Binary Promoter Improving the Moderate-Temperature Adhesion of Addition-Cured Liquid Silicone Rubber for Thermally Conductive Potting

The strong adhesion of thermally conductive silicone encapsulants on highly integrated electronic devices can avoid external damages and lead to an improved long-term reliability, which is critical for their commercial application. However, due to their low surface energy and chemical reactivity, the self-adhesive ability of silicone encapsulants to substrates need to be explored further. Here, we developed epoxy and alkoxy groups-bifunctionalized tetramethylcyclotetrasiloxane (D4H-MSEP) and boron-modified polydimethylsiloxane (PDMS-B), which were synthesized and utilized as synergistic adhesion promoters to provide two-component addition-cured liquid silicone rubber (LSR) with a good self-adhesion ability for applications in electronic packaging at moderate temperatures. The chemical structures of D4H-MSEP and PDMS-B were characterized by Fourier transform infrared spectroscopy. The mass percentage of PDMS-B to D4H-MSEP, the adhesion promoters content and the curing temperature on the adhesion strength of LSR towards substrates were systematically investigated. In detail, the LSR with 2.0 wt% D4H-MSEP and 0.6 wt% PDMS-B exhibited a lap-shear strength of 1.12 MPa towards Al plates when curing at 80 °C, and the cohesive failure was also observed. The LSR presented a thermal conductivity of 1.59 W m−1 K−1 and good fluidity, which provided a sufficient heat dissipation ability and fluidity for potting applications with 85.7 wt% loading of spherical α-Al2O3. Importantly, 85 °C and 85% relative humidity durability testing demonstrated LSR with a good encapsulation capacity in long-term processes. This strategy endows LSR with a good self-adhesive ability at moderate temperatures, making it a promising material requiring long-term reliability in the encapsulation of temperature-sensitive electronic devices.


Introduction
Optical, electronic devices and electrical modules are continuously evolving in functionality and achieving high performance while tending to be smaller and more integrated [1][2][3][4]. These devices are generally encapsulated with polymeric potting materials to achieve heat transfer and avoid environmental damages (e.g., moisture penetration and mechanical damage), thus boosting their long-term reliability [5][6][7]. Addition-cured liquid silicone rubber (LSR) possesses an amazing array of properties including a high dielectric breakdown strength, flame resistance, heat and cold resistance, stress relieving properties and long-lasting durability, which make it one of the most promising materials for the protection of electronic devices [8,9]. The poor mechanical properties of pure LSR limit its practical applications [9]. Researchers have expended great efforts to incorporate reinforcement fillers and functional fillers into LSR to improve its mechanical properties reactive groups and siloxane backbones endowed the adhesion promoters with a good compatibility with LSR, while the boron-containing groups were designed to facilitate the curing of epoxy groups. By incorporating two components, PDMS-B and D 4 H-MSEP, into LSR together, the as-prepared two-component addition-curing LSR presented good stability during storage and exhibited high reactivity and adhesion. The effect of the weight percentage of PDMS-B to D 4 H-MSEP and the content of adhesion promoters on the adhesion strength of LSR were systematically investigated. Additionally, the durability test at a temperature of 85 • C and a relative humidity of 85% was performed to evaluate the long-term stability of LSR and its encapsulation capacity.

Synthesis of Boron-Modified PDMS
Boron-modified PDMS (PDMS-B) was synthesized through a hydrosilylation reaction between ATDB and PHMS-H, as shown in Figure 1 Route A. The detailed synthesis procedure was as follows: PHMS-H (10.0 g) was dissolved in anhydrous toluene (20.0 mL) in a four-neck round bottom flask equipped with a temperature sensor, a dropping funnel, a reflux condenser and nitrogen-blow devices. The reaction solution was kept at 20 • C for 30 min to reach the pre-equilibration, followed by adding Karstedt's platinum catalyst (0.02 g) into the solution. Subsequently, excessive ATDB (10.0 g) was added dropwise into the flask through the dropping funnel in 10 min, the reaction mixture was stirred for 2 h and then temperature was increased to 60 • C until all Si-H groups were reacted. The unreacted ATDB and toluene were removed through a thin-film evaporator at 60 • C and 200 Pa to yield the desired product PDMS-B as a colorless, transparent and viscous liquid.

