Offset effects in the roesy Due to the very limited power of the r.f. spin-lock fields(a few kHz, compared to several hundred MHz for the static field), an efficient spin-lock in the transverse x,y plane occurs only close to resonance frequency of the irradiated field (usually in the center of the spectral window ) Towards edges of the spectrum, the spin-lock field becomes to weak to force the spins into a precession about its transverse axis. As a result, the spins precess about an effective axis that gradually tilts from the x, y plane(on resonance) back to the axis of the static field(far off-resonance) This has several consequences off resonance, the efficieny of the roe drops off with sin 0, sin 0, for the rOE from proton i to proton, where 0 is the angle between the spins' precession axis and the x, y plane. This angle can be calculated for each spin from the ratio between the frequency offset (causing a precessing about the z axis)and the spin- lock field(causing the x, y precession component) oc arctan where Q2, is the offset of the spin from on-resonance and y B, the field strength of the spin-lock field (both in Hz) ince we go into the spin-lock field with pure transverse magnetization of spin i(from the initial 90 pulse), we add another factor sin 0, since we can only lock the magnetization parallel to the spin-lock axis. Similarly, we have to add another sin e when we detect the resulting spin j magnetization during t2, since we can only see the x, y component of the spinlocked magnetization All this means that our measured roeexn has to be corrected to get the true roe correlated with the interproton distance A"compensated rOESY version has also been introduced, with a weaker offset dependence 90 t2 spin-lock x Here the required correction for the cross-peak integrals is104 Offset effects in the ROESY Due to the very limited power of the r.f. spin-lock fields (a few kHz, compared to several hundred MHz for the static field), an efficient spin-lock in the transverse x,y plane occurs only close to the resonance frequency of the irradiated field (usually in the center of the spectral window). Towards the edges of the spectrum, the spin-lock field becomes to weak to force the spins into a precession about its transverse axis. As a result, the spins precess about an effective axis that gradually tilts from the x,y plane (on resonance) back to the z axis of the static field (far off-resonance). This has several consequences: - off resonance, the efficieny of the ROE drops off with sin qi sin qj for the ROE from proton i to proton j, where q is the angle between the spins' precession axis and the x,y plane. This angle can be calculated for each spin from the ratio between the frequency offset (causing a precessing about the z axis) and the spin-lock field (causing the x,y precession component): q g i i B µ arctan W 1 where Wi is the offset of the spin from on-resonance and gB1 the field strength of the spin-lock field (both in Hz). - since we go into the spin-lock field with pure transverse magnetization of spin i (from the initial 90° pulse), we add another factor sin qi , since we can only lock the magnetization parallel to the spin-lock axis. Similarly, we have to add another sin qj when we detect the resulting spin j magnetization during t2 , since we can only see the x,y component of the spinlocked magnetization. All this means that our measured ROEexp. has to be corrected to get the true ROE correlated with the interproton distance: ROEtrue ROE ij ij i j = exp. sin sin 1 2 2 q q A "compensated ROESY" version has also been introduced, with a weaker offset dependence: Here the required correction for the cross-peak integrals is