Large-Scale Synthesis and Comprehensive ... - ACS Publications

May 30, 2018 - Department of Chemistry, University of Kentucky, Lexington, Kentucky ... environmentally friendly route to synthesizing homogeneous Cu-...
0 downloads 0 Views 4MB Size
Article pubs.acs.org/IC

Cite This: Inorg. Chem. XXXX, XXX, XXX−XXX

Large-Scale Synthesis and Comprehensive Structure Study of δ‑MnO2 Jue Liu,*,† Lei Yu,‡ Enyuan Hu,§ Beth S. Guiton,‡ Xiao-Qing Yang,§ and Katharine Page*,† †

Neutron Scattering Division, Oak Ridge National Laboratory (ORNL), Oak Ridge, Tennessee 37831, United States Department of Chemistry, University of Kentucky, Lexington, Kentucky 40506, United States § Chemistry Division, Brookhaven National Laboratory (BNL), Upton, New York 11973, United States ‡

S Supporting Information *

ABSTRACT: Layered δ-MnO2 (birnessites) are ubiquitous in nature and have also been reported to work as promising water oxidation catalysts or rechargeable alkali-ion battery cathodes when fabricated under appropriate conditions. Although tremendous effort has been spent on resolving the structure of natural/synthetic layered δ-MnO2 in the last few decades, no conclusive result has been reached. In this Article, we report an environmentally friendly route to synthesizing homogeneous Cu-rich layered δ-MnO2 nanoflowers in large scale. The local and average structure of synthetic Cu-rich layered δ-MnO2 has been successfully resolved from combined Mn/Cu K-edge extended X-ray fine structure spectroscopy and X-ray and neutron total scattering analysis. It is found that appreciable amounts (∼8%) of Mn vacancies are present in the MnO2 layer and Cu2+ occupies the interlayer sites above/below the vacant Mn sites. Effective hydrogen bonding among the interlayer water molecules and adjacent layer O ions has also been observed for the first time. These hydrogen bonds are found to play the key role in maintaining the intermediate and long-range stacking coherence of MnO2 layers. Quantitative analysis of the turbostratic stacking disorder in this compound was achieved using a supercell approach coupled with anisotropic particle-size-effect modeling. The present method is expected to be generally applicable to the structural study of other technologically important nanomaterials.



decades,9,14,18−21 a conclusive result has not been fully achieved. The challenge is 3-fold: First, many synthetic, natural, or biogenic layered MnO2 are very small nanoparticles, hindering traditional structure solution22−25 using powder Xray diffraction (XRD) data due to strong peak broadening and overlap. Second, the existence of strong turbostratic stacking disorder brings further complexity into structure analysis.21,26 Third, the complex interactions among interlayer cations (Li+, Na+, K+, Mg2+, Pb2+, Cu2+ etc.) and interlayer water molecules lead to complicated in-plane ordering/disordering.9,14,26 In particular, the complicated arrangements of interlayer species are reported to result in the formation of different interlayer spacings, with the three most common lateral distances being 5, 7, and 10 Å.6,20,27 A variety of local structure probes, such as extended X-ray fine structure spectroscopy (EXAFS) and X-ray pair distribution function (PDF), were thus utilized to investigate the local structure of synthetic/natural δMnO2.9,10,14,19,26 However, controversial results are often found, presumably because of either the limited probe length scale (as for EXAFS, up to a few angstroms) or the very limited sensitivity to light atoms (such as O and H) of X-ray PDF. Moreover, although it has been widely acknowledged that large amounts of interlayer water exist in almost all prototypes of synthetic or natural phyllomanganates,5,20,26,28 the influence of

