The synthesis of high quality bulk and thin film (Mg0.25(1-x)CoxNi0.25(1-x)Cu0.25(1-x)Zn0.25(1-x))O and (Mg0.25(1-x)Co0.25(1-x)Ni0.25(1-x)CuxZn0.25(1-x))O entropy-stabilized oxides is presented.
Here, we present a procedure for the synthesis of bulk and thin film multicomponent (Mg0.25(1-x)CoxNi0.25(1-x)Cu0.25(1-x)Zn0.25(1-x))O (Co variant) and (Mg0.25(1-x)Co0.25(1-x)Ni0.25(1-x)CuxZn0.25(1-x))O (Cu variant) entropy-stabilized oxides. Phase pure and chemically homogeneous (Mg0.25(1-x)CoxNi0.25(1-x)Cu0.25(1-x)Zn0.25(1-x))O (x = 0.20, 0.27, 0.33) and (Mg0.25(1-x)Co0.25(1-x)Ni0.25(1-x)CuxZn0.25(1-x))O (x = 0.11, 0.27) ceramic pellets are synthesized and used in the deposition of ultra-high quality, phase pure, single crystalline thin films of the target stoichiometry. A detailed methodology for the deposition of smooth, chemically homogeneous, entropy-stabilized oxide thin films by pulsed laser deposition on (001)-oriented MgO substrates is described. The phase and crystallinity of bulk and thin film materials are confirmed using X-ray diffraction. Composition and chemical homogeneity are confirmed by X-ray photoelectron spectroscopy and energy dispersive X-ray spectroscopy. The surface topography of thin films is measured with scanning probe microscopy. The synthesis of high quality, single crystalline, entropy-stabilized oxide thin films enables the study of interface, size, strain, and disorder effects on the properties in this new class of highly disordered oxide materials.
Since the discovery of high-entropy metal alloys in 2004, high-entropy materials have attracted significant interest due to the properties such as increased hardness1,2,3, toughness4,5, and corrosion resistance3,6. Recently, high-entropy oxides7,8 and borides9 have been discovered, opening up a large playground for material enthusiasts. Oxides, in particular, can demonstrate useful and dynamic functional properties such as ferroelectricity10, magnetoelectricity11,12, thermoelectricity13, and superconductivity14. Entropy-stabilized oxides (ESOs) have recently been shown to possess interesting, compositionally-dependent functional properties15,16, despite the significant disorder, making this new class of materials particularly exciting.
Entropy-stabilized materials are chemically homogeneous, multicomponent (typically having five or more constituents), single-phase materials where the configurational entropic contribution () to the Gibbs free energy () is significant enough to drive the formation of a single phase solid solution17. The synthesis of multicomponent ESOs, where cationic configurational disorder is observed across the cation sites, requires precise control over the composition, temperature, deposition rate, quench rate, and quench temperature7,16. This method seeks to enable the practitioner the ability to synthesize phase pure and chemically homogeneous entropy-stabilized oxide ceramic pellets and phase pure, single crystalline, flat thin films of the desired stoichiometry. Bulk materials can be synthesized with greater than 90% theoretical density enabling the study of the electronic, magnetic, and structural properties or use as sources for thin film physical vapor deposition (PVD) techniques. As the entropy-stabilized oxides considered here have five cations, thin film PVD techniques that employ five sources, such as molecular beam epitaxy (MBE) or co-sputtering, will be presented with the challenge of depositing chemically homogenous thin films due to flux drift. This protocol results in chemically homogenous, single crystalline, flat (root-mean-square (RMS) roughness of ~0.15 nm) entropy-stabilized oxide thin films from a single material source, which are shown to possess the nominal chemical composition. This thin film synthesis protocol may be enhanced by the inclusion of in situ electron or optical characterization techniques for real-time monitoring of the synthesis and refined quality control. Expected limitations of this method stem from laser energy drift which may limit the thickness of high quality films to be below 1 μm.
