SEARCH

SEARCH BY CITATION

Keywords:

  • Al;
  • Al-oxide;
  • EELS;
  • nanostructured materials;
  • nanocomposite;
  • powders

Abstract

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Experimental
  5. 3. Results and discussion
  6. 4. Conclusions
  7. Acknowledgements
  8. References

Al nanoparticles were prepared by the inert gas condensation method. After passivation with oxygen and air exposure we obtained a powdered sample of an Al-oxide/Al nanocomposite material. In the present paper we describe the use of the electron energy-loss spectroscopy (EELS) technique in a transmission electron microscope to characterize such nanostructured powders compared with a microcrystalline commercial aluminium foil. Energy-filtered images showed the presence of an alumina overlayer of ≈ 4 nm covering the aluminium nanoparticles (23 nm in diameter). EELS analysis enabled us to determine the total amount of Al2O3 and metallic Al and the structure of the alumina passivation overlayer in the sample. In particular, the extended energy-loss fine structure analysis of the data showed a major presence of Al tetrahedrally coordinated with oxygen in the alumina passivation layer of Al nanoparticles instead of the octahedral coordination found for a conventional Al foil. This surprising effect has been attributed to the nanoscopic character of the grains. The analysis of the electron-loss near-edge structure also determines the presence of a certain degree of aggregation in this kind of powdered sample as result of the coalescence of the nanocrystalline grains. The procedure presented here may have the potential to solve other problems during characterization of nanostructured materials.


1. Introduction

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Experimental
  5. 3. Results and discussion
  6. 4. Conclusions
  7. Acknowledgements
  8. References

Nanocrystalline materials with a grain or particle size below 50 nm may have special and interesting properties, as a significant number of atoms are situated at the surface or grain boundaries. In particular, the surface passivation layer of nanocrystalline metallic particles may have an important role in the final properties of the materials ( Gangopadhyay et al., 1992 ; Eckert et al., 1993 ; Apte et al., 1997 ; Sánchez-López et al., 1996 , 1997, 1998). The chemical and structural characterization of such ultrafine powders is in this respect an important field of research.

In the present paper we describe the use of the electron energy-loss spectroscopy (EELS) technique in a transmission electron microscope (TEM) to characterize such nanostructured powders. In particular, the crystallographic structure is a very important characteristic of the materials and determines many of their properties. The structure in terms of long range order can be established by X-ray diffraction (XRD) or high resolution transmission electron microscopy (HRTEM). However, these techniques provide no information for materials that are not well crystallized or for very thin and amorphous passivation layers. Extended electron energy-loss fine structure (EXELFS) is a potentially powerful technique which in principle enables us to determine the number, nature and distance of neighbouring atoms with respect to a central atom, as well as Debye–Waller factors ( Egerton, 1986). This can be done on either crystalline or amorphous materials. The principle of EXELFS relies on the behaviour of the electron ejected from a central atom after the inelastic scattering of a fast electron. In fact, provided that the collection angle of the spectrometer is small (to be consistent with the dipole approximation) it has also been shown that the EXELFS can be expressed with the same formalism as the well-known extended X-ray absorption fine structure (EXAFS) ( Schaich, 1984). The advantage of EXELFS over EXAFS is that the analysed volume can be several orders of magnitude lower, since the electron interaction is strong compared with that of photons. It is limited, however, to elements of lower Z for which the K-shells are attainable ( Hug et al., 1995 ). Also the electron-loss near-edge structure (ELNES) region of the EELS spectra may provide chemical and structural information in a way ana-logous to the X-ray absorption near-edge structure (XANES) region of the X-ray absorption spectroscopy (XAS) spectra.

