Lanthanide-ion-doped near-infrared (NIR) luminescence materials have attracted considerable attention owing to their narrow excitation and emission peaks, stable physical and chemical characteristics, as well as being promising candidates in optical and biological applications [1–5]. Unfortunately, the practical applications of NIR-II (1000–1700 nm) luminescence in some fields, including temperature sensing, bioimaging, and bioassays, have been seriously affected by the decreased emission intensity with the increasing temperature, which is known as the normal thermal quenching effect [6–8]. Multiple strategies have been devoted to combating thermal quenching of NIR-II luminescence, for example, desorption of adsorbents on lanthanide-doped NIR-II emitting materials, doping additional ions to adjust the crystal field environment, selection of hosts with moderate phonon energy, and manipulation of the design energy transfer process to populate the corresponding excited state energy level [9–11].

The energy transferred to the activator ion in a luminescent system is limited. Generally, part of the energy is emitted for luminescence through radiative transitions, while the other part is dissipated through lattice vibrations or other forms, naturally involving a multi-phonon relaxation (MPR) process [12]. Meanwhile, the phonon-assisted energy transfer (PAET) process can effectively facilitate the population of the corresponding excited state level in a mismatched energy level system, further leading to the compensation of energy dissipation for anti-thermal quenching of NIR-II luminescence [13]. More importantly, benefiting from the rich intermediate energy level of ${\mathrm{Ho}}^{3+}$ and ${\mathrm{Er}}^{3+}$, this can provide attractive possibilities for multi-phonon relaxation and phonon-assisted energy transfer processes, which further could regulate the normal thermal quenching of NIR-II luminescence and temperature sensing performance.

Specifically, lanthanide-based luminescence intensity ratio (LIR) thermometry is much encouraging for achieving accurate temperature, no matter the changes in doping concentration, excitation power, and detection efficiency. For example, the reported ${\mathrm{Sc}}_2{\mathrm{W}}_3{\mathrm{O}}_{12}\text{:}\mathrm{Yb}/\mathrm{Er}$ phosphors with negative thermal expansion have achieved the thermal enhancement of ${\mathrm{Er}}^{3+}$ for green and red emission, as well as NIR-II luminescence through introducing Zr and Nb, enabling the design of optical thermometers in the visible region [14]. Recently, ${\mathrm{K}}_3{\mathrm{ZrF}}_7\text{:}\mathrm{Yb}/\mathrm{Er}$ nanocrystals also exhibited abnormal thermal enhancement of upconversion luminescence by energy compensation from defects induced by the heterovalent doping of ${\mathrm{Yb}}^{3+}/{\mathrm{Er}}^{3+}$ into the host lattice, which results in the constructed nanothermometer possessing the highest $S_r$ of $1.55\%{\mathrm{K}}^{-1}$ based on the conventional thermal coupling energy level of ${\mathrm{Er}}^{3+}$ [15]. Encouraged by the reported progress, it is worth noting that a more detailed and systematic exploration from the perspective of phonon-based MPR and PAET still has a profound effect on temperature sensing and manipulation of NIR-II luminescence in fluoride systems beyond the traditional oxide host.

