Starting materials were three kinds of Cu thin-films (JX Advanced Metals Corporation), name of commodity, HA, HA-V2 and TPC, with the style of roll packing (0.8 m width and 5 m long). Their futures supplied form “JX” were: 1) a thickness of 18 µm, 2) the grains on the flat surface would be highly oriented, i.e. a crystal plane of (100) would be dominant after the heat treatment of 473 K for 1.8 × 10³ s (30 min) in an inert atmosphere. Their (100) plane orientation degree Q (100) is the highest for HA, and decreased a little for HA-V2, and TPC. In the present study, a square shape in size 25 × 25 mm² was cut from the roll and used. These cut thin-films, after wiping their surfaces with ethanol, were heated under various atmosphere, such as N₂ (purity 99.9%) gas flow (100 mL/min) in the cylindrical electric furnace (the inside diameter of 60 mm) up to 473 K for 1.8 × 10³ s, under conventional vacuum using the combined rotary and diffusion pumps (a degree of vacuum: 1 - 3 × 10⁻³ Pa) with a rotary & diffusion pumps, and in air at 373 to 523 K for 4.2 × 10² - 1.8 × 10³ s. The heat treatment in air was performed in quick heating and cooling.

Powder X-ray diffraction (XRD: Smartlab, Rigaku, Tokyo, Japan) analysis with CuKα radiation (a wavelength of 0.15418 nm) was utilized for identification of the crystalline phases and evaluation of both crystallite sizes and grain orientation Q(HKL) for crystal plane (HKL) using the following equation; Q(HKL) = [ΣI(HKL)/ΣI(HKL)] × 100(%). Here, ΣI(HKL) means the summation of (HKL) plane’s diffraction intensity I(HKL), and ΣI(HKL) the summation of all (HKL) plane’s diffraction intensities I(HKL). However, as the crystal structure of metal Cu is face-centered cubic (FCC) and due to the extinction law, ΣI(HKL) = I(200) and ΣI(HKL) = I(111) + I(200) + I(220) + I(311). The measuring conditions for powder XRD were as follows: accelerating electric voltage and current were 45 kV and 200 mA, respectively, scanning speed 1.667 × 10⁻¹˚/s (10˚/min), scanning angle 2θ: 40˚ - 100˚.

Microstructural observation with a field emission-type scanning electron microscope (FE-SEM; SU8020, Hitachi High-Technologies Corporation, Tokyo, Japan) and a transmission electron microscope (TEM, JEM-2100F, JEOL, Tokyo, Japan) equipped with the energy-dispersive X-ray spectroscopy (EDS, JED-2300T, JEOL) were performed. Before TEM observation, the Cu samples were cut into a small specimen using a focused ion beam (FIB, FB2200, Hitachi High-Technologies Corporation).

Chemiluminescence (CL) of Cu thin-films [sample size of 2.5 × 10⁻³ m (25 mm) square] in a 2.5 × 10⁻⁷ m³ (0.25 mL) aqueous luminol (Reagent special grade: purity > 95.0%) solution with a concentration of 5.0 × 10⁻¹ - 5.0 × 10⁻² mol·m⁻³ (5.0 × 10⁻⁴ - 5.0 × 10⁻⁵ mol·L⁻¹) mixed with 5.0 × 10⁻⁶ m³ (5.0 mL) carbonic acid buffer solution (NaOH/NaHCO₃, pH = 10.7 - 10.9) [16] was observed under dark conditions K using a CL detector (CLA-FS3, Tohoku Electronic Industrial Co., Ltd., Sendai, Japan), during measurement temperature was kept at 293 by cooling/warming system to keep a constant temperature. After dropping the luminol solution in a 6.0 × 10¹ s’ warming up of the detector, the intensity of CL was integrated between 6.1 × 10¹ and 6.0 × 10² s (9 min) [14] . The obtained summation of CL intensity for 5.4 × 10² s (9 min), ΣCL (9 min), which was calculated as follows: at first, estimate the summation of background noise, ΣCL (BG), from 0 to 6.0 × 10¹ s, and then summarize the CL intensity from 6.1 × 10¹ to 6.0 × 10² s, ΣCL (total 9 min), finally, true ΣCL (9 min) was determined to be as the true ΣCL (9 min) = ΣCL (total 9 min) – ΣCL (BG) × 9. This true ΣCL (9 min), in a simple term, ΣCL (9 min) value was utilized to evaluate the amount of ROS generated from the surface of Cu thin-films.

