**3. Advanced temperature sensors**

In addition to the well-known disadvantages mentioned above, conventional temperature sensors face limitations when it comes to measuring temperature in complex and changing harsh environments or obtaining microscopic temperature data. In order to overcome these limitations and meet the demand for temperature measurement in extreme and complex conditions, advanced temperature sensors have emerged. These include optical temperature sensors, fiber optic temperature sensors, acoustic temperature sensors, and micro and nano temperature sensors. In this section, we will provide a detailed introduction to these advanced temperature sensors based on their typical classification characteristics.

#### **3.1 Optical temperature sensor**

Optical temperature sensors utilize the principles of Stephen Boltzmann's law and Planck's radiation law to measure temperature based on the optical effects of heat. This enables noncontact temperature measurement, allowing for the measurement of the surface temperature of high-temperature objects. Due to these advantages, optical temperature sensors are widely used in industrial production and food processing.

#### *3.1.1 Infrared thermal imaging*

An infrared thermal imager is the most widely used device among optical temperature sensors. It is based on the principle of thermal radiation of infrared to construct temperature fields. The current state-of-the-art direction in infrared thermal imager technology is reflected in the micro-electro-mechanical systems (MEMS) manufacturing process. The infrared thermal imager receives infrared radiation energy and converts it into temperature gradients through the front-end detector. The temperature gradients are then converted into electrical signals by the thermopile, and digital signals are obtained after amplification, shaping, and analog-to-digital conversion. These signals are visualized as temperature clouds on the display. Infrared thermal imaging provides an effective and fast method for real-time surface temperature measurement and is suitable for surface temperature field measurement with uneven temperature distribution or surface monitoring of superheated temperature.

The performance of IR MEMS is closely related to factors such as IR absorber absorption efficiency, thermopile performance, and thermal isolation layer material [4]. As a result, significant research efforts have been made to further enhance IR MEMS performance. The schematic of IR MEMS is shown as **Figure 5**. Key performance indicators for IR MEMS include time response, responsiveness, and noise level. To improve IR absorber absorption efficiency, Hou et al. [6] used carbon nanoparticles (CNP), Si3N4, and TiN nanoparticles loaded with a porous structured carbon microparticle (CMP) coating (CMP/CNP-Si3N4-TiN) as absorption media. The experimental results demonstrated that the coating has the capability to function as highly sensitive broadband absorbers in the range of 3–5 μm to 8–14 μm, achieving absorption rates of 93.8% and 92.6%, respectively. To improve the time response rate, Li et al. [7] utilized a novel single-sided micromachining technique fabricated in 111 wafers. They found that the p-Si-Al thermocouple in series in the IR thermopile demonstrated significantly higher Seebeck coefficients and lower noise compared to the conventional polycrystalline Si-Al thermocouple. By optimizing the cross-sectional area of the two thermoelectric material layers, the signal-to-noise ratio (D\*) of the thermopile was improved, resulting in an ultrahigh responsivity of 342 V/W and an ultrashort response time of 0.56 ms.

In various fields such as fire detection, security monitoring, human body temperature measurement, drone control, and industrial automation, infrared thermal imaging technology has gained widespread application due to its fast response, low

**Figure 5.** *Infrared thermopile components schematic [5].*

*Current Status and State-of-Art Developments in Temperature Sensor Technology DOI: http://dx.doi.org/10.5772/intechopen.112877*

noise, absence of mechanical movement, wide spectral range, high stability, and strong anti-interference capabilities.

#### *3.1.2 Laser temperature sensor*

Unlike infrared thermography, a laser temperature sensor is an active temperature measurement technology that calculates the surface temperature of an object by actively emitting a laser beam onto the surface of the object and measuring the energy reflected or scattered by the laser. Since the laser is actively emitted and received for temperature measurement, it can achieve long-distance temperature measurement and avoid the influence of environmental radiation.

