3A molecular sieves are suitable for "usual" drying, and cited by David Jensen publication is good as introduction - with Karl Fisher titration and non-vacuum sieves regeneration. You can account on near 1 ppm of water
More serious investigations using tritium tracer was performed by David R. Burfield group (first publication from series http://pubs.acs.org/doi/abs/10.1021/jo00438a024?journalCode=joceah ). In comparison to 3A sieves, activated alumina should be more efficient (~ 100 times)
https://www.chem.tamu.edu/rgroup/gladysz/documents/nmr6.pdf - for polar solvents
@David Jensen - Sorry, I did not see your link text to the end. Thanks for link with all publications - I have been previously only paper copies of these articles
Thanks to David Jensen and to Andrei Rogosa for their documentation ( I havenot check every one). There is also ( Sorry if already given) : "Drying of Organic Solvents: Quantitative Evaluation of the Efficiency of Several Desiccants" by D. Bradley G. Williams* and M. Lawton, J. Org. Chem. 2010, vol 75, issue 24, pp. 8351-8354.