Ultrafast Lasers - Science topic

Lasers can process different types of materials, such as metals, semiconductors, polymers and composites. Usually, by using pulsed lasers (e.g., short and ultrashort pulses), laser micromachining via ablation (material removal) is possible. Then, you can perform laser-based materials processing such as drilling, marking, engraving, and surface texturing. I believe that for carbon fibers nanosecond, picosecond, and femtosecond lasers emitting radiation at the near Infrared or visible ranges would work properly. There are plenty of laser sources commercially available which are capable to do the job. When it comes to fast and ultrafast laser surface texturing, developments to control optical properties, biological behavior, wettability of several liquids, adhesion, and wear can be easily found in the literature.

From my point of view, there is always the need of finding the most profitable combination between the laser radiation and the material properties, accompanied to the fulfillment of the technical requirements and the cost and time associated with the process.

Best regards !!!

Hi Amr, I agree with the others. And sometimes, I find it is very practical to visualize the two spots (fixed and delay lines) with a camera at a distance - the further it is located from the fixed and delay arms, the better you see the mis-alignement. This simple camera system is very handy for your alignment.

Good luck there

Satadru Chakrabarty

I would advise you to decide where your composite can be applied in photovoltaics, microelectronics, microelectromechanics, photocatalysis, corrosion prevention or biomedicine technologies. After that, the question of the mechanism of photocatalysis should be decided. It is associated with adsorption, the formation of metal nanoparticles, the adsorption-shuttle mechanism, the structure of the monolayer and nanoparticles. To what is offered by Yurii V Geletii , V. B. Zaytsev , you can add XPS, AFM

Do you want to synchronize pulsing of two lasers? I would assume that picosecond laser is likely self-oscillating but nanosecond one can be externally triggered. in such case if you need to synchronize them, you need to extract 78Mhz and turn it into digital clock which can be digitally divide by 278571 using some CPLD/FPGA board to make 280 Hz or so to clock pulsing of the nanosecond laser.

To extract 78 MHz you can use Phtodetector --> RF amplifier (e.g. Mini-Circuits ZFL-1000LN+ ) followed by some 78Mhz centered RF bandpass filter and then some high speed comparator board to obtain digital clock.

The reason for a "cleaner" LASER cut with pico or femto-second pulsed LASERs is because the pulse of light is too fast for the molecular interaction between the incident photons and the molecular structure of the material/substrate (the photon pulse is much faster than the molecular vibrations). With the CO2 LASER it is long pulse (greater than nanosecond) or CW and therefore the incident photons are on the same time scale molecular vibrations in the material and thus interact with the molecules causing them to vibrate to the point of breakdown. The fast LASER pulses are on a quantum time scale (electronic dipoles only), which means they interact with material at a quantum level (energy is always conserved); this creates a plasma plume that ejects the obliterated material. For transparent bulk materials (crystalline or amorphous glass) if the intense pulse is below the surface there will be no obliteration, but rather a deformation in the dipoles of the material causing an index modification.

Yes QCL lasers have been used to excite phonons in solid state wide gap semiconductors as GaN.

This book chapter may help you:

1. Coherent phonon excitation in wide gap semiconductors

https://books.google.dz/books?id=6-Y8YnOfK7QC&pg=PA149&lpg=PA149&dq=quantum+cascade+laser+phonon+excitation&source=bl&ots=BHZmWt-F2G&sig=ACfU3U1B36YD6JkebqhFOUI1oRH2rZqInw&hl=fr&sa=X&ved=2ahUKEwip7b2M_-npAhUb5uAKHfcYB4EQ6AEwCXoECAoQAQ#v=onepage&q=quantum%20cascade%20laser%20phonon%20excitation&f=false

An expensive solution could be to insert an optical isolator, but keep in mind dispersion of the crystal in this case.

