1 Introduction

Lithium-ion batteries (LIBs) are widely used in portable, vehicular, backup power supplies, and other fields because of their advantages, such as high specific capacity, good cycle stability, slow self-discharge rate, and long life span. In recent years, with the development of new materials and new craft, LIBs have made great progress, becoming the most mature and popular secondary battery for commercialization. The Nobel Prize in Chemistry in 2019 was awarded to John B. Goodenough, the pioneer of LIBs [1]. However, the rapid development of high-energy-consuming components and high-power devices has increased the demand for more advanced batteries with higher performances in capacity, life span, endurance, and safety. Aside from the development of new material systems for electrodes, researchers have paid increasing attention to the optimization of structural design and manufacturing processes of battery components, aiming to open new breakthroughs for improving battery performance.

The current collector is an important part of a lithium-ion battery. It is not only used as the carrier of active electrode materials but also for electricity collection and conduction. Its surface structure affects battery performance indexes directly, including conductivity, internal impedance, charge–discharge specific capacity, and cycle life [2,3,4]. As is well known, most negative current collectors are made of deferred or electrolytic copper foil with a very smooth and simple-scale surface. This condition is likely to form a weak interface between the current collector and active materials, resulting in poor adhesion and uneven contact between two phases, as well as a lower conductivity. After numerous charge–discharge cycles, the active material tends to be pulverized and destructed, causing a sharp drop in battery performance [5]. What is worse is that this condition may even trigger a security risk. In addition, the traditional current collector with a smooth surface can load only a limited amount of active materials, causing a great loss of battery capacity. In this context, developing new methods to fabricate functional structures on the surface of a current collector is essential to enhance the contact status and bonding strength between the electrode material and current collector and finally improve the cell performance.

Many studies have been conducted on the preparation of negative current collectors with different surface structures. A variety of methods have been reported, including the in-situ growth method [6,7,8] and preparation of carbon nanotube film by electrophoresis deposition [9]. With regard to the three-dimensional (3D) structured current collector, Park et al. [10] prepared a 3D porous Sn-Cu composited electrode by using quick freezing and electroless plating techniques. Kim et al. [11] created an amorphous silicon electrode integrated with a copper-nanocolumn current collector. The above work shows that preparing a structured current collector with a complex surface is an effective way to enhance the performance of LIBs. Inspired by this strategy, people have used electrodeposition [12, 13], etching [14], electroless plating [15], rolling [16], laser processing [17], electrical discharge machining [18], and other methods to construct new current collectors. The surface is processed with regularly or randomly distributed raised-edge structures, micro/nanoarray structures, columnar cone structures, and special patterned structures. Other researchers have used dealloying [19, 20], bubble template [21], and freeze casting [10] methods to prepare a copper current collector with a 3D porous structure. The cell performance can be effectively promoted by strengthening the structural and accordingly functional properties of the current collector.

The above-mentioned methods for preparing the structured current collectors mostly depend on complicated chemical, physical, or hybrid technological means, with low efficiency, high cost, and low yield. Therefore, in this work, we propose an efficient manufacturing method based on the plowing-extrusion (P-E) processing mechanism to construct the functional surface structures of the current collector. This approach is denoted as a twice-crisscross micro-plowing (TCMP) method typical of a facile chip-free cutting process, which can efficiently and cost-effectively prepare arrayed micro-groove structures on the surface of a thin metallic sheet [22]. The crisscross grooves are formed and accompanied by an array of convex bulges, burr fins, subsidence, and other secondary microstructures. This structuralized surface has abundant morphologic elements with obvious periodicity, diversity, and multiscale properties. It is purely made by mechanical machining that is conducive to mass production. On the basis of both experimental test and modeling simulation, this study aims to reveal the processing and formation mechanisms of surface structures with cross-scale topographic characteristics and validate its competence in enhancing the functions of the current collector and the performances of the battery.

