Content uploaded by Pietro Galinetto
Author content
All content in this area was uploaded by Pietro Galinetto
Content may be subject to copyright.
Response to “Comment on ‘Enhancement of room temperature
ferromagnetism in N-doped TiO2−xrutile: Correlation with the local electronic
properties’ ” †Appl. Phys. Lett. 97, 186101„2010…‡
G. Drera,1M. C. Mozzati,2P. Galinetto,2Y. Diaz-Fernandez,3L. Malavasi,3F. Bondino,4
M. Malvestuto,5and L. Sangaletti1,a兲
1Dipartimento di Matematica e Fisica, Università Cattolica, via dei Musei 41, 25121 Brescia, Italy
2Dipartimento di Fisica “A. Volta” and CNISM, Università di Pavia, Via Bassi 6, 27100 Pavia, Italy
3Dipartimento di Chimica, Università di Pavia, Via Bassi 6, 27100 Pavia, Italy
4Laboratorio TASC, IOM-CNR, S.S.14, Km 163.5, I-34149 Basovizza, Italy
5Sincrotrone Trieste, S.S. 14 Km 163.5 Area Science Park, I-34149 Basovizza, Italy
共Received 1 October 2010; accepted 11 October 2010; published online 4 November 2010兲
关doi:10.1063/1.3509410兴
The main concerns in the comment1to our recent Letter2
seem to be 共i兲the role of oxygen vacancies 共VOs兲in deter-
mining the electronic structure and, ultimately, the extent of
magnetization, of N-doped TiO2and 共ii兲the fact that photo-
emission data measured with synchrotron radiation could not
really represent the electronic structure of bulk materials.
共i兲The first objection is based on the assumption that
N-doping of TiO2rutile creates VOs. This assumption
should not be regarded as a general rule, as careful
studies on epitaxially grown N-doped TiO2rutile
films3,4have shown that for extremely low density of
VOs, interstitial Ti can also be present, yielding sub-
stitutional N sites with a ⫺3 formal charge due to
charge transfer 共CT兲from shallow-donor interstitial Ti
levels. In any case, if we assume that we have enough
VOs to favor N incorporation, the discussion relevant
for magnetic properties is not about how many VOs
we have, rather about where the extra electrons left
from the VOs go. While there is general consensus
about the CT to Ti 3din oxygen substoichiometric
TiO2, yielding in-gap states with unpaired magnetic
moment, upon N-doping this CT mechanism is known
to be in competition with oxidation of Ti and subse-
quent filling of N states at the top of the valence band,
as shown in Fig. 7 of Ref. 5. Consistently, we do not
observe an enhancement of the in-gap 3dstates upon
N-doping, as we show in Fig. 4共c兲of Ref. 2, but new
N-related states at the top of the valence band. At this
light, we believe that the observed magnetization en-
hancement should be related to band-gap narrowing
effects, as we state in the conclusions of our Letter,2
rather than to the creation of new VOs.
共ii兲As for the second objection, the comment1focuses on
the x-ray photoemission 共XPS, Fig. 2 in Ref. 2兲and
resonant photoemission 共RESPES, Fig. 4 in Ref. 2兲
spectroscopy data, while the results of x-ray absorp-
tion spectroscopy 共XAS, Fig. 3 in Ref. 2兲are disre-
garded. We wish to remark that the probe depth of the
XAS spectra, measured by collecting the drain current
from the sample, is more bulk sensitive than the pho-
toemission data. Consistently, from the XAS analysis
we showed that additional in-gap states are found by
looking at the empty states probed by N K-edge spec-
tra, which are in good agreement with the bulk model
of substitutional N atoms in the rutile lattice, at the
basis of our electronic structure calculations. More-
over, the second objection is rather generic and
ignores the basic notions on the different probe
depths of photoemission techniques.6For example, at
the light of recent synchrotron-based hard XPS
experiments,6it is obviously not always true that syn-
chrotron radiation is a very surface sensitive tech-
nique, as claimed in Ref. 1. In our experiment, the
bulk sensitivity increases going from RESPES at the
Ti L-edge 共about 1 nm for the valence band states兲to
XPS 共about 2 nm for the Ti 2p,N 1s, and O 1score
lines兲and ultimately to XAS 关about 4 nm 共Ref. 7兲兴.
