1 Introduction

Copper, due to its unique combination of properties including high electrical and thermal conductivities, resistance to corrosion and recyclability, is used as wires, cables or tubes in many industrial applications such as heat exchangers, electrical systems, plumbing or short- and long-range communications. In most cases, coatings are previously applied to copper, either metallic to improve solderability or to reduce oxidation or polymer films for electrical insulation [1,2,3]. In such processes, adhesion and wettability properties of copper are thus crucial to ensure durable performances and homogeneity of the coating. Thus, assessing the surface properties in particular roughness and wettability of copper wires could be useful for many applications.

Relatively little work has been devoted to the study of the wetting properties of copper samples [4,5,6,7]. In these studies, copper is mainly processed in a plane geometry and its wettability measured using water and ethylene glycol (EG) as polar solvents or diiodomethane as apolar liquid. In a vertical geometry, Hsu et al. have reported a study of the improvement of water wettability of copper braids through surface modification strategies [8].

However, the characterization of the wetting properties of liquids on materials with cylindrical geometry is not straightforward when their diameter is lower than a few tens or hundreds of micrometers. Thus, wettability measurements on horizontal wires with diameters of a few tens of micrometers have been performed; however, they are not obvious and direct as the cylindrical geometry of the wire impacts the drop shape [9]. Different methods have been used to carry out measurements on vertical fibers. On the one hand, the Wilhelmy-based method is widely used for wettability characterization of fibers. It relies on the measurement of the capillary force exerted by the liquid on the wetted fiber, using a high-precision force tensiometer. In this case, the length of the liquid contact line on the fiber needs to be accurately determined in order to deduce the contact angle. Such measurement can be complicated for thin fibers with non-ideal geometry such as natural fibers, for example. Consequently, static contact angles have mainly been determined in this way, while only a few studies have reported advancing and receding contact angles on fibers [10, 11], in particular on micrometer-scale carbon fibers [12]. On the other hand, measurements of contact angles through optical microscopy analysis allow analyzing directly the meniscus shape. For example, Hansen et al. compared the static water contact angle values determined on PMMA-coated glass fibers with 20–800 µm diameters through optical microscopy imaging of the meniscus profile with those obtained through the sessile drop method on the corresponding PMMA-coated glass substrates [13]. The values obtained on the fibers, between 70° and 73°, were in agreement with the 68° value measured on the plane substrate. Advancing and receding contact angles of water on natural fibers have also been determined through the analysis of the shape of a water drop bridging two parallel fibers [14]. This method requires precision in aligning the two fibers and the control of their distance to form the bridge. Optical microscopy can also be used to measure the height of the meniscus formed by the liquid along the vertical fiber. The contact angle is then deduced from a model which depends on the fiber diameter with respect to the capillary length value [15, 16]. For example, static contact angles of water on carbon fiber tows of 65.8° ± 2.9° have been obtained in this way [12]. It should be emphasized that water is mainly used as liquid to validate the method used to determine contact angle on fibers.

The aim of this work was to determine the surface free-energy of a 700 µm diameter copper wire which is used in many industrial processes, as well as its dispersive and polar components. For this, a contact angle measurement method was first developed and validated using a well-defined PTFE wire by comparing with the contact angles obtained on a PTFE plane surface. Advancing and receding contact angles of a large variety of liquids with a wide range of surface tensions and polarities were then determined on the PTFE wire from the analysis of the meniscus profiles. Once the method approved, the same experiments have been carried out to determine the surface free-energy of a copper wire.

2 Materials and methods

2.1 Materials

Chemical Methyl benzoate (Acros, 99%), dimethyl sulfoxide (DMSO—Merck, > 99%), dimethylformamide (DMF—Merck, SeccoSolv), ethylene glycol (EG—Sigma-Aldrich, 99.8%), toluene (Merck, SeccoSolv) and tri-cresyl phosphate (TCI > 99.0%) were used without purification. Their surface tensions extracted from literature are summarized in Table 1 [17].

