First principles investigations on the electronic structure of anchor groups on ZnO nanowires and surfaces A. Dominguez, M. Lorke, A. L. Schoenhalz, A. L. Rosa, Th. Frauenheim, A. R. Rocha, and G. M. Dalpian Citation: Journal of Applied Physics 115, 203720 (2014); doi: 10.1063/1.4879676 View online: http://dx.doi.org/10.1063/1.4879676 View Table of Contents: http://scitation.aip.org/content/aip/journal/jap/115/20?ver=pdfcov Published by the AIP Publishing Articles you may be interested in First-principles investigation of the electronic and Li-ion diffusion properties of LiFePO4 by sulfur surface modification J. Appl. Phys. 116, 063703 (2014); 10.1063/1.4892018 First-principles study of electronic structures and photocatalytic activity of low-Miller-index surfaces of ZnO J. Appl. Phys. 113, 034903 (2013); 10.1063/1.4775766 C-doped ZnO nanowires: Electronic structures, magnetic properties, and a possible spintronic device J. Chem. Phys. 134, 104706 (2011); 10.1063/1.3562375 Size effects on formation energies and electronic structures of oxygen and zinc vacancies in ZnO nanowires: A first-principles study J. Appl. Phys. 109, 044306 (2011); 10.1063/1.3549131 Magnetic coupling properties of Mn-doped ZnO nanowires: First-principles calculations J. Appl. Phys. 103, 073903 (2008); 10.1063/1.2903332 [This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP: 186.217.234.103 On: Fri, 10 Oct 2014 16:49:49 http://scitation.aip.org/content/aip/journal/jap?ver=pdfcov http://oasc12039.247realmedia.com/RealMedia/ads/click_lx.ads/www.aip.org/pt/adcenter/pdfcover_test/L-37/347967005/x01/AIP-PT/JAP_ArticleDL_1014/AIP-2293_Chaos_Call_for_EIC_1640x440.jpg/47344656396c504a5a37344142416b75?x http://scitation.aip.org/search?value1=A.+Dominguez&option1=author http://scitation.aip.org/search?value1=M.+Lorke&option1=author http://scitation.aip.org/search?value1=A.+L.+Schoenhalz&option1=author http://scitation.aip.org/search?value1=A.+L.+Rosa&option1=author http://scitation.aip.org/search?value1=Th.+Frauenheim&option1=author http://scitation.aip.org/search?value1=A.+R.+Rocha&option1=author http://scitation.aip.org/search?value1=G.+M.+Dalpian&option1=author http://scitation.aip.org/content/aip/journal/jap?ver=pdfcov http://dx.doi.org/10.1063/1.4879676 http://scitation.aip.org/content/aip/journal/jap/115/20?ver=pdfcov http://scitation.aip.org/content/aip?ver=pdfcov http://scitation.aip.org/content/aip/journal/jap/116/6/10.1063/1.4892018?ver=pdfcov http://scitation.aip.org/content/aip/journal/jap/116/6/10.1063/1.4892018?ver=pdfcov http://scitation.aip.org/content/aip/journal/jap/113/3/10.1063/1.4775766?ver=pdfcov http://scitation.aip.org/content/aip/journal/jcp/134/10/10.1063/1.3562375?ver=pdfcov http://scitation.aip.org/content/aip/journal/jap/109/4/10.1063/1.3549131?ver=pdfcov http://scitation.aip.org/content/aip/journal/jap/109/4/10.1063/1.3549131?ver=pdfcov http://scitation.aip.org/content/aip/journal/jap/103/7/10.1063/1.2903332?ver=pdfcov First principles investigations on the electronic structure of anchor groups on ZnO nanowires and surfaces A. Dominguez,1,a) M. Lorke,1,a) A. L. Schoenhalz,2,a) A. L. Rosa,1 Th. Frauenheim,1 A. R. Rocha,3 and G. M. Dalpian2 1BCCMS, Universit€at Bremen, Am Fallturm 1, 28359 Bremen, Germany 2CCNH, Universidade Federal do ABC, Av. dos Estados 5001, Santo Andr�e, Brazil 3IFT, Universidade Estadual Paulista, R. Dr. Bento Teobaldo Ferraz, 271, S~ao Paulo, Brazil (Received 3 March 2014; accepted 13 May 2014; published online 30 May 2014) We report on density functional theory investigations of the electronic properties of monofunctional ligands adsorbed on ZnO-(1010) surfaces and ZnO nanowires using semi-local and hybrid exchange-correlation functionals. We consider three anchor groups, namely thiol, amino, and carboxyl groups. Our results indicate that neither the carboxyl nor the amino group modify the transport and conductivity properties of ZnO. In contrast, the modification of the ZnO surface and nanostructure with thiol leads to insertion of molecular states in the band gap, thus suggesting that functionalization with this moiety may customize the optical properties of ZnO nanomaterials. VC 2014 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4879676] I. INTRODUCTION The combination of organic and inorganic materials for the fabrication of devices with novel properties has received considerable attention in technology. In particular, devices based on nanostructured materials are highly sensitive to adsorbed compounds due to their large surface-to-volume ratios. ZnO is an important candidate for the assembly of such devices, partly because of the possibility of fabrication of a large number of ZnO nanostructures.1 Adsorption of molecular layers can modify the properties of ZnO surfaces and provide a strategy for further immobilization of dyes and biomolecules. It has been reported that the electrical and optoelectronic performance of ZnO nanobelts are improved via functionalization with carboxylic acids.2 The enhance- ment in conductivity was attributed to the passivation of the nanobelt surface, which prevents that the oxygen in the sur- rounding environment reacts with point vacancies at the sur- face. Recently, it has been demonstrated that adsorbates can also assist charge separation and suppression of back recom- bination in nanostructured ZnO/P3HT