Open Access
Add to BibSonomy
Abstract
Large-area copper layer removal is one of the essential processes in manufacturing printed circuit boards (PCB) and frequency selective surfaces (FSS). However, laser direct ablation (LDA) with one-step scanning is challenging in resolving excessive substrate damage and material residue. Here, this study proposes a laser scanning strategy based on the laser-induced active mechanical peeling (LIAMP) effect generated by resin decomposition. This scanning strategy allows the removal of large-area copper layers from FR-4 copper-clad laminates (FR-4 CCL) in one-step scanning without additional manual intervention. During the removal process, the resin decomposition in the laser-irradiated area provides the mechanical tearing force, while the resin decomposition in the laser-unirradiated area reduces the interfacial adhesion force and provides recoil pressure. By optimizing scanning parameters to control the laser energy deposition, the substrate damage and copper residue can be effectively avoided. In our work, the maximum removal efficiency with different energy densities, pulse duration, and repetition frequency are 31.8 mm2/ms, 30.25 mm2/ms, and 82.8 mm2/ms, respectively. Compared with the reported copper removal using laser direct write lithography technology combined with wet chemical etching (LDWL+WCE) and LDA, the efficiency improved by 8.3 times and 66 times. Predictably, the laser scanning strategy and the peeling mechanism are simple and controllable, which have potential in electronics, communications, and aerospace.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Multilayer materials with metal conductive layer/insulating substrates play an important role in electrical and electronics [1,2], mobile communications [3], and aerospace [4,5]. Such materials are usually composed of dielectric substrates like glass fiber epoxy laminates (FR-4) or aramids, and conductive metal layers such as copper or aluminum. General applications include printed circuit boards (PCB) [6], and frequency selective surfaces (FSS) [7] using electromagnetic functional materials. It is necessary to remove metal layers in designated areas on the surface to construct patterns to achieve specific functions, such as conductive patterns on PCB [8] or resonant unit arrays on FSS [9]. For metal removal from such multilayer material, mask lithography combined with wet chemical etching (WCE) is still widely adopted [10,11]. Recently, reports on laser direct write lithography technology combined with WCE (LDWL + WCE) have increased. Aos Alwaidh [8] and Ai [9] reported the processing of rigid and flexible PCBs and the fabrication of FSS on FR-4 copper-clad laminates (CCL) using UV laser LDWL + WCE, respectively. However, mask lithography and LDWL required a series of cumbersome processes including gluing, etching, and debonding. Moreover, eco-unfriendly chemical solutions were adopted.
As a flexible material removal method, laser direct ablation (LDA) has received much attention in recent decades [12–15]. This non-contact removal process makes the patterning of multilayer materials simple and flexible, allowing for high-resolution fine circuitry [16]. The LDA no longer seems to be the first choice for large-area material removal due to the significant differences in physical-chemical characteristics between different layers. Despite this, there is still enthusiasm for using LDA to remove the large-area metal in multilayer materials. Jaka reported an advanced pulse positioning approach that utilizes acousto-optic deflectors (AOD) to control the random location of laser pulses on the material surface, achieving large-scale material removal with low accumulated heat [6]. However, this approach requires strict synchronization control between the laser and AOD. Ai [9] and Yang [17] reported using LDA to remove the copper layer-by-layer from FR-4 CCL to fabricate FSS by galvanometer scanning, respectively. The realization of high-quality copper removal with the layer-by-layer method required lower laser energy and more scanning times, which means a decrease in efficiency.
Blaž Kmetec reported a process of large-area copper removal based on laser-induced delamination in 2009 and found that the removal mechanism is mechanical peeling caused by laser-induced phase transition of the epoxy resin substrate [13]. The laser irradiation was divided into the laser structuring and delamination Stages in their processing. Although this process effectively avoids substrate damage, it requires switching the laser and scanning patterns during processing. Due to the complicated material heating process caused by processing parameters and material properties, removing large-area copper layers by a single laser scanning without substrate damage is challenging.
Based on the laser-induced copper removal method mentioned in Ref. [13], we propose a laser scanning removal strategy that relies mainly on resin decomposition to actively peel off large-area copper layers. According to the observed and measured results, the scanning process was systematically analyzed to understand the copper peeling mechanism and the physic-chemical change. The influence of scanning parameters on the removal quality including copper residues and substrate damage was also investigated. Compared to layer-by-layer removal and laser-induced delamination, this one-step removal method is more flexible and efficient. This laser-induced active mechanical peeling (LIAMP) method can be an essential reference for the selective removal of multilayer materials.
