Optimizing the fluorescent signal for in-vivo nonlinear imaging in a completely all-fibered ytterbium chirped pulse amplifier

Summary form only given. Multiphoton microscopy, including two-photon laser scanning microscopy and second-harmonic generation is a unique technique for non-invasive imaging in biological samples. However, clinical translation and applications outside the laser laboratory have been hampered by the need to use bulky, expensive and environmentally sensitive lasers, that often require active water cooling and have high energy consumption. Femtosecond fiber lasers, which have relaxed cooling and energy consumption requirements, are very attractive as master sources for multiphoton microscopy [1]. A challenge for applications in nonlinear imaging in living tissue is to maximize the total fluorescent yield from each fluorophore. Excitation with a transform-limited pulse significantly increases the efficiency of excitation [2]. Transform limited pulses are also advantageous to minimize the thermal load in living tissue. Multiphoton microscopy would strongly benefit from the implementation of robust all-integrated fiber based concepts that allow for truly turn-key maintenance free operation.When developing high-peak power integrated fiber laser sources for multiphoton microscopy it is important to: (i) minimize the nonlinearities in the system; (ii) to implement suitable fiber-based stretching and compression schemes that will not trade-off pulse duration and pulse fidelity for all-fiber integration, and will instead allow to produce high fidelity pulses for compact multiphoton imaging systems. Ideally, the pulse duration from such a fiber source should be independent from the output pulse energy, to allow a fast change of the excitation power without the use of expensive electro-optic modulators. Hollow-core photonic bandgap fibers (HCPBF) are very attractive for the delivery of femtosecond pulses for applications in nonlinear optical microscopy. Two main approaches based on HCPBF compressors have been demonstrated: one uses a standard single mode fiber (SMF) stretcher [3,4], and the other concept is based on a DCF stretcher [5]. In this work, we compared the in-vivo imaging capabilities of two completely all-fiber single mode Yb-fiber lasers that use the above-mentioned stretcher-compressor concepts. The amplifier system consists of a broadband 76 MHz Yb-fiber oscillator, a fiber stretcher, and two single mode fiber amplifiers. Spectral filters are used after the oscillator and between the two amplifier stages. The first filter is used to tune the wavelength of the seed pulse and the second filter is used for ASE suppression. The output of the amplifier is spliced to a HCPBF compressor. For both stretcher configurations, the maximum output power at the output of the HCPBF compressor is about 120 mW. After this, the beam is guided into a fluorescence microscope adapted for laser scanning multiphoton imaging. We measured the second harmonic FROG traces of the pulses after passing through the microscope system and optimized the stretcher length for best compressibility in both cases. The system with a standard polarization maintaining SMF stretcher delivers 280 fs pulses with considerable satellites carrying more than 20% of the energy, and the system with the DCF stretcher delivers 260 fs pulses with only one small side pulse that carries less than 5% of the energy. Both systems have the same output spectrum, corresponding to a Fourier transform limited duration of 255 fs. Figure 1 shows a comparison of two imaged 2-color frames obtained with the two different configurations. For the same power and under the same imaging conditions, the fluorescence signal for the system that uses the DCF stretcher was ~45 % higher than the fluorescence signal that uses a standard SMF stretcher. This is to be expected due to the much better pulse compressibility achieved with the DCF stretcher. This result is consistent with the 50% fluorescent signal enhancement that is expected, based on the calculated second-order coherence factors of the pulses from the two different stretcher approaches.