Wettability and reactivity
of the Al/ZnO couple at a temperature of 1273 K in vacuum have been studied with application of modified sessile drop method. The results obtained for the ZnO single crystal substrates of various orientations evidenced non-wetting by liquid aluminium, showing high contact angles of 100°–107°. Detailed transmission electron microscopy characterization unambiguously proved that the creation of two interfacial layers revealing highly epitaxial growth took place at high-temperature interaction between liquid Al drop and ZnO substrate. The layers of columnar grains consisted of reactively formed Al2O3 and ZnAl2O4, the second one growing at the expense of the alumina. The alumina was identified as either cubic γ-, tetragonal δ- or monoclinic λ-Al2O3, depending on subtle changes in chemistry of the surface top layer of the ZnO substrates.
Begleitmaterial
Supplementary material 1 (MPEG 45575 kb)
Hinweise
Electronic supplementary material
The online version of this article (doi:10.1007/s10853-015-9474-2) contains supplementary material, which is available to authorized users.
Introduction
Light-weight Al–Al2O3 composites produced in situ due to redox reaction between liquid Al and oxides such as SiO2, mullite, kaolinite, CeO2, NiO, MgO and ZnO are attractive alternative for conventional light alloys used in aviation and ground transport industry [1‐6]. That is why, there is a strong need to extend the knowledge on both reactivity and wettability of these systems of interest. The sessile drop experiment is of particular usefulness in such studies and was successfully applied previously to Al/SiO2 [1], Al/NiO [3, 4], Al/MgO [5], Al/ZnO [6‐8] and Al/Y2O3 [9] systems. The last two systems reveal similarities, since the reactively formed aluminium oxide crystallites are large and surrounded by the metallic channels composed of either Al(Mg) or Al(Zn). These two systems can find an application when the higher content of alumina is desired, and larger amount of metal is predictably reduced, as one way to control the quantity of the alloying element is by using MeO, which produces Me in a vapour form by the redox reaction with Al. Zinc and magnesium oxides fulfil this requirement. The experiment performed at 1273 K leads to the formation of reaction product region (RPR) composed of alumina particles interpenetrated by the metallic channels being either Al(Mg) or Al(Zn). In the RPR, the alumina content can be increased, while the zinc or magnesium is kept low in these metallic channels because of its diffusion into the drop followed by further vaporization from the liquid Al(Zn) or Al(Mg) metal. So far, several researchers have used zinc oxide for in situ reactive synthesis of Al–Al2O3 composites [9‐13].
The structural investigation of Al/ZnOSC sessile drop couples reported in works [7, 8] revealed the formation of not only large α-Al2O3 crystals interpenetrated by the Al(Zn) channels but also an additional thin layer located next to the ZnO substrate. In study [6], the examination performed on polycrystalline ZnO substrate reacting with molten aluminium established the occurrence of thin layer of about 250 nm specified as metastable δ-Al2O3. In the next paper, the application of the same sample preparation procedure but using the single crystal ZnO substrate of 〈0001〉 orientation (further denoted as ZnO〈0001〉) was reported. The creation of the thin alumina layer at the ZnO-side interface was also detected using the transmission electron microscopy (TEM). At that moment, its growth was presumed to be an effect of cooling [8]. In order to confirm this statement, the Al/ZnO couple was produced by capillary purification (CP) procedure combined with drop pushing procedure (DP) described in [14]. Such a unique procedure comprises squeezing of the liquid aluminium from the capillary on the substrate surface at the test temperature of 1273 K, followed by pushing it into another location of the substrate after appropriate contact time. Several advantages arose from the experiment performed in such a way. At first, it allowed non-contact heating of both Al and ZnO materials and, therefore, to avoid the effect of heating history due to the interaction between them during contact heating to the test temperature, commonly used in wettability tests. Next, it caused the removal of the primary oxide film from the Al drop directly in the high-temperature chamber under vacuum. And finally, it helped to exclude or at least significantly reduce the effect of the cooling history, as the droplet was no more in contact with the substrate during the cooling. Three phases were detected within the Al/ZnO〈1−100〉 couple: locally distributed large crystals of α-Al2O3, the alumina of unknown type and ZnAl2O4 spinel placed between the unknown Al2O3 and ZnO. Both alumina and spinel took form of columnar grains, which placed one above other formed thin and continuous layer of about 200 nm in thickness [7]. The above results as well as incomplete literature data require systematic examination of the Al/ZnO contact system comprising of both the reactivity and wettability strongly correlated with the substrate surface structure, which was the aim of the research presented here. The systematic studies with application of the pushing drop procedure and various crystallographic orientations of the ZnO substrate were performed first time ever.
