Upconversion luminescence in Er3+ doped and Er3+/Yb3+ codoped zirconia and hafnia nanocrystals excited at 980 nm Luis A. Gómez, Leonardo de S. Menezes, Cid B. de Araújo, Rogeria R. Gonçalves, Sidney J. L. Ribeiro et al. Citation: J. Appl. Phys. 107, 113508 (2010); doi: 10.1063/1.3428478 View online: http://dx.doi.org/10.1063/1.3428478 View Table of Contents: http://jap.aip.org/resource/1/JAPIAU/v107/i11 Published by the AIP Publishing LLC. Additional information on J. Appl. Phys. Journal Homepage: http://jap.aip.org/ Journal Information: http://jap.aip.org/about/about_the_journal Top downloads: http://jap.aip.org/features/most_downloaded Information for Authors: http://jap.aip.org/authors Downloaded 11 Jul 2013 to 186.217.234.138. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jap.aip.org/about/rights_and_permissions http://jap.aip.org/?ver=pdfcov http://oasc12039.247realmedia.com/RealMedia/ads/click_lx.ads/www.aip.org/pt/adcenter/pdfcover_test/L-37/932441298/x01/AIP-PT/JAP_CoverPg_0513/AAIDBI_ad.jpg/6c527a6a7131454a5049734141754f37?x http://jap.aip.org/search?sortby=newestdate&q=&searchzone=2&searchtype=searchin&faceted=faceted&key=AIP_ALL&possible1=Luis A. G�mez&possible1zone=author&alias=&displayid=AIP&ver=pdfcov http://jap.aip.org/search?sortby=newestdate&q=&searchzone=2&searchtype=searchin&faceted=faceted&key=AIP_ALL&possible1=Leonardo de S. Menezes&possible1zone=author&alias=&displayid=AIP&ver=pdfcov http://jap.aip.org/search?sortby=newestdate&q=&searchzone=2&searchtype=searchin&faceted=faceted&key=AIP_ALL&possible1=Cid B. de Ara�jo&possible1zone=author&alias=&displayid=AIP&ver=pdfcov http://jap.aip.org/search?sortby=newestdate&q=&searchzone=2&searchtype=searchin&faceted=faceted&key=AIP_ALL&possible1=Rogeria R. Gon�alves&possible1zone=author&alias=&displayid=AIP&ver=pdfcov http://jap.aip.org/search?sortby=newestdate&q=&searchzone=2&searchtype=searchin&faceted=faceted&key=AIP_ALL&possible1=Sidney J. L. Ribeiro&possible1zone=author&alias=&displayid=AIP&ver=pdfcov http://jap.aip.org/?ver=pdfcov http://link.aip.org/link/doi/10.1063/1.3428478?ver=pdfcov http://jap.aip.org/resource/1/JAPIAU/v107/i11?ver=pdfcov http://www.aip.org/?ver=pdfcov http://jap.aip.org/?ver=pdfcov http://jap.aip.org/about/about_the_journal?ver=pdfcov http://jap.aip.org/features/most_downloaded?ver=pdfcov http://jap.aip.org/authors?ver=pdfcov Upconversion luminescence in Er3+ doped and Er3+/Yb3+ codoped zirconia and hafnia nanocrystals excited at 980 nm Luis A. Gómez,1,a� Leonardo de S. Menezes,1 Cid B. de Araújo,1 Rogeria R. Gonçalves,2 Sidney J. L. Ribeiro,3 and Younes Messaddeq3 1Departamento de Física, Universidade Federal de Pernambuco, 50670-901 Recife-PE, Brazil 2Departmento de Química, FFCLRP-USP, 14040-90, Ribeirão Preto-SP, Brazil 3Instituto de Química, UNESP-Univ. Estadual Paulista, 14801-970 Araraquara-SP, Brazil �Received 29 December 2009; accepted 15 April 2010; published online 1 June 2010� Frequency upconversion �UC� luminescence in nanocrystalline zirconia �ZrO2� and hafnia �HfO2� doped with Er3+ and Yb3+ was studied under continuous-wave excitation at 980 nm. Samples of ZrO2:Er3+, ZrO2:Er3+ /Yb3+, and HfO2:Er3+ /Yb3+ were prepared by the sol-gel technique and characterized using x-ray diffraction and electron microscopy. A study of the infrared-to-green and infrared-to-red UC processes was performed including the analysis of the spectral and the temporal behavior. The mechanisms contributing to the UC luminescence were identified as excited state absorption and energy transfer among rare-earth ions. © 