to-diameter ratio of 30:1. Before testing, the before the injection process, was 299.8 um, after the first injec-
samples were dried at 120
C for 4 h. The test temperature was 240, tion molding, the L became 259.6 um for RPBT-3316-1st and
250, and 260
C, respectively. The shear stress ranged from 55.1 to decreased by 13.4%, compared with PBT-3316. During injection
451 bar and shearing rate was from 50.4 to 3490 s1. molding, when PBT-3316 passed through the screw, the nozzle and
gate region of the mold, glass fiber attrition caused by fiber inter-
action, fiber contact with machine surface, and the action of flow
2.3.6. Thermogravimetric analyses (TGA) measurement stresses in the melt were responsible for the reduction of the
PBT-3316 and RPBT-3316 were crashed by high-speed knapper average glass fiber length. Moreover, the glass fibers break pro-
and dried at 120
C for 4 h. TGA analysis was carried out using a gressively resulting in a loss of reinforcement efficiency because it
PerkineElmer Pyris thermal analyzer under air atmosphere at a encountered high shear rate conditions [12]. Since the glass fibers
heating rate of 10
C/min and the range of scanning temperature suffer severe damage during the thermal-process, their ultimate
was from 50 to 500
C. length in the finished part is usually only a fraction of its initial
value [25]. As a result, the L of RPBT-3316 decreased with the in-
crease of reprocessing cycles.
2.3.7. Flame retardant measurement
Kelly and Tyson [11] proposed a fundamental relationship be-
The LOI value was measured according to ISO 4589-1984 and
tween the critical fiber length (lc) and the interfacial shear strength
the dimension of all the samples were 130  6.5  3.2 mm. The
(si) as given below:
vertical burning tests were performed according to UL-94 and the
dimension of all the samples were 127  10  0.75 mm, GWIT tests
were performed in accordance with IEC 606955-2-13:2000, Table 1
respectively, the dimension of all the samples were Effect of injection molding cycles on carboxyl content, intrinsic viscosity and length
80  80  0.75 mm, and the values reported reflected an average of glass fibers of RPBT-3316.

from five tests. Samples Carboxyl Intrinsic viscosity Average length

content (mg/kg) of GF (um)

PBT-3316 65 1.03 299.8

2.3.8. Bromine content measurement RRPBT-3316-1st 106 0.74 259.6
XRF (Shimadzu EDX-720) was applied to measure the content of RRPBT-3316-2nd 184 0.61 212.8
bromine of RPBT-3316, the values reported reflected an average RRPBT-3316-3rd 224 0.54 202.3
from ten tests and the voltage and current was 5 kV, 5 uA, RRPBT-3316-4th 236 0.48 185.3
RRPBT-3316-5th 262 0.41 167.4
respectively.
 Z. Shuidong et al. / Polymer Degradation and Stability 105 (2014) 140e149 143

was able to work as reinforcement. As it could be seen from

df d
Lc ¼ (1) Fig. 1(a), the percentage of glass fibers length higher than the
2si critical length (Vj) is 0.539, with 18.8% of the lengths larger than
This equation clearly indicates that as critical fiber length de- 450 um. The fiber length distribution enables these fibers acting as
creases interfacial shear strength increases thereby enhancing the reinforcement of the PBT-3316. Since the stress transfer between
composite strength. Where d is the fiber diameter (13.0 um fiber and matrix was also related to the glass fiber critical length, a
determined by SEM and calculated from 50 fiber diameter). Ac- decrease in the glass fiber length leads to only a smaller number of
cording to Deshmukh et al. [25], the value of si is 19.4 MPa, and the fibers been able to work as reinforcement. The fiber length
when the glass fiber was processed, the fiber fracture strength (df ) distribution of RPBT-3316-1st is presented in Fig. 1(b). Compared
of glass fiber is 750 MPa [26]. From these given parameters, it can with those of PBT-3316, the percentage of the glass fiber lengths
be calculated that the Lc is 252.3 um from Eq. (1). which were shorter than 150 um increased, meanwhile the Vj was
The degradation of glass fiber during injection molding process 0.354 resulting in 6.8% of the lengths becoming larger than 450 um.
could be characterized by fiber length distribution shown in Fig. 1. The Vj of RPBT-3316-2nd, RPBT-3316-3rd, RPBT-3316-4th, RPBT-
With the increase of reprocessing cycle, the percentage of the glass 3316-5th decreased to 0.312, 0.265, 0.254 and 0.198, respectively.
fibers length, which was higher than the critical length (Vj) Moreover, the percentage of the glass fibers with larger lengths
decreased, indicating that only a small number of the glass fiber (higher than 450 um) decreased significantly as presented in