Incidentally we observe that, bulk sensitive, optical
transmission measurements on our samples revealed a
band-gap narrowing of about 0.25 eV upon N-doping,
in agreement with what we inferred by the joint analy-
sis of XAS and photoemission data.
In conclusion, our photoemission data are not in contra-
diction with the results reported in previous studies on
N-doped rutile systems with a low content of VO, such as the
sample denoted as N-doped at RT in Fig. 2共a兲of Ref. 8,or
the B sample in Ref. 4. Nor they conflict with the data re-
ported for Fe-doped or Co-doped TiO2, since the CT mecha-
nism observed upon N doping appear to be different from
those occurring upon doping with Co and Fe cations 共i.e.,
unlike N, neither Co nor Fe become closed shell ions due to
CT effects兲. The apparently identical number of VOswe
measured upon N doping, yielding the same intensity for the
Ti 3d in-gap states 关Fig. 4共c兲of Ref. 2兴, is therefore ascribed
to CT of excess electrons to N, rather than to molecules
adsorbed on the surface. As reported in literature,8the Ti 3d
in-gap states are restored upon annealing in vacuum, a treat-
ment that produces VOs. We avoided this treatment, as the
increase of the VOdensity is known to change the saturation
magnetization of ferromagnetic TiO2−xsamples 共see, e.g.,
Ref. 9兲. With this constraint, contaminations due to air expo-
a兲Electronic mail: sangalet@dmf.bs.unicatt.it.
APPLIED PHYSICS LETTERS 97, 186102 共2010兲
0003-6951/2010/97共18兲/186102/2/$30.00 © 2010 American Institute of Physics97, 186102-1
sure could not be avoided, but the in-gap states detected in
our RESPES spectra 关Fig. 4共c兲of Ref. 2兴are extracted from
the difference between the on- and off-resonance spectra, and
therefore the possible effects of adsorbed molecules cancel
out to yield the observed in-gap, Ti 3d1, resonating states.
1Q.-H. Wu, Appl. Phys. Lett. 97, 186101 共2010兲.
2G. Drera, M. C. Mozzati, P. Galinetto, Y. Diaz-Fernandez, L. Malavasi, F.
Bondino, M. Malvestuto, and L. Sangaletti, Appl. Phys. Lett. 97, 012506
共2010兲.
3S. A. Chambers, S. H. Cheung, V. Shutthanandan, S. Thevuthasan, M. K.
Bowman, and A. G. Joly, Chem. Phys. 339,27共2007兲.
4S. H. Cheung, P. Nachimuthu, A. G. Joly, M. H. Engelhard, M. K. Bow-
man, and S. A. Chambers, Surf. Sci. 601,1754共2007兲.
5M. Batzill, E. H. Morales, and U. Diebold, Chem. Phys. 339,36共2007兲.
6C. S. Fadley, J. Electron Spectrosc. Relat. Phenom. 178-179,2共2010兲.
7See, e.g., B. H. Frazer, B. Gilbert, B. R. Sonderegger, and G. De Stasio,
Surf. Sci. 537, 161 共2003兲.
8M. Batzill, E. H. Morales, and U. Diebold, Phys. Rev. Lett. 96, 026103
共2006兲.
9A. K. Rumaiz, B. Alia, A. Ceylana, M. Boggsb, T. Beebeb, and S. I.
Shaha, Solid State Commun. 144, 334 共2007兲.
186102-2 Drera et al. Appl. Phys. Lett. 97, 186102 共2010兲