Simultaneous measurement of pressure-dependent bulk and interfacial thermal properties in thermal interface materials using square-pulsed source thermoreflectance

This paper introduces a square-pulsed source thermoreflectance method that simultaneously quantifies the pressure-dependent bulk thermal properties and interfacial thermal resistance of thermal interface materials, revealing distinct transport mechanisms and hysteresis behaviors across different material types under mechanical loading.

The Gist

Imagine you are trying to keep a high-performance computer chip cool. The chip generates a lot of heat, and you need to move that heat away to a metal cooler (a heat sink). But here's the problem: even if two surfaces look smooth to the naked eye, under a microscope, they are actually like mountain ranges full of tiny peaks and valleys. When you press them together, they only touch at the very tips of those peaks, leaving tiny air pockets in between. Since air is a terrible conductor of heat, these gaps act like a traffic jam, causing the chip to overheat.

To fix this, engineers use Thermal Interface Materials (TIMs)—think of them as "thermal glue" or "heat paste"—to fill those gaps and create a smooth highway for the heat to travel.

This paper introduces a new, super-smart way to measure exactly how well these "thermal glues" work, especially when they are being squeezed tight inside a computer.

The Problem: The "Black Box" of Heat

Traditionally, when engineers test these materials, they treat them like simple, unchanging blocks. They assume the material's ability to conduct heat (like a highway's speed limit) and its ability to store heat (like a sponge holding water) are fixed numbers.

But in reality, these materials are squishy. When you screw a computer case down, you are squeezing the glue.

• The Squeeze Effect: Squeezing changes the material's internal structure. It might get denser, or the tiny air bubbles inside might get crushed.
• The History Effect: If you squeeze it, let go, and squeeze it again, it doesn't always go back to exactly the same shape. It has a "memory" of being squished.

Old testing methods couldn't measure all the important factors at once while the material was being squeezed. They were like trying to measure a car's speed, fuel tank size, and engine friction all at once while the car was driving over a bumpy road, but only using a single snapshot.

The Solution: The "Square-Pulsed" Flashlight

The researchers developed a new technique called Square-Pulsed Source (SPS) Thermoreflectance.

The Analogy: Imagine you are trying to figure out the properties of a thick, fuzzy blanket by shining a flashlight on it.

• Old Way: You shine a steady light. You can only tell how warm the surface gets, but you can't tell if the heat is getting stuck at the surface or moving deep inside.
• The New Way (SPS): You use a flashlight that blinks on and off incredibly fast, but you can control the speed of the blinking.
  • Fast Blinking (High Frequency): The light only warms the very top layer of the blanket. This tells you how well the surface touches the wall (the Interface).
  • Slow Blinking (Low Frequency): The heat has time to soak deep into the blanket. This tells you how the material stores and moves heat inside itself (the Bulk).

By flashing the light at many different speeds (from very slow to super fast) all at once, the researchers can mathematically "unscramble" the signal to see three things simultaneously:

1. Thermal Conductivity ($k$): How fast heat moves through the material.
2. Heat Capacity ($C$): How much heat the material can "soak up" like a sponge.
3. Interfacial Resistance (ITR): How much the heat gets stuck at the boundary where the glue meets the metal.

What They Found: Three Different Personalities

The team tested three common types of thermal "glues" under pressure: a Gel, a Pad, and a Grease. They found that each behaves like a different character in a story.

1. The Gel: The "Squishy Sponge"

• Behavior: When you squeeze the gel, it gets denser. The tiny air bubbles inside get crushed.
• Result: Because it's denser, it conducts heat better and can hold more heat.
• The Twist (Hysteresis): If you squeeze it and then let go, it doesn't bounce back perfectly. It stays a little squished. So, if you measure it again at the same pressure, it performs better the second time because it "remembered" being squeezed. It's like a memory foam pillow that doesn't fully return to its original shape.

2. The Pad: The "Compressed Composite"

• Behavior: This is a solid sheet filled with heat-conducting particles. When squeezed, the particles get pushed closer together, creating better paths for heat to travel.
• Result: Like the gel, it gets better at conducting and storing heat when squeezed.
• The Twist: It also has "memory." Once compressed, it stays in a denser state even after you release the pressure, leading to better performance on the second try.

3. The Grease: The "Wetting Agent"

• Behavior: This is the odd one out. The grease itself (the liquid part) doesn't really change when you squeeze it. Its internal structure stays the same.
• Result: The grease's ability to conduct heat inside itself stays constant.
• The Twist: The magic happens at the surface. When you squeeze the grease, it gets forced into the tiny nooks and crannies of the metal surface, filling the gaps perfectly. It's like spreading butter on toast; the butter doesn't change, but the contact between the butter and the toast gets much better. The improvement comes entirely from the interface, not the bulk.

Why This Matters

This research is a big deal for computer engineers and designers.

For years, they have assumed that thermal materials are static. They would pick a glue, look up a number in a catalog, and design their cooling system around that. But this paper proves that the material changes depending on how hard you squeeze it and how you got there.

• If you use a Gel or Pad, you need to account for the fact that it gets better over time as it settles and stays denser.
• If you use Grease, you need to focus on making sure it spreads perfectly to fill the gaps, because the grease itself won't change much.

By using this new "multi-speed flashlight" method, engineers can now design computers that are more reliable and run cooler, because they finally understand the true, dynamic behavior of the materials holding their chips together. It's the difference between guessing how a car handles a turn and actually having a sensor that tells you exactly how the tires, suspension, and road are interacting in real-time.

