Temperature-dependent FTIR spectroscopy of OH defects in Verneuil-grown corundum (α-Al2O3)

Temperature-dependent FTIR spectroscopy of OH defects in Verneuil-grown corundum (α-Al2O3) | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Temperature-dependent FTIR spectroscopy of OH defects in Verneuil-grown corundum (α-Al2O3) Etienne Balan, Michael C. Jollands, Maxime Guillaumet, Keevin Béneut This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5050116/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 17 Nov, 2024 Read the published version in Physics and Chemistry of Minerals → Version 1 posted 12 You are reading this latest preprint version Abstract The temperature dependence of the infrared absorption spectra of two Verneuil-grown corundum samples is investigated in the OH stretching range. The spectra display three main bands at 3184, 3232 and 3309 cm − 1 , belonging to the so-called "3309 cm − 1 series", as well as two additional bands at 3163 and 3278 cm − 1 previously reported in some synthetic corundum samples. The anharmonic behavior of the observed bands is analyzed using the pure dephasing model of Persson and Ryberg and depends on the local geometry of the OH defects, which are all associated with Al vacancies. A weak band at 3209 cm − 1 displays anomalous intensity changes with temperature which support a revised interpretation of both the 3209 and 3232 cm − 1 bands. The two bands are interpreted as resulting from the low-temperature equilibrium between two Ti-associated OH defects, enabled by the possibility of hydrogen quantum tunneling within the Al vacancy. The temperature-dependent properties of the 3278 cm − 1 band are similar to those of the other Al-vacancy related defects and a comparison with the theoretical properties of selected OH defects suggests that this band corresponds to the association of the H atom with a non-dissociated Al Frenkel pair. Finally, the properties of the band at 3163 cm − 1 are consistent with its previously proposed association with Si for Al substitution in corundum. Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Introduction The infrared spectra of natural and synthetic corundum (a-Al 2 O 3 ) crystals frequently reveal the presence of hydroxyl groups associated with structural defects and substitutional impurities. The absorption bands belonging to the "3309 cm − 1 series" are observed in synthetic (Eigenman and Günthard 1971; Volynets et al. 1969 , 1972 ; Beran 1991 ; Moon and Phillips 1991 , 1994 ; Kronenberg et al. 2000 ; Ramírez et al. 1997 , 2004 ) and natural (Beran and Rossman 2006 ; Soonthorntantikul et al. 2019 ) corundum. They are dominated by well-resolved bands at 3184, 3232, and 3309 cm − 1 . Weaker bands at 3295 and 3365 cm − 1 are also reported as belonging to the same series. All these bands have a polarization mostly parallel to the (001) plane consistent with the bonding of protons to the basal oxygen triangle of Al vacancies (T-Thienprasert et al. 2017 ; Balan 2020 ). The variability of the "3309 cm − 1 series" reflects a diversity of charge compensation mechanisms involving substitutional tetravalent cations in the vicinity of the H-bearing Al vacancy (Moon and Phillips 1991 , 1994 ; Ramirez et al. 2004; Balan 2020 ). Some synthetic corundum samples can also display an absorption band at 3278 cm − 1 (Turner and Crawford 1975 ; Engstrom et al. 1980 ; Kronenberg et al. 2000 ; Ramírez et al. 1997 , 2004 ; Choudhary and Vijay 2018 ) with a polarization parallel to the (001) plane which does not seem to be related to a specific impurity as it occurs in diffusively hydrogenated pure corundum (Jollands 2024 ). Another major H incorporation mechanism in corundum, which was not considered in the present study, corresponds to its association with substitutional divalent cations (Jollands et al. 2023 ). Despite the numerous studies reporting the FTIR spectra of OH defects in corundum, the assignment of the absorption bands in terms of specific atomic-scale environments is still open to discussion due to the complexity of the spectra and the diversity of hydrogen incorporation mechanisms. In the present study, further information on the OH defects in corundum are obtained by analyzing the modifications of the infrared absorption spectrum as a function of temperature. The results are analyzed using an exchange mode theory which describes the anharmonic properties of localized vibrational modes in crystals and the interpretation of some of the observed bands is further discussed in the light of theoretical models of OH defects in corundum. Methods Two 3 mm thick slabs of Verneuil-grown corundum crystals produced by Hrand Djévahirdjian, Monthey, Switzerland were investigated (R-3 and R-15 samples). The flat faces of the slabs are parallel to the [001] direction. Unpolarized transmission FTIR spectra of sample R-15 were recorded from 10 to 290 K with an instrumental resolution of 1 cm − 1 using a Nicolet 6700 FTIR spectrometer set with an EverGlo source, KBr beamsplitter and DTGS detector. Low-temperature measurements were performed using an ARS CS-204 SI cryocooler fitted with KRS-5 windows. Temperature control was ensured with a Si diode fixed on the sample holder. Unpolarized transmission FTIR spectra of sample R-3 were recorded from 20 to 540 K on a Bruker IFS 66v/S Fourier transform infrared spectrometer working in vacuum with an instrumental resolution of 2 cm − 1 . The spectrometer was set with KBr beamsplitter, Globar source and liquid nitrogen-cooled MCT detector. In situ temperature-dependent measurements were performed using a Janus He-cryostat and the temperature was controlled with a thermocouple fixed on the sample holder. To analyze in more detail the changes of individual bands as a function of temperature, the spectra were decomposed into individual Lorentzian components (Tables S1 and S2) using the Fityk code (Wojdyr, 2010 ). Depending on the spectra, a baseline was subtracted before the fit using a visually anchored spline function or was treated as a polynomial function incorporated in the fit. Only the main bands, whose parameters are expected to be less sensitive to uncertainties related to band overlaps and baseline subtraction, were analyzed in detail. The analysis of the R-3 sample was restricted to temperatures below 400 K for the same reasons. Anharmonic properties of localized vibrational modes in crystals Various mechanisms, which are usually considered independently, contribute to the evolution of the vibrational spectrum of a localized vibrational mode (LVM) with temperature. The line shift is determined by the thermal expansion of the crystal structure and by the anharmonic coupling of the LVM with other vibrational modes of the system. The line broadening is determined by the decay of the macroscopic polarization due to the energy transfer from excited oscillators to the crystal phonon bath (population relaxation with time constant T 1 ), and to the loss of phase coherence induced by thermal fluctuations (pure dephasing with time constant T 2 *). The energy transfer to the phonon bath from a high frequency LVM such as the stretching of an isolated OH defect (> 3000 cm − 1 ) implies the emission of a large number (> 3) of lower frequency phonons (< 1000 cm − 1 ). In this case, the contribution of pure dephasing to the line broadening is generally dominant (e.g. Budde et al. 2001 ). Various models of pure dephasing were proposed to describe the temperature-dependent vibrational properties of molecular adsorbates at crystal surfaces (Persson and Ryberg 1985 a,b) or polyatomic molecules in condensed phase (Shelby et al. 1979 ). More sophisticated models were later proposed and successfully applied to LVM in solids (Budde et al. 2001 , Martin et al. 2006 ). The Persson and Ryberg model uses a minimal number of parameters to capture the salient features of the anharmonic system and provides analytical expressions describing the evolution of the LVM band as a function of temperature. In this model, the LVM frequency is weakly coupled to a lower frequency mode (the "exchange mode"), the strength of the coupling being characterized by a parameter dw. The exchange mode interacts in its turn with the phonon bath via a friction parameter η. In the weak coupling limit (|δω| << η), the following relations describe the temperature-dependence of the frequency shift, ΔΩ= Ω− Ω 0 , and the width G of the high-frequency LVM band: ΔW = δω/(exp(hω ex /2pkT)-1) (1) and Γ=2 (δω 2 /η) exp(hω ex /2p kT) /(exp(hω ex /2pkT)-1) 2 (2) where W and W 0 are the angular frequencies of the LVM at finite temperature and in the low temperature limit, respectively, ω ex is the harmonic angular frequency of the exchange mode, h the Planck constant, k the Boltzmann constant, and T the temperature. As discussed by Budde et al. ( 2001 ), the dephasing mechanism related to one particular exchange mode is more efficient than the dephasing induced by the direct anharmonic coupling of the LVM with the phonon bath. The weak coupling condition of the Persson and Ryberg model is however not always met by LVM in solids but equations (1) and (2) can still be used as functional expressions to fit the experimental data (e.g., Suezawa et al. 2001 ; 2002 ). Results : Temperature-dependent FTIR spectroscopy of the R-3 and R-15 samples The room temperature spectra of samples R-3 and R-15 display the bands commonly assigned to the "3309 cm − 1 " series with relatively strong 3232 and 3184 cm − 1 bands and weaker 3309, 3295 and 3209 cm − 1 bands (Fig. 1 ). The 3278 cm − 1 band is also present, with a stronger relative absorbance in the R-3 spectrum. An additional band at 3163 cm − 1 is observed in the R-15 spectrum and a minor band at 3194 cm − 1 can be inferred from the fit of the spectrum (Fig. 1 ). Overall, these spectra are similar to those of the series of Verneuil-grown corundum samples previously reported by Beran ( 1991 ). Spectral changes observed as a function of temperature on the R-15 (Fig. 2 ) and R-3 (Fig. 3 ) samples mostly consist in a narrowing and blue-shift of the absorption bands when the temperature is lowered. The minor band at 3194 cm − 1 inferred from the fit of the RT spectrum of the R-15 sample is well resolved below 150 K. In contrast, the band at 3209 cm − 1 present in the RT spectra is only observed above 150 K. Measurements performed on the R-3 sample also indicate that the spectral changes affecting the 3309 cm − 1 series are in the continuity of those observed at lower temperature when the temperature is increased up to 540 K. However, the overlap of broadened bands at high temperature makes the analysis of individual contributions above 400 K challenging (Fig. 3 ). Of note, the baseline changes at 130 and 150 K in the R-15 spectrum (Fig. 2 ) are likely related to condensation of a thin ice layer at cryogenic temperatures. Based on the spectral fits, the bands of the 3309 cm − 1 series exhibit a similar behavior in the R-15 and R-3 samples (Fig. 4 ). Their frequency decreases by ~ 3 to 4 cm − 1 and their full width at half maximum (FWHM) increases by ~ 1 to 6 cm − 1 when the temperature increases from 10 K to 290 K. For temperatures lower than 100 K, the frequency and the width of these bands are almost temperature-independent with the FWHM at saturation ranging between 2 and 4 cm − 1 . The behavior of the bands at 3278 cm − 1 (R-3 sample) and 3163 cm − 1 (R-15 sample) is also similar to that of the bands of the 3309 cm − 1 series. It is noteworthy that the 3163 cm − 1 and 3184 cm − 1 bands in the R-15 sample have almost identical temperature behaviors. Finally, the temperature dependence of the band area is weak (Fig. 4 ), as expected for fundamental infrared absorption bands. The low-temperature area of the 3232 cm − 1 band is approximately three times larger than that of the 3163 cm − 1 band. The temperature dependence of the absorption bands can be analyzed using the Persson and Ryberg model (Eqs. 1 and 2). Given the uncertainties related to the complexity of the spectra, the frequency shifts were first fitted using Eq. 1, leading to w ex and dw (Table 1 ). The exchange mode frequency ω ex is closely related to the extent of the saturation temperature domain and ranges between 180 and 280 cm − 1 . The dw parameter is negative with an absolute value ranging between 5 to 11 cm − 1 . These parameters were subsequently introduced into Eq. 2 to fit the variations in FWHM, using h only as a free parameter. For all the bands, a satisfactory fit is obtained for both the shift and the width of the lines as a function of temperature (Fig. 4 ). The h parameter ranges between 7 and 100 cm − 1 and is always superior to |dw| (Table 1 ). These results suggest that the Persson and Ryberg model provides a reasonable account of the temperature dependence of the bands belonging to the 3309 cm − 1 series, as well as of that of the bands observed at 3278 and 3163 cm − 1 . In particular, the same exchange frequency is consistent with both the shift and the broadening of the bands, even though the friction parameter is not largely superior to the anharmonic coupling parameter. Table 1 Parameters of the Persson and Ryberg model (Eqs. 1 and 2) describing the temperature-dependent spectroscopic properties of OH stretching bands in the R-15 and R-3 samples. sample w OH (RT) (cm − 1 ) w 0 (cm − 1 ) w ex (cm − 1 ) dw (cm − 1 ) G 0 (cm − 1 ) η (cm − 1 ) R-15 3309 3312 279 -8 1.6 38 3232 3234 228 -5 3.4 9 3184 3188 181 -5 2.2 12 3163 3167 191 -6 2.2 11 R-3 3309 3312 326 -11 2.8 92 3278 3283 238 -10 3.3 39 3232 3234 228 -5 3.9 7 3184 3188 205 -6 3.0 11 Discussion: Interpretation of the OH stretching spectrum of the R15 and R3 samples The bands observed at 3309, 3232 and 3184 cm − 1 in the spectra of the R-15 and R-3 samples were unambiguously related to OH defects associated with Al vacancies and nearby Ti 4+ for Al 3+ substitution. As experimentally shown by Moon and Philipps (1991, 1994) and Ramirez et al. (2004), the rapid quenching of samples annealed