Oxidation behaviour of copper nanofractals produced by soft-landing of size-selected nanoclusters

Abstract

We report the oxidation dynamics of a copper nanocluster assembled film, containing fractal islands, fabricated by the soft-landing of size-selected copper nanoclusters with an average diameter of 3 nm. The time evolution of the spontaneous oxidation of the prepared film in air at room temperature (RT) was studied. A compositional analysis of the film was carried out in an ultra-high vacuum (UHV) deposition chamber using an in situ X-ray photoelectron spectroscopy (XPS) system. The morphological aspects of the deposited film were studied with a high resolution scanning electron microscope (SEM) and an atomic force microscope (AFM). We report the spontaneous production of highly pure (∼95%) and technologically appealing nano-crystalline Cu₂O within 300 seconds of air exposure. The crystalline structure was probed using high resolution transmission electron microscopy (HRTEM) and the optical properties were studied using a cathodoluminescence (CL) device attached to a SEM.

---

1 Introduction

Nanocluster deposition at very low energy (soft-landing) is a very powerful technique that offers the possibility of fabricating size controlled nano-islands or quantum dots^1–3 with predictable spatial distributions. This technique can also produce porous^4–6 or fractal^7,8 nanostructures with incredibly high surface to volume ratios. Due to their potential applications in catalysis,^9–12 efficient solar energy conversion,^13,14 antimicrobial coatings,^15–17 magnetic memory arrays,¹⁸ etc., studies on these types of nanostructures are very promising from both a basic research and technological perspective. Copper is a material with high abundance in nature¹⁹ and is a promising candidate for use in most of the above mentioned fields. Moreover, in nanocluster deposition using magnetron based gas aggregation type sources, copper is one of the most suitable materials due to its high sputtering yield. Pure (bare) copper is very sensitive to oxidation when it is exposed to ambient atmospheric conditions and generally forms a mixture of two types of oxides, cuprous(I) and cupric(II), very quickly. In its nanodimensional form the reaction rate will be much higher but the time scale of the phenomenon is not known in air or ambient conditions. Although a few studies on the oxidation of thin continuous metallic films have been carried out,^20,21 porous and fractal structures with very high surface to volume ratio are very complicated to address and certainly deserve more attention.

In this work we study the oxidation dynamics of nanocluster assembled sub-monolayer metallic film, containing nano-fractal islands of copper due to atmospheric exposure. The fractal islands were fabricated by the near soft-landing of size-selected copper clusters on highly oriented pyrolytic graphite (HOPG). HOPG has been employed as the substrate because firstly, it is chemically inert and secondly, soft-landed clusters develop better mobility on their surfaces for diffusion and forming larger aggregates (mostly fractals).²² The chemical compositions of the as-grown, as well as air-exposed, fractal aggregates were investigated using in situ X-ray photoelectron spectroscopy (XPS), which indicated the spontaneous formation of pure Cu₂O nanocrystals within 300 s of exposure to ambient conditions. The crystalline structure was confirmed by HRTEM analysis followed by a CL study. Systematic XPS characterization of the sample after different durations of air exposure gave an idea of the time-scale of the oxidation of the copper nanofractals, as well as giving an indication of the stability of the Cu₂O nanocrystals. As Cu₂O nanocrystals can be used in catalysis,²³ this is very crucial information. It also gives experimental evidence for the asymptotic behaviour²⁰ of the time evolution of oxide growth in ultra thin and highly porous metallic films as fractal structures have very high surface areas.

