betaFeedback

WorksPhysics and Astronomy

Coupling Evaporation, Chemical Reactions, and Crystallization: A Population-Balance Framework Applied to Two-Step Perovskite Crystallization

Abstract

A limited understanding of how evaporation couples to crystallization pathways hampers control over the formation of solution-processed perovskite photoactive layers. Identifying these mechanisms-and learning how to select and tune them-is essential for achieving the desired crystal morphology and, ultimately, improving solar-cell performance and stability. To investigate crystallization pathways in evaporating thin films, we develop a simple and general zero-dimensional population-balance model that couples evaporation, chemical reactions, and crystallization. We use the model to explore a range of parameters for a solvent-solute system that follows a two-step crystallization pathway: solvent evaporation first induces formation of a crystalline precursor, which subsequently recrystallizes into the final phase. The model predicts the evolution of the film composition during drying and provides quantitative LaMer curves. Based on these results, we introduce a vocabulary of "supply limitations" and "rate limitations" that enables the interpretation of crystallization kinetics in terms of the governing process conditions.

Introduction

Solar cells are a primary energy-harvesting technology, with steadily decreasing energy costs per kilowatt-hour and, in many regions, a lower levelized cost of electricity than fossil-based generation. They contribute significantly to the renewable energy transition [1]. Solar cells with the potential for solution processing that avoid costly traditional high-vacuum, high-temperature fabrication conditions are progressing toward pilot-scale and early-stage industrial production [2]. Among the solution-processable photoactive layer materials, for which roll-to-roll production-compatible techniques, such as slot-die coating, have been demonstrated, metal halide perovskites are a promising class. This owes to their compositionally tunable optoelectronic properties and the high power conversion efficiencies of the resulting solar cells [3]. However, solution processing of thin films poses fundamental challenges for controlling their formation. For perovskites, the films’ optoelectronic properties are closely linked to film roughness, grain size, density, and grain-boundary density. Thus, achieving reproducible, durable, high-performing devices requires precise control of crystallization [4, 5]. This control, however, is challenging; in many deposition protocols, solvent evaporation dynamically controls crystallization, triggering nucleation after a critical solution concentration is reached, and determines the time evolution of the concentration-dependent nucleation and crystal growth rates. The chemical activity of the precursor solutions further amplifies this complexity.

To elucidate film formation mechanisms, a range of in situ characterization techniques has been used. Early studies relied primarily on ex situ X-ray diffraction (XRD) to analyze the final crystalline phases and infer crystallization pathways retrospectively [6]. More recently, time-resolved in situ techniques—including grazing-incidence wide-angle X-ray scattering (GI-WAXS) [7], optical laser or white light reflectometry (LR/WLR) [8, 9], and combined multimodal approaches [10]—have enabled direct observation of intermediate phases and phase transitions during drying and annealing [11]. These studies have significantly advanced the mechanistic understanding of perovskite crystallization [12, 13, 14, 15, 16, 17, 18, 19]. Strong solvent-precursor interactions can lead to the formation of colloidal pre-clusters [20] and solvent-coordinated crystalline intermediate phases prior to the final perovskite. Intermediates alter supersaturation dynamics and provide alternative nucleation pathways: The perovskite may form directly, in parallel to, or from intermediate phases [21]. Consequently, the final film morphology—including grain size distribution, grain boundary density, and defect density—is highly sensitive to the specific crystallization pathway followed during deposition [22, 23]. Despite these advances in in situ monitoring, many investigations remain incomplete and qualitative [24]. In particular, evaporation kinetics, which control crystallization, are not consistently measured. Moreover, suggested film-formation pathways are rarely quantitatively tested against the data. In parallel, modeling frameworks that integrate chemical reactions, solvent evaporation, and crystallization are lacking in photovoltaic research. Quantitative modeling is rare and typically relies on decoupled, substantially simplified process descriptions. Evaporation of two solvents from a perovskite precursor solution has been successfully modeled under realistic coating conditions in [12, 19, 25] but without accounting for the solidification of the film. Furthermore, a single secondary perovskite crystallization could be modeled as a single process in [26] and [27], but the role of evaporation and prior solid-state precursor crystallization has not been likewise considered. In other fields, for instance, chemical engineering, more holistic approaches are employed. For chemical reactions, for example, automated model discrimination provides systematic strategies for assessing reaction pathways using experimental data [28]. For particulate systems, population balance modeling (PBM) can serve as a framework to describe how multiple distinct processes, such as chemical reactions, evaporation, and crystallization, jointly affect particle populations (e.g., crystals) [29]. For instance, it has been used for direct crystallization control of a polymer [30], to understand salt crystallization in droplets [31], or to rationalize the growth of single perovskite crystals at constant supersaturation [32]. However, such integrated approaches have not yet been widely applied to the formation of perovskite thin films. Applied modeling is rare and typically relies on decoupled or overly simplified process descriptions [26, 27, 33]. A general population balance model framework to clarify the role of evaporation in crystallization and perovskite crystallization pathways is still lacking.

Therefore, given the experimental findings on hypothetical mechanisms and the modeling state outlined above, we conclude that it is useful to propose a model that can simultaneously handle multiple chemical reactions, the evaporation of multiple solvents, and multiple successive or competitive crystallization processes.

In this paper, we propose a general zero-dimensional kinetic model for the evolution of the material composition in reactive, evaporating, and crystallizing thin films. The model, in principle, allows us to quantitatively compare a variety of process systems encompassing multiple coupled evaporation, crystallization, and chemical processes. Subsequently, in a simulation study, we investigate evaporation-induced successive crystallization for perovskite formation.

Model description

In this section, we detail our general modeling framework. It allows us to describe a set of processes (chemical reactions, evaporation, crystallization, sublimation, condensation, light-, or stress-induced evolution, demixing,...), over a set of states defined, for instance, by phase (e.g., gaseous (gg), liquid (ℓ\ell), or solid (ss)) and molecular species (e.g. H2_2O).

Population balance model

First, we outline the theory’s general idea. Then, for each process type, we go through the basic building blocks and the rate expressions in turn.

Consider a process pp converting nn reagents AiA_i into mm distinct products AjA_j with all i≠ji \ne j:

a1A1+…anAn→an+1An+1+…+an+mAn+ma_1 A_1 + \ldots a_n A_n \rightarrow a_{n+1} A_{n+1} + \ldots+ a_{n+m} A_{n+m}

Here, a1,…,an+ma_1,\ldots,a_{n+m} are the reactant and product stoichiometric coefficients. In the following, we consider only chemical reactions, crystallization, and evaporation processes, but these can, in principle, be replaced by any other process (sublimation, condensation, light- or stress-induced evolution, demixing, ...). Thereby, in this work, the reagents and products are defined by their chemical nature and phase state (e.g., chemical species in the liquid phase, a gaseous solvent, atoms or molecules in the crystalline state, etc.). However, more generally, they can also be defined by additional properties, such as mass or size (e.g., a polymer chain of given molar mass, a particle of given volume, ...). Then, the population balance for the number of moles nn of all reagents and products is given by the process advancement rate ξp\xi_p of process pp:

ξp=−1a1dn1dt=…=−1andnndt=1an+1dnn+1dt=…=1an+mdnn+mdt\xi_p = -\frac{1}{a_1}\frac{dn_1}{dt} = \ldots= -\frac{1}{a_n}\frac{dn_n}{dt} = \frac{1}{a_{n+1}}\frac{dn_{n+1}}{dt} = \ldots= \frac{1}{a_{n+m}}\frac{dn_{n+m}}{dt}

This allows us to obtain the contribution of process pp to the population evolution of any reagent or product kk as

dnk,pdt=ϵpak,pξp,\frac{dn_{k,p}}{dt} = \epsilon_p a_{k,p} \xi_p,

whereby the sign ϵp\epsilon_p is negative or positive depending on whether k=1...(n+m)k = 1...(n+m) appears as a reactant or product. Finally, if there are NpN_p active processes in the system, the global population evolution of each species is determined by the evolution of all processes. More specifically, if a species kk is involved as either reactant or product in Np,kN_{p,k} of the NpN_p processes with stoichiometry ϵpak,p\epsilon_p a_{k,p}, its population changes as

dnkdt=∑p=1Np,kdnk,pdt=∑p=1Np,kϵpak,pξp.\frac{dn_k}{dt} = \sum_{p=1}^{N_{p,k}}\frac{dn_{k,p}}{dt} = \sum_{p=1}^{N_{p,k}}\epsilon_p a_{k,p}\xi_p.

The equation above expresses the population balance for all species subject to the set of active processes and enables evaluation of the time-dependent evolution of species populations.

In the following, we detail the process-specific rates ξp\xi_p for chemical reactions, evaporation, and crystallization. Thereby, the starting point for a process pp will typically be an empirical volumetric process advancement rate ξpVM,p\frac{\xi_p}{V_{M,p}}. This rate is measured with reference to the volume VM,pV_{M,p} within which the process takes place, which we refer to as the ”mother phase volume” hereafter. The definition of the mother phase depends on the process and will be detailed below.

Chemical reactions

As a kinetic model for chemical reactions, we use a generalization of the of mass-action rate equations [34] formulated through chemical potentials rather than concentrations [35]. For a chemical reaction involving mm reactants AjA_j, with stoichiometry aja_j with j=1...mj=1...m, whose chemical potentials μj(ℓ)\mu_j^{(\ell)} in the liquid phase depend on the volume fractions ϕj\phi_j, the chemical reaction rate ξχ\xi_\chi is

ξχ=VMκ∏j=1meajμj(ℓ)/RT.\xi_\chi= V_M \kappa\prod_{j=1}^{m} e^{a_j\mu_j^{(\ell)}/RT}.

For our target application, we consider reactions in a common liquid phase VM=V(ℓ)V_M = V^{(\ell)}. Here, κ\kappa is an Arrhenius-type rate prefactor, assumed independent of composition, and the reaction orders are taken to equal the stoichiometric coefficients aja_j.

