Electropolishing of surfaces: theory and applications

Chemistry basis of electropolishing

Chemically speaking, electropolishing is a specific type of electrolysis that involves a direct electric current passing through an electrolyte in an electrolytic cell (Fig. 1). ⁶ The metal piece to be electropolished serves as the anode and is connected to a positive terminal of a direct current (DC) power supply with the negative terminal being attached to the cathode. ² The piece is then immersed into a bath comprised of a specially formulated, temperature-controlled electrolytic solution. The activation of the power supply produces an electrical current that passes from the anode to the cathode, resulting in the oxidation of the metal surface and the removal of surface impurities and irregularities, which are dissolved in the electrolyte and diffuses through the film to the cathode at a controlled rate.^7,8 At the cathode, a reduction reaction occurs, which normally produces hydrogen. ¹ The amount of metal removed depends on the specific bath, temperature, current density and the particular metallic workpiece being electropolished. ⁹ Usually, current and time are the two variables that can be controlled to reach the reproducible surface finish. ¹⁰ Electropolishing can be described by the Faraday's laws of electrolysis. Faraday's first law of electrolysis and Faraday's second law of electrolysis state that the amount of a material deposited on an electrode is proportional to the amount of electricity used and the amount of different substances liberated by a given quantity of electricity is proportional to their electrochemical equivalent (or chemical equivalent weight).^11,12 OPEN IN VIEWER

Electrolytes used for electropolishing are most often concentrated acid solutions having a high viscosity, such as mixtures of sulphuric acid and phosphoric acid.^13,14 Other electropolishing electrolytes reported in the literature include mixtures of perchlorates with acetic anhydride and methanolic solutions of sulphuric acid. Electropolishing has many applications in the metal finishing industry because of its simplicity and that it can be applied to objects of complex shape. ¹⁵ Typical examples are electropolished stainless steel drums of washing machines and stainless steel surgical devices. ¹⁶ Electropolishing is also commonly applied to the preparation of thin metal samples for transmission electron microscopy because electropolishing does not cause mechanical deformation of surface layers usually observed when mechanical polishing is used.^9,17 Ultra high vacuum (UHV) components are typically electropolished in order to have a smoother surface for improved vacuum pressures, outgassing rates and pumping speed. ¹⁸

Surface phenomena occurring during electropolishing

Understanding the surface processes occurring on the metal surface during electropolishing is the key to derive its mechanism. It can also help optimise the parameters of practical operations. ¹⁹ Electropolishing may be limited by mass transport, more specifically, diffusion of the dissolving metal ion, acceptor species from the electrolyte and water molecules. Commonly the diffusion limitation is the metal ion. Subsequently, a salt film precipitates on the surface, either, porous or both compact and porous. It is believed that salt layer formation takes place at the bottom of pits and there is an ionic transport through this layer because of the presence of a high electric field across the layer. ²⁰ A compact film on the surface will be formed during electropolishing and gas bubble evolution does not occur on the surface even though the potentials are high enough to thermodynamically allow bubble formation due to the non-electron-conducting nature of the film. ²¹ The mechanism of conduction is thought to be solid state conduction in the compact layer and the high resistance of the compact film is attributed to low mobility of ions for solid state transport. ²² The thickness of the compact film is considered to be in the order of 10 nm due to the low mobility of ions by solid state transport process. ²³ The conduction in the porous region is by migration in the electric field. There is no diffusion in the porous region since the pores are filled with the electrolyte saturated with the dissolved metal ions.^24,25 The low porosity of the porous film and its large thickness on the order of several microns also contribute to the high resistance. The current density is controlled by the diffusion limited acceptor ion approach to the surface, i.e. the concentration gradient of the metal ions across the diffusion layer. ²⁶ The film formed on a surface undergoing dissolution is not a simple oxide but rather a contaminated oxide that has incorporated anions from the electrolytic solution. ²⁷

Electropolishing proceed by the dissolution or oxidation of the metal, during which cations are released and adsorbed onto the surface (Fig. 2). The blocking of the surface by these adsorbed ions may lead to overpotential for metal dissolution. ²⁸ The cations are removed from the surface by the acceptor ions that diffuse to the surface to solvate the metal ions. The process of electropolishing results in the density near the electrode different from the bulk of the solution resulting in a hydrodynamic flow. At steady state, the concentration of the acceptor ions at the surface is fixed at zero if they are consumed as soon as they reach the surface. A recent study of copper electropolishing in phosphoric acid and niobium electropolishing in sulphuric acid-hydrofluoric acid mixture verified that the acceptor species was the rate limiting species that diffuses to the surface.^29,30 The potential transients measured for copper electropolishing displayed a sharp rise in potential with applied current, followed by a potential increase owing to the activation overpotential and mass transport overpotential. ³⁰ The huge jump in potential after the transient time indicated a very high resistance. OPEN IN VIEWER

Cathode, electrolyte and viscous layers

For electropolishing, the anode is connected to a positive terminal of a DC power supply with the negative terminal being attached to the cathode. Current flows in from the power supply and the cathode is the terminal where current flows back to the power supply. Cathode polarity with respect to the anode can be positive or negative depending on the device and the way it operates. ³² Positively charged cations always move towards the cathode and negatively charged anions move away from it. In the Electropolishing setup, the anode is positive and the cathode is negative. ³³

