Introduction

Epoxies, which exhibit superior engineering properties such as high tensile strength and flexible strength, high stiffness and modulus, great adhesion, low creep, low shrinkage on cure, excellent chemical and corrosion resistance, and good thermal and electrical properties, are important thermosetting polymers [1–3]. Unfortunately, the high crosslinking density of the cured network is closely associated with poor impact resistance [4, 5]. Because of high strength and stiffness per unit weight, glass fiber and carbon fiber are incorporated into epoxy matrixes to improve the mechanical properties and also the toughness [3, 6–17]. Fiber-reinforced epoxy composites (FRECs) have been widely used in various applications such as aerospace, aircraft, defense, sporting goods, wind turbine, marine, automobile, boat, and mechanical and electronic applications [7, 18, 19] due to their high specific stiffness and strength, good dimensional stability, great modulus, low density, low creep, low dielectric constant, excellent flexural strength and tensile strength, ease of processability, and relatively low costs [6, 20, 21]. The properties of FRECs are significantly dependent on the properties of composite constituents such as epoxy matrix, fiber, and the interface between them [7, 22, 23]. However, the notable disadvantage of fiber is its weak combination with epoxy matrix [8, 24–27]. Due to this nature, an incompatibility between the fiber and the matrix exists which decreases the properties of the composite. This defect can be overcome by chemical modification of fiber surface so as to make it a strong adhesive [28–33].

Hyperbranched polymers (HBPs) are a kind of interesting materials due to their low melting points and viscosities, high-density and versatile terminal groups, and unique spherical structure. The use of suitable terminal groups, such as hydroxyl, epoxide, amine, and carboxylic groups, can enhance the compatibility between HBPs and polymer matrixes [34]. Thus, HBPs are used as a new class of modifiers for epoxy resins [35–37]. The advantages that the introduction of HBP into epoxy matrix are their high-density and versatile terminal functional groups, a lot of free spaces and free volumes in their structures [38, 39], and their peculiar structures [39–41].

The novelty of the present work is the use of a novel modified glass fiber (GF) grafted by amino-terminated hyperbranched polymer (ATHBP) in two-step polycondensation, as a filler for an epoxy resin. In this study, firstly GF was treated by a silane coupling agent, then hyperbranched polymer with hydroxyl groups (HBPH) was grown on the surface of GF, and lastly GF-HBPH was terminated by polyamines. The mechanical properties of epoxy/GF composites were studied, discussing the results from tensile, flexural, and impact tests. Field emission scanning electron microscope (FESEM) was used to examine the fracture surfaces of composites to discuss the toughening mechanism.

Experiment sections

Materials

Diglycidyl ether of bisphenol-A epoxy resin E51, which has an epoxide equivalent of 185–208 g/eq, was obtained from Hangzhou Wujingang Adhesive Co., Ltd, Hangzhou, China. A general hardener (polyamide 650), which has an amine value of 220 ± 20 mgKOH/g, was supplied by Wuxi Resin Factory, Wuxi, China. GFs, with an average length of 30 μm and diameter of 10 μm, were purchased from Corker CO., LTD, Taikang, China. γ-Aminopropyl triethoxysilane (γ-APS, KH-550) was purchased from Sinopharm Chemical Reagent Co., Ltd, Shanghai, China. Succinic anhydride, diethanolamine, aqueous solution of H2O2 (30 wt%), diethylenetriamine (DETA), and N, N-dimethylacetamide (DMA) were purchased from Chengdu Kelong Chemical Reagent Company, Chengdu, China.

Fabrication of glass fiber grafted by amino-terminated hyperbranched polymer (GF-ATHBP)

The process was similar to our former reports [35, 37]. Typically, GFs were first treated with an aqueous solution of H2O2 and a kind of silane coupling agent, γ-APS. After dried under vacuum at 80 °C for 24 h, 1 g GFs were added into the mixture of 1.57 g succinic anhydride and 1.5 g diethanolamine at 70 °C for 2 h and then heated to 120 °C for 6 h. Then the temperature was decreased to 110 °C, and 3.1 g DETA was added dropwise into the mixture under stirring for 10 h. After washed and filtered three times with DMA and dried under vacuum at 80 °C for 12 h, GF-ATHBP was obtained. The chemical structure of GF-ATHBP is shown in Scheme 1.

