The effect of surfactants on the electropolishing behavior of copper in orthophosphoric acid

Applied Surface Science 277 (2013) 155–166

Contents lists available at SciVerse ScienceDirect

Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc

The effect of surfactants on the electropolishing behavior of copper in orthophosphoric acid A.A. Taha ∗ , A.M. Ahmed, H.H. Abdel Rahman, F.M. Abouzeid Chemistry Department, Faculty of Science, Alexandria University, Alexandria, Egypt

a r t i c l e

i n f o

Article history: Received 31 January 2013 Received in revised form 29 March 2013 Accepted 2 April 2013 Available online 13 April 2013 Keywords: Electropolishing Surfactants Scanning electron microscope (SEM) Atomic force microscope (AFM) Brightness

a b s t r a c t The electropolishing behavior of copper was studied in orthophosphoric acid with Triton X-100, sodium dodecyl sulphate and cetyl pyridinium chloride as additives for improving the finish obtained on copper surface. This was investigated by measuring and comparing anode potential-limiting current relationships in solutions of gradually increasing concentration of surfactants. The addition of surfactants to the electropolishing solution results in a lower limiting current. This confirms the mass transport of dissolved species from the anode surface to the bulk of solution as the rate-determining step in the presence of three surfactants in all concentrations investigated. Scanning electron microscope (SEM), atomic force microscope (AFM) and measured brightness values were used to investigate the copper surface after electropolishing and the results were compared to polishing done in absence of surfactants. According to SEM images and brightness values, addition of Triton X-100 was effective to enhance levelling and brightening more than sodium dodecyl sulphate and cetyl pyridinium chloride. AFM analysis showed that the roughness values (Ra ) for an electropolished copper surface, in presence of surfactants, is significantly lower than in absence of surfactants. Different reaction conditions and the physical properties of solutions are studied to obtain dimensionless correlation among all these parameters. © 2013 Elsevier B.V. All rights reserved.

1. Introduction The electropolishing (EP) has been used for various purposes in research and industry for many years. The appropriate formulation of the electropolishing solutions, EP conditions and the process application itself were and are still the subject of numerous trade secrets and patents [1,2]. The practical aspects of EP have been reviewed by Mohan et al. [3], whereas the most fundamental aspects are covered in a review by Landolt [4]. The surface phenomena of EP are generally classified into two processes: anodic levelling and anodic brightening. Anodic levelling results from a difference in the dissolution rate between peaks and valleys on a rough metal/alloy surface depending on the current distribution or mass-transport conditions [5]. It is usually associated with a decrease of roughness in the micrometer or large range [4]. Anodic brightening, on the other hand, can be achieved only under the conditions in which the metal dissolution is masstransport-controlled and the formation of a precipitated salt layer at the electrode surface is possible. The presence of salt layer can suppress the influence of crystallographic orientation and surface defects on the dissolution process. This phenomenon will lead to

∗ Corresponding author. Tel.: +20 35452486; fax: +20 35452486. E-mail addresses: asia [email protected], asia [email protected] (A.A. Taha). 0169-4332/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.apsusc.2013.04.017

microfinishing in the submicrometer scale and specular reflectivity of metals/alloys can be obtained [6–9]. Copper surface seems rough and complex under microscopic investigation, due to the presence of macro and micro scales of humps and depressions. It has received much attention due to its practical and academic interests, some workers tried to attain bright and smooth surface of copper, aluminium, stainless steel and other alloys by applying EP methods [10–12]. Edson [13] studied the brightening of copper, silver and gold by using an electrolytic solution composed mainly of thiourea and some additives such as polysaccharides as reducing materials and mineral acids as conductivity improvers. Others studied EP of copper using additives such as soluble starch, ethylene glycol and methanol to achieve smooth and highly lustrous copper surface [14]. Furthermore addition of ethylene glycol and glycerol [15], tert-butanol [16], amines [17], amino acids [18] and aldehydes [19] improved planarization efficiency of copper surface, where the up and down thickness differences are eliminated. In this study, an attempt is extended to reduce the roughness of the surface by reducing the etched pits and defects formed over the surface, consequently, the surface smoothness and brightness could be increased. Accordingly, the present work is aimed to study the effect of addition of surface active agents (SAS) to orthophosphoric acid solutions used as electrolytes for EP of copper.

156

A.A. Taha et al. / Applied Surface Science 277 (2013) 155–166

Table 1 Chemical composition of copper. Element

Cu

Cd

Ag

P

Ni

Fe

Zn

Pb

Sn

wt%

99.98

0.001

0.001

Negative

Negative

Negative

Negative

0.003

0.005

The SAS chosen for examination were nonionic surfactant namely Triton X-100, anionic surfactant namely sodium dodecyl sulphate (SDS) and cationic surfactant namely cetyl pyridinium chloride (CPC) which were available in purified form. The SAS chosen offer variation in the chain length of the hydrocarbon portion, and the type of the polar portion. SAS have many advantages such as high inhibition through adsorption on the metal surface, low price, low toxicity and easy production. It is well known that SAS are characterized by critical micelle concentration (so called CMCs). The CMC is the concentration, where SAS in solution change their initial molecular solvated state. Most of the physical and chemical properties of SAS solutions undergo an abrupt variation at this concentration. This effect is

of luggin tube filled with orthophosphoric acid–organic solution similar to that in the cell. The tip of the luggin tube was placed 0.5–1 mm from anode wall. Polarization curves, from which the limiting current was determined, were plotted by increasing the applied current stepwise and measuring the corresponding steadystate potential. Two minutes were allowed for reaching the steady state potential. Before each run, the back of anode was insulated with polystyrene lacquer and the active surface was polished with fine emery paper, degreased with trichloroethylene, washed with alcohol and finally rinsed in distilled water with a measured resistivity >18 M/cm. The temperature was regulated by placing the cell in thermostatic water bath at 25 ± 0.5 ◦ C. The copper sheets used in the present work have the composition shown in Table 1. The structure of SAS is given below:

of interest for theoretical reasons as well as for practical application. In the present study, potentiodynamic polarization measurements were performed to examine the EP process of pure copper in H3 PO4 acid in the presence of these SAS, for the range of concentrations below, at and above CMC. The extent of polishing was assessed through an elaborate study relating the influence of SAS concentrations (and type) and the resulting surface morphology (investigated by SEM), surface topography (investigated by AFM) and surface brightness. Physical properties of the solutions such as density, viscosity and diffusion coefficient were studied to obtain a dimensionless correlation among all parameters.

