Skip to main content
NCBI home page
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2024 Jan 16;18(4):2972–2981. doi: 10.1021/acsnano.3c07960

Nonvolatile Isomorphic Valence Transition in SmTe Films

Shogo Hatayama †,‡,*, Shunsuke Mori †, Yuta Saito ‡, Paul J Fons ‡,§, Yi Shuang ∥, Yuji Sutou †,∥,*
PMCID: PMC10832057  PMID: 38228321

Abstract

graphic file with name nn3c07960_0008.jpg

The burgeoning field of optoelectronic devices necessitates a mechanism that gives rise to a large contrast in the electrical and optical properties. A SmTe film with a NaCl-type structure demonstrates significant differences in resistivity (over 105) and band gap (approximately 1.45 eV) between as-deposited and annealed films, even in the absence of a structural transition. The change in the electronic structure and accompanying physical properties is attributed to a rigid-band shift triggered by a valence transition (VT) between Sm2+ and Sm3+. The stress field within the SmTe film appears closely tied to the mixed valence state of Sm, suggesting that stress is a driving force in this VT. By mixing the valence states, the formation energy of the low-resistive state decreases, providing nonvolatility. Moreover, the valence state of Sm can be regulated through annealing and device-operation processes, such as applying voltage and current pulses. This investigation introduces an approach to developing semiconductor materials for optoelectrical applications.

Keywords: SmTe, nonvolatile isomorphic valence transition, electronic structure change, lanthanide monochalcogenide, opto-electrical application, valence state mixing

The electronic structure of a material, which shapes its characteristics, is influenced by the types of atoms present and their arrangement. For instance, despite their similar crystal structures, Si and Ge display differing electronic structures, resulting in distinct band gap (Eg) values.1,2 Additionally, modifications in atomic arrangement that change the symmetry of atomic position, such as those observed in materials undergoing phase transitions between amorphous and crystalline phases, can alter a material’s electronic structure.3 Ge-Sb-Te (GST) compounds are renowned for their physical property contrast due to changes between these phases.4 While drastic atomic rearrangement in GST significantly affects electrical properties such as the resistivity (ρ), it has a limited impact on Eg. The burgeoning field of optoelectronic devices would benefit from a mechanism that can offer substantial contrast in electrical and optical properties.5 In contrast, AM4X8 tetrahedral cluster compounds (A = Ga or Ge; M = V, Nb, or Ta; X = Se or S), which are classified as Mott insulators, undergo a metal–insulator transition (MIT) under hydrostatic pressure.6 These compounds can experience MIT without disrupting the symmetry of the crystal structure. The decreasing interatomic distances and increasing overlap of interatomic orbitals under pressure result in a metallic electronic structure. This implies that atomic species alternation and drastic atomic rearrangement may not be mandatory for altering the electronic structure. Instead, only external pressure is required to induce and sustain a metallic state in these compounds.

Developing a tunable, stable, and modified electronic structure under ambient conditions, devoid of atomic species and arrangement alterations, could introduce a paradigm for electronic structure change (ESC) in materials. To this effect, we focus on a valence fluctuation system (VFS) that enables atoms to mix valence states, resulting in varied electronic structures.7 Certain lanthanide monochalcogenides, including Yb, Tm, Sm, and Eu, with a NaCl structure are known to be VFSs. In these systems, trivalent and divalent states coexist under external pressure, significantly modifying ESCs along with valence states. Figure 1a illustrates ρ changes between ambient conditions (ρamb) and critical pressure (ρc), where ρ drops off in the NaCl structure due to pressure.8−16 For TmS, TmSe, YbTe, YbSe, and YbS, shown by triangle symbols in Figure 1a, the resistivity value at the highest reported pressure was taken as ρc because the pressure range for sufficient valence mixing has not been reported. As this figure suggests, VFSs can significantly alter their electronic properties without any underlying change in atomic arrangement or species. Among these compounds, SmSe and SmTe display a significant resistivity change of around 7 orders of magnitude, making them potential candidates for our research. We focus on SmTe, which exhibits the largest resistivity change, as shown in Figure 1a. Bulk SmTe material has demonstrated to undergo a pressure-induced MIT through the application of hydrostatic pressure as can be seen in Figure 1b.13,17 Under ambient conditions, SmTe, with its NaCl structure, maintains a stable Sm2+ valence state (denoted as B1(Sm2+)). Application of pressure triggers a substantial decrease in resistivity, along with a transformation to a CsCl structure with a stable Sm3+ valence state. Upon pressure release, SmTe reverts to the NaCl structure and ρ returns to its original value. Sm3+ and Sm2+ can coexist in NaCl-type SmTe below the structural transition pressure. As the applied pressure increases, the Sm3+ fraction grows, drastically reducing ρ and mirroring the ESC. Here, NaCl-type SmTe, with a fully reduced ρ under pressure, is denoted as B1(Sm2+/Sm3+). Alam et al. theoretically posited that valence mixing in Ce, a lanthanide element like Sm, can reduce the formation energy.18 This theory suggests the possibility of freezing in the mixed valence state of SmTe under ambient conditions in a metastable state. Sputter deposition, a well-established technique for electronic and optical applications fabrication, can produce residual compressive stress up to 6 GPa in thin films.19 This stress is expected to induce a compressive stress field in a film under ambient conditions. As shown in Figure 1a, SmTe falls within the compressive stress range produced by sputter deposition. This understanding motivated us to fabricate SmTe in the thin-film form to realize a mixed valence state.

Figure 1.

Figure 1

Open in a new tab

Resistivity change behavior of VFSs. (a) Resistivity change between ambient condition (ρamb) and critical pressure (ρc) as a function of critical pressure. Data taken from refs (8−16). The colored circles represent the drop in resistivity observed in the NaCl structure under varying applied pressure. ρc corresponds to the value of the resistivity at the highest pressures reported in the literature for TmS, TmSe, YbTe, YbSe, and YbS, as represented by colored triangles. The yellow hatched region indicates the compressive stress range introduced by sputter deposition. (b) Resistivity as a function of external pressure for the bulk SmTe material. The resistivity points were extracted from Figure 3 in ref (13). The dashed line represents the structural transition pressure for the structural transformation between the NaCl and CsCl structures. The black arrow indicates the critical pressure for the SmTe film.

In this study, we demonstrate the existence of mixed Sm valence states in a non-equilibrium state at ambient conditions within a SmTe thin film. The compressive stress induced during sputter deposition facilitates the stabilization of these mixed valence states in the as-deposited film. Moreover, annealing can adjust the balance of Sm3+ and Sm2+ fractions. Following a valence transition (VT), we observed notable contrasts in ρ (over 105) and Eg (approximately 1.45 eV) between the as-deposited (low ρ) and 440 °C-annealed (high ρ) films despite almost no change in the crystal structure. Furthermore, the mixed valence state can be manipulated within a nanoscale device through Joule heating induced by applying an electrical pulse. These findings propose a paradigm for ESC via VT, enabling the realization of a broad range of material characteristics without alterations in the atomic species or drastic atomic rearrangement, such as amorphous–crystal transition.

Results

SmTe Film Characteristics

Figure 2a depicts the electrical resistance of the as-deposited SmTe film in relation to temperature. The as-deposited SmTe film exhibits a low-resistance state with a room-temperature resistance of approximately 104 Ω. As the temperature rises, the resistance increases significantly at around 400 °C, continuing to increase until around 440 °C. During the cooling process, the resistance progressively increases, surpassing our measurement system’s maximum range (1.2 × 108 Ω) near room temperature. Figure 2b displays X-ray diffraction (XRD) patterns for the as-deposited and 440 °C-annealed films. Both films form in the NaCl structure with a [100] preferred orientation perpendicular to the substrate surface. Despite forming in the NaCl structure, the as-deposited film exhibits a lower resistance than the bulk material. Transmission electron microscopy (TEM) observations confirm that both the as-deposited and 440 °C-annealed films maintain the NaCl crystal structure (Supporting Information Figures S1–S3).

