You should know that;

just at the expense of a very, very lot of intensity by a set of slits.

Intensity is the currency in optics.....

Thanks sir. A slit would help to extract the partially coherent part.

Did you mean a series of nearly infinite numbers of slit?

Yes, a number of parallel horizontal sheets (slits) in the direction of light travelling will give you parallel beam (but will affect the intensity by eliminating the non parallel rays). Or you can use sets of mirrors to give direction to the beam and make them parallel. Various mirrors are available in accordance with the wavelengths.

Thanks Akeshwar!

However, do we have any more efficient ways that would greatly minimize loss from small-apature slits?

Here the output light intensity is critical and assume that >80% of the incoherent-to-coherent conversion is needed.

A pair of slits having appropriate slit widths and distance of slits is enough. It make no sense to have up to 'infinite' number of slits in direction of propagation of the light.

Perpendicular to this direction you may set up the 'next' set-up of slit system. But also here it makes no sense to go for an infinite number of slit systems while decreasing the relative distances between them to 'zero'. You cannot violate the entropy law. It is like to invent a perpetuum mobile.

Each of these individual discrete beams are partly coherent with respect to themselves but they are in general not coherent with respect to the neighbors.

In order to be coherent relative to one another they have to arise come from the 'point' of the source.

@ Akeshwar. You can try to use a lot of mirriors to have a 'parallel' beam; but the beam is not partly coherent not even with respect to itself. All individual 'beams' have to be originated from the some 'source'. Otherwise they have got no fixed phase relation relative to one another and thus cannot interfere.

Thanks Dr.Gerhard for explanation and it should help to 'extract' some coherent part from the major scattered light. However I would mean to have a system that can 'convert' nearly all of the incoherent light to creat coherent one. It is like making a laser beam from sunlight with only passive component. Do you have any idea?

If I have (or had) such an idea I would have been a rich man......

Having only passive elements without introducing energy thus bringing in a lot of entropy) you cannot take out low entropy earnings/revenues such as laser light out of sun light.

It is a dream ....

Can we couple a scattered light to a multi-mode optical fiber at one end with the other end tapering down adiabatically to support only single mode?

Gerhard Martens thanks but I partially agree. Most of dreams in long scientific history may came true later and create many high-impact technologies for mankind. And this particular dream maybe coming true very soon.

You might be interested to read some articles on "solar-pumped lasers" like:

https://www.nature.com/articles/srep14758.pdf

https://www.osapublishing.org/DirectPDFAccess/12C4027D-A105-0E72-F25D2670B5A425BF_225792/oe-19-27-26399.pdf?da=1&id=225792&seq=0&mobile=no

https://www.youtube.com/watch?v=JGGa3tycbZ4

http://iopscience.iop.org/article/10.1088/1612-202X/aa688a

Heating applications (even melting metals) using solar concentrators are known from a long time there is nothing new. This is ok for near visible wavelength beams. but as we decrease the wavelength it become more and more difficult to use and manufacture cheap reflecting surfaces.

Let us come back to the original question: converting scattered light into coherent one by passive elements;

The singlemode fiber is just one embodiment of a 'slit system'.

The intensity loss when coupling divergent light into such a fiber is dramatic. I have dealt with multi-mode 50µm down to mono-mode fibres (~ 5µm core diam.). Coupling light into these fibres is a nightmare even when you have got a laser as light source; a very lot of intensity loss. Passive coupling optics or a taper help a bit, but do not really prevent the intensity loss fiasco.

Your above mentioned laser techniques are very interesting but they belong to the active ones. That is another theme for providing coherent radiation.

I believe (but might be wrong) that whenever an optical mineralogist doing liquid-immersion RI work uses polarized, monochromatic light and looks straight down any optical axis of an anisotropic substance: When there is a condition of match between liquid and substance, resonance probably causes low-power lasing from reflections at the liquid-substance interface. Paradoxically, purely transparent fragments disappear under such conditions of matched RIs and might be seen if a polarizing "analyzer" is placed above the sample; sometimes one must rotate the stage to be able to see such fragments "at match". But it could be argued that there is no such lasing and that light under such conditions passes with no reflections at all, due to boundary conditions. I'm also not quite sure why simply polarizing monochromatic light from a given laser or diffraction grating might not meet your needs; which is why I hesitated to chime in at all. :-)

Bragg's Law pertains to light scattered in other directions, together with potential Wavelength phase shifts that I ignore, above.

I hope this helps and also hope to learn from feedback related to this response. I am not a laser expert.