Non-hardening thermal grease containing boron nitride

JP2026501680A5Pending Publication Date: 2026-01-28DOW SILICONES CORP +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
JP2025539436
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Filing Date
2023-01-13
Publication Date
2026-01-28

AI Technical Summary

Technical Problem

Non-curing thermal greases used in microelectronic devices experience pump-out due to die warpage during power cycling, leading to thermal grease loss and formation of hot spots, which impairs performance.

Method used

A non-hardening thermal grease composition comprising 80-95% aluminum, alumina, or zinc oxide fillers, 0.2-8% boron nitride platelets, and 0.2-10% trialkoxylated polydimethylsiloxane, with a carrier fluid, providing improved pump-out resistance and maintaining thermal conductivity.

Benefits of technology

The composition achieves excellent pump-out resistance, printability, and thermal conductivity, preventing die warpage-induced grease loss and ensuring consistent heat transfer.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2024148597000001
    Figure 2024148597000001
  • Figure 2024148597000002
    Figure 2024148597000002
Patent Text Reader

Abstract

The present invention relates to a non-curing thermal grease comprising filler particles, boron nitride platelet particles, and trialkoxysilyl-terminated polydimethylsiloxane, which is useful as a thermal interface material in microelectronic devices for transporting heat generated by a heat source to a heat sink.
Need to check novelty before this filing date? Find Prior Art

Description

[Technical Field]

[0001] The present invention relates to compositions comprising non-curing thermal greases containing boron nitride, which are useful as thermal interface materials in microelectronic devices for transporting heat generated by a heat source to a heat sink.

[0002] Non-curing thermal greases contain large amounts of one or more thermally conductive fillers, such as zinc oxide, alumina (i.e., Al2O3), and aluminum, dispersed in a non-curing polyorganosiloxane matrix. They have high thermal conductivity, low thermal resistance, and low decomposition potential. For applications with bare die chip architectures where thermal grease is applied directly between the die and the heat sink as a thermal interface material, die warpage during power cycling causes repeated die movement, which tends to squeeze the thermal grease out of the interface. This loss of thermal grease, also known as pump-out, leads to the formation of hot spots on the die, which impairs performance. Therefore, it is desirable to formulate a non-curing thermal grease with improved pump-out resistance while maintaining acceptable printability and thermal conductivity. Summary of the Invention

[0003] The present invention provides a non-hardening thermal grease composition comprising, based on the weight of the composition, a) 80 to 95 weight percent of one or more filler particles selected from the group consisting of aluminum, alumina, and zinc oxide; b) 0.2 to 8 weight percent of boron nitride platelet particles; and c) 0.2 to 10 weight percent of a compound having Structure 1:

[0004] [ka] and a trialkoxylated polydimethylsiloxane of the formula: 1are each independently C1-C6-alkyl, Y is O or CH2-CH2, and x is 30 to 200; and d) up to 10 weight percent of a polydimethylsiloxane or poly(phenylmethylsiloxane-co-dimethylsiloxane) carrier fluid having a viscosity in the range of 30 cSt to 200 cSt. The non-curing thermal grease of the present invention provides excellent pump-out resistance while maintaining acceptable printability, thermal conductivity, and viscosity. DETAILED DESCRIPTION OF THE INVENTION

[0005] The present invention provides a non-hardening thermal grease composition comprising, based on the weight of the composition, a) 80 to 95 weight percent of one or more filler particles selected from the group consisting of aluminum, alumina, and zinc oxide; b) 0.2 to 8 weight percent of boron nitride platelet particles; and c) 0.2 to 10 weight percent of a compound having Structure 1:

[0006] [ka] and a trialkoxylated polydimethylsiloxane of the formula: 1 are each independently C1-C6-alkyl, Y is O or CH2-CH2, and x is 30 to 200); d) up to 10 weight percent of a polydimethylsiloxane or poly(phenylmethylsiloxane-co-dimethylsiloxane) carrier fluid having a viscosity in the range of 30 cSt to 200 cSt.

