August 13th, 2020
The next-generation concepts need concentrators for which the features are, in many cases, not yet fully defined. It seems that most of the present next- generation devices will need the same features as those required for the present PV concentration cells and, thus, we will refer to them here.
The PV concentrator constitutes a specific optical design problem, with features that make it very different from other optical systems. Even designs for other optical concentration applications (like solar thermal energy, wireless optical communications, high-sensitivity sensors, etc) may not be suitable for PV. Generally speaking, PV concentrators must be 
(3) capable of producing high concentration (> 1000 x) for a high cell cost,
(4) noticeably insensitive to manufacturing and mounting inaccuracies and
(5) capable of producing sufficiently uniform illumination of the cell.
In order to be efficient and inexpensive, only a few optical surfaces can be used in the design. The present trend towards high concentration systems arises from the high cost of the present III-V solar cells (which can only be offset with a high concentration) and the promising efficiency results of these cells under concentrations as high as 1000x: over 30% with two-junction tandem cells  and over 26% with single-junction cells .
The insensitivity to inaccuracies may be important to achieve successful concentration products. This is related to the angular aperture of the concentrator, usually called the concentrator acceptance (half) angle a. Achieving a high acceptance angle (let say, a > 1°, which is close to four times the sun’s angular radius as = 0.265° relaxes the accuracy of the optical surface profiles, module assembling, installation, supporting structure, etc. Due to the inverse relation between acceptance angle and concentration (see section 13.1.2), a higher concentration for a given (sufficient) acceptance angle means a more problematic design. The present experience in concentrator systems is still too brief to understand the dependence of the electricity cost on the acceptance angle. This dependence seems to be much stronger in PV applications than in solar thermal concentration systems, as discussed in section 13.1.4, and, therefore, the longer experience with solar thermal installations is no help. Our opinion is that for mass production and low-cost installation, this parameter may be critical.
The illumination uniformity is probably the most singular aspect of PV concentrators with respect to other fields of application. However, getting uniform illumination is a problem in other classical optical designs (like the condenser design in projection optics) but in these problems no other features are necessary: cost is usually high, complexity is allowed and the optics exit aperture is oversized (which would imply a small acceptance angle a in the PV framework).