Table of Content:

CARS Facilities

Our CARS facility is part of the NRC-Olympus CARSLab. We use a Ti:Sapphire oscillator (Coherent Mira900) that produces 60-fs pulses with 80 MHz repetition rate, at a wavelength adjustable between 750-1000 nm. A variable beamsplitter (half-wave plate followed by a polarizing beam cube) sends 200-250 mW to a commercial photonic crystal fibre (PCF) module (NKT Photonics, FemtoWhite CARS) for the creation of Stokes light to be used for CARS imaging, and the remainder (>1 W) comprises the pump beam used for TPF, CARS, and SHG imaging modalities. A set of long-pass filters spectrally selects (>950 nm) the Stokes supercontinuum before it is chirped. The pump beam is chirped separately, before being sent to a computer-controlled delay stage and then recombined with the Stokes beam at the combiner. The combined beams are then routed to the microscope. Bandpass filters are used to discriminate TPEF, SHG, SFG and CARS signals from the pump and Stokes beams. Our single-femtosecond-laser-source CARS microscope relies on chirp-matching of the pump and Stokes pulses for optimal CARS signal and spectral resolution. Furthermore, the probed Raman frequency can be rapidly scanned by continual temporal tuning of the two pulses using a computer-controlled delay stage on the pump arm.

We have two microscopes available for use. The primary microscope used is an Olympus Fluoview 300 laser scanning microscope. This is an inverted microscope that we have modified for signal collection in the forward direction. While we typically use the system for nonlinear optical imaging, it is also capable of collecting DIC images (can be combined with nonlinear processes). This system can acquire images at approximately 1 frame/sec (512x512 pixels). It can also be equipped with an incubator for studying live-samples over long time periods (24+ hours). A second, home-built upright microscope is also available and can be used simultaneously with the Olympus microscope.

Velocity Map Imaging

Velocity map imaging (VMI) is a type of charged particle mass spectrometry. Time of flight spectrometers usually map ion energy to time. VMI spectrometers map ion velocity to position on a detector. This feature makes VMI a high-resolution ion imaging technique capable of determining highly differential information about the release direction and speed of ions, following a photophysical process.

Figure 1. Sketch of the VMI technique

The approach is schematically illustrated in Figure 1. Neutral products of a photodissociation, reactive event, or the neutral electronically excited molecules are ionized by a probe laser. The ions or the photoelectrons are projected by ion optics onto imaging microchannel plate detector. A CCD camera captures the 2-D image which can then be used to reconstruct the 3-D velocity distribution. The distribution contains all the the kinetic energy and angular distribution of the charged particle detected. The use of ion optics eliminates the spatial blurring of the image due to finite size of molecular beam, allowing velocity resolution on the order of 1% or better. VMI allows detecting multiple events and the detection efficiency is extremely high. It is a powerful, versatile experimental technique. Figure 2 shows the Molecular Photonics Groups’ VMI machine.

Figure 2. VMI Spectrometer

Magnetic Bottle Spectrometer

Time resolved photoelectron spectroscopy (TRPES) is a convenient method to study ultrafast nonadiabatic processes because it is sensitive to both electronic configurations and vibrational dynamics (see Ultrafast Molecular Dynamics ). In TRPES experiments, a photoelectron spectrum of electronically excited molecule is measured as a function of pump-probe delay. The pump laser pulse excites the molecules whereupon a time delayed probe laser pulse ionizes them. Emitted photoelectrons are dispersed according to their kinetic energy and/or angular distribution as a function of time delay. TRPES has several practical and conceptual advantages:

  1. Any molecular state can be ionized. There are no ‘dark states’.
  2. Charged particle detection is extremely sensitive.
  3. Higher order multiphoton processes, which can be difficult to discern in femtosecond experiments, are readily revealed.
  4. Highly detailed information can be obtained by differentially analyzing the outgoing photoelectron as to its kinetic energy and angular distribution.
  5. Photoelectron-photoion coincidence measurements can allow for studies of cluster solvation and effects as a function of cluster size.

The magnetic bottle time of flight (TOF) spectrometer is the simplest photoelectron spectrometer in our laboratory. It uses a strong inhomogeneous magnetic field (1T) to rapidly parallelize electron trajectories into a flight tube, followed by a weak homogeneous magnetic field (10G) to guide the electrons to the detector, a multi channel plate. The collection efficiency approaches 30% while maintaining an energy resolution that is in the order of the laser bandwidth (~0.02 eV).