at high temperature favors a dilute configuration of Ti 4+ cations. This results in an increased intensity of the 3232 and 3184 cm − 1 bands, which are related to negatively charged defects involving a single Ti 4+ for Al 3+ substitution. In contrast, a slower quenching or a lower annealing temperature favors more clustered and electrostically neutral configurations with two Ti 4+ for Al 3+ substitution in the vicinity of the OH-bearing Al vacancy, leading to an increase in the relative intensity of the 3309 cm − 1 band. The bands of the 3309 cm − 1 series are observed in the whole series of Verneuil-grown samples investigated by Beran ( 1991 ) as well as in the as grown high purity corundum investigated by Ramirez et al (1997). This suggests that even in nominally pure corundum samples the concentration of Ti is high enough to contribute to the charge compensation of Al vacancies. The interpretation of these bands in terms of clustering schemes of OH-bearing Al vacancies and Ti for Al substitutions was latter supported by a theoretical study of OH defects in corundum (Balan 2020 ). The most stable configuration displaying two Ti atoms approximately facing the OH group corresponds to the band at 3309 cm − 1 . Two configurations with a single Ti atom ((1H + ) Al (Ti 4+ ) Al1 and (1H + ) Al (Ti 4+ ) Al2 models) have close stabilities and OH stretching frequencies. It was thus assumed that the two corresponding bands overlap in the 3232 cm − 1 band, explaining its stronger relative intensity. The significantly weaker 3184 cm − 1 band was assigned to a third configuration with a similar stability and lower OH stretching frequency. In this case, the Ti 4+ cation is located at the site forming a pair of face-sharing octahedral sites with the vacant Al site ((1H + ) Al (Ti 4+ ) Al10 model). Finally, the weaker band at 3295 cm − 1 is ascribed to a less stable configuration with a single Ti atom ((1H + ) Al (Ti 4+ ) Al4 model). The temperature-dependent measurements of the R-15 and R-3 samples provide some insight into the anharmonic properties of the defects related to the main bands. In the two samples, the 3309 cm − 1 band corresponds to a coupling of the LVM with an exchange mode at slightly higher frequency than for the 3232 and 3184 cm − 1 bands. The friction parameter of the exchange mode is also consistently greater for the 3309 cm − 1 band than for the two other bands. In comparison, the behaviors of the 3232 and 3184 cm − 1 bands are similar, which confirms that the anharmonic properties of the OH stretching modes are sensitive to the details of the local defect geometry. However, it is noteworthy that no splitting of the 3232 cm − 1 band is observed over the whole temperature range, even though its low temperature FWHM is larger than that of the 3309 and 3184 cm − 1 bands. Assignment of the 3209 cm − 1 band to a Ti-bearing OH defect The temperature-dependent behavior of the weaker 3209 cm − 1 band is uncommon as its absorbance steadily decreases when the temperature decreases and the band vanishes below 150 K (Figs. 2 & 3 ). Consistently, this band, as well as its OD counterpart, are not observed in the spectra of hydrogen or deuterium-treated corundum samples measured at 80 K by Ramirez et al. (1997) but are present in those measured at 295 K (Figs. 1 and 2 of Ramirez et al. 1997). This behavior thus differs from that usually expected for a fundamental transition and observed for the other bands of the R-3 and R-15 spectra (Fig. 4 ). The presence of a "hot" band associated with the stronger 3232 cm − 1 band is unlikely because the frequency difference is only ~ 20 cm − 1 , while the 3209 cm − 1 band appears only above 150 K. These properties, as well as the similar behavior inferred for the corresponding OD band, are hardly reconciled with a phenomenon depending on the anharmonic coupling of the stretching mode with a hypothetical vibrational mode at very low frequency. The 3209 cm − 1 band more likely corresponds to a fundamental transition and by implication its intensity changes should find their origin in a low-temperature equilibrium between OH configurations with relatively close energies. A relation of the 3209 cm − 1 band to Ti-bearing defects is suggested by its association with the 3232 and 3184 cm − 1 bands in the spectrum of synthetic corundum samples intentionally doped with Ti (Moon and Philips 1991, 1994; Ramirez et al. 2004) or reported as nominally pure (but likely containing a low Ti concentration) (Ramirez et al. 1997; Kronenberg et al. 2000 ). Note that the band is not present in the spectra reported by Eigenman and Gunthard (1971) as they were recorded at the liquid nitrogen temperature. Also considering that the 3232 cm − 1 band does not split over the whole temperature range (10 K-540 K), which challenges its assignment to two overlapping signals, it is thus appropriate to jointly discuss the presence and the properties of the two 3232 and 3209 cm − 1 bands in the light of the theoretical properties of the Ti-bearing OH defects, in particular those of the two closely related (1H + ) Al (Ti 4+ ) Al1 and the (1H + ) Al (Ti 4+ ) Al2 configurations. Using a Boltzmann statistics, the intensity variations of the 3232 and 3209 cm − 1 bands lead to an energy difference of ~ 5.5 kJ/mol (Fig. 5 ), which is close to the theoretical energy difference between the two models (4.9 kJ/mol, Table 2 ). In addition, the difference of the theoretical OH stretching frequency between these two models, previously determined at ~ 1 cm − 1 (Balan 2020 ), increases to ~ 9 cm − 1 using larger 3x3x1 supercell models (Table 2 ), which is closer to the experimentally observed splitting (~ 20 cm − 1 ). Accordingly, the relative intensity changes of the 3232 and 3209 cm − 1 bands could reflect an equilibrium between the (1H + ) Al (Ti 4+ ) Al1 configuration and the slightly less stable (1H + ) Al (Ti 4+ ) Al2 configuration, respectively. This equilibrium between the (1H + ) Al (Ti 4+ ) Al1 and (1H + ) Al (Ti 4+ ) Al2 configurations corresponds to the hopping of the H atom from one oxygen to the other available oxygen atom of the large oxygen triangle which forms the base of the vacant Al site. The third oxygen of the triangle belongs to the coordination sphere of the neighboring Ti ion and is significantly less prone to protonation (Balan 2020 ). The possibility of H hopping in an Al vacancy is consistent with the findings of Holder et al. ( 2013 ) who predicted an efficient H tunneling between equivalent oxygen sites of the Al vacancy in corundum, while the corresponding potential energy barrier (0.612 eV) is well above the thermal energy at room temperature (0.025 eV). The quantum tunneling of the hydrogen atom within the Al vacancy could thus enable the low-temperature equilibrium between the energetically close (1H + ) Al (Ti 4+ ) Al1 and (1H + ) Al (Ti 4+ ) Al2 configurations. In contrast, the relative intensities of the other bands, i.e. the 3309 and 3184 cm − 1 bands, are mostly inherited from the high temperature history of the sample (Moon and Philips 1991, 1994, Ramirez et al. 2004). In this interpretation, the stronger intensity of the 3232 cm − 1 band is thus not explained by the overlap of two different absorption bands of similar intensities, as previously assumed, but rather results from the unquenchable conversion of the (1H + ) Al (Ti 4+ ) Al2 configuration into the slightly more stable (1H + ) Al (Ti 4+ ) Al1 configuration, both configurations occurring with an equivalent probability at high temperature. Due to the mobility of hydrogen within the vacancy, the relative intensities of the 3232 and 3184 cm − 1 bands at low temperature directly probe the Ti occupancy of the four nearest-neighbor octahedral sites surrounding the Al vacancy, which is consistent with the ~ 3:1 ratio of the band areas observed at low temperature (Fig. 4 ). Table 2 Theoretical properties of selected models of OH defects in corundum obtained using a 3x3x1 supercell and the same numerical parameters as in Balan ( 2020 )*. The .cif files of the models are provided in the Supplementary Information. model charge tot. energy (Ry) rel. energy (kJ/mol) d OH (Å) w OH (cm − 1 ) we xp (cm − 1 ) Singly protonated Al vacancy (1H + ) Al,Oc -2 -19976.6339 - 0.9929 3298 3184 Interstitial proton (1H + ) i + 1 -20162.2070 - 1.0071 2998 n.o. Proton associated with Al vacancy and one nearby Ti for Al substitution (1H + ) Al (Ti 4+ ) Al1 -1 -19998.7281 0 0.9903 3336 3232 (1H + ) Al (Ti 4+ ) Al2 -1 -19998.7244 4.9 0.9906 3327 3209 (1H + ) Al (Ti 4+ ) Al10 -1 -19998.7249 4.2 0.9928 3296 3184 Proton associated with Al vacancy and two nearby Ti for Al substitutions (1H + ) Al (Ti 4+ ) Al1 (Ti 4+ ) Al2 0 -19976.6339 - 0.9852 3418 3309 Proton associated with Al vacancy and one nearby Si for Al substitution (1H + ) Al (Si 4+ ) Al1 -1 -19891.1170 20.7 0.9911 3326 n.o. (1H + ) Al (Si 4+ ) Al2 -1 -19891.1141 24.6 0.9909 3331 n.o. (1H + ) Al (Si 4+ ) Al10 -1 -19891.1328 0 0.9937 3283 3163 Proton associated with a non-dissociated Al Frenkel pair (1H + ) Al (Al 3+ ) i1 + 1 -20161.9715 0 0.9874 3390 3278 (1H + ) Al (Al 3+ ) i2 + 1 -20161.9582 17.5 0.9941 3273 n.o. Pure corundum 0 -20160.2250 - - - - Al Frenkel pair (V) Al (Al 3+ ) i 0 -20159.8198 - - - - *Calculations were performed using the PWscf and PHonon codes (Giannozzi et al., 2009 ) with norm conserving pseudo-potentials (Hamann, 2013 ; Schlipf and Gygi, 2015 ) and PBE functional (Perdew et al. 1996 ). The electronic integration was restricted to the G point with 80 Ry and 480 Ry cutoffs for the wavefonction and electron density, respectively. n.o.: not observed Interpretation of the 3278 cm − 1 band in synthetic corundum Synthetic corundum samples (Turner and Crawford, 1975 ; Engstrom et al., 1980 ; Kronenberg et al., 2000 ; Ramírez et al., 1997 a, 2004 ; Choudhary and Vijay, 2018 ) can display an absorption band at 3278 cm − 1 , which is also observed in very pure samples (Jollands 2024 ). Its relative intensity is greater after a fast sample cooling, while a slower cooling favors the 3309 and 3232 cm − 1 bands (Ramirez et al., 2004). The relative intensities of these bands are reversible indefinitely by alternate fast and slow cooling (Ramirez et al. 2004). The 3278 cm − 1 band in the R-3 sample was strong enough to determine its temperature behavior, which is roughly similar to that of the bands of the 3309 cm − 1 series (Fig. 4 , Table 1 ). The observed shift and FWHM variation are consistent with those reported by Engstrom et al. ( 1980 ) from measurements at 77 and 300 K. As the 3278 cm − 1 band is observed in very pure corundum samples, this band likely involves an intrinsic defect of corundum, i.e. it is not related to a specific trace impurity. Its polarization almost parallel to the (001) plane (Ramirez et al. 2004; Jollands 2024 ) and its temperature-dependent changes also suggest that it is related to an OH group located in the large oxygen triangle forming the base of a vacant Al site, as for the other defects of the 3309 cm − 1 series. However, its frequency is too high to be consistent with the OH stretching frequency of a singly protonated vacancy (Table 2 ), which implies a modification of the local environment of the vacant Al site. Based on the previous theoretical study of Balan ( 2020 ), the variations in the stretching frequency of OH groups associated with Al vacancies in corundum are dominantly controlled by the occupancy of the facing sites. For example, the OH defects associated with two Ti for Al substitutions are observed at a frequency significantly higher than those associated with a single substitution, while the presence of Ti in a neighboring lateral position has almost no effect on the OH stretching frequency of the singly protonated vacancy (Table 2 ). Considering these constraints, the best candidate for the 3278 cm − 1 band is the association of the OH group with a non-dissociated Al Frenkel pair, i.e. the displacement of an Al atom from its structural site to a nearby vacant site, forming a positively charged defect (Fig. 6 ). In this intrinsic defect, the presence of an Al 3+ ion occupying the structurally vacant site facing the OH group would explain the increase in the OH stretching frequency with respect to that of the singly protonated vacancy. Excluding a third configuration leading to a strong oxygen overbonding, two configurations differing by the respective location of the hydrogen atom and interstitial Al atom can occur (Fig. 6 ). The most stable configuration (by 17.5 kJ/mol) has a theoretical stretching frequency 94 cm − 1 higher than that of the (1H + ) Al (Ti 4+ ) Al10 defect, which is assigned to the 3184 cm − 1 band. Accordingly, the frequency of the OH groups associated with the Al Frenkel pair is expected to be ~ 3278 cm − 1 , fully consistent with the experimentally observed frequency (Fig. 7 ). Furthermore, the pairing of hydrogen with the non-dissociated Al Frenkel defect is favored by 2.3 eV, as assessed by combining the total energies of pure corundum, interstitial hydrogen defect and non-dissociated Frenkel pair (Table 2 ). Of note, the theoretical formation energy of the non-dissociated Al Frenkel pair is 5.51 eV, which compares with the value of 4.5 eV previously determined by Matsunaga et al. ( 2003 ). The formation of a non-dissociated Al Frenkel pair involves a very local displacement of an Al atom, consistent with the reversible variations in defect population observed as a function of the cooling rate (Ramirez et al. 2004). At high temperature, the hydrogen is partitioned between this type of defect and other defects involving aliovalent impurities, while the latter are favored at lower temperatures or by slower cooling rates. However, it is noteworthy that the intensity changes observed by Ramirez et al. (2004) do not match a model with only two types of OH defects (excluding the minor proportion of defects related to the weak 3309 cm − 1 band); i.e., the single Ti-bearing defects and the defect associated with a