2 Experimental

The deposition of size-selected copper nanoclusters with an average diameter of 3 nm (or about 1100 Cu-atoms per cluster) at very low energy onto the HOPG substrate was performed using a state of the art ultra high vacuum compatible nanocluster source with a subsequent mass selection process.³ The nanocluster source used was a gas aggregation type source based on magnetron sputtering, which is capable of producing ionized clusters with high flux and a distribution of sizes. The produced clusters were transported by only a pressure gradient and after size selection, they were deposited on a substrate located in a large chamber, very far from the source. This arrangement ensured deposition at very low energy. It should be noted that QMF did not contribute to the kinetic energy of the clusters during size selection in the direction along the beam or the path of the clusters. The deposition system being equipped with an in situ XPS facility provided the opportunity to chemically analyze the deposited film (at a pressure of <5 × 10⁻¹⁰ mbar) after deposition without exposing the sample to the ambient atmospheric conditions, including air. So, just after deposition one set of XPS measurements was performed. The same sample was then taken out of the vacuum and placed in the laboratory atmosphere for a few different time spans and subsequent XPS measurements were carried out. The deposited material (Cu) started to oxidize when exposed to the atmospheric conditions and a gradual change in the oxygen to copper ratio on the substrate (HOPG) surface occurred (discussed quantitatively later). The relative humidity of the laboratory was maintained at 37–40% and the temperature was kept at 25 °C during the whole experiment.

Deposition was carried out for 30 min using a 2 nA cluster current measured at the quadrupole mass filter (QMF). Less than one microgram of copper was deposited in this process. So to detect appreciable changes in the oxygen to copper ratio of the sample, the substrate needed to be chosen with extreme care, as oxidation of the substrate could complicate the conclusions based on the oxidation of the desired material. Freshly cleaved HOPG was the best candidate in this respect as it remained unaffected for an extended time during exposure to the atmosphere and this was ensured before starting the experiment. XPS measurements were performed using AlKα (energy = 1486.6 eV) X-rays and the ejected photoelectrons were analysed by a Class150 (VSW Ltd., UK) hemispherical electron analyser at an operating pass energy of 20 eV and a scan rate of 0.15 eV per step with a dwell time of 1 second. We took our XPS measurements for the Cu2p and O1s peaks while keeping all other experimental conditions the same as before. The collected data were analyzed using commercially available CasaXPS software.²⁴ Our main goal was to identify the change in oxygen to copper ratio following exposure to the atmosphere for different time spans. For this purpose we needed a high signal to noise ratio. So, to increase the precision of the measurements, we took 20 sweeps for each high resolution spectrum. The morphological aspects of the cluster assembled film were characterized using high resolution scanning electron microscopy (SEM) and atomic force microscopy (AFM). The crystalline properties of the nanostructures were examined by high resolution transmission electron microscopy (HRTEM) of the concurrently deposited copper nanoclusters on an amorphous carbon film coated on a standard TEM grid. CL measurements were performed by a Gatan CL system attached to the SEM using 10 keV electron beam excitation. The details of the CL-SEM set-up can be found in ref. 25.

3 Results and discussion

As shown by the survey XPS spectra in Fig. 1 and the high resolution spectra in the inset, no signals for any oxygen adsorption onto the HOPG substrate were detected, even after 60 h of exposure to ambient conditions, which is a much longer span than the longest exposure of the deposited film. This ensures very sensitive detection of oxygen adsorption onto the intended metal structures. The as-deposited sample was investigated first (without ambient exposure) to initialize the experiment. The survey spectrum from the XPS measurement of the Cu-nanoclusters as-deposited on HOPG is shown Fig. 2(a). The Cu2p doublet peaks at ∼953 eV and ∼933 eV, designated as Cu2p_1/2 and Cu2p_3/2 respectively, are shown in Fig. 2(b) and were obtained using high resolution XPS analysis. We can clearly see from the high resolution spectrum [Fig. 2(c)] that there is no signal for O1s present in the energy range of 540–525 eV, which confirms the fact that pure metallic fractals were formed without any oxidation during deposition. Also there are no other elements present which may contribute to oxygen adsorption.

Fig. 1 XPS spectra of the HOPG substrate after exposure to atmospheric conditions for different time spans. The data are presented in increasing order of time as indicated by the arrow. High resolution spectra for the BE range of 540–525 eV are shown in the inset.

Fig. 2 (a) The survey spectrum of the as-deposited copper nanocluster film with the positions of the Cu2p and O1s peaks indicated. (b) High resolution X-ray photoelectron spectrum for Cu2p showing the splitting between Cu2p_1/2 and Cu2p_3/2 as 19.9 eV. (c) High resolution spectrum of the as-deposited sample in the 540–525 eV binding energy range.