Hi Cheng,

I suppose you would not only like to reach a high conversion efficiency, but also keep a good pulse shape in space and time. This is not so easy. I am afraid, at 1.6TW/cm2 you are ovepumping quite strongly, and the pulse shape is considerably distorted. If you really want to understand what is going on in your crystal and get an impression of an optimum pumping, I would recommend you to dounload the free SNLO software by Dr. Arlee Smith (http://www.as-photonics.com/snlo). It is quite good and easy to understand and work with. You can even make a movie of your pulse and the second harmonic propagating through the crystal.

Best regards,

Alexander

Dear Raul,

Thank you for your response. I got the point. I will find that talks and will study them.

Best regards,

Some relevant articles are as follows that you can follow.

a) eprints.whiterose.ac.uk/86590/1/paper_39.pdf

b) https://link.aps.org/doi/10.1103/PhysRevB.95.054305

c) cpb.iphy.ac.cn/EN/article/downloadArticleFile.do?attachType=PDF&id=26057

d) pubman.mpdl.mpg.de/pubman/item/escidoc:2262490/.../PhysRevX.6.021003.pdf

We have to pay attention that Nd:YAG is 4 -level laser system. In opposite 3-level dye or ruby system no absorption laser emission takes place on the case. In the same way no gain exist without pumping pumping of this part of laser element. But output energy from laser decreases because last one is proportional of pumped volume. Similar conclusion results from the main oscillation condition: gain should be equal or more than losses.

My best regards!

Hello,

You need to have a synchronous chopper and a CCD which has the capability to be externally triggered.

If you are using an amplified laser system based on a titanium sapphire laser, you may derive a trigger signal from the regenerative amplifier to synchronize the chopper and the CCD to the laser.

First, to check that your chopper is synchronized to the laser pulse train you can put the chopper in the attenuated laser beam an set a reasonably fast photodiode after the chopper . Then you observe the output of your photodiode. If the frequency of the chopper is lets say, one half of the pulse repetition rate, you should see that exactly one every other laser pulse are cleanly passing trough the chopper. The oscilloscope, trigger on every laser pulse (use the trigger output provided by the regenerative amplifier or another fast photodiode to trigger the scope). Set the scope on a time scale such that will show say 5 to 10 pulses, should show a stable comb-like pattern with no fluctuation of the photodiode signal amplitude. Any amplitude fluctuation or modulation of the amplitude of the comb means that your chopper is not synchronized (beat pattern). You may also use two channels of the scope one for a photodiode on a portion of the laser beam before the chopper (ch1), and another after the chopper (ch2). Then trigger the scope from channel 2 and watch the two channels. You will see two comb patterns. If the chopper is synchronous to the laser the two combs should be locked and stable. If one comb displaces in time relative to the other, then your chopper is out of synchronism.

Synchronizing the CCD will require more information. But basically you trigger the CCD on every laser pulse but chop the pump at half the repetition rate of your laser. Then you collect a continuous series of alternating spectra with pump on and with pump off. There are several ways to do this and it will really depend on what you have in your setup So i will defer my answer for this part until then.

Dear Eyad Abdur-Rahman,

it depends on how you define the widths (in the time and the energy domains) and what shape the pulses have. If the widths are considered at the half maximum and your pulse has a Gaussian shape, you will need about 18 eV. The central frequency is not important.

Regards,

AT

Actually in our case. Yes their is a plasma formation so I dont know the exact effect in this case on the free electrons in the surface. But thank you for your answer it open some other explanation for me ..

Dear Anabel,

counts per second is what your detector delivers, but the unit contains no information on the actual cause of the counts. In practice, you always have "false" counts, for example caused by noise or by light from the environment. Photons per second is what you want to measure- but how can you be sure photons being the cause of any signal from your detector? You see, I hope, that with “counts per second” you are on the save side. But in most cases there is no difference between the units, provided you have taken care that the number of “false” counts is small compared to the true signals.

