spark UDF issue while add ing a new column of type array by transforming one of the existing columns

Question

transformed_df.show(10) 

 +-----+--------------------+--------------------+
|   id|                text|      finished_lemma|
+-----+-------