Dear all,
Actually, looking at the 103Rh question reminded me of something I saw a long time ago. For estimating the PW of a rarely looked at nucleus, one can calculate it rather well from the known PW of another nucleus provided the signal path is the same (oscillator, amplifier, coil etc.). The formula is:
PW(x) = PW(y) * (gamma(y)/gamma(x))exp(0.75)
The formula works quite well so I am convinced of its validity but cannot find the original source. Can anyone confirm this as the correct formula, or otherwise indicate what it is ?
Regards,
Karel Klika
________________________________
From: jfxiang_at_iccas.ac.cn [jfxiang_at_iccas.ac.cn]
Sent: 20 May 2015 06:08
To: ammrl_at_ammrl.org
Subject: AMMRL: summary of 103Rh NMR
Dear friends
Thank you very much that I have obtained the most important message to set up my 103Rh experiment in my spectrometer, Clemens, Craig, Dr. Virginia W. Miner, Bob Berno, David VanderVelde, Akien, Geoffrey and Sebastian Kemper have offered me the important comments: The following are the key points:
(1) about the 1D experiment of 103Rh The paper of Laurence Carlton is very useful.
According
http://chem.ch.huji.ac.il/nmr/techniques/1d/row5/rh.html a T1 of 1 s is to be expected for Rh(acac)3 in CDCl3, so a 2 s recycle delay should be more than enough really. But in fact, in my experiment, I never succeed in obtain 103Rh spectrum with 2s delay by zg30. I can obtain some signal with 60 s delay. I didn't know why.
The papers published by Professor Laurence Carlton at Wits University; he has done extensive work on 103Rh - you will find information on chemical shifts and coupling constants.
(2) HMQC or HMBC is a goog way to find the chemical shifts of 103Rh. Nearly all of responsers agreed with this idea. The following are some answers
The best way to see 103Rh is by doing so indirectly, via coupling to a high sensitivity nucleus using an HMQC/HMBC experiment. Which nucleus is used for observation will depend on the structure of the compound.
You can run also a HMQC experiment to find the signal. For this to work you need a coupling between either the CH3 groups or the CH of the acac. It would be a 4J (Rh,H) and is probably quite small. Do you see a splitting on the 1H signals? 103 Rh is 100% abundant so you will see a nice doublet if there is a coupling. The HMQC is the best experiment as it works well even if the X nucleus pulse is not calibrated.
Seeing longer range 1H-X correlations is possible in HMQC provided you have set the delays accordingly. Newer Bruker pulse programs use the parameter CNST2 to indicate the expected 1J coupling constant, and the delays are calculated from that. However, if you instead insert a long range coupling constant, then new delays would be calculated, and then you would see the longer range correlations.
Is there some 1H signal which you can see is split into doublets because of Rh coupling (100% natural abundance) and see the actual size of the J?
I would set up an HMBC experiment starting from the HMBC15N parameter set (assumes you don't need to suppress a large 1 bond coupling as is commonly the case for carbon). Then change the second nucleus from 15N to 103Rh.
I like HMBC for this because it uses just a single 90 degree pulse on the X nucleus. The sequence will also work to some extent even if you start with a guess for the length of the 90 that isn't accurate at first. If you know exactly where the Rh peak should be, great, but it can be very efficient to find an unknown X nucleus peak by running only the first block of the HMBC. Step through various values of o2p until you find the one that gives the biggest signal, containing a 1H spectrum just of those peaks coupled to Rh. When you have a good o2p, then you can run the first block and vary the length of p3 to find the exact length of the Rh 90. I have used this approach successfully with 195Pt, and 89Y. I don't have a low gamma probe to try 103Rh, wish I did!
You can adjust your HMQC to another J coupling. If you change CNST2 to the 2J or 3J coupling you will have nearly the same as a HMBC. The only matter is to know this coupling constant, therefore you have to search the litrature. If the coupling constant is very small (eg 1 Hz) its perhaps better to chose a value higher than this (eg 5 Hz) otherwise your spins relax during the time the anti phase magnetisation should develop.
Following the precedure above, I can carried out HMBC and HMQC experiments easily. But at the beginning, Clemens offer me a good start point. This really helps especially for me, a new beginner.
You can use the 1D experiment hmqcndrd1d to search for the resonance. When O2P is too far away you should get no signal. As the Rhodium resonance gets into the excitation region you will start to see signal. When you are closest you get the biggest signal. Once you approximately know where it is you can reduce the step size.
You can easily run this optimization with popt or paropt and change O2P .
Once the signal region is found you can set up an actual HMQC 2D experiment. First with a larger sweep width and then finally with 90 degree pulses on both channels and a narrow sweep in F1.
(3) about gradient setup in HMBC or HSQC. I am using Bruker Avance 600 spectrometer with TS2.1pl6. Therefore, the following advice works very well.
There is a nice au program from Bruker called 'gradratio'. If you execute this on your changed HMQC, it will give you the right gradient ratios.
Thanks again for all friends.
Best regards
Junfeng
NMR Lab, ICCAS China
Received on Wed May 20 2015 - 08:40:53 MST