Chromatic Change in Copper Oxide Layers Irradiated with Low Energy Ions

Abstract

The color of a thin copper oxide layer formed on a copper plate was transformed from reddish-brown into dark blue-purple by irradiation with 5 keV Ar⁺ ions to a fluence as low as 1 × 10¹⁵ Ar⁺ cm⁻². In the unirradiated copper oxide layer, the copper valence state of Cu²⁺ and Cu⁺ and/or Cu⁰ was included as indicated by the presence of a shake-up satellite line in a photoemission spectrum. While for the irradiated one, the satellite line decreased in intensity, indicating that irradiation resulted in the reduction from Cu²⁺ to Cu⁺ and/or Cu⁰. Furthermore, nuclear reaction analysis using a ¹⁶O(d, p)¹⁷O reaction with 0.85 MeV deuterons revealed a significant loss of oxygen (5 × 10¹⁵ O atoms cm⁻²) in the irradiated layer. Thus, the chromatic change observed in the present work originated in the irradiation-induced reduction of a copper oxide.

1. Introduction

On ion beam experiments including materials analysis and modification with ion beams, a beam monitoring system is installed in the sample chamber to monitor the beam position and uniformity in the beam spot. Most of the beam monitor consists of a fluorescent plate, enabling real time visualization of a beam spot on the plate. A SiO₂ plate, for example, has been used for beam monitoring because of strong emission [1,2,3,4,5] in the visible range when irradiated with MeV-ion beams. A Cr-doped Al₂O₃ (e.g., AF995R, Desmarquest) is also suitable for beam profiling [6,7] for ion beams with energies larger than several hundred keV. The aforementioned materials are insulators and therefore electric charging takes places on the fluorescent plate irradiated with ion beams, resulting in deflection of ion beams in the vicinity of the fluorescent plate if their acceleration voltage is comparable to the charged potential of a few tens kilovolts [8]. This means that the fluorescent point would be different from the real position, and further the fluorescent point would not appear at all in the case of low energy ion beams with <10 keV. In addition to the fluorescent materials, phosphors such as ZnS based materials [9,10,11], which have been widely used for screens of a cathode-ray tube, are applicable to beam monitoring. Other candidates of inorganic luminescent materials can be found in the literature [12]. Powders of such materials, mixed with a conducting paste and deposited onto a conducting plate, are a candidate for ion beam monitoring materials. However, such powders cost very high, but their lifetimes are very short because radiation damage causes degradation of light-emitting efficiency. It is, therefore, not easy to view a beam spot of a low energy ion beam with energy of several keV on real-time.

On the other hand, non-real time beam monitoring can be conducted by the color change of materials irradiated with ion beams. A polyimide film is, for example, widely used to check both the position and uniformity of an ion beam, because blackening due to graphitization [13,14,15] occurs when the film is irradiated with ion beams. A polyimide film is, however, non-conducting and is inapplicable to the beam viewer for low energy ion beams with energies of several keV. The favorable beam viewer for low energy ion beams should be composed of electrically conducting materials. A metallic copper plate, even if a thin upper layer of copper oxide is present, is a good conducting material. The color of the copper plate covered with thin oxide is reddish-brown, largely different from polished metallic copper. The present authors made an attempt to fabricate a beam viewer in which the appearance of a beam spot turns shiny due to removal of the oxide layer by physical sputtering. In the irradiation apparatus with base pressure of 2 × 10⁻⁴ Pa, the shiny beam spot could be clearly recognized after irradiation with 5 keV Ar⁺ ions to a fluence of 1 × 10¹⁵ Ar⁺ cm⁻². Surprisingly, in the other irradiation equipment with base pressure of 2 × 10⁻⁶ Pa, the color of the beam spot turned dark blue-purple after irradiation under the same conditions above. In the present work, the chromatic change observed in the irradiation equipment with such a high vacuum is examined to fabricate a new type of beam viewer for low energy ion beams.