Fig. 1. Glass fiber length distribution of PBT-3316 and RPBT-3316.

144 Z. Shuidong et al. / Polymer Degradation and Stability 105 (2014) 140e149

Fig. 2. (a) SEM of PBT-3316 (100). (b) SEM of PBT-3316 (500). (c) SEM of RPBT-3316-1st (100). (d) SEM of RPBT-3316-1st (500). (e) SEM of RPBT-3316-2nd (100). (f) SEM of
RPBT-3316-2nd (500). (g) SEM of RPBT-3316-3rd (100). (h) SEM of RPBT-3316-3rd (500). (i) SEM of RPBT-3316-4th (100). (f) SEM of RGFFRPBT-5th (500). (j) SEM of RPBT-
3316-4th (500). (k) SEM of RPBT-3316-5th (100). (l) SEM of RPBT-3316-5th (500).
 Z. Shuidong et al. / Polymer Degradation and Stability 105 (2014) 140e149 145

Fig. 1(c)e(f). Factors that affected the degradation of the glass fiber Table 2
were related to the processing and it indicated that, the longer glass Effect of degradation of intrinsic viscosity and glass fiber on mechanical properties of
RPBT-3316.
fibers were cut shorter during the process of injection molding.
When the composite was subjected to injection molding, the shear Samples Notched impact Tensile Flexural
mixing in the screw preplasticization would cause breakage of the strength (J/m) strength (MPa) strength (MPa)

