Ratiometric Raman and Luminescent Thermometers Constructed from Dysprosium Thiocyanidometallate Molecular Magnets

Molecular crystals acting as temperature sensors, designed for multiple measurement techniques, can be a promising pathway for self‐calibrating thermometers. The molecular assemblies [DyxIIIY1–xIII(phen)2(μ‐OH)2(H2O)2]·[AuI(SCN)2]2·phen·0.5MeCN·0.5H2O (x = 0, 0.1, 0.02; phen = 1,10‐phenanthroline) which contain weakly bonded lanthanide(III) and Au(I) metal complexes are reported. They have Raman scattering in the low‐frequency (LF) region with sharp peaks, one of the prerequisites to design Raman thermometers. The Raman thermometric behaviors are characterized by three vibrational bands, and their thermal sensitivity is compared with their emission thermometric ability. The LF phonon linked with the Au···Au vibration helps increase the thermometric sensitivity of emission and Raman thermometers. Additionally, magnetically diluted complexes containing both Dy3+ and Y3+ ions are prepared to demonstrate the effect of temperature sensing capability through Raman and emission spectroscopy among isostructural mixed‐metal materials. Compounds are immersed inside various benchtop solvents to unravel the robustness and functioning of Raman thermometers. Furthermore, they reveal single‐molecule magnet properties, which disclose the effect of LF phonons on the spin relaxation process. Therefore, the reported molecular materials are Raman and luminescent thermometers, and they contain a molecular magnetic center with thiocyanidoaurate ions playing a critical role due to their emissive and Raman activities.

immersing in oil; materials sensitive to high energy UV radiation (e.g., photochromic materials); and for in vitro tests of biological samples in a solvent medium, important for the course of cryopreservation and the detection of temperature limit tests for biological systems (e.g., studies of extremophile organisms in their natural environment conditions). Recent discoveries have shown that Raman thermometry can be a useful tool to monitor the cellular temperature during photothermal therapy of tumors. [11] A similar application has also been realized for luminescent thermometers in the field of nanomedicine. [12] However, the application of emissive thermometers to the nonluminescent probe is impossible. On the contrary, vibrational features are more ubiquitous in the materials that feature in the research of Balaji Srinivasan et al. [13a] They have shown that the temperature of overhead power transmission cables can be monitored in real-time mode through Raman thermometry. [13] It is worth mentioning that the first attempts to develop synergistic Raman and photoluminescence thermometers have been conducted, but these studies were limited only to the detection of Raman active bands through emission spectroscopy, not direct Raman spectroscopy. [14] Raman thermometry works based on the vibrational peak shifts with temperature or by taking the ratio of the Raman band at the anti-Stokes (aS) and Stokes (S) positions. The second working principle exploring the aS/S ratio distinguishes Raman thermometry from infrared thermometry in terms of increased sensitivity. [9a,15] The efficiency of the anti-Stokes process is directly proportional to the population of excited vibrational modes which follows the Boltzmann distribution with temperature. A good Raman thermometer should possess a large Raman scattering cross-section, high Raman scattering in the low-frequency (LF) region, sharp peaks, and low absorbance of material concerning the exciting source.