2. Materials and Methods

Oxygen-free copper plates of 10 × 10 × 1 mm³ (purity 99.99%) were supplied from NILACO, Tokyo, Japan. The Cu plate was put on a laboratory hot plate setting to 473 K in ambient air for 2 min to form the Cu oxide layer. The maximum temperature measured on the surface with chromel-alumel thermocouple was 440 K, somewhat lower than the setting temperature of 473 K, as recorded in Figure 1. The color of heat-treated Cu plate was reddish-brown. The samples were irradiated with 5 keV-Ar⁺ ions up to a fluence of 1 × 10¹⁵ cm⁻² using a 5 kV-ion gun installed in a vacuum chamber with a base pressure of 2 × 10⁻⁶ Pa. The ion incidence was normal with respect to the sample surface.

X-ray photoelectron spectroscopy (XPS) using Mg Kα radiation (hν = 1253.6 eV) was performed with a JEOL 9010 X-ray photoelectron spectrometer (JEOL, Tokyo, Japan) to analyze chemical states of Cu before and after the irradiation. The XPS analysis was carried out immediately after the irradiation to avoid the change in chemical states upon the ambient air exposure. Rutherford backscattering spectrometry (RBS) and nuclear reaction analysis (NRA) using the ¹⁶O(d, p)¹⁷O reaction were conducted for chemical composition analysis with 2 MeV He ions and 0.85 MeV deuterons, respectively, produced from the Van de Graaff accelerator of Hiroshima University. The standard sample for the NRA was a SiO₂ layer formed on a Si wafer (SiO₂/Si), which contains 5.8 × 10¹⁷ O atoms·cm⁻² determined by RBS.

3. Results and Discussion

Figure 2a shows a photograph of the sample surface irradiated with Ar⁺ ions to a fluence of 1 × 10¹⁵ cm⁻² using a 5 kV ion gun. The color of the sample changed from reddish brown to dark blue-purple at an irradiation spot. This darkening was visible through a viewing port, and started at the fluence as low as 10¹⁴ cm⁻². On the sample surface irradiated to a fluence of 10¹⁴ cm⁻², the color of the edge at the beam spot was not clear as can be seen in Figure 2b, indicating that the border between irradiated and unirradiated areas was not so abrupt. Thus, the uniformity of a beam intensity inside the beam spot could be estimated by the uniformity of color. The minimum fluence to recognize the beam spot with naked eyes will be examined for further discussion of the sensitivity and applicability of the chromatic change for a beam viewer.

The mechanism of the observed chromatic change was discussed below, along with RBS, NRA, and XPS data. Figure 3 shows the backscattering spectrum of the Cu oxide layer formed on a Cu plate before irradiation. The chemical composition and thickness of the oxide layer was estimated to be CuO_0.4 and 1.9 × 10¹⁷ CuO_0.4 cm⁻², respectively, by fitting a simulation spectrum to experimental data, where the program SIMNRA 6.06 [16] was used to obtain the simulation spectrum. The thickness of 1.9 × 10¹⁷ CuO_0.4 cm⁻² can be converted into 38 nm, much larger than the ion projected range of 5 nm predicted by the SRIM simulation [17], assuming the atomic density to be 5 × 10²² CuO_0.4 cm⁻³. The chemical composition CuO_0.4 indicates that the oxide layer contains the mixture of metallic copper and copper oxides.