In this study, based on MPR and PAET processes, we have delicately proposed the phonon-based LIR thermometry in the NIR-II region through manipulating the ${\mathrm{Ho}}^{3+}$ luminescence in $\beta \text{-}{\text{NaLuF}}_4\text{:}{\mathrm{Ho}}^{3+}/{\mathrm{Er}}^{3+}/{\mathrm{Yb}}^{3+}$ hollow microtubes. Meanwhile, the mechanism of anti-thermal quenching of ${\mathrm{Ho}}^{3+}$ is attributed to the ${\mathrm{Er}}^{3+}$ as an energy trap center to facilitate the PAET process, thereby enabling a considerable energy transfer efficiency of up to 80.52% between ${\mathrm{Er}}^{3+}$ and ${\mathrm{Ho}}^{3+}$ without ${\mathrm{Yb}}^{3+}$ ions as sensitizers. Benefiting from the thermally boosted emission (${\mathrm{Ho}}^{3+}\text{:}\sim 1190\mathrm{nm}$, ${{}^5\mathrm{I}}_6\to {{}^5\mathrm{I}}_8$ transition) and thermally quenched emission (${\mathrm{Er}}^{3+}\text{:}\sim 1530\mathrm{nm}$, ${{}^4\mathrm{I}}_{13/2}\to {{}^4\mathrm{I}}_{15/2}$ transition), the phonon-based NIR-II ratiometric thermometers readily rendered the superior sensitivity and resolution ($S_r=0.57\%{\mathrm{K}}^{-1}$, $\delta T=0.77\mathrm{K}$). This work not only elucidates the phonon-assisted strategy to build an efficient NIR-II emission system but also presents a promising trajectory for the advancement of cutting-edge ratiometric temperature sensing.

Lanthanide-doped $\beta \text{-}{\text{NaLuF}}_4$ hollow microtubes were synthesized using a previously described facile hydrothermal approach [16]. The synthesis typically commenced with the dissolution of ${\mathrm{NaNO}}_3$, ${\mathrm{KNO}}_3$, $\mathrm{Lu}{({\mathrm{NO}}_3)}_3$, $\mathrm{Yb}{({\mathrm{NO}}_3)}_3$, $\mathrm{Er}{({\mathrm{NO}}_3)}_3$, and $\mathrm{Ho}{({\mathrm{NO}}_3)}_3$ in a beaker, according to the stoichiometric ratio. This was followed by stirring for 30 min to form a mixed solution. Meanwhile, 0.9260 g ${\mathrm{NH}}_4\mathrm{F}$ was dissolved in 14 mL of deionized water under continual stirring and subsequently added to the aforementioned solution. The resulting mixture was transferred into a 50 mL Teflon-lined autoclave and maintained at 200°C for 12 h. Once cooled to room temperature, the final samples were produced through a process of centrifugal precipitation, followed by three washes using deionized water and ethanol, and then drying at 60°C for 10 h.

X-ray diffraction (XRD) patterns of lanthanide-doped $\beta \text{-}{\text{NaLuF}}_4$ hollow microtubes were assessed using a Rigaku Smart-Lab diffractometer with Cu $\mathrm{K}\alpha$ radiation ($\lambda =0.15405\mathrm{nm}$). We used field emission scanning electron microscopy (FE-SEM) (Regulus-8100, Hitachi) combined with energy dispersive X-ray spectroscopy (EDS) analysis to examine the morphological properties and elemental composition of samples. The emission spectra of the samples were measured using an Andor SR-500i spectrometer (Andor Technology Co., Belfast, UK) under 980 nm laser diode (LD) excitation, which combined an SR830 DSP lock-in amplifier and a CCD detector. Fourier transform infrared (FT-IR) spectroscopy was implemented using an FT-IR spectroscope (VERTEX 80, Bruker). The temperature-dependent luminescence property of the samples was monitored by a temperature control system (313–523 K) (TAP-02, orient-KOJI).

