Critical thickness for ferroelectricity in perovskite ultrathin films

Abstract

The integration of ferroelectric oxide films into microelectronic devices, combined with the size reduction constraints imposed by the semiconductor industry, have revived interest in the old question concerning the possible existence of a critical thickness for ferroelectricity. Current experimental techniques have allowed the detection of ferroelectricity in perovskite films down to a thickness of 40 A (ten unit cells), ref. 3. Recent atomistic simulations have confirmed the possibility of retaining the ferroelectric ground state at ultralow thicknesses, and suggest the absence of a critical size. Here we report first-principles calculations on a realistic ferroelectric-electrode interface. We show that, contrary to current thought, BaTiO3 thin films between two metallic SrRuO3 electrodes in short circuit lose their ferroelectric properties below a critical thickness of about six unit cells (approximately 24 A). A depolarizing electrostatic field, caused by dipoles at the ferroelectric-metal interfaces, is the reason for the disappearance of the ferroelectric instability. Our results suggest the existence of a lower limit for the thickness of useful ferroelectric layers in electronic devices.

Ferroelectricity in ultrathin perovskite films

“…In contrast to the PbTiO 3 films, none of the BaTiO 3 structures were found to be ferroelectric for m=2 or m=4, in agreement with the results in Ref. 13.…”

“…In a continuum model by Batra et al 4 which includes the depolarization effect, a critical thickness of 100Å for perovskite films was analytically derived, assuming a Thomas-Fermi screening length of 1Å for the metal electrodes. Also, a recent first-principles calculation revealed that BaTiO 3 films with SrRuO 3 electrodes lose ferroelectricity below ∼24Å (6 unit cells), 13 thus suggesting that a minimum thickness limit exists for useful ferroelectric films. While indeed a minimum film thickness must be influenced by the polarization of the ferroelectrics and the screening length of the electrodes, 14 it is not yet clear whether the depolarizing field can ever be completely removed by realistic electrodes on ultrathin films, nor how monodomain thin film ferroelectricity is affected by the choice of electrodes and by the interactions at the metal/oxide interface.…”

Direct evidence for ferroelectric polar distortion in ultrathin lead titanate perovskite films

“…In Ref. [9], this polarization reduction was attributed to imperfect screening of the depo-larizing field [12]. With XPD, using Pb as emitter, the tetragonality was measured down to the unit cell level as shown in Fig.…”

“…In particular, "nanoscale" ferroelectrics have attracted considerable attention [1,2,3,4,5]. The question of the existence of a critical thickness, in other words whether or not ferroelectricity can be maintained at reduced dimensions, is amongst the most exciting topics of the field today, with very active experimental [6,7,8,9] and theoretical efforts [10,11,12].…”

Strain gradients in epitaxial ferroelectrics

“…1,2 The incorporation of realistic mechanical and electrical boundary conditions in the first-principles formulations is generating insight into the mechanisms limiting the ferroelectric response in thin ferroelectric layers. [3][4][5] But while the evidence suggests that ferroelectricity may indeed be stable in films of only a few monolayers, 1-3 the sharp peak in the dielectric constant usually associated with the ferroelectric transition is systematically depressed even in films of hundreds of nanometers in thickness. This obviously limits the technological impact that would arise from the ability to maintain ferroelectricity and large dielectric constants down to the nanoscale in real devices.…”