Abstract
Ratiometric luminescence thermometry with trivalent lanthanide ions and their 4fn energy levels is an emerging technique for non-invasive remote temperature sensing with high spatial and temporal resolution. Conventional ratiometric luminescence thermometry often relies on thermal coupling between two closely lying energy levels governed by Boltzmann’s law. Despite its simplicity, Boltzmann thermometry with two excited levels allows precise temperature sensing, but only within a limited temperature range. While low temperatures slow down the nonradiative transitions required to generate a measurable population in the higher excitation level, temperatures that are too high favour equalized populations of the two excited levels, at the expense of low relative thermal sensitivity. In this work, we extend the concept of Boltzmann thermometry to more than two excited levels and provide quantitative guidelines that link the choice of energy gaps between multiple excited states to the performance in different temperature windows. By this approach, it is possible to retain the high relative sensitivity and precision of the temperature measurement over a wide temperature range within the same system. We demonstrate this concept using YAl3(BO3)4 (YAB):Pr3+, Gd3+ with an excited 6PJ crystal field and spin-orbit split levels of Gd3+ in the UV range to avoid a thermal black body background even at the highest temperatures. This phosphor is easily excitable with inexpensive and powerful blue LEDs at 450 nm. Zero-background luminescence thermometry is realized by using blue-to-UV energy transfer upconversion with the Pr3+−Gd3+ couple upon excitation in the visible range. This method allows us to cover a temperature window between 30 and 800 K.
Similar content being viewed by others
Introduction
Chemical reactions, biological functionalities, and various physical phenomena are all strongly determined by the local temperature. The measurement of local temperatures on successively smaller scales requires new methods of temperature sensing that rely on a remote detection principle. Several such concepts already exist. The most commonly known example is infrared (IR) thermography or pyrometry1,2,3, which only allows temperature measurements of surfaces of objects based on their emitted thermal radiation. According to the Stefan−Boltzmann law, the intensity of the emitted radiation scales with the fourth power of temperature and is expected to be sensitive for high temperatures and objects with high emissivity.
An alternative method of remote temperature sensing is luminescence thermometry, which exploits the fact that luminescence properties such as the emission intensity or the luminescence decay time are highly dependent on the local temperature of the surroundings of the luminescent species4,5,6. It requires a simple setup consisting of a laser excitation source, luminescent micro- or nanocrystals or molecules embedded in the object of interest, and a fast and efficient light detection system7,8,9. Its promising advantages over thermography include the lower accessible temperature ranges and non-invasive temperature detection below surfaces. Its successful application in biological media has been highlighted for many different optical centres, such as nitrogen vacancy (NV−) centres in fluorescent nanodiamonds10,11,12, semiconducting quantum dots (QDs)13,14,15, bulk halides16,17, and organic fluorophores18,19,20,21.
An alternative emerging class of luminescent thermometers are micro- or nanocrystals doped with trivalent lanthanide ions22,23. Their rich electronic level structure that arises from the partly filled 4fn (n = 1–14) inner shell with characteristic energy gaps on the order of thermal energies (several 100 cm−1) allows for thermal coupling between energetically adjacent levels. Most representatives of this class of luminescent thermometers thus rely on ratiometric sensing; in ratiometric sensing, the temperature is measured by means of the intensity ratio of the emission bands from two thermally coupled excited levels. The key advantages of this approach are its simplicity and that it is self-referenced. Moreover, the narrow linewidths of the intraconfigurational 4fn−4fn transitions minimize spectral overlaps between emission lines from two thermally coupled energy levels, even when the energy separation is small. The approximate independence of their energies from the surrounding host material related to the low radial extension of the 4f orbitals allows us to select the emission energy range of interest by appropriately choosing the lanthanide ion. Recently, there has also been progress in reverting this scheme and performing thermometry by detecting a reference emission from the same excited level upon dual excitation from two thermally coupled ground levels24,25,26. Luminescence (nano)thermometry has been demonstrated to be both a precise and accurate technique for probing fundamental thermodynamic phenomena at the micro- and nanoscale27,28,29 and helps to assess the local temperature fluctuations in tissue30,31,32,33,34,35 or in chemical reactors36,37,38,39,40. Currently, there has been observable progress on how to implement and standardize this technique for applications41,42,43,44,45, the connected limitations46,47, or how to couple temperature sensing with other functionalities such as single-molecule magnetism48,49 and (meso-)porous materials50,51, e.g., theranostics52.
The intensity ratio of the emission lines originating from two thermally coupled excited levels within the same configuration follows simple Boltzmann statistics if thermalization is faster than population decay (by any radiative or nonradiative pathways)53,54. It has been recently demonstrated that despite their appealing simplicity, Boltzmann-based thermometers with two excited levels suffer from a fundamental thermodynamic limitation55. At very low temperatures compared to the energy gap ∆E21 between the two excited levels |1〉 and |2〉 (kBT ≪ ∆E21), the probability of thermally exciting the population from the lower to the higher excited emissive level is vanishingly small, which translates to a negligible intensity of the emission from the higher excited level. In addition, the nonradiative absorption rate governing the thermalization from level |1〉 to |2〉 becomes so slow that it can no longer compete with the radiative decay of level |1〉, which leads to a decoupling of the two excited states. In contrast, at very high temperatures compared to the energy gap (kBT ≫ ∆E21), the populations of the two excited levels in the thermodynamic limit effectively equalize and thus lower the sensitivity of the thermometer, which relies on a relative net change in population between the two excited levels. Both described extremes diminish the overall achievable precision of the thermometer and imply that any Boltzmann-based two-level thermometer only performs with sufficient statistical precision within a limited temperature range dependent on the energy gap ∆E21 between the two excited levels.