2 Experiment

2.1 Plowing-Extrusion Processes

In the experiment, a planer (BC6063B, GSK) was used as the machining platform to apply a unidirectional cutting motion to the workpiece. The workpiece was a square copper sheet with a side of 7.5 mm and a thickness of 0.3 mm. The plowing tool made from W18Cr4V (GB)/ T1 (ANSI) was prepared by wire cutting, electrical discharge machining, and edge grinding [22]. The specific design parameters of the cutting tool were prescribed as follows: rake angle α = 40–50°, clearance angle κ = 20–30°, cutting edge inclination angle β = 15–30°, forming angle θ = 10–20°, tool width B0 = 10–20 mm, and tool thickness Lt = 2–4 mm. A self-made stainless-steel fixture for fastening the workpiece was installed on the workbench. The fixture included a regular 12 prism (side length = 10 mm, height = 55 mm) and a cylinder (diameter = 20 mm, height = 3 mm), which respectively control the crossing angle and fix the copper sheet. This fixture also helped prevent the copper sheet from bending due to its rigid support.

Before machining, the verticality of the clamped cutting tool and the level of the workpiece were corrected with a dial gage to prevent tilting without affecting the forming accuracy of the surface structures. The working parameters and working stroke of the shaper were adjusted. After tool setting, the thin copper sheet was plowed twice in the following sequence:

  1. (1)

    First P-E: The cutting depth values and amount of feed were adjusted before plowing, which started from the edge of the copper sheet with a certain cutting pitch. This process completes the one-time formation of the surface structure with a line of grooves.

  2. (2)

    Second P-E: The workpiece fixture was rotated at a certain angle (i.e., crossing angle), and the same cutting parameters were used to perform the second-time processing on the workpiece to produce a 3D structured surface on the copper sheet. As a result, an array of grooves accompanied with convex bulges, scaly burr fins, and subsidence was formed.

The complete machining processes are shown in Fig. 1. Unless otherwise stated, the copper sheets prepared in the same batch were used as the workpiece substrate for the testing of three key processing parameters, i.e., cutting pitches, cutting depths, and crossing angles, to ensure the accuracy and reliability of the test data. The values are listed in Table 1. For each type of a single test, e.g., contact angle test, at least three samples with the same processing parameters were tested to ensure repeatability.

Fig. 1
figure 1

Machining processes of TCMP

Full size image
Table 1 Values of the three key processing parameters
Full size table

The external shape and side surface with a circular or rectangular profile may have effects on the P-E process. However, the cutting process reaches a relatively stable stage of structure formation in a short time. Thus, the above transient influence on the morphology of the copper sheets can be ignored in this study.

2.2 Structure Characterization

A field emission scanning electron microscope (SEM, Merlin, Zeiss) and a super-depth 3D electron microscope (VHX-1000, KEYENCE) were used to characterize the microscopic morphology of the sample surfaces.

2.3 Cutting Force Test

To systematically evaluate the cutting force during the machining process, a digital measurement platform was established. When the cutting tool plows the surface of the copper sheet, the copper material flows to both sides of the cutting direction (X-axis) under the effects of both cutting force and extrusion of the tool. The feed resistance Fy in the Y-axis direction can be neglected. The cutting force in the horizontal direction overcomes the axial motion (X-axis) resistance, while the vertical cutting force overcomes the depth setting motion (Z-axis) resistance. A pressure sensor (9256C1, Kistler) was placed under the workpiece during the machining process. A charge amplifier (5080A, Kistler) was used to amplify and transmit the signal to the data acquisition system (5697A, Kistler). Finally, the data were processed by a computer [23].

2.4 Contact Angle Test

A contact angle testing instrument (JC2000D, POWEREACH) was used to characterize the wettability of the copper-sheet surface. For each test, the number of water droplets was 4 μl.

2.5 Slurry Retention Rate Test

The electrode of a battery was composed of a thin copper-sheet current collector coated with graphite slurry. The electrode was treated by ultrasonic vibration, and the slurry retention rate of the active material on the surface could be calculated to calibrate the bonding degree between the current collector and graphite slurry. The effect of ultrasonic vibration on electrode shedding is much higher than the real reaction; thus, the influence degree of ultrasonic vibration can be used to assess the electrode volume change. The specific process of ultrasonic vibration can be described as follows. First, the mass of the active material on the copper surface (ma) was prescribed. Second, the electrode was placed in deionized water for ultrasonic treatment at 80 W for 15 min. Then, it was cleaned with deionized water and dried under a vacuum environment at 60 ℃. The remaining amount of active materials on the surface of the copper sheet (mb) was calculated. Finally, the slurry retention rate K of graphite slurry on the current collector could be determined by using Eq. (1).

$$ K = \frac{{m_{\text{b}} }}{{m_{\text{a}} }} \times 100\% $$
(1)

2.6 Friction Coefficient Test

The friction coefficient was used to quantify the bonding force. A commercial instrument (MPG-1G, BRUKER) was used to test the friction coefficient of the electrode surface in a point-surface friction mode. The following test parameters were used: the gravity of the weight was 7 N (700 g weight), rotating speed was 100 min−1, test time was 5 min, and friction radius was 5 mm. The test sample was a current collector coated with a slurry of active materials.