non-dissociated Al Frenkel pair. In this case, the relative intensities of the 3232 and 3184 cm − 1 bands ascribed to the three energetically close single Ti-bearing defects should not be significantly affected. This suggests that an additional OH defect contributes to the 3184 cm − 1 band, which, according to the theoretical frequencies (Table 2 ), could be the singly protonated Al vacancy. This would explain why a minor frequency shift in the 3184 cm − 1 band occurs, together with the relative intensity changes, as a function of the sample cooling rate (Ramirez et al 2004). In line with the findings of Moon and Philips (1991, 1994), a slow quenching rate favors the association of Ti with H-bearing Al vacancies, mostly leading to OH defects associated with a single Ti atom in samples containing a low-Ti concentration and to OH defects associated with two Ti atoms in samples with higher Ti concentrations. In the fast quenched samples, the population of defects is closer to that observed at high temperature and includes the positively charged H-bearing Frenkel defects ((1H + ) Al (Al 3+ ) i ) • and negatively charged H-bearing Al vacancies ((1H + ) Al )'', as well as isolated Ti ions substituted for Al ions ((Ti 4+ ) Al ) • and Ti-associated Al vacancies (V Al (Ti 4+ ) Al )'', corresponding to the following equilibria: ((Al) Al ) × + ((1H + ) Al (Ti 4+ ) Al )' = (V Al (Ti 4+ ) Al )'' + ((1H + ) Al (Al 3+ ) i ) • ((1H + ) Al (Ti 4+ ) Al )' = ((Ti 4+ ) Al ) • + ((1H + ) Al )'' Finally, it is noteworthy that the interpretation of the 3278 cm − 1 band in a-Al 2 O 3 in terms of an OH group associated with the vacant Al site of a non-dissociated Al Frenkel pair also likely holds for the OH band observed at 3269 cm − 1 (corresponding OD band at 2430 cm − 1 ) in the isostructural compound, a-Ga 2 O 3 , after proton implantation (Venzie et al. 2022 ). Interpretation of the 3163 cm − 1 band in Verneuil-grown corundum samples The band at 3163 cm − 1 observed in the R15 sample was occasionally reported in Verneuil-grown corundum samples by Beran ( 1991 ) and by Kronenberg et al. ( 2000 ). This band is sometimes dominant (Choudhary and Vijay 2018 ), especially in corundum samples grown in presence of silicon (Volynets 1972). Notably, this band differs from the band observed at 3161 cm − 1 in natural corundum samples, which is likely related to the charge compensation of the Fe 2+ for Al 3+ substitution (Jollands et al. 2023 ). The present results indicate that the temperature behavior of the 3163 cm − 1 band is almost identical to that of the 3184 cm − 1 band, supporting the idea that the local geometries of the related defects are similar. As stated above, the 3184 cm − 1 band is associated with the (1H + ) Al (Ti 4+ ) Al10 configuration (Table 2 ). Accordingly, selected models of OH defects involving a Si (instead of Ti) for Al substitution were theoretically investigated (Table 2 ). While the most stable Ti bearing defects correspond to three configurations with energies differing by less than 5 kJ/mol, the Si-bearing defect configuration involving the paired octahedral sites ((1H + ) Al (Si 4+ ) Al10 model) is more stable than the two other configurations by more than 20 kJ/mol. Compared with the Ti-bearing (1H + ) Al (Ti 4+ ) Al10 model, the contraction of the vacant Al site related to the smaller ionic radius of the nearby Si 4+ , leads to a 13 cm − 1 lower OH stretching frequency, which is consistent with the experimentally observed frequencies (Fig. 7 ). Although it is difficult to accurately determine the incorporation of Si in corundum due to mass interferences in ICPMS analysis (Emmett et al. 2017 ), the present results are a further indication that the 3163 cm − 1 band observed in the spectrum of Verneuil-grown corundum samples is related to the presence of Si traces in the corundum structure, as proposed by Volynets (1972). Declarations Author contributions. MJ and EB designed the study. EB, KB and MG performed the spectroscopic measurements. EB performed the simulations. All co-authors discussed the results and prepared the manuscript. Code availability. PWscf and PHonon codes (Giannozzi et al. 2009) are available at http://www.quantum-espresso.org/. The pseudo-potentials (Schlipf and Gygi 2015) are available at http://www.quantum-simulation.org/potentials/sg15_oncv/. Structure drawings have been done using the VESTA software (https://jp-minerals.org/vesta/en/; Momma and Izumi 2011). Acknowledgments. Calculations were performed using the HPC resources of the SACADO MeSU platform at Sorbonne Université. Support from the IMPMC spectroscopy platform is acknowledged. References Balan E (2020) Theoretical infrared spectra of OH defects in corundum (á-Al 2 O 3 ). Eur J Mineral 32:457–467. https://doi.org/10.5194/ ejm-32-457-2020 Beran A (1991) Trace hydrogen in Verneuil-grown corundum and its colour varieties - an IR spectroscopic study. Eur J Mineral 3:971-975. https://doi.org/10.1127/ejm/3/6/0971 Beran A., Rossman GR (2006) OH in naturally occurring corundum. Eur J Mineral 18:441-447. https://doi.org/10.1127/0935-1221/2006/0018-0441 Budde M, Parks Cheney C, Lüpke G., Tolk NH, Feldman LC (2001) Vibrational dynamics of bond-center hydrogen in crystalline silicon. Phys. Rev. B: Condens. Matter, 63, 195203. https://doi.org/10.1103/PhysRevB.63.195203 Eigenmann K, Günthard HH (1971) Hydrogen incorporation in doped α-Al 2 O 3 by high temperature redox reactions. Chem Phys Lett 12:12-15. https://doi.org/10.1016/0009-2614(71)80605-X Engstrom H, Bates JB, Wang JC, Abraham MM (1980) Infrared spectra of hydrogen isotopes in α-Al 2 O 3 . Phys Rev B 21:1520-1526. https://doi.org/10.1103/PhysRevB.21.1520 Emmett JL, Stone-Sundberg J, Guan Y, Sun Z (2017). The role of silicon in the color of gem corundum. Gems & Gemology 53: 42–47. Choudhary G., Vijay, S (2018) 3161cm -1 infra-red feature in synthetic sapphires, GIA International Gemological Symposium 2018. Giannozzi P, Baroni S, Bonini N, Calandra M, Car R, Cavazzoni C, Ceresoli D, Chiarotti GL, Cococcioni M, Dabo I, Dal Corso A, de Gironcoli S, Fabris S, Fratesi G, Gebauer R, Gerstmann U, Gougoussis C, Kokalj A, Lazzeri M, Martin-Samos L, Marzari N, Mauri F, Mazzarello R, Paolini S, Pasquarello A, Paulatto L, Sbraccia C, Scandolo S, Sclauzero G, Seitsonen AP, Smogunov A, Umari P, Wentzcovitch RM (2009) Quantum ESPRESSO: a modular and open-source software project for quantum simulations of materials. J Phys: Cond Mat 21:395502. https://doi.org/10.1088/0953-8984/21/39/395502 Hamann DR (2013) Optimized norm-conserving Vanderbilt pseudopotentials. Phys Rev B 88:085117. https://doi.org/10.1103/PhysRevB.88.085117 Holder A., Osborn KD, Lobb CJ, Musgrave C (2013). Bulk and Surface Tunneling Hydrogen Defects in Alumina. Phys Rev Lett 111. http://doi.org/10.1103/physrevlett.111.065901 Jollands MC (2024) Diffusivity of Al vacancies in corundum (a-Al2O3) Submitted to Journal of American Ceramic Society Jollands MC., Jin S, Curti M, Guillaumet M, Béneut M, Giura P, Balan E (2023) Vibrational properties of OH groups associated to divalent cations in corundum (α-Al 2 O 3 ). European Journal of Mineralogy, 35:873-890. https://doi.org/10.5194/ejm-35-873-2023 Kronenberg AK, Castaing J, Mitchell TE, Kirby SH (2000) Hydrogen defects in α-Al 2 O 3 and water weakening of sapphire and alumina ceramics between 600 and 1000°C - I. Infrared characterization of defects. Acta Mater 48:1481-1494. Martin KR, Blaney P, Shi G, Stavola M, Fowler WB (2006) Temperature dependence of the vibrational spectrum of a Li-OH complex in ZnO: infrared absorption experiments and theory. Phys Rev B: Condens Matter 73:232509. https://doi.org/10.1103/PhysRevB.73.235209 Matsunaga T, Tanaka T, Yamamoto T, Ikuhara Y (2003) First principles calculations of intrinsic defects in Al 2 O 3 . Phys Rev B 68:085110. https://doi.org/10.1103/PhysRevB.68.085110 Momma K, Izumi F (2011) VESTA 3 for three-dimensional visualization of crystal, volumetric and morphology data. J Appl Crystallogr 44:1272-1276. https://doi.org/10.1107/S0021889811038970 Moon AR, Phillips MR (1991) Defect clustering in H,Ti: α-Al 2 O 3 . J Phys Chem Solids 52:1087-1099. https://doi.org/10.1016/0022-3697(91)90042-X Moon AR, Phillips MR (1994) Defect clustering and color in Fe,Ti: α-Al 2 O 3 . J Amer Ceram Soc 77:356-367. https://doi.org/10.1111/j.1151-2916.1994.tb07003.x Perdew JP, Burke K, Ernzerhof M (1996) Generalized gradient approximation made simple. Phys Rev Lett 77:3865-3868. https://doi.org/10.1103/PhysRevLett.77.3865 Persson BNJ, Ryberg R (1985a) Vibrational phase relaxation at surfaces: CO on Ni(111). Phys Rev Lett 54:2119–2122. Persson BNJ, Ryberg R (1985b): Brownian motion and vibrational phase relaxation at surfaces: CO on Ni(111). Phys Rev B: Condens Matter 32:3586–3596. Ramírez R, González R, Colera I, Chen (1997) Electric-field-enhanced diffusion of deuterons and protons in a-Al 2 O 3 crystals. Phys Rev B 55: 237-242. Ramírez R, Colera I, González R, Chen Y, Kokta MR (2004) Hydrogen-isotope transport induced by an electric field in a-Al 2 O 3 single crystals. Phys Rev B 69:014302, https://doi.org/10.1103/PhysRevB.69.014302 Schlipf M, Gygi F (2015) Optimization algorithm for the generation of ONCV pseudopotentials. Comput Phys Comm 196:36. https://doi.org/10.1016/j.cpc.2015.05.011 Shelby RM, Harris CB, Cornelius PA (1979) The origin of vibrational dephasing of polyatomic molecules in condensed phases. J Chem Phys 70:34–41. https://doi.org/10.1063/1.437197 Soonthorntantikul W, Khowpong C, Atikarnsakul U, Saeseaw S, Sangsawong S, Vertriest W, Palke A (2019) Observations on the heat treatment of basalt-related blue sapphires, Gemological Institute of America Report, 60 pp. Suezawa M, Fukata N, Saito M, Yamada-Kaneta H (2001) Temperature dependences of line widths and peak positions of optical absorption peaks due to localized vibration of hydrogen Si. Physica B 308–310:220–223. https://doi.org/10.1016/s0921-4526(01)00728-1 Suezawa M, Fukata N, Saito M, Yamada- Kaneta H (2002) Infrared spectra of hydrogen bound to group-III acceptors in Si: Homogeneous line broadening and sidebands. Phys Rev B 65:075214 2002 T-Thienprasert J, Boonchun A, Reunchan P, Limpijumnong S (2017) Identification of hydrogen defects in α-Al 2 O 3 by first-principles local vibration mode calculations. Phys Rev B 95:134103. https://doi.org/10.1103/PhysRevB.95.134103 Turner TJ, Crawford JH Jr (1975) V centers in single crystal Al 2 O 3 , Solid State Comm 17:167-169. Venzie A., Portoff A, Stavola M, Fowler WB, Kim J, Jeon D-W, Park J-H, Pearton SJ (2022) H trapping at the metastable cation vacancy in α-Ga 2 O 3 and α-Al 2 O 3 . Appl Phys Lett 120:192101. https://doi.org/10.1063/5.0094707 Volynets FK, Vorob'ev VG, Sidorova YA (1969) Infrared absorption bands in corundum crystals. Zhurnal Prikladnoi Spektroskopii 10:981-984. Volynets FK, Sidorova YA, Stsepuro NA (1972) OH-groups in corundum crystals grown by the Verneuil technique. J Appl Spectr 17:1088-1091. Wojdyr M (2010) Fityk: a general-purpose peak fitting program. J Appl Crystallogr 43:1126–1128. https://doi.org/10.1107/S0021889810030499 Additional Declarations No competing interests reported. Supplementary Files Supplementaryinformation.zip Cite Share Download PDF Status: Published Journal Publication published 17 Nov, 2024 Read the published version in Physics and Chemistry of Minerals → Version 1 posted Editorial decision: Revision requested 09 Oct, 2024 Reviews received at journal 09 Oct, 2024 Reviewers agreed at journal 03 Oct, 2024 Reviews received at journal 01 Oct, 2024 Reviewers agreed at journal 24 Sep, 2024 Reviews received at journal 18 Sep, 2024 Reviewers agreed at journal 16 Sep, 2024 Reviewers agreed at journal 15 Sep, 2024 Reviewers invited by journal 14 Sep, 2024 Editor assigned by journal 14 Sep, 2024 Submission checks completed at journal 14 Sep, 2024 First submitted to journal 07 Sep, 2024 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5050116","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":361295243,"identity":"cf6cfcf2-7864-4b7b-aeaa-e0a3666406e7","order_by":0,"name":"Etienne Balan","email":"data:image/png;base64,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","orcid":"","institution":"Sorbonne University","correspondingAuthor":true,"prefix":"","firstName":"Etienne","middleName":"","lastName":"Balan","suffix":""},{"id":361295244,"identity":"591a2de7-9ab5-4377-a448-eb638291d32b","order_by":1,"name":"Michael C. Jollands","email":"","orcid":"","institution":"Gemological Institute of America","correspondingAuthor":false,"prefix":"","firstName":"Michael","middleName":"C.","lastName":"Jollands","suffix":""},{"id":361295245,"identity":"58b90a23-071d-44bf-8c69-84245ae6e097","order_by":2,"name":"Maxime Guillaumet","email":"","orcid":"","institution":"Sorbonne University","correspondingAuthor":false,"prefix":"","firstName":"Maxime","middleName":"","lastName":"Guillaumet","suffix":""},{"id":361295246,"identity":"c98bbf29-527d-4f50-b884-349c8ca03d14","order_by":3,"name":"Keevin Béneut","email":"","orcid":"","institution":"Sorbonne University","correspondingAuthor":false,"prefix":"","firstName":"Keevin","middleName":"","lastName":"Béneut","suffix":""}],"badges":[],"createdAt":"2024-09-07 18:46:19","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5050116/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5050116/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s00269-024-01301-9","type":"published","date":"2024-11-17T15:56:57+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":65801951,"identity":"e13560c9-a999-4098-ac80-04ff6086e6bd","added_by":"auto","created_at":"2024-10-03 00:27:17","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":101152,"visible":true,"origin":"","legend":"\u003cp\u003eRoom temperature infrared spectra of the Verneuil-grown corundum samples. Each spectrum was decomposed in individual bands using Lorentzian functions. The spectra are vertically shifted for clarity.