In the present study we have analysed by EELS Al nanoparticles that have been prepared by the inert gas evaporation technique ( Gleiter, 1992). Al was evaporated in a N2 atmosphere and the ultrafine powder, collected in a cold finger, was passivated with oxygen before opening the chamber to air. Under these conditions the material obtained consists of particles of metallic aluminium covered by a very thin alumina overlayer ( Sánchez-López et al., 1998 ). By compaction and sintering of this material we could prepare a nanostructured Al-oxide/Al foil that, although showing metallic shine and low ohmic resistivity, can be heated to temperatures as high as 1273 K without breaking its structure ( Sánchez-López et al., 1996 ). It has also been shown ( Apte et al., 1997 ) that this nanocomposite is a low-density high-strength material. The analysis of the EXELFS as well as the ELNES region of the EELS spectra enabled us to characterize chemically and structurally this interesting nanocomposite Al-oxide/Al material. For conventional microcrystalline aluminium, exposure to oxygen produces a dense and compact nanometric alumina overlayer ( Cabrera & Mott, 1949) which prevents further oxidation of the material (passivation). In the case of nanometric aluminium grains, the radius of curvature may induce structural limitations to the formation of the alumina passivation overlayer. The aim of this study was to investigate the effect of this parameter. It is important to emphasize that the procedure described here may have applications for solving other problems during characterization of nanostructured materials.

2. Experimental

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Experimental
  5. 3. Results and discussion
  6. 4. Conclusions
  7. Acknowledgements
  8. References

The experimental device used for the synthesis of the Al ultrafine grains consists of a small HV chamber pumped to a residual vacuum better than 5 × 10−7 Torr. A tungsten boat containing small pieces of an Al foil (from Goodfellow Metals, 99.0% pure) was heated resistively to a temperature of 1623 K in a N2 atmosphere of ≈ 1 Torr. The evaporated material loses kinetic energy by collisions with the inert gas molecules and condenses as small particles that were collected on a cold Cu substrate. After preparation, the chamber was evacuated and the powder passivated by introducing small doses of oxygen before exposure to air. The loose powder was then smoothly stripped off and stored in air. We will refer to this sample as Al2O3/Al-nano. An Al foil, an amorphous alumina sample (a-Al2O3) obtained by an ion-beam-induced deposition technique ( Caballero et al., 1996 ) and a well crystallized γ-Al2O3 sample were used as references.

TEM examination of the sample was carried out in a Philips CM10 microscope working at 100 kV. The samples were dispersed in ethanol by sonication and dropped on a copper grid coated with a holey carbon film. Particle size distribution was evaluated from several micrographs using an automatic image analyser. The number of particles selected for consideration in the statistical calculus was 425. The particle size is defined as twice the average radius measured from the centre of the area to the perimeter of the particle.

XRD analysis was carried out using Cu Kα radiation in a Siemens D5000 diffractometer.

X-ray photoelectron spectroscopy (XPS) spectra were recorded in a VG-Escalab 210 spectrometer working in the constant analyser energy mode with a pass energy of 50 eV and using Mg Kα radiation as excitation source. The binding energy (BE) reference was taken at the C1 s peak from carbon contamination of the samples at 284.5 eV. An estimated error of ± 0.1 eV can be assumed for all measurements. For quantification, the XPS spectra were subjected to background substraction ( Shirley, 1990). Data fit was carried out by a least-squares routine supplied by the instrument manufacturer, using Gaussian–Lorentzian peaks.

EELS spectra were acquired in a Philips EM420 microscope operating at 120 kV and fitted with a Gatan model 666 parallel detection electron spectrometer. In order to record the Al K-edge, the illuminated area was ≈ 1 μm in diameter and the integration time in the photodiode array was 6 s. Spectra were recorded in the diffraction mode with a camera length of 122 mm and a collection angle of ≈ 15 mrad. No condenser aperture and an EELS entrance aperture of 5 mm were used to increase the signal. Although no specific study of radiation damage was carried out, it was visually checked that during the experiment the specimen did not change. The measured energy resolution at the zero-loss peak of the coupled microscope/spectrometer system was about 1.5 eV.

Prior to analysis of EXELFS oscillations, spectra were recorded for dark current and channel-to-channel gain variation. A low-loss spectrum was also recorded with each K-edge with the same illuminated area. After subtraction of the background with a standard power-law function, the spectra were deconvoluted for plural scattering with the Fourier-ratio method. All these treatments were performed within the EL/P program (Gatan).

EXELFS oscillations were analysed with the software package developed by Bonin et al. (1989 ). The coordination numbers (N), distances (R) and Debye–Waller factors (σ) were extracted by a least-squares fitting procedure that uses the theoretical phases and amplitudes proposed by McKale et al. (1988 ) previously calibrated with the appropriate references (Al foil and γ-Al2O3).

