Promoting Thermal Conductivity of Alumina-Based Composite Materials by Systematically Incorporating Modified Graphene Oxide

Abstract

Small amounts of thermally conductive graphene oxide (GO) and modified GO are systematically introduced as a second filler to thermal interface materials (TIMs) consisting of alumina (Al₂O₃) particles and polydimethylsiloxane (PDMS). The surface of GO is covalently linked with an organic moiety, octadecylamine (ODA), to significantly improve the miscibility and dispersity of GO across the TIM matrix. Subsequently, two series of PDMS-Al₂O₃ composite TIMs are manufactured as a function of GO and ODA-GO content (0.25 wt%–2.5 wt%) to understand the effect of these second additives. The incorporation of GO into the Al₂O₃-PDMS composite materials generally increases the thermal conductivity (TC), ranging from 18% to 29%. Conversely, the use of ODA-GO further enhances the overall performance of TIMs (22–54%) by facilitating the dispersion degree of GO across the composite matrix. The great improvement in TC is presumably related to the formation of conductive pathways by uniformly integrating 2D-type GO flakes across spherical Al₂O₃ particle networks. The ability to simply regulate the polarity of the thermally conductive second filler can provide an idea for designing cost-effective and practical TIM-2-type pads that can be commercially applicable in between an integrated heat spreader and a heat sink.

1. Introduction

The advanced manufacturing process has led to the miniaturization and integration of numerous components to develop new electronic devices. Given the condensed assembly of conducting and semiconducting electronic products, increasing challenges are associated with the effective dissipation of heat flux inevitably generated during the operation [1,2,3]. Thus, thermal management has become an essential issue in guaranteeing the reliability and service life of electronic devices such as integrated computers, automobiles, and energy harvesting/storage systems. In this sense, various materials have been fabricated to design a heat sink that can effectively dissipate and eliminate heat from the electronic components, known as thermal interface materials (TIMs) [3,4,5]. Although the simplest approach for developing TIMs involves the utilization of polymer binders and thermally conductive fillers [4,5,6,7], the main concern is the proper selection of polymer precursors and additives that can meet several criteria, including environmental friendliness, cost-effectiveness, and easy preparation.

Among many types of polymer precursors for thermal conductivity (TC) applications, polydimethylsiloxane (PDMS) is of particular interest because of several advantages, including low cost, easy handling/processing at room temperature, and non-toxic characteristics [5,8,9,10]. In addition, the polymerized final products exhibit high flexibility, chemical inertness, physical/thermal stability, and environmental friendliness. However, most polymer binders, including PDMS, typically exhibit very low TC values (<0.2 W/m·K) that cannot fulfill the need for heat management in modern electronics. Due to this limitation, various inorganic powders (e.g., Al₂O₃, ZnO, MgO, SiC, BN, etc.) and carbon-based materials (graphene, graphite, carbon nanotube, etc.) as fillers are often introduced to polymer-based binders to greatly improve the heat dissipation capability of final composite materials [2,6,8,10,11]. Unlike metal-derived TIMs, metal oxide-based thermally conductive materials prepared with inorganic fillers can possess the following advantages: electrical insulation (i.e., only heat dissipation), corrosion resistance (i.e., long-term stability), and high flexibility (i.e., no air pockets between the electronic device and TIM). Although relatively higher-cost inorganic fillers (e.g., MgO, SiC, and BN) offer better thermal transport capability [2,12,13,14], developing a strategy to enhance the TC performance of readily available and low-cost fillers (e.g., Al₂O₃) is highly desirable. For example, the hybridization of several fillers to regulate the overall TC of composite materials is gaining momentum [7,11,15,16]. Specifically, developing a method to increase the contact areas of the thermally conductive fillers across the TIM networks by doping an efficient second filler synergistically allows for maximizing the TC of the original fillers. One simple approach to improving the interfacial interactions (i.e., conduction pathways) across the main fillers can be accomplished by introducing a small quantity of additional high-performance fillers in the main TIM matrix. If the high-thermal-transport materials could serve as a bridge to effectively connect the gaps of the main fillers, the resulting TIMs would display minimal loss of TC. As such, constructing thermal paths inside the composite TIMs by doping a small percentage of the second filler is a great way to improve the overall TC performance [4,5,10,12,15]. In addition, the doping material is designed to increase the attractive force between the main filler and polymer binder. The resulting TIMs can easily boost the overall efficiency by promoting the interfacial transportation of phonons (i.e., low interfacial thermal resistance). Thus, the synergistic improvement can be achieved simply by utilizing a small amount of the second filler across the inexpensive main fillers, which could be an alternative approach to developing effective and practical TIMs instead of discovering completely new TIMs.

