Formation and Characterization of Hydrogen ... - ACS Publications

Sep 19, 2017 - ABSTRACT: Two-dimensional (2D) materials are promising for applications in a wide range of fields because of their unique prop- erties...
0 downloads 0 Views 5MB Size
Article pubs.acs.org/JACS

Formation and Characterization of Hydrogen Boride Sheets Derived from MgB2 by Cation Exchange Hiroaki Nishino,†,● Takeshi Fujita,#,● Nguyen Thanh Cuong,∇,● Satoshi Tominaka,○,● Masahiro Miyauchi,◆ Soshi Iimura,¶ Akihiko Hirata,#,■ Naoto Umezawa,○ Susumu Okada,‡ Eiji Nishibori,‡,▲ Asahi Fujino,† Tomohiro Fujimori,§ Shin-ichi Ito,∥ Junji Nakamura,⊥,▲ Hideo Hosono,¶,◇ and Takahiro Kondo*,⊥,▲,◇ †

Institute of Materials Science, Graduate School of Pure and Applied Sciences, ‡Division of Physics, Faculty of Pure and Applied Sciences, §College of Engineering Sciences, ∥Technical Service Office for Pure and Applied Sciences, Faculty of Pure and Applied Sciences, and ⊥Division of Materials Science, Faculty of Pure and Applied Sciences, University of Tsukuba, Tsukuba 305-8573, Japan # WPI-Advanced Institute for Materials Research, Tohoku University, Sendai 980-8577, Japan ∇ International Center for Young Scientists, National Institute for Materials Science, Tsukuba 305-0047, Japan ○ International Center for Materials Nanoarchitectonics, National Institute for Materials Science, Tsukuba 305-0044, Japan ◆ Department of Materials Science and Engineering, Tokyo Institute of Technology, Tokyo 152-8552, Japan ¶ Laboratory for Materials and Structures, Tokyo Institute of Technology, Yokohama 226-8503, Japan ■ Mathematics for Advanced Materials-OIL, AIST-Tohoku University, Sendai 980-8577, Japan ▲ Tsukuba Research Center for Interdisciplinary Materials Science (TIMS), and Center for Integrated Research in Fundamental Science and Engineering (CiRfSE), University of Tsukuba, Tsukuba 305-8571, Japan ◇ Materials Research Center for Element Strategy, Tokyo Institute of Technology, Yokohama 226-8503, Japan S Supporting Information *

ABSTRACT: Two-dimensional (2D) materials are promising for applications in a wide range of fields because of their unique properties. Hydrogen boride sheets, a new 2D material recently predicted from theory, exhibit intriguing electronic and mechanical properties as well as hydrogen storage capacity. Here, we report the experimental realization of 2D hydrogen boride sheets with an empirical formula of H1B1, produced by exfoliation and complete ion-exchange between protons and magnesium cations in magnesium diboride (MgB2) with an average yield of 42.3% at room temperature. The sheets feature an sp2-bonded boron planar structure without any longrange order. A hexagonal boron network with bridge hydrogens is suggested as the possible local structure, where the absence of longrange order was ascribed to the presence of three different anisotropic domains originating from the 2-fold symmetry of the hydrogen positions against the 6-fold symmetry of the boron networks, based on X-ray diffraction, X-ray atomic pair distribution functions, electron diffraction, transmission electron microscopy, photo absorption, core-level binding energy data, infrared absorption, electron energy loss spectroscopy, and density functional theory calculations. The established cation-exchange method for metal diboride opens new avenues for the mass production of several types of boron-based 2D materials by countercation selection and functionalization.



electron confinement.1−5 Combining 2D materials through layer stacking in a controlled manner can also produce several novel functionalities in the form of new three-dimensional (3D)

INTRODUCTION Two-dimensional (2D) materials consisting of a single or a few layers of atoms have superior performance as compared to conventional materials or their bulk counterparts in a variety of applications, because of their unique properties, including their flexibility, high specific surface area, and quasi-2D © 2017 American Chemical Society