To characterize the mixture thermodynamics, we assume an incompressible liquid that can be described by the Flory-Huggins-Gibbs free energy density

fV(ℓ)=RTv0(∑i1Niϕiln⁡ϕi+12∑ijϕiχijϕj).f_V^{(\ell)} = \frac{RT}{v_0}\left(\sum_i \frac{1}{N_i}\phi_i\ln\phi_i + \frac{1}{2}\sum_{ij}\phi_i\chi_{ij}\phi_j\right).

where the χij=χji\chi_{ij} = \chi_{ji}, χii=0\chi_{ii} = 0 denote pairwise material volume interaction parameters and Ni=νi/ν0N_i = \nu_i/\nu_0 relative molar volumes with molar volumes νi\nu_i and reference ν0=min⁡jνj\nu_0 = \min_j \nu_j. From the free energy density, we deduce the material-resolved chemical potentials for the liquid part of the blend:

μi(ℓ)RT=(1+ln⁡ϕi−Ni∑j1Njϕj+Ni∑jχijϕj−Ni12∑kjϕkχkjϕj).\frac{\mu_i^{(\ell)}}{RT} = \left(1 + \ln\phi_i - N_i\sum_j \frac{1}{N_j}\phi_j + N_i\sum_j\chi_{ij}\phi_j - N_i\frac{1}{2}\sum_{kj}\phi_k\chi_{kj}\phi_j\right).

Thus in ideally mixing solutions χ=0\chi= 0 with equal molar volumes νi=νj\nu_i = \nu_j, the chemical rate reduces to the standard mass action product of the concentrations ξχ∼VM∏j=1meajln⁡ϕj∼VM∏jCjaj\xi_\chi\sim V_M\prod_{j=1}^{m}e^{a_j\ln\phi_j}\sim V_M\prod_j C_j^{a_j} where we used that the volume fractions are proportional to the concentrations by the molar volumes ϕj=νjCj\phi_j = \nu_j C_j.

In general, each reaction thus introduces two free kinetic parameters, the pre-exponential factor and the activation energy of κ\kappa, which reduce to a single value κ(T)\kappa(T) under isothermal conditions.

Evaporation

Consider a thin film, consisting of a mixture containing, among m−km-k non-evaporating materials (solutes and potentially some crystals), kk evaporating solvents AiA_i. We describe the evaporation process as a phase transformation taking place at the surface SS separating the condensed film from the gas phase,

Ai(ℓ)→Ai(g),A_i^{(\ell)} \rightarrow A_i^{(g)},

whereby the ii-th liquid solvent AiA_i becomes gaseous and leaves the film.

For the rates, we use one of the semi-empirical expressions derived from the Hertz–Knudsen (HK) model [36], [37]. For each solvent ii, we have

ξe,i=S(t)α2πmikBT(Pi∗(T)eμi(ℓ)/RT−Pi∞),(1)\xi_{e,i} = S(t)\frac{\alpha}{\sqrt{2\pi m_i k_B T}}\left(P_i^*(T)e^{\mu_i^{(\ell)}/RT}-P_i^\infty\right), \tag*{(1)}

which yields the expected average particle effusion rate through the surface. We approximately account for the slowdown of evaporation due to a reduction in initial surface area S0S_0 as the film becomes crystalline through the relative liquid-to-total film volume S(t)≈S0V(ℓ)V(ℓ)+V(s)S(t)\approx S_0\frac{V^{(\ell)}}{V^{(\ell)}+V^{(s)}}. The molecular escape probability α\alpha depends only on the flow conditions and surface geometry. It can be considered constant given stationary drying conditions at constant temperature TT [38, 39]. For instance, α\alpha can be used to account for the drying conditions encountered during spin coating [40, 41, 42] and gas quenching [19]. The basic assumption of HK, that the liquid-vapor interface is in equilibrium, i.e., μi(ℓ)=μi(g)\mu_i^{(\ell)}=\mu_i^{(g)}, yields the solvent pressure Pi∗eμi(ℓ)/RTP_i^*e^{\mu_i^{(\ell)}/RT} at the film surface. Thereby, Pi∗(T)P_i^*(T) is the ii-th solvent vapor pressure and μi(ℓ),μi(g)\mu_i^{(\ell)},\mu_i^{(g)} are its chemical potential in the liquid and gas phases, and RR is the gas constant. The driving force for evaporation is the pressure difference between the film surface and the environment (Pi∗(T)eμi(ℓ)/RT−Pi∞)(P_i^*(T)e^{\mu_i^{(\ell)}/RT}-P_i^\infty). For gas molecules of mass mim_i, with kBk_B the Boltzmann constant, multiplying the pressure by 12πmikBT\frac{1}{\sqrt{2\pi m_i k_B T}} yields the rate at which particles cross a unit area of a plane from one side.

All in all, for evaporation, we consider one drying-condition-dependent adjustable parameter, α\alpha, and n(n−1)/2n(n-1)/2 pairwise interaction parameters χij\chi_{ij}.

Crystallization

For the derivation of a crystallization rate, we take inspiration from classical nucleation theory (CNT) [43] and the visualizations from spatial simulation approaches, such as [44], but mostly we lean on the commonly employed kinetic approach due to Kolmogorov-Johnson-Mehl-Avrami (KJMA) [45, 46].

A single crystallization process

We conceptualize crystallization as a particular phase transition (first order) where an liquid or solid material (ℓ/s1)(\ell/s_1) transitions to a different solid phase state (s/s2)(s/s_2) through the sub-processes of nucleation and growth. Consider, for instance, the phase transition of a single material XX:

X(ℓ/s1)→X(s/s2)X^{(\ell/s_1)} \to X^{(s/s_2)}

During nucleation, according to CNT, clusters of crystalline material called nuclei form within the mother phase volume VMV_M due to thermal fluctuations. These nuclei either shrink or, when they exceed a critical size rcr_c [mm], become stable nuclei that grow into macroscopic crystals. Thus, crystals begin to form at a characteristic nucleation rate density η˙\dot{\eta} [#/md^ds] - that is, the number # of supercritical nuclei per unit time and per d-dimensional mother-phase volume. Note that with a composition-dependent nucleation rate density η˙\dot{\eta}, the expected first crystal arrival or induction time is already implicitly determined.

During crystal growth, the sizes of the existing crystals increase. With each crystal, we associate a radius

r(t,τ)=rc(t≥τ)+∫τtg(t′)f(t′) dt′,(2)r(t,\tau)=r_c(t\geq\tau)+\int_{\tau}^{t}g(t')f(t')\,\mathrm{d}t', \tag*{(2)}

whereby τ\tau [ss] is the time of birth of the crystal, and gg [m/sm/s] is the velocity at which the crystal’s interface propagates, i.e., the ”growth rate”. The ”impingement function” ff accounts for the geometrical confinement of crystals by boundaries due to other crystals that hinder further growth into the directions where crystal interfaces touch. It decreases monotonically from 1, corresponding to freely growing crystals at the beginning of the process, to 0 when the volume is fully crystalline and all crystals impinge.

The crystallization rate can be expressed as the sum of the volume increases due to nucleation and growth:

ξCr=1νX(ℓ)(BdrcdVM(t)η˙(t)+dBdr˙(t)∫0t(rc+∫τtg(t′)f(t′) dt′)d−1VM(τ)η˙(τ) dτ)(3)\xi_{\mathrm{Cr}}=\frac{1}{\nu_{X^{(\ell)}}}\left(B_d r_c^d V_M(t)\dot{\eta}(t)+dB_d\dot{r}(t)\int_0^t\left(r_c+\int_{\tau}^{t}g(t')f(t')\,\mathrm{d}t'\right)^{d-1}V_M(\tau)\dot{\eta}(\tau)\,\mathrm{d}\tau\right) \tag*{(3)}

Here, Bd=πd/2Γ(d/2+1)B_d=\frac{\pi^{d/2}}{\Gamma(d/2+1)} denotes the volume of a unit sphere, with gamma function Γ\Gamma and dd the number of linearly independent directions into which crystals grow, termed ”growth dimension”. The expression can be interpreted as follows: The first part of the right-hand side of Eq. (3) corresponds to nucleation: During a time step dt\mathrm{d}t, a number of VM(t)η˙(t)dtV_M(t)\dot{\eta}(t)\mathrm{d}t nuclei of volume BdrcdB_d r_c^d form in the mother phase volume VMV_M. The second part corresponds to growth: already-present nuclei contribute volume changes according to their time of appearance. The

volume change at time t>0t>0 for each of the n˙(τ)dτ=VM(τ)η˙(τ)dτ\dot{n}(\tau)\mathrm{d}\tau=V_M(\tau)\dot{\eta}(\tau)\mathrm{d}\tau crystals born at τ\tau is equal to dBdrd(t,τ)dt=dBdr˙(t)r(t,τ)d−1\frac{\mathrm{d}B_d r^d(t,\tau)}{\mathrm{d}t}=dB_d\dot{r}(t)r(t,\tau)^{d-1}. The volume increase due to all crystals born at all times τ\tau between 0 and tt is therefore equal to ∫0tdBdrd(t,τ)dtdτ=dBdr˙(t)∫0trd−1(t,τ)VM(τ)η˙(τ)dτ\int_0^t \frac{\mathrm{d}B_d r^d(t,\tau)}{\mathrm{d}t}\mathrm{d}\tau=dB_d\dot{r}(t)\int_0^t r^{d-1}(t,\tau)V_M(\tau)\dot{\eta}(\tau)\mathrm{d}\tau, whereby r(t,τ)r(t,\tau) is given by Eq. (2).

Eq. (3) constitutes our general crystallization model. However, to calculate ξCr\xi_{\mathrm{Cr}} for a concrete system, we need to specify further the nucleation rate density η˙\dot{\eta} and growth rate gg as well as the impingement function ff for each crystallization process. Let us now discuss these quantities in detail for a system relevant to perovskites involving multiple crystallization processes.

Example including multiple crystallization processes

Consider three materials SS, XX, and CC, where SS and part of XX are liquid (ℓ)(\ell), and CC and part of XX are crystalline (s)(s). The process system

X(ℓ)+2S(ℓ)⇌C(ℓ)X^{(\ell)}+2S^{(\ell)}\rightleftharpoons C^{(\ell)}
C(ℓ)→C(s)C^{(\ell)}\to C^{(s)}
C(s)→X(s)+2S(ℓ)C^{(s)}\to X^{(s)}+2S^{(\ell)}
X(ℓ)→X(s)X^{(\ell)}\to X^{(s)}

includes a direct crystallization (P3) competing with a two-step crystallization (P0-P2). Once X(ℓ)X^{(\ell)} combines with 2S(ℓ)2S^{(\ell)} into C(ℓ)C^{(\ell)} (P0), C(ℓ)C^{(\ell)} crystallizes (process P1) and eventually transforms to crystalline X(s)X^{(s)}, releasing S(ℓ)S^{(\ell)} (P2). Thus, if we interpret S(ℓ)S^{(\ell)} as the solvent and X(ℓ)X^{(\ell)} as the dissolved precursor, P3 represents a direct crystallization, and P0-P2 a secondary crystallization of perovskite via solid-state intermediate phase formation of X(ℓ)X^{(\ell)} from solution.