2. Experimental details
2.1 Experimental setup and materials
As shown in Fig. 1(a), the experimental system was the most common laser processing equipment. A fiber laser (JPT Corporation, China) with a maximum output power of 100 W was used to output a nanosecond laser pulse at 1064 nm. The pulse duration (2 ns∼500 ns, non-continuous), repetition frequency (10 kHz∼4000 kHz), and pulse energy (maximum: 0.04 mJ at 2 ns and 1.5 mJ at 500 ns) are adjustable. A galvanometer (SCANLAB, IntelliSCAN III 10) was adopted to guide laser scanning on the material surface. A telecentric F-theta lens (self-design) with a focal length of 100 mm was employed to focus the laser beam (with a focused spot diameter of theoretically 29 µm at 1/e2 of the maximum intensity). The maximum scanning area is approximately 50 mm × 50 mm. All experiments were conducted in an atmospheric environment.
The experimental material was commercial FR-4 copper-clad laminate (CCL) with the structure shown in Fig. 1(b). The uppermost layer is the copper layer with a thickness of 12 µm, and the substrate is glass fiber. The copper layer is glued to the substrate by epoxy resin as the binder with a thickness of 12∼15 µm. And the total thickness is 1 mm. The thermal conductivity of resin is three orders of magnitude lower than that of copper [18,19]. This study aimed to remove the copper layer from the CCL by laser ablation without copper residue and damage to the glass fiber substrate.
2.2 Laser scanning strategy
Laser hatch scanning is widely accepted as a fast and effective method for removing large-area copper layers. However, this scanning method involved a two-dimensional (2D) laser overlapping of spot-spot and line-line, which leads to an extraordinarily complicated and difficult-to-control heating process. Thus, we simplified the 2D laser overlapping to 1D spot-spot overlapping by increasing the hatch spacing (larger than the diameter of the focal spot), as shown in Fig. 2. This strategy divides the scanning area into laser-irradiated and -unirradiated areas. The geometry of region scanning was 6 mm × 6 mm square. In the processing geometry, we defined the laser pulse move direction of the first hatch line as the X-direction and the hatch direction as the Y-direction. The adjustable parameters in these experiments include scanning speed, hatch spacing D, laser pulse energy, repetition frequency, and pulse duration.
2.3 Measurements and characterization
All of the 3D morphology measurements were conducted by a laser confocal microscope (KEYENCE, VK1000). Furthermore, a thermal field emission scanning electron microscope (SEM, Czech Republic FEI, NanoSEM 450 FP2053/45) was used to observe substrate damage and copper residues.
3. Results and discussion
3.1 Copper removal mechanism
3.1.1 Explanation of the removing process
To investigate the removal process, two tests with different scanning parameters were conducted, and the results are shown in Fig. 3. In these experiments, a pulse duration of 30 ns was selected. The energy density was 28.22 J/cm2, which was twice the damage threshold of copper at 30 ns (measured to be 14.2 J/cm2) [20,21]. The repetition frequency was set to 100 kHz. All of these experiments were achieved in a single-scanning. As described in Test 1, the hatch spacing was 40 µm and the scanning speed varied from 500 mm/s to 320 mm/s. The hatch spacing must be larger than the diameter of the focal spot to construct laser-irradiated and unirradiated areas. It can be seen that with the scanning speed decreases, the copper residue in the initial segment gradually decreases. And when the speed is less than 360 mm/s, the brightness of the ablation area also decreases, which means that there may be substrate damage. In Test 2, the scanning speed was 420 mm/s and the hatch spacing varied from 60 µm to 30 µm. These results showed high removal quality that without obvious copper residue (excluding No.1) and glass fiber exposure (substrate damage).
Fig. 3. Samples of laser processing (No.1-No.10 are the laser processing results with the scanning speed varying from 500 mm/s to 320 mm/s (with the interval of 20 mm/s), the hatch spacing is 40 µm. No.11-No.26 are the laser processing results with the hatch spacing varying from 60 µm to 32 µm (with the interval of 2 µm), the scanning speed is 420 mm/s.).
In order to analyze the laser ablation process, the 3D morphology and SEM images of the initial segment and central area (No.13) were observed, and the results are shown in Fig. 4 and Fig. 5. The scanning speed and the hatch spacing were 420 mm/s and 56 µm, respectively. As can be seen in Fig. 4(a), macroscopically high removal quality exhibited copper residue in the initial segment at the microscopic level. And the residual copper has a groove structure (with three grooves, named G1, G2, and G3, respectively). The groove area was the hatch line with laser irradiation, while the protruding part was the laser-unirradiated area. From the depth measurement results in Fig. 4(a), the depth of the groove gradually increased with the laser scanning continuing until the copper residue completely disappeared. Figure 4 (b) was the cross-sectional morphology observed after rough grinding. It can be easily observed that there is still a certain thickness of copper in grooves G1, while grooves G2 and G3 show irregular pits in some areas. Figure 4(c)-(e) provide details of the three grooves, respectively. The bottom of the pits in G2 and G3 is exposed resin. Furthermore, the copper on both sides of these grooves still adhered to the resin layer.
