Persistent Luminescence Hole-Type Materials by Design - American ...

58 downloads 1745 Views 494KB Size Report
Feb 5, 2016 - might be a good host for the PL with certain transition-metal elements because ... the lonsdaleite suggests that Cr and Mn may give rise to good.
Research Article www.acsami.org

Persistent Luminescence Hole-Type Materials by Design: TransitionMetal-Doped Carbon Allotrope and Carbides Bingyan Qu,† Bo Zhang,† Lei Wang,*,†,‡ Rulong Zhou,*,† Xiao Cheng Zeng,*,§,∥ and Liang Li⊥ †

Laboratory of Amorphous Materials, School of Materials Science and Engineering, Hefei University of Technology, Hefei, Anhui 230009, P. R. China ‡ Faculty of Applied Sciences, Delft University of Technology, Mekelweg15, 2629 JB Delft, The Netherlands § Department of Chemistry and Nebraska Center for Materials and Nanoscience, University of NebraskaLincoln, Lincoln, Nebraska 68588, United States ∥ Hefei National Laboratory for Physical Sciences at Microscale and Collaborative Innovation Center of Chemistry for Energy Materials, University of Science and Technology of China, Hefei, Anhui 230026, China ⊥ State Key Laboratory of Automotive Safety and Energy, Tsinghua University, Beijing 100084, China ABSTRACT: Electron traps play a crucial role in a wide variety of compounds of persistent luminescence (PL) materials. However, little attention has been placed on the hole-trap-type PL materials. In this study, a novel hole-dominated persistent luminescence (PL) mechanism is predicted. The mechanism is validated in the night pearl diamond (NPD) composed of lonsdaleite with ultralong persistent luminescence (PL) (more than 72 h). The computed band structures suggest that the Fe ion dopant in lonsdaleite is responsible for the luminescence of NPD due to the desired defect levels within the band gap for electronic transition. Other possible impurity defects in lonsdaleite, such as K, Ca, Mg, Zn, or Tl dopants, or C vacancy can also serve as the hole-trap centers to enhance the PL. Among other 3d transition-metal-ion dopants considered, Cr and Mn ions are predicted to give rise to PL property. The predicted PL mechanism via transition-metal doping of lonsdaleite offers an exciting opportunity for engineering new PL materials by design. KEYWORDS: long afterglow, first-principle calculation, diamond, transition metals, defects, luminescence, lonsdaleite, hexagonal diamond

I. INTRODUCTION Over the past two decades, considerable effort has been devoted toward understanding mechanisms of persistent luminescence (PL) materials with prolonged afterglow beyond 10 h. The effort is justified because of promising applications of PL materials in electronic display, information storage,1 vivo targeted bioimaging,2 among others. For most PL materials studied, e.g., aluminates SrAl2O4:Eu,Dy and CaAl2O4:Eu,Nd,3−5 silicates Ba5Si8O21:Eu,Dy6 and CdSiO3:Tb,7 phosphates CaZr(PO 4 ) 2 :Eu, 8 stannite Sr 2 SnO 4 :Sm, 9 nitride oxide BaSi2O2N2:Eu,10 gallates ZnGa2O4:Cr,1 and Zn3Ga2Ge2O10:Cr,11 the electron carriers are believed to play a crucial role in PL. Important features underlying these PL materials include the excited-state levels being close to the conduction band minimum (CBM) and suitable depth of the electron-trap levels dominating the thermal release rate at room temperature. Both features can have a pronounced effect on the afterglow (Figure 1(a)). Despite successful developments of many types of PL materials, it is still an open question whether the hole traps rather than electron traps, in certain prominent PL materials, can play a major role in the PL phenomena (see Figure 1(b)). A better understanding of the hole-trap-associated PL phenomena would be very important not only to broaden © 2016 American Chemical Society

Figure 1. Illustration of possible mechanisms due to electron (a) and hole (b) traps that can directly contribute to the persistent luminescence.

the fundamental view about the physical nature of PL but also for the design of new PL materials with novel properties. To achieve the energy-level scheme as shown in Figure 1(b), the ground-state level of the emission center should be close to the valence band maximum (VBM), while the trap levels should have suitable depth to the VBM. However, for host materials Received: November 29, 2015 Accepted: February 5, 2016 Published: February 5, 2016 5439

