The quickest way to import fastq data into Galaxy from NCBI with the pairs split is with the tool
Faster Download and Extract Reads in FASTQ format from NCBI SRA. The output creates collections for you and can be used as input to
Or you can go back to the original interleaved reads and split them up – then run
Trimmomatic on them. That particular tool might not fail on interleaved inputs but is also isn’t designed to handle them.
I added some tags to your post to help with how to split up the interleaved reads. And here is an FAQ and a tutorial that explain directly what interleaved fastq data is and how to manipulate it. Some tools accept that format, but most do not. But if you are trying to build up a workflow or plan to do a similar analysis again, using the tool that splits the read data at retrieval is simpler (and “faster”).