Our lab studies the basic physics and applications of the interaction of extremely intense laser light with matter.


Below you will find an overview of our research, along with  expandable descriptions of some of the exciting physics and technology involved. There are also links to recent papers in each section, as well as a chronological list of selected publications here.

Spatio-temporal optical vortices (STOVs)


We have discovered the formation of “smoke ring” -like vortices that travel along with the laser pulse and control its global energy flow, a universal phenomenon underlying strong field laser propagation.

Our lab both predicted and demonstrated the existence of spatio-temporal vortices (STOVs), a novel type of optical vortex present in all nonlinear optical collapse arrest processes.



Prior experimental studies of optical vortices has focused exclusively on “twisted

light” beams with optical phase rotation about the propagation axis, Laguerre

Gaussian modes being a prime example. Such a vortex can be described entirely as

a spatial object in two dimensions and is not strictly embedded in the traveling field.


STOVs, on the other hand, can be tori embedded in the traveling optical field and

move with it, as described by our publication in Phys. Rev. X and displayed in

simulations on our publications page.


A simple toy model can help explain the origin of STOVs. If a one-dimensional

plane wave propagating in air has two areas with a large difference in intensity,

the intense half (core) will experience a larger nonlinear index transient than the

other half (periphery). This gives rise to phase front shear between the two

regions, culminating in a phase defect and field envelope null when the phase

difference reaches   . The defect launches two STOVs of opposite phase

circulation, as shown below. When viewed in 3-D, this circulation forms a torus

or "smoke ring," as shown at right.

















Spatio-temporal vortices appear to be a fundamental characteristic of strong field laser propagation, and after forming they continue to affect beam dynamics. Depending on the sign of the dispersion, the flow of energy (described by the Poynting flux) will be either saddle or spiral around the vortex ring. In addition, the STOV ring clearly identifies the spatial locations of the filament ‘core’ and ‘reservoir’, formerly qualitative descriptions. STOVs also explain how filaments can retain phase coherence over long distances despite a host of strongly phase shifting nonlinear processes, as we first observed in our study of multi-filament-induced air waveguides. Several articles in Physics, Optics & Photonics News, and the CMNS news site explore the details of our work, as well as possible applications in areas like optical communication.


Phase slippage between core and periphery due to nonlinear material response results in the formation of two STOVs. Image from [1].

Top,  a Laguerre Gaussian mode has phase circulation around a null at the beam axis. Center, a STOV is a ring shaped vortex with phase circulation forming a torus, as shown in the bottom figure. Top/center image credit: Nihal Jhajj/Howard Milchberg

Wakefield Acceleration in High Density Plasma


Using a high density gas jet enables relativistic acceleration of electrons in plasma wakefields with lower laser powers, leading the way to truly portable laser accelerators.

Accelerators are one of the basic tools of modern science, accelerating particles near the speed of light to probe fundamental nuclear and particle physics, or to drive medical imaging and industrial processes. The basic design for conventional electron accelerators is the radio frequency linear accelerator (RF linac). Since damage thresholds limit RF-structures to accelerating fields on the order of 100 MV/m, these systems are large, extending well beyond a kilometer for high energy experiments. Laser wakefield acceleration of electrons in plasma provides a way of shrinking the size of accelerators because lasers are capable of producing enormous fields, and there is no damage threshold in a plasma. Recent work in our lab on wakefield acceleration in dense plasma jets opens up the opportunity to shrink accelerators even further, using lower energy laser pulses.


Wakefield acceleration uses ultrashort laser pulses to excite large plasma waves, which in turn create large longitudinal electric fields capable of accelerating electrons to relativistic speeds in the laser's "wake," as shown in the figure below. While wakefield acceleration has succeeded in making high quality, high energy electron beams, multi-terawatt lasers have generally been required, and the number of accelerated electrons is relatively low. Our lab has developed a novel, cryogenically cooled, high density gas jet which enables laser wakefield acceleration of electrons with much lower laser powers, as related in our paper in Physical Review Letters.

Since gas from the jet is highly dense, the nonlinear response from electrons oscillating relativistically in the laser focus causes the pulse to self focus and drive even stronger plasma waves. This process continues until a substantial number of electrons are injected into the wake and accelerated, even when our laser power on target is well below a terawatt.


These results open the way to truly widespread laser-driven accelerators, since commercial high repetition rate femtosecond lasers with millijoule (subterawatt) pulses are common in scientific laboratories, unlike large multiterawatt systems. An article from CMNS highlighted our work and explores some of the possible uses, while other recent papers extended our work to a kHz system and mid-IR laser drivers, and studied radiation emitted during the injection process.

