Computational Predictions and Microwave Plasma Synthesis of Superhard Boron-Carbon Materials

materials

Article Computational Predictions and Microwave Plasma Synthesis of Superhard Boron-Carbon Materials

Paul A. Baker, Shane A. Catledge, Sumner B. Harris, Kathryn J. Ham ID , Wei-Chih Chen, Cheng-Chien Chen and Yogesh K. Vohra * Department of Physics, University of Alabama at Birmingham (UAB), Birmingham, AL 35294, USA; [email protected] (P.A.B.); [email protected] (S.A.C.); [email protected] (S.B.H.); [email protected] (K.J.H.); [email protected] (W.-C.C.); [email protected] (C.-C.C.) * Correspondence: [email protected]; Tel.: +1-205-934-6662

Received: 22 June 2018; Accepted: 23 July 2018; Published: 25 July 2018

Abstract: Superhard boron-carbon materials are of prime interest due to their non-oxidizing properties at high temperatures compared to diamond-based materials and their non-reactivity with ferrous metals under extreme conditions. In this work, evolutionary algorithms combined with density functional theory have been utilized to predict stable structures and properties for the boron-carbon system, including the elusive superhard BC5 compound. We report on the microwave plasma chemical vapor deposition on a silicon substrate of a series of composite materials containing amorphous boron-doped graphitic carbon, boron-doped diamond, and a cubic hard-phase with a boron-content as high as 7.7 at%. The nanoindentation hardness of these composite materials can be tailored from 8 GPa to as high as 62 GPa depending on the growth conditions. These materials have been characterized by electron microscopy, X-ray photoelectron spectroscopy, Raman spectroscopy, X-ray diffraction, and nanoindentation hardness, and the experimental results are compared with theoretical predictions. Our studies show that a signiﬁcant amount of boron up to 7.7 at% can be accommodated in the cubic phase of diamond and its phonon modes and mechanical properties can be accurately modeled by theory. This cubic hard-phase can be incorporated into amorphous boron-carbon matrices to yield superhard materials with tunable hardness values.

Keywords: boron-carbon compound; superhard materials; ab initio calculations; chemical vapor deposition

1. Introduction The first row of elemental solids (C, N, O, and B—jointly referred to as CNOB) form dense covalent solids in three-dimensional (3D) network structures that are extremely hard, have a high-energy density content, and exhibit unique electronic and optical properties. While diamond (hardness of approx. 100 GPa) and cubic-boron nitride (hardness of approx. 45 GPa) have long been established as the cornerstone of a multi-billion dollar abrasives industry, there is considerable scientific and technological interest in novel superhard materials (hardness greater than 40 GPa) based on CNOB. Low pressure/low-temperature synthesis affords the metastable development of unique superhard binary, ternary, and quaternary phases from CNOB precursors that can be quenched to form conformal coatings on a large range of substrates. In this paper, we focus on a sub-set of superhard materials based on the boron-carbon system, where the synthesis of superhard BC5 material has previously been claimed using the high-pressure high-temperature technique at a pressure of 24 GPa and temperature of about 2200 K [1]. The synthesized BC5 material had a measured hardness value of 71 GPa and high thermal stability up to 1900 K [1]. Subsequently, a significant amount of theoretical work has suggested various

Materials 2018, 11, 1279; doi:10.3390/ma11081279 www.mdpi.com/journal/materials Materials 2018, 11, 1279 2 of 12 stable and metastable superhard modifications of boron-carbon systems [2–9]. Our previous study of low-level boron incorporation in diamond by microwave plasma chemical vapor deposition showed a significant change in the plasma gas-phase chemistry and morphology of the diamond films by the introduction of boron in a methane/hydrogen/nitrogen plasma [10]. The focus of this work is on the synthesis of metastable superhard boron-carbon composites from the gas phase using low-temperature microwave plasma chemical vapor deposition. The advantage of microwave plasma chemical vapor deposition (CVD) is its ability for large area synthesis and for overcoming limitations of high-pressure, high-temperature techniques. We have employed gas phase precursors based on hydrogen (H2), methane (CH4), and diborane (B2H6) in materials synthesis using a microwave plasma source.

2. Materials and Methods The silicon substrates were obtained from pieces of a <100> oriented, N/Ph doped silicon wafer (University Wafer #1095). These were ultrasonically cleaned with solvents, scratched with diamond powder (2–4 µm particle size) for 30 s on a polishing pad, and then ultrasonically cleaned with DI water and methanol to remove any diamond particles. The B-C ﬁlms were grown using a 6 kW microwave plasma chemical vapor deposition reactor on a silicon substrate using hydrogen/methane/diborane chemistry. The deposition conditions were: 500 standard cubic centimeters per minute (SCCM), H2 as the carrier gas, 22 SCCM CH4 as the carbon source, and 0.1–0.45 SCCM B2H6 as the boron source, and the substrate temperature was carefully controlled in the range of 750–950 ◦C. The ﬁrst 45 min of each deposition was performed with only the methane as the precursor to deposit a layer of microcrystalline diamond and the diborane was added to the plasma for the synthesis of high-boron content superhard boron-carbon composites. The deposition conditions are outlined in Table1 for the four samples described in this manuscript.

Table 1. Growth conditions for the four samples described in this study. The nomenclature of samples is as follows: LBDD = lightly boron-doped diamond, HBDD = heavily boron-doped diamond, SBCC = soft boron-carbon composite, and HBCC = hard boron-carbon composite.

