The Simplest MOF Units for Effective Photodriven Hydrogen Evolution

Subscriber access provided by UNIVERSITY OF TOLEDO LIBRARIES

Article

The Simplest MOF Units for Effective Photodriven Hydrogen Evolution Reaction Ting Liao, Liangzhi Kou, Aijun Du, YuanTong Gu, and Ziqi Sun J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.8b04599 • Publication Date (Web): 29 Jun 2018 Downloaded from http://pubs.acs.org on June 30, 2018

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

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

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

Journal of the American Chemical Society

The Simplest MOF Units for Effective Photodriven Hydrogen Evolution Reaction Ting Liao†*, Liangzhi Kou†, Aijun Du†, Yuantong Gu†, Ziqi Sun†* †

School of Chemistry, Physics and Mechanical Engineering, Queensland University of Technology, QLD 4000, Australia

ABSTRACT: Metal-organic frameworks (MOFs) combining the merits of both organic and inorganic functional building structures are fundamentally important and can meet the requirement of vast scientific and technological applications. Intrigued from the fact that transition metals (TMs) are widely embedded in carbon sp2 network or strongly interact with bare graphene edge, single transition metal atom may work as linker to connect carbon chains to build nano-architectures. A new MOF building structure, [Metal-Carbon-(Benzene)i-Chain]n rings short as [M-CBiC]n (M=Ti, V, and Cr) with increasing carbon chain length i (=0, 1, 2…), was proposed as carbon chains CBiC connected by single transition metal atom M to form a ring structure with multi-edges n (=26), based on advanced computational methods. They are thermodynamically stable and chemically and physically versatile with ring shape, electronic structures, optical response, as well as hydrogen adsorption energy, varies by changing the length of carbon chain, the edge number of rings, or the type of connecting metal atoms. The optical response to incoming light of [M-CBiC]n rings can be adjustable to cover the entire visible solar spectrum range and exhibit red shift by either increasing the edge number n or filling the d bands in connecting transition metals. In combination to their ideal adsorption energy of hydrogen atoms, |∆GH*|, the proposed [M-CBiC]n building structure is attractive for photocatalytic or photoelectrochemical hydrogen evolution applications when they are extended in space to build up 1D, 2D, and 3D dimensional MOF frameworks.

INTRODUCTION Metal-organic frameworks (MOFs) materials, constructed by linking finite metal centers or metal oxides to organic linkers, present extraordinary, sometimes unexpected physicalchemical properties, which usually combine the merits of both.1-5 By changing the composition and their geometry, such as diameter, one can in principle generate millions of possible MOF materials with diverse properties, which are suitable for tremendously various applications, such as gas sorption and separation,6-7 photocatalysis,8-9 photovoltaic,10-12 and electrochemistry,13-16 and more targeted applications are being extensively discovered and reported at an ever increasing pace. Discovery and rational design of new metal/metal oxides connected MOFs is a very active field with the key challenge depending on the appropriate choice of the constituent metal/metal oxides and bridging organic linkers to form a specific network. Many efforts have been endeavored to understanding and manipulating the structural, electronic, optical, and chemical properties of MOFs, which is key to transforming the MOFs from the lab into real-world.17-19 In practice, the rational design of MOFs is hampered by the difficulty on synthesis of these materials, so it is important to take advantage of advanced computational methods which allows us to establish database for discovering new porous MOFs frameworks ideal for target applications, and to save and support the experimental efforts. A key aspect of theoretical study is its ability to predict the properties of a MOFs material or its building structure, including structural, electronic, optical, and electrochemical properties, before the material is synthesized and to highlight a novel avenue for tuning specific performance of proposed frameworks. By doing this in a controlled way, it is possible to assess what metal atom or organic

chain can be adjusted to perturb the MOFs’ properties.20-21 In turn, these modified porous materials or building structures also open up the opportunity for achieving optimized properties of MOFs. At present, the advancement of density functional theory (DFT) calculations is able to provide important insights into the energetic and related properties for a variety of applications including catalysis, energy storage, sensing, and electronic devices.22-25 Herein, we describe a new MOFs building structures based on computational design. Inspired by the natural affinity of transition metal with carbon via either σ or π bonding characters as can be easily visualized in existing nanostructures, such as metal-embedded carbon nanosheet or nanotubes,26-27 metal hybrid or interconnected carbon nanostubes,28-29 metal-welded carbon nanotubes,30 metal joined benzene ring,31 etc., carbon chains are able to be directly connected by single transition metal to form building ring for MOFs nano-architectures. The ends of isolated carbon chains are terminated with undersaturated carbon atoms, and their reactions with transition metal atoms (M), such as Ti, V, and Cr with the oxidation states of +4, +5, and +6, respectively, will remove the dangling bonds and render those formed ring structure stable. A new MOFs building ring-which, as shown in Figure 1, -built from [Metal-Carbon-(Benzene)i-Chain]n rings short as [MCBiC]n, with multi-edges n=2-6 ring structures, where M (M = Ti, V, or Cr) is a single transition metal of 3d groups and CBiC is a carbon chain with increasing carbon chain length i=0, 1, 2…, -are proposed with versatile chemistry. A vast number of this MOFs architectures in one-dimensional (1D), twodimensional (2D) or three-dimensional (3D) networks may be successfully constructed with adjustable functionalization by varying either the transition metal atoms M in the connector,