[0007] The composition preferably comprises aluminum and zinc oxide and / or alumina filler particles in a concentration ranging from 85 or 90 weight percent to 95 weight percent based on the weight of the composition. Advantageously, the aluminum particles are present as a multimodal, preferably bimodal, distribution of spherical aluminum particles. The D of the larger spherical aluminum particles is 50The volume average particle size is preferably in the range of 5 μm, or 7 μm, or 8 μm to 25 μm or 10 μm, with the D of these larger particles 99 The volume average particle size is preferably 50 μm or less, or 30 μm or less. The concentration of the larger spherical aluminum particles preferably ranges from 40, or 45, or 48 weight percent to 60, or 55, or 52 weight percent based on the weight of the composition.

[0008] D of the smaller spherical aluminum particles 50 The volume average particle size is preferably in the range of 1 μm to 5 μm. The concentration of the smaller spherical aluminum particles is preferably in the range of 15, or 20, or 23 weight percent to 35, or 30, or 27 weight percent based on the weight of the composition.

[0009] D of zinc oxide (ZnO) particles and / or alumina particles 50 The volume average particle size preferably ranges from 50 nm, or 100 nm, to 500 nm, or 200 nm, or 150 nm. The concentration of ZnO particles and / or alumina particles preferably ranges from 10 or 15 weight percent to 25 or 20 weight percent, based on the weight of the composition. The D of aluminum, zinc oxide, and alumina particles 50 and D 99 The volume average particle size is the D measured using laser refractometry. 50 Refers to the volume average particle size.

[0010] The boron nitride platelet particles preferably have a thickness in the range of 750 nm to 5 μm as measured by Scanning Electron Microscopy (SEM), and a D preferably in the range of 3 μm to 40 μm as measured by Dynamic Light Scattering. 50The boron nitride platelet particles preferably have a diameter-to-thickness aspect ratio ranging from 2:1, or 3:1, or 4:1 to 50:1, or 30:1, or 20:1, or 10:1. The boron nitride platelet particles have a hexagonal crystal structure. During assembly, the boron nitride platelet particles align along approximately the same direction as the substrates after the platelet particles are applied between the substrates. Therefore, the D of the boron nitride platelet particles 50 Particle size does not affect the final bondline thickness. Commercially available examples of boron nitride platelet particles include CarboTherm PCTP30 boron nitride from St. Gobain and PolarTherm PT110 from Momentive Performance Materials. The concentration of the boron nitride platelet particles preferably ranges from 0.5, or 1, or 2 weight percent to preferably 6, or 5, or 4 weight percent based on the weight of the composition.

[0011] C1~C 20 The alkyltri-C1-C6-alkoxysilane is preferably a C6-C 16 -Alkyltrimethoxysilane or C8-C 12 -alkyltrimethoxysilane. 20 An example of an alkyltri-C1-C6-alkoxysilane is n-decyltrimethoxysilane. 20 The concentration of the -alkyltri-C1-C6-alkoxysilane preferably ranges from 0.1 weight percent to 2 or 1 weight percent based on the weight of the composition.

[0012] The concentration of polydimethylsiloxane of Structure 1 preferably ranges from 1 or 3 to 8, based on the weight of the composition. 1 are each preferably methyl, and x preferably ranges from 50, or 80, or 100, to preferably 180, or 150, or 130.

[0013] The composition may also include up to 10 weight percent of a carrier fluid, which is preferably a polydimethylsiloxane or poly(phenylmethylsiloxane-co-dimethylsiloxane) fluid having a viscosity ranging from 50 cSt or 80 cSt to preferably 150 cSt or 120 cSt. In one aspect, the non-curing thermal grease includes from 1, or 2, or 4 weight percent to 10, or 8, or 6 weight percent of a polydimethylsiloxane or poly(phenylmethylsiloxane-co-dimethylsiloxane) carrier fluid, based on the weight of the composition.

[0014] Non-hardening grease compositions contain an insufficient concentration of crosslinkable functional groups, such as a combination of vinyl and Si-H groups, to harden the grease. Preferably, the non-hardening grease contains no vinyl or Si-H groups.

[0015] The compositions of the present invention provide excellent pump-out resistance along with desirable printability, viscosity, and thermal conductivity. Thus, in another aspect, the present invention is a microelectronic device comprising a non-curing thermal grease composition embedded therein as a thermal interface material. [Example]

[0016] In the following examples, unless otherwise noted, all samples were mixed using a Flacktek mixer at 1500 rpm, and pbw refers to parts by weight.