Nucleation rate density and growth rate

To specify nucleation rate densities and growth rates, we draw inspiration from the commonly referred to as LaMer theory [47, 48]. LaMer theory, in essence, introduces two distinct thresholds for nucleation (ϕR∗)(\phi_R^*) and growth (ϕR∗)(\phi_R^*) that determine when each phase begins and ends, which we refer to as the criticality threshold and solubility limit.

For each crystallization, we use a reaction volume fraction ϕR=VRVM\phi_R=\frac{V_R}{V_M}, where the reactant volume VRV_R is the amount of material available to the process in the mother phase. For Eq. (4) for instance, ϕR1=VC(ℓ)VC(ℓ)\phi_{R_1}=\frac{V_{C^{(\ell)}}}{V_{C^{(\ell)}}} for P1 and ϕR2\phi_{R_2} for P2 is one if the mother phase C(s)C^{(s)} is available and zero if not. Thus if for example P0 is balanced on the time scale of P1, one effectively has VR1=k+k−VS(ℓ)2VX(ℓ)V_{R_1}=\frac{k_+}{k_-}V_{S^{(\ell)}}^2V_{X^{(\ell)}} where k+k_+, k−k_- are the forward and backward rates for P0.

With ϕR\phi_R and ϕR∗,ϕRc\phi_R^*,\phi_R^c, one can then generally define composition-dependent nucleation and growth rates. In the following, we will, for instance, use

\begin{aligned} \dot{\eta} &= \dot{\eta}_0 \max\left(\phi_R-\phi_R^c,0\right),\\ g &= g_0 \max\left(\phi_R-\phi_R^*_R,0\right), \tag*{(4)} \end{aligned}

where η˙0,g0\dot{\eta}_0,g_0 are constant scalings. Eq. (4) constitutes a simple choice where both growth and nucleation accelerate as the available volume fraction of the crystallizing material increases beyond the thresholds ϕR,ϕRc\phi_R,\phi_R^c. Generally, through refined expressions, additional effects, such as changes in the mixture's diffusion properties, curvature-driven growth, etc., can be accommodated.

Mother phases and impingement functions

To handle the process system Eq. (4), we introduce the following simplifying assumptions:

A1: Solid products of a crystallization cannot extend beyond the boundaries of their mother phase.

A2: Every crystalline phase can impinge against any other crystalline phase growing in the same phase.

A3: Liquid crystallization products end up in a common liquid phase.

In Figure 1, the situation is schematically visualized, assuming that S(ℓ)S^{(\ell)}, generated by P2, reenters the liquid phase. On the one hand, both C crystals and X crystals, due to the direct crystallizations P1 and P3, are produced from the liquid phase. Therefore, they share the liquid phase as their mother phase and thus can impinge on each other. On the other hand, X crystals due to P2 are produced in the C(s)C^{(s)} phase, and therefore their mother phase

Sketch of material domains for competitive successive crystallization in a dilute liquid phase

Figure 1. Sketch of material domains for competitive successive crystallization in a dilute liquid phase (ℓ\ell). On the one hand, C(s)C^{(s)} and X(s)X^{(s)} directly crystallize from the liquid phase, and as solids, they occupy separate domains. On the other hand, X(s)X^{(s)} resulting from re-crystallization of C(s)C^{(s)} appears inside C(s)C^{(s)}'s domain and S(s)S^{(s)} thereby released reenters the common liquid phase. Because the conditions regarding material supply and crystal impingement differ for the crystallization processes between the liquid phase on the one hand (ℓ\ell) and the C(s)C^{(s)} phase on the other hand, material supply and crystal impingement need to be distinguished between the processes.

is the C(s)C^{(s)} phase, i.e., a separate solid phase. Thus, if we assume that the growing X crystals due to P2 remain within the C(s)C^{(s)} phase domains, they can only impinge on each other but not on X crystals from process P3.

Therefore, considering Figure 1, we define the impingement volume VIV_I as the volume of subdomains included in the mother-phase where crystallization is prohibited. Based on this definition, we make the following ansatz:

f=ϕIf=(VMVM+VI)γf=\phi_I^f=\left(\frac{V_M}{V_M+V_I}\right)^\gamma

Here, ϕI\phi_I represents the impingement volume fraction, and γ\gamma is a positive empirical exponent. Typically, the mother phase volume VMV_M is consumed by crystallization, whereas the impingement volume VIV_I corresponds to the volume of crystalline species.

Returning to our previous discussion of Eq. (4), for the liquid to solid transitions P1 and P3, the mother phase and impingement volumes are VM=V(ℓ)V_M=V^{(\ell)} and VI=V(s)=VC(s)+VXP2(s)+VXP3(s)V_I=V^{(s)}=V_{C^{(s)}}+V_{X^{(s)}_{P2}}+V_{X^{(s)}_{P3}}. For the solid to solid transition P2, the mother phase and impingement volume are VM=VC(s)V_M=V_{C^{(s)}} and VI=VXP2(s)V_I=V_{X^{(s)}_{P2}}, where XP2(s)X^{(s)}_{P2} and XP3(s)X^{(s)}_{P3} denote the solid products produced from processes P2 and P3, respectively.

For processes P1 and P3, the impingement volume fractions are equal and equal to the total liquid volume fraction: ϕI/3=V(ℓ)V(ℓ)+V(s)\phi_{I/3}= \frac{V^{(\ell)}}{V^{(\ell)}+V^{(s)}}. In contrast, the impingement volume fraction for P2 is given by ϕI2=VC(s)VC(s)+VXP2(s)\phi_{I2}= \frac{V_{C^{(s)}}}{V_{C^{(s)}}+V_{X^{(s)}_{P2}}}.

Here, ϕI2\phi_{I2} decreases from one to zero - similar to a single process - but with a time-dependent denominator. ϕI/3\phi_{I/3}, however, decreases monotonically from one and reaches zero only if the final state is fully crystalline, which depends on the initial amounts of 2S(ℓ)2S^{(\ell)} and X(ℓ)X^{(\ell)}.

To determine γ\gamma, we turn to a single crystallization of a pure material XX from its liquid phase X(ℓ)→X(s)X^{(\ell)}\to X^{(s)}. In this case, one can compare the crystalline volume fraction evolutions predicted by our model, Eq. (2) and Eq. (3), with those predicted by KJMA. Assuming for this constant rates η˙=η0\dot{\eta}=\eta_0 (t>0t>0), g=g0g=g_0 (t>0t>0) and rc=0r_c=0, our model is expected to recover ϕKJMA(s)(t)=1−e−Bdη˙0g0dtd+1\phi_{\mathrm{KJMA}}^{(s)}(t)=1-e^{-B_d\dot{\eta}_0g_0^d t^{d+1}}. Adjusting γ\gamma to fit our model curves to the KJMA curves, we thus choose the value γ=0.7\gamma=0.7, which we use in the following. Figure 11 shows such a fit for some choice of parameters.

Transformed crystalline volume fraction per time for a single crystallization

Figure 11. Transformed crystalline volume fraction per time for a single crystallization B(ℓ)→B(s)B^{(\ell)} \to B^{(s)} in a B(ℓ),B(s)B^{(\ell)}, B^{(s)}-blend, calculated from our crystallization cell-model and the common literature model KJMA0: ϕ(s)(t)=1−e−Bdη0dtd+1\phi^{(s)}(t)=1-e^{-B_d\eta_0^{d}t^{d+1}}. For a critical radius rc=0r_c=0, nucleation rate density η˙0=3⋅10−4\dot{\eta}_0=3\cdot10^{-4}, and growth rate g0=0.2g_0=0.2. The curves generated with our model using an impingement function f=(V(ℓ)V(ℓ)+V(s))0.7f=\left(\frac{V^{(\ell)}}{V^{(\ell)}+V^{(s)}}\right)^{0.7} closely match the curves from the common KJMA0 model.

Per crystallization, we thus have six parameters that are: The growth dimension dd, the critical radius rc [m]r_c\,[m], a growth rate and nucleation rate density scaling g0[m/s],η˙0[#/smd]g_0[m/s],\dot{\eta}_0[\#/sm^d], and two thresholds, one saturation limit ϕR∗\phi_R^* per growth rate and one criticality threshold ϕRc\phi_R^c per nucleation rate density. The latter are, however, obsolete for solid-state transitions (ϕR∗=ϕRc=0\phi_R^*=\phi_R^c=0).

Model System for Perovskite film formation

To illustrate the typically expected - but not validated-behavior of our general model, for a particular system, we consider coating of the common perovskite reference Methylammonium lead halide (MAPI) from standard solvents such as Dimethylformamide (DMF), Dimethylsulfoxide (DMSO), and N-Methyl-2-pyrrolidone (NMP). For these materials in-situ data on process mechanisms are available [14, 15, 16, 18, 49, 50, 51, 52, 53].

As reported by these studies, perovskite crystallization from a solvent-precursor solution involves several processes: solvent (S(ℓ)S^{(\ell)}) evaporation [19], solvent and precursor (AX(ℓ)AX^{(\ell)}, BX2(ℓ)BX_{2}^{(\ell)}) coordination into complexes (C(ℓ)C^{(\ell)}) eventually forming colloidal clusters [16], and a crystalline solid state precursor (D(s)D^{(s)}), as well as crystallization of perovskite (X(s)X^{(s)}), either directly or from the solid state precursor [15].

Based on this, we propose the following reference process system template for perovskite formation from solution:

S(ℓ)→S(g)nS(ℓ)+AX(ℓ)+BX2(ℓ)⇌C(ℓ)C(ℓ)→D(s)+mS(ℓ)D(s)→ABX3(s)+(n−m)S(ℓ)AX(ℓ)+BX2(ℓ)→ABX3(s)(5)\begin{aligned} S^{(\ell)} &\to S^{(g)} \\ nS^{(\ell)} + AX^{(\ell)} + BX_{2}^{(\ell)} &\rightleftharpoons C^{(\ell)} \\ C^{(\ell)} &\to D^{(s)} + mS^{(\ell)} \\ D^{(s)} &\to ABX_{3}^{(s)} + (n-m)S^{(\ell)} \\ AX^{(\ell)} + BX_{2}^{(\ell)} &\to ABX_{3}^{(s)} \tag*{(5)} \end{aligned}

Although the representation is schematic, we use effective species by summarizing the complete solution chemistry in terms of two chemical reactions. We mostly ignore stoichiometry. This makes the process system generic, enabling it to serve as a starting point for analyzing a wide range of material systems.