Good luck with your work

Max

The topological charge of a pulsed laser beam can be measured by interfering the vortex beam with a reference beam that has a flat phase. The number of nodes in the interference pattern is a measurement of the OAM in the vortex beam.

See the following article: G Gariepy et al. Creating High-Harmonic Beams with Controlled Orbital Angular Momentum

Hi Hongyan,

For a given pulsewidth, the phase of the pulse is what is going to determine the spectral bandwidth. If the phase varies linearly over time then the pulse will be transform limited (narrowest spectrum possible), if the phase varies quadratically, then the pulse is said to be positively or negatively chirped, depending on the sign of the curvature. The Fourier transform of such pulse gives a wider spectrum, the more dispersion the pulses has and the wider the spectrum gets. This kind of second order dispersion can usually be compensated using a pulse compressor (gratings, prims, chirped DBR,...). Coming out of a fiber, it is likely that the pulses will be positively chirped (easy compensation), but could also have some third order dispersion (cubic phase variation) which is more challenging to compensate.

I hope this helps.

Do you have sample geometry issue? If you work on tissue, I imagine the samples are somewhat freeform? A sample stretcher, clamp, or a jig could help you to set your samples to a prefixed form, then you will not have problems with finding focal plane each time you ablate a new sample.

Here are two option that fit your desired parameters:

Low RIN diodes you can find also from EM4 or Eblana Photonics with RIN<160 dBc/Hz. However, the output power is rather low. Seminex Corporation provides high power, but i am not sure about the RIN.

Actually i am using optical frequency comb for Absolute distance measurements. and interested in Factors that affect the accuracy of measurements 

There are a couple ways to look at this: one is that the time of a pulse is related to the energy by way of a Fourier transform. A narrow peak in time in represented by a wide band in energy (or equivalently, frequency). Also, you might simply realize that if you isolate a pulse quite tightly in time, the uncertainty principle requires that the energy of that pulse be rather broad. Energy (frequency) and time are complementary.

So this really depends on what you want. If your just want the basic information an auto-correlator will do the job, you can build your own, just need a crystal some optics and a translation stage. (simple! :-)) you can do this for an order of magnitude less money than paying Swamp Optics $10k. However, the auto-correlator will only give you symmetric information, i.e. you won't know if that feature is a prepulse or a postpulse. But, for that matter you can build your own FROG for the same $1000, but then you need software to analyze the spectral info, which might be great for a student to do, but isn't really thesis work. In any case, for more detailed info you need a cross-correlator. (Or a FROG or SPIDER). This much more involved, and is commercially available. Del Mar Photoics makes a good cross-correlator, their web site SUCKS, but their products are good. But if you need spectral and phase info, or worry about pulse stability buy the FROG. Del Mar makes one too, have no idea how it compares to the GRENOUILLE. Hope this helps.

I'd like to add three of my favorites:

1) ultrafast spectroscopy and coherent control of single molecules. This goes hand in hand with the already mentioned pulse-shaping techniques, ultrafast nanoscopy and in-depth characterization of ultrafast pulses. I'm still amazed by the fact that it is possible to detect and manipulate the optical response of single particles and to compare the result to theoretical predictions. 

2) 4D ultrafast electron microscopy. It extends the concept of ultrashort pulse generation to electrons and opens up completely new insights into structure and dynamics of matter under various physical and chemical conditions, for instance phase transitions, particles in optical near fields or spatially-corellated systems.

3) although already fairly old and not very surprising, starting out from the photon-echo experiments, I still consider the development of multi-dimensional optical spectroscopy as ground-breaking. Clear visualization of chemical bond transformation, energy transfer between molecules and relation to chemical structure is in my opinion a very important development that still holds great potential for instance in medical research and energy related science.