Despite the significant advances in the growth and characterization of thin film oxide materials10,18,19,20,21, the correlation between stereochemistry and electronic structure in oxides can lead to significant differences in the final material stemming from seemingly insignificant methodological differences. Furthermore, the field of multicomponent entropy-stabilized oxides is rather nascent, with only two current reports of thin film synthesis in the literature7,16. ESOs lend themselves particularly well to this process, circumventing challenges that would be presented by chemical vapor deposition and molecular beam epitaxy. Here, we provide a detailed synthesis protocol of bulk and thin films ESOs (Figure 1), in order to minimize materials processing difficulties, unintended property variations, and improve the acceleration of discovery in the field.
Caution: Wear necessary personal protective equipment (PPE) including close-toed shoes, full length pants, safety glasses, particulate filtration mask, lab coat, and gloves as oxide powders pose a risk for skin contact irritation and eye contact irritation. Consult all relevant material safety data sheets before beginning for additional PPE requirements. Synthesis should be done with the use of engineering controls such as a fume hood.
1. Bulk Synthesis of Entropy-stabilized Oxides
2. PLD of ESO Single Crystal Films
X-ray diffraction (XRD) spectra were taken of both the prepared (Mg0.25(1-x)CoxNi0.25(1-x)Cu0.25(1-x)Zn0.25(1-x))O (x = 0.20, 0.27, 0.33) and (Mg0.25(1-x)Co0.25(1-x)Ni0.25(1-x)CuxZn0.25(1-x))O (x = 0.11, 0.27) bulk ceramics (Figure 4a) and deposited thin films (Figure 4b). These data show that the samples are single phase and can be used in the determination of lattice constant, crystalline quality, and film thickness. X-ray photoelectron spectroscopy (XPS) (Figure 5) and atomic force microscopy (AFM) (Figure 6) data were taken to determine the nominal composition of both the targets and films and to show surface quality of the deposited thin films.
XRD spectra from the bulk ESO samples shows that the synthesized compositions are single phase rock salts with lattice parameters of 4.25 Å, 4.25 Å, and 4.24 Å for XCo = 0.20, 0.27, and 0.33, respectively. These values are in relative agreement with Vegard's law and those reported in references 7 and 16. The lattice parameters were determined using Cohen's method22. The deposited films are single crystalline and epitaxial to the (001)-oriented MgO substrate as only the 002 and 004 film peaks are observed. The Laue fringes observed about the 002 and 004 peaks are a consequence of the high crystalline quality and smooth interfaces of the deposited films. The period of the oscillations is determined by the thickness of the film and reveals a ESO thickness of ~80 nm, consistent with our nominal thickness.
XPS data show that all constituent cations in both the ESO bulk samples and thin films are in a 2+ and high spin (where applicable) state. Compositions calculated from these spectra show that all samples are of nominal composition, to within <1% error. Compositions were obtained by fitting the XPS data to a Shirley background function implemented in CasaXPS. EDS chemical maps also agree with the nominal composition and show that the deposited films are chemically homogeneous on the length scale of 10-100 m.
AFM micrographs show that the samples are flat across a 5 m x 5 m scan range with RMS roughness values of 1.1 Å, 1.2 Å, and 1.4 Å for the XCo = 0.20, 0.27, and 0.33 films, respectively. Low angle 2– XRD data agrees with these roughness numbers16. The peak-to-peak roughness of the films is approximately 3.3 Å in all cases, which is less than the respective lattice constants of the films and may be attributed to the noise of the instrument. AFM images were processed using NT-MDT nova software.
Figure 1: Flow chart showing order of operations for entropy-stabilized oxide (ESO) thin film synthesis. First, the ESO ceramic pellets are synthesized in bulk. Then, the samples are ablated with a high-power laser and adsorbed onto a substrate to deposit single crystalline thin films. The crystallinity, topography, stoichiometry and homogeneity are proven using X-ray diffraction (XRD), atomic force microscopy (AFM), X-ray photoelectron spectroscopy (XPS) and energy-dispersive X-ray spectroscopy (EDS), respectively. Please click here to view a larger version of this figure.
Figure 2: (a) Parts and (b) schematic of the pressing die. The parts are A: die, B: plunger, C: short plunger, D: top and bottom plates, E: removal piston, and F: removal sheath. (c) Picture showing the die ready for pressing. Please click here to view a larger version of this figure.
