A novel mask electrochemical additive and subtractive combined manufacturing technique for microstructures with high machining performance

In this paper, a novel mask electrochemical additive and subtractive combined manufacturing technique was proposed. This is a machining method at the atomic level, and it can be used to produce metal microstructures with high-profile accuracy and low surface roughness. Due to the accumulation of electric field lines during mask electrochemical deposition, the height of the edges of the microcolumns is usually twice or more than the height of the central position in the deposition plane. A combined machining method based on the electric-field constraint of the mask is thus proposed to improve the accuracy of the profile and its surface roughness. The feasibility of the proposed method was verified by both simulations and experiments. The height difference between the column center and the surrounding layer on the surface of nickel microcolumns was reduced from 13 to 2 µm, and the roughness of the tops of the microcolumns was also improved. Experiments to examine electrolysis leveling were carried out to verify the correctness of the results of the simulations and theoretical calculations. Finally, the parameters were optimized using orthogonal experiments, and an array of nickel microcolumns with a diameter of 200 µm and a height of nearly 50 µm was obtained using these optimal parameters. The profile accuracy and surface roughness of the high-precision microcolumn array were improved by using the mask electrochemical additive and subtractive combined machining technique, and a high-precision microcolumn array structure was manufactured.


Introduction
Microscopic metal structures with complex shapes are widely used in diverse fields including microelectronics, aerospace [1], and precision mold [2]. Such parts are often characterized by their small size, large number of feature structures, dense arrangement, and the requirement for high-accuracy machining, and this poses great challenges to processing and manufacturing technology [3]. There are currently a variety of processing methods for such metal microstructures [4]. Mask electrochemical deposition has been verified as a good technique in the field of fine manufacturing due to its high localization, low stress, lack of thermal deformation, and absence of cracks in the deposition layer [5,6]. Mask electrochemical deposition processing technology uses a hollow insulating template to constrain the electric field in the processing area so as to realize the fabrication of metal microstructures with high precision [7]. Since this processing technology relies on metal-ion reduction in an electrochemical reaction on the cathode surface, a micro-or nanomachining scale can be achieved. The design of the hollow parts of the mask is flexible, and electrochemical additive processing can achieve the processing goal of creating complex three-dimensional metal microstructures [8].
Many researchers have carried out a series of studies on electrochemical deposition machining improvement [9][10][11]. Zhao proposed a hybrid electrochemical technique combining electrolysis and electroplating to manufacture a micro nozzle with a specific shape and hydrophilic. Firstly, an initial nozzle is electrolytically machined by a microelectrode with side wall insulation, then the direction of the electric field is reversed and the microelectrode is used as an auxiliary anode for electroplating [12]. This hybrid electrochemical technique uses electrochemical deposition to improve the defects of electrochemical dissolution machining and form a specific nozzle shape. The method introduced in this paper is different from the above hybrid electrochemical technique. It takes advantage of the high surface quality of the electrochemical dissolution method and uses the electrochemical dissolution method as the posttreatment method of electrochemical deposition to improve the defects caused by the mask electrochemical deposition. Li studied the effect of pulse (PC) and pulse-reverse current (PRC) on the microstructure and nanomechanical responses of the individual micro-pillars of electrodeposited nickels. It was found that the PC produced a heterogeneous microstructure consisted of coarser columnar grains and a multitude of nanocrystals, whereas PRC resulted in a homogeneous nano-grain distribution due to disrupted grain growth and facilitated nucleation. The nickel micro-pillar with uniform nanocrystalline texture also exhibited enhanced hardness and plasticity [13]. Zhu attempted to understand the filling mechanism by PPR plating and then to explore a potential solution for void-free filling in easy electrolytes. The result implied that at low current density the reverse pulse played a strong suppression effect that contributed to the "V" shape growth. A competitive growth model between the bottom reversed "V" structure and the upper "V" structure was proposed to explain the void-free filling mechanism in PPR plating process [14]. West presented a theoretical model based on a one-dimensional approximation for the filling of high-aspect ratio trenches and vias with copper by pulse reverse electrodeposition. The results shown that the offtime should be on the order of the diffusion-time constant and that the deposition time should be smaller than the the diffusion time constant. Reverse current is shown to play an important role in diminishing void size [15]. Huang investigated the plating through hole (PTH) via direct current (DC) and pulse-reverse (PR) electroplating