Is it possible to convert sunlight to microwaves?

Glass does NOT change the frequency of light when it refracts. Frequency remains the same throughout the refraction process. It is the WAVELENGTH that is "sacrificial" to the refraction properties of the material...not the frequency. That is why the path of light geometrically compensates for it...because it is a geometric property of the light that changes...not a timing property.

Frequency is what defines color/spectral band. NOT wavelength. Red in, red through, red out. Violet in, violet through, violet out.

You CAN change frequency by putting the Doppler effect (and its Einstein-gravitational counterpart) to work. Have a source of visible light move away from the receiver, and this can convert it to a microwave.

Of course, this is greatly impractical, because it will require greater speeds than have ever been made artificially. I.e. significant fractions of the speed of light.

Another more practical manner that might work would be fluorescence. Look at a fluorescent light tube. On the interior of the glass, there is a white powdery coating. That is particles called phosphors (nothing to do with phosphorous, except historically). What these particles do, is "step-down" the frequency, by absorbing a photon, and then releasing it in several stages. That is how fluorescent lights work...the mercury vapor electrodischarge's native emission is primarily UV radiation, and then the phosphors step it down to visible.

Of course, it is greatly difficult to do this involving microwaves, because there aren't any characteristic atomic frequencies which are anywhere near microwave frequencies.

So...I guess you are stuck with plugging your microwave in to the output of your Sunnyboy inverter, which is plugged in to your sun-exposed PV array of modules.

Unless...you are comfortable with the natural emission of microwaves that the sun already emits.