2010 American Institute of Physics. �doi:10.1063/1.3428478� I. INTRODUCTION Dielectric materials doped with rare earth �RE� ions have been attracting large attention due to the possibility of appli- cations in photonic devices.1,2 In particular, a phenomenon that has been receiving especial attention is the optical fre- quency upconversion �UC�, widely used to obtain visible lu- minescence using lasers operating in the infrared region for excitation of the samples. Generally, in order to observe ef- ficient UC emissions the host material should have low cut- off phonon frequency to allow large lifetime of the RE ions electronic levels.3–5 In addition, high UC efficiencies depend on the doping strategy used, like using codopants and/or me- tallic nanoparticles.6–8 For some applications, e.g., displays and sensors, nanostructured materials play a relevant role.9 An important photonic material for UC studies is zirco- nia �ZrO2� which has high refractive index and low phonon cutoff frequency.10 Previous studies show that ZrO2 nano- crystals �NC� doped with Er3+ or codoped with Er3+ /Yb3+ under infrared excitation is luminescent at �525 nm, �550 nm, and �650 nm, due to transitions from the Er3+ states 2H11/2, 4S3/2, and 4F9/2 to the ground state �4I15/2�.11 Although electronic 4f −4f transitions are forbidden for the electric dipole mechanism, this rule is broken when the ion occupies a lattice site without inversion symmetry. In such situation, the local field associated to the crystal structure plays an important role in the luminescence process. In the ZrO2 case, one observes monoclinic, tetragonal, and cubic phases, and since the first one is the less symmetric, one expects a larger luminescence efficiency when RE ions are hosted in such structure. Moreover, it is observed that for all ZrO2 phases, the more energetic phonon modes have ener- gies comparable to or smaller than 500 cm−1.12 Another system of interest is hafnia �HfO2�, which is a high-K material that is widely used as optical coating. HfO2 is isostructural with ZrO2, and when doped with RE it pre- sents similarities in the optical properties.13,14 In a recent work, Matarelli et al.15 have performed spectroscopic studies �Raman and luminescence� of Er3+-doped HfO2 NC in sili- cate glass ceramics obtained by sol-gel-techniques. They could determine also the structure of the grown HfO2 NC by performing X-ray diffraction �XRD� on the samples. For both matrices it is observed a transparency window over a wide wavelength range, as well as high refractive index and similar Raman spectra. UC emission at �650 nm under excitation at 980 nm was reported for Er3+ doped SiO2–HfO2 films.16 However, the UC phenomenon in RE doped HfO2 NC was not studied before. In this work, we present a study of the infrared-to-green and infrared-to-red UC processes for NC of ZrO2 and HfO2 singly doped with Er3+ and codoped by Er3+ /Yb3+ ions. The luminescence spectra of the samples was analyzed and the mechanisms contributing for the UC processes were identi- fied with basis on the UC signal dependence with laser in- tensity, RE ions concentration, and the temporal response of the samples. II. EXPERIMENTAL ZrO2 NC were prepared from a suspension of ZrOCl2 �zirconium oxichloride� in ethanol. The mixture of ZrOCl2, ethanol was refluxed for 1 h leading to stable NC suspension. Finally, erbium and ytterbium were added as chloride salts to obtain molar ratios of 0.5% and 1.0% for Er3+ and 10.0% for Yb3+. HfO2 NC were prepared in the same way from HfOCl2 and ethanol