Figure 2.

Figure 2

Open in a new tab

SmTe film characteristics. (a) Temperature dependence of two-terminal resistance for an as-deposited SmTe film. (b) XRD patterns for an as-deposited film, 440 °C-annealed film, and SiN film grown on SiO2/Si substrates. The dashed lines indicate the Bragg reflections from the Si substrate. The reported peak positions for NaCl-type and CsCl-type SmTe bulk materials are displayed below the XRD patterns. Peak position analysis reveals that the lattice constant (a) values, assuming a NaCl structure, for the as-deposited (6.639 Å) and 440 °C-annealed (6.589 Å) films were nearly identical to those of NaCl-SmTe bulk material (∼6.595 Å). Data taken from ref (21). This change in a incorporates the relaxation effect of internal compressive stress introduced by sputter deposition and annealing. (c) Resistivity measured using the van der Pauw method as a function of annealing temperature for SmTe films. The dashed lines indicate the reported resistivity for B1(Sm2+) and B1(Sm2+/Sm3+). Data from refs (13 and 17). (d) Schematic image of the band structures for low- and high-resistive states.

Figure 2c displays ρ at room temperature as a function of the annealing temperature. The value of ρ of the as-deposited film, 1.3 × 10–2 Ω·cm, does not exhibit a large variation up to approximately 400 °C, a trend that aligns with the behavior observed in Figure 2a. After annealing to 440 °C, ρ increases to 1.2 × 103 Ω·cm, a value nearly equivalent to that reported for B1(Sm2+) bulk material.13,17 Typically, Sm2+ is more stable than Sm3+ in NaCl-type SmTe under ambient conditions. As illustrated in Figure 1b, NaCl-type SmTe with Sm3+ (B1(Sm3+)) can be stabilized by applying pressure below 10 GPa, which leads to a decrease in the value of ρ in NaCl-type SmTe as the Sm3+ concentration increases.13,17,20 In fact, the ρ value of the as-deposited film falls between those of B1(Sm2+) and B1(Sm2+/Sm3+), suggesting mixed Sm2+ and Sm3+ valence states. We propose that the compressive stress induced by sputter deposition stabilizes B1(Sm3+) of the as-deposited film, resulting in metastable NaCl-type SmTe with a significant B1(Sm3+) fraction. These findings suggest that in a VFS like SmTe, various film resistances can be achieved through a VT, even in the absence of atomic rearrangement.

Hall effect measurements confirmed n-type conduction across all samples, consistent with reports on B1(Sm2+).22 As the annealing temperature was increased up to 400 °C, the carrier density (n) was found to remain nearly constant, while the mobility (μ) slightly increased, resulting in a gradual decrease in the resistance. In the temperature range above 400 °C, n decreases drastically, while μ experiences only a slight decrease (Figure S4), indicating that changes in n primarily drive the ρ alteration. As previously noted, the as-deposited film, with a substantial fraction of B1(Sm3+) (low-resistivity film), is suggested to transform to a B1(Sm2+)-dominant (high-resistivity) film through annealing. This implies that electron carriers decrease as Sm3+ is reduced to Sm2+, resulting in a substantial increase in ρ. The values of Eg were estimated to be 0.35 and 1.8 eV for the low- and high-resistivity films, respectively (Figure S5). The valence-band spectrum for the 440 °C-annealed film noticeably shifts toward higher binding energies than the as-deposited film (Figure S6). The energy difference between the Fermi level (EF) and the valence-band maximum (EV) was estimated to be 0.32 eV for the as-deposited film and 1.2 eV for the 440 °C-annealed film. The location of EF was above the center of Eg in both films (Figure 2d), indicating n-type conductivity, as confirmed by the Hall effect measurements. In the low-resistive state, the energy difference between EF and the conduction band minimum (EC) was 0.03 eV, facilitating carrier generation. Conversely, the EC–EF value increased to 0.6 eV in the high-resistive state, leading to a decrease in n. These changes in the electronic structure account for the significant observed changes in the electrical characteristics of SmTe.

Residual Stress Change Behavior in SmTe Film

Figure 3a demonstrates nearly identical XRD peak positions for the as-deposited and annealed films, confirming that changes in the electrical and optical properties do not arise from a structural transition. Generally, in sputtered films anisotropic lattice strain is introduced due to the constraints imposed by the substrate. Assuming different Young’s moduli along the [hkl] directions (Yhkl), the residual stress (σ) can be estimated using the modified Williamson–Hall equation as follows:23

graphic file with name nn3c07960_m001.jpg 1

where Bhkl denotes the full width at half-maximum for Bragg reflection of the hkl plane, K is the crystalline-shape factor (assumed as 1), λ signifies the X-ray wavelength used for XRD measurement, and D represents the grain size. σ can be determined from the slope for Bhkl cos θ versus (4 sin θ)/Yhkl (see Supporting Information Section S1). Figure 3b presents the estimated σ as a function of the annealing temperature, where negative and positive signs indicate compressive and tensile stress, respectively. This estimation revealed that the as-deposited film had a compressive stress of 0.27 GPa. However, based on Figures 1b and 2c, achieving mixed valence states with the resistivity observed in the as-deposited film requires a compressive stress of 3–4 GPa, which exceeds the estimated σ. This discrepancy could be attributed to thin film effects on Young’s modulus and MIT behavior. For instance, Yamamoto et al. reported that the Young’s modulus of a Ga-doped ZnO film increased from about 100 to 300 GPa as the film thickness decreased from 500 to 100 nm due to thin-film effects.24 Moreover, the MIT behavior itself is also affected by the thin-film effect, leading to changes in transition temperature, as reported in V2O3.25 According to the theoretical prediction by Kuroda et al.,26 anisotropic strain along the [100] has a greater impact on the reduction of Eg in SmTe than isotropic strain. This also suggests that the [100] orientation of the SmTe film may reduce the stress required to change the physical properties compared to the reported values for the bulk material, leading to a discrepancy from earlier work in the required compressive stress. Although a precise estimation of σ warrants further investigation, these results indicate the successful introduction of compressive stress by sputter deposition. Additionally, σ gradually decreases until an annealing temperature of 300 °C and then abruptly increases starting at 400 °C, resulting in the tensile σ. As the temperature at which ρ begins to increase is nearly identical to the temperature at which σ changes from compressive to tensile, this might suggest that the σ change reflects the occurrence of a VT. The observed abrupt increase in ρ can be attributed to a change in σ, as evident from a comparison of the results shown in Figures 2c and 3b. This suggests that the change in the electronic structure is driven by variations in the stress field within the film.

Figure 3.

Figure 3

Open in a new tab

Residual stress estimation for SmTe films. (a) XRD patterns for the as-deposited and annealed films at different temperatures. The dashed lines indicate the Bragg reflection from the Si substrate. The peak positions remain nearly unchanged across all films, suggesting the absence of a crystal structure transition. (b) Residual stress plotted against the annealing temperature.