A way to circumvent this thermodynamic limitation and to widen the temperature range for the highest precision of a Boltzmann-based thermometer is to extend the concept of thermal coupling between more than two energetically close excited levels. This method allows us to retain both an optimized response and sensitivity of a single luminescent thermometer by monitoring emission intensity ratios from multiple higher energy excited levels separated by different energy gaps from the lowest excited emissive state. While this has already been realized on a qualitative level56,57,58, no clear guidelines have been established with regard to the most advisable temperature ranges for energy gaps for two or more higher excited levels thus far. Knowledge of the fundamental thermodynamic properties of such a thermally coupled multilevel Boltzmann-based luminescent thermometer helps decide at which temperatures a change to another luminescence intensity ratio (LIR) is advisable to achieve the highest thermometry precision and which combination of energy gaps is optimal for the (wide) temperature range of interest.
Luminescence thermometry at high temperatures also suffers from another practical problem. Any solid emits thermal or black-body radiation, which becomes more intense at high temperatures. Moreover, Wien’s displacement law (λmaxT = b) with b ≈ 2.8978 mm∙K) states that the peak wavelength of the Planckian black-body spectrum shifts to shorter wavelengths at high temperatures. While the peak maximum clearly lies in the infrared range at usual lab-accessible temperatures (T < 2000 K), the short-wavelength tail of the black-body spectrum can already interfere with the luminescence spectrum in the visible range at temperatures as low as 800 K. Thus, any ratiometric luminescent thermometer devised for high temperature sensing best relies on emission in the ultraviolet range, which is clearly unaffected by black-body background at temperatures below 1200 K.
In this work, we present such a designed multilevel luminescent thermometer that uses the three excited 6PJ (J = 7/2, 5/2, 3/2) crystal fields and spin-orbit levels of the UV B-emitting lanthanide ion Gd3+; we were motivated by thermodynamic considerations and demonstrated how a single luminescent phosphor can be optimized for precise thermometry from cryogenic (30 K) to high temperatures (800 K) with constantly high relative sensitivities above 0.5% K−1. Intense excitation with a cost-effective blue wavelength of 450 nm is interesting for practical applications. This is possible by co-doping with Pr3+ in huntite-type crystallizing YAl3(BO3)4 (YAB)59 and upconversion into its 4f15d1 configuration60,61,62,63,64, followed by efficient energy transfer to the 6IJ’ (J’ = 17/2…7/2) levels of Gd3+. Not only does this offer the possibility of background free upconversion thermometry, but it also bears potential for applications such as UV lasing64 or visible light excited photocatalysis65 combined with local temperature sensing. Similar visible-to-UV upconversion relying on triplet-triplet annihilation has otherwise been recently reported by Harada et al. for metal-organic compounds66
Results
Thermometric performance of an excited three-level system
The benefit of a third emissive level on the performance of a luminescent thermometer can be appreciated as follows. The thermal response (often also referred to as absolute sensitivity), Sa(T), of the luminescence intensity ratio (LIR) is directly related to the absolute effective thermal population of the microstate of the higher excited level from which the emission arises. The relative sensitivity, Sr(T), in contrast, relates to the relative net change in population between the two thermally coupled excited levels of interest55. As indicated above, the thermodynamically optimized performance of a two-excited level thermometer is limited to a small temperature range only. This performance is related to the low response at low temperatures due to the almost negligible thermal population in the higher excited state and a low relative sensitivity at high temperatures due to almost equalized populations (per microstate) in both thermally coupled excited levels. The optimum temperature range depends on the probed energy gap ∆E21 and is given by55
A simple way to widen this optimum temperature range is the addition of a third excited level |3〉 separated by an energy gap ∆E31 = (1 + s)∆E21 from the lowest excited level |1〉. According to Eq. (1), a widening of the optimum temperature range can be achieved if \(s \ge \sqrt 2 /2 \approx 0.707\) and ∆E31 is used as the thermometric measure above the critical temperature \(T^{\prime} = \frac{{\Delta E_{21}}}{{2k_{{{\rm{B}}}}}}\).
This proposal is schematically depicted in Fig. 1a–c. Above \(T^{\prime}\), the relative net change in population between levels |3〉 and |1〉 becomes higher than the respective net change between levels |2〉 and |1〉, resulting in a higher relative sensitivity of the thermometer upon the exploitation of the higher energy gap. This strategy can also be generalized to more than three levels in an iterative manner.
Boltzmann cryothermometry with Gd3+
The previously introduced thermodynamic concept of a multilevel thermometer can be realized with Gd3+. Here, we use the small energy splitting between different crystal field components of the 6P7/2 level for temperature sensing in the low-temperature regime. In the next section, the larger splitting between 6P7/2 and higher energetic spin-orbit levels 6P5/2 and 6P3/2 for accurate measurements at higher temperatures is exploited instead. A schematic energy level diagram illustrating the working principle of crystal field component-based Boltzmann cryothermometry with Gd3+ is depicted in Fig. 2a. For cryogenic temperatures (4 K < T < 100 K), Eq. (1) suggests that the most suitable energy gap for the performance of Boltzmann thermometry is on the order of 50−100 cm−1. This energy range is typical for the splitting of 4fn-based spin-orbit levels by the surrounding crystal field potential, although other examples, such as the R lines of Cr3+ doped into hosts with strong crystal fields, also have energy differences on this order of magnitude67,68,69.
Gd3+ is a very weak absorber70,71. In addition, efficient and intense UV light sources are not readily available, and direct UV excitation also gives rise to background emission. We thus chose a different strategy by co-doping the sample with Pr3+, which is schematically depicted in Fig. 2b. Upon excitation with a pulsed optical parametric oscillator (OPO) in the blue range at 448 nm, the Pr3+ ions are excited to the 3P2 level. After quick temperature-independent nonradiative relaxation to the 1D2 level given the high phonon energies in YAB (\(\hbar \omega _{\max }\sim 1300\,{{{\rm{cm}}}}^{ - 1}\))72, electric dipole-allowed excited state absorption (ESA) into the 4f15d1 configuration of Pr3+ at energies in the range of 40,000 cm−1 can take place72, which is followed by an energy transfer to the 6IJ’ levels of Gd3+ at energies of approximately 37000 cm−1 (270 nm). The dependence of the Gd3+-related emission intensity on the incident power of the OPO also indicates a two-photon upconversion process (see Supplementary Information, Fig. S1). Similar upconversion spectra were, however, also measured using an inexpensive continuous wave (CW) blue laser with 1 W output power (see Supplementary Information, Fig. S2), which demonstrates that this concept also works with any high-power LED.