Figure 4 presents the NRA spectra of the sample with and without irradiation. Peaks located around channel number of 260 correspond to protons emitted by the ¹⁶O(d, p)¹⁷O reaction. The peak intensities are (2.15 ± 0.05) × 10³ and (2.01 ± 0.04) × 10³ counts for the unirradiated and irradiated samples, respectively. The oxygen content in the irradiated sample was determined to be (7.1 ± 0.2) × 10¹⁶ O atoms·cm⁻² by the SiO₂/Si standard sample, smaller by 6% than that in the unirradiated sample ((7.6 ± 0.2) × 10¹⁶ O atoms·cm^–2). Thus, oxygen atoms were found to be released from the CuO_0.4 layer by the irradiation with 5 keV-Ar⁺ ions to a fluence of 1 × 10¹⁵ cm⁻². The SRIM simulation [17] predicts the sputtering yield of O in CuO_0.4 to be 4.9 O atoms·ion⁻¹, which means that the sputtered O atoms will be approximately 5 × 10¹⁵ O atoms·cm⁻², corresponding to a difference in the oxygen contents between the unirradiated and irradiated samples. The NRA results and the SRIM calculation suggest that the release of O atoms originates in physical sputtering. In the sputtering process of CuO_0.4 bombarded with 5 keV-Ar⁺ ions, approximately 1.4 × 10¹⁵ Cu atoms·cm⁻² as calculated by the SRIM will be lost, leading to the change in chemical composition from CuO_0.4 to CuO_0.38 in the layer.

Figure 5 shows XPS Cu 2p core level photoemission spectra (PS) of the CuO_0.4 layer with and without irradiation. Considering the inelastic mean free path of photoelectrons whose kinetic energy is approximately 300 eV, information about copper valence states can be obtained within 0.8 nm [18] below the surface by the photoemission spectra. As can be seen in Figure 5, a broad shake-up satellite line due to the charge transfer [19,20] appeared in the binding energy (BE) region from 937 to 941 eV of PS for the unirradiated sample and decreased in intensity after irradiation. More quantitatively, the intensity ratio of the satellite to the main line at BE of 932.4 eV was 0.16 and 0.10 for the unirradiated and irradiated samples, respectively. These results indicate that the unirradiated sample included the copper valence state Cu²⁺ and it was transformed into Cu⁺ and/or Cu⁰ (Cu⁺/Cu⁰) by irradiation. Thus the 5 keV-Ar⁺ irradiation reduced Cu²⁺ to Cu⁺/Cu⁰. The irradiation-induced reduction observed in the present work could be seen in the change in the shape of the Cu 2p_3/2 lines before and after irradiation.

Figure 6a,b depicts Cu 2p_3/2 lines for the samples before and after irradiation, respectively. Each spectral line consists of three components denoted by I, II, and III. The lines, after background subtraction by the Shirly method [21], were decomposed by three Voigt functions using a curve fitting procedure. The component I located at BE of 932.4 eV was assigned to be Cu and/or Cu₂O. The BE of Cu (932.6 eV) [22,23,24] is very close to that of Cu₂O (932.5 eV) [25,26,27,28], therefore the component I cannot be further decomposed by a curve fitting. The components II and III located at BE of 933.6 eV and 934.8 correspond to CuO [28,29,30,31] and Cu(OH)₂ [28,32], respectively. The fractions of each component obtained by the curve fitting are summarized in

Table 1. Compositions of Cu, Cu2O, CuO and Cu(OH)₂ determined by Cu 2p_3/2 photoemission spectra (PS) lines for the unirradiated and irradiated samples.

Samples	Compositions (%)
	Cu	Cu2O	CuO	Cu(OH)2
Unirradiated	73.8		16.5	9.7
	(65.7) 1	(8.1) 1
Irradiated	94.1		5.5	0.4
	(65.7) 1	(28.4) 1

¹ These values were obtained by the assumption that the composition of the analyzing layer was CuO_0.4 before irradiation and the fraction of Cu was unchanged after irradiation.

for the samples with and without irradiation. The fractions corresponding to copper valence state Cu²⁺ (CuO and Cu(OH)₂) decrease, while the fraction of Cu⁺/Cu⁰ (Cu₂O/Cu) increased by irradiation, indicating that the Ar⁺ irradiation reduced Cu²⁺ to Cu⁺/Cu⁰. This was consistent with the result deduced from the decrease in intensity of shake-up satellite line as described above.

