Theoretically, there are tremendous challenges because there is no simple way to extract potential energy surfaces from force measurements. A given potential energy surface is a function of the many coordinates of the nanoparticles. Particularly for large molecules, there might be many different potential surfaces corresponding to the same force measurement. Nevertheless, one may be able to tackle this problem with a three-pronged approach:

1. Based on modeling experimental data, one may proceed with a top-down strategy. Namely, starting from the experimental data, one can construct a model for different molecules and clusters while drawing on empirical methods to help extract the essential features and build an appropriate potential model for a given compound.

2. One may also explore a bottom-up approach, by using first-principles methods (density functional theory [41] and quantum chemical techniques [42]) to calculate the potential surfaces for simple molecules. Here we are virtually free of any input parameters by constructing the model potentials and testing them on much larger molecules. This step can be iterated many times by comparing the quality of the potential with ab initio calculations. Once a reliable potential is obtained, it may be used to compute the properties associated with different structures in real materials. At this stage, almost no ab initio calculations can be done practically since they would exhaust the present available computational power. This readily leads us to the third step.

3. One may directly compare theoretical predictions with the...