Thermal Conductivity for Polymer Composite Materials: Recent Advances in Polyimide Materials

全文

(1)Journal Review. Journal of Chemical Engineering of Japan, Vol. 54, No. 5, pp. 186–194, 2021. Thermal Conductivity for Polymer Composite Materials: Recent Advances in Polyimide Materials Masashi Haruki Faculty of Mechanical Engineering, Institute of Science and Engineering, Kanazawa University, Kakuma-machi, Kanazawa-shi, Ishikawa 920-1192, Japan Keywords: Thermal Conductivity, Polyimide, Composite, Hexagonal Boron Nitride, Graphene Derivative Polymer materials have received much attention from a wide range of industrial interests. Functional fillers are often introduced to polymers to add new functions that the pristine polymer materials show inferior values. As for thermal conductivity (TC), thermal conductive fillers such as ceramics, functional nanocarbons, and metals are often used. Among the polymers used in industrial fields, polyimide (PI) has excellent thermal stability and solvent resistance. However, the TC of PI is low as is often the case in other commonly used polymers. In the present paper, PI-based composite materials were reviewed from the viewpoint of TC. Hexagonal boron nitride, aluminum nitride, graphene, graphene derivatives (graphene oxide, reduced graphene oxide) and carbon nanotubes were the focus as representative thermal conductive fillers.. Introduction. 1. Polyimide. Thermal interface materials (TIMs) have been used in a wide range of industrial fields to achieve effective heat transfer such as that between the inside and outside of a device, as well as between components in the modules. The importance of the thermal conductivity (TC) of TIM has increased recently because the devices used for the industrial products have increasingly become highly integrated regardless of trends in downsizing. TIMs are used to reduce the interfacial resistance between components, and, therefore, polymers are often used as the base materials because of relatively superior adhesion properties. However, polymers inherently have low levels of TC. In many cases, thermal conductive materials that are functional ceramics, carbons and metals are often added to a polymer matrix to improve the TC of polymer-based materials (Ngo et al., 2016; Zhang et al., 2019). Recent progress in the development of polymer-based TIM materials has been reported by Ma et al. (2021). The polymer species reported were mainly epoxy resin and polydimethylsiloxane, and TC values of TIMs ranged widely from <1 to more than 100 W/(m·K). Many types of polymers are used as the key materials in industrial products. Among them, polyimide (PI) has received much attention as an important material in industrial products because of excellent thermal stability, solvent resistance, and electrical insulation properties. Also, PI has a much lighter weight compared with ceramics and metals. This review highlights research that has focused on enhancing the TC of PI-based materials via the addition of thermal conductive fillers.. A reaction scheme for the production of Kapton-type PI is shown in Figure 1. PI is generally produced via the ring-opening polyaddition reaction of tetracarboxylic dianhydride and diamine. Polyamic acid (PAA) solution is first produced from the polymerization of monomers at relatively low temperature in an aprotic polar solvent such as N,Ndimethylacetamide (DMAc) and N,N-dimethylformamide (DMF). The solvent in the solution is then removed, and the PAA is imidized into PI by either by thermal imidization treatment or by utilizing an imidizing agent. The characteristics of PIs strongly depend on a species of monomers such as solvent resistance. For example, Kapton-type PI that consists of 4,4′-diaminodiphenyl ether (ODA) and pyromellitic dianhydride (PMDA) cannot dissolve in a solvent. On the other hand, PI that is polymerized using 2,2-bis(3,4anhydrodicarboxyphenyl)-hexafluoropropane (6FDA) as a fluorinated aromatic tetracarboxylic dianhydride can be dissolved in solvents such as DMAc and DMF (Matsuura et al., 1991; Haruki et al., 2017). Moreover, other properties such as thermal stability also depend on the monomer species used to make up the PI structure (Sroog, 1991).. 2. Mechanisms of Heat Conduction in Polymer Composites. Received on July 15, 2020; accepted on January 4, 2021 DOI: 10.1252/jcej.20we136 Correspondence concerning this article should be addressed to M. Haruki (E-mail address: [email protected]).. Heat conduction of electrical conductive materials such as metals is based on a transfer of free electrons. On the other hand, that of non-electrical conductive materials such as many polymers is explained by phonon transport. A phonon is defined as a quasiparticle that is represented by quantized lattice vibration energy. As for the heat conduction of polymers, phonon scattering is generated during phonon transfer at the sites of defects and amorphous portions. During phonon transport in crystalline materials such as CNT,. 186. . Copyright This©article 2021 isJournal The published Society of Chemical under of Chemical a CC Engineering BY-NC-ND Engineers, oflicense. Japan.