Figure 2 shows the corresponding time-resolved composition of a film for an illustrative set of parameters: Dissolved precursor (yellow) reacts with the solvent (blue), building up complexes (purple). Solvent evaporation reduces the total film height (black) and increases the complex concentration. As the complex concentration rises, perovskite (dark red) or solid-state precursor (orange) starts crystallizing. Once an intermediate phase is present, it recrystallizes into perovskite, thereby releasing the solvent. Finally, when the solvent has evaporated and all intermediate phase has recrystallized, a pure perovskite film remains.

Figure 2. Composition of a drying perovskite ink depending on processing time: VV denotes the volumes of the species and V0V_{0} the total initial volume. Liquid solvent (blue) evaporates reducing the total film volume (black). Solvent coordinates with liquid precursor (yellow) building up complexes (purple). The increasingly concentrated complexes accumulate and reorganize into crystalline perovskite (dark red) or solid-state precursor (orange) clusters. Subsequently, the intermediate phase recrystallizes into the perovskite phase, releasing a structurally integrated solvent.

In our modeling framework, the system Eq. (5), with two chemical reactions, one evaporation, and three crystallizations, involves at least 2+1+(6+4+6)=192 + 1 + (6 + 4 + 6) = 19 free parameters, making parameter studies challenging. Therefore, to understand the basic system behavior, we

will focus on the successive crystallization subsystem:

S(ℓ)→S(g)X(ℓ)+2S(ℓ)→C(s)C(s)→X(s)+2S(ℓ)(6)\begin{aligned} S^{(\ell)} &\to S^{(g)} \\ X^{(\ell)} + 2S^{(\ell)} &\to C^{(s)} \\ C^{(s)} &\to X^{(s)} + 2S^{(\ell)} \tag*{(6)} \end{aligned}

Here, we adapted the stoichiometry to the DMF/MAPI material system Table I. S(ℓ)S^{(\ell)} corresponds to DMF, X(ℓ)X^{(\ell)} to PbI2_{2} + MAI, C(ℓ)C^{(\ell)} to MAPI-2(DMF), and X(s)X^{(s)} to MAPI. Eq. (6) can be viewed as the crystallization sector of our template Eq. (4), without direct perovskite crystallization, provided that the chemical equilibrium is balanced on the timescale of evaporation and crystallization.

Model Behaviour

We turn to studying Eq. (6). Our strategy is to build understanding from interpretable subsystems, using simulated curves to explore which subsystems — and which curve features — are informative about the process system. The extracted physical interpretations do not depend strongly on the precise parameter values but only on the parameter regimes. For instance, to vary crystallization speeds, we vary the growth rates gg rather than the nucleation rate density η˙\dot{\eta}. This choice does not affect the interpretations, since increasing either rate leads to more crystal volume appearing at earlier times. The simulation results presented are, for concreteness, based on material parameters for Dimethylformamid (DMF) as the solvent and Methylammonium lead halide (MAPI) as the perovskite. For more detailed information on the parameters, see Appendix § 1.4 B.

Preliminary study: direct crystallization of a pure material

As we use a non-standard crystallization model, we first consider a single crystallization of a material XX:

X(ℓ)→X(s)X^{(\ell)} \rightarrow X^{(s)}

Furthermore, we assume that initially all material XX is in the liquid state (ℓ)(\ell).

Consider Figure 3. In panels a) and b), the contributions from nucleated crystalline volume (the first part of the LHS of Eq. (3)) are negligible (rc≪rˉ∞r_c \ll\bar{r}_{\infty}). The standard sigmoidal shape of the crystallization curves shifts to later times as the growth-rate scaling gg decreases. Panel b) demonstrates that the parameter space is degenerate: For each of the three growth dimensions d=1,2,3d=1,2,3 (green, orange black) two distinct combinations of growth and nucleation rate density scalings gg, η˙\dot{\eta} yield identical curves. That the factor αG=gdη˙\alpha_G = g^d\dot{\eta} controls the crystallization timescale rather aligns with expectations based on KJMA theory. In contrast, panels c) and d) explore cases where a non-negligible fraction of the crystalline volume has nucleated. In panel c), for d=3d=3 and finite rc=10−8 mr_c=10^{-8}\,\mathrm{m} we adjust the relative importance of gg and η˙\dot{\eta} while keeping gdη˙g^d\dot{\eta} constant. For very large nucleation rates, the volume change due to nucleation becomes dominant, and the crystallization curve is no longer controlled by αG\alpha_G. While the curves overlap at higher growth rates for a constant αG\alpha_G (as seen in black and red), as nucleation increasingly dominates (from orange to blue), the curves change shape drastically. This transition between nucleation- and growth-dominated regimes is also reflected in the Avrami plot in panel d), where the slope is often associated with the growth dimension. The slope, however, cannot be interpreted as the growth dimension in general; rather, it ranges from 1 to 3, corresponding to the exponents of the nucleation- and growth-dominated regimes, respectively. For the ensuing investigations, we are primarily interested in the features of the curves related to the process system. Therefore, without loss of generality, we restrict variations in crystallization speeds to the standard growth-dominated regime and vary either η˙\dot{\eta} or g0g_0.

Four-panel crystallization plots

Figure 3. Crystallization of a pure material: Relative time-resolved mole numbers of X(s)X^{(s)} for a) different growth rates gg and b) at constant scaling gdη˙g^{d}\dot{\eta} with growth rate gg raised to the growth dimension d=1,2,3d=1,2,3, and nucleation rate density η˙\dot{\eta}. c) and d) Variation of relative importance of nucleation against growth gdη˙=constg^{d}\dot{\eta}=\mathrm{const} for d=3d=3 c), relative time-resolved mole numbers of X(s)X^{(s)}, d) a monotonic transformed version of c) and the volume fraction called Avrami plot. The crystallization curves in a) shift to later times, from black to blue, as the growth rate gg increases. Their sigmoid shape is typically expected. b) Several combinations of distinct nucleation rate density η˙\dot{\eta} and growth rate scalings gdg^{d} raised by the growth dimension dd yield equivalent crystallization curves. This depends on the balance between nucleation and crystallization, as illustrated in panels c) and d). As the importance of η˙\dot{\eta} relative to g3g^{3} for g3η˙=constg^{3}\dot{\eta}=\mathrm{const} increases (black to blue), the Avrami Exponent d) decreases. The degeneracy of the parameter space demonstrated in b) is reflected in the concentration of curves in c) when the growth-rate scaling is dominant.

Solvent-integrating crystallization

Now consider the primary crystallization step as in Eq. (6):

X(ℓ)+2S(ℓ)→C(s)X^{(\ell)} + 2 S^{(\ell)} \rightarrow C^{(s)}

Here, when a sufficient amount of precursor X(ℓ)X^{(\ell)} is dissolved in the solvent S(ℓ)S^{(\ell)}, the precursor crystallizes into

a distinct solid material C(s)C^{(s)} that structurally incorporates solvent. In contrast to the previously considered direct crystallization (Section IV A), the transformation now occurs in a precursor-solvent environment. Furthermore, it depends on the relative amounts of the two reactant species X(ℓ)X^{(\ell)} and 2S(ℓ)2 S^{(\ell)}. To appreciate the qualitative behavior difference, we consider variations of the initial precursor volume fraction ϕ0,X(ℓ)\phi_{0,X^{(\ell)}}.

In Figure 4 panel a), the transformed precursor mole numbers are plotted against time. The LaMer graph in panel b) characterizes the instantaneous state of the solution by the time-dependent precursor volume fraction in the liquid phase. The horizontal lines therein correspond to the implied material-specific solubility and criticality limits given by the condition σN/G=kϕB(1−ϕB)2−ϕRc/∗=0\sigma_{N/G}=k\phi_B(1-\phi_B)^2-\phi_R^{c/*}=0. The two resulting limits correspond to the liquid precursor in the solvent ϕ∗<ϕeq\phi^*<\phi^{\mathrm{eq}} and the solvent in the liquid precursor ϕeq<ϕ∗\phi^{\mathrm{eq}}<\phi^*. At the equimolar volume fraction ϕeq=1⋅νX(ℓ)1⋅νX(ℓ)+2⋅νS(ℓ)\phi^{\mathrm{eq}}=\frac{1\cdot\nu_{X^{(\ell)}}}{1\cdot\nu_{X^{(\ell)}}+2\cdot\nu_{S^{(\ell)}}}, the same number of reactant molecules are available relative to the process: nS(ℓ)=2nX(ℓ)n_{S^{(\ell)}}=2n_{X^{(\ell)}}. If the initial precursor liquid volume fraction ϕ0,X(ℓ)\phi_{0,X^{(\ell)}} lies outside the range set by the criticality limit ϕc\phi^c and the equimolar volume fraction ϕeq\phi^{\mathrm{eq}} — either below it (black, ϕ0,X(ℓ)<ϕc<ϕeq\phi_{0,X^{(\ell)}}<\phi^c<\phi^{\mathrm{eq}}), or above it (blue, ϕeq<ϕc<ϕ0,X(ℓ)\phi^{\mathrm{eq}}<\phi^c<\phi_{0,X^{(\ell)}}) crystallization does not occur. This happens since in both cases one of the two reactants, 2S(ℓ)2S^{(\ell)} and X(ℓ)X^{(\ell)}, becomes too sparse to trigger nucleation. Otherwise, if the initial precursor volume fraction ϕ0,X(ℓ)\phi_{0,X^{(\ell)}} lies within the critical region (between the dotted lines) - ϕc<ϕ0,X(ℓ)<ϕeq\phi^c<\phi_{0,X^{(\ell)}}<\phi^{\mathrm{eq}} (orange and red) or ϕeq<ϕ0,X(ℓ)<ϕc\phi^{\mathrm{eq}}<\phi_{0,X^{(\ell)}}<\phi^c (yellow and green) - crystallization occurs. It then proceeds in the direction of the limiting reactant until either the saturation limit above or below the equimolar volume fraction is reached. For instance, if there is less liquid precursor X(ℓ)X^{(\ell)} than solvent 2S(ℓ)2S^{(\ell)} (ϕc<ϕ0,X(ℓ)<ϕeq\phi^c<\phi_{0,X^{(\ell)}}<\phi^{\mathrm{eq}}), the crystallization is limited by the available precursor. The precursor liquid volume fraction approaches the solubility limit of the precursor in the solvent, ϕ∗<ϕeq\phi^*<\phi^{\mathrm{eq}}. For a series of initializations approaching the equimolar reactant initialization (within the critical region) ϕeq\phi^{\mathrm{eq}} form either below (red, orange) or above (green, yellow), therefore the final amount of crystalline material C(s)C^{(s)} increases. The process kinetics, in contrast, depends on the likelihood ϕS(ℓ)2ϕX(ℓ)\phi_{S^{(\ell)}}^2\phi_{X^{(\ell)}} of finding the three-molecule combination 2S(ℓ)+X(ℓ)2S^{(\ell)}+X^{(\ell)}, hence the process speed therefore is not necessarily maximal at the equimolar reactant initialization ϕeq\phi^{\mathrm{eq}}.