Growth Microwave Chamber CH B H Growth Time h Sample 4 2 6 Temp. (◦C) Power (W) Pressure (Torr) (SCCM) (SCCM) (Diamond/B-C) LBDD 875 900 58 22 residual 5.5 h/N-A HBDD 845 850 51 22 0.15 2/2.8 SBCC 925 900 60 22 0.15 0.75/6.3 HBCC 775 850 53 22 0.45 0.77/6.8

The films were analyzed with a Phi Electronics Versaprobe 5000, equipped with a micro-focused Al monochromatic source (λ = 1486.6 eV) and a dual anode conventional X-ray source with a neutralizer. The Mg anode provides X-rays with an energy of 1253.6 eV and the survey spectra were taken with a pass energy of 187.85 eV and a step size of 0.5 eV. Spectra were calibrated such that the C1s peak position was at 284.5 eV. A Panalytical Empyrean X-ray diffractometer was used to obtain diffraction patterns of the films. The optics used were a hybrid monochromator (Kα1 = 1.54059 Å) with a 1/8◦ divergence slit and a parallel plate collimator on the diffracted beam path with a proportional detector. The films were imaged with a FEI QuantaTM 650 FEG scanning electron microscope. The micro-Raman spectrometer was a Dilor XY with an input laser of 532 nm. Hardness and Young’s modulus were measured using a NanoIndenter XP with a Berkovich diamond tip with a nominal radius of 50 nm. A common and valid concern is in regard to blunting of the indenter diamond tip when performing hardness measurements on superhard materials. Therefore, in our measurements, we performed calibration of the indenter area function before and after hardness measurements on B-C thin films. A fused silica calibration standard (accepted Young0s modulus of 72 GPa) was tested before and after testing of each CVD-grown sample. All samples, including silica, were indented to a maximum depth of 400 nm. The measured Young0s modulus and hardness Materials 2018, 11, 1279 3 of 12 values were determined at maximum load and averaged from 10–15 indents (for silica samples), with uncertainty represented as standard deviation of the data. The Young0s modulus of the silica before and after testing all CVD-grown samples was 73.3 ± 0.6 GPa and 74.8 ± 0.9 GPa, respectively. Therefore, the Young0s modulus of the silica calibration standard remained within 4 percent of its average starting value and the indenter tip area function was not modified throughout all tests. We have performed crystal structure predictions to study stable superhard BC5 structures. In principle, one would like to find the stable structure of a compound knowing only its chemical formula by locating the minimum of the Gibbs free energy G = U + PV − TS. It is much more time-consuming to compute the entropy and temperature effects, so quite often only the enthalpy H = U + PV is minimized in practice. The minima of the potential energy surface correspond to different stable and metastable structures, which could potentially be stabilized under different experimental conditions. The crystal structure prediction is performed using the USPEX (Universal Structure Predictor: Evolutionary Xtallography) software [11–13] based on an evolutionary algorithm developed by Oganov, Glass, Lyakhov, and Zhu. This stochastic method uses concepts such as survival of the fittest and mutation inspired by biological evolution to locate the global minimum of a potential energy surface. The implementation of the algorithm features local optimization, real-space representation, and flexible physically motivated variation operators for highly efficient and accurate structure generation and prediction. For BC5, we consider unit cells containing two formula units (12 atoms) and search for the lowest-enthalpy structures in the pressure range of 0–75 GPa with a 5 GPa interval. Among the superhard phases of BC5 we found, we then performed an additional structure relaxation calculation at zero pressure to locate the lowest-energy structures, which is a widely used procedure in structure prediction calculation. The ab initio electronic structure calculation is performed using the VASP (Vienna Ab initio Simulation Package) program [14,15]. VASP adopts a plane wave basis set and a pseudopotential method. In our calculations, a plane wave cutoff energy of 600 eV was used, and the projector augmented wave method [16,17] with the PBE/GGA exchange correlation functional [18] was employed. The Γ-centered 22 × 22 × 5 k-point sampling in the Brillouin zone by the Monkhorst-Pack method [19] was used to calculate the total energy summation. For self-consistent and structure relaxation calculations, an energy difference of less than 10−6 eV/unit-cell was set for the electronic loop convergence criterion. All structures were relaxed until the forces on each ion were smaller than 10−3 eV/Å. The phonon calculations were performed with the density functional perturbation theory method [20] with a 2 × 2 × 2 supercell. An energy difference of less than 10−8 eV/unit-cell and forces on each ion of less than 10−7 eV/Å were used for the convergence criteria. The resulting interatomic force constants provided inputs for the PHONOPY [21] code to calculate the phonon dispersions and density of states. The lattice parameter calculations for boron-doped cubic diamond were carried out with 16 × 16 × 16 k-point sampling. Single unit cells were used for 0 at%, and 12.5 at% boron concentrations, and a 2 × 2 × 2 supercell was used for 1.563 at%, 3.125 at%, 4.688 at%, 6.25 at%, and 9.375 at% data points. It is noted that the 0 at% case overestimates the lattice constant of pure diamond to 3.5716 Å, which is about a 0.14% error from the accepted value of 3.5667 Å. We also used VASP to compute the elastic modulus tensor, from which the bulk and shear moduli can be calculated, and then the Vickers hardness can be estimated using Chen’s model [22].

3. Results Figure1 shows the Raman spectrum of a lightly boron-doped diamond ﬁlm (sample LBDD in Table1). The Raman spectrum is dominated by a single zone-center mode at 1327.6 cm −1 and has a downward shift from a pure cubic diamond peak at 1332.5 cm−1. The additional weak features that are observed in Figure1 are attributed to the Fano effect in boron-doped diamond literature [23,24]. Materials 2018, 11, 1279 4 of 12 Materials 2018, 11, x FOR PEER REVIEW 4 of 12 Materials 2018, 11, x FOR PEER REVIEW 4 of 12