Demonstration of Nonvolatile Valence Transition of SmTe Film

The aforementioned findings indicate that the introduction and relaxation of compressive stress are pivotal in VT. To further substantiate this hypothesis, a nanoscale two-terminal device was fabricated (Figure 4a) to demonstrate the nonvolatile control of VT through thermally induced compressive stress (see Figure S8 for structural details). This device structure permits local Joule heating by applying an electrical pulse, followed by quenching. During the heating process, the thermal expansion of the heated area and its subsequent restraint by the surroundings, as depicted in Figure 4a, induce compressive stress. In fact, it was demonstrated in a MnTe-based device that this compressive stress is preserved in ambient conditions through quenching and induces a metastable crystalline phase transformation.27 XRD measurements indicate that the VT of the SmTe film is driven by compressive stress. This suggests that the resistive states of SmTe could be modulated by applying electrical pulses to this device structure. As shown in Figure 4b, the low-resistance state (set state) was switched to the high-resistance state (reset state) by a voltage pulse. Due to the high voltage (more than several tens of volts in this architecture) needed to generate a current in the reset state originating from its high-resistance state, a current pulse was used for the set operation because it is difficult to achieve stable set operation using a voltage pulse. By applying a 500 μs width and 150 μA current pulse, a resistive state nearly equivalent to the initial set state was restored. The resistive contrast in the device is roughly the same magnitude as that in the SmTe film. At this stage, similar resistive switching behavior was obtained in three different cells (Figure S9), indicating that the observed resistive switching is not caused by irreproducible events

Figure 4.

Figure 4

Open in a new tab

Nonvolatile valence transition behavior of SmTe film, reverting from the high-resistive state to the low-resistive state. (a) Schematic of a two-terminal device demonstrating localized heating near the heater electrode surface by applying an electrical pulse (highlighted in orange). The resulting heating induces thermal expansion of the heated region represented by the accompanying red arrows, generating compressive stress (shown by cyan arrows). (b) Measurement of the cell resistance–voltage characteristic using pulse voltage with a 100 ns width and a 200 ns fall time, denoted by red circles. Transitioning to a highly resistive state is achieved by applying a pulse voltage of 10.5 V with a width of 100 ns and a fall time of 200 ns. A pulse current of 150 μA with a width of 500 μs is applied to revert to the low-resistive state. The pulse width of 500 μs is the lower limit of the pulse generator used for the pulse current mode in this study. (c) Cell resistance as a function of the number of valence transitions. The high-resistive state is achieved using a pulse voltage of 10.5 V with a width of 100 ns and a fall time of 200 ns. The low-resistive state is obtained by applying a pulse current of 150 μA with a width of 500 μs. (d) Cross-sectional transmission electron microscopy (TEM) image of the SmTe-based device in the low-resistive state. (e, f) High-resolution TEM (HR-TEM) images captured from the areas indicated by blue- and red-colored open squares, respectively, in panel d. Inset images display the corresponding fast Fourier transform (FFT) patterns obtained from the entire regions of panels e and f.

During reset operation, the SmTe layer is speculated to be heated to the transition temperature of the VT, termed TV, which can be determined by the electronic state, as we will demonstrate in the next section. This heating process induces a VT from Sm3+ to Sm2+. Conversely, while it reverts to the set state, the SmTe layer is heated to a significantly higher temperature than TV. This causes compressive stress due to the difference in thermal expansion coefficients between SmTe and the surrounding material, followed by rapid cooling, or quenching, to restore the non-equilibrium mixed valence state. In the present device structure, the material placed on the heater electrode can be heated to temperatures ranging from 1400–1600 K, as predicted by numerical calculations.28 The introduced compressive stress, σc, can be roughly estimated to be around 1.6 GPa using the formula σc = EαΔT, where E represents the Young’s modulus (89.09 GPa for SmTe bulk material), α denotes the thermal expansion coefficient (16 ppm°C–1 for SmTe bulk material), and ΔT is the temperature difference from room temperature (1100 K).29,30 The computed value of σc is slightly smaller than the value required to achieve the mixed valence states with ρ of the as-deposited film, but it is comparable with σ for the as-deposited film. Considering the thin-film effect and anisotropic strain, the stress necessary for the set operation is within a reasonable range, thus supporting the assumption that compressive stress acts as the driving force for achieving low-resistive states.

As shown in Figure 4c, resistive switching has been demonstrated 20 times. This suggests that the resistive change in the SmTe film is not predicated on irreversible alterations, such as the oxidation of composed elements. Although the high-resistive state is the stable B1(Sm2+), it is crucial to confirm the structure of the reset state. Figure 4d presents a cross-sectional TEM image of a SmTe device in the set state after one cycle of the reset–set operation. Fast Fourier transform (FFT) patterns of both states (Figure 4e,f) reveal that the structure of SmTe above the TiN electrode corresponds to a NaCl structure oriented along the [100] direction. This is because the low-resistive state SmTe largely resembles the NaCl structure, as indicated by the XRD measurement. Of particular significance, the absence of the typical dome-shaped active region commonly observed in melt-quench-type device operations, such as phase change memory, emphasizes that the drastic atomic rearrangement, such as amorphous–crystalline transition, does not occur during the VT switching in SmTe-based devices.

Subsequently, thermal stability was assessed to compare the low-resistive states in the thin film and the device. Figure 5a depicts the change in resistance behavior for the as-deposited film under varying isothermal conditions (350, 360, 370, and 380 °C). With each incremental rise in temperature, the onset time for resistance increase consistently moved toward shorter times. Figure 5b presents the failure time, defined as the time when the resistive change attains a 10% increase from its initial value (Figure S10). The activation energy was determined to be 3.08 eV by analysis of the isothermal resistive changes. The low-resistance state in the as-deposited film remained stable for 10 years at 262 °C (Figure 5b). Even under the isothermal condition of 244 °C, the low-resistance state can be sustained for 100 years. The thermal stability of the device’s low-resistive state was also evaluated through cell resistance measurement at an isothermal temperature of 380 °C (Figure 5c). At 4280 s, the cell resistance suddenly escalated, leading to failure. This failure time mirrors a similar trend observed in the thin film, as highlighted by the blue circle in Figure 5b. In addition to the low-resistive state, it has also been demonstrated that the high-resistive state exhibits a high thermal stability (Figure S11).

Figure 5.

Figure 5

Open in a new tab

Thermal stability for the low-resistive state of SmTe. (a) Resistance variation over time under different isothermal conditions (350, 360, 370, and 380 °C). (b) Failure time as a function of temperature for the as-deposited SmTe film to estimate the thermal stability of the low-resistive state under isothermal conditions. (c) Time-dependent cell resistance for the set state at an isothermal temperature of 380 °C.

These experimental results reinforce the proposition that mixed valence states can be induced by compressive stress through sputter deposition and Joule heating. However, it is important to emphasize that it is difficult to conclude at this stage that the reversible resistance change originates solely from VT, although this is a plausible cause. Further investigation of the resistance change, the corresponding stress distribution in the device, and the effect of grain size dependence with annealing on the VT mechanism is required and is the subject of a future study. It is also important to note that the current device architecture, while similar to two-terminal nonvolatile memory devices, is intended to demonstrate reversible resistive switching and not the superior device performance of SmTe-based devices.

Electronic Structural Changes in SmTe Films

Measurements of the SmTe electronic structure were conducted by using hard X-ray photoelectron spectroscopy (HAXPES). Figure 6a presents the valence-band spectra (VBS). The VBS of B1(Sm2+) bulk material markedly differs from that of the as-deposited film, as demonstrated in Figure 6b.31 Specifically, the Te 5s peak splits into two distinct peaks in the as-deposited film while remaining unified in the B1(Sm2+) bulk material. This suggests that the film’s VBS cannot be described by a single valence state. Although the experimental VBS of B1(Sm3+) has not been reported, density functional calculations have predicted a VBS closely matching that of the B1(Sm2+).32 Therefore, we propose that the VBS between 0 and 8 eV consists of Te 5p and Sm 4f states in both B1(Sm2+) and B1(Sm3+). Numerous multiplets within a narrow energy range complicate the deconvolution of the VBS between 0 and 8 eV. However, owing to its peak position, the Te 5s peak can be deconvolved into two separate contributions corresponding to Te bonding to Sm3+(-Sm3+) and Sm2+ (-Sm2+). Given that Sm3+ is more electronegative than Sm2+,33 the binding energy of Te 5s(-Sm3+) is greater than that of Te 5s(-Sm2+).