Fig. 4. (a)The SEM image of the initial segment. There are three grooves in the figure. And the depth of the three grooves is given in the upper right corner. (b)The cross-section of the three grooves. The enlarged SEM images of (c)G1; (d)G2; (e)G3.
Fig. 5. (a) The SEM image of the substrate interface after copper removal, the copper layer was removed by forcefully tearing with a clip. (b) The 5X enlarged SEM image of (a). (c)The 10X enlarged SEM image of (a). (d) The SEM image of the substrate interface after copper removal, the copper layer was removed by laser ablation. (e) The 5X enlarged SEM image of (d). (f)The 10X enlarged SEM image of (d).
Figure 5 shows the SEM images of the substrate interface after copper was removed without and with laser ablation. In Fig.5(a)-(c), the copper layer is removed by forcefully tearing with a clip, while the copper layer in Fig. 5(d)-(f) is removed by laser ablation. As shown in Fig. 5(a)-(c), it can be seen that the original substrate interface has massive micron-holes, which are formed by the fabrication of FR-4 CCL with hot-pressing. The presence of these micro-holes results in a larger contact area between the resin and copper, forming strong interfacial adhesion. After laser ablation, the number of micro-holes in Fig. 5(d) significantly decreased. The area with micro-holes is the laser-unirradiated area, while grooves are the laser-irradiated area. From the enlarged SEM images, there are many crystalline particles in the laser-irradiated area. Compared to the rough surface in Fig. 5(c), the surface of the laser-unirradiated area in Fig. 5(f) becomes smooth. In brief, the laser ablation can be divided into three types: only copper ablation was present (G1), resin exposure and copper residue were present (G2 and G3), copper completely removed (Fig. 5(d)-(f)).
3.1.2 Explanation of the removing process
An explanation was provided in Fig. 6 to understand the laser ablation removal process. The copper layer removal process included copper evaporation and caused resin decomposition caused mechanical peeling. According to the results in Fig. 4 and Fig. 5, the laser scanning process could be divided into three stages: copper evaporation stage (Stage I, groove G1 in Fig. 4), copper layer partial penetration stage (Stage II, groove G2 and G3 in Fig. 4) and mechanical peeling stage (Stage III, grooves in Fig. 5), as shown in Fig. 6(a). Here, several distances were defined. The width L1 of Stage I, L2 of Stage II, L3 of Stage III, and the hatch spacing D. The distance of d in Fig. 6(d) stands for the peeling width, which is equal to the hatch spacing minus the single-line ablation width. The height h represents the depth of resin decomposition in Stage III. Figure 6(b)-(d) shows the copper removal details in these three stages.
Fig. 6. (a) Schematic diagram of one-step scanning process for large-area copper removal. Different stages in the removal process. (b)Stage I; (c)Stage II; (d)Stage III. (The thickness scale of each layer in the figure has been adjusted to facilitate understanding of the copper layer removal process).
Stage I: the removal process mainly relies on copper evaporation, and this process only happens in the laser-irradiated area. At the beginning of the laser processing, the copper layer melts and evaporates, forming a groove in the hatch line (G1 in Fig. 4). During this stage, a thermal balance is about to be established, where the ablated groove is relatively shallow. As the laser processing continues, more laser energy is deposited on the materials. Therefore, more copper would be ablated due to the increase in energy absorption, and the groove becomes deeper. It is worth noting that there is no resin decomposition. Meanwhile, the number of grooves is inconsistent under different scanning parameters during this stage.
Stage II: the removal process includes copper evaporation, the peeling effect generated by local resin decomposition. Similar to Stage I, as energy deposits, the copper layer is still ablating. During this stage, two physical changes of the copper layer are crucial for resin decomposition. (1) The thickness of the copper layer decreases after melting and evaporation. (2) The temperature of the copper layer increases, leading to an increase in the energy absorption coefficient of copper [22]. The energy absorbed by the copper layer is converted into heat, which can directly affect the resin, leading to decomposition and gas expansion (Formation of recoil pressure) [23,24]. After resin decomposition, the copper layer can be peeled off and the substrate exposed. Consequently, the groove gradually deepens (G3 > G2 > G1).