Among 4 kinds of ROS, such as ·OH, H₂O₂, ·O₂⁻· and ¹O₂, scavengers [17] [18] , 2-propanol [19] [20] (“2-pro”, for ·OH, Nacalai Tesque Chemicals, Kyoto, Japan) [15] - [17] , sodium pyruvate [19] (“s-pyr”, for H₂O₂, Fuji-film Wako Pure Chemical Co., Ltd., Osaka, Japan), nitro blue tetrazolium [21] (“nbt”, for ·O²⁻·, Nacalai Tesque Chemicals) and sodium azide [18] (for short “NaN₃”, for ¹O₂, Fuji-film Wako Pure Chemical) were mainly used to determine which ROS was generated. At first, these scavengers were solved into the buffer solution with the concentration 5.0 × 10⁻² mol·m⁻³ (5.0 × 10⁻⁵ mol·L⁻¹) and then obtained 5.0 × 10⁻⁷ m⁻³ (0.50 mL) or 1.0 × 10⁻⁶ m⁻³ (1.0 mL) scavenger solution were used. Each solution and a 4.0 × 10⁻⁶ m⁻³ (4.0 mL) pure buffer solution were mixed. CL measurement was performed as the same as CL intensity measurement, after a 6.0 × 10¹ s’ warming up of the detector, the intensity of CL was integrated between 6.0 × 10¹ - 6.0 × 10² s. The amount of difference (Δ) for each ROS was evaluated using the equation of

Δ = ΣCL (9 min, without scavenger) – ΣCL (9 min, with scavenger) (1)

As will be stated, two kinds of Cu thin films, i.e. 1) as-obtained HA thin film and 2) antibacterial activity improved HA film which was heated under vacuum (2.0 × 10⁻³ Pa) at 473 K for 1.8 × 10³ s (30 min), followed by re-oxidation at 398 K for 4.2 × 10² s (30 min) in air were provided to biotest using Escherichia coli NBRC3301, Staphylococcus aureus NBRC 12732, and A/Hong Kong/8/68; TC adapted virus ATCC VR-1689 based on JIS (Japanese Industrial Standards) Z 2801 (Film adhesion method) by QTEC Technology Co., Ltd., Kobe City, Japan. These bacteria and viruses were adopted as follows: Escherichia coli was selected as a representative Gram-negative bacterium, Staphylococcus aureus as a representative Gram-positive bacterium, and A/Hong Kong/8/68; TC as a representative virus. Test conditions were as follows: The sample size was a square with 40 × 40 mm², E. coli 4.5 × 10⁵ CFU/mL, Staphylococcus aureus 4.0 × 10⁵ CFU/mL and A/Hong Kong virus 2.2 × 10⁷ PFU/mL, and a PE film was used as a control. Biotest for two bacterial species and one virus was performed for 8.64 × 10⁴ s (24 h) under dark condition. Furthermore, a short test of both as-obtained HA and antibacterial activity improved HA films using E. coli NBRC 3972 at 300 K for 1.8 × 10³ s (30 min) in shade. Antibacterial activity value R was calculated by the following equation:

R = (Log [control 30]-Log [control 0]) – (Log [sample 30]-Log [sample 0])