Received: June 13, 2017 Published: September 19, 2017 13761

DOI: 10.1021/jacs.7b06153 J. Am. Chem. Soc. 2017, 139, 13761−13769

Article

Journal of the American Chemical Society layered materials (van der Waals heterostructures).6 Therefore, synthesis of new 2D materials opens several pathways for the applied use of van der Waals heterostructures. Single layer boron (borophene) sheets are a new type of 2D material, theoretically predicted at first7−13 and then recently synthesized on Ag(111) surfaces by bottom-up methods.14,15 On the other hand, hydrogenated borophene (borophane, boron hydride, or hydrogen boride) sheets have only been reported from theoretical studies examining several structure types,16,17 and have not been realized experimentally. Borophane sheets are predicted to have interesting electronic and mechanical properties16 as well as favorable hydrogen storage capacity.17 In this work, we report the experimental realization of 2D hydrogen boride (HB) sheets with an empirical formula of H1B1 created by the complete cationexchange between protons and the magnesium cations in magnesium diboride (MgB2). Liquid exfoliation of 3D layered materials is widely used to obtain 2D materials at room temperature, due to its low cost and simplicity, and can be classified as a top-down approach in comparing to synthesis involving material buildup (bottom-up approach). Sheets are obtained in solution or suspension by exfoliation from 3D layered materials through sonication in a surfactant solution, ion/polymer intercalation, or functionalization followed by exfoliation.18−20 To date, numerous 2D materials including graphene, hexagonal boron nitride (h-BN), transition metal dichalcogenides (e.g., WS2 and MoSe2), metal halides (e.g., MoCl2 and PbCl4), and oxides (e.g., MnO2 and LaNb2O7) have been produced by exfoliation in liquid. As the parent material in the top-down synthesis of borophene or borophene-related 2D sheets, we have focused on MgB2, a binary compound composed of hexagonal boron sheets alternating with Mg cations. Because MgB2 inherently contains 2D boron sheets, it is of interest to determine whether borophene could be formed by simple exfoliation and deintercalation of Mg. According to a recent report, however, ultrasonication of water with MgB2 at room temperature produces Mg-deficient hydroxyl-functionalized boron nanosheets rather than pure boron sheets.21 The presence of Mg and hydroxyl species in nanosheets can be explained by the instability of charged boron sheets in water derived from MgB2 by exfoliation. In our previous study, we clarified that MgB2 is exfoliated in water not by simple Mg deintercalation, but by cation-exchange reactions between protons and Mg cations in MgB2, where the produced HB sheets subsequently react with water to form Mg-deficient hydroxylfunctionalized boron nanosheets as a result of hydrolysis.22 We thus consider that the designed ion-exchange method between Mg cations and other cations is a way to produce well-defined, borophene-related, and stable 2D materials through a top-down approach. In this work, we demonstrate the experimental formation of HB sheets by using cation-exchange between protons and Mg cations of MgB2 in nonaqueous media at room temperature.

42.3% (see Figure S1). The main byproduct was B(OH)3, which was formed by the hydrolysis reaction of HB sheets with water absorbed in the ion-exchange resin (see the Supporting Information). The HB sheets were purified by filtration after cooling the acetonitrile suspension to 255 K (see Figure S2). In the methanol suspension, B(OH)3 reacts with methanol to form trimethyl borate, which evaporates under the drying process at 343 K. Scanning electron microscopy (SEM) images of HB sheets exhibit clear bent and/or folded sections as shown in Figure 1B, indicating the flexible nature of the exfoliated 2D sheets. No Mg signal was detected from the HB sheets by X-ray photoelectron spectroscopy (XPS), as shown in Figure 1C, while a significant amount of hydrogen (H2) was detected as the major desorption species in thermal desorption spectroscopy (TDS) measurements (Figures 1D and S3). These results indicate that the Mg cations of MgB2 were completely exchanged by protons. On the basis of TDS results and weights for 10 different sample lots, the elemental ratio of the HB sheets was determined as H:B = 1:1.02 ± 0.18 (see Figure S3 for estimation details). These results show that HB sheets with a B:H ratio of 1 were formed by following a stoichiometric ion-exchange reaction at room temperature (eq 1): MgB2 + 2H+ → Mg 2 + + 2HB

(1)