Figure 6.

Figure 6

Open in a new tab

Electronic structural changes in SmTe films. (a) Valence-band spectra for the SmTe films. (b) Valence-band spectrum for SmTe bulk material. Data from ref (31). (c) Te 5s spectra for the SmTe films. The pink- and gray-colored peaks correspond to Te 5s bonding to Sm2+ and Sm3+, respectively. (d) Average valence and energy difference between Te 5s energies bonding to Sm2+ and Sm3+ as a function of the annealing temperature. The onset temperature at which zave begins to decrease was defined as Tv. (e) Te 3p core-level spectra for the SmTe films. (f) Binding energy for deconvolved Te 5s peaks as a function of carrier density (n).

Figure 6c shows the deconvolved Te 5s spectra. With increasing annealing temperature, the Te 5s(-Sm2+) peak area expands while the energy difference between the two peaks, ΔE, simultaneously decreases. The average valence of Sm, zave, can be computed using the following equation:31

graphic file with name nn3c07960_m002.jpg 2

where A2 and A3 represent the Te 5s(-Sm2+) and Te 5s(-Sm3+) peak areas, respectively. Figure 6d shows zave and ΔE as functions of the annealing temperature. zave is approximately 2.4 in the as-deposited film and slightly reduces upon annealing above 400 °C. The zave for the low-resistivity films (around 2.4) suggests the presence of a B1(Sm3+) component, consistent with the predictions based on electrical measurements. As stated in the introduction, the valence state mixing of Ce is known to contribute to phase stabilization in Ce2Fe14B.18 Therefore, intermediate zave values between 2 and 3 are believed to result in the high thermal stability of the low-resistive SmTe. zave decreases dramatically with annealing above 420 °C and reaches 2.1 for the 440 °C-annealed film. These findings indicate that annealing initiates a VT from a non-equilibrium mixed valence state to the stable Sm2+-dominant state. Figures 2c, 6d, and S4a show that changes in ρ align with variations in zave, suggesting changes in n result from a Sm VT. These observations affirm that nonvolatile Sm VT leads to significant ECS and considerable changes in electronic properties, absent a structural transition and alteration of atomic species.

Discussion

Resistive Change Mechanism via Valence Transition

Based on the aforementioned results, we proposed a plausible mechanism for reversible resistive change below. Joule heating above Tv triggers VT from the non-equilibrium state to the stable state during the reset operation. For the reversion to the set state, the B1(Sm3+) fraction must be reinstated. As discussed earlier, the B1(Sm3+) state is stabilized at ambient conditions in both the thin film and the device due to deposition and thermally induced compressive stress in the SmTe layer, which results in non-equilibrium mixed valence states. Therefore, the reversible switching in a SmTe-based device can be explained via nonvolatile and reversible Sm VT.

In addition to zave, ΔE is an important factor in determining the electronic structure. The Te 5s(-Sm2+) and Te 5s(-Sm3+) peaks are observed to independently shift in the films (Figure 6c). Moreover, the spectral shape of the Te 5p + Sm 4f peak changes in conjunction with the Te 5s shape (Figure 6a), suggesting that the evolution of the VBS shape is due to a combination of energy levels attributable to Sm2+ (Sm2+ band) and Sm3+ (Sm3+ band). Figure 6e presents Te 3p core-level spectra. The discrepancy in peak shape between the Te 3p and Te 5s peaks in the as-deposited film might arise from differences in multiplet splitting, similar to the peak shape variation among Mn 3s, Mn 3p, and Mn 2p3/2.34 As both the core-level and VBS shift toward higher binding energies with increasing annealing temperature (Figure 6a,e), a rigid-band shift occurs and ΔE changes.35 The rigid-band shift width, ΔEb, can be expressed by the following equation:35

graphic file with name nn3c07960_m003.jpg 3

where K is a constant, ΔQ is the change in valence, and ΔVM and ΔEs are shifts due to changes in Madelung potential and core-carrier screening, respectively. Because the Sm2+ and Sm3+ bands shift independently, ΔQ can be disregarded. Therefore, ΔEb is determined by the sum of ΔEs and ΔVM. Due to mobile carriers, core-carrier screening induces an energy shift toward low binding energies due to mobile carriers. Thus, the magnitude of ΔEs is proportional to n.35 While ΔVM has the opposite sign of ΔEs (see Supporting Information Section S2), it can be observed that the deconvolved Te 5s peaks shift toward lower binding energies with increasing n by comparing Figures S4a and 6c, demonstrating that ΔEb is dominated by ΔEs.

ΔEs induces a shift toward lower energies for electron and hole carriers, the shift being larger for holes than electrons.35 The above Te 5s analysis suggests that the Sm3+ band is situated at energies 1.3 eV lower than the Sm2+ band in the high-resistive state, resulting in the formation of the conduction band minimum (CBM) and valence band maximum (VBM) regions by the Sm3+ 5d and Sm2+ 4f states, respectively (Figure 7a, left). This scenario implies a VT from Sm2+ to Sm3+, denoted as 4f65d0 → 4f55d1,36 that leads to the translocation of an electron from the Sm2+ 4f state to the Sm3+ 5d state. This is due to the Sm3+ 5d level being energetically closer to the VBM than the Sm2+ 5d state (center image of Figure 7a). The electron excitation causes the Sm3+ band to serve as an electron-doped state, while the Sm2+ band transforms into a hole-doped state. If this assumption holds, then the energy values of the Sm2+ band should exhibit a stronger dependence on the value of n than that of the Sm3+ band. To validate this assumption, we plotted the deconvolved Te 5s peak positions as a function of n in Figure 6f. The figure shows that both peak positions shift toward lower energies as n increases. Furthermore, the Te 5s(-Sm2+) peak shift was larger than that of the Te 5s(-Sm3+) peak, indicating that the shifts for the Sm3+ and Sm2+ bands likely result from the core-carrier screening of electrons and holes, respectively.

Figure 7.

Figure 7

Open in a new tab

Reversible valence transition mechanism and accompanying physical properties. (a) Evolution of the averaged density of states (DOS) during the transition from a high-resistive state to a low-resistive state. In a valence-fluctuation system (see ref (37)), where valence states spatially and temporally fluctuate, the DOS for each orbital is schematically shown as an averaged quantity. The percentages of each Sm valence state were estimated using the area fraction of deconvolved Te 5s peaks. (b) Resistivity-change plot as a function of band gap change for various material systems calculated based on the reported band gap values for equilibrium and non-equilibrium states (Figure S12). The red circles indicate a nonvolatile electronic structure change (ESC) with reversible switchability (RS) through atomic rearrangement. Phase-change materials exhibiting phase transitions between amorphous and crystalline phases are plotted (Ge2Sb2Te5,38,39 GeTe,40 Ag-In-Sb-Te,41 GeSb,42,43 Sb2Te3,44,45 Cr2Ge2Te6,46−48 and In3SbTe249,50). The red-colored squares represent nonvolatile ECS without RS, featuring transition metal dichalcogenides undergoing phase transitions (MoS2:1T-2H transition51,52 and MoTe2:1T′-2H transition53,54). The blue circles indicate volatile ESC triggered and maintained by external forces, such as pressure or temperature. Representative examples include insulator and metallic phases of Ti3O5,55 VOx,56−62 and GaTa4Se8.63,64 The orange star indicates the SmTe films.