Energy filtered images were obtained by the image-EELS method using an energy-filtering TEM (Zeiss CEM 902) operating at 80 kV. Sequences of energy-filtered images were acquired from 10 eV to 90 eV with an energy step of 1 eV between each image. The width of the energy slit in the back focal plane of the spectrometer was ≈ 4 eV. The magnification of the image was 140 000 and the image size was 256 × 256. Under these conditions, one pixel corresponds to ≈ 1 nm area. The mappings were obtained by a two-image subtraction model at the plasmon energies of aluminium and alumina (20 and 27 eV, respectively) so that the image obtained at the minimum of the plasmon low-loss region at 16 eV was subtracted to the image obtained at the maximum of the characteristic plasmon peaks.

3. Results and discussion

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Experimental
  5. 3. Results and discussion
  6. 4. Conclusions
  7. Acknowledgements
  8. References

Figure 1 shows a TEM image and the particle size distribution histogram of the Al2O3/Al-nano sample. Particle sizes ranging from 10 to 40 nm and a mean particle size of 23 nm have been determined for this sample.

image

Figure 1. . TEM image (left) and particle size distribution histogram (right) for the Al2O3/Al-nano sample. Superimposed on the histogram is the lognormal distribution function.

Download figure to PowerPoint

The XRD diffractogram of the passivated Al nanoparticles ( Fig. 2, top) indicates the presence of pure aluminium. The Al 2p photoelectron spectrum of the sample ( Fig. 2, bottom) shows, however, that most of the aluminium at the surface is oxidized to Al3+ species. In fact, owing to the preparation procedure the sample is made up of aluminium metallic cores coated by an alumina overlayer. This Al2O3 passivation layer is very thin and not well crystallized so that XRD peaks from Al2O3 could not be detected.

image

Figure 2. O3/Al-nano sample.

Download figure to PowerPoint

High-magnification images of the sample have been included in Fig. 3. The alumina passivation layer can be observed as a kind of layered coverage on all the particle surfaces. This overlayer appears quite amorphous so that the electron diffraction pattern (also shown in Fig. 3) can be indexed according to pure aluminium metal peaks.

image

Figure 3. /Al-nano sample. Rings correspond to Al metal planes.

Download figure to PowerPoint

Figure 4 shows energy filtered images of the Al2O3/Al-nanosample where we can distinguish the alumina overlayer (red) covering the metallic aluminium cores (green). The high degree of interconnection or coalescence between particles can also be seen. An average thickness of ≈ 4 nm can be evaluated for the alumina passivation layer from this figure. This value is in good agreement with previous results obtained by HRTEM ( Sun et al., 1994 ), energy filtered images ( Eckert et al., 1993 ) and XPS studies ( Sánchez-López et al., 1998 ). The application of the EELS analysis in the transmission electron microscope will allow us to obtain a more exhaustive characterization of this material.

image

Figure 4. . Chemical mapping on the Al2O3/Al-nano sample by the image-spectrum EELS method. Top: zero-loss images. Bottom: Al2O3 distribution in red; Al distribution in green.

Download figure to PowerPoint

Figure 5 shows the EELS spectra recorded for the Al K edge of the Al2O3/Al-nano sample compared with that of Al foil and amorphous alumina reference samples. The signal-to-noise ratio appears acceptable and several EXELFS oscillations can be observed. Here it should be emphasized that for the Al K-edge EELS presents an important advantage over XAS. In fact, the cut-off due to the Si K-edge of the usually employed X-ray monochromators at 1839 eV allows the extraction of the EXAFS oscillations only over a window of 289 eV. In the case of EELS no limitations over this window are present.

image

Figure 5. . Al K-edge EELS spectra for the Al2O3/Al-nano sample and for the amorphous Al2O3 and Al reference samples.

