Vacuum-compatible infrared emissivity characterization of high-temperature materials using direct electric heating

1 Introduction

Refractory metals and silicide-coated refractory materials are widely used in high-temperature aerospace components because of their high melting points, favorable mechanical properties, and resistance to severe thermal environments (Perepezko and Sakidja, 2010; Zhong et al., 2016; Sato et al., 2015). During service, these components may be subjected to high heat fluxes, repeated thermal cycling, and, depending on the application, low-pressure or vacuum conditions. Such environments can lead to material degradation and variations in surface radiative properties, ultimately affecting thermal management, temperature monitoring, and structural reliability (Hai et al., 2015; Wei and Zhang, 2010; Morris, 2005; Fergus and Gall, 2013). Accurate characterization of material emissivity under representative operating conditions is therefore essential for the thermal design, performance evaluation, and lifetime prediction of aerospace thermal protection and propulsion systems.

Various thermal-analysis instruments, including simultaneous thermal-analysis systems and automated thermogravimetric analyzers, have been developed for material characterization under different operating conditions (Matuschek and Utschick, 1997; Xiong et al., 2003; Mikkelsen et al., 2005; Devaraju and Suresha, 2011). These systems are useful for evaluating mass variation and thermal behavior, but they are not specifically designed for broadband spectral emissivity measurement under combined high-temperature and vacuum conditions.

Spectral emissivity can be determined using radiance-comparison, reflectance-based, calorimetric, and multiwavelength-pyrometry methods. In recent years, several advanced measurement facilities have been developed to extend the applicable temperature and wavelength ranges. Arduini et al. developed an induction-heated apparatus in a vacuum vessel for measuring spectral directional emissivity from 1 to 20 μm at temperatures up to approximately 2400 K (Arduini et al., 2021). Urban et al. conducted an interlaboratory comparison of pulse-heating, induction-heating, radiometric, calorimetric, and polarimetric methods for refractory materials over temperatures from 1,250 to 4000 K (Urban et al., 2024). Mironov et al. established an induction-heated system for conductive materials in the 1–18 μm range at 300 °C–1700 °C, incorporating correction of reflected surrounding radiation and multiwavelength pyrometry (Mironov et al., 2024). Liu et al. recently developed a normal spectral emissivity measurement device specifically for low-temperature and vacuum conditions (Liu et al., 2024). These studies demonstrate substantial progress in high-temperature emissivity metrology, particularly in temperature determination, background-radiation correction, and uncertainty evaluation.

Recent investigations have also emphasized that the emissivity of refractory metals is strongly dependent on wavelength, temperature, and surface condition. Zhang et al. measured the normal spectral emissivity of tungsten in the 3–5 μm range using an FTIR spectrometer and quantified its dependence on temperature and wavelength (Zhang et al., 2023). He et al. systematically investigated the combined effects of surface roughness and temperature on the emissivity of molybdenum (He et al., 2024). Shu et al. measured the normal spectral emissivity of tungsten between 170 °C and 500 °C using an infrared thermometer under conditions relevant to high-temperature component monitoring (Shu et al., 2021). More recently, Kayikci et al. proposed an infrared emissivity measurement approach for solid and opaque materials, further highlighting the importance of reliable radiometric comparison and background treatment (Kayikci et al., 2025). These material-specific studies confirm the necessity of measuring emissivity under controlled and application-relevant thermal conditions rather than relying solely on room-temperature or handbook data.