In general, resonant energy transfer (RET) can occur when the energy difference is small enough, such as in the case of ${\mathrm{Yb}}^{3+}\text{-}{\mathrm{Er}}^{3+}$ pairs [Fig. 1(a)]. For the energy mismatch between donor and acceptor, nonresonant energy transfer between rare earth ions can be achieved through the assistance of phonons in a host, including annihilating and emitting phonons [Figs. 1(b) and 1(c)]. According to the Miyakawa-Dexter theory, the corresponding temperature-dependent phonon-assisted energy transfer rate $W_{\mathrm{PAET}}(T)$ can be described as Eqs. (1) and (2) [17–19]: $$W_{\mathrm{PAET}}(T)=W_{\mathrm{PAET}}(0){[\exp (\mathrm{\Delta }E/k_{B}T)-1]}^{-p},$$(1)$$W_{\mathrm{PAET}}(T)=W_{\mathrm{PAET}}(0){[1-\exp (-\mathrm{\Delta }E/k_{B}T)]}^{-p},$$(2)where $W_{\mathrm{PAET}}(0)$ is the PAET rate at 0 K; $T$ is the absolute temperature; $\mathrm{\Delta }E$ is the energy mismatch between donor and acceptor; $p$ is the phonon number involved in the PAET process; and $k_B$ is the Boltzmann constant. Moreover, the temperature-dependent MPR rates can be mathematically written as $$W_{\mathrm{MPR}}(T)=W_{\mathrm{MPR}}(0){[1-\exp (-hv/k_{B}T)]}^{-p}.$$(3)$W_{\mathrm{MPR}}(T)$ represents the MPR rates at temperatures 0 K and T, respectively. That is to say, the PAET and MPR processes are strongly correlated with phonons of hosts with increasing temperature, and further regulate the NIR-II luminescence, enabling the interesting opportunity for phonon-based LIR ratiometric temperature sensing.

By virtue of the moderate phonon energy and high disorder in structure, $\beta \text{-}{\text{NaLuF}}_4$ has been widely employed as an ideal candidate fluoride host for upconversion luminescence. Despite previous exciting achievements, the effect of phonon energy on luminescence and temperature sensing in fluoride hosts has been rarely systematically explored in detail, particularly in the NIR-II region. Herein, an energy mismatch system of ${\mathrm{Yb}}^{3+}/{\mathrm{Ho}}^{3+}/{\mathrm{Er}}^{3+}$ co-doped $\beta \text{-}{\text{NaLuF}}_4$ hollow microtubes was carefully proposed. Under 980 nm excitation, the electron pumps from the ground state to the excited state level of ${\mathrm{Yb}}^{3+}$ by absorbing excitation energy. Subsequently, it transfers a portion of the energy to adjacent ${\mathrm{Er}}^{3+}$ owing to its small energy level mismatch, achieving downshifting luminescence (${{}^4\mathrm{I}}_{13/2}\to {{}^4\mathrm{I}}_{15/2}$ transition, $\sim 1530\mathrm{nm}$) of ${\mathrm{Er}}^{3+}$ through the MPR process. Moreover, due to the energy mismatch between ${\mathrm{Yb}}^{3+}$ and ${\mathrm{Ho}}^{3+}$, the PAET process is necessary to promote NIR-II luminescence (${{}^5\mathrm{I}}_6\to {{}^5\mathrm{I}}_8$ transition, $\sim 1190\mathrm{nm}$) [20–22]. Accordingly, the phonon-based LIR ($I_{\mathrm{Ho}}/I_{\mathrm{Er}}$) of a ratiometric NIR-II thermometer can be derived by the above analysis [23]: $$\mathrm{LIR}=\frac{I_{\mathrm{Ho}}}{I_{\mathrm{Er}}}=\frac{W_{\mathrm{PAET}}{A}_1}{W_{\mathrm{MPR}}{W}_R{A}_2}=\alpha \times {[1-\exp (-\frac{hv}{KT}\left)\right]}^{-\beta },$$(4)$$\alpha =\frac{W_{\mathrm{PAET}}(0)A_1}{W_{\mathrm{MPR}}(0)W_R{A}_2},$$(5)$$\beta =\frac{\mathrm{\Delta }{E}_1-\mathrm{\Delta }{E}_2}{hv}.$$(6)$\mathrm{\Delta }{E}_1$ and $\mathrm{\Delta }{E}_2$ are the energy gaps. Herein, $\mathrm{\Delta }{E}_1$ is the energy gap between ${{}^4\mathrm{I}}_{11/2}$ and ${{}^4\mathrm{I}}_{13/2}$ of ${\mathrm{Er}}^{3+}$ ions, $\mathrm{\Delta }{E}_2$ is the energy gap between ${{}^2\mathrm{F}}_{5/2}$ of ${\mathrm{Yb}}^{3+}$ ions and ${{}^5\mathrm{I}}_6$ of ${\mathrm{Ho}}^{3+}$ ions, $W_R$ is the rate of resonant energy transfer, and $A_1$ and $A_2$ are the spontaneous emission rate of the corresponding emitting transitions, respectively.

