Electron paramagnetic resonance characterization of defects in monoclinic HfO2 and ZrO2 powders

Abstract

Electron paramagnetic resonance (EPR) measurements have been made at X-band and room temperature on monoclinic HfO2 and ZrO2 powders from several suppliers. They reveal the presence of eight main paramagnetic centers H1, H2, H3, H4, and Z1, Z2, Z3, and Z4. H1 and Z1 are analogous as H4 and Z4 and H2 and Z2 are similar as H3 and Z3. H1 and Z1 have axial symmetry with g∥<g⊥<ge, where ge is the free electron g value. H1 is found in all, and Z1 in all but one, of the samples in their as-received state but with a wide range of concentrations. However, annealing the samples in air up to 900 °C reduces the volume concentration range and the areal concentrations all become of order 1011 cm−2. Irradiation with γ-rays does not affect their concentration. The Z1 centers are found to be the same as those previously observed in ZrO2 powders that were attributed to Zr3+ ions in coordinatively unsaturated (cus) sites at and/or near the surface. Our results are consistent with this model for Z1 and with an analogous model of cus Hf3+ for H1. H4 and Z4 are centers of isotropic symmetry with g values that are both within ±0.0004 of 2.0027; they are produced in all HfO2 and ZrO2 samples, respectively, that are heated in vacuum at ≥300 °C. Their concentration reaches a maximum of order 1017 cm−3 or 1012 cm−2 in the range of 550–750 °C. They are also most likely to be mainly at and/or near the surface and to involve an electron trapped in an oxygen vacancy cluster. The EPR spectra of H2 and Z2 are consistent with those of S=1/2 centers of orthorhombic symmetry with principal g values about equal to or just less than ge suggesting that they are trapped electron centers. The electrons produced by γ-irradiation are trapped at precursors to H2 but are easily detrapped. Z2 centers also appear to be shallow electron traps. Their identity is uncertain; they have some characteristics of electrons trapped in oxygen vacancies and of CO2− radicals. H3 and Z3 are likely to involve holes trapped on oxygen, possibly as O− and O2− type centers, respectively, but their location in not known. Their concentration increases to an upper limit as the γ-ray dose is increased and this shows that their precursors are trapping charge generated by the γ-rays. Like the H2 and Z2 centers, even annealing at 100 °C releases the charge but their precursors, at least in HfO2, are not destroyed. The significance of these centers is discussed.

Electron spin resonance signature of the oxygen vacancy in HfO2

“…4,6,11,13,[15][16][17][18] There has been no fully accepted chemical identification of the oxygen vacancy in HfO 2 by, say, electron spin resonance (ESR). [19][20][21] This situation differs from the oxygen vacancy in TiO 2 22 and contrasts strongly with the vast literature on the paramagnetic E 0 and P b centers in SiO 2 . 23,24 In this paper, we explain the problems with the symmetry assignment of the oxygen vacancy in HfO 2 in previous calculations and present a calculation of the ESR signature of the O vacancy using hybrid density functional methods, to help aid this identification.…”

“…18 In ESR, Kang et al 19 have observed various paramagnetic defects such as the peroxyl radical in atomic layer deposited (ALD) HfO 2 . Wright and Barklie 20 observed a number of low symmetry defects in HfO 2 powders and thin films that they associated with under-coordinated Hf 3þ sites. Stesmans 21 observed the P b center in HfO 2 on Si, but this was associated with the SiO 2 interlayer not the HfO 2 itself.…”

Intrinsic defect related luminescence in ZrO2

“…The trap for the holes is not clear; however the valence band top is formed from the oxygen 2p states [23] and theoretical calculation and some experiments showed the holes localized on oxygen [19,24,25]. Since the STE was observed [6] the selftrapped hole could also exist and a configuration similar to the self-trapped hole localized at some defect could be a hole center participating in non-radiative recombination.…”

Lattice distortions and oxygen vacancies produced in Au+-irradiated nanocrystalline cubic zirconia

“…For the 400 K irradiation, the V 0 defect appears to be the dominant defect. For the 160 K irradiation, it is not quite so clear, with both the V 2À and V À defects producing similar distortions to the lattice; however, it suggests that these two defects may be the dominant defects at 160 K. Electron paramagnetic resonance experiments [17] performed on c-ray-irradiated monoclinic ZrO 2 has shown that the V À defect is the most likely to exist, leading to the possibility that this is the most dominant defect at 160 K. This cannot be categorically proven, however, as the difference in the crystal structures (monoclinic phase versus cubic phase) may lead to different charge states being dominant.…”

“…To date, most research has focused on radiation processes (primarily heavy ion irradiation) in either partially-or fully stabilized zirconia as single crystals or with micron-sized grains, or on nanograined ZrO 2 that are either embedded in a foreign matrix or as free-standing particles [15][16][17]. The work presented here examines the effect of heavy ion irradiation on NSZ with a view to understanding the defect production, saturation and evolution of oxygen vacancies, and how this impacts the unit cell structure.…”