SEM imaging gives accurate 2D topographical information on the nanometer length scale but for out of plane or height information, AFM measurements must be carried out. The SEM imaging and AFM measurement results are shown in Fig. 3. The coverage of the deposited film was estimated as ∼22% from SEM studies. Fractal islands were formed with a fractal dimension of ∼1.7, which was obtained using the standard box counting method.²⁶ Fractal objects of finite volume contain larger surface areas compared to a compact object of the same amount of material.²⁷ The mechanism of fractal formation by nanocluster deposition has already been discussed in reported literature.^7,8,28 Nanoclusters deposited at low energy diffuse over a substrate surface and when they encounter each other they either completely merge or just stick to each other giving rise to different final morphologies. Merging or complete coalescence lead to the formation of a compact island structure,^33,35 while just sticking together leads to ramified or fractal island formation. The merging or not merging of clusters and islands over the surface is governed by the ‘critical size of coalescence’, which is related to the liquid–solid transition of the clusters.³⁶ As observed from the AFM measurements [Fig. 3(b) and (c)], the average height of the fractal branches was ∼5 nm. It should be noted that the wavy nature seen in background of the AFM image is an artifact and during height estimation low height values were therefore ignored. As the metallic nano-structures under study have larger surface to volume ratios than other types of compact structures, they have more surface atoms. Hence very high rates of reaction are expected for such systems at ambient oxygen pressures. In this situation, in order to detect oxidation of the deposited material, we needed to examine the small changes in the oxygen signal of the sample in small temporal steps. This was achieved by choosing appropriate experimental conditions, such as a substrate which does not get oxidized after a much longer exposure to atmospheric conditions compared to the time scale of the main experiment (discussed at the beginning of this section) and an ultra high vacuum level. In Fig. 4(c) the oxygen peak (O1s: 535–525 eV range of binding energy) intensity grows (with respect to the background signal) with increasing duration of air exposure. By merely looking at the oxygen peak it is impossible to quantitatively determine the oxygen adsorption phenomenon (due to many experimental reasons, the photoelectron intensity can vary for different sets of collections) so the ratio of the number of oxygen to copper atoms was calculated from the XPS data for each step of oxidation. From the CasaXPS database and literature it was found that the relative sensitivity factor (RSF) values of the Cu2p_3/2 and O1s peaks were 16.7 and 2.93 respectively for AlKα excitation. Using these values and the ratios of the areas under the high resolution Cu2p_3/2 and O1s peak curves after Tougaard type background subtraction,²⁹ the ratio of oxygen to copper atoms has been calculated. Fig. 4(a) shows the ratio of oxygen to copper atoms on the sample surface plotted against different durations of atmospheric exposure. The increase in the ratio is entirely due to oxidation (oxygen adsorption) of the deposited copper, without any contribution from substrate oxidation, as was confirmed earlier in Fig. 1. Initially the rate of oxidation was high and within 300 s the O/Cu atomic ratio (r) assumed a saturation value. The data are fitted with a function (indicated by a red line in Fig. 4(a)) of the form:

r = r₀[1 − exp(−bt)] (1)

where r₀ should have a maximum value of 0.5 if only Cu₂O is formed and can have a maximum value of 1.0 if finally only CuO is formed and b is a constant that depends on the partial pressure of oxygen.³⁰

Fig. 3 (a) SEM image and (b) AFM image of the nanocluster assembled film on the HOPG substrate. (c) A height profile, along the line shown in (b), shows the features are on average 5–6 nm high (neglecting the uneven background). Black scale bar in the SEM image = 200 nm.

Fig. 4 (a) Atomic ratio variation with different durations of atmospheric exposure. (b) High resolution X-ray photoelectron spectra for Cu2p. The spectra here show no satellite peaks between the two main peaks (Cu2p_1/2 and Cu2p_3/2). (c) High resolution spectra in the 535–525 eV binding energy range collected from the cluster assembled film after different air exposure times.

