Skip to main content
NCBI home page
ACS Omega logoLink to ACS Omega
. 2024 Feb 26;9(10):11747–11754. doi: 10.1021/acsomega.3c09147

Atomic Layer Deposition of ScF3 and ScxAlyFz Thin Films

Elisa Atosuo †,*, Mikko J Heikkilä †, Johanna Majlund †, Leevi Pesonen †, Miia Mäntymäki †, Kenichiro Mizohata ‡, Markku Leskelä †, Mikko Ritala †,*
PMCID: PMC10938443  PMID: 38496930

Abstract

graphic file with name ao3c09147_0017.jpg

In this paper, we present an ALD process for ScF3 using Sc(thd)3 and NH4F as precursors. This is the first material made by ALD that has a negative thermal expansion over a wide-temperature range. Crystalline films were obtained at the deposition temperatures of 250–375 °C, with a growth per cycle (GPC) increasing along the deposition temperature from 0.16 to 0.23 Å. Saturation of the GPC with respect to precursor pulses and purges was studied at 300 °C. Saturation was achieved with Sc(thd)3, whereas soft saturation was achieved with NH4F. The thickness of the films grows linearly with the number of applied ALD cycles. The F/Sc ratio is 2.9:3.1 as measured by ToF-ERDA. The main impurity is hydrogen with a maximum content of 3.0 at %. Also carbon and oxygen impurities were found in the films with maximum contents of 0.5 and 1.6 at %. The ScF3 process was also combined with an ALD AlF3 process to deposit ScxAlyFz films. In the AlF3 process, AlCl3 and NH4F were used as precursors. It was possible to modify the thermal expansion properties of ScF3 by Al3+ addition. The ScF3 films shrink upon annealing, whereas the ScxAlyFz films show thermal expansion, as measured with HTXRD. The thermal expansion becomes more pronounced as the Al content in the film is increased.

Introduction

Scandium is the lightest element in the series of rare earth metals. Contrary to many lanthanides, however, scandium does not have applications in the field of luminescence due to lack of 4f–4f transitions. Instead, an interesting property of scandium is the negative thermal expansion coefficient of scandium fluoride, encountered over a wide temperature range, approximately from 10 to 1100 K.1

Materials showing negative thermal expansion can counteract the thermal expansion of other materials in delicate applications, such as optical devices and semiconductor components. The counteracting role of ScF3 has been studied for example in Cu/ScF3 core–shell structures, where the motivation was to provide integrated circuit heat sinks with a similar thermal expansion coefficient to silicon.2 ScF3 has also been modified by adding Y3+, Ti3+, Al3+, Fe3+, Zr4+, and Fe3+ + Ga3+ ions into the structure.3−8 By adjusting the ion concentration, thermal expansion coefficients close to zero have been achieved.5−8

To the best of our knowledge, ScF3 has not yet been deposited by atomic layer deposition (ALD). ALD is an advanced thin film deposition method that is based on self-limiting reactions of alternately supplied gaseous precursors. The assets of ALD over many other thin film deposition methods are the stoichiometric, uniform, and conformal films it produces with excellent thickness control and reproducibility.9 Due to these properties, ALD is suitable for uniformly coating, e.g., nanoparticles or introducing dopants into films with uniform distribution. Thus, ALD is an excellent method for producing ScF3-based negative or zero thermal expansion materials.

In this article, an atomic layer deposition (ALD) process is presented for ScF3 using Sc(thd)3 and NH4F as precursors. Upon evaporation, NH4F decomposes to HF and NH3. The use of NH4F in ALD was first presented by Ylilammi and Ranta-aho10 By using NH4F, which is a solid precursor, the handling of commonly used but toxic HF is avoided. Also the metal impurity incorporation that is often observed when using metal fluoride precursors, such as TiF4, is avoided.

The ScF3 process presented here is the first ALD process for a wide-temperature-range negative thermal expansion (NTE) material. The process was also combined with an AlF3 ALD process to deposit ScxAlyFz.