In this study, readily available and inexpensive aluminum oxide (Al₂O₃) particles and carbon-based graphene oxide (GO) were chosen as the main and supplementary doping fillers, respectively. Particularly, the highly conductive GO can be easily fabricated with an organic moiety to exhibit controlled polarity, which can serve as an attractive second filler. In addition, a non-toxic PDMS precursor was selected as the polymer binder to prepare TIMs. As the PDMS binder is relatively hydrophobic, the modification of hydrophilic GO with a long-chain hydrocarbon (e.g., ODA-octadecyl amine) would greatly increase the miscibility and dispersity of the GO filler in the polymer precursor. It is noted that the amount of inorganic-based fillers (i.e., Al₂O₃ particles) needed to prepare practical TIMs is generally much higher than that of polymer resin (>2–8 times) [8,16,17]. In addition, effectively mixing a high concentration of ceramic Al₂O₃ particles in a polymer precursor solution is somewhat challenging to establish continuous heat conduction pathways (i.e., even distribution). As carbon-based materials possess highly effective TC, the presence of compatible GO in the mixture of Al₂O₃ particles and PDMS binder can serve as an effective bridge to synergistically enhance the TC. Upon the introduction of the modified GO, further increasing the miscibility/dispersity (i.e., improving interfacial interactions) can result in the formation of compact and continuous pathways to maximize thermal conduction. As such, two series of GO and modified GO were systematically introduced as the second filler to a mixture of Al₂O₃ particles and PDMS precursor to manufacture pad-type TIMs. The TC performance of the resulting TIMs was thoroughly examined as a function of GO content to understand the importance of the second additive. Developing a simple strategy to utilize small quantities of thermally conductive materials and demonstrating the possibility of promoting the overall thermal transportation capability will allow for the design of low-cost but highly effective TIMs for practical applications.

2. Materials and Methods

2.1. Materials

Graphene oxide (GO) used in this experiment was purchased from Standard Graphene Inc. (GO-V50, Ulsan, Republic of Korea). Octadecyl amine (ODA, Sigma Aldrich, St. Louis, MO, USA), polydimethylsiloxane (PDMS) binder consisting of LSI-GL300 A and LSI-GL300 B (HRS Co. Ltd., Pyeongtaek-si, Gyeonggi-do, Republic of Korea), emulsifier (DN-XEF-20010, J-I Materials, Wanju-gun, Jeollabuk-do, Republic of Korea), PT-5000 catalyst (BRB International, Ittervoort, The Netherlands), and Al₂O₃ powder mixed with various sizes of particles (Denka, Japan) were purchased from the indicated companies.

2.2. Preparation of GO-ODA

The functionalization of GO with ODA was carried out by a slightly modified method based on the reactions between carbon-based materials and organic moieties [18,19,20,21]. Initially, GO (1.2 g) was dispersed in DI water (600 mL) using an ultrasonic bath for 5 min. ODA (1.8 g) dissolved in ethanol (180 mL) was introduced to the GO solution. The mixture was vigorously stirred and refluxed at 95 °C for 4 h. The resulting product was centrifuged (Vision Science Co. LTD., Daejeon, Republic of Korea) at 4500 rpm for 30 min (at least twice) and filtered using a vacuum filtration apparatus. The collected black powder was then dried in an oven at 50 °C before use.

2.3. Systematic Preparation of Thermal Interface Materials (TIMs) as a Function of GO and ODA-GO Content

Two series of TIMs were prepared using GO and ODA-GO as a function of amount (0.25 wt%–2.50 wt% regarding the amount of Al₂O₃ particles). For example, the pad-type TIMs containing 0.25 wt% GO were prepared using the following recipe. PDMS B (0.81 g) and emulsifier (0.05 g) were added to a glass vial containing GO, which was sonicated and stirred for at least 10 min. After the full dispersion of GO, Al₂O₃ particles (4.05 g), PDMS A (0.09 g), and the catalyst (0.02 g) were added to the mixture. After vigorous stirring for at least 20 min, the final mixture was cast on a glass slide (Matsunami, Japan) covered with a fluorine release film using a 2 mm tape. The size of TIMs was adjusted to at least 2.5 cm × 2.5 cm × 2 mm (thickness), and the mixture was firmly pressed using a glass rod to obtain flat surfaces. The TIM pads were then covered with another fluorine release film and placed in an oven at 100 °C for 30 min, which were then subjected to TC measurements. Additional TIM pads containing GO or ODA-GO were prepared using the same process simply by increasing their amount. The entire preparation process without altering other experimental conditions (e.g., mixing speed, curing time/temperature, etc.) could allow us to monitor the thermal conductivity changes as a function of GO or ODA-GO content. The modification of GO with ODA and the overall manufacturing process of TIMs are presented in Figure 1.

