You can stack / pd.to_datetime / unstack
pd.to_datetime(dte.stack()).unstack()
explanation
pd.to_datetime works on a string, list, or pd.Series. dte is a pd.DataFrame and is why you are having issues. dte.stack() produces a a pd.Series where all rows are stacked on top of each other. However, in this stacked form, because it is a pd.Series, I can get a vectorized pd.to_datetime to work on it. the subsequent unstack simply reverses the initial stack to get the original form of dte