accelerated paper Fluorescence Rejection in Resonance Raman Spectroscopy Using a Picosecond-Gated Intensified Charge-Coupled Device Camera EVTIM V. EFREMOV, JOOST B. BUIJS, CEES GOOIJER, and FREEK ARIESE* Department of Analytical Chemistry and Applied Spectroscopy, Laser Centre Vrije Universiteit Amsterdam, the Netherlands A Raman instrument was assembled and tested that rejects typically 98– 99% of background fluorescence. Use is made of short (picosecond) laser pulses and time-gated detection in order to record the Raman signals during the pulse while blocking most of the fluorescence. Our approach uses an ultrafast-gated intensified charge-coupled device (ICCD) camera as a simple and straightforward alternative to ps Kerr gating. The fluorescence rejection efficiency depends mainly on the fluorescence lifetime and on the closing speed of the gate (which is about 80 ps in our setup). A formula to calculate this rejection factor is presented. The gated intensifier can be operated at 80 MHz, so high repetition rates and low pulse energies can be used, thus minimizing photodegradation. For excitation we use a frequency-tripled or -doubled Ti : sapphire laser with a pulse width of 3 ps; it should not be shorter in view of the required spectral resolution. Other critical aspects tested include intensifier efficiency as a function of gate width, uniformity of the gate pulse across the spectrum, and spectral resolution in comparison with ungated detection. The total instrumental resolution is 7 cm 1 in the blue and 15 cm 1 in the ultraviolet (UV) region. The setup allows one to use resonance Raman spectroscopy (RRS) for extra sensitivity and selectivity, even in the case of strong background fluorescence. Excitation wavelengths in the visible or UV range no longer have to be avoided. The effectiveness of this setup is demonstrated on a test system: pyrene in the presence of toluene fluorescence (k exc ¼ 257 nm). Furthermore, good time-gated RRS spectra are shown for a strongly fluorescent flavoprotein (k exc ¼ 405 nm). Advantages and disadvantages of this approach for RRS are discussed. Index Headings: Time-resolved Raman; Gated detection; Intensified charge-coupled device; ICCD; Baeyer–Villiger monooxygenase. INTRODUCTION The objective of this work was to assemble a Raman setup that can be used over a broad range of ultraviolet (UV) and visible excitation wavelengths and that can efficiently reject background fluorescence. In many Raman applications fluo- rescence interference is a major obstacle. It may originate from matrix components or solvent impurities, from the compound of interest itself, or from optical components. A standard approach to avoid fluorescence is the use of longer excitation wavelengths such as 785 or 1064 nm. Of course, this is disadvantageous from a sensitivity point of view because the Raman signal is proportional to k 4 . In resonance Raman spectroscopy (RRS) the intensity of the Raman signal typically increases by several orders of magnitude. More importantly, selective excitation becomes possible because the Raman signal from non-resonantly excited matrix components (e.g., solvent, DNA, or protein matrix) remains low. Unfortunately, fluorescence interference will become more severe because RRS is based on excitation at or close to the absorption maximum of the molecule of interest. Consequently, the probability of fluorescence emission upon relaxation from the exited state increases substantially. Furthermore, when using visible or near-UV excitation sources, fluorescence from matrix components will also increase. Asher has shown that in RRS with excitation in the deep-UV region (,250 nm) fluorescence interference is negligible. 1 Our group has successfully followed this approach by developing deep-UV RRS as a detection method in liquid chromatography 2 and capillary electrophoresis. 3 However, an inherent disadvantage of excitation in the deep-UV region is the decrease in selectivity of RRS because most molecules absorb well in this region. For instance, in the case of a flavoprotein it would be very advantageous to selectively excite the flavin moiety (which has specific absorption in the visible range) without complicating the Raman spectrum with other protein components. However, because flavins are also strongly fluorescent, recording good-quality RRS spectra is a major challenge. For these reasons, fluorescence rejection is crucial for the development of RRS in the visible/near-UV range. An overview of available options was compiled by Pelletier. 4 Chemometric approaches using second derivatives 5 or wave- length shifts 6 have been used in order to distinguish the sharp Raman lines from the broad fluorescence background. However, if possible one would prefer to reduce fluorescence interference as much as possible before detection and have the chemometric option still available if necessary. Adding a fluorescence quencher to the sample may be an option in some cases, although of course this would imply a sample modification. The most promising option is to make use of time discrimination between instantaneous Raman photons and fluorescence that will occur after a few ns. In 1986 time- resolved detection was extensively tested by Everall and co- workers, using picosecond laser pulses and a gated diode array detector 7 or a time-to-amplitude converter to gate the photo- multiplier detector. 8 The latter reference also provides an overview of the available technology for gated detection at the time of publication. At the time, the improvement in signal-to- background ratio was limited by the nanosecond time Received 20 February 2007; accepted 26 March 2007. * Author to whom correspondence should be sent. E-mail: ariese@few.vu.nl. Volume 61, Number 6, 2007 APPLIED SPECTROSCOPY 571 0003-7028/07/6106-0571$2.00/0 Ó 2007 Society for Applied Spectroscopy