Methods

All of the depositions were done with an F120 cross-flow reactor (ASM Microchemistry Oy). 99.999% nitrogen was used as the carrier and purging gas, and the depositions were done without further gas purification. The operating pressure of the reactor was 10 mbar. Sc(thd)3 (2,2,6,6-tetramethyl-3,5-heptanedionato scandium) was purchased from Volatec Oy, AlCl3 (99%) from Acros Organics and NH4F (≥99.99) from Sigma-Aldrich. All precursors were delivered from glass boats inside the ALD reactor and pulsed with an inert gas valving. The evaporation temperatures for Sc(thd)3, AlCl3 and NH4F were 105, 80, and 70 °C, respectively. The depositions were performed on Si substrates with the native oxide.

The thickness and refractive index of the films were measured by a Film Sense FS-1 Multiwavelength ellipsometer. Cauchy model was used for fitting. X-ray diffraction was measured with a PANalytical X’pert Pro MPD diffractometer, and the diffractograms were analyzed with PANalytical HighScore Plus software (version 4.7). High temperature XRD (HTXRD) was measured in a nitrogen atmosphere with the same diffractometer connected to an Anton Paar HTK1200N furnace. The 99.999% nitrogen was further purified with a MicroTorr MC1–902F gas purifier. The data were Rietveld refined with a MAUD 2.992 software.11

A Hitachi S-4800 field-emission scanning electron microscope (FESEM) was used for the morphology investigation. The same instrument connected to the Oxford INCA 350 microanalysis system was used for qualitative and quantitative energy dispersive X-ray spectroscopy (EDS) measurements. The Sc/Al ratios were determined from the EDS data using ScKα and AlKα lines with GMRFilm software. To increase the conductivity of the samples, a Au/Pd or carbon coating was applied prior to the FESEM and EDS measurements. Morphology of some of the films was further measured with a Veeco Multimode V atomic force microscopy (AFM) instrument. AFM measurements were done in tapping mode, and images were captured in air by using silicon probes with a nominal tip radius of 10 nm and a nominal spring constant of 5 N/m (Tap150 from Bruker). Images were flattened to remove artifacts caused by scanner bow and sample tilt. Roughness was calculated as a root-mean-square value (Rq). The elemental composition was quantitatively determined with time-of-flight elastic recoil detection analyses (ToF-ERDA) using a 5 MV EGP-10-II tandem accelerator. 40 MeV 79Br7+ ions were used as bombarding ions with a detection angle of 40°.

Results and Discussion

ScF3 Films

The ScF3 deposition was first studied at deposition temperatures of 250–375 °C using 1 s pulse and purge lengths for both precursors. Film growth was observed at all temperatures, but the films deposited at 325 °C and higher had poor adhesion to the substrate and were partly flaking off. The effect was more pronounced when the deposition temperature was increased. According to the literature, Sc(thd)3 should be stable against decomposition up to at least 375 °C.12

Considering the flaking of the samples at high deposition temperatures, and on the other hand, the fact that the film purity usually increases as the deposition temperature is increased, the deposition temperature of 300 °C was chosen for the saturation tests. Figure 1a (black squares) depicts the growth per cycle (GPC) as a function of the Sc(thd)3 pulse length. The NH4F pulses and purges were 1 s. The GPC saturates to ∼0.2 Å/cycle with 1 s Sc(thd)3 pulses. The effect of the NH4F pulse length on the GPC was studied while the Sc(thd)3 pulses and purges were kept at 1 s (Figure 1b, black squares). The GPC saturates slowly toward 0.24 Å with 5.0 s of NH4F pulses. The same GPC is obtained also when both purges are 2 s, indicating that a 1 s purge time is sufficient (Figure 1b, red cross). Because the NH4F pulse length was not sufficient in the Sc(thd)3 pulse length studies, pulse lengths of 1 and 2 s were studied for Sc(thd)3 while keeping the NH4F pulses at 5 s (Figure 1a, blue stars). Only a slight increase is seen in the GPC, and 1 s Sc(thd)3 pulses can be considered sufficient.

Figure 1.

Figure 1

Open in a new tab

Growth per cycle as a function of (a) Sc(thd)3 pulse length and (b) NH4F pulse length.

The thicknesses of the films as a function of the number of ALD cycles was studied at a deposition temperature of 300 °C. Even when using a pulsing sequence of 1 s/1 s/3 s/1 s, which is slightly unsaturated with respect to the NH4F pulses, the thickness is linearly dependent on the applied ScF3 cycles (Figure 2).

Figure 2.

Figure 2

Open in a new tab

Thickness of the ScF3 films as a function of the applied ALD cycles.