Keeping the above materials research trend in mind, taking also into account the magnetic material perspective, we selected trivalent lanthanide ions (Ln(III)) as one of the precursors to include the emission and magnetic properties simultaneously. [2b,16] Ln(III) complexes have diverse luminescence colors and exhibit single-molecule magnet (SMM) properties. [16] Ln(III)-containing complexes reveal SMM behavior due to their strong magnetic anisotropy generating the energy barrier (ΔE) of spin relaxation. For such opto-magnetic lanthanide complexes, studying the vibrational signature in the LF region for Raman active modes through LF Raman spectroscopy can enhance our understanding of the dependence of physical properties (magnetic, optical) on structural changes. [17] Raman spectroscopy in the LF region has been used to detect various bonds of Au atoms in gold clusters, vibrational modes of pharmaceutical crystals, and phonons of graphene. [18] In this regard, Raman spectroscopy applied to a temperature sensor in the LF region (especially below 1 THz) has not been reported. Besides, a ratiometric thermometer (both Raman-type and emission-type) can be a great choice as a self-calibrating thermometer that uses the spectral intensity ratio for contactless thermal sensing. It is vital as many Raman thermometers explore the band shift rather than the ratio between the intensities of specific Raman peaks. Further utilization of two different spectroscopy techniques (Raman and luminescent) can provide more reliable information about the local temperature of the system by reducing measurement errors and measurement repeatability. Motivated to explore the Raman scattering for opto-magnetic molecular complexes, we prepared Dy(III)containing molecular materials with the formula of [Dy x III Y 1x III (phen) 2 (μ-OH) 2 (H 2 O) 2 ]·[Au I (SCN) 2 ] 2 ·phen·0.5MeCN·0.5 H 2 O [x= 0 (DyAu), x = 0.1 (Dy 0.1 Y 0.9 Au), x = 0.02 (Dy 0.02 Y 0.98 Au)], where phen is 1,10-phenanthroline. We selected this system as it has sharp Raman scattering peaks in the LF region revealed for the Y(III)-and Gd(III)-containing analogous compounds, one of the prerequisites for a good Raman thermometer. [5e] Further, emission properties at variable temperatures were investigated and, thanks to the presence of emissive Dy 3+ ions, luminescent thermometric properties were observed. Therefore, we compared the thermometric sensitivity of the emission and Raman spectroscopic features for all the complexes. Additionally, we demonstrated Raman thermometer behavior's applicability under various solvent conditions, in which classical fluorescent thermometers rarely operate, [6a-c] by measuring the Raman spectra for submerged solid crystals. Finally, due to the presence of paramagnetic Dy 3+ ions, we observed the SMM behavior for all compounds, which improved with magnetic dilution and has been correlated with the vibrational characteristics detectable in Raman spectroscopy.

X-Ray Crystallography
Colorless crystals of DyAu, Dy 0.1 Y 0.9 Au, and Dy 0.02 Y 0.98 Au were obtained from the mother solution containing an acetonitrile/ water mixture (see Experimental Section, Supporting Information). Single-crystal X-ray diffraction (SC-XRD) measurements for DyAu revealed that the material is centrosymmetric, crystallizing in the triclinic P-1 space group (Figure 1, Figures S1 and S2, and Tables S1 and S2, Supporting Information). Trimetallic mixed assemblies were prepared by mixing the appropriate amount of Ln(III) precursors and confirmed its perfect isostructural with DyAu by powder X-ray diffraction measurement ( Figure S3, Supporting Information). Each Dy(III) center coordinates two water molecules, two hydroxide ions, and two phen molecules. These gave four oxygen and four nitrogen atoms in the first coordination sphere to form a triangular dodecahedron geometry, confirmed by the Continuous Shape Measure (CShM) analyses (Tables S3, Supporting Information). The two cationic components {Dy III (phen) 2 (H 2 O) 2 (μ-OH) 2 } 2+ are connected through hydroxide molecular bridges to yield a dimeric unit with Dy(III) separation of 3.73 Å. The anionic {[Au I (SCN) 2 ] -} 2 -moiety consists of two Au(I) atoms, with two S-coordinated thiocyanide ions each. The Au-Au distance between two bis(thiocyanido)gold(I) units is 3.19 Å at 90 K, which is longer than the observed distance in Y(III)-and Gd(III)-based analogous matrices. [5e] This implies that as the atomic radius of lanthanide ions is shrinking within the lanthanide series, the Au-Au distance is expanding as a result, and thus the aurophillic interaction is gradually weakened. Notably, crystallographic studies revealed a significant increase in the Au-Au distance from 3.19 Å at 90 K to 3.31 Å at 300 K (Tables S2, Supporting Information), indicating that thermometric properties are associated with structural changes. Nevertheless, weak diffraction of crystals prevents determining the full temperature dependence.
