I did it once the other way round. Attach a NH2 bearing small molecule to the activated ester on the chip. A 1 mM concentration of the small molecule (check solubility!) worked fine in a HBS-N buffer system at 5ul/min with a 10-20 min. injection. The immobilization level was generally in the range of 100-200 RU. Depending on the later experiment u might have to reduce the injection time to reduce the surface capacity. I guess it should work in your set-up as well.