Table 1 Surface tension and dispersive and polar components of the model liquids used [17]
Full size table

Materials Polytetrafluoroethylene (PTFE) wire with 600 µm diameter and PTFE plate were purchased from Bohlender GmbH and Bytac™ (Saint-Gobain), respectively. Copper wires with 700 µm diameter were obtained from Nexans. All the samples were cleaned with an anti-scratch paper soaked in dichloromethane and dried in air for 15 min. Contact angle measurements were then performed within 1 h.

2.2 Contact angle measurements and surface-free energy determination

Advancing and receding contact angles were measured under atmospheric conditions at 20 °C using the drop shape analysis-profile device equipped with a tiltable plane (DSA-P, Kruss, Germany). For contact angle measurements on the PTFE plane surface, a 30 µL drop was first deposited on the sample using a variable—volume micropipette. The sample surface sustaining the drop was then tilted at a constant speed (1° s−1), and the images of the drop were simultaneously recorded. The advancing and receding contact angles were measured at the front and rear edge of the drop, respectively, just before the triple line starts moving. The angles were determined using the tangent of the drop profile at the triple line.

For contact angle measurements on a wire, a homemade experimental setup shown in Fig. 1a was developed. The wire was fixed to a manual high-precision mechanical stage enabling its movement in the z-direction.

Fig. 1
figure 1

a Experimental setup, b photograph of the wetting profile on a vertical wire and the inset presents the profile extracted from ImageJ software

Full size image

This setup was placed on the stage of the DSA-P device in order to use the illumination system and camera of the apparatus. Once the wire was immersed in the liquid, images were recorded with the DSA-P camera. The contact angles made by the liquid on the left (θL) and right (θR) sides of the wire were determined as illustrated in Fig. 1b. This method was checked by comparing the contact angles obtained through the extraction of the liquid meniscus and wire edges using ImageJ software (Fig. 1b). To assess the advancing and receding contact angles, the wire was successively dipped in and withdrawn in a cup filled to the roof of the studied liquid. The total vertical movement travel of the wire was close to 6 mm. Images were recorded every 0.3 mm, and the left and right contact angles were measured. Contact angles were then plotted against wire vertical position.

The reported contact angle values correspond to the average of six measurements with an error bar corresponding to the standard deviation.

The surface free-energies of the PTFE and copper wires were determined using the Owens–Wendt approach which is based on the decomposition of the surface tension (or energy) γ into two components due to dispersion forces (\(\gamma_{{}}^{\text{d}}\)): on the one hand, and dipole–dipole interactions and hydrogen bonding (\(\gamma_{{}}^{\text{p}}\)), on the other hand : \(\gamma = \gamma_{{}}^{\text{d}} + \gamma_{{}}^{\text{p}}\) [18]. Using liquids with known surface tension and surface tension components, the surface free-energy of the wires and its components can be deduced from the contact angle value using Eq. 1 [18]:

$$\cos \theta = - 1 + 2 \sqrt {\gamma_{\text{S}}^{\text{d}} } \left( {\frac{{\sqrt {\gamma_{\text{L}}^{\text{d}} } }}{{\gamma_{\text{L}} }}} \right) + 2 \sqrt {\gamma_{\text{S}}^{\text{p}} } \left( {\frac{{\sqrt {\gamma_{\text{L}}^{\text{p}} } }}{{\gamma_{\text{L}} }}} \right)$$
(1)

where the subscripts L and S refer to the liquid and the solid, respectively. While a minimum of two liquids is required for this determination, we used six liquids in order to fit the data with a better accuracy, the slope and intercept of the regression line giving the surface-free energy components.

2.3 Optical microscopy and Atomic force microscopy

Optical microscopy images were obtained with a BX60M microscope from Olympus.

The atomic force microscopy (AFM) experiments were performed in peak force tapping mode with a Dimension Icon microscope from Brüker. Measurements were carried out in air at room temperature, using ScanAsyst-air cantilevers with a spring constant of 0.4 N m−1 (Brüker). 50 µm × 50 µm and 5 µm × 5 µm images were obtained with the height mode and a 256 × 256 dots resolution, at a 0.3 Hz scan rate. Data processing was performed with a NanoScope software version 1.40 from Brüker.