(2) Fig. 1 Scheme for the synthesis of polyimide (Kapton type polyimide). graphene and crystalline ceramics, phonon transfer ideally proceeds because the atoms are closely and uniformly arranged, which produces a high level of thermal conduction (Zhang et al., 2019; Guo et al., 2020a). The heat conduction in thermal conductive filler/polymer composites is characterized by both smooth phonon transport inside the filler particles and by non-straightforward phonon transport in a bulk polymer. Moreover, phonon scattering also occurs at the filler/polymer and filler/filler interfaces. Therefore, not only the intrinsic TC of fillers but also the geometric parameters of fillers such as size and aspect ratio are important factors that determine the TC of polymer-based composite materials (Zhang et al., 2019). Furthermore, the dispersion behaviors of fillers inside the polymer matrix are also a dominant factor that determines the TC. There are three acceptable theories for the heat transfer of composite materials: thermal conduction path theory, thermal percolation theory and thermoelastic coefficient theory (Guo et al., 2020a). Among them, the thermal conduction path theories seem most used for discussions in research papers. In thermal conductive path theory, heat conduction occurs mainly through thermal conductive paths or through networks constructed by the fillers inside the polymer matrix. When the filler content is low, the fillers are isolated from each other and conductive paths are not constructed. In that case, the TC increases incrementally with increases in the filler content. By continuously adding fillers, conductive paths and networks are gradually formed and TC clearly increases (Guo et al., 2020a). Therefore, one of the key factors to prepare a highly thermal conductive PIbased composite involves the efficient formation of a thermal conductive paths, and much research has been reported thus far. Moreover, the orientation of the fillers is also important for the TC of the composite materials that are composed of fillers with anisotropic TC and high-aspect-ratio shapes. As for the preparation of PI-based composites, the fillers ordinarily should be added not to the to PI matrix but to the PAA solution because many types of PI do not dissolve in a solvent and also are not thermoplastic polymers. The concentration of PAA usually makes up approximately 20 wt% of the solution. Therefore, the volume of the mixture of the filler, PAA and solvent significant decreases during the drying process, and the fillers with anisotropic shapes inside the Vol. 54 No. 5 2021. Table 1 Thermal conductivities of representative thermal conductive fillersa Material hBN AlN Graphene SWCNT MWCNT. TC [W/(m·K)]b in. out. 600 30 320 4,000–5,000 6,000 2,000. Ref. Lin et al. (2013) Yoshioka et al. (2006) Zhang et al. (2019) Ma et al. (2010) Ma et al. (2010). a. Abbreviation of the name of chemicals are listed in Nomenclature. : In-plane; out: Out-of-plane. b in. PI matrix assume to lie down when sheet shape composite materials are prepared. This leads to a low TC in the out-ofplane direction for PI-based composite sheets. Therefore, the special treatments are required to enhance the out-ofplane TC of the sheet shape composites.. 3. Recent Research for Thermal Conductivity of PI-Based Composites 3.1 Overview This section introduces recent advanced studies on enhancement of the TC of thermal conductive filler/PI composite materials. Many types of functional ceramics fillers and carbon fillers have been used as thermal conductive fillers to be combined with PI. In a survey using Scopus®, hexagonal boron nitride (hBN) was most examined ceramics to enhance the TC of PI. Moreover, aluminum nitride (AlN) has also been the focus of improvements in the TC of PI. As for the addition of functional carbons to PI, graphene and graphene derivatives such as graphene oxide (GO) and reduced-graphene oxide (rGO) have been well researched. First, we review the studies of composites using ceramic fillers, and then reviews of research using functional carbon fillers follow. In addition to the explanations of the composite materials, the TCs of the fillers and the composites introduced in the present review are summarized in Tables 1 and 2, respectively. The sizes of fillers and polyimide species are also listed in Table 2 along with the TC. Moreover, how to accomplish surface treatment of the fillers is also listed in the table.. 187.

(3) 188. Journal of Chemical Engineering of Japan. hBN_MTES hBN. hBN_ γ-MPS hBN. hBN. BN. 1,4DB/PMDA. ODA/PMDA. ODA/PMDA. PI. AlN AlN/SiC whisker. AlN/hBN. PI. hBN. 65 µm-AlN (70%)+1 µm-hBN (30%). 4 µm 4 µm-AlN (75%)+SiC (25%) 4 µm-AlN (50%)+SiC (50%) 4 µm-AlN (25%)+SiC (75%). 50 vol% 60 vol% 70 vol%. 50 vol% 50 vol% 50 vol% 50 vol%. 14.2 vol% 14.2 vol% 14.3 vol% 14.1 vol%. 12.4 vol%. BNNS. 0.25 wt%. 30 wt%. 30 vol% 20 vol% 30 vol%. 40 wt% 40 wt%. 50 wt% 50 wt%. 60 vol%. 60 vol%. 12.4 vol%. 11 µm 11 µm 0.5 µm 0.5 µm. 30 µm. 11 µm 5 µm 0.5 µm. 8 µm. <0.5 µm of hBN. 30 wt% 30 wt% 30 wt% 30 wt%. Loading. PI_BNNS. PI. ODA/6FDA. PI. BNNS. 3–5 µm 3–5 µm. hBN. ODA/sBPDA. ODA/PMDA. powder powder. hBN_PI. 6F-OH/ODPA. 1 µm 1 µm (70%)+70 nm (30%) 1 µm (30%)+70 nm (70%) 70 nm. hBN_MPA. ODA/PMDA. Filler size, composition. Filler typeb. PI matrix. DC electrical field. DC electrical field. Press of BNNS-coated PI microspheres. PI-coated hBN powder. Specific treatment. Table 2 Thermal conductivities of representative PI-based compositesa. ,in. 3.2 6.4 9.3. 1.76 2.23 1.99 1.75. 0.59out 0.40out 0.49out 0.32out. Kume et al. (2009). Li and Chung (1994). Haruki et al. (2020). Cao et al. (2020) 4.25in 0.4out 1.38in 0.83*,out. Wang et al. (2019) 2.81in 0.73out. Liu et al. (2019). Haruki et al. (2018) 1.126out 0.625out 0.84out. 0.291. Yang et al. (2016). Chen et al. (2015). Tanimoto et al. (2013). Kuo et al. (2013). Li and Hsu (2010). Ref.. 0.748 0.68*. 1.583 1.323. 17.5* 2.3*,out. 2.3. 0.57* 1.2 0.69* 0.46*. TC [W/(m·K)]c.