Scheme 1
scheme 1

The chemical structure of GF-ATHBPs

Full size image

Preparation of epoxy/GF composites

Epoxy and GF-ATHBP (GF-ATHBP/epoxy, 0–5 wt/wt%) were blended and sonicated for 10 min. Then polyamide 650 (epoxy/polyamide 650, 1:1.2 wt/wt), which was pre-heated to 60 °C, was added and the resulting mixture was stirred for another 10 min. Finally, the mixture was poured into stainless steel templates and cured at 60 °C for 48 h for testing in an oven. The whole preparation process is shown in Scheme 2.

Scheme 2
scheme 2

The routes of synthesis of GF-ATHBP and fabrication of composite

Full size image

Characterization

Attenuated total internal reflectance infrared (ATR-IR) spectroscopic analysis was conducted with a Tensor 37 instrument (Bruker, Germany) within the range of 400–4000 cm−1 and a resolution of 4 cm−1. 1H nuclear magnetic resonance (1H NMR) spectrum was recorded on an AVANCE 500 NMR spectrometer (Bruker, Switzerland) with deuterated dimethyl sulfoxide (DMSO-d6) as the solvent at 293 K. Thermal gravimetric analysis (TGA) was performed with a SHIMADZU DTG 60 TG/DTA simultaneous measuring instrument (SHIMADZU, Japan) at a heating rate of 20 °C min−1 from 30 to 800 °C under a nitrogen atmosphere. Tensile tests were performed with an SHIMADZU AG-X plus (SHIMADZU, Japan) test machine. The composite samples were prepared according to GB/T 2567-2008. Flexural tests were characterized with a UTM4000 electromechanical universal testing machine (SANS, China) according to GB/T 2567-2008 standard of dimensions 80 * 15 * 4 mm. Impact resistance measurements were performed with an ZBC 50 pendulum impact testing machine (New SANS, China) without a notch in the specimen according to GB/T 2567-2008 standard of a thickness of 4 mm and a width of 10 mm. Field emission scanning electron microscopy (FESEM) images were recorded using a Hitachi SU8000 field emission scanning electron microscope (Hitachi, Japan), and the fracture surfaces of the composites were sputter-coated with gold before observation.

Results and discussions

Characteristics of the GF-ATHBP

The ATR-IR spectrum of GF-ATHBP is shown in Fig. 1. The broad peak at 3295.9 cm−1 is ascribed to the symmetric stretching of –NH. The two weak peaks at 2943.5 and 2842.1 cm−1 are attributed to the C–H stretching. The strong peak at 1647.8 cm−1 is assigned to the C=O stretching. The weak peak at 1554.8 cm−1 is attributed to the N–H bending vibration. The peak at 1434.8 cm−1 belongs to the C–H bending vibration. The weak peak at 1267.5 cm−1 is ascribed to the C–O–C stretching vibration. The strong peak at 1063.8 cm−1 is assigned to the C–N stretching. These peaks indicate that ATHBP may be successfully grafted onto the surface of silanized GFs.

Fig. 1
figure 1

ATR-IR spectrum of GF-ATHBP

Full size image

NMR spectroscopy can provide detailed information about the structure, dynamics, reaction state, and chemical environment of molecules. The 1H NMR spectrum of GF-ATHBP is shown in Fig. 2. As can be seen in Fig. 2, the 1H NMR spectrum clearly displays the characteristic peaks of ATHBP, the peaks appearing at 1.779 ppm (N–CH 2–CH2–O, protons a), 2.533 ppm (N–CH2–CH 2–O, protons b), 2.188 ppm (C–CH 2–CH2–, protons c), 2.665 ppm (CH2–CH 2–C, protons d), 3.222 ppm (N–CH 2–CH2–N, protons e), 3.414 ppm (N–CH2–CH 2–N, protons f), 1.554 ppm (N–CH 2–CH2–NH2, protons g), 3.503 ppm (N–CH2–CH 2–NH2, protons h), and 7.933 ppm (N–CH2–CH2–NH 2, protons i). And the spectrum shows three kinds of proton peaks, which can be attributed to linear, dendritic, and terminal units. For instance, the peaks at 1.779, 2.533, 2.188, and 2.56 ppm are attributed to the linear units, the peaks at 3.222 and 3.414 ppm are assigned to the dendritic units, and the peaks at 1.554, 3.503, and 7.933 ppm are ascribed to the terminal units.

