INVERSE DETECTION ON THE AM-400/500

Ad Bax and S. Subramanian published an excellent paper (J. Magn. Reson., 565-569,67 [1986]) describing a new pulse sequence for proton-carbon correlation by observing the protonspectrum. Their example shows a complete 1-bond 2D correlation of 12 mg of a compound (MW=1044!) obtained in 29 minutes! The experiment was done using INVERSE DETECTION. This technique will increasingly be used for new experiments.

On the AM series spectrometers, the REVERSE MODE must be set up:

A switch has been added on the console to select NORMAL or REVERSE Mode. REVERSE Mode activates the electronics in the console (a switch on the 451 receiver unit was formerly used) and relays in a home-built unit bolted to the side of the Bruker Pre-amp housing. These relays interchange the pulse and decoupler amplifier outputs, and switch the X-amplifier directly to the X-port on the probe. Currently, an extra amplifier still needs to be connected if GARP decoupling of the X-nucleus is desired.

The REVERSE switch interchanges some of the electronics units; O1, SF01 and O2, SF02 are now reversed. The decoupler is used to generate the proton pulses. O2 must be set to the values normally used for O1. In addition, the pre-amplifier range now must be set to H 1 instead of 1 1. The Inverse Detection experiments require that the X-nucleus (for instance carbon) be pulsed and the resulting proton spectrum be acquired. SF02 needs to be set to e.g. 500.13 and O2 to the middle of the proton spectrum (e.g. 8000). SF0 now needs to be set to the X-frequency and O1 needs to be set to the middle of the F1 region to be pulsed (e.g. the center of a carbon spectrum).

Note that the usual ZG command cannot be used. Use one of the AU programs, e.g. INVH1.AU, to pulse with the decoupler. GS will also not be working, use INVH1.GS for setting the receiver gain etc. Note that P1 is used to pulse at power S1!


Implementation

We have modified the AM-400/500 for Inverse Detection and (hardware wise) related experiments. The changes provide fast level switching of the observe rf power. An external power amplifier can be used for pulsing and decoupling (using Reverse Mode) at frequencies up to 200 MHz. The pulse power is lower than normal. This results in a longer 90 deg pulse width of between a factor 3 to 3.5. For decoupling, up to 5 watts of power can be used. With a GARP decoupling sequence, the power should be sufficient for carbon decoupling with proton detection.

A separate low power output is available for homonuclear experiments on both instruments. The output can be adjusted from 0 to 100 dB. 0 dB should correspond to about 6L of the normal decoupler output. This output can be used in place of the decoupler for phase coherent solvent suppression or long, low power pulses. Compared to the TLO output, the level is approximately 3 times lower. With additional gear, this feature could be extended to a selective excitation system.

The additional output is enabled with a " :T:C7 " suffix in a pulse program. " :T " enables the transmitter, " :C7 " activates the low power channel. TLO should be set for X-nuclei as a precaution (although this is not strictly necessary, there is some built-in protection. Note that the switching speed is about 2 ´ sec. A dummy pulse (e.g. P0:C) of 2 to 5 ´ sec MUST be used after the pulse with a " :C7 " suffix only. If this is not done, a 2 ´ sec hard pulse occurs! If the length of the FTLO duration is critical, it also must be preceded by a dummy pulse.

An additional problem occurs with the :T :C7 statement. A decoupler pulse with a phase statement will not work unless preceded by a dummy delay with a phase statement.

I suggest using P7 for the low power pulses, P0 for the dummy pulse.

Examples are shown below:

;EXAMPLE.INV  Pulse X-Nucleus with low power (P7) followed
;by High Power (P5) observe transmitter pulses, delay,
;pulse protons with decoupling transmitter and acquire proton signal.
	
1   TL        ; switch to TLO mode
    ZE        ; Zero memory
2   D1 DO S3  ; decoupler off, select power level S3
3   P0 :C7    ; set to about 5 ´ sec to allow switching to low power mode
4   P7 :T:C7  ; turn transmitter on in low power mode
5   P0 :C7    ; Transmitter off, switch back to high power
6   P5        ; 90 deg transmitter pulse X
7   D3        ; A delay
8   P1 :D     ; A proton 90 deg pulse from the decoupler
9   GO=2      ; Acquire proton signal in inverse mode 
10  D1 DO S1  ; Quit with decoupler gated off and at low power.
EXIT

 ;INVXHD.AU
;2D H-1/X correlation via heteronuclear zero and double 
;quantum coherence using BIRD sequence in inverse mode, phase sensitive using TPPI
;decoupling during acquisition using GARP1 sequence. This requires BSV7/8 with
;fast TLO (RCP7) and additional amplifier (Kalmus). See also INVH1.AU
;A.Bax and S. Subramanian, J.Mag.Res. 67, 565-569 (1986)
;Modified for standard parameter sets, 3/31/89, R. Nunlist. 