In many cases, the emissivity of the object being measured is unknown, prompting the UK's National Physical Laboratory (NPL) [8] to study and propose the laser emissivity free thermometry (LEFT) technique such as **Figure 6** shown. This laserbased method for measuring target surface temperature does not require knowledge of the object's emissivity. The LEFT method is achieved by analyzing the ratio of intensities of two different laser wavelengths absorbed by the target material being tested. While the laser temperature measurement method performs well on surfaces with high emissivity, it may encounter difficulties when applied to surfaces with low emissivity. An et al. [9] constructed a noncontact temperature measurement system based on the principle of infrared laser radiation temperature measurement shown in **Figure 7**. The experimental results revealed that, for surface temperatures between

**Figure 6.** *LEFT device overall structure schematic [8].*

*Active dual wavelength infrared laser measurement schematic [9].*

873 K and 1173 K, the relative deviation of the results from the reference value was within 0.5% for high emissivity samples. For low emissivity samples, the relative deviation from the reference temperature was within 0.8%, with an average absolute deviation of 3.3 K. The researchers also measured the surface temperature of low emissivity samples from 873K to 1173K.

#### **3.2 Optical fiber temperature sensor**

The fiber optic temperature sensor is a type of sensor that utilizes optical fiber transmission to measure temperature. Unlike optical temperature sensors, which operate on a different principle, fiber optic temperature sensors rely on the thermal effect and spectral properties of optical fiber to perform temperature measurements. When subjected to varying temperatures, the refractive index within the fiber structure changes, leading to alterations in the optical transmission characteristics. Due to the small and flexible structure of optical fibers, this type of temperature sensor is particularly well-suited for use in complex environments with high temperatures, pressures, and radiation levels. Additionally, the number of temperature measurement points can vary, with fiber optic temperature sensors typically categorized as point, quasi-distributed, or distributed.

#### *3.2.1 Point temperature sensor*

The point-type fiber optic temperature sensor is capable of measuring the temperature at a single point in space using its temperature measurement method. Virginia Tech [10] first explored the sapphire fiber Fabry-Perot (F-P) sensor as a point-type fiber optic temperature sensor due to its high heat resistance and measurement resolution. This sensor utilizes the thermal expansion or thermogenic effect at the temperature measurement endpoint to generate an interference phenomenon between two light beams *via* refraction or reflection. This interference causes a change in the phase and amplitude of the optical signal, which is used to evaluate the temperature at the measurement point. The basic structure of this sensor is illustrated in **Figure 8**.

Jiang's group [11] conducted a series of studies on sapphire fiber optic sensors, which resulted in the development of an all-sapphire fiber optic temperature sensor [12]. This sensor is packaged in an all-sapphire structure, as shown in **Figure 9**, which effectively eliminates the mismatch in the coefficient of thermal expansion of the material under high-temperature conditions. The device is capable of operating at high temperatures ranging from 1000 to 1500°C for extended periods while maintaining a linear sensitivity growth trend with temperature. The evaluation conducted at room temperature indicates a measurement accuracy of 0.15 μm with an error of less than 2% of the full scale. Furthermore, the team proposed a higher-order mode suppression technique based on the sapphire fiber sensor

**Figure 8.** *Conventional sapphire fiber optic Fabry-Perot sensors [10].*

*Current Status and State-of-Art Developments in Temperature Sensor Technology DOI: http://dx.doi.org/10.5772/intechopen.112877*

#### **Figure 9.**

*All-sapphire fiber optic temperature sensor [12].*

using multimode [13] or tapered multimode [14], which effectively mitigates the irregular coupling between higher-order modes and fundamental modes, thereby improving the signal-to-noise ratio of interference fringes. Due to their high accuracy, sensitivity, and long measurement distance, point-type fiber optic temperature sensors find wide application in industrial, medical, and aviation fields for temperature measurement in extreme environments characterized by high temperature, high pressure, and high radiation.

## *3.2.2 Quasi-distributed temperature sensor*

The quasi-distributed temperature sensor employs a periodic grating fabricated on an optical fiber structure to enable temperature measurement over a distance where the grating is arranged. As the optical signal enters the optical fiber grating structure and passes through a location with a large change in refractive index between the gratings, reflection and transmission phenomena occur. The wavelength of the reflected light is influenced by the grating period and the amount of refractive index change. Therefore, by measuring the optical properties, specifically the wavelength, of the reflected light signal, the temperature value at the location of the fiber grating can be inferred, as depicted in **Figure 10**.