The concentration ratio of Cu₂O to Cu can be indirectly determined by the atomic ratio O/Cu of analyzing layer using the fractions of three components. Of course, the O/Cu can be calculated by the intensity ratio of O 1s to Cu 2p PS lines. However, it is impossible to accurately determine the atomic ratio O/Cu because of the presence of adventitious carbon contamination that includes oxygen atoms in the outermost layer. Therefore, the atomic ratio O/Cu in the analyzing layer was assumed to be 0.4 that was determined by RBS, resulting in the concentration ratio Cu₂O/Cu of 8.2/65.7 for the sample without irradiation. These values are presented with parenthesis in Table 1. For the sample with irradiation, the atomic ratio O/Cu in the analyzing layer was definitely different from that in the oxide layer, and thus, the composition CuO_0.38 determined by the combination of RBS and NRA could not be used to calculate the fraction of Cu₂O/Cu. Therefore, the fraction of metallic copper was assumed to be unchanged after irradiation. In fact, Panzner et al. [33] demonstrated that the oxide CuO was reduced to Cu₂O, while the oxide Cu₂O was much more stable and no more reduction to Cu occurred under sputtering with 3–5 keV Ar⁺ to a low fluence. Then the concentration ratio Cu₂O/Cu was calculated to be 28.4/65.7 for the sample with irradiation. These values were also presented inside parenthesis in Table 1.

As described above, XPS analysis revealed that the 5 keV-Ar⁺ irradiation reduced Cu²⁺ to Cu⁺/Cu⁰ in the cuprate. This result is consistent with the previous studies that CuO thin films were transformed into Cu₂O by ion irradiation [33,34]. Next, the relationship between the reduction and the chromatic change was discussed briefly.

It is well known that the color of cuprous oxide (Cu₂O) powder is red. In addition, Cu₂O in the form of both the thin film [35,36] and nanoparticle [37,38] would be reddish considering their optical absorption spectra. On the other hand, the color of the irradiated layer was found to be dark blue-purple, different from the color of pure Cu₂O. Fredj and Burleigh [39] showed that the color of a copper oxide layer in which Cu₂O is primarily included varied from bare copper color to green depending on its thickness. Thus, they demonstrated the possibility that the Cu₂O layer exhibited the color other than red. One possibility for the chromatic change is the presence of Cu₂O phase in a metallic Cu phase. Another possibility is the change in a refractive index of the oxide layer accompanied by the change in chemical states of Cu and/or radiation damage by Ar ions. Unfortunately, the dominant effect on the chromatic change is unclear at present.

As mentioned in the Introduction, the beam spot in the sample irradiated with 5 keV-Ar⁺ ions using the other machine turned bare copper color due to the removal of an oxide layer by sputtering, different from the present work. The reason for this difference is unknown but may result from the initial thickness and composition of the oxide layer. Further studies with various thicknesses and compositions are needed to clarify the mechanism underlying the irradiation-induced chromatic change in copper oxide layers. In addition, the minimum fluence at which the chromatic change can be recognized is necessary to examine for the applicability of the new type of a beam viewer. Such investigations are now in progress.

4. Conclusions

The color of a thin copper oxide layer formed on a copper plate turned from reddish-brown to dark blue-purple by irradiation with 5 keV Ar⁺ ions to a fluence of 1 × 10¹⁵ Ar⁺ cm⁻². Nuclear reaction analysis revealed that a significant amount of oxygen (5 × 10¹⁵ O atoms·cm⁻²) was released from the irradiated layer. The reduction of cupric oxide (CuO) to cuprous oxide (Cu₂O) occurred in the layer after the irradiation as confirmed by the decrease in intensity of a shake-up satellite line and the change in the shape of a Cu 2p_3/2 line in photoemission spectra. The reduction led to the compositional change in the mixture of Cu/Cu₂O/CuO, which would result in the chromatic change.