Up to the quadratic order of dispersion, the effect of propagation through material on the temporal duration of an initially-transform-limited Gaussian laser pulse is given by:

T_out=T_in*\sqrt[1+(4*(\alpha^2)*(z^2)/((T_in)⁴)]

where T_in is the starting pulse duration (T_in=T_FWHM/(2*Sqrt[ln(2)])), T_out is the final pulse duration, z is the propagation length, and \alpha is the quadratic dispersion term given by

\alpha = (n'(\omega)/c) + (\omega * n''(\omega)/(2c))

where n is the index of refraction as a function of \omega, \omega is the angular frequency (evaluated after differentiation at the pulse's central frequency), and c is the speed of light.

For BK7 glass, and assuming that your Gaussian pulse has a central wavelength of 800 nm (the answer would be slightly different otherwise):

Remember, again, that the 1/e pulse duration rather than full-width-half-maximum (FWHM) pulse duration is used in this calculation.

To answer the question of "What is the difference between temporal dispersion of 6fs,18 fs and 30fs Gaussian pulses in glass?", the difference is the bandwidth: because a 6 fs pulse has a much broader bandwidth, the opposite ends of its spectrum will experience significantly different indices of refraction. Therefore, shorter pulses see their pulse durations increase more than longer pulses going through equal amounts of material.

See the link for more information.

Following Cameron: Based on Raman scattering measurements A. Compaan thought that pulsed laser irradiated Silicon melted at "low" temperature.

- Raman Measurement of Lattice Temperature during Pulsed Laser Heating of Silicon

H. W. Lo and A. Compaan, Phys. Rev. Lett. 44, 1604 (1980)

But after careful experimental calibrations by him and others, he later came up to publicly retract and acknowledge that laser heated silicon melts at the same temperature as adiabatically heated silicon.

Still later on, femtosecond experiments with photon energies larger than the atoms' binding energy of silicon, showed that:

- Time-Resolved Reflectivity Measurements of Femtosecond-Optical-Pulse-Induced Phase Transitions in Silicon. C. V. Shank, R. Yen, and C. Hirlimann 

Phys. Rev. Lett. 50, 454 (1983)

- Femtosecond-Time-Resolved Surface Structural Dynamics of Optically Excited Silicon. C. V. Shank, R. Yen, and C. Hirlimann

Phys. Rev. Lett. 51, 900 (1983)

Coming to your question : from the above arguments, for large light intensities matter is ionized on a picosecond time scale and ions and electrons are expelled outwards the focal volume being at the origin of the plume one observes in laser ablation.

For ultrashort light pulses inertia of matter is too long for droplets to be expelled, while for longer pulses, light energy still impinges the molten matter once it has formed leading to mechanical instabilities in the liquid at the origin of matter projections.

Melting or not melting is a second order question as the mechanism involved in the ablation process is more complex.

Hope it helps.

 

Zsuzsanna:

SHG will never occur in liquids, due to inversion symmetry liquids posses.

In random ensemble of nano particles, SHG will probably not exist either, for the same reason. SHG require an-isotropic  medium. HyperRayleigh scattering can be produced in virtually any medium (there are restrictions on micro symmetry, but it is hard to find a material that would violate those for all spatial axes). 

Moreover, SHG is nearly always polarization dependent, while HyperRayleigh is not.

Hope that helps.

Thank you all for your helpful comments.

Dear Dr Pendyala,

You can read more about TCSPC in the book that I link you below, but I can give you a general description about the technique.

TCSPC actually measures the time-delay between two signals, the signal of photoluminescence coming from the sample and the signal of an electrical pulse generator which triggers the laser (or LED) source. After being triggered by the electrical pulse generator, the laser (LED) produces pulsed light with the same time profile as the electrical pulse. The same electrical pulse was sent simultaneously both to the laser (LED) and to the detector and that time was set as zero. So, the detector compares the time-delay between the two signals, the signal which triggered the laser and the photoluminesce from the sample, and then it puts a count at the specific point which corresponds to the specific time delay of the photon relative to the electrical pulse. The above mentioned procedure is repeated many times and so more photons are added on the histogram.