fibers and induced further fiber attrition through fiberefiber in- PBT-3316 75  2.6 115.6  9.5 161.9  10.3
teractions and fracture at the solidemelt interface during the RRPBT-3316-1st 67  3.5 109.3  8.6 143.4  8.5
RRPBT-3316-2nd 54  3.9 96.1  6.9 128.1  7.4
compression/solidification phase [23]. The degradation of L and Vj
RRPBT-3316-3rd 42  3.2 85.9  7.4 115.1  5.6
has a negative effect on the polymer composite [25] because they RRPBT-3316-4th 36  3.2 63.7  4.6 105.2  5.7
increase the amount of fiber ends, which are sites for stress con- RRPBT-3316-5th 33  2.5 54.9  3.5 95.2  5.1
centration where initiation and propagation of interfacial cracks
could occur, leading to tensile failure of the composite.
higher values compared to PBT-3316. Good interfacial adhesion
3.3. Morphology of RPBT-3316 between glass fiber and PBT matrix of RPBT-3316-1st was
observed, pulled-out-glass fibers and craze remnants on the
In order to study the difference in reinforcing effect of glass fiber fracture surface occurred scarcely. Compared with those of PBT-
to PBT matrix, the fracture surface of tensile strength specimens for 3316, the intrinsic viscosity, L and Vj of RPBT-3316-2nd
PBT-3316 and RPBT-3316 were analyzed by SEM. It could be seen decreased significantly, its mechanical properties decreased. The
from Fig. 2 that the fracture surface had little deformation, indi- notched impact strength, tensile and flexural strength of RPBT-
cating brittle fracture behavior. Furthermore, Fig. 2(a and b) pre- 3316-2nd was 54 J/m (72% of PBT-3316), 96.1 MPa (83.1% of PBT-
sented a feature of major fiber breakage with several glass fibers 3316) and 128.1 MPa (79.1% of PBT-3316), respectively. The shear
pulled-out, which was an indication of a good adhesion between by screw was responsible for the degradation of the PBT molecules
glass fiber and PBT matrix. It was noted that the surfaces of the glass and L, the quantity of pulled-out-glass fibers and craze remnants
fibers became rougher and seem to be covered with a thin layer on the fracture surface was enhanced for RPBT-3316-2nd. When
(the PBT matrix or coupling agent), indicating that most of the glass the injection molding cycles of PBT-3316 was increased up to five,
fibers were well trapped by PBT matrix. Compared to the both L and Vj of RPBT-3316-5th exhibited the lowest value,
morphology of PBT-3316, RPBT-3316 had an increased ratio of implying that reinforced efficiency of glass fiber was lowest.
pulled-out-glass fiber. With an increase of screw shear, the ductile Moreover, glass fiber pullout had been observed in various short
failure of the PBT matrix changed to a brittle one, which was fiber reinforced thermoplastics as the dominant failure [30,31]
associated with the formation of crazing. More craze remnants on revealing that the RPBT-3316-5th specimen showed a brittle
the fracture surface were clearly visible both in PBT matrix (Fig. 2ce fracture mechanism proved by Fig. 2(k and l).
l), suggesting that the matrix continuity was destroyed. It was In summary, with the increase of injection molding cycles, the
found that the pulled-out of glass fiber dominated fracture surface shear of screw acting on the PBT and glass fiber increased, causing
of RPBT-3316-5th, parallel to this, matrix brittleness, and the the decrease in intrinsic viscosity, L and Vj. Thomason [32]
moderate adhesion between glass fiber and PBT was further demonstrated that, Charpy impact strength increases initially
reduced [27]. During the injection molding process, the shear by with fiber length but levels out above 6 mm. The mechanical
the screw would destroy the initially good interfacial adhesion properties of RPBT-3316-5th decreased significantly, though RPBT-
between glass fiber and PBT matrix, the glass fiber was cut down 3316-1st and RPBT-3316-2nd obtained acceptable mechanical
accompanying the molecular degradation of PBT. From 3.2, it could properties and could be reused in the electronic and electrical field
be observed that, L and Vj of RPBT-3316 decreases with increasing such as relay shell, insulate switches and so on.
the injection molding cycle, thus decreasing the interfacial adhe-
sion. As a result, it was easier to pullout the glass fiber from the PBT
matrix and craze remnants occurred. RPBT-3316 would suffer a 3.5. Rheological behavior of RPBT-3316
decreased mechanical and thermal properties once poor fiber-
matrix adhesion, voids and other flaws occurred [12]. In order to further investigate the effect of degradation of mo-
lecular weight and glass fiber on rheological behavior of RPBT-3316,
3.4. Mechanical properties of RGFFRPBT the transient stress responses of samples with the start-up of
steady shear flows with various shear rates and temperature were
It is well known that tensile behavior [28] and impact strength studied. Shear viscosity versus shear rate for PBT-3316 and RPBT-
[29] are dependent on the fiber length and the fiber orientation in 3316 at 240
C, 250
C, 260
C are presented in Fig. 3. All samples
relation to the loaded direction. Practically, once the Vj decreases, exhibited shear-thinning behavior with the increase of shear rate,
the part of the glass fiber with length shorter than lc could not and the magnitude of the shear viscosity was strongly dependent
reinforce the thermoplastic, resulting a decrease of the mechanical on both the temperatures and the shear rates. With the increase of
properties of the composites [28,29]. temperature and shear rates, all samples show lower shear vis-
The notched impact strength, tensile and flexural strength of cosity and elucidated the characterization of Non-Newtonian Fluid:
RPBT-3316 are shown in Table 2. With increasing the injection shear-thinning and nonlinear viscoelastic properties. A significant
molding cycle, the intrinsic viscosity, L and Vj of RPBT-3316 de- reduction in viscosity of RPBT-3316 at higher shear rate was
creases, resulting in a reduction of the mechanical properties of observed. At low shear rate, the fibers were disoriented and the
RPBT-3316. RPBT-3316-1st obtained acceptable mechanical prop- probability of a fiberefiber collision was much higher. The collision
erties, though the properties had a small decrease. The notched increased with fiber loading resulting in an increase of the shear
impact strength, tensile and flexural strength of RPBT-3316-1st viscosity of RPBT-3316. However, as the shear rate increased, the
was 67 J/m (89.3% of PBT-3316), 109.3 MPa (94.5% of PBT-3316), stress was high enough to overcome the percolation as the formed
and 143.4 MPa (88.6% of PBT-3316), respectively. From Table 1, network by fiberematrixefiber was destroyed, the molten chains
Fig. 1(b) and Fig. 2(c and d), it can be seen that, both the average were free to move [33], resulting in decreasing the shear viscosity
length of glass fiber and intrinsic viscosity could obtain rather of RPBT-3316.
146 Z. Shuidong et al. / Polymer Degradation and Stability 105 (2014) 140e149

