The challenges and focuses on plastic welding by picosecond laser in

ed
the field of camera module assembly
Lin Zhang1, Jincheng Wen1, Nannan Wang2, Han Wu*1, Xiuquan Ma1
1 The State Key Laboratory of Digital Manufacturing Equipment and Technology, School of

iew
Mechanical Science and Engineering, Huazhong University of Science and Technology, Wuhan
430074, China
2 GZ Photonics Technology Co., Ltd., Dongguan, China
* Corresponding author: hanwu@hust.edu.cn.

v
Abstract

re
Polycarbonate (PC) and cyclic olefin copolymers (COC) are extensively employed in optical lenses
due to their exceptional optical properties. However, mechanical clamping forces connecting lenses to
each other and barrel within the camera module assembly (CMA) can lead to image quality degradation.
er
Laser welding of plastics has emerged as an innovative technique poised for addressing this issue.
Picosecond laser welding has successfully achieved joints between PC to PC, PC to COC, and COC
to COC under optical contact (OC) and non-optical contact (NOC) conditions. Comprehensive
pe
analyses of weld zone morphology and mechanical properties have revealed diffusion bonding as the
mechanism for joint formation and brittle fracture as the mechanism for fracture. By contrasting the
requirements of camera module assembly with these plastic joints, the joint bonding strength is deemed
adequate. However, numerous challenges and limitations hinder the advancement of laser plastic
ot

welding within the CMA field. Pertinent suggestions have been provided to address these obstacles,
including the optimization of welding parameters, utilization of a high-NA focal system, and surface
tn

polishing of materials. The implementation of improved laser welding techniques is expected to

significantly contribute to the future development of CMA.
Keywords: plastic welding, polycarbonate, cyclic olefin copolymers, picosecond laser welding,
camera module assembly
rin
ep
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4854305
 1. Introduction
Plastics play vital roles in optics, medical devices, food and packaging field, etc. [1, 2], owing to

ed
their lightweight nature, high specific strength, and other outstanding properties. In the optics sector,
polycarbonate (PC) and cyclic olefin copolymers (COC) are frequently employed due to their high
chemical resistance, optical clarity, and mechanical toughness [3, 4], making them popular choices for

iew
optical lenses. These lenses are typically integrated into optical systems to mitigate spherical,
chromatic, and other aberrations, particularly in camera modules [5, 6] utilized in smartphones,
automobiles, and machine vision applications. As illustrated in Figure 1, a smartphone camera module
comprises barrels, spacers, and lenses. The depicted stacked die configuration of lenses demonstrates

v
how the lenses are held in place by spacers and a barrel. However, this method of lens attachment can
induce stress and distortions on lens surfaces, ultimately leading to degraded image quality and reduced

re
reliability. Therefore, there is a pressing need for innovative techniques to connect lenses, necessitating
further exploration.
In the camera module assembly (CMA), two types of joints, lens to lens and lens to spacers, as
er
depicted in Figure 1, require careful consideration. In recent years, lasers have demonstrated efficacy
in welding various materials including glasses [7-9], metals [10-12], and ceramics [13, 14]. Extensive
investigations into laser welding of plastic components have also been conducted [1, 3, 15-21].
pe
However, industrial products are subject to stringent standards, especially those intended for
specialized environments. To our knowledge, the joints in CMA impose specific requirements:
i) The joint strength must exceed 20 N (at room temperature) to ensure reliability.
ii) The joint should withstand mechanical vibration.
ot

iii) The joint area should be minimized, typically around 150-300 µm or smaller.
iv) Surface distortions induced after joining should be limited to approximately 0.5 μm or less.
tn

Addressing these requirements poses significant challenges and has hindered the widespread
adoption of laser welding techniques in CMA. The laser source plays a pivotal role, with different
types offering distinct functions. Currently, four primary types [17] are employed in welding
rin

thermoplastic components, categorized by wavelength:

1) CO2 lasers (10600 nm) are commonly utilized for thermoplastic laser welding, typically forming
butt joint configurations [22], albeit with limitations in part design applications.
2) Diode lasers (780-980 nm) offer a wide range of radiation wavelengths (808 to 980 nm) and are
ep

suitable for welding various thermoplastics. However, they lack efficient radiation beam
adjustment capabilities, making precise spot size control challenging and limiting their
effectiveness over long working distances [23].
Pr

3) Nd: YAG lasers (532/1064 nm) are often used with absorbing additives in thermoplastic welding
due to their lower absorption by thermoplastic materials.

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4854305
 4) Fiber lasers (1000-21000 nm) are preferred for precision welding of sheets, films, molded parts,
and fabrics, owing to their higher efficiency and superior beam quality [17].

ed
In summary, different laser sources are applied in plastic welding under varying conditions, each
offering specific advantages and limitations. Additionally, pulse duration serves as another critical
parameter influencing the morphology and mechanical properties of plastic welding. Non-ultrafast

iew
lasers [1, 24-26] are absorbed directly along the propagation length of thermoplastic components,
resulting in a widely spread heat-affected zone and imposing thickness limitations on the upper joining
component. In contrast, ultrafast lasers [3, 18-20] typically produce a minute heat-affected zone due
to their ultrashort interaction times and high peak intensities within the focal volume.

v
In our study, we focus on laser transmission welding of COC-PC, PC-PC, and COC-COC using
picosecond laser pulses to achieve lens-lens and lens-spacers joints. Detailed examinations of

re
microstructures, element distribution, and mechanical properties are conducted to assess the feasibility
of laser welding in CMA, thereby elucidating the challenges and focal points encountered during the
laser welding process of lenses to lenses and lenses to spacers.
er
pe
ot

Fig. 1. The schematic diagram of a cross-section of the camera module.

