Dear all,

I'm trying to simulate stmas experiments with simpson 1.1.2 (linux) but
I run into synchronisation problems.
I tracked the problem down to the use of 2nd order quadrupalor interaction.

When using 1st order, I can simulate stmas and reproduce roughly
experimental efficiency Vs pulse length.

To understand better I simulated a one pulse experiment (23Na with
Cq=2MHz, MAS=10kHz, pulse 2us, 40kHz, 1H larmor frequency 400MHz). I
record satellite coherences only (matrix set detect elements {{1 2} { 3
4 } }) and try to track the position of the first rotational echo one
rotor period after the pulse.
Using 1st order only I can observe the echo centered at 99us after the
pulse that is exactely one rotor period after the center of the pulse.
Using 2st order the echo is centered at 101us after the pulse meaning it
is shifted by the pulse length.

This is definitely an error in 2nd order hamiltonian handling. I raised
1H larmor frequency to 400GHz to make 2nd order interaction negligeable
but I still observe the center shifted by 2us.


Julien TREBOSC

testing simpson input file :

spinsys {
channels 23Na
nuclei 23Na
shift 1 0 0 0 0 0 0
quadrupole 1 2 2e6 0.0 0 0 0
}

par {
method direct
np 100
crystal_file zcw986
start_operator I1z
proton_frequency 400e9
detect_operator I1p
variable dw 0.2
spin_rate 10000
sw 1e6/dw
verbose 1001
use_cluster 0
}

proc pulseq {} {
global par
maxdt 0.1
set dw $par(dw)

matrix set detect elements {{1 2} { 3 4 } }

reset
pulse 2.0 20000 0
delay 90
for {set i 0} {$i<$par(np)} {incr i} {
acq 90
delay $dw
}
}

proc main {} {
global par

foreach par(Order) {1 2 } {
set f [fsimpson {{quadrupole_1_order $par(Order)}}]
fsave $f $par(name)$par(Order).fid
funload $f
}
}