As a proof of concept, we first synthesized the $\beta \text{-}{\text{NaLuF}}_4\text{:}{\mathrm{Yb}}^{3+}$, ${\mathrm{Er}}^{3+}$, ${\mathrm{Ho}}^{3+}$ hollow microtubes by a moderate hydrothermal approach. The XRD patterns of $\beta \text{-}{\text{NaLuF}}_4\text{:}18\%{\mathrm{Yb}}^{3+}$, $x\%{\mathrm{Er}}^{3+}$ ($x=1$, 3, 5, 7) and $\beta \text{-}{\text{NaLuF}}_4\text{:}18\%{\mathrm{Yb}}^{3+}$, $y\%{\mathrm{Ho}}^{3+}$ ($y=0.5$, 1, 1.5, 2) with diverse doping concentrations are shown in Figs. 2(a) and 2(b), which are well indexed to the standard data with JCPDS#27-0726, confirming the crystal phase purity of all prepared samples. Given the significant influence of host crystalline phonon energy on NIR-II emission efficiency, FT-IR and Raman spectra emerge as valuable tools for investigating material phonon modes. As presented in Fig. 2(c) and Appendix A, there are five obvious Raman peaks of 239, 294, 350, 497, and $630{\mathrm{cm}}^{-1}$, which are characteristic of the hexagonal phase structure and exhibit similarities to those observed in reported ${\mathrm{NaYF}}_4$ powders. SEM and EDS element mapping pictures of $\beta \text{-}{\text{NaLuF}}_4\text{:}18\%{\mathrm{Yb}}^{3+}$, $5\%{\mathrm{Er}}^{3+}$ hollow microtubes are displayed in Fig. 2(d) to demonstrate the microscopic morphology and elemental composition, revealing that Na, Lu, Yb, Er, and F are uniformly distributed on the surface of the hollow microtubes, which further demonstrates that ${\mathrm{Yb}}^{3+}$ and ${\mathrm{Er}}^{3+}$ were successfully doped into the crystal lattice. In addition, as shown in Figs. 2(e) and 2(f), the sample is composed of high-quality hollow microtubes with a length of $\sim 6.85\mathrm{\mu m}$ and a width of $\sim 2.25\mathrm{\mu m}$. As such, ${\mathrm{Yb}}^{3+}/{\mathrm{Er}}^{3+}/{\mathrm{Ho}}^{3+}$ co-doped $\beta \text{-}{\text{NaLuF}}_4$ samples have been successfully synthesized, offering an essential foundation for the investigation of luminescence properties and temperature sensing.