Thermally regenerative Bragg fiber grating sensors (RFBG) are widely used in fiber grating temperature sensors due to their high sensitivity and stability, especially their exceptional performance in high-temperature environments. The Hong Kong Polytechnic University [16] achieved the first secondary thermally regenerated Bragg fiber grating sensor (R<sup>2</sup> FBG) by continuously ramping the temperature, enabling temperature measurements of up to 1400°C. The device exhibits not only a temperature sensitivity of 13.7 pm/°C and excellent linearity at 250–900°C but also 15.3 pm/°C at high temperatures of 900–1370°C, and even more outstanding linearity (R2 = 99.9%). In addition to its excellent high-temperature measurement performance, RFBG is also suitable for long-term temperature measurements. F.J. Dutz [17] employed four six-element RFBG arrays in a chemical test stack with the package structure shown in **Figure 11**. The device operated for two years from 150°C to 500°C and exhibited no failures or significant wavelength drift.

**Figure 10.** *Fiber Bragg grating principal schematic [15].*

**Figure 11.** *Six-element regenerative Bragg grating package diagram [17].*

Thermal RFBG has several advantages such as high-temperature resolution, accuracy, real-time response, and strong resistance to electromagnetic interference. Therefore, it finds extensive applications in industrial automation, forest fire monitoring, and other fields.

### *3.2.3 Distributed temperature sensor*

Compared to quasi-distributed temperature sensors, distributed fiber optic temperature sensors enable temperature measurement over the entire length of the fiber optic path rather than just at local points. These sensors transmit optical signals through temperature-sensitive materials, such as optical fibers, and obtain photon energy and phase changes at various locations along the fiber line using Raman scattering or Brillouin scattering, ultimately allowing for the temperature distribution to be measured along the fiber. Among these sensors, the distributed temperature sensor-Raman (DTS-R) system is typically based on the optical time-domain reflectometer (OTDR) principle as shown in **Figure 12**, which involves emitting short pulses in the fiber and using the optical time difference between the round-trip to provide temperature variation and spatial location information along the entire fiber, as shown schematically below.

Marianne Stely Peixoto e Silva [19] conducted a comprehensive evaluation of the wide temperature performance of commercial time-domain optical reflectometry (OTDR) and erbium-doped fiber amplifiers (EDFA) on optical fibers up to 6 km in length. The results showed that the sensor sensitivity was 0.01 dB/°C@100 ns, providing temperature measurement from −196 to 400°C with a system resolution of 5°C. The sensor accuracy was 5°C in the range from −196 to 187°C, while it could reach 11.5°C at higher temperatures. Furthermore, Liu's group [20] explored the performance of temperature distributors based on Raman scattering at high

*Current Status and State-of-Art Developments in Temperature Sensor Technology DOI: http://dx.doi.org/10.5772/intechopen.112877*

#### **Figure 12.**

*Schematic diagram of Raman distributed temperature sensor system based on OTDR technique [18].*

temperatures. They proposed a Raman scattering temperature distributor system based on sapphire fibers under experimental conditions, using a high-power picosecond pulsed laser with a wavelength of 532 nm, which exhibited stability at a temperature resolution of up to 1200°C, with a spatial resolution of 14 cm and a temperature resolution of 3.7°C. In a subsequent experiment [21] with a fiber length of 2 m, they used a sub-nanosecond pulsed laser to increase the temperature detection limit to 1400°C and spatial resolution to 12.4 cm. Distributed optical fiber temperature sensors are commonly used in places such as oil and gas pipelines or transmission lines that require long-distance temperature measurements due to their characteristics.

#### **3.3 Acoustic temperature sensor**

An acoustic temperature sensor detects temperature by measuring changes in the speed and frequency of sound waves. Based on the principles of acoustics, the sensor uses the relationship between the speed, propagation characteristics, and temperature of sound waves propagating in a medium to measure temperature. As the temperature in the medium changes, the speed and frequency of sound waves also change in accordance with the thermal properties of the medium. The sensor is capable of measuring these changes and calculating the temperature. Due to its unique monitoring characteristics, acoustic temperature sensors are commonly used for the health monitoring of complex structures.

## *3.3.1 Acoustic time-of-flight*

The velocity of sound propagation is closely linked to the density, elastic modulus, and pressure of the medium it travels through. Temperature changes can be detected by observing the effect of temperature on the time-of-flight (TOF) in the medium. Typically, the speed of sound is directly proportional to temperature, such that as temperature increases, so too does the speed of sound in the medium. By measuring changes in sound velocity, temperature changes can be inferred. Wang [22] completed ultrasonic TOF measurements to determine temperature using a waveguide device. The measurement process requires a small region of effective

**Figure 13.** *Waveguide temperature measurement schematic [22].*

thermal inertia to be placed within the temperature field, as illustrated in **Figure 13**. This device can achieve high-precision temperature measurement of 0.015°C in constant temperature water bath environments below 100°C (using PT100 RTD as a reference). As such, this approach is a cost-effective and reliable alternative to contact measurement methods.