From the fitted curve, the value of r₀ was found to be ≈0.478 ± 0.026, a value very near to 0.5 as obtained after 300 s of air exposure. Therefore, in the probing depth of the XPS study, all the copper atoms in the fractal nanostructures were oxidized to Cu₂O and achieved saturation. Oxidation of metallic copper in the presence of oxygen occurs through the formation of Cu₂O and CuO. Thermodynamically, when the partial pressure of oxygen is greater than the dissociation pressure of the oxides, only oxides are formed. The kinetics of any chemical reaction depend on a trade off between the changes in the free energy of formation of the different products in that reaction. The free energy changes associated with the formation of Cu₂O and CuO are 122 187 J Mol⁻¹ and 99 311 J Mol⁻¹ respectively and from these values the oxygen dissociation pressures can be calculated as ∼10⁻¹⁷ and ∼10⁻¹³ in atmospheric pressure units.³¹ So if the oxygen pressure is high enough (>10⁻¹³ atmospheric pressure), there is a high chance of CuO formation quickly after Cu₂O formation. However, if the oxygen pressure is lower than ∼10⁻¹³ but higher than ∼10⁻¹⁷ only Cu₂O can form. In the present case, the oxidation of fractal structured nanoislands composed of copper nanoclusters in air, only the formation of Cu₂O was observed and it remained stable for at least 30 min. From the HRXPS spectra of Cu2p (Fig. 4(b)), it was confirmed that no CuO was formed. The presence of CuO generates shake up peaks in the observed binding energy range, along with broadening of the Cu2p_1/2 and Cu2p_3/2 peaks.³² The oxide growth on the metallic non continuous film comprised of fractal nanostructures with a branch height of ∼5–6 nm cannot be explained by the theories applicable to continuous thick as well as thin films. This is because nanofractals have a much higher surface to volume ratio and hence a much higher number of surface atoms. Moreover, the soft-landing of nanoclusters produces highly porous film.⁴ In this scenario, if we consider that the rate of formation of oxide molecules is directly proportional to the remaining unoxidised Cu atoms then, dN/dt = b(N₀ − N). So the number of oxidized atoms (N), at an instant of time (t), is N = N₀(1 − exp(−bt)), where N₀ is the total number of Cu atoms in that volume. The atomic ratio of oxygen to copper for Cu₂O formation is: . The crystallinity of the nanostructures produced by concurrent low energy cluster deposition on amorphous carbon film along with the HOPG substrate, was examined by HRTEM. The HRTEM image in Fig. 5(a) shows a lattice spacing of 2.4 Å, which corresponds to the (111) interplanar spacing of Cu₂O. It should be noted that for presentation purposes the lattice fringe on only one nano-structure is zoomed in, although all of the nanostuctures gave fringes with the same value of spacing.

Fig. 5 (a) HRTEM image of deposited copper nanocluster amorphous carbon film showing a lattice spacing of 2.4 Å. (b) CL spectrum obtained from the material deposited on the HOPG substrate.

A cathodoluminescence study [Fig. 5(b)] of the deposited film on the HOPG substrate revealed a clear peak at ∼560 nm, which corresponds to an energy of 2.21 eV. Bulk Cu₂O is a direct band gap semiconductor with a band gap of 2.17 eV (570 nm) but we measured a wider gap due to the nanometer dimensions of the nanocrystals. It must be noted that energy bands can only form in crystalline materials. Thus, the CL data support the TEM and XPS data in confirming that Cu₂O was formed.

4 Conclusions

In summary, the oxidation behavior of a nanofractal structure, formed by soft-landed copper nanoclusters, was investigated on a time scale of seconds. Quick (<300 s) formation of crystalline Cu₂O with a very high production yield (∼95%) was observed just by atmospheric exposure in the laboratory. In the signature cathodoluminescence spectrum, band gap widening was detected due to the nanometer dimensions of the Cu₂O crystals, unlike our previous investigation where the formed islands were 3 dimensionally large.³³ The studies presented in this paper will help to achieve a better understanding of the time scale of the oxidation of Cu, which is important both technologically and fundamentally. Cu₂O crystals of nanometer size are a technologically promising material,^23,34 they can be obtained in pure form just by oxidation in atmospheric conditions and remain stable for more than 30 minutes in ambient conditions.