The molecules of crystallization in DyAu include phen, water, and acetonitrile. They help in stabilizing the supramolecular assemblies through various hydrogen bonds involving hydrogens of water and nitrogens of thiocyanide ions, and between water and non-coordinated phen molecule. Along the crystallographic c axis, one of the coordinated phen, and crystallization phen molecules form π-π stacking, with the closest distance between their centroids of 3.62 Å. The composition and isostructurality of the whole series of materials, DyAu, Dy 0.1 Y 0.9 Au, and Dy 0.02 Y 0.98 Au, are confirmed through thermogravimetric analyses, indicating the amount of coordinated and crystallization solvent molecules calculated from the CHNS elemental analyses ( Figure S4, Supporting Information). Infrared and Raman spectroscopy were performed for all compounds, showing the presence of the cyanido stretching vibrations as well as the broad O-H vibrational bands ( Figures S5-S6, Supporting Information).

Raman Thermometry
Further, to understand the vibrational features in the LF region for all the compounds, we performed Raman spectroscopy for the single crystals of DyAu, Dy 0.1 Y 0.9 Au, and Dy 0.02 Y 0.98 Au (Figure 2 and Figures S7-S11, Supporting Information). The 532 nm laser light was used to probe the Raman active modes by observing the Stokes and anti-Stokes Raman shifts ( Figure 2). The spectra have several peaks assigned to the vibrations of bis(thiocyanido)gold(I) moieties and the phen molecule, present in the crystal structure (Figures S7 and S8, Table S4, Supporting Information). The peak centered around 108 cm -1 correlates with the vibrational modes of Au···Au stretching vibration, which shrinks upon cooling, as confirmed by the shift of this peak in the high-frequency region. The peak intensity of these phonons with temperature decreases at different rates depending upon the energy of vibrations. The peak intensity of high-energy Raman shift (>500 cm -1 ) saturates at higher temperatures (close to room temperature). However, the peak intensity of low-energy stretching frequency decreases following the Boltzmann distribution. [5e] This makes vibrational peaks in the LF region ideal for thermometric characterization in the wide temperature range. Therefore, we selected three vibrational peaks centered at 27, 108, and 304 cm -1 to represent the energy spectrum of vibration from below 33 cm -1 (below 1 THz) through around 99 cm -1 (3 THz) up to ≈303 cm -1 (9 THz) to investigate its thermometric behavior.
To precisely determine the temperature, the ratio of scattered anti-Stokes and Stokes Raman signal (also referred to as ρ) corresponding to the abovementioned vibrational frequency is calculated and fitted with the Equation (1). [9a,10a] where I aS and I S are the peak intensity of anti-Stokes and Stokes Raman shifts, respectively, C is the calibration constant, ν L (s -1 ) is the frequency of the laser, ν P (s -1 ) frequency of the vibrational band, h (J s −1 ) is Planck's constant, k B (J K -1 ) is the Boltzmann's constant, and T (K) is the temperature. Equation (1) suggests that ρ depends on the frequency of the incident laser as well as on the use of the detection method. The ρ is proportional to the quadruple of the ratio of incident laser frequency and frequency of vibrational mode when energy-based detection systems are used.
[10a] However, the charge-coupled device based detector proportionality relation takes the expression of Equation (1). The fitted results are presented in Figure 2 and  large C for 304 cm -1 (1.03) vibrational peak in the compound with a high concentration of Y(III) can be correlated with the smaller Au···Au bond distance (3.105 Å).