(4) Vol. 54 No. 5 2021. 189. MWCNT_BN MWCNT. GO_APTSi. Graphene. GO_KH550. GO_VTES. ODA/BPDA. ODA/BPDA. ODA/PMDA. ODA/BPDA. ODA/PMDA. GO GO/hBN. GF GF/AlN. rGO and BNNS used each layer. ODA/PMDA. ODA/PMDA. ODA/PMDA. rGO (50%)+BNNS (50%). GF (9%)+AlN (91%). GO (5%)+hBN (95%). 200 nm 200 nm 200 nm-AlN (33%)+1 µm-BN (67%) 200 nm-AlN_KH560 (33%)+1 µm-BN_KH560 (67%) 200 nm-AlN (30%)+1 µm-BN (70%) 200 nm-AlN_KH560 (30%)+1 µm-BN_KH560 (70%). powder. 3.8 wt%. 1 wt% 11 wt%. 1 wt% 21 wt%. 4 wt%. 18.8 vol% 7.6 vol%. 0.8 wt% 3 wt% 0.8 wt% 3 wt%. 2.5 wt%. 12 wt%. 1.5 wt%. 3 wt% 3 wt%. 20 wt% 20 wt%. 9 vol% 9 vol% 9 vol% 9 vol% 10 vol% 10 vol%. 30 wt%. 30 wt%. 60 vol%. 60 vol%. Loading. BNNS/PI aerogel+rGO/PI sandwich structure. Using GWFs. BN-coated MWCNT. GMA-grafted GO. Specific treatment. Song et al. (2019). 4.019in 0.44*,out 2.404in 0.40*,out. in. 1.49in. 7.21in 11.19in. 6.118 11.203in. 0.63. 1.06*,out 0.52*,out. 0.89 0.9* 0.17* 0.25*. 1.11*. Guo et al. (2020b). He et al. (2020). He and Wang (2020). Chao et al. (2019). Haruki et al. (2019). Fazil et al. (2018). Wu et al. (2017). Gong et al. (2016) in. 3.73 0.41out. Qian et al. (2015) 0.32*. Yan et al. (2014). Tseng et al. (2013) 0.92out 0.27out 0.388 0.24*. Liu et al. (2020) 0.675 0.70* 0.711 0.77* 0.82* 0.89*. 3.17*,out. Tanimoto and Ando (2014). Ref.. 2.32*,out. TC [W/(m·K)]c. Abbreviation of the name of chemicals are listed in Nomenclature. b: Mixture of fillers; _: Surface-modified filler. c *: Reading value from the figure; in: In-plane; out: Out-of-plane. MWCNT_ODA. ODA/PMDA. a. CNF SWCNT. ODA/PMDA. GO. GO_GMA GO. ODA/PMDA. AlN. AlN AlN_KH560 AlN/BN AlN_KH560/BN_KH560 AlN/BN AlN_KH560/BN_KH560. 10 µm. AlN/BN. ODA/PMDA. TPE-Q/BPDA. 10 µm-AlN (50%)+5 µm-BN (50%). AlN. PPD/sBPDA. powder. AlN. ODA/sBPDA. Filler size, composition. Filler typeb. PI matrix. Table 2 Continued Table 2 Thermal conductivities of representative PI-based compositesa.

