A phenomenological theory of nonphotochemical laser induced nucleation

Abstract

Our analysis of the experimental data related to nonphotochemical laser induced nucleation in solutions leads to the inevitable conclusion that the phase transformation is initiated by particles that are metallic in nature. This conclusion appears paradoxical because the final products are dielectric crystals. We show that the experimental results are well accounted for by the theory of electric field induced nucleation of metallic particles that are elongated in the direction of the field. However, new physical and chemical insights are required to understand the structure of the metallic precursor particles and the kinetics of subsequent dielectric crystallization.PACS numbers: 05.70. Fh, 64.60.qj, 64.70.dg, 82.60.Nh There is growing experimental support for electric field induced nucleation of solute particles in supersaturated solutions. First reported by Garetz et. al. [1], the phenomenon referred to as nonphotochemical laser induced nucleation (NPLIN) has been observed with both oscillating [1][2][3][4][5][6][7] and static [8] fields. The term 'nonphotochemical' emphasizes that there is no light absorption; hence, underlying changes in electronic structure capable of chemical reactions are ruled out. In all of the above cases, the final product of nucleation was found to be small dielectric particles.A clear indication that the field is the primary phase change driver is the alignment of nucleated particles along the direction of the applied field (or laser beam polarization). That phenomenon has led to a type of 'polarization switching' wherein the polymorph (crystal structure) of the nucleated crystal can be controlled by applying either linear or circular polarized light [2,3]. The underlying mechanism remains an open question with many practical implications [9].Our summary of NPLIN data from the literature is presented in Fig. 1 where the peak laser intensity (or applied field) and exposure time are provided at which crystallization was eventually observed in solutions at various supersaturation levels. For our purposes, the most important results are that: i) the field reduces the nucleation time by 13 orders of magnitude or even larger; and ii) it can do so at optical frequencies (∼ 10 14 Hz). Other observations include the existence of a threshold field, below which nucleation did not occur, and a linear type correlation between the cumulative fraction of samples nucleated and laser intensity [2,6], as well as the solute supersaturation [6].The anomalous strength of the observed field effect can be conveniently expressed in the terms of nucleation barrier W 0 that determines the nucleus induction time, τ = τ 0 exp (W 0 /kT ). Here, τ 0 > ∼ 10 −13 s is the characteristic atomic vibration or diffusion time, k is Boltzmann's constant, and T is temperature. The observed reduction of τ by a factor of 10 −13 requires decreasing the nucleation barrier by approximately ∆W (E) = W 0 −W (E) ∼ 30kT . The ordinate is the peak applied field (unless labeled as the threshold field) at which nucleation occurred...

Electrostatic Theory of Metal Whiskers

“…The region d = d 0 can be approximated by a potential wall. Following substantiation in the previous work, [29][30][31][32][33][34][35][36][37] we consider nucleation along the path d = d 0 ; alternative paths that start from the origin introduce only insignificant numerical factors. Correspondingly, we will use d < ∼ 1 nm in the above Eqs.…”

“…Based on data for other types of systems undergoing field induced nucleation, it was estimated that a reasonable minimum diameter is in the sub-nanometer range. [29][30][31][32][33][34][35][36][37] Lacking more concrete information, we assume here d 0 < ∼ 1 nm. In the region d > d min , the free energy is substantially larger than described by Eq.…”

Comment on “Threshold switching via electric field induced crystallization in phase change memory devices” [Appl. Phys. Lett. 100, 253105 (2012)]

“…We find the conclusion of Diosdado 1 and the underlying approaches 2,3 to be unjustified and misrepresenting our proposed mechanism of FIN, [4][5][6][7][8][9][10] as explained below.…”

“…The work by Diosdado 1 is limited to spherical particles, thereby ignoring the main content of FIN theory [4][5][6][7][8][9][10] that the field effect is dominated by elongated particles. As a result, the estimate by Diosdado 1 is related to E 0 and not to the experimentally relevant E $ E c ( E 0 .…”

“…Furthermore, we note that simulation results should not overshadow the experimental evidence that supports FIN while opposing electronic switching models, such as the observations of "underthreshold" switching, threshold voltage dependence on temperature and pulse width, and the statistical distributions of the delay time and threshold field. [4][5][6][7][8][9][10] Realistic simulation techniques must be able to reproduce those observations as well. We hope this comment will trigger a discussion useful to the entire community of phase change memory and the related topic of field effects in phase transitions.…”