[5e] The calculated C values for the 108 cm -1 phonon modes are closer to reported values for the 143 cm -1 vibrational peak in TiO 2 . [12a] Interestingly, the determined ρ value is the largest for the 27 cm -1 vibrational band for all temperature ranges, followed by the 108 cm -1 . This observation implies that the population of scattered light with high energy photons (resulting in the anti-Stokes Raman shift) after coupling LF molecular vibrations with incident light is larger than for the higher energy stretching vibrations within the same system. The experimental ρ value lies in the range of 0.89-0.62, 0.58-0.10, and 0.22-0.02 for the vibrational band at 27, 108, and 304 cm -1 , respectively, for all compounds. The ratio (ρ) starts to saturate as the temperature of the surroundings approaches 80 K for the 304 cm -1 . We evaluated the relative thermal sensitivity (S r ) using Equation (2) by utilizing the fitted curve from the ρ versus T dependence to further examine the thermometric accuracy: The evaluated sensitivity is listed in  a) The scheme depicts the Raman spectra measurement method, using the cryogenic stage. b) Temperature-dependent Raman scattering for DyAu in the temperature range of 300-80 K. c) Experimental anti-Stokes/Stokes ratios in the range of 300-80 K for the Raman vibrational modes at 27 cm -1 (top), 108 cm -1 (middle), and 304 cm -1 (bottom) were collected after being excited by a 532 nm laser source. d) The calculated thermal sensitivity and e) temperature uncertainty for the temperature range of 300-80 K. The dotted plots in (c) represent experimental results and are fitted with Equation (1), thermal sensitivity in (d) is calculated with Equation (2), and the dotted-dashed line in (e) is calculated with Equation (3). The inset of (b) depicts the vibrational modes associated with bis(thiocyanido)gold(I) moieties for peaks centered around 108 and 304 cm -1 calculated based on density functional theory. Details about the Raman active vibrational modes are given in Table S4 (Supporting Information).
stretching frequencies as temperature decreases from 300 to 80 K, reaching maxima at the lowest temperature. Maximum sensitivity (6.84% K -1 ) is observed for the 304 cm -1 , followed by 108 cm -1 (2.42% K -1 ), and lowest for the 27 cm -1 (0.65% K -1 ) at 80 K. The opposite trend for the sensitivity compared with temperature-dependent of ρ can be explained by the fact that the rate of depopulation of the vibrationally excited state is faster for the 304 cm -1 stretching frequency when compared with the lower energy vibrational bands. The observed sensitivity is higher than the recently reported Raman thermometer (based on TiO 2 ) in the temperature range of 280-325 K, which lies in the range of 0.26-0.1% K -1 . [10a] We also calculated the temperature uncertainties using Equation (3), which suggests that they lie below 1 K for the broad temperature region ( Figure S9, Supporting Information): where δρ ρ is a relative error in the determination of ρ calculated after considering the spectral resolution and its background noise. Similar values of thermometric parameters were found among all assemblies thanks to their isostructurality, while the tuning of these parameters can be realized by the selection of vibrational peak frequency (Table S6, Supporting  Information).
Additionally, to understand the repeatability of these parameters, we performed similar experiments on the crystals of Dy 0.1 Y 0.9 Au and Dy 0.02 Y 0.98 Au under the same measurement conditions (Figures S10 and S11, Supporting Information). Both complexes have ρ lying in a similar range for all three vibrational modes as well as their relative thermal sensitivity (Table S6, Supporting Information). This implies that the isostructural compounds' vibrational signature in the LF region that uniquely identifies their crystal lattice packing follows a similar trend of thermal cooling. For the first time, we present a systematic study of molecular assemblies with similar crystal packing exhibiting an analogous thermal-dependent Raman scattering process.

