Exact Solutions for Bimodal Distributions under Stochastic Plasma Irradiation in Thin Films

Joel Saucedo, Uday Lamba, Hasitha Mahabaduge, [4pt] Department of Chemistry, Physics, & Astronomy, Georgia College & State University, GA, 31061, USA, [-2pt] Department of Physics & Astronomy, Ithaca College, NY, 14850, USA · 2025

Read Exact Solutions for Bimodal Distributions under Stochastic Plasma Irradiation in Thin Films by Joel Saucedo, Uday Lamba, Hasitha Mahabaduge, [4pt] Department of Chemistry, Physics, & Astronomy, Georgia College & State University, GA, 31061, USA, [-2pt] Department of Physics & Astronomy, Ithaca College, NY, 14850, USA (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-07268.txt.

Introduction

Joel Saucedo, Uday Lamba, and Hasitha Mahabaduge settle a standing puzzle about films blasted by plasma: why their grain sizes split into two humps, and which scaling law really governs them. Their 2025 paper builds the first exact analytical theory for the system, uniting noise-driven evolution, defect saturation, and nucleation-loss balance in one formalism.

They derive the closed-form steady-state grain area distribution. A precise dimensionless threshold marks where the single hump splits into two. Defect saturation physics forces a universal scaling law linking mean grain area to flux and an Arrhenius temperature factor. The picture is competition: random impingement fragments the microstructure while steady growth heals it. The authors offer the result as a predictive base for plasma materials processing.

Catalog record

CREATORJoel Saucedo, Uday Lamba, Hasitha Mahabaduge, [4pt] Department of Chemistry, Physics, & Astronomy, Georgia College & State University, GA, 31061, USA, [-2pt] Department of Physics & Astronomy, Ithaca College, NY, 14850, USA
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE41 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2507-07268