	IIy
1	ZE
2	D1 S3 DO      ;relaxation delay
	P1:D PH1      ;90 deg proton Decoupler pulse
	D2  	      ;1/(2J)XH
	(P1*2 PH1):D (P5*2 PH7) ;180 deg proton and X pulse
	D2
	P1:D PH9y
	D7  	      ;recovery delay
	P5 PH11	      ;90 deg X pulse
3	P1:D PH1  	  ;90 deg proton pulse
4	D2            ;1/(2J) X, H  refocussing dela
5	P5 PH3        ;90 deg X pulse
	D0            ;Evolution
6	P1*2:D PH2    ;180 deg proton pulse
	D0	;Evolution
7	P5 PH4        ;90 deg X pulse
	D2	          ;1/(2J)XH refocussing delay
	D5	          ;DE/2
8	D5:D PH8
9	D8 ADC	      ;D8 = 2 usec
	P0:C7         ;To prevent pulses when loop recycles
10	P8*0.339:C7:T:C3 PH10;GARP1 decoupling
	-----------
	long GARP sequence follows here
	-----------
	P8*0.729:C7:T:C3 PH10
	P8*0.593:C7:T:C3 PH12
	L1 TO 11 TIMES 2
	P0:C7	;(IMPORTANT)
12 L2 TO 10 TIMES UPR
13 RCYC=2 PH5
	WR #1
	IF #1
	IP3
	IN=1
	EXIT

PH1=0
PH7=0
PH9=2
PH11=0 0 0 0 2 2 2 2
PH2=0
PH3=0 2
PH4=0 0 2 2
PH5= R0 R2 R2 R0
PH8=0
PH10=0
PH12=2

;Set spectrometer for reverse mode, Connect external amp to BSV-7(8), output
;via low pass filter to probe
;Add 400/500 MHz High Pass Filter, if necessary,.BETWEEN probe (1H channel)
;AND pre-amp! (VERY IMPORTANT!).
;SF01, O1 for x-Nuke, SF02, O2 for protons
;D1 = (0.85 * (shortest proton T1)) - AQ.
;S3 = 0H
;AM400: Set FTLO attenuation to about 15 dB (gray box on top of console)
;AM500: Set FTLO attenuation to 26dB.
;P1 = 90 deg Proton decoupler pulse at S3
;D2 = 1/(2J)XH
;P5 = 90 deg X pulse with external amplifier
;D7 = optimize to null protons not bound to C-13 (T1 dependent)
;(about 0.46*T1 )
;D5 = DE/2 (in usec)
;D8 = 2 usec
;P0 = 5 usec. Important!!
;P8 = 90 deg pulse for X decoupling w. ext. amp., set for 60 usec.
;L2 => L2 * 31.75 * 4 * P8 = AQ. 
;L2 = AQ/(127*P8).
;DS = 2 or 4. NS = 8 * n
;IN = 1/4SW(X) = (1/2) DW(X), SW1 = SW(X) /2! 
;ND0 = 4,  MC2 = W (TPPI)
;Requires phase correction in F2 and possibly in F1.
; PASC  GARPLOOP calculates L2. 

Parameter Examples for AM-500/400:

Use the same basic approach as described in the Routine 2D Manual. First, obtain a proton spectrum to establish the proper sweep width etc. If a carbon spectrum is available, use it to retrieve SW, DW, SF0, and O1 (to set O1 !!).

One-Bond Correlation, 13C GARP Decoupled, INVXHD.AU

The following parameters are set up in INVCH.500/400:
The F1 window is set for about 180 ppm to -19 ppm
The F2 window is set for about 10.5 to -0.5 ppm.

Be sure all 2D parameters are set as follows:

AM-500

                 F2              F1

   TD=            2048W            512W
   SI=            2048W           1024W
   SF=           500.130          125.735426
   SW=          5555 	      12500 (200 ppm)
   HZ/PT=     © 1 to 4               © 20 to 50
   WDW1=            Q                Q
   LB=              i.               i.
   GB=              i.               i.
   SSB=             2                2
   HZ/PT2/HZ/PT1=   i.

   SIZE.SER IN K=                 REDF= N
   SIZE.SMX IN K=   i.            REV=  Y
   MC2=W                 MAXM= >400
   NE=512                 ND0=  4
Set/check other Acquisition Parameters: PW = RD = 0, P0 = 5 ´s.