ACS Paragon Plus Environment

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

edge numbers, or the length of the carbon chains (Figure 1). Their explored features, such as

Figure. 1: Schematic of assembling of carbon chains and single transition metals (left panel) into [M-CBiC]n (M=Ti, V, and Cr, i=0, 1, 2…, n=2-6) metal-organic building rings (middle panel) and into 1D, 2D and 3D MOFs framework (right panel).

thermodynamic stability, electronic and optical properties, and hydrogen evolution reactions, are obtained from DFT and time-dependent DFT (TDDFT) calculations, which might shed insight into attentively promising application of this kind novel MOFs as photocatalytic and photoelectrochemical hydrogen evolution materials. COMPUTATIONAL METHODS All electronic and structural calculations were performed within the spin-polarized density functional theory (DFT) framework as implemented in the Quantum-Espresso package.32 Ultrasoft pseudopotentials are introduced to describe the electron-ion interactions.33 Perdew-Burke-Ernzerhof (PBE) functional in generalized gradient approximation (GGA) was employed to describe the exchange-correlation functional.34 The Kohn-Sham (KS) orbitals and the charge density are represented using basis sets consisting of plane waves (PWs) up to a maximal kinetic energy of 90 Ry and 500 Ry, respectively. The proposed MOFs building structures with different edge numbers and carbon chains length were modeled using the supercell. A vacuum space of 15 Å in the non-periodic direction is used to avoid interaction between the neighboring images. Γ-point sampling of the Brillouin zone was adopted for the relaxation of each supercell, which is sufficient considering their large real-space dimensions. Denser k-point grid were then applied for the electronic structure analysis and optical properties study.35 The convergence in energy and force are set to 10-7 eV and 10-4 eV/Å. The dispersion interactions, which is not accounted for by local exchange and correlation functionals, are critical to the hydrogen binding energy for the open metal sites in [MCBiC]n structures, so we use the semi-empirical London-type correction (DFT-D2)36-37 to treat the long-range Coulomb interactions in the periodic structures. To better describe the electronic structures and optical properties of the open metal sites with localized d electrons, the Hubbard-type corrections U are also employed with the value of 3.5 eV.38 Such corrections are important for capturing the correct spin ground state of each building ring and improving the accuracy of the d band energies. The time-dependent density functional theory (TDDFT) calculations were carried out to obtain the full absorption spectra of [M-CBiC]n building rings using the same PBE functional.

Page 2 of 8

The Liouville–Lanczos algorithm in adiabatic approximation was employed to predict the absorption spectrum σ, which is implemented in turboTDDFT program as part of the Quantum Expresso package.39-41 The range of wavelength λ was selected to be comparable to the experimental UV-Vis region. The linear charge-density response to a perturbing electromagnetic radiation of wavelength λ was also obtained. RESULTS AND DISCUSSION Metal-organic frameworks (MOFs) can be regarded as extended structures that are built into destined nanoscale architectures using molecular building structures. In order to truly obtain MOFs structures by design, one had first to identify the principal building structures for network, which consists of metal or metal-oxide centers that are held together by organic linkers. Figure 1 presents the schematic of assembling of carbon chains and single transition metals into a new class of ring-shape MOFs building structures [M-CBiC]n then to highly porous periodic metal organic frameworks in diverse dimensions. In this work, the connecting center was chosen as single transition metal M (M=Ti, V, or Cr), which creates unsaturated “open” sites known to have a strong affinity for carbon network, coordinating with the intended carbon chains CBiC (i=0, 1, 2 …). Here, n is the number of carbon chains forming the edges of the building ring structures (n=2-6), and the number i along each edge of the carbon chains is used to define the number of benzene ring. The size and shape of this ring structures is adjusted accordingly to meet the requirement of different pore sizes ranging to nm as a common feature of MOFs. Transition metal atom Ti, V, and Cr have at least four 3d valence electrons that may form covalent bonds with the two edge carbon atoms of each carbon chain. The energy penalty due to unsaturated edge may be significantly reduced if the carbon chain is covalently terminated by a transition metal atom, similar to the dome termination of carbon nanotubes.42 These [M-CBiC]n building structures were identified as those with high point symmetry at the vertices and can be extended to space periodic structures, like one-dimensional single chains, two-dimensional plane with square or hexagonal nanodot, or three-dimensional architecture with one-dimensional pores (Figure 1). For example, [M-CBiC]n-based (i=0, n=4) three-dimensional MOFs architectures displayed in Figure 1 is in high symmetry of R3M. A unit cell representing a twodimensional [M-CBiC]n-based (i=0, n=4) MOFs plane has a symmetry of P422. The primitive unit cell of one-dimensional [M-CBiC]n-based (i=0, n=2) MOFs chain is in a symmetry of PMMM. To study the edge number dependence, the optimized structure of [M-CBiC]n (M = Ti, V and Cr) rings at fixed carbon chain length (i=0) are plotted in Figure 2, where each ring contains two to six carbon chains (from top to bottom) designated as n, from 2 to 6, respectively. These candidates of MOFs building rings consist of either an oval, triangular, square, pentagonal, or hexagonal shapes. As shown in Figure 2, when the linking transition metal M changes from Ti to V to Cr, the angles between the ends of two carbon chains, θ, increase in the same order with increasing 3d valence electrons numbers. For [Ti-CB0C]n ring, the carbon chains remain nearly in a straight line when they are linked by Ti atoms, especially for the greater edge-number rings. The corresponding angles are 93.74° for the two-edge ring and 107.86° for the six-edge ring, as shown in Figure 2. However, when V and Cr atoms are used as the linkers, carbon chains bend outward from the