INTRODUCTION MnO2 is a polymorph-rich metal oxide; at least six different polymorphs, namely, α, β, γ, λ, R, and δ, have been synthesized in atmospheric conditions (Figure S1).1−7 Among these different polymorphs, the layered δ-MnO2 (or birnessite) has attracted tremendous attention because of its structural similarity to various biogenic MnOx, which play important roles in Mn recycling in oceanic and terrestrial environments.8−11 It has also been well-recognized that porous δMnO2 has a strong capability of absorbing heavy-metal ions, such as Pb2+, As3+, Zn2+, and Cu2+, and thus has long been considered as a promising candidate for heavy-metal-ion remediation.12,13 More recently, layered MnOx nanoparticles or thin films have been reported to show excellent electric/ photocatalytic water oxidation capabilities when fabricated under appropriate conditions.14,15 In particular, it has been reported that the catalytic capabilities of these chemically or electrochemically synthesized MnOx nanoparticles are strongly correlated with the specific types of interlayer cations.16,17 It has also been reported that the interlayer cations have a strong influence on the rechargeable battery performance of δ-MnO2.5 Thus, it is clear that an accurate structure determination will help to better understand the relationship between the structure and functionality of this promising multifunctional metal oxide. Although tremendous efforts have been spent on determining the structure of synthetic or natural δ-MnO2 in the last few © XXXX American Chemical Society

Received: February 20, 2018

A

DOI: 10.1021/acs.inorgchem.8b00461 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

samples were prepared by sonication of the as-synthesized MnO2 nanoparticles in isopropyl alcohol, followed by dropcasting onto laceyC-coated Cu TEM grids and then drying at 60 °C for 1 h. Imaging processes were carried out in ImageJ.35 3. Thermogravimetric Analysis (TGA) and Differential Scanning Calorimetry (DSC). TGA was carried out between 25 and 600 °C on a Q5000IR (TA Instruments) system with a constant flow of nitrogen gas (25 mL/min). The sample was placed in an aluminum pan and stabilized at 25 °C for 2 h, then heated to 100 °C at 2 °C/min, and held at this temperature for 4 h. It was then heated to 600 °C using a 2 °C/min ramp rate. DSC was carried out on a DSC 404 F1 Pegasus system (NETZSCH, Germany). During the measurement, the temperature was scanned from 25 to 700 °C at a heating rate of 1 °C/min with a constant Ar flow of 20 mL/min. 4. XRD. Powder XRD patterns were collected on a PANalytical X’Pert MPD Pro powder X-ray diffractometer with Cu Kα radiation (λ = 1.54187 Å) equipped with an Anton Paar XRK-90 linear detector. The XRD experiments were carried out with a step size of 0.016711° and at a scan rate of 100 s/step in a Bragg−Brentano configuration. The primary and secondary radii were set to 240 mm. High-resolution synchrotron XRD data were collected at beamline 11-BM at Argonne National Laboratory (ANL) with a wavelength of 0.41398 Å.36 Room temperature synchrotron X-ray total scattering data were collected on beamline 11-ID-B at the Advanced Photon Source (APS), ANL. The rapid-acquisition PDF method was used with an X-ray energy of 58.6 keV (λ = 0.2114 Å).37 A PerkinElmer amorphous Si two-dimensional image-plate detector (2048 × 2048 pixels and 200 × 200 mm pixel size) was used for two-dimensional data collection with a sample-to-detector distance of ∼180 mm. The two-dimensional data were converted to one-dimensional XRD data using the Fit2D software.38 PDF data were obtained from Fourier transformation of the background and Compton scattering corrected data [S(Q)] in PDFgetX2 with a Q range of 0.1−26 Å−1.39 5. Neutron Total Scattering. Neutron total scattering data of deuterated samples were collected at the NOMAD beamline at the Spallation Neutron Source (SNS) at ORNL.40 For the current experiment, about 1 g powder sample was loaded into a 6 mm V can. Two 30 min scans were collected for each powder sample and then summed together to improve the statistics. The detectors were calibrated using scattering from a diamond powder standard prior to the measurements. Neutron powder XRD data were normalized against a V rod, the background was subtracted, and the total scattering structure factor S(Q) data were transformed to PDF data G(r) using the specific IDL codes developed for the NOMAD instrument with a Q range of 0.2−31.4 Å−1. 6. Mn and Cu K-edge X-ray Absorption Spectroscopy (XAS). Mn and Cu K-edge XAS spectra were collected at beamline 8-ID at the National Synchrotron Light Source II at BNL. Data were collected in transmission mode, with the beam intensity detuned to 60% to minimize high-order harmonics. A reference spectrum of a Cu/Mn foil was collected simultaneously as an energy calibrant. The X-ray absorption near-edge structure (XANES) and EXAFS spectra were first processed using the Athena software package.41 The AUTOBK code was then used to normalize the corresponding absorption coefficients and separate the EXAFS signal, χ(k), from the isolated atom-absorption background. The corresponding EXAFS signals χ(k) were weighted by k3 to emphasize the signal at high k and then Fourier-transformed into real space.42 7. Modeling and Refinement Methods. Local structure refinements using X-ray and neutron PDF were carried out in TOPAS Academic, version 6.33 The instrumental related dampening [dQ, instrumental fwhm of S(Q)]43 and broadening (Qb)44 were refined from the PDF data of standard NIST Si (neutron) or Ni (Xray) powder. dQ and Qb were refined as 0.053 Å (1−30 Å) and 0.026 Å−2 for NOMAD and 0.067 Å (1−30 Å) and 0.025 Å−2 for 11-ID-B. These two values were fixed during further structure refinements. A sine function [sin(Qr)/Qr] was convoluted to the calculated PDF to account for the termination effect due to the finite Qmax used for Fourier transform.45 An empirical PDFgui-type Δ2 (Δ2/r2) term was