2.2. Surface tension measurements

2. Materials and methods 2.1. Potentiodynamic measurements The cell used in the present work consists of rectangular container having the dimensions of 5 cm × 10 cm with electrodes fitting the whole cross section. The electrodes were rectangular copper sheets of 10 cm height and 5 cm width. Cathode–anode distance was 5 cm. The electrical circuit consisted of a 6 V D.C. power supply, a variable resistance and a multi range ammeter connected in a series with cell. A high impedance voltammeter was connected in parallel with the cell to measure its potential. Five concentrations of orthophosphoric acid (6, 8, 10, 12 and 14 M) were prepared from analar grade H3 PO4 (85%). Ten concentrations of surfactant solutions (Triton X-100, sodium dodecyl sulphate and cetyl pyridinium chloride) with 8 M H3 PO4 are used ranging from 5 × 10−7 M to 1 × 10−2 M. The steady state anode potential was measured against reference electrode consisted of copper wire immersed in a cup

The surface tension was measured at 25 ◦ C using Du Nouy tensiometer (Kruss type 8451) for various concentrations of the surfactants additives. The temperature (±0.1 ◦ C) was kept constant by circulating the thermostatted water through a jacketed vessel containing the solution. 2.3. Density and viscosity measurements The density () and viscosity () of different solutions were used to calculate the dimensionless groups Sherwood (Sh), Schmidt (Sc) and Grashof number (Gr) (Table 2). The density was measured by using DA-300 Kyoto electronic instrument. The viscosity was measured by using Koehler viscosity bath (model K23400). The accuracy of the instrument is 2%. 2.4. Diffusion coefficients measurements The diffusion coefficient of blank solution and of solutions in the presence of different concentrations of SAS used in this study were calculated by measuring the limiting current of the anodic dissolution of a copper rotating disc in H3 PO4 and applying the Levich equation [15]: 2/3

iL = 0.62nFDeff  −1/6 (Csat,Cu − Cb,Cu )˝1/2

(1)

where iL = IL /A. iL is the limiting current density (A cm−2 ), A is the cross-sectional area of Cu disc, n is the number of electrons involved in the reaction, F is Faraday constant, Deff is the effective diffusion coefficient of dissolving species,  is the kinematic viscosity, ˝ is

6.67E+08 4.29E+08 1.66E+08 1.71E+08 4.52E+08 163,630.65 44,619.96 13,384.16 3959.51 4503.82 1520.38 1414.08 1024.02 2366.33 2000.00

Sc Gr Sh

Gr

A.A. Taha et al. / Applied Surface Science 277 (2013) 155–166

157

the electrode rotation rate, Csat,Cu is the saturation concentration of dissolved species and Cb,Cu is the bulk concentration which is zero. The saturation solubility of copper phosphate in different H3 PO4 concentrations was determined by using Perkin Elmer 2380 atomic absorption spectrophotometer. The mass transfer coefficient (K) of the polishing process, which was used in data correlation, was calculated (Table 2) from the limiting current using the equation K=

IL zAFCCu2+

(2)

4.17 3.55 3.33 1.01 0.89 1.879 4.203 5.732 6.196 14.326

4074.14 9617.74 12,365.81 43,171.38 100,290.52

D (×106 cm2 s−1 )

The scanning electron microscope images were taken using (JEOL, JSM-5300, scanning microscope, OXFORD instrument). For this purpose the copper sheet anode was (1 cm × 1 cm). 2.6. Atomic force microscope (AFM)

1.106 1.231 1.392 1.421 1.605

b (g cm−3 )

 (g cm−1 s−1 )

Sc

2.5. Scanning electron microscope (SEM)

The average surface roughness factor (Ra ) was measured at two points near the centers of the test sheet with an area of 25 ␮m × 25 ␮m by an atomic force microscope (AP-0100 Autoprobe CP-Research, Thermomicroscopes). The resolution is 256 × 256 lines, scan rate is 1 Hz and the AFM images were analyzed using ProScan 1.8 software while the software used for image processing is IP2.1. 2.7. Degree of brightness The degree of brightness of the surface was measured using BRIGHTNESS & COLOR METER; Model No. 68-50-00-0002.

1.162 1.300 1.425 1.432 1.666 6.34 5.02 3.41 2.39 1.78 0.120 0.092 0.056 0.036 0.024 0.98 0.95 0.85 0.78 0.70 6 8 10 12 14

0.600 0.460 0.280 0.180 0.120

k (×103 cm s−1 ) iL (A cm−2 ) IL (A) CCu2+ (×103 mol cm−3 ) Conc. of H3 PO4 (M)

Table 2 Physical properties of H3 PO4 solutions at 25 ◦ C.

i (g cm−3 )

3. Results and discussion 3.1. Electropolishing of copper in H3 PO4 electrolyte without SAS Fig. 1a shows a set of a typical current potential curves (with a well defined limiting current plateau) obtained at different H3 PO4 acid concentrations. It is obvious that the limiting current decreases with increasing H3 PO4 acid concentrations within the range studied (6–14 M) as represented in Table 2. This is in agreement with the finding of other authors who worked within the same range of concentrations using other anode geometry [20–22]. The effect of H3 PO4 acid concentration on the value of the limiting current can be explained on the basis of the mass transfer equation [4]: IL =

ZFD C 2+ ı Cu

(3)

The decrease in the limiting current with increasing H3 PO4 acid concentration is attributed to three effects, first the saturation solubility of copper phosphate in H3 PO4 acid decreases, which makes the onset of limiting current decreases. Second, the viscosity of the solution increases with increasing H3 PO4 acid concentration with consequent decrease in the diffusivity of Cu2+ according to Stock–Einstein equation: D = constant T

(4)

where D is the diffusivity of Cu2+ (cm2 s−1 ),  is the viscosity of solution (g cm−1 s−1 ) and T is the absolute temperature. Third, the increase in solution viscosity with increasing H3 PO4 acid concentration (Table 2) results in an increase in the diffusion layer thickness which represent the resistance to the rate of mass transfer of Cu2+ from anode surface to bulk solution.