Figure 1. The measured Raman spectrum from a lightly boron‐doped diamond (sample LBDD) Figure 1. TheThe measured measured Raman Raman spectrum spectrum from from a a lightly boron boron-doped‐doped diamond (sample LBDD) recorded with the 532 nm laser excitation. A zone‐center Raman mode at 1327.6 cm−1 −is1 accompanied recorded with thethe 532 nmnm laserlaser excitation.excitation. AA zone-centerzone‐center RamanRaman mode mode at at 1327.6 1327.6 cm cm−1 isis accompanied accompanied by weak bands attributed to the Fano effect. by weak bands attributed to the Fano effect. The higher boron‐carbon films can be divided into two distinct categories: the ones that are The higherhigher boron-carbon boron‐carbon films films can can be be divided divided into into two two distinct distinct categories: categories: the ones the thatones are that grown are grown at temperatures higher than 900 °C and the ones that are grown below 850 °C. The high grownat temperatures at temperatures higher than higher 900 than◦C and 900 the °C ones and thatthe areones grown that beloware grown 850 ◦ belowC. The 850 high °C. temperature The high temperature films grown above 900 °C contain amorphous boron‐doped graphite (soft‐phase), while temperaturefilms grown films above grown 900 ◦ Cabove contain 900 amorphous°C contain amorphous boron-doped boron graphite‐doped (soft-phase), graphite (soft while‐phase), the while films the films grown below 850 °C mostly contain superhard boron‐carbon phases. Figure 2a shows the thegrown films below grown 850 below◦C mostly 850 °C contain mostly superhard contain superhard boron-carbon boron phases.‐carbon Figurephases.2 aFigure shows 2a the shows Raman the Raman spectrum from the film grown at 925 °C (sample SBCC in Table 1), showing Raman peaks Ramanspectrum spectrum from the from film grownthe film at grown 925 ◦C at (sample 925 °C SBCC (sample in Table SBCC1), in showing Table 1), Raman showing peaks Raman labeled peaks “D” labeled “D” and “G” attributed to amorphous boron‐doped graphitic carbon and a peak attributed labeledand “G” “D” attributed and “G” to attributed amorphous to boron-doped amorphous boron graphitic‐doped carbon graphitic and a peakcarbon attributed and a peak to amorphous attributed to amorphous carbon (AC), as well as broad bands attributed to heavily boron‐doped diamond. tocarbon amorphous (AC), as carbon well as (AC), broad as bands well attributed as broad to bands heavily attributed boron-doped to heavily diamond. boron Figure‐doped2b showsdiamond. the Figure 2b shows the Raman spectrum of a sample grown at 775 °C (sample HBCC), where boron‐ FigureRaman 2b spectrum shows the of a Raman sample spectrum grown at 775of a ◦sampleC (sample grown HBCC), at 775 where °C (sample boron-doped HBCC), graphitic where carbonboron‐ doped graphitic carbon and amorphous carbon peaks are completely absent. dopedand amorphous graphitic carbon peaksand amorphous are completely carbon absent. peaks are completely absent.

Figure 2. (a) Raman spectrum from sample SBCC showing “D” and “G” bands attributed to Figure 2. (a) Raman spectrum from sample SBCC showing “D” and “G” bands attributed to microcrystallineFigure 2. (a) boron Raman‐doped spectrum graphite, from and sample amorphous SBCC showingcarbon (AC); “D” and(b) Raman “G” bands spectrum attributed from to microcrystalline boron‐doped graphite, and amorphous carbon (AC); (b) Raman spectrum from samplemicrocrystalline HBCC showing boron-doped a hard cubic graphite,‐phase. and The amorphous vertical bars carbon show (AC); the ( blocation) Raman of spectrum phonon frommodes sample in sample HBCC showing a hard cubic‐phase. The vertical bars show the location of phonon modes in cubicHBCC diamond showing as determined a hard cubic-phase. by neutron The scattering vertical experiments bars show the [25]. location The observed of phonon Raman modes modes in cubic in cubic diamond as determined by neutron scattering experiments [25]. The observed Raman modes in HBCCdiamond sample as determinedare considerably by neutron shifted scattering from the experimentscubic diamond [25 ].positions. The observed Raman modes in HBCC HBCC sample are considerably shifted from the cubic diamond positions. sample are considerably shifted from the cubic diamond positions. In Figure 2, we have also indicated the location of phonon modes observed in cubic diamond by In Figure 2, we have also indicated the location of phonon modes observed in cubic diamond by neutron scattering [25]; however, our observed modes are shifted downward in frequency due to the neutron scattering [25]; however, our observed modes are shifted downward in frequency due to the addition of boron in the lattice. The measured Raman frequencies are shown in Table 2 and are addition of boron in the lattice. The measured Raman frequencies are shown in Table 2 and are compared with the theoretical calculations described later in this paper. compared with the theoretical calculations described later in this paper.

Materials 2018, 11, 1279 5 of 12

In Figure2, we have also indicated the location of phonon modes observed in cubic diamond by neutron scattering [25]; however, our observed modes are shifted downward in frequency due to Materials 2018, 11, x FOR PEER REVIEW 5 of 12 the addition of boron in the lattice. The measured Raman frequencies are shown in Table2 and are compared withTable the theoretical2. Measured calculations Raman Frequencies described of spectra later inshown this paper.in Figure 1 and Figure 2.

PeakTable 1 2. MeasuredPeak 2 RamanPeak 3 Frequencies Peak 4 of spectraPeak shown 5 inPeak Figures 6 1 andPeak2. 7 Peak 8 Sample (cm−1) (cm−1) (cm−1) (cm−1) (cm−1) (cm−1) (cm−1) (cm−1) Peak 1 Peak 2 Peak 3 Peak 4 Peak 5 Peak 6 Peak 7 Peak 8 SampleSBCC 463.9 ± 2.0 684.8 ± 17.4 1084.8 ±14.4 1209.6 ± 2.9 1287.1 ± 0.5 1346.1 ± 5.8 1530.0 ± 1.1 1587.3 ± 1.0 (cm−1) (cm−1) (cm−1) (cm−1) (cm−1) (cm−1) (cm−1) (cm−1) SBCCHBCC 463.9466.7± 2.0± 2.1 684.8705.8± 17.4 ± 29.0 1084.81078.5±14.4 ± 8.3 1209.61208.2± ±2.9 0.8 1287.11290.9± ± 0.50.5 1346.1—± 5.8 1530.0— ± 1.1 1587.3— ± 1.0 HBCC 466.7 ± 2.1 705.8 ± 29.0 1078.5 ± 8.3 1208.2 ± 0.8 1290.9 ± 0.5 — — — LBDDLBDD —— —— —— —— 1327.61327.6± ± 0.20.2 — — — — — Theory 508.6 702.2 1174.7 1209.3 1294.7 — — — Theory 508.6 702.2 1174.7 1209.3 1294.7 — — —

FigureFigure3 3shows shows X-ray X‐ray Photoelectron Photoelectron Spectroscopy Spectroscopy (XPS) (XPS) of the of boron-carbonthe boron‐carbon film film grown grown with lowwith andlow highand boronhigh boron content. content. XPS is primarilyXPS is primarily used to quantifyused to quantify the boron-content the boron in‐content the film in by the comparing film by thecomparing intensity the of intensity the B 1s peakof the to B the1s peak C 1s to peak. the C In 1s Figure peak. 3Ina, Figure the boron-carbon 3a, the boron film‐carbon contains film contains 2.9 at% boron2.9 at% and boron 97.1 and at% 97.1 carbon at% (sample carbon (sample HBDD in HBDD Table 1in), Table and in 1), Figure and in3b, Figure the sample 3b, the contains sample 7.7contains at% boron7.7 at% and boron 92.3 and at% 92.3 carbon at% (sample carbon (sample HBCC in HBCC Table 1in). Table The value 1). The of value 2.9 at% of boron-doping 2.9 at% boron‐ fordoping HBDD for isHBDD consistent is consistent with the with heavily the heavily boron-doped boron‐ samplesdoped samples in the literature, in the literature, while thewhile 7.7 the at% 7.7 boron at% boron is the highestis the highest level of level boron of boron incorporation incorporation in our in experiments. our experiments.