Molecular beams are typically generated by Even-Lavie valves governing molecular pulses which are synchronized with the laser pulses. This allows for high target densities with gaseous, liquid and even some solid samples.

Coincidence Imaging Spectrometer (CIS)

The CIS machine is similar in concept to the PEPICO apparatus: electrons and ions are recorded in coincidence in order to obtain as much information as possible. The difference is that the CIS machine also allows imaging of ions and electrons through the use of position-sensitive detectors. In this way we can record energy, time and angle-resolved coincidence data, allowing us to fully disentangle nuclear and electronic evolution of excited states of molecules (see Time-Resolved Concidence Imaging Spectroscopy.)

Figure 1: Schematic illustration of the photoelectron-photoion imaging spectrometer

The basic principles of the coincidence imaging spectrometer are illustrated in Figure 1. At the center of the spectrometer, a molecular beam is crossed at right angles by two femtosecond laser pulses: a pump pulse which triggers the photochemical process to be studied and a probe which at adjustable time delays ionizes the molecule or the created fragments. Both the photoelectron and photofragment velocities carry valuable information about reaction, but measuring the two in coincidence can in some cases be the key to obtaining the full picture. A coincident measurement of photoelectron and photofragment velocities can be obtained using several approaches. The NRC CIS machine is using imaging Wiley-McLaren time-of-flight spectrometers, where both electrons and ion are extracted using homogeneous electric fields, to accelerate the charged particles onto time- and position-sensitive detectors. By measuring both the time-of-flight and the position on the detector the velocities are readily reconstructed.

Figure 2: Crossed delay line anode for measuring the position of ions and electrons on the detector

Each detector consists of a multi-channel plate (MCP) backed by a crossed delay line detector. The time-of-flight is directly measured by looking at the timing signal generated as the electron cloud leaves the MCP. The position of the ions impact on the MCP is measured with the crossed delay line detector as illustrated in Figure 2.

From the backside of the MCP a localized cloud of electrons is extracted onto two sets of charge collection fingers, one for each dimension. The electron pulses propagate down the fingers to serpentine delay lines where they bifurcate and travel towards both ends of their respective delay lines; the difference in arrival time at the two ends of the delay line is directly proportional to the position along that dimension. Using fast preamplifiers and carefully tuned constant-fraction discriminators to measure the arrival times leads us to high spatial resolution and fast read-out. Since all data are acquired and stored on an event-by-event basis, filtering of coincident electron and ion events is easily accomplished during data analysis. In summary, we accomplish simultaneous measurement of a photoelectron and a recoiling photofragment originating from a single molecule.

Figure 3. Pictures of the Coincidence Imaging Spectrometer

KHz Ablation Source

Not every molecule can be easily brought to the gas phase. Some melt at thousand of Kelvin (i.e. metals like Nb), others would rather decompose than melt upon heating (bio-molecules like guanine). A nice, common, elegant method to “evaporate” sample is by laser ablation. A laser beam focused onto the surface of a bulk sample deposits serious amount of energy within a tiny focus area. Small volume/portion of material gets instantly molt within femtoseconds-to-hundreds of nanoseconds time – the temporal width of the applied laser pulse. “Vapors” are picked up by a stream of supersonically expanding buffer gas (usually, but not exclusively, helium). This results in a directed and cool beam of sample molecules (translationally – down to sub-Kelvin; rotationally - down to tens to tenths of Kelvin; vibrationally - less coherent but still some cooling effect presents). Cooling not only makes experimental data look nicer but also allows creation of molecular clusters. Homomolecular clusters are very interesting and are must study as the intermediate/transition between a single molecule and its bulk phase. Heteromolecular clusters can be of use, for instance, when studying influence of hydrogen bonds between two understood molecules on their spectrum and/or intramolecular dynamics. Good example would be DNA-base pairs (in the gas phase!)

Figure 1. Schematic layout of the ablation source

The design of the ablation source is presented in the figure (can be found also in ref1). The target sample, a cylindrical rod of 6 mm diameter and up to 3 cm length, is rotated and translated so that a fresh surface is continuously exposed to the laser (10 mJ/150 ns pulse of 523.5nm Nd:YLF laser, focused into 0.1 mm spot). The motor is kept outside the vacuum chamber and rotary motion transmitted into the machine through special feedthrough and bevel gears. At the endpoints of the screw drive, the direction of the stepper motor is reversed and helical motion of the opposite sense occurs. Vaporized sample and sample clusters are swept through a “growth channel” – a tube (for metal clusters it is 2 mm dia. and ca.1 cm long) with a buffer gas under stagnation pressure of ~10 torr at the throat of the nozzle - and expanded into a vacuum chamber. The buffer gas provides conditions for the cooling and clustering of the ablated/desorbed material by collisions in the source exit channel.