tn

2. Experimental procedure
2.1 Materials
In this study, APEL 5014 (COC: size of 30×15 mm) and EP 6000 (PC: size of 30×15 mm), both
rin

with a thickness of 1 mm, are selected for processing due to their wide applications in optical lenses.
Additionally, the commonly used material for optical lens spacers, PC 1225Y, is employed in this
work with the same dimensions. As depicted in Figure 2, these materials are covered by different color-
protective films, with PC 1225Y serving as the black substrate. Considering the application scenarios
ep

of COC lenses, PC lenses, and PC spacers, the following joint groups are established: APEL 5014 to
EP 6000 (A5E6), APEL 5014 to APEL 5014 (A5A5), EP 6000 to EP 6000 (E6E6), EP 6000 to PC
1225Y (E6P1), and APEL 5014 to PC 1225Y (A5P1).
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4854305
 ed
iew
Fig. 2. The picture of plastics used in the experiments.
2.2 Laser system

v
The picosecond laser operates at a wavelength of 1064 nm, a maximum average power of 50 W,

re
100 kHz Gaussian laser pulses with width < 15 ps, and the beam quality M2<1.3. The output laser
beam has a diameter of approximately 1.5 mm and is collimated through a tenfold beam expander
before reaching the focal system.
2.3 Process setup er
The welding system's schematic configuration, depicted in Fig. 3, comprises a picosecond laser, a
pressure sensor, a lifting platform, an arch-shaped setup, and XYZ-translation stages. The top side of
the arch-shaped setup features machined recesses, within which a designed glass is securely mounted.
pe
After passing through a 1-inch plano-convex lens (with a focal distance of 25 mm), the picosecond
laser is directed through the glass to focus on the interface of the welding materials.
As illustrated in the inset of Fig. 3, an APEL 5014 is stacked atop an EP 6000. The A5E6 sample
rests on a 30x30x1 mm silicone pad, followed by a ceramics plate. The silicone pad serves to mitigate
ot

pressure unevenness on the A5E6 surfaces, while the ceramics plate isolates thermal diffusion from
the sample to the pressure sensor. The lifting platform exerts significant clamping force between the
tn

APEL 5014 and EP 6000, with force monitored by the pressure sensor. These components are mounted
on an XYZ-translation stage, with scan speeds ranging from 1 to 10 mm/s. Ultimately, picosecond
pulses pass through the transparent APEL 5014 and focus on the interface between the APEL 5014
rin

and EP 6000, with the zero position for the focal plane defined at the APEL 5014-EP 6000 interface.
ep
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4854305
 ed
v iew
re
Fig. 3. The schematic diagram of the picosecond laser welding system, inset shows a welding
case.
2.4 Mechanical testing and characterization of bonding areas.
er
The welding samples underwent characterization utilizing optical microscopy, scanning electron
microscopy (SEM) equipped with energy dispersive X-ray spectroscopy (EDS) for microstructure
analysis and elemental composition determination of the welding seams. Shear-strain tests were
pe
conducted using a shear test machine (AG-IC 100kN, SHIMADZU, Japan). Following the shear test,
the morphologies of fracture sample surfaces were analyzed using SEM and laser confocal microscopy
(VK-X200; Keyence Co.; Japan). Mechanical vibration was measured using an all-electric dynamic
test instrument (E1000, Instron, America).
ot

3. Results
3.1 Morphology and characteristics of weld zones
tn

The A5E6/A5A5/E6E6 samples were produced using a laser power of 2 W, a focal plane of 0 mm,
a scan speed of 1.5 mm/s, and a clamping force of approximately 140 N. The E6P1/A5P1 samples
were achieved by increasing the scan speed to 4 mm/s to account for the enhanced absorption of laser
rin

energy due to the presence of carbon black doping in PC 1225Y [2, 27-30]. Representative samples of
A5E6 and E6P1 are presented in Fig. 4a and b, respectively.
For the A5E6 samples, the formation of a keyhole was prioritized upon a laser pulse irradiation,
followed by the formation of multiple keyholes that created a continuous ablated region in the weld
ep

center in Fig. 4 c. Surrounding the ablated regions were molten regions with a width of 260 μm, and
the molten borders were distinguishable from the base materials. These molten regions were
recognized as domain-connected areas. Similarly, the welding seam of E6P1 comprised ablated
Pr

regions and molten regions with a diameter of 670 μm. The molten region in the E6P1 samples was
more pronounced than in the A5E6 samples, indicating stronger bonding strength in E6P1. Thus, the

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4854305
 use of a laser absorber is common practice to achieve high-strength weld joints between two
transparent plastic parts [30, 31]. Additionally, such absorbers can be tailored by incorporating suitable

ed
additives, such as pigments or absorbers, depending on specific application requirements [17].

v iew
re
Fig. 4. (a) A5E6 samples; (b) E6P1 samples; Optical images of the welding seam. (c) A5E6; (d)
er E6P1.
To further explore the characteristics of the joints, cross-sectional analyses of the weld seams of
A5E6 and E6P1 were conducted, as depicted in Fig. 5 and Fig. 6, respectively. In Fig. 5 a, SEM
pe
imaging revealed ablation occurring on both the APEL 5014 and EP 6000 sides, resulting in an oval-
shaped ablated region with a width of 56 μm and a length of 74 μm. Molten regions were observed
surrounding the ablated region, as shown in Fig. 5 a, with elemental analysis indicating a composition
rich in C and O elements throughout the entire area in Fig. 5 b and c. Notably, the molten regions
ot

exhibited complete mixing of the two different polymers without any defects such as cracks or pores.
Furthermore, EDS analysis in Fig. 5 d of a specific point in Fig. 5a confirmed the presence of carbon-
containing fragments within the ablated region. This carbonization is attributed to the photochemical
tn

interactions between the polymers and the laser, with the decomposition of the polymer generating
small gaseous molecules such as CO and CO2 [32].
The cross-section of the weld seam in E6P1 (Fig. 6 a) revealed numerous porous defects within the
rin

weld area. These defects were characterized by a central ablated region surrounded by micropores,
which were formed due to laser ablation and the trapping of gases within the molten region. The
viscosity of the melted polymer hindered the overflow and fusion of these pores. Elemental analysis
ep

in Fig. 6 b and c demonstrated the distribution of C and O elements throughout the weld seam cross-
section, indicating appropriate bonding between EP 6000 and PC 1225Y. Additionally, an EDS line
scan from the EP 6000 side to the PC 1225Y side in Fig. 6 d revealed a consistent C:O ratio across the
Pr

base materials, approximately 9:2. At the edges of the pores (Fig. 6 d), there was a sudden increase in
carbon concentration while the oxygen concentration remained unchanged, indicating the presence of
carbonization in these areas.