viscosity of RRPBT-3316-1st was 0.74 lower than that of PBT-3316.

Intrinsic viscosity of RPBT-3316-4th and RPBT-3316-5th compared
to that of PBT-3316, decreased significantly. Secondly, composites
with shorter glass fiber length have lower shear viscosity. The L of
RRPBT-3316-1st was 259.6 um, while the RPBT-3316-4th decreased
to 185.3 um, as the glass fiber length decreased, both the chances
for fiber orientation and fiber-matrix interaction decreased [34].
Increasing the injection molding cycles, the disentanglement of
molecular chains and misalignment of glass fiber, as well as the
melt flow properties of RPBT-3316 would be enhanced. According
to the results, RPBT-3316 achieved lower viscosity than that of PBT-
3316, which was beneficial to reduce the injection defect such as
weldline, short molding, and glass fiber emergence on the surface
and so on.

3.6. Thermal stability of RPBT-3316

The decomposition of pure PBT has been established [34,35] and

it was reported that the decomposition temperature of 2% mass loss



(T2%) and maximum mass loss (Tmax%) of PBT was 350 C and 400 C
[36e38], respectively. TG and DTG of RPBT-3316 is shown in Fig. 4,
both TG and DTG curves of RPBT-3316 were shifted to lower tem-
peratures with the increasing of injection molding cycles. It could
be illustrated that the thermal decomposition of RPBT-3316 pro-
ceeds in two-step process between 325
C and 450
C. The first
process (from 325
C to 350
C) resulted in the brominated poly-
styrenics (BPS) decomposition, while the second thermal decom-
position step after 380
C was attributed to the mass loss of PBT. For
glass fiber reinforced flame retardant system, BPS was first evolved
at slightly lower temperatures than those of the flammable product

Fig.3. Rheological behavior of PBT-3316 and RPBT-3316 at different temperature,

240
C (a), 250
C (b), and 260
C (c).

At the same temperature and shear rate, PBT-3316 obtained a

higher shear viscosity, while the shear viscosity of RPBT-3316-1st
decreased significantly. With the increasing of injection molding
cycles, the shear viscosity of RPBT-3316 decreased, RPBT-3316-4th
showed the lowest shear viscosity (the shear viscosity of RPBT-
3316-5th was too low to be tested at 240
C). In general, two rea-
sons were responsible for the reduction. Firstly, polymer with lower Fig. 4. Thermal decomposition of PBT-3316 and RPBT-3316 (a: TG curves; b: DTG
intrinsic viscosity shows lower shear viscosity. The intrinsic curves).
 Z. Shuidong et al. / Polymer Degradation and Stability 105 (2014) 140e149 147

