Pulse Measurement Techniques Using an Acousto-Optic Programmable Dispersive Filter

Abstract

Programmable spectral filters, such as acousto-optic programmable dispersive filters (AOPDF, Tournois (1997)) or spatial light modulators inserted in the Fourier plane of zerodispersion lines (Froehly et al., 1983; Weiner, 2000), have opened up the field of ultrafast pulse shaping and given the ability to manipulate spectral amplitude and phase of broadband ultrashort pulses. These devices have found a great number of applications, among which are phase compensation in chirped-pulse amplification laser chains (Seres et al., 2003; Ohno et al., 2002; Verluise et al., 2000), coherent control experiments within atomic or molecular systems (Tkaczyk et al., 2008; Murphy et al., 2007; Veshapidze et al., 2007; Ogilvie et al., 2006; Yamada et al., 2005) and complex pulse shaping for photo-injectors (Garzella et al., 2006). Surprisingly, pulse shapers have mainly been used to control already well characterized ultrashort pulses but seldom to characterize these pulses themselves. And yet, pulse shapers provide a convenient way to perform quantitative, reliable and versatile pulse measurements. A few pioneering works have already demonstrated that several of the existing pulse measurement techniques could be implemented with pulse shapers (Galler & Feurer, 2008; Sung et al., 2008; Forget et al., 2007; Oksenhendler et al., 2003; Monmayrant et al., 2003) and that new pulse characterization methods could even be invented (Forget et al., 2007; Grabielle et al., 2009; Lozovoy et al., 2004). The use of pulse shapers for ultrashort pulse metrology has already found practical applications, such as in situ pulse compression at the focus of high NA objectives in twophoton microscopy (von Vacano et al., 2006; 2007). Beyond specific applications, pulse shapers are expected to extend the robustness and dynamic range of the existing pulse characterization techniques. Indeed, all techniques suffer from specific weaknesses and cross-check can help to overcome the drawbacks, limitations or ambiguities related to a particular technique and to eliminate spurious results. Spectral phase interferometry for direct electric-field reconstruction (SPIDER) (Iaconis & Walmsley, 1998), for example, requires a precise calibration and determination of the relative delay between the two replicas. Such a requirement is much less stringent for the second-harmonic (SH) frequency resolved optical gating (FROG) technique. Conversely, SHFROG suffers from time direction ambiguity and phase retrieval is not straightforward. Finally, these two methods are not equivalently robust with respect to complex pulse shapes. Cross-check between SH-FROG and SPIDER results is, however, often made difficult since it involves separate measurement devices, which multiplies the causes of