A Anyone knows the slow light in EIT？

[wangvivi]

TL;DR Summary

I build a EIT system, but the pulse delay between the reference and probe is only several nanoseconds, I read many papers and in their system, the delay can up to several microseconds, I have tried many ways to improve it, but it doesn't work

I built a EIT system in rubidium gas cell and measured the pulse delay between the reference and probe light. Here, I used a zero-order diffracted light of an AOM as my probe light, but there is just several nanoseconds between these two pulse. I read some papers, like adjusting the Rabi frequency of control light or the temperature of my gas cell to improve the pulse delay, but it doesn't work.
Can anyone who is familiar with electromagnetic induced transparency help me or give me some advice? Thank you so much for your reply.