Solvent integrating crystallization for different initial precursor volume fractions

Figure 4. Solvent integrating crystallization for different initial precursor volume fractions ϕ0,X(ℓ)\phi_{0,X^{(\ell)}}. a) Relative mole number of intermediate phase (C(s)C^{(s)}) molecules against time. b) Precursor volume fraction per time (LaMer Curve) with solubility ϕ∗\phi^{*} and criticality ϕc\phi^{c} thresholds for precursor (X(ℓ)X^{(\ell)}) and solvent (S(ℓ)S^{(\ell)}), and equimolar reactant n0,X(ℓ)=2n0,S(ℓ)n_{0,X^{(\ell)}} = 2n_{0,S^{(\ell)}} volume fraction ϕeq\phi^{\mathrm{eq}}. The process starts only if the initial precursor concentration ϕ0,X(ℓ)\phi_{0,X^{(\ell)}} lies in the critical region between ϕc<ϕeq\phi^{c}<\phi^{\mathrm{eq}} and ϕc>ϕeq\phi^{c}>\phi^{\mathrm{eq}} (between the dotted lines). If the process starts, it proceeds until its supply is exhausted, which occurs when a saturation limit ϕ∗\phi^{*} is reached. For instance, if precursor is sparse compared to solvent ϕ0,X(ℓ)<ϕeq\phi_{0,X^{(\ell)}}<\phi^{\mathrm{eq}} the precursor X(ℓ)X^{(\ell)} will be exhausted first and therefore the lower solubility limit ϕ∗<ϕeq\phi^{*}<\phi^{\mathrm{eq}} is approached. The final amount of intermediate phase C(s)C^{(s)} formed therefore is higher for initial precursor volume fractions ϕ0,X(ℓ)\phi_{0,X^{(\ell)}} that are closer to the equimolar volume fraction ϕeq\phi^{\mathrm{eq}} (comparing only volume fractions on the same side, e.g., either below or above ϕeq\phi^{\mathrm{eq}}).

Successive Crystallization of a pure material

As a first combination of two crystallizations, consider the two successive steps:

X(ℓ)→C(s)X^{(\ell)} \rightarrow C^{(s)}
C(s)→X(s)C^{(s)} \rightarrow X^{(s)}

{#eq:P\gamma}

Here, initially, liquid XX crystallizes into solid CC, which recrystallizes into solid XX. Thereby, the progress of the first process referred to as Pγ depends on the supply of solid CC provided by the process Pβ. Therefore, we will investigate the system’s behavior with respect to the relative speeds of the two processes. Without loss of generality, we will vary the speed of Pγ relative to the speed of Pβ, varying Pγ’s growth rate. Initially, there is an unlimited consumption of C(s)C^{(s)} by Pγ, where it proceeds independently of Pβ. However, as Pγ becomes process-limited by Pβ, the production of X(s)X^{(s)} converges to the up-slope of the C(s)C^{(s)} rise.

In Figure 5, the X(ℓ)X^{(\ell)} curve in a) remains unchanged upon variations of Pγ’s speed. If Pγ is so slow that the production of C(s)C^{(s)} completes before the production of X(s)X^{(s)} begins (black to orange), the kinetics of X(s)X^{(s)} are independent of the kinetics of C(s)C^{(s)}. The C(s)C^{(s)} curves reflect both the crystallization kinetics of C(s)C^{(s)} (the build-up phase) and the crystallization kinetics of X(s)X^{(s)}. As the speed of Pγ increases (from yellow to green), the kinetics of Pγ become partly limited by Pβ.

Successive crystallization for different growth rates of the secondary crystallization step

Figure 5. Successive crystallization for different growth rates of the secondary crystallization step. Mole numbers per material relative to the total initial number of moles over time. a) The single collapsed X(ℓ)X^{(\ell)} curve (blue) corresponds to the up-flank of the C(s)C^{(s)} curves in b). c) As the speed of the X(s)X^{(s)} crystallization increases with higher growth rates gg (black to blue), the maximal cumulative amount of C(s)C^{(s)} decreases. This, in the extreme, leads to a process-limited secondary X(s)X^{(s)} crystallization step (yellow to blue).

In extreme cases, for the green C(s)C^{(s)} curve, this can lead to an extended tail after the peak: until Pγ becomes limited by Pβ, the consumption of C(s)C^{(s)} by Pγ outpaces the production of C(s)C^{(s)} by Pβ. Once Pγ becomes limited by Pβ, its consumption matches Pβ’s production, resulting in a plateau that eventually ends when Pγ stops.

Finally, if Pγ is significantly faster than Pβ (as shown by the blue curves), the global transformation process is completely limited by Pβ. In this scenario, the production of X(s)X^{(s)} is entirely controlled by the production of C(s)C^{(s)}. The production of C(s)C^{(s)} becomes nearly unobservable, and the observed production of X(s)X^{(s)} ultimately reflects the kinetics of C(s)C^{(s)} formation.

More importantly, if the kinetics of X(s)X^{(s)} production are entirely dictated by the crystallization of C(s)C^{(s)}, the absolute height of the C(s)C^{(s)} peak will be very low. Therefore, in the case of perovskite formation, the absence of a signal (for example, from in-situ GIWAXS) corresponding to an intermediate phase does not provide sufficient evidence to conclude that formation occurs through direct crystallization.

Successive solvent-coupled crystallization

As compared to the successive crystallization Section IV C, the successive subsystem appearing in Eq. (6) and given by

X(ℓ)+2S(ℓ)→C(s)(7)X^{(\ell)} + 2S^{(\ell)} \rightarrow C^{(s)} \tag*{(7)}
C(s)→X(s)+2S(ℓ)(8)C^{(s)} \rightarrow X^{(s)} + 2S^{(\ell)} \tag*{(8)}

additionally involves the solvent S(ℓ)S^{(\ell)}. The release of S(ℓ)S^{(\ell)} by P3 therefore always couples the processes, and this feedback from P3 to P2 can alter P2. Therefore, as previously, we will analyze the role of the S(ℓ)S^{(\ell)} release by P3 by varying the initial precursor volume fraction.

Figure 6 shows the LaMer curve for various initial precursor volume fractions ϕ0,X(ℓ)\phi_{0,X^{(\ell)}} (black to blue). The effect of the solvent coupling on the crystallization P2 can be assessed by comparison with Figure 4. If there is an oversupply of either S(ℓ)S^{(\ell)} or X(ℓ)X^{(\ell)}, such that X(ℓ)X^{(\ell)} is fully miscible in S(ℓ)S^{(\ell)} or S(ℓ)S^{(\ell)} is fully miscible in X(ℓ)X^{(\ell)}, the process cannot proceed. In contrast to Figure 4, however, when S(ℓ)S^{(\ell)} is abundant, and X(ℓ)X^{(\ell)} is sparse, the process always proceeds to fully exhaust X(ℓ)X^{(\ell)}. This is possible since P3 resupplies S(ℓ)S^{(\ell)}. The effect of the solvent coupling, therefore, is that P3 can effectively lift any solvent supply limitation on P2.

P2, as compared to the solvent, integrating crystallization Section IV B, therefore depends only weakly on the amount of solvent. Because of this, we expect the material amount curves based on P2 and P3 to show few qualitative differences from those based on the successive system without the solvent coupling Figure 5. To test this, we also vary the speed of the secondary crystallization, P3, relative to that of the first, P2. In particular we consider a fixed initial volume ratio of S(ℓ)S^{(\ell)} to X(ℓ)X^{(\ell)} within the region where S(ℓ)S^{(\ell)} is the limiting reactant ϕc<ϕ0,X(ℓ)<ϕeq\phi^{c}<\phi_{0,X^{(\ell)}}<\phi^{\mathrm{eq}}.

Compare Figure 7 with Figure 5 based on the successive system Figure 5. Here, in the solvent-coupled successive crystallizations (P2 and P3), even if the secondary step (P3) is slower than the first (P2), the steps do not decouple. This occurs because, initially, there is less solvent reactant 2S(ℓ)2S^{(\ell)} than there is of the precursor reactant X(ℓ)X^{(\ell)}. Consequently, P2 eventually requires P3 to resupply the solvent to proceed (yellow process).

In the extreme, when P3 is much slower than P2 (black C(s)C^{(s)} curve), P2 can become fully P3-limited, yielding a flat C(s)C^{(s)} peak. Two other differences are evident when comparing the amounts of the final product X(s)X^{(s)}. First, the X(ℓ)X^{(\ell)} conversion is incomplete as the X(ℓ)X^{(\ell)} curve does not reach zero. This is the expected influence of the finite solubility threshold of the precursor in the solvent, which dictates that the precursor can only be consumed above ϕ∗<ϕeq\phi^{*}<\phi^{\mathrm{eq}}. Second, while the amounts of final product X(s)X^{(s)} are constant for the successive system (Figure 5), they can vary for the solvent-coupled successive system (Figure 7). One might again suspect the influence of P2’s solubility threshold. However, according to Figure 4, for a single crystallization from solution that starts at constant initial volume fraction, the amount of final product should not change with the speed of the process. Instead, the differences in the amount of final product arise not from varying initial amounts of solvent, but from differences in the final amount of solvent. If P3 is sufficiently slow (black to orange X(ℓ)X^{(\ell)} and X(s)X^{(s)} curves), P3 will continue after P2 has stopped because P2’s precursor supply is exhausted. As P3 continues to resupply solvent S(ℓ)S^{(\ell)}, the final solution will therefore contain less precursor X(ℓ)X^{(\ell)} than dictated by the solubility limit. Conversely, if P3 is sufficiently fast (yellow to purple), the time lag between the stopping of P2 and P3 becomes small. Consequently, both processes effectively come to a halt when the solubility limit of the precursor in the solvent is reached. The final amount of the perovskite X(s)X^{(s)}, therefore, again depends only on the initial amounts of the precursor and solvent.