Cu. The dependence of the Cu microstructure on the PR frequency and the underlying mechanism of the crack formation in Cu have been discussed [16]. Tschulik presents a new technique for electrodeposition of separated three-dimensional metallic structures solely by the superposition of magnetic gradient fields. Separate columns and stripes of Cu are generated by pulse reverse plating controlled by tailored magnetic field gradients [17]. Narayanasamy using pulsed reverse current technique by varying its duty cycle and frequency at the fixed concentration, the Ni-MoSe2 nanocomposite coating was successfully prepared on mild steel specimen. These results revealed that the incorporation of MoSe2 in the nickel matrix has improved mechanical and corrosion resistance properties [18]. In this report, the combined method is divided into two steps, one is electrochemical deposition and another is electrolysis. When the electrochemical deposition is all completed, use different working fluids for one-step electrolytic processing. The method is to use the advantages of electrochemical dissolution as a post-treatment method of electrochemical deposition to directly remove lots of materials to improve the processing effect of electrochemical deposition. Electrodeposition followed by electrochemical machining (achieved by a reverse polarity) is obviously different from the combined method in this paper in terms of processing way and effect. In electrodeposition followed by electrochemical machining (achieved by a reverse polarity), reverse current serves two purposes. One is to use reverse pulse to reduce concentration polarization and improve the effect of electrochemical deposition. The other is to remove a few of material and refining the deposited grains through reverse current. All processes are carried out in the same working fluids, and the reverse pulse is applied in multiple sections during the electrochemical deposition process. Yang adopted the annular-assisted cathode method to solve the problem of the thickness of the electrodeposition at the center of the mask being inconsistent with the thickness of the edge; a decrease in this thickness ratio without affecting the deposition rate was achieved [19]. Mekaru examined the edge effects of electric field lines, and they added a leveling agent to the plating solution to eliminate defects; they obtained smooth electroplating performance in their experiments [20]. Flynn proposed a method combining additive and subtractive processes within a single workstation, harnessing the relative merits of each process [21].
Despite the advantages noted earlier, mask electrodeposition also has some limitations [22]. In the mask electrodeposition process, photographic film is usually used to constrain the electric field, but the electric field lines tend to gather at the edges of the hollow cavity and are distributed unevenly within it [23]. As a result, the metal deposition rate at the edge of the cavities is higher than that in their middle regions, resulting in uneven microstructure thickness machining and a "saddle" shape at the top of the cast layer section [24], and the surface quality of a single electrodeposited metal microstructure is poor.
Mask electrochemical machining is a subtractive material-processing technology that is based on electrochemical anodic dissolution at the atomic level [25,26]. Hollow insulating masks are used to limit the electrochemical dissolving of anodes, thus realizing local profile and surface finishing of microstructures [27]. Therefore, regarding the saddle morphology problem in single-mask electrochemical deposition, the mask electrochemical dissolution method is proposed to repair the contour defects and finish the surface. This aims to remove the saddle contour defects at the top of the cast layer section caused by the uneven distribution of electric field lines during mask electrodeposition and realize the high-precision machining of metal microstructures.
A novel mask electrochemical additive and subtractive combined manufacturing technique (MEASM) is proposed in this paper. This method combines mask electrochemical deposition and electrochemical dissolution into one process. Mask electrochemical deposition is used to form a preliminary microstructure profile, and then electrochemical dissolving is used as a post-processing method to improve the shape accuracy and surface roughness of the metal microstructure. Finally, a high-precision and high-quality metal microcolumn array structure can be manufactured using this MEASM technique.
In this study, a series of experiments were carried out using the MEASM process. Firstly, the principles of electrochemical composite machining were analyzed, and mechanisms for improving the effects of electrochemical deposition machining were explored. Secondly, the results of simulations of mask electrochemical deposition and MEASM were compared to illustrate the improvement provided by the technique in the morphology of the saddle and the surface roughness of the metal microcolumn structure. The feasibility and performance of MEASM were verified by comparisons with pure electrochemical deposition in both theoretical and experiments. Aiming to further improve the machining performance of the technique, machining parameters including the voltage, duty cycle, and frequency were examined using orthogonal experiments, and optimal machining parameters were selected. Finally, a microcolumn array structure with high precision and surface quality across a large area was fabricated by using these optimal machining parameters.

