The thermal expansion coefficient of PC silicon carbide (SiC) single crystals is a critical parameter governing wafer processing, epitaxial growth, device packaging, and high-temperature structural performance. This article provides a comprehensive technical overview of PC SiC single crystal thermal expansion coefficient determination, covering sample preparation, measurable quantities, standard test methods such as ASTM E228, ASTM E831, and high-temperature X-ray diffraction, together with typical applications and practical guidance. Use this guide to select the appropriate dilatometry technique and to interpret anisotropic expansion data with confidence.
Technical Overview
PC silicon carbide (PC SiC) single crystals are high-purity, semiconductor-grade crystals valued for their exceptional thermal conductivity, wide bandgap, high hardness, and outstanding dimensional stability at elevated temperatures. Because most PC SiC crystals belong to the hexagonal polytypes (4H or 6H), their thermal expansion behavior is anisotropic: the linear thermal expansion coefficient measured along the crystallographic c-axis [0001] differs measurably from that measured along the a-axis [11-20] or [10-10].
The linear thermal expansion coefficient (CTE) is defined as the fractional change in length per unit change in temperature. The mean (secant) coefficient between two temperatures T1 and T2 is expressed as αm = (L2 − L1) / (L0 × (T2 − T1)), where L0 is the reference length at the starting temperature, while the instantaneous (tangent) coefficient is α(T) = (1/L0) × (dL/dT). For hexagonal SiC single crystals, typical mean values between room temperature and 1000 °C fall in the range of approximately 4.0 × 10⁻⁶ K⁻¹ to 4.8 × 10⁻⁶ K⁻¹, with the c-axis coefficient generally a few percent higher than the a-axis coefficient.
Accurate CTE determination is essential for numerous engineering decisions. In epitaxial growth, thermal expansion mismatch between SiC substrates and GaN or SiC epilayers generates stress during heating and cooling cycles. In device packaging, the CTE difference between SiC dies and copper, ceramic, or solder materials controls thermomechanical fatigue life. In wafer manufacturing, CTE data are required to predict warpage and stress during high-temperature processes such as ion implantation annealing. A reliable determination therefore demands controlled orientation, precise dilatometry, traceable calibration, and protection against high-temperature oxidation.
Testing Samples
Standard Dilatometer Specimens
For push-rod dilatometry, PC SiC single crystal samples are prepared as rectangular bars or cylinders, typically 20 mm to 25 mm in length with a cross-section of 4 mm to 8 mm. The two end faces are ground flat and parallel to within 5 µm to 10 µm, and the side faces are ground to a surface roughness of Ra ≤ 1.6 µm. Specimen dimensions are measured at 23 °C ± 2 °C using calibrated micrometers or length metrology instruments before testing.
Orientation-Controlled Specimen Pairs
To capture the anisotropy of the hexagonal structure, specimens are fabricated in matched pairs: one with its longitudinal axis parallel to the c-axis [0001] and one parallel to the a-axis [11-20]. Crystallographic orientation is verified by X-ray diffraction or Laue back-reflection prior to machining, with a permitted deviation of no more than ±1 degree, because misorientation directly biases the measured anisotropy ratio.
Specimens for TMA and High-Temperature XRD
Thermomechanical analysis (TMA) accepts smaller specimens, typically 2 mm to 10 mm in length, including sections cut from wafers. For high-temperature X-ray diffraction, flat wafer pieces of approximately 10 mm × 10 mm with a thickness of 0.3 mm to 1 mm are suitable, as the technique measures lattice parameter changes rather than bulk length changes.
Preparation and Pre-Treatment
All specimens are cleaned with solvents, dried, and, where machining-induced residual stress is suspected, subjected to a mild stress-relief anneal below the temperature at which surface degradation occurs. Each sample is documented with polytype identification (4H or 6H), doping type and level, growth method, and orientation certificate so that results can be traced to a specific crystal lot.
Testing Items
The thermal expansion coefficient determination for PC SiC single crystals typically includes the following deliverables:
- Mean (secant) linear CTE over specified temperature intervals, for example 25–200 °C, 25–500 °C, 25–1000 °C, and 25–1500 °C.
- Instantaneous (tangent) CTE as a continuous function of temperature, reported as an α(T) curve.
- Thermal strain curves (ΔL/L0 versus temperature) for the c-axis and a-axis orientations.
- Anisotropy ratio (αc / αa) as a function of temperature.
- Identification of nonlinear expansion regions and transition behavior at elevated temperatures.
- Measurement uncertainty budgets, repeatability checks, and calibration traceability records.
Testing Methods
Push-Rod Dilatometry (ASTM E228)
Push-rod dilatometry is the primary method for bulk PC SiC single crystals. The specimen is placed in an alumina or fused-silica holder with a low, constant contact force applied through a push rod, and length changes are detected by a linear variable differential transformer (LVDT). Testing is conducted in high-purity argon or vacuum at heating rates of 3 K/min to 5 K/min, covering temperatures from room temperature up to 1500 °C or 1600 °C. The system is calibrated against certified reference materials, such as sapphire or fused silica, and the instrument expansion is subtracted from the raw signal. With proper calibration, repeatability better than ±0.1 × 10⁻⁶ K⁻¹ can be achieved.