Since Pr3+ is very prone to the resonant cross-relaxation process [Pr1, Pr2]: [3P0, 3H4] → [1G4, 1G4], it was mandatory to keep the Pr3+ concentration at only 0.7 mol% (see also Supplementary Information, Fig. S3). The large energy gap between the 6PJ (J = 7/2…3/2) levels and the 8S7/2 ground level of Gd3+, which does not have any intermediate electronic levels, as well as the high mutual lanthanide distances in the YAB host structure (~ 5.9 Å)59, allow an increase in the Gd3+ concentration to 20 mol% without any observable concentration quenching effects (see Supplementary Information, Fig. S4). The high energy gap increases the probability of efficient energy transfer from Pr3+. Fig. 2c depicts a portion of the luminescence spectra of YAB: 0.7% Pr3+ and 20% Gd3+ appear in the UV-B range and clearly show that the strategy to excite Gd3+ via Pr3+-based upconversion and subsequent energy transfer does work. The absence of a 4f15d1 → 4f2 broad-band emission of Pr3+ also shows that the energy transfer from Pr3+ to Gd3+ must be very efficient, which is in agreement with earlier findings73.
Boltzmann thermometry with Gd3+ at cryogenic temperatures becomes possible if the two clearly resolved components of the 6P7/2 → 8S7/2-based group of radiative transitions are chosen (see Fig. 2a). These transitions give rise to two intense emission peaks located at 310.6 nm and 311.3 nm (see Fig. 2b), whose intensity ratio (higher energetic to lower energetic emission) is used as the temperature-correlated signal. The LIR at temperature T, R(T), normalized to its value at the highest measured reference temperature, T0, thus follows the law23,74
where kB is the Boltzmann constant and ∆Em1 (m > 1) is the energy gap between the two thermally coupled levels and is the only fitting parameter. A fit to a linearized version of Eq. (2) gave an effective energy gap of ∆ECF = (69 ± 5) cm−1, in excellent agreement with the value of 72 cm−1 determined from the emission spectra. It should be noted that a corresponding splitting of the 8S7/2 level into Kramers’ doublets is much smaller (on the order of only 1 cm−1) if spin-orbit coupling is only intermediate because of the lack of orbital degeneracy given the quantum number L = 075. This makes the 8S7/2 ground level an effective single level in the case of Gd3+.
The LIR deviates from the expected Boltzmann behaviour at temperatures below 25 K. This observation is related to a kinetic limitation. The nonradiative absorption rate from the lower to the higher excited level, \(k_{{{{\rm{nr}}}}}^{{{{\rm{abs}}}}}\left( T \right)\), is given by76,77
with gm as the degeneracy of the higher excited level (gm = 2 for Kramers’ doublets), knr(0) as an intrinsic nonradiative rate governed by the properties of the electronic transition, and
as the thermally averaged phonon occupation number of an acoustic or optical phonon mode with energy \(\hbar \omega _{{{{\rm{ph}}}}}\) and p as the number of phonons consumed during the thermalization between the two excited levels (∆Em1 = \(p\hbar \omega _{{{{\rm{ph}}}}}\), here it is p = 1). This temperature-dependent nonradiative absorption rate competes with the radiative decay (and eventually other quenching) rates from level |1〉. At sufficiently low temperatures, the nonradiative absorption rate decreases than the intrinsic radiative decay rate, and thus, Boltzmann thermalization is kinetically inhibited. We have recently shown that a very accurate estimate for the expected onset temperature of Boltzmann behaviour, Ton, is given by55
With the spectroscopic value of ∆ECF = 72 cm−1, the expected onset temperature for Boltzmann thermalization between the two excited levels is thus approximately 23 K, very close to the observed deviation from Boltzmann behaviour below 25 K. A similar deviation has also been observed in the case of Cr3+67,78, which can be explained by the same kinetic effect. In turn, the optimum temperature range for the most precise thermometry that exploits this energy gap between the two Kramers’ doublets of the 6P7/2 level of Gd3+ is between 30 and 51 K according to Eq. (1). The connected relative sensitivities, Sr(T), as defined by
in the case of a Boltzmann thermometer that varies between Sr(30 K) = 11.6% K−1 and Sr(51 K) = 3.98% K−1 in this optimum range, which is clearly above the usually desired threshold of Sr = 1% K−1 in practice23. In fact, the relative sensitivity at 30 K is the highest reported value for luminescent cryothermometers at that particular temperature thus far50,67,68,79,80. It is important to note, however, that it is both the balance between reliably detectable luminescence signals and relative sensitivity that determine the overall precision of a luminescent thermometer (see also below)55,81. Another important point is that for accurate cryothermometry with small energy gaps, a high spectral resolution spectrometer is also required to separate two emission lines that are close in wavelength. Finally, that high relative sensitivities at cryogenic temperatures are not surprising, as can already be appreciated from the T−2 dependence in Eq. (6) that dominates at low temperatures and stems from the strong relative net change in population between the two thermally coupled excited levels in this domain (see Fig. 1b).