It is well-known that XPS core-level binding energies reflect the charge states of surface elements in terms of chemical shift, and from the XPS B 1s spectra we can determine the charge states of surface boron atoms before and after the ion-exchange process. As shown in Figure 1C, there are two peaks in the B 1s region for both MgB2 and the HB sheets. In the case of the starting material (MgB2), the peak at the lower binding energy (188.2 eV) corresponds to the negatively charged boron species in MgB2, while the peak at the higher binding energy (193.2 eV) corresponds to the positively charged boron forming boron oxides such as B2O3 on the MgB2 surface, as reported previously.23 Mg 2p peak from MgB2 at 51.3 eV is also in accordance with the value reported in literature, which is known to reflect Mg positive charge greater than 0 but less than +2 (B is thus negatively charged).23−25 We note here that the negatively charged boron species can be clearly observed after the ionexchange process (HB sheets) as a peak at 187.4 eV. The presence of a negatively charged boron peak without the Mg peak in Figure 1C further indicates that the Mg cations were exchanged with protons by this ion-exchange process to form HB sheets. The absence of Mg and other species was further shown by energy dispersive X-ray spectrometry (EDS, Figure S4) and the wide-scan XPS data shown in Figure S5. The small peak at 193.2 eV in the sample after the ion-exchange process arises from the presence of B(OH)3 as a byproduct (see the Supporting Information). Similar but different nonvolatile yellow boron hydride solids were reported in previous research on boron hydride done by Stock in 1923−1924,26 where two types of yellow boron hydrides were reported. (i) A yellow boron hydride powder (H/B = ∼0.8) was obtained by thermal treatment of B4H10 gas prepared by HCl treatment of magnesium boride; only 6 g of B4H10 gas can be obtained from 4200 g of magnesium boride, but it is a major boron hydride gas produced by this method. This product retained hydrogen even after being heated at 300 °C. (ii) Another yellow boron hydride powder was obtained by holding B6H12 gas at room temperature to form a solid product through a spontaneous decomposition reaction. Although the



RESULTS AND DISCUSSION HB Sheet Formation by Cation-Exchange with MgB2. Ion-exchange was conducted by adding MgB2 powder to methanol (or acetonitrile) in the presence of an ion-exchange resin; this entire process was conducted under nitrogen at room temperature and ambient pressure with stirring (Figure 1A). After 3 days, black precipitates were removed by filtration, and the filtrate was dried on a hot plate at 343 K under nitrogen. HB sheets were then obtained as a yellow powder as the ionexchange reaction product at an average reproducible yield of 13762

DOI: 10.1021/jacs.7b06153 J. Am. Chem. Soc. 2017, 139, 13761−13769

Article

Journal of the American Chemical Society

Figure 1. Ion exchange between protons and Mg cations of MgB2 produces HB sheets. (A) Synthesis scheme for HB sheets. (B) SEM image of HB sheets. Left: powders of HB sheets mounted on Si wafer. Right: HB sheets obtained by drying supernatant of the solution shown in panel A on a TEM grid. (C) XPS B 1s and Mg 2p core level spectra for MgB2 and HB sheets. (D) TDS of HB sheets, where the elemental ratio of B:H is approximately 1:1 based on TDS data and weights for 10 different sample lots.

as 1200 °C. The difference is probably caused by the different synthesis route; that is, HB sheets were formed by top-down approach of ion-exchange and exfoliation with keeping boron framework, while boron hydride solids were formed by bottomup synthesis using small boron hydride molecules. HB Sheets Consist of Boron with an sp2 Bonding Configuration. Typical SEM, transmission electron microscopy (TEM), and scanning transmission electron microscopy (STEM) images of the HB sheets are shown in Figure 2A. The observed stable sheets can not only form larger sheets (as much as a few tens of micrometers in size) but also cages or tubes, indicating the flexible nature of these HB sheets (see also Figure S6). The HB sheet thickness can be roughly estimated in the range of 0.5 nm to several nanometers on the basis of the line profile of the edges as indicated by “L” in Figure 2B, suggesting

composition of this product was reported as identical to that of the former yellow solid, this latter product completely decomposed to boron and hydrogen at 220 °C. In 1959, Shapiro and Williams reported that the reaction of diborane and decaborane in the liquid phase results in the formation of nonvolatile yellow solids with a BH stoichiometry.27 Their infrared analysis indicated the loss of so-called bridge hydrogens, which differs from the case of our HB sheets. Also, their solid loses approximately one-half of its hydrogen as a function of temperature up to 180−200 °C, which also differs from our HB sheets. We thus consider that the HB sheets produced by ion-exchange in this work are different from previously reported boron hydride solids synthesized from small boron hydride molecules, especially in terms of the decomposition feature as shown by TDS in Figure 1D, where H2 desorption starts from about 150 °C and continues up to as much 13763

DOI: 10.1021/jacs.7b06153 J. Am. Chem. Soc. 2017, 139, 13761−13769

Article

Journal of the American Chemical Society

Figure 2. SEM, TEM, electron diffraction, STEM, and EELS of HB sheets. (A) SEM, TEM, and STEM images. (B) Electron diffraction and TEM image. The line profile at position “L” is also shown. (C) EELS spectrum recorded during TEM measurements.