SF0 = 125.730 , SF02 = 500.13, O2 = {O1 for normal proton, e.g. 8000},

O1 = O11 = {O1 for Carbon, e.g. 39287}, IN = 1/2 DW(x) = 10

AM-400

                 F2              F1

   TD=            2048W            512W
   SI=            2048W           1024W
   SF=            400.130         100.615
   SW=           4424.78        10000 (200 ppm)
   HZ/PT=       © 1 to 4        © 20 to 50
   WDW1=             Q              Q
   LB=               i.             i.
   GB=               i.             i.
   SSB=              2              2
   HZ/PT2/HZ/PT1=    i

   SIZE.SER IN K=        REDF= N
   SIZE.SMX IN K= i.     REV=  Y
MC2=W                 MAXM= >400
NE=512                ND0= 4
Set/check other Acquisition Parameters: PW = RD = 0, P0 = 5 ´s

SF0 = 100.62, SF02 = 400.13, O2 = {O1 for normal proton, e.g. 6300},
O1 = O11 = {O1 for Carbon, e.g. 229000}, IN = 1/2 DW(x) = 12.5

The SI2(1), TD2(1) and NE values may need to be adjusted to obtain digital resolution of 1 to 5 Hz for protons and about 20 to 50 Hz for carbon, depending on the sweep width SW and SW1! This is done by setting new SI2, TD2. Check and correct SI2(1), TD2(1)1. Type ST2D to update the relevant 2D parameters.

           F2        F1

   TD      2048W      TD1    512W
   SI      2048W      SI1   1024W
           SSB1=     SSB2= 2  90 deg shifted sine-bell
           WDW2=WDW1=  Q      Squared Sine Bell window for both dimensions.

   MC2 = W  Phase sensitive, TPPI
   REV = Y  reversed diagonal
   REDF= N
   ND0 = 4	   For IN calculation
   I2D =      Not useful
   NE  = 512  number of increments (serial files)
One parameter NOT read by 'RJ' or 'RJ2D' is the loop count L2! If SW2 and TD2 are the same as in the parameter file, L2 should be 38 on the AM-500. If either SW2 or TD2 are changed, L2 needs to be re calculated: L2 > or = AQ / (127*P8). An error message might appear if L2 is too short. If SW2 is changed, D5 must be changed to D5 = DE/2. PASC GARPLOOP sets D5 and activates a calculator to determine L2. L2 must then be typed!

The D1 and D7 delay are T1 dependent. Set

D1 = (0.84 * T1) - AQ,

D7 = 0.46 * T1 where T1 is the shortest Proton T1.

Values of D1 = 0.5 s and D7 = 0.4 s assume that the shortest relaxation time is about 1 second which may or may not be the case with your sample.

Values for the pulses are listed in the "Measurements" section. Please note that the exact numbers may change with time!


Long-Range Correlation, 13C Coupled, INV4LPLR.AU

Parameter files: INVCHLR.400/500

Be sure all 2D parameters are set as follows (AM-400 Values):

                  F2              F1

   TD=            2048W            512W
   SI=            2048W           1024W
   SF=            400.130         100.600
   SW=            4237            12500
   HZ/PT=       © 1 to 4       © 20 to 50

WDW1= S Q LB= i. i. GB= i. i. SSB= 2 2 HZ/PT2/HZ/PT1= i. SIZE.SER IN K= REDF= N SIZE.SMX IN K= i. REV= Y MC2=M MAXM= >400 NE=512 ND0= 2

Set/check other Acquisition Parameters: PW = RD = 0, P0 = 5 us AM-400:

SF0 = 100.602, SF02 = 400.13, O2 = {O1 for normal proton, e.g. 6300},
O1 = O11 = {O1 for normal Carbon, e.g. 22990}, IN = DW(X) =

AM-500:

SF0 = 125. , SF02 = 500.13, O2 = {O1 for normal proton, e.g. 8000},
O1 = O11 = {O1 for normal Carbon, e.g. 43000}, IN = DW(X) = 17


Processing in uxnmr:

Uxnmr does not know that the spectra were aqcuired in reverse mode. This will lead to an impossible SF for the F1 13C dimension.

Uxnmr on X32

To fix this, type '2s bf1', enter 500.13, type '2s o1', enter 8000. (400.13 and 6300 an AM-400) or whatever values you used. The macro "aminv" prompts for these paramters. Type (xmac) 'aminv'. to run the macro.

Xspec (Uxnmr/p) on Sun or SGI

Use 'xmac fixsf' correct the parameters.

MEASUREMENTS

All measurements use the decoupler for the proton pulses while observing protons. The carbon pulses are generated by the BSV-7/8 or the external amplifier. Carbon pulses are measured by observing the response of the carbon satellites as a function of carbon pulse width.