The ESC during the transition from a high-resistive state to a low-resistive state is schematically depicted as Sm2+ and Sm3+ bands in Figure 7a. The VT promotes an increase in the B1(Sm3+) fraction, causing the Sm2+ 4f electron to be excited into the Sm3+ 5d state (Sm2+ → Sm3+ + e–). This results in hole-doped Sm2+ and electron-doped Sm3+ bands. The excited electron serves as a conduction-band carrier, increasing n. As the VT progresses, hole- and electron-doping intensify, causing an increase in ΔEs due to core-carrier screening. The Sm2+ band undergoes a more substantial shift than the Sm3+ band due to hole-doping (Figure 7a center). Consequently, in the low-resistive state, i.e., the as-deposited film, the Eg value drops to 0.35 eV due to the significant 2.5 eV ΔE value (Figure 7(a) right). The disparities in Eg (∼1.45 eV) and ΔE (∼1.2 eV) between the high- and low-resistive states are comparable. Therefore, the ESC in SmTe films can be interpreted based on a VT and an accompanying rigid-band shift. During the transition from the low-resistive state to the high-resistive state, the Sm3+ state captures an electron from the conduction band, resulting in the valence state reverting to Sm2+. In this transition, the excited electron reverts to the Sm2+4f state, annihilating the doped states. Therefore, the band shift due to core-carrier screening is diminished, inducing a high-resistivity state. This discussion can help elaborate the possible mechanism for the valence state change, consistent with the experimental results. In conclusion, the reversible ESC in SmTe may be VT driven by the above mechanism.

Conclusions

In summary, we examined the VT behavior of the SmTe film. The compressive stress induced in the thin-film form was utilized to realize a mixed valence state in a metastable form, resulting in a low-resistive state. The resistive state was transformed to a high-resistive state during VT facilitated by annealing or Joule heating by an electrical pulse. The possible mechanisms underlying the electronic structure change and reversible resistive switching were elucidated, taking into account the VT and the associated rigid-band shift. Figure 7b delineates the changes in ρ and band gap across various transition systems.38−64 For reversible phase transitions, such as the Mott transition and the amorphous-crystalline phase transition, a significant change is observed in ρ, while the band gap remains relatively unaffected. Conversely, in two-dimensional materials undergoing structural transformations, such as MoS2 and MoTe2, ρ and Eg can experience substantial changes. However, achieving reversible switches in these materials poses a challenge. In contrast, the SmTe thin film examined in this study exhibited substantial changes in electrical resistivity and band gap due to VT while simultaneously enabling reversible switching. This showcases the mechanism of nonvolatile isomorphic VT as a pioneering approach to induce changes in the electronic structure. Furthermore, our findings highlighted the potential for using VT dynamic control in applications such as nonvolatile optoelectronic applications.

Experimental Section

SmTe Film Characterization

SmTe films were deposited on SiO2(100 nm)/Si or quartz substrates using radiofrequency magnetron cosputtering of Sm and Te pure metal targets at room temperature. The base pressure in the chamber was maintained below 5.0 × 10–5 Pa under an Ar atmosphere. Prior to SmTe, a 2 nm thick SiN layer was deposited on the substrates and the SmTe surface was capped with a 5 nm thick SiN layer to inhibit Sm oxidation. The film composition was confirmed as Sm56Te44 through energy dispersive X-ray spectrometry (EDS) conducted with a scanning electron microscope (JSM-7100F, JEOL, Japan). The deviation of the measured composition from the stoichiometric value may be attributed to a relative error of a few percent inherent in the EDS measurement.65 The temperature dependence of the resistance of the as-deposited film was measured by using a two-point probe. The resistance was recorded during annealing up to 440 °C, at a heating rate of 10 °C/min in an Ar atmosphere, followed by cooling to room temperature.

The crystal structure of a 200 nm thick SmTe film, deposited on a SiO2(100 nm)/Si substrate, was determined through XRD with a Cu Kα source at room temperature, utilizing the conventional 2θ/θ Bragg–Brentano geometry (ULTIMA, Rigaku, Japan). For the XRD measurements, the films were annealed at predetermined temperatures of 300, 400, 420, 430, and 440 °C at a heating rate of 10 °C/min and then cooled to room temperature.

The electrical properties of the films were evaluated at room temperature using a Hall effect measurement apparatus (Resitest 8400, Toyo Corp., Japan) in AC Hall effect measurement mode. The SmTe films that were previously assessed for XRD measurements were employed for these electrical property measurements.

HAXPES measurements were performed at beamline BL47XU at SPring-8 (JASRI) by using the SmTe films employed in the XRD measurements. The samples were introduced into the experimental chamber, which maintained a pressure below 1.0 × 10–5 Pa. Photoelectron spectra were collected with an R4000 electron analyzer (VG Scienta, U.K.). The HAXPES measurements were performed at a photon energy of 7.94 keV at room temperature. The binding energy scale was calibrated by measuring the energy position of the Au 4f level from a gold reference sample. The spectra background was subtracted using Shirley’s method, and the Te 5s spectra were fitted with Voigt functions.66

The relative reflectance was measured against an Al reference mirror, and the transmittance of the SmTe films was recorded at room temperature within a wavelength range from 400 to 1100 nm by using a spectrophotometer (V-730, JASCO, Japan). For these measurements, SiN (5 nm)/SmTe (200 nm)/SiN (2 nm) stacked layers were deposited on a quartz substrate. We calculated the absorption coefficient based on the reflectance and transmittance values we gathered. Subsequently, we determined the optical band gap for the SmTe films using this absorption coefficient.

A cross-section of the films and the device in the low-resistive state was analyzed using high-resolution TEM (HR-TEM; ARM200F, JEOL, Japan) to examine the SmTe microstructure using an acceleration voltage of 200 kV. The TEM samples were prepared by using ion milling (PIPS, Gatan, USA) or a focused ion beam (FIB; JIB-4600F, JEOL, Japan).

Two-Terminal Device Cell Fabrication and Electrical Measurements

The SmTe-based two-terminal device cell was constructed as depicted in Figure S8, with the bottom electrode being a square-shaped TiN with an area of 218 × 218 nm2. The TiN electrode surface was subjected to reverse sputtering for 1.5 h to eliminate the native oxide layer prior to the deposition of the SmTe layer. After the SmTe layer deposition, a TiN layer, 15 nm thick, and a W layer, 130 nm thick, were grown on top of the SmTe layer. The constructed device was then annealed to 440 °C to achieve a high-resistive NaCl-type SmTe. Resistance (R)–voltage (V) characteristics were determined using a semiconductor parameter analyzer (B1500A, Keysight, USA). Electrical pulses were directed from the top electrode to the bottom electrode. The device’s resistance was gauged via a current (I)–voltage (V) curve obtained using a voltage sweep of 0.1 V. The high-resistive state was achieved using a pulse shape with a 100 ns pulse width and a 200 ns trailing edge, and the voltage amplitude was recorded using an oscilloscope (TBS1202B, Tektronix, USA). A current pulse with a width of 500 μs was employed to reach the low-resistive state.

Acknowledgments

Part of this work was supported by JSPS KAKENHI (Grant Nos. 18H02053, 21H05009, and 21K18805, to Y. Sutou). The HAXPES measurement was performed at beamline BL47XU at SPring-8, Japan as parts of the proposal of 2020A1240. The authors thank Prof. Y. -H. Song and Prof. J. Hong (Hanyang University, Korea) for the experiment support.