Despite these advances, simultaneously combining high-temperature operation, high-vacuum conditions, rapid thermal response, and broadband in-situ spectral measurement remains technically challenging. Some facilities achieve high-temperature emissivity measurements under vacuum conditions but rely on induction or pulse heating and separate reference or optical-switching arrangements (Arduini et al., 2021; Urban et al., 2024). Other systems operate in air (Mironov et al., 2024), focus on low-temperature emissivity measurements under vacuum conditions (Liu et al., 2024), or characterize individual materials over restricted temperature or spectral ranges (Zhang et al., 2023; He et al., 2024; Shu et al., 2021; Kayikci et al., 2025; Dai et al., 2005; Zhang and Dai, 2013). Although these configurations are effective for their intended applications, the use of separate heating elements, reference radiators, or optical switching mechanisms may increase structural complexity and background-radiation correction requirements (Arduini et al., 2021; Urban et al., 2024; Mironov et al., 2024). Therefore, there remains a need for a vacuum-compatible platform that directly and rapidly heats conductive refractory samples while enabling in-situ FTIR-based emissivity measurement over a wide temperature range.

To address this requirement, a vacuum-compatible infrared radiation measurement system for high-temperature materials is developed in this work. Two electric-heating concepts, namely directly clamped electric heating and conductive thermoelectric heating, are first investigated, and the directly clamped configuration is adopted in the final measurement device because of its rapid response and convenient optical access. The novelty of the proposed system lies in the integration of direct Joule heating, a sample-integrated cavity reference, and in-situ FTIR spectral measurement under high-vacuum conditions. In the directly clamped configuration, the electrically conductive sample itself serves as the heating element, avoiding the need for a separate external heater and shortening the heat-transfer path. Following the radiance-comparison principle (Arduini et al., 2021; Kayikci et al., 2025), the cavity fabricated in the sample serves as a near-blackbody reference (Yousef et al., 2000), allowing the radiation from the sample surface and the cavity to be measured under the same thermal and vacuum conditions. This configuration enables rapid heating and real-time broadband emissivity measurement within a single experimental platform. The performance of the system is evaluated using molybdenum, tungsten, and tantalum samples at 800, 1,000, and 1,200 °C, together with an uncertainty analysis. The developed system is designed to support heating from 500 °C to above 3,000 °C, whereas the emissivity experiments reported in this study were conducted within the allowable temperature range of the selected samples.

2 Methods

2.1 Electric heating methods

Based on the principles of heat transfer in a vacuum environment, two electric heating methods are employed to heat the sample: the directly clamped electric heating method and the conductive thermoelectric heating method. Correspondingly, two types of heating structures have been designed in this study. The directly clamped electric heating structure heats the sample directly by using the sample itself as a conductor. In contrast, the conductive thermoelectric heating structure relies on an intermediate material, such as graphite, to form a high-temperature isothermal zone, thereby enabling conductive heating of the sample.

2.1.1 Directly clamped electric heating method

The thermoelectric heating system based on the directly clamped electric heating method is shown in Figure 1.

A double-layer design is adopted for the experimental bench: the upper layer accommodates the high-vacuum furnace body and instruments for temperature measurement, such as pyrometers, while the lower layer houses the vacuum pumping unit and high-power low-voltage (LV) transformer. In this setup, the sample serves as the heating element due to its electrical conductivity. Both ends of the sample are clamped by hollow brass electrodes, allowing direct Joule heating. A window is integrated into the furnace wall adjacent to the sample position, enabling real-time temperature monitoring and visual observation via a pyrometer. The power output is adjusted to ensure the sample reaches the target temperature within a specified timeframe. The hollow electrodes are water-cooled to prevent overheating. Graphite felt is used as an insulating material to protect both equipment and operators from radiant heat, as well as to minimize heat loss within the system.

2.1.2 Conductive thermoelectric heating method

The thermoelectric heating system based on the conductive method is shown in Figure 2, which also employs a double-layer structure. The upper layer contains the high-vacuum furnace and temperature measurement instruments, while the lower layer includes the vacuum pump and high-power LV transformer. In this method, a tubular graphite heater generates a uniform high-temperature isothermal zone. The sample is positioned in the center of this zone using a graphite carrier, allowing it to be heated both conductively and radiatively. Graphite electrodes at both ends of the heater ensure electrical conduction and thermal insulation. The electrodes extend outside the furnace body and are water-cooled for protection. High-temperature ceramic insulation is applied to both ends of the heating assembly, and a graphite insulation barrel is added to reduce radiative heat loss.