Anzeige
Experimental
Aluminium of 99.999 % purity and ZnO single crystal substrates were subjected to the sessile drop wettability tests at 1273 K using experimental facility described in details in work [14]. All samples were ultrasonically cleaned in the isopropanol before the test. Zinc oxide single crystals ZnOSC of 〈1−100〉, 〈1−120〉, and 〈0001〉 orientations and of two variants of atomic termination of the ZnO surface (denoted as ZnO〈0001〉Zn or ZnO〈0001〉O for Zn-terminated and O-terminated surfaces, respectively) are grown by hydrothermal method under high pressure with the surface roughness less than 5 Å.
As the experimental conditions have significant influence on the wetting test results [15], two testing procedures were applied in the sessile drop wettability tests: (1) CP and (2) drop pushing in order to expose the Al/ZnO interface at high temperature directly in vacuum chamber after appropriate interaction time and uncover freshly formed reaction products for further SPM and TEM investigation. It consisted of squeezing the aluminium droplet through an alumina capillary at the test temperature and its deposition on the ZnO substrate. After 20 min of Al–ZnO interaction, the drop was transferred to a new location of the same ZnO substrate for another 5 min contact and then finally pushed away from the substrate to the alumina support placed next to the ZnO single crystal [14]. After wettability tests, the Al/ZnO couples were cooled at the rate of ∼10 K/min. All the manipulations were undertaken under dynamic vacuum produced by the continuously working turbomolecular pump.
The TEM lamellas were obtained using focused ion beam technique (Quanta 3D, FEI) and in situ lift out technique, which guaranteed the precise probing of the place of analyses. The TEM techniques comprising of bright-field (BF), high-resolution observation and also selected area electron diffraction (SAED) were carried out using TECNAI G2 FEG super TWIN (200 kV) microscope. Additionally, high-angle annular dark-field detector allowed the observation in the scanning transmission mode and an energy dispersive X-ray spectroscopy system manufactured by EDAX company integrated with the microscope let to carry out the local chemical analyses as well as to collect the maps of element distribution in the particular location in the sessile drop sample.
Results and discussion
Wettability tests
The contact angle measurements of aluminium on ZnOSC substrates of 〈1−100〉, 〈1−120〉 and 〈0001〉 (Zn or O faced) orientations were carried out at 1273 K for 20 min. The obtained results (Fig. 1a and Table 1) indicate that interaction of liquid aluminium with all ZnO substrates exhibits non-wetting behaviour (θ > 90°), reaching the contact angle values of θ = 100–107°.
Fig. 1
Wetting kinetics of Al/ZnOSC system measured at first location of the Al drop on the ZnO substrate at 1273 K lasting for 20 min, in which ZnO substrates of various orientations, 〈0001〉 of two variants O or Zn faced, 〈1−100〉 and 〈1−120〉, were applied (a). Comparison of the wetting angle changes for the drop placed at first (20 min at 1273 K) and second (5 min at 1273 K) location of the same ZnO 〈0001〉Zn substrate (b)
Table 1
Approximate thickness values of alumina and spinel layers formed due to the interaction between aluminium and ZnO single crystal of various orientation after 5 and 20 min in 1273 K as well as the type of formed Al2O3
Couple
Testing conditions
Contact angle (°)
Time of interaction (min) at 1273 K
Thickness (nm) of Al2O3
Thickness (nm) of ZnAl2O4
Type of A2O3
Time (min)
Temp (K)
T = 0 min
T = 20 min
θ left
θ right
θ left
θ right
Al/ZnO〈0001〉Zn
20 + 5
1273
105.1
103.4
103.6
97.5
20
150
600
δ οr γ
5
1273
96.3
101.6
95.0
103.3
5
350
400
δ οr γ
Al/ZnO〈0001〉O
20 + 5
1273
101.0
101.3
102.0
101.7
20
300
500
δ
5
1273
95.9
100.0
91.8
95.1
5
400
400
λ
Al/ZnO〈1−100〉 Zn
20 + 5
1273
101.9
99.0
96.0
93.1
20
200
550
γ, δ, λ
5
1273
1*
1*
1*
1*
5
150
200
γ, δ, λ
Al/ZnO〈11−20〉Zn
20 + 5
1273
90.6
101.1
100.6
98.6
20
180
1000
δ οr γ
5
1273
2*
2*
2*
2*
5
350
280
δ οr γ
1* sticking of the drop to the pusher making impossible to calculate the contact angles
2* drop too big to push it to another place on substrate
×
Anzeige
