Thanks a lot @Ratna Chaitanya Raju Bandaru​ , I picked up your code and I did this:

final_df = df.withColumn("CreateDateNew", concat(from_unixtime((F.regexp_extract(col("CreateDate"), "(\d+)", 1)/1000),"yyyy-MM-dd HH:mm:ss"), lit(" "), regexp_replace(regexp_extract("CreateDate", re, 2), "(\\d{2})(\\d{2})" , "$1:$2" )))

and I got the desired output:

Once again, thanks and kind regards 👍