Synthesis of Epoxy and Alkoxy Groups-Bifunctionalized D 4 H
The synthetic pathway of epoxy and alkoxy groups-bifunctionalized D 4 H is illustrated below (Figure 1 Route B): typically, ATMS (6.48 g, 40.0 mmol) and EPHE (3.92 g, 40.0 mmol) were dissolved in anhydrous toluene (20.0 mL) in the aforementioned four-neck round bottom flask. After adding Karstedt's platinum catalyst (0.01 g), the flask was pre-equilibrated with N 2 for 30 min at 20 • C. D 4 H (9.6 g, 40 mmol) was added dropwise into the solution for 10 min under vigorous stirring. After all of the -CH = CH 2 groups from ATMS and EPHE were reacted, which was monitored by Fourier-transform infrared spectroscopy (stretching vibration of -CH = CH 2 groups: 1640 cm −1 ), the solution was mixed with activated carbon for 2 h and filtrated through a 0.22 µm filter to remove the residual platinum catalyst. The unreacted monomers and toluene were removed through the thin-film evaporator to yield the epoxy and alkoxy bifunctional D 4 H-MSEP as a colorless and transparent liquid.

Synthesis of Epoxy and Alkoxy Groups-Bifunctionalized D4H
The synthetic pathway of epoxy and alkoxy groups-bifunctionalized D4H is illustrated below (Figure 1 Route B): typically, ATMS (6.48 g, 40.0 mmol) and EPHE (3.92 g, 40.0 mmol) were dissolved in anhydrous toluene (20.0 mL) in the aforementioned fourneck round bottom flask. After adding Karstedt's platinum catalyst (0.01 g), the flask was pre-equilibrated with N2 for 30 min at 20 °C. D4H (9.6 g, 40 mmol) was added dropwise into the solution for 10 min under vigorous stirring. After all of the -CH = CH2 groups from ATMS and EPHE were reacted , which was monitored by Fourier-transform infrared spectroscopy (stretching vibration of -CH = CH2 groups: 1640 cm −1 ), the solution was mixed with activated carbon for 2 h and filtrated through a 0.22 μm filter to remove the residual platinum catalyst. The unreacted monomers and toluene were removed through the thin-film evaporator to yield the epoxy and alkoxy bifunctional D4H-MSEP as a colorless and transparent liquid.

Preparation of Self-Adhesive Two-Component Addition-Curing LSR
PDMS-Vi, PDMS-mVi and Al2O3 were kneaded with a planetary mixer for 30 min in the designed amounts, as shown in Table 1, followed by a heat treatment at 150 °C and a vacuum of 500 Pa for 1 h to promote the uniform dispersion of Al2O3 in the PDMS matrix. Subsequently, the Karstedt's platinum catalyst and PDMS-B were added in the cooled PDMS/Al2O3 matrix and kneaded for 30 min under vacuum conditions to obtain the component A of the two-component LSR. PDMS-Vi, PDMS-mVi and Al2O3 were treated as above to obtain uniform liquid silicone rubber blends. PHMS-H, D4H-MSEP and 3methyl-1-pentyn-3-ol (platinum inhibitor) were mixed with a cooled PDMS/Al2O3 matrix under vacuum conditions to obtain the silicone composite termed as the component B of the two-component LSR. Component A and component B were thoroughly stirred and then mixed in equal proportions to afford the two-component LSR. The LSR was deaired at a vacuum of 500 Pa for 2 min before using.

Preparation of Self-Adhesive Two-Component Addition-Curing LSR
PDMS-Vi, PDMS-mVi and Al 2 O 3 were kneaded with a planetary mixer for 30 min in the designed amounts, as shown in Table 1

Fourier Transform Infrared (FTIR) Analysis
FTIR measurements were performed on a Thermofisher Scientific Nicolet iS10 spectrometer (USA) to identify the chemical structures of the synthesized PDMS-B and D 4 H-MSEP. Samples (0.2 mL) were coated on KBr slices (25 mm × 2 mm) to form thin films for FTIR measurements.