Successive crystallization with solvent coupling

Figure 7. Successive crystallization with solvent coupling. Variation of the growth rates gg of the X(s)X^{(s)} crystallization with fixed solvent-limited C(s)C^{(s)} crystallization. Relative material-resolved mole numbers over time. a,b,c) Precursor X(ℓ)X^{(\ell)} and solvent S(ℓ)S^{(\ell)} are consumed to build up intermediate phase C(s)C^{(s)} crystals, which then recrystallize into the perovskite X(s)X^{(s)}. d) For larger growth rate scalings gg, X(s)X^{(s)} crystallization becomes faster than C(s)C^{(s)} crystallization and is limited by it (yellow to purple). For smaller growth rate scalings gg (orange to black), X(s)X^{(s)} crystallization from C(s)C^{(s)} is slower than C(s)C^{(s)} crystallization. As initially there are fewer S(ℓ)S^{(\ell)} than X(ℓ)X^{(\ell)} molecules available for integration into C(s)C^{(s)}, the crystallization of C(s)C^{(s)} is limited in its final stages by the release of solvent from C(s)C^{(s)} due to the crystallization of X(s)X^{(s)}. For the black C(s)C^{(s)} curve, this leads to an extended plateau at its maximum height.

Evaporation-induced, solvent-integrating crystallization

In real-life processing of perovskite films for photovoltaics, crystallization is induced by solvent evaporation. Within our modeling framework, the description of the intermediate phase crystallization thus requires an evaporation process coupled to a single concentration-dependent crystallization:

S(ℓ)→S(g)S^{(\ell)} \to S^{(g)}
{#eq:P1}

X^{(\ell)} + 2S^{(\ell)} \to C^{(s)}

Equation not extracted
(9)

In contrast to the concentration-dependent crystallization discussed in the absence of evaporation (Section IV B), here the precursor concentration also changes due to the solvent evaporation P1. Evaporation increases the precursor concentration over time. Hence we expect the crystallization P2 to occur eventually—even starting from sub-critical initial concentrations ϕ0,X(ℓ)<ϕc∗<ϕeq\phi_{0,X^{(\ell)}} < \phi^*_{c} < \phi^{\mathrm{eq}}. Once crystallization begins, its progress depends on the relative rates of evaporation and crystallization. We therefore fix the evaporation-rate parameters while varying the growth rate of the crystallization process, P2.

Successive crystallization with solvent coupling for different initial precursor volume fractions

Figure 6. Successive crystallization with solvent coupling for different initial precursor volume fractions ϕ0,X(ℓ)\phi_{0,X^{(\ell)}}. Liquid precursor volume fraction per time. Crystallization happens only if ϕ0,X(ℓ)\phi_{0,X^{(\ell)}} lies in the critical region between the dotted curves defined by the criticality limits ϕc\phi^{\mathrm{c}} for the liquid precursor in the solvent and solvent in the liquid precursor, respectively (for green to red but not for black and blue). Crystallization ends only when the precursor saturation limit ϕ∗<ϕeq\phi^{*}<\phi^{\mathrm{eq}} is reached (lower dotted line) and proceeds towards this limit even starting from initial precursor volume fractions close to the solvent in precursor solubility limit (green).

Consider the LaMer graphs in Figure 8 panel b). There are two asymptotes corresponding to two final states that we identify from the X(ℓ)X^{(\ell)} and C(s)C^{(s)} curves: As S(ℓ)S^{(\ell)} evaporates, first, the precursor concentration becomes larger than ϕc<ϕeq\phi^{\mathrm{c}}<\phi^{\mathrm{eq}} and triggers nucleation. Afterward, if the crystallization is fast (blue to yellow), C(s)C^{(s)} buildup proceeds (see C(s)C^{(s)} curve) until the precursor volume fraction in the liquid reaches the saturation limit ϕ∗<ϕeq\phi^{*}<\phi^{\mathrm{eq}}. Subsequently, since the crystallization P2 lowers supersaturation faster than evaporation builds it up, P2 is limited by the rate at which evaporation builds supersaturation. This results in a single maximum in the LaMer curve, followed by a constant regime at the solubility limit. Otherwise, if crystallization is slow (orange to black), the solvent fully evaporates before the precursor is exhausted, leaving liquid residues of the precursor.

Figure 8. Evaporation-induced crystallization for various crystal growth rates gg. a,b,c) Relative material-resolved molecule numbers over time. d) Precursor volume fraction in liquid phase per time. Black horizontal lines indicate characteristic thresholds for solubility ϕ∗\phi^*, criticality ϕc\phi^c and equimolar reactants ϕeq\phi^{\mathrm{eq}}. a,b,c) S(ℓ)S^{(\ell)} evaporation builds supersaturation, inducing C(s)C^{(s)} crystallization that exhausts the precursor. c) For large growth-rate scalings gg (yellow to blue), crystallization occurs much faster than evaporation. The C(s)C^{(s)} crystallization exhausts the supersaturation and then becomes kinetically limited by the evaporation. Otherwise, for small growth rates (orange to black), S(ℓ)S^{(\ell)} is completely removed before all X(ℓ)X^{(\ell)} is integrated into the C(s)C^{(s)}, leaving precursor residues (2). b) The LaMer curve correspondingly shows two distinct behaviors (blue to yellow) and (orange to black).

Both regimes are also visible in the precursor X(ℓ)X^{(\ell)} and the intermediate phase C(s)C^{(s)} curves. Consider, for instance, C(s)C^{(s)}: If the crystallization P2 is fast (blue to yellow) after a steep rise corresponding to the free P2 kinetics, a sudden slowdown becomes visible towards the end. This secondary slope in the C(s)C^{(s)} curve is entirely determined by evaporation; we therefore say that P2 becomes evaporation-limited. Despite P2’s evaporation-limited kinetics, however, in this regime, crystallization proceeds until all precursor X(ℓ)X^{(\ell)} available to the process (up to the solubility limit) is finally consumed. In contrast, when crystallization (P2) is significantly slower than evaporation (P1) (orange to black), the solvent can evaporate before the precursor is fully converted to the intermediate phase; therefore, crystallization stops as the solvent is exhausted.

In summary, slow evaporation relative to crystallization leads to the complete conversion of X(ℓ)X^{(\ell)} into C(s)C^{(s)} through two subsequent kinetic regimes. In contrast, fast evaporation leads to partial conversion through a single kinetic regime.

Evaporation-induced successive solvent-coupled crystallization

Finally, we add the secondary intermediate phase-to-perovskite crystallization step to the evaporation-induced intermediate phase crystallization system:

S(ℓ)→S(g)S^{(\ell)} \rightarrow S^{(g)}
X(ℓ)+2S(ℓ)→C(s)X^{(\ell)} + 2S^{(\ell)} \rightarrow C^{(s)}
C(s)→X(s)+2S(ℓ)C^{(s)} \rightarrow X^{(s)} + 2S^{(\ell)}

As before, we restrict the X(ℓ)X^{(\ell)} volume fraction to below the criticality limit, which ensures that crystallization of C(s)C^{(s)} (P2) is evaporation-induced. From our analysis of evaporation-induced S(ℓ)S^{(\ell)}-integrating crystallization Section IV E, we expect P2 kinetics to fall into two distinct regimes, depending on its speed relative to the evaporation (P1). If the evaporation is slow compared with the crystallization of C(s)C^{(s)} (“P1 < P2”), P2 kinetics become evaporation-limited. If evaporation is fast (“P1 > P2”), P2 kinetics become solvent-limited. In addition, the relative speed of the secondary crystallization of X(s)X^{(s)} (P3) compared with P1 and P2 may also matter. In principle, each of the six possible speed orderings of P1, P2, and P3 (e.g., “P1 < P2 < P3”, “P2 < P1 < P3”, ...) could yield distinct behavior. To scan these, we first fix both the evaporation P1 and the C(s)C^{(s)} crystallization P2, and then consider the effects of accelerating P3 by increasing its growth rate.

P2 evaporation-limited

The C(s)C^{(s)} curves in panel a) of Figure 9 show the kinked rise characteristic of evaporation-limited kinetics. This occurs when C(s)C^{(s)} crystallization is faster than evaporation, as in the yellow-to-blue curves of Figure 8 panel (a). Likewise, and for the same reason, the conversion of X(ℓ)X^{(\ell)} to C(s)C^{(s)} is complete in all cases here. The collapsed LaMer curves in panel b) (blue to black) essentially correspond to the single LaMer curve for evaporation-induced-limited crystallization in Figure 8 panel b) (blue and green). Therefore, the influence of S(ℓ)S^{(\ell)} recycling on the global kinetics is negligible regardless of the relative speed of P3: whether P3 is slower than P2 (black to red), of the same order as P2 (orange), or faster than P2 (yellow to blue). This might also be expected since S(ℓ)S^{(\ell)} is always present in excess. Still, the global C(s)C^{(s)} evolution and X(s)X^{(s)} formation depend on the kinetic balance between P2 and P3. This manifests as a transition between ”kinetics only weakly coupled by S(ℓ)S^{(\ell)} recycling” and ”P3 fully P2-limited”.

Four-panel time-dependent molecule-number plots and precursor volume-fraction plot for coupled crystallization

Figure 9. Evaporation-induced successive solvent coupled crystallization for various growth rates gg of the secondary X(s)X^{(s)} crystallization, starting from an evaporation-induced primary C(s)C^{(s)} crystallization that is faster than the evaporation. a,b,c) Relative material resolved time-dependent molecule numbers. b) Precursor volume fractions in the liquid phase. a,b,c) Solvent evaporation induces the C(s)C^{(s)} crystallization, which exhausts the precursor X(ℓ)X^{(\ell)}. c) The C(s)C^{(s)} crystallization is limited by the evaporation, as it drains supersaturation faster than it is built up by the evaporation (blue to black). d) Simultaneously, X(s)X^{(s)} crystallizes from C(s)C^{(s)}, and this process can either be limited by the production of C(s)C^{(s)} (yellow to blue) or proceed independently (black to orange). In all cases, within the significant crystallization region, the LaMer curves in panel b) collapse to a single curve. Initially, the supersaturation threshold ϕc<ϕeq\phi^{c} < \phi^{\mathrm{eq}} for the liquid precursor in the solvent is reached. Subsequently, it is consumed by the C(s)C^{(s)} crystallization, which continues at the saturation limit ϕ∗<ϕeq\phi^{*} < \phi^{\mathrm{eq}}.

