Figure 3A displays the schematic diagram of the ITO/PLZTN/Ag pyroelectric device. The core sensing material, PLZTN ceramic, covered with the top electrode of transparent ITO and a bottom Ag electrode, forms an almost freestanding state as shown in Fig. 3B. The scanning electron microscope (SEM) micrograph of the fracture surface in Fig. 3C demonstrates that the ceramic has uniform crystal grains and a dense structure. Figure 3D shows the x-ray diffraction pattern of the PLZTN ceramic, revealing a rhombohedral R3m phase of lattice parameters of a = b = c = 4.09 Å and α = β = γ = 89.91°, which belongs to an ABO₃-type perovskite structure. For further structure information, Fig. 3E shows the Raman spectra of the PLZTN ceramic at different temperatures in the range between 100 and 1,000 cm⁻¹. These Raman peaks correspond to BO₆ stretching/bending vibration modes and the A–O stretching mode [26,27]. It can be clearly seen that as the temperature increases, the Raman vibration modes gradually widen and weaken. It originates from the temperature-driven ferroelectric–paraelectric phase transition.

Figure 3F displays a well-developed saturated ferroelectric hysteresis loop of the PLZTN ceramic and strain-electric field loops under strain drive (bipolar and unipolar). The coercive field (E_c) is 11 kV cm⁻¹ and the remnant polarization (P_r) is 31 μC cm⁻². The positive strain value can reach about 0.28%. The typical butterfly-like curve reflects the rotation of the ferroelectric domain driven by the bipolar field and also represents a good converse piezoelectric performance. The calculated piezoelectric constant from the unipolar strain curve is about 615 pm V⁻¹. Figure 3G shows the temperature-dependent dielectric properties in the PLZTN ceramic. It can be clearly seen that the highest dielectric peak corresponding to Curie temperature indicating ferroelectric–paraelectric phase transition shifted from 117 °C at 10³ Hz to 125 °C at 10⁵ Hz, and this dielectric frequency dispersion represents a typical relaxor behavior. In addition, the transition from a nonergodic relaxor to an ergodic relaxor state occurs at the second dielectric peak, whose temperature (89 °C) is the depolarization temperature (T_d) [28]. Below T_d, the PLZTN material is in a nonergodic relaxor state, in which the electric field can induce an irreversible long-range ferroelectric order, so that high remnant polarization (Fig. 3F) and no obvious dielectric frequency dispersion (Fig. 3G) can be observed. When PLZTN is in an ergodic relaxor state above T_d, in which the external field can induce a reversible ferroelectric phase, low remnant polarization in P–E curves and a sprout shape instead of a butterfly shape in S–E curves are displayed (Fig. S1), indicating the disappearance of long-range ordered ferroelectric domains in the relaxor state. The pyroelectric effect describes the change in spontaneous polarization induced by the relatively small temperature change, and for most ferroelectrics, the precondition is the polarized ferroelectric materials. Therefore, in principle, T_d should be the inflection point where the pyroelectric effect of relaxor ferroelectrics gradually disappears. The FOM for pyroelectric energy conversion is defined as p_i²/ε₀ε_r (where p_i is the pyroelectric coefficient, ε₀ is the permittivity of free space, and ε_r is the relative permittivity) [25]. That is to say, a high pyroelectric coefficient and a low dielectric constant are required for excellent pyroelectric energy harvesting, which are difficult to achieve simultaneously in general ferroelectric materials. The pyroelectric coefficient of the PZLTN ceramic in the suspended state calculated from Eq. 5 and Fig. 4E is 1,122 μC m⁻² K⁻¹ at 1 atm and the FOM is 58 J m⁻³ K⁻² (ε_r = 2,449 at 1 kHz), showing an excellent pyroelectric performance. In a word, the PLZTN ceramic has good ferroelectric, piezoelectric, and dielectric properties, making it a candidate for applications in pyroelectric photoelectric detectors, energy harvesting, and so on [29,30].