The NIR-II emission spectra of $\beta \text{-}{\text{NaLuF}}_4\text{:}18\%{\mathrm{Yb}}^{3+}$, $5\%{\mathrm{Er}}^{3+}$ and $\beta \text{-}{\text{NaLuF}}_4\text{:}18\%{\mathrm{Yb}}^{3+}$, $1.5\%{\mathrm{Ho}}^{3+}$ hollow microtubes were collected under 980 nm excitation [Fig. 3(a)]. The optimal doping concentration of ${\mathrm{Er}}^{3+}$ in the NIR-II region is 5%, while it remains at 1% for visible emissions in $\beta \text{-}{\text{NaLuF}}_4\text{:}{\mathrm{Yb}}^{3+}$, ${\mathrm{Er}}^{3+}$ samples, which can be attributed to the accelerated cross relaxation process, promoting the population of a lower excited state level with increased ${\mathrm{Er}}^{3+}$ doping concentration for enhanced downshifting emission (Appendix B) [24,25]. In contrast, the NIR-II emission intensity of ${\mathrm{Ho}}^{3+}$ exhibited first an increase and then decreased owing to the concentration quenching effect with doping content from 0.5 to 2%. Furthermore, the superior characteristic emission peaks, corresponding to the ${{}^4\mathrm{I}}_{13/2}\to {{}^4\mathrm{I}}_{15/2}$ transition of ${\mathrm{Er}}^{3+}$ and ${{}^5\mathrm{I}}_6\to {{}^5\mathrm{I}}_8$ transition of ${\mathrm{Ho}}^{3+}$, were obviously observed in co-doped ${\mathrm{Er}}^{3+}/{\mathrm{Ho}}^{3+}$ samples, as shown in Figs. 3(b) and 3(c). More importantly, elaborately selecting a doping strategy including the ${\mathrm{Yb}}^{3+}/{\mathrm{Er}}^{3+}/{\mathrm{Tm}}^{3+}$ and ${\mathrm{Yb}}^{3+}/{\mathrm{Ho}}^{3+}/{\mathrm{Tm}}^{3+}$ system in $\beta \text{-}{\text{NaLuF}}_4$ simultaneously enables the outstanding NIR-II emissions, which far exceeds the reported NIR-II emitting materials (Appendix C).

To shed light on the luminescence mechanism, we further measured the pump-power-dependent NIR-II emission spectra of $\beta \text{-}{\text{NaLuF}}_4\text{:}18\%{\mathrm{Yb}}^{3+}$, $5\%{\mathrm{Er}}^{3+}$, $z\%{\mathrm{Ho}}^{3+}$ ($z=1$, 3, 5) samples under 980 nm excitation (Appendix D). As displayed in Fig. 3(d), there is a linear relationship between the logarithm of excitation power and the logarithm of the emission intensity, enabling the slopes of the fitted line corresponding to the transition process of ${\mathrm{Ho}}^{3+}$ and ${\mathrm{Er}}^{3+}$. From the obtained slopes ($n\sim 1$), it confirmed that the NIR-II emissions originating from the ${{}^4\mathrm{I}}_{13/2}\to {{}^4\mathrm{I}}_{15/2}$ transition ($\sim 1530\mathrm{nm}$) of ${\mathrm{Er}}^{3+}$ and ${{}^5\mathrm{I}}_6\to {{}^5\mathrm{I}}_8$ transition ($\sim 1190\mathrm{nm}$) of ${\mathrm{Ho}}^{3+}$ occurred via one-photon absorption processes [Fig. 3(e)] [26–28]. Interestingly, by keeping an ${\mathrm{Er}}^{3+}$ concentration of 5%, the slope value was slightly decreased with the increased ${\mathrm{Ho}}^{3+}$ doping content of 3%, indicating that a higher ${\mathrm{Ho}}^{3+}$ concentration could accelerate the PAET process between ${\mathrm{Er}}^{3+}$ and ${\mathrm{Ho}}^{3+}$ ions.