Left, Schematic of laser wakefield acceleration. Image from [2]. Right, experimental concept. Image credit: George Hine.


Waveguide creation from thermal expansion after filamentation. Image credit: APS/Alan Stonebraker


Filament waveguides from thin air


Heated air expanding in the wake of an array of laser filaments can guide light signals in a long-lived atmospheric “optical fiber.”

Femtosecond filaments arise from interplay between nonlinear self-focusing of ultrashort pulse laser beams in transparent media (such as air) and defocusing by field ionized plasma, resulting in a region of high laser intensity which propagates over long distances. Research into femtosecond filaments has explored applications such as triggering of high voltage breakdowns, remote spectroscopy such as LIDAR and LIBS, and generation of super broadband coherent spectra.  We have shown that by exploiting the strongly nonlinear absorption of propagating filament energy by the atmosphere, extremely long-lived waveguide structures can be ‘laser-written’ into the air.



As a filament propagates, the energy deposited in the air cause thermo-acoustic expansion, producing a density and refractive index minimum on axis that lasts for several milliseconds. By arranging multiple filaments in an azimuthal ‘cage’ shape, we imprint an effective waveguide 'cladding' surrounding an air 'core' of unperturbed density. We demonstrated that this long-lasting air waveguide can guide high peak and average power laser beams.. Our results, published in Phys. Rev. X and highlighted in Physics, pave the way for projecting high energy/power laser pulses over long distances through the atmosphere.


















Just as air waveguides can guide laser pulses, they can also collect remote optical signals. In later work, we demonstrated collection of the emission spectrum from a plasma spark with a ~50% increase in signal: scaling this result to collection distances of ~100 m would result in a signal improvement of 104. To fully characterize the waveguides, our lab also published papers on the shock waves which form them, and the energy deposition during filament propagation.

Combination of four individual filaments to guide a high power heater pulse. Image credit: APS/Alan Stonebraker

Microstructured Plasma Waveguides for Direct Laser Acceleration


Corrugated plasma channels enable electron acceleration that scales directly with laser field strength, making relativistic electron bunch generation possible with smaller, high repetition rate laser systems.

Applications of the interaction of intense laser pulses with matter, which include laser-driven particle acceleration and light sources such as x-ray lasers, coherent EUV/x-ray generation and terahertz generation, rely on achieving high intensities over a long interaction length.  However, diffraction imposes a trade-off between peak intensity and focal volume, since focusing a laser pulse more tightly in space causes more rapid divergence to a larger spot with lower intensities.


Waveguides, such as optical fibers, are the solution to this problem, extending the range over which the laser has a high intensity.  However, conventional glass fiber optics have a damage threshold set by the critical power for self-focusing, ~ 3 MW, rendering such optics useless for pulses that can reach 30 TW in our lab.  The only option for guiding extremely intense pulses is to make a waveguide composed of plasma—which requires a channel with a low plasma density on axis and higher density on the sides. Our lab first demonstrated the plasma waveguiding of ultraintense laser pulses, and is now investigating modifications that will increase the versatility of these plasma micro-optics.


Scaling down plasma accelerators with direct acceleration by a laser field


In conventional laser wakefield acceleration (LWFA), electrons are not directly accelerated by the laser field, but are instead pushed along by the electrostatic fields in the nonlinear plasma wake of an intense pulse. For a given plasma density, this demands a threshold ultrashort pulse laser energy—which can be rather high—to generate a plasma wake of sufficient amplitude. Alternative schemes that try to accelerate electrons directly with the laser field are limited to very short interaction regions near a material boundary or at a tight focus, strongly limiting the acceleration. Our lab has developed a variation on the plasma waveguide that enables direct acceleration of electrons by the laser field. By adding periodic modulations in the channel density or diameter, components of a guided, radially polarized laser pulse can propagate at less than the speed of light. This allows acceleration of electrons directly by the laser field, even with very low laser powers.


Direct acceleration is possible because a component of the radially polarized electric field points along the pulse propagation direction (along the waveguide).  In a vacuum or a straight waveguide, such a field would provide no net acceleration, since any forward push in the propagation direction would immediately be canceled  by a backward push when the electric field points in the opposite direction on the other half of the optical cycle.  The role of the modulated plasma waveguide is to emphasize the forward pushes and suppress the backward pushes. This “quasi phase matching” results in net acceleration which is directly proportional to the laser field and has no threshold—unlike LWFA—needed to achieve acceleration. A paper describing our acceleration mechanism is found here, while additional papers simulating improvements to the scheme are found here and here.