The GPC values as a function of the deposition temperature are depicted in Figure 3 for films deposited with 1500 cycles. Similar to LiF and GdF3 ALD processes using NH4F as the fluorine source, the GPC increases as a function of the deposition temperature.13,14 The GPC is similar to the GdF3 ALD process where the GPC was 0.22–0.26 Å.14

Figure 3.

Figure 3

Open in a new tab

GPC as a function of the deposition temperature.

Figure 4a shows the grazing incidence X-ray diffractograms (GIXRD) of ∼25–35 nm films deposited at 250–375 °C. All the deposition temperatures resulted in crystalline films. The diffraction patterns of the cubic, hexagonal, and rhombohedral ScF3 overlap, complicating the analysis. The film deposited at 300 °C was measured also in the 2θ-ω mode with 4° offset, which means that only the planes nearly parallel to the film surface are probed and Si substrate peaks are avoided. As seen in Figure 4b, the diffractogram obtained with the 2θ-ω scan is not similar to the reference pattern, indicating a preferred orientation. Thus, the relative intensities of the GIXRD reflections can not be used to identify the phase.

Figure 4.

Figure 4

Open in a new tab

(a) GIXRD of films deposited at 250–375 °C and (b) 2θ-ω measurement (offset 4°) of a film deposited at 300 °C.

The GIXRD data of the film deposited at 300 °C were Rietveld refined using the MAUD software package. Cubic, hexagonal, and rhombohedral crystal forms were attempted, and the best fit to the data was achieved using the cubic form. However, some of the reflections are shifted with respect to the reference patterns, which is an indication that the phase is either strained cubic ScF3 or contains an additional phase. The films were therefore annealed to see whether some relaxation would occur. When kept at 300 °C for 4 h in nitrogen atmosphere, shifts in the positions of some of the reflections occurred improving the fit to the cubic phase though still not matching perfectly. Figure 5 shows the diffractogram, the Rietveld fit, and the quality of the fit (Rwp value) of the annealed film.

Figure 5.

Figure 5

Open in a new tab

Rietveld fit of the XRD data of the ScF3 film. The black circles present the measured data, whereas the red line presents the fitting. The plot below shows the deviation of the fitting from the data.

The composition and stoichiometry of the films deposited at 250–350 °C were determined with ToF-ERDA (Table 1). The F/Sc ratio ranges from 2.9 to 3.1 as expected from the XRD results. The hydrogen content is moderate, max. 3.0 at %, whereas the carbon and oxygen contents are low. Only the carbon content is clearly temperature dependent: the higher the deposition temperature the lower the carbon content. No nitrogen was found in the films, despite NH3 forming upon NH4F decomposition.

Table 1. Stoichiometry and Impurity Contents of ScF3 Films Deposited at 250–350 °C as Measured by ToF-ERDA.

Tdep (°C) Sc F O N C H stoichiometry
250 23.6 ± 0.4 72.0 ± 0.9 1.64 ± 0.10   0.49 ± 0.06 2.3 ± 0.5 3.1
275 24.2 ± 0.5 71.0 ± 1.0 1.33 ± 0.09   0.40 ± 0.06 3.0 ± 0.7 2.9
300 23.9 ± 0.4 73.5 ± 0.9 1.26 ± 0.08   0.23 ± 0.04 1.2 ± 0.3 3.1
325 23.7 ± 0.4 72.7 ± 0.8 1.31 ± 0.09   0.15 ± 0.02 2.2 ± 0.5 3.1
350 23.9 ± 0.5 73.0 ± 0.9 1.22 ± 0.08   0.12 ± 0.02 1.8 ± 0.4 3.1
Open in a new tab

As in previous studies on ALD metal fluoride films, also the ScF3 films eroded fast during the ToF-ERDA measurements.14,15 Interestingly, there was a clear trend in the erosion rates of the measured ScF3 films. The fastest erosion was observed with the film deposited at the lowest temperature and the slowest with the film deposited at the highest temperature. The reasons for this remain unknown and might be related to morphology or roughness. The stoichiometries and impurity contents are similar and are therefore not likely to affect the erosion rates. The densities of the films were measured by XRR from films deposited at 300 and 325 °C. The density of these films is the same, 2.6 g/cm3, and therefore, density is not a probable reason for the different behavior.