The term single photon counting comes from the fact that in the time interval between two consecutive excitations (between two consecutive pulses) only one photon, emitted by the sample, can be detected. When this photon is detected it is put on the histogram and is added to the previous photons which had the same time-delay with it. That is why in every cycle the counts on the histogram grow.

Of course, as you will see in the book, the technique is more complicated than I described you as there are many electronic elements involved and some key steps such as how exactly is the time-delay between the two signals recorded.

I hope this answer to be helpfull to you.

The conquest of knowledge has not limit, we must go everywhere we can go, explore every corner of the Universe.

I agree, this answer is too general. Here are some more practical arguments.

- If you want to picture a moving car with a camera you need to expose the film no more than about 1/100th of a second for the picture to be sharp. If you would like to picture the flying bullet of a firearm you would live the camera open in a dark room and shine the bullet with a nanosecond flash of light at the time it passes in front of the camera. When exploring the fastest dynamics of electrons in semiconductors you need femtosecond light pulses, and you want to explore the fastest dynamics of electrons because you want computer to run ever faster.

- This argument does not apply to attosecond light pulses because even the light objects like electrons do move too slowly at an attosecond time scale. The second strong argument favouring short pulses goes as follows. If you can put as much energy in an attosecond light pulse than in a nanosecond light pulse then the peak power of the attosecond pulse is a billion times larger than the peak power of the nanosecond pulse. When we will be able to produce pulses with peak powers larger than 10²⁰ W/cm² then we will be able to observe the interaction of two visible photons. In the mean time we will explore the field of photo induced nuclear reactions.

Hello Donggyu,

Have a look at this article by Durfee et al.

C. G. Durfee, S. Bera, A. J. Sabbah, J. A. Squier, and M. Ellison, “Spectral shaping filter for broadband amplifiers,” Optics Communications, vol. 263, no. 2, pp. 256–260, Jul. 2006.

Using these filters and techniques, I have been able to produce <25fs, >4mJ pulses from a Ti:sapphire regen.

Best regards,

Alan

fs lasers have very small energy per pulse, making it only possible to ablate very small spots, making very small plasma with little emission. You need a good optical setup and high sensitivity spectrometer, compared to ns LIBS. But the small ablation per pulse can be advantage in some cases like thin layer depth profiling etc. I dont have experience with attosecond laser but I guess its the same

Hello,

not me.

Have you tried to search for it on Addgene ?

They sell all kinds of plasmids for very cheap...

Clem

Before you send back your laser, check if your  TCSPC  detection is, in your case, slow enough, some of the devices have only a detection window for 10ns or 100ns (longest scale one can measure) the rep rates in his cases are not crucial, but if you are sure that it works try to get a  laser with tunable rep rate form 2MHz to some kHz.   Check also the following paper, might be useful: "Size- and temperature-dependence of exciton lifetimes in CdSe quantum dots" in Phys. Rev. B 74, 085320 (2006) by C. de Mello Donegá, M. Bode, and A. Meijerink, they measured similar CdSe QDots

Dear Marco,

May be an article published in Quantum Electronics v.42(12), p.1097 (2012) will allow you to get answer.

I'll add a nuance to the perfectly correct answer above. Where you find a Cooper minimum experimentally depends somewhat on how you look for it. If I remember correctly, the copper minimum in argon appears near 42eV if one looks for a minimum in the photo ionization cross section by counting argon ions. It shows up closer to 39eV if one resolves photoelectron energy and counts only the 3s electrons. This is the interference mentioned above. The sign change mentioned above is related to the radial node of the initial orbital and how it affects the overlap with an outgoing wave.

I seem to remember it moves around a bit in high harmonic generation, being found in different places for the long paths and short paths. There should be some papers on this from the Bucksbaum group, I think, with Limor Spector as one of the students.

Well, optical tables, clean room accessories, and laser safety stuff to start with !