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-5050116/v1/71e9226f645a76e50ec100f6.png"},{"id":65802066,"identity":"9d7bb2e2-091b-459d-aafe-3a7c890b2b52","added_by":"auto","created_at":"2024-10-03 00:35:18","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":197158,"visible":true,"origin":"","legend":"\u003cp\u003eTransmission FTIR spectra of the R-15 sample recorded at temperatures ranging from 10 to 290 K. The spectra are vertically shifted for clarity.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-5050116/v1/1c0dd67018ff7d6b1aca3e75.png"},{"id":65801952,"identity":"4acb20f7-6b84-4dc6-afbe-caf1a4be846a","added_by":"auto","created_at":"2024-10-03 00:27:18","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":160565,"visible":true,"origin":"","legend":"\u003cp\u003eTransmission FTIR spectra of the R-3 sample recorded at temperatures ranging from 20 to 540 K. The spectra are vertically shifted for clarity.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-5050116/v1/a923a8cef65c6b55250ee85b.png"},{"id":65801954,"identity":"3258b045-83c5-48d9-9850-4e7cf4d7e126","added_by":"auto","created_at":"2024-10-03 00:27:18","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":43795,"visible":true,"origin":"","legend":"\u003cp\u003eRelative shift \u003cstrong\u003eΔΩ\u003c/strong\u003e, FWHM and area of the main bands in the R-15 (left) and R-3 (right) samples reported as a function of temperature: 3309 cm\u003csup\u003e-1\u003c/sup\u003e (full squares), 3278 cm\u003csup\u003e-1\u003c/sup\u003e (open squares), 3232 cm\u003csup\u003e-1\u003c/sup\u003e (circles), 3184 cm\u003csup\u003e-1\u003c/sup\u003e (diamonds), and 3163 cm\u003csup\u003e-1\u003c/sup\u003e (triangles) bands. The dotted lines correspond to the Persson and Ryberg model (Eqs. 1 and 2) using the parameters reported in Table 1.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-5050116/v1/5327cb536f16f594c4594bba.png"},{"id":65801956,"identity":"2353f702-683b-4949-a710-3f16342b8b35","added_by":"auto","created_at":"2024-10-03 00:27:18","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":56796,"visible":true,"origin":"","legend":"\u003cp\u003eRatio of the area of the 3209 and 3232 cm\u003csup\u003e-1\u003c/sup\u003e bands as a function of temperature in samples R-15 (full circles) and R-3 (open circles). The lines correspond to the ratio expected for a Boltzamnn distribution of the corresponding defects, assuming identical absorption coefficients. The solid line corresponds to an energy difference of 5.5 kJ/mol while the upper and lower dotted lines correspond to energy differences of 4 and 6 kJ/mol, respectively.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-5050116/v1/043f344aae8d8f86f9e3926b.png"},{"id":65801955,"identity":"91789023-3de3-4cb0-b1dc-e58c69fd84f4","added_by":"auto","created_at":"2024-10-03 00:27:18","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":209993,"visible":true,"origin":"","legend":"\u003cp\u003eViews of the OH defects associated with a non-dissociated Al Frenkel pair in corundum: (1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Al\u003csup\u003e3+\u003c/sup\u003e)\u003csub\u003ei1\u003c/sub\u003e (left), (1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Al\u003csup\u003e3+\u003c/sup\u003e)\u003csub\u003ei2\u003c/sub\u003e (right). Light blue: octahedral site occupied by the interstitial Al atom, deep blue: structural octahedral Al sites, light pink sphere and red line: H atom and OH bond, respectively.\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-5050116/v1/b84f719d66f03fec3d75c6f3.png"},{"id":65801958,"identity":"ada031ec-5d9b-46b8-9918-ebb8e9773549","added_by":"auto","created_at":"2024-10-03 00:27:18","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":54576,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of theoretical (Table 2) and observed frequencies for the relevant defect models. The experimental frequencies are indicated in the figure. The nearly constant difference between the theoretical and experimental frequencies (~115 cm\u003csup\u003e-1\u003c/sup\u003e) is slightly lower than that previously reported by Balan (2020) (~125 cm\u003csup\u003e-1\u003c/sup\u003e) due to the larger 3x3x1 supercell used in the present study.\u003c/p\u003e","description":"","filename":"7.png","url":"https://assets-eu.researchsquare.com/files/rs-5050116/v1/3e72c08ef128790fa94294c5.png"},{"id":69274746,"identity":"c1cc125d-030b-4745-a220-af0b41af98c7","added_by":"auto","created_at":"2024-11-18 16:21:32","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1589296,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5050116/v1/7dd29481-d104-4470-a00b-bddeb28dfba6.pdf"},{"id":65801957,"identity":"33be49a4-1d77-44a0-aa85-c29e15f0699f","added_by":"auto","created_at":"2024-10-03 00:27:18","extension":"zip","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":526701,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementaryinformation.zip","url":"https://assets-eu.researchsquare.com/files/rs-5050116/v1/ad7d3770637d901026f73942.zip"}],"financialInterests":"No competing interests reported.","formattedTitle":"Temperature-dependent FTIR spectroscopy of OH defects in Verneuil-grown corundum (α-Al2O3)","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThe infrared spectra of natural and synthetic corundum (a-Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e) crystals frequently reveal the presence of hydroxyl groups associated with structural defects and substitutional impurities. The absorption bands belonging to the \"3309 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e series\" are observed in synthetic (Eigenman and G\u0026uuml;nthard 1971; Volynets et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e1969\u003c/span\u003e, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e1972\u003c/span\u003e; Beran \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e1991\u003c/span\u003e; Moon and Phillips \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e1991\u003c/span\u003e, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e1994\u003c/span\u003e; Kronenberg et al. \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2000\u003c/span\u003e; Ram\u0026iacute;rez et al. \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e1997\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2004\u003c/span\u003e) and natural (Beran and Rossman \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Soonthorntantikul et al. \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) corundum. They are dominated by well-resolved bands at 3184, 3232, and 3309 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. Weaker bands at 3295 and 3365 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e are also reported as belonging to the same series. All these bands have a polarization mostly parallel to the (001) plane consistent with the bonding of protons to the basal oxygen triangle of Al vacancies (T-Thienprasert et al. \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Balan \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). The variability of the \"3309 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e series\" reflects a diversity of charge compensation mechanisms involving substitutional tetravalent cations in the vicinity of the H-bearing Al vacancy (Moon and Phillips \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e1991\u003c/span\u003e, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e1994\u003c/span\u003e; Ramirez et al. 2004; Balan \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Some synthetic corundum samples can also display an absorption band at 3278 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e (Turner and Crawford \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e1975\u003c/span\u003e; Engstrom et al. \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e1980\u003c/span\u003e; Kronenberg et al. \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2000\u003c/span\u003e; Ram\u0026iacute;rez et al. \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e1997\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Choudhary and Vijay \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) with a polarization parallel to the (001) plane which does not seem to be related to a specific impurity as it occurs in diffusively hydrogenated pure corundum (Jollands \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Another major H incorporation mechanism in corundum, which was not considered in the present study, corresponds to its association with substitutional divalent cations (Jollands et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eDespite the numerous studies reporting the FTIR spectra of OH defects in corundum, the assignment of the absorption bands in terms of specific atomic-scale environments is still open to discussion due to the complexity of the spectra and the diversity of hydrogen incorporation mechanisms. In the present study, further information on the OH defects in corundum are obtained by analyzing the modifications of the infrared absorption spectrum as a function of temperature. The results are analyzed using an exchange mode theory which describes the anharmonic properties of localized vibrational modes in crystals and the interpretation of some of the observed bands is further discussed in the light of theoretical models of OH defects in corundum.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eTwo 3 mm thick slabs of Verneuil-grown corundum crystals produced by Hrand Djévahirdjian, Monthey, Switzerland were investigated (R-3 and R-15 samples). The flat faces of the slabs are parallel to the [001] direction.\u003c/p\u003e \u003cp\u003eUnpolarized transmission FTIR spectra of sample R-15 were recorded from 10 to 290 K with an instrumental resolution of 1 cm\u003csup\u003e− 1\u003c/sup\u003e using a Nicolet 6700 FTIR spectrometer set with an EverGlo source, KBr beamsplitter and DTGS detector. Low-temperature measurements were performed using an ARS CS-204 SI cryocooler fitted with KRS-5 windows. Temperature control was ensured with a Si diode fixed on the sample holder.\u003c/p\u003e \u003cp\u003eUnpolarized transmission FTIR spectra of sample R-3 were recorded from 20 to 540 K on a Bruker IFS 66v/S Fourier transform infrared spectrometer working in vacuum with an instrumental resolution of 2 cm\u003csup\u003e− 1\u003c/sup\u003e. The spectrometer was set with KBr beamsplitter, Globar source and liquid nitrogen-cooled MCT detector. In situ temperature-dependent measurements were performed using a Janus He-cryostat and the temperature was controlled with a thermocouple fixed on the sample holder.\u003c/p\u003e \u003cp\u003eTo analyze in more detail the changes of individual bands as a function of temperature, the spectra were decomposed into individual Lorentzian components (Tables S1 and S2) using the Fityk code (Wojdyr, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). Depending on the spectra, a baseline was subtracted before the fit using a visually anchored spline function or was treated as a polynomial function incorporated in the fit. Only the main bands, whose parameters are expected to be less sensitive to uncertainties related to band overlaps and baseline subtraction, were analyzed in detail. The analysis of the R-3 sample was restricted to temperatures below 400 K for the same reasons.\u003c/p\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eAnharmonic properties of localized vibrational modes in crystals\u003c/h2\u003e \u003cp\u003eVarious mechanisms, which are usually considered independently, contribute to the evolution of the vibrational spectrum of a localized vibrational mode (LVM) with temperature. The line shift is determined by the thermal expansion of the crystal structure and by the anharmonic coupling of the LVM with other vibrational modes of the system. The line broadening is determined by the decay of the macroscopic polarization due to the energy transfer from excited oscillators to the crystal phonon bath (population relaxation with time constant T\u003csub\u003e1\u003c/sub\u003e), and to the loss of phase coherence induced by thermal fluctuations (pure dephasing with time constant T\u003csub\u003e2\u003c/sub\u003e*). The energy transfer to the phonon bath from a high frequency LVM such as the stretching of an isolated OH defect (\u0026gt; 3000 cm\u003csup\u003e− 1\u003c/sup\u003e) implies the emission of a large number (\u0026gt; 3) of lower frequency phonons (\u0026lt; 1000 cm\u003csup\u003e− 1\u003c/sup\u003e). In this case, the contribution of pure dephasing to the line broadening is generally dominant (e.g. Budde et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2001\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eVarious models of pure dephasing were proposed to describe the temperature-dependent vibrational properties of molecular adsorbates at crystal surfaces (Persson and Ryberg 1985 a,b) or polyatomic molecules in condensed phase (Shelby et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e1979\u003c/span\u003e). More sophisticated models were later proposed and successfully applied to LVM in solids (Budde et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2001\u003c/span\u003e, Martin et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). The Persson and Ryberg model uses a minimal number of parameters to capture the salient features of the anharmonic system and provides analytical expressions describing the evolution of the LVM band as a function of temperature. In this model, the LVM frequency is weakly coupled to a lower frequency mode (the \"exchange mode\"), the strength of the coupling being characterized by a parameter dw. The exchange mode interacts in its turn with the phonon bath via a friction parameter η. In the weak coupling limit (|δω| \u0026lt;\u0026lt; η), the following relations describe the temperature-dependence of the frequency shift, ΔΩ= Ω− Ω\u003csub\u003e0\u003c/sub\u003e, and the width G of the high-frequency LVM band:\u003c/p\u003e \u003cp\u003eΔW = δω/(exp(hω\u003csub\u003eex\u003c/sub\u003e/2pkT)-1) (1)\u003c/p\u003e \u003cp\u003eand\u003c/p\u003e \u003cp\u003eΓ=2 (δω\u003csup\u003e2\u003c/sup\u003e/η) exp(hω\u003csub\u003eex\u003c/sub\u003e /2p kT) /(exp(hω\u003csub\u003eex\u003c/sub\u003e/2pkT)-1)\u003csup\u003e2\u003c/sup\u003e (2)\u003c/p\u003e \u003cp\u003ewhere W and W\u003csub\u003e0\u003c/sub\u003e are the angular frequencies of the LVM at finite temperature and in the low temperature limit, respectively, ω\u003csub\u003eex\u003c/sub\u003e is the harmonic angular frequency of the exchange mode, h the Planck constant, k the Boltzmann constant, and T the temperature. As discussed by Budde et al. (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2001\u003c/span\u003e), the dephasing mechanism related to one particular exchange mode is more efficient than the dephasing induced by the direct anharmonic coupling of the LVM with the phonon bath. The weak coupling condition of the Persson and Ryberg model is however not always met by LVM in solids but equations (1) and (2) can still be used as functional expressions to fit the experimental data (e.g., Suezawa et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2001\u003c/span\u003e; \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2002\u003c/span\u003e).