60 120

(a) Element Position B/C (b) C1s Element Position B/C 50 100 C1s C 1s 284.5 eV 97.1% C 1s 284.5 eV 92.3% 80 40 B 1s 187.8 eV 2.9% B 1s 188.2 eV 7.7% (thousands) (thousands)

30 60 second second sat.

* per per 40 π‐π 20

sat. * C1s sat. π‐π B1s Counts Counts 10 20 B 1s Mg source

0 0 400 200 0 400 200 0 Binding energy (eV) Binding Energy (eV)

FigureFigure 3.3. TheThe X-rayX‐ray PhotoelectronPhotoelectron SpectroscopySpectroscopy (XPS)(XPS) analysisanalysis ofof thethe twotwo boron-carbonboron‐carbon ﬁlmsfilms wherewhere RamanRaman spectraspectra havehave beenbeen presentedpresented inin FigureFigure2 2.. The The composition composition of of the the ﬁlm film is is determined determined from from the the intensitiesintensities of of BB 1s1s andand C1sC1s emissionemission intensities. intensities. ((aa)) HBDDHBDD samplesample andand ((bb)) HBCCHBCC sample.sample. TheThe satellitesatellite (sat)(sat) peaks peaks associated associated with with C1s C1s emission emission are are also also labeled. labeled.

FigureFigure4 4 shows shows the the nanoindentation nanoindentation hardness hardness measurements measurements on on samples samples grown grown at at high high and and low low temperatures.temperatures. Figure Figure 44aa showsshows aa load-displacementload‐displacement curvecurve forfor soft-sample soft‐sample SBCC SBCC grown grown at at high high temperaturestemperatures containing microcrystalline microcrystalline boron boron-doped‐doped graphite graphite indented indented to toa depth a depth of 400 of 400nm. nm.The Themeasured measured hardness hardness (H) and (H) Young and Young′s modulus0s modulus (E) of the (E) soft of‐sample the soft-sample are H = 7.8 are GPa H and = 7.8 E = GPa 174 GPa, and Erespectively. = 174 GPa, respectively.Figure 4b shows Figure a4 bload shows‐displacement a load-displacement curve for curve hard for‐sample hard-sample HBCC HBCCgrown grownat low attemperatures low temperatures indented indented to a depth to a depth of 400 of nm. 400 The nm. measured The measured hardness hardness and andYoung Young′s modulus0s modulus of the of thehard hard-sample‐sample are are H H= 62 = 62GPa GPa and and E E= =532 532 GPa, GPa, respectively. respectively. The The relative relative contribution contribution of elastic andand plasticplastic deformation deformation can can bebe calculatedcalculated fromfrom thethe ﬁnalfinal unloading depth of the load load-displacement‐displacement curves. TheThe softsoft samplesample (SBCC)(SBCC) showsshows anan elasticelastic contributioncontribution ofof 35%35% andand thethe hardhard samplesample (HBCC)(HBCC) showsshows anan elasticelastic contributioncontribution ofof asas highhigh asas 79%.79%.

Materials 2018, 11, 1279 6 of 12 Materials 2018, 11, x FOR PEER REVIEW 6 of 12

FigureFigure 4. The 4. loadThe loaddisplacement displacement curves curves for for the the two samplessamples to to a deptha depth of 400 of nm.400 (nm.a) SBCC (a) SBCC sample sample showingshowing considerable considerable plastic plastic deformation deformation and and yielding yielding a a hardness hardness value value of H of = H 7.8 = GPa 7.8 andGPa Elastic and Elastic b ModulusModulus of E = of 174 E = GPa 174 GPa based based on on two two such such indents indents andand ( ()b HBCC) HBCC sample sample showing showing minimal minimal plastic plastic deformation and yielding a hardness value of H = 61.7 GPa and Elastic Modulus of E = 532 GPa based deformation and yielding a hardness value of H = 61.7 GPa and Elastic Modulus of E = 532 GPa based on seven such indents. on seven such indents. Figure5 shows the high-resolution thin-film X-ray diffraction pattern of the B-C samples recorded Figure 5 shows the high‐resolution thin‐film X‐ray diffraction pattern of the B‐C samples with monochromatic Cu Kα1 radiation with a wavelength of λ = 1.54059 Å. The diffraction patterns are recordedindexed with to monochromatic a mixture of two cubicCu K phases,α1 radiation as indicated with a by wavelength the splitting of λ (111)= 1.54059 diffraction Å. peakThe withdiffraction patternsincreasing are indexed boron-content. to a mixture For the of HBCCtwo cubic samples, phases, cubic as phases indicated with latticeby the parameters splitting of a1 (111)= 3.5743 diffraction Å peak withand aincreasing2 = 3.5917 Å boron are recorded.‐content. The For lattice the parameter HBCC samples, a1 corresponds cubic to phases the underlying with lattice cubic-diamond parameters a1 phase and the lattice parameter a corresponds to the boron-doped hard phase. It should be added that = 3.5743 Å and a2 = 3.5917 Å are recorded.2 The lattice parameter a1 corresponds to the underlying no additional super-lattice X-ray diffraction peaks were observed in addition to the two cubic-phases, cubic‐diamond phase and the lattice parameter a2 corresponds to the boron‐doped hard phase. It as documented above. should be added that no additional super‐lattice X‐ray diffraction peaks were observed in addition Figure6 shows scanning electron micrographs (SEM) of various morphologies observed in our to the twofilms. cubic Figure‐phases,6a shows as the documented typical morphology above. of an HBDD sample (2.9 at% boron), where (100) cubic growth morphology is apparent in the micrograph. Figure6b shows a needle-like morphology amorphous boron-doped graphitic carbon sample (sample SBCC) where X-ray diffraction does not show crystalline graphite, confirming the amorphous nature of the deposit. Figure6c shows the surface morphology of the sample containing 7.7 at% boron (sample HBCC). This morphology shows a less faceted structure than the HBDD sample, possibly indicative of increased crystalline defects (e.g., twinning).