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4854305
 ed
v iew
re
Fig. 5. SEM images and EDS results of A5E6. (a) Weld cross-section; (b) C elemental distributional
map; (c) O elemental distributional map; (d) EDS of the point one in (a).
er
pe
ot
tn
rin

Fig. 6. SEM images and EDS results of E6P1. (a) Weld cross-section; (b) C elemental distributional
map; (c) O elemental distributional map; (d) EDS of the red line.
3.2 Mechanical properties of joints
The shear strengths were determined using a compressive shear method, as illustrated in Fig. 7 a,
ep

with the samples vertically clamped to ensure that the pressure direction remained parallel to the
interface. Shear stress was calculated as the ratio between the compressive force and the welding area
(width × length). The shear strengths of A5E6, A5A5, E6E6, E6P1, and A5P1 were depicted in Fig.
Pr

7b, with each group measured at least three times. E6E6 exhibited a higher shear strength compared
to A5E6 and A5A5, measuring 17.63 MPa, 14.42 MPa, and 11.97 MPa, respectively. This disparity

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4854305
 can be attributed to the stronger tensile strength of polycarbonate (PC) compared to cyclic olefin
copolymers (COC) [33], which likely explains why the E6E6 (PC-PC) joint surpassed the A5A5

ed
(COC-COC) joint significantly. The shear strength of E6P1 closely matched that of A5P1, but both
were twice that of the transparent-transparent welding samples (A5E6/A5A5/E6E6), demonstrating
that the laser absorber in PC 1225Y can modify the molten regions and enhance bonding strengths.

iew
The mechanical vibrations test in Fig. 7 c was conducted under 10 Hz and at least 20 N load cycling.
This test simulated real-world scenarios where joints must withstand mechanical vibrations in
applications of CMA. The load-time curves of A5E6 and E6P1 presented in Fig. 7 d demonstrated the
excellent tolerance of joints to alternating loads and mechanical vibration, meeting the bonding

v
strength requirements for CMA.

re
er
pe
ot
tn
rin

Fig. 7. (a) The schematic of the compressive shear test; (b) the shear strengths of A5E6, A5A5,
ep

E6E6, E6P1, and A5P1; (c) the schematic of mechanical vibration test; (d) the load-time curves of
A5E6 and E6P1 under a 10 Hz force cycling.
3.3 Fracture analysis of weld samples
Fig. 8 presented the fracture surfaces of A5E6 and E6P1 after tensile testing, revealing ablated
Pr

craters and molten regions comprising the entire welded area in Fig. 8 a and b. The multi-ablated
craters induced by laser pulses were continuously connected to form an ablated region. On the fractured

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4854305
 surface of APEL 5014 in Fig. 8 a and c, fibrils and microfibrils were observed around the ablated
craters. These fibrils exhibited an elongated-growing direction parallel to the shear force direction,

ed
indicating that the A5E6 joints underwent plastic deformation before fracture [34]. Additionally, rib
marks [35] were observed around the ablated craters, representing cleavage planes separated from the
EP 6000 surfaces. The presence of numerous cleavage planes suggests brittle fracture in the A5E6

iew
samples.
Meanwhile, the fracture surface of PC 1225Y separated from E6P1 was shown in Fig. 8 b and d,
with a few microfibrils observed at the molten regions in Fig. 8 d. However, numerous dimples
appeared on the molten regions in Fig. 8 d, indicating superior ductility of the E6P1 samples compared

v
to the A5E6 samples. Moreover, the presence of dimples with numerous cleavage planes suggests that
E6P1 experienced plastic deformation at the beginning of the shear test followed by brittle fracture.

re
Ultimately, the fracture mechanism of both A5E6 and E6P1 samples is identified as brittle fracture.

er
pe
ot
tn

Fig. 8. SEM images of fracture morphology. (a) APEL 5014 surface of A5E6; (b) PC 1225Y
surface of E6P1; (c) high-magnification image of red square area in (a); (d) high-magnification
rin

image of red square area in (b).

To further examine the fractured positions of the joint samples, 3D height maps of the fractured
surfaces of A5E6 and A5P1 are illustrated in Fig. 9. The fractured positions of both A5E6 and A5P1
were found to be adjacent to the bottom layers (EP 6000 and PC 1225Y), suggesting an interface failure
ep

[2] as the joint strength is weaker than the tensile strength of the base materials [33]. This indicates a
preference for a diffusion bonding mechanism in the joints.
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4854305
 ed
v iew
re
er
Fig. 9. Height map of the fracture surfaces. (a) Top surface of A5E6; (b) bottom surface of
pe
A5E6; (c) Top surface of A5P1; (d) bottom surface of A5P1.
3.4 Effect of welding parameters
Numerous welding parameters can influence the heat input to the interface during the process,
including laser power, welding/scan speed, focal plane, laser spot size, laser working distance, and
ot

others [2, 16, 17, 36, 37]. Due to the small joint region in camera module assemblies (CMA), it is
essential to control weld widths as small as possible. In our study, the heat input was determined by
tn

laser power, scan speed, and focal plane, with the clamp force maintained around 140 N.
Fig. 10 illustrates the effects of the focal plane, laser power, and scan speed on the weld widths of
A5E6, with weld widths identified by ablated and molten regions. As the focal planes moved from
rin

negative to positive positions in Fig. 10 a, the diameters of the ablated and molten regions decreased
and then increased, resulting in a small and stable welding width of around 420 μm within a focal plane
of -20 to 20 μm. Outside this regime, the widths of the ablated and molten regions showed higher
standard deviations, indicating a less stable process. Furthermore, the focal plane has a more significant
ep

effect on the molten region than the ablated region.