under reflux for 1h to get a fully transparent suspension. The HfO2 NC suspension was filtered through a 0.2 �m filter and then erbium and ytterbium were added as chloride salts to obtain molar ratios of 1.0% for Er3+ and 1.0% for Yb3+. a�Author to whom correspondence should be addressed. Present address: Escola Politécnica de Pernambuco—Universidade de Pernambuco, 50750- 470 Recife, PE, Brazil. Electronic mail: lagomezma@poli.br. JOURNAL OF APPLIED PHYSICS 107, 113508 �2010� 0021-8979/2010/107�11�/113508/5/$30.00 © 2010 American Institute of Physics107, 113508-1 Downloaded 11 Jul 2013 to 186.217.234.138. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jap.aip.org/about/rights_and_permissions http://dx.doi.org/10.1063/1.3428478 http://dx.doi.org/10.1063/1.3428478 http://dx.doi.org/10.1063/1.3428478 All suspensions were dried at room temperature and then annealed at 900 °C during 2 h. As a result of the synthesis procedure fine and homogenous powders are obtained. The preparation of the samples for the optical experiments was made by pressing the powders in a sample holder which allowed maintaining the same excitation and optical signal collection conditions for all samples. The composition and the sizes of the NC, as well as samples’ nomenclature, are summarized in Table I. The crystalline phases of the calcinated powders were identified by analysis of the XRD pattern. Morphology and particle sizes were determined using a transmission electron microscope. Laser excitation was made using a continuous wave �cw� diode laser ��30 mW� operating at 980 nm, which was modulated by a chopper at 288 Hz, located in the focus of a 1:1 telescope. After passing through the telescope, the excitation beam was focused by a 5 cm focal length lens on the samples. Fluorescence spectra were collected using a multimode optical fiber connected to a monochromator to which a pho- tomultiplier tube with gallium arsenide photocathode was at- tached. A digital oscilloscope was used to record the UC signals that were obtained with the samples at room tempera- ture. The excitation of the samples and the signal collection were carried out under the same conditions for all samples studied. III. RESULTS AND DISCUSSION The XRD patterns are shown in Fig. 1. Monoclinic phase was observed for the zirconia samples doped only with erbium �samples A and B� and tetragonal phase for the sample codoped with Er3+ /Yb3+ �sample C�. The HfO2 sample codoped with Er3+ /Yb3+ �sample D� shows tetrago- nal and monoclinic phases with the latter phase being domi- nant. A similar result was reported for sol-gel derived HfO2 films.17 Figure 2 shows the emission spectra under cw excitation at 980 nm. Green and red UC emissions are observed at �525, �550, and �650 nm, corresponding to transitions from states 2H11/2, 4S3/2, and 4H9/2 to the Er3+ ground state �4I15/2�, respectively. It is observed in Fig. 2 that the green emission increases with the Er3+ concentration but the red emission remains weak. The relevant energy levels for Er3+ and Yb3+, as well as the UC emissions, represented by the solid arrows are shown in Fig. 3. The UC emission in the spectral region from 520 to 570 nm is due to the successive optical transitions 4I15/2→ 4I11/2 and 4I11/2→ 4F7/2, followed by nonradiative decay to levels 2H11/2 and 4S3/2. Then, radia- tive transitions from these levels to the ground state multiplet originate the green emissions. This UC pathway was previ- ously identified for Er3+ hosted in different bulk crystals and glasses.18,19 The red emission