The electrolyte is the substance that produces an electrically conducting solution when dissolved in a polar solvent, such as water. ⁴ The dissolved electrolyte separates into cations and anions, which disperse uniformly through the solvent. During electropolishing, the cations of the solution would be drawn to the electrode that has an abundance of electrons, while the anions would be drawn to the electrode that has a deficit of electrons. ⁸ The movement of anions and cations in opposite directions within the solution amounts to a current. Different metals may need different types of electrolyte. The composition and properties of the electrolyte is very important for the electropolishing quality. ¹ In addition to metal salts, Electropolishing electrolytes usually contain a number of additives for various purposes. Some agents are used to increase electrolyte conductivity (supporting electrolytes). Others may be used for increasing bath stability (stabilisers), activating the surface (surfactants or wetting agents), improving levelling or metal distribution (levelling agents), or optimising the chemical, physical or technology properties of the surface.^34,35 These properties include corrosion resistance, brightness or reflectivity, hardness, mechanical strength, ductility, internal stress, wear resistance or solderability.^36,37 The properties of electrolyte are usually characterised by electrolytic conductance, which is different from electrical conductance in metal. Both inorganic and organic salts, acids and alkalis can be used to increase the electrolytic conductance. The conductivity of an electrolyte is a function of the degree of dissociation, the mobility of the individual ions, the temperature and viscosity, and the electrolyte composition. ³¹

During Electropolishing, a viscous layer on the workpiece is formed.^5,38 The non-uniform thickness of the viscous layer over the material surface results in a different ohmic resistance from the cathode to the anode, which causes greater dissolution of the protruded parts compared to the depressed part, thus creating a uniform surface profile. ³⁹ A varying concentration gradient of the dissolved metal ions over the protrusions and valleys occurs (Fig. 3). The thickness of the diffusion layer over the protrusion is smaller than the value over the valley. Therefore, the value of the limiting current would be larger over the protrusion compared to that over the valley, resulting in the metal surface being levelled out as electropolishing proceeds. ³¹ Mass transport control has been established as the reason behind electropolishing, which is associated with the voltage range and is characterised by a constant current density, independent of the voltage applied.^1,40 OPEN IN VIEWER

Relation between anode potential and anode current density

Faraday's laws provide the theoretical basis of electrode potential and current density.^11,41 However, in a real application, many factors influence the eletropolishing quantity and quality. Faraday's laws state that the amount of chemical charge at an electrode is exactly proportional to the total quantity of the electricity. However, side reactions may consume the product if taking place simultaneously at the electrode. ⁴² Therefore, inefficiencies may arise from the side reactions other than the intended reaction takes place at the electrodes. In electropolishing, sufficient voltage should be provided by the power source. The voltage–current relationship follows the Ohm's law, i.e. the current is driven by a potential difference, or voltage through the conducting medium, either electrolytic or metallic. The local current density is a very important variable in electroplating operations and affects the character of the electropolishing. The local current density on an electrode, defined as the current in amperes per unit area of the electrode, is a function of the position on the electrode surface and the current passing through the electrode that accomplishes the desired electropolishing is expressed in terms of current efficiency.^40,43 The current distribution over an electrode surface is complicated and tends to concentrate at edges and points, and unless the resistance of the solution is very low, the current will flow to the workpieces near the opposite electrode more readily than to the more distant workpieces. It is desired to operate processes with uniform current distribution, i.e. the current density is the same at all points on the electrode surface. ⁴⁴

When the workpieces are immersed into a solution, equilibrium is reached between the tendency of the metal to enter solution as ions and the opposing tendency of the ions to lose their charge and deposit on the cathode. The equilibrium is dynamic with metal being ionised and discharged as well as being deposited and reduced. These two effects cancel each other and there is no net change in the system. For the realisation of metal deposition at the cathode and metal dissolution at the anode, the equilibrium is broken by an external potential, which makes the electrode reactions take place at a practical rate. The extra energy needed to force the electrode reaction to proceed at a required rate (or its equivalent current density) is quantified in terms of overpotential or electrode polarisation, the difference in the electrode potential of an electrode between its equilibrium potential and its operating potential when a current is flowing. ⁴⁵ The operating potential of an anode is always more positive than its equilibrium potential, while the operating potential of a cathode is always more negative than its equilibrium potential. ⁴⁶ The value of the overpotential depends on the inherent speed of the electrode reaction, which increases with increasing current density. A slow reaction (with small exchange current density) will require a larger overpotential for a given current density than a fast reaction (with large exchange current density). Since an electrode reaction commonly occurs in more than one elementary step, there is an overpotential associated with each step. Even for the simplest case, the overpotential is the sum of the concentration overpotential and the activation overpotential. The terms overpotential is sometimes called overvoltage, the difference between the cell voltage (with a current flowing) and the open-circuit voltage. ⁴⁷ Consequently, the cell voltage of the electrolytic cell is always more than its open-circuit voltage. The overvoltage is the sum of the over potentials of the two electrodes of the cell and the ohmic loss of the cell.^48,49

For electropolishing, the current source is a power source in the form of a battery or rectifier that converts alternating current electricity to regulated low-voltage DC current. The geometric shape and contour of a workpiece to be polished affect the quality of the eletropolishing. The DC current flows more densely to sharp edges than to the less accessible recessed areas, making electropolishing possible. In other words, the current distribution is not uniform. Therefore, a judicial placement of the anode(s) as well as modifications of the current density is required to obtain quality eletropolishing results.^50,51

Effect of pulsed current

The conventional electropolishing is normally performed in media with high viscosity and/or low conductivity such as concentrated acid (phosphoric acid, sulphuric acid and their mixtures or in perchloric acid-acetic acid solutions) and non-aqueous solutions (ethylene glycol, methanol-sulphuric acid). ¹ Sometimes methanol, instead of water is used as a solvent.^13,14,52 In the last century late fifties and early sixties, the development of electrochemical machining showed that fast dissolution of metals in natural electrolytes under vigorous convection conditions may lead to electropolished surfaces, ⁵³ with the controlling factors being convective mass transfer.^54,55 Soon after, it was found that the electropolishing in neutral salt solutions could be reached at relatively low average current densities and flow rates if pulse current with high density was utilised. ⁵⁶ More tables and electropolishing electrolytes for various metals can be obtained in the books by W. J. McTegart ⁵ and P. V. Shigolev. ⁵⁷

