Hi. So,

a) Should be the unpaired two sample t-test.

b) Mann–Whitney test should be appropriate.

c) If you expect non-normality in the population and because of smaller sample Mann-Whitney can be better. But under the normality assumption, t-test should be fine.