3D-Printed Conformal Thin Film Thermocouple Arrays for Distributed High-Temperature Measurements

Abstract

Conformal thin film sensing represents a cutting-edge technology capable of precisely measuring complex surface temperature fields under extreme conditions. However, fabricating high-temperature-resistant conformal thin film thermocouple arrays remains challenging. This study reports a method for manufacturing conformal thin film thermocouple arrays on metal spherical surfaces using a printable paste composed of silicates and Ag. Specifically, the use of silicate glass phases enhances the high-temperature performance of the silver printable paste, enabling the silver ink coatings to withstand temperatures up to 947 °C and survive over 25 h at 900 °C. The thermocouples, connected to Pt thin films, exhibited a Seebeck coefficient of approximately 17 μV/°C. As a proof of concept, an array of six Ag/Pt thin film thermocouples was successfully fabricated on a metal spherical surface. Compared to traditional wire-type thermocouples, the conformal thin film thermocouple arrays more accurately reflect temperature variations at different points on a spherical surface. The Ag/Pt conformal thin film thermocouple arrays hold promise for monitoring temperature fields in harsh environments, such as aerospace and nuclear energy applications.

1. Introduction

High-quality development of aerospace equipment requires accurate acquisition of operating condition parameters during its service life, which enables further technical improvements [1,2,3,4]. However, the working environment is characterized by high pressure, high temperature, high-speed airflow impingement, and strong vibration, resulting in the surface conditions of components being significantly influenced by the surrounding environment [5,6]. The installation of traditional embedded sensors may disrupt the structural integrity an’d cause aerodynamic problems, and, compared to traditional sensors, our developed conformal thin film thermocouple arrays are smaller in size, conformal, and easily arrayable [7,8]. Therefore, accurately obtaining the surface operating conditions without damaging the structure, performance, and flow field environment has become a global technical challenge. With the development of modern thin film deposition technology, integrated thin film sensing technology has been formed by combining the advantages of traditional contact-based temperature measurement techniques while avoiding the drawbacks of discrete devices [9,10,11,12,13]. Thin film thermocouples utilize in situ thin film deposition and surface patterning techniques to directly deposit the traditional sensing materials and multilayer thin films on the measured surface, achieving integrated thin film thermocouple structures [11,14,15]. However, for the state monitoring of complex curved surfaces, such as aircraft engines and hypersonic vehicle aerodynamic surfaces, the traditional physical vapor deposition is limited by its working principle and cavity space constraints, making it difficult to manufacture high-temperature thin film thermocouples on large and complex curved surfaces [16,17]. Additionally, current thin film thermocouples are mostly single-point sensing elements unable to meet the distributed temperature measurement requirements in practical applications [14,15]. Therefore, thin film thermocouple arrays have become an important future development direction [14,18,19]. Recent advances in 3D printing technology have shown numerous advantages in the fabrication of high-temperature thin film sensors, including the ability to accurately pattern complex curved surfaces, control pattern thickness, improve manufacturing efficiency, and reduce material waste [20,21,22,23,24,25,26]. Consequently, the selection of suitable printable materials has become a key factor in the manufacturing of thin film thermocouple arrays. While the current approach of using printed ceramic pastes can enable the development of high-temperature thin film thermocouples, the drawback of ceramic thermocouples lacking backside compensation wires hinders their large-scale practical application [11,14,27]. Metal-based thin film thermocouples are a promising solution to this problem, but the conventional vapor-deposited thin film thermocouples currently in use have not been able to exceed 600 °C in terms of high-temperature resistance [28,29]. Silver in metallic materials offers a good cost–performance ratio, with a melting point of up to 961.8 °C and excellent thermal stability [30]. However, currently reported silver inks suffer from poor thermal stability due to agglomeration and volatilization at high temperatures, rendering them unsuitable for use above 600 °C [31,32]. Despite silver’s high melting point, thin films made from existing silver inks fall significantly short of this melting point, failing to realize the potential of silver in high-temperature sensing applications. Therefore, there is an urgent need to develop printable Ag-based pastes. Silicate compounds can form amorphous phases during the sintering process [33], which not only enhance the film’s integrity and improve the interface bonding but also encapsulate the Ag particles, minimizing their volatilization and enhancing their high-temperature stability.