Table 3 PBT-3316R was observed, it could exhibit good thermal stability

Effect of degradation of intrinsic viscosity and glass fibers on flame retard properties during the injection molding and was independent to injection
of RPBT-3316.
molding cycles. As a result, RPBT-3316 could maintain the original
Samples LOI UL-94 Br content（%） flame retardant properties as PBT-3316.
PBT-3316 32.8  0.3 V-0 4.85  0.05 GWIT is defined as the temperature, which is 25
C hotter than
RRPBT-3316-1st 32.6  0.3 V-0 4.80  0.04 the maximum temperature of the glow wire, and does not cause
RRPBT-3316-2nd 32.5  0.2 V-0 4.80  0.05 ignition of the material during three subsequent tests. Ignition is
RRPBT-3316-3rd 32.4  0.4 V-0 4.75  0.06
defined as the appearance of flames on the surface of the spec-
RRPBT-3316-4th 32.4  0.4 V-0 4.75  0.05
RRPBT-3316-5th 32.1  0.1 V-0 4.75  0.04 imen for at least 5s [40,41]. Francesco et al. [42] proposed a model
(presented in Fig. 5) to explain the polymer behavior during glow
wire test: in the part of the specimen which was in contact with
evolution from the substrate polymer [34], causing the initial the wire, 3 zones can be recognized: A first zone at temperature
flame-retarding effect and decreasing the T2% of composites. The T2% above the decomposition temperature (Pyrolysis Zone); A second



of samples ranged from 340 C to 350 C, Tmax% was between 380
C zone where the material is just melted, and the degradation is
and 390
C, and the total mass loss reached over 60% at its final slower (Melting Zone); A third zone where the temperature is not
point (500
C). Compared with those of PBT-3316, T2% and Tmax of high enough to melt the polymer, but the material can be
RPBT-3316-1st and RPBT-3316-2nd decreased mildly. With the in- deformed because it is above its heat distortion temperature
crease of injection molding cycles, RPBT-3316 showed both lower (Softening Zone). In their opinion, the processes taking place in
molecular weight and higher CC. The intrinsic viscosity of PBT the pyrolysis zone mainly determine the test results. Furthermore,
decreased from 0.61 to 0.41, meaning that molecular weight of PBT both mechanical and rheological properties can have an important
decreased synchronously, while the CC of PBT increased from influence on GWIT properties [42]. The data of GWIT is listed in
224 mg/kg to 262 mg/kg, with thermal stability of RPBT-3316 Table 4, it shows that, the GWIT of RPBT-3316 increased with an
decreasing significantly. Compared with those of PBT-3316, the increase in the injection molding cycle. The ignition time, extin-
values of T2% and Tmax of RPBT-3316-5th were reduced by 7.8
C and guish time, ignition duration and weight loss of PBT-3316 was
6.6
C, respectively. According to research of Bikiaris et al. [39], 5.5 s, 24.3 s 18.8 s and 16.5%, respectively. Compared with those of
higher CC may contribute to lower stability of PBT. Thus, the ther- PBT-3316, the ignition time of RPBT-3316 decreased from 5.5 s to
mal stability of RPBT-3316 decreased with increasing injection 0.5 s, while the extinguish time, ignition duration and weight loss
molding cycles. Fortunately, the lowest T2% was 342.8
C, it guar- increased to a higher value. From Section 3.4, 3.5 and 3.6, it is seen
anteed the RPBT-3316 to obtain good thermal stability as the melt that the heat flow properties of RPBT-3316 is improved, while the



processing temperature of RPBT-3316 ranged from 265 C to 285 C. thermal stability decreases. According to the Francesco model [42],
the Pyrolysis Zone, Melting Zone and Softening Zone of RPBT-3316
3.7. Flame retardant properties of RPBT-3316 were longer than those of PBT-3316, respectively. The mechanical
properties and shear viscosity of RPBT-3316-4th and RPBT-3316-
LOI values and UL-94 rating tests were used to investigate the 5th were much lower than those of PBT-3316, implying that
flame retardance of the composites, the results were summarized they were easier to soften and melt during the GWIT test, causing
in Table 3, it could be seen that RPBT-3316 maintained the original the hot wire to penetrate and ignite the specimen more easily,
LOI value of PBT-3316. The LOI value of PBT-3316 was 32.8, while corresponding to the lengthened ignition time. Moreover, the
the LOI value of RPBT-3316R was higher than 32.0. Meanwhile, weight of melt RPBT-3316-4th and RPBT-3316-5th wrapped
RPBT-3316 could maintain V-0 classification in the UL-94 test. From around the hot tip of the wire increased, prolonging the flame
the TG analysis, it could be seen that RPBT-3316 obtained good time, as a result their extinguish time, ignition duration and
thermal stability at melt process temperature (265
Ce285
C) and weight loss increased. This kind of behavior was observed for all
thermal degradation of BPS hardly occurred. The result was the materials tested. Despite that, the RPBT-3316 could reach a
confirmed by XRF determination: the Br content of PBT-3316 was GWIT higher than 775
C and pass the test [41], indicating that,
4.85  0.05%, the Br content of RPBT-3316 was higher than 4.71, and the flame retardant properties of RPBT-3316 could match the
only little decrease was observed. Though a decrease BPS content of requirement for electrical insulation application.