Stage III: the removal process includes copper evaporation and gas expansion-induced burst in the laser-irradiated area, and mechanical peeling in the laser-unirradiated area. Different from Stage I and Stage II, the copper layer in the laser-unirradiated area is also removed. To further understand the mechanical peeling process of the copper layer, the forces at the interface between the copper layer and resin in the laser-unirradiated area are briefly analyzed (as described in Table 1). There are three forces at the interface, which are the interfacial adhesion force (Fa), the recoil pressure (Fr) [23,24] generated by the decomposition of a small amount of resin, and the mechanical tearing force (Fm) formed by the gas expansion and peeling effect generated by the extensive resin decomposition in the laser-irradiated area. The directions of each force are shown in Fig. 6(d).
Table 1. Forces in the interface of resin and copper
During this stage, the heat accumulation is further enhanced and the resin decomposed area also increased. The substrate resin in the laser-unirradiated is also affected by the heat conduction of the copper layer. In this area, because of the thicker copper layer and less resin decomposition, the substrate resin decomposes and then re-solidifies (smooth resin surface in Fig. 5(f)). Then, the resin decomposition and re-solidification results in a reduction in the number of holes, reducing the contact area of copper-resin and the interfacial adhesion (Fa). At the same time, the gas expansion generated by the resin decomposition will also create a recoil pressure at the interface (Fr). The copper layer is gradually separated from the substrate by the combined action of these two forces.
On the other hand, Large-area gas expansion in the laser-irradiated area can cause a strong mechanical tearing force to act on the copper layer of the laser-unirradiated area (Fm). The mechanical force can peel off the copper layer from the substrate. Because of the huge difference in thermal conductivity coefficient between resin and copper [18,19,23], heat conduction and accumulation occur mainly in the copper layer. The heat conduction and accumulation inside the resin layer are limited to a very small range (the red arrow in Fig. 6(d)). Then, the mechanical peeling effect became the main copper layer removal mechanism.
3.2 Improvement of the removal quality
High-quality copper layer removal means no copper residues while avoiding damage to the substrate. According to the analysis of the laser ablation process in 4.1, the residual copper is mainly present in Stages I and II, while the damage to the substrate is mainly present in Stage III. In Stages I and II, the copper residue is expressed as the width of Stages I and II (L1 and L2, a larger width means more copper residue). On the other, the influence on the substrate is analyzed by the contrast ratio (CR) of the grooves in Stage III. For the purpose of improving the removal quality, the influence of scanning parameters on the removal process of Stages I, II and III was investigated. Therefore, high removal quality should ensure that the entire laser ablation process is dominated by the mechanical peeling effect of Stage III.
3.2.1 Influence of scanning speed on the removal quality
Figure 7(a) shows the widths of Stages I and II with different scanning speeds, respectively (with a laser pulse duration of 30 ns, a repetition frequency of 100 kHz, and an energy density of 28.22 J/cm2, a hatch spacing of 40 µm). In Fig. 7(a), with the scanning speed decreasing, the width L1 of Stages I decreased to 0 at 420 mm/s, and the width L2 of Stage II decreased to 0 at 380 mm/s. Owing to laser pulse overlapping, lower scanning speed means higher laser-irradiated and copper-absorbed energy, extending the range of heat conduction. It indicates that the resin can easily decompose, making Stages II and Stages III appear earlier in the scanning process. A simple conclusion is that the copper residue in the initial segment can be eliminated by decreasing the scanning speed.
Fig. 7. (a) The width of Stage I and Stage II with different scanning speeds. (b) The CR of grooves in Stage III with different scanning speeds. The SEM images of Stage III with the scanning speed of (c)460 mm/s; (d)380 mm/s; (e)320 mm/s.
However, a lower scanning speed means that more energy is absorbed by the resin layer, causing excessive resin decomposition in Stage III. Figure 7(b) is the CR of grooves in Stage III with different scanning speeds. And Fig. 7(c)-(e) are several SEM images of the surface after laser ablation. All of the CR were the average of 10 measured lines. It can be seen that the CR shows a small-scale fluctuation when the scanning speed is larger than 400 mm/s, and the rapidly increases as the speed decreases to 380 mm/s. As the speed further decreases, the contrast value shows a significant drop. These contrast fluctuations can be attributed to different degrees of resin decomposition in the laser-irradiated and -unirradiated areas.
At a larger scanning speed, the resin in the laser-irradiated area decomposes in large quantities, while the resin in the laser-unirradiated area experiences localized decomposition, resulting in a minor change in h. As shown in the SEM image of 460 mm/s in Fig. 7(c), the groove structure is not obvious. Although these results indicate high copper layer peeling quality, there is a high amount of copper residue in the initial segment.