DOI: 10.1021/acsami.5b11609 ACS Appl. Mater. Interfaces 2016, 8, 5439−5444

Research Article

ACS Applied Materials & Interfaces

Figure 2. Computed band structure, total and projected density of states (DOS and PDOS) of (a) perfect lonsdaleite, (b) Fe-doping lonsdaleite, and (c) imaginary part of the dielectric function of Fe-doped lonsdaleite. The defect levels in (b) are denoted by the blue solid lines. The Fermi level is set to 0 eV.

hole-dominated PL as well. We also find that the cubic diamond in general is not a good host compared to lonsdaleite due to the relatively higher occupied energy levels of the dopants.

with a wide band gap, such as fluoride and metal oxide discussed in ref 12, the emission centers are far from the VBM, and thus the energy-level scheme cannot meet the requirement for hole-type-based PL. Recently, Dorenbos systematically studied energy levels of the emission-center ions of Eu3+ and Ce3+ in various PL materials, based on numerous experimental data, and found that the location of the energy level (relative to the vacuum level) of 4f levels is more stable, compared to the variation of VBM for oxide, fluoride, sulfide, or iodide.12−19 This may be a general trend for all 4f elements and 3d elements. As for the VBM of a compound, from the molecular orbital view, its position is mainly contributed by the valence electronic orbital of the ligand which has the highest valence electronic energy level. According to the nephelauxetic effects, for a given cation, the sequence of the energy level of the valence electronic orbital of different ligands roughly follows

II. COMPUTATIONAL DETAILS All the calculations are performed using the VASP 5.3 program,23,24 where the projector augmented wave (PAW) pseudopotentials and the Perdew−Burke−Ernzerhof (PBE)25 exchange-correlation functional within the generalized gradient approximation (GGA) are adopted. To better describe 3d elections, the GGA+U method with the formulation suggested by Dudarev et al.26 is also employed. The electronic wave functions are described by a set of plane wave functions with an energy cutoff of 400 eV. To prohibit the spurious interaction among the periodic images of the defect, we select supercells containing 4 × 4 × 3, 4 × 4 × 4, and 3 × 3 × 2 primitive cells of lonsdaleite, cubic diamond, and wurtzite ZnS, respectively. The Brillouin zones are sampled by the 3 × 3 × 3 Γ-centered k-point mesh within the Monkhorst−Pack scheme27 in all calculations. Since the ion radii of the dopants considered in this work are much larger than that of the C ion, only the substitution doping is considered.

Transition metal (3dn) > Tl > Al > Sn > Si > Ge > C > Se > S > N > Br > O > F

in the covalent semiconductors.20 This trend is consistent with the measured binding energies of VBM for both wide-band compounds and semiconducting minerals whose data are collected in refs 18, 21, 12, and 19. Thus, one can simply shift the VBM upward to a suitable level by selecting a suitable host material and then predict the relative location of the acceptor level (dopant in the host) with this information. More importantly, this sequence can offer a useful guide for screening certain hosts and dopants. We notice that C is in the middle of the sequence, ahead of N and O. As such the VBM of the carbide may be higher than oxide and nitride compounds and lower than the transition metal. The ground-state levels of transition metals acting as the emission center are more likely close to the VBM of C. As such, the hole-dominated PL may be realized through the mechanism illustrated in Figure 1(b). We noticed a news report on the blue-green-emitting natural night pearl diamond (NPD) with a long decay time of more than 72 h.22 The news suggests that lonsdaleite (hexagonal diamond) might be a good host for the PL with certain transition-metal elements because lonsdaleite is the main component of NPD. In this article, we show that lonsdaleite in NPD can be a viable host material to realize hole-dominated PL with Fe as dopant, thereby confirming the hole-dominated PL mechanism. Moreover, we show that many other defects in the NPD, such as Zn dopant and C vacancy, can yield acceptor levels close to the VBM, thereby serving as hole traps to prolong the PL. Our systematic study of other 3d transition-metal ions as dopants in the lonsdaleite suggests that Cr and Mn may give rise to good

