Thanks for the responses guys!
I've tried finding the general form for the turning points in terms of 'a'. I solved f'(x)=0 and got T.Ps at x=2, x=(1-a)^1/2 -1 and a=-(1-a)^1/2 -1. For the function to be one to one, the T.P at 2 must be the last turning point on the right side of the graph, since the domain is [2, infinity), and if there is another T.P past x=2 then it will not pass the horizontal line test. So I solved for (1-a)^1/2 -1 to be less than or equal to 2, since this T.P x co-ordinate must be to the left of the x=2 T.P, and i got -8<(or equal to) a < 1 (or equal to). However the answer is a is from -8 (inclusive) to infinity.
I'm thinking that a>1 is included because the x co-ordinates of the T.Ps are undefined for a>1(in second line from solving f'(x)=0) and hence the only T.P is at x=2, and therefore has inverse for a>1. Therefore combine these values (a>1) with -8<a<1 to get, -8<a<infinity. I'm not sure if this thinking is correct or where I've gone wrong - it's still quite confusing. Do you guys have a better way of doing this (assuming that what I said makes any sense

)?