the interlayer water, such as hydrogen bonding, on the structure and morphology has not been well understood. In contrast to X-ray/electron scattering, where the atomic scattering power increases monotonically as the atomic number increases, the (coherent) neutron scattering length is isotope-dependent and very sensitive to light atoms such as O (b = 5.8 fm) and 2H (D) (b = 6.67 fm). Therefore, neutron total scattering (both Bragg diffraction and diffuse scattering) provides the best opportunity to fully resolve the complex structure of δ-MnO2 nanoparticles. Here, we report a new environmentally friendly route to synthesizing Cu-rich δ-MnO2 nanoflowers in large scale. This heavily stacking faulted δ-MnO2 nanoflower was used as a model compound (of layered δ-MnO2) to carry out a thorough local and average structure investigation. In contrast to the traditional structure determination where the average structure is often first solved and refined while the local disorder is treated as a perturbation of the ideal periodic structure,29,30 to circumvent the challenges of the particle size, stacking disorder, and interlayer spacing, a different approach was used here; i.e., the local structure was determined first, and the intermediate and long-range structure was later constructed and refined by incorporating stacking disorder and nanoparticle size/shape effects. Combined X-ray and, for the first time, neutron PDF data were used to identify the accurate local structure of the synthetic Cu-rich δ-MnO2. The results are compared to analysis of the Cu and Mn K-edge EXAFS data. The intermediate-tolong-range structure was refined with combined X-ray and neutron scattering data using a numerical stacking disorder model in TOPAS (version 6).31−33 This local-to-average structure determination approach does not only enable the identification and quantification of both interlayer Cu2+ and water; it also enables the determination of average stacking sequences and (average) nanoparticle dimensions. It is expected that the current local-to-average structure probe can be directly adopted to study other layered oxides/hydroxide nanoparticles with heavy/turbostratic stacking disorder.34 It may also provide valuable hints on the rational design/ optimization of layered oxide materials for rechargeable battery and catalyst applications.



EXPERIMENTAL METHODS

1. Materials Synthesis. A total of 0.01 mol of KMnO4 (Sigma Aldrich, 99%) was dissolved in 100 mL of Milli-Q water, and 0.015 mol of Cu powder (Sigma Aldrich; particle size ∼ 100 μm) was added to the KMnO4 solution under vigorous magnetic stirring. A total of 0.04 mol of H2SO4 was first diluted in 20 mL of solution and then added to the mixed KMnO4 solution under vigorous stirring. The mixture was further heated to 40 °C on a hot plate. The color of the solution changed from purple (MnO4−) to dark green (Cu2+) at the end of the reaction (∼6 h). The product was centrifuged at a speed of 7000 rpm for 10 min, and the supernatant was removed. The black precipitate was further rinsed, washed, and ultrasonicated with Milli-Q water and then filtered. This process was repeated four times before drying the final black precipitates at 100 °C in a vacuum furnace. Elemental analysis using scanning electron microscopy (SEM)− energy-dispersive X-ray (EDX) reveals that the Cu-to-Mn ratio is about 1:8, while only trace amounts of K+ (