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

Journal of the American Chemical Society ring center and the corresponding angles increase to around 99.12° to 122.24° and 100.63° to 128.21°, respectively, for the two-edge and six-edge rings. This trend is not affected by the length of the carbon chains by increasing the number i in the [M-CBiC]n rings and

To explore the thermodynamic possibility of assembling proposed [M-CBiC]n building rings from the constituent elements, we employ a general approach to determine their free energy of formation. The relative stability order from free energy of formation calculations can refer to a compound with different component and structure. It is worth noted that here

Figure. 3: Calculated Gibbs free energy of formation, δGf, as a function of the carbon molar fraction for [M-CBiC]n (M=Ti, V, and Cr) Metal-Organic framework building rings with increasing carbon chain lengths i and edge numbers n. The results for B80 (Ref. 35) and Ferrocene are included for comparison. Calculated δGf values for each carbon chain CBiC (i=0, 1, and 2) are plotted for reference as black, red and purple solid line, respectively, so are their optimized atomic structures.

Figure. 2: Optimized structures of [M-CBiC]n (M=Ti, V, and Cr, i=0) Metal-Organic framework building rings with increasing edge numbers n (top to bottom). The calculated angles of each ring are also highlighted. The carbon, titanium, vanadium, and chromium atoms are indicated by green, orange, purple, and blue spheres, respectively.

the interpretation to this phenomenon will be discussed by electronic structure analysis in later section. The single 3d transition metal atom, M (M = Ti, V, and Cr), introduced between the carbon chains is able to mediate their magnetic coupling. For each building rings of [M-CBiC]n, two spin coupling configurations, FM and AFM, were calculated to determine the preferred magnetic state. Ti connected [MCBiC]n rings show a non-magnetic property in regardless of the edge numbers. It is interesting to see that all the constructed [V-CBiC]n rings are found to have AFM ground states, and the total magnetic moment per ring increase consistently with increasing edge number, as shown in Table S1. The [CrCBiC]n rings clearly prefers AFM coupling state for less edge numbers (n=2-5) but change to FM magnetic state for more edge number (n=6). The calculated energy difference between the FM and AFM coupling structures, defined as ∆E=E(AFM)–E(FM), for each V connected [M-CBiC]n building rings also increase accordingly with increasing edge number similar to the total magnetic moment per unit cell. Among them, the [V-CBiC]n rings has the largest energy difference around 0.82 eV between the FM and AFM states, suggesting that the V atom is more effective in regulating the magnetic coupling between the carbon chains. The [M-CBiC]n building ring with the most stable magnetic states are used in the following discussions.

we are more interested to give a qualitative evaluation on the relative stability of hypothetical structures and to assess if they are energetically feasible to be synthesized. Here, we compute the free energy of formation δGf of the [M-CBiC]n building rings and compare with other existing nanostructures.43-47 δGf([M-CBiC]n) = Et([M-CBiC]n) – nMµM – mCµC –qHµH (M=Ti, V, and Cr, i=0-2, n=2-6) (1) where Et([M-CBiC]n) is the total energy of [M-CBiC]n building ring and µM, µC and µH are the chemical potential of each transition metal, carbon, and hydrogen, respectively, in their ground state structures. nM, mC and qH are the molar fractions of metal atom, carbon, and hydrogen in the building rings, respectively, satisfying nM + mC + qH = 1. Figure 3 shows the calculated free energy of formation of [M-CBiC]n building rings as a function of the carbon concentration, and those for CBiC carbon chains as well for reference denoted as solid lines. Similar to carbon bucky balls and carbon nanotubes, the calculated free energy of formation is positive and decreases with ring size by either increasing the carbon chain or the edge number, as expected from curvature effects.30 In details, for each transition metal, the increasing length of carbon chain lead to an overall increase in the free energy of formation of [M-CBiC]n rings, so are the CBiC carbon chains themselves as well, as shown in Figure 3. On the other hand, the stability of [M-CBiC]n rings become lower with reduced carbon ring edge number which can be better visualized as zoomed in the inset of Figure 3. As a result, the thermodynamical stability and binding strength is maximized at large sized [M-CBiC]n rings. However, for small [M-CBiC]n (i=0) rings, the calculated free energy of formation is also smaller than that of existing Ferrocene (2.6 eV/atom), which is competitively small to render [M-CBiC]n ring stable and high possibility of successful syn-



thesis. This can be realized utilizing advanced chemical methods, which allows either metal atom substitutions of selected carbon atoms on the ends of carbon chains or via a direct stepwise synthesis.48-49 Additionally, to check that the optimized structure was resistant to external deformations, the [M-CBiC]n rings were relaxed again after tensile strain being applied in the direction from ring center to the linking metal atom until 20%. The ring

Figure. 4: The ring shape change of Ti- to V- to Cr-connected [MCBiC]n (i=0) rings (acute to round ring) is interpreted by progressive filling of delocalized d bands. Schematic representation of the evolution of electronic configuration of connecting metal atoms to the crystalline field model for MC4 configurations to the ring shape change of [M-CBiC]n rings.