158

A.A. Taha et al. / Applied Surface Science 277 (2013) 155–166

6M 8M 10M 12M 14M

a

Current (A)

Current (A)

0.6

0.6

0.3

b

Triton x-100 Conc.(M) 0.00 5 × 10 -6(before CMC) 5 × 10 -5(at CMC) 1 × 10 -2(after CMC)

0.4

0.2

0.0 0.0

0.5

1.0

1.5

2.0

0.0 0.0

Anode poterntial / V

0.4

0.8

1.2

Anode potential /V SDS.(M) 0.00 5 × 10 -6(before CMC) 5 × 10 -4(at CMC) 1 × 10 -2(after CMC)

0.6

c

Current (A)

Current (A)

0.6

0.4

0.2

0.0 0.0

0.4

0.8

1.2

Anode potential /V

d

CPC Conc.(M) 0.00 5 × 10 -6(before CMC) 1 × 10 -3(at CMC) 1 × 10 -2(after CMC)

0.4

0.2

0.0 0.0

0.4

0.8

1.2

Anode potential /V Fig. 1. Typical polarization curves for the electropolishing of vertical plate in presence of different concentrations of (a) H3 PO4 acid, (b) 8 M H3 PO4 + Triton X-100, (c) 8 M H3 PO4 + SDS and (d) 8 M H3 PO4 + CPC (T = 298 K).

tension  values obtained for different concentrations of SAS in 8 M H3 PO4 is represented in Fig. 2. Sharp decrease in surface tension values is observed as the concentration increases and then the curves break rather rapidly at still relatively low concentrations and continue to decrease slowly as the concentration increase. The critical micelle concentration (CMC) was determined from the intersection points in the –log C curves. The reproducibility of the surface tension versus concentration curve was checked by performing at least three separate experiments. The obtained (CMCs) values of the studied surfactants are 5 × 10−5 , 5 × 10−4 and 1 × 10−3 mol/l for Triton X-100, SDS and CPC respectively. The anodic polarization curves for the copper anode that were electropolished in 8 M H3 PO4 electrolytes with the addition of SAS of different concentrations (before, at and after CMC) are shown in Fig. 1b–d. Clearly, a limiting current plateau could be found in each anodic polarization curve. An obvious decrease in the

Triton x-100 SDS CPC

80

γ(mN/m)

Mechanistic studies of EP have revealed that during EP of copper in H3 PO4 (Fig. 1a), when a current passes across the electrolyte, the surface of the anode become covered with a viscous layer. This layer has higher viscosity and higher electrical resistivity than the bulk of the solution. The layer thickness differ from site to site: above protrusions the film is thinner than above the valleys. Hence peaks dissolve more rapidly than valleys. Also the diffusion of metal ions out of a scratch or cavity is slow compared with that over a peak. Due to the non-uniform current density metal from projections dissolves more rapidly than from crevices which produce a surface levelling effect. Brightening is usually attributed to the presence of a compact solid copper oxide film on the anode through which metal dissolution takes place as revealed by mechanistic studies of copper EP [23–28]. According to Hoar and Rothwell [23], Kojima and Tobias [24] electropolishing (brightening) occurs due to the formation of a solid copper oxide film, absent during etching, covering the surface of copper electropolished anode in orthophosphoric acid. Also an oxide film has been detected by Poncet et al. by the electroluminescence technique [25–28], Thus, the observed behavior of a limiting current is due to the fact that the dissolution is under mass transport control [4,29], in which diffusion of the reaction products is limited and is in turn determining the overall reaction rate. The salt-film precipitated mechanism involves rate limiting diffusion of cations of the dissolving metal from the anode into the bulk [30,31]. At the limiting current, a thin salt film with a saturated concentration of metallic cation is present on the anode surface and limits the rate at which metal ions leave the surface.

60

40

20

3.2. Electropolishing of copper in H3 PO4 electrolyte with SAS added

-6

-4

-2

log C

Before discussing the effect of the three studied SAS on dissolution process, the CMC must be determined. Variation of the surface

Fig. 2. Variation in surface tension regarding concentration for Triton X-100, SDS and CPC at 25 ◦ C.

A.A. Taha et al. / Applied Surface Science 277 (2013) 155–166

159

Table 3 Values of limiting current IL (A), inhibition efficiency percentage (IE%) for dissolution of copper in 8 M H3 PO4 in absence and presence of SAS at 25 ◦ C. Triton X-100

0.00 5.0E−07 1.0E−06 5.0E−06 1.0E−05 5.0E−05 1.0E−04 5.0E−04 1.0E−03 5.0E−03 1.0E−02

SDS IE%

IL (A)

IE%

IL (A)

IE%

0.460 0.300 0.285 0.280 0.272 0.305 0.290 0.282 0.274 0.250 0.220

– 34.74 38.04 39.13 40.87 33.70 37.00 38.70 40.43 45.65 52.17

0.460 0.341 0.333 0.312 0.303 0.294 0.280 0.271 0.255 0.244 0.180

– 25.87 27.61 32.17 34.13 36.09 39.13 41.09 44.57 46.96 60.87

0.460 0.350 0.338 0.330 0.320 0.311 0.300 0.290 0.322 0.336 0.350

– 23.91 26.52 28.26 30.22 32.39 34.78 36.96 30.00 26.95 23.91

limiting current was detected when the three studied SAS were added (Table 3). This suggests that the anodic polarization behavior of copper is very sensitive to small concentrations of SAS. The observed decrease in limiting current with the addition of the studied SAS (Fig. 1b–d) is consistent with a salt-film mechanism. When SAS was added, it is probable that adsorption of these compounds on and close to the anode surface could help to control ion diffusion process in three ways namely: (i) filling up of surface cavities and depressions. (ii) The looser packing of the adsorbed film at peaks facilitates the removal of the metal at faster rate than loss from valleys. Therefore surface levelling takes place because the passivation of crevices is more stable and it inhibits etching. Peaks are instead dissolving more rapidly. (iii) By replacement of water molecules in this region. This process may affect the dielectric constant of the medium and lower the number of water molecules available for solvating cations, hence lower the rate of diffusion of cations away from the anode surface. If (IL )blank is the limiting current in absence of SAS and (IL )SAS in the presence of SAS, then IE% can be calculated from the following equation IE% =

IL(blank) − IL(SAS) IL(blank)

× 100

CPC

IL (A)