In Fig. 10 b, the widths of the ablated and molten regions continued to increase with increasing laser
power, and both low and high laser powers can lead to instability (as indicated by increasing standard
Pr

deviation). Very low laser power may fail to heat the polymers to their glass transition or melting range
temperatures, resulting in a lack of fusion and unstable welding. Conversely, very high laser power

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4854305
 can induce overheating, burning, and decomposition of the base materials [2]. The appropriate range
of laser power in our experiments was determined to be 1.4 to 1.6 W. In Fig. 10 c, the widths of the

ed
ablated and molten regions decreased as scan speed increased from 2 to 4 mm/s, suggesting that higher
scan speeds reduce interaction time and heat input to the polymer component, resulting in width
reductions. Therefore, weld widths can be optimized by controlling the focal plane at the interface,

iew
selecting appropriate laser power, and increasing welding speed.

v
re
er
pe
ot
tn
rin
ep
Pr

Fig. 10. Effect of welding parameters on weld widths of A5E6. (a) Focal plane, at a scan speed of
3mm/s and a laser power of 1 W; (b) laser power, at a scan speed of 3mm/s and the focal plane of -20

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4854305
 μm; (c) scan speed, at a laser power of 1 W and the focal plane of +20 μm.
4. Discussion

ed
4.1 Influences of optical and non-optical contact
The gaps between two base materials have been a critical factor directly affecting joint strength and
joint area morphologies [2, 38]. It is essential for the base materials to be in good contact to ensure

iew
that both components can melt and touch each other forming a weld seam. Both optical contact (OC)
and non-optical contact (NOC) were achieved, and the cross-sections of A5E6 and A5P1 weld areas
were exhibited in Fig. 11.
In NOC welding, as shown in Fig. 11 a inset, the erupted-ablated products at the borders of the

v
molten region were clearly observed. Conversely, in OC welding in Fig. 11 b inset, the erupted-ablated
products, and even the molten regions, were hardly discernible. It was also noticeable that the ablated

re
regions are oval-shaped for NOC and rod-shaped for OC (in Fig. 11 a and b, respectively), indicating
that the laser power is more concentrated in OC than in NOC welding. Additionally, OC tends to limit
ablation within the focal volume, and the contact regions can transfer more heat than NOC, leading to
er
wider and deeper molten regions [38]. In Fig. 11 c, the NOC of A5P1 exhibited an air layer at the
interface, which may cause laser scatter and numerous bubbles/pores during the welding process.
Conversely, OC welding in Fig. 11 d only generated a main rod-shaped ablation without pores, and
pe
the molten region was wider and greater than in NOC welding.
ot
tn
rin
ep
Pr

Fig. 11. Cross-section of the weld area with inserting an optical image of the weld seam. (a)
A5E6 with interface gap; (b) A5E6 without interface gap; (c) E6P1 with interface gap; (d)

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4854305
 E6P1 without interface gap.
4.2 Forming process of the joint

ed
To comprehensively analyze the joint formation under OC and NOC conditions, schematic diagrams
of the melting and solidification process in welding are shown in Fig. 12. As depicted in Fig. 12 a1,
the A5E6 samples initially have an air gap between two base material surfaces, and a picosecond laser

iew
pulse is transmitted through an APEL 5014 and focused on the interface. Within a time scale of
hundreds of picoseconds [39], ablations occur on both the top and bottom surfaces through
photochemical and photothermal processes in Fig. 12 a2 [40]. These ablations are attributed to
ultrahigh-pressure waves, micro-explosions [20], gas eruption, and pyrolysis products, indicating that

v
the laser beam raises the temperature above the degradation temperature of the base material [41, 42].
Subsequently, within a time scale of nanoseconds, the laser energy absorbed in both the top and bottom

re
parts is transformed into heat energy, locally increasing the temperature above the melting point of the
base material [28]. The melting materials then start flowing around the ablated region radially and
erupting from top to bottom for APEL 5014 and in reverse for EP 6000 (Fig. 12 a2). Some pyrolysis-
er
product carbon (observed in Fig. 5 d) also erupts with the melting materials and spreads at the melting
region borders under a large air layer condition (Fig. 12 a2). Then, the melting materials of APEL 5014
and EP 6000 contact each other, allowing segmental motion, intermixing of polymer chains, and the
pe
formation of entanglements [1, 2, 17]. Eventually, within a time scale of microseconds upon cooling
(Fig. 12 a3), the air gap is bridged by the solidification of the melting materials, resulting in an oval-
ablated region inside the molten regions.
Regarding the OC (small gap) welding depicted in Fig. 12 b, the joint processes are remarkably like
ot

NOC welding. However, the laser energy is more concentrated on the focal volume and less scattered
by the air gap, leading to a deeper ablated region and broader molten region in Fig. 12 b3. Besides, the
tn

OC condition hinders the eruption of melting materials and pyrolysis products outside the molten
region (Fig. 12 b2), explaining the disappearance of the molten borders observed in Fig. 11 b inset.
Similarly, the NOC and OC welding processes of A5P1/E6P1 can also be presented using Fig. 12 a
rin

and b, respectively. Overall, the NOC welding of the A5P1 sample, as demonstrated in Fig. 11 c, has
resulted in numerous pores due to an air layer and strong laser absorption, and the laser scattering
among the gap may have an important effect on the porous structures.
ep
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4854305
 ed
v iew
re
Fig. 12. Schematic diagram for plastic joint processes. (a1) initial status of NOC laser
welding; (a2) the ablation and meltdown of NOC laser welding; (a3) solidification of NOC
laser welding; (b1) initial status of OC laser welding; (b2) the ablation and meltdown of OC
laser welding; (b3) solidification of OC laser welding.
er
4.3 Challenges and focuses of laser welding in CMA
The success of transparent-to-transparent and transparent-to-nontransparent plastic joints in our
pe
experiments is noteworthy. However, further discussion is warranted regarding the applicability of
these joints to industrial products in CMA.
4.3.1 Mechanical property versus applications in CMA
Investigating bonding strength is imperative, particularly for lens-lens and lens-spacer joints, which
ot

require a minimum of 20 N. Considering that lens diameters typically range from 5 to 8 mm, with an
assumed weld width of 150 μm, the estimated bonding strengths required range from 5.3 to 8.5 MPa.
In Fig. 7 b, the shear strength of A5A5 exhibits a minimum value among the joint groups, at 11.97
tn

