(a) Drawings of the MONSTR setup, left lower (“bottom”) and upper (“top”) decks are shown. The ultrafast laser together with the metrology laser (HeNe or the 488.5 nm laser) enter the bottom deck on the side as indicated by the red arrow. The first beam splitter (BS) divides the beam between the bottom and top deck. The beams are further split and pass through compensation plates (CP) for dispersion compensation. The green top deck is folded on top of the yellow bottom deck as indicated by the orange arrow. The entire resulting setup is shown on the right side. The metrology laser is reflected back by the large dichroic mirror (DCM) and the error signals due to the interference of beam Ref. and C (“Bottom deck error”), A* and B (“Top deck error”), and all four beams (“Between deck error”) emerge on the side and are detected by the photodiodes (PD). The error signals serve as feedback for high speed loop filter which compensate any drifts by controlling the voltage on the piezoelectric transducers (PZT) (b) Top view of the external 2DFT setup. In order to maintain phase stability between the FWM signal and the reference beam, the HeNe or the 488.5 nm laser propagates through the same path. The metrology laser light is reflected back by dichroic mirrors and the interference pattern is detected with the photodiode. Any mechanical drifts are compensated using a piezoelectric transducer driven by a fourth high speed loop filter. (c) The four phase stabilized linearly polarized beams obtained from the MONSTR instrument are focused on the sample, which is held in the cryostat at 5 K. The cryostat is mounted at a small angle with the perpendicular to the excitation beams in order to allow for the FWM signal to be collected in reflection geometry. A replica of the focus is monitored with the CCD camera. The interference pattern between the beams is used to obtain the phase difference between laser pulse pairs. (d) The sequence of the laser pulses used in the experiments, where A* corresponds to the phase conjugate pulse.
(a) The error signals for the “top deck,” “between deck,” and “bottom deck” interferometers recorded for approximately 5 min with and without active stabilization engaged. The active stabilization was achieved using the 488.5 nm single longitudinal mode laser. The “top deck” error signal only is shown for both cases. The measured mean interferometer displacement is ±1.62 nm for “top” and “bottom deck,” and ±1.55 nm for the “between deck,” respectively. (b) and (c) The “top deck” error signal during scan. In (b) the correction due to the stepping inaccuracy of the delay stages has been applied after every laser fringe (after 488.5/4 nm), whereas in (c) the correction is applied after four fringes (after 488.5 nm) (d) Photograph of the MONSTR and external optics using the 488.5 nm laser for active phase stabilization and providing four orange beams at 593 nm for the 2DFT experiment.
Flow chart of the data acquisition. In order to ensure identical steps during the scan, the error signals are read and the correction distance in nm is calculated. The calculation of the correction distance δ and sign is straightforward based on the deviation of the error voltage from zero, the sign, and the initial voltage slope. The correction distance is applied to the next step of the delay stages to ensure that the stepping error never surpasses λ/16. Larger errors in the stepping of the delay stages can cause the feedback loop filters and PZTs to skip an entire fringe.
(a) The real part of 2DFT spectra from CdSe/ZnS core/shell colloidal quantum dots at 4.5 K using an excitation density of 0.5 mW per excitation beam. (b) The real part of 2DFT spectra from GaAs quantum wells at 4.5 K using an excitation density of 100 μW per excitation beam.
Article metrics loading...
Full text loading...