structure of each ring appears to be uniaxially stretched in the direction normal to the tensile direction. But none of them showed any sign of big changes in the structural shape and geometry until bonds being broken. It is anticipated that the ring structure may be uniformly stretched more easily than by breaking the ring symmetry. This situation demonstrates that the minimum of [M-CBiC]n ring is deep enough to retain its stability at moderate conditions. The particular geometry of [M-CBiC]n building ring can be rationalized to attribute to the nature of the pd hybridized bonds of the fourfold bonded metal atom and carbon atoms, which holds the key to understanding the structural integrity and bonding characters of the [M-CBiC]n building ring. In Figure 4 we analyzed the valence electronic states of [MCBiC]n building ring, for example i=0, to understand their variable shape change. The filled and unfilled electronic states are shaded with purple and green, respectively. From the band structures for all the [M-CBiC]n rings the gap states mainly comes from the d orbitals of metal atoms at the linking center. For the linkers of Ti, V or Cr atoms, they are pd hybridized

with carbon atoms in the assembled structures and have progressive filling of localized d orbitals that lie in the bandgap of bonding (σ) and anti-bonding states (σ*) as illustrated in Figure 4. The effect of carbon atoms on the electronic structure is minor, except for broadening of the d bands. So C-p orbitals are excluded form Figure 4 for an easy visualization. In an ideal four-coordinated tetrahedral symmetry (Td), the d electron levels of transition metal centers are split into degenerate dyz, dzx, dxy (t2g) and dz2, dx2–y2 (eg) levels. However, in Figure 4 the d orbitals from the Ti centers in [M-CBiC]n rings undergo crystalline field distortion towards trigonal pyramidal symmetry and split into three groups, dzx, (dz2, dx2–y2), and (dyz, dxz), with a sizeable gap (~1 eV) between the second two groups of orbitals (more details in regards of energy band gaps are in Figure S1 of the Supporting Information). For the under-coordinated open-metal sites, like V and Cr in [M-CBiC]n rings, the degenerated (dz2, dx2–y2) orbitals are further split into two, as shown in Figure 4. The most interesting thing is the (dyz, dzx) and dz2 orbtials change their orders in [V-CBiC]n and [Cr-CBiC]n rings, so are their crystal field symmetry changed accordingly from quasi-square-pyramidal to square planar symmetry, respectively. Indeed, the characteristic ring shape change of [M-CBiC]n building rings from a nearly acute ring shape for the isolated [Ti-CBiC]n rings to round ring shape in forming [Cr-CBiC]n rings (Figure 2) can be interpreted by the order of d electron bands, so is the crystal field models formed by transition metals and their four-coordinated carbon atoms, MC4. In details, Figure 4 displays the evolution of electronic configuration of connecting metal atoms to the crystalline field model for MC4 configurations to the ring shape change of [MCBiC]n rings. Due to the evolving electronic structures of d2s2 to d3s2 to d5s1 in connecting metal atoms, the crystal field of metal center and surrounding carbon atoms change accordingly from trigonal pyramidal to quasi-square-pyramidal to square planar, so is the symmetry of MC4 bonding configurations. Depending on the choice of the metal elements, one of the three coordination modes is thermodynamically preferred for the MC4 configurations. As shown in Figure 4, trigonal pyramidal (D3h) is preferred for TiC4 configurations in energy than the others. In the structure with lowest energy predicted for TiC4, the coordinated carbon atoms are more preferred to be aligned in an out-of-plane space, allowing a tilted space arrangement in according with the acute edge shaped ring structures. In contrast, the ground-state geometry that we have obtained for VC4 and CrC4 are quasi-square-pyramidal and square planar geometry, belonging to the C4v and D4h groups, respectively. The four carbon atoms surrounding the metal center are more likely to be in an in-plane alignment corresponding to round edge shaped ring structures. It should be noted that this simple configuration model assumes ideal coordination and is helpful to give an idea on the characteristics of local coordination environment of studied clusters, but structural distortion often seen in many specific crystalline fields which lead to deviation in the electronic structure is not considered. The diversity of the predicted [M-CBiC]n rings based MOFs with tunable fundamental properties offers many opportunities for a wide range of applications from opto-electronics to energy storage to catalysis. Except for the exceptionally high surface areas as other porous materials, the proposed [M-CBiC]n rings based MOFs are also characterized with a number of under-coordinated metal sites, which is naturally reactive and have high affinity to accommodate a variety of guest species,


Page 4 of 8

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

Journal of the American Chemical Society for example H+. Then the presence of the metallic sites in the [M-CBiC]n based MOFs, in addition to an electronic potential with an electron affinity that is lower than the water reduction potential, should be substantially promising to enhance the catalytic or photocatalytic activity, similar to a variety of other nanomaterials.50-53 The conduction band (CB) edge of [MCBiC]n rings were also evaluated on an absolute energy scale with respect to the vacuum level (Figure S2 in the Supporting Information). Our results clearly show that the conduction band edges of M-CC MOFs units are in the positions higher to the reduction potential for H+/H2. Separated electrons can then migrate to the reaction sites where hydrogen is adsorbed and perform visible light response H2 evolution activity. The

studies of the optical properties for [M-CBiC]n are also of interest in view of potential uses as an active material or in the buffer layer between the electrodes and the inorganic active materials in solar cell applications.54 The coordination frameworks usually present unexpected optical properties as compared to their constituents, and this constitutes an effective means to obtain functional materials with desired properties.