(5) 3.2 Ceramics fillers 3.2.1 Hexagonal boron nitride hBN is the most widely used ceramic thermal conductive material for PI-based composite materials. hBN is made up of a layered sp2bonded hexagonally packed sheet similar to graphite (Yu et al., 2018; Guerra et al., 2019), and the TC of hBN represents anisotropy, and 600 and 30 W/(m·K) for in-plane and out-of-plane directions, respectively (Lin et al., 2013). Techniques used for the addition of hBN to a PI matrix have been reported over the past few decades. Sato et al. (2010) achieved a 60 vol%-loading of hBN into a polyimide matrix that maintained its flexibility, and significantly increased the TC of the composite material. Moreover, Haruki et al. (2018) systematically investigated the effect of the size of hBN on the out-of-plane TC of the hBN/PI composite sheets. The TC was increased with an increase in the average size of hBN that ranged from 0.5 to 11 µm. To enhance the affinity of hBN for a PI matrix, the surface of hBNs are often modified to gain organic characteristics. Li and Hsu (2010) used two different average sizes of hBNs (1 µm and 70 nm) and the surfaces of both were. modified using 3-mercaptopropionic acid (MPA). Investigation into the effects of hBN sizes using mixed hBNs (1 µm-hBN/70 nm-hBN=3/7 and 7/3) showed that hBNs of 1 µm-hBN/70 nm-hBN=7/3 was best for the TC because of the appropriate ratio for structuring random conductive bridges. Moreover, Kuo et al. (2013) used an organosoluble PI that consisted of 2,2-bis(3-amino-4-hydroxyphenyl)hexafluoropropane (6F-OH) as a fluorinated diamine and 4,4′-oxydiphthalic dianhydride (ODPA) as a tetracarboxylic dianhydride, the surfaces of hBN grains were previously coated with PI. PI-coated hBNs were milled to obtain powdery fillers, and they were mixed with PI to prepare the composite film. The hBN content achieved 60 vol% and the TC reached 2.3 W/(m·K). Moreover, methyltriethoxysilane (MTES) (Chen et al., 2015) and 3-glycidyloxypropyltrimethoxy silane (γ-MPS) (Yang et al., 2016) were also used as the surface modifier for hBN to enhance the affinity for the interface between hBN and the PI matrix. In either case, the TCs of modified hBN/PI were higher than that of pristinehBN/PI composites with the same content of hBN. Based on the considerable anisotropy of the TC of hBN,. Fig. 2 In-plane (a) thermal conductivity and out-of-plane (b) thermal conductivity of BN/PI composites with different BN loading. Reprinted with permission from Wang et al. (2019). Copyright (2019) Elsevier Ltd.. Fig. 3 Schematic illustration of preparing PI/oriented BNNSs composites. Reprinted with permission from Cao et al. (2020). Copyright (2020) Elsevier Ltd. 190. Journal of Chemical Engineering of Japan.

(6) anisotropic thermal conductive PI-based materials were developed and evaluated. Tanimoto et al. (2013) investigated the effects of hBN size, shape, concentration and orientation on both the thermal diffusivity and TC of PI-based composites for two types of PIs with different levels of flexibility. The use of hBNs with a large size and high-aspect ratio resulted in a significant in-plane orientation in the PI matrix. Consequently, large anisotropies were found in both the thermal diffusivity and TC. Furthermore, the composites consisting of a rigid PI matrix showed a larger anisotropy than that of a flexible PI matrix. Moreover, Wang et al. (2019) prepared the hBN/PI composites with an anisotropic TC via simple methods that consisted of ball milling, high-pressure compression, and low-temperature sintering. The hBNs were well attached in the in-plane direction, and the TCs of the in-plane and out-of-plane directions of the composite materials were 3.2 and 2.4 times higher than those of pure PI, respectively (Figure 2). The structure of hBN is similar to that of graphite, and bulk hBN is formed by stacking sheets of hBN via sp2 bonding. Recently, nano-sheet-shaped hBN exfoliated from bulk hBN has been used as thermal conductive material (Liu et al., 2019). Cao et al. (2020) used a hot-press method to prepare hBN nano-sheet/PI composites that were highly thermal conductive. First, self-assembled hBN nanosheet/PI microspheres were prepared, and they were then pressed to effectively obtain thermal conductive paths made of hBN nano-sheets on the surface of PI microspheres (Figure 3). The in-plane TC reached 4.25 W/(m·K) for a 12.4 vol% hBN nano-sheet. Research into control of the orientation of the hBN in the sheet-shaped PI matrix using electrical field treatment was recently reported by Haruki et al. (2020). The electrical field formed with direct current was loaded with the hBN/PAA solution, and the change in the orientation of the hBN fillers in the PI matrix was confirmed via X-ray diffraction. Namely, the ratio was increased for the hBN with an a-axis that faced the out-of-plane direction of the composite sheet. As a result, the TC in the out-of-plane direction was improved by 50% compared with that of a sheet without electrical field treatment for the 14.2 vol%-hBN/PI composite sheet when the 11 and 0.5 µm hBNs were used. 3.2.2 Aluminum nitride The TC of AlN is as high as 320 W/(m·K), and the electrical insulation and mechanical strength of AlN are also excellent. Moreover, the value of the coefficient of the thermal expansion of AlN approximates that of silicon. Therefore, AlN has been used in the field of electronics for devices such as a heat sink and as a high thermal conductive substrate (Lu et al., 2019; Yoshioka et al., 2006). Moreover, AlN has also received much attention for use as the thermal conductive ceramics fillers to be mixed with polymer materials including PI. Tanimoto and Ando (2014) investigated the relationship between morphological structure and thermophysical properties such as the TC of PI-based composites that consist of micrometer-sized AlN particles. Two types of PIs with either rigid chains (3,3′,4,4′-biphenyltetracarboxylic dianhydride Vol. 54 No. 5 2021. (sBPDA)-p-phenylene diamine (PPD), sBPPD) or flexible chains (sBPDA-ODA, sBPOD) were used. The out-of-plane TC of the composite film of the AlN/sBPPD showed a higher value than that estimated using the theoretical model, unlike the AlN/sBPOD composite film. The reason was speculated that the orientation of the rigid chains of sBPPD was significantly distorted by the AlN particles. Li and Chung (1994) investigated the TC of the composites of particulate AlN/PI and particulate AlN+silicon carbide (SiC) whisker/PI. The TC of AlN/PI showed 1.76 W/(m·K) for an AlN of 50 vol% compared with 0.128 W/(m·K) for that of pristine PI. Moreover, a TC of 2.23 W/(m·K) was achieved when SiC whisker and AlN were co-added at a volume ratio of 25 : 75 to make up 50 vol% of the total filler content. Moreover, coadditions of AlN and hBN have been carried out by Kume et al. (2009), Song et al. (2019), and Liu et al. (2020). 