(5)

The decrease in limiting current is a function of SAS concentration and type. In case of Triton X-100 (Table 3), the limiting current decreases as concentration increases up to 1 × 10−5 mol/l, at CMC (5 × 10−5 mol/l), the limiting current IL increases and after CMC, as Triton X-100 concentration increases, the limiting current decreases. Triton X-100 is a non ionic surfactant composed of a fairly polar hydrophilic head of O (CH2 CH2 O)10 H and a hydrophobic part from C8 H17 C6 H4 (the t-octyl-ph group). The inhibition behavior on dissolution rate of copper in 8 M H3 PO4 by Triton X-100 can be explained on the basis of molecular adsorption and the CMC makes effective boundary conditions. Below CMC, Triton X-100 is typically below the monolayer level, and above which adsorption can consist of multiple layers of physically adsorbed Triton X-100 molecules. Above the CMC, increasing Triton X-100 concentration leads to the gradual formation of multilayers that further reduce the rate of dissolution beyond what can be achieved with monolayer coverage below the CMC. It is likely that additional surface coverage in the form of multilayers and an associated increase in surface and boundary layer viscosity are responsible for the additional increase in dissolution inhibition above CMC [32]. The exceptional increase in dissolution rate of copper that occurs at 5 × 10−5 mol/l Triton X-100 can be attributed to the large micelle size of Triton X-100 which can’t form adsorbed film so limiting current increases and inhibition percentage decreases (Table 3). For SDS which has (C12 H25 SO4 −1 ), it is observed that, the anodic dissolution rate decreases as SDS concentration increases at all concentrations studied (before, at and after CMC). It was found

Table 4 Effect of additives on the IE% value after electropolishing process. Types of additives

Viscosity (mPa s)

IE%

No additives 1 × 10−2 M Triton X-100 1 × 10−2 M SDS 1 × 10−2 M CPC

4.20 5.66 6.52 4.66

– 52.17 60.87 23.91

that, the dissolution retardation action of SDS in 8 M H3 PO4 results from physical (electrostatic) adsorption of negatively charged sulfate ions to positively charged copper surface forming a barrier on copper surface. In early stages of adsorption (low surface coverage), i.e., at the low SDS concentration, the adsorption of the negatively charged sulphate group at copper surface takes place. When the concentration of SDS increases, more SDS was found to adsorb electrostatically on the copper surface. A hemi-micelle barrier composed of SDS ions was formed over the whole surface due to the interaction between the oxygen via Vander Waals forces. The barrier becomes more compact and protective with adsorption of more highly negatively charged SDS micelle (when the SAS concentration is increased above the CMC, the number of micelles increases but the free SAS concentration stays constant at the CMC values). In this way, more decrease in limiting current and higher IE% will observe after CMC of SDS [33]. Further analysis of data in Table 3 clarify that the limiting current of copper dissolution in 8 M H3 PO4 in presence of CPC

inhibition efficiency percentage (IE %)

Conc. of SAS (M)

60

Triton x-100 SDS CPC

50

40

30

1×10-6

1×10-5

1×10-4

1×10-3

1×10-2

Conc.(M) Fig. 3. Inhibition efficiency percentage – concentration dependence curve at 25 ◦ C for the three studied surfactants.

160

A.A. Taha et al. / Applied Surface Science 277 (2013) 155–166

Fig. 4. SEM images: (a) raw sample before polishing, (b) after electropolishing without addition (blank), (c) after EP + 5 × 10−6 M Triton X-100 (before CMC), (d) after EP + 5 × 10−5 M Triton X-100 (at CMC), (e) after EP + 1 × 10−2 M Triton X-100 (after CMC), (f) after EP + 5 × 10−6 M SDS (before CMC), (g) after EP + 5 × 10−4 M SDS (at CMC), (h) after EP + 1 × 10−2 M SDS (after CMC), (i) after EP + 5 × 10−6 M CPC (before CMC), (j) after EP + 5 × 10−4 M CPC (around CMC) and (k) after EP + 1 × 10−2 M CPC (after CMC)

decreases and IE% increases up to 5 × 10−4 M. At and after CMC, limiting current increases and IE% decreases. The net effect of CPC addition on copper dissolution rate is inhibitory or enhancement, depending on the level of CPC concentration. CPC only up to 5 × 10−4 M reduces dissolution rate but when the CPC concentration exceeds this level, the copper dissolution is enhanced [34]. On other hand, the addition of CPC will introduce Cl− anion to

the H3 PO4 acid solution. Cl− anion will be the first adsorbed at the electrode/solution interface through electrostatic attraction and create an excess negative charge toward the solution phase and favor more adsorption of positively charged cetyl pyridinium cation. Thus the CPC will be electrostatically adsorbed on the electrode surface covered with primarily adsorbed chloride ions [35]. The decrease in inhibition efficiency after CMC can be

A.A. Taha et al. / Applied Surface Science 277 (2013) 155–166

161

Table 5 Measured roughness (Ra ), Rq (RMS), peak-valley ratio and brightness of copper samples. Sample

EP conditions

Brightness

Peak-valley ratio (nm)

Ra (nm)

Rq (RMS) (nm)

A B C D E F G H I J K

Before electropolishing After electropolishing without additives After EP + 5 × 10−6 M Triton X-100 (before CMC) After EP + 5 × 10−5 M Triton X-100 (at CMC) After EP + 1 × 10−2 M Triton X-100 (after CMC) After EP + 5 × 10−6 M SDS (before CMC) After EP + 5 × 10−4 M SDS (at CMC) After EP + 1 × 10−2 M SDS (after CMC) After EP + 5 × 10−6 M CPC (before CMC) After EP + 5 × 10−4 M CPC (around CMC) After EP + 1 × 10−2 M CPC (after CMC)