Extension of high-precision thermometry with Gd3+ to higher temperatures
The small crystal field splitting between the Kramers’ doublets stemming from the 6P7/2 spin-orbit level of Gd3+ is not well suited for temperature measurements above 100 K since the excited state populations in the two thermally coupled crystal field states equalize at elevated temperatures. This fact in turn lowers both the response and sensitivity of the chosen radiative transitions for temperature sensing. For thermometry in the range of room temperature, it is necessary to employ two levels with a higher energy gap. According to Eq. (1), a suitable energy gap to measure temperatures between 200 and 400 K is on the order of 500–700 cm−1, in the range of spin-orbit interaction-induced splitting for the 4fn-based electronic levels of lanthanides. Consequently, we used the LIR of the radiative transitions from the 6P7/2 and 6P5/2 spin-orbit levels to the 8S7/2 ground level (see Fig. 3a) of Gd3+ to demonstrate its applicability for temperature sensing in the range of room temperature (see Fig. 3b). The LIR of the 6P5/2 to 6P7/2 emission lines was measured as a function of temperature as determined from the emission spectra shown in Fig. 3b and fit to Eq. (2) (Fig. 3c). The results indicated an effective energy gap between the 6P7/2 and 6P5/2 levels of ∆E21 = (502 ± 11) cm−1, again in excellent agreement with the value of 506 cm−1 that was spectroscopically deduced from the emission spectra. With this energy gap, the most suitable temperature range for high-precision luminescence thermometry is between 213 and 364 K with corresponding relative sensitivities of Sr(213 K) = 1.60% K−1 and Sr(364 K) = 0.55% K−1. These relative sensitivities are almost one order of magnitude higher than the corresponding sensitivities for thermometry with crystal field splitting of 72 cm−1 in this temperature range. This result demonstrates that the concept of multiple excited states with different energy separations could realize high-temperature sensing accuracy over a wider temperature range, given sufficiently high photon counts of the emission bands of interest. As in the case of cryothermometry, the LIR deviates from Boltzmann behaviour below 200 K, which is a consequence of the very slow nonradiative absorption in that temperature range53,54,55. This observation also coincides very well with the estimated onset temperature for Boltzmann behaviour at Ton = 163 K according to Eq. (5) for an energy gap of 506 cm−1.
An extension of the concept of optimized luminescence thermometry with Gd3+ to even higher temperatures is finally possible by the exploitation of thermal coupling between the excited 6P3/2 and 6P7/2 levels (Fig. 3d). The corresponding LIR is clearly only a useful temperature measure at very high temperatures due to the generally low intensity of the 6P3/2 → 8S7/2-based transition located at 295 nm (inset of Fig. 3b). In particular, this application also demonstrates why Gd3+ is a particularly useful choice as a luminescent thermometer for high-temperature sensing. Since there are no lower-lying energy levels between the 6P7/2 level located approximately 32000 cm−1 above the 8S7/2 ground state, thermal quenching of the Gd3+-based luminescence in a wide bandgap host compound such as YAB cannot take place. Fitting the temperature dependence of the corresponding LIR that stems from the two excited levels 6P3/2 and 6P7/2 to Eq. (2) (Fig. 3d) gives an effective energy gap of ∆E31 = (1136 ± 81) cm−1, again in excellent agreement with the spectroscopically determined value of 1120 cm−1. This energy gap is optimally suited to measure temperatures between 472 and 805 K with corresponding relative sensitivities of Sr(472 K) = 0.73% K−1 and Sr(805 K) = 0.25% K−1, which is an improvement by a factor of 2.2 compared to thermometry with the 6P5/2−6P7/2 gap in this temperature range. Here, again, thermalization and the sustainment of a Boltzmann equilibrium between the 6P3/2 and 6P7/2 spin-orbit levels of Gd3+ is only observed above temperatures of at least 450 K, while Eq. (5) suggests an expected onset temperature of approximately 360 K. The discrepancy between observed and estimated onset is related to the weak intensity of the 6P3/2 → 8S7/2 transition that makes it difficult to accurately determine the LIR below 450 K.
The photoluminescence spectra reveal that the intensity of the radiative emission from the 6P3/2 spin-orbit level of Gd3+ is very weak compared to the emission from the 6P7/2 level. According to calculations by Detrio on Gd3+, the 6P3/2 → 8S7/2 transition is generally characterized by very low intensities irrespective of the chosen host material given the very low reduced matrix elements governing the intensities of dipole-allowed electronic transitions in Judd−Ofelt theory82. This trend raises the question of whether the LIR between the radiative transitions from the thermally coupled 6P3/2 and 6P7/2 levels actually offers a more precise temperature measure based on the higher relative sensitivity Sr(T) or whether continuous exploitation of the LIR between the transitions from the 6P5/2 and 6P7/2 levels up to higher temperatures is more favourable based on the higher luminescence intensity of the 6P5/2 → 8S7/2 radiative transition. For that purpose, we calculated the minimum theoretical relative temperature uncertainty for both thermometry measures within the optimum temperature range of the 6P3/2−6P7/2 gap (472−805 K). The relative temperature precision of a ratiometric Boltzmann-based thermometer is given by55
where I10 denotes the intensity (in integrated photon counts) of the lower energetic emission and \(R\left( \infty \right)\) is the extrapolated LIR at infinite temperatures that can be obtained from the respective Boltzmann fits. Graphs depicting the evolution of both the relative sensitivity and the relative temperature uncertainty at different temperatures are shown in Fig. 4a, b. They clearly indicate that upon setting the same photon count number of I10 = 107 counts of the 6P7/2 → 8S7/2-based emission, the LIR between the radiative transitions from 6P5/2 and 6P7/2 still gives rise to a lower relative temperature uncertainty (ca. 0.08%) based on the higher relative intensity of the 6P5/2 → 8S7/2 transition than the corresponding LIR involving the radiative transition from the higher energetic 6P3/2 level (relative temperature uncertainty of ca. 0.2%). In that case, it is advisable to still retain the LIR stemming from the adjacent 6P5/2 and 6P7/2 levels. Thus, it depends on the trade-off between practically achievable intensities and thermodynamically guided optimum temperature range, which is the more advisable choice as a temperature measure to probe high temperatures with Gd3+ in YAB: Pr3+, Gd3+. Irrespective of the choice, it is possible to keep the absolute temperature uncertainty below a threshold of 1.3 K even at 820 K. We do stress, however, that these relative temperature uncertainties are purely statistical and do not include any additional systematic error that could occur in a real-case application. It is also important that these low uncertainties require a high integrated intensity (~107 cts) of the 6P7/2 → 8S7/2-related emission of Gd3+, which may be difficult to achieve in an application based on the proposed upconversion.