interfaces.3 Several molecules have been investigated to determine the most appropriate anchor groups for the functionalization of ZnO. Compounds that has been employed as ligands include carboxylic acids,4–12 amines,13–15 silanes,4,16–18 phosphonic acids,19 and thiols.14,19–22 Aside from a required covalent binding between the anchor group and the substrate, the electronic structure of the adduct determines the suitabil- ity of the ligand for a desired function of the modified nano- structure. In ZnO nanoparticles, for instance, the presence of molecular states in the energy gap may affect the emission efficiency in ZnO-based ultraviolet (UV) lasers. However, this can be a desired feature for detection of water contami- nants in photocatalytic sensors.23 Fourier-transform infrared (FT-IR) and UV spectros- copy indicated that carboxylic acids are promising anchor groups for ZnO. However, it appears that the amount, adsorption position, and acidity of these ligands as well as the solution pH should also play a role in the binding proper- ties.24 On the other hand, recent atomic force microscopy (AFM) measurements have refuted the rightness of carbox- ylic acids for the modification of ZnO, showing that alkyl carboxylic acids etch the surface.25 Thiols have also been suggested to form strong bonds on ZnO.16,26 Alkanethiols were found to bind to the surface, forming highly uniform monolayers with some etching detected after long immersion times in an alkanethiol solution.25 However, earlier investi- gations reported non-uniform coverages or even physisorbed layers of these functional groups.24 Several chemisorbed27–31 and physisorbed molecules32–37 on ZnO have been theoretically investigated. The chemical stability of carboxyl (-COOH) groups on ZnO has been con- firmed by theoretical tight-binding27 and first-principles calcu- lations.28,29 A combined investigation using first-principles and FT-IR experiments concluded that molecules containing acetylacetone are as good anchors as carboxylic acid on ZnO (1010).28,30 In a previous work, we showed that -COOH, thiol (-SH), amino (-NH2),27,29,38 and phosphonic acid (PO(OH)2)39 groups chemisorb onto the ZnO (1010) surfaces with binding energies typically within 1–3 eV. In most theoretical works on ZnO functionalization thus far, the generalized-gradient approximation (GGA) within the Perdew-Burke-Ernzerhof (PBE) formalism40,41 has been employed for the study of the structural and electronic proper- ties of the modified systems. Whereas geometries are reliably given within this approach, electronic structures suffer from the well known band-gap problem of semi-local density func- tionals. In the case of ZnO, this error is striking as GGA yields a band gap of 0.7 eV, which highly underestimate the experimental value of 3.4 eV. One of the reasons for this underestimation is given by the wrong description of the a)A. Dominguez, M. Lorke, and A. L. Schoenhalz contributed equally to this work. 0021-8979/2014/115(20)/203720/9/$30.00 VC 2014 AIP Publishing LLC115, 203720-1 JOURNAL OF APPLIED PHYSICS 115, 203720 (2014) [This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP: 186.217.234.103 On: Fri, 10 Oct 2014 16:49:49 http://dx.doi.org/10.1063/1.4879676 http://dx.doi.org/10.1063/1.4879676 http://dx.doi.org/10.1063/1.4879676 http://crossmark.crossref.org/dialog/?doi=10.1063/1.4879676&domain=pdf&date_stamp=2014-05-30 location of Zn-3d states, which lie around 2 eV higher in energy compared to experiment.42 Accurate many-body tech- niques such as GW are found to correct the band gap and the position of Zn-3d states in bulk ZnO.43 However, its applic- ability to more complex systems such as surfaces and interfa- ces is still cumbersome. Hybrid density functional approaches, where a fraction of Hartree-Fock exchange is admixed, improves upon physical and chemical properties of solids and molecules. In general, the description of bond lengths, ionization energies, cohesive properties, and band gaps are in much better agreement with experiment when employing hybrid functionals.44 In this work, we use PBE0,45,46 to investigate the electronic properties of ZnO/organic interfaces. We employ substituted methane mol- ecules of the type CH3-X with X¼ -COOH, -SH, and -NH2 as prototype adsorbates. In Sec. II, the computational details of our calculations are described. In Sec. III A, we report on the electronic structure of the ZnO (1010) surface modified with the aforementioned moieties. The nonpolar (1010) is the most stable surface of ZnO and is the majority facet in many grown ZnO nanostructures.1 Afterwards, in Sect. III B we address the adsorption of the anchor groups on the nonpolar facets of an ultrathin ZnO nanowire. The analysis of the electronic structure was done using the averaged electrostatic potential between the bulk and the surfaces, and the vacuum level between the surfaces and the wires. We show that neither the valence band maximum (VBM) nor the conduction band min- imum (CBM) of the bare surface shift significantly compared to the bulk value. This allowed us to use the vacuum level of the bare surface as a reference to align the energy levels of the surface with those of the ligands. We find that the molecu- lar state follows the band gap opening for nanostructures, instead of being pinned to a specific energetic range. II. COMPUTATIONAL