\u003c/p\u003e \u003cdiv id=\"Sec4\" class=\"Section3\"\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e "},{"header":"Results : Temperature-dependent FTIR spectroscopy of the R-3 and R-15 samples","content":"\u003cp\u003eThe room temperature spectra of samples R-3 and R-15 display the bands commonly assigned to the \"3309 cm\u003csup\u003e− 1\u003c/sup\u003e\" series with relatively strong 3232 and 3184 cm\u003csup\u003e− 1\u003c/sup\u003e bands and weaker 3309, 3295 and 3209 cm\u003csup\u003e− 1\u003c/sup\u003e bands (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The 3278 cm\u003csup\u003e− 1\u003c/sup\u003e band is also present, with a stronger relative absorbance in the R-3 spectrum. An additional band at 3163 cm\u003csup\u003e− 1\u003c/sup\u003e is observed in the R-15 spectrum and a minor band at 3194 cm\u003csup\u003e− 1\u003c/sup\u003e can be inferred from the fit of the spectrum (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Overall, these spectra are similar to those of the series of Verneuil-grown corundum samples previously reported by Beran (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e1991\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eSpectral changes observed as a function of temperature on the R-15 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e) and R-3 (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e) samples mostly consist in a narrowing and blue-shift of the absorption bands when the temperature is lowered. The minor band at 3194 cm\u003csup\u003e− 1\u003c/sup\u003e inferred from the fit of the RT spectrum of the R-15 sample is well resolved below 150 K. In contrast, the band at 3209 cm\u003csup\u003e− 1\u003c/sup\u003e present in the RT spectra is only observed above 150 K. Measurements performed on the R-3 sample also indicate that the spectral changes affecting the 3309 cm\u003csup\u003e− 1\u003c/sup\u003e series are in the continuity of those observed at lower temperature when the temperature is increased up to 540 K. However, the overlap of broadened bands at high temperature makes the analysis of individual contributions above 400 K challenging (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Of note, the baseline changes at 130 and 150 K in the R-15 spectrum (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e) are likely related to condensation of a thin ice layer at cryogenic temperatures.\u003c/p\u003e\u003cp\u003eBased on the spectral fits, the bands of the 3309 cm\u003csup\u003e− 1\u003c/sup\u003e series exhibit a similar behavior in the R-15 and R-3 samples (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Their frequency decreases by ~ 3 to 4 cm\u003csup\u003e− 1\u003c/sup\u003e and their full width at half maximum (FWHM) increases by ~ 1 to 6 cm\u003csup\u003e− 1\u003c/sup\u003e when the temperature increases from 10 K to 290 K. For temperatures lower than 100 K, the frequency and the width of these bands are almost temperature-independent with the FWHM at saturation ranging between 2 and 4 cm\u003csup\u003e− 1\u003c/sup\u003e. The behavior of the bands at 3278 cm\u003csup\u003e− 1\u003c/sup\u003e (R-3 sample) and 3163 cm\u003csup\u003e− 1\u003c/sup\u003e (R-15 sample) is also similar to that of the bands of the 3309 cm\u003csup\u003e− 1\u003c/sup\u003e series. It is noteworthy that the 3163 cm\u003csup\u003e− 1\u003c/sup\u003e and 3184 cm\u003csup\u003e− 1\u003c/sup\u003e bands in the R-15 sample have almost identical temperature behaviors. Finally, the temperature dependence of the band area is weak (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e), as expected for fundamental infrared absorption bands. The low-temperature area of the 3232 cm\u003csup\u003e− 1\u003c/sup\u003e band is approximately three times larger than that of the 3163 cm\u003csup\u003e− 1\u003c/sup\u003e band.\u003c/p\u003e\u003cp\u003eThe temperature dependence of the absorption bands can be analyzed using the Persson and Ryberg model (Eqs.\u0026nbsp;1 and 2). Given the uncertainties related to the complexity of the spectra, the frequency shifts were first fitted using Eq.\u0026nbsp;1, leading to w\u003csub\u003eex\u003c/sub\u003e and dw (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The exchange mode frequency ω\u003csub\u003eex\u003c/sub\u003e is closely related to the extent of the saturation temperature domain and ranges between 180 and 280 cm\u003csup\u003e− 1\u003c/sup\u003e. The dw parameter is negative with an absolute value ranging between 5 to 11 cm\u003csup\u003e− 1\u003c/sup\u003e. These parameters were subsequently introduced into Eq.\u0026nbsp;2 to fit the variations in FWHM, using h only as a free parameter. For all the bands, a satisfactory fit is obtained for both the shift and the width of the lines as a function of temperature (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). The h parameter ranges between 7 and 100 cm\u003csup\u003e− 1\u003c/sup\u003e and is always superior to |dw| (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). These results suggest that the Persson and Ryberg model provides a reasonable account of the temperature dependence of the bands belonging to the 3309 cm\u003csup\u003e− 1\u003c/sup\u003e series, as well as of that of the bands observed at 3278 and 3163 cm\u003csup\u003e− 1\u003c/sup\u003e. In particular, the same exchange frequency is consistent with both the shift and the broadening of the bands, even though the friction parameter is not largely superior to the anharmonic coupling parameter.\u003c/p\u003e\u003cdiv class=\"gridtable\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eParameters of the Persson and Ryberg model (Eqs.\u0026nbsp;1 and 2) describing the temperature-dependent spectroscopic properties of OH stretching bands in the R-15 and R-3 samples.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e\u003ccolgroup cols=\"7\"\u003e\u003c/colgroup\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003esample\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003ew\u003csub\u003eOH\u003c/sub\u003e (RT)\u003c/p\u003e \u003cp\u003e(cm\u003csup\u003e− 1\u003c/sup\u003e)\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003ew\u003csub\u003e0\u003c/sub\u003e\u003c/p\u003e \u003cp\u003e(cm\u003csup\u003e− 1\u003c/sup\u003e)\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003ew\u003csub\u003eex\u003c/sub\u003e\u003c/p\u003e \u003cp\u003e(cm\u003csup\u003e− 1\u003c/sup\u003e)\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003edw\u003c/p\u003e \u003cp\u003e(cm\u003csup\u003e− 1\u003c/sup\u003e)\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eG\u003csub\u003e0\u003c/sub\u003e\u003c/p\u003e \u003cp\u003e(cm\u003csup\u003e− 1\u003c/sup\u003e)\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eη\u003c/p\u003e \u003cp\u003e(cm\u003csup\u003e− 1\u003c/sup\u003e)\u003c/p\u003e \u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eR-15\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3309\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3312\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e279\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-8\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1.6\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e38\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3232\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3234\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e228\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-5\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e3.4\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e9\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3184\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3188\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e181\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-5\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e2.2\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3163\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3167\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e191\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-6\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e2.2\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e11\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eR-3\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3309\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3312\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e326\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-11\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e2.8\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e92\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3278\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3283\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e238\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-10\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e3.3\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e39\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3232\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3234\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e228\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-5\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e3.9\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e7\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3184\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3188\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e205\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-6\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e3.0\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e11\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/table\u003e\u003c/div\u003e"},{"header":"Discussion: Interpretation of the OH stretching spectrum of the R15 and R3 samples","content":"\u003cp\u003eThe bands observed at 3309, 3232 and 3184 cm\u003csup\u003e− 1\u003c/sup\u003e in the spectra of the R-15 and R-3 samples were unambiguously related to OH defects associated with Al vacancies and nearby Ti\u003csup\u003e4+\u003c/sup\u003e for Al\u003csup\u003e3+\u003c/sup\u003e substitution. As experimentally shown by Moon and Philipps (1991, 1994) and Ramirez et al. (2004), the rapid quenching of samples annealed at high temperature favors a dilute configuration of Ti\u003csup\u003e4+\u003c/sup\u003e cations. This results in an increased intensity of the 3232 and 3184 cm\u003csup\u003e− 1\u003c/sup\u003e bands, which are related to negatively charged defects involving a single Ti\u003csup\u003e4+\u003c/sup\u003e for Al\u003csup\u003e3+\u003c/sup\u003e substitution. In contrast, a slower quenching or a lower annealing temperature favors more clustered and electrostically neutral configurations with two Ti\u003csup\u003e4+\u003c/sup\u003e for Al\u003csup\u003e3+\u003c/sup\u003e substitution in the vicinity of the OH-bearing Al vacancy, leading to an increase in the relative intensity of the 3309 cm\u003csup\u003e− 1\u003c/sup\u003e band. The bands of the 3309 cm\u003csup\u003e− 1\u003c/sup\u003e series are observed in the whole series of Verneuil-grown samples investigated by Beran (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e1991\u003c/span\u003e) as well as in the as grown high purity corundum investigated by Ramirez et al (1997). This suggests that even in nominally pure corundum samples the concentration of Ti is high enough to contribute to the charge compensation of Al vacancies.\u003c/p\u003e\u003cp\u003eThe interpretation of these bands in terms of clustering schemes of OH-bearing Al vacancies and Ti for Al substitutions was latter supported by a theoretical study of OH defects in corundum (Balan \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). The most stable configuration displaying two Ti atoms approximately facing the OH group corresponds to the band at 3309 cm\u003csup\u003e− 1\u003c/sup\u003e. Two configurations with a single Ti atom ((1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl1\u003c/sub\u003e and (1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl2\u003c/sub\u003e models) have close stabilities and OH stretching frequencies. It was thus assumed that the two corresponding bands overlap in the 3232 cm\u003csup\u003e− 1\u003c/sup\u003e band, explaining its stronger relative intensity. The significantly weaker 3184 cm\u003csup\u003e− 1\u003c/sup\u003e band was assigned to a third configuration with a similar stability and lower OH stretching frequency. In this case, the Ti\u003csup\u003e4+\u003c/sup\u003e cation is located at the site forming a pair of face-sharing octahedral sites with the vacant Al site ((1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl10\u003c/sub\u003e model). Finally, the weaker band at 3295 cm\u003csup\u003e− 1\u003c/sup\u003e is ascribed to a less stable configuration with a single Ti atom ((1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl4\u003c/sub\u003e model).