For years, laser manufacturers have promised the ability to produce very high-power, cost-effective lasers that are stable outside of a laboratory environment. Recent requirements in aerospace and defense applications, such as LIDAR (Light Detection and Ranging), IR countermeasures, and laser targeting, are now calling on these manufacturers to deliver.

Figure 1. A beamshaper (top) redistributes the energy in the wavefront of a highly coherent beam to generate the desired, speckle-free profile over a limited depth of focus. In contrast, a controlled angle diffuser (bottom) overlaps multiple copies of the input beam to form the desired profile in the far-field. The diffuser is much more tolerant to changes in the input beam shape and modal structure, but the level of speckle in the output will be a function of the input beam coherence and quality.

Although some companies have stepped up to the challenge and successfully developed these devices, pushing the current limits of manufacturing and design does not come without a price. Generally, beam quality is sacrificed for higher output, either because multiple laser cavities are being used or because many modes are generated within a single cavity. Though occasionally the output will combine to form a beam that appears Gaussian-like in profile, more often than not these lasers have a structure that is neither Gaussian nor the frequently desired Top-Hat intensity profile. In addition, these intensity profiles often vary during operation.

Figure 2. An input beam with high, and possibly changing, structure can be remapped by a controlled angled diffuser to both homogenize and reshape the basic angular profile. In this image, an input beam with a Gaussian-like intensity profile has been transformed by a diffuser to a rectangular profile with uniform intensity in the far-field.

If the structure of the output beam intensity profile remains stable, and if the application only needs one particular profile at a limited projection range, customized refractive or diffractive beam-shaping solutions can provide high-quality and high-efficiency results. Essentially, beam shaping is the process of remapping the output angle for sections of the beam in such a way that the energy will overlap favorably at a specific distance. From Figure 1 (top), it is apparent that the distribution of energy is dependent upon the propagation length of the beam. As the projection length varies from that of the intended design, the device no longer functions properly. However, any variation in laser-to-laser output may require a unique design for each system. Furthermore, if the structure changes over time or during operation, these beam-shaping solutions may no longer provide adequate homogenization or profile redistribution.

Figure 3. Unlike a more traditional refractive element, which continuously reshapes the entire incident wavefront, diffractive elements typically break up each wavefront and interfere it with successive wavefronts to form the desired output pattern. This SEM image of the surface structures of a diffractive element shows small regions imparting discrete phase delays to sections of the wavefront. These structures are generally of micron, or even sub-micron, scale.

Fortunately, a solution exists that is much less dependent on the actual structure of the beam profile and more dependent on the nature of the structure. The solution comes in the form of diffractive diffusers and homogenizers that can produce very precise and controlled outputs, which change little from laser to laser. Basically, diffractive diffusers function by forming multiple replications of the input beam that overlap and are propagated forward in an array of defined angles to create a specific output geometry (Figure 1 bottom).

For the aggressive, high-power lasers in discussion, coherence is generally reduced to the point where speckle’s contribution to non-uniformity in the output beam profile is often negligible. This reduction in coherence is due to one of two factors. One factor is that independent cavities add together to reduce the speckle contrast. The reduction is defined as:


The other factor is that a single cavity often generates enough modes to result in an M2 value on the range of 10 or higher. Beam qualities of this level generally result in speckle contrasts of significantly less than 10%. For these cases, if the structure has a high enough frequency, or if the intensity contrast of the structure is low, the beam can often be homogenized with little penalty in divergence.

Unfortunately, there is no ideal solution. Every technological approach has trade-offs. Very high efficiency diffusers generally suffer from low quality uniformity and limitations to the control of the profile. High control over the output generally means a diffuser technology that results in an 8% to 20% reduction in efficiency. While this is not what the system designer, who has just spent much of his budget driving a laser solution to generate 15% more power, hopes to hear, this generally does not translate into a direct loss in overall system efficiency. Traditional methods of defocusing or clipping a beam to limit the degree of non-uniformity to the intensity profile almost always result in far lower efficiencies. For example, if a beam has a Gaussian profile (either to start with, or after some level of blur), and is clipped to hit a uniformity spec of +/-10%, only 20% of the light will be transmitted.

In general, most beams can be diffused such that well-controlled uniformity is provided with a roll-off fairly similar to the diffraction limit of the original beam. For lasers with structure that is very low frequency and high contrast, higher divergence angle copies of the original beam have to be overlapped to properly homogenize the beam. Because of the relationship of divergence to beam diameter, much of the optical train needs to be increased in diameter to reduce the divergence back to its original value.

As alluded, not only can these diffractive technologies homogenize a beam, they can also reshape angular distribution in an almost arbitrary fashion to accurately match that of the system designer’s target (Figure 2). For example, a circular beam may be remapped to a rectangular field to match an imaging sensor, at a loss of efficiency of only around 20%. This angular remapping function alone is generally worth the inclusion of the technology, but the homogenization of the beam is also built into the functionality.

One method of fabrication for these diffractive diffusers is through photolithography, also known as optical lithography. This process uses a series of coating, patterning and etching steps. A reticle or contact mask is populated with the appropriate design information. Through UV exposure, that design information is transferred into a photo-resist, a light-sensitive polymer, which coats a wafer surface. The wafer is then developed and this pattern is transfer-etched directly into the substrate. Depending on the specific requirements of these diffractive diffusers, these wafers can go through the photolithographic cycle many times. The resulting structures are generally less than twice the wavelength, and the substrates into which they are patterned are typically a millimeter thick (Figure 3). The driving factors for thickness of the substrate are mechanical stability and transmitted wavefront error.

This technology has been used in many aerospace and defense applications, such as target illumination, counter measures, mine detection and LADAR (Laser Detection and Ranging), where efficiency and uniformity are critical. At Tessera North America, these diffractive optical elements have been successfully designed, fabricated and tested in Fused Silica, Silicon and Germanium, for wavelengths ranging from 157nm-14um.

This article was written by Robert Hutchins, Principal Optical Engineer, and Jessica Wargats, Sales Engineer, Tessera North America (Charlotte, NC). For more information, contact Mr. Hutchins at This email address is being protected from spambots. You need JavaScript enabled to view it., Ms. Wargats at This email address is being protected from spambots. You need JavaScript enabled to view it., or visit .


  1. J. W. Goodman, “Some Fundamental Properties of Speckle, “ J. opt. Soc. Am. 66, pp. 1145-1149, 1976.
  2. J. W. Goodman, “Statistical Properties of Laser Speckle Patterns,” Topics in Applied Physics volume 9 (edited by J. C. Dainty), pp. 9-75, Springer-Verlag, Berlin Heidelberg, 1984.