Basic control parameters governing the performance

The key parameters related to voltage–current curve include applied voltage range, the current density range. The factors related to electrolytes include the component ratio of the species, temperature, the electrolyte bath age, the convection of the electrolyte bath and duration of the whole electropolishing time.^29,32,73

As many factors are involved and have to be optimised in an electropolishing process, there is no one-fit-all parameter set for all electropolishing setups. First, the electrode-electrolyte system should be specified. The choice of electropolishing electrolyte is highly dependent on the anodic metal type. Second, the parameters related the electropolishing process are normally optimised against available current–voltage relationship, which has normally been documented in literature for specific electrode-electrolyte systems. Many electropolishing systems exhibit general current–voltage curves akin to the one given in Fig. 6.^61,74-76 As aforementioned, in Regime I, what primarily occurs is the direct dissolution of the metal (i.e. etching occurs). In Regime II, the slight decrease current density with increase applied voltage normally indicates the generation of the passive layer on the anodic surface. In Regime III, the passivation layer is stabilised and anions dissolved from the metal surface diffuse through the passivated layer. In Regime IV, with the further increment of the voltage, the passivated later breaks down on anodic surface. Also in this area, the anodic metal surface will begin pitting. It should be mentioned that, despite the well-established current–voltage relationship, the refinement to the current–voltage curve is necessary for a given electropolishing cell. OPEN IN VIEWER

Normally, the electrolytes suitable for electropolishing are acids (e.g. perchloric, sulphuric, phosphoric, etc.) or their mixtures in a certain solvent, such as water, alcohol and acetic acid. In order to increase the solution viscosity, glycerol, butyl glycol or urea, etc. are usually added.^43,73

Copper is one of the most intensively studied workpieces for electropolishing system. Many types of electrolytes^32,75 or electrolytes mixtures^61,76-80 meet the requirement in polishing copper. Some of the previous work, under which conditions the electropolishing process was performed has been summarised in Table 1.

OPEN IN VIEWER

Table 1

Summary of major parameters of electropolishing process on copper workpiece

Electrolyte	Temperature	Current density range	Voltage range	Literature reference
Anhydrous orthophosphoric acid	25°C	Not specified	Not specified	P.A. Jacquet 81
Orthophosphoric acid solution (Non-agitated Solution)	−5.6°C to 53°C	0.2 to 1.5 A/dm2	0.7 to 0.9 V	P.A. Jacquet 61
Orthophosphoric acid/water solution (Agitated solution)	−5.6°C to 53°C	0.2 to 1.5 A/dm2	Not specified	P.A. Jacquet 61
Orthophosphoric acid/water solution	25°C	0.0625 A/cm2	0.8 to 1.2 V	J. A. Allen 77
Phosphoric acid/water solution with/without glycerol, ethylene glycol	17 to 25°C	2 to 4.2 mA/cm2	0.25 to 1.5 V	H.F Walton 78
Phosphoric acid/copper phosphate/water solution	25°C	50 to 1000 mA/cm2	0.2 to 0.8 V	S.H. Glarum J.H. Marshall79,82
Orthophosphoric acid/sulphuric acid/water solution	25°C	0.1 A/cm2	1.8 to 2.2 V	E.C.W Perryman 80

Another type of metal that is a good raw model to illustrate the importance of the electropolishing condition optimising is the stainless steel. Stainless steel has found vast applications in nowadays society. However, their standard colour-light grey does not always fit in the surroundings where they are applied. Surface treatments, in this sense, are crucial to their versatility. Before the surface treatments executed, a surface polishing step is a must. ⁸³ Among many surface polishing methods, such as mechanical polishing or grind, chemical polishing and electropolishing, the latter method has many advantages versus its counterparts.^84,85 For example, the dissolution rate of the metallic surface can be removed at a significantly increased rate compared with chemical polishing, with an enhanced-quality of the metal surface finish.

The stainless steel surface smoothness also plays a key role for some biomedical applications. For example, it is known that the substrate roughness affects the adhesion of the bacteria.^{86 87 –88} With a careful electropolishing process, the stainless steel would have a microscopically smoother surface, leading to a decreasing bacterial attachment. ⁸⁹

The selection of the electropolishing electrolyte of the stainless steel has some prerequisites:

Being capable of dissolving metallic surface layer on some preferential parts on an atomic level;

The major processing parameters significantly affect the properties of the stainless steel, such as the surface smoothness and anticorrosion performance.^43,90 Table 2 is a brief summary on many work done on electropolishing stainless steel previously.

OPEN IN VIEWER

Table 2

Summary of major parameters of electropolishing process on stainless steel workpiece

Electrolyte	Stainless steel type	Temperature	Current density range	Voltage range	Literature reference
Phosphoric acid, sulphuric acid and glycerol mixture	304	30 to 70°C	0.5 to 1 A/cm2	Not specified	Tai-Chou Lee et al. 43
Phosphoric acid, sulphuric acid and glycerol mixture	304	5 to 80°C	0.5 to 1 A/cm2	Not specified	Chi-chang Hu et al. 91
Phosphoric acid, sulphuric acid and water solution	Fe13Cr	25 to 70°C	0.01 to 0.225 A/cm2	−0.81 to 1.45 V	D. Landolt, et al. 92
Four electrolytes (Mixtures of sulphuric acid, phosphoric acid and glycerol)	316L	50 to 80°C	Not specified	Voltage scanning rate 5 mV/s	Herzog, Grégoire et al. 93
Mixtures of Sulphuric acid, phosphoric acid and water with/without glycerine	316L	50 ± 95°C	0.5 to 2.5 A/cm2	Not specified	Y. F. Chen, et al. 90
Mixtures of sulphuric acid, phosphoric acid, water and glycerol	316L	65 to 70°C	Not specified	2.5 to 10 V	Sasha Omanovic et at. 94