When the speed is reduced to 380 mm/s, the resin decomposed area in the laser-irradiated area increases, while the resin decomposed in the laser-unirradiated area occurs nearside of the laser-irradiated area. In that case, the width of d decreases while the height of h increases. At the same time, obvious groove structures were observed in the SEM image (Fig. 7(d)).
When the speed further decreases, the resin in the laser-unirradiated area shows a large amount of decomposition, leading to a drastic decrease in the height of h. As shown in the SEM image of 320 mm/s (Fig. 7(e)), while the resin in the laser-irradiated area is excessive decomposed, exposing the glass fiber of the substrate.
3.2.2 Influence of hatch spacing on the removal quality
Another scanning parameter that has a significant effect on the peeling effect is the hatch spacing. Figure 8 shows the widths of Stages I and II and the CR of Stage III with different hatch spacing, respectively. The laser conditions are the same as the experiments of scanning speed. The scanning speed was 420 mm/s. Since the independent variable is hatch spacing, the width of L1 and L2 are replaced by the number of grooves. From Fig. 8 (a), it can be seen that the grooves of Stage I remain at 0. The grooves of Stage II are reduced from 4 to 2 when the hatch spacing is reduced from 60 µm to 52 µm and remain at 2 as the hatch spacing continues to decrease. In the scanning speed of 420 mm/s, there is no Stages I during the scanning process (shown in Fig. 7 (a)). That is mainly because a smaller hatch spacing accelerates the process of heat accumulation, causing the position of resin decomposition moving to the initial segment. However, due to the unidirectional laser scanning on the Y-axis, the first hatch line cannot be affected by the subsequent hatch line. Based on this analysis, it could be simply concluded that the copper residue in the initial segment only can be eliminated by decreasing the scanning speed.
Fig. 8. (a) The number of grooves of Stage I and Stage II with different hatch spacing. (b) The CR of grooves in Stage III with different hatch spacing. The SEM images of Stage III with the hatch spacing of (c)56 µm; (d) 42 µm; (e)32 µm.
Figure 8(b)-(e) show the influence of hatch spacing on Stage III, including the CR and SEM images. As the hatch spacing decreases, the CR of Stage III also shows a small-scale fluctuation. Since the scanning speed remains constant, the ablation depth caused by resin decomposition in the laser-irradiated area does not change significantly. While the decrease in hatch spacing only causes a decrease in the width of d in laser-unirradiated areas. In that case, the CR of the grooves does not change drastically. Unlike the scanning speed, the SEM images show that there is no exposed glass fiber on the substrate with a low hatch spacing, indicating that the change of hatch spacing does not significantly impact the ablation depth. From the SEM images, despite a large hatch spacing that can achieve a high-quality peeling surface and high removal efficiency, it will prevent the copper layer in the laser-unirradiated area from being completely peeled off.
3.2.3 Optimization of scanning parameters
As mentioned, the scanning speed is crucial for the improvement of large-area copper layer removal quality. The optimal scanning speed not only eliminates copper residues in the initial segment but also avoids damage to the substrate. From the analysis of the removal process in Sec.3.1, the first scanning line is required to enter Stage II directly to eliminate the edge residue. The optimal scanning speed can be 380 mm/s for the laser condition used in 3.1 and 3.2. For the purpose of quickly finding the optimal scanning speed under different laser conditions, experiments of single-line scanning with different speeds were conducted (with a laser pulse duration of 30 ns, a repetition frequency of 100 kHz, and an energy density of 28.22 J/cm2).
Within a certain speed range, the depth of single-line ablation gradually increases as the scanning speed decreases. This speed range was named as “depth change speed range” (> 420 mm/s, Fig. 9(a)). When the scanning speed decreases, the laser-irradiated energy and the copper layer per absorbed energy per unit area increases. With the presence of heat accumulation and conduction, the high temperature of copper at the resin-copper interface causes resin decomposition and local gas expansion. The shock wave generated by local gas expansion will peel off copper in a small area and expose the substrate. This speed range was defined as “local penetration speed range” ((340 mm/s – 420 mm/s), Fig. 9(b-e)). As the scanning speed continues to decrease, the peeled area in the single-line ablation increases until the copper of the entire scanning line is completely removed. And this speed range at which the copper layer on a single line is completely removed means “complete removal speed range” ((< 340 mm/s), Fig. 9(f)). When the first hatch line scans with the penetration speed, the occurrence of Stage I can be avoided, and the width of Stage II can be diminished. Further, Stage II can be avoided entirely with a lower scanning speed. The optimal scanning speed can be set to the middle of the local penetration speed range.