Technical Summary

1. Problem Statement

Thermal Interface Materials (TIMs) are critical for managing heat dissipation in electronic packaging, yet their thermal performance is complex and often misunderstood under real-world assembly conditions.

• The Gap: Existing methods (e.g., steady-state ASTM D5470 or transient Laser Flash Analysis) typically fail to simultaneously determine three key parameters: thermal conductivity ($k$), volumetric heat capacity ($C$), and interfacial thermal resistance (ITR).
• The Complication: TIM properties are not intrinsic constants; they evolve with mechanical pressure and loading history (compression/unloading cycles).
  • Soft TIMs (gels, pads, greases) undergo structural changes (densification, void collapse, wetting) that alter $k$ and $C$.
  • The contact state at the interface changes, altering ITR.
  • Current techniques often assume $C$ is constant or cannot decouple bulk properties from interfacial resistance, leading to inaccurate thermal modeling in electronic systems.

2. Methodology: Square-Pulsed Source (SPS) Thermoreflectance

The authors developed a novel experimental framework combining Square-Pulsed Source (SPS) thermoreflectance with a pressure-controlled mechanical loading stage.

• Principle: The method uses a pump-probe laser setup where the pump beam is modulated by a 50% duty-cycle square wave.
• Frequency Range: A key innovation is the exceptionally broad modulation frequency range (1 Hz to 10 MHz).
  • Low Frequencies (1 Hz – 100 kHz): Probe deep thermal penetration depths, making the signal sensitive to bulk heat storage ($C$) and diffusion ($k$).
  • High Frequencies (100 kHz – 10 MHz): Probe shallow depths, making the signal highly sensitive to interfacial thermal resistance (ITR).
• Experimental Setup:
  • A transparent substrate (glass) coated with a thin Aluminum (Al) transducer layer is pressed against the TIM sample.
  • A spring-loaded fixture applies controlled mechanical pressure (loading and unloading cycles).
  • The transient thermoreflectance signal is recorded and fitted against a heat-transfer model to extract $k$, $C$, and ITR simultaneously.
• Materials Tested: Three representative TIM classes were evaluated:
  1. Thermally conductive gel (RS382).
  2. Thermal pad (Laird HD90000).
  3. High-vacuum grease (Dow Corning).

3. Key Contributions

• Simultaneous Decoupling: The study demonstrates the first method to simultaneously and independently resolve $k$, $C$, and ITR in a single transient experiment under varying mechanical loads.
• Path-Dependent Characterization: It introduces a framework to track the hysteresis (history dependence) of thermal properties during compression and unloading cycles, revealing that thermal states are governed by both instantaneous pressure and prior mechanical history.
• Mechanistic Differentiation: The work distinguishes between two distinct physical mechanisms of pressure response: bulk-state evolution (changes in material density/porosity) vs. interface-dominated evolution (changes in wetting/contact area).

4. Key Results

The study revealed distinct pressure-dependent behaviors for the three TIM types:

A. Thermally Conductive Gel & Thermal Pad

• Behavior: Both exhibited strong pressure dependence in all three parameters ($k$, $C$, and ITR).
  • Loading: As pressure increased, $k$ and $C$ increased significantly (due to bulk densification and void collapse), while ITR decreased (due to improved contact).
  • Hysteresis: A pronounced hysteresis loop was observed. At the same nominal pressure, the unloading branch showed higher $k$, higher $C$, and lower ITR compared to the loading branch.
• Physical Interpretation:
  • Gel: Modeled as a compressible porous medium. Compression collapses entrapped gas pockets (reducing porosity), increasing the effective density of the solid matrix. The hysteresis is attributed to incomplete structural recovery (residual porosity) and irreversible viscoelastic deformation.
  • Pad: Modeled as a compliant filled composite. Compression reduces specific volume, increasing $C$ and improving internal conductive pathways ($k$). Hysteresis arises from permanent compression set.

B. High-Vacuum Grease

• Behavior: Exhibited a fundamentally different response.
  • Bulk Properties: Thermal conductivity ($k \approx 1.5$ W/m·K) and volumetric heat capacity ($C \approx 0.65-0.70$ MJ/m³·K) remained nearly constant across the pressure range.
  • Interfacial Properties: ITR decreased significantly with pressure (from 0.08 to 0.03 mm²·K/W).
• Physical Interpretation: The grease acts as an invariant bulk medium under these conditions. The thermal improvement is driven almost exclusively by interface evolution—better wetting and gap filling at the Al/TIM interface—rather than changes in the bulk material structure.

5. Significance and Impact

• Design Accuracy: The findings challenge the industry practice of treating TIM properties as fixed constants. Accurate thermal management in electronic packaging requires accounting for the state-dependent nature of $k$, $C$, and ITR, especially under varying assembly pressures and stress relaxation.
• Reliability Analysis: Understanding the hysteresis (history dependence) is crucial for predicting transient thermal impedance during power cycling, rework, or thermal shock, where the TIM may not return to its initial state.
• Optimization Strategy:
  • For gels and pads, optimization must address both bulk densification and interface conformity.
  • For greases, optimization should focus primarily on interfacial wetting and spreading, as bulk properties are less sensitive to pressure.
• Methodological Advance: The SPS technique provides a robust, non-contact tool for characterizing complex, multi-physics thermal systems, bridging the gap between material science and practical electronic packaging design.