Toward ordered mesoporous rare-earth sesquioxide thin films via polymer templating: high temperature stable C-type Er₂O₃ with finely-tunable crystallite sizes

Abstract

Facile polymer templating enables the production of cubic C-type Er₂O₃ thin films with a 3D honeycomb structure of 15–17 nm diameter pores and crystalline domain sizes ranging from 3–4 nm at 600 °C to ∼13 nm at 900 °C. These novel nanomaterials not only have a thermally stable and robust framework but are also well-defined at both the nano- and microscale.

---

Mesoporous metal oxides with nanocrystalline walls and uniform pores have been the focus of much research over the past two decades. Two of the main reasons for this “lasting trend” are that (1) physical properties are often altered due to the high surface area-to-volume ratio and that (2) many materials have seen benefits from structuring at the nanometre length scale in that they were able to outperform their bulk counterparts. While such oxide materials could certainly pave the way for the development of novel “nano-enabled” devices and technologies, their production remains challenging.

In recent years, it has been shown that polymer templating provides access both to mesoporous metal oxides and to nanocomposites with long-range periodicities.¹ Their formation relies on the coassembly of either (molecular) sol–gel precursors or preformed building blocks with an amphiphilic polymer, which acts as the structure-directing agent.² Materials in thin film format can be produced by the same methods but using an evaporation-induced self-assembly (EISA) process.³ This process was introduced by Brinker and coworkers more than 10 years ago. Since then, numerous important metal oxide films, including TiO₂, α-Fe₂O₃ and so forth, with mesoporous morphologies and thicknesses ranging from a few nanometres to several hundreds of nanometres have been reported and tested for various potential applications.⁴ However, to date the major weakness of EISA and related sol–gel routes has been the difficulty of controlling the conversion of the initially amorphous (glassy) solids to highly crystalline compounds while retaining both nanoscale periodicity and porosity. For this reason, the majority of the polymer-templated materials described in the literature exhibit only a partially crystalline framework.

In the present work, we report the polymer templating synthesis and material properties of 3D mesoporous erbium oxide (C-type Er₂O₃) thin films with open pores averaging 16 nm in diameter, tunable domain sizes of the crystallites and a high thermal stability. Er₂O₃ was chosen as an example of a rare-earth (or lanthanide) sesquioxide because, to our knowledge, only a few such materials with both a well-defined pore structure and nanocrystalline walls have been reported so far.⁵ However, we note that the synthesis procedure shown here is not only simple but also applicable to many other rare-earth sesquioxides in the form of thin films.

In a typical synthesis, both hydrated erbium chloride and a KLE-type⁶ (poly(ethylene-co-butylene)-block-poly(ethylene oxide)) diblock copolymer are dissolved in a mixed solvent of ethanol, 2-methoxyethanol, and glacial acetic acid (see the experimental procedure in the ESI† for details). Thin films can be produced via dip-coating on virtually any polar substrate. During this work we found that the use of glacial acetic acid is crucial to achieving mesoporous materials. At this point, we believe that it is capable of suppressing the recrystallization of erbium chloride species presumably by complexation. If no glacial acidic acid is used, the initially clear films became opaque during evaporation of the volatile constituents at annealing temperatures below 150 °C.

The pore structure of the KLE-templated Er₂O₃ thin films at the top surface and in the bulk was examined by scanning electron microscopy (SEM) and transmission electron microscopy (TEM). Fig. 1a–c show both the top view and cross-sectional SEM images of a 600 nm thick film after thermal annealing at 800 °C (see also Fig. S1 in the ESI†). The low- and high-magnification top view SEM images in panels (a) and (b) reveal a distorted cubic network of 15–17 nm diameter pores and further indicate that the pore cavities at the solid–air interface are open. In addition, it can be clearly seen that the oxide material studied in this work is crack-free at the micrometre level and that the walls have a thickness of 10–14 nm. The latter observation helps explain why the nanoscale structure is preserved after crystallization.⁷ From cross-sectional SEM in panel (c) we are able to establish that the pore network observed at the top surface persists throughout the bulk of the films.

