Polarization Control via He-Ion Beam Induced ... - ACS Publications

Feb 26, 2016 - (20) Vighter, I.; Ievlev, A. V.; Vlassiouk, I.; Jesse, S.; Kalinin, S. V.;. Joy, D. C.; Rondinone, A. J.; Belianinov, A.; Ovchinnikova,...
0 downloads 0 Views 2MB Size
Subscriber access provided by LAURENTIAN UNIV

Article

Polarization control via He-ion beam induced nanofabrication in layered ferroelectric semiconductors Alex Belianinov, Vighter Iberi, Alexander Tselev, Michael A. Susner, Michael A. McGuire, David C. Joy, Stephen Jesse, Adam Rondinone, Sergei V. Kalinin, and Olga S. Ovchinnikova ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.5b12056 • Publication Date (Web): 26 Feb 2016 Downloaded from http://pubs.acs.org on March 2, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Applied Materials & Interfaces is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Polarization Control via He-ion Beam Induced Nanofabrication in Layered Ferroelectric Semiconductors

Alex Belianinov,1, 2* Vighter Iberi,2,4 Alexander Tselev, 1, 2, 3 Michael A. Susner, 5 Michael A. McGuire,5 David Joy,2,4 Stephen Jesse,1, 2 Adam J. Rondinone,2 Sergei V. Kalinin,1, 2 Olga S. Ovchinnikova1, 2*

1 The Institute for Functional Imaging of Materials and the Center for Nanophase Materials Sciences, Oak Ridge National Laboratory, Oak Ridge, TN 37831 2. Center for Nanophase Materials Sciences, Oak Ridge National Laboratory, Oak Ridge, TN 37831 3. Department of Physics and Astronomy, University of Tennessee, Knoxville, Knoxville TN 37996 4. Department of Materials Science and Engineering, University of Tennessee, Knoxville, Knoxville TN 37996 5. Materials Sciences and Technology Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831

*Corresponding Authors Alex Belianinov, [email protected] Olga S. Ovchinnikova, [email protected] This manuscript has been authored by a contractor of the U.S. Government under contract DEAC05-00OR22725. Accordingly, the U. S. Government retains a paid-up, nonexclusive, irrevocable, worldwide license to publish or reproduce the published form of this contribution, prepare derivative works, distribute copies to the public, and perform publicly and display publicly, or allow others to do so, for U.S. Government purposes.

Corresponding author: [email protected]

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Keywords: Helium Ion Microscopy, Atomic Force Microscopy, Layered Materials, Ferroelectricity, 2D Crystals Abstract Rapid advances in nanoscience rely on continuous improvements of material manipulation at near-atomic scales. Currently, the workhorse of nanofabrication is resist-based lithography and its various derivatives. However, the use of local electron, ion, and physical probe methods is expanding, driven largely by the need for fabrication without the multi-step preparation processes that can result in contamination from resists and solvents. Furthermore, probe based methods extend beyond nanofabrication to nanomanipulation and to imaging which are all vital for a rapid transition to the prototyping and testing of devices. In this work we study helium ion interactions with the surface of bulk copper indium thiophosphate CuMIIIP2X6 (M = Cr, In; X= S, Se), a novel layered 2D material, with a Helium Ion Microscope (HIM). Using this technique, we are able to control ferrielectric domains and grow conical nanostructures with enhanced conductivity whose material volumes scale with the beam dosage. Compared to the copper indium thiophosphate (CITP) from which they grow, the nanostructures are oxygen rich, sulfur poor, and with virtually unchanged copper concentration as confirmed by Energy Dispersive Xray spectroscopy (EDX). Scanning Electron Microscopy (SEM) imaging contrast as well as Scanning Microwave Microscopy (SMM) measurements suggest enhanced conductivity in the formed particles, whereas Atomic Force Microscopy (AFM) measurements indicate that the produced structures have lower dissipation and a are softer as compared to the CITP.