DETAILS Periodic density functional theory (DFT) calculations were performed as implemented in the Vienna ab-initio Simulation Package (VASP).47–50 Plane wave basis func- tions with an energy cutoff of 400 eV and 300 eV have been employed for the surface and nanowire systems, respec- tively. The projector-augmented wave (PAW) method51,52 has been used throughout. The geometries were optimized at the PBE level. The electronic structures for the optimized geometries were calculated within the hybrid PBE0 formal- ism. For the Brillouin zone integration, we employed Monkhorst-Pack (MP)53,54 meshes of (4� 4� 4) for bulk, (1� 4� 4) for the surfaces, and (1� 1� 4) for the nanowires (NWs). For the calculation of the density of states (DOS), MP grids of (1� 10� 10) for the surfaces and (1� 1� 10) for the NWs were used. All atomic positions have been relaxed until the interatomic forces were smaller than 0.01 eV/Å. III. RESULTS A. Surfaces The use of PBE0 substantially improves upon the band gap of bulk ZnO, yielding a value of 3.2 eV, which is much more in line with the experimental value.55 To uniquely compare our PBE and PBE0 results for bulk ZnO, we employ the method proposed in Ref. 56, where the band edges of the solid are aligned using the electrostatic potential averaged along the y-z plane, resulting from the PBE and PBE0 calculations. We find that for bulk ZnO the VBM as obtained with PBE0 is shifted downwards by 1.6 eV with respect to the PBE value, in very good agreement with the results reported in Ref. 57. The Zn 3d states have a binding energy of 7.5 eV. The CBM, mainly composed of Zn-4s states, shifts upwards by 0.7 eV within PBE0, compared to the PBE results. To model the bare surface, we consider a tetragonal supercell consisting of a wurzite ZnO slab containing 8 Zn- O atomic layers and a vacuum region of approximately 17 Å along the [1010] direction. For the modified surfaces, the same supercell dimensions were employed. The correspond- ing optimized structures are depicted in Fig. 1. Two mole- cules per supercell (one at each side of the slab) with equivalent configurations were considered. This leads to one ligand per surface unit cell. As we have shown in previous studies,27,29,58 this surface coverage is the thermodynami- cally most stable one. Our results show that both -COOH and -SH groups adsorb on the surface in a dissociative manner, while -NH2 does not dissociate (see Fig. 1). This fact plays an important role in the resulting electronic structure as will be discussed in what follows. The adsorption energies, which provide in- formation about how the surface and the molecules interact, are calculated as Eads ¼ 1 n ðET � EBare � EmolÞ, where ET is the total energy of the hybrid system (surfaceþmolecule), EBare is the total energy of bare surface, n is the number of molecules adsorbed on the surface, and Emol is the energy of a neutral (isolated) molecule in gas phase. Using the PBE0 functional, the obtained adsorption energies are �1.35 eV, �0.76 eV, and �0.90 eV, for -COOH, -NH2, and -SH cases, respectively. These energies compare well to previously obtained values at the PBE level (�1.39 eV, �0.88 eV, and �1.03 eV for -COOH, -NH2, and -SH, respectively.29) The distributions of the optimized atomic bond distances as obtained with either PBE or PBE0 are shown in Fig. 2. We can conclude that there are no significant changes between PBE0 and PBE results, thus demonstrating that the FIG. 1. Optimized structures of the ZnO (1010) surface covered with one monolayer of (a) CH3-COOH, (b) CH3-NH2, and (c) CH3-SH molecules. The spheres represent the following atoms with the corresponding color within parenthesis: oxygen (red), zinc (gray), carbon (brown), nitrogen (light blue), sulfur (yellow), and hydrogen (light pink). 203720-2 Dominguez et al. J. Appl. Phys. 115, 203720 (2014) [This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP: 186.217.234.103 On: Fri, 10 Oct 2014 16:49:49 latter functional provides an accurate structural description for these systems. In the center of the slab, all systems fea- ture Zn-O interatomic distances similar to those found for bulk ZnO (around 2.0 Å). The main difference in bond lengths can be observed at the surface, where the bond dis- tances are longer than for bulk for the -COOH and -SH cases, whereas they are shorter for the bare and - NH2-modified surfaces. The latter two cases are the ones with the presence of dangling bonds at the surface (dangling bonds at the oxygen binding sites are passivated for -COOH and -SH due to dissociation of the ligands). The alignment of the energy levels shows that the VBM of the bare surface does not shift significantly (see Appendix) compared to the bulk value. Subsequently, the band align- ment has been done using the vacuum level of the bare sur- face and the adsorbed systems as in Refs. 59–62. We have ensured that the vacuum region of the supercell is large enough, so that the electrostatic potential reaches its vacuum value and is flattened out. Fig. 3 shows the total and projected DOS for the bare surface and surfaces with adsorbed molecules using the band alignment described above. The DOS are aligned in such way that the zero on the x-axis corresponds to the VBM for the bare surface. The band structures for the bare and modi- fied surfaces are presented in Fig. 4. The band alignment corresponds to that employed for the DOS. The calculated band gap of the bare surface in this work is �3.15 eV. Here, quantum confinement effects due to reduction of the dimen- sionality are counteracted by the presence of surface levels FIG. 2. Bond lengths for (a) the bare and modified ZnO surfaces using (b) CH3-COOH, (c) CH3-NH2, and (d) CH3-SH ligands obtained at the PBE and PBE0 levels of theory. FIG. 3. Projected DOS for (a) the bare and modified surfaces with (b)- COOH, (c)-SH, and (d)-NH2 groups. The black and green lines represent, respectively, the total DOS and its projection onto the ligand states. The dashed line denotes the Fermi energy. 