Fig. 5. Scheme showing the three-stages of a glow wire test and zones interested in the process.
148 Z. Shuidong et al. / Polymer Degradation and Stability 105 (2014) 140e149

Table 4
Effect of degradation of intrinsic viscosity and glass fibers on flame retard properties of RPBT-3316 (GWIT 775
C).

Samples Ignition time (s) Extinguish time (s) Ignition duration (s) Weight loss (wt)%

PBT-3316 5.5  1.3 24.3  0.5 18.8  0.5 16.5  1.2

RRPBT-3316-1st 2.6  0.5 25.1  1.2 22.5  1.1 23.1  0.9
RRPBT-3316-2nd 2.1  0.3 25.9  0.8 23.8  0.9 25.6  1.3
RRPBT-3316-3rd 1.4  0.4 26.2  0.8 24.7  0.6 26.6  1.5
RRPBT-3316-4th 0.8  0.3 26.6  0.9 25.6  0.5 28.7  2.1
RRPBT-3316-5th 0.5  0.2 27.1  1.1 25.7  0.4 29.1  1.8

4. Conclusions [7] Gao F, Tong LF, Fang ZP. Effect of a novel phosphorous-nitrogen containing
intumescent flame retardant on the fire retardancy and the thermal
behaviour of poly(butylene terephthalate). Polym Degrad Stab 2006;91(6):
In this paper, the relationship between structure and properties 1295e9.
of GFFRPBT was studied: when the PBT-3316 was processed by [8] Braun U, Bahr H, Sturm H, Schartel B. Flame retardancy mechanisms of metal
injection molding, due to the screw shear, the degradation of phosphinates and metal phosphinates in combination with melamine cya-
nurate in glass-fiber reinforced poly(1,4-butylene terephthalate): the influ-
intrinsic viscosity of PBT and the length of glass fiber were ence of metal cation. Polym Adv Technol 2008;19:680e92.
observed. The variance of the components would affect the prop- [9] Yang W, Hu Y, Tai QL, Lu HD, Song L, Richard K, et al. Fire and mechanical
erties of RPBT-3316 in different trends. performance of nanoclay reinforced glass-fiber/PBT composites containing
aluminum hypophosphite particles. Compos Part A 2011;42:794e800.
[10] Chen L, Luo Y, Hu Z, Lin GP, Zhao B, Wang YZ. An efficient halogen-free flame
(1). By increasing the injection molding cycles of RPBT-3316, it retardant for glass-fibre-reinforced poly(butylene terephthalate). Polym
reduces the PBT molecular weight and L, especially the Vj Degrad Stab 2012;97:158e65.
[11] Denault J, Vu-Khanh T, Foster B. Tensile properties of injection molded long
decreases, which was responsible for lowering the mechan- fiber thermoplastic composites. Polym Compos 1989;10:313e21.
ical properties. It seemed that, reprocessing was possible, if [12] Joshi M, Maiti SN, Misra A. Influence of fiber length, fiber orientation and
the PBT-3316 obtained enough long in glass fiber length. interfacial adhesion on poly (butylene terephthalate)/polyethylene alloys
reinforced with short glass fibers. Polym Compos 1994;15:349e58.
(2). Flow property of RPBT-3316 was improved, the decrease in [13] Fung CP, Chung KP. Multi-response optimization in friction properties of PBT
molecular weight of PBT and the length of glass fiber with the composites using Taguchi method and principle component analysis. J Mater
increasing of the injection molding cycle was responsible for Process Technol 2005;170:602e10.
[14] Jeng MC, Fung CP, Li TC. The study on the tribological properties of fiber-
decreasing the shear viscosity of RPBT-3316.
reinforced PBT composites for various injection molding process parame-
(3). The thermal stability of RPBT-3316 decreased with increasing ters. Wear 2002;252:934e45.
the degradation of the molecular weight of PBT, and the [15] Fung CP, Hwang JR, Hsu CC. The effect of injection molding process parame-
length of the glass fiber. As the temperature of injection ters on the tensile properties of short glass fiber-reinforced PBT. Polym Plast
Technol Eng 2003;42:45e63.
molding (265
Ce285
C) was much lower than T2% of RPBT- [16] Wyzgoski Michael G, John AK, Glen EN. Fatigue of fiber-reinforced injection
3316, BPS did not decompose during the melt process. This molded plastics. Polym Compos 2004;25:489e98.
enabled RPBT-3316 to achieve a UL-94 V-0 classification, [17] Khamsehnezhad A, Hashemi S. Mechanical properties of single- and double-
gated injection moulded short glass fibre reinforced PBT/PC composites.
obtain a higher LOI value (>32), and pass the GWIT (775
C) J Mater Sci 2008;43:6344e52.