The ToF-ERDA depth profiles are not reliable due to fast erosion. However, it is possible to prevent the erosion to some extent by using a capping layer on top of the films.14Figure 6 shows depth profiles of ScF3 films deposited at 275, 300, and 325 °C and capped with ALD-alumina after breaking the vacuum. The capping was performed at 300 °C using the AlCl3 + H2O process. As seen, the depth distributions of scandium and fluoride match well.

Figure 6.

Figure 6

Open in a new tab

Depth profiles of ex situ alumina capped ScF3 films deposited at 275 (∼65 nm), 300 (∼80 nm), and 325 °C (∼80 nm).

Figure 7 shows field-emission scanning electron microscope (FESEM) images of ∼60–85 nm thick films deposited at 250–325 °C. The films contain lamellar-like grains (Figure 7), which is visible especially in the sample deposited at 275 °C. Similar lamellar structures have been observed in other ALD metal fluorides, such as yttrium fluoride and holmium fluoride.16,17

Figure 7.

Figure 7

Open in a new tab

FESEM images of films deposited at 250–325 °C. The thicknesses of the films are 71, 63, 78, and 85 nm for 250, 275, 300, and 325 °C, respectively.

The same films were measured with an atomic force microscope (AFM). The lamellar structure is seen also in AFM as shown in Figure 8 for the film deposited at 275 °C. The root-mean-square roughness (Rq) values are similar in all the films, 5.1 nm for the film deposited at 275 °C and 5.9 nm for the films deposited at 300 and 325 °C (Figure 9). Since the film deposited at the lowest temperature was the thinnest and the film deposited at the highest temperature was the thickest, the roughness seems to decrease with the increasing deposition temperature.

Figure 8.

Figure 8

Open in a new tab

AFM image of a ScF3 film deposited at 275 °C.

Figure 9.

Figure 9

Open in a new tab

AFM images of films deposited at 275–325 °C. The thicknesses of the films are 63, 78, and 85 nm for 275, 300, and 325 °C, respectively.

The thermal expansion of the ScF3 films was studied by HTXRD in a nitrogen atmosphere. The film was preheated for 4 h at 300 °C to relax any internal film stress, as explained earlier. The HTXRD measurements were done in the temperature range of 25–345 °C, because according to preliminary tests, the films oxidize at ∼425 °C and one even at 345 °C. This occurred even though the furnace was preheated at 300 °C in nitrogen atmosphere for 1 h while simultaneously pumping with a turbomolecular pump to remove any moisture remaining in the porous ceramic insulation material.

Figure 10a shows the HTXRD measurements of a 94 nm thick film. The positions of the reflections do not change much. The unit cell parameter was determined at each temperature with the MAUD software (Figure 10b). The lattice constant at room temperature is 4.0087 Å, and it decreases to 4.0054 Å as the temperature is increased, indicating negative thermal expansion, i.e., contraction. The diffractogram, the Rietveld fit and the Rwp value at 185 °C are shown in Figure 11.

Figure 10.

Figure 10

Open in a new tab

(a) HTXRD of a preheated ScF3 film and (b) temperature dependence of the lattice parameter a of the film as modeled using the cubic ScF3 model.

Figure 11.

Figure 11

Open in a new tab

Rietveld fit of the XRD data of the ScF3 film measured at 185 °C. The black circles present the measured data, whereas the red line presents the fitting. The plot below shows the deviation of the fitting from the data.

To our knowledge, there are no previous reports on the thermal expansion properties of ScF3 in thin film form. In this work, the unit cell parameter versus temperature data were fitted with a second order polynomial: a(T) = a0 + a1T + a2T2, where a0, a1 and a2 are the fitting parameters. From the definition of the linear thermal expansion coefficient

graphic file with name ao3c09147_m001.jpg

we obtain the thermal dependence of the lattice constant a as

graphic file with name ao3c09147_m002.jpg

As an example, the linear thermal expansion coefficients of an undoped ScF3 film are −2.4 × 10–6 K–1 at 25 °C and −2.3 × 10–6 K–1 at 52 °C.