35.80 40.11 53.99 53.11 70.93 56.95 62.88 67.20 54.00 59.51 43.38

133.00 101.00 67.43 77.10 42.00 80.50 54.10 54.10 69.06 43.82 61.58

141.9 112.3 81.16 70.16 47.57 64.32 60.48 56.27 59.32 52.71 65.37

179.5 141.3 103.00 80.16 58.28 85.57 73.35 72.58 76.12 65.57 83.24

Fig. 5. AFM images (two dimensional): (a) raw sample before polishing, (b) after EP without addition (blank), (c) after EP + 5 × 10−6 M Triton X-100 (before CMC), (d) after EP + 5 × 10−5 M Triton X-100 (at CMC), (e) after EP + 1 × 10−2 M Triton X-100 (after CMC), (f) after EP + 5 × 10−6 M SDS (before CMC), (g) after EP + 5 × 10−4 M SDS (at CMC), (h) after EP + 1 × 10−2 M SDS (after CMC), (i) after EP + 5 × 10−6 M CPC (before CMC), (j) after EP + 5 × 10−4 M CPC (around CMC) and (k) after EP + 1 × 10−2 M CPC (after CMC)

162

A.A. Taha et al. / Applied Surface Science 277 (2013) 155–166

Fig. 6. AFM images (three dimensional): (a) raw sample before polishing, (b) after EP without addition (blank), (c) after EP + 5 × 10−6 M Triton X-100 (before CMC), (d) after EP + 5 × 10−5 M Triton X-100 (at CMC), (e) after EP + 1 × 10−2 M Triton X-100 (after CMC), (f) after EP + 5 × 10−6 M SDS (before CMC), (g) after EP + 5 × 10−4 M SDS (at CMC), (h) after EP + 1 × 10−2 M SDS (after CMC), (i) after EP + 5 × 10−6 M CPC (before CMC), (j) after EP + 5 × 10−4 M CPC (around CMC) and (k) after EP + 1 × 10−2 M CPC (after CMC).

interpreted as, the bilayer adsorption occur so the second layer inverse direction of hydrophilic group adsorption of CPC and increase repulsion forces between the adsorbed molecules of CPC [36]. Generally, the viscosity in the presence of three studied SAS play an important role in their effect on the EP process. The conductivity () of electrolytic solution is highly influenced by the difference of viscosity (), of each type of additives according to the equation:  = const./. The case of SDS addition to the electrolytic solution composed of orthophosphoric acid (85%, w/w) showed a high viscosity more than the case of both Triton X-100 and CPC addition, where the addition of SDS leads to the increase in viscosity to 6.52 mPa s; while the measured viscosity was 5.66, 4.66, and 4.20 mPa s in case of Triton X-100, CPC addition and without any addition, respectively (Table 4).

This phenomenon demonstrates the enhancement of inhibition by addition of Triton X-100, SDS and CPC where the increase in viscosity of electrolytic solution leads to the increase in electric resistance to the high velocity of ionic motion through the diffusion layer. Consequently, the aggressive attack of ionic species is decreased, reducing pitting and defects formed over the surface. From the calculated values of inhibition efficiency percentage (IE%) (Table 3 and Fig. 3), it is clear that IE% of each studied SAS is concentration dependant. The difference between inhibition behavior of the three studied SAS due to their different structure and their CMC is best seen in the low and high concentration range. At low concentration, 1 × 10−5 M, (before CMC), the surfactant IE ranked as follows: Triton X-100 > SDS > CPC

A.A. Taha et al. / Applied Surface Science 277 (2013) 155–166

while at high concentration, 1 × 10−2 M, (after CMC) the surfactant IE% ranked as follows:

2.24 Triton x-100 R2 =0.98

SDS > Triton X-100 > CPC.

2.20

log Sh

3.3. Surface analysis

163

Surface analysis include three techniques (1) scanning electron microscope, SEM (2) degree of brightness (3) atomic force microscope, AFM

2.16

2.12 8.0

8.2

8.4

8.6

log (Sc.Gr)

log Sh

2.4

SDS R2 =0.99

2.3

2.2

8.4

8.6

8.8

log (Sc.Gr)

2.32

log Sh

3.3.1. SEM analysis As shown in Fig. 4a–k, the surface morphology before and after EP in the absence of additives and in the presence of different concentrations of SAS. Fig. 4a shows the morphology of copper before EP. A large number of pits with large size and high depth distributed over the surface are seen. In contrast, after EP in 8 M H3 PO4 , the specimen surface was smoother and no pits were observed, where levelling and brightening occur and deep cavities were filled up but grain boundaries are still observed (Fig. 4b). Fig. 4c represents the SEM image of copper in presence of 5 × 10−6 M Triton X-100 [before CMC]. It shows uniform, smooth and bright surface to some extent more than blank sample (Fig. 4b). Deep cavities are eliminated by filling up, also, grain boundaries are decreased gradually. This behavior may be due to involvement of Triton X-100 molecules in the cavities of copper surface so appear more uniform than blank. By increasing Triton X-100 to 5 × 10−5 M (at CMC), the SEM micrograph (Fig. 4d) shows that smoothness, brightness increases but uniformity decreases while grain boundaries increase more than blank (Fig. 4b). This behavior may be attributed to formation of micelles by Triton X-100 and releasing of it from peaks of copper surface in order to aggregate and form micelles. Addition of 1 × 10−2 M Triton X-100 (after CMC) to the electrolyte revealed the enhancement of surface. Grain boundaries are completely diminished (Fig. 4e). The well polished surface may be due to increase in the adsorption ability (52.17% – Table 3) of Triton X-100 molecules, leading to filling up of all deep cavities. Fig. 4f shows the SEM micrograph of copper specimen in presence of 5 × 10−6 M SDS (before CMC). Only a slight difference was observed more than blank, where grain boundaries are still observed. This may be due to weak adsorption of SDS molecules at this concentration. In presence of 5 × 10−4 M SDS (at CMC) (Fig. 4g), electropolished surface appear uniform, smooth and bright more than Fig. 4f. This seems to be due to micelle formation; the micelles are highly charged and because of the higher electrostatic attraction, the adsorption of SDS increases and consequently the grain boundaries are reduced [37]. On addition of 1 × 10−2 M SDS (after CMC), levelling and brightening are obviously occurred (Fig. 4h). The grain boundaries grooves disappeared and a considerable smooth surface was revealed. This may be due to the increasing of adsorption power of SDS molecules on copper surface. This behavior confirms the higher inhibition efficiency of SDS at this concentration (60.87% – Table 3). The micrograph of the specimen in presence of 5 × 10−6 M CPC [before CMC] is shown in Fig. 4i, where levelling and brightening are occurred. Only a slight difference was observed compared to blank, where the grain boundaries are still represented on the surface of copper but it appears uniform more than blank. After EP in presence of 5 × 10−4 M CPC (around CMC) (Fig. 4j) the brightness and uniformity of the surface was better than image 4i, where grain boundaries are completely disappeared.