203720-3 Dominguez et al. J. Appl. Phys. 115, 203720 (2014) [This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP: 186.217.234.103 On: Fri, 10 Oct 2014 16:49:49 (dangling bonds) resulting in a band gap only slightly larger than that of bulk ZnO (�3.09 eV). From Figs. 3 and 4, we can observe that only in the case of modification with -SH, molecular levels are inserted in the band gap. In contrast, - COOH molecular levels contribute to the valence band edge due to hybridization of states of the carboxyl oxygen and the substrate. For the -NH2 case, occupied molecular levels are located deeper in the valence band. B. Nanowires We now investigate the effects of the change in dimen- sionality on the electronic properties of organic-ZnO interfa- ces by considering the same adsorbates on thin ZnO NWs. The NWs have been modeled using infinitely long ZnO wires with hexagonal cross sections, grown along the [0001] direction. The bare NW exhibits nonpolar (1010) and (1210) facets as shown in Fig. 5. We consider a supercell containing 48 atoms63 and a sufficiently large vacuum region in order to avoid spurious interactions with the supercell images. For the modified NWs a monolayer (ML) of ligands (one ligand per surface Zn-O pair) was considered. This amounts to 12 molecules per supercell. We have shown in a previous inves- tigation that a monolayer coverage is energetically favored over lower coverages for the modification of ZnO surfaces.29 We chose alternate orientation of the molecules in such a way that on the (1010) facet the ligands point to the same direction whereas the orientation is exchanged in neighbor- ing (1010) facets (see Fig. 5). We first investigate the bare NW. To test the reliability of the PBE functional for the description of the investigated structure, we performed a geometry optimization calculation using the PBE0 functional. The optimized geometry at the PBE0 level of theory is shown in Fig. 5. Both O and Zn atoms of the surface layer relax towards the bulk region. However, the relaxation for Zn is larger than for O atoms, thus leading to formation of a buckled surface Zn-O dimer. This relaxation shrinks the surface Zn-O bond length from the bulk value (2.01 Å) to 1.88 Å. The distance between near- est neighbor Zn atoms along the wire growth direction varies from 3.2 Å in the bulk region to 3.0 Å close to the surface. The tilt angles of the surface Zn-O dimer are x1¼ 10.48 and x2¼ 3.78, respectively, according to the definitions used in Ref. 63. The good agreement between these results and ear- lier PBE findings,63 together with the experience gained from the surface calculations, give us confidence to select the PBE functional for the structural relaxation of the modi- fied NWs. In Figs. 6(a) and 9(a), the projected DOS and the band structures along the C-A direction are shown for the bare NW. The energy levels of the different structures have been aligned using the electrostatic potential of the vacuum region as in Refs. 59–62, setting the VBM of the bare surface as zero of the energy scale in all cases. The system has a direct band gap of 3.7 eV at the C-point. The increase of the band gap compared to the bulk case is due to quantum confinement effects on the band edges. As expected from a simple particle-in-a-box-like pic- ture, the CBM is more affected by quantum confinement than the VBM, due to the difference in effective masses of electrons at the CBM and holes at the VBM. From the pro- jected DOS, we can infer that the VBM is composed of OA-p states, shifted 0.1 eV from the bulk-like OC-p states (data not shown). In contrast, the CBM is made of ZnC-s states belonging to the bulk region of the NW. Compared to bulk ZnO, Zn-d and O-p states have a stronger overlap. The effec- tive masses for the highest valence band and lowest conduc- tion band are �0.048 me and 0.008 me, respectively, where me is the electron mass. Next, we discuss on the modified ZnO NWs. The opti- mized structure for the -COOH-covered wire is shown in Fig. 7. The -COOH group adsorbs via two asymmetric bonds between the carboxyl oxygens, O1 and O2, and the surface Zn atoms. This is accompanied by a proton transfer from the anchor group to the surface OA atom. The intra-molecular angle O1-C-O2 of the moiety is 1208. After adsorption of the ligand, the surface Zn-O bond length is nearly restored to its bulk value (2.05 Å). Similarly, the distance between nearest neighbors Zn atoms along the wire axis is reduced to 3.1 Å. FIG. 4. Band structure for the bare sur- face (a) and the surface with adsorbed molecules: -COOH (b), -SH (c), and - NH2 (d). The picture on the right side represents the path sampling in the Brillouin