As described in experimental procedures 2.1 and 2.2.1, the as-obtained Cu thin-films, HA, HA-V2, and TPC, were heated in a N₂ gas flow. Figure 1 shows powder XRD patterns of Cu thin-films (A) the as-obtained and (B) after the vacuum heating, respectively. In (A), all diffraction peaks such as (111), (200), (220), and (311) were confirmed, however, in (B), only (200) peak was recognized, indicating almost all grains were oriented to (200) plane, that is (100)-orientation (due to the extinction rule for FCC, (100) diffraction peak cannot be observed). Then, all thin-films were heated at 373, 423, 473 and 523 K in N₂ and by their powder XRD measurement, Q(100) was evaluated. Figure 2 shows their Q(100) values as a function of heating temperatures, with the Q(100) values of as-obtained conventional Cu plates (25 × 25 × 1.0 × 10⁻⁹ m³) [1] and 473 K-heated as a reference. Q(100) of HA and HA-V2 started from 10.0/6.5% at 300 K (as-obtained) and increased rapidly to 99.8/96.2% at 423 K, finally saturated to 100/98.8% at 473 K, respectively. However, TPC increased Q(100) from 43.7% to 98.7% at 423 K, on the contrary, conventional Cu plate revealed 85.3% even at 473 K. Afterward, both HA and TPC films were focused on evaluation. To investigate the microstructural difference in crystallite size X_s at each crystal plane, which was determined by Scherrer’s equation [22] (X_s = K·λ/(β·cosθ, here K: constant 1.0, λ: wave length 0.15418 nm, β: full width at half maximum, θ: diffraction angle) were estimated. Figures 3-5 reveal X_s of as-obtained films, heated HA and TPC under various conditions, respectively. From Figure 3 , it is worth noting that only X_s of (220) for as-obtained HA is very small at 16.3 nm, under that almost all X_s are 35 - 50 nm except for X_s of (311) are 30 nm. It should be noted that after various heat treatments (the precise conditions will be described later), HA showed a very large X_s of (200) about 160 nm, suggesting high Q(100) orientation ( Figure 4 ). On the other hand, as shown in Figure 5 , all X_s of TPC were the same from 100 to 120 nm except for (222) plane after heating. This behavior is thought to be common phenomena observed for conventional metals.

Next, the chemiluminescence (CL) intensities of thin-films, which were heat-treated under various conditions, were measured for 9 min and summated. Figure 6 summarized the results as simple overview, that is ΣCL (9 min) as a function of heating temperature; some intercalary vertical comments indicate the heating condition. Starting from the left hand, as-obtained thin-films revealing the ΣCL (9 min) values around 200 × 1000 counts (200 k·count), “473 K for 30 min in N₂” heated samples gave a small amount reduction, however, on the other hand, when a heating method was changed to “vacuum” [“473 K for 30 min in a vacuum”] ΣCL (9 min) increased to 512.6 (TPC), 701.8 (HA-V2), and 907.5 (HA) k·count. Between thin-films prepared in “N₂” and “vacuum” heating, crystallite sizes X_s and orientation degree Q(100) were the same. Therefore, much difference in ΣCL (9 min) values might come from the difference of outermost surface physicochemical condition; “vacuum” heating might bring the clean surface, on the other hand, “N₂” heating might not. Then, based on the idea that CL intensity, i.e. amount of reactive oxygen species ROS is generated on the Cu oxidative very thin films such as Cu₂O and CuO, two kinds of thin-film were re-oxidated at 373 K for 7 min in air after “vacuum” heating, this process could be considered to clean up the outermost surface once and reconstruct Cu-oxide thin-films. HA film showed the highest ΣCL (9 min) around 1600, followed by 775 (HA-V2) and 540 (TPC) k·count as shown in Figure 6 . However, further re-oxidation at the higher temperatures resulted in the decrease of ΣCL (9 min). Therefore, the combination of the vacuum (473 K/30 min) and re-oxidation (373 K/7 min/air) was thought to be the optimum process to give the maximum ΣCL (9 min). Other HA-V2 and TPC thin-films also showed the best values in this combination process among them.