Figure. 5: Absorption spectra for [M-CBiC]n (i=0, M=Ti, V, and Cr) rings with increasing edge number (n=2-6). Inset is the spatial response charge density for the specified absorption peak of λ, λ', and λ'', respectively. The electric field is polarized as designated by the arrows. Positive charge density is in red color and negative charge density in yellow color.

Hence, engineering of the valence band energy of [M-CBiC]n building rings through linking metal or carbon chain functionalization could be an effective way to produce a promising photocatalyst active in the visible solar spectrum range. Our results for the optical absorption spectra of a series of selected [M-CBiC]n (M=Ti, V, and Cr, i=0, n=2-6) rings are presented in Figure 5. The modulation of absorption originates from the bonding nature between the carbon chains and the linking metal atoms. We predict a broad visible light absorption for [M-CBiC]n rings by changing the metal center or the edge numbers. We can see clear differences between the computed absorption spectra of the [M-CBiC]n rings by changing metal linkers and edge numbers. Different linkers of the same ring lead to very different photo-absorption cross sections, which is clearly reminiscent of the assigned conductive elec-

trons. As shown in Figure 5, the [Ti-CBiC]n rings exhibit quite narrow absorption bands centered in the blue-light adsorption region. Whereas the [V-CBiC]n and [Cr-CBiC]n rings give rise to broad and red-shifted bands up to the IR region. [V-CBiC]n rings shows a reduction of the absorption in the blue-light region but more on green-light adsorption. The computed spectra of [Cr-CBiC]n rings exhibit two major characters: an absorption band with intensity increasing with increasing edge numbers in the visible range around 500-600 nm, and a broad band with lower intensity up to the IR (Figure 5). Comparing the adsorption spectra for [M-CBiC]n with increasing edge numbers shows that electron-electron interactions induce a red-shift and a continuous increase of the absorption intensity. In particular, [Ti-CBiC]n rings exhibit the absorption band with the highest intensity at 400 nm for six-edge ring, while that of reduced edge number exhibits a narrower band between 350 and 400 nm, slightly blue-shifted and with lower intensity with respect to the electron band. The spectra of the [V-CBiC]n and [Cr-CBiC]n rings evolves into broad bands whose intensities reach a maximum at wavelength of 550 and 700 nm, respectively, for six-edge rings. While linking metal atoms are responsible for the appearance of absorptions in the optical range, the finite-size quantum-confinement effects with variant edge number play an important role in the intensities of absorption peak. The nature of the absorption peaks can be studied by examining the electronic response of selected wavelengths. The inset of Figure 5 shows the imaginary (absorptive) part of the response charge density F(r, λ), for the [M-CBiC]n rings (M = Ti, V, and Cr) at each specific wavelengths, λ, λ', and λ'', respectively. The response charge density exhibit the spatial distribution of the induced dipole stemmed from the incoming electric field at given frequency. For each adsorption peaks at λ, the main contribution to F(r, λ) derives from metal atom excitations, in consistent with the delocalized d orbitals over the metal atoms. Additionally, direct intra-carbon-chain excitations partially contribute to the absorption spectra in the visible range. Notably, the more edge numbers the ring is the more involved in the charge density response to light excitation (inset in Figure 5), in agreement with its highest peak of six-edge [M-CBiC]n rings at each specific wavelength. Coupling the wide-spectrum optical response of predicted [M-CBiC]n rings with highly sensitive hydrogen adsorption should lead to promising photoelectron catalytic MOFs materials. The elementary discharge step (Volmer reaction), which is involved in hydrogen evolution reaction (HER), is the bonding of hydrogen to the active site on the catalysis surface as process H+ + e– + * → H* (* denotes bonding site). This process is inherently correlated with the catalyst itself. Therefore, an energy picture of the H* bonding, including bonding structure and bonding energy, on the [M-CBiC]n rings is essential to understand the chemical reactivity that takes place during HER. The free energy of hydrogen bonding ∆GH*, which is an ideal descriptor of hydrogen evolution activity for a wide variety of metals and alloys,55-58 is used in this study. ∆GH*=∆EH*+∆EZPE-T∆SH* (2) The free energy of hydrogen bonding ∆GH* includes the changes on the total energies, ∆EH*, zero point energy vibrational correction, ∆EZPE, and thermal contribution T∆SH*, between the hydrogen bonded states and the gas phase. In details, the total energies of hydrogen bonding ∆EH* can be calculated as



∆EH*=Et([M-CBiC]n +H*)-Et([M-CBiC]n)-Et(H2)/2 (3) The calculated zero point energy of hydrogen bonded states is 0.27 eV in regardless of the linking transition metal atoms and edge numbers, so its difference to the gas phase, ∆EZPE, is 0.14 eV. The vibrational entropy of hydrogen bonded state is assumed to be small, so the change of entropy can be evaluated as ∆SH*=-1/2SH2, where SH2 is the entropy of H2 in the gas phase at the standard conditions.59 The overall correction to ∆GH* can be simplified as ∆GH*=∆EH*+0.34 eV (4) A hydrogen bonding free energy |∆GH*| ~ 0 is key to an efficient catalytic reaction with high HER activity. In Figure 6, we plotted the calculated free energy of hydrogen bonding ∆GH* of the [M-CBiC]n rings with variant linking atoms (M=Ti, V, and Cr, i=0) and edge numbers (n=2-6). The H atoms bound to the [M-CBiC]n rings all yield stable structures. We found that the [Ti-CBiC]n and [V-CBiC]n rings can more strongly binds the H atoms with a |∆GH*| less than 0.3 eV, as shown in Figure 6. Especially two-edge [Ti-CC]2 ring results in negative ∆GH* and its absolute value is almost close to zero

Figure. 6: The free energy diagram of hydrogen evolution at zero potential (U=0V) and zero pH for the [M-CBiC]n (M=Ti, V, and Cr, i=0, n=2-6) rings.