In the bulk form, ScF3 is reported to have a linear thermal expansion coefficient α = −3.88 × 10–6 K–1 in the temperature range of 325–675 K (52–402 °C).18 Thermal properties of thin films can differ from the bulk, for example, due to stress caused by the mismatch between the film and the substrate. In addition, the thickness of the film might affect the thermal properties. ScF3 particle size has been shown to affect the thermal expansion coefficient.19

ScxAlyFz Films

From the application point of view, the main aim in studying negative thermal expansion materials is to obtain a zero thermal expansion material. In the literature, these kinds of materials have been obtained for example by the addition of Al3+ ions in ScF3 powder.5 Our aim was to investigate whether Al3+ ions could be incorporated in the ScF3 films by combining the ScF3 ALD process with an AlF3 ALD process.

The original idea was to combine ScF3 cycles with Al(thd)3 + NH4F cycles in a supercycle manner. However, the precursor combination of Al(thd)3 + NH4F did not result in film growth. Earlier it has been shown by Mäntymäki et al. that the combination of Al(thd)3 and TiF4 does not produce any film on Si, and poor-quality films were grown on LiF.20 Therefore, the combination of AlCl3 and NH4F was chosen to be studied. To our knowledge, the combination of AlCl3 and NH4F has not been used in AlF3 ALD processes before. In literature, AlF3 has been deposited by using the precursor combinations of AlCl3 + TiF4, AlMe3 + TaF5, AlMe3 + HF/pyridine, and Al(NMe2)3 + HF.20−24

The growth was studied at a deposition temperature of 300 °C on both Si and ScF3 surfaces. The ScF3 films were approximately 30 nm thick and were deposited at 300 °C. The GPC was ∼0.3 Å on Si and ∼1.3 Å on ScF3 according to ellipsometry. According to GIXRD, the AlF3 films grow amorphous on silicon but are crystalline on ScF3 substrates. The film on ScF3 contains both hexagonal (ICDD PDF 43-435) and rhombohedral (ICDD 44-231) AlF3 (Figure 12). There is an additional unknown reflection present, however. A 2θ-ω scan with a 4° offset was measured to investigate whether the films were orientated. As seen in Figure 12, the rhombohedral phase is orientated whereas the hexagonal phase is hardly seen, indicating that it may be located on the surface.

Figure 12.

Figure 12

Open in a new tab

GIXRD and 2θ-ω measurements on an AlF3 film deposited on ScF3.

ScxAlyFz films were deposited at 300 °C by varying the ratio of the binary cycles of ScF3 and AlF3, e.g., 1000(10(Sc(thd)3 + NH4F) + 1(AlCl3 + NH4F)). In the AlF3 binary cycle, a 1 s/1.5 s/3 s/3 s pulsing sequence was used. The Al content was studied with EDS by comparing the Sc and Al atomic percentages (Figure 13).

Figure 13.

Figure 13

Open in a new tab

Al content in atomic percents as measured by EDS as a function of the Sc/Al binary cycle ratio.

The X-ray diffractograms of the ScxAlyFz films are otherwise similar to those of the ScF3 film, but there is an additional unidentified reflection at the 2θ position of ∼38°. A 116 nm thick film with the Al content of 4.3 at % (Sc/Al ratio 0.83:0.17) was measured with HTXRD after preheating it at 300 °C for 4 h (Figure 14a). The film with the Sc/Al ratio of 0.83:0.17 was chosen, because in the literature close to zero thermal expansion material was obtained when the Sc/Al ratio of a powder was 0.85:0.15.5

Figure 14.

Figure 14

Open in a new tab

(a) HTXRD of a preheated ScxAlyFz film and (b) temperature dependence of the lattice parameter a of the film as modeled using the cubic ScF3 model.

The intensity of the unidentified reflection decreases upon annealing but increases again upon cooling. This might indicate that the additional reflection belongs to rhombohedral ScxAlyFz which changes to cubic ScxAlyFz as has been reported in literature.5 In general, the positions of the other reflections shift to lower angles, indicating that the film expands during annealing (Figure 14a). Attempts were made to fit the data with either cubic or rhombohedral ScF3 or a combination of both. The data was best fitted with the cubic ScF3, and the modeled unit cell parameters are shown in Figure 14b. For example, the lattice parameter at room temperature is 3.9464 Å, and it increases to 3.9586 Å as the temperature is increased to 345 °C. The Rietveld fit of the data at 185 °C is shown as an example (Figure 15).

Figure 15.