Principle of MEASM
The MEASM approach makes full use of the advantages of both mask electrochemical deposition and mask electrochemical dissolution to realize the fabrication of a metal microarray structure with high localization, high shape accuracy, high surface quality, and high efficiency. The whole machining process includes two stages, as illustrated in

Stage 1: Electrochemical additive manufacturing(EAM)
In this process, mask electrochemical deposition is carried out to achieve the initial formation of the microstructure. The mask deposition process is also an additive manufacturing method. As shown in Fig. 1a, it is necessary to model the shape of the microcolumn by CAD, and then use the model established by CAD to prepare the insulating mask. Mask electrodeposition completes the additive manufacturing of microstructures through layer-by-layer deposition in the mask cavity. Nickel sulfamate was used as the main salt in the electrodeposition process. Under the action of a pulse current, nickel ions are deposited on the cavity structure of the mask surface from the electrodeposition solution. The Fig. 1 Principle of combined electrochemical addition and subtraction machining electrocrystallization process causes nickel atoms to be adsorbed onto the surface of the cathode, and the insulation mask shields the electric field in the non-machining area. The electric field is always limited in the machining area, which greatly improves the localization of the processing and realizes the efficient preparation of the metal microstructure. However, due to the effects of electric-field aggregation, the deposition rate will be higher at the edges of the columns than in their centers. The top surfaces of the metal microcolumns prepared by mask electrochemical deposition thus present a saddle morphology, as shown in Fig. 1b. To obtain microcolumns with the same thickness as the film, their height in pure deposition must exceed that of the film, and then the material above the film must be removed. If the lowest point of the profile (L) does not exceed that of the film after deposition, microcolumns with a height equal to the film thickness cannot be obtained after electrochemical machining, as shown in Fig. 1f.

Stage 2: Electrochemical subtractive manufacturing(ESM)
In this process, mask electrochemical dissolution machining is carried out on the microstructure deposited in the previous step to achieve precision dressing of the profile. The working liquid tank is divided into two separate parts, one part is EAM working fluid, and the other part is ESM working fluid. When the EAM process is completed, EAM working fluid circulation system shut down. Then start ESM working fluid circulation system, working fluid is replaced with a new HCl solution, and in situ ESM is carried out by applying an inverse pulse current between the electrodes. Under the action of the electric field force, the Ni 2+ deposition in the anodic nickel microcolumn is oxidized and dissolved into the electrolyte. By this electrochemical anode dissolution, the anode materials are removed in the form of ions. The electrochemical dissolution rate of the metal microcolumn is different due to the potential difference between the concave and convex parts on the surface of the microcolumn. The saddle morphology caused by electric-field aggregation in the electrochemical deposition step gets removed by the electric-field aggregation in ESM. Thus, high-quality surface finishing and high-profile precision can be quickly achieved, as shown in Fig. 1c.

Simulation of EAM
The two-dimensional electroplating process of nickel microcolumn was simulated by using COMSOL Multiphysics simulation software. The electroplating module in the electrochemical physics field, the deformation geometry module in the mathematical physics field and the third current under charged neutral condition were used in the model. The deformation geometry is used to simulate the growth of cathode boundary during electrodeposition of nickel. It is assumed that the deposition efficiency at the cathode and the dissolution current efficiency at the anode are 100%, and the occurrence of side reactions is not considered. Ignore the natural convection effect.