Materials 2018, 11, x FOR PEER REVIEW 6 of 11

Figure 4: The load displacement curves for the two samples to a depth of 400 nm. (a) SBCC sample showing considerable plastic deformation and yielding a hardness value of H = 7.8 GPa and Elastic Modulus of E = 174 GPa based on two such indents and (b) HBCC sample showing minimal plastic deformation and yielding a hardness value of H = 61.7 GPa and Elastic Modulus of E = 532 GPa based on seven such indents.

Figure 5 shows the high-resolution thin-film x-ray diffraction pattern of the B-C samples recorded with monochromatic Cu Kα1 radiation with a wavelength of λ = 1.54059 Å. The diffraction Materials 2018, 11, x FOR PEER REVIEW 7 of 12 patterns are indexed to a mixture of two cubic phases, as indicated by the splitting of (111) diffraction peak with increasing boron-content. For the HBCC samples, cubic phases with lattice parameters a1 = 3.5743 Å and a2 = 3.5917 Å are recorded. The lattice parameter a1 corresponds to the underlying cubic-diamond phase and the lattice parameter a2 corresponds to the boron-doped hard phase. It should beMaterials added2018 that, 11, 1279 no additional super-lattice x-ray diffraction peaks were observed in7 of addition 12 to the two cubic-phases, as documented above.

Figure 5. High resolution X‐ray diffraction pattern recorded on various B‐C films using a hybrid monochromator showing increased splitting of the cubic‐diamond (111) diffraction peak with increasing boron‐content. The boron‐doped diamond peak shift to lower 2θ angles, as indicated by arrows, showing an increase in lattice parameter with increasing boron content.

showFigure crystalline 5. High graphite, resolution confirming X-ray diffraction the amorphous pattern recorded nature on of various the deposit. B-C ﬁlms Figure using 6c a hybridshows the surfacemonochromator morphology showing of the sample increased containing splitting of 7.7 the at% cubic-diamond boron (sample (111) diffractionHBCC). This peak morphology with increasing shows a lessboron-content. faceted structure The boron-dopedthan the HBDD diamond sample, peak possibly shift to lowerindicative 2θ angles, of increased as indicated crystalline by arrows, defects (e.g., showingtwinning). an increase in lattice parameter with increasing boron content.

(a) (b) (c) )

10 µm 4 µm 10 µm

Figure 6. Scanning ElectronElectron Micrograph Micrograph (SEM) (SEM) of of various various boron-carbon boron‐carbon composites composites synthesized synthesized in this in thisstudy. study. (a) SEM (a) SEM of the of HBDD the HBDD sample sample (2.9 at% (2.9 boron) at% boron) showing showing (100) morphology;(100) morphology; (b) SEM (b) of SEM the of SBCC the SBCCsample sample containing containing microcrystalline microcrystalline boron-doped boron‐doped graphite; graphite; (c) SEM (c of) SEM the HBCCof the HBCC sample. sample.

Among the superhard structures we found, the orthorhombicorthorhombic BCBC55 (containing two formula units) with with Pmma Pmma symmetry, symmetry, as as shown shown in inFigure Figure 7, 7has, has the the lowest lowest energy energy [6]. Its [ 6 ].fully Its relaxed fully relaxed lattice parameterslattice parameters a, b, anda, cb ,are, and respectively,c are, respectively, 2.5103, 2.5103,2.5251, 2.5251,and 11.373 and Å. 11.373 The Å.volume The volume is thereby is thereby 6.0076 Å6.00763/atom. Å3 The/atom. bulk The and bulk shear and moduli shear moduliare computed are computed to be 378 to and be 378 382 and GPa, 382 respectively, GPa, respectively, which whichyields ayields Vickers a Vickers hardness hardness of H = 63 of GPa H = for 63 the GPa structure for the structurein Figure 7 in using Figure the7 usingChen′s the model Chen [22].0s modelIn general, [22]. similarIn general, hardness similar values hardness are obtained values are using obtained other using hardness other models hardness [26,27]. models [26,27]. The corresponding simulated X-ray diffraction pattern is shown in Figure8 and compared with that of the HBCC sample. We note that there exists another orthorhombic BC5, also with Pmma symmetry [6,9]; this structure with direct boron-boron bonding has a slightly higher energy and a predominant X-ray diffraction peak below 10◦, which is not observed in our experiments.

Materials 2018, 11, x FOR PEER REVIEW 7 of 11

Figure 5: High resolution x-ray diffraction pattern recorded on various B-C films using a hybrid monochromator showing increased splitting of the cubic-diamond (111) diffraction peak with increasing boron-content. The boron-doped diamond peak shift to lower 2θ angles, as indicated by arrows, showing an increase in lattice parameter with increasing boron content.

Figure 6 shows scanning electron micrographs (SEM) of various morphologies observed in our films. Figure 6a shows the typical morphology of an HBDD sample (2.9 at% boron), where (100) cubic growth morphology is apparent in the micrograph. Figure 6b shows a needle-like morphology amorphous boron-doped graphitic carbon sample (sample SBCC) where x-ray diffraction does not show crystalline graphite, confirming the amorphous nature of the deposit. Figure 6c shows the surface morphology of the sample containing 7.7 at% boron (sample HBCC). This morphology shows a less faceted structure than the HBDD sample, possibly indicative of increased crystalline defects (e.g., twinning).

(a) (b) (c) )

10 µm 4 µm 10 µm

Figure 6: Scanning Electron Micrograph (SEM) of various boron-carbon composites synthesized in this study. (a) SEM of the HBDD sample (2.9 at% boron) showing (100) morphology; (b) SEM of the SBCC sample containing microcrystalline boron-doped graphite; (c) SEM of the HBCC sample.