2.4. Characterization

The morphological and structural features of ingredients and final TIMs were examined by an optical imager equipped with a polarized light source (Zeiss Axio Imager 2 Pol, Carl Zeiss AG, Baden-Württemberg, Germany) and a field emission scanning electron microscope installed with a Gemini column (FESEM, Zeiss Sigma 300 VP, Carl Zeiss AG, Baden-Württemberg, Germany). The wettability of the GO, ODA-GO, and TIMs was examined by a water contact angle analyzer (Contact Angle Goniometer, Ossila Limited., Sheffield, U.K.). Raman scattering measurements were carried out using a ProRaman-L Analyzer (Enwave Optronics, Irvine, CA, USA) equipped with a 785 nm laser (~10 mW of laser power, Thor Labs, Newton, NJ, USA). A Fourier-transform infrared spectrophotometer equipped with an attenuated total reflection (ATR) sampling device (Spectrum 100 FT-IR Spectrometer, PerkinElmer, Hopkinton, MA, USA) was used to examine the several functional groups of representative GO, ODA-GO, and TIMs in the scan range of 600–4000 cm⁻¹. The thermal degradation patterns of three fillers (i.e., GO, ODA-GO, and Al₂O₃ powder) were also examined with a thermogravimetric analyzer (TGA, SDT Q600, TA Instrument, New Castle, DE, USA). A small quantity (3–5 mg) of samples was loaded in an alumina pan for the measurements under the following conditions: pre-heating at 80 °C for 10 min, ramping temperature of 20 °C/min, and purging N₂ gas at 50 mL/min. A PXRD system equipped with a Cu Kα X-ray (MiniFlex 600, Rigaku Corp., Tokyo, Japan) was employed to examine the structural information of GO, ODA-GO, Al₂O₃ powder, and TIMs (scan range: 3–80° and scan rate: 5°/min). A vertical type of thermal conductivity meter with a resolution of 0.001 °C detecting temperature (TCM-100V, HAN Tech, Seoul, Republic of Korea) was used to evaluate the thermal conductivity of TIMs following ASTM D5470. All TIM samples (a minimum of 2.2 mm × 2.2 mm) were prepared on a glass slide covered with a fluorine release film.

3. Results

3.1. Morphological Features and Wettability of TIM Components

Prior to examining the pad-type TIMs, the two main fillers, Al₂O₃ particles and graphene oxide (GO), before and after modification with octadecyl amine (ODA), were thoroughly characterized as these components primarily influence the overall thermal conductivity (TC). The Al₂O₃ particles appeared to be highly polydisperse to serve as an ideal filler for TIMs because relatively small-sized particles can easily fill in among large particles to increase the probability of thermal conduction via lattice vibrations (phonons) for TC (Figure 2). It has been reported that the different sizes of ceramic particles are purposely varied to achieve higher packing density at the interface of larger fillers to maintain continuous thermal pathways [22]. The GO powder appeared to be similar to ODA-GO, and SEM images also did not show notable morphological changes except for slightly curvy/twisted features before the modification. However, the degree of dispersity for GO and ODA-GO in water was distinctively different, as shown by digital photos. In addition, the SEM images of GO and ODA-GO displayed different morphological features after suspending in water and an organic solvent (Figure S1). Specifically, GO showed rough surfaces with sharp edges in water but an aggregated form with smooth surfaces in isopropanol, respectively. However, ODA-GO displayed completely opposite behaviors, implying that the miscibility and dispersity of ODA-GO could be better in slightly hydrophobic PDMS polymer resin than GO. Furthermore, the contact angle of a water droplet on the ODA-GO (~78°) was far greater than that of the GO (~24°) upon preparing film-type sheets using the MSE PRO Pellet Pressing Die set (MSE Supplies). This wettability change evidently suggests the successful surface modification of GO from hydrophilic to relatively hydrophobic properties.