MPa, exceeding the required bonding strength. Furthermore, the dynamic-load test of weld samples in
Fig. 7 d demonstrates their tolerance to mechanical vibration. Overall, the mechanical properties of
plastic joints produced by picosecond laser welding are found to be suitable for applications in CMA.
rin

4.3.2 Weld width versus applications in CMA

The weldable areas of lens-to-lens or lens-to-spacer joints are constrained by the design of the lens
and barrel, with the diameter of clamping areas typically around 500 μm or even smaller, as illustrated
ep

in the dotted areas of Fig. 1. Therefore, a suggestion to optimize the weld width, mentioned in section
1, is around 150-300 µm or less. In ultrafast laser welding of plastics [3, 18-20], the weld seam contains
ablated and molten regions, corresponding to the results in Fig. 8 and Fig. 9. The ablated regions are
directly influenced by the laser spot size, with the diameter of the ablated crater generally smaller than
Pr

the laser spot size [43-45]. Additionally, the overlap of laser beams [46] and pulse energy [47, 48] can
also determine the widths of the ablated region. The widths of the ablated region in Fig. 10 are

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4854305
 identified to be around 200 μm, limited by the focal system (a 25 mm focal distance plano-convex lens
used in our experiments). To minimize the ablated region widths, it is recommended to utilize a high-

ed
NA objective lens [49, 50] in the welding system due to its strong focusing capability. Furthermore,
the effect of welding parameters on ablated widths has been discussed in detail in Fig. 10, revealing
that ablated widths are affected by laser power and welding speed. The results in Fig. 10 demonstrate

iew
that less energy input (low laser power and high scan speed) during the welding process decreases the
ablated width. Accordingly, the ablated width obtained in our experiments may not meet the
requirements in CMA, and further studies on shrinking the ablated regions would be conducted by
using lower laser power, higher welding speed, and a high-NA focal system.

v
The molten region plays a crucial role in both directly impacting joint strength and determining weld
width. While expanding the molten region can notably enhance bonding strength [51], the

re
requirements of CMA mandate keeping the molten region width around 150-300 µm or even smaller.
Therefore, achieving an optimal balance between molten width and joint strength is imperative for
CMA application. Furthermore, it is noted in section 3.4 that molten region widths are highly sensitive
er
to laser power, welding speed, and focal plane. However, as observed in Fig. 10, the minimum molten
region width is approximately 420 μm, posing challenges in meeting CMA requirements. Similar to
optimizing ablated width, further refinement of molten region widths can be achieved through
pe
meticulous adjustment of welding parameters and leveraging a high-NA focal system. Additionally,
some research [3] has explored the use of femtosecond laser systems for plastic welding to attain higher
bonding strength and smaller weld widths.
4.3.3 Stability and surface distortion versus applications in CMA
ot

Welding stability and surface distortions are critical considerations for applications in CMA. Fig.
13 showed joint samples with rectangular welding seams under different conditions. In Fig. 13 a,
tn

closed-loop welding of A5E6 with a gap layer (NOC) resulted in a weld seam break. Welding with a
gap layer proved challenging to control and monitor due to difficulties in maintaining uniform gap
width and uniform-distributed clamping pressure. This instability led to breaks and unsuccessful welds.
rin

In contrast, OC welding of A5E6 and A5P1 achieved closed-loop welding, as depicted in Fig. 13 b
and c, respectively. In these experiments, OC welding was ensured by applying a clamping force
exceeding 300 N to reduce the interface gap to less than λ/4 [52-54], where λ represents the wavelength.
OC welding emerged as a reliable and valuable technique for CMA applications.
ep

However, OC interface distortion occurred upon releasing the strong clamping force, as shown in
Fig. 13 d (similarly observed in [55]). To mitigate the adverse effects of strong clamping forces causing
surface distortions and NOC conditions resulting in weld instability, it is proposed to polish the
Pr

welding surfaces of base materials and clamp the two polished surfaces with soft forces [2, 55]. It is
suggested that clamping forces may not be necessary for laser welding of plastics, as Van der Waals

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4854305
 forces [56, 57] between polished surfaces could facilitate OC welding without inducing surface
distortions. Further research in this direction could lead to improved welding techniques for CMA

ed
applications.

v iew
re
er
Fig. 13. Optical images of welding samples. (a) NOC welding of A5E6 with breaks;(b) OC
pe
welding of A5E6; (c) OC welding of A5P1; (d) cross-section with distortions of A5E6 by OC
welding.
5. Conclusions
In this study, successful joints between PC to PC, PC to COC, and COC to COC were achieved
ot

utilizing a picosecond laser under both OC and NOC conditions. Extensive analysis of the mechanical
properties of joints and morphologies of welding areas, aligned with requirements in CMA, facilitated
tn

a comprehensive investigation and discussion of the welding process mechanism, challenges, and
focuses of plastic laser welding. The key conclusions are summarized as follows:
1) Welding of APEL 5014 to EP 6000 (A5E6), APEL 5014 to APEL 5014 (A5A5), EP6000 to EP
6000 (E6E6), APEL 5014 to PC 1225Y (A5P1), and EP 6000 to PC 1225Y (E6P1) using
rin

picosecond laser pulses was achieved through optimization of laser power, focal plane, scan
speed, and clamping pressure. Analysis revealed the formation of an oval-ablated region
surrounded by molten regions, with pyrolysis-product carbon present. The joint mechanism
ep

involves melting polymer chains, intermixing, and entanglements, leading to diffusion bonding
upon solidification.
2) OC and NOC welding exhibit distinct differences in weld seam morphology and ablated region
Pr

characteristics. OC welding shows a rod-shaped ablated region and a wider, deeper molten
region due to concentrated laser energy at the interface. In contrast, NOC welding displays an
oval-shaped ablated region and visible molten border, attributed to the eruption of pyrolysis