In addition to contact-based temperature measurement, this principle also allows for contactless temperature restoration at the measurement point [23], as depicted in **Figure 14**. Wang [24] proposed a temperature field reconstruction algorithm that enables two-dimensional temperature field reconstruction, reducing the relative error to a maximum of 2.881%, and improving the anti-noise interference capability. This contactless acoustic measurement method provides real-time measurement, high accuracy, a wide measurement range, and environmental adaptability, making it a promising solution for temperature measurement in harsh environments.

## *3.3.2 Resonant frequency*

This type of temperature sensor combines the principle of piezoelectric effect. The piezoelectric element is placed at the measurement position, and by applying an electric field to the piezoelectric element, the inverse piezoelectric effect of the piezoelectric element causes it to vibrate and produce sound waves, and the piezoelectric effect of the piezoelectric element generates an RF signal. When the temperature of an object is at a certain temperature, the gap width changes due to the thermal expansion effect of the piezoelectric element, and the resonance frequency changes. The detector analyzes the RF signal (the RF signal of the upper and lower electrodes is strongest at the resonance frequency) to obtain the temperature.

*Current Status and State-of-Art Developments in Temperature Sensor Technology DOI: http://dx.doi.org/10.5772/intechopen.112877*

**Figure 14.** *TOF-based temperature field reconstruction method [23].*

Despite the promising theoretical basis for thin film bulk acoustic resonator (FBAR) temperature sensors, the technology is still in the experimental study stage and has not been commercially promoted. Zhang [25] proposed an FBAR temperature sensor shown in **Figure 15** with a patterned support layer that demonstrated a differential pressure linearity error of only 0.35%, significantly smaller than existing sensors. However, the sensitivity parameter of the FBAR temperature sensor is critical, as demonstrated by the slow temperature stability achieved within 110 s and an accuracy of 0.015°C when the resonator temperature was controlled at 75°C and the ambient temperature was 25°C. To improve the temperature sensitivity, Zhao [26] proposed a dual-mode thin film bulk acoustic resonator (DM-FBAR) temperature sensor that uses different phosphorus-doped silica insertion layers. The FBAR with a SiO2 insertion layer doped with 4 sccm PH3 achieved a high-temperature sensitivity of up to 64.8 kHz/°C (resonant frequency magnitude of GHz).

#### **3.4 Micro/nano temperature sensor**

Micro/nano temperature sensors are typically fabricated using special treatment of materials on a micron or nano scale. Common materials used for fabrication include metals, semiconductors, polymers, and nanoparticles. Due to their small size, high sensitivity, fast response time, and ability to achieve local measurements, they are widely used for temperature monitoring in microelectronic devices, biomedical research, the automotive industry, and other fields.

#### **Figure 15.**

*Structure of the designed FBAR sensor chip: (a) cross section and (b) exploded view. 1-Si substrate, 2-silicon nitride support film, 3-bottom electrode, 4-ZnO piezoelectric film, 5-top electrode, 6-resistor heater, and 7-resistor temperature sensor [25].*

#### *3.4.1 Micro/nanoscale thermocouple*

Micro/nano thermocouples use the same measurement principle as traditional thermocouples, which is based on the thermoelectric effect of materials for temperature measurement. The difference lies in the use of modern micro or nano processing technology to produce a micro and nanoscale probe as a measurement tool. This allows for the measurement of temperature at a very small scale and with high sensitivity, making them useful in various fields including microelectronics, biomedicine, and nanotechnology.

