Oxygen-free copper (99.99% purity, C1020) plate, a square shape in size 25 × 25 × 1.0 mm³ (Kikukawa Industry Co., Ltd, Tokyo, Japan), were used as starting material. These Cu plates, after wiping their surfaces with ethanol, were heated in air at 373, 473, 523, 573, 623, 673 K for 4.2 × 10² s; heat treatment was performed in quick heating and cooling.

As described at 2.2.2 in 2. Experimental procedure, chemiluminescence (CL) of Cu plates was measured using the luminol solution. Figure 2 shows dependence of CL intensity of as-received Cu plate on the concentration of luminol solution. Based on the previous experimental data [11] , their concentrations were varied from 5.0 × 10⁻⁴ to 5.0 × 10⁻⁹ mol/L (M). Except for extraordinary high CL intensity curve obtained using luminol’s concentration of 5.0 × 10⁻⁴ M, the CL intensity was decreased gradually with decreasing luminol’s concentration from 5.0 × 10⁻⁵ to 5.0 × 10⁻⁹ M. By considering the suitable intensity and linearity, the aqueous luminol solutions with concentration of 5.0 × 10⁻⁵ or 5.0 × 10⁻⁶ M were utilized in the present study. Then the Cu plates were rapidly heated for 4.2 × 10² s and quenched in air to form copper thin-film oxides.

Figure 3 exhibits the CL curves of these heated Cu plates and the summation of CL intensity SCL (9 min) is displayed; here the luminol’s concentration of 5.0 × 10⁻⁵ M was adopted. SCL (9 min) increased from 111.6 × 10³ count, i.e., 111.6 k·count, to 736.6 k·count with increasing temperature up to 573 K and decreased rapidly to 48.5 k·count at 673 K, suggesting that SCL (9 min) depends much on the amount of copper oxides generated during the heat treatment in air and there might be the most suitable heating temperature. Luminol can emit light in pH value around 11.0 [16] , then in order to eliminate pH effect we searched the light emission reagent which can perform in pH around 7. MPEC [25] is a reagent which can perform light emission in neutral pH, however, its intensity is very weak, so we used MPEC with the higher concentration than Luminol. Figure 4 shows SCL (9 min) of Cu plates prepared under various temperatures, which were determined using luminol (the concentration and the amount of instillation, 5.0 × 10⁻⁵ M and 2.50 × 10⁻⁷ m³, 0.25 mL) and MPEC (5.0 × 10⁻¹ M, 0.25 mL) under pH = 10.8 and 7.5 buffer solutions (4.0 mL), respectively. Both SCL (9 min) exhibited the highest value around 573 K; the best heating temperature is around 573 K. However, it was found that the top temperature sometimes shifted to the lower temperature, for example, 373 K or even room temperature, depending on the surface conditions of Cu plates as-received. Then, the surfaces of Cu plates heated at various temperatures were observed using an SEM under the magnification around 20,000 and compared. Figure 5 shows their images of various Cu plates; as-received (No. 1), and heated at 373 K (No. 2), 473 K (No. 3), 573 K (No. 4), and 673 K (No. 5) for 4.2 × 10² s in air. Among the Cu plates from as-received to 473 K, little change has been recognized, however, 573 K-heated Cu plate showed the granular surface. Furthermore, at 673 K a small amount of needle-like granular were observed. As the highest SCL (9 min) value was attained on the Cu plate heated at 573 K, the outermost surfaces of Cu plates heated at 50 K lower and higher than the top temperature (573 K) were observed using the high-resolution TEM. Figure 6 displays the cross-section images of their outermost surfaces, in the top of each photograph upper left gray portion is bulk Cu, along to the right lower direction white thin film and black mass were observed on the surface of bulk Cu. Here, the magnifications of 1) 523 K and 2) 573 K are the same (see the scale bar of 100 nm), however, the right 3) 623 K is a little lower magnification (see the scale bar of 200 nm) because of the thick oxide film of Cu₂O, which will be described latter. Here we notice that the heating temperature increased from 523 to 573 K, “island-shaped, nub” film was first formed and grew along the base; “a hetero growth thin film (Cu oxide film on metal Cu)” was observed. This growth might be caused by “Volmer-Weber mode” [33] mechanism, that is, in the case that the surface tension of grown substance (Cu₂O) is larger than adhesion force and poor wettability between Cu₂O and Cu.