3.3 Carbon fillers 3.3.1 Graphene and graphene derivatives Graphene was discovered by Novoselov et al. (2004). It has a honeycomb-like structure that incorporates a monolayer of sp2hybridized carbon atoms (Kumar et al., 2020). Graphene has received much attention since it’s discovery because of its excellent mechanical, electrical and thermophysical properties (TC: 4,000–5,000 W/(m·K)) (Zhang et al., 2019). Graphene and graphene derivatives such as GO and rGO, are often used as the functional fillers for the preparation of PI-based composites. The dispersions of GO and rGO in a polymer matrix are better than that of pristine graphene, and processing is also easier, although the TCs of the graphene derivatives are much lower than that of graphene (Kumar et al., 2020). The graphene derivatives were often modified via the use of functional chemicals. Tseng et al. (2013) prepared a composite of PI and glycidyl methacrylate-grafted GO (GO_GMA), and the TC of the GO_GMA/PI was much higher than that of GO/PI with a filler content of as much as 20 wt%. On the other hand, the TC of GO_GMA/PI was lower than that of CNT_GMA/PI at a filler content up to 2 wt%. Qian et al. (2015) also prepared functionalized GO nanosheets with 3-aminopropyltriethoxysilane (APTSi) to fabricate high-performance PI-based composite material. Moreover, Wu et al. (2017) modified the surface of GO using a coupling agent (KH-550) to improve the interface interaction between the PI matrix and GO. Furthermore, GO nanosheets modified with vinyltriethoxysilane (VTES) as the functionalized filler was used to enhance the properties of the PI-based composite (Fazil et al., 2018). Graphene derivatives and highly thermal conductive ceramic fillers have often been co-added to a PI matrix. He and Wang (2020) prepared a PI-based composite material that consisted of both GO and hBN as thermal conductive fillers. The in-plane TC was increased to 11.203 W/(m·K) at 1 wt%GO and 20 wt%-hBN compared with 0.223 W/(m·K) of pristine PI and 6.391 W/(m·K) of 20 wt%-hBN/PI. GOs worked as bridges even at 1 wt%, which drastically increased the TC. Furthermore, He et al. (2020) also used the technique of co191.

(7) Fig. 4 The schematic illustration of the fabrication process of GWFs/PI composite films. Reprinted with permission from Gong et al. (2016). Copyright (2016) Elsevier Ltd.. addition of multiple thermal conductive fillers. Graphene flakes (GFs) and AlN were added to PI matrix material and ceramics, and the in-plane TC reached 11.19 W/(m·K) at 1 wt% GFs and 10 wt% AlN. Moreover, Guo et al. (2020b) prepared multilayer structured composite materials that consisted of micro-sandwich structures of reduced-GO/PI composite and hBN nanosheet/PI composite, and the in-plane TC was 11-fold higher than that of pristine PI. As for graphene, Gong et al. (2016) developed PI-based composite sheets using graphene woven fabric (GWF) as a core frame. As for the preparation of GWF, the nickel mesh was coated with graphene via chemical vapor deposition using CH4 and H2, and the nickel wire was then removed via etching treatment to obtain GWF. The 3D graphene structure consisted of 10 layers of GWF was impregnated with PAA solution, and the GWF/PAA was converted to the GWF/PI composite via thermal imidization (Figure 4). The in-plane TC of the GWF/PI sheet was increased to 3.73 W/(m·K) even at approximately 12 wt% of GWF, which corresponded to 1,418% that of pristine PI. 3.3.2 CNTs Many researches for application of CNTs have been reported since Sumio Iijima first reported (Iijima, 1991). There are mainly single- and multi-walled types of CNTs (SWCNT, MWCNT), they have high levels of mechanical strength, electrical properties and TC as well as flexibility and less weight. The reported values for TCs are 6,000 and 2,000 W/(m·K) for SWCNT and MWCNT, respectively. CNTs also have high-aspect ratio, and, therefore, CNTs are often used as thermal conductive fillers for PI-based composite materials (Ma et al., 2010; Ning et al., 2018). Recently, Ning et al. (2018) prepared a PI-based composite material using super-aligned CNTs. The out-of-plane TCs of SWCNT/PI and carbon nanofiber (CNF)/PI composites were also investigated by Haruki et al. (2019). Moreover, Chao et al. (2019) prepared MWCNTs with COOH (MWCNT-COOH), and amino MWCNT (MWCNT_ODA) was produced from MWCNT-COOH and ODA. MWCNT_ 192. ODA was mixed with PAA solution of PMDA-ODA type and the MWCNT_ODA/PAA solution was then cast onto a substrate to obtain MWCNT_ODA/PI composite material. Good chemical interaction and compatibility was obtained between PI matrix and MWCNT_ODA, which led to an enhancement of both the thermal and mechanical properties, and to stability of the composite material. CNTs have excellent electrical conductivity as well as TC. Therefore, the electrical insulation of PI disappears even with the addition of a small amount of CNTs. Yan et al. (2014) synthesized BN-coated MWCNT to enhance the TC of a PI-based composite that retained its electrical insulation properties. Both volume and surface resistances of the PIbased composite material were decreased with an increase in the content of the BN-coated MWCNT. However, the decrease in the rate of the PI-based material using the BNcoated MWCNT was lower than that when using pristine MWCNT. Moreover, the BN-coated MWCNT improved dispersion in the PI matrix, which also led to an improvement in the ability to form a thermal conductive network.. Conclusions Recent studies focused on enhancing the TC of PI-based composites were reviewed. Hexagonal boron nitride, aluminum nitride, graphene and graphene derivatives, and carbon nanotubes were the primary thermal conductive fillers. The construction of thermal conductive paths inside the PI matrix is important in order to obtain a high level of TC, as well as increase in the amount of fillers loaded. The addition of multi species of fillers is also an effective technique to obtain thermal conductive paths. Highly thermal conductive PI-based composites will extend the range of applications for PI materials. Nomenclature AlN. = aluminum nitride. Journal of Chemical Engineering of Japan.