Data Availability Statement

All data are available in the main text and/or the Supporting Information.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsnano.3c07960.

  • HR-TEM observation for SmTe films; Hall properties as a function of annealing temperature; optical properties for low-resistive and high-resistive states/Tauc plots; valence-band spectra; anisotropic Young’s modulus values/modified Williamson–Hall plot; device structure; definition of resistive change in the isothermal measurement; Thermal stability for the high-resistive state; Madelung potential calculation; Band gap values for various material systems (PDF)

Author Contributions

Conceptualization: S.H.; Y. Sutou. Methodology: S.H.; S.M.; P.F. Investigation: S.H., S.M., Y. Saito, P.F., Y. Shuang. Visualization: S.H. Supervision: Y. Sutou. Writing—original draft: S.H.; Y. Saito; Y. Sutou. Writing—review and editing: S.H., S.M., Y. Saito, P.F., Y. Shuang, Y. Sutou.

The authors declare no competing financial interest.

Notes

Preprint: Hatayama S.; Mori S.; Saito Y.; Fons P.; Shuang Y.; Sutou Y.. An isomorphic valency transition in SmTe film enabling nonvolatile resistive change. Research Square Preprint, 2022. 10.21203/rs.3.rs-2318820/v110.21203/rs.3.rs-2318820/v1 (posted November 29, 2022).

Supplementary Material

nn3c07960_si_001.pdf (3.5MB, pdf)