2.4 Scanning Electron Microscopy coupled with Energy-dispersive X-ray spectroscopy (SEM–EDX)

SEM observations were performed using a ZEISS Gemini SEM 300 with a backscattered electron detector (BSD) at a voltage of 15 keV. The elementary analysis (EDX) was made using Silicon Drift Detector technology and a Brüker Quantax Spectrometer.

3 Results and discussion

In order to validate the contact angle measurement method of liquids on wires presented in this article, polytetrafluoroethylene (PTFE) a very common and inert material which can easily be obtained as plates or wires was used. Once validated, the method was then applied to the determination of the surface free-energy of a copper wire.

3.1 Validation of the contact angle measurement method using a PTFE substrate

To compare the values of the advancing and receding contact angles measured on the PTFE plane surface and on the PTFE wire, the sample surface topography was first analyzed. This step is important as these contact angle values can strongly be affected by the surface roughness of samples.

3.1.1 Surface topography

5 µm x 5 µm AFM images of PTFE plane surface and wire are presented in Fig. 2. Both materials display a quite smooth surface at this length scale. The root-mean-square roughness deduced from these images reaches 13 nm for PTFE plane surface and 38 nm for PTFE wire. These close surface states indicate that their advancing and receding contact angles should be comparable.

Fig. 2
figure 2

5µm x 5 µm AFM images of surface topographies of a PTFE wire, and b PTFE plate

Full size image

3.1.2 Wettability

3.1.2.1 Wettability of PTFE wire and plane surface

An illustration of the measurement of the advancing and receding contact angles of DMSO as liquid on a PTFE wire is shown in Fig. 3. For that, a PTFE wire was dipped into DMSO and the contact angle was measured every 0.3 mm of vertical wire displacement. One can notice that the contact angle first increases until a plateau is reached corresponding to the value of the advancing contact angle. Then, in the same way, the wire was withdrawn from the liquid and the contact angle measured every 0.3 mm again. In this way, the contact angle decreases and reaches a plateau associated with the value of the receding contact angle.

Fig. 3
figure 3

Advancing (open circle) and receding (filled circle) contact angles of DMSO on a PTFE wire

Full size image

Such method was applied to several model liquids having different surface tensions. The values of the advancing and receding contact angles are reported in Fig. 4. They are compared to those obtained on the PTFE plane surface using the tiltable plane method.

Fig. 4
figure 4

Advancing (a) and receding contact angles (b) of different model liquids on a PTFE wire (scratch) and a PTFE plane surface (full)

Full size image

Considering the advancing contact angles reported in Fig. 4a for the all set of model liquids, the values measured on the wire and the plate are very close to each other. In addition, the contact angles obtained on the PTFE plane surface are in agreement with those reported in the literature [19,20,21,22]. However, few studies of PTFE wire wetting by known liquids can be found in the literature. For example, advancing and receding contact angles were measured on PTFE fibers in contact with different liquids including ethylene glycol, and very close values to those found in this article have been reported [23]. In the case of the receding contact angles (Fig. 4b), similar values are obtained on the PTFE wire and plate. This is in agreement with the close surface roughness of both PTFE materials as determined by AFM (Fig. 2).These results obtained using a wide range of liquids with various properties allow the validation of the method used for the measurement of contact angles on wires. It should, however, be specified that this method is limited to contact angles lower than 90° and rather 85° in practice for a good precision.

Finally, one can also notice that for the whole liquids, the contact angle hysteresis varies between 1 and 10° approximately. This small hysteresis is in agreement with the smooth surfaces detected by AFM, with a root-mean-square roughness lower than 40 nm on a 5 µm × 5 µm scale (Fig. 2). The contact angles are therefore independent of the substrate geometry in the studied configuration.

3.1.2.2 PTFE surface free-energy and its polar and dispersive components

The PTFE surface free-energy as well as its polar and dispersive components was deduced by considering the all set of contact angle measurements using liquids of different surface tensions and polarities (Experimental section). Several “surface tension components” approaches have been developed in the literature to model the liquid–solid surface tension such as the Fowkes, the Owens–Wendt and the acid–base approaches [18, 24, 25]. Being aware of the inherent imperfections of each of these different models, we chose the Owens–Wendt approach as it is widely used and has been considered in the literature to determine the surface free-energy of PTFE and copper surfaces, thus facilitating comparisons.

