Are you trying to isolation carbapenem-resistant strains? Be aware that imipename is not a natural product and many bacteria are susceptiable to it. You may need to use extremely low concentration in your screen system as well.
Isolation and Purification of specifcalyliCyanobacteria: Some General Principles : https://link.springer.com/content/pdf/10.1007%2F978-3-662-13187-9_8.pdf
@Adel: In general, I do agree with Christopher's recommendation to use protocols of R. Rippka and others. However, their protocols were mainly addressed to the isolation of so called unialgal cultures of cyanobacteria. My multiyear experience showa that there is nothing better than manual separation perfectly described by Rippka et al.
However, the obtaining of so called axenic cultures defenitely requires the application of antibiotics. Unfortunately, "contemporary" cyanobacteria (CB) and their HETEROTROPHIC satellites and parasites have got more resistant to different antibiotics than they were 30-35 years ago. It is not simple now to choose the most appropriate antibiotics right away because a researcher must test the sensitivity of both a CB and its "partners" to a big number of antibiotics. To speed up the selection of appropriate antibiotics, I, first of all, identified the nature of heterotrophic contaminants, e.g. gram-positive or negative, phylogenetic position, e.g. Bacilli, etc. If they were, for instance, a Bacillus, I tested their susceptibility to antibiotics with high efficiency to gram-positive bacteria. After that I tested the susceptibility of a CB culture of my interest to selected antibiotics. Eventually, I verified doses of antibiotics which were able to kill contaminants at minimal inhibition of a valuable culture of CB. Only after that I was able to proceed to Rippka's et al protocol. My approach helped me to avoid the catching of a black cat in a dark room.
Let me know please if something needs additional explanation. IB