CPC R2 =0.98

2.28

2.24

2.20 8.2

8.3

log (Sc.Gr) Fig. 7. The overall mass transfer correlations in presence of different concentrations of surfactants.

In presence of higher concentration of CPC (after CMC) (Fig. 4k) non-uniform surface, small pits, and grain boundaries return to appear clearly. This indicates that specific adsorption of catalytically active chloride counter ion on the copper electrode promotes the dissolution [38].

3.3.2. Degree of brightness The results obtained from SEM were ascertained by measuring the degree of surface brightness after EP treatment. The increase in degree of surface brightness was clearly observed (Table 5), where the brightness value before EP is 35.80. After EP in 8 M H3 PO4 is 40.11 while after EP by addition of Triton X-100 (after CMC), SDS (after CMC) and CPC (around CMC), the degree of surface brightness increased to 70.93, 67.20 and 59.51 respectively. It is considered that, in the EP process, the polishing action is set up by one anode layer and anode film. The anode layer of electrolyte, whose thickness is considerably greater than the wave length of the electrode profile for successful EP to take place. The anode film is a thin layer of solid material or adsorbed atoms and molecules. The polishing action is considered to be due to a diffusion-controlled

164

Table 6 General correlation of free mass transfer in presence of surfactants at 25 ◦ C. Conc. (M) Blank

CCu2+ (×103 mol cm−3 ) 0.95

IL (A)

iL (A cm−2 )

initial (g cm−3 )

bulk (g cm−3 )

 (g cm−1 s−1 )

 (cm2 s−1 )

k (×104 cm s−1 )

D (×106 cm2 s−1 )

Sh

Sc

Gr

0.46 0.61 0.66 0.76

0.092 0.122 0.132 0.152

1.3000 1.2952 1.2922 1.2895

1.2312 1.2300 1.2288 1.2267

0.04203 0.03620 0.03249 0.03126

0.03414 0.02943 0.02644 0.02548

5.02 6.65 7.20 8.29

3.55 4.47 5.56 6.55

141.344 148.857 129.484 126.567

9615.948 6583.911 4755.766 3890.286

44507.20 56957.89 68769.38 73505.31

1.10 1.20 1.30 1.37 1.47 1.58 1.75 1.92 2.00 2.00

0.300 0.285 0.280 0.272 0.305 0.290 0.282 0.274 0.250 0.220

0.0600 0.0570 0.0560 0.0544 0.0610 0.0580 0.0564 0.0548 0.0500 0.0440

1.2832 1.2875 1.2900 1.2936 1.3025 1.3046 1.3057 1.3059 1.3068 1.3118

1.2516 1.2602 1.2702 1.2763 1.2831 1.2947 1.2989 1.3053 1.3111 1.3172

0.05044 0.05121 0.05230 0.05305 0.05055 0.05106 0.05300 0.05423 0.05506 0.05667

0.04030 0.04063 0.04118 0.04157 0.03939 0.03944 0.04080 0.04154 0.04200 0.04302

2.83 2.46 2.23 2.06 2.15 1.90 1.67 1.48 1.30 1.14

1.75 1.55 1.30 1.11 1.31 1.24 1.13 1.06 0.96 0.72

161.496 158.783 171.689 185.352 164.128 153.388 147.776 139.513 134.931 158.318

22974.47 26198.6 31626.55 37514.66 30140.52 31728.46 36043.00 39377.92 43675.46 59724.12

14861.71 12585.25 8871.069 7607.241 9401.062 6683.053 4845.57 3931.239 2426.016 582.2885

SDS 5.0E−07 1.0E−06 5.0E−06 1.0E−05 5.0E−05 1.0E−04 5.0E−04 1.0E−03 5.0E−03 1.0E−02

0.90 1.00 1.15 1.25 1.35 1.50 1.65 1.75 1.90 2.20

0.341 0.333 0.312 0.303 0.294 0.280 0.271 0.255 0.244 0.180

0.0682 0.0666 0.0624 0.0606 0.0588 0.0560 0.0542 0.0510 0.0488 0.0360

1.2566 1.2742 1.2935 1.3256 1.329 1.3366 1.34 1.3507 1.355 1.3646

1.2331 1.2508 1.2695 1.3012 1.3043 1.3116 1.3146 1.325 1.3292 1.3386

0.04291 0.04388 0.04606 0.04647 0.05230 0.05267 0.05312 0.05759 0.06141 0.06522

0.0348 0.03508 0.03629 0.03571 0.0401 0.04016 0.04041 0.04346 0.0462 0.04872

3.93 3.45 2.81 2.51 2.26 1.93 1.70 1.51 1.33 0.85

3.04 2.89 2.52 2.03 1.83 1.56 1.34 1.18 1.05 0.91

129.325 119.611 111.433 123.557 123.119 123.998 127.204 128.510 127.105 93.181

11461.68 12159.7 14381.75 17567.12 21877.78 25742.68 30200.1 36990.77 44127.48 53543.86

14972.47 14624.64 13806.92 14145.68 11326.85 11365.2 11375.68 9872.357 8742.236 7866.463

CPC 5.0E−07 1.0E−06 5.0E−06 1.0E−05 5.0E−05 1.0E−04 5.0E−04 1.0E−03 5.0E−03 1.0E−02

0.90 0.90 1.00 1.10 1.25 1.366 1.483 1.516 1.516 1.533

0.350 0.338 0.330 0.320 0.311 0.300 0.290 0.322 0.336 0.350

0.0700 0.0676 0.0660 0.0640 0.0622 0.0600 0.0580 0.0644 0.0672 0.0700

1.3002 1.3029 1.31907 1.31973 1.31982 1.32332 1.32338 1.32594 1.32729 1.3287

1.2817 1.2857 1.2932 1.3057 1.3077 1.3085 1.3108 1.3148 1.3157 1.3173

0.0507 0.05184 0.05273 0.05373 0.05436 0.05519 0.05615 0.05101 0.05058 0.04660

0.03955 0.04032 0.04078 0.04115 0.04157 0.04218 0.04283 0.03880 0.03844 0.03537

4.03 3.89 3.42 3.01 2.58 2.28 2.03 2.20 2.30 2.36

2.65 2.29 1.96 1.57 1.33 1.10 0.88 1.01 1.13 1.23

152.073 169.946 174.473 192.013 193.853 206.89 230.799 217.925 203.252 192.351

14943.14 17639.35 20782.67 26212.36 31301.51 38411.45 48774.93 38606.03 33898.71 28852.46