III. RESULTS AND DISCUSSION First, we compute the band structures of the perfect lonsdaleite. As shown in Figure 2(a), lonsdaleite exhibits a wide band gap of ∼3.65 eV (PBE level). The partial density of states (PDOS) shows that the VBM and CBM are primarily attributed by the C 2p orbital, as in the case of cubic diamond. It is known that the PBE functional tends to underestimate the band gap of semiconductors and insulators. So, the realistic band gap of lonsdaleite should be much wider, although no experimental measurement of the band gap has been reported in the literature. Here, we also utilized the Heyd−Scuseria−Ernzerhof (HSE06) hybrid functional28−30 to obtain a more accurate band gap, i.e., 4.6 eV. The latter gap is in good agreement with the value (4.5 eV) estimated by Salehpour and Satpathy31 and that (4.68 eV) based on the GW method.32 Compared to the band gap of the cubic diamond (5.47 eV), the lonsdaleite’s band gap is 0.87 eV smaller.33 For the cubic diamond, both the CBM and VBM are 3-fold degenerate, while these states are split into a 1fold and a 2-fold state owing to the crystal field of hexagonal lonsdaleite.31 This splitting causes the upward shift of VBM and downward shift of CBM. Next, we examine a variety of dopant ions and seek the suitable one as a luminescent center in lonsdaleite. Generally, the optical excitation and light emission for the long-persistent materials are related to the 4f−5d or 4f−4f transition of rareearth ions or d−d transition of transition-metal ions with partially occupied d levels. In Table 1, the composition of the 5440

DOI: 10.1021/acsami.5b11609 ACS Appl. Mater. Interfaces 2016, 8, 5439−5444

Research Article

ACS Applied Materials & Interfaces Table 1. Element Composition in NPD22 composition

Zn

S

Si

Al

Mg

content composition

22.35% Ca

12.99% Tl

5.64% Cl

3.79% K

2.35% Fe

content

1.56%

1.20%

0.83%

0.58%

1.10%

NPD is listed, where only the Fe ion meets the requirement for the optical excitation. The electronic properties of Fe-doped lonsdaleite are shown in Figure 2(b). Compared to the band structures of perfect lonsdaleite (see Figure 2(a)), one can see that the Fe ion dopant induces some defect levels in the band gap. The computed PDOS indicates that the defect levels stem from the Fe 3d orbitals. A few defect levels are below the Fermi level and occupied by electrons, while the others are empty. The energy difference between the occupied and unoccupied levels are about 2.2 eV (PBE computation without including the Hubbard U parameter), slightly smaller than that corresponding to the luminescent wavelength of NPD (about 2.4 eV). To better describe the strong interaction of the localized 3d electrons of Fe, we also used the GGA+U method to compute the electronic structure of the Fe-doped lonsdaleite, with U value ranging from 1 to 4 eV. The introduction of U correction can slightly move the occupied levels of Fe downward but has little effect on the unoccupied levels. For U = 1 eV, the energy gap between the occupied and unoccupied 3d levels of Fe is consistent with the luminescent wavelength of NPD. The PDOS shows a very weak hybridization between Fe 3d and C 2p states. Such a small hybridization unlikely affects the electron transition probabilities between Fe 3d levels based on computed imaginary part of dielectric function (at PBE level) (see Figure 2(c)). We conclude that the observed blue-green emitting is most likely attributed to the electron transition between the Fe 3d orbitals. Through counting the occupied energy levels of Fe ion, we obtain the number of valence electrons bound to Fe in the ground state, which is 4; i.e., Fe is in the +4 valence state. This is reasonable because each of the four surrounding C atoms will draw one electron from Fe. However, it is still essential to check the stability of Fe4+ since Fe3+ and Fe2+ are the more common forms of Fe ions. So, we calculate the formation energies of Fe ions with +4, +3, and +2 valence states in the lonsdaleite, respectively, according to the widely accepted formula34

Figure 3. Calculated formation energies of the Fe ion at different valence states as a function of the Fermi level. The VBM is taken as reference.