is weak because the nonradia- tive transition �2H11/2 , 4S3/2�→ 4F9/2 is not efficient, since the energy gap of �2000 cm−1 would require the simultaneous creation of at least four phonons. The excitation pathway can be checked by measuring the dependence of the UC lumi- nescence intensity with the pump power. The results shown in Fig. 4 indicate a slope is �2.0 for samples A and B, corresponding to the absorption of two photons. The excitation pathway in samples C and D is different than in samples A and B due to the presence of Yb3+ ions. TABLE I. Characteristics of the studied samples, as well as their compositions �mol %�. Sample Matrix Structure NC average size �nm� Dopant concentration �mol %� Er3+ Yb3+ A ZrO2 Monoclinic 82 0.5 0.0 B ZrO2 Monoclinic 71 1.0 0.0 C ZrO2 Tetragonal 43 1.0 1.0 D HfO2 Monoclinic �dominant� and tetragonal 39 1.0 1.0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 D C B In te ns ity (a rb itr ar y un its ) 2θθθθ (degrees) A FIG. 1. �Color online� X-ray diffractograms. Samples: A �ZrO2:Er3+ 0.5%�, B �ZrO2:Er3+ 1.0%�, C �ZrO2:Er3+ 1.0%/Yb3+ 1.0%�, and D �HfO2:Er3+ 1.0%/Yb3+ 1.0%�. The data were displaced in the vertical scale for a better visualization. 500 520 540 560 580 640 660 680 0 20 40 60 80 100 120 4F9/2 4I15/2 4S3/2 4I15/2 In te ns ity (a rb itr ar y un its ) Wavelength (nm) A B C D x100 2H11/2 4I15/2 FIG. 2. �Color online� UC spectra emitted by the samples when excited at 980 nm. 113508-2 Gómez et al. J. Appl. Phys. 107, 113508 �2010� Downloaded 11 Jul 2013 to 186.217.234.138. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jap.aip.org/about/rights_and_permissions Because the laser wavelength is in resonance with the Yb3+ transition 2F5/2→ 2F7/2 and because its oscillator strength is greater than the Er3+ transition 4I15/2→ 4I11/2, the excitation of levels 2H11/2 and 4S3/2 is mainly due to energy transfer �ET� from Yb3+ to Er3+ ions. The ET efficiency depends on the relative concentration of Er3+ and Yb3+ and the lifetime of the levels participating in the process. Because the triva- lent RE ions substitute tetravalent Zr4+ and Hf4+ and because of the higher concentration in samples C and D, the ET be- tween Yb3+ and Er3+ ions may be efficient and thus the Er3+ levels’ lifetimes may change by large amount in comparison with samples A and B. Moreover, clusters of Yb3+ and Er3+ ions can be formed due to charge compensation. To obtain a better understanding of the UC behavior the time evolution of the luminescence signal was investigated chopping the laser beam. The results shown in Figs. 5 and 6 are summarized in Table II. A reference signal proportional to the laser intensity is presented in both figures to indicate the smallest rise-time and decay-time that could be measured after the interruption of the excitation laser by the chopper. The characteristic times given in Table II were determined using a simple exponential function matched between 33% and 1%, and 67% and 100% of the normalized laser intensity for decay and rise times, respectively. Concerning samples A and B, the UC signal rise time given in Table II ��0.18 ms� confirms that the UC pathway corresponds to two steps of one-photon absorption �4I15/2 → 4I11/2→ 4F7/2�. Transition 4I11/2→ 4F7/2 is the so-called ex- cited state absorption �ESA�. The decay signals illustrated in Fig. 5�a� show that the characteristic decay time for sample A is smaller than that for sample B, meaning that there are two dynamical regimes for these samples. The two regimes can be understood considering the