(~0.04 eV), which is highly competitive with commonly used noble metal Pt. Its low H* binding energy |∆GH*| can lead to a high activity for the HER of the proposed [Ti-CBiC]n or [VCBiC]n rings based porous MOFs structures. A less stable binding structure of H atoms to the [Cr-CBiC]n ring is formed with a |∆GH*| in the range of 0.34 ~ 0.70 eV. In our study the linking metal atoms play a leading role in bonding or activating the H atoms, in addition to stabilizing and modifying the ring itself. The metallic states associated with the metal sites are more under-saturated in the Ti or V-connected [M-CBiC]n rings compared to Cr-connected cases, therefore lead to more easily release of H in the next step of electrochemical hydrogen evolution in the [Ti/V-CBiC]n. The other difference in trapping H between the isolated [M-CBiC]n rings is their variant edge number. When the linking atom is fixed, the numbers of edges of each [M-CBiC]n rings show different bonding orders for attracting H atoms. For the even-edge numbered [TiCBiC]n rings, the H can bind to the metal site with higher stability at a lower |∆GH*| compared to the cases in the odd-edge numbered rings. In contrast, odd-edge numbered [V-CBiC]n and [Cr-CBiC]n rings can entrap H atoms with a lower energy than that obtained for the even-edge numbered rings. Generally speaking, the linking metal atoms play a more direct role on its catalytic performance for hydrogen evolution than the edge numbers. The reactivity of the [Ti-CBiC]n and [V-CBiC]n rings might be a better catalyst for the HER in the blue and green-

light adsorption region with a much reduced |∆GH*|. The H coverage studied in this work depends on the edge number. When the edge number increases from 2 to 6, the H coverage reduces from 50% to 17%. By increasing H coverage to 100% in [Ti-CC]2 MOFs unit, the calculated hydrogen adsorption energy for the second hydrogen atom only decreased by 0.02 eV, suggesting a stability for uploading more hydrogen atoms to the MOFs units. CONCLUSION In summary, we propose a basic building structure for MOFs architecture, [M-CBiC]n rings (M=Ti, V, and Cr, number of benzene ring in carbon chain i =0, 1, 2…, ring edge number n =2-6), and highlight some of their interesting structural, electronic, and photocatalytic properties and how they are influenced by their constituent elements and geometry. All the [M-CBiC]n species possess a ring-shape structure containing a metal cation M interacting with a carbon chain CBiC. This gives rise to diverse electronic structures, such as lowest energy magnetic configurations and electrical conductivity, for each building ring by changing linking metal atom and edge numbers. We also note that as the bonding d bands progressively filled from Ti to V to Cr, [M-CBiC]n building rings evolve from a nearly acute ring shape to a round ring shape, as the crystal field of metal center and surrounding carbon atoms change accordingly from trigonal pyramidal to quasi-squarepyramidal to square planar. The thermodynamical stability and binding strength is maximized at large sized [M-CBiC]n rings based on the calculated free energy of formation. The manipulation on the type of linking transition metals and the edge numbers of carbon chains produce broad visible light absorption spectra for [MCBiC]n rings in addition to a preferred hydrogen bonding free energy ∆GH*. [Ti-CBiC]n and [V-CBiC]n based MOFs structures exhibit promising photocatalytic performance for hydrogen evolution reactions (HERs) combining the merits of the predominant absorption spectrum at blue- and green-light adsorption region, respectively, and an ideal |∆GH*| within 0.3 eV. While the Cr atoms on the [Cr-CBiC]n rings which have less unsaturated states will be less trapping of the H atoms even at their intensive red-light adsorption region. In combination with an extremely large surface area, the proposed [MCBiC]n-based materials, when it extend to 1D/2D/3D nanoarchitectures, have very specific customization of their chemistry and promise as high-performance MOFs materials for flexible photo-electrochemical devices and photo-excited catalyst.

ASSOCIATED CONTENT Supporting Information Calculated magnetic states, phonon dispersion, density of states, band edges in relative to the vacuum level, and absorption spectra of [M-CBiC]n MOFs units. This material is available free of charge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION Corresponding Author [email protected], [email protected]

ACKNOWLEDGMENT


Page 6 of 8

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

Journal of the American Chemical Society The authors acknowledge financial support from the Australian Research Council (ARC) through a Future Fellowship grant (FT160100281), a Discovery Early Career Researcher Award (DECRA) grant (DE150100280), and a Discovery Project grant (DP160102627). T.L. also appreciates the generous grants of CPU time from the Australian National Computational Infrastructure Facility.