The idea and basic design of an ablation source is not new. The sample’s area exposed to a laser is ejected to the gas phase shot after shot and soon the “drilled” channel becomes deep and narrow enough so that material is not fanned with buffer gas sufficiently and ablated sample deposits back on and around the surface it was lifted from a moment ago. Common solution is to rotate and translate the disk (ref2) or rod (ref3) made of a target material.

The unique thing about Ottawa design ablation source (ref1) is that it works in kHz repetition rate (as opposite to usual 5-20 Hz). Repetition rate is vital for the PEPICO type of experiments where for the correct data interpretation the count rate have to be well below the single event per laser shot (see the PEPICO machine page). To make this (high repetition rate) possible the classic design was revised. To protect the laser input window from debris from the target surface and to avoid clogging up of the channels in the source body (which would occur much faster at 1kHz) the buffer gas is supplied through the same opening the laser beam access the sample rod. Thus only one low-pressure hole is left in the source body, ensuring full and stable gas flow through the exit channel. Another feature is that just rotation and translation of a rod is not enough for long stable operation. Laser removes material at the high rate and draws a helical track on the rod surface. Upon driving the rod back and forth this track is reproduced and as the result a deep thread is cut into the rod leading to a strong diminution and eventual loss of the signal.

To avoid this uneven exposure, the rotation and up-and-down movement has to be decoupled. This is done in an asynchronous manner by dithering the laser focus with a ~0.1 Hz period and a ~1 mm amplitude of travel. With this approach, the pattern of laser spots on the rod surface does not produce regular helices. This simple but crucial modification allows ablation sources operation at kHz rate for over twelve continuous hours (i.e. > 4x107 laser shots).

Figure 2. An ablation source inside the PEPICO machine. A Campargue-type design implies differentially pumped source chamber.
Figure 3. A dummy rod and an ablation pattern on its surface. Obtained by simultaneous rotating, translating the rod and asynchronous dithering the focusing lens.

Transient Absorption Spectrometer

Transient absorption is a widely used form of pump-probe spectroscopy, where a white light continuum pulse serves as probe, in order to simultaneously interrogate the dynamics of all states in a broad energy range, whereby detailed information on molecular dynamics can be extracted, either directly, through looking at individual sections of the data, or through global fitting.

In a typical experiment, the pump pulse will be derived either straight from the Non-collinear OPA (NOPA), or from some harmonic or sum-frequency scheme, yielding typically ~ μJ of energy, in pulses as short as ~ 20fs.

Figure 1. Shown here is a sketch of the Transient absorption setup. Special care is taken to reduce dispersion in the white light continuum after it has been formed, using exclusively UV enhanced Aluminum mirrors for transport and focussing. The liquid sample can be either circulated through a thin flow cell, or for larger volumes and when better time resolution is required a jet can be used.

A small fraction of the 800nm fs pulse train is focussed into a 2mm thick CaF2 window, creating a single filament white light continuum (WLC), spanning the wavelength range 350nm to the Near Infra Red (NIR). This WLC is now simply passed through a thin 2% output coupler, in order to cut out the intens residual fundamental, and the very structured continuum around it. The continuum is then split into a probe and reference beam, each passing through the sample. The signal is overlapped with the pump, and both signal and reference are averaged on the CCD of an imaging f=300m spectrometer (Acton Research).

Varying the delay between pump and probe is accomplished using a folded delay stage.

PEPICO Spetrometer

Figure 1. The PEPICO-laser ablation station

PEPICO stands for Photo-Electron Photo-Ion COincidence spectroscopy. The idea is to do femto-second time resolved Photo-Electron Spectroscopy on clusters of different molecular compositions by post-selection of the corresponding ion mass. The fs probe laser pulse ionizes the molecule, creating one electron and a corresponding singly charged ion. The electron’s kinetic energy release is measured by a ‘Magnetic bottle’ TOF spectrometer, where a strongly divergent magnetic field (plus a weak guiding magnetic field) steers the electrons towards the detector without changing their speed. The ion’s mass is measured by a conventional TOF spectrometer, where the ion extraction electric field is pulsed after the electron has left. Both TOF spectrometers are designed for maximum detection efficiency, a property crucial for PEPICO.