In the lower images with the higher magnification, the green line indicates the intensity of oxygen content along with a horizontal thin red line. As shown in Table 1, TEM observation gave the thickness of films about 40 to 100 nm for 1) 523 K and 2) 573 K samples, respectively, at 623 K-heated samples “non-uniform” layer. The former thickness values (40 and 100 nm) correspond well to those (40 and 100 nm) of Cu₂O formed on oxygen-free (<5 ppm) Cu heated at 473 K for 25 and 100 min in air, respectively, which is reported by M. Honkanen et al. [12] . Of course in the present study the heating conditions were much different: the temperature (523 K and 573 K) was much higher, and the soaking time (7 min) was very short. M. Honkanen et al. [12] also, described that below 473 K in air Cu₂O was first formed, and then above 473 K, Cu₂O changed into CuO by reacting with O.

(b) Lattice parameters of Cu₂O and CuO

*Thin film XRD was measured on the surface layers from the top to around 200 nm under.

By using the EDX analysis on thin films, the compositional ratios of Cu and O were determined. From the results of thin-film XRD analysis on Cu plates heated, which will be described in next Figure 7, the mass% of Cu₂O and CuO and their crystallite sizes were estimated (Table 1(a)). The content of Cu₂O was increased from 6.0 to 22.6 mass% gradually with increasing heating temperature up to 623 K, however, that of CuO increased suddenly from 0.40 to 6.05 mass% between 573 and 623 K. On the other hand, the crystallite size of Cu₂O reduced in size from 19.9 to 9.2 nm with increasing temperature, which may be reflected by the abrupt growth of CuO between 573 and 623 K. Figure 7 shows the thin-film XRD patterns of Cu plates as-received (300 K) and heated at 473, 573, and 673 K for 4.2 × 10² s in air. XRD peaks of Cu₂O and CuO are not recognized on as-received Cu plate, however, heated at 473 K only a small Cu₂O peak appeared around 2q = 36.6˚, and at 573 K both Cu₂O and CuO peaks are observed. Furthermore, at 673 K many Cu oxides peaks were recognized. Based on these patterns, mass% of Cu₂O and CuO and crystallite sizes were estimated [32] . Table 1(b) shows the lattice parameters of cubic Cu₂O and monoclinic CuO formed on the Cu plate surfaces, with reported PDF data. At-a-glance, the lattice parameter a (042874 to 0.42654 nm) and unit volume V (0.078810 to 0.077605 nm³) of Cu₂O phase are larger than those of the reported PDF data (a = 0.42520 nm and V = 0.07884 nm³) and decreased gradually with increasing temperature, however, their changing degrees are very small. On the other hand, those of CuO phase except b values are almost the same with margin for error. Figure 8 displays the Cu₂O and CuO contents (mass%) near the outmost surface of the Cu plates heated at various temperatures. Cu₂O gradually increased from 373 K and rapidly rose at 573 K, on the other hand, CuO showed little change up to around 573 K, which corresponding to the first Cu₂O formation at lower temperature than CuO on the Cu surface during heating in air. From this, temperature of 573 K might be the stage of sudden alternation, suggesting that copper oxides are active for generation of ROS.

From the preliminary experiments, as we noticed that there must be the most suitable amount of each scavenger to suppress the generation of ROS, so investigation

of the relationship between SCL values and the amount of scavengers was performed. Figure 9 displays the variations of SCL as a function of the amount of various scavengers. In this figure, the amount of luminol is constant (5.0 × 10⁻⁵ M, 0.25 mL) and, for example, the amount of instillation of nbt solution with the same concentration (5.0 × 10⁻⁵ M) as luminol was changed; the bottom point for SCL is achieved around 0.50 mL, suggesting that the most suitable amount of nbt is 0.50 mL. And as the same manner, the most suitable amount of other scavengers were determined; for 2-propanol 0.25 mL, for s-pyr 0.50 mL and for NaN₃ 0.50 mL tentatively. In the case of the last scavenger NaN₃, the 0.50 mL could reduce SCL value than that without scavenger and 1.0 mL (4 times higher mol ratio) addition might be too much.

