Final Report Summary - UPDUS (Understanding photoprotection mechanisms in DNA by two-dimensional UV spectroscopy)
When ultraviolet (UV) radiation is absorbed by the DNA, the electronic energy acquired by the molecule is efficiently converted into vibrational energy (heat) on an ultrafast timescale (from a few tens to a few hundreds of femtoseconds) thus preventing DNA damage. If the UV energy were not dissipated, it would lead to the generation of free radicals and initiate a variety of photoreactions, eventually corrupting the information encoded in the base sequence. However, the underlying photochemical mechanisms behind the fate of photoexcitations in DNA are not clear yet. To track these ultrafast processes taking place in DNA, we proposed to implement two-dimensional (2D) electronic spectroscopy in the UV spectral range.
This goal presents some technological and scientific challenges that constituted the main objectives of this project:
(1) Generation and characterization of ultrashort pulses in the UV range: ultra-broadband pulses are mandatory to fully exploit the benefits of the 2D spectroscopy and allow us to address the full spectrum of the excited states with the highest possible time resolution.
(2) Development of a 2D setup in the near UV, using as pump the pulses generated in the objective 1. To this end, we chose a scheme based in partially collinear pump-probe geometry.
(3) Spectroscopy of DNA monomers to map the full transition from the excited state to the ground state. The high temporal resolution and broad spectral tuning of the setup enables to track the transition of the wavepacket from the Franck-Condon region to the conical intersections and the ground state.
The generation of UV pulses is not straightforward, since there are no gain media that produce femtosecond pulses in this spectral region and optical parametric amplifiers (OPAs) are hampered by the occurrence of two-photon absorption of the short-wavelength pump pulse. For this reason, we need to shift ultrashort visible pulses to the UV range using nonlinear frequency conversion processes. In particular, we have used second-harmonic (SH) generation of a visible non-collinear OPA, and sum-frequency (SF) mixing with a portion of the fundamental beam. With these schemes we have generated 300-nJ deep-UV pulses (265-280 nm) and 1-µJ in the near-UV range (320-380 nm), respectively (panels b and c in Fig. 1.1).
The temporal characterization of the so-generated ultrashort UV pulses is also challenging due to the short-wavelength absorption of nonlinear materials and the unfavorable phase-matching conditions, which prevent the use of traditional setups (FROG, SPIDER…). To this end, we have demonstrated a modified scheme of the two-dimensional spectral shearing interferometry (2DSI) that has allowed us to characterize sub-10 fs pulses in three spectral regions (UV, visible and infrared). 2DSI relies on the interference between the test pulse and two quasi-monochromatic pulses (ancillae). In our approach the ancillae are generated by means of a pulse shaper (Fig. 1.1 a), improving the accuracy of the measurement with respect to previous methods.
The so-generated deep-UV pulses were used for pump-probe spectroscopy of DNA nucleosides and proteins, whereas the near-UV pulses were employed for spectroscopy of graphene, azobenzenes and polymers. To the best of our knowledge, these are the first pump-probe experiments in DNA nucleosides with resolution better than 20 fs. The samples (guanosine, uridine, adenosine, cytidine and 5-methyluridine) were prepared in aqueous solutions and flowed with peristaltic pump in free sample jet configuration. They were pump with sub-20 fs pulses centered at 270 nm and probed in the visible/infrared spectral region. We have found two decay times for all the nucleosides (one in the femtosecond scale and another in the picosecond scale), compatible with a four-level system model (Fig. 1.2). The interpretation of the experimental results has been done in close collaboration with Prof. Garavelli (Computational Chemistry, University of Bologna) and Dr. Ines Delfino (Biological Sciences department, Tuscia University).
Finally, we have implemented a 2D spectroscopy setup in the near-UV. We chose a pump-probe geometry with white-light pulses generated in sapphire as probe and two variable delayed UV pulses as pump. The main drawback of 2D spectroscopy concerning UV pulses is the demanding broadband pulse shaping, which falls outside the transparency window of the commonly used liquid crystal spatial light modulators (SLM). As alternative, we have used a translating wedge-based identical pulses encoding system (TWINS) for the generation of two variable delayed pump pulses in the visible range. The so-generated collinear pulses were then shifted to the UV range by SF with a narrowband pulse at 800 nm. To account for the dispersion suffered by the UV due to propagation, we imparted a slightly negative chirp on the visible pulses, which is transferred to the UV. With this configuration, we have demonstrated control of the pulse delay with attosecond precision and stability better that λ ∕ 360 (Fig. 1.3).
In summary, we have generated ultra-broadband pulses in the UV range, which have allowed us to study the dynamics of the DNA nucleosides with unprecedented temporal resolution by means of pump-probe experiments. We have also developed an approach for the temporal characterization of these ultrashort UV pulses. Finally, we have demonstrated one of the first 2D spectroscopy setups in the world operating at the UV range. We expect that the results from these studies will enable for a better understanding of the DNA photoprotection mechanism, and pave the way for further investigations on the role of UV light in proteins and other biomolecules.