Figure 15

Open in a new tab

Rietveld fit of the XRD data of the ScxAlyFz film measured at 185 °C. The black circles present the measured data, whereas the red line presents the fitting. The plot below shows the deviation of the fitting from the data.

Similar HTXRD results were obtained when the film was not preheated. Films with different Sc:Al ratios were therefore measured with HTXRD without preheating. Figure 16a presents the modeled lattice parameters as a function of the temperature for these films. The lattice constant of the film with the largest Al content (Sc/Al ratio 0.65:0.35 and total Al content 8.8 at %) is much smaller than that of the other films. The film with the smallest Al contents 4.3 and 3.2 at % (0.83:0.17 and 0.87:0.13) have the largest lattice constants and they are very close to each other. The third largest lattice constant is measured in a film with 6.2 at % Al content (0.75:0.25) and in close proximity is the film with 5.6 at % Al content (0.78:0.22). The lattice constants of the films with Al contents of 5.6 and 6.2 at % are thus in reverse order than what was expected. This might be explained by the uncertainty in the Sc/Al ratios. The much smaller lattice constant in the film with the largest Al content might in turn be explained by the residual stress in the film.

Figure 16.

Figure 16

Open in a new tab

(a) Lattice constants of films with different Sc/Al ratios as a function of temperature and (b) the change (%) in the lattice constant as compared to room temperature lattice constant in ScF3 film and films with different Sc/Al ratios.

Figure 16b shows the percentual change in the lattice constant as compared to the room temperature lattice constant in films with different Sc/Al ratios. As expected, the percentual change in the lattice constant is the largest in the film with the largest Al content, approximately 0.48% at 345 °C. The films with Al contents of 5.6 and 6.2 at % have very similar values with each other, 0.41 and 0.40%. As expected, the lowest percentual change 0.27% is in the film with the lowest Al content. The percent change in the lattice constant of the undoped ScF3 film is included in the figure for comparison.

The calculated linear thermal expansion coefficients at 25 °C range from 4.0 × 10–6 K–1 for the film with a 3.2 at % Al content to 10.3 × 10–6 K–1 for the film with an 8.8 at % Al content.

Conclusions

An ALD process for ScF3 was studied by using Sc(thd)3 and NH4F as precursors. The films grow crystalline at the studied deposition temperatures of 250–375 °C. The GPC increases along the deposition temperature from 0.16 to 0.23 Å. The F/Sc ratio is 2.9–3.1 as measured by ToF-ERDA. Small hydrogen, carbon and oxygen contents were found in the films, and their maximum contents are 3.0, 0.5, and 1.6 at %, respectively. Nitrogen was not found in the films.

The saturation of the GPC with respect to precursor pulses and purges was studied at 300 °C. Saturation of the GPC was achieved with Sc(thd)3 pulses, but soft saturation was achieved with the NH4F pulses. Linear thickness increase with the number of applied ALD cycles was observed. The film morphology is lamellar type, especially at lower deposition temperatures, as investigated by FESEM. The films are rough; for example, for a 78 nm film deposited at 300 °C the Rq is 5.9 nm.

ScF3 films were also doped with Al3+ by combining Sc(thd)3 + NH4F and AlCl3 + NH4F binary ALD cycles at 300 °C. The ScF3 films were measured with HTXRD and showed a negative thermal expansion. In turn, the ScxAlyFz films expand upon annealing, and the expansion is more pronounced as the Al content increases. It is thus possible to tune the thermal expansion properties of ScF3 by the addition of Al3+ ions to the structure. It is assumed that by finding the right Sc/Al ratio, obtaining a zero thermal expansion material with ALD is possible.

Acknowledgments

The use of ALD Center Finland research infrastructure is acknowledged.

Author Present Address

§ ASM Microchemistry Oy, Pietari Kalmin katu 3, Helsinki 00560, Finland

Author Present Address

∥ Beneq Oy, Olarinluoma 9, Espoo 02200, Finland

The authors declare no competing financial interest.