Governing equation and boundary condition setting
The flux of each ion in the working fluid is calculated by the Nernst-Planck equation: in the formula, N i represents the transfer vector (mol m −2 s −1 ), c i is the electrolyte concentration, z i is the charge number of the ion, u i is the mobility of the charged substance, F is Faraday constant, φ l is the electrolyte potential. The mass conservation equation is: i = 1, 2 for different substances. The electroneutral condition is expressed by the following expression: the boundary conditions of cathode and anode are determined by the Butler-Volmer equation. The electrochemical deposition process can be described in terms of the following simplified equation: the first step is the speed control step, and the second step is assumed to be in equilibrium. Based on this assumption, the local current density related to potential and nickel ion concentration can be described by the following formula: where η represents overpotential and is defined as follows: other insulation boundary conditions:

Simulation basic settings
(1)The geometric model Figure 2 is the Simulation geometry model of EAM, 1 is the anode boundary, 2 is the cathode boundary, 3 is the working fluid boundary, and 4 and 5 are the mask boundary. In the figure, the maximum grid cell size is 1.5E-5 m, and the minimum grid cell size is 3E-8 m.

Analysis of EAM simulation results
The electrolyte current density distribution in the electrochemical deposition machining area was obtained by simulation, as shown in Fig. 3a. It can be seen that the electrolyte current density is larger at the bulge of the deposition part due to the clustering effect of the electric field lines. From these simulation results, the distribution of the current density in the machining area in a pure deposition process was extracted, and this is shown in Fig. 3b. Due to the accumulation of the electric field lines at the edges of the microholes, the current density is larger there. Because the electric field line distribution is sparse in the middle of the holes, there is lower current density in these regions. The current density in positions on the microcolumns close to the edge of the holes is larger than the other positions, causing the microcolumn saddle defect shape to be more obvious. Figure 4a shows the profiles of deposition layers at different times in the electrochemical deposition process. Due to the accumulation of electric field lines at the uplift, the distribution of current density during the processing is uneven. The top profile (K) of the deposition layer presents a saddle shape, and this becomes more obvious with increasing processing time; the height difference of the top eventually reaches 13 µm. Figure 4b shows a comparison between the actual final pure electrochemical deposition profile and the ideal flat electrochemical dissolution profile. The highlighted region above the red dotted line is the saddle defect that needs to be removed. The area above this line is the amount of allowance that needs to be removed by ESM if a smooth profile is desired. Under the action of electrochemical dissolution, the material above the ideal smooth profile should be removed. The saddle morphology at the top of the microcolumns was removed, and their height was kept at 50 µm.

Simulation of ESM
In order to explore the improvement effect of ESM on "saddle" defect formed after EAM, the current (EC) module under AC/DC physical field in COMSOL Multiphysics finite element simulation software. The deformation geometry module model in the deformation grid under mathematical physics field was selected to simulate the process of removing materials from the top of the contour after EAM.

Governing equation and boundary condition setting
It is assumed that the electrolyte is isotropic and that its kinematic viscosity remains unchanged during the electrochemical dissolution process. The effects of bubbles, electrolytic products, and temperature on the conductivity of the working fluid are ignored and assumed to be constant. Ignoring the influence  other boundaries are insulation boundaries: where n represents the normal vector of the boundary and U represents the anode voltage.
The calculation of electrochemically dissolved material consumption is based on the assumption that the loss rate is proportional to the normal current density on the electrode surface. In the deformation geometry, the velocity S perpendicular to the electrode surface grid is set as: Based on the above, the electric field distribution in the electrochemical dissolution machining region can be calculated through electrochemical simulation, and then the current density distribution on the top surface of the anode "saddle" morphology can be obtained, which provides a theoretical basis for the dynamic removal process of "saddle" morphology.

Simulation basic settings
(1) The geometric model Table 2 The geometric profile after EAM as shown in Fig. 5 was used as the electrochemical dissolution simulation model to simulate the saddle structures being electrochemically dissolved and smoothed. 1 is the cathode boundary, 2 is the anode boundary, 3 is the working fluid boundary, and 4 and 5 are the mask boundary. The maximum grid cell size is 1.36E-4 m, and the minimum grid cell size is 6.11E-7.