3.2. Compositional and Thermal Properties of TIM Components

The presence of ODA on the surface of GO was confirmed by FT-IR and Raman spectra (Figure 3). The GO layer exhibited several strong peaks at ~3200 cm⁻¹ (–OH), ~1640 cm⁻¹ (aromatic C=C), and ~1050 cm⁻¹ (C-O-C epoxide) [23,24]. After the modification with ODA, additional peaks appeared at 3000–3700 cm⁻¹, 2920 cm⁻¹/2850 cm⁻¹, and 1680 cm⁻¹/1213 cm⁻¹ corresponding to -NH, C-H (symmetric and asymmetric stretching), and C(O)-N in the -C(O)NHR structure of amide, respectively [18,25]. The peak at ~1740 cm⁻¹ should be the C=O group but could also be associated and overlapped with adventitious CO₂. Separately, typical Al₂O₃ powder shows a peak at ~1150 cm⁻¹ for Al-O-Al, and additional peaks (1650 cm⁻¹ and 1470 cm⁻¹) are physically adsorbed water moisture [26,27]. Similarly, Raman spectra were obtained to examine the characteristics of 2D carbon materials for defects (D band) and graphitic domains (G band). The D band centered at 1330 cm⁻¹ is associated with the structural defects of GO, whereas the G peak centered at 1596 cm⁻¹ reflects the vibration model of graphitic carbon (C=C bond). The ratio of D peak to G peak (I_D/I_G) is often employed to predict the relationship between the sp³ and sp² hybrid carbon atoms. The I_D/I_G ratio of GO and ODA-GO is around 1.01 and 1.07, respectively. This slight increase in the peak intensity suggested the electronic structural changes of GO upon the introduction of ODA (i.e., chemical grafting), which were in suitable agreement with the literature reports [20,21,25].

Thermogravimetric analysis (TGA) was utilized to investigate the thermal stability and the presence of ODA around GO (Figure 4). While the Al₂O₃ powder showed negligible weight loss (1.5 wt%), GO and ODA-GO displayed notably different decomposition patterns as a function of temperature. As these samples were pre-dried at 80 °C for 10 min, no detectable evaporation of water moisture was observed below 100 °C where the loss of water around this temperature is often ascribed to the evaporation of moisture coming from the layer structure of GO [28,29]. The rapid weight loss started over 120 °C, which was presumably caused by the decomposition of oxygen-containing functional groups of GO (e.g., hydroxyl, carboxyl, and epoxy). In contrast, the ODA-modified GO initially exhibited gradual decomposition at 120 °C and a significant weight loss over 350 °C. The gradual weight loss in the early stage could be due to the surface functionalization between ODA and carboxylic and/or epoxy groups (i.e., reducing the number of these groups to undergo the deoxygenation process). Further decline could be due to the gradual loss of covalently bonded ODA from the surface of GO. The significant weight loss (far more than bare GO) over 350 °C was closely matched with the boiling point of ODA (349 °C). These distinctively different thermal degradation patterns clearly indicated the successful modification of GO with ODA. The utilization of highly miscible ODA-GO can be effectively implemented between two phases (i.e., liquid PDMS and solid Al₂O₃ particles) to provide suitable conduction pathways throughout the TIM networks.

The X-ray diffraction (XRD) patterns of Al₂O₃ powder, GO, and ODA-GO were also obtained to examine their structural information (Figure 4). The Al₂O₃ powder displayed many diffraction peaks at 25.9 (012), 35.5 (104), 38.1 (110), 43.6 (113), 52.8 (024), 57.8 (116), 61.6 (018), 66.8 (214), 68.5 (300), and 77.3° (119), which corresponded to a typical rhombohedral structure of pure α-alumina [30]. The diffraction patterns of bare GO and ODA-GO were distinctively different. Specifically, GO shows a strong peak at 2θ = 11.9° (001) for the preferential orientation of GO basal planes parallel to the sample plane and a weak/broad peak at slightly over 21° for somewhat disordered graphitic domains [19,31]. Based on the characteristic peak of GO, the interlayer space was calculated to be 0.74 nm (an example of the interlayer spacing calculation shown in Figure 4) [32,33]. The ODA-modified GO displayed a small new peak at 2θ = 5.25°, often interpreted as the covalent interactions between the basal plane oxygen functional groups and the amino group of ODA. The calculated interlayer space was around 1.68 nm, which implied successful functionalization by intercalating the alkyl chains of ODA into the GO spaces [20,34]. In addition, a strong and broad characteristic peak at 2θ = 21.3° corresponding to the interlamellar spacing of 0.42 nm is related to the graphite structure (i.e., reduced GO). This feature is often explained by the partial elimination of oxygen-containing functional groups from GO and the rearrangement of graphitic domains during the ODA modification, which was confirmed by wettability changes from hydrophilic to hydrophobic characteristics above [33,35,36]. Thus, the presence of ODA around GO could play an important role during the preparation of thermal interface materials (TIMs) with PDMS and Al₂O₃ powder by improving the dispersity/miscibility of GO [37,38].

