[en] Superparamagnetic iron oxide nanoparticles are used as contrast agents for molecular and cellular MRI. The dipolar interaction between their huge magnetic moment and those of water protons causes a drastic decrease of the proton transverse relaxation time T2. Different theoretical models have been developed to describe the relaxation process, namely the Motional Averaging Regime, the Static Dephasing Regime and the Partial Refocusing Model [1]. The adequate model should be chosen according to the size and magnetization of the particles but also according to the diffusion properties of protons. The transition between these different models has been investigated thanks to computer simulations of transverse relaxation [2] and comparison of T2 values obtained with particles of increasing size [3]. However, this latter method suffers from the influence of the size distribution of the particles on T2. We chose to change the diffusion properties of protons instead of using particles of varying sizes. This allows to explore the domains of validity of the theoritical models, with only one type of particles whose size is well defined.
Suspensions of 3nm iron oxide particles were prepared in different solvents: water, ethanol and dipropylène glycol. Suspensions of the particles in binary glycerol-water systems were also used. This allows to scan a large range of diffusion coefficients D. Then the transverse relaxation rate was measured at different temperatures and plotted (after correction from the solvent diamagnetic contribution) versus 1/D (figure 1). The obtained results are in qualitative agreement with the theoretical predictions even if some unexpected behavior has been noticed for fast diffusing systems. Further work will include T1 Nuclear Magnetic Relaxation Dispersion profiles and their fittings as well as the fitting of the data of figure 1 with the empirical model of Vuong et al [1]. Beyond the fact that the understanding of relaxation allows the optimization of the particles, the results presented here also show that the slowing down of water diffusion occuring in vivo (for example in the intracellular medium) will for sure influence the tranverse relaxation.