(8) APTSi = 3-aminopropyltriethoxysilane BN = boron nitride BNNS = boron nitride nanosheet BPADA = 2,2-bis(4-(3,4-dicarboxyphenoxy) phenyl) propane dianhydride BPDA = 4,4′-biphthalic anhydride CNF = carbon nanofiber 1,4DB = 1,4-diaminobenzene 6FDA = 2,2-bis(3,4-anhydrodicarboxyphenyl)hexafluoropropane 6F-OH = 2,2-bis(3-amino-4-hydroxyphenyl)hexafluoropropane GF = graphene flakes GMA = glycidyl methacrylate GO = graphene oxide GWF = graphene woven fabric hBN = hexagonal boron nitride KH550 = silane coupling agent used in Wu et al. (2017) KH560 = silane coupling agent used in Liu et al. (2020) MPA = 3-mercaptopropionic acid γ-MPS = 3-glycidyloxypropyltrimethoxy silane MTES = methyltriethoxysilane MWCNT = multi-walled carbon nanotube ODA = 4,4′-diaminodiphenyl ether ODPA = 4,4′-oxydiphthalic dianhydride PAA = polyamic acid PI = polyimide PMDA = pyromellitic dianhydride PPD = p-phenylene diamine rGO = reduced graphene oxide sBPDA = 3,3′,4,4′-biphenyltetracarboxylic dianhydride sBPOD = PI produced from sBPDA and ODA sBPPD = PI produced from sBPDA and PPD SiC = silicon carbide SWCNT = single-walled carbon nanotube TC = thermal conductivity TIM = thermal interface material TPE-Q = 1,4-bis(4-aminophenoxy)benzene VTES = vinyltriethoxysilane. Literature Cited Cao, L., J. Wang, J. Dong, X. Zhao, H.-B. Li and Q. Zhang; “Preparation of Highly Thermally Conductive and Electrically Insulating PI/BNNSs Nanocomposites by Hot-Pressing Self-Assembled PI/ BNNSs Microspheres,” Compos., Part B Eng., 188, 107882 (2020) Chao, M., Y. Li, G. Wu, Z. Zhou and L. Yan; “Functionalized Multiwalled Carbon Nanotube-Reinforced Polyimide Composite Films with Enhanced Mechanical and Thermal Properties,” Int. J. Polym. Sci., 2019, 9302803 (2019) Chen, Y., X. Gao, J. Wang, W. He, V. V. Silberschmidt, S. Wang, Z. Tao and H. Xu; “Properties and Application of Polyimide-Based Composites by Blending Surface Functionalized Boron Nitride Nanoplates,” J. Appl. Polym. Sci., 132, 41889 (2015) Fazil, S., S. Saeed, M. Waseem, W. Rehman, M. Bangesh and K. Liaqat; “Improving Mechanical, Thermal, and Electrical Properties of Polyimide by Incorporating Vinyltriethoxysilane Functionalized Graphene Oxide,” Polym. Compos., 39(S3), E1635–E1642 (2018) Gong, J., Z. Liu, J. Yu, D. Dai, W. Dai, S. Du, C. Li, N. Jiang, Z. Zhan and C. T. Lin; “Graphene Woven Fabric-Reinforced Polyimide Films with Enhanced and Anisotropic Thermal Conductivity,” Compos., Part A Appl. Sci. Manuf., 87, 290–296 (2016) Guerra, V., C. Wan and T. McNally; “Thermal Conductivity of 2D Nano-Structured Boron Nitride (BN) and Its Composites with. Vol. 54 No. 5 2021. Polymers,” Prog. Mater. Sci., 100, 170–186 (2019) Guo, Y., K. Ruan, X. Shi, X. Yang and J. Gu; “Factors Affecting Thermal Conductivities of the Polymers and Polymer Composites: A Review,” Compos. Sci. Technol., 193, 108134 (2020a) Guo, F., X. Shen, J. Zhou, D. Liu, Q. Zheng, J. Yang, B. Jia, A. K. T. Lau and J. K. Kim; “Highly Thermally Conductive Dielectric Nanocomposites with Synergistic Alignments of Graphene and Boron Nitride Nanosheets,” Adv. Funct. Mater., 