Fig. 2
figure 2

1H NMR spectrum of GF-ATHBP

Full size image

TGA is commonly used to determine selected characteristics of materials that exhibit either mass loss or gain due to decomposition, oxidation, or loss of volatiles (such as moisture). Figure 3 shows the TGA curves of hydroxylated GF, silanized GF, ATHBP, and GF-ATHBP measured under nitrogen atmosphere. As the weight loss diagrams illustrate, the final weight losses of hydroxylated GF, silanized GF, ATHBP, and GF-ATHBP at 800 °C are 18.6, 33.4, 77.3, and 50.5 %, respectively. Thus, the grafting ratio of silane on GF is 14.8 %. Furthermore, the temperature for 5 and 10 wt% weight losses (T 5 % and T 10 %) of silanized GFs is 103.8 and 223.2 °C, respectively. In contrast, the GF-ATHBP is stable until 200 °C and no meaningful weight loss is detected. The T 5 % and T 10 % values of GF-ATHBP are 229.3 and 320.7 °C, respectively. The higher decomposition temperature of GF-ATHBP may be due to that the grafted ATHBP reduces the moisture content owing to hydrophobicity on the surface of GFs [42]. However, the presence of ATHBP makes GF-ATHBP to be less stable than silanized GF under high temperature. As the weight loss diagram of GF-ATHBP illustrates, the GF-ATHBP has a higher weight loss at 800 °C than the corresponding sample of silanized GF. It should be noticed that the weight loss of ATHBP increases slowly after 600 °C, which indicates that ATHBP decomposes completely at that temperature. The weight losses of GF-ATHBP and silanized GF at 600 °C are 37.1 and 29.5 %, respectively. So the grafting ratio of ATHBP on GF is 7.6 %. The TGA curves confirm the ATHBP molecules being successfully attached to the surface of the silanized GF.

Fig. 3
figure 3

TGA curves of hydroxylated GF, silanized GF, ATHBP, and GF-ATHBP

Full size image

Figure 4 shows the surface characterizations of the silanized GFs and GF-ATHBPs. The surfaces of silanized GFs are rough and the average diameter is about 11.3 μm (Fig. 4a), calculated from the inset image with high magnification. The surfaces of GF-ATHBPs, which are shown in Fig. 4b, are rougher than those of silanized GFs. And the average diameter is about 13.2 μm, which is calculated according to the insetting image of a single GF-ATHBP in Fig. 4b. The diameter change, about 1.9 μm, may be another evidence to indicate the successful growth of ATHBP onto the surfaces of GF.

Fig. 4
figure 4

FESEM images of silane-functionalized GFs (a) and GF-ATHBPs (b)

Full size image

Mechanical properties of composites

The tensile strength of epoxy/GF composites as a function of GF content is presented in Fig. 5. Compared to the silanized GFs, the tensile strength of the composites in the presence of GF-ATHBPs steadily enhances with increase of GF-ATHBPs content. As can be seen from Fig. 5, the highest tensile strength is 49.43 MPa, obtained for 2 wt% GF-ATHBPs, which is 31.1 % higher than that of the neat epoxy. With further addition of GF-ATHBPs, the tensile strength reduces dramatically and even becomes the lowest when it reached 5 wt% loading. The tensile modulus of the neat epoxy and epoxy composites is shown in Fig. 6. The change in the trend of the tensile modulus graph is similar to the tensile strength graph. The tensile modulus increases steadily from 1571 MPa for neat epoxy to 2427 MPa for composite with 2 wt% GF-ATHBPs, corresponding to the increase of 54.4 %; after that, the trend alters and decreases to 1496 MPa for composite with 5 wt% GF-ATHBPs. Similar to the effect of GF-ATHBPs, the tensile modulus increases with the increase of silanized GF content, reducing the increasing trend. But the increase in epoxy composite modified with silanized GF is lower than that in the composite with same content of GF-ATHBP. Elongation at break is another index to evaluate the tensile property of epoxy-based composites and is shown in Fig. 7. It can be seen from Fig. 6, the elongation at break increases sharply from 2.43 % for the neat epoxy to 5.24 % for the composite with 2 wt% GF-ATHBPs, increasing by about 115.6 %. And then the trend alters and the elongation value decreases to 2 % for 5 wt% GF-ATHBPs. In contrast, the elongation at break is independent on the addition of silanized GFs except the formulation with 1 wt% GFs. The growth of ATHBP has affected the affinity between GFs and epoxy matrixes effectively, leading to a better dispersion of GFs and also improving the compatibility and interaction between GFs and epoxy matrixes.