ACS Paragon Plus Environment

Page 2 of 24

Page 3 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Introduction Continuous development of nanoscience and nanotechnology depends on new ways to control and pattern matter at all scales. While classical, resist-based lithography methods remain the mainstay of nanofabrication, local probe methods are becoming progressively better-suited for the same applications, largely due to their lack of contamination-inducing multi-step preparation processes. As such, electron and ion beams for direct nanopatterning via e-beam lithography,1 electron beam induced deposition (EBID),2 focused ion beam (FIB) milling,3 and ion beam induced deposition (IBID)2 are entering the mainstream realm of nanofabrication. Furthermore, local physical probes, by the means of scanning probe microscopy (SPM), are also capable of electrical, mechanical, and thermal writing.4,

5

In contrast to electron or ion beam fabrication,

nanofabrication through the use of SPM has the distinction of utilizing a material’s microstructure as the basis for the nanofabrication as has been demonstrated in ferroelectric lithography6 and charge lithography.7 Particularly interesting nanofabrication applications involve layered materials with unique physical properties, atomically-defined structures, and scalability in the bulk. For these reasons graphene, prototypical in the class of 2D materials,8,

9

is at the forefront of intense

research.10 The fabrication of graphene-based devices has been heavily explored using electron probe techniques11,

12

as well as SPM13,

14

lithography methodologies. Issues particular to

graphene manipulation with e-beam lithography include processing with a resist that contaminates the sample and degrades its properties such as transport. On the other hand, SPM based lithography can be time consuming and unreliable, due to physical changes of the SPM probe during the lithographic process. Therefore, recent attention has been focused on the direct nano-patterning of graphene using a scanning helium ion beam.15,

ACS Paragon Plus Environment

16

Use of a helium ion beam

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

for nanofabrication has the benefits of clean, resist-free processing; the He ion has a smaller mean free path in the material when compared to standard e-beam lithography, where charge injection may result in either strain or relaxation of the sample; and finally He ion use over Ga reduces ion implantation, common in standard Focused Ion Beam (FIB) lithography applications.17,

18

The use of a helium ion, a noble gas, in a helium ion microscope (HIM) as a

source for the introduction of localized defects has already been explored for graphene through controlled ion dose exposure.19 It has been demonstrated that by introducing local defects by HIM, it is possible to tune the conductivity of graphene nanoribbons.16 We further explore the use of HIM for the patterning and the introduction of localized defects into copper indium thiophosphate (CITP) CuMIIIP2X6 (M = Cr, In; X= S, Se). CITP is a layered van der Waals crystal that exhibits ferrielectricity, ionic conductivity, and is a member of a rich library of layered materials with a wide array of properties. 20 21 22 CITP is best described as a sulfur framework where the Cu and In fill the octahedral voids and the P–P pairs form a triangular pattern within the interlinked sulfur cages.21

22

A first-order phase transition from the

paraelectric to the ferrielectric state, occur at Tc = 315 K and is driven by the ordering in the copper sublattice as well as the displacement of cations from their centrosymmetric positions in the indium sublattice (C2/c to Cc symmetry). Based on X-ray diffraction data the direction of the spontaneous polarization at the phase transition into the ferrielectric phase is perpendicular to the layered plane.23

24 25

Preparation of clean-surfaced and thin layers of this material is rather

straightforward and is described in the Methods section. CITP has been extensively studied using neutron scattering, Raman spectroscopy, and single frequency and band excitation piezoresponse force microscopy (PFM).23 26 27 28 22, 24 25 20

ACS Paragon Plus Environment

Page 4 of 24

Page 5 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

In this work, we examine the tuning of ferroelectric behavior and the fabrication of surface microstructures as a first step to the manufacturing of 2D heterostructured devices through the controlled helium-ion beam exposure. We then probe the properties of these fabricated structures using Band-Excitation Atomic Force Microscopy (BE-AFM), Scanning Microwave Microscopy (SMM), Energy Dispersive X-ray spectroscopy (EDX), and Scanning Electron Microscopy (SEM) .29

30 31

Surprisingly, we observe precisely controlled growth of

microstructures atop CITP surface as well as local control of the surface ferrielectricity as a function of beam energy control in HIM. Furthermore, our results indicate that the grown particles exhibit enhanced ionic conductivity on an otherwise insulating surface.