Download figure to PowerPoint

As a first step we will analyse the near-edge structure of the EELS spectra (ELNES). This feature also contains information about the type of short range structure in a way analogous to the XANES region of the XAS spectra. Although the simulation of the ELNES spectra is complicated one can use this as a fingerprint technique. In this case we have carried out linear combinations of the reference spectra for the Al, γ-Al2O3 and amorphous alumina (a-Al2O3) (see Fig. 6). The best reproduction of the experimental ELNES spectra for the nanocrystalline Al2O3/Al-nano sample was obtained for the values of 35% of amorphous Al2O3 and 65% of metallic Al ( Fig. 6, left). Other linear combinations on the basis of different amounts of the oxide component are also included. It is also worth mentioning that a good agreement could not be obtained by using the spectrum of the crystalline γ-Al2O3 reference compound ( Fig. 6, right). This is in accordance with the amorphous character of the alumina overlayer in our nanocrystalline sample. For an isolated spherical particle 23 nm in diameter with an alumina overlayer 4 nm in thickness (see Fig. 7) we should obtain values of 66.5% of Al2O3 and 33.5% of metallic Al, corresponding to the bulk analysis character of the EELS. However, as shown in Fig. 1, the texture of the nanoparticles presents a high interconnection or coalescence of grains. A series of clusters of four spherical particles interconnected assuming the geometry represented in Fig. 7 gives a value of 36% of Al2O3 and 64% as metallic Al. The agreement of these values with the experimental data obtained from the ELNES linear combination is in good accordance with the TEM observed morphology of the powdered sample and with the energy filtered images ( Fig. 4).

image

Figure 6. . ELNES spectra at the Al K-edge of the Al2O3/Al-nano sample in comparison to an Al metallic foil and the a-Al2O3 (left) or γ-Al2O3 (right) samples. The spectrum for the sample (full line) is compared with the linear combination (dashed line) of the spectra for Al and the corresponding alumina reference.

Download figure to PowerPoint

image

Figure 7. . Schematic geometry of an isolated spherical particle and of a series of interconnected spherical grains of passivated nanostructured aluminium.

Download figure to PowerPoint

In principle, at the near-edge structure, we should distinguish between octahedrally and tetrahedrally coordinated Al3+ ( Brydson et al., 1994 ). However, owing to the amorphous character of the alumina overlayer and the low resolution of the EELS in the TEM microscope this is not clear. This resolution (2–3 eV at the Al K edge) is, however, good enough to record the EXELFS oscillations where we should obtain information about the type of coordination of aluminium in the oxide passivation layer.

Figure 8 shows the EXELFS oscillations extracted from the spectra after background removal (edge background followed by spline function). A simple comparison of the pattern of the Al2O3/Al-nano sample with those of the γ-Al2O3 and the Al foil references confirms that aluminium nanoparticles are really a nanocomposite material formed of an alumina passivation layer and Al cores. A more careful analysis of these EXELFS spectra can be carried out by Fourier transformation (FT) and fitting. Figure 9 shows the FT curves for the samples whose EXELFS oscillations were reported in Fig. 8, together with their corresponding FT curves obtained by fitting analysis. In this figure the first peak located around 1.65–1.9 Å corresponds to Al–O distances while the peak at ≈ 2.8 Å corresponds to Al–Al distances. It is clear from this figure that the Al foil reference has also a passivation alumina component (Al–O distances) owing to the air exposure of the sample. This alumina component is of course more intense for the aluminium nanoparticles owing to their high surface area. The fitting curves of the EXELFS oscillations have been also included in Fig. 8 and the best fitting parameters are displayed in Table 1. For the γ-Al2O3 reference sample we found the typical Al–O distances in both tetrahedral and octahedral coordination at 1.65 and 1.90 Å, respectively. The coordination numbers obtained for the reference compound correspond to the γ-Al2O3 structure, where two-thirds of the total aluminium atoms are in octahedral coordination (apparent coordination number: 6 × 2/3 = 4) and one-third in tetrahedral coordination (apparent coordination number: 4 × 1/3 = 1.3). The analysis of the Al2O3/Al-nano sample shows that the passivation layer of alumina can be fitted only with Al–O distances of 1.70 Å, indicating the preferential formation of tetrahedrally coordinated aluminium. This is a singular behaviour of the nanostructured material compared with that of a conventional Al foil which shows Al–O distances of 1.90 Å in its oxide passivation component. This is typical of an α-Al2O3 phase with all the Al octahedrally coordinated with oxygen. The formation of tetrahedral Al–O coordination spheres in the Al2O3/Al-nano sample may be attributed to structural limitations due to the nanometric character of this sample. A structure with Al in tetrahedral coordination may be stabilized in small nanometric particles with a high radius of curvature. It is also likely that the alumina formed is quite amorphous, showing no long-range order, as stated from the absence of X-ray diffraction peaks. Similar behaviour has been reported for erbium ( Li et al., 1988 ) and yttrium oxides ( Skandan et al., 1992 ) where high-pressure polymorphic phases can be stabilized in grains a few nanometres in diameter. Also, the γ-Fe2O3 has been found ( Sethi & Thölen, 1993; Sánchez-López et al., 1997 ) to be formed by air exposure of nanometric Fe particles instead of the α-Fe2O3 structure. It should be emphasized here that heating of the Al2O3/Al-nano sample in an Ar atmosphere leads to the crystallization of the γ-Al2O3, which contains Al in both tetrahedral and octahedral coordination, and not to the formation of an α-Al2O3 phase.