2.7 Battery Performance Test

A charge/discharge system (CT2001A, LAND) was used to measure the cycle performances of LIBs. The current density was 0.2 C, the voltage range was 0.02–3 V, and the cycle number was 50. Cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS) tests were conducted on an electrochemical workstation (PGSTAT 302 N, Metrohm). The CV curves were obtained under the following conditions: a scanning speed of 1 mV s−1, voltage range of 0.02–3 V, high potential of 3 V, low potential of 0.02 V, and sampling interval of 0.001 s. The EIS curves were tested in a constant voltage scanning mode, with a scanning frequency of 10−2–105 Hz. The voltage was 3 V with an amplitude of 5 mV.

3 Simulation

The proposed P-E processes must involve the plastic forming mechanism and material flow law of the copper. In this study, the commercial software Deform-3D was used to simulate the P-E processes of creating the surface microstructures on a thin copper sheet, which can be used as the negative current collector in the LIB. When the material of the workpiece and tool, cutting speed, and other processing parameters were determined, Deform-3D would give credible results for revealing the mechanism of TCMP. All simulation parameters are the same as the actual experimental parameters. The simulation results are in agreement with the actual test results, thus verifying the reliability of the simulation.

3.1 First P-E

3D models of the P-E tool–workpiece system were established, as shown in Fig. 2. The workpiece and the tool were partially simplified to reduce the amount of calculation during the simulation process, and the size of the workpiece was 5 mm × 5 mm × 0.3 mm. The cutting depth was set to 160 μm. In the initial state, the tool did not contact the workpiece and was therefore set as a rigid body without meshing. The workpiece was meshed because it was regarded as a deformed part. The simulation parameter settings for the first P-E process are shown in Table 2. The tool was a moving body, and the XYZ axes of the bottom surface of the workpiece were all in a fixed and constrained state.

Fig. 2
figure 2

Finite element model of the first P-E process

Full size image
Table 2 Simulation parameters of the first P-E process
Full size table

3.2 Second P-E

The second P-E process was simulated to derive an in-depth understanding of the forming mechanism of structures, such as subsidence, burr fins, and convex bulges, as shown in Fig. 3. The surface structure of the workpiece was designed according to the surface morphology after the first P-E process. The original base surface was raised by the force of flow and extrusion of the grooved metal, and the raised height was 0.06 mm. The raised fins on both sides of the groove had a height of 0.09 mm, the spacing between the grooves was 0.75 mm, the cutting depth was 0.16 mm, and the crossing angle was 90°. The simulation parameter settings are listed in Table 3. To reduce the amount of calculation, the grid is divided using the local grid window method, and the other settings were kept the same as in the first P-E simulation.

Fig. 3
figure 3

Finite element model of second P-E process

Full size image
Table 3 Simulation parameters of the second P-E process
Full size table

4 Results and Discussion

4.1 Formation Mechanisms of the Surface Structures

Figure 4 shows the forming law of the base metal in the first P-E process. In this process, first, the cutting-edge splits and separates the metal. Then, the metallic material flows along the extrusion surface of the tool, and the matrix forms a V-shape grooved structure along the cutting direction, while the raised-edge structures are formed on both sides of the groove.

Fig. 4
figure 4

Forming law of first P-E process

Full size image

To further reveal the load distribution, metal flow, and stress distribution during the plowing process, Fig. 5 provides the simulation images at a cutting step of No. 300. Figure 5a shows that the cutting load on the surface of the base metal is mainly concentrated on the front of the V-shape groove. This condition is due to the unidirectional plowing. The force-receiving parts are concentrated on the cutting-edge, extrusion surface, and forming surface of the tool. Figure 5b shows that, due to the impact of the cutter, the metal flows in parallel with the cutting direction and then becomes perpendicular to the cutting direction with the extrusion of the surface material. Figure 5c shows that the equivalent stress is mainly distributed in the zone where the cutting edge and the extrusion surface are in contact with the workpiece because of more severe metal deformation. This finding can also be inferred from the image of the metal flow velocity field in Fig. 5b.