Effect of resistance (Jacquet theory)

Jacquet was the first researcher who investigated the importance of the ohmic resistance of the viscous liquid layer between the anode and the electrolyte during electropolishing. ⁷⁴ The non-uniform thickness of the viscous layer on the material surface leads to different resistance between the cathode and the anode, which causes larger dissolution of the protruded parts versus the depressed part. On this basis, a uniform surface profile can be created. In a later experiment led by Jacquet and others, ⁶¹ they showed that in a plot of current density, J as a function of cell potential (or anode potential), E, a horizontal plateau existed for many solutions used in electropolishing processes. The J-E plot was found that whenever the diffusion of one reactant consumed by an electrode process was the controlling factor. The concentration of the electrode surface reactant is much less than its bulk counterpart, and the effective thickness of hydrodynamic boundary layer has a definite value determined by either forced or natural convection. Particularly, a limiting current density is observed for the electrodeposition of metals at the cathodic side. In a parallel study, Kolthoff and Miller ¹⁰⁹ also observed a limiting current density for the anodic dissolution of mercury if the diffusion of the complexing ions such as S₂O₃ ⁻. SO₃ ⁻ CN⁻ and SCN⁻ was set as the limiting factor.

In a later study, ¹¹⁰ Hickling and Higgins showed that, in the anodic dissolving rate determining step, the current density of a 0.15 mm asperity could be 2.5 times that of a depression. In their study, an inverse relationship of current density and viscosity was also observed with an increased current density with stirring. The earliest descripting on the resistance of the viscous film formed between anode and the electropolishing liquid described by Jacquet is somewhat obscure.

Diffusion phenomena (Elmore theory)

The Jacquet's resistance theory on electropolishing was later challenged by Elmore. ⁶² It has been observed that the factors which might facilitate to destroy the electrolytic layer between the anode and electrolyte led to uneven polishing of the metal surface.^78,111 This case, however, could be altered by modifying the electropolishing conditions (e.g. temperature, agitation, electrolyte viscosity and anode position from being horizontal to vertical). ⁷⁵ Based on these observations, Elmore^62,112 postulated the importance of diffusion in the electropolishing process. He used the copper-in-orthophosphoric as a model system and studied the current–voltage relationship, and eventually attributed the polishing effect to a varying concentration gradient of the dissolved metal ions over the projections and crevices.

The basic assumptions of Elmore theory of electropolishing are: (a) metal ions are dispersed from the anode by diffusion and convection instead of electrolytic migration; (b) the interface of anode and electrolyte is saturated with the products of solution. On this basis, Elmore believes that in the case of copper polished in orthophosphoric acid, the concentration of anodic copper ions in the electrolyte increases with current density, until it reaches a maximum value. Dissolution of the metal continues beyond this point only to the extent that copper ions are allowed to diffuse into the bulk electrolyte from the boundary layer. The value of the Nernst diffusion layer thickness, δ over the projection is smaller than the value over the crevices,^97,113 which results in a larger limiting current, as defined by Equation 6, ¹¹⁴ over the projection part than the crevices. (6) n is the total number of ions involved; F the Faraday constant; D the diffusion coefficient of the rate limiting species; C_s the surface concentration, i.e. saturation concentration of metal ions in the solution and C_b the bulk concentration of the ions.

The larger limiting current density around projection part results in the metal surface being levelled out as electropolishing proceeds. In the Nernst diffusion layer model (Equation 6), a concentration gradient is assumed to be at the interface, which varies linearly at the immediate vicinity of the metal surface and then has a curvature until it merges with the final constant electrolyte bulk concentration of the species. ⁹⁶ It is a mathematical construct that artificially separates the electrolyte region near the anode surface from rest of the electrolyte. This has become the generally accepted mechanism for electropolishing and the mass transport control has been established as the reason behind electropolishing.

Elmore further concluded that the ratio could be defined as (7) where i_o is the applied current, t_o is the transition time to build the polishing conditions, C_m is the solubility of metal in the electrolyte, A is the surface area of the anode, F is the Faraday constant and D is the diffusion coefficient of dissolved metal (constant). It is worth noting that, the ratio in Equation (7) was developed based upon Smuluchowski's equation ¹¹⁵ correlating the motion of charged species to the surrounding media viscosity. And it does not consider the diffusion rate decrease resulted from the viscosity increase. Thus, Elmore's theory has its own limitation, considering the viscosity is capable of answering: (a) why many polishing baths are viscous liquids; (b) why the character of polishing increases with viscosity ⁷⁸ ; and (c) why dilution or increase in temperature hinder the polishing of copper in orthophosphoric acid. ³²

Later, Williams and Barrett ²⁰ offered the experimental details to confirm Elmore's theory.

Focused on the film formed on copper during electropolishing, they showed that the film is composed of a phosphate of copper characterised using electron diffraction techniques. This facilitated to confirm the basic assumption in Elmore's theory that a solubility limit of copper existed in the boundary layer of the electrolyte. A further study led by Hoar and Farthing ¹¹⁶ showed that, since the copper phosphate film is soluble in phosphoric acid even in the absence of applied potential, the solubility limit could be attributed to the a decrease in hydrogen ion concentration at the anodic surface during polishing process.

