End Cap Splicing of Photonic Crystal Fibers With Outstanding Quality For High Power Applications
End Cap Splicing of Photonic Crystal Fibers With Outstanding Quality For High Power Applications
End Cap Splicing of Photonic Crystal Fibers With Outstanding Quality For High Power Applications
ABSTRACT
The application of photonic crystal fibers (PCF), especially in high power fiber laser systems, requires special
preparation technologies with some significant differences compared to standard fibers. Features, like air-clad structures,
highly rare-earth doped cores with low NA and stress applying parts of the PCFs, require additional steps in fiber
preparation and innovative splicing technologies to gain optical properties. Here we discuss a contamination- free carbon
dioxide laser splicing device, which is used for defined air-clad collapsing and end cap splicing to get a stable and sealed
fiber end face with preserved high beam quality and additional functionality. The special design of the computer-
controlled laser splicing process provides a versatile tool with high reproducibility for joining different geometries with
an adjustable well-balanced heat distribution. A wide range of PCFs with different diameters, air-clad structures and
doped materials up to ~2 mm have been spliced. For selected PCF-end cap splices cleave or polishing requirements as
well as results on beam quality, tensile strength and further splicing features are presented.
Keywords: CO2-laser, fusion splicing, end cap, PCF
1. INTRODUCTION
Fiber laser exhibit high beam quality with high efficiency up to the multi kW-range for continuous wave output and mJ
pulses in pulse regimes. A promising technology for further scaling is the utilization of photonic crystal fibers (PCF)
with high numerical aperture in pump cladding and large mode field diameters of the fundamental mode in the laser core
for the development of high-power fiber lasers and amplifier systems. For the integration of this fiber type with air holes
in a laser system special care must be taken. Cleaving and fusion splicing can become a time and material consuming
issue. But the sealing of the PCF-end facet is absolutely essential in order to avoid that the same particles or humidity
enter air holes, which would lead to system failure in high power applications.
Commercial fusion splicers utilize a hydrogen/oxygen flame, a resistively heated filament, electrical arc or plasma
discharge with two or more electrodes as a heat source. Nozzles and electrodes are worn down and might generate debris
on the fibers during the glass process.
1.1 CO2-laser fusion splicing device
In this chapter we demonstrate the fusion splicing of PCF on end caps based on a water-cooled 70W CO2-laser as the
heat source. The high absorption at the wavelength of 10.6µm offers a broad range of glass processing such as fiber
fusion splicing, tapering, cleaning etc. First successful applications of this laser fusion splicing technology with a special
optical design of the device have been made already [1]. In the following investigations we use some improvements in
terms of processing and changes to the device itself and demonstrate PCF-end cap splices with very short collapse range
of the air holes. In front of the CO2- laser a telescope is used for laser focus adjustment in order to distribute the heat
either on the end cap or the fiber. An annular (collimated or non-collimated) laser beam is generated by a tilted turning
mirror. After two off-axes parabolic mirrors, the generated conical annular laser beam enables the easy adaptation for
different fiber diameters via a height-adjustable z-stage (Fig. 1).
Figure 2. Fiber with partially viewable end cap and fiber captured by the side view camera (left) view through the end cap
with the second camera
The overlay cross hair lines serve as indispensable references for the invisible annular laser beam.
1.2 Properties of the applied PCFs
There is as huge number of different PCFs depending on the target application. The following fiber types have been used
during the investigations:
Figure 3. Passive PCF (outer cladding diameter: 830 µm) for tensile strength investigations
Secondly, in order to investigate the impact of the CO2-laser fusion splicing on the beam quality we used an active PCF
with stress applying rods (DC-400/40-PZ-Yb, single-mode, polarizing double-clad Ytterbium-fiber with large mode
area, M²: 1.4) having an outer cladding diameter of 690µm (Fig. 4). The pump core is again defined by an air-clad with a
diameter of 400 µm (NA~0.57) and the core is defined by the inner hole array with d/Λ of 0.178 and a period of 10.1µm.
The dimensions of the applied end caps are 10 mm for the diameter and 5 mm thick.
Figure 4. Active PCF (outer cladding diameter: 690 µm) for beam quality measurements
Figure 6. Impact of air holes collapse for the passive PCF when using different parameter sets
Fusion splicing of components with diverse glass composition and additional different diameters with high yield is hard
to achieve with conventional splice equipment. In contrast to all conventional fusion splicers the off-axis angle as well as
the controlled defocus of the incident CO2-laser beams enables a more precise melting of the end cap. The air hole
microstructure in the PCF, especially the holes defining the single mode core, need to be preserved for light guiding and
mode quality, which will be achieved by decreasing the heat. Thereby the question of the achievable tensile strength
comes up.
A standard fusion strength test of a fiber-fiber splice is performed with two rotary mandrels. If the fibers are coiled on
the mandrels with the splice in a straight line a tensile force is applied [2]. But PCFs are very susceptible to small
bending radii. Additionally difficulties are expected to grip PCF and end caps without any torsion or bending. So we
used the successfully tested approach as described in [1] again. The following picture shows a short piece [~300mm] of
PCF with two spliced end caps arranged in a pull strength test machine (Instron). The range of measurements is for the
small load cell up to 100N or 2000N for the large load cell. In some cases a second test was necessary with the larger
load cell. However the achieved forces were a lot lower than in the first test. We assume a preliminary damage caused by
the first pull test.
Figure 10. After the pull test fiber tips remaining on the end cap
2.2 Beam quality investigations
The beam quality has been analyzed with an M² measurement device from Spiricon Inc. Now we used our active PCF in
an assembly with a narrow band fiber laser at 1080 nm. The laser was coupled in a short piece (~0.5 m) of active PCF
equipped with a spliced end cap. As mentioned above this fiber has an M² of 1.4. A PCF-end cap splice with different
collapse width should decrease the beam quality. For our first experiment the collapse width was close to zero with a
measured value M² of around 1.42 and subsequent degradation with larger collapse. Detailed results will be presented.
4. SUMMARY
The paper presents investigations of a CO2-laser fusion splicing device applied on two different PCFs. The link between
collapse width, mechanical stability and resulting beam quality has been demonstrated. A defined collapse process with
the same splicing equipment supports the following cleave and joining tasks. The off-axis angle of the defocused
incident laser beam together with large working offers an excellent heat distribution on the joining components without
any contaminations for different geometries or cross sections.
REFERENCES
[1] Boehme, S., Beckert, E., Eberhardt, R., Tuennermann, A., “Laser splicing of end caps – process requirements in
high power laser applications“, Proc. SPIE 7202, 2009
[2] Yablon, A. D., “Optical Fiber Fusion Splicing”, Springer-Verlag Berlin Heidelberg, 2005
[3] Bourliaguet, B. Pare, C., Edmond, F., Croteau, A., "Microstructured fiber splicing" Opt. Express Vol. 11, No. 25,
3412 – 3417, 2003
[4] S. Richter, S. Döring, A. Tünnermann und S. Nolte, „Bonding of glass with femtosecond laser pulses at high
repetition rates,” Applied Physics A: Materials Science & Processing, Vol.103, 2, 257-261