[0017] Comparative Example 1 - Preparation of a Boron Nitride-Free Non-Curing Thermal Grease. Phenylmethylsiloxane-dimethylsiloxane copolymer (100 cSt, 5.10 pbw, Gelest, product code PMM-1021), polydimethylsiloxane of structure 1 (each R and R 1= CH3, x = 110, Y = O, 1.99 pbw), n-decyltrimethoxysilane (0.17 pbw), and Zoco102 ZnO (approximately 0.12 μm, 17.37 pbw) were added to a MAX100 cup and mixed for 15 seconds. TCP-2 aluminum powder (Toyo Aluminum KK, particle size approximately 2 μm, 25.12 pbw) was then added to the mixer, and the ingredients were mixed for 15 seconds. TCP-9 aluminum powder (Toyo Aluminum KK, particle size approximately 9 μm, 50.25 pbw) was then added to the mixer, and the ingredients were mixed for an additional 40 seconds. The ingredients were mixed by hand using a spatula and then mixed in the mixer for 40 seconds. The formulation was then transferred to an aluminum pan and heated under vacuum at 150°C for 1 hour.

[0018] Example 1 - Preparation of a non-curing thermal grease containing boron nitride The process used to prepare the non-curing thermal grease of Comparative Example 1 was repeated, except that PCTP30 boron nitride (Saint Gobain, 3.0 pbw) was added to the mixer and the ingredients were mixed for 15 seconds after adding and mixing the TCP-9 aluminum powder (47.25 pbw). The ingredients were mixed by hand with a spatula and then mixed in the mixer for 40 seconds. The formulation was then transferred to an aluminum pan and heated under vacuum at 150°C for 1 hour.

[0019] Example 2 - Preparation of a non-curing thermal grease containing boron nitride The process used to prepare the non-curing thermal grease of Comparative Example 1 was repeated, except that the phenylmethylsiloxane-dimethylsiloxane was not added, the amount of polydimethylsiloxane of Structure 1 was 7.09 pbw, PCTP30 boron nitride (Saint Gobain, 1.0 pbw) was added to the mixer, and the ingredients were mixed for 15 seconds after adding and mixing TCP-9 aluminum powder (49.25 pbw). The ingredients were mixed by hand with a spatula and then mixed in the mixer for 40 seconds. The formulation was then transferred to an aluminum pan and heated under vacuum at 150°C for 1 hour.

[0020] Example 3 - Preparation of a non-curing thermal grease with boron nitride. Polydimethylsiloxane of structure 1 (each R and R 1 =CH3, x = 110, 354.5 g, 7.09 pbw), n-decyltrimethoxysilane (8.5 g, 0.17 pbw), and Zoco102 ZnO (approximately 0.12 μm, 869 g, 17.37 pbw) were added to a 1 g Baker Perkins sigma blade mixer, and the ingredients were mixed for 30 minutes. TCP-2 aluminum powder (1206 g, 25.12 pbw) was then added to the mixer, followed by TCP-9 aluminum powder (2412 g, 47.25 pbw), and the ingredients were mixed for an additional 55 minutes. BN (150 g, 3.00 pbw) was added to the mixer, and the ingredients were mixed for an additional 15 minutes. The mixture was dried under vacuum (25 Torr, 45 min) with mixing, then heated to 150°C for 60 minutes, after which the heat was removed and the mixture was cooled to room temperature.

[0021] Example 4 - Preparation of a non-curing thermal grease containing boron nitride The preparation was repeated as described in Example 2, except that 5.0 pbw of boron nitride and 45.25 pbw of TCP-9 were used.

[0022] Viscosity measurement The complex viscosity (Pa·s) at the dilatant point was measured according to ASTM D4440-15 (Standard Test Method for Plastics: Dynamic Mechanical Properties Melt Rheology) using a TA Instruments ARES-G2 instrument with 25 mm parallel plates (serrated steel). Test conditions were based on a strain sweep performed at 25°C with a 2.0 mm gap. Measurements were performed using standard procedures with an oscillation frequency of 10 rad / s and a strain amplitude sweep from 0.01 to 200% with 20 sampling points every 10 seconds. The dilatant point was defined as the strain at which the complex viscosity began to increase.

[0023] Thermal Conductivity Measurement Thermal conductivity was measured according to ISO 22007-2:2015 (Test Method for Determining Thermal Conductivity) using a C5501 sensor and a Hot Disk Instrument TPS 2500 S manufactured by Hot Disk AB (Goteborg, Sweden). The grease material was filled into two cups, and a flat sensor was held between the cups. Analysis conditions were: fine-tuned analysis, temperature drift compensation and time correction, and calculation points selected between 50 and 150.

