Experimental study of the proposed super-thermal-conductor - BAs

At a Glance

Metadata	Details
Publication Date	2015-02-01
Journal	DSpace@MIT (Massachusetts Institute of Technology)
Authors	Bing Lv, Yucheng Lan, Xiqu Wang, Qian Zhang, Allan J. Jacobson
Analysis	Full AI Review Included

Technical Documentation & Analysis: Super-Thermal Conductor BAs

Executive Summary

Research Focus: Experimental synthesis and characterization of Boron Arsenide (BAs) single crystals, a proposed super-thermal conductor for advanced thermal management applications in microelectronics.
Key Achievement: Measured thermal conductivity ($\kappa$) of $\sim 200 \text{ Wm}^{-1} \text{ K}^{-1}$, which is high for non-carbon insulators (e.g., SiC), but an order of magnitude lower than the theoretically predicted value ($\sim 2000 \text{ Wm}^{-1} \text{ K}^{-1}$).
Methodology: Two-step synthesis involving solid-state reaction followed by Chemical Vapor Transport (CVT) using Iodine ($\text{I}_2$) as the transport agent to grow small, cubic zinc blende crystals ($\sim 300-500 \text{ µm}$).
Limiting Factors: The significant reduction in $\kappa$ is attributed to material imperfections, primarily $\sim 2.8%$ Arsenic (As) vacancies (acting as mass defects), crystal twinning, and grain boundaries, which severely increase phonon scattering.
6CCVD Value Proposition: While BAs shows promise, 6CCVD Single Crystal Diamond (SCD) already delivers thermal conductivity up to $4000 \text{ Wm}^{-1} \text{ K}^{-1}$ reliably, commercially, and in large formats (up to $125 \text{ mm}$), providing an immediate, high-performance solution for microelectronic thermal challenges.
Path Forward: Achieving the predicted BAs performance requires extreme purity and defect control. 6CCVD offers high-purity SCD materials and advanced polishing (Ra < 1 nm) that directly address the defect and surface quality limitations encountered in this BAs research.

Technical Specifications

Parameter	Value	Unit	Context
Predicted Thermal Conductivity ($\kappa$)	$\sim 2000$	$\text{Wm}^{-1} \text{ K}^{-1}$	Theoretical maximum for perfect BAs crystal
Measured Thermal Conductivity ($\kappa$)	$\sim 200$	$\text{Wm}^{-1} \text{ K}^{-1}$	Experimental value obtained via TDTR on $\sim 300 \text{ µm}$ sample
Lattice Parameter ($a$)	$4.7830(7)$	Å	Refined from X-ray diffraction (Zinc Blende structure)
B:As Ratio (XPS)	$50.7:49.3$	N/A	Indicates $\sim 2.8%$ As-deficiency (p-type doping)
Sample Size (Grown Crystals)	$300 - 500$	$\text{µm}$	Typical size of as-grown single crystals
BAs Decomposition Temperature	$\sim 920$	$\text{°C}$	Irreversible decomposition to $\text{B}_{12}\text{As}_2$
Source Zone Temperature (CVT)	$900$	$\text{°C}$	Used to prevent irreversible decomposition
Growth Zone Temperature (CVT)	$\sim 650$	$\text{°C}$	Temperature for BAs crystal nucleation and growth

Key Methodologies

The BAs single crystals were synthesized using a two-step procedure designed to manage the high volatility and decomposition issues of arsenic and boron arsenide.

Polycrystalline Powder Precursor Synthesis (Solid State Reaction)
- Reactants: Pure As ($99.999%$) and B ($99.99%$) mixed at a B:As ratio of $1:1.8$.
- Container: Sealed quartz ampoules under vacuum.
- Heating Profile: Slow ramp to $500 \text{°C}$ ($10 \text{ h}$), followed by reaction at $800 \text{°C}$ for $3$ days.
- Refinement: Repeated grinding and reheating cycles were necessary to achieve a homogeneous, close-to-stoichiometric powder.
Single Crystal Growth (Chemical Vapor Transport - CVT)
- Assembly: BAs powder precursor, excess As, and Iodine ($\text{I}_2$) transport agent were sealed in a fused silica tube ($25 \text{ cm}$ length).
- Temperature Gradient: Tube placed in a two-temperature-zone furnace.
  - High Temperature Zone (Source): $900 \text{°C}$.
  - Low Temperature Zone (Growth): $\sim 650 \text{°C}$.
- Duration: $2-3$ weeks.
- Post-Growth Cleaning: Crystals were cleaned using moderately concentrated $\text{HCl}$ acid (to remove excess As) followed by washing with de-ionized water and ethanol.
Characterization Techniques
- Thermal Measurement: Time-Domain Thermoreflectance (TDTR) was used to measure $\kappa$.
- Structural Analysis: X-ray Powder Diffraction (XRD), X-ray Single Crystal Diffraction, and Convergent Beam Electron Diffraction (CBED) confirmed the cubic zinc blende structure and identified complex twinned structures.
- Chemical Analysis: X-ray Photoelectron Spectroscopy (XPS) determined the B:As ratio, confirming the presence of $\sim 2.8%$ As-deficiency.