Acknowledgements

Authors are grateful to Prof. Tapas K Chini, Dr P. Das and Mr Achyut Maity for SEM and CL studies, to Dr B. Satpati for TEM study, to Prof. D. Ghose and Mrs D. Chowdhury for AFM measurements and finally to Mr Debraj Dey for technical help concerning the maintenance of the nanocluster deposition and XPS facilities.

References

1. N. Grönhagen, T. T. Järvi, N. Miroslawski, H. Hövel and M. Moseler, Decay Kinetics of Cluster-beam-deposited Metal Particles, J. Phys. Chem. C, 2012, 116, 19327–19334.
2. S. Mondal, B. Satpati and S. R. Bhattacharyya, Formation of Monodispersed Films from Size-Selected Copper Nanoclusters, J. Nanosci. Nanotechnol., 2015, 15, 611–615 (also: arXiv preprint arXiv:1307.0427, 2013).
3. S. Mondal and S. R. Bhattacharyya, Performance of a size-selected nanocluster deposition facility and in situ characterization of grown films by X-ray photoelectron spectroscopy, Rev. Sci. Instrum., 2014, 85, 065109.
4. G. Palasantzas, S. A. Koch and J. T. M. de Hosson, Growth front roughening of room-temperature deposited copper nanocluster films, Appl. Phys. Lett., 2002, 81, 1089–1091.
5. V. N. Popok, I. Barke, E. E. B. Campbell and K.-H. Meiwes-Broer, Cluster–surface interaction: from soft landing to implantation, Surf. Sci. Rep., 2011, 66, 347–377.
6. C. Binns, Nanoclusters deposited on surfaces, Surf. Sci. Rep., 2001, 44, 1–49.
7. L. Bardotti, P. Jensen, A. Hoareau, M. Treilleux and B. Cabaud, Experimental observation of fast diffusion of large antimony clusters on graphite surfaces, Phys. Rev. Lett., 1995, 74, 4694–4697.
8. P. Jensen, Growth of nanostructures by cluster deposition: experiments and simple models, Rev. Mod. Phys., 1999, 71, 1695–1735.
9. T. Ressler, B. L. Kniep, I. Kasatkin and R. Schlogl, The microstructure of copper zinc oxide catalysts: bridging the materials gap, Angew. Chem., Int. Ed., 2005, 44, 4704–4707.
10. A. Murugadoss, E. Sorek and M. Asscher, Structure and Composition of Au–Cu and Pd–Cu Bimetallic Catalysts Affecting Acetylene Reactivity, Top. Catal., 2014, 57, 1007–1014.
11. M. B. Thathagar, J. Beckers and G. Rothenberg, Copper-catalyzed Suzuki cross-coupling using mixed nanocluster catalysts, J. Am. Chem. Soc., 2002, 124, 11858–11859.
12. M. Nesselberger, M. Roefzaad, R. Fayal Hamou, P. Ulrich Biedermann, F. F. Schweinberger, S. Kunz, K. Schloegl, G. K. H. Wiberg, S. Ashton, U. Heiz, K. J. J. Mayrhofer and M. Arenz, Nat. Mater., 2013, 12, 919–924.
13. H. M. Al Dosari and A. I. Ayesh, Nanocluster production for solar cell applications, J. Appl. Phys., 2013, 114, 054305.
14. C. Clavero, Plasmonic conversion of solar energy, TBD, Lawrence Berkeley National Laboratory, LBNL Paper LBNL-6249E, 2014, http://escholarship.org/uc/item/7w32p1t6.
15. N. Cioffi, L. Torsi, N. Ditaranto, G. Tantillo, L. Ghibelli, L. Sabbatini, T. Bleve-Zacheo, M. D’Alessio, P. Giorgio Zambonin and E. Traversa, Copper nanoparticle/polymer composites with antifungal and bacteriostatic properties, Chem. Mater., 2005, 17, 5255–5262.
16. M. S. Usman, M. E. E. Zowalaty, K. Shameli, N. Zainuddin, M. Salama and N. A. Ibrahim, Synthesis, characterization, and antimicrobial properties of copper nanoparticles, Int. J. Nanomed., 2013, 8, 4467.