For understanding the PAET process from ${\mathrm{Er}}^{3+}$ to ${\mathrm{Ho}}^{3+}$ more deeply, a series of co-doped samples $\beta \text{-}{\text{NaLuF}}_4\text{:}2\%{\mathrm{Ho}}^{3+}$, $\beta \text{-}{\text{NaLuF}}_4\text{:}5\%{\mathrm{Er}}^{3+}$, and $\beta \text{-}{\text{NaLuF}}_4\text{:}5\%{\mathrm{Er}}^{3+}$, $m\%{\mathrm{Ho}}^{3+}$ ($m=1$, 2, 3, 5) was synthesized. The emission spectra of the samples under 980 nm excitation were recorded in Figs. 4(a) and 4(b). Specifically, benefitting from the self-sensitization of ${\mathrm{Er}}^{3+}$, dominant green emissions (centered at 525 and 545 nm), red emission (centered at 655 nm), and downshifting emission (centered at 1530 nm) were clearly observed in $\beta \text{-}{\text{NaLuF}}_4\text{:}5\%{\mathrm{Er}}^{3+}$, which are attributed to ${{}^2\mathrm{H}}_{11/2}$, ${{}^{4}S}_{3/2}$, ${{}^4\mathrm{F}}_{9/2}$, and ${{}^4\mathrm{I}}_{13/2}\to {{}^4\mathrm{I}}_{15/2}$ transitions, respectively. By contrast, $\beta \text{-}{\text{NaLuF}}_4\text{:}2\%{\mathrm{Ho}}^{3+}$ exhibited no emission owing to its extremely weak absorption cross-section with the excitation energy at 980 nm. The new emission bands of ${\mathrm{Ho}}^{3+}$ can be observed in the co-doped ${\mathrm{Er}}^{3+}$ and ${\mathrm{Ho}}^{3+}$ samples, originating from ${{}^5\mathrm{F}}_4$, ${{}^{5}S}_2$, ${{}^5\mathrm{F}}_5$, and ${{}^5\mathrm{I}}_6\to {{}^5\mathrm{I}}_8$ transitions, indicating the existence of energy transfer from the ${{}^2\mathrm{H}}_{11/2}$, ${{}^4\mathrm{S}}_{3/2}$ level of ${\mathrm{Er}}^{3+}$ to the ${{}^5\mathrm{F}}_4$, ${{}^5\mathrm{S}}_2$ level of ${\mathrm{Ho}}^{3+}$ as well as the PAET process from the ${{}^4\mathrm{I}}_{11/2}$ of ${\mathrm{Er}}^{3+}$ to the ${{}^5\mathrm{I}}_6$ of ${\mathrm{Ho}}^{3+}$. With the increase ${\mathrm{Ho}}^{3+}$ doping concentration from 1% to 5%, the NIR-II emission intensity of ${\mathrm{Ho}}^{3+}$ was gradually enhanced; meanwhile the downshifting emission intensity of ${\mathrm{Er}}^{3+}$ first increased and then decreased seriously, as depicted in Fig. 4(c). These results show that the back PAET process from ${\mathrm{Ho}}^{3+}$ to ${\mathrm{Er}}^{3+}$ also has a significant effect on NIR-II downshifting emission. More importantly, this was further supported by the downshifting lifetime of the ${{}^4\mathrm{I}}_{13/2}$ level of ${\mathrm{Er}}^{3+}$, which gradually decreased from 13.6 to 2.67 ms with the increasing ${\mathrm{Ho}}^{3+}$ contents from 1% to 5% [Figs. 4(d) and 4(e)]. Correspondingly, the experimental ${\mathrm{Er}}^{3+}\text{-}\text{to}\text{-}{\mathrm{Ho}}^{3+}$ energy transfer efficiency can be calculated using the following equation [29]: $$\eta =1-\frac{\tau }{{\tau }_0},$$(7)where $\tau$ and ${\tau }_0$ are the lifetime of the donor (${\mathrm{Er}}^{3+}$) in the presence and absence of the accepter (${\mathrm{Ho}}^{3+}$). The quantitative analysis revealed the rapid increase of energy transfer efficiency up to 80.52% for PAET pathways with increasing ${\mathrm{Ho}}^{3+}$ doping concentration [Fig. 4(f)], further validating that the PAET process plays a critical role in the energy mismatch system of co-doped ${\mathrm{Er}}^{3+}$ and ${\mathrm{Ho}}^{3+}$ [Fig. 4(g)].