MaterialsAmong 2018, 11 the, x FORsuperhard PEER REVIEW structures we found, the orthorhombic BC5 (containing two formula units)8 of 12 with Pmma symmetry, as shown in Figure 7, has the lowest energy [6]. Its fully relaxed lattice parameters a, b, and c are, respectively, 2.5103, 2.5251, and 11.373 Å. The volume is thereby 6.0076 Å3/atom. The bulk and shear moduli are computed to be 378 and 382 GPa, respectively, which yields a VickersMaterials hardness2018, 11, of 1279 H = 63 GPa for the structure in Figure 7 using the Chen′s model [22]. 8In of general, 12 similar hardness values are obtained using other hardness models [26–27].

Figure 7. The lowest-energy structure of superhard BC5 (containing two formula units) predicted by the evolutionary algorithm as implemented in USPEX [13–15]. The unit cell is orthorhombic with Pmma symmetry. The lattice parameters are described in the text. The predicted hardness is 63 GPa.

The corresponding simulated X-ray diffraction pattern is shown in Figure 8 and compared with that of theFigure HBCC 7. The sample lowest-energy. We note structure that of there superhard exists BC5 another(containing orthorhomb two formula units)ic BC predicted5, also by with Pmma symmetry the[6,9]; evolutionary this structure algorithm with as implemented direct boron in USPEX-boron [13 bonding–15]. The has unit cella slightly is orthorhombic higher with energy and a Pmma symmetry. The lattice parameters are described in the text. The predicted hardness is 63 GPa. predominant X-ray diffraction peak below 10°, which is not observed in our experiments.

FigureFigure 8. Theoretical 8. Theoretical X-ray X-ray diffraction diffraction patterns patterns with 1.54061.5406 wavelength wavelength angstrom angstrom (Cu K-alpha1)(Cu K-alpha1) for for

orthorhombicorthorhombic BC5 with BC5 with Pmma Pmma symmetry symmetry compared compared toto thethe experimental experimental X-ray X-ray diffraction diffraction pattern pattern for for the HBCCthe HBCC sample sample.. Each Each spectrum spectrum is isnormalized normalized by itsits highest highest peak peak intensity. intensity.

We haveWe havealso performed also performed DFT DFT calculations calculations to to investigate the the vibrational vibrational frequencies frequencies of superhard of superhard BC5 phasesBC5 phases at zero at zeropressure pressure.. Figure Figure 9 9shows shows thethe phononphonon dispersion dispersion and and phonon phonon density density of states of states (DOS) of the orthorhombic BC5 structure in Figure7. In Figure9, there is no negative frequency mode (DOS) of the orthorhombic BC5 structure in Figure 7. In Figure 9, there is no negative frequency mode identiﬁed in the spectra, showing that the theoretical BC5 structure in Figure7 is dynamically stable. identified in the spectra, showing that the theoretical BC5 structure in Figure 7 is dynamically stable. The phonon DOS of cubic diamond is also shown for comparison. The existence of boron atoms The phononsoftens theDOS chemical of cubic bonds diamond and shifts is the also vibrational shown modes for comparison. to lower frequencies The existence compared of to those boron in atoms softensdiamond. the chemical We note bonds that the and pure shifts GGA the functional vibration underestimatesal modes to thelower vibrational frequenc frequenciesies compared by a few to those in diamondpercent;. We a more note accurate that the determination pure GGA functional of the phonon underestimates energies can be the obtained vibrational using, for frequencies example, by a few percent;the B3LYP a hybridmore accuratefunctional determination[28], which is computationally of the phon muchon energies more expensive. can be Figure obtained 10 shows using, for examplethe, theoreticalthe B3LYP Raman hybrid scattering functional spectra [28 for], which diamond is andcomputationally BC5 at zero pressure. much The more single expensive theoretical. Figure −1 Raman peak for diamond is located at 1294.7 cm . For BC5, two main theoretical Raman peaks are 10 shows the theoretical Raman scattering spectra for diamond and BC5 at zero pressure. The single theoretical Raman peak for diamond is located at 1294.7 cm−1. For BC5, two main theoretical Raman peaks are located at 1209.3 and 1174.7 cm−1. Additional low-energy theoretical Raman peaks are located at 702.2 and 508.6 cm−1. As shown in Table 2, overall, the theory and experiment are in good agreement.

Materials 2018, 11, 1279 9 of 12

Materials 2018located, 11, atx FOR 1209.3 PEER and REVIEW 1174.7 cm −1. Additional low-energy theoretical Raman peaks are located at 702.2 9 of 12 Materialsand 2018 508.6, 11, x cm FOR−1 .PEER As shown REVIEW in Table 2, overall, the theory and experiment are in good agreement. 9 of 12

5 FigureFigure 9Figure. 9Phonon. Phonon 9. Phonon dispersion dispersion dispersion and and anddensity density density ofof of statesstates states (DOS) (DOS) of ofof orthorhombic orthorhombicorthorhombic BC 5BC BCwith5 with with Pmma Pmma Pmma symmetry. symmetry. symmetry. TheThe phonon phononThe phonon DOS DOS of DOSof cubic cubic of cubic diamond diamond diamond isis isalsoalso also shown.shown. shown. EachEach phonon phononphonon DOS DOSDOS spectrum spectrum spectrum is normalized is is normalized normalized by its by by its its peakpeak intensity. intensity.peak intensity.

Figure 10. Comparison of the Raman spectrum for the cubic-diamond phase without boron Figure 10Figure. Comparison 10. Comparison of the of the Raman Raman spectrum spectrum for the the cubic-diamond cubic-diamond phase phase without without boron boron incorporation and the orthorhombic BC5 with Pmma symmetry. incorporationincorporation and the and orthorhombic the orthorhombic BC BC5 with5 with Pmma Pmma symmetry.symmetry.