Fig. 1 The morphology of KLE-templated C-type Er₂O₃ thin films with nanocrystalline walls after thermal annealing at 800 °C. (a, b) Top view SEM images at different magnifications. (c) A cross-sectional SEM image.

Fig. 2a–d show the morphology data from the TEM studies (see also Fig. S2 in the ESI†). These data are generally consistent with the SEM observations. Panels (a–c) represent bright-field TEM images at different magnifications, which reveal an interconnected pore stucture for samples heated at 800 °C. From these images, it is also evident that the pore walls consist of randomly oriented, sub-10 nm crystallites. This result is further verified by both high-resolution TEM and selected-area electron diffraction (SAED). The HRTEM image in panel (d) shows the (222) lattice planes of a single crystallite and the SAED pattern in panel (e) displays Debye-Scherrer rings. Lattice spacings calculated from the diffraction rings match with the JCPDS reference card no. 8-0050 for cubic C-type Er₂O₃ (see Fig. S3 in the ESI† for details).⁸

Fig. 2 The nanoscale structure of KLE-templated C-type Er₂O₃ thin films with nanocrystalline walls after thermal annealing at 800 °C. (a–c) Bright-field TEM images at different magnifications showing open pore cavities averaging 16 nm in diameter. (d) HRTEM image showing the (222) lattice planes of cubic Er₂O₃. (e) The SAED pattern of the same film.

Overall, the electron microscopy data in Fig. 1 and 2, S1 and S2† collectively demonstrate that Er₂O₃ can be templated to produce mesoporous thin films that are not only well-defined but also thermally very stable. It should be noted that these features are prerequisites for many applications.

The nanoscale structure of the KLE-templated Er₂O₃ thin films was also probed using synchrotron-based grazing incidence small-angle X-ray scattering (GISAXS). Panels (a) and (b) of Fig. 3 show that amorphous material produces patterns with distinct maxima. These maxima are characteristic of a distorted cubic pore structure. The elliptical shape of both patterns further indicates that the samples undergo significant lattice contraction in the direction normal to the substrate (approx. 70% after 500 °C). In contrast, the in-plane contraction is negligible because the films are bound to the substrate. At first glance, such contraction seems rather large but is not surprising in view of the sol–gel processing. From the GISAXS data in panels (c) and (d) it can be observed that annealing temperatures above 700 °C lead to the complete loss of out-of-plane scattering. This loss is a direct result of both the form anisotropy of the thin films with only a few repeat units normal to the plane of the substrate and the crystallization process, which begins at approximately 580 °C (see Fig. 4b). However, the results with GISAXS confirm that the distorted 3D honeycomb network of pores is retained up to annealing temperatures as high as 900 °C. This, in turn, implies that the sol–gel derived Er₂O₃ material can readily accommodate the crystallites that form at the onset of crystallization and that the walls are sufficiently thick to allow for the uniform growth of the crystalline phase without severely distorting the pore-solid architecture.

Fig. 3 GISAXS patterns at an angle of incidence β = 0.2° obtained on a KLE-templated Er₂O₃ thin film heated at 300 °C for 12 h (a) and 500 °C (b), 700 °C (c) and 900 °C (d) for 5 min, respectively.

Fig. 4 (a) N₂-adsorption–desorption isotherms of 600 nm thick, KLE-templated C-type Er₂O₃ films with a total area of 50 cm² heated at 800 °C. (b) XRD patterns recorded after thermal annealing at 600 °C (red), 700 °C (blue), 800 °C (green) and 900 °C (brown). (c, d) XPS detail spectra of the Er 4d and O 1s core level regions, respectively. Solid lines in red are fits to the data. The Shirley method was applied to subtract the background.