As a proof of concept, we further establish a relationship between the PAET process and temperature sensing of the ${\mathrm{Er}}^{3+}\text{-}{\mathrm{Ho}}^{3+}$ co-doped system. Figure 5(a) illustrates the temperature-dependent emission spectra of the $\beta \text{-}{\text{NaLuF}}_4\text{:}18\%{\mathrm{Yb}}^{3+}$, $5\%{\mathrm{Er}}^{3+}$, $1\%{\mathrm{Ho}}^{3+}$ over a temperature range from 313 to 523 K under 980 nm excitation. Notably, the NIR-II emission of ${\mathrm{Ho}}^{3+}$ displays abnormal thermally enhanced performance along a temperature increase, reaching the 1.45-fold enhancement at 523 K, while the downshifting emission (centered at 1530 nm) of ${\mathrm{Er}}^{3+}$ shows serious thermal quenching [Fig. 5(b)], which could be attributed to the efficient PAET process between ${\mathrm{Er}}^{3+}$ and ${\mathrm{Ho}}^{3+}$ with the increased temperature [Fig. 5(d)]. This was further confirmed by the opposite trend of integrated emission intensities of ${\mathrm{Ho}}^{3+}$ ($\sim 1190\mathrm{nm}$) and ${\mathrm{Er}}^{3+}$ ($\sim 1530\mathrm{nm}$) emission with elevated temperature in a series of $\beta \text{-}{\text{NaLuF}}_4\text{:}18\%{\mathrm{Yb}}^{3+}$, $5\%{\mathrm{Er}}^{3+}$, $z\%{\mathrm{Ho}}^{3+}$ ($z=1$, 2, 3, 5) hollow microtubes, which further validated that the existence of a back PAET process between ${\mathrm{Ho}}^{3+}$ and ${\mathrm{Er}}^{3+}$ ions plays a considerable role in NIR-II temperature sensing [Fig. 5(c)].

In summary, we have proposed the phonon-based LIR thermometry in the NIR-II region through manipulating the ${\mathrm{Ho}}^{3+}$ luminescence in $\beta \text{-}{\text{NaLuF}}_4\text{:}{\mathrm{Ho}}^{3+}/{\mathrm{Er}}^{3+}/{\mathrm{Yb}}^{3+}$ hollow microtubes. The MPR and PAET processes between ${\mathrm{Er}}^{3+}$ and ${\mathrm{Ho}}^{3+}$ are strongly attributed to the dominant phonon of the $\beta \text{-}{\text{NaLuF}}_4$ host lattice, which corresponds to the strongest peak in the Raman spectrum, further achieving the definition of LIR. Meanwhile, the mechanism of anti-thermal quenching is attributed to the ${\mathrm{Er}}^{3+}$ as an energy trap center to facilitate the PAET process, thereby enabling a considerable energy transfer efficiency up to 80.52% between ${\mathrm{Er}}^{3+}$ and ${\mathrm{Ho}}^{3+}$ without ${\mathrm{Yb}}^{3+}$ ions as sensitizers. Benefiting from the thermally boosted emission (${\mathrm{Ho}}^{3+}\text{:}\sim 1190\mathrm{nm}$, ${{}^5\mathrm{I}}_6\to {{}^5\mathrm{I}}_8$ transition) and thermally quenched emission (${\mathrm{Er}}^{3+}\text{:}\sim 1530\mathrm{nm}$, ${{}^4\mathrm{I}}_{13/2}\to {{}^4\mathrm{I}}_{15/2}$ transition), the phonon-based NIR-II ratiometric thermometers readily rendered superior sensitivity and resolution ($S_r=0.57\%{\mathrm{K}}^{-1}$, $\delta T=0.77\mathrm{K}$). The present work expands a fascinating perspective for overcoming the thermal quenching of NIR-II luminescence, and holds enormous potential for high-performance optical temperature sensing.

Acknowledgment. The authors would like to acknowledge the Instrument and Equipment Sharing Platform, College of Physics (Jilin University), for testing assistance.