Fig. 5
figure 5

Tensile strength of composites as a function of GF content

Full size image
Fig. 6
figure 6

Tensile modulus of composites as a function of GF content

Full size image
Fig. 7
figure 7

Elongation of composites as a function of GF content

Full size image

The flexural strength of composites with different contents of GFs is shown in Fig. 8. Similar to the results of tensile test, the flexural strength increases with the increase of GF-ATHBP content, reducing the increasing trend. The highest flexural strength is 92.34 MPa, obtained for 2 wt% GF-ATHBPs, which is 30.5 % higher than that of the neat epoxy. It also can be observed that the effect of silanized GFs on the flexural strength is less obvious than that of GF-ATHBPs and it is difficult to summarize the patterns. Figure 9 shows the impact strength of composites as a function of GF content. As can be seen from Fig. 9, the impact strength suddenly increases from 7.667 kJ/m2 for the neat epoxy to 13.375 kJ/m2 for the composite with 2 wt% GF-ATHBPs, which increases by about 74.4 %. Then it starts to decrease after the loading exceeds 2 wt%. Generally, this phenomenon occurs with an inhomogeneous distribution, while the quantity of GF-ATHBPs increases. The results of impact tests can be explained as follows: firstly, grafted GFs perform a better improving effect at lower loading due to their enhanced compatibility and interaction with epoxy matrixes. The former may optimize the distribution, whereas the latter may guarantee an effective stress transfer from the matrixes to the GF-ATHBPs [43]. Secondly, the incorporation of GF-ATHBPs may decrease the crosslinking density of epoxy composite, which is due to that the growth of ATHBP breaks the stoichiometric balance between epoxide and amine values and then leads to the excess of curing agents. Lastly, an inhomogeneous distribution and self aggregation cause negative effects on the mechanical properties of composites. The impact strength of composites decreases with increasing silanized GF content. On the whole, ATHBP is helpful for our epoxy-based composites by improving GF dispersion, compatibility, and interfacial interaction and thus gives remarkable improvements on mechanical properties.

Fig. 8
figure 8

Flexural strength of composites as a function of GF content

Full size image
Fig. 9
figure 9

Impact strength of composite as a function of GF content

Full size image

Morphologies of GF-ATHBPs and the impact fracture surfaces of composites

The change in mechanical properties can possibly be explained by observing the differences in the impact fracture surfaces of the specimens. The impact fracture surface morphologies of specimens are shown in Fig. 10. The neat epoxy shows a typical smooth surface with a little crack propagation (Fig. 10a), which leads to poor resistance to mechanical damage, especially for impact resistance. The fracture surface image of composite with silanized GFs shows several large shearing deformations (Fig. 10b), which may be attributed to the improvement in the mechanical properties.However, the smooth surface of GF, which indicate the poor compatibility between GFs and epoxy matrixes, may be the weaknesses of composites and thus decreases the mechanical properties. The impact fracture surface images of composites with GF-ATHBPs (Fig. 10c, d) show more crack propagations, crack deflections, and shearing deformations, which support the improved mechanical properties. Moreover, the surface of exposed GF-ATHBP is rough and the exposed part is short, which indicate the excellent compatibility and strong interaction between GF-ATHBP and epoxy matrix. Furthermore, the embedded part of GF-ATHBP can transfer stress from GF-ATHBP to the epoxy matrix and thus improve the mechanical properties.

Fig. 10
figure 10

FESEM morphologies of the impact fracture surfaces of neat epoxy (a), composites with GFs (b), and 2 wt% GF-ATHBPs (c, d)

Full size image

Conclusions

GF, grafted by an ATHBP, named as GF-ATHBP, was incorporated into epoxy resin to fabricate epoxy-based composites. The structure of GF-ATHBP was characterized by ATR-IR, 1H NMR, TGA, and FESEM. The effects of GF-ATHBP content on the mechanical properties of composites were studied, discussing the results from tensile, flexural, and impact tests. For comparison, composites with silanized fibers were also fabricated and investigated. The impact fracture surfaces of composites were observed by FESEM to explain the change in the mechanical properties.

The results of ATR-IR and 1H NMR indicate the successful growth of ATHBP onto the surfaces of GFs. TGA curves illustrate that the grafting ratio is about 7.6 %. FESEM images show that the grafting layer on the surface of GF is about 1.9 μm.

Compared to the neat epoxy and composites with silanized GFs, the introductions of GF-ATHBPs can favorably improve the mechanical properties of epoxy-based composites. For instance, the tensile strength, elongation at break, flexural strength, and impact strength for the composite with 2 wt% GF-ATHBPs increase by about 31.1, 115.6, 30.5, and 74.4 %, compared to the neat epoxy, respectively.

FESEM morphologies show that the excellent compatibility and interaction between GF-ATHBP and epoxy matrix may be ascribed to the enhancement of mechanical properties of epoxy/GF-ATHBP composites, especially for impact strength.