The porous properties were further analyzed by N₂-physisorption. Fig. 4a shows typical isotherms for 600 nm thick films heated at 800 °C. These data provide a Brunauer-Emmett-Teller (BET) surface area of ∼230 m² cm⁻³ and a total porosity of ∼53%, which gives ample evidence for the accessibility of the pore cavities in the interior. Higher annealing temperatures lead to a reduction in the BET surface area while the porosity remains virtually unaltered (see Fig. S4 in the ESI†).

Apart from the nanoscale structure, we also examined the optical properties because Er₂O₃ and other rare-earth sesquioxides hold promise as potential materials for antireflective coatings and high-k gate dielectrics.⁹ For both of these applications, knowledge of the band gap energy, E_g, is indispensable. Optical absorption measurements (see Fig. S5 in the ESI†) indicate a band gap at room temperature of (5.35 ± 0.05) eV due to oxygen–metal charge transfer (i.e., excitation from the O 2p conduction band to the Er 5d valence band). This value is in agreement with measured ones for single- and polycrystalline Er₂O₃ materials.¹⁰

Lastly, a series of X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS) measurements were carried out to more fully characterize the KLE-templated Er₂O₃ thin films, in particular the crystallization behavior and the electronic bonding configuration. Fig. 4b shows XRD patterns as a function of annealing temperature. From these data it is apparent that erbium oxide crystallizes in a purely cubic C-type rare-earth sesquioxide structure, otherwise referred to as bixbyite (see also the structure visualization in Fig. S3 in the ESI†).⁸ In this fluorite-derived structure with space group (T⁷_h) erbium atoms are located on two crystallographically non-equivalent six-fold positions to oxygen, namely 8b and 24d Wyckoff sites with S₆ and C₂ symmetries, respectively. In contrast, all of the oxygen atoms are equivalent and occupy the 48e sites (C₁). The crystallization begins at approximately 580 °C in a narrow temperature interval. Applying the Scherrer equation to the line broadening of the most intense (222) and (400) peaks at 2θ = 29.33° and 33.97°, respectively, provides an average crystalline domain size of 3.5 nm after 600 °C. This dimension is assumed to be the stable crystallite size. The Scherrer analysis further reveals that the nanocrystalline C-type Er₂O₃ domains in the pore walls gradually grow with increasing annealing temperatures. Interestingly, we find an almost perfect linear relationship between the crystallite size and temperature (see Fig. S6 in the ESI†); the average domain sizes of the crystallites can be finely-tuned over a range from 3–4 nm at 600 °C to ∼13 nm at 900 °C. This finding demonstrates that the dimensions of both the crystallites and the inorganic walls are of the same order after thermal annealing at 900 °C, which therefore provides an explanation for the fact that no significant restructuring of the 3D pore structure is observed up to such high temperatures.⁷

Panels (c) and (d) of Fig. 4 show XPS detailed spectra of the Er 4d and O 1s core level regions of KLE-templated Er₂O₃ heated at 800 °C in air. A Shirley background was subtracted from both spectra and the peaks were fitted using Gaussian-Lorentzian functions. The Er 4d spectrum contains a doublet due to spin–orbit splitting with binding energies of (169.26 ± 0.05 eV) and (167.25 ± 0.05 eV) for the 4d_3/2 and 4d_5/2 core levels, respectively. These positions are in agreement with reported measured values for Er₂O₃.¹¹ The broad peak on the higher energy side of the Er 4d lines is due to plasmon loss. In contrast, deconvolution of the O 1s spectrum indicates two different oxygen binding states. The main peak at (529.21 ± 0.05 eV) can be attributed to oxygen in Er₂O₃, while the shoulder peak at (531.28 ± 0.05) eV can be assigned to Er–OH.¹¹ The presence of hydrated erbium oxide at the top surface is presumably a result of interactions with ambient moisture.

Overall, the XRD and XPS data lead us to conclude that (1) larger fluctuations in composition can be ruled out and that (2) all characteristics of the material studied here can be associated with the cubic C-type phase of erbium sesquioxide.