Results The surface morphology of a freestanding CITP crystal prepared by exfoliation32 and mounted onto carbon tape is illustrated in Figure 1(a). As can be seen in panel (a), through the use of the standard graphene preparation methods it is possible to create large, clean terraces without major defects that are well suited for microscopy. Previous single frequency PFM and Band Excitation30 PFM (BE-PFM) studies have shown domain networks consisting of two types of domains.20 These dendritic-like domains vary in size and shape, but are on the order of 0.1-3 µm in the long axis. The phase of the piezoresponse flips by π radians across well-defined domain walls; domains remain continuous over monolayer steps, as well as step bunches up to 30 nm in height. These results have been duplicated using BE-PFM in ultra-high vacuum, with the sample prepared in situ indicating the high quality and cleanliness of the cleaved CITP surface.

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

An enlarged optical image of a corner of the crystal (Figure 1(b)), highlighted with a green box in Figure 1a, shows visible changes on the surface of the CITP crystal post helium-ion beam exposure on 1 × 1 µm2 areas in the imaging mode of the HIM. These appear as faint, regularly arranged squares indicating a physical, as well as a chemical, change of the sample in the beam exposed zones. We have used an Atomic Force Microscope (AFM) to systematically explore the surface morphology of the HIM exposed areas. The topographical changes due to the helium-ion beam exposure of 5 × 1014 to 1 × 1015 ions/cm2 were mapped by contact mode AFM and are shown in Figure 1c. It is immediately apparent that an increase in ion beam dose resulted in continuous growth of the surface structures. The relationship between helium-ion beam dose, structure volume, structure area, and the average normalized PFM signal was investigated for beam doses of 1 × 1014 to 1 × 1018 ions/cm2 across 1 × 1 µm2 area with a 0.5 µs per pixel irradiation dwell time (Figure 1(d)). Both volume and area increase rapidly for ion beam dose in the 5 × 1014 ions/cm2 to 1 × 1015 ions/cm2 regime after which there is an onset of saturation between 1 × 1015 and 1 × 1016 ions/cm2. It is important to note that we did not observe any topographical changes on the CITP sample for beam doses lower than 5 × 1014 ions/cm2. This observation suggests that at lower ion beam dosage, before the growth onset, < 5 × 1014 ions/cm2, we are able to suppress the surface ferrielectric states only, as seen in the normalized Piezoresponse Force Microscopy (PFM) plot, Figure 1(d), whereas at larger doses, 1 × 1015 ions/cm2, we transition to particle formation and then growth following saturation at 1 × 1018 ions/cm2. Supplementary Information Figure 1 illustrates particle stability atop CITP up to 100 °C. It is important to note here that since PFM is not a quantitative technique, we normalized the response to the overall average Amplitude × cos(Phase) of each PFM image, with the pristine domain areas giving a value of 0.5 for the normalized PFM signal.

ACS Paragon Plus Environment

Page 6 of 24

Page 7 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

A possible mechanism for such a self-limiting process may be related to the mean free path of the helium ions in the CITP crystal. Using the SRIM Monte-Carlo simulation package we are able to calculate a roughly 239 nm penetration depth into a 100 µm thick CITP material of 20 keV He ions generated by a point source. It should be noted that penetration depth here refers to the average implantation depth of the helium ions in CITP. Therefore, we assume that the helium ions are penetrating only the top several hundred layers of the bulk CITP; with the secondary electrons, generated due to helium-ion CITP lattice collisions, driving the chemical process responsible for the particle formation and ferrielectricity suppression. To establish the underlying mechanisms of the changes in the ferrielectric material, as well as the mechanical behavior in the exposed areas of the CITP surface, we employed Band Excitation Piezoresponse Force Microscopy (BE-PFM).30 Figure 2 illustrates an aggregation of amplitude, phase and topography results for beam doses of 5 × 1014 to 1 × 1015 ions/cm2. Dashed green lines in the amplitude and phase panels of Figure 2 (top two rows respectively) outline the particle edge in the exposed regions shown by white dashed lines; the topography of the particle is shown in the bottom row. Amplitude images illustrate the disappearance of domain structures in the entire exposed regions, with some partial domains remaining at the lowest dosage, Figure 2(a). This is corroborated by the continuous contrast disappearance in the phase images Figure 2(e – h) in the 5 × 1014 to 7 × 1014 ions/cm2 range with nearly no coherent phase signal remaining at 1 × 1015 ions/cm2. Topography images in Figure 2 (i – l) exemplify particle growth in the PFM imaged areas with the smallest particle being 19 nm, and the largest 124 nm tall. Clearly the beam induces chemical, as well as physical, changes on the surface in the He structuring process. The two transformations are coupled: a higher ion beam dose results in a larger microstructure as well as nullifies the piezoresponse.