Figure 7 shows various grain orientation degrees, such as Q(110), Q(100) and Q(111), with the summation of CL intensity, Σ (CL) (9 min) of HA thin-film prepared under DP vacuum (2.0 × 10⁻³ Pa) at various temperatures for 1.8 × 10³ s (30 min). The CL measurement conditions were the same as those mentioned before, such as buffer’s pH: 10.7, 5.0 mL and 5.0 × 10⁻⁵ M luminol solution, 0.25 mL.

It is safe to say that Σ (CL) (9 min) value depends much on Q(100). From 325 K to 400 K, Q(100) increased a little from 10.2 to 24.2% and on the contrary, Q(110) decreased a little from 82.8 to 70.0%. If we could presume that Q(100) and Q(110) have positive and negative effects on Σ (CL) (9 min), respectively, a little increment/decrement could bring a gradual increase of Σ (CL) (9 min). Furthermore, at temperatures around 400 to 425 K, with increasing Q(100), at the same time with decreasing Q(110), Σ (CL) (9 min) increased gradually, being independent on Q(111). At higher than about 425 K, nevertheless, the constant value of Q(100), Σ (CL) (9 min) increased rapidly. This might be explained by the grain growth of (100)-plane at these temperatures. As far as counter part of TPC thin-film is concerned, as shown in Figure 8 , the same trend as HA with increasing Q(100) and decreasing Q(110) could be recognized. Therefore, it might be probably true that Σ (CL) (9 min) value depends much on Q(100).

This phenomenon could be explained as follows:

1) It has been reported that the structure of Cu oxide thin-film is a stacked layer from bottom to top Cu/Cu₂O/CuO [6] [23] [24] .

2) The growing rate of Cu oxide thin-film on Cu(100) plane is much higher than those of Cu(111) and Cu(110) [25] . Here, it should be noted that these results were pointed by Iijima [25] using Cu single crystal.

3) The high growing rate of Cu oxide films on Cu(100) might introduce easily some defects in thin-film, such as facets, steps, and/or kinks.

4) As shown in Figure 9 , ROS may be produced at these defects at the Cu₂O/CuO thin-film layer formed on the outermost surface of Cu, because these defects are very active physically [26] . Generation of reactive oxygen species (ROS) around above mentioned “steps”, and/or “kink” in the Cu₂O/CuO thin-film layer was explained using chemical reaction equations shown in Figure 9 .

As the amount of ROS reflects the Σ (CL) values, therefore, Cu(100) plane can give high Σ (CL). As mentioned, the high growing rate of Cu oxide thin-film, here, we consider Cu oxide thin-film could be Cu₂O, on Cu(100) could be explained qualitatively.

Figure 10 shows the crystal structure of (A) face-centered-cubic (FCC) Cu [27] and (B) simple cubic Pn3m Cu₂O (cuprite) [28] . Oxygen ions in Cu₂O occupy the eight corners of cubic cell and inside it, and four Cu⁺ ions form a regular tetrahedron, i.e. Cu⁺: (1/4, 1/4, 1/4), (1/4, 3/4, 3/4), (3/4, 1/4, 3/4), (3/4, 3/4, 1/4) [29] : O₂⁻: (0, 0, 0), (1/2, 1/2, 1/2). It should be noted that Cu₂O unit cell contains a regular tetrahedron of Cu₄O.

