Unit of plank constant in the calculation of tunneling probability

[Flyyoung]

I'm calculating the band-to-band probability using the WKB approximation given by T=exp(-4*lamda*sqrt(2*m*Eg^1.5)/(3*q*hbar*(deltaPhi+Eg))). I am wondering which unit of the plank constant is used here? The one in J-s which is 6.62e-34 or in eV-s which is 4.13e-15? However, the former one results in 0 and the latter gives 1, and the change of lamda (the relevant length scale) or deltaPhi or Eg doesn't really make an impact on the probability. Why is this?

Many thanks!

 

[Flyyoung]

Actually (q*hbar) gives 1.05e-34, however, still it doesn't give any clue why the probability is almost insensitive to the change of deltaPhi/relevant length scale/bandgap :( Anyone can help please?