they are only 0.10 eV above the VBM. So, when excited by light (hv), one electron of Fe4+ enters into the high 3d energy level of Fe4+, creating an excited state (Fe4+)* with the low levels not fully occupied. The electron in the VBM can enter into these low levels by thermal exciting (Eth), leaving a hole at the VBM and the unstable Fe3+ ion. Equally, this process can be considered as a hole jumps from the Fe4+ level into the valence band, Fe4+ + hv + Eth → [Fe3+, h], just as we always consider for dopants in semiconductors. Since Fe3+ is unstable in lonsdaleite, it may change back to (Fe4+)* when a hole in the valence band appears at the position of Fe, and then (Fe4+)* transits back to the ground Fe4+ state accompanied by light emission. Due to the delocalizing of the valence bands, the hole at the valence bands can transfer to any other bands and can also be trapped by the hole-trap levels with suitable depth for a while when the hole reaches the position of a trap center and then move back to valence bands, so that the longpersistent luminescence can occur. Analysis of the electronic structures thus far indicates that the proximity of the trap levels to the VBM is critical to the longpersistent luminescence. To gain more insight into which defects are responsible for the trap levels, we compute electronic properties with considering every dopant listed in Table 1 in lonsdaleite. In addition, the intrinsic defect of C vacancy (VC) is also considered. The computed band structures are shown in Figure 4. When a C atom is substituted by a potassium atom (K), the defect levels are induced in the band gap for both spin channels. Some defect levels are occupied, while the others are empty. The occupied defect levels are distributed within an energy range of ∼1.0 eV. The lowest occupied level is approximately 0.5 eV above the VBM. These occupied levels can act as the hole traps to prolong the luminescence. Besides the K dopant, Ca, Mg, Zn, Tl, S, Cl, and the C vacancy (VC) can also induce the defect levels in the band gap. Except S and Cl, all other defects can introduce occupied defect levels closely above the VBM, meaning that these defects can also serve as hole-trap centers like K. For the S dopant, the defect energy levels are too close to the CBM and thus cannot serve as trap hole carriers. In other words, the S dopants in the lonsdaleite may actually be adverse to the PL. As shown in Table 1, the composition of S is quite high in NPD. The S atoms are likely combined with some other elements to form compounds (see below) rather than as stand-alone dopant atoms in the lonsdaleite. For the Cl dopant, the occupied defect levels are also quite high above the VBM. Among all the favorable defects, the Ca dopant gives the shallowest trap level, nearly overlapping with the VBM at some k-points. The K dopant gives the deepest trap level with the depth of 1.5 eV. All these defects are expected to contribute to

E F(Feq + 4) = E q(C191Fe) − E(C192) − μFe + μC + qεf

where Eq(C191Fe) and E(C192) represent the total energy of the compound with and without Fe; q is the charge of the doped system; μC and μFe are the chemical potentials of the bulk C and Fe; and εf is the Fermi level of the system. Since the valence state of the Fe ion in the neutral system is +4, the q corresponding to Fe3+ and Fe2+ are −1 and −2, respectively. The calculation results (at the PBE level) are shown in Figure 3 where the VBM is taken as reference. Fe4+ gives lower formation energy when εf is lower than 3.12 eV, while Fe3+ is more stable when εf is higher. The calculated band gap of lonsdaleite is only 3.65 eV (PBE level). It means that Fe3+ can only be stable at electron-doping conditions. According to the calculated energy levels of all the other dopants in NPD (see below), no dopants can donate high density of electron carriers, while some of them , e.g., Al, can supply high density of holes. So, in NPD, Fe4+ is stable, while Fe3+ is unstable. Note that without the U correction the computed occupied d levels are only 0.30 eV above the VBM, while with U = 1 eV, 5441

DOI: 10.1021/acsami.5b11609 ACS Appl. Mater. Interfaces 2016, 8, 5439−5444

Research Article

ACS Applied Materials & Interfaces

Figure 4. Computed band structures of lonsdaleite with (a) K, (b) Ca, (c) Mg, (d) Tl, (e) Zn, (f) single C vacancy, (g) S, (h) Cl, (i) Al, or (j) Si defect. The defect levels are denoted by the blue lines in the band structures. The Fermi level is set to 0 eV.

Figure 5. Computed band structures of (a) perfect and (b) the Fe-doped ZnS. The defect levels are denoted by the blue lines in the band structures. The Fermi level is set to 0 eV.