Fig. 5
figure 5

Simulation results of the first P-E process: a load distribution, b metal flow velocity field, c equivalent stress distribution

Full size image

Figure 6a shows the shape of the workpiece after the first P-E simulation. The groove formed by the first P-E process in the picture shows obvious periodic characteristics in the structure. Thus, only a single groove was used to analyze the forming mechanism. Figure 6b shows the image of material movement at the end of the simulation. The protrusions on both sides of the groove have a height of about 0.05–0.15 mm.

Fig. 6
figure 6

First P-E: a morphology after completion of the simulation, b metal displacement

Full size image

With the periodicity of the formed structures taken into consideration, only a single unit was used to reveal the flow and deformation mechanisms of the metallic material under unconstrained conditions during the second P-E process. Figure 7 shows the states of the workpiece at different steps. The characteristic structure produced at each stage is shown in Fig. 8. The cutting edge contacts the raised edge of the groove top (65 steps), and subsequently, the extrusion surface extrudes the raised edge (95 steps). The extrusion surface extrudes the top of a groove and the raised edge at 135 and 195 steps, respectively. The above actions and behaviors repeat to form the surface structure on the copper sheet. The forming process can be divided into three stages.

Fig. 7
figure 7

Forming stage division of the second P-E process

Full size image
Fig. 8
figure 8

Structure formation images: a chopping stage, b forming stage of the subsidence at the top of the groove, c forming stage of the convex bulge

Full size image

The first stage (I) is the chopping stage, which generates the burr fins on the edge of the groove. Figure 9 shows the structure formation and appearance in different steps, the metal flow, and equivalent stress distribution diagram while chopping, which are analyzed from the 70th step when the cutting-edge touches the raised edge on the edge of a groove. As the number of simulation steps increases, the cutting-edge splits the raised-edge structure on the edge of the groove of the workpiece first, and then the metal flows along the extrusion surface of the tool to form a burr fin and a raised edge. The metal flows parallel and perpendicular to the cutting direction. The flow process in the parallel direction generates burr fins, while the flow in the vertical direction forms raised edges on both sides of the groove. At the same time, a higher metal flow rate occurs in the place where the substrate contacts the cutting edge and the extrusion surface because of the relatively large plastic deformation of the metal in this part. Figure 9c shows the equivalent stress distribution in the splitting stage. The stress is mainly concentrated in the front part of the groove and changes with the movement of the tool because the metal is severely extruded during the cutting process, resulting in more severe deformation.

Fig. 9
figure 9

Chopping stage: a forming structure, b metal flow velocity filed, c equivalent stress distribution

Full size image

The second stage (II) is the formation of the subsidence at the top of a groove to generate the raised edge and subsidence. The forming law of subsidence is shown in Fig. 10. As the tool moves in the feed direction, the cutting-edge separates the material first, and then the metal flows along the tool extrusion surface. The forming surface then squeezes the flowing metal to form the raised edge on both sides of the grooves. The raised edges on both sides of the first P-E groove are enclosed to form a subsidence structure on the top of the groove.

Fig. 10
figure 10

Structure forming law images: a, b subsidence, and c, d convex bulge

Full size image

The third stage (III) is the formation of the burr fin and convex bulge structure. The forming law of the convex bulge is shown in Fig. 10. In the 150th step, the cutting-edge splits the raised edge on the top of the groove. As the tool moves, the metal flows along the extrusion surface of the tool and bends. Finally, it wraps the grooves to form a convex bulge, and the metal flows out, which forms burr fins in the interlaced spaces of the grooves. At the same time, the cutting-edge contacts the raised edge of a groove and enters the next forming period.

4.2 Influence of Processing Parameters on the Cutting Force

The cutting force mainly overcomes the resistance of metal materials to elastic and plastic deformation and also the friction between the metal material and the plowing tool.

The average cutting force can be calculated from the testing results. For the convenience of numbering samples, this study named each sample with different processing parameters. Here the letters S, D, and A represent the variables of cutting pitch (μm), cutting depth (μm), and crossing angle (°), respectively.