\u003c/p\u003e\u003cp\u003eThe temperature-dependent measurements of the R-15 and R-3 samples provide some insight into the anharmonic properties of the defects related to the main bands. In the two samples, the 3309 cm\u003csup\u003e− 1\u003c/sup\u003e band corresponds to a coupling of the LVM with an exchange mode at slightly higher frequency than for the 3232 and 3184 cm\u003csup\u003e− 1\u003c/sup\u003e bands. The friction parameter of the exchange mode is also consistently greater for the 3309 cm\u003csup\u003e− 1\u003c/sup\u003e band than for the two other bands. In comparison, the behaviors of the 3232 and 3184 cm\u003csup\u003e− 1\u003c/sup\u003e bands are similar, which confirms that the anharmonic properties of the OH stretching modes are sensitive to the details of the local defect geometry. However, it is noteworthy that no splitting of the 3232 cm\u003csup\u003e− 1\u003c/sup\u003e band is observed over the whole temperature range, even though its low temperature FWHM is larger than that of the 3309 and 3184 cm\u003csup\u003e− 1\u003c/sup\u003e bands.\u003c/p\u003e\u003ch2\u003eAssignment of the 3209 cm\u003csup\u003e− 1\u003c/sup\u003e band to a Ti-bearing OH defect\u003c/h2\u003e\u003cp\u003eThe temperature-dependent behavior of the weaker 3209 cm\u003csup\u003e− 1\u003c/sup\u003e band is uncommon as its absorbance steadily decreases when the temperature decreases and the band vanishes below 150 K (Figs.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e \u0026amp; \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Consistently, this band, as well as its OD counterpart, are not observed in the spectra of hydrogen or deuterium-treated corundum samples measured at 80 K by Ramirez et al. (1997) but are present in those measured at 295 K (Figs.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e and \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e of Ramirez et al. 1997). This behavior thus differs from that usually expected for a fundamental transition and observed for the other bands of the R-3 and R-15 spectra (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). The presence of a \"hot\" band associated with the stronger 3232 cm\u003csup\u003e− 1\u003c/sup\u003e band is unlikely because the frequency difference is only ~ 20 cm\u003csup\u003e− 1\u003c/sup\u003e, while the 3209 cm\u003csup\u003e− 1\u003c/sup\u003e band appears only above 150 K. These properties, as well as the similar behavior inferred for the corresponding OD band, are hardly reconciled with a phenomenon depending on the anharmonic coupling of the stretching mode with a hypothetical vibrational mode at very low frequency. The 3209 cm\u003csup\u003e− 1\u003c/sup\u003e band more likely corresponds to a fundamental transition and by implication its intensity changes should find their origin in a low-temperature equilibrium between OH configurations with relatively close energies.\u003c/p\u003e\u003cp\u003eA relation of the 3209 cm\u003csup\u003e− 1\u003c/sup\u003e band to Ti-bearing defects is suggested by its association with the 3232 and 3184 cm\u003csup\u003e− 1\u003c/sup\u003e bands in the spectrum of synthetic corundum samples intentionally doped with Ti (Moon and Philips 1991, 1994; Ramirez et al. 2004) or reported as nominally pure (but likely containing a low Ti concentration) (Ramirez et al. 1997; Kronenberg et al. \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2000\u003c/span\u003e). Note that the band is not present in the spectra reported by Eigenman and Gunthard (1971) as they were recorded at the liquid nitrogen temperature. Also considering that the 3232 cm\u003csup\u003e− 1\u003c/sup\u003e band does not split over the whole temperature range (10 K-540 K), which challenges its assignment to two overlapping signals, it is thus appropriate to jointly discuss the presence and the properties of the two 3232 and 3209 cm\u003csup\u003e− 1\u003c/sup\u003e bands in the light of the theoretical properties of the Ti-bearing OH defects, in particular those of the two closely related (1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl1\u003c/sub\u003e and the (1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl2\u003c/sub\u003e configurations. Using a Boltzmann statistics, the intensity variations of the 3232 and 3209 cm\u003csup\u003e− 1\u003c/sup\u003e bands lead to an energy difference of ~ 5.5 kJ/mol (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e), which is close to the theoretical energy difference between the two models (4.9 kJ/mol, Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). In addition, the difference of the theoretical OH stretching frequency between these two models, previously determined at ~ 1 cm\u003csup\u003e− 1\u003c/sup\u003e (Balan \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), increases to ~ 9 cm\u003csup\u003e− 1\u003c/sup\u003e using larger 3x3x1 supercell models (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e), which is closer to the experimentally observed splitting (~ 20 cm\u003csup\u003e− 1\u003c/sup\u003e). Accordingly, the relative intensity changes of the 3232 and 3209 cm\u003csup\u003e− 1\u003c/sup\u003e bands could reflect an equilibrium between the (1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl1\u003c/sub\u003e configuration and the slightly less stable (1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl2\u003c/sub\u003e configuration, respectively.\u003c/p\u003e\u003cp\u003eThis equilibrium between the (1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl1\u003c/sub\u003e and (1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl2\u003c/sub\u003e configurations corresponds to the hopping of the H atom from one oxygen to the other available oxygen atom of the large oxygen triangle which forms the base of the vacant Al site. The third oxygen of the triangle belongs to the coordination sphere of the neighboring Ti ion and is significantly less prone to protonation (Balan \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). The possibility of H hopping in an Al vacancy is consistent with the findings of Holder et al. (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2013\u003c/span\u003e) who predicted an efficient H tunneling between equivalent oxygen sites of the Al vacancy in corundum, while the corresponding potential energy barrier (0.612 eV) is well above the thermal energy at room temperature (0.025 eV). The quantum tunneling of the hydrogen atom within the Al vacancy could thus enable the low-temperature equilibrium between the energetically close (1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl1\u003c/sub\u003e and (1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl2\u003c/sub\u003e configurations. In contrast, the relative intensities of the other bands, i.e. the 3309 and 3184 cm\u003csup\u003e− 1\u003c/sup\u003e bands, are mostly inherited from the high temperature history of the sample (Moon and Philips 1991, 1994, Ramirez et al. 2004). In this interpretation, the stronger intensity of the 3232 cm\u003csup\u003e− 1\u003c/sup\u003e band is thus not explained by the overlap of two different absorption bands of similar intensities, as previously assumed, but rather results from the unquenchable conversion of the (1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl2\u003c/sub\u003e configuration into the slightly more stable (1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl1\u003c/sub\u003e configuration, both configurations occurring with an equivalent probability at high temperature. Due to the mobility of hydrogen within the vacancy, the relative intensities of the 3232 and 3184 cm\u003csup\u003e− 1\u003c/sup\u003e bands at low temperature directly probe the Ti occupancy of the four nearest-neighbor octahedral sites surrounding the Al vacancy, which is consistent with the ~ 3:1 ratio of the band areas observed at low temperature (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e\u003cdiv class=\"gridtable\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eTheoretical properties of selected models of OH defects in corundum obtained using a 3x3x1 supercell and the same numerical parameters as in Balan (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2020\u003c/span\u003e)*. The .cif files of the models are provided in the Supplementary Information.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e\u003ccolgroup cols=\"8\"\u003e\u003c/colgroup\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003emodel\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003echarge\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003etot. energy\u003c/p\u003e \u003cp\u003e(Ry)\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003erel. energy\u003c/p\u003e \u003cp\u003e(kJ/mol)\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003ed\u003csub\u003eOH\u003c/sub\u003e\u003c/p\u003e \u003cp\u003e(Å)\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003ew\u003csub\u003eOH\u003c/sub\u003e\u003c/p\u003e \u003cp\u003e(cm\u003csup\u003e− 1\u003c/sup\u003e)\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003ewe\u003csub\u003exp\u003c/sub\u003e\u003c/p\u003e \u003cp\u003e(cm\u003csup\u003e− 1\u003c/sup\u003e)\u003c/p\u003e \u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSingly protonated Al vacancy\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1H\u003csup\u003e+\u003c/sup\u003e )\u003csub\u003eAl,Oc\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-2\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-19976.6339\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.9929\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e3298\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e3184\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInterstitial proton\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003ei\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+ 1\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-20162.2070\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e1.0071\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e2998\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003en.o.\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProton associated with Al vacancy and one nearby Ti for Al substitution\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl1\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-1\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-19998.7281\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.9903\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e3336\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e3232\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl2\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-1\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-19998.7244\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e4.9\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.9906\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e3327\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e3209\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl10\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-1\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-19998.7249\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e4.2\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.9928\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e3296\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e3184\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProton associated with Al vacancy and two nearby Ti for Al substitutions\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl1\u003c/sub\u003e(Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl2\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-19976.6339\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.9852\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e3418\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e3309\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProton associated with Al vacancy and one nearby Si for Al substitution\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Si\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl1\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-1\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-19891.1170\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e20.7\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.9911\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e3326\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003en.o.\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Si\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl2\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-1\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-19891.1141\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e24.6\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.9909\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e3331\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003en.o.\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Si\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl10\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-1\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-19891.1328\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.9937\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e3283\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e3163\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProton associated with a non-dissociated Al Frenkel pair\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Al\u003csup\u003e3+\u003c/sup\u003e)\u003csub\u003ei1\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+ 1\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-20161.9715\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.9874\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e3390\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e3278\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Al\u003csup\u003e3+\u003c/sup\u003e)\u003csub\u003ei2\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+ 1\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-20161.9582\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e17.5\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.9941\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e3273\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003en.o.