P2 solvent-limited

Now, we fix the evaporation kinetics and keep the primary crystallization P2 slower than the evaporation, as for orange and red in Figure 8. Subsequently, we vary the speed of the secondary crystallization, P3, via the growth-rate scaling. The X(s)X^{(s)} curves in Figure 10 show three regimes.

Five-panel plot showing evaporation-induced successive solvent-coupled crystallization

Figure 10. Evaporation-induced successive solvent-coupled crystallization for various growth rates gg of the secondary X(s)X^{(s)} crystallization and starting from an initially evaporation-induced primary C(s)C^{(s)} crystallization that is slower than the evaporation. a,b,c) Relative material-resolved molecule numbers over time. e) Precursor volume fractions in the liquid phase per time. a,b,c) Evaporation of the solvent S(ℓ)S^{(\ell)} induces C(s)C^{(s)} crystallization, which drains X(ℓ)X^{(\ell)} but may not completely consume it (red to black in b). d) X(s)X^{(s)} forms from C(s)C^{(s)} releasing solvent. If the X(s)X^{(s)} crystallization is sufficiently fast (orange to purple), it lifts the solvent limitation from the C(s)C^{(s)} crystallization. If it is slow (red to black), however, it can only partially counterbalance solvent evaporation. The C(s)C^{(s)} crystallization therefore remains solvent-limited, which results in X(ℓ)X^{(\ell)} residues. The LaMer curves in panel e) show two basic behaviors in the significant crystallization regions. For the faster X(s)X^{(s)} crystallizations (purple to orange), the concentration of the supersaturated precursor decreases to the saturation limit of the precursor in the solvent, ϕ∗<ϕeq\phi^{*}<\phi^{\mathrm{eq}}, at which the C(s)C^{(s)} crystallization subsequently continues. Conversely, for slower X(s)X^{(s)} crystallizations (red to black), the solvent S(ℓ)S^{(\ell)} is exhausted before the precursor X(ℓ)X^{(\ell)}. Consequently, the liquid precursor volume-fraction curves approach 1.

If P3 is very slow (black), the recycled S(ℓ)S^{(\ell)} evaporates before it can be consumed by P2. P2’s kinetics therefore remain solvent-limited: C(s)C^{(s)} crystallization stops early, and X(ℓ)X^{(\ell)} residues may remain. As a result, X(ℓ)X^{(\ell)} to X(s)X^{(s)} conversion is incomplete.

If P3 is slightly faster (red), the solvent limitation on P2 can be partly lifted, resulting in distinct crystallization phases. The LaMer curve in panel b) crosses ϕ∗>ϕeq\phi^{*} > \phi^{\mathrm{eq}} three times: crystal growth pauses after the first crossing, then resumes between the second and third crossings. Finally, if P3 is sufficiently fast (orange to purple), solvent recycling due to P3 lifts the solvent limitation on P2. Consequently, X(ℓ)X^{(\ell)} is fully converted to C(s)C^{(s)} and the X(ℓ)X^{(\ell)}-to-X(s)X^{(s)} conversion is complete. Within this regime, for P3 even faster (purple), P3 is P2-limited, and the X(s)X^{(s)} curve follows the C(s)C^{(s)} production—as also visible in the corresponding X(ℓ)X^{(\ell)} curve.

Conclusions

In summary, we developed a general zero-dimensional population balance framework coupling crystallization, evaporation, and chemical reactions. For perovskite films formed from solution, we proposed an effective reaction template informed by multichannel in situ studies of MAPI in common solvents, intended as a reasonable mechanistic prior for these systems. Within that framework, we used parameter sweeps to investigate the evaporation-driven two-step (successive) crystallization subsystem, building up from subsystems with fewer coupled processes.

The resulting quantitative evaporation, phase-resolved crystallization, and LaMer curves yield several findings and their consequences:

  1. Allowing finite-size nuclei, crystallization admits a nucleation-dominated regime alongside the classical growth-dominated one, lowering the Avrami exponent without changing the growth dimension. The Avrami exponent is therefore ambiguous as a growth dimension.

  1. When a secondary step is kinetically limited by the primary one, the intermediate phase can be practically invisible, although the mechanism is indirect. The absence of an intermediate signal is insufficient evidence for direct formation.

  1. With limited solvent, the primary and secondary solvent-coupled steps are mutually coupled regardless of their relative rates. In the evaporation-driven case, intermediate crystallization is therefore either incomplete, from solvent depletion when evaporation is fast, or rate-limited by evaporation itself rather than by its own kinetics when it is slow; perovskite can also form at the evaporation rate with no intermediate detectable at all. Evaporation therefore shapes both phase-specific crystallization curves — directly the intermediate’s, and, through it, the perovskite’s — so neither curve identifies the mechanism on its own, and evaporation must be measured with them.

These findings concern the subsystem studied; because they depend mainly on relative timescales, they should hold irrespective of the rate parameters, and hence of material system and processing conditions. Varying evaporation rates rather than crystallization rates, for instance, should give equivalent results [54].

Where diffusion limitations or preferential crystallization at the substrate or film surface dominate, spatially resolved modeling will be required. Because the template retains the direct and successive channels in parallel, their competition can be examined without further model construction, and because mechanistic specificity enters only through the rate parameters, the same framework applies to other materials and to the effects of temperature and precursor composition. Ultimately, these mechanistic signatures offer a route to distinguishing crystallization pathways from the features of measurable drying and phase-resolved composition curves.

Acknowledgements

We acknowledge the financial support from the Helmholtz Association through the Solar Technology Acceleration Platform for emerging Photovoltaics (Solar-TAP) and the Deutsche Forschungsgemeinschaft (DFG) via the Collaborative Research Center (‘ChemPrint’, Project No. 538767711, CRC 1719).

Appendix

Model Description

Model Behaviour

a. Simulation Setup Each run is specified by the material properties Table I, its initial and boundary conditions, the temperature TT, Flory parameter χ\chi and evaporation prefactor α\alpha (Table II), and the parameters of each crystallization, C1C_1 and, where present,

DMFPbI2_2MAIMAPI
chemical formulaC3_3H7_7NOPbI2_2CH3_3NH4_4I[CH3_3NH3_3]+^+Pb+^+I3−_3^-
density ρ\rho [kg/m3\mathrm{kg}/\mathrm{m}^3]944616022804118
molar mass MM [kg/mol\mathrm{kg}/\mathrm{mol}]0.07310.46100.15900.6200
vapor pressure P∗P^* [Pa\mathrm{Pa}]490.9

caption: TABLE I. Material characteristics at T0=298T_0 = 298 [K\mathrm{K}], P0=105P_0 = 10^5 [Pa\mathrm{Pa}].

:::

Fig.Processesϕ0\phi_0TT [K\mathrm{K}]χ\chiα\alpha
3C1C_1(1, 0, 0, 0)(1,\ 0,\ 0,\ 0)
4C1C_1(sw, sw, 0, 0)(sw,\ sw,\ 0,\ 0)
5C1,C2C_1,C_2(1, 0, 0, 0)(1,\ 0,\ 0,\ 0)
6C1,C2C_1,C_2(sw, sw, 0, 0)(sw,\ sw,\ 0,\ 0)
7C1,C2C_1,C_2(0.25, 0.75, 0, 0)(0.25,\ 0.75,\ 0,\ 0)
8E1,C1E_1,C_1(0.068, 0.932, 0, 0)(0.068,\ 0.932,\ 0,\ 0)351010−5.357810^{-5.3578}
9E1,C1,C2E_1,C_1,C_2(0.068, 0.932, 0, 0)(0.068,\ 0.932,\ 0,\ 0)352010−5.357810^{-5.3578}
10E1,C1,C2E_1,C_1,C_2(0.068, 0.932, 0, 0)(0.068,\ 0.932,\ 0,\ 0)352010−5.357810^{-5.3578}

Table 2. TABLE II. Initial and boundary conditions. ϕ0\phi_0 is the initial composition with solute, solvent, solid-state intermediate, perovskite components; χ\chi are the pairwise Flory–Huggins interaction parameters and α\alpha is the prefactor appearing in the rate expression Eq. (1) for the evaporation E1E_1, absent where no evaporation is active.

Fig.Cryst.ϕRi\phi_{R_i}η˙0′\dot{\eta}'_0 [#/m3^3]g0′g'_0 [m/s]ddrcr_c [m]ϕR∗\phi_R^*ϕRc\phi_R^c
One crystallisation
3 a)C1C_1ϕ1\phi_1101810^{18}swsw310−1010^{-10}00
3 b)C1C_1ϕ1\phi_1swswswswswsw10−1010^{-10}00
3 c,d)C1C_1ϕ1\phi_1swswswsw310−1010^{-10}00
4C1C_12.25ϕ1ϕ22.25\phi_1\phi_2101310^{13}10−610^{-6}310−810^{-8}0.10.25
Two crystallisations
5C1C_12.25ϕ1ϕ22.25\phi_1\phi_210410^410−810^{-8}110−810^{-8}00
5C2C_2ϕ3\phi_3101010^{10}swsw310−810^{-8}00
6C1C_12.25ϕ1ϕ22.25\phi_1\phi_210510^510−610^{-6}110−810^{-8}0.050.15
6C2C_2ϕ3\phi_3101210^{12}3⋅10−63\cdot10^{-6}310−810^{-8}0.10.25
7C1C_12.25ϕ1ϕ22.25\phi_1\phi_210410^410−610^{-6}110−810^{-8}0.10.25
7C2C_2ϕ3\phi_3101210^{12}swsw310−810^{-8}0.10.25
Evaporation active (E1E_1, Table II)
8C1C_12.25ϕ1ϕ22.25\phi_1\phi_210510^5swsw310−810^{-8}0.050.15
9C1C_12.25ϕ1ϕ22.25\phi_1\phi_2101110^{11}5⋅10−55\cdot10^{-5}310−810^{-8}0.050.15
9C2C_2ϕ3\phi_3101110^{11}swsw310−810^{-8}0.050.15
10C1C_12.25ϕ1ϕ22.25\phi_1\phi_2101110^{11}4⋅10−64\cdot10^{-6}310−810^{-8}0.050.15
10C2C_2ϕ3\phi_3101110^{11}swsw310−810^{-8}0.050.15

Table 3. TABLE III. Crystallisation parameters η˙0′=η˙0⋅(1−ϕRc)\dot{\eta}'_0=\dot{\eta}_0\cdot(1-\phi_R^c), g0′=g0⋅(1−ϕR∗)g'_0=g_0\cdot(1-\phi_R^*) are the rescaled nucleation rate density and growth rate scalings, dd is growth dimension, rcr_c the critical radius and ϕR∗,ϕRc\phi_R^*,\phi_R^c the growth and nucleation thresholds. ϕRi\phi_{R_i} is the reactant volume fraction given in the third column.