3.3. Structural Features of Various TIMs

After thorough characterizations of these two fillers, TIMs were systematically prepared in the form of a thin pad as a function of GO and ODA-GO contents (Figure 5). Although the use of a higher amount of fillers generally increases the TC due to more dense packing (i.e., decreased defects) [12], a fixed amount of Al₂O₃ particles was introduced in the PDMS precursor to examine the role of thermally conductive GO and ODA-GO. Thus, one can easily monitor the influence of the GO additives and their distribution on the TC of the final TIMs as a function of the content. In addition, the overall TC property can be compared based on the dispersion degrees of GO and ODA-GO across the TIM pads. The color of the TIM samples gradually darkened with the increase in GO content (shown by digital photos). While the TIM samples containing ODA-GO generally showed an even color distribution, the GO-containing TIMs displayed a somewhat irregular color with the presence of random dark spots. Upon the use of GO amounts at 2.5% and above, the color of the samples was completely black, which made it difficult to visually examine the general distribution of GO across the TIMs. It appears that the use of ODA-GO greatly improved the dispersity and compatibility in the PDMS polymer matrix. SEM images show the smooth top surface and dense distribution of Al₂O₃ particles across the TIM samples (side views). The corner view generally shows how Al₂O₃ particles are arranged from top to bottom of a typical TIM layer. Unfortunately, it was very difficult to identify the distribution patterns of carbon-based GO and ODA-GO across the layer (Figure S2). Although examining the clear features of the PDMS resin and GO is challenging, very small dots (a dotted red circle) randomly appearing throughout the cross-section, particularly for ODA-GO-containing TIM, could be due to the presence of ODA-GO. It is noted that the use of over 2.5 wt% GO and ODA-GO required a much longer curing time, and the resulting TIMs were somewhat brittle, easily creating cracks during the handling process under our preparation conditions. This observation implied the need for a higher amount of the PDMS precursor resin as a binder with the increase in the second filler content.

To evaluate the distribution of GO and ODA-GO, typical optical and polarized light images were collected from the top surface (Figure 6 and Figure S3). Unlike brightfield optical images, polarized light images can provide clearer information on absorption color and optical path boundaries between substances with differing refractive indices. These images evidently presented the dense packing of Al₂O₃ particles and the slightly different distribution of GO and ODA-GO across the surface. Specifically, the optical image and the polarized image of the TIM pad prepared with only Al₂O₃ particles appear to have very similar patterns, presumably due to the same composition and uniformity. The small dark spots could be associated with surface defects (e.g., gaps) that affect the homogeneity of the pad. However, the TIM pads containing GO and ODA-GO displayed somewhat different images. The conventional optical images revealed a uniform distribution of Al₂O₃ particles, whereas the polarized images mainly show very dark spots, possibly coming from GO and ODA-GO, caused by much less reflection of the light source. The distribution of GO and ODA-GO across TIM pads was distinctively different, implying that the way the GO integration pattern into a mixture of Al₂O₃ and PDMS is highly dependent on its surface modification.