If the outermost surface of Cu is oxidized, it can be questioned which crystal plane is suitable for the first formation of Cu₂O. Figure 10 also shows a schematic configuration of regular tetrahedron of Cu₄O positioned on (i) (100), (ii) (110), and (iii) (111) planes of Cu; large brown circles in each plane are Cu atoms (atomic radius = 0.128 nm) [30] , and small brown and large broken-line circles of regular tetrahedron Cu₄O present Cu⁺ (ionic radius with the coordination number 2, 0.046 nm) [31] and O₂⁻ (ionic radius with the coordination number 4, 0.138 nm) [31] ions, respectively. In addition, the diagonal length of regular tetrahedron I_Cu-Cu(Cu2O) is 2^−1/2·a_(Cu2O) = 0.30066 nm and the side length of square (100) is a_(Cu) = 0.35819 nm. Here, the lattice parameters a_(Cu) = 0.35819 nm (PDF#4105040) and a_(Cu2O) = 0.42520 nm (PDF#1000063) were used for calculation. Here, it should be noticed that the density of Cu-Cu pair (“red star symbol”) with the shortest distance per unit area, Den(Cu-Cu); (i) Den(Cu-Cu) on Cu(100) plane: 4/a_(Cu)², (ii) Den(Cu-Cu) on Cu(110) plane: 2^1/2/a_(Cu)², (iii) Den(Cu-Cu) on Cu(111) plane: 3^1/2/a_(Cu)². Therefore, Den(Cu-Cu) for C(100) is the highest and followed by those of Cu(111) and Cu(110) planes. This order agrees well with the growth rate of Cu oxide thin-films on Cu planes. Furthermore, Cu atom in the center of Cu(100) plane might have some influence on making Cu-O bonding at first. Based on these configurations and relaxed structure metal surface of Cu, we presumed that Cu(100) plane might provide better stable situation for the formation of thin-film of Cu₂O readily than others.

Then, the microstructure of most upper (outermost) surface of HA was observed using SEM and TEM. Figure 11 shows the micrographs of thin films, No. 1 is as-obtained HA with the Σ (CL) (9 min) = 111.6 k·count, No. 2 after heating at 473 K for 1.8·10³ s (30 min) in N₂ the with Σ (CL) (9 min) = 396.3 k·count, No. 3 after vacuum heating (DP) at 473 K for 30 min with the Σ (CL) (9 min) = 907.5 k·count, and No. 4 after vacuum heating (DP) and re-oxidation at 373 K for 7 min in air with the Σ (CL) (9 min) = 1597.9 k·count. At a glance, the surface of No. 2 is different from others, i.e. consisting of granularity, and others are cloudlike textures. Then, the most upper surfaces of No. 2 and No. 4 samples were observed using TEM. Figure 12 shows the cross-section of outermost surface of No. 2, indicating the line-analysis position and intensities of Cu/Pt, respectively. By comparing (A) and (B), it is clear that black and grey regions on (A) are Cu and Cu₂O_1−x, and black bulk in right side is Pt passivating film. And it is characteristic that a few accentuated cylinders consisting of Cu₂O_1−x covered with a hat-like Pt are formed on the Cu base. In Figure 12(C) , atomic intensities are presented as bar graph. Figure 13 shows TEM images and elemental analysis for the outermost surface of No. 4 heated at 473 K for 1.8 × 10³ s (30 min) in a vacuum, followed by heating 373 K for 7 min in air (re-oxidation); (A) is the cross-section image with a horizontal line indicating point-analysis position, (B) and (C) are elemental-analysis bar graph and atomic density on each position, respectively. It should be noted that in Figure 13 , no accentuated cylinders are recognized and only two kinds of thin layers are observed. From the position number on the blue line in (A) and each point analysis data shown as bar graph in (B), numbers from “1” to “3” represent mainly Cu base, and “4” and “5” suggest Cu₂O_1−x and CuO, respectively. This change in composition corresponds to the color change of image shown in (A). Here, we should pay attention that the CL summation ΣCL of the latter of No. 4 (vacuum and re-oxidized) is much higher than that of former sample No. 2 (N₂ heating) in Figure 11 . That is, the outermost surface, which consists of graded double thin layers of Cu₂O_1−x and CuO could produce more ROS than that of the accentuated cylinders consisting of Cu₂O_1−x.