In this study, we have developed a printable composite material comprising terpineol, silicate, and Ag, which is used for 3D printing conformal thin film thermocouple arrays with six thermocouples on a metal spherical head. This composite material utilizes the SAC silicate compound as a sintering aid, effectively improving the high-temperature thermal stability of the Ag composite thin film and reducing the porosity to enhance the film’s integrity. Through experimental testing, the thin film thermocouple has demonstrated excellent performance and stability at up to 947 °C under harsh conditions. Furthermore, the developed conformal thin film thermocouple arrays on the metal spherical surface can accurately reflect the temperature changes on the complex surface, exhibiting great application potential in engineering practice. This work provides an innovative and practical technical approach for high-temperature resistant conformal distributed sensors and their related applications.

2. Materials and Methods

2.1. Material

The silver (Ag) powder used in this study had an average particle size of 1 μm and the purity was 99.5% (Chengdu Hehe Micro-Nano Technology Co., Ltd., Chengdu, China). The SiO₂-Al₂O₃-CaO powder used in this study had an average particle size of 1 μm and was amorphous (Chengdu Hehe Micro-Nano Technology Co., Ltd., China) the purity was 99.5%. Terpineol (organic solvent) was purchased from Guangzhou Mingen Petroleum Products Co. (Guangzhou, China). The high-temperature-resistant Pt inks were purchased from Chengdu Hehe Micro-Nano Technology Co., Ltd., China.

2.2. Fabrication Methods

Fabrication Method Figure 1a shows the preparation process of the Ag/SAC (SiO₂-Al₂O₃-CaO) composite paste. Terpineol, SAC powder, and Ag powder were mixed at a weight ratio of 1:1:2. A ceramic magnet was added to the mixture. The mixing process was then conducted using a magnetic stirrer set at 600 RPM for one to two hours to ensure sufficient mixing, producing the Ag/SAC composite paste suitable for 3D printing. Figure 1b portrays the fabrication process of the conformal thin film thermocouple arrays using a Ag/SAC composite slurry. First, the metal spherical surface was cleaned with alcohol and water, and then it was left in a ventilated place for half an hour until the surface was completely dry. Subsequently, the metal spherical surface was completely immersed in the electrical insulating paste so that the electrical insulating paste completely covered the metal spherical surface, and then the metal spherical surface was left standing for a while. After the insulation coating had fully flattened, it was placed into a muffle furnace for heat treatment at 850 °C for half an hour. Then, using the Weissenberg Direct Writing technique [34], the prepared Ag/SAC paste and Pt ink were used to print chip-shaped conformal film thermocouple arrays on the metal spherical surface with an insulating coating, as shown in Figure 1b. This method ultimately fabricates a conformal film thermocouple array on a metal spherical surface.

2.3. Characterization Methods

SEM and EDS elemental mapping images were obtained using a Zeiss Sigma 300 (Oberkochen, Germany) scanning electron microscope. X-ray diffraction (XRD) was conducted using a Shimadzu XRD-6100 instrument (Kyoto, Japan). The resistances of the Ag/SAC composite thermistors were measured in a tube furnace using a Keysight 34972A data acquisition system (Santa Rosa, CA, USA). Temperatures were recorded using commercially available type K thermocouples. The infrared thermal imaging camera was a Fluke Ti480 PRO (Everett, WA, USA) with a maximum temperature measurement of 1000 °C.