REFERENCES (1) Howarth, A. J.; Liu, Y. Y.; Li, P.; Li, Z. Y.; Wang, T. C.; Hupp, J. T.; Farha, O. K. Nature Rev. Mater. 2016, 1, 15018. (2) Wang, H. L.; Zhu, Q. L.; Zou, R. Q.; Xu, Q. Chem 2016, 2, 52. (3) Reed, D. A.; Keitz, B. K.; Oktawiec, J.; Mason, J. A.; Runčevski, T.; Xiao, D. J.; Darago, L. E.; Crocellà, V.; Bordiga, S.; Long, J. R. Nature 2017, 550, 96. (4) Talin, A. A.; Centrone, A.; Ford, A. C.; Foster, M. E.; Stavila, V.; Haney, P.; Kinney, R. A.; Szalai, V.; Gabaly, F.; Yoon, H. P.; Léonard, F.; Allendorf, M. D. Science 2014, 343, 66. (5) Shimizu, G. K. H.; Taylor, J. M.; Kim, S. Science 2013, 341, 354. (6) Nijem, N.; Wu, H. H.; Canepa, P.; Marti, A.; Balkus, Jr. K. J.; Thonhauser, T.; Li, J.; Chabal, Y. J. Am. Chem. Soc. 2012, 134, 15201. (7) Rodenas, T.; Luz, I.; Prieto, G.; Seoane, B.; Miro, H.; Corma, A.; Kapteijn, F.; Xamena, F.; Gascon, J. Nature Mater. 2015, 14, 48. (8) Wang, W.; Xu, X. M.; Zhou, W.; Shao, Z. P. Adv. Sci. 2017, 4, 1600371. (9) Tian, J.; Xu, Z. Y.; Zhang, D. W.; Wang, H.; Xie, S. H.; Xu, D. W.; Ren, Y. H.; Wang, H.; Liu, Y.; Li, Z. T. Nature Commun. 2016, 7, 11580. (10) Gallis, D. F. S.; Rohwer, L. E. S.; Rodriguez, M. A.; Nenoff, T. M. Chem. Mater. 2014, 26, 2943. (11) Kaur, R.; Kim, K. H.; Paul, A. K.; Deep, A. J. Mater. Chem. A 2016, 4, 3991. (12) Haider, G.; Usman, M.; Chen, T. P.; Perumal, P.; Lu, K. L.; Chen, Y. F. ACS Nano 2016, 10, 8366. (13) Morozan, A.; Jaouen, F. Energy Environ. Sci. 2012, 5, 9269. (14) Hod, I.; Deria, P.; Bury, W.; Mondloch, J. E.; Kung, C. W.; So, M.; Sampson, M. D.; Peters, A. W.; Kubiak, C. P.; Farha, O. K.; Hupp, J. T. Nature Commun. 2015, 6, 8304. (15) He, L. H.; Duan, F. H.; Song, Y. P.; Guo, C. P.; Zhao, H.; Tian, J. Y.; Zhang, Z. H.; Liu, C. S.; Zhang, X. J.; Wang, P. Y.; Du, M.; Fang, S. M. 2D Mater. 2017, 4, 025098. (16) Liu, J. L.; Zhu, D. D.; Guo, C. X.; Vasileff, A.; Qiao, S. Z. Adv. Energy Mater. 2017, 7, 1700518. (17) Mingabudinova, L. R.; Vinogradov, V. V.; Milichko, V. A.; HeyHawkins, E.; Vinogradov, A. V. Chem. Soc. Rev. 2016, 45, 5408. (18) Loera-Serna, S.; Zarate-Rubio, J.; Medina-Velazquez, D. Y.; Zhang, L.; Ortiz, E. Surf. Innov. 2016, 4, 76. (19) Mandel, K.; Granath, T.; Wehner, T.; Rey, M.; Stracke, W.; Vogel, N.; Sextl, G.; Müller-Buschbaum, K. ACS Nano 2017, 11, 779. (20) Nazarian, D.; Ganesh, P.; Sholl, D. S. J. Mater. Chem. A 2015, 3, 22432. (21) Dzubak, A. L.; Lin, L. C.; Kim, J. H.; Swisher, J. A.; Poloni, R.; Maximoff, S. N.; Smit, B.; Gagliardi, L. Nature Chem. 2012, 4, 810. (22) Liao, T.; Sun, Z. Q.; Sun, C. H.; Dou, S. X.; Searles, D. J. Sci. Rep. 2014, 4, 6256. (23) Dou, Y. H.; Liao, T.; Ma, Z. Q.; Tian, D. L.; Xiao, F.; Sun, Z. Q.; Kim, J. H.; Dou, S. X. Nano Energy 2016, 30, 267. (24) Liao, T.; Sun, C. H.; Du, A. J.; Sun, Z. Q.; Hulicova-Jurcakova, D.; Smith, S. J. Mater. Chem. 2012, 22, 13751. (25) Liao, T.; Sun, Z. Q.; Dou, S. X. ACS Appl. Mater. & Interfaces 2017, 9, 8255. (26) Zhao, J.; Deng, Q. M.; Bachmatiuk, A.; Sandeep, G.; Popov, A.; Eckert, J.; Rümmeli, M. H. Science 2014, 343, 1228.