Acceptor theory – distribution of anions

While Elmore and others proposed that the diffusion of metal ions determined the dissolution of the metallic anode to the liquid electropolishing agent, Halfway suggested that, instead of ionic diffusion, the distribution of the anions along the anodic surface controlled the dissolution of the metal in electropolishing. ¹¹⁷ This point of view was considerably developed by Edwards ³⁹ both experimentally and theoretically. Using a composite copper anode, which had projections and recesses on the surface, in phosphoric acid solutions, Edwards found that the current distribution on the projections and recesses endured very little change when the total current fell and polishing conditions were established. Thus, a quantitative investigation of the smoothing efficiency under various conditions on the surface of standard roughness were performed. The resulted efficiencies in Edward's experiment was compared with those calculated from a wholly diffusion-controlled electropolishing process. The conclusion drawn in that study was that the smoothing mode in electrolytic polishing was not specific to processes, but caused only by the variations in concentration gradient through the diffusion layer on anode surface. ³⁹

More specifically, the basic assumption in Edward's theory is that the polishing effect was determined by the depletion of the anolyte layer of ‘acceptors’. The term ‘acceptor’ means the anions and molecules (e.g. water molecules) which has the potential to combine or complex with the anodic metal. In this framework, the basic electropolishing mechanism is one of the migration of ‘acceptors’ to the anode instead of the diffusion of metallic ions in the opposite direction. Accordingly, the limiting factor of the current density is related to a maximum concentration gradient when almost all the acceptors approach the anode and react with the surface metallic ions on the anode. Due to the continuous consumption of the acceptors on the anode surface, the acceptor concentration is much lower around the electrode than that in the bulk solution.

Later, Carl Wagner ²⁶ used the acceptor theory to mathematically analyse an ideal electropolishing process. He assumed that the electrode surface could be best described by a sing wave profile, as described below (8) where a is the wavelength, b is the amplitude, y is the distance from the ‘average surface plane’ of the anode and x is the coordinate parallel to the average surface plane. The ionic concentration profile versus y is schematically illustrated in Fig. 8. This profile, based upon Fick's second law, was eventually derived as ²⁶ (9) OPEN IN VIEWER

Here the is the average value of the concentration gradient at the anode. A linear dependence of the concentration on the distance, y could only be found at the inner most part of the discussion layer.

Taking into consideration of the change in the shape of the surface profile as a function of time, the sine wave amplitude at time t could be derived as (10) where b_o is the amplitude of the sine wave surface profile at time zero, and b is the surface profile at time, t; u is the corresponding displacement of the average surface plane, and could be further derived as (11)

Thus, the loss of metal per unit area, Δm/A could be calculated as (12) where ρ is the density of the metal.

Using a 2-dimmensional Laplace equation inside the electrolyte, the local potential gradient could be written as (13)

The local potential gradient, is linearly related to the local current density by Ohm's law. Through the set of formulae (from Equation 9 to 13), Carl Wagner mathematically solved the profiles for the decrease of surface roughness as a function of the recess of the average surface, amount of metal dissolved per unit surface area, and product of current density and time. It should be noted that, the development of the larger surface roughness profile (i.e. macroprofile) from Wagner's formula agrees well with Edward's theory. Also in his analysis, Wagner rejected Elmore's theory regarding the inconsistence of the metal ion concentration at the constant current density at wide variation of anodic potential.

Passivation theory (Hoar and Farthing theory)

Hoar and others^98,99 suggested that the thin layer formed between the electrode and the electrolyte during electropolishing was featured as a passivation film.

A metal is considered to be in a passive state when the metal ions cease to dissolve to the surrounding electrolyte from the metal surface, even if a more positive potential than its oxidation potential is applied to the electrode. Usually, a fast increase in anodic potential with a sharp drop in current density at the onset of the electrochemical passivity can be observed. ³²

The electrochemical passivation has been known since the Faraday age. Early interpretations of this phenomena suggested that the metal surface was oxidised to stop the metal from continuous dissolving from the anodic side. In the case of the oxide films on some metals,^118,119 it has been shown that the thickness was the order of a few Angstroms, and their growth was enhanced by oxidising agents while diminished or prevented by halides or other reducing agents.

On the basis of the similarity between the current–voltage curves of the electropolishing and those in many passivation reactions, researchers postulated the formation of the transient passivation film, either salt or oxide, to partially explain the phenomena in electropolishing. ³² The analogy between the oxidation of an easily oxidised metal, such as aluminium or zinc, and the electropolishing has been illustrated by Napier and others. ¹²⁰ They showed that the aluminium could be polished by placing it in an anodic oxidation bath, or a bath which dissolved the alumina film. A similar study on the formation of the passive layer on iron was performed by Tegart and others. ¹²¹ They showed that the anodic processes involved alternate passivation for the polishing of iron in sulphuric acid at around 60–70°C and 4.5 volts.

Hoar and Mowat ⁹⁹ made nickel anodic in a 50/50 sulphuric acid/water bath. They found that there was a short interval of time when a slow rise of anodic single potential showed up, followed by a large rapid one at its end right before polishing happened. They compared the phenomena with those established by Muller for the onset of the passivity of nickel, iron and other metals in less concentrated sulphuric acid, ¹²² and postulated that the rapid rise of potential at the onset of electropolishing was due to the similar formation of a compact oxide film on the metal surface to the anodic passivation. However, this film dissolved as fast as it was formed anodically from the metal. ⁹⁹

Direct experimental observations on the existence of the oxide film at the onset of electropolishing was given by Hoar and Farthing. ¹¹⁶ An electropolishing system composed of copper and α-brass anodes, the orthophoric acid/water bath (50/50 by volume) and a cathode with holes was used in their study. Mercury was dropped onto the copper or α-brass anodes through the holes on cathode, and wetted the anodic surface immediately, and then spread over it, if the passive film had not formed on the anodic surface. By telling the different wettability of the mercury droplet on the anodic surface, they showed whether or not that anodic surface was covered by the passivation film.