Our PEPICO spectrometer will be used for studies on involatile metal clusters and fragile bio-relevant clusters such as solvated DNA bases or DNA basepairs. To this end, we will apply an NRC developed 1kHz laser ablation molecular beam source, which has been demonstrated to produce a stable beam of niobium clusters and of Guanine clusters respectively.

Figure 2. Pump-probe fs laser beam- molecular beam interaction zone of the PEPICO spectrometer inside the UHV chamber.

Time resolved Photo-Electron Spectroscopy has the advantage of disentangling electronic from vibration motion in the molecule and is hence well suited for studies of ultrafast relaxation dynamics in large molecules. By adding molecule after molecule to build up clusters of different size and composition, one can study how the ultrafast dynamics is affected when going from single isolated gas phase molecules towards more complicated systems. In a way we try to bridge gasphase studies with liquid phase or solid state studies.

Figure 3. Sketch of the interaction.

Figure 4. 20-tuple permanent magnet assembly work in the clean room, to be used for the electron magnetic bottle TOF part of the PEPICO spectrometer.
High Energy Oscillator

A modern Ti:sapphire femtosecond oscillator is now a common tool for use in many labs. There are commercially available oscillators that achieve an average power of 3 W with a repetition rate of 80 MHz. This is sufficient for many purposes and these types of lasers are commonly used for nonlinear optical microscopy; however, there are some drawbacks of this approach. Nonlinear processes increase as the square or the cube of the intensity. This means for nonlinear optical microscopy we want a laser with higher energy per pulse and/or a shorter duration. However, higher average power (i.e. using a 6 W laser to get 75 nJ per pulse of energy) is not feasible for nonlinear optical microscopy of cells because it will overheat the cells (literally cook them). Using shorter pulses is a possibility, but there are limits to how far this can be pushed.

Femtosecond amplifiers have much higher energy per pulse at similar average powers. However, the energy per pulse is so high we would have to attenuate substantially to avoid damaging the cell (laser machining would occur otherwise). Furthermore, we need several laser pulses per pixel to achieve good contrast in our image. In a standard confocal laser scanning microscope, images are 256 x 256 pixels or bigger. With a repetition rate of 5 KHz and requiring 10 pulses per pixel, it would take more than 2 minutes to scan one image, not very practical for observing any live cell processes.

For nonlinear optical microscopy, the ideal system is one that exists between these two extremes, combining moderate energy per pulse with high repetition rate and relatively low average power. Two routes exist to such a system, a high repetition rate amplifier or a low repetition rate oscillator. We have chosen to use a low repetition, high energy oscillator since in general, amplifiers are more complicated than oscillators and the repetition rate can still be too low for larger images.

The high energy oscillator
In order to achieve our goal of high energy per pulse at comparable average power, we chose to reduce the repetition rate of a Ti:sapphire oscillator. This can be done by greatly increasing the length of the laser cavity. A normal oscillator cavity is about 3 m long (3x108/3 = 100 MHz). For our laser, we use a cavity which is approximately 55 m long, giving a repetition rate of 5.5 MHz. With an average power of 2 W, this gives energy per pulse of 360 nJ. By ensuring the pulse remains short, we are able to generate the intensities we need to get significant nonlinear signal while still operating at high repetition rate.

To achieve the necessary path length while still having a laser of reasonable size, it is necessary to introduce many folds into the cavity. We do this by using a Herriot multipass cell that allows us to fold the beam twelve times, greatly increasing the path length. The long path length and high pulse energy can introduce instabilities during mode-locking. To inhibit these, we use a saturable Bragg reflector to ensure mode-locking and reduce the possibility of double pulsing. The overall layout is illustrated below.

Figure 1. Layout of the high energy oscillator.

It should be noted that this laser runs in the positive dispersion regime, so the output pulse is not transform limited (to ensure this, chirped mirrors are used). The intracavity pulse is chirped to reduce nonlinear effects inside the crystal. After the output coupler, the pulse is recompressed using chirped mirrors; however, because of higher order chirp, the pulse cannot be compressed to transform limited. The final output of the laser is expected centered at 800 nm with 2 W of average power at 5.5 MHz repetition rate and pulse duration of 50 fs (although bandwidth would support a shorter pulse). Below is a schematic of the laser.

Figure 2. Schematic of the laser.