Figure 3: Schematics of pulsed laser deposition (PLD) optical path and vacuum chamber. (a) Illustration of the optical path of the PLD system and (b) cutaway view of the vacuum chamber. The beam is focused onto the target, where it excites a plasma plume that then adsorbs onto the substrate. Please click here to view a larger version of this figure.
Figure 4: 2θ-ω X-ray diffraction (XRD) spectra from as-prepared ESO samples. (a) 2– XRD spectra from as-prepared ESO bulk samples. The labeled peaks correspond to the ideal rock salt structure, showing the presence of no secondary phases. (b) XRD spectra of ESO thin films grown on (001)-oriented MgO substrates. The spectra reveal the presence of the film 001 peaks, demonstrating phase purity and epitaxy. (Inset) High resolution 2– XRD scan around the 002 film and substrate peaks, clearly showing the Laue oscillations about the film peak, demonstrating films are flat and of excellent crystalline quality. * indicates the 002 reflection from the MgO substrate. Please click here to view a larger version of this figure.
Figure 5: Composition and homogeneity of thin film ESO samples. (a) XPS of the Co variant and Cu variant ESO targets and thin films, showing that all samples are of nominal composition. (b) Energy dispersive X-ray spectroscopy (EDS) compositional maps, showing that the films are chemically homogeneous. Scale bars = 30 µm. Please click here to view a larger version of this figure.
Figure 6: Contact AFM images of the deposited Co variant (top) and Cu variant (bottom) films, showing that all films have sub-unit cell root-mean-square (RMS) roughness. The periodic pattern from top left to bottom right is an artifact of the measurement. Please click here to view a larger version of this figure.
Table 1: Masses of constituents in Co variant ESO.
Table 2: Masses of constituents in Cu variant ESO.
Table 3: XPS determined compositions of Co variant ESO targets and thin films.
Table 4: XPS determined compositions of Cu variant ESO targets and thin films.
We have described and shown a protocol for the synthesis of bulk and high-quality, single crystalline films of (Mg0.25(1-x)CoxNi0.25(1-x)Cu0.25(1-x)Zn0.25(1-x))O (x = 0.20, 0.27, 0.33) and (Mg0.25(1-x)Co0.25(1-x)Ni0.25(1-x)CuxZn0.25(1-x))O (x = 0.11, 0.27) entropy-stabilized oxides. We expect these synthesis techniques to be applicable to a wide range of entropy-stabilized oxide compositions as more are discovered in the development and expansion of the field. Further, the synthesis of compositionally varied entropy-stabilized oxides offers a platform for studying the roles of structural and chemical disorder on functional properties.
While our protocol leads to single phase and high-quality entropy-stabilized oxides, limitations to the technique exist and modifications to the synthesis can be envisioned for advanced understanding of the material and higher reproducibility of the synthesis. Below, we outline critical steps within the protocol, possible modifications, troubleshooting and limitations of the technique, the significance with respect to existing methods, and envisioned future applications for this technique. Critical steps to this process are sintering, quenching, the dihydroxylation of the MgO surface, and the determination and monitoring of the laser fluence. For the bulk samples to be single phase, it is essential that they are sintered for at least 24 h and rapidly quenched from the sintering temperature. If the bulk targets are not single phase, or of the desired density, they can be reground and repressed to reach a higher density. The dihydroxylation of the MgO surface is also a critical step, as attempts to grow on (001)-oriented MgO without this result in amorphous films. Another key issue and limitation of the technique includes the drifting of the laser energy leading to deviation from the intended deposition conditions. This is discussed more in depth below.
The modifications to the technique increase reproducibility and enable real-time troubleshooting. In particular, in situ analysis, such as reflection high-energy electron diffraction23 (RHEED), low-angle X-ray spectroscopy24, X-ray reflectivity25, X-ray diffraction26,27, second harmonic generation28,29, or ellipsometry30, can be added to the thin film deposition procedure. This would enable structural and compositional characterization for real-time monitoring of the thin film growth conditions. As our protocol does not involve any in situ diagnostics, our report of potentially critical surface kinetics and structural evolution, that may occur during synthesis, is lacking. Further, our protocol calls for the constituent powders to be hand mixed and grounded with an agate mortar and pestle. Others in the community, however, have reported the use of shaker7 and ball milling15, using YSZ or agate media, which may give more consistent results by eliminating the physical demand of hand grinding.