Viscosity Measurement
The viscosity of PDMS/Al 2 O 3 composites was measured using a Brookfield DV2T viscometer at 25 • C according to the ASTM D-445 standard. Three specimens were measured for each measurement, and the mean value was calculated.

Mechanical Properties Test
The tensile strength and elongation at break of the cured LSR samples were measured by a Gotech AI-7000S universal testing machine at a stretching rate of 50 mm min −1 at 25 • C according to the GB/T 528 standard. Under 3000 MPa pressure, LSR samples were cured at 80 • C for 2 h and then at 25 • C for another 24 h to form a 2 mm-thick rectangular sheet. The samples were tailored into dumbbell shapes before mechanical tests. Five specimens were measured for each measurement, and the mean value was calculated.

Thermal Conductivities Test
The thermal conductivities of the LSR sheets (25 mm × 25 mm, 2 mm-thick) were tested by a thermal analyzer (TC3000E, Xiatech, Xi'an, China) according to the ASTM D-5930 standard. Five specimens were measured for each measurement, and the mean value was calculated

Adhesion Performance
Tensile lap-shear strength tests were performed using a Gotech AI-7000S universal testing machine to evaluate the adhesion performance of the LSR towards Al and printed circuit board (PCB) plates ( Figure 2), both of which are the major substrate materials for electronic devices. The Al and PCB plates were rinsed with isopropanol in an ultrasonic bath for 10 min at room temperature and then dried at 60 • C. Two-component LSR was potted into the tailored gap between two adherends and cured at 80 • C without any pressure followed by 25 • C for 24 h. The effective bonding surface of the LSR and adherends was 25 × 25 mm 2 , and the thickness of the LSR was 2 mm. The lap-shear strength between the LSR and adherends was calculated as the following equation: where τ s , F m and w are the lap-shear strength, maximum force and effective bonding surface, respectively. Five specimens were measured for each measurement, and the mean value was calculated.

Structural Characterization of PDMS-B and D4H-MSEP
The chemical structures of the raw materials and resultant PDMS-B and D4H-MSEP were identified by FT-IR, as shown in Figure 3. In Figure 3a, the characteristic band of 1300-1500 cm −1 , with the peak centered at 1370 cm −1 and the sharp peak at 1640 cm −1 , corresponded to the stretching vibration of B-O bonds and -CH = CH2 groups in ATDB, respectively [36][37][38]. The band in the range of 2080-2200 cm −1 was ascribed to the stretching vibration of -Si-H groups in PHMS-H. In the FT-IR spectrum of PDMS-B, the appearance of a new absorption peak for B-O bonds and the absence of an absorption peak for the -Si-H groups and -CH = CH2 groups indicated the complete hydrosilylation

Structural Characterization of PDMS-B and D 4 H-MSEP
The chemical structures of the raw materials and resultant PDMS-B and D 4 H-MSEP were identified by FT-IR, as shown in Figure 3. In Figure 3a, the characteristic band of 1300-1500 cm −1 , with the peak centered at 1370 cm −1 and the sharp peak at 1640 cm −1 , corresponded to the stretching vibration of B-O bonds and -CH = CH 2 groups in ATDB, respectively [36][37][38]. The band in the range of 2080-2200 cm −1 was ascribed to the stretching vibration of -Si-H groups in PHMS-H. In the FT-IR spectrum of PDMS-B, the appearance of a new absorption peak for B-O bonds and the absence of an absorption peak for the -Si-H groups and -CH = CH 2 groups indicated the complete hydrosilylation reaction between PHMS-H and ATDB and the successful synthesis of boron-modified PDMS. In Figure 3b, the sharp absorption peak at 2170 cm −1 was assigned to -Si-H groups in D 4 H. The characteristic absorption peaks at 2840 cm −1 and 910 cm −1 were ascribed to the stretching vibration of -Si-OCH 3 groups in the ATMS and epoxy groups in EPHE, respectively [39][40][41]. The appearance of new absorption peaks for -Si-OCH 3 groups and epoxy groups, the reduced absorption intensity of the peak for -Si-H groups and the disappearance of the absorption peak for -CH = CH 2 groups in the yielded D 4 H-MSEP demonstrated that the epoxy groups and alkoxy groups-bifunctionalized D 4 H was synthesized. Notably, partial Si-H groups were intentionally retained, as observed in the spectrum of D 4 H-MSEP, in order to bind to the LSR matrix through hydrosilylation during the further curing process.