role played by the cross- relaxation �CR� corresponding to ��4S3/2→ 4I9/2�+ �4I15/2 → 4I13/2�� and ET ��4I11/2→ 4I15/2�+ �4I11/2→ 4S3/2�� pro- cesses. In order to describe this behavior, an appropriate rate equation model was derived as follows: 0 5000 10000 15000 20000 25000 CR CR 98 0 nm E S A 2F 5/2 2F 7/2 65 0 nm 55 0 nm 52 5 nm 4F 3/2 4F 7/22H 11/24S 3/2 4F 9/2 4I 9/2 4I 11/2 4I 13/2 4I 15/2 E n er g y (c m -1 ) Er3+ Yb3+ ET 98 0 nm E S A FIG. 3. �Color online� Energy levels scheme for the Er3+ and Yb3+ ions. The solid arrows represent photons absorbed or emitted by the ions. Dashed and curly arrows represent cross relaxation and ET processes, respectively. 1 10 0.1 1 10 100 1 10 In te ns ity (a rb itr ar y u ni ts ) Pump Power (mW) Slopes A 1.9 B 1.8 C 1.4 D 2.0 (a) Slope C 1.4 (b) FIG. 4. �Color online� Dependence of the UC intensity with the pump power: �a� emission at 550 nm �b� emission at 650 nm. 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.01 0.1 1 (a) Laser B A N o rm al iz ed in te n si ty (a rb itr ar y un its ) Time (ms) 550 nm -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.1 1 N or m al iz ed in te ns it y (a rb itr ar y un its ) (b)550 nm Time (ms) Laser C D -0.6 -0.4 -0.2 0.0 0.1 1 N or m al iz ed in te ns ity (a rb itr ar y un it s) (c)650 nm Time (ms) Laser C FIG. 5. �Color online� �a� Decay of the luminescence signal originated from the 4S3/2 level �550 nm�. Sample A �triangles� and sample B �circles�. The scattered laser signal is represented by the stars. �b� Rise time of the 550 nm signal for samples C �diamonds� and D �squares�. The dynamics is clearly limited by the temporal resolution of the apparatus in the case of sample D. �c� Rise time of the 650 nm signal for sample C. 113508-3 Gómez et al. J. Appl. Phys. 107, 113508 �2010� Downloaded 11 Jul 2013 to 186.217.234.138. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jap.aip.org/about/rights_and_permissions ṅ1 = − R1n1 − Wn1n5 + W1n3 2 + �51n5 + �41n4 + �31n3 + �21n2, ṅ2 = Wn1n5 + �52n5 + �42n4 + �32n3 − �21n2, ṅ3 = R1n1 − R2n3 − 2W1n3 2 + �53n5 + �43n4 − ��32 + �31�n3, ṅ4 = Wn1n5 + �54n5 − ��43 + �42 + �41�n4, ṅ5 = R2n3 − Wn1n5 + W1n3 2 − ��54 + �53 + �52 + �51�n5, n1 + n2 + n3 + n4 + n5 = 1, �1� where the index i�i=1–5� is related to the Er3+ energy levels 4I15/2, 4I13/2, 4I11/2, 4I9/2, �4F7/2 , 2H11/2 , 4S3/2�, respectively. The pump rates are Ri=�iI /h�, where �i is the absorption cross section, I is the intensity of the pump laser on the sample, h� is the photon energy, �i-j�i , j=1–5� are the radia- tive plus nonradiative decay rates from level i to level j, W is the CR rate, and W1 is the ET rate. An analysis of the rate equation system for level 5, which originates the green emission from levels 2H11/2 and 4S3/2, shows the UC intensity dependence with the CR rate, ET probability and the radiative and nonradiative decay rates. A nonsteady state solution for the proposed rate equa- tion model may describe the results shown in Fig. 5�a�. How- ever, as the set of Eqs. �1� contains the product between the populations of two distinct levels, an analytical solution is given only assuming low population of level 5, where the product n1n5�n5. However, since this approximation is not enough to describe the luminescence signal from level 5, an analysis was made to investigate the role played by the CR, ET, and radiative and nonradiative decays for n5. The radia- tive decay rate was taken from the experimental data for ZrO2 bulk crystal and the nonradiative decay rate was esti- mated using the energy