Analysis of ESM simulation results
The electric field line distribution in the electrochemical dissolution machining area was obtained by simulation, and this is shown in Fig. 6a. It can be seen from the overall diagram and the local enlarged region that the current density is larger at the top of the convex contour than in other areas. The current density distribution in the ESM area will be affected differently at different times due to the effects of electric-field aggregation.
The current density distribution of the machining area at the initial time of electrochemical dissolution in was obtained by simulation, and this is shown in Fig. 6b. As the electric field lines gather at the edge of the microcolumn, the current density at the top point of the profile (K) is large, while the current density in the middle of the microcolumn is small due to the sparse distribution of the electric field lines. According to the current density distribution of the machining area at the initial and subsequent times, it can be concluded that during the electrochemical dissolution  Fig. 6 Simulation results of electrochemical dissolution: a current density distribution in the initial state of ESM and b distribution of current density at different ESM times process, the removal rate of the material at the edge bulge is higher than that in the middle area. It can thus be concluded that the electrochemical dissolution can achieve the effect of leveling the saddle defect morphology.
The simulation results show that the height difference of the top profile decreases to about 3 µm after 7.4 s. It can be seen that the concentration of electric field lines at the bulge leads to an uneven distribution of current density at different positions during the process. With increasing electrochemical dissolution processing time, the top point of the profile (T) of the saddle morphology is gradually removed and finally reaches a flat state. Figure 7 shows a flat microstructure with a highest point of 53.7 µm and a lowest point of 50.75 µm after ESM, which is consistent with the expected hypothesis. The problem of saddle morphology after deposition has thus been verified and solved by simulation.

Experimental procedures
A scanning electron microscope (JSM-IT300) was used to examine the top morphology of the microcolumns, and a confocal laser microscope (LEXT OLS400) was used to measure the profile of the top of the microcolumns. To obtain the required initial workpiece, the film was coated onto a copper substrate followed by photoresist exposure and development.
In this study, MEASM was used to process a metal microcolumn array. The machining method was divided into two steps, and two separate working fluids were used for each: the first was electrochemical deposition processing working fluid, and the second was electrochemical dissolution processing working fluid. A nickel sulfamate working solution was selected for the electrochemical deposition experiments, and a new HCl solution was used in the electrochemical dissolution experiments. The composition and contents of the working solutions are shown in Table 3, and the experimental parameters of the MEASM process are shown in Table 4.

Performance verification of MEASM
Experiments were conducted using mask electrochemical deposition and MEASM to obtain microcolumns from the two different processing methods. Figures 8a and b indicate that compared to pure mask electrochemical deposition, MEASM can be used to produce microcolumns with smooth   It can be seen from the height-difference bar chart in Fig. 9 that the height difference at the top of the Comparison of surface roughnesses and height differences produced by the different methods microcolumns produced by MEASM is reduced from 13 to 4.48 µm when compared with those produced by single electrochemical deposition. The highest height of the microcolumn processed by the MEASM is 52.465 µm. With overelectrochemical machining, the saddle morphology problem is solved, but the height difference of the top profile is increased again. It can be seen from the histogram that the surface roughness of the top of the microcolumns is greatly improved by MEASM and the saddle morphology is also removed. The surface roughness is reduced from Sa 1.726 to 0.441 µm.
The three groups of comparative experiments confirmed that MEASM is effective for profile leveling, and the parameter selection in the ESM step has a significant impact on the profile and the leveling effect of the tops of the microcolumns. Therefore, further optimization of electrochemical leveling parameters is needed to achieve optimal machining.