zone. FIG. 5. Optimized structure of the bare NW: (a) top view and (b) side view. 203720-4 Dominguez et al. J. Appl. Phys. 115, 203720 (2014) [This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP: 186.217.234.103 On: Fri, 10 Oct 2014 16:49:49 Our findings agree with Fourier transform infrared attenuated total reflectance measurements on ZnO nano tips functional- ized with carboxylic acid, which show a loss of the carbonyl bond (C¼O). This is in line with a bidentate attachment of the -COOH group on the surface as found in our calculations. The DOS and electronic structure for the -COOH-ZnO system are shown in Figs. 6(b) and 9(b), respectively. We find strong changes in the band structure around the VBM. Specifically, the surface OA p-states hybridize with O1 and O2 p-states of the functional group, resulting in localized states at �0.5 eV. The electronic structure reveals that the band gap becomes indirect, as the localized states have a slight dispersion to higher energies at the A-point. This resembles the behavior for the adsorption on the surface, where also an indirect gap is found (see Fig. 4). In general terms, the electronic structures for the -COOH-modified sur- face and NW are similar. As for the bare NW, the p-states of the OA atoms have a strong hybridization with Zn-d states in the valence band. From the comparison of the DOS projected onto OA and OB atoms we can infer that the band bending is significantly reduced to about 0.05 eV, leading to almost flat-band conditions. C-p states are mainly located deep in the VB and overlap strongly with the Zn-d states. The CBM remains composed of Zn s-states of the bulk region. The band gap of the modified wire increases to 4.1 eV. Such modifications of the band gap with respect to the bare nano- structure have been observed for dye-functionalized TiO2 (Ref. 60) and Si NWs (Ref. 64). They appear due to the inter- action between the ligand LUMO and the CBM of the sub- strate for the former and between the ligand HOMO and the VBM for the latter. Previous GGAþU investigations have already indicated that ZnO (1010) surfaces modified with acetic acid do not feature band gap states. However, for other carboxylic acids, such as benzoic acid, intragap states are observed. The states of the -COOH group are rather localized and should not change the electron mobility in ZnO wires. On the contrary, the hybridization at the VBM even reduces the curvature at the C-point, enhancing the hole mobility. Our results confirm the experimental evidences of Ref. 65, where the saturation electron mobility in ZnO thin film FETs functionalized with stearic acid (CH3(CH2)16COOH) was enhanced by about one order of magnitude. We found that the band bending and curvature of the -COOH-ZnO wire at the C-point are reduced compared to the bare NW. We now turn to the modification of the NW with -SH. The optimized structure is shown in Fig. 8. The most striking difference from the previous cases is the asymmetry between the different adsorbates. Namely, the moieties containing the C1 and C3 atoms relax towards each other, while the ones with C2 and C4 atoms repel each other. The -SH groups attach to the NW via monodentate S-Zn bonds with a bond length of 2.23 Å for S2 and 2.26 Å for S1. The Zn-S-C angle is 1068 for the ligands with C1 and C3 and 1088 for the ligands with C2 and C4. Again, the hydrogen atom of the -SH group is transferred to the surface OA atoms. The NW geometry relaxes significantly with a Zn-O bond length of 2.17 Å. AFM, FT-IR, and X-ray photoelectron spectroscopy (XPS) measurements have indicated a strong S-Zn covalent bond in thiol-functionalized ZnO (1010) surfaces. Moreover, thiol adsorption has also been found experimentally on polar (0001) ZnO surfaces22 and confirmed via XPS and Raman spectroscopy on ZnO NWs.22 The DOS and electronic structure for the -SH-ZnO NW are shown in Figs. 6(c) and 9(c), respectively. Compared to FIG. 6. Projected DOS for (a) the bare and modified NWs with (b)-COOH, (c)-SH, and (d)-NH2 groups. The dashed line denotes the Fermi energy. 203720-5 Dominguez et al. J. Appl. Phys. 115, 203720 (2014) [This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP: 186.217.234.103 On: Fri, 10 Oct 2014 16:49:49 the modification with -COOH, the electronic structure under- goes more significant changes. Two molecular states per ad- sorbate appear in the gap region. These states can be grouped into two subsets. The first one, located energetically at �0.3 eV and 0.7 eV has contributions mainly from the p-states of S1 and C1, while the second group at 0.5 eV and 1.0 eV stems from the p-states of S2 and C2. This distinction can be traced back to geometry differences between the two non-equivalent ligands configurations with indexes 1 and 2. The Zn-S1 and Zn-S2 bond lengths are different. Also, the distance between C3 and C1 is smaller than both C1-C2 and C2-C4 distances, thus changing the strength of the inter-molecular interaction for the two configurations. The band structure in Fig. 9(c) reveals that the intra-gap states are substantially more localized compared to the surface case (see Fig. 4). Additionally, we find localized states stemming from the carbon atoms around �4 eV. In contrast to -COOH, the molecular states of the thiol group do not hybridize with the VBM states. In general, it can be stated that the