The wetting kinetics of the drop pushed away to another location at the same substrate has been also studied for 5 min. As can be seen in Fig. 1b for the ZnO 〈0001〉 substrate, no change in the contact angle value took place in comparison with the first position of the drop on the substrate. The contact angle formed on ZnO 〈0001〉 surface was lower by 6°, reaching the value of 99°. The dissimilar wetting behaviour of the Al drop in the first and second positions on the same substrate might be related to the effect of liquid metal saturation with oxygen and zinc during first 20 min contact, resulting in a suppressed interaction between the fresh substrate surface and the Al(O, Zn) drop in its second position after it was moved.
The tendency to lower the contact angle value with time of interaction can be associated with either the evaporation from the drop or due to chemical reaction, which takes place at the drop/substrate interface under experimental conditions used. First phenomenon can be excluded by the measurements of the diameter (dmax) and height (hmax) of the drop at the beginning and at the end of the contact angle measurements. During the performed experiment, dmax increased with decreasing hmax, while the corresponding drop volume changed up to 0.2 % only, which clearly evidenced that although the metal evaporation takes place, its effect on the drop size is negligible. Therefore, it was presumed that lowering of the contact angle is caused by the formation of the new compound at the drop/substrate interface.
The literature data suffer from very limited information concerning both the wettability and reactivity in Al/ZnO system followed by the detailed microstructure and phase composition characterization. The presented results can be compared with those reported by Sobczak et al. [1, 16] showing high reactivity accompanied with a lack of wettability in the Al/ZnO couple. High contact angles (θ = 111 for single crystalline substrate and θ = 140 for polycrystalline substrate) were measured at 1273 K. At that temperature, the Al2O3 reaction product, suggested by the electron microscopy characterization coupled with EDS analysis, is known to be wetted by molten Al, as reported in the analysis of literature data on contact angle measurements of Al/Al2O3 system performed by Sobczak et al. [17]. Since the removal of oxide film from Al drop did not improve the wettability in Al/ZnO couple, its non-wetting behaviour was explained either by secondary roughening of initially smooth ZnO single crystal surface or by secondary oxidation of the Al drop surface [1]. The first explanation came from the observation of the fine alumina precipitates detected at the substrate surface around the Al drop and probably created by evaporative-reactive deposition caused by enlarged transport of Al vapour under UHV and its reaction with ZnO. The idea of secondary oxidation of the drop surface was based on the observations of fine alumina precipitates at the drop surface, particularly densely localized close to the triple line. Their formation was explained by enhanced transfer of oxygen from the ZnO substrate caused by copious evaporation of Zn. It should be highlighted also that in previous reports, no information about the type of reactively formed alumina was provided, while the results obtained were compared with experimental data showing wetting in the Al/Al2O3 couple at 1273 K when α-Al2O3 was used. Following more recent reports by Shen et al. [18‐21] concerning the influence of the α-Al2O3 single crystal orientation on wetting its surface by molten aluminium in the temperature range of 800–1500 °C, it can be found that this effect is very strong. The wetting of molten Al on differently oriented single crystals R〈0112〉, A〈1120〉 and C〈0001〉 showed that the adhesion of the molten aluminium on the R and A surfaces is much stronger in comparison with the C surface.
Therefore, in this study, the detailed structural investigation was performed using two advanced techniques of materials characterization, i.e. the scanning probe microscopy (SPM) technique (in order to examine the subtle changes in substrate surface topography after high-temperature contact with liquid Al), followed by the transmission electron microscopy examination of the microstructure and phase composition of the compounds present at the drop/substrate interface.