Discussion
Single ion luminescence thermometry based on a Boltzmann equilibrium between two emitting excited levels is both thermodynamically and kinetically limited to work with optimized precision only within a certain temperature range. If a luminescent thermometer is supposed to cover wider temperature ranges, thermal coupling between more than two excited states must be exploited. In this work, we offer quantitative guidelines based on a recently developed theoretical framework that allows us to extend the usability of a Boltzmann-based luminescent thermometer over wider temperature ranges by carefully choosing the energy gaps between multiple thermally coupled levels. We demonstrate the concept practically for the case of Gd3+, which exhibits three energetically well-isolated 6PJ (J = 7/2, 5/2, 3/2) levels in the UV range that can additionally rely on splitting in a local crystal field. Gd3+ is thus a prime choice for especially high-temperature ratiometric luminescence thermometry, as the black-body background does not interfere in this emission range. Moreover, it absorbs and emits in the solar-blind spectral range, and thus, any background signal from ambient sunlight is also avoided. Since Gd3+ is a weak absorber, we demonstrated an alternative excitation scheme with a cost-effective and well-established excitation wavelength of 450 nm available from blue LEDs via Pr3+ in a blue-to-UV upconversion process. By ground state absorption into the 3P2 level, followed by efficient excited-state absorption into the 4f15d1 configuration of Pr3+, it is possible to indirectly excite Gd3+ by energy transfer into its 6IJ’ (J’ = 17/2…7/2) levels, from which the excited state population finally undergoes nonradiative relaxation into the 6PJ levels. Optimized Boltzmann cryothermometry between 30 and 50 K was possible with two crystal field components of the excited 6P7/2 level separated by 72 cm−1. Temperature sensing in the range of room temperature is better performed by exploiting the LIR of the radiative transitions stemming from the 6P5/2 and 6P7/2 spin-orbit levels of Gd3+ separated by 506 cm−1. Finally, high-temperature thermometry between 470 and 800 K is also possible using Boltzmann thermalization between the 6P3/2 and 6P7/2 levels of Gd3+, with an energy separation of 1120 cm−1. Both the 6P3/2 → 8S7/2- and the 6P5/2 → 8S7/2-based emissions were compared in terms of the overall achievable temperature precision of the respective ratiometric thermometry approaches. While the 6P3/2 → 8S7/2-based emission is very low in intensity, it offers the necessary high relative sensitivity at high temperatures. It was shown that the final choice for high-temperature thermometry with Gd3+ depends on the practically achievable integrated intensity of the lowest energetic 6P7/2 → 8S7/2 transition.
In summary, a theoretical framework is presented to quantitatively understand and design single ion thermometers relying on multiple thermally coupled excited states to widen the window for accurate thermal sensing. The concept is shown to be feasible using a Pr3+, Gd3+-coactivated YAl3(BO3)4 phosphor with Gd3+ as a Boltzmann-based single ion luminescent thermometer. UV upconversion emission from the 6PJ levels of Gd3+ is sensitized via two-step excitation of Pr3+ at 450 nm to the states of the 4f15d1 configuration, followed by energy transfer to Gd3+, which is practically feasible with the use of high-power blue LEDs. Analysis of the temperature-dependent emission demonstrates how simple thermodynamic principles can be applied to circumvent the fundamental limitations in temperature precision of a conventional two-level single ion luminescent thermometer to finally make one thermometry system applicable to a wide range of temperatures while retaining high precision. This work paves the way for the targeted design of luminescent thermometers tailored towards applicational requirements for temperature accuracy within specific temperature ranges.
Methods
Chemical reagents
Y(NO3)3 ∙ 6 H2O (Alfa Aesar, Germany, 99.99%), Al(NO3)3 ∙ 9 H2O (Sigma–Aldrich, Germany, ≥ 98%), Pr(NO3)3 ∙ 6 H2O (Strem Chemicals, France, 99.9%), Gd(NO3)3 ∙ 6 H2O (Sigma–Aldrich, Germany, 99.99%), H3BO3 (Merck, Germany, 99.8%) and urea (Sigma–Aldrich, Germany, ≥ 98%) were used without further purification.
Synthesis of YAl3(BO3)4 (YAB): x% Pr3+, y% Gd3+ microcrystalline powder
Microcrystalline powder samples of YAB: x% Pr3+, y% Gd3+ (x = 0.1, 0.3, 0.5, 0.7, 1.0, and 2.0; y = 5, 10, 20, 40, and 60) were prepared by a modified urea–nitrate solution-based combustion route83. Stoichiometric amounts of Ln(NO3)3 ∙ 6 H2O (Ln = Y, Pr, and Gd) and Al(NO3)3 ∙ 9 H2O were dissolved in distilled H2O (approximately 40 ml) under vigorous stirring at room temperature. Subsequently, solid urea (molar ratio urea/lanthanide ions = 3:1) and H3BO3 (5 mol% excess) were added to the transparent nitrate-containing solution. The solution was heated to 80 °C to improve the solubility of the H3BO3 and constantly stirred for 30 min at that temperature. The beaker was covered with parafilm to prevent excessive solvent evaporation. The resulting solution was quickly transferred to an alumina crucible and placed into a preheated furnace at 500 °C in air for 10 min to initiate combustion. The thus-formed colourless solid precursor was carefully ground with additional solid H3BO3 (half of the previous stoichiometric amount) in a mortar to account for the losses that occurred during the combustion step. The solid mixture was finally sintered at 1100 °C for 6 h in air. After naturally cooling to room temperature, the obtained colourless residue was ground to a fine powder, and its phase purity was verified by powder X-ray diffraction (Philips PW391, Cu Kα1 radiation (λ = 1.54056 Å), U = 40 kV, I = 20 mA, reflection mode). The X-ray diffraction pattern was scanned in a 2θ range between 10° and 80° with a step size of 0.02° (see Supplementary Information, Fig. S5). Morphology and energy dispersive X-ray spectroscopy characterizations were performed using scanning electron microscopy (SEM; Nova, NANO SEM 430; see Supplementary Information, Fig. S6).