3.4. Thermal Conductivity of Various TIMs as a Function of GO and ODA-GO Contents

Based on the previous report [8], the TC values can be theoretically calculated by the Maxwell model using a two-phase system (PDMS and Al₂O₃ particles) [1]. This model can effectively predict the gradual increase in TC values of composite pads with a relatively low loading range of fillers (10 wt%–50 wt%). However, the experimental TCs were lower than 1 W/m·K at this content range when using only Al₂O₃ particles as the main filler, which was also reported by another group [17]. As such, an additional step is taken to significantly improve the overall TC of the TIM pads consisting of PDMS and Al₂O₃ particles, simply by introducing a small amount of thermally conductive second filler at a fixed amount of the main filler. Thus, the TCs of two series of TIM pads were examined as a function of GO and ODA-GO content. Initially, the TC value for a pure PDMS polymer matrix without a thermally conductive filler (i.e., Al₂O₃ particles) was examined to be ~0.263 W/m·K, which was comparable to typical polymer-based resins reported by others [8,17]. Negligible changes were observed upon the addition of GO and ODA-GO (e.g., 0.5 wt%), implying the necessity of the main filler, Al₂O₃ particles, to bridge a continuous heat conduction path. After the introduction of Al₂O₃ particles, the TC value notably increased to ~1.175 W/m·K due to the establishment of a heat transfer path where the phonons (i.e., thermal energy carriers) can be easily delivered from one end to the other end. It is noted that no heat transfer by electrons is expected because of the dielectric Al₂O₃ filler. Upon the introduction of additional fillers (i.e., GO and ODA-GO), all TIM pads displayed greater TC values, presumably because GO and ODA-GO have unique structural (i.e., flat) and thermally conductive features (e.g., 72 W/m·K–670 W/m·K for GO depending on the oxidation degree vs. ~30 W/m·K for high-purity Al₂O₃ [11,39]) can serve as a bridge by filling in the gaps among Al₂O₃ particles. The schematic drawing in Figure 7 illustrates speculated heat pathways in the presence of Al₂O₃, GO, and ODA-GO. Enabling the continuous packing of Al₂O₃ particles can greatly help improve the heat conduction channels, thereby potentially maximizing the heat transfer capability. In other words, the 2D structure of GO-based materials can serve as an effective bridge to establish more thermally conductive networks across the pads. Generally, the incorporation of ODA-GO into Al₂O₃ particles displayed slightly higher TC values than those with unmodified GO. This observation is probably due to the better miscibility of the ODA-modified GO in a mixture of PDMS and Al₂O₃ particles (e.g., even distribution with a less agglomerated form of GO). Interestingly, the TC values gradually decreased when the amount of GO increased, possibly due to the uneven dispersion of GO, causing relatively poor contact with Al₂O₃ in the PDMS matrix, which was also reported by others [4,7,12].

A detectable decrease in the TC as a function of GO and ODA-GO content could be associated with voids and defects caused by a slightly excess two-dimensional layered structure during the spherical Al₂O₃ powder spreading process. As the GO and ODA-GO contents increase, their dispersity and/or compatibility could cause the formation of these voids and defects across the pads, which could potentially induce interface thermal resistance and result in phonon scattering, thereby reducing the TC. Under our preparation conditions, the TIM pad prepared with 0.25 wt% of ODA-GO displayed notably greater TC (~55% increase) than that of pure Al₂O₃ powder and ~20% higher than the TIM pad prepared with an equivalent GO content. Furthermore, the TIMs were prepared on a large scale (>12 cm × 12 cm) to demonstrate their practicability (Figure S4). The TIMs prepared only with Al₂O₃ and a mixture of Al₂O₃ and 0.5 wt% GO appeared to possess more defects (caused by air pockets) than those prepared with 0.5 wt% ODA-GO, suggesting the importance of the miscibility of fillers. Vacuuming out entrapped air bubbles from the mixture before the casting and curing steps could easily eliminate this problematic issue, which will be reported in the future. Overall, the introduction of a small amount of thermally conductive GO filler can increase the heat transfer capability of conventional Al₂O₃-based TIM pads. Particularly, the modification of GO with ODA readily improves their miscibility across the TIM pad to further enhance the overall TC performance.

4. Conclusions

The simple modification of GO with ODA notably improved its compatibility in the systematic preparation of TIMs using a mixture of PDMS polymer precursor and Al₂O₃ particles. The successful surface fabrication of GO was confirmed by water wettability and vibrational spectrophotometers, thermal stability, and X-ray-based interlayer distance. Upon the introduction of thermally conductive GO and ODA-GO as additives, two series of TIMs prepared with a fixed amount of the main filler, Al₂O₃ particles, offered the possibility of evaluating the relationships between the TC properties and the amounts of additives. The TC values of TIMs prepared only with PDMS and a mixture of PDMS and Al₂O₃ particles were much lower than those prepared with GO and ODA-GO, where these additives served as a bridge by possibly filling up the gaps across TIM networks. Particularly, the TIMs containing ODA-GO showed detectably higher TC efficiency due to the enhanced miscibility and dispersion of GO across the TIM networks. However, increasing the amount of these fillers may have induced uneven packing and created defects across the pads to elevate interface thermal resistance (i.e., a detrimental effect). Investigating the TC properties of conventional TIMs upon adding a small quantity of GO could provide a better understanding of the role of the second additive in thermal management systems. As such, the proper modification of thermally conductive fillers to enhance their dispersity across the polymer and main filler matrices can lead to the development of inexpensive but highly efficient and practical TIM-2-type pads.