30, 1910826 (2020b) Haruki, M., A. Wasada, Y. Hosokawa, S. Kihara and S. Takishima; “Deposition of PMDA-TFDB Type Polyimide onto Microscale Trenches Patterned on Silicon Wafer Using Supercritical Carbon Dioxide,” J. Supercrit. Fluids, 127, 200–207 (2017) Haruki, M., J. Tada, K. Tanaka, H. Onishi and Y. Tada; “Enhancing the Effective Thermal Conductivity of Kapton-Type Polyimide Sheets via the Use of Hexagonal Boron Nitride,” Thermochim. Acta, 662, 1–7 (2018) Haruki, M., K. Tanaka, J. Tada, H. Onishi and Y. Tada; “Effective Thermal Conductivity for Nanocarbon/Polyimide and Carbon Nanofiber/ Hexagonal Boron Nitride/Polyimide Composites,” Polym. Compos., 40, 3032–3039 (2019) Haruki, M., J. Tada, R. Funaki, H. Onishi and Y. Tada; “Enhancing Thermal Conductivities of Hexagonal Boron Nitride/Fluorinated Polyimide Composite Materials Using Direct Current Electrical Fields,” Thermochim. Acta, 684, 178491 (2020) He, X. and Y. Wang; “Highly Thermally Conductive Polyimide Composite Films with Excellent Thermal and Electrical Insulating Properties,” Ind. Eng. Chem. Res., 59, 1925–1933 (2020) He, X., X. Yu and Y. Wang; “Significantly Enhanced Thermal Conductivity in Polyimide Composites with the Matching of Graphene Flakes and Aluminum Nitride by in Situ Polymerization,” Polym. Compos., 41, 740–747 (2020) Iijima, S.; “Helical Microtubules of Graphtic Carbon,” Nature, 354, 56–58 (1991) Kumar, A., K. Sharma and A. R. Dixit; “Carbon Nanotube- and GrapheneReinforced Multiphase Polymeric Composites: Review on Their Properties and Applications,” J. Mater. Sci., 55, 2682–2724 (2020) Kume, S., I. Yamada, K. Watari, I. Harada and K. Mitsuishi; “HighThermal-Conductivity AlN Filler for Polymer/Ceramics Composites,” J. Am. Ceram. Soc., 92, S153–S156 (2009) Kuo, D. H., C. Y. Lin, Y. C. Jhou, J. Y. Cheng and G. S. Liou; “Thermal Conductive Performance of Organosoluble Polyimide/BN and Polyimide/(BN+ALN) Composite Films Fabricated by a SolutionCast Method,” Polym. Compos., 34, 252–258 (2013) Li, L. and D. D. L. Chung; “Thermally Conducting Polymer-Matrix Composites Containing both AIN Particles and SiC Whiskers,” J. Electron. Mater., 23, 557–564 (1994) Li, T. L. and S. L. C. Hsu; “Enhanced Thermal Conductivity of Polyimide Films via a Hybrid of Micro- and Nano-Sized Boron Nitride,” J. Phys. Chem. B, 114, 6825–6829 (2010) Lin, Z., Y. Liu, S. Raghavan, K. Moon, S. K. Sitaraman and C. Wong; “Magnetic Alignment of Hexagonal Boron Nitride Platelets in Polymer Matrix: Toward High Performance Anisotropic Polymer Composites for Electronic Encapsulation,” ACS Appl. Mater. Interfaces, 5, 7633–7640 (2013) Liu, X., T. Ji, N. Li, Y. Liu, J. Yin, B. Su, J. Zhao, Y. Li, G. Mo and Z. Wu; “Preparation of Polyimide Composites Reinforced with Oxygen Doped Boron Nitride Nano-Sheet as Multifunctional Materials,” Mater. Des., 180, 107963 (2019) Liu, L., C. Cao, X. Ma, X. Zhang and T. Lv; “Thermal Conductivity of Polyimide/AlN and Polyimide/(AlN+BN) Composite Films Prepared by in-Situ Polymerization,” J. Macromol. Sci. A: Pure Appl. Chem., 57, 398–407 (2020) 193.