Structural Characterization of PDMS-B and D4H-MSEP
The chemical structures of the raw materials and resultant PDMS-B and D4H-MSEP were identified by FT-IR, as shown in Figure 3. In Figure 3a, the characteristic band of 1300-1500 cm −1 , with the peak centered at 1370 cm −1 and the sharp peak at 1640 cm −1 , corresponded to the stretching vibration of B-O bonds and -CH = CH2 groups in ATDB, respectively [36][37][38]. The band in the range of 2080-2200 cm −1 was ascribed to the stretching vibration of -Si-H groups in PHMS-H. In the FT-IR spectrum of PDMS-B, the appearance of a new absorption peak for B-O bonds and the absence of an absorption peak for the -Si-H groups and -CH = CH2 groups indicated the complete hydrosilylation reaction between PHMS-H and ATDB and the successful synthesis of boron-modified PDMS. In Figure 3b, the sharp absorption peak at 2170 cm −1 was assigned to -Si-H groups in D4H. The characteristic absorption peaks at 2840 cm −1 and 910 cm −1 were ascribed to the stretching vibration of -Si-OCH3 groups in the ATMS and epoxy groups in EPHE, respectively [39][40][41]. The appearance of new absorption peaks for -Si-OCH3 groups and epoxy groups, the reduced absorption intensity of the peak for -Si-H groups and the disappearance of the absorption peak for -CH = CH2 groups in the yielded D4H-MSEP demonstrated that the epoxy groups and alkoxy groups-bifunctionalized D4H was synthesized. Notably, partial Si-H groups were intentionally retained, as observed in the spectrum of D4H-MSEP, in order to bind to the LSR matrix through hydrosilylation during the further curing process.