(27) Krasheninnikov, A. V.; Lehtinen, P. O.; Foster, A. S.; Pyykkö, P.; Nieminen, R. M. Phys. Rev. Lett. 2009, 102, 126807. (28) Li, E. Y.; Marzari, N. ACS Nano 2011, 5, 9726. (29) Yang, C. K.; Zhao, J. J.; Lu, J. P. Phys. Rev. Lett. 2003, 90, 257203. (30) Liao, T.; Sun, Z. Q.; Kim, J. H.; Dou, S. X. Nano Energy 2017, 32, 209. (31) Zhang, X. Y.; Wang, J. L.; Gao, Y.; Zeng, X. C. ACS Nano 2009, 3, 537. (32) Giannozzi, P.; Baroni, S.; Bonini, N.; Calandra, M.; Car, R.; Cavazzoni, C.; Ceresoli, D.; Chiarotti, G. L.; Cococcioni, M.; Dabo, I.; Dal Corso, A.; Fabris. S.; Fratesi, G.; De Gironcoli, S.; Gebauer, R.; Gerstmann, U.; Gougoussis, C.; Kodalj, A.; Lazzeri, M.; Martin-Samos, L.; Marzari, N.; Mauri, F.; Mazzarello, R.; Paolini, S.; Pasquarello, A.; Paulatto, L.; Sbraccia, C.; Scndolo, S.; Sclauzero, G.; Seitsonen, A. P.; Smogunov, A.; Umari, P.; Wentzcovitch, R. M. J. Phys.: Condens. Matter 2009, 21, 395502. (33) Vanderbilt, D. Phys. Rev. B 1990, 41, 7892. (34) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett. 1996, 77, 3865. (35) Monkhorst, H. J.; Pack, J. D. Phys. Rev. B 1976, 13, 5188. (36) Grimme, S. J. Comput. Chem. 2006, 27, 1787. (37) London F. Z. Phys. Chem., Abt. B 1930, 11, 222. (38) Anisimov, V. I.; Aryasetiawan, F.; Lichtenstein, A. I. J. Phys.: Condens. Matter 1997, 9, 767. (39) Rocca, D.; Gebauer, R.; Saad, Y.; Baroni, S. J. Chem. Phys. 2008, 128, 154105. (40) Walker, B.; Gebauer, R. J. Chem. Phys. 2007, 127, 164106. (41) Walker, B.; Gebauer, R.; Saitta,A. M.; Baroni, S. Phys. Rev. Lett. 2006, 96, 113001. (42) Sanchez-Valencia, J. R.; Dienel, T.; Gröning, O.; Shorubalko, I.; Mueller, A.; Jansen, M.; Amsharov, K.; Ruffieux, P.; Fasel, R. Nature 2014, 512, 61. (43) Li, Y. F.; Zhou, Z.; Shen, P. W.; Chen, Z. F. ACS Nano 2009, 3, 1952. (44) Kan, E. J.; Li, Z. Y.; Yang, J. L.; Hou, J. G. J. Am. Chem. Soc. 2008, 130, 4224. (45) Hod, O.; Barone, V.; Peralta, J. E.; Scuseria, G. E. Nano Lett. 2007, 7, 2295. (46) Yang, X. B.; Ding, Y.; Ni, J. Phys. Rev. B 2008, 77, 041402(R). (47) Dumitrică, T.; Hua, M.; Yakobson, B. I. Phys. Rev. B 2004, 70, 241303(R). (48) Shultz, A. M.; Sarjeant, A. A.; Farha, O. K.; Hupp, J. T.; Nguyen, S. T. J. Am. Chem. Soc. 2011, 133, 13252. (49) Burnett, B. J.; Barron, P. M.; Hu, C. H.; Choe, W. Y. J. Am. Chem. Soc. 2011, 133, 9984. (50) Butler, K. T.; Hendon, C. H.; Walsh, A. J. Am. Chem. Soc. 2014, 136, 2703. (51) Ran, J. R.; Guo, W. W.; Wang, H. L.; Zhu, B. C.; Yu, J. G.; Qiao, S. Z. Adv. Mater. 2018, 30, 1800128. (52) Ran, J. R.; Gao, G. P.; Li, F. T.; Ma, T. Y.; Du, A. J.; Qiao, S. Z. Nat. Commun. 2017, 8, 13907. (53) Wei, Y. N.; Ma, Y. D.; Wei, W.; Li, M. M.; Huang, B. B.; Dai, Y. J. Phys. Chem. C 2018, 122, 8102. (54) Lopez, H. A.; Dhakshinammorthy, A.; Ferrer, B.; Atienzar, P.; Alvaro, M.; Garcia, H. J. Phys. Chem. C 2011, 115, 22200. (55) Nørskov, J. K.; Bligaard, T.; Logadottir, A.; Kitchin, J. R.; Chen, J. G.; Pandelov, S.; Stimming, U. J. Electrochem. Soc. 2005, 152, J23. (56) Greeley, J.; Jaramillo, T. F.; Bonde, J.; Chorkendorff, I.; Nørskov, J. K. Nature Mater. 2006, 5, 909. (57) Seo, B.; Joo, S. H. Nano Convergence 2017, 4, 19. (58) Zheng, Y.; Jiao, Y.; Zhu, Y. H.; Li, L. H.; Han, Y.; Chen, Y.; Du, A. J.; Jaroniec, M.; Qiao, S. Z. Nature Commun. 2014, 5, 3783. (59) Stull, D. R.; Prophet, H. JANAF thermochemical tables 1971, DTIC Document.



Page 8 of 8

Table of Contents


8

The Simplest MOF Units for Effective Photodriven Hydrogen Evolution

Recommend Documents