That's probably because of magnetic field instability.
Environmental stability of high resolution spectrometers is crucial for their intrinsic performance and sensitivity. The most common causes of spectrometer instability are variations of room temperature, mechanical vibrations, and fluctuations of external magnetic field. In the urban areas, the motion of massive ferromagnetic objects, such as elevators, and fluctuating DC power lines are major sources of magnetic field noise. The noise originating from DC power supply lines of the city transport system is especially difficult to avoid for NMR facilities located in downtown areas. This is quite popular problem and may result in non-lorentzian lineshapes.
How can I check is it a magnetic field instability?
The simplest way to check it is to run a serie of 1D experiments using sample giving single, narrow line "Lineshape sample" - 1% CHCl3 in deuterated acetone). Field stabilizing systems, such as "lock" should be turned off. Then you can observe variations of Larmor frequency of signal proportional to magnetic field disturbances. Like this:
But when lock is on, it should be OK ...?
No, not always. Lock system is often fast enough to keep signal at right position:
But not so effective in keeping good lineshape:
Why "lock" systems are not fast enough ?
"Lock" systems measure Larmor frequency of some nuclei (most often deuterium). One of basic properties of measurment of period signals is the inverse relation between time of measuement and precision. For example if you want to measure frequency with precision of 0.1 Hz, you have to acquire signal for 10 seconds. That's why measurement of magnetic field with enough precision requires tens-of-seconds long time constants of lock system.
How can I help my "lock" ?
Additional correction, working faster than lock, would help it to deal with such disturbances. We would like to offer you Field Stabilizing Device measuring field disturbances with so-called fluxgate sensor, which is much faster than a lock system. Correction is external and does not require any ingerention in spectrometer hardware. After installation and calibration it works automatically, so it can be used by non-professionals. The effects can be impressive:
Lineshape stability is increased:
t1-noise in multidimensional experiments is reduced even ten-fold
In deuterium experiments this system may be used instead standard lock.
These are pictures from your lab only, so is this solution general?
We have installed this device in a couple of NMR laboratories in Europe and it always helps. In fact, the problem of magnetic field disturbances is quite common in urabn areas, so check if you can make your spectrometer more stable and get better signal-to-noise ratio!
How to order this device?
Just click the link on the right side of this page and email us. Price depends on local conditions, but is very attractive comparing to all standard spectrometer hardware.