References

  1. Bludau W.; Onton A.; Heinke W. Temperature dependence of the band gap of silicon. J. Appl. Phys. 1974, 45, 1846–1848. 10.1063/1.1663501. [DOI] [Google Scholar]
  2. Chopra K. L.; Bahl S. K. Structural, electrical, and optical properties of amorphous germanium films. Phys. Rev. B 1970, 1, 2545. 10.1103/PhysRevB.1.2545. [DOI] [Google Scholar]
  3. Kim J.-J.; Kobayashi K.; Ikenaga E.; Kobata M.; Ueda S.; Matsunaga T.; Kifune K.; Kojima R.; Yamada N. Electronic structure of amorphous and crystalline (GeTe)1–x(Sb2Te3)x investigated using hard x-ray photoemission spectroscopy. Phys. Rev. B 2007, 76, 115124 10.1103/PhysRevB.76.115124. [DOI] [Google Scholar]
  4. Wuttig M.; Yamada N. Phase-change materials for rewriteable data storage. Nat. Mater. 2007, 6, 824–832. 10.1038/nmat2009. [DOI] [PubMed] [Google Scholar]
  5. Zhu C.; Sun X.; Liu H.; Zheng B.; Wang X.; Liu Y.; Zubair M.; Wang X.; Zhu X.; Li D.; Pan A. Nonvolatile MoTe2 p–n diodes for optoelectronic logics. ACS Nano 2019, 13, 7216–7222. 10.1021/acsnano.9b02817. [DOI] [PubMed] [Google Scholar]
  6. Wang Y.; Kang K.-M.; Kim M.; Lee H.-S.; Waser R.; Wouters D.; Dittmann R.; Yang J. J.; Park H.-H. Mott-transition-based RRAM. Mater. Today 2019, 28, 63–80. 10.1016/j.mattod.2019.06.006. [DOI] [Google Scholar]
  7. Lawrence J. M.; Riseborough P. S.; Parks R. D. Valence fluctuation phenomena. Rep. Prog. Phys. 1981, 44, 1. 10.1088/0034-4885/44/1/001. [DOI] [Google Scholar]
  8. Adhikari D. R.; Adhikari S. K.; Sungh V. K. Electrical properties of rare earth chalcogenides under the effect of pressure. Kathmandu Univ. J. Sci. Eng. Technol. 2007, 1 (III), 1. [Google Scholar]
  9. Li Y.; Liu J.; Zhang P.; Jing Q.; Liu X.; Zhang J.; Xiao N.; Yu L.; Niu P. Electrical transport properties of eute under high pressure. J. Mater. Chem. C 2021, 9, 17371. 10.1039/D1TC02941J. [DOI] [Google Scholar]
  10. Shchennikov V. V.; Gizhevskiy B. A.; Chebotayev N. M.; Klinkova L. A. Electrical properties of europium chalcogenides under ultrahigh pressure. Phys. Met. Metall. 1990, 69, 78–84. [Google Scholar]
  11. Ohashi M.; Kondo D.; Sato N.; Suzuki T.; Komatsubara T.; Takahashi H.; Môri N. Physical properties of SmS metallic phase at low temperature. Rev. High Pressure Sci. Technol. 1998, 7, 611–613. 10.4131/jshpreview.7.611. [DOI] [Google Scholar]
  12. Ohashi M.; Takeshita N.; Mitamura H.; Matsumura T.; Uwatoko Y.; Suzuki T.; Goto T.; Ishimoto H.; Môri N. Physical properties of TmS under pressure. Phys. B: Condens. Matter 2000, 281, 264–266. 10.1016/S0921-4526(99)01210-7. [DOI] [Google Scholar]
  13. Tsiok O. B.; Khvostantsev L. G.; Smirnov I. A.; Golubkov A. V. Electron and lattice stages in the valence transition in SmTe under a high hydrostatic pressure. J. Exp. Theor. Phys. 2005, 100, 752–759. 10.1134/1.1926436. [DOI] [Google Scholar]
  14. Matsumura T.; Kosaka T.; Tang J.; Matsumoto T.; Takahashi H.; Môri N.; Suzuki T. Pressure induced semiconductor to metal transition in TmTe. Phys. Rev. Lett. 1997, 78, 1138. 10.1103/PhysRevLett.78.1138. [DOI] [Google Scholar]
  15. Ribault M.; Flouquet J.; Haen P.; Lapierre F.; Mignot J. M.; Holtzberg F. Pressure study of the metal-insulator transition in TmSe. Phys. Rev. Lett. 1980, 45, 1295. 10.1103/PhysRevLett.45.1295. [DOI] [Google Scholar]
  16. Welber B.; Jayaraman A. Pressure-induced changes in the reflectivity of SmSe and some comments on Nd-substituted SmSe. J. Appl. Phys. 1979, 50, 462–466. 10.1063/1.325634. [DOI] [Google Scholar]
  17. Jayaraman A.; Narayanamurti V.; Bucher E.; Maines R. G. Pressure-induced metal-semiconductor transition and 4 f electron delocalization in Sm Te. Phys. Rev. Lett. 1970, 25, 368. 10.1103/PhysRevLett.25.368. [DOI] [Google Scholar]
  18. Alam A.; Johnson D. D. Mixed valency and site-preference chemistry for cerium and its compounds: A predictive density-functional theory study. Phys. Rev. B 2014, 89, 235126 10.1103/PhysRevB.89.235126. [DOI] [Google Scholar]
  19. Greczynski G.; Lu J.; Jensen J.; Petrov I.; Greene J. E.; Bolz S.; Kölker W.; Schiffers Ch.; Lemmer O.; Hultman L. Strain-free, single-phase metastable Ti0. 38Al0. 62N alloys with high hardness: metal-ion energy vs. momentum effects during film growth by hybrid high-power pulsed/dc magnetron cosputtering. Thin Solid Films 2014, 556, 87–98. 10.1016/j.tsf.2014.01.017. [DOI] [Google Scholar]
  20. Schmiester G.; Wortmann G.; Kaindl G.; Bach H.; Holtzberg F. Pressure-induced valence changes in EuS and SmTe. High Pressure Res. 1990, 3, 192–194. 10.1080/08957959008246073. [DOI] [Google Scholar]
  21. Chatterjee A.; Singh A. K.; Jayaraman A. Pressure-induced electronic collapse and structural changes in rare-earth monochalcogenides. Phys. Rev. B 1972, 6, 2285. 10.1103/PhysRevB.6.2285. [DOI] [Google Scholar]
  22. Batlogg B.; Kaldis E.; Schlegel A.; Wachter P. Electronic structure of Sm monochalcogenides. Phys. Rev. B 1976, 14, 5503. 10.1103/PhysRevB.14.5503. [DOI] [Google Scholar]
  23. Jayasankar K.; Pandey A.; Mishra B. K.; Das S. Evaluation of microstructural parameters of nanocrystalline Y2O3 by X-ray diffraction peak broadening analysis. Mater. Chem. Phys. 2016, 171, 195–200. 10.1016/j.matchemphys.2016.01.005. [DOI] [Google Scholar]
  24. Yamamoto N.; Makino H.; Yamamoto T. Young’s modulus and coefficient of linear thermal expansion of ZnO conductive and transparent ultra-thin films. Adv. Mater. Sci. Eng. 2011, 2011, 136127 10.1155/2011/136127. [DOI] [Google Scholar]
  25. Valmianski I.; Ramirez J. G.; Urban C.; Batlle X.; Schuller I. K. Deviation from bulk in the pressure-temperature phase diagram of V2O3 thin films. Phys. Rev. B 2017, 95, 155132 10.1103/PhysRevB.95.155132. [DOI] [Google Scholar]
  26. Kuroda M. A.; Jiang Z.; Povolotskyi M.; Klimeck G.; Newns D. M.; Martyna G. J. Anisotropic strain in SmSe and SmTe: Implications for electronic transport. Phys. Rev. B 2014, 90, 245124 10.1103/PhysRevB.90.245124. [DOI] [Google Scholar]
  27. Mori S.; Hatayama S.; Shuang Y.; Ando D.; Sutou Y. Reversible displacive transformation in MnTe polymorphic semiconductor. Nat. Commun. 2020, 11, 85. 10.1038/s41467-019-13747-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Hatayama S.; Yamamoto T.; Mori S.; Song Y.-H.; Sutou Y. Understanding the origin of low-energy operation characteristics for Cr2Ge2Te6 phase-change material: enhancement of thermal efficiency in the high-scaled memory device. ACS Appl. Mater. Interfaces 2022, 14, 44604–44613. 10.1021/acsami.2c13189. [DOI] [PubMed] [Google Scholar]
  29. Dubey R.; Singh S. High pressure phase transition in samarium monochalcogenides SmX (X= S, Se, Te) and their alloy SmS1– xSex. J. Mol. Struct. 2013, 1052, 204–213. 10.1016/j.molstruc.2013.08.062. [DOI] [Google Scholar]
  30. Rooymans V. C. J. M. Berichte der Bunsengesellschaft/Physical Chemistry Chemical Physics 1966, 70, 1036. 10.1002/bbpc.19660700919. [DOI] [Google Scholar]
  31. Campagna M.; Bucher E.; Wertheim G. K.; Longinotti L. D. Valence mixing and semiconductor-metal transition in the Sm monochalcogenides. Phys. Rev. Lett. 1974, 33, 165. 10.1103/PhysRevLett.33.165. [DOI] [Google Scholar]
  32. Petit L.; Svane A.; Lüders M.; Szotek Z.; Vaitheeswaran G.; Kanchana V.; Temmerman W. M. Phase transitions in rare earth tellurides under pressure. J. Phys.: Condens. Matter 2014, 26, 274213 10.1088/0953-8984/26/27/274213. [DOI] [PubMed] [Google Scholar]
  33. Portier J.; Campet G.; Etourneau J.; Tanguy B. A simple model for the estimation of electronegativities of cations in different electronic states and coordinations. J. Alloys Compd. 1994, 209, 285–289. 10.1016/0925-8388(94)91115-0. [DOI] [Google Scholar]
  34. Ilton E. S.; Post J. E.; Heaney P. J.; Ling F. T.; Kerisit S. N. XPS determination of Mn oxidation states in Mn (hydr)oxides. Appl. Surf. Sci. 2016, 366, 475–485. 10.1016/j.apsusc.2015.12.159. [DOI] [Google Scholar]