Thermomechanical Analysis (ASTM E831 / ISO 11359-2)
TMA applies a small static force (typically 0.02 N to 0.2 N) to a small specimen and records dimensional change during a controlled temperature ramp, usually from room temperature to 1000 °C. The method is well suited to wafer sections, rapid material screening, and quality control comparisons between crystal lots, with typical CTE precision of about 5 percent for samples a few millimeters long.
High-Temperature X-Ray Diffraction
High-temperature XRD measures the lattice parameters a(T) and c(T) of the single crystal directly in a heated stage under vacuum or inert gas. The crystallographic expansion coefficients are calculated from the relative changes in lattice spacing along each axis. This method provides true single-crystal anisotropic coefficients, confirms the polytype simultaneously, and eliminates uncertainties associated with macroscopic specimen geometry, making it an excellent cross-check for dilatometry results.
Laser Interferometric Dilatometry (ASTM E289)
For reference-grade or certification-level measurements, Fizeau or laser interferometric dilatometry measures displacement optically with sub-nanometer resolution, achieving uncertainties on the order of 0.01 × 10⁻⁶ K⁻¹ to 0.05 × 10⁻⁶ K⁻¹. This technique is used when PC SiC crystals serve as reference materials or when very small differences between crystal lots must be resolved.
Data Reduction and Reporting
All raw displacement data are corrected for the expansion of the instrument using reference material runs performed under identical conditions. Mean CTE values are computed over agreed temperature intervals, tangent coefficients are obtained by differentiating smoothed strain curves, and results are reported with the specimen orientation, polytype, atmosphere, heating rate, calibration reference, and expanded uncertainty. Round-robin verification between dilatometry and XRD is recommended for critical programs.
Applications
Thermal expansion coefficient data for PC SiC single crystals support a wide range of industrial and scientific uses:
- Power semiconductor wafer manufacturing, where CTE data predict wafer warpage and stress during high-temperature annealing and oxidation processes.
- Epitaxial growth of SiC and GaN layers, enabling calculation of thermal mismatch stress and bow control in heteroepitaxy.
- Device packaging and die-attach engineering, including solder joint fatigue analysis and selection of substrates and heat spreaders with matched expansion behavior.
- Brazing and ceramic-to-metal joining, where joint design depends on accurate CTE values across the service temperature range.
- Optical instrumentation and space-borne mirror substrates, where dimensional stability under thermal cycling is a primary design requirement.
- Aerospace and high-temperature structural components, including thermal management modules and wear-resistant parts.
- Crystal growth quality control, where CTE deviations can indicate polytype mixing, doping anomalies, or residual stress in as-grown boules.
FAQ
What is the typical thermal expansion coefficient of PC SiC single crystal?
The mean linear CTE of hexagonal SiC single crystals is typically between 4.0 × 10⁻⁶ K⁻¹ and 4.8 × 10⁻⁶ K⁻¹ over the range from room temperature to 1000 °C, increasing gradually with temperature. The c-axis value is generally slightly higher than the a-axis value, and exact results depend on polytype, doping, and the specified temperature interval.
Why must the measurement be performed in an inert atmosphere?
Silicon carbide begins to oxidize in air at elevated temperatures, forming a SiO₂ surface scale that changes the apparent specimen dimensions and distorts the measured CTE. High-purity argon or vacuum prevents oxidation and ensures that the recorded expansion reflects the crystal itself.
Which standards apply to the determination?
Push-rod dilatometry is performed per ASTM E228, thermomechanical analysis per ASTM E831 or ISO 11359-2, and interferometric dilatometry per ASTM E289. High-temperature XRD lattice parameter measurements follow generally accepted crystallographic procedures with instrument calibration standards.
Why are measurements made along both the a-axis and c-axis?
Hexagonal 4H and 6H SiC single crystals are thermally anisotropic, so a single coefficient cannot describe expansion in all directions. Measuring oriented specimen pairs along [0001] and perpendicular to it provides the two principal coefficients needed for stress modeling and anisotropy assessment.
What sample dimensions and orientations are required?
For dilatometry, bars of 20 mm to 25 mm in length with ends flat and parallel within about 10 µm are preferred; TMA accepts samples from 2 mm to 10 mm long. Orientation must be verified by XRD or Laue method within approximately ±1 degree of the intended crystallographic axis.
What temperature range and accuracy can be achieved?
Measurements typically cover room temperature to 1500 °C, extendable to about 1600 °C in argon with suitable hardware. Expanded uncertainties are usually within ±3 percent relative or ±0.1 × 10⁻⁶ K⁻¹ for dilatometry, and interferometric methods can reach an order of magnitude better precision for certification work.