I would then have a look at some metrology tools: spectrometers, autocorrelators, cameras, photodiodes, oscilloscopes. And then look for optics and opto-mechanics suppliers.

Hello,

1) There are commercially available ultrafast fiber lasers that provide the tunable output.

2) There is no reason why any of the very compact and reliable ultrafast fiber lasers (there are models that are smaller than typical computer keyboard with 50-100 fs and 50-100 mW) would not be used for 'seeding', or as Marco puts it 'driving' regen amplifiers. Few commercial companies have been doing that for quite some time.

3) There are commercial fs fiber lasers with < 20 fs pulses

4) As no one commented on price yet: fiber fs lasers should be 3-5 times cheaper than similar Ti:Sapphire oscillators. Of course depending what are you comparing it to.

However, as valid for most things, go for high quality and not lowest price.

With increased reliability and smaller footprint (push-button operation for most fiber fs lasers) you are sacrificing bit of the power. For 50-100 MHz that can be compared to most often used T:Sapphs you will be getting few 100 mW for <100 fs vs 1-4 W with ~100-150fs from Ti:Sapphire.

But for fiber - no pump laser, not chiller, no tweaking the cavity...

In short fs fiber lasers can cover most of the ~450-2300 nm range and you can also access MID IR (2um -~15um) with very simple arrangement.

The Kerr effect depend on the full structure and adding a dielectric layer should have an impact on the reflectivity (depending on the extra optical path added).

Assuming that the incident light is s-polarised, in order to measure the Kerr angle, you probably measure both reflected components (not rotated s and rotated p). The Kerr angle will be given by their ratio rsp/rss.

Adding a dielectric layer on top of a magnetic layer should modify mainly the not rotated part. A usual trick is to choose the thickness of the dieletric layer to a quarter of the wavelength in order to make it an ARC (anti reflecting coating), so rss very small. The Kerr angle becomes then giant even if rsp does not change so much. Giant Kerr angle does not mean more light in the rotated polarisation.

In your case, if you really measure the angle, i m surprised that it does not change when you add PDMS (2 microns I guess, not 2 mm ?), except if you are close to half a wavelength thickness. If you measure only rsp (the intensity of the rotated component), then it would change less with PDMS.

@Nguen - The answers depend on what exactly you are talking about. E.g. for discharging (or recharging) an electret, it might be enough to heat only electrons. For lattice changes, at least some of lattice temperatures need a possibly temporal increase.

Plasma does absorb light. And reflects it, too. Therefore, for a transparent material plasma can be a mediator of energy transfer from light to heat in the material, like in water at visible light. And plasma can shield the material from light, like plasma at a metal surface...

Do not forget, that microsecond and femtosecond pulses are as different as a heart beat and a year season, and effects might be very different. E.g. in fs optics, self-focusing is rarely as big problem as with micro-nanosecond lasers.

At low energy you can try to mask progressively the spectrum between the gratings. If doing so you observe a displacement of the focal spot on your camera, this means that you have an angular chirp caused by non parallel gratings (the different wavelength are focused on different positions on the camera) . Then you slightly turn one of the gratings with respect to the other and you should round the focal spot down to a reasonnable size, mainly limited by the aberrations of your beam (+ that of your imaging system).

If the mark on the grating is at the last pass, this could be pollution (are you in vaccuum ?) and this could be cleaned (don't do it by yourself, ask to your grating provider). If it's at the input, this could be burned and this is not reversible. However this is probably not the cause of your problem.

I would like to add in this thread. The question is about calculating intrinsic non-radiative decay rate of luminescent materials. The total non-radiative decay rate can be considered mainly to be composed of three factors Knr=Kni+Kne+Knisc. These terms are :

Kni=intrinsic non-radiative decay rate and is basically an internal conversion process in which electronic energy is converted into the vibrational energy of fluorophore and since vibrational energy is driven by thermal process, it increases with the increase in temperature. Due to this, fluorescence intensity decreases with the increase in temperature.