Raman Thermometry in Solvents
The observation of a temperature-dependent Raman signal related to the various vibrational modes of the single crystalline sample motivated us to look for its potential application. Various forms of application can be envisioned, such as detecting the first-order phase transitions for other compounds in a wide temperature range (e.g., liquid-solid boundary of azeotropic mixtures) by mixing a Raman thermometer with the material having an unknown transition temperature. Another form of application can be to test the workability of these Raman thermometers in the presence of various gases, liquid solvents, solid mixtures, etc. Therefore, to demonstrate the introductory applications scheme, we tested the Raman thermometric effect for the crystals immersed in a few benchtop solvents like water, methanol, ethanol, isopropanol, acetone, hexane, and acetonitrile. The colorless crystals of DyAu were placed inside a thin cuvette, followed by the addition of ≈4 mL of solvent to fill up the container, as depicted in Figure 3. This study, for the first time, shows the interference of solvents on the LF Raman signal of known materials in the presence of various solvents while detecting the phase transition of liquids around their freezing point. The phase changes are also confirmed by the picture of frozen solvents near their known freezing temperatures (Figure 3, Figures S12-S18, and Table S7, Supporting Information). The Raman spectrum measured after the freezing point suggests that solidified solvent inhibits crystal excitation under the same optical path, leading to the disappearance of scattering from DyAu.
The detected Raman signal in the LF region for the different solvents demonstrates the same sharp peak with a similar background as the pristine material except for the measurement in acetonitrile. The DyAu in acetonitrile exhibits visible peaks related to the known vibrations. However, it has a very large background below 50 cm -1 ( Figure S17, Supporting Information). Therefore, we performed full thermometric characterization of DyAu in all other solvents by evaluating the anti-Stokes to Stokes ratio as a function of temperature. The Raman signal decreases with temperature in all solvents. The ratio ρ is fitted with Equation (1), indicating the non-intervention of the solvents. The maximum thermal sensitivity corresponding to all the stretching vibrations is observed close to the freezing temperature, and the minimum sensitivity is observed at 300 K, comparable with the original material. Maximum sensitivity observed is 2.44% K -1 at 273, 176, 170, 180, 180, and 180 K for the DyAu in the water, methanol, ethanol, isopropanol, acetone, and hexane, respectively, for the 108 cm -1 vibrational band related to the Au···Au motion (Table S8, Supporting Information). In conclusion, DyAu is a thermometer that works under different solvent conditions to sense the surrounding temperature before freezing the corresponding solvent.

Emission Thermometry and Its Comparis on with Raman Thermometry
The solid-state UV-vis-NIR absorption spectra indicate strong absorption in the UV region limited to the 400 nm ascribable to Au(I) centers ( Figure S19, Supporting Information). In the visible and NIR region, very weak sharp peaks are observed for the DyAu related to the direct excitation of Dy(III) ion, which diminishes upon dilution with Y(III) ions for Dy 0.1 Y 0.9 Au and Dy 0.02 Y 0.98 Au. The absence of absorption near 532 nm also helped preserve the crystalline sample under intense laser irradiation for Raman spectrum collection.
Apart from vibrational thermometry, these assemblies exhibit visible emission originating from Dy(III) ions and excimer Au(I) entities, observed over a broad temperature region of 10-300 K (Figure 4). The excitation by 325 nm light yielded two sets of peaks in all three complexes (Figure 4a-c). The first set of broad peaks centered around 460 and 510 nm originates from the excimers of {[Au(SCN) 2 ] -} 2, and the second set of sharp peaks located at 480, 575, and 660 nm comes from the f-f electronic transitions of Dy III centers, 4 F 9/2 → 6 H 15/2 , 6 H 13/2 , and 6 H 11/2 transitions, respectively. [5e] The excitation spectrum corresponding to different emissive peaks, within the whole series of materials, consists of a similar broad peak centered around 365 nm with variable broadening. The excitation is assigned mainly to the singlet-to-singlet (S 0 → S 1 ) and singletto-triplet (S 0 → T 1 ) transitions within dimers of bis(thiocyanido) gold(I) entities ( Figures S20-S22 (1), and the thermal sensitivity in (d) is calculated with Equation (2). The inset of (a) contains the picture of the cuvette with the indicated solvents utilized to drown the crystals of DyAu visible at the bottom of the container. The inset of (b) shows the picture of a single-crystal taken just before the measurement by the microscopic camera attached to the equipment at the indicated temperature. Black circles reveal the point of measurement. The dotted vertical line in (c) points to the freezing temperature of acetone.