Figure 6. Computed band structures of lonsdaleite doped with (a) Sc, (b) Ti, (c) V, (d) Cr, (e) Mn, (f) Co, (g) Ni, and (h) Cu ions. The defect levels in the band gap originated from 3d orbitals are denoted by the blue lines.

the hole-trap levels distributed over a wide energy range from 0 to 1.5 eV with respect to the VBM, thereby leading to the remarkably long decay time of PL. For the Al dopant, as shown in Figure 4(i), the defect levels merge with the valence band so

that the generation of hole carriers should be very easy. Thus, high density of hole carriers is expected in the high-valence bands. The hole carriers also endow the NPD with high luminescent intensity due to the high rate of hole-carrier 5442

DOI: 10.1021/acsami.5b11609 ACS Appl. Mater. Interfaces 2016, 8, 5439−5444

Research Article

ACS Applied Materials & Interfaces

Figure 7. Computed band structures of the cubic diamond doped with (a) Sc, (b) Ti, (c) V, (d) Cr, (e) Mn, (f) Fe, (g) Co, (h) Ni, (i) Cu, and (j) Zn ions. The defect levels in the band gap originated from 3d orbitals are denoted by the blue lines.

IV. CONCLUSION We have proposed a novel hole-dominated persistent luminescence mechanism. In supporting this mechanism, we show that in Fe-doped lonsdaleite the Fe ions act as the luminescent center, while other K, Ca, Mg, Zn, and Tl dopants or C vacancy can induce widely distributed hole traps that can contribute to the remarkably prolonged decay time. We also predict that Cr and Mn ions in lonsdaleite may even give more favorable persistent luminescence property. Our results also point to future fabrication of new hole-dominated PL materials by design, for example, in 3d element-doped carbon allotropes or carbides. We hope this prediction can promote future experimental studies of this new class of PL materials.

recombination. Lastly, the Si dopant has little influence on the high and low conduction bands and thus little contribution to the PL of the lonsdaleite. Note that because the compositions of Zn and S are quite high in NPD (Table 1), ZnS is likely a common form of compound. We therefore also compute the electronic structures of Fe-doped ZnS (Figure 5). It is found that the Fe ions in ZnS cannot emit blue-green light because the energy difference between the occupied and unoccupied levels is only ∼0.8 eV. By now, our calculation results resolve the puzzle of the longtime PL of NPD. The Fe dopant in lonsdaleite is predicted to act as a luminescent center. Can other 3d transition metal ions in lonsdaleite also serve as luminescent centers for the persistent luminescence? To address this question, we compute electronic properties of the lonsdaleite with other 3d ion dopants. The computed band structures (without U correction) are shown in Figure 6. Under the crystal field of lonsdaleite, the 3d levels are split into two parts for both spin channels. In Figure 6, only the defect levels in the band gap originated from 3d orbitals are denoted by the blue lines; the higher 3d levels of Sc, which are located within the conduction bands and the lower 3d levels of Ni and Cu, which enter into the valence bands, are not marked in the figure. Apparently, the depth of the occupied energy levels of the 3d ions to the VBM decreases with increasing atomic number. The Cr, Mn, and Co dopants’ occupied levels are also close to the VBM, implying that these dopants may give hole-dominated PL properties as well. With the U correction, the occupied 3d levels of Co would be deeply into the valence bands, while those of Cr and Mn would be closer to the VBM. Hence, we predict that Cr and Mn are more acting as shallower traps compared with that of Fe. For the purpose of comparison, we also compute electronic structures of the cubic diamond with various 3d dopants. As shown in Figure 7, the 3d energy levels of the transition metals are very similar to those in lonsdaleite except that the occupied 3d levels are all consistently higher compared to those in lonsdaleite. Cr, Mn, and Fe dopants in cubic diamond may still yield PL. However, because of their relatively higher occupied 3d levels (0.65 eV for Fe), the PL intensity is expected to be much weaker than that based on lonsdaleite. This is a possible reason why no night pearl made from cubic diamond is seen in nature.



AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. *E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors are indebted to Professor P. Dorenbos of Delft University of Technology for discussions of PL mechanism of lonsdaleite which are very valuable in the interpretation of our calculation results from the view of experimental aspect. This work is supported by the National Natural Science Foundation of China (Grant No: 51302059, 51171055, 51322103, 11404085, and 11104056), the Funding of the State Key Laboratory of Automotive Safety and Energy (KF14062), and University of Science and Technology of China Qian-ren B fund for summer research. The numerical calculations in this paper have been done in the supercomputing system in the Supercomputing Center of University of Science and Technology of China.