Thermal Conductivity and Mechanical Properties of LSR
The Al 2 O 3 loading amount plays a key role in the thermal conductivity and viscosity of LSR, which directly affects the heat dissipation ability and potting processability. As shown in Figure 4, the thermal conductivity and viscosity of LSR monotonically increased with the increase in the Al 2 O 3 content, and such an increasing trend is more obvious at a higher Al 2 O 3 loading amount. In detail, the thermal conductivity of LSR increased from 0.78 W m −1 K −1 to 1.81 W m −1 K −1 as the Al 2 O 3 content increased from 75 wt% to 87.5 wt%. The viscosity of LSR increased from 2622 mPa·s to 9185 mPa·s as the Al 2 O 3 content increased from 75 wt% to 85.7 wt%. However, the viscosity increased dramatically to 14,452 mPa·s when the Al 2 O 3 content increased to 87.5 wt%. The aggregation of Al 2 O 3 and the intermolecular interaction between the Al 2 O 3 agglomerates resulted in the rapidly increased viscosity of LSR, which made it unsuitable for potting applications in some complex devices where the viscosity should be less than 10,000 mPa·s. Therefore, the LSR with an 85.7 wt% Al 2 O 3 loading amount, showing a thermal conductivity of 1.59 W m −1 K −1 and a viscosity of 9185 mPa·s, was selected for the further improvement of the adhesion performance.
14,452 mPa•s when the Al2O3 content increased to 87.5 wt%. The aggregation of Al2O3 and the intermolecular interaction between the Al2O3 agglomerates resulted in the rapidly increased viscosity of LSR, which made it unsuitable for potting applications in some complex devices where the viscosity should be less than 10,000 mPa•s. Therefore, the LSR with an 85.7 wt% Al2O3 loading amount, showing a thermal conductivity of 1.59 W m −1 K −1 and a viscosity of 9185 mPa•s, was selected for the further improvement of the adhesion performance. The effect of the Al2O3 content on the mechanical properties of the LSR were investigated and are shown in Figure 5. It was observed that the tensile strength and the elongation at break of the LSR were improved by the incorporation of the semi-reinforcing filler Al2O3. The optimal tensile strength of LSR reached up to 1.37 MPa with 83.3 wt% Al2O3 content, while the highest elongation at break reached up to 144.7% with 80.0 wt% Al2O3 content. However, the excessive incorporation of Al2O3 led to their aggregation, and the Al2O3 agglomerates act as defects in the LSR, reducing the effective interaction area between the particle and the matrix and hence the deterioration of the LSR mechanical properties. The LSR loaded with 85.7 wt% Al2O3 content, showing a tensile strength of 1.31 MPa and an elongation at break of 114%, was selected for the following studies.     The effect of D4H-MSEP content on the lap-shear strength of LSR and adherends joints was investigated and is shown in Figure 7. Without the addition of an adhesion promoter, the LSR showed almost no adhesion to adherends due to its low surface energy and low surface activity, leading to the easy removal of LSR from the adherend surfaces. In contrast, the lap-shear strength between the LSR and Al (PCB) plates significantly increased to 1.04 MPa as the D4H-MSEP content increased to 1.5 wt% (mass percentage of PDMS-B to D4H-MSEP was kept at 30%), followed by a smooth enhancement approaching 1.12 MPa with the further increase in D4H-MSEP content to 2.0 wt%. Increasing the adhesion promoter content would facilitate the enrichment of PDMS-B and D4H-MSEP molecules on the LSR surface, hence improving the adhesion strength of LSR. The concentration of the adhesion promoter molecules at the interface was limited owing to the steric hindrance; however, the lap-shear strength showed a limited enhancement when the adhesion promoters content exceeded 1.5 wt%. Overall, these results confirmed that the high reactivity and interface enrichment of the binary adhesion promoters PDMS-B and D4H-MSEP provided a synergistic effect on the enhancement of LSR adhesion performance and endowed the LSR with a great adhesion ability at a moderate temperature of 80 °C. The effect of D 4 H-MSEP content on the lap-shear strength of LSR and adherends joints was investigated and is shown in Figure 7. Without the addition of an adhesion promoter, the LSR showed almost no adhesion to adherends due to its low surface energy and low surface activity, leading to the easy removal of LSR from the adherend surfaces. In contrast, the lap-shear strength between the LSR and Al (PCB) plates significantly increased to 1.04 MPa as the D 4 H-MSEP content increased to 1.5 wt% (mass percentage of PDMS-B to D 4 H-MSEP was kept at 30%), followed by a smooth enhancement approaching 1.12 MPa with the further increase in D 4 H-MSEP content to 2.0 wt%. Increasing the adhesion promoter content would facilitate the enrichment of PDMS-B and D 4 H-MSEP molecules on the LSR surface, hence improving the adhesion strength of LSR. The concentration of the adhesion promoter molecules at the interface was limited owing to the steric hindrance; however, the lap-shear strength showed a limited enhancement when the adhesion promoters content exceeded 1.5 wt%. Overall, these results confirmed that the high reactivity and interface enrichment of the binary adhesion promoters PDMS-B and D 4 H-MSEP provided a synergistic effect on the enhancement of LSR adhesion performance and endowed the LSR with a great adhesion ability at a moderate temperature of 80 • C.  