We employed the substrate-free structure of Fig. 3A to measure the short-circuit currents and open-current voltages by 405-nm LED (light-emitting diode) excitation at a low pressure of 0.1 Pa and 1 atm, as shown in Fig. 4A and B and Fig. S2. It can be seen that under light intensities of 10.4, 38.3, and 60.5 mW cm⁻², the output current at 0.1 Pa is higher than at 1 atm, but not much. For instance, the average of the current peaks under 38.3 mW cm⁻² illumination is 63.8 nA at 0.1 Pa, while it is 63.7 nA at 1 atm (Fig. 4A). The pyroelectric current mainly depends on the change rate of sample temperature, as shown in Eq. 4, which is very small under different degrees of vacuum in our experimental conditions; thus, the influence of ambient thermal conductivity on the pyroelectric current is limited, resulting in small changes in the output current at different pressures. As shown in Fig. 4B, the average of the voltage peak is 12.2 V at 0.1 Pa, and the peak is 9.6 V at 1 atm. The influence of air pressure on the output voltage is greater than that of current, but the increase in voltage is obviously suppressed due to the load resistance. Furthermore, the output power of the ITO/PLZTN/Ag device loaded with a series of resistance at 1 atm and 0.1 Pa under 3 light intensities of 405-nm illumination is shown in Fig. 4C and D and Figs. S3 to S5. It can be seen that under 38.3 mW cm⁻² illumination, the maximum output power can reach 1,271 nW by a loading resistance of 1 GΩ at 0.1 Pa, which is increased by about 1.3 times compared with that at 1 atm (Fig. 4C and D). Similar phenomena were also observed under 10.4 and 60.5 mW cm⁻² illumination (Figs. S3C and D and S5C and D). Then, we explored the charge density generated from the ITO/PLZTN/Ag device under the 405-nm illumination with different intensities (Fig. 4E). The charge density of the ITO/PLZTN/Ag device at 0.1 Pa can reach 0.55 μC cm⁻² under 10.4 mW cm⁻² illumination, 2.51 μC cm⁻² under 38.3 mW cm⁻² illumination, and 4.48 μC cm⁻² under 60.5 mW cm⁻² illumination. Charge density at 0.1 Pa is increased by about 1.6 times under 10.4 mW cm⁻² illumination, 2.1 times under 38.3 mW cm⁻² illumination, and 3.0 times under 60.5 mW cm⁻² illumination, compared with that at 1 atm, respectively. Obviously, the charge density of the ITO/PLZTN/Ag device under a low pressure of 0.1 Pa is much higher than that under atmosphere pressure, because the increase in charge density is mainly related to the temperature change of the sample. These experimental results reveal that in the photoinduced pyroelectric output signals, the charge density is the most sensitive to low pressure, and in detail, the enhancement of vacuum (1 atm ➔ 0.1 Pa) results in a marked increase in charge density.

A detailed study on the charge density under different degrees of vacuum was conducted. Figure 5A to C displays the charge densities of the ITO/PLZTN/Ag device irradiating by 405 nm with 3 light intensities at a low pressure ranging from 10⁵ to 0.1 Pa. Extracted from Fig. 5A to C and the corresponding charge in Fig. S6, the maximum charge density as a function of the degree of vacuum is shown in Fig. 5D. In the case of 405-nm illumination with a constant light intensity, the charge density is related to the degree of vacuum with the most sensitive range between 100 and 0.1 Pa, which conforms to thermal conductivity vacuum gauges [31]. In the photoinduced pyroelectric output, the determinant temperature variation of a pyroelectric element is closely linked to the thermal conductivity of the element to the surrounding environment. Hence, we adopted the suspended pyroelectric element to make the heat transfer medium around the element mainly air. The change in the charge density of the ITO/PLZTN/Ag device with the degree of vacuum is mainly realized by the variation of thermal conductivity of the heat transfer medium air. It is well known that the change in thermal conductivity of gas with pressure is the physical basis for thermal conductivity vacuum gauges.

Figure 6 shows the thermal conductivity and simplified electrical circuit sections of a photoinduced pyroelectric detector in a specific space. If radiation W(t) with periodic frequency ω is incident normal to the surface of the pyroelectric element, whose effective emissivity is η, and induces a temperature change ΔT, it is described aswhere H and G_T are the element thermal capacity and the thermal conductance to the surrounding environment (or called ambient thermal conductivity), respectively [29,32]. The solution of Eq. 1 is

The pyroelectric effect refers to the spontaneous polarization P_s with the change of temperature (T), which is expressed by the pyroelectric coefficient p (a vector) as [29]

When a thin plate of polarized ferroelectric covered with electrode area A on the upper and lower surfaces is selected as the pyroelectric element, the pyroelectric charge (Q) and the pyroelectric current i_p generated in an external circuit has the formwhere p_i is a component of vector p normal to the sample surface (Fig. 1) [29]. Integrating Eq. 4 with respect to t (time) generates charge density

The substitution of Eq. 2 into Eq. 5 gives

Since H and p_i are the physical properties of the pyroelectric, the charge density generated in the photoinduced pyroelectric effect is mainly affected by G_T in the case of the given element, ensuring the same radiation. Thus, in theory, the charge density can be enhanced by lowering ambient thermal conductivity.

