LARGE OPTICS: Petawatt high-LIDT coatings don't sacrifice performance

Over the past decade, an increasing number of (mainly government-financed) large petawatt laser systems have been built or are planned for construction. The world map of ultrahigh-intensity laser capabilities compiled by C.P.J. Barty (ICUIL 2012) at the National Ignition Facility (Livermore, CA) lists a total of 71 sites that are involved in high-power laser systems, with 16 sites in North America, 30 in Europe, and 25 in Asia.¹

In the past, development of such powerful laser systems was hampered by the lack of suitable laser materials, but since the development of chirped pulse amplification (CPA) in the mid-1980s this issue has been overcome for short pulse lengths. Rapid development has taken place in this field, with applications including studies in high-energy-density physics, flash x-radiography, and laser-ignited nuclear fusion.

According to some predictions (ICUIL 2011), the generation of laser intensities of up to 10²⁴ W/cm² and the generation of gamma-ray pulses of about 100 yoctoseconds (10^-22 s) is possible.² As of 2010, the peak power of the world's most powerful laser (again according to ICUIL 2011) was approximately 11.5 PW, but by the end of 2015 this is planned to increase to approximately 127 PW. In addition, several exawatt-scale projects are currently in the planning stage.

To meet these demands, it is critical to concurrently develop high-performance laser optics that offer high laser-induced damage threshold (LIDT) without sacrificing spectral or phase performance. Although the optics can be made larger to lower the power density, this approach inevitably increases both the size and cost of the system. A compromise must be found, with the LIDT being the key parameter when considering the most appropriate beam size.

Most large-scale petawatt systems are first evaluated in the form of tabletop or smaller-scale demonstration units. As a result, it is of vital importance that all optical components used in the demonstration unit be scalable to the size of the final system, ideally at a nonprohibitive cost. Being able to use the same type of coating for the beam-handling optics throughout the system, regardless of the size of the optics, also makes system modeling much easier.

Evolving needs

CVI Melles Griot's manufacturing facility on the Isle of Man focuses on the manufacture of both small and large high-LIDT precision optics, concentrating on the design and manufacture of large femtosecond optics. To eliminate absorption centers that can cause catastrophic damage, high-LIDT laser optics require both substrate and coatings to be manufactured to the highest standards. In addition, the larger an optic becomes, the more important issues like coating stress also become, especially when dealing with coatings that consist of a large number of layers and are required to function over a wide range of temperatures in air-as well as in vacuum-without performance degradation. Most optics manufactured for these applications require a surface flatness or transmitted wavefront distortion of at least λ/10.

The range of components required when building a petawatt laser encompasses mirrors, windows, beamsplitters, and polarizers; as ever more powerful laser systems have been designed, the coating specifications for all these types of components have changed considerably over time.

For ultrashort (femtosecond or picosecond) laser pulses, the minimum pulse length Δt at a center wavelength λ₀ is related to the spectral bandwidth Δλ (at full-width half-maximum, or FWHM) as follows:

One solution has been the development of a new class of high-reflectivity (reflectivity greater than 99%), broadband (680-900 nm), low-group-delay-dispersion (GDD) mirrors that exhibit LIDT values capable of supporting the next generation of 10 PW lasers (see Fig. 2). In experiments conducted with 130 ps laser pulses, the LIDT was measured to be higher than 6 J/cm². Laser damage threshold values for several large-area femtosecond mirrors are noted in the table.

Another area of continued development is in polarizers designed for use with femtosecond lasers. For these applications, polarizers must be able to operate over a large range of wavelengths, ideally from 700–900 nm. Standard Brewster-angle plate polarizers typically only operate over a limited bandwidth (a few nanometers) and are not suitable.

In contrast, optically contacted polarization cubes polarize the incoming beam at an extinction ratio of approximately 500:1 or better over the full 700-900 nm range (see Fig. 3). The p-polarized beam exits the polarizer without any beam deviation, making the component easy to align. These polarizers are also more compact than plate polarizers. For example, a rectangular plate polarizer with dimensions measuring 2 × 6 in. would have to be made approximately 1 in. thick to allow a surface flatness of l/10 and when used at a 70° angle of incidence would only be able to handle a beam smaller than 1.5 in. diameter. In contrast, polarization cubes in sizes up to 76.2 mm can accommodate beams up to approximately 70 mm in diameter (smaller polarization cubes are available for tabletop petawatt laser systems).

REFERENCES

1. ICUIL 2012, web site of the International Committee of Ultra-High Intensity Lasers (accessed June 2012); http://www.icuil.org/events-a-activities/laser-labs.html.

2. IUIL 2011: The International Committee on Ultra-High Intensity Lasers, ICUIL News, 2 (April 2011).