Solution phase high repetition rate laser pump x-ray probe picosecond hard x-ray spectroscopy at the Stanford Synchrotron Radiation Lightsource

Abstract

We present a dedicated end-station for solution phase high repetition rate (MHz) picosecond hard x-ray spectroscopy at beamline 15-2 of the Stanford Synchrotron Radiation Lightsource. A high-power ultrafast ytterbium-doped fiber laser is used to photoexcite the samples at a repetition rate of 640 kHz, while the data acquisition operates at the 1.28 MHz repetition rate of the storage ring recording data in an alternating on-off mode. The time-resolved x-ray measurements are enabled via gating the x-ray detectors with the 20 mA/70 ps camshaft bunch of SPEAR3, a mode available during the routine operations of the Stanford Synchrotron Radiation Lightsource. As a benchmark study, aiming to demonstrate the advantageous capabilities of this end-station, we have conducted picosecond Fe K-edge x-ray absorption spectroscopy on aqueous [Fe^II(phen)₃]²⁺, a prototypical spin crossover complex that undergoes light-induced excited spin state trapping forming an electronic excited state with a 0.6-0.7 ns lifetime. In addition, we report transient Fe Kβ main line and valence-to-core x-ray emission spectra, showing a unique detection sensitivity and an excellent agreement with model spectra and density functional theory calculations, respectively. Notably, the achieved signal-to-noise ratio, the overall performance, and the routine availability of the developed end-station have enabled a systematic time-resolved science program using the monochromatic beam at the Stanford Synchrotron Radiation Lightsource.

FIG. 1.

Schematic of the laser pump x-ray probe setup at SSRL beamline 15-2.

FIG. 2.

(a) Geometry optimized calculated structures of the low-spin (LS) and high-spin (HS) states of aqueous [Fe^II(phen)₃]²⁺. (b) LS and HS electronic configurations of the iron d-orbitals in octahedral symmetry. (c) Schematic of the photoinduced relaxation dynamics of aqueous [Fe^II(phen)₃]²⁺ as described in the main text.

FIG. 3.

Fluence dependence of the photoinduced difference signals of aqueous [Fe^II(phen)₃]²⁺. Vertical dashed lines indicate the fluences used in the experiment. (a) X-ray absorption difference signal measured from a 2.5 mM aqueous solution at a pump probe delay of ∼100 ps with the incident energy fixed to ∼7125.6 eV. (b) X-ray emission difference signal measured from a 10 mM aqueous solution with the incident energy fixed to 8800 eV, the x-ray emission energy fixed to ∼7057.3 eV, and the pump probe delay set to ∼200 ps.

FIG. 4.

Picosecond Fe K-edge x-ray absorption spectroscopy of aqueous [Fe^II(phen)₃]²⁺. (a) Difference signal (pumped minus unpumped) as a function of pump probe delay and incident x-ray energy. (b) Difference signal (red) at a fixed pump probe delay of ∼100 ps. The inset highlights transient features in the pre-edge/edge region. The shaded area represents the standard error of all averaged scans. (c) Kinetic trace recorded at a fixed incident energy of ∼7126 eV shown together with a mono-exponential fit (R² = 0.9989). Error bars represent the standard error of all averaged scans.

FIG. 5.

Picosecond Kβ main line x-ray emission spectroscopy of aqueous [Fe^II(phen)₃]²⁺. (a) Pumped (purple) and unpumped (blue) Kβ main line spectra recorded at a time delay of ∼60 ps. Shaded areas represent the standard error of all averaged scans. (b) Kβ main line difference spectrum compared with the scaled difference of an independently measured Fe(II) high-spin reference spectrum from Zhang et al. and the laser off spectrum [black line, Eq. (2)]. The shaded area represents the standard error of all averaged scans. (c) Kinetic trace of the Kβ main line difference signal recorded with the x-ray emission energy fixed to ∼7056 eV. Error bars represent the standard error of all averaged scans. The black line represents a fit using Eq. (1).

FIG. 6.

Picosecond valence-to-core x-ray emission spectroscopy of aqueous [Fe^II(phen)₃]²⁺. (a) Pumped (purple) and unpumped (blue) valence-to-core spectra recorded at a time delay of ∼60 ps. Shaded areas represent the standard error of all averaged scans. (b) Valence-to-core difference spectrum (red) compared with the scaled calculated high-spin minus low-spin difference spectrum (black line). The shaded area represents the standard error of all averaged scans. (c) Calculated low-spin (blue) and high-spin (purple) valence-to-core spectra. (d) Evolution of the mean standard error per energy point of the difference spectrum recorded at 60 ps with increasing measurement time (number of identical scans $N$). The black line represents a fit proportional to $N^{-1/2}$. The inset shows the mean difference spectrum after different numbers of recorded scans. The acquisition of a single scan containing 66 energy points takes ∼5.6 min.