gap law with parameters of ZBLAN glass20,21 that are not expect to be much different than for ZrO2 and HfO2. The estimated value of the total transition rate from level 5 is 1.6�104 s−1 corresponding to a decay time of 62.5 �s. Since this time is very short and does not find correspondence to the measured data, the dynamical re- sponse observed in Fig. 5�a� has to be attributed to the CR and ET. During the time interval that the chopper is blocking the laser beam, the pump rate decreases and a monoexponen- tial decay of the luminescence is observed remaining frac- tions of milliseconds after the beam is blocked. As the laser wavelength is resonant with level 3, we expect that the ET process feeds efficiently level 5, even at low pump rate, and the CR removes this population efficiently. This process con- tinues until the population of level 3 starts to decrease. In this case, ET is not an efficient channel to feed level 5 and the CR is dominant, contributing to the long decay time of the green emission. It means that that the temporal response observed in Fig. 5�a� is a competitive process between the CR and ET and depends strongly of the level 3 population. The fluorescence spectra of ZrO2 codoped with 1.0% Er3+ and 1.0% Yb3+ �sample C� and HfO2 codoped with 1.0% Er3+ and 1.0% Yb3+ grains �sample D� are also shown in Fig. 2. For these two samples, weak red emission is ob- served at �650 nm. As is illustrated in Fig. 3, besides the direct absorption of two laser photons starting from the ground state �4I15/2� to state 4F7/2, another mechanism con- tributing to the green and red emissions is the ET from an excited neighbor Yb3+ ion to the Er3+ ion. For the green emission, the process occurs in two steps. In the first step the 4I11/2 level is populated via ground state absorption �4I15/2 +h�laser→ 4I11/2� and ET via �2F5/2�Yb3+�+ 4I15/2�Er3+� → 2F7/2�Yb3+�+ 4I11/2�Er3+��. In the second step the Er3+ ions in the 4I11/2 level are excited to the 4F7/2 level via ESA �4I11/2+h�laser→ 4F7/2� and ET �2F5/2�Yb3+�+ 4I11/2�Er3+� 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0.01 0.1 1 N or m al iz ed in te ns ity (a rb itr ar y un its ) (a) 550 nm Time (ms) Laser C D 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.01 0.1 1 N or m al iz ed in te ns ity (a rb itr ar y un it s) (b) 650 nm Time (ms) Laser C FIG. 6. �Color online� �a� Decay of the fluorescence signal at 550 nm. Sample C �diamonds� and sample D �squares�. Two different regimes are observed: a fast relaxation with a characteristic time of 0.21 ms and a slow decay with characteristic time of 0.70 ms. �b� Fluorescence decay originated from the 4F9/2 level for sample C �diamonds�—decay time: 0.90 ms. TABLE II. Summary of the rise-times and decay times ��5%� for Er3+ green and red emissions. Sample Emission at 550 nm Emission at 650 nm Rise time �ms� Decay time �ms� Rise time �ms� Decay time �ms� A �0.18 Short times 0.29 ¯ ¯ Long times 0.13 B �0.18 0.31 ¯ ¯ C 1.48 0.70 0.40 0.90 D �0.18 0.21 ¯ ¯ 113508-4 Gómez et al. J. Appl. Phys. 107, 113508 �2010� Downloaded 11 Jul 2013 to 186.217.234.138. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jap.aip.org/about/rights_and_permissions → 2F7/2�Yb3+�+ 4F7/2�Er3+��. Nonradiative relaxation from the 4F7/2 to 2H11/2 and 4S3/2, and radiative decay from those levels to the 4I15/2 will give origin to the emissions centered at 525 and 550 nm. The red emission is due to the radiative relaxation from the 4F9/2 to 4I15/2. The 4F9/2 level is popu- lated via nonradiative relaxation from upper levels, via ESA �4I13/2+h�laser→ 4F9/2� and ET �2F5/2�Yb3+�+ 4I13/2�Er3+� → 2F7/2�Yb3+�+ 4F9/2�Er3+��. The dependence of the fluorescence intensity at 550 and 650 nm with the input power is shown in Figs. 4�a� and 4�b�. Slopes of 1.4 and 2.0 for samples C and D for the green emission, and 1.4 for sample C for the red emission were obtained. These results indicate that for sample D, two pho- tons of 980 nm are involved in the excitation of the Er3+ ions and the ET mechanism from Yb3+ is poorly efficient, that is evidenced observing the fast rise time of the 4S3/2 level, as is shown in the Fig. 5�b�. The decay time of 0.21 ms obtained from Fig. 6�a� is similar to the decay time obtained for sample B, then the mechanism of UC should be explained by CR and ET between Er3+ ions. The behavior of sample C is different. The influence of ET from Yb3+ to Er3+ ions is clearly observed in the slopes indicated in Fig. 4 and the different rise times shown in Figs. 5�b� and 5�c�. These times are 1.48 and 0.40 ms for the green and red signals, respectively, indicating that the ET process �2F5/2�Yb3+�+ 4I11/2�Er3+�→ 2F7/2�Yb3+�+ 4F7/2�Er3+�� is more efficient than the process �2F5/2�Yb3+�+ 4I13/2�Er3+� → 2F7/2�Yb3+�+ 4F9/2�Er3+��. On the other hand, the temporal response of the green emission shown in Fig. 6�a� is affected by CR, ET between Er3+ ions, and ET between Er3+ and Yb3+ ions. The decay time of the signal from sample C is larger than for sample B, indicating that other mechanism like the ET �2F5/2�Yb3+�+ 4I11/2�Er3+�→ 2F7/2�Yb3+� + 4F7/2�Er3+�� is contributing to increase the population of the level 5. In order to understand the long lifetime of the level 2F5/2�Yb3+� we have to consider contribution from energy back-transfer as described by �4S3/2�Er3+�+ 2F7/2�Yb3+� → 4I13/2�Er3+�+ 2F5/2�Yb3+��. The measured decay time of the red UC signal was 0.90 ms. The contribution of the radiative and nonradiative decay rates from the level 4F9/2 was obtained following the same procedure for the level 4S3/2. In this case, the total transition rate was 2.4�103 s−1, corresponding to a decay time of 0.40 ms, indicating that the contribution of upper levels, ESA �4I13/2+h�laser→ 4F9/2�, and ET �2F5/2�Yb3+�+ 4I13/2�Er3+� → 2F7/2�Yb3+�+ 4F9/2�Er3+�� are important to increase the population of the level 4F9/2. A balance between the mechanisms described above is determining the temporal response shown in the Figs. 6�a� and 6�b�. The slope of 1.4 observed for the green and red emissions indicates saturation of the 4I13/2 level attributed to the energy back-transfer mechanism �4S3/2�Er3+� + 2F7/2�Yb3+�→ 4I13/2�Er3+�+ 2F5/2�Yb3+��.22 IV. CONCLUSIONS The infrared-to-green and infrared-to-red frequency UC processes were studied in ZrO2 and HfO2 NC doped with Er3+ and Er3+ /Yb3+. Red UC emission was observed only for the samples with Yb3+ as codopant. The decay times for ZrO2 doped only with erbium are different due to changes in the crystal size and doping concentration. Although HfO2 and ZrO2 NC are isostructural, a comparison between HfO2 and ZrO2 NC codoped with Er3+ /Yb3+ reveals that the inclu- sion of Yb3+ strongly affects the crystalline structure and the luminescence properties in the visible spectrum of the hafnia NC. For the codoped samples studied, the hafnia matrix pre- sents four times higher UC efficiency in the green region than zirconia, while the latter shows two times higher UC efficiency than the hafnia in the red region. However, the signal intensity in this region is �50 times weaker than that in the green. ACKNOWLEDGMENTS We acknowledge the financial support from the Brazilian agencies Conselho Nacional de Desenvolvimento Científico e Tecnológico �CNPq� and Fundação de Amparo à Ciência e Tecnologia de Pernambuco �FACEPE�. This work was per- formed under the Nanophotonics Network Project and INCT de Fotônica Project �CNPq�. 