2.4. Test Methods

The R-T test system consisted of a quartz tube furnace (OTF-1200X, MTI KJ GROUP, Hefei, China), a data acquisition unit (KEYSIGHT 34972A), and a type K thermocouple (KPS-K, Taizhou China). The thermocouple and Ag/SAC composite thermistor were placed together in the constant-temperature zone of the tube furnace to obtain signals from both the thermocouple and the thermistor. Application verification of Ag/SAC conformal thin film thermocouple arrays included a data acquisition unit (KEYSIGHT 34972A), a 1064 nm laser, a Type K thermocouple, and a Ag/Pt thermocouple, which were placed together in the constant-temperature zone of the tube furnace to obtain signals from both the type K thermocouple and the Ag/Pt thermocouple.

3. Results

3.1. Sintering Process and Characterization

Sintering plays a crucial role in enhancing the overall performance of Ag/SAC composite films. Initially, we utilized Scanning Electron Microscopy (SEM) to observe the morphological changes of Ag/SAC films at different temperatures (Figure 2a) and measured their thickness variations (Figure 2b), analyzing and calculating the corresponding conductivity changes (Figure 2c). At 30 °C, SEM showed particulate distribution of silver and SAC with minimal conductivity. As the temperature reaches 800 °C, the surface morphology of Ag/SAC undergoes significant changes, the thickness decreases, and the conductivity increases substantially. This is due to the vitrification transformation of SAC, forming a dense glass phase that encapsulates the silver particles, resulting in a high-density and highly conductive Ag/SAC composite film. As the temperature increases to 900 °C, the sample’s thickness further reduces, and the conductivity increases. However, at 1000 °C, despite a continuous decrease in thickness, conductivity begins to decline. These changes are likely due to the agglomeration and volatilization of silver particles and because the phase transformation within the SAC glass phase at high temperatures impedes the formation of effective conductive pathways. XRD results (Figure 2d) at 30 °C, 800 °C, and 900 °C show only characteristic peaks of Ag, as SAC remains amorphous, which is critical for the densification of the Ag/SAC film. At 1000 °C, significant crystallization and phase transformations occur, forming compounds, such as CaSiO₃ and Al₂O₃, but no Ag compounds are detected, indicating good high-temperature stability of the Ag/SAC composite film. The introduction of the SAC composite significantly enhances the high-temperature stability of the Ag film. Based on SEM, thickness, conductivity, and XRD results, the optimal sintering temperature for Ag/SAC films is 800 °C to 900 °C.

Further analysis of the EDS images of the Ag/SAC films annealed at 900 °C (Figure 3a) shows a uniform distribution of elements, such as O, Ca, Si, Ag, and Al, on the film surface, indicating the successful embedding of the SAC silicate glass phase around the Ag film, forming a highly dense composite conductive film. The cross-sectional SEM and EDS image (Figure 3b) reveal that the main elements are similar to those in the planar results, with no apparent pores, further confirming the high density of the film and its tight integration with the alumina substrate. The formation of the SAC glass phase enhances adhesion and exhibits excellent thermal expansion matching with alumina. These results are attributed to the formation of the SAC glass phase, which not only enhances the film’s density but also improves the bonding performance with the substrate.