Figure 10 summarized CL curves of as-received Cu plates added with the most suitable amounts of scavengers, using a constant amount of luminol (5.0 × 10⁻⁵ M, 0.25 mL) and the buffer solution (pH = 10.9, 5.0 mL); the values of SCL are also shown in this figure. Scavengers such as 2-propanol, nbt, s-pyr revealed the CL reduction, i.e., among ROS which is formed on the Cu plate surface, hydroxyl radial ·OH, superoxide anion ·O₂ and hydrogen peroxide H₂O₂ were contained, however, singlet oxygen ¹O₂ was not recognized. From the difference (D) between SCL values measured without and with scavengers, for example, D·SCL (2-propanol) = SCL (without 111.6) – SCL (with 53.1) = 58.5 k·count might be corresponding to the amount of ·OH. These amounts of each ROS formed on the heated Cu plates will be summarized later. Figures 11-14 are the CL curves measured without and with each scavenger, using Cu plates heated at 373, 473, 573 and 673 K, respectively, for 4.2 × 10² s in air. In Figure 13 and Figure 14, the scavenger’s effect for singlet oxygen ¹O₂ disappeared. Here, we have to mention that each scavenger cannot reduce only one ROS, i.e., one scavenger might eliminate more than one ROS at the same time. Then, we calculate main ROS by normalizing. Figure 15 is the summation of results showing the components of ROS formed on the heated Cu plates. At 573 K the highest amount of ROS was obtained and their main ROS were H₂O₂, ⋅ O 2 − and ·OH, however, we should consider the amount of H₂O₂ and ·OH are combined, because the formation of H₂O₂ is strongly related with the amount of ·OH due to the following equation of ·OH + ·OH = H₂O₂. Furthermore, even though some amounts of singlet oxygen ¹O₂ are recognized at both 373 and 473 K, we might ignore its formation.

Finally, the mechanism for generation of hydroxyl radical ·OH, hydrogen per oxide H₂O₂, and superoxide anion on the copper surface are proposed in Figure 16. As already described, a very fine and small amount of thin copper oxide films, Cu₂O and CuO are formed on the outermost surface in air. Cu bulk plate is drawn as a rectangular shape and placed in water or air. In the case of former, water contacts with air and the latter, air contains water vapor. Based on some chemical knowledge, i.e., standard electrode potential E_o(Cu) = 0.34 V, SHE [standard hydrogen electrode]) [34] , copper’ stability in oxygen dissolved acidity or neutral aqueous solution, and a potential-pH diagram of Cu-H₂O at room temperature in air [35] , Cu can be oxidized in the alkaline aqueous solution

with pH = 7 - 14 by dissolved oxygen O₂; that is Cu --> Cu₂O or CuO. However, these products are supposed to be minute quantity due to a short reaction time for CL measurement of 6.0 × 10² s at 293 K. Therefore, in Figure 16 (i) Cu in the bulk copper coelute into water at the surface; Cu --> Cu⁺ + e⁻ (or Cu²⁺ + 2e⁻), (ii) O₂ in air dissolves into H₂O, (iii) and through the reaction (1/4·O₂ + 1/2·H₂O), it (OH) reacts with e⁻ (or 2e⁻) at metal surface, (iv) then ·OH is formed. Furthermore, (v) two ·OH changes into H₂O₂ + 2e⁻. On the other hand, [* Cu⁺ and Cu²⁺ in (i)] reacts with [H₂O₂ in (v)] into (vi) hydroxyl radical ·OH or (vii) superoxide anion. These (vi) and (vii) equations are called “Fenton Reaction” [13] . Here, at the first thin-film layer Cu₂O and the second layer CuO formed on the outermost surface of bulk Cu, as presented in upper side of Figure 16, (2Cu⁺ + 1/2·O₂ + 2e⁻) and (Cu²⁺ +1/2·O₂ + 2e⁻) are produced, respectively. These Cu⁺ and Cu²⁺ react with H₂O₂, as presented by the equation (vi) and (vii); “Fenton Reaction” [13] takes place. It should be noted that these reactions tend to take place at “active step” of the edge place near islands of Cu₂O/CuO, not continuous thin-film.