REFERENCES

(1) Bessière, A.; Sharma, S. K.; Basavaraju, N.; Priolkar, K. R.; Binet, L.; Viana, B.; Bos, A. J. J.; Maldiney, T.; Richard, C.; Scherman, D.; Gourier, D. Storage of Visible Light for Long-Lasting Phosphorescence 5443

DOI: 10.1021/acsami.5b11609 ACS Appl. Mater. Interfaces 2016, 8, 5439−5444

Research Article

ACS Applied Materials & Interfaces in Chromium-Doped Zinc Gallate. Chem. Mater. 2014, 26, 1365− 1373. (2) Abdukayum, A.; Chen, J.-T.; Zhao, Q.; Yan, X.-P. Functional Near Infrared-Emitting Cr3+/Pr3+ Co-Doped Zinc Gallogermanate Persistent Luminescent Nanoparticles with Superlong Afterglow for in Vivo Targeted Bioimaging. J. Am. Chem. Soc. 2013, 135, 14125− 14133. (3) Clabau, F.; Rocquefelte, X.; Jobic, S.; Deniard, P.; Whangbo, M. H.; Garcia, A.; Le Mercier, T. Mechanism of Phosphorescence Appropriate for the Long-Lasting Phosphors Eu2+-Doped SrAl2O4 with Codopants Dy3+ and B3+. Chem. Mater. 2005, 17, 3904−3912. (4) Hölsä, J.; Laamanen, T.; Lastusaari, M.; Malkamäki, M.; Welter, E.; Zajac, D. A. Valence and environment of rare earth ions in CaAl2O4:Eu2+,R3+ persistent luminescence materials. Spectrochim. Acta, Part B 2010, 65, 301−305. (5) Qu, B.; Zhang, B.; Wang, L.; Zhou, R.; Zeng, X. C. Mechanistic Study of the Persistent Luminescence of CaAl2O4:Eu,Nd. Chem. Mater. 2015, 27, 2195−2202. (6) Wang, P.; Xu, X.; Zhou, D.; Yu, X.; Qiu, J. Sunlight Activated Long-Lasting Luminescence from Ba5Si8O21: Eu2+,Dy3+ Phosphor. Inorg. Chem. 2015, 54, 1690−1697. (7) Rodrigues, L. C. V.; Brito, H. F.; Hölsä, J.; Stefani, R.; Felinto, M. C. F. C.; Lastusaari, M.; Laamanen, T.; Nunes, L. A. O. Discovery of the Persistent Luminescence Mechanism of CdSiO3:Tb3+. J. Phys. Chem. C 2012, 116, 11232−11240. (8) Zhang, J.-C.; Xu, C.-N.; Long, Y.-Z. Elastico-mechanoluminescence in CaZr(PO4)2:Eu2+ with multiple trap levels. Opt. Express 2013, 21, 13699−13709. (9) Xu, X.; Wang, Y.; Gong, Y.; Zeng, W.; Li, Y. Effect of oxygen vacancies on the red phosphorescence of Sr2SnO4:Sm3+ phosphor. Opt. Express 2010, 18, 16989−16994. (10) Botterman, J.; Eeckhout, K. V. d.; Baere, I. D.; Poelman, D.; Smet, P. F. Mechanoluminescence in BaSi2O2N2:Eu. Acta Mater. 2012, 60, 5494−5500. (11) Pan, Z.; Lu, Y.-Y.; Liu, F. Sunlight-activated long-persistent luminescence in the near-infrared from Cr3+-doped zinc gallogermanates. Nat. Mater. 2012, 11, 58−63. (12) Dorenbos, P. Lanthanide 4f-electron binding energies and the nephelauxetic effect in wide band gap compounds. J. Lumin. 2013, 136, 122−129. (13) Dorenbos, P.; Rogers, E. G. Vacuum Referred Binding Energies of the Lanthanides in Transition Metal Oxide Compounds. ECS J. Solid State Sci. Technol. 2014, 3, R150−R158. (14) Dorenbos, P. The electronic structure of lanthanide doped compounds with 3d, 4d, 5d, or 6d conduction band states. J. Lumin. 2014, 151, 224−228. (15) Dorenbos, P. A Review on How Lanthanide Impurity Levels Change with Chemistry and Structure of Inorganic Compounds. ECS J. Solid State Sci. Technol. 2013, 2, R3001−R3011. (16) Dorenbos, P. Ce3+ 5d-centroid shift and vacuum referred 4felectron binding energies of all lanthanide impurities in 150 different compounds. J. Lumin. 2013, 135, 93−104. (17) Dorenbos, P. Electronic structure engineering of lanthanide activated materials. J. Mater. Chem. 2012, 22, 22344−22349. (18) Rogers, E. G.; Dorenbos, P. Vacuum Referred Binding Energy of the Single 3d, 4d, or 5d Electron in Transition Metal and Lanthanide Impurities in Compounds. ECS J. Solid State Sci. Technol. 2014, 3, R173−R184. (19) Dorenbos, P. Determining binding energies of valence-band electrons in insulators and semiconductors via lanthanide spectroscopy. Phys. Rev. B: Condens. Matter Mater. Phys. 2013, 87, 035118. (20) Hjalmarson, H. P.; Vogl, P.; Wolford, D. J.; Dow, J. D. Theory of Substitutional Deep Traps in Covalent Semiconductors. Phys. Rev. Lett. 1980, 44, 810−813. (21) Xu, Y.; Schoonen, M. A. A. The absolute energy positions of conduction and valence bands of selected semiconducting minerals. Am. Mineral. 2000, 85, 543−556.

