Self-Organized Al2Cu Nanocrystals at the Interface of Aluminum

Article

Subscriber access provided by SUNY DOWNSTATE

Self-organized Al2Cu nanocrystals at the interface of aluminumbased reactive nanolaminates to lower reaction onset temperature Lorena Marín, Benedicte Warot-Fonrose, Alain Esteve, Yves J. Chabal, Luis Alfredo-Rodriguez, and Carole Rossi ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.6b02008 • Publication Date (Web): 04 May 2016 Downloaded from http://pubs.acs.org on May 5, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Applied Materials & Interfaces is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Self-organized Al2Cu nanocrystals at the interface of aluminum-based reactive nanolaminates to lower reaction onset temperature Lorena Marín†, Bénédicte Warot-FonroseΨ, Alain Esteve†, Yves J. Chabal♦, Luis Alfredo RodriguezΨ, Carole Rossi†* †

LAAS-CNRS, The University of Toulouse, 7 Avenue du colonel Roche, F-31077 Toulouse, France Ψ

CEMES-CNRS, The University of Toulouse, 29 rue Jeanne Marvig, F-31055 Toulouse, France

♦

Department of Materials Science and Engineering, The University of Texas at Dallas, Richardson, Texas 75080, United States

KEYWORDS: Al2Cu, reactive nanolaminates, Al-Cu, Al-CuO, reactive interfaces, nanoenergetics.

ABSTRACT

Nanoenergetic materials are beginning to play an important role in part because they are being considered as energetic components for materials, chemical, and biochemical communities (e.g. micro-thermal sources, micro-actuators, in-situ welding and soldering, local enhancement of chemical reactions, nano-sterilization and controlled cell apoptosis) and because their fabrication/synthesis raises fundamental challenges that are pushing the engineering and scientific frontiers. One such challenge is the development of processes to control and enhance the reactivity of materials such as energetics of nanolaminates, and the understanding of

ACS Paragon Plus Environment

1


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 30

associated mechanisms. We present here a new method to substantially decrease the reaction onset temperature and in consequence the reactivity of nanolaminates based on the incorporation of a Cu nanolayer at the interfaces of Al/CuO nanolaminates. We further demonstrate that control of its thickness allows accurate tuning of both the thermal transport and energetic properties of the system. Using High Resolution Transmission Electron Microscopy, X-ray diffraction and Differential Scanning Calorimetry to analyze the physical, chemical and thermal characteristics of the resulting Al/CuO + interfacial Cu nanolaminates, we find that the incorporation of 5 nm Cu at both Al/CuO and CuO/Al interfaces lowers the onset temperature from 550 to 475 °C due to the lower-temperature formation of Al-Cu intermetallic phases and alloying. Cu intermixing is different in the CuO/Cu/Al and Al/Cu/CuO interfaces and independent of total Cu thickness: Cu readily penetrates into Al grains upon annealing to 300 °C, leading to Al/Cu phase transformations, while Al does not penetrate into Cu. Importantly, θAl2Cu nanocrystals are created below 63% wt Cu/Al, and coexist with the Al solid solution phase. These well-defined θ-Al2Cu nanocrystals seem to act as embedded Al+CuO energetic reaction triggers that lower the onset temperature. We show that ~ 10 nm thick Cu at Al/CuO interfaces constitutes the optimum amount to increase both reactivity and overall heat of reaction by a factor of ~20%. Above this amount, there is a rapid decrease of the heat of reaction.

INTRODUCTION Nanomaterials and nanotechnology are nowadays, in many important aspects of the research and industry, very fertile fields

1,2

. A very promising and growing field of research concerns the

nanoengineering of advanced reactive nanolaminates capable of long-term chemical energy storage. Reactive nanolaminates also termed nanostructured metastable intermolecular


2

Page 3 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


composites, are stacked layers of metal and oxide that undergo exothermic, self-propagating reactions when layer mixing is induced by an external energy source

3–7

. These nanolaminates

exhibit mobile, high temperature reaction zones where atoms of layers diffuse across the interfaces, wherein velocity and temperature can be manipulated by composition and geometry of the component materials. They are used as customized heat sources for rapid fuses and microinitiators

6–8

biological neutralization

molecular delivery

4,5

9–12

, brazing of materials

13

and pressure-mediated

. Reactive nanolaminates of various chemistries have been explored to

adjust the reaction properties to the requirements given by the application, but Aluminum is the most studied metals because of its high oxidation enthalpy