Optical spectroscopy and temperature-dependent measurements
Excitation was performed with an external pulsed wavelength tuneable Opotek Opolette HE 355 II (Carlsbad, CA, USA) OPO pumped by a frequency-tripled Nd:YAG laser at a repetition rate of 20 Hz and a temporal pulse width of approximately 6 ns. Emission spectra were acquired on an Edinburgh FLS920 spectrofluorometer (Livingston, UK) equipped with a 0.25 m double Littrow-configuration grating monochromator blazed at 300 nm and a Hamamatsu R928 (Shizuoka, Japan) photomultiplier tube (PMT) for photon detection. All emission spectra were corrected for grating efficiency and detector sensitivity. Temperature-dependent measurements below room temperature were performed with an Oxford Instruments liquid He flow cryostat (Oxford, UK) and an external temperature control unit, which measured the temperature by means of a thermocouple in direct contact with the powder sample holder. High-temperature luminescence spectra were acquired by placing the sample into an externally water-cooled Linkam (Surrey, UK) THMS600 microscope stage (±0.1 °C temperature stability). The temperature was externally controlled by a thermocouple in immediate contact with the sample holder. Photoluminescence decay curves were acquired by pulsed excitation with the OPO and detection of the time-resolved signal with a multichannel scaler (MCS) attached to a Hamamatsu H7422 PMT (Shizuoka, Japan) for minimized background. For the demonstration of the excitation of the upconverted emission of Gd3+ with blue light, a CW laser of 450 nm and maximum output power of 1 W (Changchun New Industries Optoelectronics Technology, Changchun, China) was employed as an excitation source.
References
Lahiri, B. B. et al. Medical applications of infrared thermography: a review. Infrared Phys. Technol. 55, 221–235 (2012).
Bagavathiappan, S. et al. Infrared thermography for condition monitoring—a review. Infrared Phys. Technol. 60, 35–55 (2013).
Xiao, Y. Z. et al. Precision measurements of temperature-dependent and nonequilibrium thermal emitters. Laser Photonics Rev. 14, 1900443 (2020).
Brites, C. D. S. et al. Thermometry at the nanoscale. Nanoscale 4, 4799–4829 (2012).
Jaque, D. & Vetrone, F. Luminescence nanothermometry. Nanoscale 4, 4301–4326 (2012).
Dramićanin, M. Luminescence Thermometry: Methods, Materials, and Applications (Woodhead Publishing, 2018).
Allison, S. W. & Gillies, G. T. Remote thermometry with thermographic phosphors: instrumentation and applications. Rev. Sci. Instrum. 68, 2615–2650 (1997).
Allison, S. W. et al. Nanoscale thermometry via the fluorescence of YAG:Ce phosphor particles: measurements from 7 to 77 °C. Nanotechnology 14, 859–863 (2003).
Allison, S. W. A brief history of phosphor thermometry. Meas. Sci. Technol. 30, 072001 (2019).
Kucsko, G. et al. Nanometre-scale thermometry in a living cell. Nature 500, 54–58 (2013).
Neumann, P. et al. High-precision nanoscale temperature sensing using single defects in diamond. Nano Lett. 13, 2738–2742 (2013).
Wang, N. et al. Magnetic criticality enhanced hybrid nanodiamond thermometer under ambient conditions. Phys. Rev. X 8, 011042 (2018).
Walker, G. W. et al. Quantum-dot optical temperature probes. Appl. Phys. Lett. 83, 3555–3557 (2003).
Ortega-Rodríguez, A. et al. 10-fold quantum yield improvement of Ag2S nanoparticles by fine compositional tuning. ACS Appl. Mater. Interfaces 12, 12500–12509 (2020).
Santos, H. D. A. et al. Ultrafast photochemistry produces superbright short-wave infrared dots for low-dose in vivo imaging. Nat. Commun. 11, 2933 (2020).
Yakunin, S. et al. High-resolution remote thermometry and thermography using luminescent low-dimensional tin-halide perovskites. Nat. Mater. 18, 846–852 (2019).
Morad, V. et al. Hybrid 0D antimony halides as air-stable luminophores for high-spatial-resolution remote thermography. Adv. Mater. 33, 2007355 (2021).
Gota, C. et al. Hydrophilic fluorescent nanogel thermometer for intracellular thermometry. J. Am. Chem. Soc. 131, 2766–2767 (2009).
Okabe, K. et al. Intracellular temperature mapping with a fluorescent polymeric thermometer and fluorescence lifetime imaging microscopy. Nat. Commun. 3, 705 (2012).
Sakaguchi, R., Kiyonaka, S. & Mori, Y. Fluorescent sensors reveal subcellular thermal changes. Curr. Opin. Biotechnol. 31, 57–64 (2015).
Uchiyama, S. & Gota, C. Luminescent molecular thermometers for the ratiometric sensing of intracellular temperature. Rev. Anal. Chem. 36, 20160021 (2017).
Brites, C. D. S., Millán, A. & Carlos, L. D. Lanthanides in luminescent thermometry. Handb. Phys. Chem. Rare Earths 49, 339–427 (2016).
Brites, C. D. S., Balabhadra, S. & Carlos, L. D. Lanthanide-based thermometers: at the cutting-edge of luminescence thermometry. Adv. Opt. Mater. 7, 1801239 (2019).
Souza, A. S. et al. Highly-sensitive Eu3+ ratiometric thermometers based on excited state absorption with predictable calibration. Nanoscale 8, 5327–5333 (2016).
Trejgis, K., Bednarkiewicz, A. & Marciniak, L. Engineering excited state absorption-based nanothermometry for temperature sensing and imaging. Nanoscale 12, 4667–4675 (2020).
Trejgis, K. et al. Nd3+ doped TZPN glasses for NIR operating single band ratiometric approach of contactless temperature readout. J. Lumin. 224, 117295 (2020).
Brites, C. D. S. et al. Instantaneous ballistic velocity of suspended Brownian nanocrystals measured by upconversion nanothermometry. Nat. Nanotechnol. 11, 851–856 (2016).
Bastos, A. et al. Thermal properties of lipid bilayers determined using upconversion nanothermometry. Adv. Funct. Mater. 29, 1905474 (2019).