Figure 8 presents the effect of the curing temperature on the adhesion strength of LSR with adherends. The LSR exhibited a gradually enhanced lap-shear strength towards both Al and PCB plates as the curing temperature increased from 60 to 90 °C, resulting from the faster diffusion of the adhesion promoters to the LSR surface and the higher reactivity of the epoxy groups in D4H-MSEP at higher curing temperatures. When the curing temperature reached 80 °C, the lap-shear strength of the LSR towards both of the two adherends was higher than 1.0 MPa, and the cohesive failure was observed at the interface of the LSR/adherends joints after lap-shear strength testing. This result indicated that the obtained LSR could be used for the encapsulation of temperature-sensitive materials such as plastic seals and sensitive electronic components. The LSR encapsulant prepared in this work is compared against the addition-cured silicone encapsulants reported in the literature and commercial products ( Table 2). The LSR potting compound described in this paper can achieve effective adhesion at lower temperatures.   Figure 8 presents the effect of the curing temperature on the adhesion strength of LSR with adherends. The LSR exhibited a gradually enhanced lap-shear strength towards both Al and PCB plates as the curing temperature increased from 60 to 90 • C, resulting from the faster diffusion of the adhesion promoters to the LSR surface and the higher reactivity of the epoxy groups in D 4 H-MSEP at higher curing temperatures. When the curing temperature reached 80 • C, the lap-shear strength of the LSR towards both of the two adherends was higher than 1.0 MPa, and the cohesive failure was observed at the interface of the LSR/adherends joints after lap-shear strength testing. This result indicated that the obtained LSR could be used for the encapsulation of temperature-sensitive materials such as plastic seals and sensitive electronic components. The LSR encapsulant prepared in this work is compared against the addition-cured silicone encapsulants reported in the literature and commercial products ( Table 2). The LSR potting compound described in this paper can achieve effective adhesion at lower temperatures.  Figure 8 presents the effect of the curing temperature on the adhesion strength of LSR with adherends. The LSR exhibited a gradually enhanced lap-shear strength towards both Al and PCB plates as the curing temperature increased from 60 to 90 °C, resulting from the faster diffusion of the adhesion promoters to the LSR surface and the higher reactivity of the epoxy groups in D4H-MSEP at higher curing temperatures. When the curing temperature reached 80 °C, the lap-shear strength of the LSR towards both of the two adherends was higher than 1.0 MPa, and the cohesive failure was observed at the interface of the LSR/adherends joints after lap-shear strength testing. This result indicated that the obtained LSR could be used for the encapsulation of temperature-sensitive materials such as plastic seals and sensitive electronic components. The LSR encapsulant prepared in this work is compared against the addition-cured silicone encapsulants reported in the literature and commercial products ( Table 2). The LSR potting compound described in this paper can achieve effective adhesion at lower temperatures.   The encapsulation with silicone potting materials can improve the device operation reliability and durability, which is critical for optical and electronic applications including fiber optic bundles, sensors, insulated gate bipolar transistors and so on [44]. The temperature of 85 • C and the relative humidity of 85% were set to carry out the durability testing so as to accelerate the evaluation of the device performance. The dependence of lap-shear strength on testing time was investigated to evaluate the long-term reliability of LSR for device encapsulation [45]. As shown in Figure 9, the lap-shear strength of LSR towards both Al and PCB plates was almost unchanged during the 200 h testing period. This good, durable adhesion would provide long-term protection for a variety of optical and electronic devices.  The encapsulation with silicone potting materials can improve the device operation reliability and durability, which is critical for optical and electronic applications including fiber optic bundles, sensors, insulated gate bipolar transistors and so on [44]. The temperature of 85 °C and the relative humidity of 85% were set to carry out the durability testing so as to accelerate the evaluation of the device performance. The dependence of lap-shear strength on testing time was investigated to evaluate the long-term reliability of LSR for device encapsulation [45]. As shown in Figure 9, the lap-shear strength of LSR towards both Al and PCB plates was almost unchanged during the 200 h testing period. This good, durable adhesion would provide long-term protection for a variety of optical and electronic devices.

Conclusions
Binary adhesion promoters, boron-modified PDMS-B and epoxy and alkoxy groupsbifunctionalized D 4 H-MSEP, were synthesized in order to provide an addition-cured thermally conductive LSR encapsulant with a strong self-adhesion ability at moderate temperatures. The thermal conductivity, viscosity and mechanical and adhesion properties of the LSR encapsulant were systematically investigated. In detail, the LSR containing 2.0 wt% D 4 H-MSEP and 0.6 wt% PDMS-B showed a lap-shear strength of 1.12 MPa to-wards the Al plate when curing at 80 • C. The cohesive failure occurred at the interfaces of both the LSR/Al and LSR/PCB joints. The LSR with an 85.7 wt% Al 2 O 3 loading content provided a sufficient heat dissipation ability and fluidity for potting applications. Importantly, durability testing at the temperature of 85 • C and at 85% relative humidity suggested that LSR features a good encapsulation capacity during a long-term operation. This design strategy endows LSR with long-term reliability for further potential applications in the encapsulation of temperature-sensitive optical and electronic devices at specific working temperatures.