As shown in Figures 1, 2, in the two electric heating structures, the conductive thermoelectric heating method involves a longer isothermal zone, which contributes to heating the sample uniformly. However it involves a relatively complicated structure, making it difficult to observe the sample. The directly clamped electric heating method is simple in structure and easy for sample observation. However, the heating body will be oxidized under high temperature and aerobic conditions. Therefore, how to protect the heating body becomes a major problem.

2.2 Radiation characteristics measurement method

In this study, the radiance-comparison principle (Arduini et al., 2021; Kayikci et al., 2025) is adopted, with the sample cavity serving as a near-blackbody reference (Yousef et al., 2000). The sample is designed with a circular cavity structure, as illustrated in Figure 3. Given that the sample is relatively thin (approximately 2 mm in thickness), the temperature difference between its front and back surfaces can be considered negligible.

According to Planck’s law, the spectral radiance L λ , T of a blackbody at a given temperature T and wavelength λ is known quantity. The cavity region of the sample approximates a near blackbody due to multiple internal reflections, making it reference for comparison. To determine the emissivity ε λ , T of the material surface, we measure the spectral radiance of both the outer sample surface L s λ , T and the cavity L c λ , T using a Fourier Transform Infrared (FTIR) spectrometer under identical temperature conditions. The emissivity of the material is then determined by calculating the ratio of the measured radiance values of the sample surface and the cavity:where L s λ , T is the spectral radiance of the sample surface, L c λ , T is the spectral radiance measured inside the cavity, λ is the wavelength, T is the uniform temperature of the sample.

ε λ , T = L s λ , T L c λ , T (1)

This method is based on the assumption that the cavity behaves as a near-perfect blackbody at the same temperature, as discussed in previous radiance-comparison and blackbody-cavity studies (Arduini et al., 2021; Kayikci et al., 2025; Yousef et al., 2000). It allows emissivity determination without directly relying on absolute radiance values. Equation 1 is valid under the conditions that the sample is optically opaque, the temperature is uniformly distributed, and the cavity exhibits high effective emissivity due to its geometry. In our case, the measurements are performed within the mid-infrared range, ensuring the reliability of this method.

2.3 Test materials and specimen fabrication

Commercially supplied molybdenum (Mo), tungsten (W), and tantalum (Ta) sheets were obtained from Baoji Belong Metal Co., Ltd. (Baoji, Shaanxi, China). Mo, W, and Ta were selected because of their high melting temperatures, good thermal stability, and electrical conductivity. These properties make them suitable for high-temperature and vacuum applications, including heating elements, vacuum-furnace components, thermal protection structures, and aerospace propulsion components. Their electrical conductivity also makes them compatible with the directly clamped Joule-heating method used in this study.

The specimens were fabricated from rectangular metal sheets with dimensions of approximately 50 mm × 25 mm × 2 mm. A rectangular notch measuring approximately 3 mm × 3 mm was produced at one end of each sheet by laser cutting. The sheet was subsequently roll-formed into a cylindrical specimen, and the adjoining longitudinal edges were joined by laser welding. After forming, the rectangular notch constituted a small aperture in the cylindrical wall. No additional polishing or surface coating was applied after fabrication, and the specimens were tested in the as-fabricated and unpolished condition. The specimen geometry and the actual figure are illustrated in Figure 3.

