The Datastore still appears because there is an active process or an iso mounted/snapshot.
First select your cluster -> right click -> datastore -> and rescan
Second, go to vCenter select the datastore and check Check which host you didn’t unmount on, also you can go to the VM tab of this datastore and check if you have a VM.
Login to ESXi over ssh and check this outpu
lsof | grep name of datastore
Run this esxcli storage filesystem list find storage UUID and send me this output.
lsof | grep storage uuid
Replace the naa.
Run this command vsish -e ls /storage/scsifw/devices/<“naa. id of your datastore”>/worlds/ |sed ‘s:/::’ |while read i;do ps |grep $i;done
Check the lock file/ process file to unmount the datastore