Caixeta, F. J. et al. High-quantum-yield upconverting Er3+/Yb3+-organic–inorganic hybrid dual coatings for real-time temperature sensing and photothermal conversion. J. Phys. Chem. C 124, 19892–19903 (2020).
del Rosal, B. et al. Nd3+ ions in nanomedicine: perspectives and applications. Optical Mater. 63, 185–196 (2017).
del Rosal, B. et al. In vivo luminescence nanothermometry: from materials to applications. Adv. Opt. Mater. 5, 1600508 (2017).
Piñol, R. et al. Real-time intracellular temperature imaging using lanthanide-bearing polymeric micelles. Nano Lett. 20, 6466–6472 (2020).
Ximendes, E. C. et al. Thulium doped LaF3 for nanothermometry operating over 1000 nm. Nanoscale 11, 8864–8869 (2019).
Skripka, A. et al. Advancing neodymium single-band nanothermometry. Nanoscale 11, 11322–11330 (2019).
Hazra, C. et al. Erbium single-band nanothermometry in the third biological imaging window: potential and limitations. Adv. Opt. Mater. 8, 2001178 (2020).
Geitenbeek, R. G. et al. In situ luminescence thermometry to locally measure temperature gradients during catalytic reactions. ACS Catal. 8, 2397–2401 (2018).
van Ravenhorst, I. K. et al. In situ local temperature mapping in microscopy nano-reactors with luminescence thermometry. ChemCatChem 11, 5505–5512 (2019).
Hartman, T. et al. Operando monitoring of temperature and active species at the single catalyst particle level. Nat. Catal. 2, 986–996 (2019).
Geitenbeek, R. G. et al. Luminescence thermometry for in situ temperature measurements in microfluidic devices. Lab Chip 19, 1236–1246 (2109).
Hartman, T. et al. Operando nanoscale sensors in catalysis: all eyes on catalyst particles. ACS Nano 14, 3725–3735 (2020).
Bednarkiewicz, A. et al. Standardizing luminescence nanothermometry for biomedical applications. Nanoscale 12, 14405–14421 (2020).
Dramićanin, M. D. Trends in luminescence thermometry. J. Appl. Phys. 128, 040902 (2020).
Bednarkiewicz, A. et al. Luminescence based temperature bio-imaging: status, challenges, and perspectives. Appl. Phys. Rev. 8, 011317 (2021).
Ximendes, E. et al. Quo vadis, nanoparticle-enabled in vivo fluorescence imaging? ACS Nano 15, 1917–1941 (2021).
van Swieten, T. P. et al. Mapping elevated temperatures with a micrometer resolution using the luminescence of chemically stable upconversion nanoparticles. ACS Appl. Nano Mater. 4, 4208–4215 (2021).
Labrador-Páez, L. et al. Reliability of rare-earth-doped infrared luminescent nanothermometers. Nanoscale 10, 22319–22328 (2018).
Shen, Y. L. et al. In vivo spectral distortions of infrared luminescent nanothermometers compromise their reliability. ACS Nano 14, 4122–4133 (2020).
Errulat, D. et al. A luminescent thermometer exhibiting slow relaxation of the magnetization: toward self-monitored building blocks for next-generation optomagnetic devices. ACS Cent. Sci. 5, 1187–1198 (2019).
Gálico, D. A. et al. Triplet-state position and crystal-field tuning in opto-magnetic lanthanide complexes: two sides of the same coin. Chem. A Eur. J. 25, 14625–14637 (2019).
Kaczmarek, A. M. et al. Developing luminescent ratiometric thermometers based on a covalent organic framework (COF). Angew. Chem. Int. Ed. 59, 1932–1940 (2020).
Gomez, G. E. et al. Tunable energy-transfer process in heterometallic MOF materials based on 2,6-naphthalenedicarboxylate: solid-state lighting and near-infrared luminescence thermometry. Chem. Mater. 32, 7458–7468 (2020).
Kaczmarek, A. M. et al. Visible and NIR upconverting Er3+−Yb3+ luminescent nanorattles and other hybrid PMO-inorganic structures for in vivo nanothermometry. Adv. Funct. Mater. 30, 2003101 (2020).
Geitenbeek, R. G., de Wijn, H. W. & Meijerink, A. Non-Boltzmann luminescence in NaYF4:Eu3+: implications for luminescence thermometry. Phys. Rev. Appl. 10, 064006 (2018).
Suta, M. et al. Making Nd3+ a sensitive luminescent thermometer for physiological temperatures—an account of pitfalls in Boltzmann thermometry. Nanomaterials 10, 543 (2020).
Suta, M. & Meijerink, A. A theoretical framework for ratiometric single ion luminescent thermometers—thermodynamic and kinetic guidelines for optimized performance. Adv. Theory Simul. 3, 2000176 (2020).
Tian, X. N. et al. Temperature sensor based on ladder-level assisted thermal coupling and thermal-enhanced luminescence in NaYF4:Nd3+. Opt. Express 22, 30333–30345 (2014).
Ćirić, A. et al. Comparison of three ratiometric temperature readings from the Er3+ upconversion emission. Nanomaterials 10, 627 (2020).
Li, L. P. Three-energy-level-cascaded strategy for a more sensitive luminescence ratiometric thermometry. Sens. Actuators A: Phys. 304, 111864 (2020).
Belokoneva, E. L. et al. The crystal structure of YAl3(BO3)4. Zh. Strukturnoi Khimii 22, 196–198 (1981).
Hu, C. H. et al. Visible-to-ultraviolet upconversion in Pr3+:Y2SiO5 crystals. Chem. Phys. 325, 563–566 (2006).
Cates, E. L. & Kim, J. H. Upconversion under polychromatic excitation: Y2SiO5:Pr3+, Li+ converts violet, cyan, green, and yellow light into UVC. Optical Mater. 35, 2347–2351 (2013).
Cates, E. L., Wilkinson, A. P. & Kim, J. H. Visible-to-UVC upconversion efficiency and mechanisms of Lu7O6F9:Pr3+ and Y2SiO5:Pr3+ ceramics. J. Lumin. 160, 202–209 (2015).
Cates, E. L. & Li, F. F. Balancing intermediate state decay rates for efficient Pr3+ visible-to-UVC upconversion: the case of β-Y2Si2O7:Pr3+. RSC Adv. 6, 22791–22796 (2016).
