This phenomenon is not well understood.

http://pubs.rsc.org/en/content/articlelanding/2012/nr/c2nr30640a#!divAbstract

Think in terms of large surface to volume ratios and partially filled d-orbitals!!!!!!!!!!!!!!!!

Dear Abbas

The important role of size in determining the electronic, chemical, optical and magnetic properties of nanomaterials is well established. A significant recent revelation is ferromagnetism of nanoparticles of metal oxides and other inorganic materials which are otherwise diamagnetic in bulk form.[1] Of particular interest are the magnetic properties of nanoparticles of metals such as Au and Ag.As early as 1999, Hori et al.[2] observed magnetism in approximately 3 nm Au and Pd nanoparticles protected by polyvinylpyrrolidone (PVP) with unexpectedly large magnetic moments of about 20 spins per particle. Since then, a few other authors

have reported magnetism in Au nanoparticles,[3–8] although the origin of ferromagnetism in Au nanoparticles has remained somewhat unclear. In most of the reported studies, the Au nanoparticles were covered with capping agents such as thiols and amines. Crespo et al.[3] reported that gold nanoparticles

capped with weakly interacting reagents like tetraalkyl ammonium

bromides are diamagnetic, but those protected by strongly interacting thiols are ferromagnetic because of the 5d localized holes generated through Au–S charge transfer. The origin of orbital magnetism arising out of spin–orbit coupling

as a result of this charge transfer has been discussed.[4, 5] Hori and co-workers,[8] however, observed greater ferromagnetic spin coupling in gold nanoparticles weakly protected with polyacrylonitrile, poly(allylamine hydrochloride) and PVP. They attribute the particle-size dependence of ferromagnetism to

the ferromagnetic nature of the surface Au atoms due to the

so-called Fermi hole effect. On the basis of an X-ray circular dichroism

study Negishi et al.[9] concluded that localized holes created by AuS bonding are responsible for the spin polarization of gold clusters. While it would seem that the presence of thiol or amine capping agents is necessary for the magnetism

of Au nanoparticles on the basis of some of the reports, Reich et al.[10] observed magnetism in 27 nm thin films of Au deposited on Pyrex glass. More recently, Wu et al.[11] reported that icosahedral gold nanoparticles prepared by the gas-condensation method are ferromagnetic, whereby the moments of the core

and surface atoms point in opposite directions.

[1] a) J. M. D. Coey, S. A. Chambers, MRS Bull. 2008, 33, 1053; b) A. Sundaresan,

C. N. R. Rao, Solid State Commun. 2009, 149, 1197; c) A. Sundaresan,

C. N. R. Rao, Nano Today 2009, 4, 96.

[2] H. Hori, T. Teranishi, Y. Nakae, Y. Seino, M. Miyake, S. Yamada, Phys. Lett.

A 1999, 263, 406.

[3] P. Crespo, R. Litrn, T. C. Rojas, M. Multigner, J. M. de La Fuente, J. C.

Snchez-Lpez, M. A. Garca, A. Hernando, S. Penads, A. Fernndez,

Phys. Rev. Lett. 2004, 93, 087 204.

[4] A. Hernando, P. Crespo, M. A. Garca, Phys. Rev. Lett. 2006, 96, 057206.

[5] E. Guerrero, M. A. Mrquez, M. A. Garca, P. Crespo, E. F. Pinel, A. Hernando,

A. Fernndez, Nanotechnology 2008, 19, 175701.

[6] C. Gonzalez, Y. Simon-Manso, M. Marquez, V. Mujica, J. Phys. Chem. B

2006, 110, 687.

[7] Y. Yamamoto, T. Miura, M. Suzuki, N. Kawamura, H. Miyagawa, T. Nakamura,

K. Kobayashi, T. Teranishi, H. Hori, Phys. Rev. Lett. 2004, 93,

116801.

[8] H. Hori, Y. Yamamoto, T. Iwamoto, T. Miura, T. Teranishi, M. Miyake, Phys.

Rev. B 2004, 69, 174411.

[9] Y. Negishi, H. Tsunoyama, M. Suzuki, N. Kawamura, M. M. Matsushita, K.

Maruyama, T. Sugawara, T. Yokoyama, T. Tsukuda, J. Am. Chem. Soc.

2006, 128, 12034.

[10] S. Reicha, G. Leitus, Y. Feldman, Appl. Phys. Lett. 2006, 88, 222502.

[11] C.-M. Wu, C.-Y. Li, Y.-T. Kuo, C.-W. Wang, S.-Y. Wu, W.-H. Li, J. Nanopart.

Res. 2010, 12, 177.

The origin of magnetism in noble metal nanoparticles is not very clear till now. Local enhancement in the density of states and the highly enhanced surface-to-volume ratio of these nanoscale systems are the major reasons for which ferromagnetism is observed in nanoscale clusters of systems that do not exhibit ferromagnetic ordering in the bulk. This has to do with the confinement effects associated with reduced coordination number, local symmetry changes due to the effect of vacancies or by increasing the percentage of atoms at the grain boundaries or interfaces and  changes of the lattice constant at the surface that induce a narrowing of the d band. You can have a look at http://iopscience.iop.org/article/10.1088/0953-8984/25/48/484006/meta;jsessionid=D98011B361EE9A495A921B265E03F3DE.c1