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Exact Solutions for Bimodal Distributions under Stochastic Plasma Irradiation in Thin Films Joel Saucedo, Uday Lamba, Hasitha Mahabaduge, [4pt] Department of Chemistry, Physics, & Astronomy, Georgia College & State University, GA, 31061, USA, [-2pt] Department of Physics & Astronomy, Ithaca College, NY, 14850, USA Abstract A persistent paradox complicates the study of plasma-irradiated thin films, where bimodal grain distributions and ambiguous scaling laws—roughly shifting between Φ − 1 / 2 \Phi^{-1/2} and Φ − 1 \Phi^{-1} , or a general inverse dependence on plasma flux—are observed yet remain theoretically unreconciled. Existing models fail to unify noise-driven evolution, defect saturation kinetics, and nucleation-loss balance within a single, self-consistent formalism. This work resolves these discrepancies by developing the first exact analytical theory for this system. We derive the closed-form steady-state grain area distribution, P s ​ s ​ ( A ) P_{ss}(A) , establish the precise dimensionless threshold for bimodality onset at Π c = 4 / ( 3 ​ 3 ) \Pi_{c}=4/(3\sqrt{3}) , and demonstrate that defect saturation physics mandate a universal ⟨ A ⟩ ∝ κ 2 Φ − 1 e + 2 E b / k B T s \langle A\rangle\propto\kappa^{2}\Phi^{-1}e^{+2E_{b}/k_{B}T_{s}} scaling law. The framework reveals how competition between stochastic impingement and deterministic growth triggers microstructure fragmentation, resolving long-standing ambiguities in irradiation-induced surface evolution and providing a predictive foundation for materials processing. I. Introduction Plasma-based processing is a cornerstone of modern materials science, enabling tailored functional coatings and thin films [ 1 ] . Despite its widespread use, a predictive, first-principles understanding of microstructure evolution under irradiation is hindered by a fundamental experimental paradox. Across a range of physical vapor deposition (PVD) systems, two unresolved and seemingly contradictory phenomena are consistently reported. First is the emergence of bimodal grain size distributions, suggesting two distinct stabilization pathways coexisting under a single set of processing conditions [ 29 ] . Second is the conflicting scaling relationship between mean grain area \langle A\rangle and plasma flux \Phi , which reflects fundamental energy-driven transitions in microstructure evolution. At low normalized energy E^{*} and generalized temperature T^{*} , growth/coarsening dominates [ 16 , 20 ] , and at the low-energy edge of high-entropy [ 7 ] . Conversely, high E^{*}/T^{*} promotes nucleation dominance, producing \langle A\rangle\propto\Phi^{-1} scaling [ 17 ] , high-entropy at high energy [ 7 , 9 ] . This transition aligns with Anders’ extended structure zone diagram, where critical E^{*} thresholds (e.g., substrate bias \gtrsim 100-200 V) shift the balance from coarsening to nucleation regimes [ 1 ] , yet the exact bimodality threshold and scaling origins remained unresolved. This discrepancy points not merely to an incomplete model but to a foundational gap in our understanding of the interplay between stochastic and deterministic driving forces. The theoretical landscape has thus far been fragmented, with models addressing these complexities only in isolation. Classical theories of curvature-driven grain growth, for instance, provide a deterministic baseline but entirely neglect the stochastic nature of plasma bombardment, a simplification that proves untenable in any regime where ion-induced defect generation is significant [ 21 ] . Conversely, pure noise-driven models capture stochastic fragmentation but ignore essential saturation effects that limit grain growth at later stages [ 26 ] . More sophisticated frameworks incorporating defect-mediated recrystallization [ 32 ] have offered deeper insight but do not self-consistently integrate the kinetics of nucleation and loss. No existing theory has successfully unified three concurrently active phenomena: (i) the experimentally verified saturation of defect density at high flux [ 33 ] , (ii) the suppression of nucleation arising from direct particle impingement, and (iii) the noise-induced fragmentation that dominates the evolution of smaller grains. This paper presents an analytical framework that resolves these long-standing limitations. By synthesizing concepts from stochastic differential geometry with a self-consistent Fokker-Planck population balance formalism, we move beyond numerical approximations to derive exact, closed-form solutions for the steady-state microstructure. This approach yields three primary advances. First, we derive the complete grain size distribution, P_{ss}(A) , where the coupling between noise and impingement effects appears explicitly. Second, through a discriminant analysis of this distribution’s potential function, we establish the exact, dimensionless threshold for the onset of bimodality, \Pi_{c}=4/(3\sqrt{3}) . Third, our formalism demonstrates how the physics of defect saturation naturally and necessarily mandate the \langle A\rangle\propto\Phi^{-1} scaling law by governing the balance between grain nucleation and loss rates. Collectively, these results provide the first complete theoretical resolution of the plasma microstructure paradox, replacing empirical ambiguity with predictive capability. II. Theory To construct a predictive theory for microstructural evolution, one must first establish the fundamental stochastic processes that orchestrate the system’s dynamics. We posit that the evolution of a polycrystalline thin film, when subjected to a continuous ion flux \Phi , can be effectively captured within a continuum description. In this picture, the area A of an individual grain is not static but evolves through a competition between two primary mechanisms: a deterministic, curvature-driven growth and stochastic fluctuations induced by discrete plasma-surface interactions. This competition is mathematically embodied in the following stochastic differential equation (SDE): \mathop{}\!\mathrm{d}A=\kappa\sqrt{A}\left(1-\frac{A}{A_{\text{max}}}\right)\mathop{}\!\mathrm{d}t+\sigma\sqrt{A}\mathop{}\!\mathrm{d}W_{t} (1) Before proceeding, a note on the physical interpretation of these terms is warranted. The drift term is governed by the parameter \kappa=\pi M\gamma , which amalgamates the grain boundary mobility M and the boundary energy \gamma . Its proportionality to \sqrt{A} (i.e., the grain perimeter) reflects the classic theory of curvature-driven growth. However, this growth is not unbounded; the saturation term (1-A/A_{\text{max}}) serves as a crucial mean-field correction, modeling the impingement of growing grains against their neighbors. The effective maximum area, A_{\text{max}} , is not a fixed constant but is itself determined by the state of the system, typically scaling with the mean grain area, A_{\text{max}}=k\langle A\rangle , establishing a necessary self-consistency [ 15 , 21 ] . The second term represents the stochastic drive from the plasma. Its amplitude, \sigma=(2\Phi)^{1/2}a^{2} , where a is the atomic spacing, correctly captures the physics of a Poissonian arrival of ions, whose cumulative effect on the boundary manifests as a Wiener process W_{t} . \kappa\sqrt{A}\left(1-\frac{A}{A_{\text{max}}}\right)\mathop{}\!\mathrm{d}t \sigma\sqrt{A}\mathop{}\!\mathrm{d}W_{t} Figure 1: Conceptual illustration of the competing forces on grain boundaries. (a) Deterministic, curvature-driven growth causes large grains to consume smaller ones. (b) Stochastic fluctuations from plasma bombardment introduce noise that disrupts this process. \Pi<\Pi_{c} \Pi>\Pi_{c} Figure 2: Illustration of localized noise-induced fragmentation. (a) A stable, unimodal microstructure with a few large grains, characteristic of a low-noise environment ( \Pi<\Pi_{c} ). (b) The same microstructure where high, localized noise ( \Pi>\Pi_{c} ) has shattered the central and lower grains into a new population of small, gray-shaded cells, creating a distinctly bimodal structure. Concurrent with the evolution of existing grains is the birth of new ones. Nucleation is not a simple, memoryless Poisson process. Instead, it exhibits self-exciting behavior, where the formation of one grain can promote the formation of others nearby, often mediated by local strain fields. A Hawkes process provides a natural and powerful mathematical framework for this phenomenon, with a conditional intensity given by: \lambda(t)=\Gamma_{0}+\int_{-\infty}^{t}\Gamma_{0}e^{-(t-s)/\tau}\cos^{2}(2\theta)\mathop{}\!\mathrm{d}N_{s} (2) Here, the basal nucleation rate \Gamma_{0}\propto\exp(-E_{b}/k_{B}T_{s}) is set by the substrate temperature T_{s} and a fundamental energy barrier E_{b} . The integral term, however, is what endows the process with memory; it sums over all past nucleation events N_{s} , with

Read the complete text