of the sample, we conducted angstrom-beam electron diffraction measurements31 using STEM at a low current (approximately 0.1 pA). As shown in Figure 3A, electron diffraction patterns were present when measured at the local position (approximately 1 nm beam spot area) by STEM (spots are indicated by arrows in Figure 3A, right-top), while electron diffraction scanning over the entire sample did not indicate any diffraction spots (Figure 3A, right-bottom). From these results, we can conclude that HB sheets do not have any long-range order but have local structural order. To further investigate the short-range order observed by STEM, we conducted X-ray atomic pair distribution function (XPDF) analysis based on our XRD data (Figure 3B) as shown in Figure 3C (details of the analysis are summarized in Figure S7). We were careful with whether or not the contribution from hydrogen exists, and found that in our medium Q range data, the contribution from H in the B−H compounds is not negligible. The ratio of atomic scattering factors of H and B is calculated to be 0.25 (Q = 0 Å−1), 0.023 (Q = 5 Å−1), or 0.0084 (Q = 10 Å−1) using previously reported methods.32 Thus, the H atoms do contribute to these data, and we can obtain a reasonable structure function (Figure S7D). We carefully converted the XRD data into XPDF data, because the PDF data suggest the presence of atom pairs below 1 Å. PDFs in this range often suffer from termination ripples caused by Fourier transform, but we believe that the features in Figure 3C do not represent error because the ripples decayed in the r range shorter than the peaks. Experimentally obtained XPDF patterns were thus fitted well by a simulated pattern for the model local structure shown in Figure 3D, where the hydrogen position was optimized by first-principles calculations based on density functional theory (DFT). This XPDF-derived structure consists of a hexagonal boron sheet with hydrogen atoms at the bridge sites of both faces of the sheet,

that the HB sheets consist of a single or at most a few layers. We note here that further examination is required to exactly determine the sheet thickness and to establish a thickness histogram of the nanosheets in the product by using atomic force microscopy (or some other method). The elements and bonding configuration of the HB sheets were directly analyzed during TEM measurements by using electron energy loss spectroscopy (EELS). EELS confirmed that boron is the dominant species in the sheet (Figure 2C), as signals are only detected above the background level within the boron energy region (∼190 eV), while signals for carbon, nitrogen, and oxygen were not detected. This is consistent with XPS (Figures 1C and S5) and EDS (Figure S4) results. In the boron energy region, two clear peaks are observed at 187.5 and 197.1 eV, which can be assigned as EELS peaks originating from the K-shell excitation from B 1s to π* and B 1s to σ*.28,29 This indicates that the boron atoms of the HB sheets have a planar sp2 bonding configuration rather than sp3 bonding. Possible Structure of the HB Sheets. X-ray diffraction (XRD) patterns of HB sheets show only broad peaks (Figure 3), suggesting that the HB sheets do not feature long-range order. Atomically resolved TEM and electron diffraction measurements also confirm the absence of long-range order (Figure 2B). In the case of 2D sheets, however, XRD peaks are not always observed or are very weak even for sheets having crystalline structures due to the flexible nature of the sheet; XRD data of reduced graphene oxide do not show any peaks as well.30 For electron diffraction and TEM observations, electron irradiation may cause perturbations within the sample due to the low-mass elements like boron and hydrogen being present, preventing observation of the ordered atomic structures. In this work, to distinguish the amorphous nature of the sample from the electron beam damage 13764

DOI: 10.1021/jacs.7b06153 J. Am. Chem. Soc. 2017, 139, 13761−13769

Article

Journal of the American Chemical Society

Figure 3. STEM, electron diffraction, XRD, XPDF, FTIR, and UV−vis of HB sheets. (A) Left: STEM image and EELS spectrum taken at the position indicated by the arrow. Right: Electron diffraction patterns at the local regime with angstrom-beam (top) and whole regime (bottom). (B) XRD pattern of HB sheet, where both background and byproduct B(OH)3 signals are subtracted (see Figure S7). (C) XPDF of HB sheet. Experimental results (blue curve) were fitted by simulated results (red curve) using the model structure shown in panel D, where the green curve shows the difference between experimental and simulated results. (D) Model local structure of HB sheet used for the fitting simulation of the XPDF experimental results, where hydrogen positions were optimized using DFT calculations. (E) FTIR spectra of HB sheets and the starting material MgB2. (F) UV−vis spectra of HB sheets and the starting material MgB2.