At a semi-quantitative level, our results show that the main vibrational frequency in BC5 softens At a semiAt- aquantitative semi-quantitative level, level, our our results results show show thatthat the the main main vibrational vibrational frequency frequency in BC5 softensin BC5 softens compared to that in cubic diamond due to slightly elongated atomic bonds, and that broad low- comparedcompared to that to in that cubic in cubic diamond diamond due to to slightly slightly elongated elongated atomic bonds,atomic and bonds, that broad and low-energy that broad low- energypeaks peaks emerge emerge due due to theto the presence presence of boron of boron atoms atoms [9]. These [9]. These generic generic features features are also are observed also observed in energy peaks emerge due to the presence of boron atoms [9]. These generic features are also observed in otherother superhard superhard BC BC5 phases5 phases and are and in goodare agreementin good with agreement the experimental with the Raman experimental measurements, Raman in other superhard BC5 phases and are in good agreement with the experimental Raman measurementsas shown,in as Figure shown2. in Figure 2. measurementsFurtherFurther DFT, as shown DFTcalculations calculations in Figure were were 2. performed performed to to estimate the the lattice lattice parameter parameter of the of cubic the cubic diamond diamond Furtherstructure DFT as calculations a function of at% were boron performed doping. The to calculation estimate predictsthe lattice that parameter the lattice parameter of the cubic linearly diamond structure as a function of at% boron doping. The calculation predicts0 that the lattice parameter structureincreases as a function with increasing of at% boron boron content, doping. in accord The with calculation the Vegard predictss law [29 ]. that Figure the 11 latticeshows the parameter linearlycalculated increases points, with as increasing well as a linear boron fit with content equation, iny accord= 0.003859 withx + 3.5712. the Vegard′s law [29]. Figure 11 linearlyshows increasesthe Thecalculated DFT with calculations points increasing, as allow well boron supercellsas a linear content of fit boron-doped ,with in accord equation diamond with y = the0.003859 to relax Veg whileardx + ′3.5712s maintaining law. [29]. Figure a 11 shows thecubic calculated diamond points structure, as when well the as borona linear content fit with does equation not exceed y 12.5 = 0.003859 at%. Thex maximum+ 3.5712. lattice parameter measured by XRD of the thin films is 3.5917 Å. Using this value with the linear fit equation, the calculations predict a boron content of 5.3 at%. It should be added that our measured value of the lattice parameter of 3.5917 Å is similar to 3.589 Å, which is the value reported for the cubic phase of BC3 [30], and 3.59 Å, the value reported for the cubic phase of BC5 [1]. This indicates that our

Materials 2018, 11, x FOR PEER REVIEW 9 of 12

Figure 9. Phonon dispersion and density of states (DOS) of orthorhombic BC5 with Pmma symmetry. The phonon DOS of cubic diamond is also shown. Each phonon DOS spectrum is normalized by its peak intensity.

Figure 10. Comparison of the Raman spectrum for the cubic-diamond phase without boron

incorporation and the orthorhombic BC5 with Pmma symmetry.

At a semi-quantitative level, our results show that the main vibrational frequency in BC5 softens compared to that in cubic diamond due to slightly elongated atomic bonds, and that broad low- energy peaks emerge due to the presence of boron atoms [9]. These generic features are also observed in other superhard BC5 phases and are in good agreement with the experimental Raman measurements, as shown in Figure 2. Materials 2018, 11, 1279 10 of 12 Further DFT calculations were performed to estimate the lattice parameter of the cubic diamond structure as a function of at% boron doping. The calculation predicts that the lattice parameter linearlyHBCC increases ﬁlm consists withof increasing highly doped boron diamond, content containing, in accord 5.3 with at% boron the Veg in aard matrix′s law of amorphous [29]. Figure 11 showsboron-doped the calculated graphitic points carbon, as well which as acceptsa linear the fit remaining with equation 7.7 at% y boron,= 0.003859 as measuredx + 3.5712 by. XPS.

Figure 11. Plot of DFT calculation showing lattice parameters for cubic diamond structure as a function of the boron content. The linear behavior is predicted by Vegard0s Law. The inset structure shows a

2 × 2 × 2 supercell of boron-doped cubic diamond with a boron content of 9.375 at%.

4. Discussion Aided by evolutionary algorithm predictions, we have synthesized a novel series of boron-carbon materials using a microwave plasma chemical vapor deposition technique employing hydrogen/methane/diborane gas-phase chemistry. The hardness of the ﬁlm can be varied from 8 GPa to as high as 62 GPa depending on the growth conditions, thus opening the door for superhard materials synthesis suitable for high temperature operations. We have used DFT calculations to explain our experimental results, achieving overall good theory-experiment agreements on the measured hardness and vibrational spectra recorded by Raman spectroscopy. The theoretically predicted stable orthorhombic structure for stoichiometric BC5 (16.7 at% boron) was not observed in our metastable synthesis from the gas-phase. Instead, we document a metastable cubic diamond-like phase all the way to the highest boron concentration of 7.7 at%. Our data analysis also indicates that the lattice parameter of the cubic hard phase synthesized by the microwave plasma method is similar to the hard phases reported in BC3 and BC5 materials that have been synthesized by high-pressure high-temperature techniques. Low pressure/low-temperature plasma synthesis from the gas phase of the superhard boron-carbon composites offers advantages over high-pressure high-temperature methods in terms of large area deposition on a variety of substrates.

Author Contributions: P.A.B. and Y.K.V. conceived the plasma synthesis process for boron-carbon composites and carried out the growth experiments. W.-C.C. and C.-C.C. contributed to theoretical calculations, including evolutionary algorithm predictions and the density functional theory of boron-carbon system. S.B.H. carried out SEM, X-ray diffraction analysis, and DFT calculations to predict boron-doped diamond lattice parameters. S.A.C. performed the nanoindentation hardness measurements and the related analysis. K.J.H. carried out the Raman spectroscopy on all samples and the related analysis. Y.K.V. coordinated the development of the manuscript for publication. Funding: This material is based upon work supported by the National Science Foundation (NSF) EPSCoR RII-Track-1 Cooperative Agreement OIA-1655280. We also acknowledge support from the NSF Major Research Instrumentation (MRI) Grant No. DMR-1725016. Any opinions, ﬁndings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reﬂect the views of the National Science Foundation. Materials 2018, 11, 1279 11 of 12

Conﬂicts of Interest: The authors declare no conﬂict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, and in the decision to publish the results.