According to reports [33,34], the thermal conductivity of air in closed space can be approximately expressed aswhere κ_o is the thermal conductivity of air at atmosphere pressure and room temperature, T_avg represents the average temperature of enclosed space, D is the distance between the element and heat sink, and P denotes the pressure of air. If the suspended element within a closed cavity is irradiated with a constant light power W for a period time at room temperature (ω = 0), substituting the thermal conductivity of air κ_air given by Eq. 7 when the thermal conductance G_T only considers κ_air (G_T = κ_air) into Eq. 6 generates

At room temperature and atmosphere pressure, the corresponding charge density is

Substituting Eq. 9 into Eq. 8, the charge density is now expressed as

Therefore, through the abbreviated Eq. 10, the charge density changes with air pressure according to the functional relationship of \sigma ={\sigma }_0\cdot \left(1+\frac{N}{P}\right) (N is a constant) at a specified temperature. It is the theory basis of pyroelectric vacuum sensors and vacuum enhancement of charge density in the photoinduced pyroelectric effect. Additionally, the relationship between output voltage, current, and power with air pressure is discussed in Supplementary Note 1.

Incorporating the detector parameters (in Materials and Methods) into Eq. 10, the charge density of the ITO/PLZTN/Ag device as a function of the degree of vacuum by irradiating at 405 nm with light intensities of 10.4, 38.3, and 60.5 mW cm⁻² can be respectively expressed as

The curves in Fig. 5D, obtained from Eqs. 11 to 13, and the experimental data of the maximum charge density (Fig. 5A to C and Fig. S6) are good consistent with each other in the range from 10⁵ to 0.1 Pa.

To evaluate the performance of the detector, sensitivity and response speed are important parameters. It is known that the sensitivity of a linear sensor is a constant, but for a nonlinear sensor like the device in this work, the sensitivity S is a variable in the form of S = dy/dx, and the sensitivity of the ITO/PLZTN/Ag device is calculated from the derivative of Eqs. 11 to 13 with respect to pressure P, expressed as

Figure 5E shows the sensitivity S as a function of pressure P according to Eqs. 14 to 16 at 3 irradiation intensities, indicating that the higher the degree of vacuum, the greater the sensitivity. In addition, the response speed of charge density under 405-nm illumination is shown in Fig. 5F and Fig. S7. Here, the response speed is expressed by the rise time and decay time, that is, the time required for the maximum charge density rising from 10% to 90% under radiation, and the time needed for the maximum charge density declining from 90% to 10% after the light is removed [35]. The charts in Fig. S7C to E show that the rise time is slightly affected by the degree of vacuum and is within a certain range of values in this experimental condition. As previously mentioned, the pyroelectric signal is produced under sudden irradiation. When constant irradiation continues, the temperature of the sample will eventually reach a new steady value, and the charge density in the circuit will tend to be stable in the circuit. According to Eqs. 5 and 2, the charge density of a sample is related to the temperature change induced by the radiation, so the corresponding response speed depends on the thermal relaxation time of the pyroelectric element [36], where H/G_T is a thermal time constant [29]. It is reported that a BaTiO₃-based pyroelectric device has a faster response time on the substrate than those in suspended state, which is mainly due to the difference in the thermal conductance to the surrounding environment (G_T) [9]. The response time of the Pb(Zr,Ti)O₃ (PZT) film device is 10 times shorter than that of the PZT ceramic device, because the thermal capacity (H) of the PZT film device is less than that of the ceramic device [37]. In addition, different radiation sources also have an impact on the response speed. When the PLZTN device is exposed to a laser irradiation with a faster response time than LED, the response speed has been significantly improved, as shown in Fig. S8. It can also be seen that the response speed at 1 atm is obviously faster than that at 0.1 Pa, due to the influence of G_T.

Furthermore, as shown in Fig. S1, PLZTN ceramics display good and stable ferroelectric and converse piezoelectric properties below a depolarization temperature of 89 °C, which is a good prerequisite for pyroelectric output. It can be seen that the charge density of the PLZTN device has good temperature stability from room temperature to about 50 °C (Fig. S9). From the perspective of expanding the range of temperature stability, we think that ferroelectric materials with higher Curie temperature (such as LiTaO₃ with a T_C of 665 °C [29]) are good alternatives as components of a pyroelectric vacuum detector. Besides, as shown in Fig. S10, when the temperature is above the depolarization temperature of 89 °C, the pyroelectric effect of PLZTN disappears. Thus, the PLZTN pyroelectric vacuum sensor cannot work beyond depolarization temperature. In addition, Fig. S11 records the charge density of the PLZTN device at different times under periodic irradiation lasting 12 h, demonstrating good performance stability. As mentioned above, the charge density has a functional relationship with the degree of vacuum of air in the form of \sigma ={\sigma }_0\cdot \left(1+\frac{N}{P}\right) from 10⁵ to 0.1 Pa in the photoinduced pyroelectric effect. Therefore, the charge density of the pyroelectric element under radiation can be used to detect the degree of vacuum, and its essence is to sense ambient thermal conductivity at different low pressures. In a pyroelectric vacuum sensor, the lighting equipment can be installed outside the vacuum system, which is easy to disassemble and carry, and the device can realize self-powered electrical signal output without requiring additional power during operation. In actual use, it is necessary to further optimize the pyroelectric device, correct the parameters of vacuum detection, and perform calibration.