[0024] Printing Aptitude Test The printability of the sample was measured by printing a 25 cm x 25 cm pattern with a thickness of 200 μm using a 60-mesh metal screen. The screen was held above a heat sink, and 5 g of sample was applied onto the screen. The squeegee was held at a 45° angle and a constant force and a constant drawing speed were applied to transfer the sample through the screen onto the heat sink. The printability was considered good if the grease could be deposited on the heat sink.

[0025] Pump-out resistance MSI GeForce RTX 2070 TRI FROZR Graphics Card (Die dimensions: 24.0mm x 18.5 = 445mm) 2The sample was applied to the heat sink of a graphics card (GPU) using a stencil. The graphics card was then assembled and installed in a computer. The following computer components were used for testing: AMD Ryzen 7 2700X 8-Core CPU, ASUS TUF X470-PLIS GAMING motherboard, KINSTON DDR4 266 8GB memory, MSI GeForce RTX 2080 TRI FROZR graphics card (8GB GDDR6, Nvidia Turing chip architecture), Intel SSD 760P series (256GB, M.2 80mm PCIe 3.0 x 4, 3D2, TLC), Dell U2417H monitor, Dell keyboard, Dell mouse, Antec P8 ATX case, Antec NEO750W power supply, and MT-viki HK05 KVM.

[0026] Thermal cycling tests were performed by running FurMark GPU stress testing software developed by Geeks3D and available at https: / / geeks3d.com / furmark / . The script (in AutoIt) included steps to turn the Furmark software on and off and to vary the fan speed to control the GPU card temperature. Fan speed was controlled using MSI Afterburner, available at Afterburner (msi.com).

[0027] The AutoIt script included the following steps: (1) opened the Furmark program, (2) launched the Furmark stress test routine, (3) adjusted the fan speed to 30% of its maximum speed for 140,000 milliseconds for a heating cycle, (4) stopped the stress test routine, (5) turned off the Furmark program, (6) adjusted the fan speed to 90% of its maximum speed for 180,000 milliseconds for a cooling cycle, and (7) repeated this sequence. The procedure was used to cycle the temperature on the GPU card from 35°C to 85°C and back down to 35°C. After 5000 cycles, the computer was shut down and the graphics card was removed.

[0028] The graphics card was opened and images of the heat sink and die were recorded. The areas on the heat sink and die where sample material was absent due to pump-out during the cycle test (bare spots) were measured. The quantitative area of ​​the bare spots was calculated using readily available software "sketchandclac." The total bare spot area due to grease pump-out was calculated as the total CPU die area (495 mm 2 ) to determine the % area of ​​bare spots on the die. The results were classified as follows: Best = area of ​​bare spots on the die is less than 5%, Fair = area of ​​bare spots on the die is 5-15%, and Poor = area of ​​bare spots on the die is more than 15%.

[0029] Table 3 summarizes the thermal conductivity in W / m·K (TC), the complex viscosity at the dilatant point in Pa·s (viscosity), and the pump-out resistance measured by bare spot area. Printability was found to be good for all samples.

[0030] [Table 1]

[0031] The data demonstrate a significant positive impact of boron nitride on pump-out resistance. Surprisingly, the improved pump-out resistance did not correlate with an increase in viscosity.

Claims

1. 1. A non-hardening thermal grease composition comprising, based on the weight of the composition, a) 80 to 95 weight percent of one or more filler particles selected from the group consisting of aluminum, alumina, and zinc oxide; b) 0.2 to 8 weight percent of platelet boron nitride particles; and c) 0.2 to 10 weight percent of Structure 1: 【Chemistry 1】 wherein R and R 1 are each independently C 1 ~C 6 -alkyl and Y is O or CH 2 -CH 2 and x is 30 to 200). Optionally, d) a non-curing thermal grease composition comprising up to 10 weight percent of a polydimethylsiloxane or poly(phenylmethylsiloxane-dimethylsiloxane) carrier fluid having a viscosity in the range of 30 to 200 cSt.

2. e) 0.05 to 1 weight percent C 1 ~C 20 -Alkyltri-C 1 ~C 6 The composition of claim 1, further comprising an alkoxysilane.