electronic properties of the SH-modified NW are very similar to those for the surface case. Obviously, nondegenerate molecular states due to the two different ligand configurations on the wire are not observed in the surface case, where all adsorbate geometries are equivalent by translational symmetry. Finally, we investigate the adsorption of the -NH2 group on the ZnO NW. The corresponding optimized structure is shown in Fig. 10. Similar to the case of -SH group, for the -NH2-ZnO system we have two nonequivalent ligand dimer configurations. Adsorbates with carbon atoms C1 and C3 relax towards each other while those with C2 and C4 atoms bend away from each other. The ligands bind to the NW via a monodentate mode with Zn-N1 and Zn-N2 bond lengths of 2.15 Å and 2.23 Å, respectively. The Zn-N-C angle is 1228 for the ligands with indices 1 and 3 and 1368 for those with indexes 2 and 4. Unlike the functional groups investigated so far, -NH2 does not undergo dissociative adsorption on the wire. A molecular adsorption has been shown to be favorable over ligand dissociation on ZnO nonpolar surfaces in Ref. 29. As an outcome of this binding mode, Zn-O bond lengths relax to about 1.97 Å at the NW surface. The chemisorption FIG. 7. Optimized structure for the -COOH-modified NW: (a) top view and (b) side view. FIG. 8. Optimized structure of the SH-modified NW: (a) top view and (b) side view. FIG. 9. Band structure along the C-A direction for the bare NW (top left) and the NWs modified with -COOH (bottom left), -NH2 (top right) and -SH (bottom right) groups. The dashed line denotes the Fermi energy. 203720-6 Dominguez et al. J. Appl. Phys. 115, 203720 (2014) [This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP: 186.217.234.103 On: Fri, 10 Oct 2014 16:49:49 of amines on ZnO films has been confirmed by XPS66 and AFM measurements. The DOS and band structure for the -NH2-ZnO system are shown in Figs. 6(d) and 9(d), respectively. Our calcula- tions show that no intra-gap states exist for this system. The VBM is mainly composed of OA and OB p-states. The DOS shows a N-p component along the entire VB whereas the pro- jection onto the C-p states indicates localization in the same energy range as for the Zn-d states (data not shown). In con- trast to the case of -SH, we find only insignificant differences in the DOS projected on the two nonequivalent groups of adsorbates (see Fig. 6). The most striking feature of the band structure is the large overall shift to higher energies. This behavior is also seen for the -NH2-modified surface (Fig. 4). This could be explained by noticing that the surface oxygen atoms are non-saturated and the required energy to remove an electron from the dangling bond is therefore reduced com- pared to the cases where the molecule binds in a dissociative manner. The main feature of the -NH2-ZnO system is hence a surface passivation that leads to almost perfect flat band con- ditions in conjunction with a reduced work function. IV. CONCLUSIONS We have investigated the electronic properties of ZnO surfaces and NWs modified by several functional groups. We found that the behavior of the investigated moieties is very similar on both surfaces and NWs. Our results suggest that functionalization with carboxylic acids or amines does not alter the transport and conductivity properties of ZnO nanostruc- tures due to the presence of almost flat band conditions. In con- trast, the functionalization with thiols might offer a route for modification of optical properties of ZnO nanomaterials due to the appearance of molecular states in the energy gap. Our results provide new insights to improve the physical properties of oxide nanostructures and surfaces for device applications. ACKNOWLEDGMENTS We acknowledge financial support from the Deutsche Forschungsgemeinschaft under the program FOR1616, the University of Bremen, and the CAPES/DAAD/PROBRAL international partnership program. We also thank computational resources from HLRN (Hannover/Berlin- Germany) and CENAPAD (S~ao Paulo-Brazil). APPENDIX: ALIGNMENTS The alignment between ZnO bulk and surface is pre- sented in Fig. 11. Fig. 12 shows the band alignment for the surface with different functional groups. We can conclude there is again a different trend for the non-dissociated case (-NH2): the energy difference from vacuum level of bare sur- face and with adsorbed molecule, defined as the D parameter, is positive (�0.3 eV). For the surface with dissociated adsorbed molecules (-COOH and -SH cases), the D has nega- tive values around �1.3 eV. In case of nano wires, the alignment of the band edges for adsorbed -COOH is shown in Fig. 13. As discussed before, the band edge alignment has been performed with respect to the electrostatic potential in the vacuum region. We find that the surface modification in NWs shifts the CBM upwards in energy by 0.4 eV, while the VBM remains almost unchanged and only the band bending is reduced. This means the beneficial properties of ZnO are preserved, while the surface is stabilized by the functional group. Compared to the isolated molecule, we noted that the surface modification shifts the HOMO of the CH3-COOH molecule about 0.7 eV upwards and the LUMO upwards by 0.3 eV. In Fig. 14, the band edge alignment for SH modification of the NW is shown. Now the HOMO of the isolated mole- cule lies above the VBM of the bare wire. This already indi- cates a qualitative difference to other functional groups that FIG. 10. Optimized structure of the NH2-modified NW: (a) top view and (b) side view. FIG. 11. Band alignment scheme for the studied hybrid interfaces using the vacuum level of the bare surface as the common level. FIG. 12. Band alignment scheme for the studied hybrid interfaces using the vacuum level of the bare surface as the common level. 