Micro/nano thermocouples can be prepared using electrochemical etching techniques to create a micro/nanoscale probe. However, a new method was proposed by Huang [27] that involves depositing carbon and platinum metals on the inner and outer surfaces of quartz nanopipettes such as **Figure 16** shown. This method achieved an average temperature sensitivity of 1.98 ± 0.07 μV/K during calibration in the normal temperature range (−10–20°C), with a temperature resolution of 0.08–0.24°C. While the structural stability of this thermocouple preparation method has yet to be verified, Gu's team [28] successfully prepared micro and nanoscale thermocouples (W-Pt, 100 nm probe) with good structural stability. They obtained a temperature resolution of less than 0.1°C and a temperature response time of about 400 ns and demonstrated its effectiveness in measuring cell temperature in subsequent experiments [29]. Due to their stability and smaller thermal inertia compared to traditional thermoelectric devices, micro/nano thermocouples have great potential for applications in microelectronics and microbiological industries.

#### *3.4.2 Magnetic nanoparticles*

Magnetic nanoparticles possess unique properties that distinguish them from traditional magnetic materials, mainly due to their nanometer scale. These properties *Current Status and State-of-Art Developments in Temperature Sensor Technology DOI: http://dx.doi.org/10.5772/intechopen.112877*

**Figure 16.**

*(a) Schematic diagram of a nano thermocouple probe for intracellular temperature sensing and (b) Nano thermocouple probe [27].*

include magnetic transparency and superparamagnetic effects. The former characteristic is particularly advantageous for minimizing potential harm to biological organisms during temperature measurement. Additionally, according to Langevin's magnetic model [30], the magnetization of superparamagnetic particles is temperature-dependent. Thus, the theoretical foundation for magnetic nanoparticle temperature measurement (MNT) has been established, and a corresponding measurement device schematic is presented in **Figure 17**.

Xu's group [32] has developed a temperature measurement technique utilizing magnetic nanoparticles (MNPs) based on a real-time DC magnetization model. The MNP temperature-magnetic response rate was analyzed using theories related to heat transfer and ultrafast magnetic dynamics, leading to an innovative approach that utilizes both the frequency-domain response and time-domain response of MNPs. The frequency-domain response is utilized for parameter calibration, while the time-domain response is utilized for solving the temperature variation, thus achieving semi-invasive transient temperature measurement. Direct heating of the MNP with single pulses of varying pulse widths resulted in a rapid temporal resolution of 14.4 ns. In addition to its fast response time, this method also boasts a satisfactory

**Figure 17.** *MNT process diagram [31].*

temperature resolution. Meanwhile, Wang [33] enriched the harmonic signal by applying a dual-frequency magnetic field, reducing the thermal noise level through the principle that mixed-frequency harmonic information can reduce the sampling bandwidth [31], and detecting the magnetization signal with a tunnel magnetoresistance (TMR) element to obtain the analog signal of temperature. The four-harmonic model allows the temperature measurement error to be controlled within 0.015K.

Given the high accuracy of measurement, small particle size, and excellent temporal resolution associated with magnetic nanoparticles, magnetic particle imaging (MPI) [34] is a highly promising technique for a wide range of applications in industry and biomedicine, including the precise and noninvasive measurement of temperature.

#### *3.4.3 Micro/nanoscale fluorescent particles*

Fluorescent particle micro-clusters [35], such as organic dyes, lanthanide chelates, quantum dots, and lanthanide-doped nanoparticles, utilize the thermal properties of the material to emit light with different intensities, spectral distributions, and decay lifetimes at different internal electron energy levels and under different temperature conditions.

Erving C [36] injected a solution containing fluorescent nanoparticles (NPs) into a mouse and placed them in an imaging chamber at a constant temperature environment. Two wavelengths of laser light were used to achieve tissue heating (808 nm) and optical excitation of the particles, respectively. The results were recorded by a camera and the fluorescence intensity was converted to temperature using a software algorithm. This measurement method effectively demonstrated the phenomenon of local heating in mice, with a temperature resolution of 0.6°C, as expected (**Figure 18**).

This technology offers a combination of high thermal sensitivity (>1%/K) and spatial resolution (<10 μm). It has a short collection time (<1 ms) and can be operated

#### **Figure 18.**

*(a) Subcutaneous injection of thermal sensitivity NPs in a mouse; temperature profiles as b) heating and c) cooling take place; (d) schematic representation of the in vivo SDTI experiment; fluorescence images of (e) heating and (f) cooling process [36].*

remotely, making it suitable for use in biological fluids, with rapidly moving objects, and in strong electromagnetic fields. Due to its real-time observation capabilities and intuitive visualization, fluorescent particle clusters have unique advantages in monitoring the process of cellular life activities.