Interestingly, the emission pattern for all of the assemblies is different in the context of the strength of emission originating from the Au(I) excimers (Figure 4a-c). The temperature-dependent spectrum of DyAu indicates that the energy absorbed by Au(I) centers is partially transferred to neighboring Dy(III) ions. At low temperatures, the Dy(III) emission peaks become sharper and stronger due to the deactivation of nonradiative thermal relaxation pathways. This is also supported by the lack of emission peak originating from triplet-to-singlet electronic states of {[Au(SCN) 2 ] -} 2 , which was visible in the analogous Y(III) or Gd(III) matrices. [5e] Therefore, the dilution of Dy(III) ions with Y(III) ions in the excess amount allows the Au(I) centers to emit light in the high energy region (460 nm) and low energy region (around 490 and 510 nm) for the Dy 0.1 Y 0.9 Au and Dy 0.02 Y 0.98 Au. Overall, such changes impact the bluish emission color at 300 K and changes to light blue  and cyan for the Dy 0.1 Y 0.9 Au and Dy 0.02 Y 0.98 Au, respectively, at 100 K (as shown in the inset of Figure 4a -c). The further highresolution emission spectrum of Dy 0.02 Y 0.98 Au helped to understand the splitting of different multiplets for the 6 H 15/2 and 6 H 13/2 ( Figure S23, Supporting Information).
The varying luminescence spectra with temperature motivated us to fully characterize the thermometric properties for all three complexes by taking the ratio of emission intensity centered around 575, 480, and 460 nm. The three different thermometric parameters (Δ(T)) are defined by the ratio of emission intensities at 575 to 480 nm, 480 to 460 nm, and 575 to 460 nm (Figure 4d-f). The experimental results reveal different dependencies with temperature for three assemblies. Δ(T) corresponding to the I 480 /I 460 ratio is larger than other thermometric parameters for a wide temperature range. The apparent anomaly can be due to the overlapped luminesce from Dy(III) as well as from the Au(I) centers around this emission region. Since all three luminescence peaks originate from different emissive energy levels, we utilized the Mott-Seitz model given in Equation (4) to fit the experimental data: [19] T where Δ 0 is the thermometric parameter at T = 0 K, α 1 , α 2 are the ratio of non-radiative (W 0 at T = 0 K) and radiative (W R ) rates, and ΔE 1 , ΔE 2 are the activation energies for involved electronic transitions (subscript 1 denotes the nominator, and 2 implies the denominator of the Δ(T)). The fitted parameters are presented in Table S9 (Supporting Information) for all complexes. Fitted results suggest that α 2 and ΔE 2 corresponding to the 460 nm emissive peak are larger than for the 575 and 480 nm within the same thermometric parameter, signifying the differences in photoluminescence origin.
Additionally, the relative thermal sensitivity of the molecular assemblies is evaluated with Equation (5) (Table S10, Supporting Information): The maximum thermal sensitivity for the luminescence thermometer is found in Dy 0.02 Y 0.98 Au, which is 3.5% K -1 at 55 K, 2.4% K -1 at 55 K, and 1.8% K -1 at 63 K, for the Δ(480/460), Δ(575/460), and Δ(575/480), respectively. Larger sensitivity is associated with the photoluminescence of Au(I) centers varying strongly with temperature due to a closer AuAu bond at low temperature. For the same reason, thermal sensitivity increases going from DyAu to Dy 0.02 Y 0.98 Au through Dy 0.1 Y 0.9 Au. Temperature uncertainty lies below 1 K for all compounds calculated with the help of δT = (δΔ/Δ)/S r ( Figure S24 and Table  S10, Supporting Information). The repeatability of Δ(T) for all complexes was confirmed by the three cycles of measurement in two different temperature regions ( Figures S25-S27, Supporting Information). After considering all the fitted parameters, Dy 0.02 Y 0.98 Au is found to be the best temperature sensor among the three complexes; thus, better-performing emission thermometers can be designed by linking structural changes with luminescence.