A B:H ratio of one also matches the experimentally derived ratio (Figure 1D). Both the presence of planar sp2 bonding in boron (as shown by EELS in Figure 2C) and the presence of BH stretching for B−H−B linkage33,34 at 1619 cm−1 in the FTIR data (Figure 3E) match this XPDF-derived structure in terms of bonding characteristics. The experimental photo absorption of HB sheets is roughly estimated as ∼2.9 eV from the onset of the absorption (Figure 3F), which is similar but slightly larger than the onset energy at 2.3 eV in the DFT photo absorption spectrum in the xy plane (Figure 4) for the XPDF-derived structure; the excitation at Γ between the bands shown by the arrow in Figure 4A corresponds to this absorption (see also Figures S9 and S10). The optical absorption spectrum reasonably explains the yellow color of the HB sheets (Figure 1A), because energy levels over 2.9 eV correspond to blue to

forming B−H−B linkages (similar to the three-center twoelectron bonds of diborane11,17 in terms of H delocalization) with 1:1 HB stoichiometry. The expected diffraction spots of this derived structure are superimposed with the experimentally observed local electron diffraction pattern in Figure 3A by yellow circles, and are consistent with observed diffraction positions. We note here that the observed electron diffraction cannot be explained by known boron hydrides; we compared this result with simulated diffraction patterns of B2H6, B5H9, B9H15, B10H13, B10H14, B12H16, B13H19, B14H20, B16H20, B18H22, and B20H26, regardless of whether they are solid or not at room temperature and atmospheric pressure (Figure S8). The hexagonal boron structure in the XPDF-derived HB sheet (Figure 3D) is nearly identical to that of the MgB2 starting material, resembling the flat hexagonal structure of graphene. 13765

DOI: 10.1021/jacs.7b06153 J. Am. Chem. Soc. 2017, 139, 13761−13769

Article

Journal of the American Chemical Society

Table 1. B 1s Core Level Binding Energies of HB Sheets sample experiment DFT calculation

Figure 4. DFT results of the electronic states, absorption, and IR spectrum. (A) Electronic energy band of infinite monolayer HB sheet shown in Figure 3D. (B) Optical absorption spectra of monolayer HB along the xy and z directions. (C) IR spectrum of monolayer HB, where IR intensity is plotted versus wavenumber (v), and the vibrational mode at 1613 cm−1 associated with the depicted motion.

B(OH)3 hydrogen boride sheets B(OH)3 XPDF-derived structure Borophane I (B4H4, Cmmm) Borophane II (B2H2, P3̅m1) Borophane III (B2H2, Pmmn) Borophane IV (B4H4, CPbcm) Borophene

B 1s (eV)

relative energy difference (eV)

193.2 187.7 202.4 196.2 197.3

0 −5.8 0 −6.2 −5.1

198.7

−3.7

199.1

−3.3

199.4

−3.0

201.2

−1.2

agreements between experiment and theory, H1B1 empirical formula, sp2 bonding configuration, presence of B−H−B linkage, photo absorption, and binding energy, we can conclude that the XPDF-derived structure (Figure 3D) is the most plausible local structure of the obtained HB sheets as compared to the predicted borophane sheets16 and reported boron hydride clusters (Figure S8). Next, we discuss the absence of long-range order in these sheets. As shown in Figure 3D, the XPDF-derived HB sheets have 2-fold symmetry around the hydrogen positions, but 6-fold symmetry with regards to the boron networks. Thus, three different anisotropic domains can be formed on the same hexagonal boron framework; hydrogen does not have any longrange order because ion-exchange between Mg cations and protons should occur simultaneously at the MgB2 positions, where the B−B distance differs when H is present and when H is absent in each domain. More specifically, in the ion-exchange process, protons are released to the polar solvent from SO3H groups in the ion-exchange resin used in this work (Figure 1A) and exchange with magnesium cations in MgB2. This ionexchange should start at surfaces and edges of MgB2 crystals simultaneously, rather than inside of the MgB2 material because surface and edges are exposed to the suspension including ionexchange resin. The exfoliation of the sheets is then considered to occur together with further ion-exchange in the inside regions of the MgB2 crystals. Each ion-exchange process does not break the resultant charge neutrality, because of the reaction as expressed in eq 1. In this process, protons randomly occupy three of six possible bridge sites in each hexagonal boron-network, which results in breakage of the periodicity and lattice symmetry, although the reaction is a kind of topotactic conversion. Therefore, the anisotropic domain of the HB sheets results in the absence of long-range ordered structures in the experimentally obtained HB sheets. If the hydrogen positions are selected as one of the three domain structures in the whole region, the structure would be the same as that from previous theoretically predicted sheets such as Borophane I16,17 with B−B distances of 1.82 and 1.71 Å for the bonds with H and without H, respectively,17 differing slightly from experimentally derived distances (1.775 and 1.579 Å). This difference between the experimental and theoretical B−B distances is likely caused by the strain induced by the coexistence of three different domains in the experimentally obtained sheets. Indeed, the presence of the strain is consistent with a large IR absorption intensity ratio between 2509 and 1619 cm−1 peaks (Figure 3E) as described above (see also Figure S11). The absence of long-range structural order can thus be ascribed to the coexistence of the three types of different anisotropic domains formed on each hexagonal boron sheet.