The second term Kne is external conversion rate and it depends on the conversion of internal electronic energy to external species. This term is often referred as external quenching. In this process, the fluorophore loses its energy by collisions to solute and it also includes concentration quenching. This term is the major factor to be separated from internal conversion rate.

The third term Knisc is the inter system crossing from singlet electronic state to triplet electronic state. This term is generally negligible in dyes and this is only substantial for heavy metal complexes.

Now the question comes that how can we distinguish between Kni and Kne. We can determine Kne by a series of fluorescence experiments. If the quenching process is dynamic i.e. if it occurs from excited state, you can see a change in excited state life time by changing the concentration of the fluorophore and this change is given by the equation t0/t=1+k0t0Q, where t and t0 are the life time of the flurophore normally and without any non-radiative decay process, respectively. Therefore t0 is very difficult to find out because we can not have a situation where we do not have non-radiative decays. Stern-Volmer equation gives a relationship between the fluorescence intensity and the concentration [F0/F=1+k0t0Q]. By measuring the fluorescence intensity at different concentrations, we can find out the the product of k0 and t0. By measuring the excited state life time at different concentrations we can evaluate ko and kec is koQ. This way we can determine the external quenching rate.

Now if Kne is due to the ground state then this process is referred as steady state quenching and we will not get an change in excited state life time with concentration. In this case the equation of quenching is F0/F=1+KsQ and it does not include t. therefore, it is very easy to determine Ks and then Kne=KsQ.

So we can evaluate Kne and then we can calculate Kni. Kne does not depend on temperature, so we can determine Kni at different temperatures and have some conclusion.

This might help.

This is my thinking according to my understanding. There may be some things incorrect, any corrections will be helpful for me also.

I think you are talking about ripping apart the vacuum? This is known as the Schwinger limit (after Julian Schwinger 's paper 1951 http://dx.doi.org/10.1103%2FPhysRev.82.664 ) This is essentially quantum electrodynamics, and such an experiment would test the fundamental validity of the theory. Plans are underway to build a laser capable of exceeding this limit, which is 1.3E18 V/m electric field, this field is equivalent to the rest mass energy of the electron over a Compton wavelength, or equivalently 5E29 W/cm^2 laser intensity, but effects should start around 1E27 W/cm^2, and vacuum birefringence becomes measurable at 1E22 W/cm^2 (inelastic photon scattering has been measured at SLAC by C. Bamber et al. PRD 60 092204 (1999)) The laser is part of the European ELI consortium, ( http://www.extreme-light-infrastructure.eu/ ), which seeks to build 4 large laser facilities in Europe for different physics reasons. The facility to test QED will be the ELI-Ultra High Field laser to be determined this year. QED theory predicts (unlike classical Maxwell equations) that photons will interact elastically with each other (photon-photon scattering), which creates virtual electron-positron pairs (Feynman diagrams describe this). ELI will still likely have to use tricks to reach the Schwinger limit like attosecond pulse generation, but can reach the 1E22 W/cm^2 turn on. This is explored in section B.VI.1 of the ELI science document. ( http://www.extreme-light-infrastructure.eu/pictures/ELI-scientific-case-id17.pdf ) These experiments could resolve the issue of dark matter and dark energy, or lead to new physics beyond the standard model by measuring the expected yield of particles and anti-particles from the focal spot particle showers.

References:

S.S. Bulanov et al PRL 105 220407 (2010)http://prl.aps.org/pdf/PRL/v105/i22/e220407

ELI Document : http://www.extreme-light-infrastructure.eu/pictures/ELI-scientific-case-id17.pdf

J. Schwinger PR 82 664 (1951) http://dx.doi.org/10.1103%2FPhysRev.82.664

C. Bamber et al., "Studies of nonlinear QED in collisions of 46.6 GeV electrons with intense laser pulses", Phys. Rev. D, 60 (1999) 092204. doi:10.1103/PhysRevD.60.092004