test. [18] Hashemi S. Temperature, strain rate and weldine effects on strength and
(4). The results demonstrates that, the properties of RPBT-3316- micro-mechanical parameters of short glass fibre reinforced polybutylene
terephthalate(PBT). Polym Test 2011;30:801e10.
1st and RPBT-3316-2nd could match the demand of the [19] Villalobos M, Awojulu A, Greeley T, Turco G, Deeter G. Oligomeric chain ex-
electronic and electric field and could be reused solely in tenders for economic reprocessing and recycling of condensation plastics.
fields such as relay, electrical connectors, switch and so on. Energy 2006;31:3227e34.
[20] Passalacqua V, Pilati F, Zamboni V, Fortunato B, Manaresi P. Thermal degra-
dation of poly(butylene terephthalate). Polymer 1976;17:1044e8.
Acknowledgments [21] Firas A, Daver F. Recycled poly(ethylene terephthalate) chain extension by a
reactive extrusion process. Polym Eng Sci 2004;44:1579e87.
[22] Brgeret A, Bozec MP, Quabtin JC, Crespy A, Gasca JP, Arpin M. Study of
The authors wish to acknowledge the financial support of the interphase in glass fiber reinforced poly(butylene terephthalate) composites.
National Science Fund for 50903032 and U1333126, Fund Research Polym Compos 2004;25(1):12e25.
[23] Chen F, Jones FR. Injection moulding of glass fibre reinforced phenolic com-
Grant for Science and Technology Bureau of Jinjiang, Fujian (Jin- posites: 1-Study of the critical fibrelength and the interfacial shear strength.
jiang Financial Grant No. 2010314, No. 2011200), and Program of Plast Rubber Compos Process Appl 1995;23:241e8.
the Pearl River Young Talents of Science and Technology in [24] Singh R, Chen F, Jones FR. Injection moulding of glass fibre reinforced phenolic
composites: 1-Study of the critical fibre length and the interfacial shear
Guangzhou, China (2013J2200013) and the Fundamental Research strength. 2: study of the injection molding process. Polym Compos
Funds for the Central Universities (2013ZZ0011). 1998;19(1):37e47.
[25] Deshmukh GS, Peshwe DR, Pathak SU, Ekhe JD. Effect of functionalized elas-
tomer addition on mechanical and interfacial properties of poly (butylene
References terephthalate)/glass fiber composites. Polym Compos 2012;33(1):58e67.
[26] Zou NY, Ye DQ, Jiang ZZ, editors. Application technologies of glass fibres.
[1] Fung CP. Fibre orientation of fibre-reinforced PBT composites in injection Beijing: China Petrochemical Press; 2004.
moulding. Plast Rubber Compos 2004;33:170e6. [27] Kӧppl T, Brehme S, Wolff-Fabris F, Altstädt V, Schartel B, Dӧring M. Structure-
[2] Shen HB, Nutt S, Hull D. Direct observation and measurement of fiber archi- property relationships of halogen-free flame-retarded poly(butylene tere-
tecture in short fiber-polymer composite foam through micro-CT imaging. phthalate) and glass fiber reinforced PBT. J Appl Polym Sci 2012;124:9e18.
Compo Sci Tech 2004;61:1529e38. [28] Kelly A, Tyson WR. Tensile properties of fibre-reinforced metals: copper/
[3] Chisholm BJ, Fong PM, Zimmer JG, Hendrix R. Properties of class-filled ther- tungsten and copper/molybdenum. J Mech Phys Solids 1965;13:329e50.
moplastic polyesters. J Appl Polym Sci 1999;74:889e97. [29] Cottrell AH. Proceedings of the royal society of London. Mathem Phys Sci A
[4] Yoshioka T, Tsuji M, Kawahara Y, Kohjiya S, Manabe N, Yokota Y. Morpho- 1964;282:2e9.
logical study by TEM on uniaxially oriented thin films of PBT. Polymer [30] Mohd ZAI, Lim NC. Effect of moisture absorption on the tensile properties of
2005;46:4987e92. short glass fiber reinforced poly(butylene terephthalate). Polym Eng Sci
[5] Tjahjadi M, Gallucci RR. Tracking resistance of flame retardant glass reinforced 1994;34:1645e55.
PBT. J Elas Plast 1999;31(3):205e12. [31] Yow BN, Ishiaku US, Mohd IZA, Karger-kocsis J. Fracture behavior of rubber-
[6] Park JM, Park YH. Synthesis and properties of novel flame retardant poly(- modified injection molded poly(butylene terephthalate) with and without
butylene terephthalate). Macrom Res 2005;13(2):128e34. short glass fiber reinforcement. J Appl Polym Sci 2002;84:1233e44.
 Z. Shuidong et al. / Polymer Degradation and Stability 105 (2014) 140e149 149

