I would like know how we can compress microsecond duration laser pulses by a factor of 1000 for a commercial aviation industry application.

Dr. Stephen Roberson Ph.D. and Dr Paul Pelligrino Ph.D. wrote in a 2016 U.S Army Research Library unclassified technical report titled "Compression of Ultrafast Laser Beams" that

A pulsed laser is normally considered an ultrafast pulsed laser when the pulse duration of that laser is in the realm of picoseconds and below. Ultrashort laser pulses deliver a very high peak power to their targets because of their short durations.

Professor Aaron Lindenburg, et al. from Stanford University wrote a 2016 Physics Review Letter article titled "Picosecond electric-field-induced threshold switching in phase-change materials" that

It is well known that the electric field required for threshold switching increases for short pulse durations.

If it is possible, I would like to use pulse compression to generate $1\ \text{nanosecond}$ duration pulses from $1\ \text{microsecond}$ laser pulses in order to reduce the required $340\ \text{kilovolts per centimeter}$ DC or AC electric field strengths.

In Dr. Arbore's article, the Stanford University researchers use second harmonic generation in chirped quasi-phase matching diffraction gratings to compress $17\ \text{picosecond}$ duration pulses to $110\ \text{femtosecond}$ duration pulses. I wish to find out why all the literature searches I have done discuss pulse compression of ultrafast pulsed lasers only when the pulse duration of that laser is in the realm of picoseconds and below.

We are using a simple pulsed CO2 laser with long pulse duration of millisecond order in a tube at a low pressure of less than $30\ \text{torr}$. The power supply for our laser system switches the voltage of the AC power line ($60\ \mathrm{Hz}$) directly. The power supply does not need elements such as a rectifier bridge, energy-storage capacitors, or a current-limiting resistor in the discharge circuit. In order to control the laser output power, the pulse repetition rate is adjusted up to $60\ \mathrm{Hz}$ and the firing angle of the silicon controlled rectifier (SCR) gate is varied from $30°$ to $150°$. The maximum laser output of $35\ \mathrm{W}$ is obtained at a total pressure of $18\ \text{torr}$, a pulse repetition rate of $60\ \mathrm{Hz}$, and a SCR gate firing angle of $90°$. In addition, the resulting laser pulse width is approximately $3\ \mathrm{ms}$ (full width at half maximum). This is a relatively long pulse width, compared with other repetitively pulsed CO2 lasers.

The amorphous semiconductor used by the Stanford University researchers was first used by Dr Stanford R. Ovshinsky Ph.D. The problem that held back large scale usage of the Ovshinsky diode was poor reliability. It would be very difficult to obtain F.A.A approval for using such an amorphous semiconductor device in the cockpit.

Time-bandwidth product

The time-bandwidth product of a pulse is the product of its temporal duration and spectral width (in Frequency space). In ultrafast Laser physics, it is common to specify the full width at half maximum (FWHM) both in time and frequency domain. The minimum possible time-bandwidth product is obtained for bandwidth-limited pulses. For example, it is ≈0.315 for bandwidth-limited sech2-shaped Pulses or ≈0.44 for Gaussian-shaped pulses. This means e.g. that for a given spectral width, there is a lower limit for the pulse duration. This limitation is essentially a property of the Fourier transform.

Here is the bandwidth specification and calculation I was hoping could be considered further if we can ignore the laser amplifier gain bandwidth product.

For a 1 microsecond laser pulse duration,

$$\text{bandwidth} = \frac{0.44}{1\times 10^{-6}\ \mathrm{s}} = 4.4 \times 10^7\ \mathrm{Hz}\ \text{FWHM}$$

so full width half maximum is 44 Megahertz (MHz).

  • $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. $\endgroup$
    – David Z
    Aug 15, 2016 at 16:52
  • $\begingroup$ @Moderator, Please protect this article for the Community. $\endgroup$
    – Frank
    Aug 23, 2016 at 1:43
  • $\begingroup$ It doesn't need protection as no bad answers have been posted. Anyway, I did edit it to improve the formatting and remove some bits that seemed unrelated to the question. When you edit, don't mark your changes as you had been doing. Just rewrite the question to make it look as if you'd written it that way in the first place. If anyone wants to see what changes you've made, we have the revision history for that; they shouldn't be identified in the question. $\endgroup$
    – David Z
    Aug 23, 2016 at 6:39

3 Answers 3


So this is a discharge pumped laser.

At 18 Torr the laser lines will be quite narrow. Do you know how much bandwidth you have in your current output pulses? I think this is important.

You may have the needed bandwidth (~1/1us). If so, then I'll look from a crystal with anomalous dispersion for the compression.

If you do not have the bandwidth then you need to create the bandwidth (see https://en.m.wikipedia.org/wiki/Bandwidth-limited_pulse). I am thinking pressure broadening the spectral lines at 10um is the way to go. If you do not have the bandwidth then compression isn't possible.