3 Device architecture and working principle

3.1 Overall structure and components

A wide-temperature-range vacuum radiation characteristics measuring device has been developed based on the directly clamped electric heating method. This system is capable of testing the radiative properties of materials in a vacuum environment over a broad temperature range, from 500 °C up to more than 3,000 °C. The schematic diagram and physical layout of the device are shown in Figure 4. To simulate the actual service conditions of materials in space, such as repeated thermal cycling, the device enables rapid heating and cooling of the sample. Through comparative analysis, the directly clamped electric heating method was selected over the conductive thermoelectric heating method due to its higher heating efficiency and better suitability for transient thermal tests. The complete system comprises an experimental bench, a lifting-type vacuum chamber, a directly clamped electric heating unit, a vacuum subsystem, a temperature measurement module, an emissivity measurement system, and a control unit. The vacuum subsystem includes a vacuum measurement device, a molecular vacuum pump (model F250/1,500), and a mechanical vacuum pump (model 2XZ-15D), which together ensure a stable low-pressure environment. The sample temperature is monitored by an optical fiber pyrometer (model SMART-FG-3522), and its spectral emissivity is measured in real time using a Fourier Transform Infrared Spectrometer (model FT-IR6000), both of which observe the sample through a designated optical window. The directly clamped electric heating system is mounted on the experimental bench, enabling stable and repeatable high-temperature operation throughout the entire testing process.

The working principle of the device is as follows: the sample itself serves as the heating element, utilizing its electrical conductivity to generate heat through direct current flow. Both ends of the sample are clamped and conductively heated by water-cooled hollow brass electrodes, ensuring stable electrical contact and preventing overheating of the clamping components. A high-power, low-voltage transformer, along with a control cabinet installed behind and beside the experimental bench, provides the necessary large current and low voltage required for rapid heating. The output power is precisely adjusted to enable the sample to reach the target temperature within a predetermined time frame, while real-time temperature monitoring is carried out using an optical fiber pyrometer. To achieve high-vacuum conditions, a specially designed vacuum cover is employed. A sight glass is positioned on the cover at the same height as the sample, allowing simultaneous visual observation and spectral measurement. This configuration facilitates real-time monitoring of the sample temperature, surface condition, and spectral emissivity using the optical fiber pyrometer and the Fourier Transform Infrared Spectrometer. The vacuum pumping system, located beneath the experimental bench, includes a motorized lifting mechanism that lowers the vacuum cover into position during testing. Once sealed, the system initiates vacuum pumping to establish the required high-vacuum environment for radiation measurements. The control cabinet, positioned on one side of the bench, is responsible for automated regulation of the experimental process, including heating control, vacuum management, and data acquisition.

3.2 Vacuum pumping system

To establish the required high-vacuum environment, the system employs a two-stage pumping arrangement consisting of a mechanical backing pump (model 2XZ-15D) and a molecular pump (model F250/1,500). The mechanical pump first reduces the chamber pressure to the operating range required by the molecular pump, after which the molecular pump is activated to establish the high-vacuum condition. A schematic of the vacuum system is provided in Figure 5. This combination of turbo molecular and mechanical pumps offers significant advantages: it enables the system to reach a vacuum level of approximately 10⁻³ Pa without the need for baking, and up to 10⁻⁵ Pa with baking. Moreover, the system is free from contamination and is capable of achieving oil-free ultra-high vacuum conditions, making it suitable for sensitive material property testing. For pressure measurement within the vacuum chamber, a dual-range vacuum gauging approach is adopted. A resistance-type gauge is employed for the low-vacuum range (from atmospheric pressure down to 10⁻¹ Pa), while an ionization gauge is used for the high-vacuum range (from 10⁻¹Pa to 10⁻⁸Pa). This configuration ensures accurate and reliable monitoring of the vacuum environment throughout the experimental process.

3.3 Control system

The control system of the wide-temperature-range vacuum radiation measuring device comprises several integrated components, including a vacuum chamber, industrial personal computer, temperature control instruments, a high-power low-voltage transformer, vacuum pumping system, digital temperature display units, power output modules, and vacuum measurement and control devices. This system enables real-time acquisition and regulation of sample temperature, precise monitoring of vacuum level, and automated vacuum pumping operations throughout the experimental process. Both manual and fully automated control modes are supported, allowing flexibility depending on experimental requirements. A schematic block diagram of the control system is presented in Figure 6, illustrating the functional architecture and interconnection of the various subsystems.