Hoar's theory could also quantitatively explain the ‘smoothing’ effect on preferable dissolution of asperities on anode during the electropolishing process. ¹¹⁶ The transient mobile layer of saturated solution and solid formed on the anodic surface before the ‘passive’ film had been produced, and dissipated afterwards. This layer would be both thicker and less disturbed in the depression than the asperities. Thus, one may expect conditions for passivity to be reached sooner in the depression part than on the asperities, and the resulted great increase in anodic polarisation in the depression would divert most current to the asperities. This statement, as drawn by Hoar and Fathering, was also in agreement with Evan's theory. ¹¹⁸

Interestingly, a later study led by Rowland ¹²³ indicated that the observations in Hoar's experiment could not be confirmed. In fact, a copper phosphate film was detected by electron diffraction afterwards. ²⁰ Therefore, the exact nature of the film formed on the copper still remained open to question then.

The periodic oscillation of the current–voltage curve before electropolishing conditions were established, as observed by Hoar and Mowat ⁹⁹ was attributed to an alternate activation–passivation phenomena. Using silver anode in cyanide plating bath,^124,125 Francis and colner recorded the periodic oscillations and observed the formation and dissolution of the film on silver surface. It should be noted that the theory of alternate activation–passivation is actually inconsistent with Edward's acceptor theory. The viscous anodic layer, serving to screen the acceptors, was still basic and may determine the nature of the film observed in Hoar's study.

Ionic adsorption (Darmois theory)

Rowland found that, the noble metals (e.g. gold, platinum and palladium) could be anodically polished in the molten salt, such potassium chloride or sodium chloride as electrolyte under certain circumstance. ¹²³ He considered this phenomenon not explainable by postulating the anodic layer of oxide, hydroxide, etc. on the anode, ⁹⁹ because the substances corresponding to oxides would be chlorides of the metals. In this case, the possibility of oxide films due to dissolved oxygen was ruled out by the experiments in hydrogen and by the experiments in which polishing could occur at temperatures well above those at which oxide could survive. He also made the comment on Hoar and Farthing study, ¹¹⁶ that the non-adhesive of mercury droplet to the copper anode was not due to the formation of the anodic oxide film, but to the failure to establish a contact-adsorbed film of mercury at higher current density than required by electropolishing.

Later, Darmois and others ¹²⁶ presented a theory of polishing which involves the adsorption on the anode of the anions from the electrolyte. As a specific example, ¹²⁷ in the case of aceto-perchloric baths, the anode was proposed to become covered with a thin layer of ClO₄ ⁻ ions that were capable of establishing an electrostatic field intense enough to screen the metal ions. The authors also eliminated any mechanism which involved the formation of an oxide film.

Optical microscope

In order to characterise and evaluate the surface morphology before/after the electropolishing process, many microscope techniques have been utilised. Optical microscopes are the oldest design among its category, and has been intensively used to observe the electropolishing surface finishes. For example, in some of the Jacquet's pioneer electropolishing work,^61,76 optical microscope was employed to show the surface brightness and crystal grains of the copper anode (using polarised light source) after electropolishing process.

In a study on the mechanism of electropolishing, Edwards made a copper composite anode with projections and recesses on the surface which were isolated with each other. By using a binocular optical microscope, he could observe the smoothing effect preferably on the projections during electropolishing process, thus being able to quantitatively evaluate the electropolishing smooth efficiency. ³⁹

The optical microscope is also an important tool to estimate the sample quality used for more advanced electron microscope. For instance, R. D. Schoone et al. develop an electropolishing unit for preparing the thin metal specimens for transmission electron microscope (TEM). The electropolishing finishes were evaluated using optical microscope. It was determined in their study that, under certain electropolishing conditions, the dislocations and stacking fault in pure cobalt existed, whereas fine metallic grain structure was observed for Fe–Ni alloy. ¹³³

Despite the intensive employment of optical microscope, its resolution is limited by the interference of the visible light, which is about one half of the minimum visible light wavelength (200 nm). ¹³⁴ Thus, microsmoothing surface finishes after electropolishing process cannot be effectively evaluated by solely the optical microscope. Thus, other techniques should be resorted to if one needs to perform characterisation on the microelectropolishing finishes below 200 nm.

Electron microscope

Restricted to light diffraction issue, traditional optical microscope could not be used to evaluate the surface feature smaller than the light diffraction limit. However, if the light source is changed by something with much higher energy, the irradiation wavelength should be decreased accordingly. Based on this point, electron microscope became to play an important role in detecting structures much smaller than light diffraction limit. ¹³⁴ With improved beam energy, the wavelength can be further brought down to sub-nanometre range. For instance, an electron microscope equipped with a 5 KeV electron source is capable to produce the electron irradiation of 0.248 nm. Atomic-resolution image could even be generated with a sub-50 pm electro probe. ¹³⁵ Two types of electron microscopes are often used in electropolishing finishes evaluation. The first one is TEM, using which the high energy electron beam transmits through a thin specimen that is partially transparent to electron beam. The electron beam carrying the specimen structural information, and the spatial variation of the sample structural information can be viewed by either a fluorescent viewing screen or recorded by a charge-coupled device (CCD) camera. ¹³⁶ The second type of electron microscope is scanning electron microscopy, often abbreviated as SEM. It produces the images by probing the specimen with a focused electron beam which scans across a rectangular area on top of the sample surface, and it has been vastly applied to characterising the nanometre surface morphology.^{137 138 139 –140}

