One of the major problems when measuring diffusion coefficients by NMR is the presence of convection artefacts. When NMR diffusion experiments are performed at temperature different from ambient temperature, temperature gradients due to probe design can cause thermal convection and therefore significantly affect the signal amplitude. Several works have studied this aspects and approaches to Remove, compensate and measuring such convection effects have been reported:

• 96JMRA50-118
• 97JMR372-125
• 98JMR13-132
• 99JMR334-139
• 00JMR323-142
• 00JMR125-145
• ( 98JMR126-131 )
• Convection compensation in electrophoretic NMR ( 01JMR126-150 )
• Measurement of the convection velocity in the presence of diffusion ( 01JMR35-150 )
• Flow convection profiles from MRI ( 00JMR125-145 )
• Sample rotation using a multiple integer of the spinning rate compared to the diffusion time ( 01JMR48-153 )

Other very important experimental aspects to be mentioned:

• Non-linear Gradients ( 96JMRA285-119 , 97JMR343-124 , 99JMR108-137 , 01JMR343-148 )
• Background gradients due to the magnetic susceptibility differences at sample interfaces ( 00JMR14-146 , 01JMR49-150 )
• Radiation damping effects ( 01JMR49-150 )