3.2. Ag/SAC Composite Slurry Film Thermal Resistance Test

Assessing the resistance–temperature relationship of Ag/SAC composite materials is crucial to their application in high-temperature sensors. To analyze the performance of the sensitive layer created from the Ag/SAC paste, we compared it with a standard K-type thermocouple. For easy testing of the sensitive layer data, we fabricated a grid-shaped sensitive layer. Due to its good electrical conductivity, its resistance was measured using a four-wire method, and silver wires were attached to its four sides. The sensitive layer and K-type thermocouple were then placed in the same constant temperature zone of the tubular furnace. The K-type thermocouple temperature data were recorded, and the four-wire method was used to measure the resistance data of the sensitive layer. Initially, four rounds of heating and cooling cycles (from 30 °C to 900 °C) were conducted, and the resulting curves were plotted (Figure 4a). Upon analysis of Figure 4a, it was found that the resistance of the sensitive layer increased with the temperature and recovered to its initial value after cooling, demonstrating its good high-temperature electrical stability and positive temperature coefficient. Such stability may be attributed to the vitreous phase formed by SAC during sintering. This has established a solid foundation for Ag/SAC composite material to replace precious metals in high-temperature sensors. A polynomial function fitting showed a strong correlation (R² = 0.99978) between resistance and temperature (Figure 4b). To validate its high-temperature stability, we conducted a stepwise heating test from room temperature to 900 °C. Each 200 °C increase was followed by a one-hour hold, and data were recorded using the K-type thermocouple and sensitive layer. Analysis of Figure 4c reveals that the output resistance of the sensitive layer is consistent with the output temperature trend of the K-type thermocouple and maintains stability during the platform period of each step, manifesting its remarkable heat resistance. Finally, the sensitive layer was subjected to an alternating cycle of six-hour testing between 700 °C and 900 °C, and the data were plotted (Figure 4d). Results reveal that the change in resistance is synchronous with temperature change without any time delay, further validating its exemplary stability at high temperatures. Then, a long-term stability test was conducted at a constant temperature of 900 °C for 30 h, with a resistance change of only 0.46% per hour. Through the highest-temperature test, the temperature was found to be around 947 °C, where obvious changes in the sensitive layer’s resistance began to appear (Figure 4f). These experimental results show that the Ag/SAC sensitive layer possesses excellent thermo-resistive properties, offering robust support for its application in high-temperature sensors.

3.3. Application Verification of Ag/Pt Conformal Thin Film Thermocouple Arrays

To demonstrate the advantages of our conformal thin film thermocouple arrays for distributed monitoring in extreme conditions, we first tested a single thermocouple (Figure 5). Using 3D printing, we deposited Ag/SAC composite paste and Pt ink on a metal strip with an insulation layer, forming the pattern shown in Figure 5a. This was then annealed at a high temperature of 850 °C for an hour in a muffle furnace. Subsequently, welding points were made at the ends of the Ag and Pt wires, and Pt wires were connected to them. The individual Ag/Pt thermocouple was thus fabricated, and a standard K-type thermocouple was placed in the constant-temperature zone of a tubular furnace. Type K thermocouples are commonly used sensors that can be reliably operated at high temperatures up to 1100 °C. As commercial products, their measurement accuracy is generally dependable. Thus, Type K thermocouples can be used as a control group to calibrate Ag/Pt thin film thermocouples. When calibrating the Ag/Pt thermocouples, the hot junctions of both the Type K thermocouple and the Ag/Pt thermocouple need to be placed in the same location to ensure temperature consistency and synchronization between the two types of thermocouples. By doing so, the Ag/Pt thin film thermocouples can be accurately calibrated, ensuring precise measurements. The voltage data of the Ag/Pt thermocouple and the temperature data of the K-type thermocouple were recorded simultaneously, and Figure 5b was plotted. Curve fitting with a quadratic equation yielded an R² of 0.99855 (Figure 5c), confirming the thermocouple’s feasibility.

