Addition to “On the Charge State of Titanium in Titanium Dioxide”

Abstract

W e provide here additional context to support the conclusion that we reached 1 that "Ti 3+ " (if one is to use an integer although charge quantization is not required here) is a better description of Ti in TiO 2 than "Ti 4+ ". Raebiger et al. 2 showed that the charge on a transition metal atom in a compound is very stable with respect to changes of the formal oxidation state (FOS) due to the so-called negative-feedback charge regulation. In ref 3, charge reporter molecules were used to reach a similar conclusion; specifically for Ti, a charge state of about +3 |e| was estimated. 3 Earlier, Christensen and Carter reported significant covalent character of bonds in ZrO 2 and concluded that in ZrO 2 "Zr is likely to be Zr(II) like" despite the FOS of Zr(IV). 4 In ref 1., we directly analyzed valence electron densities to ascertain that one electron charge resides on Ti in TiO 2 molecules and solids within less than half of the Ti−O bond length, independently of definitions of common charge assignment schemes. Specifically, valence densities around Ti in TiO 2 were compared to those in Ti 0 ...Ti 4+ ions. Such ions are directly detectable in plasma, 5,6 as opposed to, for example, XPS on solids (often cited as evidence for Ti 4+ ) where changes in charge surrounding the ionic core are detected indirectly. We confirmed that Bader charges provide a reasonable partitioning of space and Bader charges on the order of +2.5 |e| were most reasonable. We note that Bader charges do show full oxidation where it is known to happen (e.g., alkali atoms in semiconductors). 7−9 We also note that there is decent agreement between charge densities around Ti in a TiO 6 environment measured by X-ray/electron diffraction and computed by DFT. 10,11 Phillips fractional ionicity (based on Pauling electronegativity) of a Ti−O bond is 0.6, which implies a significant covalent character (compared, e.g., to 0.9 for LiF). Recent ionic potential models 12 that are able to reproduce properties of various TiO 2 polymorphs use charges of +2.4 and −1.2 |e| on Ti and O, respectively. 13 (Reference 13 reported structures and elastic properties of rutile TiO 2 , while we confirmed that a potential using these charges can reproduce structures of anatase, rutile, and bronze TiO 2 .) The "Ti 4+ " language implies no further oxidation of Ti and no further reduction of O. However, a degree of oxygen reduction is easily seen in ab initio calculations of doped oxides including TiO 2 , 7,14 and oxygen redox has finally been embraced in such systems. 15 That it took so long to recognize it 15 may have to do with the ubiquitous use of FOS to rationalize the mechanism. Given the above, "Ti 4+ " following from the ideology of FOS 16 appears to be a rather unphysical description. The FOS concept, which itself relies on the ionic approximation 17 that clearly does not hold in TiO 2 , predates modern quantum chemistry 18 and the knowledge gained from it, does not seem to be useful here. 19 ■ AUTHOR INFORMATION

Oxygen Redox Activity in Cathodes: A Common Phenomenon Calling for Density-Based Descriptors

“…Due to the pronounced covalent character of the transition metal–oxygen bonds, the atomic charges of the transition metal ions obtained from the Bader analysis are significantly lower than what would be expected from their formal oxidation states (FOS). This is consistent with previous theoretical reports which also show that this is not an error of the Bader partitioning itself, but it is clearly caused by a valence charge accumulation around the transition metal ions. , Oxygen contributions to the redox activity are also consistent with the well-established mechanism of charge self-regulation on transition metal centers whereby the TM center remains mostly redox-inactive while net charge density gains and losses around oxygen are mostly responsible for the redox process …”

“…This is consistent with previous theoretical reports which also show that this is not an error of the Bader partitioning itself, but is clearly caused by a valence charge accumulation around the transition metal ions. [14][15] Oxygen contributions to the redox activity are also consistent with the well-established mechanism of charge self-regulation on transition metal centers whereby changes in valence change are smaller than changes in the FOS. 16…”

Oxygen Redox Activity in Cathodes – a Common Phenomenon Calling for Density-Based Descriptors

“…This is consistent with previous theoretical reports which also show that this is not an error of the Bader partitioning itself, but is clearly caused by a valence charge accumulation around the transition metal ions. [14][15] Oxygen contributions to the redox activity are also consistent with the well-established mechanism of charge self-regulation on transition metal centers whereby changes in valence change are smaller than changes in the FOS. 16 The reason for the oxygen contributions to redox activity being reflected differently in the PDOS and the Bader charge analysis lies in the way these measures are computed: Bader charges rely on the topological partitioning of the electron density in real space (for details see e.g.…”

Titanium-Based Alloy Surface Modification with TiO2and Poly(sodium 4-styrenesulfonate) Multilayers for Dental Implants

“…This behavior has been observed in other LbL films containing sulfonic groups, for instance, sulfonated phthalocyanines interacting electrostatically with polycations demonstrated splitting varying from 35 to 47 cm –1 . , In addition, the LbL film spectrum had the bands related to the benzene ring vibrations red-shifted from 1130 to 1128 cm –1 and from 1010 to 1007 cm –1 . Because of the small shifts on the Ti–O stretching band positions, all these findings corroborate to indicate a supramolecular-level interaction between the multilayers, possibly through the electrostatic interactions between sulfonic groups and Ti (with the charge state +3 in titanium dioxides). , The complete assignments and band positions are listed in Table S1 in Supporting Information.…”

“…Because of the small shifts on the Ti−O stretching band positions, all these findings corroborate to indicate a supramolecular-level interaction between the multilayers, possibly through the electrostatic interactions between sulfonic groups and Ti (with the charge state +3 in titanium dioxides). 69,70 The complete assignments and band positions are listed in Table S1 in Supporting Information.…”