TEM has become a key tool to correlate the nature of the microstructure of metals and the alloys directly with their important physical and chemical properties. The metal specimen for TEM has to be delicately prepared in order to make it thin enough to let the electron transmit through. ¹³³ The thin metal foils are usually of a thickness of a few hundred nanometres or less.^141,142 In addition, the successful TEM analysis highly depends on the quality of the thin metal foils prepared. Although pure mechanical polishing technique exists to make the TEM thin specimen, the sample quality usually is hard to be fully controlled.^143,144 Intensive studies have demonstrated that an electropolishing step is necessary for TEM sample preparation.^{133,143,145,146} Normally, the preparation of the thin TEM specimen is composed of three steps, (a) obtaining a sample piece less than 2 mm thick, (b) mechanically thinning the sample down to about 0.2 mm and (c) further decreasing the sample thickness down to nanometre range so that sufficient electron beam penetration can occur. ¹⁴⁷

The key factors of electropolishing, including voltage, current density, temperature, electropolishing duration and convective flow rate need to be carefully optimised in order for one to obtain an ideal TEM sample for reproducible observation. Necip ¹⁴⁸ prepared high quality pure aluminium TEM samples through a double-jet electropolishing technique. Some useful hints to obtain the least failures in preparing the TEM specimen with optimum electropolishing parameters were given in his study.

Besides pure metal, the alloy microstructures are critically important for insight understanding of their mechanical properties. For example, two phase TiAl alloy has a great potential for aerospace applications due to its excellent specific mechanical properties. ¹⁴⁹ Considerable research effort has been dedicated to detailed characterisation of TiAl on the microstructure-related mechanical properties. A. Couret et al. ¹⁵⁰ used a jet electropolishing technique with a polishing solution composed of perchloric acid and methanol to prepare the TEM thin specimen. A quantitative TEM analysis of the lamellar microstructure in the TiAl alloy could be performed due to the delicate electropolishing specimen preparation process. Using a range of advanced methods combined with region specific replication entails and the electropolishing method, S. Sulaiman et al. ¹⁵¹ prepared TEM replica specimens of P91 steel and carried out quantitative microanalysis on the P91 steel specimen.

Despite its versatility in determining the thin specimen microstructure, TEM cannot be used to determine the sample surface morphology. Thus, in order to determine the surface morphology after electropolishing, other techniques should be used. SEM is such a tool, which utilises a high energy electron bean to scan the sample surface. After interaction with the sample, the electrons which carry sample structural information are probed by electron detectors. As such, the sample surface structure can be revealed.

Didier Pribat et al. ¹⁵² used SEM to evaluate the electropolishing effect on copper foil in H₃PO₄ bath at elevated temperature. The surface of the electropolishing finishes was analysed by a JEOL JSM-7600F field emission SEM. The copper grain size on the specimen surface was analysed by the electron backscattering diffraction (EBSD) images. They found that the copper foil electropolished at a small window near 65°C in 2.17 M H₃PO₄ bath, exhibited a larger grain size than samples obtained under other conditions. In a similar study, T.M. Dahy et al. ¹⁵³ removed the tarnishing and roughness of copper surface with electropolishing treatment. The copper foil surface contamination and morphology were analysed by an SEM equipped with an energy-dispersive X-ray (EDX) spectroscopy.

In order to evaluate the corrosion behaviour of the 304 stainless steel at high temperature, hydrogenated water, Paul C. Sander et al. ¹⁵⁴ used SEM to observe the surface nanocrystallinity of the stainless steel oxide layer. They found that the stainless steel surface had small size (ranging between 10 and 26 nm) chromite-based crystals relative to the machined (i.e. cold worked) process. The electropolishing process, which removed surface macrostrain by cold working, as stated by Paul C. Sander et al., exhibited a preferred reference surface condition. In another study on AISI-314 stainless steel electropolishing, Sonia R. Biaggio et al. ⁸³ employed the SEM at magnification of 1000× to determine surface structure of the electropolished samples. The surface of the samples, which were classified as opaque, semi-bright and bright exhibited different surface morphology according to the corresponding SEM sample images.

Atomic force microscope

Despite the capability of determining small morphological features, SEM only provides two-dimension (2D) information of the surface finishes. In order to get through knowledge on the height profile, or roughness of the workpiece after electropolishing, AFM is often used. Contrary to optical microscope and SEM, AFM does not use optical/electrical lenses or light/electron irritation. Instead, the surface morphology information of the sample is gathered by AFM tip (with the diameter in the order of 10 nm). By scanning on the sample surface, the mechanical interaction between the tip and the sample is converted to electric signal through a piezoelectric element, thus the sample topographical information is recorded. The resolution of the AFM is not restricted by light diffraction limit, and can reach the order of 0.1 nm. Therefore, it becomes an important tool to study the microsmoothing of the workpiece after electropolishing.^43,155

One of many important applications of AFM in evaluating the electropolishing metal surface is to directly offer the information of the sample surface roughness, which cannot be fulfilled by electron microscopes. For example, Chi-Chang Hu et al. ⁹¹ studied the factors that affect the surface roughness (Ra) of 304 stainless steel (304 SS) electropolished in a mixture of phosphoric acid, sulphuric acid and glycerol. The factor Ra was characterised by using AFM height profile on the electropolished 304 SS surface. They found that the bath temperature and polishing time were two key factors that affect the 304SS Ra. In a sister study of the same group, they found that the surface roughness of 304 SS could be tuned from 7 to 45 nm by controlling electropolishing variables in glycerol-containing baths. The surface roughness, again was determined by the AFM height profile. ⁴³ T.M. Dahy et al. ¹⁵³ adjusted current density, duration, temperature and viscosity in electropolishing process on commercialised copper foil, in order to remove the tarnishing and roughness. The surface roughness of the electropolished copper foil was characterised by using AFM height images. Jerzy Zak et al. ¹⁵⁶ studied the electropolishing conditions on NiTi alloy. The quantity of the electropolishing ally surface was evaluated by AFM.