Du, Y. Y. et al. Blue-pumped deep ultraviolet lasing from lanthanide-doped Lu6O5F8 upconversion nanocrystals. Adv. Opt. Mater. 8, 1900968 (2020).
Sahu, S. P. et al. The myth of visible light photocatalysis using lanthanide upconversion materials. Environ. Sci. Technol. 52, 2973–2980 (2018).
Harada, N. et al. Discovery of Key TIPS-naphthalene for efficient visible-to-UV photon upconversion under sunlight and room light. Angew. Chem. Int. Ed. 60, 142–147 (2021).
Back, M. et al. Effective ratiometric luminescent thermal sensor by Cr3+-doped mullite Bi2Al4O9 with robust and reliable performances. Adv. Optical Mater. 8, 2000124 (2020).
Back, M. et al. Pushing the limit of Boltzmann distribution in Cr3+-doped CaHfO3 for cryogenic thermometry. ACS Appl. Mater. Interfaces 12, 38325–38332 (2020).
Mykhaylyk, V. et al. Multimodal non-contact luminescence thermometry with Cr-doped oxides. Sensors 20, 5259 (2020).
Sytsma, J., Imbusch, G. F. & Blasse, G. The spectroscopy of Gd3+ in yttriumoxychloride: Judd–Ofelt parameters from emission data. J. Chem. Phys. 91, 1456–1461 (1989).
Sytsma, J. & Blasse, G. Comparison of the emission of Eu2+ in MFCl (M = Sr, Ba) and Gd3+ in YOCl. J. Lumin. 51, 283–292 (1992).
Bartl, M. H. et al. Growth, optical spectroscopy and crystal field investigation of YAl3(BO3)4 single crystals doped with tripositive praseodymium. Spectrochim. Acta Part A: Mol. Biomol. Spectrosc. 57, 1981–1990 (2001).
de Vries, A. J. & Blasse, G. On the possibility to sensitize Gd3+ luminescence by the Pr3+ ion. Mater. Res. Bull. 21, 683–694 (1986).
Balabhadra, S. et al. Upconverting nanoparticles working as primary thermometers in different media. J. Phys. Chem. C 121, 13962–13968 (2017).
Sytsma, J. et al. Spectroscopic studies and crystal-field analysis of Cm3+ and Gd3+ in LuPO4. Phys. Rev. B 52, 12668–12676 (1995).
Riseberg, L. A. & Moos, H. W. Multiphonon orbit-lattice relaxation in LaBr3, LaCl3, and LaF3. Phys. Rev. Lett. 19, 1423–1426 (1967).
Riseberg, L. A. & Moos, H. W. Multiphonon orbit-lattice relaxation of excited states of rare-earth ions in crystals. Phys. Rev. J. Arch. 174, 429–438 (1968).
Back, M. et al. Boltzmann thermometry in Cr3+-doped Ga2O3 polymorphs: the structure matters! Adv. Opt. Mater. 9, 2100033 (2021).
Liu, X. et al. Mixed-lanthanoid metal-organic framework for ratiometric cryogenic temperature sensing. Inorg. Chem. 54, 11323–11329 (2015).
Ananias, D. et al. Near-infrared ratiometric luminescent thermometer based on a new lanthanide silicate. Chem. A Eur. J. 24, 11926–11935 (2018).
Abram, C. et al. ScVO4:Bi3+ thermographic phosphor particles for fluid temperature imaging with sub-°C precision. Opt. Lett. 45, 3893–3896 (2020).
Detrio, J. A. Line strengths for Gd3+ at a C4v site in SrF2. Phys. Rev. B 4, 1422–1427 (1971).
Yu, D. C. et al. Understanding and tuning blue-to-near-infrared photon cutting by the Tm3+/Yb3+ couple. Light.: Sci. Appl. 9, 107 (2020).
Acknowledgements
D.Y. is grateful to Prof. Wei Wei from Nanjing University of Posts and Telecommunication for useful suggestions. Moreover, Q.Z. acknowledges funding from the National Science Foundation of China (U1601205, 51472088, and 51125005), while D.Y. and A.M. acknowledge funding from the Nederlandse Organisatie voor Wetenschappelijk Onderzoek (NWO, Grant No. 731.017.302). A.M. and M.S. additionally gratefully acknowledge funding from the EU-Horizon 2020 FET-Open project NanoTBTech (grant agreement no.: 801305).
Funding
Open Access funding enabled and organized by Projekt DEAL.
Author information
Authors and Affiliations
Contributions
A.M. and M.S. conceived the thermometry approach, while D.Y., A.M., and M.S. gave the input for the indirect upconversion excitation scheme. D.Y., H.L., and M.S. performed the synthesis, characterization, and optical experiments. D.Z. and Q.Z. contributed to the interpretation of the analysis results. M.S. provided the theoretical model and thermodynamic guidelines and performed the thermometric data analysis. D.Y., A.M., and M.S. wrote the manuscript. All authors contributed to the interpretation and discussion of the results.
Corresponding authors
Ethics declarations
Conflict of interest
The authors declare no conflict of interest.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Yu, D., Li, H., Zhang, D. et al. One ion to catch them all: Targeted high-precision Boltzmann thermometry over a wide temperature range with Gd3+. Light Sci Appl 10, 236 (2021). https://doi.org/10.1038/s41377-021-00677-5
Received:
Revised:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41377-021-00677-5
This article is cited by
-
Self-optimized single-nanowire photoluminescence thermometry
Light: Science & Applications (2023)
-
Lanthanide-based ratiometric luminescence nanothermometry
Nano Research (2023)
-
Bandwidth-based flexible optical fiber thermometry based on sharp line emission regulation of Cr3+ ions
Science China Materials (2023)
-
Hollow nanoparticles synthesized via Ostwald ripening and their upconversion luminescence-mediated Boltzmann thermometry over a wide temperature range
Light: Science & Applications (2022)
-
Less is more: dimensionality reduction as a general strategy for more precise luminescence thermometry
Light: Science & Applications (2022)