Systematic errors associated with the CPMG pulse sequence and their effect on motional analysis of biomolecules

A theoretical approach to calculate the time evolution of magnetization during a CPMG pulse sequence of arbitrary parameter settings is developed and verified by experiment. The analysis reveals that off-resonance effects can cause systematic reductions in measured peak amplitudes that commonly lie in the range 5-25%, reaching 50% in unfavorable circumstances. These errors which are finely dependent upon frequency offset and CPMG parameter settings, are subsequently transferred into erroneous T-2 values obtained by curve fitting, where they are reduced or amplified depending upon the magnitude of the relaxation time. Subsequent transfer to Lipari-Szabo model analysis can produce significant errors in derived motional parameters, with tau(e) internal correlation times being affected somewhat more than S-2 Order parameters. A hazard of this off-resonance phenomenon is its oscillatory nature, so that strongly affected and unaffected signals can be found at various frequencies within a CPMG spectrum. Methods for the reduction of the systematic error are discussed. Relaxation studies on biomolecules, especially at high field strengths, should take account of potential off-resonance contributions. (C) 1997 Academic Press.