violet wavelengths. In low-dimensional systems, the exciton effect is significant because electrostatic screening is limited. The relatively good agreement of the onset of the absorption between simulated spectrum (Figure 4B) and experimental spectrum (Figure 3F) is thus considered to be due to the error cancelation between underestimation of the fundamental band gap and the absence of the exciton effect in our DFT calculations. The theoretically calculated energy of the B−H stretching vibrational mode of B−H−B linkage for XPDF-derived structure is 1613 cm−1 as shown in Figure 4C, which is in good agreement with experimental results for HB sheets (1619 cm−1, Figure 3E). In addition, we have also simulated an IR spectrum of small boron hydride B20H32 clusters, where new peaks appear at ∼2500 cm−1 corresponding to absorption by the nonbridging B−H groups (stretching vibration modes) at cluster edges (Figure S11). Therefore, the absorption at 2509 cm−1 in Figure 3E can be assigned to nonbridging B−H stretching vibrations33,34 at HB sheet defect sites, domain boundaries, and/ or edges, which are absent in the simulated infinite sheet structures. Furthermore, we also found by DFT calculation that the ratio between nonbridging B−H and bridging B−H−B peaks in the B20H32 clusters is sensitive to the strain (Figure S11). Therefore, the large IR absorption intensity ratio between the 2509 and 1619 cm−1 peaks in the synthesized HB sheets is reflected not only in the difference of the amount of species present with a corresponding vibrational mode, but also in the sheet strain. We note here that the DFT calculations used in this work usually underestimate both photo absorption and vibrational frequencies as compared to experimental results. In addition, the absolute value of the core level binding energy is not accurately computed by DFT due to the difficulty in treatment regarding the finite screening effect of core-hole pairs during photo emission. Thus, we have compared the experimental and theoretical results of the B 1s binding energy for HB sheets referenced to that of B(OH)3. The computed B 1s binding energy difference for the XPDF-derived HB structure (6.2 eV) is in good agreement with the experimental result (5.8 eV), while other theoretically predicted structures of borophane sheets,16 except for Borophane I (B4H4, Cmmm, Figure S9), yield much smaller values of 3.1−3.7 eV as summarized in Table 1. The Borophane I structure is similar to the XPDF-derived structure and thus shows a relatively higher value of 5.1 eV, which is still lower than the experimental value. The difference between the XPDF-derived and Borophane I structures will be discussed later. On the basis of these 13766

DOI: 10.1021/jacs.7b06153 J. Am. Chem. Soc. 2017, 139, 13761−13769

Article

Journal of the American Chemical Society



inaccurate, but we confirmed that this difference was within the range of PDF change due to possible H defects. The process of this analysis is shown in Figure S7. The S(Q) was then converted into reduced PDF, G(r), where r is the interatomic distance.

CONCLUSIONS We have realized 2D hydrogen boride sheets with an empirical formula of H1B1 created by exfoliation and complete ionexchange between protons and Mg cations of MgB2 at room temperature. The sheets lacked long-range order and were composed of sp2-bonded boron, suggesting hexagonal boron networks with bridge hydrogens as a local structure. The lack of long-range order was ascribed to the anisotropic domains present in the sheets, and different degrees of symmetry in different sheet structural motifs (local structures). Although further research is needed, the cation-exchange method established here is a promising route for creating boron-based 2D materials.