In summary, this research shows that mesoporous thin films of highly crystalline Er₂O₃ can be fabricated by facile coassembly strategies using a KLE-type diblock copolymer as the structure-directing agent. These novel nanomaterials not only have a very robust framework but are also well-defined at both the nanoscale and the microscale after thermal annealing at 800 °C. Future work will be dedicated to extending our synthesis method to other rare-earth sesquioxides and to the potential applications of these materials.

Acknowledgements

The authors thank Jan Perlich and Thomas Leichtweiss for assistance in materials characterization and HASYLAB at DESY for beamtime. Financial support by the German Research Foundation (DFG, grant no. BR 3499/3-1) is gratefully acknowledged.

References

1. (a) P. D. Yang, D. Y. Zhao, D. I. Margolese, B. F. Chmelka and G. D. Stucky, Nature, 1998, 396, 152; (b) S. Y. Choi, M. Mamak, N. Coombs, N. Chopra and G. A. Ozin, Adv. Funct. Mater., 2004, 14, 335; (c) B. L. Kirsch, E. K. Richman, A. E. Riley and S. H. Tolbert, J. Phys. Chem. B, 2004, 108, 12698; (d) D. Grosso, C. Boissiere, B. Smarsly, T. Brezesinski, N. Pinna, P. A. Albouy, H. Amenitsch, M. Antonietti and C. Sanchez, Nat. Mater., 2004, 3, 787; (e) F. Hoffmann, M. Cornelius, J. Morell and M. Froba, Angew. Chem., Int. Ed., 2006, 45, 3216; (f) C. Sanchez, B. Boissiere, D. Grosso, C. Laberty and L. Nicole, Chem. Mater., 2008, 20, 682; (g) C. Sanchez, P. Belleville, M. Popall and L. Nicole, Chem. Soc. Rev., 2011, 40, 696.
2. (a) C. T. Kresge, M. E. Leonowicz, W. J. Roth, J. C. Vartuli and J. S. Beck, Nature, 1992, 359, 710; (b) H. Yang, A. Kuperman, N. Coombs, S. Mamiche-Afara and G. A. Ozin, Nature, 1996, 379, 703; (c) U. Ciesla and F. Schuth, Microporous Mesoporous Mater., 1999, 27, 131; (d) G. J. D. Soler-illia, C. Sanchez, B. Lebeau and J. Patarin, Chem. Rev., 2002, 102, 4093; (e) A. Corma, P. Atienzar, H. Garcia and J. Y. Chane-Ching, Nat. Mater., 2004, 3, 394; (f) J. H. Ba, J. Polleux, M. Antonietti and N. Niederberger, Adv. Mater., 2005, 17, 2509.
3. (a) M. Ogawa, Langmuir, 1997, 13, 1853; (b) Y. F. Lu, R. Ganguli, C. A. Drewien, M. T. Anderson, C. J. Brinker, W. L. Gong, Y. X. Guo, H. Soyez, B. Dunn, M. H. Huang and J. I. Zink, Nature, 1997, 389, 364; (c) C. J. Brinker, Y. F. Lu, A. Sellinger and H. Y. Fan, Adv. Mater., 1999, 11, 579.