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4854305
 products and melting materials into the air gap. Additionally, NOC welding of A5P1/E5P1
results in numerous pores around the ablated region due to laser scattering and strong laser

ed
energy absorption.
3) The shear strength of E6E6 samples is approximately 17.63 MPa, stronger than A5E6 and A5A5
samples. A5P1 and E6P1 exhibit high bonding strengths (~33 MPa) attributed to increased laser

iew
energy absorption by carbon black in PC 1225Y, leading to a wider molten region. Besides, the
fracture mechanism is characterized by brittle fracture.
4) For application in CMA, bonding strength is sufficient for securing lenses in the barrel and
withstanding mechanical vibrations. However, the minimum weld width (~400 μm) is

v
inadequate for utilization in CMA weldable areas (~150-300 μm, or smaller). Recommendations
include optimizing laser power, and welding speed, and utilizing high-NA focal systems or even

re
femtosecond laser systems to ensure minimal weld width. NOC welding is unstable and leads to
weld breaks, while OC welding is reliable but may cause surface distortion upon the release of
strong clamping force. Surface polishing before welding, with or without soft clamping pressure,
er
is recommended to mitigate distortion around weld zones, as polished surfaces facilitate optical
contact through Van der Waals force.
Despite successful plastic welding in CMA using picosecond lasers, numerous challenges hinder
pe
further developments in laser welding for CMA applications. Extensive research and technical
breakthroughs are urgently needed to address these challenges in the future.
ot
tn
rin
ep
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4854305
 Funding
National Key R&D Program of China [2023YFB4606300]; National Science Foundation of China

ed
(NSFC) [81927805]; National Natural Science Foundation of China (NSFC) Basic Science Center
Program [52188102].

iew
Declarations
Conflict of Interest 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.

v
re
er
pe
ot
tn
rin
ep
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4854305
 Reference
[1] Z. Chen, Z. Wu, S. Lu, G. Zhang, H. Yan, A new laser welding method of two transparent plastic parts based on

ed
reflector assisted method, Journal of Materials Research and Technology 24 (2023) 5309-5320,
https://doi.org/10.1016/j.jmrt.2023.04.108.
[2] B. Acherjee, Laser transmission welding of polymers – A review on welding parameters, quality attributes, process
monitoring, and applications, Journal of Manufacturing Processes 64 (2021) 421-443,

iew
https://doi.org/10.1016/j.jmapro.2021.01.022.
[3] G.-L. Roth, S. Rung, R. Hellmann, Ultrashort pulse laser micro-welding of cyclo-olefin copolymers, Optics and
Lasers in Engineering 93 (2017) 178-181, https://doi.org/10.1016/j.optlaseng.2017.02.006.
[4] G. Khanarian, Optical properties of cyclic olefin copolymers, Optical Engineering 40 (2001) 1024-1029,
https://doi.org/10.1117/1.1369411.

v
[5] A. Chowdhury, R. Darveaux, J. Tome, R. Schoonejongen, M. Reifel, A.D. Guzman, S.S. Park, Y.W. Kim, H.W. Kim,
Challenges of megapixel camera module assembly and test, IEEE 2005 Electronic Components and Technology

re
Conference (2005) 1390-1401.
[6] J.-e. Luan, X. Baraton, W. Wong, K.-y. Goh, D. Yap, Camera module design, assembly and test challenges, IEEE
2008 10 th Electronics Packaging Technology Conference (2008) 663-668.
[7] I. Miyamoto., K. Cvecek., M. Schmidt., Advances of Laser Welding Technology of Glass -Science and Technology,
er
Journal of Laser Micro/Nanoengineering 15 (2020), https://doi.org/10.2961/jlmn.2020.02.1001.
[8] X. Jia, K. Li, Z. Li, C. Wang, J. Chen, S. Cui, Multi-scan picosecond laser welding of non-optical contact soda lime
glass, Optics & Laser Technology 161 (2023), https://doi.org/10.1016/j.optlastec.2023.109164.
[9] K. Sleiman, M. Legutko, K. Rettschlag, P. Jäschke, L. Overmeyer, S. Kaierle, CO2 laser based welding of
pe
borosilicate glass by Laser Glass Deposition, Procedia CIRP 111 (2022) 466-469,
https://doi.org/10.1016/j.procir.2022.08.068.
[10] X. Lv, L. Liu, Microstructure and mechanical performance of AZ31/6061 lap joints welded by laser-TIG hybrid
welding with Zn-Al alloy filler metal, Journal of Magnesium and Alloys (2023),
https://doi.org/10.1016/j.jma.2023.01.009.
ot

[11] Z. Zhang, J. Huang, J. Fu, P. Nie, S. Zhang, Microstructure and mechanical properties of laser welded-brazed
titanium/aluminum joints assisted by titanium mesh interlayer, Journal of Materials Processing Technology 302
(2022), https://doi.org/10.1016/j.jmatprotec.2022.117502.
tn

[12] T. Xu, S. Zhou, X. Ma, H. Wu, L. Zhang, M. Li, Significant reinforcement of mechanical properties in laser welding
aluminum alloy with carbon nanotubes added, Carbon 191 (2022) 36-47,
https://doi.org/10.1016/j.carbon.2022.01.033.
[13] E.H. Penilla, L.F. Devia-Cruz, A.T. Wieg, P. Martinez-Torres, N. Cuando-Espitia, P. Sellappan, Y. Kodera, G.
rin

Aguilar, J.E. Garay, Ultrafast laser welding of ceramics, Science 365 (2019) 803-808,
https://doi.org/10.1126/science.aaw6699.
[14] Q. Zuo, Y. Zhang, P. Li, C. Chen, Effect of process parameters on crack characteristics in laser welding of Al2O3
ceramic, Ceramics International 49 (2023) 32440-32447, https://doi.org/10.1016/j.ceramint.2023.07.190.
ep

[15] M.S. Haque, K. Moeed, M.Z. Zaka, Laser welding of thermoplastics – A review, Materials Today: Proceedings 64
(2022) 1479-1485, https://doi.org/10.1016/j.matpr.2022.04.900.
[16] J. Wu, S. Lu, H.-J. Wang, Y. Wang, F.-B. Xia, W. Jin, A review on laser transmission welding of thermoplastics,
The International Journal of Advanced Manufacturing Technology 116 (2021) 2093-2109,
Pr

https://doi.org/10.1007/s00170-021-07519-z.
[17] L.F.F.F. Gonçalves, F.M. Duarte, C.I. Martins, M.C. Paiva, Laser welding of thermoplastics: An overview on lasers,
materials, processes and quality, Infrared Physics & Technology 119 (2021),