Fig. 9. (a)-(f)Single-line ablation morphology at different scanning speeds. (g)The width and depth of a single-line scanning with a scanning speed of 420 mm/s.
The optimization of hatch spacing was aimed at avoiding damage to the substrate with a low scanning speed. With the analysis in Sec.3.2.2, the optimal hatch spacing can be selected to 52 µm. Actually, the selection of the hatch spacing needs to be adjusted appropriately based on the scanning results at the optimal speed. Although a large hatch spacing can reduce the damage of the substrate and improve the ablation efficiency, copper will be peeled off as a strip, blocking out the subsequent laser scanning. Generally, considering the removal efficiency and quality, the optimal hatch spacing can be set to twice the width of a single-line scanning (as the results in Fig. 9(g)).
3.3 Analysis of the removal efficiency
3.3.1 Influence of laser conditions on the optimal scanning parameters
The scanning parameters directly affect the process of heat accumulation and conduction after energy absorption. Correspondingly, the laser parameters directly determine the way and amplitude of energy supply. This section focuses on the influence of different laser parameters on the optimal scanning parameters, exploring the possibility of improving ablation efficiency. Figure 10 is the local penetration speed range ((a), (b), and (c), the green speed range) with different laser energy densities, pulse durations, and repetition frequencies.
From Fig. 10(a), with the increase of laser energy density, the local penetration speed range increases, and the width of the range also increases accordingly (with a pulse duration of 30 ns and a repetition frequency of 100 kHz). When the energy density of 19.14 J/cm2, 24.1 J/cm2, 28.22 J/cm2, 33 J/cm2, 36.82 J/cm2, 40.88 J/cm2, the optimal speed can be selected to 230 mm/s, 330 mm/s, 380 mm/s, 450 mm/s, 480 mm/s, 530 mm/s, respectively. As the energy density increases, the energy irradiated and material absorbed increases, causing excessive resin decomposition. Therefore, it is required to reduce the pulse overlapping by increasing the scanning speed. Figure 10(d) shows that as the energy density increases, the single-line width increases and the optimal hatch spacing increases accordingly (The single-line ablation width is not necessarily lower than the focal spot diameter). It should be noted that although the single-line ablation size increases with increasing energy density, it remains almost constant when the energy density is much larger than the damage threshold [25,26]. Moreover, the range of heat conduction with laser irradiation is also limited [27]. As a result, high energy densities have a reduced effect on the optimal hatch spacing.
Fig. 10. Variations of penetration scanning speed with (a)energy density; (b)pulse duration; (c) repetition frequency (The blue speed range represents ablation depth variation, the green speed range represents local penetration, and the red speed range represents complete copper removal.). The single-line ablation width and optimal hatch spacing with different (d)energy density; (e)pulse duration; and (f) repetition frequency. The copper removal results with the optimized process in different (g)energy density; (h)pulse duration; (i)repetition frequency.
The three speed ranges with pulse duration are shown in Fig. 10(b) (with an energy density of 28.22 J/cm2 and a repetition frequency of 100 kHz). As the laser pulse duration increases from 20 ns to 100 ns, the speed range increases from 260 mm/s ∼ 340 mm/s to 560 mm/s ∼ 620 mm/s, respectively. Then, the speed range decreases to approximately 340 mm/s when the pulse duration increases to 200 ns. The laser-irradiated energy in the speed range of different pulse durations shows the opposite trend. This is because the larger the pulse duration, the lower the peak power for the same pulse energy. Furthermore, single line penetration speeds of 150 ns and 200 ns are more likely to cause damage to the substrate due to the increased irradiated time of a single-pulse. Reference [14] also explained this phenomenon: short pulses were more likely to generate plasma that absorbed laser-irradiated energy and led to a decrease in threshold speed; long pulses easily generated vapor that dissipated away the heat in the conductive layer and caused a decrease in threshold speed. In Fig. 10(e), at the same energy density, the pulse duration has less effect on the single-line ablation width, which means that the pulse duration has almost no effect on the optimal hatch spacing.
Figure 10(c) is the speed range with different repetition frequencies (with a pulse duration of 30 ns and an energy density of 28.22 J/cm2). As the laser pulse duration increases from 50 kHz to 300 kHz, the speed range increases from 140 mm/s ∼ 180 mm/s to 1300 mm/s ∼ 1660 mm/s, respectively. The higher repetition frequency increases the number of pulses acting per unit of time, and the energy absorbed by the material increases accordingly. More pulses acting means the enhancement of heat accumulation and heat conduction effect, increasing the penetration speed range and range size. As shown in Fig. 10(f), the repetition frequency does not affect the single-line ablation width. It indicates that the repetition frequency has a limited effect on the range of heat transfer in the hatch direction. Therefore, the optimal hatch spacing remains constant for different repetition frequencies. Figure 10(g) and (h) and (i) show the copper removal results with the optimized process in different energy density, pulse duration and repetition frequency.
