Exploratory Factor Analysis and Confirmatory Factor Analysis


I'm developing an instrument with 27 items and I need to conduct Exploratory Factor Analysis and later, Confirmatory Factor Analysis.
A sample of 26 participants (data collection interrupted by Covid-19...) answered the 27 items.
A sample of 342 participants answered the 27 items and other questionnaires that can be used to assess discriminant and convergent validity of this new instrument.
So, considering this:
How can I split participants for EFA and CFA (e.g, equal number for each analysis?, the 26 should mandatorily be part of one or other analysis?)
Should I join both samples, and then randomly assign participants for each analysis?

Thank you very much.