The method outlined produces films of excellent quality, however, there are several inherent limitations to the technique. PLD makes it challenging to grow films significantly beyond 1 μm of thickness due to the drift of the laser energy. Drift of the laser energy can occur due to the passivation of F2 gas in the excimer laser tube with time and excitation of the gas. Further, laser energy drift can occur from material deposition on the UV transparent laser window on the chamber (Figure 3b). Our protocol reports the deposition on ~80 nm thick films using an oxygen pressure of 6.7 Pa; before and after this growth, we do not observe a change in the transmission of the UV laser window, which has ~10% intrinsic attenuation31. This may be a result of the relatively high oxygen pressure used during deposition, the relatively low number of shots, and the geometry of our deposition chamber. The depositions of films of appreciably larger thicknesses, with lower reactive gas pressure, or different laser window position with respect to the target may encounter difficulties stemming from laser energy drift. Energy drift due to the passivation of F2 gas in the laser tube can be minimized by routinely filling with fresh gas and monitoring the energy reported by the internal laser energy meter during deposition to compensate a decrease of laser energy with an increase in excitation voltage.
Entropy-stabilized oxides tend to have five or more cations where the entropy and bulk stabilization temperature depend drastically on composition. While currently no other deposition method has been reported to successfully grow entropy-stabilized oxide thin films, the stoichiometric evaporation and transfer from the target in PLD16,32 may provide the best chemical homogeneity. For instance, MBE and sputtering are alternative physical vapor deposition techniques that can be used to deposit high quality thin films33,34, however, MBE and multiple target co-sputtering would require accurate calibration and stable flux from five individual sources. This task is cumbersome and illustrates the anticipated difficultly in establishing the chemical homogeneity throughout the deposition of the film, assuming these techniques can be proven to deposit entropy-stabilized materials. Further, as the bulk materials require quenching from high temperature to maintain the entropy-stabilized phase, a significant density of thermodynamic point defects may prevent accurate determination of lattice parameter, resistivity, and dielectric properties. In theory, PLD should provide the ability to control the density of such defects35 and enable the accurate determination of functional and structural properties. Thus, the methodology presented here is significant for the investigation of the novel and giant functional properties of these materials.
The authors have nothing to disclose.
This work was funded in part by National Science Foundation grant No. DMR-0420785 (XPS). We thank the University of Michigan's Michigan Center for Materials Characterization, (MC)2, for its assistance with XPS, and the University of Michigan Van Vlack laboratory for XRD. We would also like to thank Thomas Kratofil for his assistance with bulk materials preparation.
MAGNESIUM OXIDE 99.95% | Fisher | AA1468422 | |
COBALT(II) OXIDE, 99.995% | Fisher | AA4435414 | |
NICKEL(II) OXIDE 99.998% | Fisher | AA1081914 | |
COPPER(II) OXIDE 99.995% | Fisher | AA1070014 | |
ZINC OXIDE 99.99% | Fisher | AA8781230 | |
TRICHLROETHLENE SEMICNDTR 9 | Fisher | AA39744K7 | |
ACETONE SEMICNDTR GRD 99.5% | Fisher | AA19392K7 | |
2-PROPANOL ACS 99.5% | Fisher | A416S4 | |
Mineral oil, pure | Acros Organics | AC415080010 | |
alumina crucible | MTI Corporation | eq-ca-l50w40h20 | |
ZIRCONIA (YSZ) GRINDING MEDIA | Inframat Advanced Materials | 4039GM-S010 | |
SiC paper 320/600/800/1200 | South Bay Technology | SDA08032-25 | |
MgO (100) substrate, 5x5x0.5 mm, 1SP | MTI Corporation | MGa050505S1 | |
OXYGEN COMPRESSED ULTRA HIGH PURITY GRADE, 99.999% | Cryogenic Gases | OXYUHP | |
NITROGEN COMPRESSED EXTRA DRY GRADE | Cryogenic Gases | NITEX |