3.3.2 Optimal hatch spacing and laser parameters
In order to analyze the copper layer removal efficiency, the time required to remove the copper layer per unit area was calculated, and the results are shown in Fig. 11. Take the rectangular pattern (a mm × b mm) as an example, the total time t required to remove the copper layer of area A can be expressed as:
where v is the scanning speed, and D is the hatch spacing. Neglect the time required for the galvanometer to accelerate and decelerate, the removal efficiency can be represented by:Fig. 11. The removal efficiency with different energy densities, pulse duration, and repetition frequency.
As shown in Fig. 11, with the increase in energy density, pulse duration, or repetition frequency, the removal efficiency of the copper layer can be effectively improved. When the energy density increases from 19.14 J/cm2 to 40.88 J/cm2, the copper layer removal efficiency increases by 3.22 times. When the pulse duration increases from 20 ns to 100 ns, the removal efficiency increases by 1.9 times. When the repetition frequency increases from 50 kHz to 300 kHz, the removal efficiency increases by 9.2 times. Compared to the reported copper removal method of LDWL + WCE and LDA, the removal efficiency of LIAMP is increased by 8.3 times and 66 times, respectively. Furthermore, the copper layer removal efficiency can be further improved by increasing the energy density and repetition frequency at a pulse duration of 100 ns.
4. Conclusions
Conclusively, a laser scanning strategy was proposed for large-area copper removal of FR-4 CCL in one-step scanning. The copper removal mechanism involves a laser-induced active mechanical peeling (LIAMP) generated by resin decomposition. Based on the experimental phenomena and measurements, the copper removal mechanism has been systematically analyzed to clarify the physical-chemical change in the laser scanning process. According to the scanning strategy, the scanning area is divided into laser-irradiated and -unirradiated areas. During the scanning, the laser energy in the irradiated area penetrates the copper layer and induces resin heating and decomposition. Then the copper layer in the laser-unirradiated area is peeled off by the gas micro-explosion generated by resin decomposition. Moreover, it was found that the scanning speed is crucial for the improvement of the copper removal quality. A large scanning speed can lead to significant copper residue in the initial segment of the scanning area, while a low scanning speed will cause severe damage to the substrate. In addition, the optimization of scanning parameters is proposed. Based on this, the influence of different laser conditions on the removal efficiency of this laser-induced peeling mechanism is explored and analyzed. The copper layer removal efficiency is up to 82.8 mm2/ms, and it can be further increased by optimizing laser conditions. So far, the efficiency is 8.3 and 66 times higher than the reported methods of LDWL + WCE and LDA, respectively. The scanning strategy has great promise for the selective removal of electromagnetic functional materials in aerospace applications. Further work will focus more on the applicability of this mechanism to thicker layers, larger pattern sizes, and other complex patterns.
Funding
Wuhan National Laboratory for Optoelectronics (2021WNLOKF017); Natural Science Foundation of Hubei Province (2020CFB423); National Natural Science Foundation of China (52175405).
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. M. Nowak, A. Antończak, P. Kozioł, et al., “Laser prototyping of printed circuit boards,” Opto-Electron. Rev. 21(3), 320–325 (2013). [CrossRef]
2. R. Berenyi, M. Ruszinko, and I. Szekely, “UV laser based direct structuring,” in 2016 39th International Spring Seminar on Electronics Technology (ISSE) (IEEE, 2016), pp. 367–371.
3. O. Klemp, M. Hueske, I. Rolfes, et al., “On the Application of a Novel Laser-Based Rapid Prototyping Process for Planar Antenna Elements,” in 2006 European Microwave Conference (IEEE, 2006), pp. 878–881.