image

Figure 8. . EXELFS oscillations at the Al K-edge of the Al2O3/Al-nano sample and the references Al and γ-Al2O3. Full lines: experimental curves, dashed lines: curves obtained by fitting.

Download figure to PowerPoint

image

Figure 9. . FT spectra corresponding to the oscillations in Fig. 8. Full lines: experimental curves, dashed lines: curves obtained by fitting.

Download figure to PowerPoint

Table 1.  . Best fitting parameters of Al K-edge EXELFS oscillations for the Al2O3/Al-nano sample and for the Al and γ-Al2O3 reference samples. *Number of oxygen atoms in tetrahedral (Ot) or octahedral (Oh) coordination in the first coordination sphere of alumina. Number of Al atoms in the first coordination sphere of metallic aluminium. †Bond lengths. ‡Debye–Waller factors referred to the appropriate references. §The reported values of Al–Ot and Al–Oh distances in γ-Al2O3 are in the ranges 1.65–1.70 Å and 1.85–1.98 Å, respectively. The reported value of Al–Al distances in metallic aluminium is 2.86 Å. ¶Apparent coordination numbers corresponding to two-thirds of total aluminium atoms in octahedral coordination and one-third in tetrahedral coordination. Thumbnail image of

A final insight to the data summarized in Table 1 shows that the Al–O coordination number for the Al2O3/Al-nano sample is 1.3. This corresponds to 33% of the expected value of 4 for an alumina phase with all the Al atoms in tetrahedral coordination. For the aluminium component of the sample, the number of neighbours in the Al–Al coordination sphere is only 9.40, which corresponds to 78% of the expected value of 12 for a pure metallic aluminium phase. These values are in good agreement with the values obtained from the ELNES data of 35% of Al2O3 and 65% of Al for the Al2O3/Al-nano sample.

4. Conclusions

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Experimental
  5. 3. Results and discussion
  6. 4. Conclusions
  7. Acknowledgements
  8. References

In the present work we have shown the capabilities of EELS spectroscopy in the TEM for the characterization of nanostructured materials. A technique like XPS gives information only about the top-most layers of the sample. XRD or HRTEM, although they are bulk analysis techniques, cannot be used in the case of amorphous or very thin overlayers. EELS appears to be an important technique to complement the information obtained by XPS, XRD and TEM being in some aspects of fundamental importance. The chemical and structural characterization in the absence of long range order is of great interest. The similarity between EELS and XAS in this respect has been emphasized in the present work.

In the particular case presented here we have found the formation of an Al2O3 passivation overlayer on metallic aluminium grains of nanometric size in which Al appears tetrahedrally coordinated with oxygen. Although the Al2O3 passivation layer of a conventional Al foil shows the typical octahedral coordination of the α-Al2O3 phase, the formation in the Al2O3/Al-nano sample of tetrahedral Al-O coordination spheres may be attributed to the nanostructured character of the sample. Finally, it is important to emphasize that the EELS analysis can also predict the textural behaviour of powdered samples. An average particle aggregation of 4 has been estimated.

Acknowledgements

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Experimental
  5. 3. Results and discussion
  6. 4. Conclusions
  7. Acknowledgements
  8. References

We thank the DGICYT (projects PB93-0183 and PB96-0863-C02-02) and the Fundación Domingo Martínez for financial support.

