Analysis of plasma degreased aluminium foil with XPS

The detailed spectrum for C 1s shows at least four peaks (Fig. 2). The main peak (C₁) is always centred at a binding energy of 286 eV. It points to acyclic or cyclic carbon compounds (C–C, C–H). The second compound (C₂) is ∼1·5 eV next to the main peak and shows single oxidised carbon compounds (C–O), which are a sign for hydroxyl groups. The next fraction (C₃) at ∼289·7 eV is an evidence for carbonyl groups (C = O) as a second step of carbon oxidation. Further oxidation of these carbonyl groups leads to carboxyl groups or carboxylic acid, which are shown in the last chemical shift (C₄) at ∼291 eV. These values are confirmed in several other studies with similar results (Table 2).14,15

Owing to the plasma treatment, all peaks of the C 1s orbital are reduced (Fig. 3). The lower oxidation states (C₁, C₂) experience a greater reduction in intensity than higher states. The plasma treatment shows only a small effect on the carbonyl peak. This may lead to the conclusion that hydrocarbons, which are too long or too heavy to be removed by the plasma treatment, get a higher oxidation state or polymerise.

The spectra of Al 2p allows to distinguish two peaks (Fig. 4). The main peak (Al₁) is centred at an electron binding energy of ∼73·5 eV and is associated with metallic aluminium. With a shift of ∼3 eV, there is a second peak (Al₂), which points to oxidised aluminium.16 This is the native oxide layer that typically forms on aluminium surfaces when exposed to air. The maximum depth in the sample from which the XPS gains information is <10 nm. Thus, the amount of aluminium detected by XPS is a first clue for the organic layer to have a thickness of <10 nm. There is nearly no change in the relative intensities of aluminium before and after the plasma treatment (Fig. 5).

The spectrum of oxygen shows one peak at an electron binding energy of ∼533·9 eV. It is mainly associated with aluminium oxide and with a small amount of aluminium hydroxide.16,17 Oxygen can be found in the native oxide layer as well as in the hydrocarbon layer. The amount of oxygen in the hydrocarbon layer is negligibly small compared to the aluminium oxide layer.

After plasma treatment, at least four peaks are recorded in the spectrum of the N 1s region. The first peak (N₁) is centred at an electron binding energy of 399·2 eV; compounds of nitrogen with aluminium form nitrides (AlN) most likely at these binding energies. The next peak (N₂) is centred at a shift of 2·2 eV. The best assignment is a nitric oxide radical (NO⁻) that connects to chains of the hydrocarbon layer. The third peak (N₃) has a chemical shift of 5·4 eV to the first peak (N₁) and can be identified as a nitrite compound ( ). The fourth peak (N₄), at an electron binding energy of 408·7 eV, points to nitrate compounds ( ).

This production of NO ions depends on the heat, power and treatment time of the DBD. Nitrite is an intermediate for the oxidation of a nitrogen radical to nitrate. Therefore, a huge amount of nitrate but only a small amount of nitrite can be detected after plasma treatment times of ∼2 min. A study of Helmke et al. revealed the presence of these compounds by analysing the acidification of plasma treated organic surfaces.18 The assignments of nitrogen compounds in this study are in accordance with other publications (Table 3).19 ^– 21 Total nitrogen bound on the aluminium surface is ∼3 at.-% (mainly nitrate).

Most metals show a native oxide layer which protects them from further corrosion. In the case of aluminium, this layer is a thin but dense oxide film. Assuming an ideal layer, the thickness of this film can be determined from the relative intensities of the Al⁰ peak and the Al³⁺ peak. The appropriate relationship is described by other publications17,22,23 as followswhere d_mo is the thickness of the oxide layer, θ is the electron take off angle, λ_{mo,Al 2p} and λ_{m,Al 2p} are the inelastic mean free paths of electrons in the metal layer and the oxide layer, D_{mo,Al 2p} with 4·61×10²² atoms cm⁻³ and D_{m,Al 2p} with 6·02×10²² atoms cm⁻³ are the atomic densities23 of atoms in the oxide layer and in the metal, and I_{mo,Al 2p} and I_{m,Al 2p} are the relative intensities of the Al³⁺ peak and the Al⁰ peak. The inelastic mean free paths λ_m and λ_o can be determined using the NIST Standard Reference Database 7·1.10 Taking into account the take-off angle of 45°, λ_{m,Al 2p} = 2·57 nm and λ_{mo,Al 2p} = 2·68 nm, we find oxide layers of ∼3·2 nm before plasma treatment and 3·3 nm after plasma treatment for the aluminium foil. As the thickness of the oxide layer is at least 3 nm, the simplified formula of McCafferty and Wightman23 can be used to calculate the thickness of the hydrocarbon layer by the ratio of the intensities of the O 1s peak generated by the oxide layer to the C 1s peakwhere λ_{mo,O 1s} and λ_{c,C 1s} are the inelastic mean free paths of electrons in the oxide layer and in the hydrocarbon layer respectively. σ_{O 1s} and σ_{C 1s} are the cross-sections of photoionisation of oxygen and carbon, D_{mo,O 1s} is the atomic density of oxygen atoms in the oxide layer, D_{c,C 1s} is the atomic density of carbon atoms in the hydrocarbon layer, I_{O 1s} is the intensity of the O 1s peak (oxide layer) and I_{C 1s} is the intensity of the C 1s peak (hydrocarbon layer). The atomic density of carbon is D_c,C1s = 1·13×10²³ atoms cm⁻³. The atomic density of oxygen atoms in the oxide layer has a value of D_{mo,O 1s}≈6·91×10²² atoms cm⁻³.23 The cross-sections of photoionisation are given by Scofield (for Al K_α: σ_{O 1s} = 2·95, σ_{C 1s} = 1·00).8 The values for inelastic mean free paths can be obtained from the NIST Standard Reference Database 7·110 (λ_{mo,O 1s} = 1·96 nm, λ_{c,C 1s} = 3·67 nm). Applied to this study, assuming an ideal layer system, the carbon layer thickness amounts to d_c = 3·3 nm before plasma treatment. After plasma treatment, there is a residual carbon layer thickness of d_c = ∼1·3 nm left. The carbon layer thickness of a conventional cleaned aluminium foil is d_c = ∼1 nm.

Summarising the information of these results, it is apparent that the hydrocarbon layer of aluminium foils is not only reduced but also partially oxidised by the plasma treatment. What remains are fragments of oxidised hydrocarbon chains, which do not evaporate at the used plasma energy density of ∼5 J cm⁻³. Heating the aluminium coils to clean the foil surfaces takes a lot of time, but a short treatment of the foil surface by an atmospheric pressure plasma in ambient air removes the organic contaminations sufficiently.

Comparing the plasma degreased foil with standard household aluminium foil using test inks, both foils have similar surface tensions. The XPS analysis shows a reduction in the hydrocarbon layer of about 3·3–1·3 nm, which confirms the result of the test inks that the foil has been degreased. Compared to the hydrocarbon layer of ∼1 nm on a customary aluminium foil, the plasma degreased foil has nearly the same carbon layer thickness.