4. (a) S. Y. Choi, M. Mamak, N. Coombs, N. Chopra and G. A. Ozin, Nano Lett., 2004, 4, 1231; (b) J. C. Yu, X. C. Wang and X. Z. Fu, Chem. Mater., 2004, 16, 1523; (c) G. De, R. Kohn, G. Xomeritakis and C. J. Brinker, Chem. Commun., 2007, 1840; (d) I. Bannat, K. Wessels, T. Oekermann, J. Rathousky, D. Bahnemann and M. Wark, Chem. Mater., 2009, 21, 1645; (e) K. Brezesinski, R. Ostermann, P. Hartmann, J. Perlich and T. Brezesinski, Chem. Mater., 2010, 22, 3079; (f) J. M. Szeifert, J. M. Feckl, D. Fattakhova-Rohlfing, Y. J. Liu, V. Kalousek, J. Rathousky and T. Bein, J. Am. Chem. Soc., 2010, 132, 12605; (g) K. Brezesinski, J. Haetge, J. Wang, S. Mascotto, C. Reitz, A. Rein, S. H. Tolbert, J. Perlich, B. Dunn and T. Brezesinski, Small, 2011, 7, 407; (h) E. Ortel, T. Reier, P. Strasser and R. Kraehnert, Chem. Mater., 2011, 23, 3201; (i) J. M. Wu, I. Djerdj, T. von Graberg and B. M. Smarsly, Beilstein J. Nanotechnol., 2012, 3, 123.
5. (a) M. Lundberg, B. Skarman, F. Cesar and L. R. Wallenberg, Microporous Mesoporous Mater., 2002, 54, 97; (b) T. Brezesinski, M. Antonietti, M. Groenewolt, N. Pinna and B. Smarsly, New J. Chem., 2005, 29, 237; (c) Y. Castro, B. Julian-Lopez, C. Boissiere, B. Viana, D. Grosso and C. Sanchez, Microporous Mesoporous Mater., 2007, 103, 273; (d) J. Hierso, O. Sel, A. Ringuede, C. Laberty-Robert, L. Bianchi, D. Grosso and C. Sanchez, Chem. Mater., 2009, 21, 2184; (e) T. Brezesinski, J. Wang, R. Senter, K. Brezesinski, B. Dunn and S. H. Tolbert, ACS Nano, 2010, 4, 967.
6. A. Thomas, H. Schlaad, B. Smarsly and M. Antonietti, Langmuir, 2003, 19, 4455.
7. (a) D. L. Li, H. S. Zhou and I. Honma, Nat. Mater., 2004, 3, 65; (b) Y.-J. Cheng, P. Muller-Buschbaum and J. S. Gutmann, Small, 2007, 3, 1379.
8. (a) D. Lupascu, A. Bartos, K. P. Lieb and M. Uhrmacher, Z. Phys. B: Condens. Matter, 1994, 93, 441; (b) E. N. Maslen, V. A. Streltsov and N. Ishizawa, Acta Crystallogr. B, 1996, 52, 414.
9. (a) V. Mikhelashvili, G. Eisenstein and F. Edelmann, J. Appl. Phys., 2001, 90, 5447; (b) T. M. Pan, C. L. Chen, W. W. Yeh and S. J. Hou, Appl. Phys. Lett., 2006, 89, 222912; (c) M. Losurdo, M. M. Giangregorio, P. Capezzuto, G. Bruno, R. G. Toro, G. Malandrino, I. L. Fragala, L. Armelao, D. Barreca, E. Tondello, A. A. Suvorova, D. Yang and E. A. Irene, Adv. Funct. Mater., 2007, 17, 3607.
10. (a) A. V. Prokofiev, A. I. Shelykh and B. T. Melekh, J. Alloys Compd., 1996, 242, 41; (b) S. Kimura, F. Arai and M. Ikezawa, J. Phys. Soc. Jpn., 2000, 69, 3451; (c) H. S. Kamineni, V. K. Kamineni, R. L. Moore, S. Gallis, A. C. Diebold, M. B. Huang and A. E. Kaloyeros, J. Appl. Phys., 2012, 111, 013104.
11. (a) G. T. K. Swami, F. E. Stageberg and A. M. Goldman, J. Vac. Sci. Technol., A, 1984, 2, 767; (b) N. Guerfi, O. Bourbia and S. Achour, Materials Science Forum, 2005, 193, 480; (c) H. S. Kamineni, V. K. Kamineni, R. L. Moore, S. Gallis, A. C. Diebold, M. B. Huang and A. E. Kaloyeros, J. Appl. Phys., 2012, 111, 013104.

---

Footnote

† Electronic Supplementary Information (ESI) available: experimental procedures, SEM/(HR)TEM images, crystallographic data, physisorption isotherms, UV-vis data. See DOI: 10.1039/c2ra21204h

---