\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePure corundum\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-20160.2250\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAl Frenkel pair\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(V)\u003csub\u003eAl\u003c/sub\u003e(Al\u003csup\u003e3+\u003c/sup\u003e)\u003csub\u003ei\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-20159.8198\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/table\u003e\u003c/div\u003e\u003cp\u003e*Calculations were performed using the PWscf and PHonon codes (Giannozzi et al., \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2009\u003c/span\u003e) with norm conserving pseudo-potentials (Hamann, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Schlipf and Gygi, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) and PBE functional (Perdew et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e1996\u003c/span\u003e). The electronic integration was restricted to the G point with 80 Ry and 480 Ry cutoffs for the wavefonction and electron density, respectively.\u003c/p\u003e\u003cp\u003en.o.: not observed\u003c/p\u003e\u003ch2\u003eInterpretation of the 3278 cm\u003csup\u003e− 1\u003c/sup\u003e band in synthetic corundum\u003c/h2\u003e\u003cp\u003eSynthetic corundum samples (Turner and Crawford, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e1975\u003c/span\u003e; Engstrom et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e1980\u003c/span\u003e; Kronenberg et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2000\u003c/span\u003e; Ramírez et al., \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e1997\u003c/span\u003ea, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Choudhary and Vijay, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) can display an absorption band at 3278 cm\u003csup\u003e− 1\u003c/sup\u003e, which is also observed in very pure samples (Jollands \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Its relative intensity is greater after a fast sample cooling, while a slower cooling favors the 3309 and 3232 cm\u003csup\u003e− 1\u003c/sup\u003e bands (Ramirez et al., 2004). The relative intensities of these bands are reversible indefinitely by alternate fast and slow cooling (Ramirez et al. 2004). The 3278 cm\u003csup\u003e− 1\u003c/sup\u003e band in the R-3 sample was strong enough to determine its temperature behavior, which is roughly similar to that of the bands of the 3309 cm\u003csup\u003e− 1\u003c/sup\u003e series (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The observed shift and FWHM variation are consistent with those reported by Engstrom et al. (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e1980\u003c/span\u003e) from measurements at 77 and 300 K.\u003c/p\u003e\u003cp\u003eAs the 3278 cm\u003csup\u003e− 1\u003c/sup\u003e band is observed in very pure corundum samples, this band likely involves an intrinsic defect of corundum, i.e. it is not related to a specific trace impurity. Its polarization almost parallel to the (001) plane (Ramirez et al. 2004; Jollands \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) and its temperature-dependent changes also suggest that it is related to an OH group located in the large oxygen triangle forming the base of a vacant Al site, as for the other defects of the 3309 cm\u003csup\u003e− 1\u003c/sup\u003e series. However, its frequency is too high to be consistent with the OH stretching frequency of a singly protonated vacancy (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e), which implies a modification of the local environment of the vacant Al site. Based on the previous theoretical study of Balan (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), the variations in the stretching frequency of OH groups associated with Al vacancies in corundum are dominantly controlled by the occupancy of the facing sites. For example, the OH defects associated with two Ti for Al substitutions are observed at a frequency significantly higher than those associated with a single substitution, while the presence of Ti in a neighboring lateral position has almost no effect on the OH stretching frequency of the singly protonated vacancy (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eConsidering these constraints, the best candidate for the 3278 cm\u003csup\u003e− 1\u003c/sup\u003e band is the association of the OH group with a non-dissociated Al Frenkel pair, i.e. the displacement of an Al atom from its structural site to a nearby vacant site, forming a positively charged defect (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). In this intrinsic defect, the presence of an Al\u003csup\u003e3+\u003c/sup\u003e ion occupying the structurally vacant site facing the OH group would explain the increase in the OH stretching frequency with respect to that of the singly protonated vacancy. Excluding a third configuration leading to a strong oxygen overbonding, two configurations differing by the respective location of the hydrogen atom and interstitial Al atom can occur (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). The most stable configuration (by 17.5 kJ/mol) has a theoretical stretching frequency 94 cm\u003csup\u003e− 1\u003c/sup\u003e higher than that of the (1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl10\u003c/sub\u003e defect, which is assigned to the 3184 cm\u003csup\u003e− 1\u003c/sup\u003e band. Accordingly, the frequency of the OH groups associated with the Al Frenkel pair is expected to be ~ 3278 cm\u003csup\u003e− 1\u003c/sup\u003e, fully consistent with the experimentally observed frequency (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e). Furthermore, the pairing of hydrogen with the non-dissociated Al Frenkel defect is favored by 2.3 eV, as assessed by combining the total energies of pure corundum, interstitial hydrogen defect and non-dissociated Frenkel pair (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Of note, the theoretical formation energy of the non-dissociated Al Frenkel pair is 5.51 eV, which compares with the value of 4.5 eV previously determined by Matsunaga et al. (\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2003\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe formation of a non-dissociated Al Frenkel pair involves a very local displacement of an Al atom, consistent with the reversible variations in defect population observed as a function of the cooling rate (Ramirez et al. 2004). At high temperature, the hydrogen is partitioned between this type of defect and other defects involving aliovalent impurities, while the latter are favored at lower temperatures or by slower cooling rates. However, it is noteworthy that the intensity changes observed by Ramirez et al. (2004) do not match a model with only two types of OH defects (excluding the minor proportion of defects related to the weak 3309 cm\u003csup\u003e− 1\u003c/sup\u003e band); i.e., the single Ti-bearing defects and the defect associated with a non-dissociated Al Frenkel pair. In this case, the relative intensities of the 3232 and 3184 cm\u003csup\u003e− 1\u003c/sup\u003e bands ascribed to the three energetically close single Ti-bearing defects should not be significantly affected. This suggests that an additional OH defect contributes to the 3184 cm\u003csup\u003e− 1\u003c/sup\u003e band, which, according to the theoretical frequencies (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e), could be the singly protonated Al vacancy. This would explain why a minor frequency shift in the 3184 cm\u003csup\u003e− 1\u003c/sup\u003e band occurs, together with the relative intensity changes, as a function of the sample cooling rate (Ramirez et al 2004). In line with the findings of Moon and Philips (1991, 1994), a slow quenching rate favors the association of Ti with H-bearing Al vacancies, mostly leading to OH defects associated with a single Ti atom in samples containing a low-Ti concentration and to OH defects associated with two Ti atoms in samples with higher Ti concentrations. In the fast quenched samples, the population of defects is closer to that observed at high temperature and includes the positively charged H-bearing Frenkel defects ((1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Al\u003csup\u003e3+\u003c/sup\u003e)\u003csub\u003ei\u003c/sub\u003e)\u003csup\u003e•\u003c/sup\u003e and negatively charged H-bearing Al vacancies ((1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e)'', as well as isolated Ti ions substituted for Al ions ((Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e)\u003csup\u003e•\u003c/sup\u003e and Ti-associated Al vacancies (V\u003csub\u003eAl\u003c/sub\u003e(Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e)'', corresponding to the following equilibria:\u003c/p\u003e\u003cp\u003e((Al)\u003csub\u003eAl\u003c/sub\u003e)\u003csup\u003e×\u003c/sup\u003e + ((1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e)' = (V\u003csub\u003eAl\u003c/sub\u003e(Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e)'' + ((1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Al\u003csup\u003e3+\u003c/sup\u003e)\u003csub\u003ei\u003c/sub\u003e)\u003csup\u003e•\u003c/sup\u003e\u003c/p\u003e\u003cp\u003e((1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e)' = ((Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e)\u003csup\u003e•\u003c/sup\u003e + ((1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e)''\u003c/p\u003e\u003cp\u003eFinally, it is noteworthy that the interpretation of the 3278 cm\u003csup\u003e− 1\u003c/sup\u003e band in a-Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e in terms of an OH group associated with the vacant Al site of a non-dissociated Al Frenkel pair also likely holds for the OH band observed at 3269 cm\u003csup\u003e− 1\u003c/sup\u003e (corresponding OD band at 2430 cm\u003csup\u003e− 1\u003c/sup\u003e) in the isostructural compound, a-Ga\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e, after proton implantation (Venzie et al. \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e\u003ch2\u003eInterpretation of the 3163 cm\u003csup\u003e− 1\u003c/sup\u003e band in Verneuil-grown corundum samples\u003c/h2\u003e\u003cp\u003eThe band at 3163 cm\u003csup\u003e− 1\u003c/sup\u003e observed in the R15 sample was occasionally reported in Verneuil-grown corundum samples by Beran (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e1991\u003c/span\u003e) and by Kronenberg et al. (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2000\u003c/span\u003e). This band is sometimes dominant (Choudhary and Vijay \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), especially in corundum samples grown in presence of silicon (Volynets 1972). Notably, this band differs from the band observed at 3161 cm\u003csup\u003e− 1\u003c/sup\u003e in natural corundum samples, which is likely related to the charge compensation of the Fe\u003csup\u003e2+\u003c/sup\u003e for Al\u003csup\u003e3+\u003c/sup\u003e substitution (Jollands et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). The present results indicate that the temperature behavior of the 3163 cm\u003csup\u003e− 1\u003c/sup\u003e band is almost identical to that of the 3184 cm\u003csup\u003e− 1\u003c/sup\u003e band, supporting the idea that the local geometries of the related defects are similar. As stated above, the 3184 cm\u003csup\u003e− 1\u003c/sup\u003e band is associated with the (1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl10\u003c/sub\u003e configuration (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Accordingly, selected models of OH defects involving a Si (instead of Ti) for Al substitution were theoretically investigated (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). While the most stable Ti bearing defects correspond to three configurations with energies differing by less than 5 kJ/mol, the Si-bearing defect configuration involving the paired octahedral sites ((1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Si\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl10\u003c/sub\u003e model) is more stable than the two other configurations by more than 20 kJ/mol. Compared with the Ti-bearing (1H\u003csup\u003e+\u003c/sup\u003e)\u003csub\u003eAl\u003c/sub\u003e(Ti\u003csup\u003e4+\u003c/sup\u003e)\u003csub\u003eAl10\u003c/sub\u003e model, the contraction of the vacant Al site related to the smaller ionic radius of the nearby Si\u003csup\u003e4+\u003c/sup\u003e, leads to a 13 cm\u003csup\u003e− 1\u003c/sup\u003e lower OH stretching frequency, which is consistent with the experimentally observed frequencies (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e). Although it is difficult to accurately determine the incorporation of Si in corundum due to mass interferences in ICPMS analysis (Emmett et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), the present results are a further indication that the 3163 cm\u003csup\u003e− 1\u003c/sup\u003e band observed in the spectrum of Verneuil-grown corundum samples is related to the presence of Si traces in the corundum structure, as proposed by Volynets (1972).\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eAuthor contributions. MJ and EB designed the study. EB, KB and MG performed the spectroscopic measurements. EB performed the simulations. All co-authors discussed the results and prepared the manuscript.