8912.425 7956.567 11559.9 6151.586 5208.196 6169.973 5077.406 5469.468 5790.97 6719.817

A.A. Taha et al. / Applied Surface Science 277 (2013) 155–166

Triton X-100 5.0E−07 1.0E−06 5.0E−06 1.0E−05 5.0E−05 1.0E−04 5.0E−04 1.0E−03 5.0E−03 1.0E−02

A.A. Taha et al. / Applied Surface Science 277 (2013) 155–166

anodic dissolution process being set up through the anode layer and film. To decrease the micro-irregularities and increase the brilliance of the finish, it is considered that it is necessary to achieve uniform formation and diffusion of cations from the anode surface, on a micro scale. It is probable that the addition of the critical concentration of the correct type of surfactant contribute to this control for the system copper in H3 PO4 acid. It is probable that the adsorption of the three studied surfactants supplements the action of the normal anodic film in improving the microstructure of the surface. An alternative cause could be that, the adsorbed molecules of the SAS on the metal surface being polished thus eliminates the difference between different planes of the crystal facing the solution, i.e., all planes are electropolished at the same rate. This role is similar to the role played by the oxide film which leads to brightness of the surface [23–28]. 3.3.3. Atomic force microscope, AFM The characterization of microstructure of the copper surface has been achieved using the atomic force microscope (AFM). The two dimensional topography of copper before and after EP without and with additives (surfactants) are given in Fig. 5a–k. The corresponding three dimensional AFM images are represented in Fig. 6a–k. Raw sample (Fig. 5a) looked uneven and appeared to have potholes with random hill like structure. The non uniformity and non homogeneity is the general appearance of the specimen. After EP in 8 M H3 PO4 acid in absence of additives. AFM images (Fig. 5b), revealed that the roughness (Ra) decreased from 141.9 nm to 112 nm without addition of any material (Table 5). The image is relatively uniform and some parts had a random hill like structure. On the other hand Ra was decreased from 141.9 nm to 47.57 nm by addition of Triton X-100. In presence of different concentrations of Triton X-100, copper surface (Fig. 5c–e and Table 5) appears flat, compact and homogenous. Random hills structure disappeared and replaced with uniform surface. From Table 5, it is clear that, at CMC of Triton X-100, the peak-valley ratio (R P-V) exhibits a slight increase than that before CMC. This may be explained as a result of the change of adsorption behavior of Triton X-100 molecules on the surface of copper. At 5 × 10−6 M, there is a slight surface coverage by Triton particles including peaks and valleys which lead to uniform surface levelling but at CMC the adsorbed layer became loose, where Triton X-100 tend to self assembled to form micelle, so we may conclude that the molecules of Triton X-100 released from peaks which lead to localized difference in the dissolution rate of copper and lead to formation of small protrusions. While after CMC, the adsorption of Triton X-100 on copper surface increases which lead to decreasing in the R P-V values and smoothening and flattening of the surface. In presence of SDS (Fig. 5f–h and Table 5), the surface appear uniform homogenous, compact and even. Random hills structures are diminished and replaced with the grain boundaries structures. The surface looked relatively flat, the Ra decreases from 141.9 nm to 64.32, 60.48 and 56.27 nm in the presence of the three studied concentrations of SDS. This behavior of SDS molecules can be attributed to the adsorption ability of SDS on copper surface. It increases by increasing SDS concentration. Inspection of AFM micrographs of electropolished copper surface in presence of 5 × 10−6 , 5 × 10−4 M CPC (Fig. 5i and j) reveal that there is no potholes and random hills in the structure of the surface, the surface looked uniform, compact, while in presence of 1 × 10−2 M CPC, the obtained copper surface (Fig. 5k) look bright and smooth on macro scale. But, the AFM analysis revealed the non uniform, non homogenous surface, irregular random hill return to appear, protrusions and potholes are represented in surface and additional roughening (Ra = 65.73 nm) as compared to

165

the values obtained in presence of 5 × 10−6 (Ra = 59.32 nm) and 5 × 10−4 (Ra = 52.71 nm) M CPC. Ra decreases from 141.9 nm to 59.32, 52.71 and 65.37 nm for 5 × 10−6 , 5 × 10−4 and 1 × 10−2 M CPC respectively [36]. 3.4. Data correlation The mass transfer correlation for natural convection can be described as: Sh = a(Sc Gr)b

(6)

where Sh is the Sherwood number (Sh = kL/D), Sc is the Schmidt number (Sc = /D), Gr is the Grashof number (Gr = (gL3 (initial − bulk )/initial 2 )), and a and b are constants. K is mass transfer coefficient (cm s−1 ), L is the length of electrode. D is the diffusion coefficient (cm2 s−1 );  is the kinematics viscosity, cm2 s−1 , g is the acceleration gravity 980 cm s−2 , initial , is the solution density at the interface, bulk , is the solution density at the bulk in g cm−3 . The values of dimensionless groups Sh, Sc and Gr used in obtaining correlation (6); the physical properties ,  and D used in calculating these dimensionless groups were measured as above (Sections 2.3 and 2.4) and were given in Table 6. By plotting log Sh against log (Sc Gr), straight lines were obtained; their slopes gave the constant b and their intercept gave the constant a. Fig. 7 shows that mass transfer data at vertical plates in presence of surfactants for the conditions 8.14 × 107 < (Sc Gr) < 1.01 × 109 fit the following equations: Triton X-100 SDS CPC

Sh = 11.22(Sc Gr)0.36 ± 0.0248 Sh = 0.457(Sc Gr)0.30 ± 0.0311 Sh = 0.174(Sc Gr)0.36 ± 0.0381