1Erbium-doped Fiber Amplifiers, edited by E. Desurvire �Wiley, New York, 1994�. 2M. Yamane and Y. Asahara, Glasses for Photonics �Cambridge University Press, Cambridge, 2000�. 3T. T. Basiev, Y. V. Orlovskii, K. K. Pukhov, and F. Auzel, Laser Phys. 7, 1139 �1997�. 4C. B. de Araújo G. S. Maciel, L. de S. Menezes, N. Rakov, E. L. Falcão- Filho, V. A. Jerez, and Y. Messaddeq, C. R. Chim. 5, 885 �2002�. 5B. M. Tissue, Chem. Mater. 10, 2837 �1998�. 6W. Lozano B., C. B. de Araújo, C. Egalon, A. S. L. Gomes, B. J. Costa, and Y. Messaddeq, Opt. Commun. 153, 271 �1998�. 7L. P. Naranjo, C. B. de Araújo, O. L. Malta, P. A. S. Cruz, and L. R. P. Kassab, Appl. Phys. Lett. 87, 241914 �2005�. 8T. A. A. Assumpção, D. M. da Silva, L. R. P. Kassab, and C. B. de Araújo, J. Appl. Phys. 106, 063522 �2009�. 9T. Ohgaki, A. Higashida, K. Soga, and A. Yasumori, J. Electrochem. Soc. 154, J163 �2007�. 10R. Reisfeld, M. Zelner, and A. Patra, J. Alloys Compd. 300, 147 �2000�. 11A. Patra, C. S. Friend, R. Kapoor, and P. Prasad, J. Phys. Chem. B 106, 1909 �2002�. 12X. Zhao and D. Vanderbilt, Phys. Rev. B 65, 075105 �2002�. 13C. Carlone, Phys. Rev. B 45, 2079 �1992�. 14P. E. Quintard, P. Barbéris, A. P. Mirgorodsky, and T. Merle-Méjean, J. Am. Ceram. Soc. 85, 1745 �2002�. 15M. Mattarelli, M. Montagna, F. Rossi, C. Tosello, N. D. Afify, M. Beti- nelli, A. Speghini, C. Armellini, Y. Jestin, F. Rocca, and S. Gialanella, Opt. Mater. 31, 1362 �2009�. 16R. R. Gonçalves, G. Carturan, M. Montagna, M. Ferrari, L. Zampedri, S. Pelli, G. C. Righini, S. J. L. Ribeiro, and Y. Messaddeq, Opt. Mater. 25, 131 �2004�. 17T. Nishide, S. Honda, M. Matsuura, and M. Ide, Thin Solid Films 371, 61 �2000�. 18F. Vetrone, J. Boyer, J. A. Capobianco, A. Speghini, and M. Bettinelli, Appl. Phys. Lett. 80, 1752 �2002�. 19C. B. de Araújo, L. S. Menezes, G. S. Maciel, L. H. Acioli, A. S. L. Gomes, Y. Messaddeq, A. Florez, and M. A. Aegerter, Appl. Phys. Lett. 68, 602 �1996�. 20E. Greenberg, G. Katz, R. Reisfeld, N. Spector, R. C. Marshall, B. Ben- dow, and R. N. Brown, J. Chem. Phys. 77, 4797 �1982�. 21M. D. Shinn, W. A. Sibley, M. G. Drexhage, and R. N. Brown, Phys. Rev. B 27, 6635 �1983�. 22G. Chen, G. Somesfalean, Y. Liu, Z. Zhang, Q. Sun, and F. Wang, Phys. Rev. B 75, 195204 �2007�. 113508-5 Gómez et al. J. Appl. Phys. 107, 113508 �2010� Downloaded 11 Jul 2013 to 186.217.234.138. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jap.aip.org/about/rights_and_permissions http://dx.doi.org/10.1016/S1631-0748(02)01460-1 http://dx.doi.org/10.1021/cm9802245 http://dx.doi.org/10.1016/S0030-4018(98)00268-5 http://dx.doi.org/10.1063/1.2143135 http://dx.doi.org/10.1063/1.3211300 http://dx.doi.org/10.1149/1.2712834 http://dx.doi.org/10.1021/jp013576z http://dx.doi.org/10.1103/PhysRevB.65.075105 http://dx.doi.org/10.1103/PhysRevB.45.2079 http://dx.doi.org/10.1016/j.optmat.2008.10.027 http://dx.doi.org/10.1016/S0925-3467(03)00261-1 http://dx.doi.org/10.1016/S0040-6090(00)01010-5 http://dx.doi.org/10.1063/1.1458073 http://dx.doi.org/10.1063/1.116481 http://dx.doi.org/10.1063/1.443720 http://dx.doi.org/10.1103/PhysRevB.27.6635 http://dx.doi.org/10.1103/PhysRevB.27.6635 http://dx.doi.org/10.1103/PhysRevB.75.195204 http://dx.doi.org/10.1103/PhysRevB.75.195204