The average cutting forces of samples with different cutting parameters are shown in Fig. 11. The cutting force in both horizontal and vertical directions, generated by the samples with different cutting pitches in the first P-E process, is almost the same. This situation is also the same for the second P-E process, in which the cutting force of samples with different cutting pitches is less than that in the first P-E. This condition occurred because, during the second P-E, the part of the workpiece surface removed by the first P-E will not produce cutting force. In addition, the horizontal cutting force is always greater than the vertical cutting force because of the relatively larger flow resistance of the copper material in the horizontal direction [23]. In general, under the condition of a certain cutting depth, the cutting force is not sensitive to the change of cutting pitch.

Fig. 11
figure 11

Average cutting force of samples with different cutting pitches, cutting depths, and crossing angles

Full size image

Results suggest that the cutting force generated by the second P-E process is much lower than the first, regardless of the change of cutting depths. In addition, the cutting force in the horizontal direction is greater than that in the vertical direction during the machining process. The metal materials understandably suffer from the greatest resistance in parallel with the cutting direction. The cutting force perpendicular to the cutting direction is mainly limited by metal flow and extrusion, which is smaller. The cutting force is positively related to the value of cutting depth. When the same cutting depth and cutting pitch are used, the average cutting force of samples with different crossing angles does not change obviously, which is consistent with the results of previous studies [23].

4.3 Influence of Processing Parameters on the Surface Morphology of the Current Collector

The surface structure of the copper-sheet current collector made by the P-E processes is composed of various microstructures, such as burr fins, array grooves, subsidence, and convex bulges, which are formed stably, as shown in Fig. 12. These morphological features directly affect the functional performances of the copper-sheet current collector, depending on the cutting pitch, cutting depth, and crossing angle. TCMP provides an effective method for active control of the shape, morphology, property, and performance of the structured current collector. Therefore, this section focuses on the morphological characteristics of the composite microstructures based on different cutting parameters to reveal the influential mechanisms of the key processing parameters.

Fig. 12
figure 12

SEM images of samples with different cutting pitches: (a), (c), (e), (g), and super-depth of field images: (b), (d), (f), (h)

Full size image

Figure 12 shows that the surface structure characteristics of samples with different cutting pitches are mostly similar. The height differences of the S-250-120-90, S-500-120-90, and S-750-120-90 samples are 165.7, 162.8, and 185.8 μm, respectively. The actual height difference of each sample is obviously greater than the designed cutting depth because the plowing tool squeezes the surface of the workpiece, causing part of the material to flow upward to form burr fins. With the existence of various surface microstructures, the height difference of the structured samples is larger than the fluctuation of the surface height difference of the smooth copper sheet [24].

The S-250-120-90 sample in Fig. 12c shows obvious multidimensional and multiscale complex surface features. This appearance helps extend the longitudinal depth of the sample surface, thereby enhancing the conductive contact between the current collector and the active material, as well as increase the loading of active materials. The burr fins and convex bulges mainly exist at the boundary of the square bump, while the subsidence is found at the center of the bump. The number of different structures can be evaluated indirectly based on the density of the bumps, as shown in Fig. 12c, e and g. In general, using a larger cutting pitch may lead to a smaller number of microstructures on the surface of the copper sheet. Optimal parameters for the cutting pitch derive from the actual functions and effects of various structures. For example, the use of a smaller cutting pitch helps generate more burr fins and convex bulges. A detail that must be noted is that when the cutting pitch is too small, the grooves formed by plowing are very likely to overlap due to the limitation of the tool geometry. This condition may cause many chips to fly out, which is not conducive to maintaining and stabilizing the surface morphology of the current collector.

Figure 13 shows the SEM images and super-depth images of samples with different cutting depths. A comparison shows that the D-250-80-90 sample has more burr fins with a more complete formation of convex bulges and small bumps. With the use of a smaller cutting depth, D-250-40-90 squeezes less on the edge of the small bump, and its subsidence and convex bulges formation are imperfect. D-250-120-90 and D-250-160-90 adopt a larger cutting depth; thus, the edges of the small bump are squeezed more strongly. As a result, the burr fins are large and thin and easily fall from the substrate because they are weakly connected to the substrate. Although the depth of the subsidence part is greater, the total number of microstructures is smaller.