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Mechanical properties of the fabricated structures, aimed at understanding the mechanical changes of the helium-ion beam induced surface growth, were explored by atomic force acoustic microscopy33 via an on-resonance blue laser mechanical excitation. Results for the frequency shift and 1/Q, where Q is the quality factor of the cantilever, are shown in Figure 3 for the particle exposed to 8 × 1014 ions/cm2. Figure 3(a) shows the resonance frequency shift map painted onto the 3D topographical surface. Note a resonance frequency shift from 342 kHz in the surrounding areas to 339 kHz on the particle. Typically, a downward shift is associated with a softening of the underlying material, in this case surface amorphization to yield the growth of the particle.34 The 1/Q map painted onto the topographical 3D surface is shown in Figure 3(b). Here, the Q-factor corroborates the BE-PFM result as well the frequency shift results where the Q factor increased in the areas of the grown particle and decreased in the ferroelectric domain areas. This increase indicates that the cantilever energy dissipation is lower when scanned over the helium-ion beam induced structure.35 Figure 3(c) is a capacitance map measured by Scanning Microwave Microscopy (SMM) painted over the 3D topographical data of a nanostructured resulting from 8 × 1014 ions/cm2.31 The drop in the capacitance signal on the particle, as compared to the uniform background of the surface, suggests a small enhancement in the conductivity. However, the conductivity channel of the SMM system did not show significant variations of the signal with respect to the background, which indicates that effect of the ion beam exposure on the material conductivity is localized at the sample surface. As mentioned above, based on SRIM calculations the beam penetration length is about 293 nm, while the sample thickness is ~100 µm. Without a conduction path to the substrate, the sample conductivity contribution to the real part of the tip-sample system admittance (i. e. conductivity) measured by the SMM system is too small to detect. The

ACS Paragon Plus Environment

Page 8 of 24

Page 9 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

insulating nature of the bulk sample also precluded us from corroborating these results with conductive AFM (c-AFM). Detailed examination of the nanoparticle BE-PFM data coupled with SEM images of the surface illustrates interplay between the particle growth and preexisting ferrielectric domains shown in Figure 4. In Fig. 4(a, b) the phase and amplitude of the BE-PFM, painted over the topography images, show impressions in the particle over regions that contained domains prior to helium-ion exposure of 1 × 1016 ions/cm2. The faint domain outline can still be seen in the phase image, albeit heavily distorted, and is similar to the weak phase signal observed in Figure 3 of domains which disappear with higher dosage. Figure 4(c) is a SEM image of the same particle as in panels (a) and (b), with the black arrow pointing to the same remaining domain at the particle edge. Interestingly, domains connected to the particle have much higher contrast as compared to the remaining domains disconnected, or far from the particle. SEM and SMM capacitance results are then in accord, with regards to the enhanced conductivity at the nanostructured site as compared to the bulk. However, the origin of the difference between the surface morphology in the grown nanoparticle (as illustrated by the white arrows in Figure 4(a, b)), the areas that used to contain ferrielectric domains, and the dead areas, is still unclear. Energy Dispersive X-ray (EDX) results are shown in the Figure 5, for the nanostructure after 1 × 1018 ions/cm2 helium-ion beam exposure together with secondary electron images of the structure Fig. 5(a). Supplementary Information Figure 2 shows full EDX spectra. Figure 5(b) is the copper signal, panels (c) and (d) are sulfur and oxygen signals respectively. The EDX results indicate that the domains are copper rich compared to the PFM dead regions. Additionally, we note strong enhancement of the oxygen signal from the EDX result in Figure 5(c), as well as the reduction of the sulfur signal in the exposed area, Figure 5(b). The rupture of the sulfur-copper