(22) CCTV. Night pearl of meteorite diamond. http://tv.cntv.cn/ video/C20912/ff5ab27ca0924571a4924faaccada88e (accessed Sep 9, 2010). (23) Kresse, G.; Hafner, J. Ab initio molecular dynamics for liquid metals. Phys. Rev. B: Condens. Matter Mater. Phys. 1993, 47, 558−561. (24) Kresse, G.; Furthmüller, J. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. Rev. B: Condens. Matter Mater. Phys. 1996, 54, 11169−11186. (25) Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996, 77, 3865−3868. (26) Dudarev, S. L.; Botton, G. A.; Savrasov, S. Y.; Humphreys, C. J.; Sutton, A. P. Electron-energy-loss spectra and the structural stability of nickel oxide: An LSDA+U study. Phys. Rev. B: Condens. Matter Mater. Phys. 1998, 57, 1505−1509. (27) Monkhorst, H. J.; Pack, J. D. Special points for Brillouin-zone integrations. Phys. Rev. B: Condens. Matter Mater. Phys. 1976, 13, 5188−5192. (28) Heyd, J.; Scuseria, G. E.; Ernzerhof, M. Hybrid functionals based on a screened Coulomb potential. J. Chem. Phys. 2003, 118, 8207− 8215. (29) Heyd, J.; Scuseria, G. E.; Ernzerhof, M. Erratum: “Hybrid functionals based on a screened Coulomb potential” [J. Chem. Phys.118, 8207 (2003)]. J. Chem. Phys. 2006, 124, 219906. (30) Heyd, J.; Scuseria, G. E. Efficient hybrid density functional calculations in solids: Assessment of the Heyd−Scuseria−Ernzerhof screened Coulomb hybrid functional. J. Chem. Phys. 2004, 121, 1187− 1192. (31) Salehpour, M. R.; Satpathy, S. Comparison of electron bands of hexagonal and cubic diamond. Phys. Rev. B: Condens. Matter Mater. Phys. 1990, 41, 3048−3052. (32) Gao, S.-P. Band gaps and dielectric functions of cubic and hexagonal diamond polytypes calculated by many-body perturbation theory. Phys. Status Solidi B 2015, 252, 235−242. (33) Clark, C. D.; Dean, P. J.; Harris, P. V. Intrinsic Edge Absorption in Diamond. Proc. R. Soc. London, Ser. A 1964, 277, 312−329. (34) Kohan, A.; Ceder, G.; Morgan, D.; Van de Walle, C. Firstprinciples study of native point defects in ZnO. Phys. Rev. B: Condens. Matter Mater. Phys. 2000, 61, 15019−15027.

5444

DOI: 10.1021/acsami.5b11609 ACS Appl. Mater. Interfaces 2016, 8, 5439−5444