Then, it was investigated which ROS was produced on the outermost surface using corresponding scavengers during CL measurement, as shown in Table 1 . In order to determine the most optimum additive amount of each scavenger, ΣCL for Cu thin film HA, which was heated under vacuum and then re-oxidized at 373 K/7min in air was measured. Figure 14 presents their results; as the most optimum additive amount could give the minimum ΣCL, the optimum additive amounts for scavenger of 2-propanol (“2-pro”) for ·OH, sodium pyruvate (“s-pyr”) for H₂O_2, nitro blue tetrazolium (“nbt”) for ·O₂⁻ were 5.0 × 10⁻⁵ M with 0.5 mL. However, the optimum amount for sodium azide (“NaN₃”), for ¹O₂ was 5.0 × 10⁻⁵ M with 1.0 mL under the conditions of 5.0 × 10⁻⁵ M luminol 0.25 mL.

Figure 15 shows the component of ROS generated on the surface of Cu plate prepared under various conditions. It is interesting that “DP” (vacuum heating at 473 K for 1.8 × 10³ s (30 min)) treatment brings ever-increasing ROS, such as, ·OH < H₂O₂ < O₂⁻ < ¹O₂. However, once after oxidizing, ·OH and H₂O₂ were recovered. On the contrary, as shown in Figure 16 , it is extreme that H₂O₂ and O₂⁻ have disappeared from HA thin-film after DP. As the same as Cu plate, once after oxidizing, all ROS have been generated again. These phenomena can suggest that: 1) DP treatment can bring passive oxidation [32] to both Cu outermost surfaces, and 2) its effect is stronger on HA thin-film than conventional Cu plate, it might be due to the hard/severe processing condition to fabricate thinner films from the thicker sheet.

For further study, the optimum vacuum heating and re-oxidation conditions were investigated. Figure 17 shows ΣCL of the Cu thin-film HA after vacuum heating (at 473 K for 1.8 × 10³ s (30 min)), followed by re-oxidation for 0 - 60 min at 373 and 398 K in air. ΣCL values plotted with ultramarine blue circles are those of HA re-oxidized at 373 K and other shapes such as triangle (back), square (top) and red circles (top) present are those of HA re-oxidized at 398 K for 1.8 × 10³ s (30 min). Here, “top” or “bottom” means that HA films were stacked in top and bottom in layers during re-oxidation. The latter three kinds of shape indicate the dispersion of measured values. From this graph, it might be concluded that the best ΣCL was achieved for HA thin-film heated under vacuum and re-oxidized at 398 K for 1.8 × 10³ s (30 min) in air.

As stated previously, antibacterial activity value R (a-a-R) of: 1) as-obtained and 2) antibacterial activity improved HA films were evaluated. Figure 18 shows their (a-a-R) vs. lapse time: evaluated for 8.64 × 10⁴ s (24 h) and 1.8 × 10³ s (30 min) at 393 K using Escherichia coli (E. coli), Staphylococcus aureus (S. aureus) and A/Hong Kong virus. After long lapse time 8.64 × 10⁴ s (24 h), there is little difference between improved HA and as-obtained thin films in (a-a-R) against E. coli, S. aureus and A/Hong Kong virus. It is clear that these bacteria and viruses are strong in this order. However, at shorter lapse time of 1.8 × 10³ s (30 min), it should be noted that: 1) (a-a-R) of the former (3.6) is a little higher than that of the latter (2.6) against for E. coli, and 2) scattering of (a-a-R), i.e. the length of vertical arow, of the former is small in comparison with that of the latter against for both E. coli and S. aureus. This means that at long lapse time of 8.64 × 10⁴ s (24 h) required lapse time by JIS (Japan Industrial Standard), the former HA showed the almost same antibacterial activity as the commercially available HA; little improvement was recognized. However, at shorter lapse time, the former overcomes the latter in a viewpoint of (a-a-R), and a small scattering (that is antibacterial activity level is stable).