The two graphs in Figure 2 clearly demonstrate the distortion that occurs when an ultrashort pulse is reflected by a standard dielectric mirror (left) as compared to the same pulse reflected by a mirror optimized for use with an ultrafast laser (right).
- Scratch-Dig: This parameter characterizes the light scatter losses throughout an optical system. When there is light scatter in a system, the optical power reaching the target is significantly reduced. An optical train with a low scratch-dig figure can also produce wavefront degradation. High quality lenses or mirrors suitable for ultrafast laser systems typically specify a scratch-dig value of 10-5.
- Flatness: The wavefront flatness of plano-optics (mirrors) is an important specification, especially for imaging systems. Plano-optics with specifications greater than the standard 1/10 wave flatness may exhibit unacceptable image distortions. This wavefront flatness specification applies equally to applications in which high intensity is required (i.e., attosecond pulse generation), since poor wavefront quality degrades the focus of the beam, reducing the intensity at the focal spot.
It’s important to note that any of the reflective optics discussed in this section may be used; however, the unique needs of your particular application should be carefully considered in the decision process.
A typical ultrafast laser optics configuration may have, in addition to polarizing and focusing elements, at least 10 reflective mirrors between the laser light source and the target. The high number of mirrors makes it imperative that the specific reflectivity of the mirrors be maximized since any optical losses due to reflection are cumulative and these losses reduce the optical power at the target.
It is important to consider that standard metallic mirrors such as gold or silver work well in the ultrafast regime only so long as the protective over-coat is designed for ultrafast laser system use. A bare metal coating has minimal GDD because the light does not penetrate deeply into the metal before being reflected, thereby minimizing dispersion. However, bare metals are not recommended because of degradation over time.
Most metallic mirrors have dielectric over-coats for environmental protection and reflectivity enhancement; however, improper design of these coatings may significantly distort the GDD at certain wavelengths within the high reflectance range.
Also, metallic mirrors typically have poor reflectivity at specific wavelengths as compared to dielectric coatings. Ultimately, the use of a dielectric or metallic mirror with no GDD specifications risks distortion of the pulse or reduced broadband reflectivity within the optical train.
As a model, consider some different options shown here for a mirror optimized for 800 nm laser light pulses in an ultrafast spectroscopy experiment. The different mirrors in this list are rated from “OK” to “Best” for this application.
Ultimately, the investment made depends upon your particular application and needs.
- The 10D20ER.2 standard silver coated mirror, one-inch diameter, 1/10 wave flatness, 15-5 scratch dig,reflectivity R = 93-96% at 800 nm, no specification for GVD – Lowest cost: OK.
- The 10B20EAG.1 femtosecond optimized silver mirror, one-inch diameter, 1/10 wave flatness, 10-5 scratch dig, reflectivity R = 99% at 800 nm, GVD less than 5 fs2 – Low to intermediate cost: GOOD.
- 10B20UF.25 low GVD ultrafast mirror, one-inch diameter, 1/10 wave flatness, 15-5 scratch dig, reflectivity R > 99% at 800 nm, near zero GVD at 800 nm – Intermediate to higher cost: BETTER. (Even though the reflectivity is similar to #2, this is an all dielectric coating, so this mirror will wear longer.)
- Newport 10Q20UF.55P ultrafast broadband turning mirror, one-inch diameter, 1/10 wave flatness, 20-10 scratch dig, Reflectivity Rp > 99.6%, near zero GVD – Higher cost: BEST (This mirror is ideal for the Spectra-Physics Mai Tai® or Insight® lasers which can be tuned to the whole 690-1040 nm, 680-1300 nm spectrum.)
Legend:
R = absolute reflectivity, typically at normal angle of incidence
Rp = reflectivity of P-polarized light, typically at 45° angle of incidence
Choosing the best optics for an ultrafast laser application can be a challenging proposition. The choices made for the ultrafast laser, particularly for the optical components, will be critical for determining the success or failure of research experiments and other ultrafast laser applications.
It is always prudent to take the time to understand the optical requirements of the application and to study the specifications of the optics selected to meet these specifications prior to purchasing the system components. The least expensive choice may not always be the most cost-effective or technologically suitable one, resulting in higher costs – and more failures – in the long run