Fig. 13
figure 13

SEM images of samples with different cutting depths: (a), (c), (e), (g), and super-depth of field images: (b), (d), (f), (h)

Full size image

Figure 14 shows the SEM images and super-depth scanning profiles of the copper sheet surface topography based on different crossing angles. The subsidence structures of A-250-80-30, A-250-80-60, and A-250-80-90 are all approximately parallelograms (squares at 90° crossing angles). A-250-80-30 and A-250-80-60 employ larger crossing angles of cutting. Thus, the small areas formed after twice plowings become long bars. As a result, the subsidence becomes narrow and not neat enough, accompanied by a wider top opening of the groove but fewer burr fins and other structures. Compared with other samples, A-250-80-90 has a more complete surface microstructure and a more regular morphology. In the range of 0–90°, the abundance of microstructures increases with the increase in the crossing angle.

Fig. 14
figure 14

SEM images of samples with different crossing angles: (a), (c), (e), (g) and super-depth of field images: (b), (d), (f), (h)

Full size image

4.4 Influence of Processing Parameters on the Surface Characteristics of the Current Collector

A critical problem of industrial copper current collectors that hinders their application for LIBs is that they are too smooth to gain sufficient surface functional properties in interfacial contact, slurry retention, and surface friction. To this end, the presented TCMP method of producing microstructures on the surface of a copper sheet illuminates its great potential to be used for LIBs. Therefore, the surface characteristics and functional properties of prepared copper sheets need to be characterized and analyzed under different processing conditions.

Figure 15 shows the contact angles of samples with different cutting pitches, cutting depths, and crossing angles. In the same group, the contact angles of the TCMP-based samples are all smaller than the smooth copper sheet. This finding confirms that the microstructures produced by TCMP have a positive effect on improving the hydrophilicity of the copper sheet surface. The microstructures can increase the surface roughness of the substrate, enhancing its wettability and capillary performance. This feature is supposed to benefit the spreading behavior of the slurry. The difference in the contact angles of the three groups of smooth copper sheets is caused by the wear of the cutting tools during the machining process.

Fig. 15
figure 15

Result of contact angle tests

Full size image

The contact angles of S-250-120-90, S-500-120-90, and S-750-120-90 with different cutting pitches are 0°, 54.0°, and 55.7°, respectively, which are all smaller than the 73.4° contact angle of the smooth surface sample. S-250-120-90 exhibits the highest hydrophilicity with a contact angle of 0° because of the presence of more microstructures with a smaller bump density. Despite the increase in the area of the subsidence structure on the surface of the samples based on the larger cutting pitch (S-500-120-90 and S-750-120-90), the number of grooves, burr fins, and convex bulges is reduced. Evidently, the increase in the area of the subsidence has a smaller impact on the surface hydrophilicity than the decrease of other structures, so the use of a larger cutting pitch does not bring sufficient hydrophilicity.

The contact angles of the D-250-40-90, D-250-80-90, D-250-120-90, and D-250-160-90 samples with different cutting depths are 62.8°, 0°, 0°, and 0°, respectively. This result indicates that the structured copper sheet with a larger cutting depth has a better capillary performance, which is mainly due to the deeper grooves on its surface. In addition, the surface of the structured workpiece fails to gain enough hydrophilicity at a very small cutting depth. An optimal value must be reached for actual operation.

For the samples with different crossing angles, the contact angles of the A-250-80-30 and A-250-80-60 samples are almost the same, indicating that the effect of crossing angle on the hydrophilicity of structured copper sheets can be neglected. The contact angle of A-250-80-90 is as small as 48.1°; thus, A-250-80-90 is more hydrophilic than other samples in the same group [16].

Figure 16 compares the slurry retention rate of samples based on different processing parameters. After the ultrasonic test, the slurry quality of each sample is weakened, but the retention rate is different.

