|
下面是操作方法：
0 ~) t+ H; \* S! {+ C' C5 ]3 t
q* V0 w3 U( v) e* k; eNMR Facility + M' c# @, j# `2 ~& X
3 i/ A& a& O: { i9 R
Equipment Protocols - NMR . G" C8 a& q$ y" g U8 D
' a5 O4 {) R8 W% j/ `
INOVA-399 " N& X0 E, W& a
& D( q# d7 S- o: cCollect a 1D 1H spectrum
; ?* l) o# q! l, ~Log in
Click the Vnmr icon
Lock, shim and collect a 1D 1H spectrum
type gain? and write down the gain value - ^) i$ x2 u5 U W$ \% s6 |
Set temperature and spin
$ L; b. G0 Z- w# ngo to start tab and spin/temp
move slider to desired temperature
(must be greater than room temperature)
move spin slider to 0 and press spin off
(make sure spin goes off)
* @; r" D$ x+ z* [
Get pulse sequence and optimize parameters
9 p; ?* t% i8 U u
type Doneshot (or any of the other DOSY experiments)
go to start tab and give file name
check solvent
change gain to the value found above by typing gain=#
type gzlvl1=5000,15000,23000
type ga
type nt=8
set delay to 10s
when data collection is finished type dssh to see all data
type ds(1)
type f full
******WARNING Poor signal to noise can result in incorrect diffusion values (high)*******
tupe aph
type dssh
make sure that the last data set has about 5% -15% of the S/N as the first data set
if there is not sufficient delay increas gt1 or del
(type del? to find current value then increase a little- type ga to try again) : R: W+ w* a7 m0 d$ P R
Collect dosy data
9 x* s8 Z4 q! Z6 b; r" E; Mtype setup_dosy
answer: 30 >5000>23000
make sure spinner is off (press acqi >press lock> press spin off)
type ga
wait for data collection to finish
SAVE THE DATA by pressing the acquire tab, future actions and save fid now
# B$ m5 C% }& l4 ]* {Data processing
4 b5 \' H1 l1 \* H' kPress Main menu >Data
and find and load the data
type fn=100000
type wft
type ds(1)
type f full
type aph0 (if spectrum needs phase correction)
integrate spectrum
type fbc
type dosy or press calculate entire dosy
note down the values for each resonance. The diffusion coefficient reported is the center of the peak
and the error bar is half the width of the widest peak- up to 20% error is acceptable
7 W3 Z) A3 V$ ]+ L; z4 N5 v
# M2 G* V5 y* }# o! ~+ D8 X- XTrouble shooting and hints
& ~( ^. {' ~ l6 K1 x# s
# ^0 F6 S7 A! a5 O7 I' ^$ U( zCommon data collection problems: # ?, j, n6 x* W6 C: H0 r
1) do not get decay down to 15%:
adjust diffusion delay or gradient diffusion length (***careful adjusting the length- too long of a length can fry the probe) adjust these two in very increments (increase gt1 by .001 and del by .01) if you do not get all the way down to 15% you will not have a well defined curve and your signal will be broad
r, i; P- r* J9 G2) data does not decay in a linear fashion:
maybe too much gt- try lowering gt and raising delay (so you still get 15%) but be careful because raising the delay introduces additional error
9 U8 e( {9 Y& A5 Y$ @6 |- x" a5 r: b% A* d6 P8 C
- signals have weird phase problems
' F# `; J* w( M7 [2 _- there may be a temperature gradient across the sample- try turning heater off
. ^/ m* N! G" @: G y- you may have a system in chemical exchange- this make DOSY very tricky and usual fails
: T+ o7 P, j0 }2 L1 e- gt1 and del are wrong- try lowering del and increasing gt1
R% j2 n7 Q4 b7 p! ]
4) too wide a spread of diffusion rates from the same compound:
increase d1 value (calculated rate depends on T1) but keep the D1 delay as short as possible- too long of a delay the greater the band of diffusion rates for different protons if you need a tighter band you need to try greater gradient strength spread or try using inept (very insnesitive)
note: a compound moving in and out of a gel will have a wider band for different protons on the same molecule could be .4 m2/s difference in diffusion rates- not surprising- should be smaller difference if the same compound were free in solution
5) final signal too broad (too big of an error):
sample volume may be too much- the more sample volume the less EVEN the gradient along the Z axis will be
Common data processing problems:
$ v% K6 g0 l4 i2 n+ l+ ]1) Program bombs:
2 T, c+ D+ X. T& I5 l6 z) l1 o
7 t) l) R4 s0 O7 b, ~- check line broadening- it cannot be zero- should be 3-5
. c5 `5 V3 k8 D# W- Z4 m) x- program will bomb if there is insufficient signal to noise
+ W4 S3 e4 X, H, n2 @& A
- peaks do not phase well and there is significant signal below the baseline-see "Spectrum-does not phase" (above)
! I H2 G6 }* t4 r, X
2) Need to run fiddle
place cursor on reference peak ( a nice clean singlet)and type rl
press Ref and give it the correct reference value for that peak
press Th and move threashold close to baseline
press full
place cursors on either side of the reference peak
type lb=-4 gf=.05
make sure full spectrum is on the screen
type fiddle('writefid','filename') and wait till it corrects every spectrum
'writefid' is types exactly and 'filename' is a file you are going to write the data to
integrate each individual peak you want to include in the DOSY analysis
make sure full spectrum is on the screen
make sure that threashold line is on screen and cursors suround reference peak
type fbc
type dosy or press calculate entire dosy
2 k$ ]; Z8 s- Z, Q& r. i
Uncommon problems: / t' ]/ j0 R% n( O/ h
1) full range of rates is not covered:
you many need to recalibrate gcal for the probe
2) one of the spectra does not phase well:
remove that spectrum
3) signal do not phase well:
check to see that the solvent is correct- it must know the correct solvent since the 90s for each solvent are different in the probe file and the 90 being off will effect whether you can phase the spectra.
4) resulting 2D spectrum is too crowded:
consider running a DosyInept instead- diffusion based on carbon signals instead of proton
5) full range of rates is not covered:
you many need to recalibrate gcal for the probe ! g' T s0 T7 N. L3 ^' X U
| 
发表于 2012-5-8 08:29:30
发表于 2012-5-10 15:26:57