[32] Thomason JL, Vlug MA. Influence of fibre length and concentration on the [38] Botelho G, Queiros A, Gijsman P. Studies on thermal and thermo-oxidative
properties of glass fibre-reinforced polypropylene. 4: impact properties. degradation of poly (ethylene terephthalate) and poly (butylene tere-
Compos Part A 1997;28:277e88. phthalate). Polym Degrad Stab 2001;73:431e5.
[33] Nair KCM, Kumar RP, Thomas S, Schit SC, Ramamurthy K. Rheological [39] Bikiaris DN, Karayannidis GP. Effect of carboxylic end groups on thermoox-
behavior of short sisal fiber-reinforced polystyrene composites. Compos Part idative stability of PET and PBT. Polym Degrad Stab 1999;63:213e8.
A 2000;31:1231e40. [40] Vendramini J, Mele P, Merle G. Mechanical modeling of the transcrystalline
[34] Levchik SV, Weil ED. A review on thermal decomposition and combustion of interphase behavior in commingled PBT/glass fiber composites. J Appl Polym
thermoplastic polyesters. Polym Adv Technol 2004;15:691e700. Sci 2000;77:2513e24.
[35] Arii T, Masuda Y. The effect of humidity on thermal decomposition of tere- [41] IEC60695-2-10, Fire hazard testing. Part2: Test methods-Glow-wire test and
phthalate polyester. J Anal Appl Pyrol 2004;71:525e36. guidance.
[36] Botelho G, Queiros A, Gijsman P. Thermooxidative studies of poly(ether- [42] Acquasanta F, Berti C, Colonna M, Fiorini M, Karanam S. Study of glow wire
esters) 1. Copolymer of poly(butylene terephthalate) and poly(ethylene ox- ignition temperature (GWIT) and comparative tracking index (CTI) perfor-
ide). Polym Degrad Stab 2000;67:13e20. mances of engineering thermoplastics and correlation with material proper-
[37] Botelho G, Queiros A, Gijsman P. Thermo-oxidative studies of poly(ether- ties. Polym Degrad Stab 2011;96:566e73.
esters) 2. Copolymer of poly(butylene terephthalate) and polybutylene ox-
ide. Polym Degrad Stab 2000;68:35e42.