Substrate surface topography after interaction with Al drop
Figure 2a shows the image of the ZnO〈11−20〉 surface exposed at high temperature after interaction with aluminium. There are five characteristic areas that can be distinguished. Regions 1 and 2 correspond to the place of the first contact of the Al drop with ZnO substrate. Region 3 represents the triple line region formed after 20 min of interaction. Region 4 is just outside of the contact place, while the fifth region is far away from it. During dropping the aluminium from the capillary, the volume of the squeezed drop slightly increases. Each ring marked with arrows visible in Fig. 2a corresponds to the new volume.
Fig. 2
SPM images of the ZnO〈11−20〉 surface exposed at high temperature after interaction with aluminium: general view of the sample surface with particular areas of interest (a). Region 1 and 2 denotes the first contact of the Al drop with ZnO substrate (b, c), region 3 represents the triple line region formed after 20 min of interaction (d), region 4–surface area close but beyond the contact with the drop (e) and region 5 is far away from it (f)
×
Figure 2b–f shows the series of measurements performed consequently for all mentioned above regions. First region (Fig. 2b) reveals the dramatic change of the substrate structure although the contact time was extremely short and lasted only few seconds. It resulted in the immediate reduction of the ZnO by liquid Al. The presence of the triple line, of about 100 nm depth, can be also noted here. The region in contact with the Al is composed of many craters along which the liquid aluminium can infiltrate the monocrystalline substrate and reactively formed products can appear.
The examination of region 2 (Fig. 2c) confirmed the presence of large channels through which the transfer of liquid Al to the reaction front and reactively formed zinc to the drop took place. The maximum depth of these channels was 140 nm. Region 3 visualized in Fig. 2d is the surface topography after extra dropping of aluminium. This structure is different from the previous one because of lack of large craters and the occurrence of fine columnar crystals within this area. Also the RPR is located much lower than the place outside the drop. The dissolution of the ZnO substrate in the drop results in the formation of the characteristic crater about 700 nm deep in the triple line area. Furthermore, in region 4 (Fig. 2e), representing the area beside the drop/substrate contact, the secondary roughening of the surface can be observed contrary to the flat surface of the zinc oxide (Ra = 2 nm) measured far away from the place of Al/ZnO contact (Fig. 2f). This phenomenon takes place at 1273 K due to chemical reaction between Al vapour and ZnO substrate surface, and it can contribute to the formation of high apparent contact angles as explained in [1].
Microstructure and phase composition of the products reactively formed at the drop/substrate interface
The surface of the ZnO substrates after high-temperature interaction with molten aluminium was next observed with scanning electron microscopy (Fig. 3). Figure 3a shows the overall view of the sample obtained in DP. The places where the aluminium drop contacted the substrate for 20 min (the first position) and 5 min (the second position) are well distinguishable as the circles of brighter contrasts. The local existence of large crystals, previously identified as α-Al2O3 [7], can be noticed (Fig. 3b). They appeared due to the interaction of ZnO substrate with so-called “daughter” droplets, which remained after pushing off the “mother” drop.
Fig. 3
Overall view of the Al/ZnO sessile drop sample on alumina support after using drop pushing procedure. Al drop was removed from ZnO substrate onto alumina support (a). SEM-SE top view image of the substrate surface with visible large dendrites of α-Al2O3 locally distributed on the substrate surface corresponding to the first position of the drop contacting ZnO for 20 min (b) and area of contact close to large dendrites (c)
×
In all cross-sectioned Al/ZnO couples, the formation of two layers revealing strong epitaxial growth was observed in the following order: ZnO/ZnAl2O4/Al2O3. Both the spinel and aluminium oxide took the shape of long, columnar crystals, of which the acquisition of either the electron diffraction or high-resolution image was not a trivial task because of their mutual overlapping.