References

  1. Greve B. K.; Martin K. L.; Lee P. L.; Chupas P. J.; Chapman K. W.; Wilkinson A. P. Pronounced Negative Thermal Expansion from a Simple Structure: Cubic ScF3. J. Am. Chem. Soc. 2010, 132, 15496–15498. 10.1021/ja106711v. [DOI] [PubMed] [Google Scholar]
  2. Xiao N.; Qiao Y. Q.; Shi N. K.; Song Y. Z.; Deng S. Q.; Chen J. Realization of High Thermal Conductivity and Tunable Thermal Expansion in the ScF3 @Cu Core-shell Composites. Sci. China Technol. Sc. 2021, 64, 2057–2065. 10.1007/s11431-020-1758-5. [DOI] [Google Scholar]
  3. Morelock C. R.; Greve B. K.; Gallington L. C.; Chapman K. W.; Wilkinson A. P. Negative Thermal Expansion and Compressibility of Sc1-xYxF3 (x ≤ 0.25). J. Appl. Phys. 2013, 114, 213501. 10.1063/1.4836855. [DOI] [Google Scholar]
  4. Morelock C. R.; Gallington L. C.; Wilkinson A. P. Evolution of Negative Thermal Expansion and Phase Transitions in Sc1-xTixF3. Chem. Mater. 2014, 26, 1936–1940. 10.1021/cm5002048. [DOI] [Google Scholar]
  5. Morelock C. R.; Gallington L. C.; Wilkinson A. P. Solid Solubility, Phase Transitions, Thermal Expansion, and Compressibility in Sc1–xAlxF3. J. Solid State Chem. 2015, 222, 96–102. 10.1016/j.jssc.2014.11.007. [DOI] [Google Scholar]
  6. Qin F.; Chen J.; Aydemir U.; Sanson A.; Wang L.; Pan Z.; Xu J.; Sun C.; Ren Y.; Deng J.; Yu R.; Hu L.; Snyder G. J.; Xing X. Isotropic Zero Thermal Expansion and Local Vibrational Dynamics in (Sc,Fe)F3. Inorg. Chem. 2017, 56, 10840–10843. 10.1021/acs.inorgchem.7b01234. [DOI] [PubMed] [Google Scholar]
  7. Wang T.; Xu J.; Hu L.; Wang W.; Huang R.; Han F.; Pan Z.; Deng J.; Ren Y.; Li L.; Chen J.; Xing X. Tunable Thermal Expansion and Magnetism in Zr-doped ScF3. Appl. Phys. Lett. 2016, 109, 181901. 10.1063/1.4966958. [DOI] [Google Scholar]
  8. Hu L.; Chen J.; Fan L.; Ren Y.; Rong Y.; Pan Z.; Deng J.; Yu R.; Xing X. Zero Thermal Expansion and Ferromagnetism in Cubic Sc1-xMxF3 (M = Ga, Fe) Over a Wide Temperature Range. J. Am. Chem. Soc. 2014, 136, 13566–13569. 10.1021/ja5077487. [DOI] [PubMed] [Google Scholar]
  9. Leskelä M.; Ritala M. Atomic Layer Deposition Chemistry: Recent Developments and Future Challenges. Angew. Chem. Int. Edition 2003, 42, 5548–5554. 10.1002/anie.200301652. [DOI] [PubMed] [Google Scholar]
  10. Ylilammi M.; Ranta-aho T. Metal Fluoride Thin Films Prepared by Atomic Layer Deposition. J. Electrochem. Soc. 1994, 141, 1278–1284. 10.1149/1.2054910. [DOI] [Google Scholar]
  11. Lutterotti L.; Chateigner D.; Ferrari S.; Ricote J. Texture, Residual Stress and Structural Analysis of Thin Films Using a Combined X-ray Analysis. Thin Solid Films 2004, 450, 34–41. 10.1016/j.tsf.2003.10.150. [DOI] [Google Scholar]
  12. Putkonen M.; Nieminen M.; Niinistö J.; Niinistö L.; Sajavaara T. Surface-controlled Deposition of Sc2O3 Thin Films by Atomic Layer Epitaxy Using β-diketonate and Organometallic Precursors. Chem. Mater. 2001, 13, 4701–4707. 10.1021/cm011138z. [DOI] [Google Scholar]
  13. Kvalvik J. N.; Kvamme K. B.; Almaas K.; Ruud A.; Sønsteby H. H.; Nilsen O. LiF by Atomic Layer Deposition—Made Easy. J. Vac. Sci. Technol. A 2020, 38, 050401 10.1116/6.0000314. [DOI] [Google Scholar]