The Owens–Wendt method (experimental section) requires the contact angle measurements with a minimum of two liquids, a polar one and an apolar one. However, the large set of used liquids allows a better accuracy in the surface free-energy determination by plotting of \(\frac{{\gamma_{\text{L}} (1 + \cos \left. { \theta } \right)}}{{2\sqrt {\gamma_{\text{L}}^{\text{d}} } }}\) as a function of \(\sqrt {\frac{{\gamma_{\text{L}}^{\text{p}} }}{{\gamma_{\text{L}}^{\text{d}} }}}\) (see Eq. 2 deduced from Eq. 1) [18].

$$\frac{{\gamma_{\text{L}} (1 + \cos \left. { \theta } \right)}}{{2\sqrt {\gamma_{\text{L}}^{\text{d}} } }} = \sqrt {\gamma_{\text{S}}^{\text{p}} } \sqrt {\frac{{\gamma_{\text{L}}^{\text{p}} }}{{\gamma_{\text{L}}^{\text{d}} }}} + \sqrt {\gamma_{\text{S}}^{\text{d}} }$$
(2)

According to Eq. 2, the regression line should be obtained with a slope and an intercept giving access to the polar (\(\gamma_{\text{S}}^{\text{p}}\)) and dispersive components \(\left( {\gamma_{S}^{\text{d}} } \right)\) of the PTFE surface free-energy, respectively.

As the advancing contact angle values are similar on the PTFE plane surface and wire, both values were used for the Owens–Wendt representation as presented in Fig. 5. The uncertainty curves were determined using the least squares method. A good agreement is obtained between the straight line describing the Owens–Wendt approach and the experimental values corresponding to a large range of liquid properties. The dispersive and polar components of PTFE deduced from the fit are \(\gamma_{S}^{\text{d}}\) = 17.3 ± 1.0 mN m−1 and \(\gamma_{\text{S}}^{\text{p}}\) = 1.4 ± 0.6 mN m−1, respectively. The surface free-energy of PTFE is thus \(\gamma_{{{\text{S}} \left( {{\text{PTFE}}\,{\text{wire}}} \right)}}^{{}}\) = 18.7 ± 1.6 mN m−1, in agreement with the values reported, for example, by Lee et al. [20], using the van Oss, Chaudhury and Good method with three model liquids (γS = 15–16 mN m−1) or by Owens et al. [18], using the same method as in this article with two-probe liquids (γS = 14 mN m−1). The studies reported in the literature also highlight the dispersive character of PTFE, as the polar component is in the range 0.1–1.5 mN m−1 [20, 25, 26], which is consistent with the value obtained in this work.

Fig. 5
figure 5

Owens–Wendt representation for PTFE. The dotted hyperbole curves correspond to uncertainties

Full size image

The method for measuring contact angles of liquids on a wire has thus been validated. It can be applied to the determination of the surface free-energy of a copper wire.

3.2 Copper wire

Copper wires with 700 µm-diameter were then studied. Their topography was first analyzed using optical microscopy and AFM. Then, advancing contact angles were measured and the surface free-energy and its components were deduced according to the same method than that used for a PTFE wire.

3.2.1 Topography

The optical microscopy image in Fig. 6a shows the presence of longitudinal stripes due to the manufacturing process. The surface of the copper wire was then imaged by AFM (Fig. 6b–d). The 50 µm × 50 μm image shows a smooth surface of the copper wire with the presence of the longitudinal stripes observed, with Rq = 43 ± 10 nm on a 20 µm × 20 µm scale.

Fig. 6
figure 6

Surface topography of a cleaned copper wire a 870 µm × 660 µm optical microscopy image, b 50 µm × 50 µm AFM image, c image b in 3D representation, d height profile of image b. e 200 µm × 150 µm SEM–EDX oxygen mapping image

Full size image

As shown in Fig. 6d, these stripes have depths ranging from a few nanometers to several tens of nanometers and up even to 100 µm depth. We can also notice that the altitude on the center of the AFM image is higher than the sides due to the curvature of the wire (Fig. 6c).