6CCVD Solutions & Capabilities

The BAs research confirms that achieving super-thermal conductivity is fundamentally a materials purity and defect control problem. 6CCVD specializes in producing MPCVD diamond materials that inherently solve these challenges, offering immediate, reliable, and scalable solutions for high-power thermal management.

Applicable Materials: Superior Thermal Performance

The paper targets a $\kappa$ of $2000 \text{ Wm}^{-1} \text{ K}^{-1}$. 6CCVD’s SCD materials routinely meet or exceed this target, offering a proven alternative to experimental BAs.

Application Requirement	6CCVD Material Recommendation	Rationale & Performance
Highest Thermal Conductivity	Optical Grade SCD (Single Crystal Diamond)	SCD provides $\kappa$ up to $4000 \text{ Wm}^{-1} \text{ K}^{-1}$. This material is grown with extremely low nitrogen incorporation, minimizing point defects that scatter phonons (the primary limitation cited in the BAs paper).
Cost-Effective High $\kappa$ Substrates	High-Purity PCD (Polycrystalline Diamond)	Available in large formats (up to $125 \text{ mm}$ wafers). Offers excellent thermal performance ($\sim 1000-2000 \text{ Wm}^{-1} \text{ K}^{-1}$) suitable for large-area heat spreading in microelectronics.
Electrodes/Sensors	Heavy Boron Doped Diamond (BDD)	For applications requiring metallic conductivity combined with diamond’s chemical inertness, leveraging the same high-quality CVD growth platform.

Customization Potential

The BAs crystals were limited to $\sim 500 \text{ µm}$ fragments, necessitating complex handling and limiting device integration. 6CCVD offers the scalability and precision required for commercial and advanced research applications:

Large Dimensions: We provide SCD plates up to $10 \text{ mm}$ thick and PCD wafers up to $125 \text{ mm}$ in diameter, far exceeding the size limitations of the BAs synthesis.
Surface Quality: To ensure optimal thermal contact and minimize surface scattering (critical for TDTR and device bonding), 6CCVD offers industry-leading polishing:
- SCD: Surface roughness Ra < 1 nm.
- Inch-size PCD: Surface roughness Ra < 5 nm.
Device Integration (Metalization): We offer internal metalization services for direct integration into thermal stacks, including standard and custom layers of Ti, Pt, Au, Pd, W, and Cu.
Thickness Control: Precise control over material thickness, from ultra-thin films ($0.1 \text{ µm}$) to robust substrates (up to $10 \text{ mm}$).

Engineering Support

The BAs study clearly demonstrates that achieving high thermal conductivity is dependent on eliminating defects (vacancies, twins, and grain boundaries). 6CCVD’s in-house PhD team are experts in controlling the MPCVD environment to minimize these exact phonon scattering mechanisms in diamond.

We offer consultation services to assist researchers and engineers in selecting the optimal diamond grade (SCD or PCD) and specifications (purity, orientation, surface finish) required to maximize thermal performance for similar High Thermal Management projects, bypassing the synthesis difficulties associated with BAs.

Call to Action: For custom specifications or material consultation, visit 6ccvd.com or contact our engineering team directly.

View Original Abstract

Recent calculations predict a super-thermal-conductivity of ∼2000 Wm-1K-1, comparable to that of diamond, in cubic boron arsenide (BAs) crystals, which may offer inexpensive insulators with super-thermal-conductivity for microelectronic device applications. We have synthesized and characterized single crystals of BAs with a zinc blende cubic structure and lattice parameters of a = 4.7830(7) Å. A relatively high thermal conductivity of ∼200 Wm-1K-1is obtained, close to those of best non-carbon crystal insulators, such as SiC, although still an order of magnitude smaller than the value predicted. Based on our XPS, X-ray single crystal diffraction, and Raman scattering results, steps to achieve the predicted super-thermal conductivity in BAs are proposed.

Tech Support

Original Source

DOI: None