In addition to the electropolishing surface roughness evaluation, AFM is also an important tool to characterise the surface morphology of the electropolished metal surface. For instance, R.M. Metzger et al. ¹⁵⁷ created highly ordered nanotopographies on aluminium single crystals by electropolishing method. The ordered nanoscale morphology on the single Al crystals was studied in details by AFM. Muhammad Aslam et al. ¹⁵⁸ use the porous anodic alumina pretreated by electropolishing process. Under different electropolishing voltage, they were able to create 100–150 nm random features (4–5 nm high) and 50 nm ordered nano stripes (1–2 nm high), which was confirmed by the AFM images.

Finish

Electropolishing is a reverse plating procedure that entails the electrochemical removal of metal impurities (including carbon, silica and free iron) from a stainless steel surface. ⁴⁰ The goal is a smooth surface, devoid of burrs or crevices that attract and trap contaminants. During electropolishing, the workpiece is fitted with electrodes, immersed in an electrolyte solution and subjected to a direct electrical current. During this electrolytic process, the metallic surface of the anodic part is preferentially dissolved ion by ion, yielding a smooth surface. ¹ Optimal results depend on careful control over the current density, the precise chemical composition of the electrolytic solution, the temperature and agitation of the bath and the duration of current exposure. ⁹ Finally, selected surfaces are mechanically buffed to a smooth finish.

Corrosion resistance

Electropolishing is widely specified to enhance corrosion resistance on a wide variety of metal alloys. ³⁶ Although most commonly used on stainless steels, electropolishing offers corrosion resistance on other alloys as well. ¹⁵⁹ Due to the versatility and superior effectiveness, electropolishing is fast becoming a replacement process for passivation, a chemical process that has been used for years to help restore contaminated stainless steel to original corrosion specifications. The intent of passivation is to remove free iron or other foreign matter from the surface of the metal and to create a chromium rich surface that is resistant to corrosion. However, passivation is generally not effective in removing imbedded free iron and contaminants and will not remove heat tint or oxide scale on stainless steel. ⁸

The electropolishing process dissolves the outer skin of metal, removing deeply imbedded contamination. ¹⁶⁰ Heavily contaminated surfaces such as machined parts, welded or brazed assemblies or other components that typically respond poorly to passivation alone, are good candidates for electropolishing. ¹⁶¹ Unlike passivation, all stainless alloys including the 400 series and precipitating grades can be processed without distortion, flash attack or hydrogen embrittlement. Just as electropolishing is used to enhance corrosion resistance on stainless steel, it offers corrosion resistance on other alloys as well. Electropolishing can be used to delay or retard the corrosion properties of copper, brass, aluminium and carbon steels. On these and other alloys, the removal of surface skin and impurities enhances the corrosion resistant properties of the component. ¹⁶²

The following figure shows type 303 stainless steel (Fig. 9). After a series of machining operations, the parts needed to be decontaminated to remove imbedded steel and other impurities. OPEN IN VIEWER

Under 40× magnification using the SEM, the passivated part is actually rougher, due to the chemical attack of the 303 stainless steel by normal passivation. In contrast, the electropolished part is smooth and clean. By carefully monitoring the amount and rate of metal removal, electropolishing dissolves the surface skin and its impurities, including impinged steel flecks and other contaminants.

Surface improvements

The most noticeable benefit of electropolishing is that it provides an immediate brightening of a metal finish. ¹ By eliminating or minimising stains, heat discolouration and minor scratches, electropolishing can bring out the metal's natural shine, which enhances the appearance of the piece. In addition, electropolishing offers remarkable ultracleaning capabilities through the removal of rust, imbedded scale and other foreign debris. ¹⁶³ This can be accomplished without stressing the piece or negatively affecting surface hardness and integrity. By removing surface imperfections, electropolishing can increase the piece's resistance to corrosion. That is because the elimination of imperfections can prevent the moisture that leads to corrosion from accumulation. In fact, some salt spray tests have indicated that electropolished pieces experience greater corrosion resistance than passivated and raw parts. Electropolishing can also effectively remove the microscopic peaks resulting from processes such as welding, forming and stamping. ¹⁶⁴ The reduction of microfinish values and accompanying increases in surface smoothness can be extremely beneficial in the production of valves and gears. ¹⁶⁵ The ability of electropolishing to remove surface metal from a part makes it extremely effective in the elimination of edge breaks. That is why electropolishing is often preferable to vibratory finishing or tumbling for use on delicate parts. Because electropolishing creates a smooth, clean surface that promotes adhesion, it is sometimes used to prepare surfaces for processes such as e-coating, plating, welding and anodising. ¹⁶⁶

Conventional mechanical finishing systems tend to smear, bend, stress and even fracture the crystalline metal surface to achieve smoothness or lustre. Electropolishing offers the advantages of removing metal from the surface producing a unidirectional pattern that is both stress- and occlusion-free, microscopically smooth and often highly reflective. Additionally, improved corrosion resistance and passivity are achieved on many ferrous and nonferrous alloys. The process micro- and macro-polishes the metal part and micro-polishing accounts for the brightness and macro-polishing accounts for the smoothness of the metal part. ¹⁶⁷

Deburring is accomplished quickly because of the higher current density on the burr, and because oxygen shields the valleys, enabling the constant exposure of the tip of the burr. Because the metal part is bathed in oxygen, there is no hydrogen embrittlement to the part. ¹⁶⁸ In fact, electropolishing is like a stress-relieve anneal. It will remove hydrogen from the surface. ¹⁶⁹ This is important to parts placed under torque. Another benefit is that bacteria cannot successfully multiply on a surface devoid of hydrogen, therefore, electropolishing is ideal for medical, pharmaceutical, semiconductor and food-processing equipment and parts. ¹⁷⁰ The combination of no directional lines due to mechanical finishing, plus a surface relatively devoid of hydrogen, results in a hygienically clean surface where no bacteria or dirt can multiply or accumulate.