17. A. Gupta and S. Silver, Silver as a biocide: Will resistance become a problem?, Nat. Biotechnol., 1998, 16, 888.
18. S. Sun, Recent advances in chemical synthesis, self-assembly, and applications of FePt nanoparticles, Adv. Mater., 2006, 18, 393–403.
19. http://www.copper.org/.
20. K. R. Lawless, The oxidation of metals, Rep. Prog. Phys., 1974, 37, 231–316.
21. N. Cabrera and N. F. Mott, Theory of the oxidation of metals, Rep. Prog. Phys., 1949, 12, 163–184.
22. K. A. Wepasnick, X. Li, T. Mangler, S. Noessner, C. Wolke, M. Grossmann, G. Gantefoer, D. H. Fairbrother and K. H. Bowen, Surface morphologies of size-selected Mo100 ± 2.5 and (MoO3) 67 ± 1.5 clusters soft-landed onto HOPG, J. Phys. Chem. C, 2011, 115, 12299–12307.
23. P. de Jongh and J. Kelly, Cu₂ O: a catalyst for the photochemical decomposition of water?, Chem. Commun., 1999, 1069–1070.
24. http://www.casaxps.com.
25. P. Das and T. K. Chini, An advanced cathodoluminescence facility in a high-resolution scanning electron microscope for nanostructure characterization, Curr. Sci., 2011, 101, 849–854.
26. R. Thouy, N. Olivi-Tran and R. Julien, Fragmentation of two-dimensional mass fractals by surface-diffusion sintering, Phys. Rev. B: Condens. Matter Mater. Phys., 1997, 56, 5321–5327.
27. T. Tel, Fractals, multifractals, and thermodynamics, Z. Naturforsch., A: Phys. Sci., 1988, 43, 1154–1174.
28. P. Jensen and N. Combe, Understanding the growth of nanocluster films, Comput. Mater. Sci., 2002, 24, 78–87.
29. M. P. Seah, Background subtraction: I. General behaviour of Tougaard-style backgrounds in AES and XPS, Surf. Sci., 1999, 420, 285–294.
30. J. Kruger and H. T. Yolken, Room temperature oxidation of iron at low pressures, Corrosion, 1964, 20, 29t–33t.
31. C. H. Xu, C. H. Woo and S. Q. Shi, The effects of oxidative environments on the synthesis of CuO nanowires on Cu substrates, Superlattices Microstruct., 2004, 36, 31–38.
32. M. C. Biesinger, L. W. M. Lau, A. R. Gerson and R. S. C. Smart, Resolving surface chemical states in XPS analysis of first row transition metals, oxides and hydroxides: Sc, Ti, V, Cu and Zn, Appl. Surf. Sci., 2010, 257, 887–898.
33. S. Mondal and S. R. Bhattacharyya, Morphological and optical properties of soft-landed supported nanoclusters: effect of rapid thermal annealing, Appl. Phys. A, 2014, 116, 1621–1626.
34. K. Akimoto, S. Ishizuka, M. Yanagita, Y. Nawa, G. K. Paul and T. Sakurai, Thin film deposition of Cu₂O and application for solar cells, Sol. Energy, 2006, 80, 715–722.
35. G. M. Francis, L. Kuipers, J. R. A. Cleaver and R. E. Palmer, Diffusion controlled growth of metallic nanoclusters at selected surface sites, J. Appl. Phys., 1996, 79, 2942–2947.
36. P. Milani and S. Iannotta, Cluster Beam Synthesis of Nanostructured Materials, Springer Series in Cluster Physics, Springer-Verlag, Berlin Heidelberg, 1999.

---

Footnote

† Present address: Maharaja Manindra Chandra College, 20 Ramkanto Bose Street, Kolkata-700003; E-mail: E-mail: smondal.xray@gmail.com

---