203720-7 Dominguez et al. J. Appl. Phys. 115, 203720 (2014) [This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP: 186.217.234.103 On: Fri, 10 Oct 2014 16:49:49 is also reflected in the resulting band structure. In the modi- fied NW, the HOMO of the molecule is shifted up by 0.5 eV while the VBM is shifted downwards by 0.4 eV with respect to the bare NW. The CBM is shifted up by 0.3 eV. In Fig. 15, the band edges alignment in shown for the CH3-NH2 modified NW. We find a pronounced upward shift of all states. The CBM and VBM shift upwards by 2.7 eV and 3.0 eV, respectively, resulting in a band gap of 4 eV. The HOMO states of the molecular groups also shift upwards by 2.9 eV. The reason for this different behavior is probably due to the stronger hybridization of the N-p states with all VB states. 1Z. L. Wang, ZnO Bulk, Thin Films and Nanostructures (Elsevier, Oxford, 2006). 2C. Lao, Y. Li, C. P. Wong, and Z. L. Wang, Nano Lett. 7, 1323 (2007). 3Y.-Y. Lin, Y.-Y. Lee, L. Chang, J.-J. Wu, and C.-W. Chen, Appl. Phys. Lett. 94, 063308 (2009). 4O. Taratula, E. Galoppini, D. Wang, D. Chu, Z. Zhang, H. Chen, G. Saraf, and Y. Lu, J. Phys. Chem. B 110, 6506 (2006). 5S. Kuehn, S. Friede, S. Sadofev, S. Blumstengel, F. Henneberger, and T. Elsaesser, Appl. Phys. Lett. 103, 191909 (2013). 6Y. Cao, E. Galoppini, P. I. Reyes, and Y. Lu, Langmuir 29, 7768 (2013). 7K. Yao, L. Chen, Y. Chen, F. Li, and P. Wang, J. Phys. Chem. C 116, 3486 (2012). 8S. K. Hau, Y.-J. Cheng, H.-L. Yip, Y. Zhang, H. Ma, and A. K.-Y. Jen, ACS Appl. Mater. Interfaces 2, 1892 (2010). 9X. Tian, J. Xu, and W. Xie, J. Phys. Chem. C 114, 3973 (2010). 10A. Lenz, L. Selegård, F. S€oderlind, A. Larsson, P. O. Holtz, K. Uvdal, L. Ojam€ae, and P.-O. K€all, J. Phys. Chem. C 113, 17332 (2009). 11O. Taratula, E. Galoppini, R. Mendelsohn, P. I. Reyes, Z. Zhang, Z. Duan, J. Zhong, and Y. Lu, Langmuir 25, 2107 (2009). 12A. Gupta, B. C. Kim, E. Edwards, C. Brantley, and P. Ruffin, Adv. Mater. Res. 567, 228 (2012). 13Y. K. Gao, F. Traeger, O. Shekhah, H. Idriss, and C. W€oll, J. Colloid Interface Sci. 338, 16 (2009). 14G. Jayalakshmi, K. Saravanan, and T. Balasubramanian, J. Lumin. 140, 21 (2013). 15G. Jayalakshmi and T. Balasubramanian, J. Mater. Sci.-Mater. Electron. 24, 2928 (2013). 16K. Ogata, K. Koike, S. Sasa, M. Inoue, and M. Yano, J. Vac. Sci. Technol. B 27, 1834 (2009). 17C. G. Allen, D. J. Baker, J. M. Albin, H. E. Oertli, D. T. Gillaspie, D. C. Olson, T. E. Furtak, and R. T. Collins, Langmuir 24, 13393 (2008). 18J. A. Soares, J. E. Whitten, D. W. Oblas, and D. M. Steeves, Langmuir 24, 371 (2008). 19C. L. Perkins, J. Phys. Chem. C 113, 18276 (2009). 20K. Ogata, K. Koike, S. Sasa, M. Inoue, and M. Yano, Appl. Surf. Sci. 241, 146 (2005). 21P. W. Sadik, S. J. Pearton, D. Norton, E. Lambers, and F. Ren, J. Appl. Phys. 101, 104514 (2007). 22J. Singh, J. Im, E. J. Watters, J. E. Whitten, J. W. Soares, and D. M. Steeves, Surf. Sci. 609, 183 (2013). 23C. Hariharan, Appl. Catal. A: Gen. 304, 55 (2006). 24E. Galoppini, J. Rochford, H. Chen, G. Saraf, Y. Lu, A. Hagfeldt, and G. Boschloo, J. Phys. Chem. B 110, 16159 (2006). 25J. Chen, R. E. Ruther, Y. Tan, L. M. Bishop, and R. J. Hamers, Langmuir 28, 10437 (2012). 26H. Yip, S. K. Hau, N. S. Baek, H. Ma, and A. K. Y. Jen, Adv. Mater. 20, 2376 (2008). 27N. H. Moreira, A. L. Rosa, and T. Frauenheim, Appl. Phys. Lett. 94, 193109 (2009). 28T. L. Bahers, T. Pauport�e, F. Labat, and I. Ciofini, Langmuir 27, 3442 (2011). 29N. H. Moreira, A. Dominguez, T. Frauenheim, and A. L. da Rosa, Phys. Chem. Chem. Phys. 14, 15445 (2012). 30T. L. Bahers, T. Pauport�e, F. Odobel, F. Labat, and I. Ciofini, Int. J. Quantum Chemistry 112, 2062 (2012). 31A. Calzolari, A. Ruini, and A. Catellani, J. Am. Chem. Soc. 133, 5893 (2011). 32S. Blumstengel, H. Glowatzki, S. Sadofev, N. Koch, S. Kowarik, J. Rabe, and F. Henneberger, Phys. Chem. Chem. Phys. 12, 11642 (2010). 33A. L. Briseno, T. W. Holcombe, A. I. Boukai, E. C. Garnett, S. W. Shelton, J. J. M. Fr�echet, and P. Yang, Nano Lett. 10, 334 (2010). 34A. Neubauer, J. M. Szarko, A. F. Bartelt, R. Eichberger, and T. Hannappel, J. Phys. Chem. C 115, 5683 (2011). 35A. Amat and F. D. Angelis, Phys. Chem. Chem. Phys. 14, 10662 (2012). 36C. Caddeo, G. Malloci, G.-M. Rignanese, L. Colombo, and A. Mattoni, J. Phys. Chem. C 116, 8174 (2012). 37P. Ruankham, L. Macaraig, T. Sagawa, H. Nazakumi, and S. Yoshikawa, J. Phys. Chem. C 115, 23809 (2011). 38A. Dominguez, N. H. Moreira, G. Dolgonos, A. L. Rosa, and T. Frauenheim, J. Phys. Chem. C 115, 6491 (2011). 39X. Q. Shi, H. Xu, M. V. Hove, N. Moreira, A. Rosa, and T. Frauenheim, Surf. Sci. 606, 289 (2012). 40J. Perdew, K. Burke, and M. Ernzerhof, Phys. Rev. Lett. 77, 3865 (1996). 