For further information about the project please contact Dr. Rocío Borrego Varillas (rocio.borrego@polimi.it) or Prof. Giulio Cerullo (giulio.cerullo@polimi.it).
This goal presents some technological and scientific challenges that constituted the main objectives of this project:
(1) Generation and characterization of ultrashort pulses in the UV range: ultra-broadband pulses are mandatory to fully exploit the benefits of the 2D spectroscopy and allow us to address the full spectrum of the excited states with the highest possible time resolution.
(2) Development of a 2D setup in the near UV, using as pump the pulses generated in the objective 1. To this end, we chose a scheme based in partially collinear pump-probe geometry.
(3) Spectroscopy of DNA monomers to map the full transition from the excited state to the ground state. The high temporal resolution and broad spectral tuning of the setup enables to track the transition of the wavepacket from the Franck-Condon region to the conical intersections and the ground state.
The generation of UV pulses is not straightforward, since there are no gain media that produce femtosecond pulses in this spectral region and optical parametric amplifiers (OPAs) are hampered by the occurrence of two-photon absorption of the short-wavelength pump pulse. For this reason, we need to shift ultrashort visible pulses to the UV range using nonlinear frequency conversion processes. In particular, we have used second-harmonic (SH) generation of a visible non-collinear OPA, and sum-frequency (SF) mixing with a portion of the fundamental beam. With these schemes we have generated 300-nJ deep-UV pulses (265-280 nm) and 1-µJ in the near-UV range (320-380 nm), respectively (panels b and c in Fig. 1.1).
The temporal characterization of the so-generated ultrashort UV pulses is also challenging due to the short-wavelength absorption of nonlinear materials and the unfavorable phase-matching conditions, which prevent the use of traditional setups (FROG, SPIDER…). To this end, we have demonstrated a modified scheme of the two-dimensional spectral shearing interferometry (2DSI) that has allowed us to characterize sub-10 fs pulses in three spectral regions (UV, visible and infrared). 2DSI relies on the interference between the test pulse and two quasi-monochromatic pulses (ancillae). In our approach the ancillae are generated by means of a pulse shaper (Fig. 1.1 a), improving the accuracy of the measurement with respect to previous methods.
The so-generated deep-UV pulses were used for pump-probe spectroscopy of DNA nucleosides and proteins, whereas the near-UV pulses were employed for spectroscopy of graphene, azobenzenes and polymers. To the best of our knowledge, these are the first pump-probe experiments in DNA nucleosides with resolution better than 20 fs. The samples (guanosine, uridine, adenosine, cytidine and 5-methyluridine) were prepared in aqueous solutions and flowed with peristaltic pump in free sample jet configuration. They were pump with sub-20 fs pulses centered at 270 nm and probed in the visible/infrared spectral region. We have found two decay times for all the nucleosides (one in the femtosecond scale and another in the picosecond scale), compatible with a four-level system model (Fig. 1.2). The interpretation of the experimental results has been done in close collaboration with Prof. Garavelli (Computational Chemistry, University of Bologna) and Dr. Ines Delfino (Biological Sciences department, Tuscia University).
Finally, we have implemented a 2D spectroscopy setup in the near-UV. We chose a pump-probe geometry with white-light pulses generated in sapphire as probe and two variable delayed UV pulses as pump. The main drawback of 2D spectroscopy concerning UV pulses is the demanding broadband pulse shaping, which falls outside the transparency window of the commonly used liquid crystal spatial light modulators (SLM). As alternative, we have used a translating wedge-based identical pulses encoding system (TWINS) for the generation of two variable delayed pump pulses in the visible range. The so-generated collinear pulses were then shifted to the UV range by SF with a narrowband pulse at 800 nm. To account for the dispersion suffered by the UV due to propagation, we imparted a slightly negative chirp on the visible pulses, which is transferred to the UV. With this configuration, we have demonstrated control of the pulse delay with attosecond precision and stability better that λ ∕ 360 (Fig. 1.3).
In summary, we have generated ultra-broadband pulses in the UV range, which have allowed us to study the dynamics of the DNA nucleosides with unprecedented temporal resolution by means of pump-probe experiments. We have also developed an approach for the temporal characterization of these ultrashort UV pulses. Finally, we have demonstrated one of the first 2D spectroscopy setups in the world operating at the UV range. We expect that the results from these studies will enable for a better understanding of the DNA photoprotection mechanism, and pave the way for further investigations on the role of UV light in proteins and other biomolecules.
For further information about the project please contact Dr. Rocío Borrego Varillas (rocio.borrego@polimi.it) or Prof. Giulio Cerullo (giulio.cerullo@polimi.it).