The comparison of selected materials revealing Raman and/or emission thermometric properties indicates that both temperature sensing methods provide similar sensitivity among high-performing thermometers ( Table 1) and other fluorescent thermometers (Table S11, Supporting Information). Although some of the luminescence thermometers reveal double-digit sensitivity, they are limited to low-temperature (usually below 10 K) regions. Temperature sensors based on Raman scattering can be characterized sometimes up to Adv. Optical Mater. 2022, 10,2201675  Ln-HL (Ln = Eu, Tb) 4-290 31% K -1 @4 K Two intensities ratio [20d] aS: anti-Stokes; S: Stokes.
2000 K, providing a very broad working temperature range. In general, Raman thermometers are rarely explored, and other investigated materials include TiO 2 , silicon, diamond, etc. Therefore, there is a great need to develop this field with new materials to increase its sensitivity and explore its utility in a wide temperature range.

Magnetic Properties
Apart from the photophysical properties, the reported materials also exhibit paramagnetic behavior thanks to the Dy(III) ions revealed through direct current (dc) and alternate current (ac) measurements. The magnetic susceptibility-temperature product (χ M T) versus temperature plot for DyAu indicates that the χ M T at 300 K is 13.93 and decreases slowly up to 50 K followed by a sharp drop to 8.25 cm 3 mol -1 K at 1.85 K. The observed value for the χ M T signal at high temperature is close to theoretically predicted for the 6 H 15/2 ground state of Dy(III) ion. It deviates from the ab initio-calculated curve slightly at low temperatures, which may be due to interatomic interactions ( Figure S28, Supporting Information). Magnetization (M) versus field (H) plots at 1.85 K shows the highest value of 5.16 μ B at 50 kOe which lies in the range typical for paramagnetic Dy(III) ions ( Figure S28, Supporting Information). The magnetically diluted samples follow the stochiometric synthetic ratio to resemble the magnetization of pure DyAu. The ac magnetic studies of DyAu reveal SMM behavior; however, at the zero dc field, the out-of-plane χ M " signal is out of the measurement range suggesting the dominance of quantum tunneling magnetization (QTM) for spin relaxation ( Figure S29, Supporting Information). The proximity of two Dy(III) ions bridged through hydroxide ions may be disadvantageous as it sometimes promotes disturbing antiferromagnetic interaction. Therefore, magnetically diluted samples Dy 0.1 Y 0.9 Au and Dy 0.02 Y 0.98 Au were examined and the SMM properties under the applied external dc field were detected. (Figure 5, Figures S30-S32, and Tables S12-S14, Supporting Information). Relaxation time (τ) versus inverse temperature (T -1 ) curve for both assemblies is fitted with an Equation describing the Raman, QTM, and Orbach spin relaxation processes ( Figure 5, the detailed explanation is given in the Supporting Information). Fitting results suggest that spin-lattice interaction dominates in the magnetization reversal process, which might correlate with the Raman signal observed in the LF region with the vibrational energy of 60 cm -1 or lower, suggesting the presence of phonon required to help the spin relaxation to happen through the Raman mechanism. [17] The obtained Orbach energy barrier for Dy 0.02 Y 0.98 Au is 90.83 K (63.1 cm -1 ) following Equation (E3), corroborating the energy gap between the m J states as predicted through ab initio calculation and underestimating by 63 cm -1 calculated with high-resolution emission spectra ( Figure S23 and Table S15, Supporting Information). The ground state composition for Dy(III) ions computed by replacing neighboring Dy(III) ions with Y(III) ions reveals that it has an anisotropic magnetic easy axis (g z = 18.56); however accompanied by nonnegligible transversal components hampering the zero-dc-field slow magnetic relaxation ( Figure S33, Table S15, Supporting Information).

