# In what applications are lasers necessary vs. just convenient?

I’m trying to understand, given the wide array of current laser applications:

1. Which applications actually require laser characteristics, as opposed to just use lasers because they are the most practical/efficient source of light with characteristics not unique to lasers
2. What alternatives exist, or may exist, to lasers for applications that don’t require laser characteristics

My key assumption here is that the essential characteristic of a laser is that it produces temporally coherent light. Often, but not necessarily, lasers produce light with two other useful characteristics:

1. Monochromaticity
2. Collimation

But these convenient characteristics can be produced using non-lasers. For example, LEDs produce monochromatic light. And lenses can take diffuse light and produce a beam collimated to near the diffraction limit, right?

If this is the case, then the only applications that require lasers are those that depend on interferometry, and that’s pretty much just holography and measurement, right? All of the other applications – directed energy, medical therapies, optical discs, optronics – could be accomplished with non-coherent collimated and/or monochromatic light, right?

Presumably lasers are used in these cases only because they are currently the most efficient or practical way of producing light with those characteristics. But now I am wondering: What other technologies could supersede lasers in practice (and efficiency) for applications that require light with high energy, collimation, and/or monochromaticity?

Here are a few things. Not a complete list.

Coherence - This makes holography possible.

Stability - Lasers are monochromatic because the light is produced by reflecting back and forth in a cavity. It is possible to select just one mode. With care, it is possible to keep that wavelength very stable over long periods of time. This is the basis of atomic clocks.

Power - Lasers come in a wide variety of power levels. Some have lots of power in the beam. It is possible to focus the beam to a small spot.

Efficiency - Currently, the most efficient light sources are LEDs. OK, an LED isn't necessarily a laser. You have to polish the ends of the crystal to make a cavity.

Switching rate - It is possible to turn an LED on and off very quickly. This is good for transmitting information at high rates of speed.

Plus a laser makes a much better cat toy than a flashlight does.

• Coherence was the only thing I thought was unique. The various pumping methods of high-power lasers might be unique, except I wonder about one: chemical light output. Do all scalable high-power chemical lights produce coherent light? Regarding switching and efficiency it sounds like LEDs are the best options, and you can turn an LED into a laser if you want to add coherence to those properties, but they're properties of solid-state light, not laser light. – feetwet May 5 '14 at 15:33
• +1 for the cat toy. Which also has motivated the tiny robotics industry. – Jiminion Feb 26 '15 at 14:18

It's true that laser's traditional applications are related to their monochromatic (one wavelength), collimated (one direction) and coherence ("one phase" or phase matching) characteristics. Laser beams have low dispersion, can be amplified and focused to reach very high photon densities and pulse shape can be sculpted at will.

Many applications are only possible thanks to one or a combinaison of these characteristics. Here are some specialized applications using strong field lasers:

• Filamentation allows us to probe targets at long distances (kilometers !)
• High intensity - up to more than $10^{18} W/cm^2$ - allows us to accelerate electrons in the relativistic domain.
• Less but still very intense field - about $10^{13} W/cm^2$ - allows us to distord atoms and molecules. Non-linear response of the molecule can be used to probe intra-molecular motions and to generate High Harmonics radiation (HHG).
• HHG can be used to produce isolated XUV pulses as short as $67\times 10^{-18}s$, which is comparable to the characteristic period of an electron.
• Can you elaborate in the context of the original question? I.e., would those applications be feasible with high-intensity non-coherent light? And does there appear to be any non-laser method of generating such high-intensity light? – feetwet Apr 6 '15 at 21:20

Pieced together from other discussions:

The only way to get large amounts of monochromatic and/or collimated light is through stimulated emission, which happens to also produce coherent light.

For example, if we want a more collimated LED then we we end up with a diode laser. And if we want to more collimate the output of an arc lamp we have to introduce a resonator chamber and once again we have a laser.