41J. P. Perdew, K. Burke, and M. Ernzerhof, Phys. Rev. Lett. 78, 1396 (1997). FIG. 14. Band edge alignments for SH-CH3 modified ZnO NWs. FIG. 15. Band edge alignments for CH3-NH2 modified ZnO NWs. FIG. 13. Band edges alignment for COOH-CH3 modified ZnO NWs. 203720-8 Dominguez et al. J. Appl. Phys. 115, 203720 (2014) [This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP: 186.217.234.103 On: Fri, 10 Oct 2014 16:49:49 http://dx.doi.org/10.1021/nl070359m http://dx.doi.org/10.1063/1.3080203 http://dx.doi.org/10.1063/1.3080203 http://dx.doi.org/10.1021/jp0570317 http://dx.doi.org/10.1063/1.4829466 http://dx.doi.org/10.1021/la4006949 http://dx.doi.org/10.1021/jp207686c http://dx.doi.org/10.1021/am100238e http://dx.doi.org/10.1021/jp908517j http://dx.doi.org/10.1021/jp905481v http://dx.doi.org/10.1021/la8026946 http://dx.doi.org/10.4028/www.scientific.net/AMR.567.228 http://dx.doi.org/10.4028/www.scientific.net/AMR.567.228 http://dx.doi.org/10.1016/j.jcis.2009.06.008 http://dx.doi.org/10.1016/j.jcis.2009.06.008 http://dx.doi.org/10.1016/j.jlumin.2013.02.057 http://dx.doi.org/10.1007/s10854-013-1193-3 http://dx.doi.org/10.1116/1.3155824 http://dx.doi.org/10.1116/1.3155824 http://dx.doi.org/10.1021/la802621n http://dx.doi.org/10.1021/la702834w http://dx.doi.org/10.1021/jp906013r http://dx.doi.org/10.1016/j.apsusc.2004.09.032 http://dx.doi.org/10.1063/1.2736893 http://dx.doi.org/10.1063/1.2736893 http://dx.doi.org/10.1016/j.susc.2012.12.006 http://dx.doi.org/10.1016/j.apcata.2006.02.020 http://dx.doi.org/10.1021/jp062865q http://dx.doi.org/10.1021/la301347t http://dx.doi.org/10.1002/adma.200703050 http://dx.doi.org/10.1063/1.3132055 http://dx.doi.org/10.1021/la103634v http://dx.doi.org/10.1039/c2cp42435e http://dx.doi.org/10.1039/c2cp42435e http://dx.doi.org/10.1002/qua.23134 http://dx.doi.org/10.1002/qua.23134 http://dx.doi.org/10.1021/ja1101008 http://dx.doi.org/10.1039/c004944c http://dx.doi.org/10.1021/nl9036752 http://dx.doi.org/10.1021/jp109615b http://dx.doi.org/10.1039/c2cp41616f http://dx.doi.org/10.1021/jp212283z http://dx.doi.org/10.1021/jp204325y http://dx.doi.org/10.1021/jp107576g http://dx.doi.org/10.1016/j.susc.2011.10.006 http://dx.doi.org/10.1103/PhysRevLett.77.3865 http://dx.doi.org/10.1103/PhysRevLett.78.1396 42L. Ley, R. A. Pollak, F. R. McFeely, S. P. Kowalczyk, and D. A. Shirley, Phys. Rev. B 9, 600 (1974). 43M. Shishkin and G. Kresse, Phys. Rev. B 75, 235102 (2007). 44J. Paier, M. Marsman, K. Hummer, G. Kresse, I. C. Gerber, and J. G. �Angy�an, J. Chem. Phys. 124, 154709 (2006). 45C. Adamo and V. Barone, J. Chem. Phys. 110, 6158 (1999). 46M. Ernzerhof and G. E. Scuseria, J. Chem. Phys. 110, 5029 (1999). 47G. Kresse and J. Hafner, Phys. Rev. B 47, 558 (1993). 48G. Kresse and J. Hafner, Phys. Rev. B 49, 14251 (1994). 49G. Kresse and J. Furthm€uller, Comput. Mater. Sci. 6, 15 (1996). 50G. Kresse and J. Furthm€uller, Phys. Rev. B 54, 11169 (1996). 51P. E. Bl€ochl, Phys. Rev. B 50, 17953 (1994). 52G. Kresse and D. Joubert, Phys. Rev. B 59, 1758 (1999). 53H. J. Monkhorst and J. D. Pack, Phys. Rev. B 13, 5188 (1976). 54J. D. Pack and H. J. Monkhorst, Phys. Rev. B 16, 1748 (1977). 55Semiconductors—Basic Data, 2nd ed., edited by O. Madelung (Springer, 1996). 56A. Carvalho, A. Alkauskas, A. Pasquarello, A. K. Tagantsev, and N. Setter, Phys. Rev. B 80, 195205 (2009). 57J. Wr�obel, K. J. Kurzydłowski, K. Hummer, G. Kresse, and J. Piechota, Phys. Rev. B 80, 155124 (2009). 58A. Dominguez, S. grosse Holthaus, S. Koppen, T. Frauenheim, and A. L. da Rosa, Phys. Chem. Chem. Phys. 16, 8509 (2014). 59M. C. Toroker, D. K. Kanan, N. Alidoust, L. Y. Isseroff, P. Liao, and E. A. Carter, Phys. Chem. Chem. Phys. 13, 16644 (2011). 60P. Umari, L. Giacomazzi, F. D. Angelis, M. Pastore, and S. Baroni, J. Chem. Phys. 139, 014709 (2013). 61K. Noori and F. Giustino, Adv. Funct. Mater. 22, 5089 (2012). 62J. Kang, S. Tongay, J. Zhou, J. Li, and J. Wu, Appl. Phys. Lett. 102, 012111(2013). 63W. Fan, H. Xu, A. L. Rosa, T. Frauenheim, and R. Q. Zhang, Phys. Rev. B 76, 073302 (2007). 64A. Miranda, X. Cartoixa, E. Canadell, and R. Rurali, Nanoscale Research Letters 7, 308 (2012). 65J. W. Spalenka, P. Gopalan, H. E. Katz, and P. G. Evans, Appl. Phys. Lett. 102, 041602 (2013). 66G. Jayalakshmi, N. Gopalakrishnan, and T. Balasubramanian, J. Alloys Compd. 551, 667 (2013). 203720-9 Dominguez et al. J. Appl. Phys. 115, 203720 (2014) [This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP: 186.217.234.103 On: Fri, 10 Oct 2014 16:49:49 http://dx.doi.org/10.1103/PhysRevB.9.600 http://dx.doi.org/10.1103/PhysRevB.75.235102 http://dx.doi.org/10.1063/1.2187006 http://dx.doi.org/10.1063/1.478522 http://dx.doi.org/10.1063/1.478401 http://dx.doi.org/10.1103/PhysRevB.47.558 http://dx.doi.org/10.1103/PhysRevB.49.14251 http://dx.doi.org/10.1016/0927-0256(96)00008-0 http://dx.doi.org/10.1103/PhysRevB.54.11169 http://dx.doi.org/10.1103/PhysRevB.50.17953 http://dx.doi.org/10.1103/PhysRevB.59.1758 http://dx.doi.org/10.1103/PhysRevB.13.5188 http://dx.doi.org/10.1103/PhysRevB.16.1748 http://dx.doi.org/10.1103/PhysRevB.80.195205 http://dx.doi.org/10.1103/PhysRevB.80.155124 http://dx.doi.org/10.1039/c4cp00667d http://dx.doi.org/10.1039/c1cp22128k http://dx.doi.org/10.1063/1.4809994 http://dx.doi.org/10.1002/adfm.201201478 http://dx.doi.org/10.1063/1.4774090 http://dx.doi.org/10.1103/PhysRevB.76.073302 http://dx.doi.org/10.1186/1556-276X-7-308 http://dx.doi.org/10.1186/1556-276X-7-308 http://dx.doi.org/10.1063/1.4790155 http://dx.doi.org/10.1016/j.jallcom.2012.11.006 http://dx.doi.org/10.1016/j.jallcom.2012.11.006