References

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Experimental
  5. 3. Results and discussion
  6. 4. Conclusions
  7. Acknowledgements
  8. References
  • 1
    Apte, P., Suits, B.H., Siegel, R.W. (1997) Hardness measurements of nanophase Al/Al-Oxide consolidated composites. Nanostruct. Mater. 9, 501 504.
  • 2
    Bonin, D., Kaiser, P., Fretigny, C., Desbarres, J. (1989) Structures Fines d'Absorption des Rayous X en Chemie, Vol. 3, Logiciels d'Analyse EXAFS (ed. by H. Dexpert, A. Michalowizc and M. Verdaguer). Société Française de Chimie, Paris.
  • 3
    Brydson, R., Richardson, I.G., Groves, G.W. (1994) Determining the local coordination of Aluminium in cement using electron-energy-loss near-edge structure. Mikrochim. Acta, 114, 221 229.
  • 4
    Caballero, A., Leinen, D., Fernandez, A., Gonzalez-Elipe, A.R. (1996) Preparation of Al2O3 thin films by ion-beam-induced CVD: structural effects of the bombardment with accelerated ions. Surf. Coat. Technol. 80, 23 26.
  • 5
    Cabrera, N. & Mott, N.F. (1949) Theory of the oxidation of metals. Rept. Prog. Phys. 12, 163 184.
  • 6
    Eckert, J., Holzer, J.C., Ahn, C.C., Fu, Z., Johnson, W.L. (1993) Melting behavior of nanocrystalline aluminum powder. Nanostruct. Mater. 2, 407 413.
  • 7
    Egerton, R.F. (1986) Electron Energy-Loss Spectroscopy in the Electron Microscope. Plenum Press, New York.
  • 8
    Gleiter, H. (1992) Nanostructured materials. Adv. Mater. 4, 474 481.
  • 9
    Gangopadhyay, S., Hadjipanayis, G.C., Dale, B., Sorensen, C.M., Klabunde, K.J. (1992) Magnetism of ultrafine particles. Nanostruct. Mater. 1, 77 81.
  • 10
    Hug, G., Blanche, G., Jaoen, M., Flank, A., Rehr, J.J. (1995) Simulation of the extended fine structure of K-shell edges in intermetallic ordered alloys. Ultramicroscopy, 59, 121 136.
  • 11
    Li, Z., Hahn, H., Siegel, R.W. (1988) New phases of erbium oxides. Mater. Lett. 6, 342 346.
  • 12
    McKale, A.G., Veal, B.W., Paulikas, A.P., Chan, S.K., Knapp, G.S. (1988) Improved ab initio calculations of amplitude and phase functions for extended X-ray absorption fine-structure spectroscopy. J. Am. Chem. Soc. 110, 3763 3768.
  • 13
    Sánchez-López, J.C., Fernández, A., Conde, C.F., Conde, A., Morant, C., Sanz, J.M. (1996) The melting behavior of passivated nanocrystalline aluminum. Nanostruct. Mater. 7, 813 822.
  • 14
    Sánchez-López, J.C., Justo, A., Fernández, A., Conde, C.F., Conde, A. (1997) Preparation and thermal evolution of vapour condensed nanocrystalline iron. Philos. Mag. B, 76, 663 667.
  • 15
    Sánchez-López, J.C., González-Elipe, A.R., Fernández, A. (1998) Passivation of nanocrystalline Al prepared by the gas phase condensation method: an XPS study. J. Mater. Res. 13, in press.
  • 16
    Schaich, W.L. (1984) Derivation of single-scattering formulas for X-ray-absorption and high-energy electron-loss spectroscopies. Phys. Rev. B, 29, 6513 6519.
  • 17
    Sethi, S.A. & Thölen, A.R. (1993) A TEM study of oxides formed on ultrafine Fe, Cr and Fe–Cr particles. Nanostruct. Mater. 2, 615 622.
  • 18
    Shirley, D.A. (1990) Pratical Surface Analysis, 2nd edn, Vol. 1, Auger and X-Ray Photoelectron Spectroscopy (ed. by D. Briggs & M. P. Seah), pp. 233 236. John Wiley & Sons Ltd, Chichester.
  • 19
    Skandan, G., Foster, C.M., Frase, H., Ali, M.N., Parker, J.C., Hahn, H. (1992) Phase characterization and stabilization due to grain size effects of nanostructured Y2O3. Nanostruct. Mater. 1, 313 322.
  • 20
    Sun, X.K., Xu, J., Chen, W.X., Dei, W.D. (1994) Preparation of Al nanoparticles in a controlled enviroment. Nanostruct. Mater. 4, 337 344.