Fig. 16
figure 16

Slurry retention rate of samples with different cutting pitches, cutting depths, and crossing angles

Full size image

The S-250-120-90 has a higher slurry retention rate, which means that the bonding force between the active material and current collector is stronger in this case. With this sample, the volume expansion of the active material during the cycle of charging and discharging is expected to be well suppressed by the surface microstructures. As the cutting pitch increases, the slurry retention rate of the samples shows a downward trend. This downward trend is due to the gradual decrease in the overall density of the microstructures with the increase in the cutting pitch, reducing the combination strength between the copper substrate and electrode slurry materials. In terms of the bonding effect on the slurry, the grooves and convex bulges play more important roles than the subsidence. In summary, the structured copper sheet is endowed with a higher slurry retention rate than the smooth one because of the coupling effects from different microstructures.

The slurry retention rate of A-250-80-90 with the largest crossing angle reaches 89.2% because its surface has more burr fins, subsidence, and convex bulges, thereby mitigating the destructive effects of ultrasonic shock on slurry attachment.

Figure 17 shows the electrode morphology of D-250-80-90 after ultrasonic testing. The burr fin does not restrict the slurry. After the ultrasonic testing, structures such as grooves and convex bulges can still store a large amount of slurry, indicating that grooves and convex bulges have the function of inhibiting the shedding of active materials. In addition, the sample’s surface exhibits many cracks after the test, but the active material layer remains intact, which shows that the electrode based on the structure of grooves has good surface characteristics and can remain stable during charge and discharge cycles.

Fig. 17
figure 17

SEM images of composite microstructure copper sheet after ultrasonic testing

Full size image

Figure 18 compares the average friction coefficients of samples based on different cutting parameters. The friction coefficient of the smooth copper sheet (0.05) is significantly lower than that of most of the structured copper sheets because of the lack of surface structures with a smaller roughness. The TCMP process introduces composited microstructures into the surface of the current collector, thereby increasing the roughness and the friction coefficient. Among the samples with different cutting pitches, S-250-120-90 has the largest friction coefficient (0.064). The use of graphite as the active material may more or less reduce the surface friction degree due to its natural function of lubrication and friction reduction.

Fig. 18
figure 18

Average friction coefficient of copper sheets based on different cutting pitches, cutting depths, and crossing angles

Full size image

For the samples with different cutting depths, the friction coefficient first increases and then decreases with the increase in the cutting depth. This phenomenon is related closely to the surface microstructures of the copper sheet. The friction coefficient of D-250-80-90 is the largest (0.072) because it has the greatest number of surface microstructures and the highest degree of roughness. This finding is consistent with the rationale from Fig. 13. The friction coefficient of D-250-40-90 is 0.048, which is slightly lower than the friction coefficient (0.050) of a smooth copper sheet. This result shows that under this condition, the graphite slurry embedded in the groove has a greater influence on the coefficient of friction than the surface microstructure.

As mentioned previously, the crossing angle has a great influence on the surface morphology of the structured copper sheet. Evidently, as the crossing angle increases, the friction coefficient of the copper surface increases gradually. This result can be inferred from the above description with regard to its morphological features.

4.5 Application and Validation of Structured Copper Current Collector in Lithium-Ion Batteries

In this study, a bare-surface sample and the as-prepared A-250-80-90 with better surface functional properties were selected as the copper current collector and coated with graphite slurry to make an integrated negative electrode assembled into lithium-ion button batteries for electrochemical performance testing and comparative analysis.

The battery cycle curves based on the smooth sample and A-250-80-90 are shown in Fig. 19a, b, respectively. The specific capacity of the former battery decays quickly, whereas the latter not only has a larger specific capacity and a higher capacity retention rate but also produces a better cycle performance. This performance is due to the abundant microstructures on its surface, which enables the whole electrode structure to maintain good stability.

Fig. 19
figure 19

Electrochemical performance of smooth copper sheets and A-250–80-90 as current collectors for LIBs

Full size image

Figures 19c, d show the 1st, 10th, 20th, 30th, 40th, and 50th voltage-specific capacity curves of the batteries assembled with bare-surface copper sheets and A-250-80-90 at a constant current of 0.2 C. The voltage-specific capacity curve of the bare-surface sample does not overlap at all, indicating that the specific capacity of the battery decays quickly with very poor reversibility. This is also the case for Fig. 19a. The battery charging and discharging platform assembled with bare-surface copper sheets is relatively obvious, indicating that the internal electrochemical reaction of the battery is relatively complete. This finding further shows that the battery capacity declines mainly because the active material falls off the substrate due to volume expansion [25]. The cycle voltage-specific capacity curve of the A-250-80-90 battery has a very high overlap, indicating that the battery has good reversibility and stability of the internal electrochemical reaction [26]. The separation of the curve between the first cycle and the subsequent cycle is attributed to the formation of the solid electrolyte interface inside the battery during the first cycle [5].

