From the statement of your question it appears that you are doing (or are familiar with doing) bulk calculations, to be contrasted with (quantum-chemical) cluster calculations. If so, then the short answer to your question is: do super-cell calculations. You choose a Bravais lattice with sufficiently large lattice constant and place the atoms of the nanoparticle at the points characterized by the basis vectors {t_1, t_2, ....}. For the super cell centred at the lattice point R_i, the atoms are located at R_i + t_1, R_i + t_2, ... (for the central super cell, centred at the origin of the coordinate system, that is R_1 = 0, the atoms are located at {t_1,t_2, ...}). For the concept of 'lattice with a basis', see Ashcroft & Mermin, p. 75. Now, so long as the adopted super cell is sufficiently large, so that the interactions between the atoms in a super cell with those in the neighbouring super cells are negligible, the properties one calculates are to a high accuracy those of the nanoparticle one intends to study.
Dear Hicham, you are welcome. The formalism I described above is independent of the type of potential one uses. It works both for the cases where one uses empirical pair potentials, and for the cases where one uses real or pseudopotentials, as in ab initio quantum-mechanical calculations.
Since your main question on this page is about "molecular dynamics simulations", all the relevant references are given in the above-mentioned review article. You may also wish to consult the following two books:
R.M. Martin, Electronic Structure - Basic Theory and Practical Methods (Cambridge University Press, 2008)
E. Kaxiras, Atomic and Electronic Structure of Solids (Cambridge University Press, 2003)