References

1. Solozhenko, V.L.; Kurakevych, O.O.; Andrault, D.; Godec, Y.L.; Mezouar, M. Ultimate Metastable Solubility of Boron in Diamond: Synthesis of Superhard Diamondlike BC5. Phys. Rev. Lett. 2009, 102, 015506. [CrossRef] [PubMed]

2. Calandra, M.; Mauri, F. High-Tc Superconductivity in Superhard Diamondlike BC5. Phys. Rev. Lett. 2008, 101, 016401. [CrossRef][PubMed]

3. Yao, Y.; Tse, J.S.; Klug, D.D. Crystal and electronic structure of superhard BC5: First-principles structural optimizations. Phys. Rev. B 2009, 80, 094106. [CrossRef]

4. Jiang, C.; Lin, Z.; Zhao, Y. Superhard diamondlike BC5: A ﬁrst-principles investigation. Phys. Rev. B 2009, 80, 184101. [CrossRef]

5. Lazar, P.; Podloucky, R. Mechanical properties of superhard BC5. Appl. Phys. Lett. 2009, 94, 251904. [CrossRef] 6. Li, Q.; Wang, H.; Tian, Y.; Xia, Y.; Cui, T.; He, J.; Ma, Y.; Zou, G. Superhard and superconducting structures of

BC5. J. Appl. Phys. 2010, 108, 023507. [CrossRef] 7. Zhao, W.J.; Wang, Y.X. Mechanical properties of superhard diamondlike BC5. Solid State Commun. 2011, 151, 478–481. [CrossRef] 8. Li, M.M.; Xiaofeng, F.; Zheng, W.T. First-principle calculations on the structural stability and electronic properties of superhard BxCy compounds. J. Phys. Condens. Matter 2013, 25, 425502. [CrossRef][PubMed]

9. Alp, I.O.; Ciftci, Y.O. Physical Properties of Superhard Diamond-Like BC5 from a First-Principles Study. J. Electron. Mater. 2018, 47, 272–284. [CrossRef] 10. Baker, A.P.; Goodloe, R.D.; Vohra, K.Y. Morphological Transition in Diamond Thin-Films Induced by Boron in a Microwave Plasma Deposition Process. Materials 2017, 10, 1305. [CrossRef][PubMed] 11. Oganov, A.R.; Glass, C.W. Crystal structure prediction using ab initio evolutionary techniques: Principles and applications. J. Chem. Phys. 2006, 124, 244704. [CrossRef][PubMed] 12. Glass, C.W.; Oganov, A.R.; Hansen, N. USPEX—Evolutionary crystal structure prediction. Comput. Phys. Commun. 2006, 175, 713–720. [CrossRef] 13. Lyakhov, A.O.; Oganov, A.R.; Stokes, H.T.; Zhu, Q. New developments in evolutionary structure prediction algorithm USPEX. Comput. Phys. Commun. 2013, 184, 1172–1182. [CrossRef] 14. Kresse, G.; Furthmüller, J. Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set. Comput. Mater. Sci. 1996, 6, 15–50. [CrossRef] 15. Kresse, G.; Furthmüller, J. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. Rev. B 1996, 54, 11169–11186. [CrossRef] 16. Blöchl, P.E. Projector augmented-wave method. Phys. Rev. B 1994, 50, 17953–17979. [CrossRef] 17. Kresse, G.; Joubert, D. From ultrasoft pseudopotentials to the projector augmented-wave method. Phys. Rev. B 1999, 59, 1758–1775. [CrossRef] 18. Perdew, J.P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996, 77, 3865–3868. [CrossRef][PubMed] 19. Monkhorst, H.J.; Pack, J.D. Special points for Brillouin-zone integrations. Phys. Rev. B 1976, 13, 5188–5192. [CrossRef] 20. Gonze, X.; Lee, C. Dynamical matrices, Born effective charges, dielectric permittivity tensors, and interatomic force constants from density-functional perturbation theory. Phys. Rev. B 1997, 55, 10355. [CrossRef] 21. Togo, A.; Tanaka, I. First principles phonon calculations in materials science. Scripta Mat. 2015, 108, 1–5. [CrossRef] 22. Chen, X.Q.; Niu, H.; Li, D.; Li, Y. Modeling hardness of polycrystalline materials and bulk metallic glasses. Intermetallics 2011, 19, 1275–1281. [CrossRef] 23. Ager, J.W.; Walukiewicz, W.; McCluskey, M.; Plano, M.A.; Landstrass, M.I. Fano interference of the Raman phonon in heavily boron-doped diamond films grown by chemical vapor deposition. Appl. Phys. Lett. 1995, 66, 616–618. [CrossRef] Materials 2018, 11, 1279 12 of 12

24. May, P.W.; Ludlow, W.J.; Hannaway, M.; Heard, P.J.; Smith, J.A.; Rosser, K.N. Raman and conductivity studies of boron-doped microcrystalline diamond, facetted nanocrystalline diamond and cauliﬂower diamond ﬁlms. Diam. Relat. Mater. 2008, 17, 105–117. [CrossRef] 25. Warren, J.L.; Yarnell, J.L.; Dolling, G.; Cowley, R.A. Lattice Dynamics of Diamond. Phys. Rev. 1967, 158, 805–808. [CrossRef] 26. Lyakhov, A.O.; Oganov, A.R. Evolutionary search for superhard materials: Methodology and applications to forms of carbon and TiO2. Phys. Rev. B 2011, 84, 092103. [CrossRef] 27. Šimu˚nek, A.; Dušek, M. Generalized bond-strength model of Vickers hardness: Application to Cr4B, CrB, CrB2, CrB4, Mo2B, MoB2, OsB2, ReB2, WB2, WB3 and Ti1.87B50. Mech. Mater. 2017, 112, 71–75. [CrossRef] 28. Baima, J.; Zelferino, A.; Olivero, P.; Erba, A.; Dovesi, R. Raman spectroscopic features of the neutral vacancy in diamond from ab initio quantum-mechanical calculations. Phys. Chem. Chem. Phys. 2016, 18, 1961–1968. [CrossRef][PubMed] 29. Vegard, L. Die Konstitution der Mischkristalle und die Raumfüllung der Atome. Z. Phys. 1921, 5, 17–26. [CrossRef] 30. Zinin, P.V.; Ming, L.C.; Ishii, H.A.; Jia, R.; Acosta, T.; Hellebrand, E. Phase transition in BCx system under

high-pressure and high-temperature: Synthesis of cubic dense BC3 nanostructured phase. J. Appl. Phys. 2012, 111, 114905. [CrossRef]

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).