4 Experiment and uncertainty analysis

4.1 Vacuum radiation characteristics experiment

The emissivity measurements of aerospace materials are essential for accurate thermal modeling and temperature control in aerospace thermal protection systems and high-temperature processing environments. To validate the performance and effectiveness of the radiation measurement device developed in this study, the spectral radiation properties of molybdenum, tungsten, and tantalum materials were experimentally measured. These materials are commonly used in rocket nozzles, heat shields, and propulsion components, where emissivity data enable more precise prediction of radiative heat loss and surface temperatures during operation. The samples used in the experiments had flat surfaces and were unpolished to reflect realistic application conditions. The experimental procedure was as follows: First, the sample was mounted into the vacuum testing chamber. Then, the vacuum system and the water-cooling system were activated to establish a stable high-vacuum environment. Subsequently, the target experimental temperatures were set to 800 °C, 1,000 °C, and 1,200 °C, respectively. During the tests, the radiance data of both the sample surface and the cavity (formed by the cavity effect) were recorded. Based on the radiative output curves, the emissivity of each material was calculated according to the ratio between cavity and surface radiation intensity. Prior to formal measurements, the FTIR spectrometer was calibrated using a certified blackbody source to ensure spectral accuracy. Each test was repeated three times under identical conditions to verify the reproducibility of the results, and the standard deviation of measured emissivity values remained below 2% across the spectral range. Moreover, the system’s accuracy was cross-validated by comparing the measured emissivity of a reference material (oxidized molybdenum) with literature values, showing good agreement. These steps collectively confirm the reliability and robustness of the proposed methodology for high-temperature emissivity characterization under vacuum, supporting its applicability in industrial and aerospace contexts. The emissivity results for molybdenum, tungsten, and tantalum are presented in Figures 7–9, respectively.

As shown in Figures 7–9, all three refractory metals exhibit non-gray spectral behavior within the measured wavelength range. The spectral emissivity decreases relatively rapidly in the short-wavelength region, particularly between approximately 2 and 5 μm, and then changes more gradually at wavelengths above approximately 6 μm. For each material, the curves obtained at 800, 1,000, and 1,200 °C show similar spectral shapes, indicating that the main wavelength-dependent characteristics were consistently captured at the three temperature levels.

A positive temperature dependence is observed for the three materials over most of the measured spectral range. For example, at 10μm, the emissivity of Mo increases from 0.191 at 800 °C to 0.247 at 1,200 °C, while the corresponding values for W and Ta increase from 0.094 to 0.142 and from 0.152 to 0.228, respectively. The increase is relatively gradual, but it is consistently observed at the selected temperatures. Similar temperature-dependent behavior has been reported for refractory metals in previous studies (Zhang et al., 2023; Shu et al., 2021).

The measured emissivity also differs among the three materials. In the long-wavelength region of approximately 8–14μm, Mo generally exhibits the highest emissivity, W the lowest emissivity, and Ta an intermediate value under the investigated conditions. The differences among the three materials become more evident as the temperature increases. These results demonstrate that the proposed system can distinguish the spectral radiation characteristics of different refractory metals. However, absolute emissivity values are also affected by material purity, surface roughness, surface contamination, and oxidation state. Because the specimens were tested in an unpolished condition, the comparison with previously published values should be regarded as qualitative unless their surface conditions are comparable.

Local fluctuations and sharp peaks are observed at several wavelengths, particularly in the short- and mid-infrared regions. Similar features appear in the spectra of more than one material and at different temperatures, suggesting that they may be related mainly to the measurement chain rather than to intrinsic material behavior. Possible sources include reduced signal-to-noise ratio, residual atmospheric absorption along the external optical path, background-radiation correction, and the wavelength-dependent transmission of the optical window. Therefore, these local peaks were not used as the main basis for evaluating the temperature dependence of the materials. Further background correction and optical-path calibration will be considered in future measurements.