Experimental realization of corrugated channels


This goal has driven several advances in our lab towards a variety of modulated plasma waveguides. Initially, we explored imaging a ring grating onto the channel forming pulse to form these density perturbations, as pictured below. We then demonstrated direct modulation in density from wires placed over an elongated gas jet and explored the effect of shocks in this setup as the gas is cooled. Recently, we published results on using a programmable spatial light modulator in an interferometric setup to introduce arbitrary periodic modulations in the gas jet, as shown below and described here.


Left/center, radial modulations are focused by an axicon to create axial modulations. Image from [4]. More flexiblity can be obtained with wires obstructing the jet (not pictured), and even more so by a programmable interferometric setup, as shown at right. Image from [5]

Image from [3]

Measuring Nonlinear Optics in the Strong Field Regime


Spectrally interfering broadband, coherent white light pulses in a pump probe experiment enables single shot measurements of refractive index transients on a few femtosecond time scale.

Ultrafast laser pulses can strongly perturb any material they’re propagating through. Atoms and molecules can be distorted, stretched, rotated, and have their electrons removed by the laser field. These fundamental microscopic dynamics introduce strong feedback on the laser propagation, so details about these processes are fundamental to understanding and modeling nonlinear optics in intense laser fields. Using high bandwidth pulses, our lab pioneered the use of single-shot supercontinuum spectral interferometry (SSSI) as a way to precisely measure the response of atoms, molecules, and plasmas to ultrafast laser fields.


Our technique uses a chirped broadband, coherent supercontinuum pulse (generated by filamentation in a high pressure gas cell), splits it in two copies and sends it through a test region. The first pulse is a reference, while the second one (the ‘probe’ pulse) is overlapped with a pump beam that changes the local index of refraction and thus the phase fronts on the probe. Even though the reference and probe pulses are separated in time, they can still interfere in an imaging spectrometer, since a diffraction grating will stretch the pulses till they overlap as shown in the figure below. Shifts in the spectral interference fringes measured on a CCD camera in the spectrometer’s focal plane then enable extraction of the pump-induced cross phase modulation on the probe, and thus the pump-induced refractive index transient. Since each frequency of light in the chirped supercontinuum pulse arrives at the target region at a different time, the spectral information also gives a complete time history of the index over the length of the chirped pulse, with temporal resolution on the order of femtoseconds. A recent paper from our lab explores the absolute limits of temporal resolution with SSSI.















 Images from [6].              a


The technique is a big improvement over conventional multi-shot pump-probe experiments, since it captures the full transient dynamics in a single shot. This makes SSSI especially well-suited for single shot measurements of highly nonlinear events with shot-to-shot fluctuations, since a multi-shot pump-probe experiment would average over potentially important shot to shot variations in signal temporal structure.


Using this powerful measurement tool, we have been able to determine the absolute refractive index response arising from the electronic and rovibrational nonlinearity of a variety of materials and gases (such as air constituents, molecular hydrogen and water), including the response driven at the mid-infrared and long-wave infrared wavelengths new to ultrashort laser systems, and even showed how SSSI measurements can be a powerful diagnostic for spatial and temporal beam profiles. We also showed that the linear increase in this bound electron nonlinearity with increasing intensity continues well into the nonperturbative regime, where electrons are significantly ionized by the strong pump pulse. These fundamental measurements have in turn enabled related experiments in filament propagation and quantum revivals of rovibrational states in air. Studies of these quantum revivals further enabled demonstration of quantum molecular lensing in filament wakes, trapping and destruction of filaments, extension of filament propagation, and even coherent control of gas hydrodynamics. A review paper of the application of SSSI to studying filaments and their applications can be found here.


References (images)


[1] N. Jhajj et al., Phys. Rev. X 6, 031037 (2016).

[2] F. Albert et al., Plasma Phys. Control. Fusion 56, 084015 (2014).

[3] A. G. York et al., Phys. Rev. Lett. 100, 195001 (2008).

[4] A. G. York, et al., J. Opt. Soc. Am. B, 25, B137 (2008).

[5] G. A. Hine, et al. Optics Letters 41, 3427 (2016).

[6] K. Y. Kim, PhD Dissertation (University of Maryland, 2002).



Laboratory for Intense Laser Matter Interactions

Institute for Research in Electronics and Applied Physics

Energy Research Facility, Bldg. #223

University of Maryland



Copyright Laboratory for Intense Laser Matter Interactions © 2020