Performance optimization of MEASM parameters
To obtain a high-precision profile and microcolumn heights consistent with the film thickness, the ESM machining allowance obtained by simulation and theoretical calculation was used. By optimizing the parameters of the ESM, the optimal parameters to match the single electrochemical deposition were obtained. Thus, the optimized additive and subtractive combined manufacturing were formed together.
To further improve the machining performance of MEASM, four key parameters-machining voltage, machining time, working-fluid concentration, and duty cycle-were optimized using orthogonal experiments. The experimental parameters and levels used are shown in Table 5. In these experiments, the top diameter, the top-profile flatness, and the height of the microcolumn structure were taken as the optimization objectives. The optimal combination of the four key factors was finally obtained through comprehensive evaluation and analysis.
The variation trends in the microcolumn height, topheight difference, and top diameter of the microcolumns under different processing voltages are shown in Fig. 10a.
It can be seen that the height of the microcolumns and the top diameter decrease gradually with increasing processing voltage, and the top-height difference first decreases and then increases with increasing processing voltage. This is because at low voltages, the anode is in the passivation region, the metal-dissolution rate is low, and the leveling effect is weak. When the voltage is increased, the height and top-diameter of the microcolumns decrease, the anode is in the over-passivation region, and the material-removal rate increases. Based on the above trend chart and range chart analysis, A3 is recommended.
The variation trends in the microcolumn height, topheight difference, and top diameter of the microcolumns after different processing times are shown in Fig. 10b. It can be seen that the heights of the microcolumns and their top diameters decrease gradually with increasing processing time, while the top-height difference first decreases and then increases with increasing processing time. This is because the removal of electrolytic material increases with increasing processing time. In this experiment, a diameter of 200 µm, a height of 50 µm, and a tip flatness as low as possible were sought, so level B2 was considered to be appropriate.
The variation trends in the microcolumn height, topheight difference, and top diameter of the microcolumns with different working-fluid concentrations are shown in Fig. 10c. As can be seen the height of the microcolumns, their top diameter, and the difference in the top height all first decreased and then increased with increasing workingfluid concentration. This may be because increasing the electrolyte concentration leads to an increase in the electrolysis reaction rate, and reaction products are difficult to exclude, affecting material removal. Therefore, level C3 could be selected as being appropriate.
The variation trends in the microcolumn height, topheight difference, and top diameter of the microcolumns with different duty cycles are shown in Fig. 10d. It can be seen that the height of the microcolumns and their top diameter gradually decrease with the increasing duty cycle, and the top-height difference first decreases and then increases with the duty cycle. This is because with increasing duty cycle, there is a longer total pulse width per unit time and a greater average current, meaning more material is removed. Level D4 was duly adopted.
To summarize, from these experiments, the optimal combination can be obtained as A3B2C3D4; this represents a processing voltage of 12 V, a processing time of 9 s, a working-fluid concentration of 2 mol L −1 , and a duty cycle of 60%. These optimal parameters were used to process a metal microcolumn array structure, and the SEM images in Fig. 11 show sections of the 100 × 100 microcolumn array that was prepared. In the combined machining, the single electrochemical deposition time was 5400 s, the diameter of the microcolumns was 200 µm, their height was 50 ± 2 µm, and their top-height difference was 2 µm. The MEASM approach can thus be used for high-efficiency, high-precision, highsurface-quality, and high-consistency machining of largearea microstructural arrays.

Conclusions
In this paper, a high-precision metal microcolumn array with microcolumn diameters of 200 µm and height of 50 µm was fabricated by MEASM. The causes of saddle defects in mask electrodeposition and the electrochemical mechanisms of MEASM have been described. The formation and removal of the saddle morphology were verified by both theory and experiment. Finally, experiments contrasting MEASM with simple deposition were carried out, and orthogonal experiments were conducted to obtain optimal parameters. The following summary conclusions can be drawn.
1) The accumulation of electric field lines in the process of mask electrochemical deposition will lead to the uneven distribution of current density, resulting in the saddleshaped morphology defect of the microstructure. By taking advantage of the non-uniformity of electric field line distribution, a mask electrochemical dissolution method is proposed to smooth the saddle morphology. Through finite element simulation and concrete experiment, the feasibility of this method is verified.  be reduced from 13 to 4.48 µm by MEASM and that the influence of the ESM parameters on the profile change is great. Orthogonal experiments were conducted to optimize the ESM parameters. Finally, a machining voltage of 12 V, a machining time of 9 s, a working-fluid concentration of 2 mol L −1 , and a duty cycle of 60% were obtained. An optimal array with a diameter of 200 µm and a height of nearly 50 µm and a top-height difference of 2 µm was obtained after machining using these parameters.