The results obtained in our system are in good agreement with the previous study [39,40]. Exponents in all equations denote a turbulent flow mass transfer mechanism in presence of surfactants molecules. 4. Conclusions • The addition of the correct type and concentration of the studied SAS improved both smoothness and brightness of electropolished copper surface. • EP treatment of copper in a bath composition containing the three studied SAS could increase the ability of the bath to produce continuous polishing over an anode. • The degree of brightness of the polished surface increases from 40.11 in absence of SAS to 70.93, 67.20 and 59.51 in presence of 1 × 10−2 M Triton X-100, 1 × 10−2 M SDS and 5 × 10−4 M CPC respectively. • Optimum EP was obtained after CMC (at 1 × 10−2 M) in case of Triton X-100 and SDS, while in presence of CPC the optimum EP was obtained at 5 × 10−4 M (around CMC). • Improvement produced in EP by the studied SAS was due to the adsorption of such agents on the anode surface. • These SAS played the part of inhibitors, reducing the etching action of the acid, and increasing the brightness of the surface. • The enhancement of the surface morphology, topography and brightness was achieved by addition of three SAS to the electrolytic solution. • The values of roughness (Ra ) indicated that EP of copper surface is more effective in case of addition of Triton X-100 (Ra = 47.57 nm) more than both SDS (Ra = 56.27 nm) and CPC (Ra = 52.71 nm). • In the presence of the three studied SAS, the solution appeared promising, and a distinct improvement in the finish was noted.

166

A.A. Taha et al. / Applied Surface Science 277 (2013) 155–166

References [1] J.D. Reid, US Patent #6,653,226, November 2003. [2] W.D. Chandler, US Patent #6,579,439, June 2003. [3] S. Mohan, D. Kanagaraj, R. Sindhuja, S. Vijayalakshmi, N.G. Renganathan, Transactions of the Institute of Metal Finishing 79 (2001) 140–142. [4] D. Landolt, Electrochimica Acta 32 (1987) 1–11. [5] M. Matlosz, D. Landolt, Journal of Electrochemical Society 136 (1989) 919–929. [6] R. Sautebin, H. Froidevaux, D. Landolt, Journal of Electrochemical Society 127 (1980) 1096–1100. [7] C. Clerc, M. Datta, D. Landolt, Electrochimica Acta 29 (1984) 1477–1486. [8] M. Datta, D. Landolt, Journal of Electrochemical Society 122 (1975) 1466–1472. [9] M. Datta, D. Landolt, Electrochimica Acta 25 (1980) 1255–1263. [10] B. Dua, I.I. Suani, Electrochemical and Solid-State Letters 8 (10) (2005) 283–285. [11] V. Vignal, J.C. Roux, S. Flandrois, A. Fevrier, Corrosion Science 42 (2000) 1041–1051. [12] G.L. Wynick, C.J. Boehlert, Materials Characterization 55 (2005) 190–202. [13] G.I. Edson US Patent #4,663,005, 1987. [14] A.M. Awad, N.A. Abdel Ghany, T.M. Dahy, Applied Surface Science 256 (2010) 4370–4375. [15] A.A. Taha, Anti-Corrosion Methods and Materials 47 (2) (2000) 94–104. [16] G.M. El-Subruiti, A.M. Ahmed, Portugaliae Electrochimica Acta 20 (2002) 151–166. [17] S.M. Abd El-Haleem, A.M. Ahmed, A.E. El-Nagger, Journal of Dispersion Science and Technology 31 (4) (2010) 512–528. [18] H.H. Abdel Rahman, A.H.E. Moustafa, M.K. Awad, International Journal of Electrochemical Science 7 (2012) 1266–1287. [19] N.M. Elmalah, S.M. Abd Elhaliem, A.M. Ahmed, S.M. Ghozy, International Journal of Electrochemical Science 7 (2012) 7720–7739. [20] I. Nirdosh, G.H. Sedahmed, International Journal of Computational Heat Transfer 23 (5) (1996) 633–642. [21] M.M. Zaki, I. Nirdosh, International Journal of Computational Heat Transfer 27 (3) (2000) 435–442.

[22] A.M. Ahmed, S.M. Abd El-Haleem, M.A. Darweesh, A.E. El-Nagger, Bulletin of Electrochemistry 19 (11) (2003) 489–504. [23] T.P. Hoar, G.P. Rothwell, Electrochimica Acta 9 (1964) 135–150. [24] K. Kojima, C.W. Tobias, Journal of the Electrochemical Society 142 (1973) 1202–1210. [25] P. Poncet, M. Braizaz, B. Pointu, J. Rousseau, Journal de Chimie Physique 4 (1977) 74–78. [26] P. Poncet, M. Braizaz, B. Pointu, J. Rousseau, Journal de Chimie Physique 3 (1978) 75–80. [27] P. Poncet, M. Braizaz, B. Pointu, J. Rousseau, N. Muhlstein, Journal of Electroanalytical Chemistry 122 (1981) 111–131. [28] B. Pointu, P. Poncet, Thin Solid Films 79 (1981) 125–131. [29] D. Landolt, P.F. Chauvy, O. Zinger, Electrochimica Acta 48 (2003) 3185–3201. [30] R.D. Grimm, D. Landolt, Corrosion Science 36 (1994) 1847–1868. [31] R.D. Grimm, A.C. West, D. Landolt, Journal of Electrochemical Society 139 (1992) 1622–1629. [32] M.L. Free, Corrosion Science 44 (2002) 2865–2870. [33] S.S. Abd El Rehim, M.A. Amin, S.O. Moussa, A.S. Ellithy, Materials Chemistry and Physics 112 (2008) 898–906. [34] M. Baghalha, M.K. Ahmadi, Corrosion Science 53 (2011) 4241–4247. [35] S.A. Abd El-Maksoud, Journal of Electroanalytical Chemistry 565 (2004) 321–328. [36] N. Tantawy, Evaluation of new cationic surfactant as corrosion inhibitor for carbon steel in a metal working fluid, The Annals of University “Dun˘area De Jos” of Galat¸i Fascicle, vol. VIII, 2005, ISSN 1221-4590, Tribology. [37] R.F. Godec, Electrochimica Acta 52 (15) (2007) 4974–4981. [38] C.M. Murira, C. Punkt, H.C. Scniepp, B. Khusid, I.A. Aksay, Langmuir 24 (2008) 14269–14275. [39] I. Nirdosh, G.H. Sedahmed, International Communications in Heat and Mass Transfer 22 (2) (1995) 261–270. [40] M. Zaki, I. Nirdosh, G.H. Sedahmed, Chemical Engineering and Processing 42 (2003) 977–984.