14

, abundance, low cost and

technological maturity. Many published reactive nanolaminates are based on Al/CuO multilayers 6,7,15–23

which

have

high-energy

release

associated

with

the

highly

exothermic

reduction/oxidation reactions to form suboxides. Such structures have tunable thermal properties. For example in Al/CuO multilayers, self-propagation rate ranges from 80 to 0 m/s by varying the thickness of the bilayer from 50 nm to 1.5µm

15

. The main technological challenge to fully

exploit these nanostructures as customized MEMS heat sources is that their overall reactivity (reflected by the onset temperature and the self-sustained combustion velocity) remains lower than those of the nanopowders mixture. Recently, the incorporation of a thin layer of Cu on the aluminum layer at the interfaces was shown to notably increase Al based nanolaminate reactivity 19. This work demonstrated that the deposition of only 5 nm of metallic Cu on the 100 nm thick Al layer prior to the deposition of the 200 nm thick CuO layers in Al/CuO nanolaminates increases the flame propagation velocity from 44 ± 1 to 72 ± 1 m/s for a 7-bilayers nanolaminates without affecting the material exothermicity. Because copper has a higher thermal conductivity than Al, incorporating pure


3


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 30

copper into Al/CuO nanolaminate systems may enhance the overall thermal conductivity leading to an increase of self-propagation velocities. This effect has been demonstrated by Shen et al. 24 by adding Ag nano-particles into Al/CuO nanothermites. On the contrary, in the case of Cu atomic diffusion, Al-Cu alloys are formed upon heating and the overall metallic film thermal conductivity can be penalized since the Cu-Al alloys feature lower conductivity that pure Al 25. Another positive effect is that sputtering pure copper on Al layer prevents the intermixing of the Al and CuO layers during deposition and the subsequent loss of heat when Al+CuO intermix. Differential Scanning Calorimetry (DSC) analysis of Al/CuO nanolaminates diluted with pure copper also suggested that solid phase transformations observed in thermal analysis and occurring between 150 and 300 °C could play an important role for such reactivity increase. These solid phase transformations result in the decomposition of supersaturated solid solutions and formation of Al-Cu intermetallic phases. The enthalpy of such phase transformations is much less than that of oxidation but this additional heat release could significantly accelerate ignition. A more detailed study of metallurgical reactions and phase evolution at the Al-Cu interfaces is absolutely required to establish which of the phase transformations occurring in such systems upon heating affects the overall onset and combustion velocity as well as to find the optimized Cu content relative to Al. In this study, we combine Differential Scanning Calorimetry (DSC), Structural and Chemical Analysis by Scanning Transmission Electron Microscopy (STEM), Electron Energy-Loss Spectroscopy (EELS), High Resolution Transmission Electron Microscopy (HR-TEM) and Xray diffraction (XRD) techniques to characterize the interface reactions and phase evolutions in Al/Cu/CuO/Cu/Al reactive nanolaminates as a function of sputtered copper thickness. The Al and CuO films thicknesses are kept constant at 90 and 180 nm, respectively. The pure copper


4

Page 5 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


interfacial film thicknesses vary from 5 to 100 nm (same thickness as Al) corresponding to a mass ratio of interfacial pure Cu over Al from 15% to 78%. These samples make it possible to investigate the influence of the relative interfacial Cu amount on the diffusion and phase transformation processes that regulate the Cu enriched Al/CuO nanolaminates reaction properties. The microstructure and composition of all layers are investigated before and after annealing at 300 °C, with a focus on the spatial distribution of Cu and the thermally induced phase transformations and the formation of different Al-Cu intermetallic phases. The behavior of pure Cu in contact with Al thin films in Al/CuO nanothermite systems is described for the first time and shown to depend on the specific interface (Al onto Cu or Cu onto Al) and on the heat treatment. Therefore, this work provides a foundation for understanding and tailoring the addition of Cu into Al-based nanolaminates, which are a novel class of interesting energetic layers, for a range of applications such as component joining 26 and ignition 7. MATERIALS AND METHODS Materials. Several Al/Cu/CuO/Cu/Al multilayers were deposited on silicon wafer using a magnetron sputtering chamber, previously described in

15

. Multilayer films were fabricated by

alternating the Al and CuO depositions. The sputter targets (Al and Cu) were acquired from Neeco with a purity of 99.999%. Aluminum was sputtered using a DC power supply at 800 Watts power in an argon background pressure of 1×10-3 mbar. CuO was sputtered from Cu target using a DC power supply with 400 Watts power in an oxygen pressure (O2/Ar flow ratio: 100/30). The successive deposition of CuO and Al were carried out without venting the chamber. To deposit Cu at each interface between each Al and CuO cycle, Cu was sputtered from a Cu target using a DC power supply with a power of 400 Watts under an argon pressure of 5×10-3