Next, we printed conformal thin film thermocouple arrays (Figure 6a), as shown in Figure 1b, on a metallic spherical surface coated with an insulating layer using Ag/SAC composite paste and Pt ink. To test their performance under extreme conditions, we irradiated the arrays with a 1064 nm fiber continuous laser. First, we aligned the center of the laser to the approximate center near the top of the spherical head. The laser power was gradually increased from 0 W to 135 W, with an increase of 15 W every approximately 40 s. Voltage data from the six Ag/Pt thermocouples were converted to temperature using the equation from Figure 5c and plotted (Figure 6c). We also recorded infrared imaging data before each change in laser power using an infrared thermal imager (Figure 6d). Analysis of Figure 6c reveals that the temperature outputs follow the same increasing trend, rising stepwise with the laser power. However, there is a noticeable difference in the temperature data output by the six thermocouple nodes. This discrepancy is due to the laser being manually aligned to the center of the spherical head, not being strictly controlled, and not being perfectly aligned with the top of the head, resulting in certain errors. Observing Figure 6d, the yellow temperature in the image represents the temperature at the top of the spherical head. It can be seen that the temperature at the top of the spherical head is higher than the temperature of the thermocouple node, and there is also a certain temperature difference between the thermocouple nodes, supporting the aforementioned notion. Under short-term laser irradiation, the surface at the top of the spherical head heats up faster than the thermocouple nodes. Given the significant temperature gradient, the temperature at the nodes will also increase correspondingly. Further, the laser was manually adjusted to one of the heat node’s locations, as shown in Figure 6e, and the laser power was increased to 165 W. We recorded the data and plotted the temperature and time graph (Figure 6e), and we captured the thermal image (Figure 6f). Analysis of Figure 6e shows that different heat nodes’ output temperatures have significant differences. Combined with Figure 6f, the yellow temperature in the image represents the temperature at the location of the laser. It can be seen that the output temperature of each heat node is close to the temperature shown in the thermal imaging.

Finally, to demonstrate the differences in temperature measurement between discrete thermocouples and conformal thin film thermocouple arrays in a high-temperature airflow environment, we affixed a K-type thermocouple to the top of the metallic spherical surface. We directed a flame spray gundirectly at the top of the metallic spherical surface, recorded the output temperature data, and illustrated it in Figure 7a, while also capturing its thermal image (Figure 7b). Analysis of Figure 7a reveals significant differences in the temperature outputs between the K-type thermocouple and the conformal thin film thermocouple arrays, with the outputs from the conformal array initially being consistent but diverging over time. Combining this with the analysis from Figure 7b, under the impact of the flame, the smaller heat capacity of the K-type thermocouple causes it to quickly heat up to above 800 °C. The metallic spherical surface, having a larger heat capacity, heats up more slowly. Under the flame impact, the highest temperature at the top is only around 300 °C, with even lower temperatures at the heat nodes. It can be observed that the flame from the torch is distributed quite uniformly over the surface of the metallic spherical, and the voltage outputs from the various heat nodes are initially consistent. However, as time passes, differences in temperature at the heat nodes lead to disparities in the voltage outputs. The discrete thermocouples struggle to accurately reflect the actual temperature of the metallic spherical surface, while the measurements from the conformal thin film thermocouple arrays are closer to reality. Traditional discrete thermocouples struggled to accurately reflect surface temperatures, while conformal arrays provided more precise in situ measurements without structural alterations, ensuring continuity and consistency in thermal conductivity, and the influence of cold/hot spot effects is reduced.

4. Conclusions

In summary, this study has developed a composite paste containing terpineol, SAC powder, and Ag powder and introduced a method for designing and fabricating conformal thin film thermocouple arrays on metal spherical surfaces. We utilized SAC as a sintering aid to lower the sintering temperature and form silicate glass phases, thereby enhancing the density of the films and their adhesion to the substrate. Additionally, these glass phases encapsulate the silver particles, significantly reducing the volatilization of silver at high temperatures and enhancing its stability. Through morphological and thermal resistance analyses, we found that films made using this composite paste exhibited excellent thermal stability and a high degree of correlation with temperature. Building on this, we printed six Ag/Pt-based thermocouples on a metal spherical cap following the designed scheme and tested them under extreme conditions using a high-power laser and a blowtorch flame. The results demonstrated that, compared to discrete sensors, the conformal thin film thermocouple arrays made with Ag/Pt accurately reflect temperature variations at different locations on the metal spherical surface. Consequently, this study offers an effective technological approach to manufacturing high-temperature-resistant conformal thin film thermocouples and related applications, providing valuable insights for practical engineering applications.