( this is the beginning of a full answer. I didn't want the comments to keep expanding)


There are several pulsed lasers available in the market ranging from milisecond to femtosecond (attosecond ??). They all have different purposes and different technology. The CO2 laser you are using will be probably the circulating gas High power CO2 laser. Usually employed in cutting welding drilling etc. First of all you can make the pulse duration shorter just by chopping (you will loose rest of the energy hence only loss no gain). CO2 lasers are available (as far as my knowledge goes) few picosecond short but not at 10.6 micron.

The duration of the laser in your case is limited by the duration of the discharge and not by the bandwidth of the laser. If you can decrease the duration of the discharge you can compress the laser pulse duration (I dont know if it is at all possible).

The duration of your laser is too long to be compressed externally. The basic principle of the external pulse compression is to device a way such that the earlier portion travel larger optical path than the latter portion and then synchronize them to make a short pulse. This method can be used to compressed picosecond pulses to femtoseconds, microsecond pulses to submicrosecond or nanoseconds but I could not think any external method to generate the short pulse in your case by 'external' means.

you may employ the technique known as 'Q' switching to decrease the pulse duration. In q switching you increase the loss in the laser cavity during the pumping and then suddenly decrease the loss (when the population inversion is at its peak) in such case the population inversion is depleted in few round trips and you get the short pulse.

Here you may note that in order to successfully employ the Q switching the duration of the pumping must be smaller than the lifetime of laser cycle or the population inversion will be deexcited via spontaneous emission route.

EDIT: The time bandwidth product hold true only for mode locked laser pulses. Usually in other pulsed lasers the bandwidth is much higher than required minimum. For ex for CO2 laser the required bandwidth for ms pulses is $\Delta\lambda \sim3.7\times10^{-10}$ micron for microsecond pulses it become $\Delta\lambda \sim3.7\times10^{-7}$ micron (no limitation posed by bandwidth) for picosecond pulses this will become $0.37$ micron (you should now be concerned). Note that above numbers are just for indication as the actual time bandwidth product depends on pulse shape.


  1. It is certainly possible that one can generate 1 ps CO2 laser pulses infect the theoretical limit is ~35 fs (single cycle pulse) but it is not possible to compress a millisecond pulse and make it microsecond. One may argue that if we can compress 1 nanosecond (stretched) pulse to 30-50 femtosecond femtosecond duration (factor of $10^{4-5}$) then why the same ratio can not be achieved for millisecond pulses, in this situation one must know that the length of the stretcher/compressor increased linearly with the length of the maximum pulse duration.
  2. It may be noted that any conventional stretcher/compressor that can be used to compress millisecond pulses must have dimensions ~100-1000 km.


There is only one solution to this problem, change the design of the laser such that the discharge timings decreases to microsecond levels. People have used RF discharge to generate microsecond CO2 lasers.

  • $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. $\endgroup$
    – David Z
    Aug 15, 2016 at 16:52

Compression of a pulse requires bandwidth, which is absent in continous, narrow band lasers. The one millisecond pulses will thus not be compressible unless additional bandwidth is generated.

If I understand your situation correctly, additional bandwidth will reduce the toal energy available significantly.

Chirped pulse amplification involves two steps: stretching a weak, ultrafast pulse over space and time so that the amplifier does not burn out, and then compressing the amplified bits. Successful compression requires phase matching in order to obtain a single output pulse.

Also see here.

Note: Fifteen terawatt picosecond CO2 laser system, OSA Publishing > Optics Express > Volume 18 > Issue 17 > Page 17865, describes a CO2 laser system providing ultrafast pulses with very high power levels. This is a MOPA system. Open access: (http://dx.doi.org/10.1364/OE.18.017865)

  • 1
    $\begingroup$ Dear sir CPA is for femtosecond pulses and he is talking about milisecond pulses. I dont think that there is any issue of gain bandwidth here. We have a picosecond laser chain (25 ps) having MOPA (master oscillator power amplifier) configuration without CPA. $\endgroup$
    – hsinghal
    Aug 13, 2016 at 1:01
  • $\begingroup$ The OP opens with the request to compress a ms pulse by a gactor of 1,000. He refers to pulse compression literature. I addressed his apparent misconceptions. Perhaps he needs a laser systems consultant. $\endgroup$ Aug 13, 2016 at 2:02
  • $\begingroup$ @Peter Diehr, Thank you for your great answer. I meant to say that I wanted to compress a microsecond pulse to a nanosecond duration pulse which hsinghal states is possible in his answer to this question. $\endgroup$
    – Frank
    Aug 13, 2016 at 21:05
  • $\begingroup$ @Frank: pulse compression implies that the total energy is preserved, while the pulse duration is much shorter - in your case 1,000 times shorter if compressed from ms to us; 1,000,000 times shorter if compressed from ms to ns. The approach used with CPA is to start with a naturally short pulse, and amplify it through a crystal with lots of bandwidth -- typically Ti:sapphire, which gives a pulse centered near 800 nm. This makes it a compound laser system (usually three lasers), including the CPA as the final step. This won't work for your existing laser system. $\endgroup$ Aug 13, 2016 at 22:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.