5


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 30

mbar. A delay of 15 minutes was imposed between each deposition cycle to allow the sputtering heat to dissipate. The chamber was fully pumped out (base pressure ~ 2×10-7 mbar) after each cycle (Al or CuO deposition) to minimize cross contamination as well as aluminium or copper oxidation by residual oxygen. Prior to deposition, the initial surfaces of the silicon wafers were cleaned by a 5-minute plasma treatment (800 Watts) to remove impurities and contaminants. To obtain free-standing nanolaminates, the silicon wafers were spin-coated with a layer of photoresist (NLOF-5µm) and baked at 110 °C for 90 sec prior to nanolaminate sputter deposition. After the nanolaminates deposition, the photoresist was dissolved in acetone to release the multilayers. Two different types of multilayer samples were prepared: (a) a three-layer Al/CuO/Al nanolaminate, and (b) a five-layer Al/Cu/CuO/Cu/Al nanolaminate obtained by adding a Cu deposition at both Al and CuO interfaces. In all samples, the thicknesses of the Al and CuO layers were fixed at 90 and 180 nm, respectively, so that foils were Al rich (oxygen to aluminum mass equivalence ratio, ∅ equal to 2). This choice emanates from a previous experimental study [9] showing that the maximum enthalpy in Al/CuO nanolaminate reaction is obtained for ∅=2. For samples (b), different thicknesses of pure Cu were sputtered at the interfaces (5 nm, 10 nm, 20 nm, 50 nm and 100 nm) to probe the dependence of the overall nanolaminate properties on the Cu/Al ratio. All samples were deposited starting with the Al metal layer first and ended with an Al top layer. Thermal Analysis. The exothermic reactions of nanolaminates were characterized by DSC on a Mettler-Toledo device equipped with a HSS8 sensor in the temperature range of 30 to 700 °C under a constant heating rate (10 °C/min) in an Ar atmosphere. Ar is purified by passing through


6

Page 7 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


an oxygen trap (Supelco) giving a purity > 99.999%. About 3-4 mg of a self-standing foil, prepared as described in the previous section and released from substrate, was placed in a 30µL platinum pan. The samples were weighted before and after the analysis and have a mass loss < 0.0006 mg. For DSC, after the first heating cycle, the sample was cooled down to room temperature and then heated again at the same heating rate. This second analysis was used to correct the baseline. It was assumed that the bulk heat capacity of the sample did not change between the first and the second heating. Structural and chemical characterization. Grazing Incidence X-Ray Diffraction (GI-XRD) experiments were performed in a Bruker D8 Discover system using the k-alpha1 line of a Cu source. The grazing angle was fixed to 1.2° and two-theta collection angle was ranged from 10 to 80° with a 0.02° step and a dwell time of 0.1s per point in all cases. XRD experiments were carried out on as-deposited (just after sputtering) and annealed samples at 300 °C, to detect the formation of Al-Cu phases below 300 °C and their dependence with initial interfacial Cu content. Scanning Transmission Electron Microscopy (STEM) and Electron Energy-Loss Spectroscopy (EELS) combined with High Resolution Transmission Electron Microscopy (HR-TEM) techniques were used to analyze the morphology, chemical composition and phase evolutions of Al/Cu/CuO/Cu/Al multilayers with temperature. The spatial resolution is estimated at 1 nm corresponding to the scan step of the EELS profile. The samples were heated up to 300 °C at a rate of 10 °C/min in-situ thanks to a Gatan 652 double-tilt heating holder. TEM experiments were carried out in cross-section samples prepared by Focused-Ion Beam (FIB) process in a FEI Helios Nanolab. STEM – EELS experiments were performed in a Tecnai F20 FEG microscope, operated at 200 kV, equipped with an image-aberration corrector. GIF Tridiem and HR-TEM images were carried out in a Hitachi HF3300 microscope operated at 300 kV.