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 24

bond can be expected to happen first in irradiated CITP material given that it has the smallest bond dissociation energy of 2.8 eV. A subsequent copper oxide and sulfur oxide formation occurs during exposure of the irradiated area to atmosphere, or due to some residual oxygen in the HIM chamber, leading to an increase signal for oxygen in the beam exposed areas.

Conclusions In this work we have demonstrated that the helium-ion interaction with the surface of bulk copper indium thiophosphate results in the controlled loss of ferrielectric domains and growth of conical structures with material volumes scaling to the dosage of the beam. These resulting nanostructures are oxygen rich, sulfur poor, and have a virtually unchanged copper concentration as compared to the CITP crystal from which it was grown. SEM image contrast as well as SMM measurements suggest small enhancement in the conductivity of the formed particle, whereas AFM based measurements indicate that the resulting structures have lower dissipation most likely associated with amorphization of the surface material. We suspect that the enhanced conductivity in the nanostructured site, as well the domain interconnectedness seen in SEM contrast, stems from simultaneous copper and sulfur oxidation. Both copper and sulfur oxidation are likely triggered by the secondary electron generation as a byproduct of the helium-ion colliding with the CITP surface. The impinging energy of the 20 keV helium-ion, can rupture the sulfur-copper bond which has a bond dissociation energy of 2.8 eV, thereby exposing the copper layer to oxidation. The copper oxide and sulfur oxide energy of formation are 0.4 eV and 3.8 eV respectively. Furthermore, particle volume saturation effects shown in Figure 1(d) is most likely controlled by the exposed surface area for the interaction of

ACS Paragon Plus Environment

Page 11 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

the oxygen with the CITP surface which may occur either due to low oxygen concentration in the high vacuum environment of HIM or upon sample exposure to atmosphere during transfer to AFM. We also note the noticeable effect of ferrielectric polarization on particle growth, namely higher growth rates on “negatively” polarized domains ( navy in the amplitude channel) with a positive phase value, than on the “positively” polarized domains (red in amplitude) with a negative phase value. More details are shown in Supplementary Information Figure 3. We attribute this behavior to the intrinsic field effect in ferroelectric, in which the presence of the negative polarization charge induces downward band bending that enhances transport of the positively charged Cu ions to the surface. This mechanism is similar to that considered by Giocondi and Roher36 for photoinduced reactions on ferroelectric surfaces.37 These results lay the groundwork for exploring the layered CITP family of compounds as a building block for 2D heterostuctured devices coupled with other low dimensional materials. We suspect that further studies of these van der Waal layered crystal compounds will reveal mechanisms for finer control of surface as well as bulk properties, like conductivity, ferroelectricity and precision manufacturing for a smooth transition to functional layered devices.

Methods Sample Growth The CuInP2S6 crystals were grown by chemical vapor transport in evacuated quartz ampoules. The chemical precursor In2S3 was prepared from In (Alfa Aesar Puratronic 99.999%) and S (JMC Puratronic 99.999%) by reacting these elements in the appropriate stoichiometric mixture

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 24

at 950oC for 48 hrs. The resultant In2S3 was ground into powder, checked for purity via X-ray diffraction, and added together with the stoichiometric quantities of Cu, P, and S (Alfa Aesar, Puratronic 99.999%+). These powders were ground together in a mortar and pestle inside of a He-filled glovebox and subsequently loaded into quartz ampoules and sealed under vacuum. The ampoules were then loaded into a box furnace and brought to 750-775oC over 15 hrs, held at this reaction temperature for 96 hrs, and then cooled to 30oC at a rate of 20oC/hr unless otherwise noted. After cooling, the ampoules were sliced open and the crystals were extracted. The crystals were platelets ~2 x1 mm2 or smaller in area and