INVP90LO.AU is used to determine the FTLO 90 deg X-nucleus pulse, P8. INVP90.AU is the equivalent for the external amplifier (if connected) or the regular BSV-7/8, using P5.
The Decoupler proton pulse P1 can be determined with INVH1.AU.

AM-500: (7/25/91)

	Inverse probe

90 deg Reverse Mode Decoupler pulse @0H = 8.2 usec. @1H = 17 usec. @3H = 21 usec (May 95) 90 deg Carbon pulse BSV-8 = 7.5 usec (Dec (94) 90 deg Carbon, ext amp: THI = 19 usec FTLO @04 dB = 19 usec @24 dB = 44 usec @26 dB = 59 usec @27 dB = 65 usec

Use about 26 dB for GARP decoupling.

	5 mm Broad Band Probe:

90 deg  Decoupler pulse                  @0H =   usec
90 deg  Decoupler pulse                  @0H =   usec (reverse mode)
                                         @1H =   usec
                                         @2H =   usec

90 deg  Carbon pulse, BSV-8     =

90 deg  Carbon, ext amp:        THI          = 19 usec
                               FTLO   @10 dB = ?? usec
                                      @11 dB = ?? usec
                                      @12 dB = ?? usec
For GARP decoupling, use ?? dB.
AM-400:

	QNP Probe: (5/28/91)

90 deg  Reverse  mode Decoupler pulse @0H = 10.3 usec
                                      @1H =      usec
                                      @2H =      usec
                                      @6H = 37.5 usec.

90 deg  Carbon pulse  BSV-7               =  5.2  usec
                (10/3/88)

90 deg  Carbon, ext amp:  THI         © 10-12    usec
                      FTLO            @12 dB = 48 usec
				      @15 dB = 76 usec
                                      @18 dB = 96 usec
Use about 15 dB For GARP decoupling.

Inverse Broad Band Probe

For the BB probe the 90 deg pulse is 22 usec for carbon (BSV-7), estimated 46 usec with Kalmus amplifier.
The 90deg decoupler pulse @ 1H is about 9 usec.


WIRING CHANGES

For O2 = O1:

PTS Synthesizer: Change cable from BU 9 to splitter

For Pre-Saturation using O1 (e.g. LSAT.AU)

On Pre-amp Housing: Replace Decoupler cable with FTLO cable

For REVERSE MODE (Proton pulses from Decoupler) On Console: Flip switch to REVERSE (up) position


INVERSE DETECTION With/Without X-Nucleus Decoupling:

Prepare for REVERSE MODE. In addition:
Only necessary if X-nucleus decoupling is used:
BSV-7/8: Remove black cable labelled "BSV OUT" at the output, replace it with the cable labelled "EXTRA AMP". Connect the "BSV OUT" cable to the output filter on Kalmus amplifier.

Pre-amp Housing:

AM-500: A 500/400 Band Pass (or 500/400 High Pass) Filter MUST be installed between the proton cable and to "PROTON" input! (This might eventually become unnecessary.)

A 2H Pass Filter can be installed between cable and input to reduce interference with Lock channel.

PULSE WIDTH CALIBRATION

Calibrating the Proton pulses using the Decoupler:

After setting up REVERSE mode, the Measurements section.

Calibrating The 90deg X-Pulse:

To set up the experiment for the first time, one needs to calibrate the 90 deg pulse for the X-nucleus. We can use a pulse program to observe the effect of the X pulse on the X satellites in the proton spectrum.

The pulse program INVP90.AU is shown schematically below:

     H:  90 - 1/2J -    - acquire<
     X:               90
To calibrate the 90 deg X-pulse for 13C, we can use e.g. CHCl3. The 1/2J delay D6 is set to about 2.3 ms. P1 is the 90 deg observe pulse at the power selected by S3, P5 is the X-pulse. By varying P5, we can note the change of intensity of the anti-phase satellite doublet in the proton spectrum. For P5 = 0, the satellites are of normal intensity. As P5 is increased, the intensity decreases. At P5 = 90 deg , the doublet is nulled. For P5 > 90 deg , the doublet inverts and the intensity increases again.

The proton parameters are set up as usual. For pulsing carbon, it is best to set SF0 and O1 (!) to be near 77 ppm (the carbon shift of CHCl3, adjust for other samples) to reduce errors due to off-resonance effects. This is particularly important if low power pulses need to be calibrated!


CHECKLIST FOR NON-STANDARD EXPERIMENTS

If you are setting up experiments which require the hardware to be re-configured, check:

BEFORE THE EXPERIMENT:

AFTER THE EXPERIMENT

:
Click Here to return to the NMR Facility Home Page.
Questions, comments to: Rudi Nunlist rnunlist@bloch.cchem.berkeley.edu

Last Update: 5/9/95 RN.