Continuing Struggle for Larger Fiber Mode Areas

Posted on 2007-02-04. Permanent link: http://www.rp-photonics.com/spotlight_2007_02_04.html

Author: Dr. Rüdiger Paschotta, RP Photonics Consulting GmbH

Ref.: encyclopedia articles on fibers, large mode area fibers, high-power fiber lasers and amplifiers; spotlight articles of 2006-12-03, 2006-10-04, 2006-08-10

High power fiber lasers and amplifiers have of course been a very important topic at the conferences Photonics West in San Jose, January 20-25, and Advanced Solid-State Photonics (ASSP) in Vancouver, January 28-31. Quite a number of talks addressed the quest for larger mode areas, as this issue has been recognized as the central bottleneck which has to be overcome in order to continue the recent enormous performance enhancements. (See the previous spotlight articles on this topic as mentioned above.) Quite some creativity has been unleashed in the context of attempts to tackle this problem:

• Sandia National Labs is optimizing refractive index profiles for very large mode areas (talk 6453-48 at Photonics West). Some combined power-law design was found to exhibit better performance than e.g. square-law or triangular profile fibers.
• Fitel reported on distributed mode filtering in a cladding-pumped amplifier (talk 6453-15 at Photonics West). This technique makes it possible to suppress higher-order modes as well as to filter out ASE.
• Researchers at the University of Michigan, collaborating with the company Nufern, have demonstrated a fiber with chirally-coupled core (talk ME2 at ASSP). This contains a second core would around the central core. The idea is basically to obtain phase-matched coupling e.g. of the LP₁₁ mode to the lossy chiral core, hoping to suppress all higher-order modes of the multimode core even at large mode areas.
• IMRA reported an impressive mode area of about 3000 μm² in an ytterbium-doped leakage channel fiber (talk 6453-42 at Photonics West and ME3 at ASSP). This is a photonic crystal fiber made so that the core supports several modes, but higher-order modes are strongly attenuated by leakage through gaps between the air holes.
• Femlight uses a 60-cm long rod-type photonic crystal fiber which has to be kept straight (talk ME4 at ASSP). The concept is essentially to live with a very weakly guiding core by avoiding any bending. Quite short pulses with up to 50 W average power are generated with high efficiency. However, such a device is actually no more used like a normal fiber; it is more like a long bulk crystal with some built-in weak guiding mechanism.
• The probably most radical approach is based on a concept developed by the famous Anthony E. Siegman: abandon the generally used method of index guiding altogether and replace it with gain guiding! Talk ME1 at ASSP presented the first experimental demonstration of this concept, using a neodymium-doped fiber which is anti-guiding when being unpumped but nevertheless exhibits well-behaved modes due to gain guiding. This was demonstrated with the rather unconventional experimental approach of pumping the fiber with a Xenon flashlamp. While the results are certainly encouraging, with an apparently already rather large mode area, the researchers are not yet able to fully assess the potential of their method.

It remains to be seen how far we will get with all these approaches. So far, it is not clear whether we will obtain a practical approach for realizing more areas e.g. of 10'000 μm² in fibers with reasonable handling properties. Fore sure, many people will continue to work hard in this direction, and obviously a breakthrough can be very rewarding.