7


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 30

RESULTS Thermal properties. All nanolaminate foils release heat upon heating, as observed in the DSC traces shown in Figure 1, with a major exothermic peak rising after 400 °C. The DSC traces clearly indicate that the thermal decomposition varies substantially as a function of the interfacial Cu thickness. In other words, the Al/CuO nanolaminates thermal decomposition greatly depends on the amount of pure Cu added into the system. The Al + CuO heat of reaction generated below 700°C, calculated by integrating the exothermic peaks over time in the 400 - 690 °C range and normalizing with respect to the foil mass, is plotted as a function of the interfacial Cu film thickness in Figure 2a. Interestingly, we can see that the maximum heat of reaction is obtained when 10 nm Cu is deposited, i.e. 26% wt of Cu. Above this amount, there is an expected decrease in heat of reaction with increasing Cu thickness as illustrated by theoretical heat of reaction plotted in the right axis of Figure 2a. Without copper between Al and CuO, the main exothermic peak is observed at 563 °C with an onset at around 550 °C. This peak is attributed to the aluminothermic reduction of CuO (3CuO + 2Al → 3Cu + Al2O3). In contrast, the incorporation of pure copper at both interfaces induces a shift of the onset (~ 475 °C) and reaction (~ 535 °C) temperatures (see Figure 2b). In addition, the incorporation of pure copper induces the apparition of a weak exothermic peak at low temperature between 150 and 300 °C (see arrows in Figure 1) and different endotherms. As the Cu thickness increases, the intensity of the low-temperature peak is also increased and shifted to higher temperature, from 146 °C (for 15% wt Cu) to 256 °C (for 78% wt Cu). This weak lowtemperature peak can be attributed to the formation of Al2Cu alloys

27

leading to different

endotherms which are in accordance with the phase diagram of the Cu-Al binary system given in


8

Page 9 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3. These solid phase transformations result in the decomposition of supersaturated solid solutions and the formation of Al2Cu alloy and are consistent with the lowering of the onset temperatures to 475 °C in samples with interfacial Cu. This temperature (475 °C) is associated with the start of the precipitation of the Al2Cu phase. As seen in the Cu-Al binary phase diagram of Figure 3, the first endothermic peak at ~ 548 °C seen for Al/5-Cu/CuO, Al/10-Cu/CuO and Al/20-Cu/CuO can be associated with the eutectic point. The stacks Al/5-Cu/CuO and Al/10-Cu/CuO are hypoeutectic, characterized by a mixing of solid Al:Cu and liquid Al:Cu solutions. In turn, the hypereutectic Al/20-Cu /CuO stacks involve the mixing of the Al2Cu alloy with a liquid Al:Cu solution. The second endotherm at ~ 591 °C is observed in all samples with interfacial copper, except samples with 100 nm of pure Cu, and is attributed to the complete dissolution of the Al2Cu alloy into the liquid phase. It corresponds to the peritectic reaction, which is in good agreement with the observation that the Al2Cu phase is stable up to 591 °C 28. The last endotherm (~ 628°C) corresponds to the melting of remaining Al in all fuel rich samples (∅=2), as is usually observed in conventional fuel-rich Al/CuO nanolaminates. Table 1 summarizes the endotherms assignments for the different samples. In summary, the deposition of thin layers of Cu at both Al/CuO interfaces induces a dilution of Cu into Al and in consequence a decrease of the Al + CuO reaction onset temperature with a clear apparition of two aluminum-copper related endotherms at 548 and 591 °C. However, for Cu thickness > 50 nm, the overall system thermal response is completely different from all Al/CuO systems, more similar than those of Al/Cu/Al systems. Considering our multilayer


9


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 30

design, the optimized Al + CuO aluminothermic reaction is found for a Cu thickness of 10 nm corresponding to 26 % wt of pure Cu with respect to Al. The nature of phases responsible for the reduced onset is examined next with X-ray diffraction. XRD Analysis. X-ray diffraction analysis was performed on all Al/Cu/CuO/Cu/Al nanolaminates deposited on silicon, comparing samples freshly sputtered and annealed at 300 °C at the heating rate of 10 °C/min under N2 atmosphere. The diffraction patterns before and after annealing are shown in Figure 4 and a summary of the XRD assignments is given in Table 2. After Sputtering: the X-ray diffraction pattern of the reference sample (Al/CuO/Al) indicates that there is an overall preferential CuO and Al grain orientation along the (111) direction corresponding to the growth direction of the Al/CuO/Al multilayer. After annealing, a similar pattern was obtained with an adjustment of the peak towards the bulk material canonical values. This slight shift can be attributed to the interfacial strain release induced by the Al and CuO lattice mismatch