Figure 4 shows the microstructure of the cross-sectioned area within the places where Al drop reacted for particular time−the first lasting 20 min and the second for another 5 min (total of 20 + 5 min) with 〈0001〉-oriented ZnO- and Zn-faced substrates. As it can be noticed in Fig. 4a, e, a layer of columnar grains can be distinguished above the ZnO〈0001〉Zn. Observations in scanning transmission mode (Fig. 4f, g) revealed various contrast within this layer. The EDS analysis showed the presence of alumina and ZnAl2O4 spinel. This last compound was observed between Al2O3 and ZnO. For the Zn-terminated 〈0001〉 orientation of ZnO single crystal, the approximate thicknesses of reactively formed layers after interaction at the first location of the drop on the substrate (20 min time) were 150 and 600 nm, and in the second location (another 5 min), they were 350- and 400-nm thick, for the alumina and spinel, respectively. In contrary to this, when the top layer of the single crystal consisted of the oxygen atoms (ZnO 〈0001〉O), the thicknesses of both products after 20 min of the interaction were 300 nm (Al2O3) and 500 nm (ZnAl2O4), while at the second location of the drop (5 min of reaction at 1273 K), they were similar and both of about 400 nm (Fig. 5).
Fig. 4
TEM-BF image of the RPR/ZnO〈0001〉Zn with visible layers of Al2O3 and ZnAl2O4 formed after 20 min of drop/substrate contact at 1273 K (a) and the one created at second location of the drop staying in contact with the substrate for another 5 min time (e) of interaction with Al at 1273 K. b–d shows the SAED patterns of the ZnO, ZnAl2O4 and Al2O3 crystals, respectively. f and g present the STEM images together with the map of element distribution Al, O and Zn recorded for the RPR formed at first and second location of the drop on the ZnO〈0001〉Zn substrate, respectively
Fig. 5
TEM-BF image of the RPR/ZnO〈0001〉O with Al2O3 and ZnAl2O4 crystals formed after 5 min of interaction of ZnO with Al drop at 1273 K (a). b and c show the SAED patterns of the ZnO together with ZnAl2O4 (b) and ZnAl2O4 with Al2O3 crystals (c)
×
×
Particular attention was paid to the determination of the type of formed aluminium oxide. The analysis of the SEAD patterns showed that the crystal structure of Al2O3, formed after 20 min interaction of aluminium with ZnO〈0001〉Zn, was either γ- or δ-Al2O3 phase, while for the ZnO〈0001〉O, the columnar grains of tetragonal phase delta prevailed. At the second location of the drop, the drop/substrate interaction was different for the applied two variants of the single crystal. In the case of the top surface layer composed of zinc atoms, the appearance of γ- or δ-Al2O3 was observed, wherein the orientation of the gamma oxide was the same as that of spinel (Fig. 4b–d). Conversely, the monoclinic λ-Al2O3 phase was mostly formed for the samples with the top layer of ZnO single crystal built of oxygen atoms (Fig. 5a–c).
Interaction of molten aluminium and ZnO of 〈1−100〉 orientation (Fig. 6a, b) resulted in the appearance of the aluminium oxide and the spinel of 200 and 550 nm in thickness (after 20 min) and 150 and 200 nm (after 5 min). However, the most spectacular difference in the thickness of these two layers was observed for the ZnO substrate of 〈1−120〉 orientation (Fig. 4c, d), where after 20 min of interaction, the aluminium oxide thickness was only 180 nm, while that of spinel increased up to 1000 nm. After 5 min of interaction, the widths were similar and reached 350 nm (aluminium oxide) and 280 nm (spinel). The reaction of 〈1−120〉-oriented ZnO single crystal and aluminium at 1273 K for both 20 and 5 min led to the formation of γ- or δ-Al2O3 but the gamma alumina dominated for the second interaction (Fig. 7). On the other hand, the research performed on sample Al/ZnO〈1−100〉 has shown the presence of all three of the mentioned variations of metastable alumina γ-, δ- or λ-Al2O3. The comparison of thickness of both alumina and spinel layers as well as the type of formed Al2O3 has been presented in Table 1.