G(r ) =

2 π

∫Q

Q max min

Q (S(Q ) − 1) sin(Qr ) dQ

(2)

The PDF data were analyzed by curve fitting using the PDFfit2 program for different structure models consisting of B and H (B:H ratio = 1) with hexagonal or nonhexagonal B networks. The model described in the main text consists of (i) stacked layers of Borophane I type (Figure S9) for simulating density only,33 and (ii) an isolated HB layer for simulating short-range order.34 Each H atom is considered to bond to two different B atoms to form three-center two-electron bonds. X-rays are scattered by electrons, and the positions of H atoms determined by X-ray measurements are known to reflect the electron locations; electrons in the H atoms should be observed between B and H atoms. In this light, the H atoms are modeled by using two electrons located between the H core (with no electron) and B atoms, rather than a single H atom. We note here that the possibility of a stacking sheets case was also examined on the basis of the comparison with DFT results (Figure S13). XRD measurements were also performed at room temperature using a two circle diffractometer (PW 3050 Philips X’pert Pro, PANalytical, Almelo, The Netherlands) with Cu Kα radiation (1.5418 Å). X-rays were generated using the line focus principle. A reflection-free Si plate was used as the sample stage. Cu Kα radiation was obtained by reflection from a singly bent highly ordered pyrolytic graphite crystal. The diffraction pattern was recorded using a solid-state detector (X’Celerator, PANalytical) with a scan speed of 0.05° 2θ/s up to 90°. High-resolution XRD data were also collected on the BL02B2 beamline at SPring-8 with an incident wavelength of λ = 0.4440351(8) Å in the Debye−Scherrer geometry with samples in spinning glass capillaries.35−38 XPS Measurements. XPS measurements were carried out at room temperature using a JPS 9010 TR (JEOL, Ltd., Japan) with an ultrahigh vacuum chamber and an Al Kα X-ray source (1486.6 eV). The pass energy was 10 eV, the energy resolution (estimated from the Ag 3d5/2 peak width of a clean Ag sample) was 0.635 eV, and the uncertainty in the binding energy was ±0.05 eV. The sample was placed on the sample holder with graphite tape with a metal contact and introduced to the ultrahigh vacuum chamber for measurement. The Shirley background was subtracted from the spectrum using SpecSurf software (version 1.8.3.7, JEOL, Ltd., Japan). For sample measurements after the ionexchange process (e.g., Figure 2B), charge-buildup in the sample caused higher binding energy shifts for those spectra. Thus, we calibrated the charge-up amount as −2.2 eV on the basis of a B(OH)3 binding energy value of 193.2 eV (see also Figure S2). TEM, STEM, EELS, EDS, and Electron Diffraction Measurements. Measurements were performed at room temperature using a JEM-2100F TEM/STEM (JEOL, Ltd., Japan) with double spherical aberration (Cs) correctors (CEOS GmbH, Heidelberg, Germany) to obtain high contrast images with point-to-point resolution of 1.4 Å. The lens aberrations were optimized by evaluating the Zemlin tableau of amorphous carbon. The residual spherical aberration was almost zero (Cs = −0.8 ± 1.2 μm with 95% certainty). The acceleration voltage was set to 120 kV, which is the lowest voltage that is effective with the Cs correctors in this system. A coherent electron beam for STEM nanodiffraction was produced using a specially designed condenser aperture with a diameter of 3.5 μm (Daiwa Techno Systems Co., Ltd.). The convergence angle and beam diameter were estimated to be 1.0 mrad and 0.8 nm, respectively. A low beam current (70 times smaller than that of the normal STEM setting) and a short exposure time (0.1 s) were used to avoid possible beam damage. All diffraction patterns were recorded with a CCD camera (Gatan, US1000). EELS measurements were carried out using a Gatan GIF Tridiem, and EDS spectra were acquired using a JEOL JED-2300T.

