It’s hard to be specific without a look at the design file. You might have chosen off-the-shelf lenses which are not providing acceptable results. But with the current lenses I would try reversing the third lens.

At 0.7 NA it will be very hard to compensate for spherical with just a few lenses and especially with off-the-shelf (OTS) lenses. Trying to compensate spherical and other aberrations, over the field (1-mm) makes it even harder. Furthermore, doing this at 283 nm is even more difficult. As you are probably aware, there are few OTS substrates that will be suitable for that part of the UV. You could use fused silica, CAF2, and sapphire.

You may consider a reflective system like an Offner relay that are commonly used in spectrometers. Dave Shafer has some good slides on those.

Finally, in a shameless act of self promotion, if you are committed to a refractive solution you may consider my paper on designing objective lenses from OTS parts. This could be done with fused silica. Optosigma provides many fused silica singlets and an AR coating in the UV that would be suitable (A51). For the first lens in the system, Edmund Optics has high-NA glass objectives in fused silica with 285-nm AR coats that would be suitable.

Thank you so much! Adding more lenses helped! 

Note: ZAR and ZMX files are not supported. Upload these as ZIP instead. There is a 25 Mb limit on file attachments. Be sure to wait for your attachment to finish uploading before you submit, or you may lose the content of your post.

@David Thank you for the suggestion.