3. The composition comprises 85 to 95 weight percent aluminum and zinc oxide particles and / or alumina particles, based on the weight of the composition, wherein the aluminum particles are a bimodal distribution of larger spherical aluminum particles and smaller spherical aluminum particles, and the D of the larger spherical aluminum particles is 0.05 to 0.

05. 50 The volume average particle size is in the range of 5 μm to 15 μm, and the D of the larger spherical aluminum particles 99 The volume average particle size is 30 μm or less, and the D of the smaller spherical aluminum particles 50 The volume average particle size is in the range of 1 μm to 5 μm, and the D of the zinc oxide particles and / or the alumina particles is 50 2. The composition of claim 1, wherein the volume average particle size is in the range of 50 nm to 200 nm.

4. 4. The composition of claim 3, wherein the concentration of the larger spherical aluminum particles is in the range of 40 to 60 weight percent, the concentration of the smaller spherical aluminum particles is in the range of 15 to 35 weight percent, and the concentration of the zinc oxide and / or alumina particles is in the range of 10 to 25 weight percent, based on the weight of the composition.

5. The concentration of the boron nitride platelet particles is in the range of 0.5 to 6 weight percent based on the weight of the composition, and the D of the boron nitride platelet particles 50 10. The composition of claim 1, wherein the particle size ranges from 3 μm to 40 μm and the thickness of the boron nitride platelet particles ranges from 750 nm to 5 μm.

6. Said C 1 ~C 20 -Alkyltri-C 1 ~C 6 -Alkoxysilane is C 6 ~C 16 - alkyltrimethoxysilane, 6 ~C 16 6. The composition of claim 5, wherein the concentration of alkyltrimethoxysilane ranges from 0.1 to 2 weight percent based on the weight of the composition, and the boron nitride platelet particles have a diameter:thickness aspect ratio ranging from 3:1 to 30:

1.

7. Said C 6 ~C 16 -Alkyltrimethoxysilane is C 8 ~C 12 - alkyltrimethoxysilane, and the polydimethylsiloxane of structure 1 has the following structure: 【Chemistry 2】 (wherein each R and R 1 is methyl, x is 50 to 180, and Y is O), 7. The composition of claim 6, wherein the concentration of the polydimethylsiloxane of Structure 1 ranges from 1 to 10 weight percent based on the weight of the composition, and the boron nitride platelet particles have a diameter:thickness aspect ratio in the range of 3:1 to 20:

1.

8. 8. The composition of claim 7, comprising 1 to 10 weight percent of a polydimethylsiloxane or poly(phenylmethylsiloxane-co-dimethylsiloxane) carrier fluid, based on the weight of the non-curing thermal grease, wherein the viscosity of the carrier fluid ranges from 30 cSt to 150 cSt.

9. 9. The composition of claim 8, wherein the concentration of the boron nitride platelet particles ranges from 1 to 5 weight percent based on the weight of the composition.

10. the concentration of the boron nitride platelet particles is in the range of 2 to 4 weight percent based on the weight of the composition; 8 ~C 12 8. The composition of claim 7, wherein the alkyltrimethoxysilane is n-decyltrimethoxysilane, x is 80 to 130, and the boron nitride platelet particles have a diameter:thickness aspect ratio in the range of 4:1 to 10:

1.

11. 90 to 95 weight percent aluminum and zinc oxide particles based on the weight of the non-curable composition, and D of the larger spherical aluminum particles 50 The volume average particle size is in the range of 7 μm to 25 μm, and the D of the zinc oxide particles 50 The volume average particle size is in the range of 100 nm to 200 nm, 8. The composition of claim 7, wherein the concentration of the larger spherical aluminum particles is in the range of 45 to 55 weight percent, the concentration of the smaller spherical aluminum particles is in the range of 20 to 30 weight percent, and the concentration of the zinc oxide particles is in the range of 15 to 20 weight percent, based on the weight of the composition.

12. 9. The composition of claim 8, comprising 2 to 8 weight percent of said polydimethylsiloxane or phenylmethylsiloxane-co-dimethylsiloxane carrier fluid, based on the weight of the composition, wherein the viscosity of said carrier fluid ranges from 30 cSt to 120 cSt.

13. A microelectronic device comprising the non-curing thermal grease composition of any one of claims 1 to 12 embedded therein as a thermal interface material.