EXPERIMENTAL SECTION

Materials. MgB2 powder (60 mg, 99%, Rare Metallic Co., Ltd., Tokyo, Japan) was added to 200 mL of acetonitrile (99.5%, Wako Pure Chemical Industries Ltd., Osaka, Japan) or methanol (99.8%, Wako Pure Chemical Industries Ltd., Osaka, Japan), followed by ultrasonication for 30 min. The prepared MgB2−acetonitrile (or methanol) suspension was then added to another acetonitrile (or methanol) suspension (100 mL) containing the sulfur-containing ion-exchange resin (30 mL, Amberlite IR120B hydrogen form, Organo Corp., Tokyo, Japan) under nitrogen at room temperature (Figure 1A). After being stirred (250 rpm) for 3 days, black precipitates were removed by filtration (1.0 or 0.1 μm pore filter, Omnipore Membrane Filters, Merck Millipore, Billerica, MA), and the filtrate was dried on a hot-plate at 343 K under nitrogen, yielding the product as a yellow powder. Here, we note that we could obtain almost the same product by drying the supernatant of the suspension, although it has a little bit brownish color rather than yellow possibly due to the inclusion of large size sheets in this particular case. We used commercial B(OH)3 (99.5%, Wako Pure Chemical Industries Ltd., Osaka, Japan) as a reference sample for the byproduct in our sample. For the titration to determine the Hammett acidity function (H0, acid strength), we used methyl red (99.5%, Kishida Chemical Co., Ltd., Osaka, Japan), p-(dimethylamino)azobenzene (99.5%, Kishida Chemical Co., Ltd., Osaka, Japan), 2-aminoazotoluene (>97%, Tokyo Chemical Industry Co., Ltd., Japan), benzeneazodiphenylamine (99.5%, Wako Pure Chemical Industries Ltd., Osaka, Japan), crystal violet (99.5%, Kishida Chemical Co., Ltd., Osaka, Japan), and 4-nitrodiphenylamine (99% Sigma-Aldrich Japan) as indicators, while benzene (dehydrated, 99.5%, Wako Pure Chemical Industries Ltd., Osaka, Japan) or ethanol (99.8%, Wako Pure Chemical Industries Ltd., Osaka, Japan) was used as a suspension or solvent, respectively (Figure S12). H0 of the HB sheet is below 1.5 and above 0.43 as shown in Table S1, indicating that the HB sheets are a solid acid; that is, H in HB sheets are proton (H+) rather than hydride (H−), which is consistent with eq 1. XRD Measurements and XPDF Analysis. The X-ray total scattering patterns for the pair distribution function analysis were recorded on a Rigaku Rapid-S diffractometer with Ag Kα (λ = 0.560883 Å) operated at 50 kV and 40 mA at room temperature (27 °C). The samples were sealed in Cole-Parmer Kapton tubes (ID = 1 mm) under argon. Corrections were carried out for background intensity (air, a blank capillary, and boric acid crystals), absorption, fluorescence, X-ray polarization, and Compton scattering. The subtraction of the intensity associated with the boric acid crystals, such as the peak at 10°, was performed using scattering data obtained for another sample containing a greater amount of the boric acid byproduct (the origin of byproducts is described in the Supporting Information). These corrected intensities were normalized by the Faber−Ziman-type scattering form factors calculated using atomic scattering factors to obtain structure functions, S(Q).31 The S(Q) in the scattering vector range of Qmax = 20.0 Å−1 was treated with a revised Lorch function (Δ = 0.75).32 Note that the form factor calculated for B(0):H(0) = 1:1 does not fit the total scattering data, likely due to the unique electronic states of B and H, and thus we used B(0):H(−1) = 1:3 to minimize the termination ripples observed below 0.4 Å in the PDF data. Thus, the partial PDF may be slightly 13767

DOI: 10.1021/jacs.7b06153 J. Am. Chem. Soc. 2017, 139, 13761−13769

Article

Journal of the American Chemical Society UV−Vis Measurements. UV−vis absorption spectra of the powders were measured using a spectrophotometer (V-660, Jasco Ltd., Japan) with an integration sphere unit in diffuse reflectance mode. Fourier Transform Infrared Spectroscopy (FTIR) Measurements. FTIR spectra were measured at room temperature using a FT/IR-300 (JASCO Analytical Instruments, Easton, MD) with KBr pellet samples. The background signal was subtracted using Spectra Manager Software (JASCO Analytical Instruments). Scanning Electron Microscopy (SEM) Measurements. SEM measurements were performed on a JSM-521 (JEOL, Ltd., Japan) operating at 10 kV. Samples were placed on Si wafers, Cu-TEM grids, or Mo-TEM grids. Gas Analysis. The gas species desorbed from heating the samples were analyzed using a gas chromatograph (GC-8A, Shimadzu, Kyoto, Japan) equipped with a Molecular Sieves 5A and a Porapak Q. Thermal Gas Desorption Spectroscopy (TDS). Quantitative analysis of the gas species desorbed from samples was performed on a TDS-1400TV equipped with a quadrupole mass spectrometer (ESCO, Ltd., Japan). Specimens were heated to 1050 °C at a rate of 60 °C/min in an evacuated (