Fig. 6
TEM-BF images of the RPR/ZnO〈1−100〉 cross section with Al2O3 and ZnAl2O4 crystals formed after 20 min (a) and 5 min (b) of interaction with Al at 1273 K together with BF images of the RPR/ZnO〈1−120〉 with Al2O3 and ZnAl2O4 crystals formed after 20 min (c) and 5 min (d) of interaction time
×
As it can be easily noticed, subtle changes of the top surface chemistry of the ZnO substrate influence the type of created alumina phase as well as the thickness of both alumina and spinel. The largest thickness was obtained in the case of spinel formed after the reaction with ZnO of 〈11–20〉 orientation. However, it should be also pointed that these phase composition changes do not cause the changes in the contact angle values, which are almost the same in all the cases and are in fact the contact angle between the aluminium and metastable alumina. The morphology of the alumina takes shape of longitudinal columnar grains in all the cases—also visible under SPM. The potential of the C4 morphology creation (co-continuous ceramic composite composed of mutually interpenetrated metallic and ceramic phases) was previously attributed to the volume mismatch between the substrate (ZnO) and reactively formed alumina that was believed to be α-Al2O3, being as large as 40 % [22]. Such a type of morphology occurs after long time of interaction, when the reactively formed product is dominated by alumina interpenetrated with Al(Zn). Short time of interaction results in creation of two products: alumina and spinel in the form of layers one above the other. The question that arises is which route of creation of the products among the two proposed below takes place in this study? The analysis of experimental results suggests that the reactively changed interfaces can be formed by two possible sequences (single slash denotes stable and double slash denotes unstable interface):
The results in Table 1 confirm that the growth of the spinel (especially pronounced in the case of 〈11–20〉 orientation) is favoured during the first contact of the aluminium with zinc oxide, when the drop is not saturated with the oxygen and zinc. After interaction in the first location, the chemical composition of the drop changes and in its second location, the drop becomes less reactive in contact with fresh ZnO surface. Therefore, the reactively formed alumina and spinel are of more or less of the same thickness. Taking into account the time of interaction at the first location (20 min of contact plus another 5 min, when the drop is placed and kept at the second location before it is pushed away) and the second one (5 min only) and comparing the microstructures, it can be concluded that alumina is formed as the first layer and later the spinel layer grows at the expense of Al2O3.
Fig. 7
SAED patterns of the ZnAl2O4 and Al2O3 crystals formed in the RPR of the Al/ZnO〈1−120〉 couple after 20 min of interaction with aluminium at 1273 K
×
Conclusions
For the first time, the effects of heating and cooling history on the interface transformation in highly reactive Al/ZnO system were dramatically reduced by combined application of two testing procedures allowing non-contact heating of a couple of materials as well as exposing of the interface at high temperature directly in the UHV chamber during the sessile drop wettability tests. Under conditions of this study, the liquid Al did not wet the ZnO substrate (θ > 90°) independently from its crystallographic orientation.
Detailed characterization of exposed substrate-side interface formed between liquid Al and ZnO single crystals of different crystallographic orientations showed that non-wetting behaviour in highly reactive Al/ZnO system may be related with the formation of non-wettable reaction product since the α-Al2O3 phase, widely accepted to be wettable by liquid Al at 1273 K and expected in previous reports as a main reaction product in the Al/ZnO couple, was not found in this study. It was evidenced that the high-temperature interaction in Al/ZnO couple led to the creation of two reactively formed layers corresponding to the metastable Al2O3 and ZnAl2O4 spinel, both characterized with highly epitaxial growth. The thicknesses of those layers varied with respect to the drop content and time of interaction: longer time promoted the spinel growth at the expense of the alumina. The alumina was identified as either cubic γ-, tetragonal δ- or monoclinic λ-Al2O3.
Moreover, the type of created alumina was affected by the crystallographic orientation and chemistry of the top surface layer of the ZnO substrates showing dissimilar alumina formed even for the substrates of the same 〈0001〉 orientation but of various Zn- or O-terminated surfaces.
Acknowledgements
The Ministry of Science and Higher Education of Poland within the Project Iuventus Plus No. IP2011061071 has supported this study. Samples were prepared in the Centre for High-Temperature Studies at the Foundry Research Institute in Cracow. The SEM and TEM studies were performed in the Accredited Testing Laboratories at the Institute of Metallurgy and Materials Science of the Polish Academy of Sciences in Cracow.
Compliance with ethical standards
Conflict of interest
The authors declare that they have no conflict of interest.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
Die im Laufe eines Jahres in der „adhäsion“ veröffentlichten Marktübersichten helfen Anwendern verschiedenster Branchen, sich einen gezielten Überblick über Lieferantenangebote zu verschaffen.