4. B. Lin, W. Huang, J. Guo, et al., “Wideband polarization rotating frequency selective surface composed of antenna–filter–antenna arrays,” Opt. Commun. 530, 129202 (2023). [CrossRef]
5. J. Y. Kim, D. S. Kim, D. W. Woo, et al., “Electromagnetic wave characteristics of second-order composite frequency-selective surfaces fabricated by e-beam evaporator,” Electron. Lett. 50(3), 189–190 (2014). [CrossRef]
6. J. Mur, B. Podobnik, and I. Poberaj, “Laser beam steering approaches for microstructuring of copper layers,” Opt. Laser Technol. 88, 140–146 (2017). [CrossRef]
7. F. Guan, H. Xiao, M. Shi, et al., “The novel frequency selective fabric and application research,” J. Ind. Text. 46(1), 143–159 (2016). [CrossRef]
8. A. Alwaidh, M. Sharp, and P. French, “Laser processing of rigid and flexible PCBs,” Opt. Lasers Eng. 58, 109–113 (2014). [CrossRef]
9. J. Ai, H. Xiang, Q. Du, et al., “Rapid and high-performance processing technology for large-area frequency selective surfaces,” J. Mater. Process. Technol. 264, 129–133 (2019). [CrossRef]
10. R. F. Pease and S. Y. Chou, “Lithography and Other Patterning Techniques for Future Electronics,” Proc. IEEE 96(2), 248–270 (2008). [CrossRef]
11. M. Moallem and K. Sarabandi, “Miniaturized-Element Frequency Selective Surfaces for Millimeter-Wave to Terahertz Applications,” IEEE Trans. Terahertz Sci. Technol. 2(3), 333–339 (2012). [CrossRef]
12. F. Dupont, S. Stoukatch, P. Laurent, et al., “355 nm UV laser patterning and post-processing of FR4 PCB for fine pitch components integration,” Opt. Lasers Eng. 100, 186–194 (2018). [CrossRef]
13. B. Kmetec, D. Kovačič, J. Možina, et al., “High-speed energy efficient selective removal of large area copper layer by laser induced delamination,” Appl. Surf. Sci. 255(20), 8598–8602 (2009). [CrossRef]
14. R. Rozman, B. Kmetec, B. Podobnik, et al., “Optimisation of direct laser structuring of printed circuit boards,” Appl. Surf. Sci. 254(17), 5524–5529 (2008). [CrossRef]
15. F. Dupont, S. Stoukatch, P. Laurent, et al., “Fine pitch features laser direct patterning on flexible printed circuit board,” Opt. Lasers Eng. 126, 105869 (2020). [CrossRef]
16. M. P. Raele, L. R. De Pretto, and D. M. Zezell, “Soldering mask laser removal from printed circuit boards aiming copper recycling,” Waste Manage. (Oxford, U. K.) 68, 475–481 (2017). [CrossRef]
17. G. Yang, Y. Ding, L. Liu, et al., “Mechanism and technology of laser selective removal of multilayer materials,” J. Mater. Process. Technol. 312, 117868 (2023). [CrossRef]
18. B. Yu, V. Till, and K. Thomas, “Modeling of thermo-physical properties for FRP composites under elevated and high temperature,” Compos. Sci. Technol. 67(15-16), 3098–3109 (2007). [CrossRef]
19. F. Sarvar, N. J. Poole, and P. A. Witting, “PCB glass-fibre laminates: Thermal conductivity measurements and their effect on simulation,” J. Electron. Mater. 19(12), 1345–1350 (1990). [CrossRef]
20. J.-Y. Natoli, L. Gallais, H. Akhouayri, et al., “Laser-induced damage of materials in bulk, thin-film, and liquid forms,” Appl. Opt. 41(16), 3156 (2002). [CrossRef]
21. B. Jaeggi, B. Neuenschwander, M. Schmid, et al., “Influence of the Pulse Duration in the ps-Regime on the Ablation Efficiency of Metals,” Phys. Procedia 12, 164–171 (2011). [CrossRef]
22. R. Fang, D. Zhang, H. Wei, et al., “Improved two-temperature model and its application in femtosecond laser ablation of metal target,” Laser Part. Beams 28(1), 157–164 (2010). [CrossRef]
23. N. Tao, G. Chen, L. Fan, et al., “Temperature-dependent material removal during pulsed laser processing of CFRP composites,” Opt. Laser Technol. 144, 107445 (2021). [CrossRef]
24. N. C. Herr, A. E. Gonzales, and G. P. Perram, “Kinetics, evolving thermal properties, and surface ignition of carbon fiber reinforced epoxy composite during laser-induced decomposition,” Polym. Degrad. Stab. 152, 147–161 (2018). [CrossRef]
25. D. V. Tran, H. Y. Zheng, Y. C. Lam, et al., “Femtosecond laser-induced damage morphologies of crystalline silicon by sub-threshold pulses,” Opt. Lasers Eng. 43(9), 977–986 (2005). [CrossRef]
26. K. Venkatakrishnan, P. Stanley, N. R. Sivakumar, et al., “Effect of scanning resolution and fluence fluctuation on femtosecond laser ablation of thin films,” Appl. Phys. A 77(5), 655–658 (2003). [CrossRef]
27. D. Bäuerle, Laser Processing and Chemistry (Springer Berlin Heidelberg, 2011).