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4854305
 https://doi.org/10.1016/j.infrared.2021.103931.
[18] R. Gian-Luca, E. Cemal, H. Ralf, A New Approach to Seal Polymer Microfluidic Devices Using Ultrashort Laser

ed
Pulses, Journal of Laser Micro/Nanoengineering 14 (2019), https://doi.org/10.2961/jlmn.2019.01.0009.
[19] G.-L. Roth, S. Rung, R. Hellmann, Welding of transparent polymers using femtosecond laser, Applied Physics A
122 (2016), https://doi.org/10.1007/s00339-016-9605-x.
[20] A. Volpe, F. Di Niso, C. Gaudiuso, A. De Rosa, R.M. Vazquez, A. Ancona, P.M. Lugara, R. Osellame, Welding of

iew
PMMA by a femtosecond fiber laser, Opt Express 23 (2015) 4114-4124, https://doi.org/10.1364/OE.23.004114.
[21] N. Amanat, C. Chaminade, J. Grace, D.R. McKenzie, N.L. James, Transmission laser welding of amorphous and
semi-crystalline poly-ether–ether–ketone for applications in the medical device industry, Materials & Design 31
(2010) 4823-4830, https://doi.org/10.1016/j.matdes.2010.04.051.
[22] G. Casalino, E. Ghorbel, Numerical model of CO2 laser welding of thermoplastic polymers, Journal of Materials

v
Processing Technology 207 (2008) 63-71, https://doi.org/10.1016/j.jmatprotec.2007.12.092.
[23] G.V. Moskvitin, A.N. Polyakov, E.M. Birger, Laser welding of plastics (review), Welding International 27 (2013)

re
725-734, https://doi.org/10.1080/09507116.2012.753282.
[24] S. Ruotsalainen, P. Laakso, V. Kujanpää, Laser Welding of Transparent Polymers by Using Quasi-simultaneous
Beam Off-setting Scanning Technique, Physics Procedia 78 (2015) 272-284,
https://doi.org/10.1016/j.phpro.2015.11.038.
[25] I. Mingareev, F. Weirauch, A. Olowinsky, L. Shah, P. Kadwani, M. Richardson, Welding of polymers using a 2μm
er
thulium fiber laser, Optics & Laser Technology 44 (2012) 2095-2099,
https://doi.org/10.1016/j.optlastec.2012.03.020.
[26] X. Wang, D. Guo, G. Chen, H. Jiang, D. Meng, Z. Yan, H. Liu, Thermal degradation of PA66 during laser
pe
transmission welding, Optics & Laser Technology 83 (2016) 35-42, https://doi.org/10.1016/j.optlastec.2016.02.022.
[27] B. Geißler, T. Laumer, A. Wübbeke, T. Frick, V. Schöppner, M. Schmidt, Analysis of the Weld Seam Morphology
of Polypropylene in Laser Transmission Welding, Journal of Manufacturing Science and Engineering 140 (2018)
111017, https://doi.org/10.1115/1.4040876.
[28] B. Acherjee, Laser transmission welding of dissimilar plastics: 3-D FE modeling and experimental validation,
ot

Welding in the World 65 (2021) 1429-1440, https://doi.org/10.1007/s40194-021-01079-2.

[29] P.J. Bates, T.B. Okoro, M. Chen, Thermal degradation of PC and PA6 during laser transmission welding, Welding
in the World 59 (2014) 381-390, https://doi.org/10.1007/s40194-014-0209-9.
tn

[30] T.B. Juhl, J.d. Christiansen, E.A. Jensen, Investigation on high strength laser welds of polypropylene and
high‐density polyethylene, Journal of Applied Polymer Science 129 (2013) 2679-2685,
https://doi.org/10.1002/app.39000.
[31] M. Aden, V. Mamuschkin, A. Olowinsky, Influence of carbon black and indium tin oxide absorber particles on laser
rin

transmission welding, Optics & Laser Technology 69 (2015) 87-91, https://doi.org/10.1016/j.optlastec.2014.12.015.

[32] T. Lippert, J.T.J.C.R. Dickinson, Chemical and spectroscopic aspects of polymer ablation: Special features and
novel directions, Chemical Reviews 103 (2003) 453-486, https://doi.org/10.1021/cr010460q.
[33] A. Agha, W. Waheed, N. Alamoodi, B. Mathew, F. Alnaimat, E. Abu‐Nada, A. Abderrahmane, A. Alazzam, A
ep

Review of Cyclic Olefin Copolymer Applications in Microfluidics and Microdevices, Macromolecular Materials and
Engineering 307 (2022), https://doi.org/10.1002/mame.202200053.
[34] W. Grellmann, B. Langer, Deformation and fracture behaviour of polymer materials, Cham: Springer International
Publishing., 2017.
Pr

[35] E. Ghorbel, I. Hadriche, G. Casalino, N. Masmoudi, Characterization of Thermo-Mechanical and Fracture Behaviors
of Thermoplastic Polymers, Materials (Basel) 7 (2014) 375-398, https://doi.org/10.3390/ma7010375.
[36] N. Kumar, N. Kumar, A. Bandyopadhyay, A State-of-the-Art Review of Laser Welding of Polymers - Part I:

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4854305
 Welding Parameters, Welding Journal 100 (2021) 221-228, https://doi.org/10.29391/2021.100.019.
[37] B. Acherjee, A.S. Kuar, S. Mitra, D.J.L.i.M. Misra, M. Processing, Empirical modeling and multi-response

ed
optimization of laser transmission welding of polycarbonate to ABS, Lasers in Manufacturing and Materials
Processing 2 (2015) 103-123, https://doi.org/10.1007/s40516-015-0009-0.
[38] B. Acherjee, Laser transmission welding of polymers – A review on process fundamentals, material attributes,
weldability, and welding techniques, Journal of Manufacturing Processes 60 (2020) 227-246,

iew
https://doi.org/https://doi.org/10.1016/j.jmapro.2020.10.017.
[39] M.V. Shugaev, C. Wu, O. Armbruster, A. Naghilou, N. Brouwer, D.S. Ivanov, T.J.Y. Derrien, N.M. Bulgakova, W.
Kautek, B. Rethfeld, L.V. Zhigilei, Fundamentals of ultrafast laser–material interaction, MRS Bulletin 41 (2016)
960-968, https://doi.org/10.1557/mrs.2016.274.
[40] V. Srinivasan, M.A. Smrtic, S.V. Babu, Excimer laser etching of polymers, Journal of Applied Physics 59 (1986)

v
3861-3867, https://doi.org/10.1063/1.336728.
[41] P. Augustyn, P. Rytlewski, K. Moraczewski, A. Skibicki, A. Mazurkiewicz, Ablation of selected thermoplastic

re
polymers using an Nd:YAG laser, Journal of Materials Science 58 (2023) 9073-9086, https://doi.org/10.1007/s10853-
023-08566-5.
[42] S. Ravi‐Kumar, B. Lies, X. Zhang, H. Lyu, H. Qin, Laser ablation of polymers: a review, Polymer International 68
(2019) 1391-1401, https://doi.org/10.1002/pi.5834.
[43] P. Pan, P. Ji, G. Lin, Spatiotemporal probe into the femtosecond laser processing of fused silica, Applied Physics A
er
128 (2022), https://doi.org/10.1007/s00339-022-05969-0.
[44] B. Chimier, O. Utéza, N. Sanner, M. Sentis, T. Itina, P. Lassonde, F. Légaré, F. Vidal, J.C. Kieffer, Damage and
ablation thresholds of fused-silica in femtosecond regime, Physical Review B 84 (2011) 094104,
pe
https://doi.org/10.1103/PhysRevB.84.094104.
[45] A. Ben-Yakar, A. Harkin, J. Ashmore, R.L. Byer, H.A. Stone, Thermal and fluid processes of a thin melt zone
during femtosecond laser ablation of glass: the formation of rims by single laser pulses, Journal of Physics D:
Applied Physics 40 (2007) 1447-1459, https://doi.org/10.1088/0022-3727/40/5/021.
[46] M. Sun, U. Eppelt, C. Hartmann, W. Schulz, J. Zhu, Z. Lin, Damage morphology and mechanism in ablation cutting
ot

of thin glass sheets with picosecond pulsed lasers, Optics & Laser Technology 80 (2016) 227-236,
https://doi.org/10.1016/j.optlastec.2016.01.023.
[47] H. Shin, D. Kim, Cutting thin glass by femtosecond laser ablation, Optics & Laser Technology 102 (2018) 1-11,
tn

https://doi.org/10.1016/j.optlastec.2017.12.020.
[48] I. Mirza, N.M. Bulgakova, J. Tomastik, V. Michalek, O. Haderka, L. Fekete, T. Mocek, Ultrashort pulse laser
ablation of dielectrics: Thresholds, mechanisms, role of breakdown, Sci Rep 6 (2016) 39133,
https://doi.org/10.1038/srep39133.
rin

[49] J. Zhan, Y. Gao, J. Sun, W. Zhu, S. Wang, L. Jiang, X. Li, Mechanism and optimization of femtosecond laser
welding fused silica and aluminum, Applied Surface Science 640 (2023) 158327,
https://doi.org/10.1016/j.apsusc.2023.158327.
[50] M. Chambonneau, Q. Li, V.Y. Fedorov, M. Blothe, K. Schaarschmidt, M. Lorenz, S. Tzortzakis, S. Nolte, Taming
ep

Ultrafast Laser Filaments for Optimized Semiconductor–Metal Welding, Laser & Photonics Reviews 15 (2020)
2000433, https://doi.org/10.1002/lpor.202000433.
[51] B. Acherjee, A.S. Kuar, S. Mitra, D. Misra, S. Acharyya, Experimental investigation on laser transmission welding
of PMMA to ABS via response surface modeling, Optics & Laser Technology 44 (2012) 1372-1383,
Pr

https://doi.org/10.1016/j.optlastec.2011.12.029.
[52] W. Watanabe, S. Onda, T. Tamaki, K. Itoh, J. Nishii, Space-selective laser joining of dissimilar transparent
materials using femtosecond laser pulses, Applied Physics Letters 89 (2006) 021106,

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4854305
 https://doi.org/10.1063/1.2221393.
[53] R.M. Carter, M. Troughton, J. Chen, I. Elder, R.R. Thomson, M.J. Daniel Esser, R.A. Lamb, D.P. Hand, Towards

ed
industrial ultrafast laser microwelding: SiO2 and BK7 to aluminum alloy, Appl Opt 56 (2017) 4873-4881,
https://doi.org/10.1364/AO.56.004873.
[54] H. Chen, J. Duan, Z. Yang, W. Xiong, L. Deng, Picosecond laser seal welding of glasses with a large gap, Opt
Express 27 (2019) 30297-30307, https://doi.org/10.1364/OE.27.030297.

iew
[55] G. Anwer, B. Acherjee, Laser polymer welding process: Fundamentals and advancements, Materials Today:
Proceedings 61 (2022) 34-42, https://doi.org/10.1016/j.matpr.2022.03.307.
[56] J.H. Burge, F. Lacroix, O.W. Fähnle, D. Hélie, R. Vallée, R. Williamson, Optical bonding reinforced by
femtosecond laser welding, in: Optical Manufacturing and Testing IX, SPIE, 2011, pp. 304-316,
https://doi.org/10.1117/12.892465.

v
[57] D. Hélie, M. Bégin, F. Lacroix, R.J.A.o. Vallée, Reinforced direct bonding of optical materials by femtosecond laser
welding, Applied optics 51 (2012) 2098-2106, https://doi.org/10.1364/ao.51.002098.

re
er
pe
ot
tn
rin
ep
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4854305