29

upon heating conditions. When a thin Cu layer is added at Al and CuO

interfaces, three additional peaks (referenced as + in Figure 4, blue curves) appear at ~ 43.4°, ~50.4° and ~74.3°. Their intensity increases with increasing Cu thickness. After annealing: after annealing at 300 °C, there is a clear evolution in the diffraction patterns mainly associated with the Cu and Al reactions at interfaces. An Al2Cu phase is clearly observable (labeled ♠ in Figure 4, red curves) in all samples with pure Cu sputtered at interfaces. As expected, the main Al2Cu peak at 20.7° strengthens with increasing Cu thickness except for sample Al/100-Cu/CuO, where more complex transformations occur as can be anticipated from the phase diagram at high copper concentration (see Figure 3). In Al/50Cu/CuO, AlCu (labeled ♣ in Figure 4) is also formed in addition to Al2Cu.


10

Page 11 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


XRD analyses of the different phases formed after annealing at 300 °C are totally consistent with the Al-Cu phase diagram in which Al2Cu nanocrystals should be formed in all conditions where Cu/Al is lower that 70% wt at 300 °C. It confirms that the weak exotherms seen in the DSC diagram below 300 °C can be attributed to alloying 19,28. Along this line, it is to be noted that for Al/100-Cu/CuO sample, XRD indicates the presence of a Cu-rich Al:Cu alloy, namely Al4Cu9. An important remaining question is the size of these alloys since we are dealing with nanoscale interfaces. Imaging with a technique like transmission electron microsocopy is the most direct way to examine this issue. TEM Analysis. Each interface of all the samples prepared in this work was characterized by TEM techniques. Here, we only report the main observations obtained for two samples only for clarity, as they are representative of the results: Al/20-Cu/CuO (41% wt Cu) and Al/50-Cu/CuO (63% wt Cu). To analyze the phase and structure evolution with temperature, we performed a systematic study using a dedicated heating TEM holder. A detailed analysis and a subsequent comparison was made using STEM-EELS and HR-TEM images of the samples right after sputtering at room temperature (RT) and the samples after heating to 300 °C with a heating rate of 10 °C/min and cooled down to RT for imaging and energy analysis. The results are reported for both the bottom interface (Al/Cu) formed by depositing Cu on the Al layer and the top interface (Cu/Al) formed by depositing an Al layer on the thin interfacial Cu layer. As deposited: Figure 5 shows the cross section images just after sputtering (at room temperature) for two samples with different thicknesses of interfacial Cu: (a) Al/20-Cu/CuO and (b) Al/50-Cu/CuO. Images are acquired in High Angle Annular Dark Field (HAADF) – STEM mode. HAADF imaging is an incoherent mode of imaging whereby the image contrast is


11


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 30

approximately proportional to the atomic number (~Z2) of constituent atoms due to Rutherfordlike scattering 30. The CuO layer (Figure 5, light grey zone) is clearly polycrystalline with a coarsened morphology with disordered columnar grains. The Al layer (Figure 5, black zone) features very well defined rectangular shape grains (optical contrasts seen in Al layer) with sizes ranging from 18 to 23 nm. In both images, the Cu interfaces are clearly visible as the brightest thin layer. The optical contrast of the Cu interfacial region separating the top CuO layers from aluminum film is attenuated due to the roughness of the columnar CuO. In the Al layers, the black regions in the STEM images (Figure 5) correspond to pure Al and the brighter regions (marked by red arrows) correspond to the Al:Cu solid solution phase with Cu content of ~9 %. Figure 5c and 5d are the EELS analyses giving the chemical profile across both interfaces for Al/20-Cu/CuO and Al/50-Cu/CuO sample, respectively. The blue curve represents the relative composition of aluminum relative to copper and red curve represents the relative composition of copper relative to aluminum. As seen in Figure 6a, the bottom Al/Cu and top Cu/Al interfaces are scanned starting in a lightdark Al grain, moving through the interface into the Cu layer. In both interfaces, we detect both Cu and Al signal over 12 nm thick, which clearly indicates atomic intermixing. The Cu intermixing penetration is 12 ± 3 nm into the top Al film (Al/Cu interface) and 18 ± 5 nm into the bottom Al film (Cu/Al interface). This penetration depth is independent of the interfacial Cu thickness. The apparent difference in intermixing depth profile can be attributed to two contributions. The main contribution is the difference in the interface roughness. The bottom interface (copper onto aluminum) is flat (roughness

Self-Organized Al2Cu Nanocrystals at the Interface of Aluminum

Recommend Documents