4.2 Uncertainty analysis

When testing the spectral radiation characteristics of samples using the vacuum-compatible infrared emissivity measurement system developed in this study (at a wavelength of 10 µm), the sources of measurement uncertainty were categorized into two classes: Class A and Class B. Class A uncertainty sources include the stability of the sample surface temperature measurement, the stability of the sample cavity (blackbody) temperature measurement, and the stability of the spectral radiation measurement. Class B uncertainty sources include the uncertainty of the sample surface temperature measurement, the influence of background radiation reflected from the sample surface, spectral noise, the effective emissivity of the reference blackbody, and the nonlinearity of the detector or sensor response (Monte and Hollandt, 2010; González-Fernández et al., 2010). The combined standard uncertainty was calculated by root-sum-square method, and the expanded uncertainty was obtained by applying a coverage factor of k = 2, as summarized in Table 1.

TABLE 1. Analysis of uncertainty of radiation characteristics measurement.

As shown in Table 1, the combined standard uncertainties at 800, 1,000, and 1,200 °C are 0.0097, 0.0097, and 0.0099, respectively. With a coverage factor of k = 2, the corresponding expanded uncertainties are 0.0194, 0.0196, and 0.0198. The effective emissivity of the sample cavity is the largest individual uncertainty component at all three temperatures. At 800 °C, reflected background radiation also makes a relatively large contribution, whereas the uncertainty associated with sample-surface temperature measurement becomes more significant as the temperature increases. The detector linearity is another important component throughout the investigated temperature range. These results indicate that further improvement of the measurement accuracy should focus mainly on cavity-emissivity evaluation, background-radiation correction, and high-temperature measurement of the sample surface.

5 Conclusion

In this study, a vacuum-compatible infrared emissivity measurement system based on directly clamped electric heating was developed. The electrically conductive specimen itself serves as the Joule-heating element, while an FTIR spectrometer measures the radiation from the specimen surface and the sample-integrated cavity reference under vacuum conditions. The system is designed to support heating from 500 °C to above 3,000 °C, although the emissivity measurements reported in this work were conducted between 800 °C and 1,200 °C.

Three types of tests were carried out to evaluate the developed system. First, spectral emissivity measurements were performed on Mo, W, and Ta at 800, 1,000, and 1,200 °C. These materials were selected because they are representative electrically conductive refractory metals used in high-temperature and vacuum applications and are compatible with the direct Joule-heating configuration. The three temperature levels were selected to examine the temperature dependence of emissivity within the investigated operating range. Second, repeated measurements were conducted under the same conditions to evaluate measurement consistency. Third, an uncertainty analysis was performed at 10 μm to assess the reliability of the measured emissivity and identify the main sources of measurement uncertainty.

The measured spectra show that the emissivity of all three materials decreases relatively rapidly in the short-wavelength region and varies more gradually at longer wavelengths. The emissivity generally increases with temperature, and measurable differences are observed among Mo, W, and Ta. At 10 μm, the emissivity values estimated from the plotted curves were approximately 0.191–0.247 for Mo, 0.094–0.142 for W, and 0.152–0.228 for Ta over the temperature range from 800 °C to 1,200 °C. The expanded uncertainty in emissivity ranged from 0.0194 to 0.0198 under the investigated conditions. These results support the ability of the developed system to distinguish the spectral emissivity characteristics of different refractory metals over the investigated temperature range.

The present experiments were intended as an initial validation of the proposed system rather than the establishment of a complete emissivity database for refractory materials. The current validation was limited to three unpolished conductive metals, three temperature levels, and an uncertainty assessment at one representative wavelength. Future work will extend the measurements to wider temperature ranges, additional refractory materials and coatings, different surface conditions, and comparisons with certified reference data.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. Generative AI was used to improve the language and readability of the manuscript. All scientific content, experimental design, data analysis, and conclusions were reviewed and verified by the authors, who take full responsibility for the final manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Pankaj Thakur, ICFAI University, Himachal Pradesh, India

Reviewed by: Zoran Radaković, University of Belgrade, Serbia

Royal Madan, Graphic Era Deemed to be University, India