(9) Lu, H., M. Qin, H. Wu, Q. He, C. Liu, X. Mu, Y. Wang, B. Jia and X. Qu; “Effect of AlN Powders on the Debinding and Sintering Behavior, and Thermal Conductivity of Injection Molded AlN Ceramics,” Ceram. Int., 45, 23890–23894 (2019) Ma, H., B. Gao, M. Wang, Z. Yuan, J. Shen, J. Zhao and Y. Feng; “Strategies for Enhancing Thermal Conductivity of Polymer-Based Thermal Interface Materials: A Review,” J. Mater. Sci., 56, 1064–1086 (2021) Ma, P. C., N. A. Siddiqui, G. Marom and J. K. Kim; “Dispersion and Functionalization of Carbon Nanotubes for Polymer-Based Nanocomposites: A Review,” Compos., Part A Appl. Sci. Manuf., 41, 1345–1367 (2010) Matsuura, T., Y. Hasuda, S. Nishi and N. Yamada; “Polyimide Derived from 2,2′-bis(trifluoromethyl)-4,4′-diaminobiphenyl. 1. Synthesis and Characterization of Polyimides Prepared with 2,2-bis(3,4dicarboxyphenyl)hexafluoropropane Dianhydride or Pyromellitic Dianhydride,” Macromolecules, 24, 5001–5005 (1991) Ngo, I. L., S. Jeon and C. Byon; “Thermal Conductivity of Transparent and Flexible Polymers Containing Fillers: A Literature Review,” Int. J. Heat Mass Transf., 98, 219–226 (2016) Ning, W., Z. Wang, P. Liu, D. Zhou, S. Yang, J. Wang, Q. Li, S. Fan and K. Jiang; “Multifunctional Super-Aligned Carbon Nanotube/ Polyimide Composite Film Heaters and Actuators,” Carbon, 139, 1136–1143 (2018) Novoselov, K. S., A. K. Geim, S. V. Morozov, D. Jiang, Y. Zhang, S. V. Dubonos, I. V. Grigorieva and A. A. Firsov; “Electric Field Effect in Atomically Thin Carbon Films,” Science, 306, 666–669 (2004) Qian, Y., H. Wu, D. Yuan, X. Li, W. Yu and C. Wang; “In situ Polymerization of Polyimide-Based Nanocomposites via Covalent Incorporation of Functionalized Graphene Nanosheets for Enhancing Mechanical, Thermal, and Electrical Properties,” J. Appl. Polym. Sci., 132, 42724 (2015) Sato, K., H. Horibe, T. Shirai, Y. Hotta, H. Nakano, H. Nagai, K. Mitsuishi and K. Watari; “Thermally Conductive Composite Films of Hexagonal Boron Nitride and Polyimide with Affinity-Enhanced Interfaces,” J. Mater. Chem., 20, 2745–2752 (2010) Song, H., B. G. Kim, Y. S. Kim, Y.-S. Bae, J. Kim and Y. Yoo; “Synergistic Effects of Various Ceramic Fillers on Thermally Conductive Polyimide Composite Films and Their Model Predictions,” Polymers (Basel), 11, 484 (2019) Sroog, C. E.; “Polyimides,” Prog. Polym. Sci., 16, 561–694 (1991). 194. Tanimoto, M., T. Yamagata, K. Miyata and S. Ando; “Anisotropic Thermal Diffusivity of Hexagonal Boron Nitride-Filled Polyimide Films: Effects of Filler Particle Size, Aggregation, Orientation, and Polymer Chain Rigidity,” ACS Appl. Mater. Interfaces, 5, 4374–4382 (2013) Tanimoto, M. and S. Ando; “Effects of Orientational Relaxation of Polymer Chains Induced by Isotropic Particles on the Enhanced Thermal Conductivity of AlN-Filled Polyimide Films,” J. Photopolym. Sci. Technol., 27, 193–198 (2014) Tseng, I. H., J. C. Chang, S. L. Huang and M. H. Tsai; “Enhanced Thermal Conductivity and Dimensional Stability of Flexible Polyimide Nanocomposite Film by Addition of Functionalized Graphene Oxide,” Polym. Int., 62, 827–835 (2013) Wang, H., D. Ding, Q. Liu, Y. Chen and Q. Zhang; “Highly Anisotropic Thermally Conductive Polyimide Composites via the Alignment of Boron Nitride Platelets,” Compos., Part B Eng., 158, 311–318 (2019) Wu, G., Y. Cheng, Z. Wang, K. Wang and A. Feng; “In situ Polymerization of Modified Graphene/Polyimide Composite with Improved Mechanical and Thermal Properties,” J. Mater. Sci. Mater. Electron., 28, 576–581 (2017) Yan, W., Y. Zhang, H. Sun, S. Liu, Z. Chi, X. Chen and J. Xu; “Polyimide Nanocomposites with Boron Nitride-Coated Multi-Walled Carbon Nanotubes for Enhanced Thermal Conductivity and Electrical Insulation,” J. Mater. Chem. A Mater. Energy Sustain., 2, 20958– 20965 (2014) Yang, N., C. Xu, J. Hou, Y. Yao, Q. Zhang, M. E. Grami, L. He, N. Wang and X. Qu; “Preparation and Properties of Thermally Conductive Polyimide/Boron Nitride Composites,” RSC Advances, 6, 18279– 18287 (2016) Yoshioka, T., Y. Makino, S. Miyake and H. Mori; “Properties and Microstructure of Aluminum Nitride Sintered by Millimeter-Wave Heating,” J. Alloys Compd., 408–412, 563–567 (2006) Yu, C., J. Zhang, W. Tian, X. Fan and Y. Yao; “Polymer Composites Based on Hexagonal Boron Nitride and Their Application in Thermally Conductive Composites,” RSC Advances, 8, 21948–21967 (2018) Zhang, Y., Y. J. Heo, Y. R. Son, I. In, K. H. An, B. J. Kim and S. J. Park; “Recent Advanced Thermal Interfacial Materials: A Review of Conducting Mechanisms and Parameters of Carbon Materials,” Carbon, 142, 445–460 (2019). Journal of Chemical Engineering of Japan.

(10)