High‑Temperature Stability of ITO/PtRh Thin‑Film Thermocouples for Aeroengine Applications

Abstract

Thin‑film thermocouples (TFTCs) enable accurate, non‑intrusive temperature monitoring in aerospace propulsion systems. We report an ITO/PtRh:PtRh TFTC, fabricated on alumina ceramic via magnetron sputtering and post‑annealed in a N₂‑air cycle. The device was calibrated through multiple cycles up to 1,000 °C, demonstrating exceptional EMF stability and repeatability. Oxygen diffusion barriers formed by PtRh oxidation, coupled with Schottky barriers at ITO grain boundaries, suppress carrier‑concentration drift. The TFTC maintains operation for >30 h in harsh environments, positioning it as a reliable sensor for aeroengine hot‑component temperature measurement.

Background

Accurate temperature data are essential for validating aeroengine thermo‑mechanical models, monitoring operating conditions, and conducting diagnostics [1–3]. Compared with conventional wire thermocouples, infrared cameras, or thermal spray gauges, TFTCs offer rapid response, minimal gas‑flow disturbance, and negligible influence on the measured surface temperature distribution [4,5].

High‑temperature TFTCs have been constructed from Pt‑PtRh and In₂O₃‑ITO systems [6–9]. However, these thin‑film architectures often suffer from stability and repeatability issues at the 800–1,000 °C range typical of aeroengines. Rhodium oxidation causes drift in Pt‑PtRh devices [10,11], while oxygen‑vacancy dynamics lead to output drift or failure in In₂O₃‑based TFTCs [12,13]. Various strategies—high‑temperature annealing, nitrogen doping [14–16], or nanocomposite approaches [8]—have improved ITO performance, yet the thermoelectric output typically declines due to oxygen diffusion. A semiconductor/metal multilayer thermoelement had not yet been reported.

In this work we introduce an ITO/PtRh composite multilayer as the thermoelement. After sputter deposition and N₂‑air annealing, we investigate the microstructure, resistivity, and high‑temperature thermoelectric response of the resulting ITO/PtRh:PtRh TFTC.

Methods

Sample Preparation

ITO (In₂O₃:SnO₂ = 90:10, 99.99 wt%, 100 mm diameter) and Pt‑13 %Rh (99.99 wt%, 100 mm diameter) targets were sputtered onto alumina and Si(100) substrates in a 7 × 10⁻⁴ Pa chamber at 110 mm target‑substrate distance. Substrates were sequentially cleaned with acetone, ethanol, and deionized water. Alternating deposition of ITO and Pt‑Rh layers produced a ~1 µm thick composite film, with the ITO layer ≈400 nm and the Pt‑Rh layer ≈100 nm. Post‑annealing involved 5 h at 1,000 °C in N₂ followed by 2 h at 1,000 °C in air (N₂‑Air) [15].

ITO/PtRh:PtRh TFTC Fabrication

The 63 mm × 1 mm × 1 µm TFTC was patterned on a 75 mm × 12 mm × 0.5 mm alumina substrate using stenciled masks. After N₂‑Air annealing, the device was calibrated in a furnace from 300 to 1,000 °C. Each setpoint was held ≥1 h to reach thermal equilibrium. Standard S and K type wire thermocouples measured hot (T₁) and cold (T₂) junction temperatures; a digital multimeter recorded the EMF from the cold junction lead.

Characterization

X‑ray diffraction (XRD) examined the ITO crystallinity; scanning electron microscopy (SEM) revealed the cross‑section of the composite film; four‑point probe measured film resistivity.

Calibration Procedure

The calibration setup is illustrated in Figure 1b: a circulating water bath creates a temperature gradient between the hot and cold junctions. EMF is recorded while T₁ and T₂ are monitored with reference thermocouples.

Results and Discussion

Microstructure and Resistivity

XRD of the annealed ITO film shows polycrystalline cubic bixbyite In₂O₃ peaks without preferred orientation; no Sn or Sn‑oxide peaks appear, confirming solid‑solution formation [17].

SEM cross‑sections (Figure 3) confirm the intended ~1 µm total thickness, with the ITO layer roughly four times the Pt‑Rh thickness.

Resistivity measurements (Table 2) reveal that the as‑deposited ITO/PtRh composite is an order of magnitude more conductive than pure ITO. Post‑annealing reduces ITO resistivity slightly (8.52 × 10⁻² Ω cm → 7.55 × 10⁻² Ω cm), likely due to densification and defect reduction. Conversely, the composite’s resistivity rises (1.68 × 10⁻³ Ω cm → 7.61 × 10⁻³ Ω cm) because surface Rh oxidation during annealing impedes charge transport [18].

Thermoelectric Performance

Static calibration (Figure 4a) shows a nonlinear increase of EMF with temperature difference, yet EMF values remain virtually unchanged across multiple cycles up to 1,000 °C, indicating superior stability.

Figure 4b displays Seebeck coefficients that rise rapidly with ΔT, reflecting the temperature‑dependent carrier concentration in degenerate ITO. The average Seebeck coefficient over three cycles is 2.19 µV/°C, slightly lower than conventional S‑ or R‑type thermocouples. This reduction is attributed to Schottky barriers at ITO grain boundaries, which limit carrier concentration variations and stabilize the device [21–23].

The EMF‑ΔT relationship fits a third‑order polynomial (Equation 3), with coefficients nearly identical across cycles (Table 3). The fitted curves confirm the device’s repeatability. The TFTC remains functional for >30 h at 1,000 °C, underscoring its suitability for aeroengine surface temperature monitoring.

Conclusions

We successfully deposited ITO and ITO/PtRh composite films on alumina via room‑temperature magnetron sputtering and achieved a stable, multilayer ITO/PtRh:PtRh TFTC after N₂‑air annealing. The composite’s resistivity increase upon annealing is attributed to Rh oxidation, while the ITO resistivity slightly improves due to densification. The device’s high‑temperature stability stems from oxygen diffusion barriers formed by PtRh oxidation and Schottky barriers at ITO grain boundaries, which suppress carrier‑concentration drift. With an average Seebeck coefficient of 2.19 µV/°C and a lifespan exceeding 30 h under harsh conditions, this TFTC is a promising sensor for precise aeroengine component temperature measurement. The multilayer strategy offers a new avenue to enhance ITO stability without relying solely on high‑temperature annealing or nitrogen doping.

Abbreviations

EMF

Electromotive force

S

Seebeck coefficient

SEM

Scanning electron microscopy

TFTC

Thin film thermocouple

XRD

X‑ray diffraction