Conclusion
We prepared Dy(III)-containing molecular materials with hydroxide bridged bis(lanthanide) moieties crystallizing with dimeric thiocyanido-{Au I 2 } units forming emissive excimers. These molecular crystals reveal several sharp peaks in Raman spectra in the LF region related to the various vibrational signatures of the crystal packing. The intensity of the temperature-dependent Raman spectrum decreases with decreasing temperature. The decrease in the Raman scattering is related to the reduction of the phonon mode population following the Boltzmann distribution. Further, to impart temperature sensing abilities to these materials, we selected three peaks of the Raman spectrum located near 27, 108, and 304 cm -1 to take the ratio of their anti-Stokes and Stokes Raman shifts. The ratio ρ is fitted to show the good thermometric ability. The maximum sensitivity of 6.84% K -1 for vibrational thermometry is observed, corresponding to the highest energy phonon of 304 cm -1 as its ρ starts to saturate at the lower temperature region. Additionally, the DyAu crystals are placed inside different benchtop solvents to test their functioning ability as a temperature indicator. For the first time, we presented the ability of these solvents to interfere in the LF region of Raman spectrum measurement. The experiment with acetonitrile resulted in a large background, especially below 50 cm -1 . Water, methanol, ethanol, isopropanol, hexane, and acetone did not affect the raw data of the obtained DyAu Raman thermometers. The maximum thermal sensitivity in solvents is observed to be 2.44% K -1 relating to the 108 cm -1 phonon suggesting the importance of the metal-metal vibrational feature in enhancing the thermometric ability. The vibrational spectroscopy for all the assemblies is also compared with their respective emission thermometric abilities. For these sets of complexes, the observed thermal sensitivity of the luminescence thermometers is lower than the corresponding sensitivity of the Raman thermometer in the entire temperature region. Moreover, observed thermometric ability through Raman and emission spectroscopy for magnetically diluted compound Dy 0.02 Y 0.98 Au is better than pure DyAu which also exhibits magnetic relaxation. The improved emission thermometric parameter in Dy 0.02 Y 0.98 Au is again related to the emission properties originating from the Au(I) centers, which change exponentially with thermal cooling. Therefore, the dilution of such magnetic matrix improves optical and magnetic properties also observed earlier for Ho(III)-cyanidometallate compounds. [21] In the future, the Raman thermometers' applications, such as the detection of phase-transition of several types of materials, will be explored by mixing the unknown compound with fully characterized Raman thermometers. More extended research should be dedicated to designing high-performing Raman-based temperature sensors functioning in a broad temperature range. This can be done by judiciously selecting the vibrational band with appropriate energy. Moreover, the reported materials reveal SMM behavior that improves upon the magnetic dilution. Spinlattice relaxation is detected for the magnetically diluted samples, correlating with the vibrational bands of Raman active modes. Therefore, generating or hampering some vibrational features in the LF region should be one of the preconditions for the design of high-performance SMMs, while their presence adds some new functionalities (combination of Raman and luminescent thermometry) as presented here by us. and out-of-plane, b) χ M " components of the complex magnetic susceptibility for the indicated temperatures of the 1.85-4.00 K range, c) the related χ M "-χ M ' Argand plots, and d) the temperature dependence of relaxation times (τ) presented as ln(τ) versus T -1 plot. Solid lines in (a)-(c) were fitted using the generalized Debye model (Table S13, Supporting Information), whereas the solid line in (d) is fitted with Equation (E3) describing QTM, Orbach, and Raman relaxation process (Detailed explanation is provided in Table S14, Supporting Information).