\u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eCode availability. PWscf and PHonon codes (Giannozzi et al. 2009) are available at http://www.quantum-espresso.org/. The pseudo-potentials (Schlipf and Gygi 2015) are available at http://www.quantum-simulation.org/potentials/sg15_oncv/. Structure drawings have been done using the VESTA software (https://jp-minerals.org/vesta/en/; Momma and Izumi 2011).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAcknowledgments. Calculations were performed using the HPC resources of the SACADO MeSU platform at Sorbonne Universit\u0026eacute;. Support from the IMPMC spectroscopy platform is acknowledged.\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003cp\u003eBalan E (2020) Theoretical infrared spectra of OH defects in corundum (\u0026aacute;-Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e). Eur J Mineral 32:457\u0026ndash;467. https://doi.org/10.5194/ ejm-32-457-2020\u003c/p\u003e\n\u003cp\u003eBeran A (1991) Trace hydrogen in Verneuil-grown corundum and its colour varieties - an IR spectroscopic study. Eur J Mineral 3:971-975.\u0026nbsp;https://doi.org/10.1127/ejm/3/6/0971\u003c/p\u003e\n\u003cp\u003eBeran A., Rossman GR (2006) OH in naturally occurring corundum. Eur J Mineral 18:441-447. https://doi.org/10.1127/0935-1221/2006/0018-0441\u003c/p\u003e\n\u003cp\u003eBudde M, Parks Cheney C, Lüpke G., Tolk NH, Feldman LC (2001) Vibrational dynamics of bond-center hydrogen in crystalline silicon. Phys. Rev. B: Condens. Matter, 63, 195203. https://doi.org/10.1103/PhysRevB.63.195203\u003c/p\u003e\n\u003cp\u003eEigenmann K, Günthard HH (1971) Hydrogen incorporation in doped \u0026alpha;-Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e by high temperature redox reactions. Chem Phys Lett 12:12-15. https://doi.org/10.1016/0009-2614(71)80605-X\u003c/p\u003e\n\u003cp\u003eEngstrom H, Bates JB, Wang JC, Abraham MM (1980) Infrared spectra of hydrogen isotopes in \u0026alpha;-Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e. Phys Rev B 21:1520-1526.\u0026nbsp;https://doi.org/10.1103/PhysRevB.21.1520\u003c/p\u003e\n\u003cp\u003eEmmett JL, Stone-Sundberg J, Guan Y, Sun Z (2017). The role of silicon in the color of gem corundum. Gems \u0026amp; Gemology 53: 42\u0026ndash;47.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eChoudhary G., Vijay, S (2018) 3161cm\u003csup\u003e-1\u003c/sup\u003e infra-red feature in synthetic sapphires, GIA\u003c/p\u003e\n\u003cp\u003eInternational Gemological Symposium 2018.\u003c/p\u003e\n\u003cp\u003eGiannozzi P, Baroni S, Bonini N, Calandra M, Car R, Cavazzoni C, Ceresoli D, Chiarotti GL, Cococcioni M, Dabo I, Dal Corso A, de Gironcoli S, Fabris S, Fratesi G, Gebauer R, Gerstmann U, Gougoussis C, Kokalj A, Lazzeri M, Martin-Samos L, Marzari N, Mauri F, Mazzarello R, Paolini S, Pasquarello A, Paulatto L, Sbraccia C, Scandolo S, Sclauzero G, Seitsonen AP, Smogunov A, Umari P, Wentzcovitch RM (2009) Quantum ESPRESSO: a modular and open-source software project for quantum simulations of materials. J Phys: Cond Mat 21:395502. https://doi.org/10.1088/0953-8984/21/39/395502\u003c/p\u003e\n\u003cp\u003eHamann DR (2013) Optimized norm-conserving Vanderbilt pseudopotentials. Phys Rev B 88:085117. https://doi.org/10.1103/PhysRevB.88.085117\u003c/p\u003e\n\u003cp\u003eHolder A., Osborn KD, Lobb CJ, Musgrave C (2013). Bulk and Surface Tunneling Hydrogen Defects in Alumina. Phys Rev Lett 111. http://doi.org/10.1103/physrevlett.111.065901\u003c/p\u003e\n\u003cp\u003eJollands MC (2024) Diffusivity of Al vacancies in corundum (a-Al2O3) Submitted to Journal of American Ceramic Society\u003c/p\u003e\n\u003cp\u003eJollands MC., Jin S, Curti M, Guillaumet M, B\u0026eacute;neut M, Giura P, Balan E (2023) Vibrational properties of OH groups associated to divalent cations in corundum (\u0026alpha;-Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e). European Journal of Mineralogy, 35:873-890. https://doi.org/10.5194/ejm-35-873-2023\u003c/p\u003e\n\u003cp\u003eKronenberg AK, Castaing J, Mitchell TE, Kirby SH (2000) Hydrogen defects in \u0026alpha;-Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e and water weakening of sapphire and alumina ceramics between 600 and 1000\u0026deg;C - I. Infrared characterization of defects. Acta Mater 48:1481-1494.\u003c/p\u003e\n\u003cp\u003eMartin KR, Blaney P, Shi G, Stavola M, Fowler WB (2006) Temperature dependence of the vibrational spectrum of a Li-OH complex in ZnO: infrared absorption experiments and theory. Phys Rev B: Condens Matter 73:232509.\u0026nbsp;https://doi.org/10.1103/PhysRevB.73.235209\u003c/p\u003e\n\u003cp\u003eMatsunaga T, Tanaka T, Yamamoto T, Ikuhara Y (2003) First principles calculations of intrinsic defects in Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e. Phys Rev B 68:085110.\u0026nbsp;https://doi.org/10.1103/PhysRevB.68.085110\u003c/p\u003e\n\u003cp\u003eMomma K, Izumi F (2011) VESTA 3 for three-dimensional visualization of crystal, volumetric and morphology data. J Appl Crystallogr 44:1272-1276. https://doi.org/10.1107/S0021889811038970\u003c/p\u003e\n\u003cp\u003eMoon AR, Phillips MR (1991) Defect clustering in H,Ti: \u0026alpha;-Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e. J Phys Chem Solids 52:1087-1099. https://doi.org/10.1016/0022-3697(91)90042-X\u003c/p\u003e\n\u003cp\u003eMoon AR, Phillips MR (1994) Defect clustering and color in Fe,Ti: \u0026alpha;-Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e. J Amer Ceram Soc 77:356-367. \u0026nbsp;https://doi.org/10.1111/j.1151-2916.1994.tb07003.x\u003c/p\u003e\n\u003cp\u003ePerdew JP, Burke K, Ernzerhof M (1996) Generalized gradient approximation made simple. Phys Rev Lett 77:3865-3868. https://doi.org/10.1103/PhysRevLett.77.3865\u003c/p\u003e\n\u003cp\u003ePersson BNJ, Ryberg R (1985a) Vibrational phase relaxation at surfaces: CO on Ni(111). Phys Rev Lett 54:2119\u0026ndash;2122.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003ePersson BNJ, Ryberg R (1985b): Brownian motion and vibrational phase relaxation at surfaces: CO on Ni(111). Phys Rev B: Condens Matter 32:3586\u0026ndash;3596.\u003c/p\u003e\n\u003cp\u003eRam\u0026iacute;rez R, Gonz\u0026aacute;lez R, Colera I, Chen (1997) Electric-field-enhanced diffusion of deuterons and protons in\u0026nbsp;a-Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e crystals. Phys Rev B 55: 237-242.\u003c/p\u003e\n\u003cp\u003eRam\u0026iacute;rez R, Colera I, Gonz\u0026aacute;lez R, Chen Y, Kokta MR (2004) Hydrogen-isotope transport induced by an electric field in\u0026nbsp;a-Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e single crystals. Phys Rev B 69:014302, https://doi.org/10.1103/PhysRevB.69.014302\u003c/p\u003e\n\u003cp\u003eSchlipf M, Gygi F (2015) Optimization algorithm for the generation of ONCV pseudopotentials. Comput Phys Comm 196:36. https://doi.org/10.1016/j.cpc.2015.05.011\u003c/p\u003e\n\u003cp\u003eShelby RM, Harris CB, Cornelius PA (1979) The origin of vibrational dephasing of polyatomic molecules in condensed phases. J Chem Phys 70:34\u0026ndash;41. https://doi.org/10.1063/1.437197\u003c/p\u003e\n\u003cp\u003eSoonthorntantikul W, Khowpong C, Atikarnsakul U, Saeseaw S, Sangsawong S, Vertriest W, Palke A (2019) Observations on the heat treatment of basalt-related blue sapphires, Gemological Institute of America Report, 60 pp.\u003c/p\u003e\n\u003cp\u003eSuezawa M, Fukata N, Saito M, Yamada-Kaneta H (2001) Temperature dependences of line widths and peak positions of optical absorption peaks due to localized vibration of hydrogen Si. Physica B 308\u0026ndash;310:220\u0026ndash;223. https://doi.org/10.1016/s0921-4526(01)00728-1\u003c/p\u003e\n\u003cp\u003eSuezawa M, Fukata N, Saito M, Yamada- Kaneta H (2002) Infrared spectra of hydrogen bound to group-III acceptors in Si: Homogeneous line broadening and sidebands. Phys Rev B 65:075214 2002\u003c/p\u003e\n\u003cp\u003eT-Thienprasert J, Boonchun A, Reunchan P, Limpijumnong S (2017) Identification of hydrogen defects in \u0026alpha;-Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e by first-principles local vibration mode calculations. Phys Rev B 95:134103. https://doi.org/10.1103/PhysRevB.95.134103\u003c/p\u003e\n\u003cp\u003eTurner TJ, Crawford JH Jr (1975) V centers in single crystal Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e, Solid State Comm 17:167-169.\u003c/p\u003e\n\u003cp\u003eVenzie A., Portoff A, Stavola M, Fowler WB, \u0026nbsp; Kim J, \u0026nbsp;Jeon D-W, Park J-H, Pearton SJ (2022) H trapping at the metastable cation vacancy in \u0026alpha;-Ga\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e and \u0026alpha;-Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e. Appl Phys Lett 120:192101. https://doi.org/10.1063/5.0094707\u003c/p\u003e\n\u003cp\u003eVolynets FK, Vorob\u0026apos;ev VG, Sidorova YA (1969) Infrared absorption bands in corundum crystals. Zhurnal Prikladnoi Spektroskopii 10:981-984.\u003c/p\u003e\n\u003cp\u003eVolynets FK, Sidorova YA, Stsepuro NA (1972) OH-groups in corundum crystals grown by the Verneuil technique. J Appl Spectr 17:1088-1091.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWojdyr M (2010) Fityk: a general-purpose peak fitting program. J Appl Crystallogr 43:1126\u0026ndash;1128. https://doi.org/10.1107/S0021889810030499\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"physics-and-chemistry-of-minerals","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"pcmi","sideBox":"Learn more about [Physics and Chemistry of Minerals](http://link.springer.com/journal/269)","snPcode":"269","submissionUrl":"https://submission.nature.com/new-submission/269/3","title":"Physics and Chemistry of Minerals","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-5050116/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5050116/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe temperature dependence of the infrared absorption spectra of two Verneuil-grown corundum samples is investigated in the OH stretching range. The spectra display three main bands at 3184, 3232 and 3309 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, belonging to the so-called \"3309 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e series\", as well as two additional bands at 3163 and 3278 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e previously reported in some synthetic corundum samples. The anharmonic behavior of the observed bands is analyzed using the pure dephasing model of Persson and Ryberg and depends on the local geometry of the OH defects, which are all associated with Al vacancies. A weak band at 3209 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e displays anomalous intensity changes with temperature which support a revised interpretation of both the 3209 and 3232 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e bands. The two bands are interpreted as resulting from the low-temperature equilibrium between two Ti-associated OH defects, enabled by the possibility of hydrogen quantum tunneling within the Al vacancy. The temperature-dependent properties of the 3278 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e band are similar to those of the other Al-vacancy related defects and a comparison with the theoretical properties of selected OH defects suggests that this band corresponds to the association of the H atom with a non-dissociated Al Frenkel pair. Finally, the properties of the band at 3163 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e are consistent with its previously proposed association with Si for Al substitution in corundum.\u003c/p\u003e","manuscriptTitle":"Temperature-dependent FTIR spectroscopy of OH defects in Verneuil-grown corundum (α-Al2O3)","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-10-03 00:27:13","doi":"10.21203/rs.3.rs-5050116/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-10-09T16:59:01+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-10-09T16:37:25+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"130368563837254010187002979112632320049","date":"2024-10-03T14:11:43+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-10-01T21:17:28+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"210111112117927844541871540265877582492","date":"2024-09-24T08:56:42+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-09-18T09:50:08+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"1022845280273738584317685200551195282","date":"2024-09-16T06:22:42+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"26912679431292383330423451495005200558","date":"2024-09-15T12:43:23+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-09-14T19:35:24+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-09-14T07:53:31+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-09-14T07:53:24+00:00","index":"","fulltext":""},{"type":"submitted","content":"Physics and Chemistry of Minerals","date":"2024-09-07T18:45:03+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"physics-and-chemistry-of-minerals","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"pcmi","sideBox":"Learn more about [Physics and Chemistry of Minerals](http://link.springer.com/journal/269)","snPcode":"269","submissionUrl":"https://submission.nature.com/new-submission/269/3","title":"Physics and Chemistry of Minerals","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"b6df8e09-5bd6-4936-8705-7d9a344ce8b5","owner":[],"postedDate":"October 3rd, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2024-11-18T15:59:14+00:00","versionOfRecord":{"articleIdentity":"rs-5050116","link":"https://doi.org/10.1007/s00269-024-01301-9","journal":{"identity":"physics-and-chemistry-of-minerals","isVorOnly":false,"title":"Physics and Chemistry of Minerals"},"publishedOn":"2024-11-17 15:56:57","publishedOnDateReadable":"November 17th, 2024"},"versionCreatedAt":"2024-10-03 00:27:13","video":"","vorDoi":"10.1007/s00269-024-01301-9","vorDoiUrl":"https://doi.org/10.1007/s00269-024-01301-9","workflowStages":[]},"version":"v1","identity":"rs-5050116","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5050116","identity":"rs-5050116","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}