4366
Appl. Phys. Lett., Vol. 85, No. 19, 8 November 2004
Xu et al.
ing temperature, the luminescence peak position blueshifts
below 700 °C. This is in agreement with the observations by
Du et al.8 and Li et al.,9 who have attributed the blue emis-
sion band to the singly ionized oxygen vacancies (F+ centers)
and the blueshift to the gradual release of internal stress by
the volume expansion of aluminum during oxide formation
in PAMs. However, a clear redshift of the blue emission band
can be observed with a further increase of the annealing tem-
perature above 800 °C, which cannot be explained definitely
by the above mechanism.
the further decomposition of carboxylic impurities. Conse-
quently, the luminescence due to the transformed carboxylic
impurities (around 470 nm) dominates the PL band. The
domination will be much clearer at higher annealing tem-
peratures, leading to the observation of redshift of the blue
emission band with the increase of annealing temperatures.
Of course, due to the high annealing temperature, the PL
intensity of the transformed carboxylic impurities is much
weaker than that of the F+ centers in the PAMs annealed
ഛ700 °C. The PL results in Fig. 4 clearly reveal the evolu-
tion of the luminescence characteristics with the PAM struc-
tures, demonstrating the effects of the high-temperature an-
nealing.
In summary, PAMs with highly ordered pore arrays were
fabricated through a typical two-step anodizing electro-
chemical procedure. We have demonstrated the clear struc-
tural transition in PAMs from the amorphous (for as-
prepared and low-temperature annealed ones) to ␥- (under
the annealing temperatures of 800–950 °C) and then to ␣-
(with an annealing temperature of 1100 °C) Al2O3. The
structure has played an important role in the optical proper-
ties of PAMs. The formation of stable carboxylic impurities
has been observed in the crystallized PAMs. The proposal
that the blue emission band in PAMs originates from the
coactions of the singly ionized oxygen vacancies (F+ centers)
and the luminescent centers transformed from oxalic impuri-
ties can well explain the observed annealing temperature de-
pendent luminescence peak position and intensity.
On the other hand, Yamamoto et al.10 have proposed, as
early as in 1981, that the oxalic impurities incorporated in
the PAMs can be transformed into luminescent centers by a
high electric-field setup inside the pores or by the Joule heat-
ing, showing a blue PL band around 470 nm. The recent
studies of Gao et al.11 further suggest that the transformed
oxalic impurities (rather than the F+ centers) could be re-
sponsible for the blue PL emission in PAMs through X-band
electron paramagnetic resonance (EPR) and infrared trans-
mission measurements. According to that origin, the lumi-
nescence intensity would be enhanced with the increase of
the annealing temperature. Nevertheless, our PL results in
Fig. 4 do not support this conclusion at high annealing tem-
peratures ͑500–1100 °C͒.
Based on our self-consistent results in Figs. 2–4, we sug-
gest that the observed blue PL band could be attributed to the
coactions of both the F+ centers and the luminescent centers
transformed from oxalic impurities. From the overall anneal-
ing temperature results in Fig. 4, we can actually observe
two PL bands, centered at ϳ400 nm and ϳ470 nm, corre-
sponding to the luminescence structures of F+ centers and
transformed carboxylic impurities, respectively. In fact, these
two emission bands can be clearly deconvoluted by two
Gaussian functions in PL spectrum of the as-prepared PAM
(dotted curves in Fig. 4) with the integrated emission contri-
bution of ϳ50% to 50%. With the annealing temperature
up to 500 °C, it has been confirmed that the formed oxygen
vacancies (F+ centers) due to the oxidation of the remaining
aluminum in PAMs under an oxygen-poor atmosphere are
more than the annihilation of F+ centers caused by the com-
bination with oxygen diffusing into the PAMs,9 while the
oxalic impurities in alumina membranes decrease because of
the decomposition. In this case, the F+ centers are predomi-
nant in the PL spectra, resulting in the increase of the PL
intensity around 400 nm and a blueshift of the PL peak po-
sition.
This work was supported by the Natural Science Foun-
dation of China under Contract No. 10125416, the Shanghai
Major Project of 03DJ14003, and the Shanghai Nanotechnol-
ogy Fundamental Research Project of 0352nm013.
1H. Masuda and K. Fukuda, Science 268, 1466 (1995).
2X. Mei, D. Kim, H. E. Ruda, and Q. X. Guo, Appl. Phys. Lett. 81, 361
(2002).
3P. R. Evans, G. Yi, and W. Schwarzacher, Appl. Phys. Lett. 76, 481
(2000).
4K. Nielsch, F. Müller, A.-P. Li, and U. Gösele, Adv. Mater. (Weinheim,
Ger.) 12, 582 (2000).
5M. J. Zheng, L. D. Zhang, G. H. Li, X. Y. Zhang, and X. F. Wang, Appl.
Phys. Lett. 79, 839 (2001).
6J. S. Suh and J. S. Lee, Appl. Phys. Lett. 75, 2047 (1999).
7J. Liang, S.-K. Hong, N. Kouklin, R. Beresford, and J. M. Xu, Appl. Phys.
Lett. 83, 1752 (2003).
8Y. Du, W. L. Cai, C. M. Mo, J. Chen, L. D. Zhang, and X. G. Zhu, Appl.
Phys. Lett. 74, 2951 (1999).
Above 500 °C, previous experimental EPR results9
show that the F+ centers in PAMs begin to decrease with the
increase of the annealing temperature. The decomposition of
oxalic impurities keeps increasing with the annealing tem-
perature until the largest degree of decomposition into
trapped CO2 in the lattice at 700 °C, as shown in Fig. 3. As
a result, the intensity of the PL band decreases under the
annealing temperature above 500 °C. The F+ centers will
continue to decrease with the further increase of the anneal-
ing temperature ͑ജ800 °C͒, as demonstrated by the very
weak EPR signal observed in 800 °C annealed PAMs.9 In
contrast, stable carboxylic impurities are formed in the crys-
tallized alumina, we therefore expect a relative slowdown in
9Y. Li, G. H. Li, G. W. Meng, L. D. Zhang, and F. Phillipp, J. Phys.:
Condens. Matter 13, 2691 (2001).
10Y. Yamamoto, N. Baba, and S. Tajima, Nature (London) 289, 572 (1981).
11T. Gao, G. W. Meng, and L. D. Zhang, J. Phys.: Condens. Matter 15, 2071
(2003).
12R. Krishnan, S. Dash, C. B. Rao, R. V. S. Rao, A. K. Tyagi, and B. Raj,
Scr. Mater. 45, 693 (2001).
13A. Misra, H. D. Bist, M. S. Navati, R. K. Thareja, and J. Narayan, Mater.
Sci. Eng., B 79, 49 (2001).
14T. Rremkumar, S. Govindarajan, W.-P. Pan, and R. Xie, J. Therm. Anal.
Calorim. 74, 325 (2003).
15F. Y. Li, L. Zhang, and R. M. Metzger, Chem. Mater. 10, 2470 (1998).
16G. E. Thompson and G. C. Wood, Nature (London) 290, 230 (1981).
17A. Heilmann, P. Jutzi, A. Klipp, U. Kreibig, R. Neuendorf, T. Sawitowski,
and G. Schmid, Adv. Mater. (Weinheim, Ger.) 5, 10 (1998).