  35. Neupane M.; Richard P.; Xu Y.-M.; Nakayama K.; Sato T.; Takahashi T.; Federov A. V.; Xu G.; Dai X.; Fang Z.; Wang Z.; Chen G.-F.; Wang N.-L.; Wen H.-H.; Ding H. Electron-hole asymmetry in the superconductivity of doped BaFe2As2 seen via the rigid chemical-potential shift in photoemission. Phys. Rev. B 2011, 83, 094522 10.1103/PhysRevB.83.094522. [DOI] [Google Scholar]
  36. Batlogg B.; Kaldis E.; Schlegel A.; von Schulthess G.; Wachter P. Optical and electrical properties of the mixed valence compound Sm3S4. Solid State Commun. 1976, 19, 673–676. 10.1016/0038-1098(76)91102-9. [DOI] [Google Scholar]
  37. Wortmann G.; Nowik I.; Perscheid B.; Kaindl G.; Felner I. Critical evaluation of Eu valences from L III-edge x-ray-absorption and Mössbauer spectroscopy of EuNi2Si2–xGex. Phys. Rev. B 1991, 43, 5261. 10.1103/PhysRevB.43.5261. [DOI] [PubMed] [Google Scholar]
  38. Kato T.; Tanaka K. Electronic properties of amorphous and crystalline Ge2Sb2Te5 films. Jpn. J. Appl. Phys. 2005, 44, 7340. 10.1143/JJAP.44.7340. [DOI] [Google Scholar]
  39. Adnane L.; Dirisaglik F.; Cywar A.; Cil K.; Zhu Y.; Lam C.; Anwar A. F. M.; Gokirmak A.; Silva H. High temperature electrical resistivity and Seebeck coefficient of Ge2Sb2Te5 thin films. J. Appl. Phys. 2017, 122, 125104 10.1063/1.4996218. [DOI] [Google Scholar]
  40. Bahl S. K.; Chopra K. L. Amorphous versus crystalline GeTe films. III. Electrical properties and band structure. J. Appl. Phys. 1970, 41, 2196–2212. 10.1063/1.1659189. [DOI] [Google Scholar]
  41. Shukla K. D.; Sahu S.; Manivannan A.; Deshpande U. P. Direct evidence for a systematic evolution of optical band gap and local disorder in Ag, In doped Sb2Te phase change material. Phys. Status Solidi RRL 2017, 11, 1700273 10.1002/pssr.201700273. [DOI] [Google Scholar]
  42. Liu C.; Cao X.; Wang J.; Yuan Y.; Su J.; Liu C.; Cheng L.; Zhang X.; Li J.; Zhang X. Investigation on the optical phase change properties of intrinsic GeSb and Ti-doped GeSb. Opt. Mater. Express 2018, 8, 936–947. 10.1364/OME.8.000936. [DOI] [Google Scholar]
  43. Raoux S.; Cabral C.; Krusin-Elbaum L.; Jordan-Sweet J. L.; Virwani K.; Hitzbleck M.; Salinga M.; Madan A.; Pinto T. L. Phase transitions in Ge–Sb phase change materials. J. Appl. Phys. 2009, 105, 064918 10.1063/1.3091271. [DOI] [Google Scholar]
  44. Park J.-W.; Eom S. H.; Lee H.; Da Silva J. L. F.; Kang Y.-S.; Lee T.-Y.; Khang Y. H. Optical properties of pseudobinary GeTe, Ge2Sb2Te5, GeSb2Te4, GeSb4Te7, and Sb2Te3 from ellipsometry and density functional theory. Phys. Rev. B 2009, 80, 115209 10.1103/PhysRevB.80.115209. [DOI] [Google Scholar]
  45. Zhu M.; Wu L.; Rao F.; Song Z.; Ren K.; Ji X.; Song S.; Yao D.; Feng S. Uniform Ti-doped Sb2Te3 materials for high-speed phase change memory applications. Appl. Phys. Lett. 2014, 104, 053119 10.1063/1.4863430. [DOI] [Google Scholar]
  46. Hatayama S.; Sutou Y.; Shindo S.; Saito Y.; Song Y.-H.; Ando D.; Koike J. Inverse resistance change Cr2Ge2Te6-based PCRAM enabling ultralow-energy amorphization. ACS Appl. Mater. Interfaces 2018, 10, 2725–2734. 10.1021/acsami.7b16755. [DOI] [PubMed] [Google Scholar]
  47. Hatayama S.; Sutou Y.; Ando D.; Koike J.; Kobayashi K. Electrical transport mechanism of the amorphous phase in Cr2Ge2Te6 phase change material. J. Phys. D: Appl. Phys. 2019, 52, 105103 10.1088/1361-6463/aafa94. [DOI] [Google Scholar]
  48. Hatayama S.; Yagi T.; Sutou Y. Mixed-conduction mechanism of Cr2Ge2Te6 film enabling positive temperature dependence of electrical conductivity and seebeck coefficient. Results Mater. 2020, 8, 100155 10.1016/j.rinma.2020.100155. [DOI] [Google Scholar]
  49. Saxena N.; Persch C.; Wuttig M.; Manivannan A. Exploring ultrafast threshold switching in In3SbTe2 phase change memory devices. Sci. Rep. 2019, 9, 19251. 10.1038/s41598-019-55874-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Pathak A.; Pandey S. K. Unraveling the optical bandgap and local structural change during phase transition in In3SbTe2 material through UV-Vis-NIR and XPS studies. J. Appl. Phys. 2022, 131, 205305 10.1063/5.0089399. [DOI] [Google Scholar]
  51. Venkata Subbaiah Y. P.; Saji K. J.; Tiwari A. Atomically thin MoS2: a versatile nongraphene 2D material. Adv. Funct. Mater. 2016, 26, 2046–2069. 10.1002/adfm.201504202. [DOI] [Google Scholar]
  52. Geng X.; Zhang Y.; Han Y.; Li J.; Yang L.; Benamara M.; Chen L.; Zhu H. Two-dimensional water-coupled metallic MoS2 with nanochannels for ultrafast supercapacitors. Nano Lett. 2017, 17, 1825–1832. 10.1021/acs.nanolett.6b05134. [DOI] [PubMed] [Google Scholar]
  53. Sankar R.; Narsinga Rao G.; Muthuselvam I. P.; Butler C.; Kumar N.; Senthil Murugan G.; Shekhar C.; Chang T.-R.; Wen C.-Y.; Chen C.-W.; Lee W.-L.; Lin M.-T.; Jeng H.-T.; Felser C.; Chou F. C. Polymorphic layered MoTe2 from semiconductor, topological insulator, to Weyl semimetal. Chem. Mater. 2017, 29, 699–707. 10.1021/acs.chemmater.6b04363. [DOI] [Google Scholar]
  54. Ruppert C.; Aslan B.; Heinz T. F. Optical properties and band gap of single-and few-layer MoTe2 crystals. Nano Lett. 2014, 14, 6231–6236. 10.1021/nl502557g. [DOI] [PubMed] [Google Scholar]
  55. Tanaka K.; Nasu T.; Miyamoto Y.; Ozaki N.; Tanaka S.; Nagata T.; Hakoe F.; Yoshikiyo M.; Nakagawa K.; Umeta Y.; Imoto K.; Tokoro H.; Namai A.; Ohkoshi S. Structural phase transition between γ-Ti3O5 and δ-Ti3O5 by breaking of a one-dimensionally conducting pathway. Cryst. Growth Des. 2015, 15, 653–657. 10.1021/cg5013439. [DOI] [Google Scholar]
  56. Shao Z.; Cao X.; Luo H.; Jin P. Recent progress in the phase-transition mechanism and modulation of vanadium dioxide materials. NPG Asia Mater. 2018, 10, 581–605. 10.1038/s41427-018-0061-2. [DOI] [Google Scholar]
  57. Rogers K. D.; Coath J. A.; Lovell M. C. Characterization of epitaxially grown films of vanadium oxides. J. Appl. Phys. 1991, 70, 1412–1415. 10.1063/1.349550. [DOI] [Google Scholar]
  58. Homm P.; Menghini M.; Seo J. W.; Peters S.; Locquet J.-P. Room temperature Mott metal–insulator transition in V2O3 compounds induced via strain-engineering. APL Mater. 2021, 9, 021116 10.1063/5.0035865. [DOI] [Google Scholar]
  59. Szymanski N. J.; Liu Z. T. Y.; Alderson T.; Podraza N. J.; Sarin P.; Khare S. V. Electronic and optical properties of vanadium oxides from first principles. Comput. Mater. Sci. 2018, 146, 310–318. 10.1016/j.commatsci.2018.01.048. [DOI] [Google Scholar]
  60. Andreev V. N.; Klimov V. A. Specific features of electrical conductivity of V3O5 single crystals. Phys. Solid State 2011, 53, 2424–2430. 10.1134/S106378341112002X. [DOI] [Google Scholar]
  61. Kachi S.; Kosuge K.; Okinaka H. Metal-insulator transition in VnO2n–1. J. Solid State Chem. 1973, 6, 258–270. 10.1016/0022-4596(73)90189-8. [DOI] [Google Scholar]
  62. Kawashima K.; Ueda Y.; Kosuge K.; Kachi S. Crystal growth and some electric properties of V6O13. J. Cryst. Growth 1974, 26, 321–322. 10.1016/0022-0248(74)90265-6. [DOI] [Google Scholar]
  63. Guiot V.; Janod E.; Corraze B.; Cario L. Control of the electronic properties and resistive switching in the new series of Mott insulators GaTa4Se8–yTey (0≤ y≤ 6.5). Chem. Mater. 2011, 23, 2611–2618. 10.1021/cm200266n. [DOI] [Google Scholar]
  64. Camjayi A.; Acha C.; Weht R.; Rodríguez M. G.; Corraze B.; Janod E.; Cario L.; Rozenberg M. J. First-order insulator-to-metal Mott transition in the paramagnetic 3D system GaTa4Se8. Phys. Rev. Lett. 2014, 113, 086404 10.1103/PhysRevLett.113.086404. [DOI] [PubMed] [Google Scholar]
  65. Newbury D. E.; Ritchie N. W. M. Performing elemental microanalysis with high accuracy and high precision by scanning electron microscopy/silicon drift detector energy-dispersive X-ray spectrometry (SEM/SDD-EDS). J. Mater. Sci. 2015, 50, 493–518. 10.1007/s10853-014-8685-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Shirley D. A. High-resolution X-ray photoemission spectrum of the valence bands of gold. Phys. Rev. B 1972, 5, 4709. 10.1103/PhysRevB.5.4709. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

nn3c07960_si_001.pdf (3.5MB, pdf)

Data Availability Statement

All data are available in the main text and/or the Supporting Information.

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

RESOURCES