The MLM-matching for the QCD process

hi,

i tried to generate some multiple b-jets processes as SM background for my current study. let's say i am interested in events with b-jets plut di-photon final states, hence i run:

generate p p > b b~ a a @0
add process p p > b b~ a a j @1

in addition, i also set:

1 = ickkw
10 = xqcut

in the run_card.dat. while i found obvious discontinuities in the differential jet rate plots. may i know what is the proper xqcut choice here for this type of QCD process where the final states do not involve any heavy objects like W/Z/top/higgs ? thank you very much for the help.

best,

ning