  14. Atosuo E.; Mizohata K.; Mattinen M.; Mäntymäki M.; Vehkamäki M.; Leskelä M.; Ritala M. Atomic Layer Deposition of GdF3 Thin Films. J. Vac. Sci. Technol. A 2022, 40, 022402 10.1116/6.0001629. [DOI] [Google Scholar]
  15. Atosuo E.; Ojala J.; Heikkilä M. J.; Mattinen M.; Mizohata K.; Räisänen J.; Leskelä M.; Ritala M. Atomic Layer Deposition of TbF3 Thin Films. J. Vac. Sci. Technol. A 2021, 39, 022404 10.1116/6.0000790. [DOI] [Google Scholar]
  16. Pilvi T.; Puukilainen E.; Munnik F.; Leskelä M.; Ritala M. ALD of YF3 Thin Films from TiF4 and Y(thd)3 Precursors. Chem. Vapor Depos. 2009, 15, 27–32. 10.1002/cvde.200806721. [DOI] [Google Scholar]
  17. Atosuo E.; Mäntymäki M.; Pesonen L.; Mizohata K.; Hatanpää T.; Leskelä M.; Ritala M. Atomic Layer Deposition of CoF2, NiF2 and HoF3 Thin Films. Dalton Trans. 2023, 52, 10844–10854. 10.1039/D3DT01717F. [DOI] [PubMed] [Google Scholar]
  18. Hu L.; Qin F.; Sanson A.; Huang L.-F.; Pan Z.; Li Q.; Sun Q.; Wang L.; Guo F.; Aydemir U.; Ren Y.; Sun C.; Deng J.; Aquilanti G.; Rondinelli J. M.; Chen J.; Xing X. Localized Symmetry Breaking for Tuning Thermal Expansion in ScF3 Nanoscale Frameworks. J. Am. Chem. Soc. 2018, 140, 4477–4480. 10.1021/jacs.8b00885. [DOI] [PubMed] [Google Scholar]
  19. Yang C.; Tong P.; Lin J. C.; Guo X. G.; Zhang K.; Wang M.; Wu Y.; Lin S.; Huang P. C.; Xu W.; Song W. H.; Sun Y. P. Size Effects on Negative Thermal Expansion in Cubic ScF3. Appl. Phys. Lett. 2016, 109, 023110 10.1063/1.4959083. [DOI] [Google Scholar]
  20. Mäntymäki M.; Heikkilä M. J.; Puukilainen E.; Mizohata K.; Marchand B.; Räisänen J.; Ritala M.; Leskelä M. Atomic Layer Deposition of AlF3 Thin Films Using Halide Precursors. Chem. Mater. 2015, 27, 604–611. 10.1021/cm504238f. [DOI] [Google Scholar]
  21. Jackson D. H. K.; Laskar M. R.; Fang S.; Xu S.; Ellis R. G.; Li X.; Dreibelbis M.; Babcock S. E.; Mahanthappa M. K.; Morgan D.; Hamers R. J.; Kuech T. F. Optimizing AlF 3 Atomic Layer Deposition Using Trimethylaluminum and TaF5: Application to High Voltage Li-ion Battery Cathodes. J. Vac. Sci. Technol. A 2016, 34, 031503 10.1116/1.4943385. [DOI] [Google Scholar]
  22. Lee Y.; DuMont J. W.; Cavanagh A. S.; George S. M. Atomic Layer Deposition of AlF3 Using Trimethylaluminum and Hydrogen Fluoride. J. Phys. Chem. C 2015, 119, 14185–14194. 10.1021/acs.jpcc.5b02625. [DOI] [Google Scholar]
  23. DuMont J. W.; George S. M. Competition Between Al2O3 Atomic Layer Etching and AlF3 Atomic Layer Deposition Using Sequential Exposures of Trimethylaluminum and Hydrogen Fluoride. J. Chem. Phys. 2017, 146, 052819 10.1063/1.4973310. [DOI] [PubMed] [Google Scholar]
  24. Jones J.-P.; Hennessy J.; Billings K. J.; Krause F. C.; Pasalic J.; Bugga R. V. Communication—Atomic Layer Deposition of Aluminum Fluoride for Lithium Metal Anodes. J. Electrochem. Soc. 2020, 167, 060502 10.1149/1945-7111/ab7c39. [DOI] [Google Scholar]

Articles from ACS Omega are provided here courtesy of American Chemical Society

RESOURCES