The CV curve can intuitively disclose the electrochemical reaction process of the battery. The CV curves of the battery assembled with the smooth copper sheet and A-250-80-90 are shown in Figs. 19e, f, respectively. The oxidation–reduction reaction process presented by the CV curves of the two batteries is basically the same. With the smooth copper sheet battery in Fig. 19e taken as an example, the reduction peak of 0.002–0.2 V corresponds to the electrochemical reaction process of the Li+ embedded active material graphite in the electrolyte during the scanning process of the reduction reaction (discharge process). In the oxidation scanning, the oxidation peak around 0.25 V is caused by the extraction of Li+ from the graphite. The overlap of the second and third CV curves of the A-250-80-90 battery is significantly better than that of the smooth copper sheet battery. This finding indicates that the use of A-250-80-90 as a current collector improves the stability of de-intercalating lithium of the negative electrode, thereby improving the reversible performance of the battery.

Figure 20 shows the EIS curves of the battery assembled with the smooth copper sheet (Bare-Cu) and A-250-80-90 before cyclic charging and discharging. The value at the intersection of the semicircle and the real axis corresponds to the ohmic resistance when electrons are transferred in the battery. The two batteries generally have the same ohmic resistance. The diameter of the semicircle of the intermediate frequency range curve is positively related to the charge transfer impedance in the active material. The semicircle radius of the smooth copper sheet battery is obviously larger than that of the battery assembled in A-250-80-90, thereby suggesting that the charge transfer impedance of this battery is smaller [27]. The low-frequency straight line is related to the Warburg impedances that Li+ diffuses into the electrode material [28,29,30]. The intercepts of the two curves on the horizontal axis are almost the same, which shows that the internal resistance of the battery caused by the LiPF6 electrolyte is identical [31]. The slope of the straight-line section of the low-frequency curve is negatively correlated with the lithium-ion migration coefficient inside the battery [32]. The slope of the low-frequency straight section of the A-250-80-90 battery is significantly smaller than that of the smooth copper sheet battery, and the lithium-ion transference number is greater. In this case, the internal electrochemical reaction of the A-250-80-90 battery is more sufficient, which can effectively improve the performance of the battery.

Fig. 20
figure 20

Results of EIS tests

Full size image

5 Conclusions

This study proposes an in-house TCMP method for machining composited functional structures on the surface of the copper current collector for LIBs. A finite element simulation of TCMP was performed for comparison purposes. With their functional structures, such as burr fins, the as-prepared current collectors have excellent surface characteristics. The performance of the battery assembled with this structured current collector was tested, and the following conclusions were made:

  1. (1)

    The effects of the P-E mechanism, metal deformation flow law, and main processing parameters on the plowing force were analyzed by simulation. Results show that the size of the cutting force is positively correlated with the cutting depth, while the cutting pitch and crossing angle have little effect on the cutting force. The vertical component of the cutting force is always smaller than the horizontal component, and the cutting force of the second P-E is less than that of the first P-E.

  2. (2)

    The formation of different structures such as burr fins, raised edges, convex bulges, and subsidence is closely related to the different forming stages. The characterization results confirm the periodicity and stability of TCMP. The most abundant burr fins and convex bulges can be formed when the cutting pitch is 250 μm, the cutting depth is 80 μm, and the crossing angle is 90°.

  3. (3)

    The as-prepared current collector’s superior hydrophilicity and bonding strength with slurry are proven by the contact angle, slurry retention rate, and coefficient of friction tests. The A-250-80-90 sample has a contact angle of 0°, a slurry retention rate of 89.2%, and a friction coefficient of 0.074, thus having the best surface properties.

  4. (4)

    Compared with the smooth copper sheet, the A-250-80-90 sample, when used as a current collector in the LIB, shows a better cycle performance. The specific capacity of 318.6 mAh/g is still maintained after 50 cycles at a current density of 0.2 C. Overall, better capacity reversibility, smaller overall impedance, higher lithium-ion transference number, and better battery performance are demonstrated.