3.2.2 Wettability and surface-free energy of a copper wire

The copper wire was immersed in five different model liquids. The values of the advancing contact angles are reported in Table 2.

Table 2 Advancing contact angles of model liquids on a copper wire
Full size table

As for the PTFE wire, the Owens–Wendt approach was then applied and is presented in Fig. 7. The surface free-energy of the copper wire γCu is thus estimated to 25.7 mN m−1 with a dispersive component γd of 25.1 ± 0.5 mN m−1 and a polar one γp = 0.6 ± 0.2 mN m−1.

Fig. 7
figure 7

Owens–Wendt curve for the copper wire using some model liquids. The dotted hyperboles curves correspond to uncertainties

Full size image

First by comparing the ethylene glycol (EG) contact angle measured on the copper wire in this work with those found in the literature, similarities and differences can be highlighted. They can probably be explained by different oxidation states of the copper surfaces. For example, on the one hand, when copper plates were treated with HCl, known for stripping the oxidized upper layers, the EG contact angle was 54° ± 1°, much lower than that found here (72.9° ± 3.7°) with the copper wire without any acid treatment [4]. On the other hand, when copper substrates were prior oxidized under an oxygen flow, the EG contact angle was about 76° ± 2°, very close to our measurement [5]. In addition, the authors report a much higher contact angle when using a high polar solvent such as water (~ 100°), which reflects the hydrophobic character of the oxidized copper surface [5]. This value is in agreement with other water wettability literature studies [5, 7, 8], and a water contact angle exceeding 90° evidenced in this work. Therefore, copper surface free-energy changes depending on its oxidation state. In fact, a surface free-energy of 39.2 mN m−1 was determined for a non-oxidized copper surface (treated with HCl), using the Van Oss–Good method with four model liquids [4]. This value is consistent with other similar studies using either the Van Oss–Good approach (γCu = 34 mN m−1) [6], or the Owens–Wendt approach (γCu = 34 mN m−1) [7]. On the other hand, when the copper surface was prior oxidized, the copper surface free-energy was much lower, in the range of 22–27 mN m−1 depending on the used approach [4]. These results are very close to the surface free-energy obtained in this article (γCu = 25.7 mN m−1), which would suggest that the copper wire would have a rather oxidized surface [4]. This result is in good agreement with the oxygen mapping by SEM–EDX (Fig. 6e) which also shows the good homogeneity of oxygen on the copper wire surface after its cleaning.

Moreover, and whatever the approach used, the authors find a dispersive component that is always much higher (γd ≈ 25–37 mN m−1) than the polar one (γp ≈ 1.5–6 mN m−1) in agreement with the components values calculated for the copper wire (γd ≈ 25.1 ± 0.5 mN m−1 and γp ≈ 0.6 ± 0.2 mN m−1) [5,6,7].

4 Conclusions

In this paper, a contact angle measurement method on vertical wires was developed by comparing the wettability of a PTFE wire and a PTFE plane substrate. The wire was either immersed or withdrawn in the liquid giving the advancing and receding contact angles, respectively. The measured contact angles were similar for both geometries allowing the validation of the method. Then, the Owens–Wendt method allowed the determination of the PTFE wire surface free-energy. The value obtained, γPTFE = 18.7 ± 1.6 mN m−1 including a dispersive component γd ≈ 17.3 ± 1.0 mN m−1 and a polar component γp ≈ 1.4 ± 0.6 mN m−1, is similar to the literature values.

Then, the same approach was applied to a 700 µm-diameter copper wire, with a large range of model liquids having various surface tensions and polarities, compared to the literature in which methods with two or three liquids are used. We thus determined a surface free-energy γCu = 25.7 mN m−1 with a dispersive component γd ≈ 25.1 ± 0.5 mN m−1 and a polar component γp ≈ 0.6 ± 0.2 mN m−1. By comparison with the literature values of both EG contact angles and surface free-energies, the surface of the studied copper wire appears to be rather oxidized.