C2C_2 (Table Table III) according to the rate laws:

η˙=η˙0′max⁡(ϕR−ϕRc1−ϕRc,0),\dot{\eta}=\dot{\eta}'_0\max\left(\frac{\phi_R-\phi_R^c}{1-\phi_R^c},0\right),
g=g0′max⁡(ϕR−ϕR∗1−ϕR∗,0).g=g'_0\max\left(\frac{\phi_R-\phi_R^*}{1-\phi_R^*},0\right).

For ϕRc<1\phi_R^c<1 and ϕR∗<1\phi_R^*<1, and using η˙0′=η˙0(1−ϕRc)\dot{\eta}'_0=\dot{\eta}_0(1-\phi_R^c) and g0′=g0(1−ϕR∗)g'_0=g_0(1-\phi_R^*), these are identical to Eq. (4). In the main text, the factors (1−ϕRc)−1(1-\phi_R^c)^{-1} and (1−ϕR∗)−1(1-\phi_R^*)^{-1} have been absorbed into the scalings η˙0\dot{\eta}_0 and g0g_0, respectively, for ease of presentation. The reaction volume fraction ϕRi\phi_{R_i} entering Eq. (4) is supplied by one of three closures: ϕRi=ϕ1\phi_{R_i}=\phi_1 for the pure-material runs, ϕR1=2.25ϕ1ϕ2\phi_{R_1}=2.25\phi_1\phi_2 where crystallisation C1C_1 is solvent-coupled, and ϕR2=ϕ3\phi_{R_2}=\phi_3 for C2C_2. A cell marked swsw indicates a parameter that is varied across the curves of the figure/panel if specified, with the per-curve values given in the figure legend. Quantities are dimensionless if no unit is noted.

Parameters were generally chosen so that the effects discussed in the main text are visible within a single observation window.

b. Numeric All evolution equations were solved using a first-order Euler discretization with rate-clipping safeguards to ensure positivity while preserving population number conservation. Domain constraints for physical quantities have generally been enforced explicitly. Simulation results have been cross-checked for several time resolutions. Physical consistency against the expectations based on the reaction systems has been assessed in all cases. Where possible, implementations have been validated against analytical results, e.g., Figure 11.

References

References

  1. [1]International Renewable Energy Agency, Renewable Power Generation Costs in 2021, Tech. Rep. (International Renewable Energy Agency, Abu Dhabi, 2022).
  2. [2]O. Almora, G. C. Bazan, C. I. Cabrera, et al., Device Performance of Emerging Photovoltaic Materials (Version 5), Advanced Energy Materials 15, 2404386 (2025).
  3. [3]S. Lee, N.-G. Park, and H. S. Jung, Cutting-Edge Advances in Scalable Perovskite Photovoltaic Devices, MRS Bulletin 49, 1265 (2024).DOI
  4. [4]F. Yang, D. Jang, L. Dong, et al., Upscaling Solution-Processed Perovskite Photovoltaics, Advanced Energy Materials 11, 2101973 (2021).DOI
  5. [5]S. Qiu, M. Majewski, L. Dong, et al., Over one-micron-thick void-free perovskite layers enable highly efficient and fully printed solar cells, Energy & Environmental Science 18, 5926 (2025).
  6. [6]J. Wang, W. Wang, Y. Chen, et al., Growth and Degradation Kinetics of Organic–Inorganic Hybrid Perovskite Films Determined by In Situ Grazing-Incidence X-Ray Scattering Techniques, Small Methods 5, 2100829 (2021).DOI
  7. [7]J. A. Steele, E. Solano, D. Hardy, et al., How to GI-WAXS: Grazing Incidence Wide Angle X-Ray Scattering Applied to Metal Halide Perovskite Thin Films, Advanced Energy Materials 13, 2300760 (2023).
  8. [8]B. Schmidt-Hansberg, M. Baunach, J. Krenn, et al., Spatially Resolved Drying Kinetics of Multi-Component Solution Cast Films for Organic Electronics, Chemical Engineering and Processing: Process Intensification 50, 509 (2011).
  9. [9]J. Rappich, C. Kaspari, C. Camus, and N. H. Nickel, Fast Optical Reflectance Measurements during Spin Coating and Annealing of Organic–Inorganic Perovskite Precursor Solutions, physica status solidi (b) 258, 2000479 (2021).
  10. [10]V. Da Silva, J. Just, Áureo. Freitas, et al., Multimodal Data Acquisition System at MAX IV, in Proceedings of the 20th International Conference on Accelerator and Large Experimental Physics Control Systems (ICALEPCS 2025) (2025) pp. 1279–1284.DOI
  11. [11]R. Sztosak, A. de Souza Gonçalves, J. N. de Freitas, et al., In Situ and Operando Characterizations of Metal Halide Perovskite and Solar Cells: Insights from Lab-Sized Devices to Upscaling Processes, Chemical Reviews 123, 3160 (2023).
  12. [12]S. Ternes, T. Börnhorst, J. A. Schwenzer, et al., Drying Dynamics of Solution-Processed Perovskite Thin-Film Photovoltaics: In Situ Characterization, Modeling, and Process Control, Advanced Energy Materials 9, 1901581 (2019).DOI
  13. [13]A. Merdasa, C. Rehermann, K. Hirselandt, et al., Eye in the Process: Formation of “Triple Cation” Perovskite Thin Films Rationalized by in Situ Optical Monitoring (2020), preprint.DOI
  14. [14]N. Alhazmi, E. Pineda, J. Rawle, et al., Perovskite Crystallization Dynamics during Spin-Casting: An In Situ Wide-Angle X-ray Scattering Study, ACS Applied Energy Materials 3, 6155 (2020).DOI
  15. [15]O. Shargaeva, H. Nässt rö m, J. Li, et al., Temperature-Dependent Crystallization Mechanisms of Methylammonium Lead Iodide Perovskite From Different Solvents, Frontiers in Energy Research 9 (2021).
  16. [16]S. Pratap, F. Babbe, N. S. Barchi, et al., Out-of-Equilibrium Processes in Crystallization of Organic-Inorganic Perovskites during Spin Coating, Nature Communications 12, 5624 (2021).DOI
  17. [17]C. Rehermann, V. Schröder, M. Flatken, et al., Role of Solution Concentration in Formation Kinetics of Bromide Perovskite Thin Films during Spin-Coating Monitored by Optical in Situ Metrology, RSC Advances 12, 32765 (2022).DOI
  18. [18]M. A. Reus, L. K. Reb, A. F. Weinzierl, et al., Time-Resolved Orientation and Phase Analysis of Lead Halide Perovskite Film Annealing Probed by In Situ GIWAXS, Advanced Optical Materials 10, 2102722 (2022).DOI
  19. [19]S. Ternes, J. Mohacsi, N. Lüdtke, et al., Drying and Coating of Perovskite Thin Films: How to Control the Thin Film Morphology in Scalable Dynamic Coating Systems, ACS Applied Materials & Interfaces 14, 11300 (2022).DOI
  20. [20]S. Pratap, Hybrid Perovskites : Colloidal Precursors to Thin Films, Ph.D. thesis, Technische Universität München (2021).
  21. [21]L. Schmidt-Mende, V. Dyakonov, S. Olthof, et al., Roadmap on Organic–Inorganic Hybrid Perovskite Semiconductors and Devices, APL Materials 9, 109202 (2021).
  22. [22]I. Hwang, Challenges in Controlling the Crystallization Pathways and Kinetics for Highly Reproducible Solution-Processing of Metal Halide Perovskites, The Journal of Physical Chemistry C 127, 24011 (2023).DOI
  23. [23]A. G. Ortoll-Bloch, H. C. Herbol, B. A. Sorenson, et al., Bypassing Solid-State Intermediates by Solvent Engineering the Crystallization Pathway in Hybrid Organic-Inorganic Perovskites, Crystal Growth & Design 20, 1162 (2020).
  24. [24]S. Ternes, C. J. Brabec, L. A. Castriotta, et al., Process Parameter Specification and Control in Solution Processing of Hybrid Perovskite Photovoltaics: From Domain-Specific Jargon to Evidence-Based, Unambiguous Description of Experimental Workflows, Advanced Energy Materials 15, e03187 (2025).
  25. [25]S. Ternes, In Situ Characterization and Modelling of Drying Dynamics for Scalable Printing of Hybrid Perovskite Photovoltaics, Ph.D. thesis, KIT (2022).DOI
  26. [26]K. Suchan, J. Just, P. Becker, et al., Optical in Situ Monitoring during the Synthesis of Halide Perovskite Solar Cells Reveals Formation Kinetics and Evolution of Optoelectronic Properties, Journal of Materials Chemistry A 8, 10439 (2020).DOI
  27. [27]N. Mrkyvkova, V. Held, P. Nádasdy, et al., Combined in Situ Photoluminescence and X-ray Scattering Reveals Defect Formation in Lead-Halide Perovskite Films, The Journal of Physical Chemistry Letters 12, 10156 (2021).DOI
  28. [28]C. J. Taylor, M. Booth, J. A. Manson, et al., Rapid, automated determination of reaction models and kinetic parameters, Chemical Engineering Journal 413, 127017 (2021).DOI
  29. [29]J. B. Rawlings, S. M. Miller, and W. R. Witkowski, Model Identification and Control of Solution Crystallization Processes: A Review, Industrial & Engineering Chemistry Research 32, 1275 (1993).DOI
  30. [30]M. T. Verki, R. H. Haldadj, S. Habibzehadeh, and F. M. Shalmani, Crystallization and Particle Size Distribution of Hydrothermally Synthesized SAPO-34: An Experimen tal and Population Balance Study, Scientific Reports 15, 6301 (2025).
  31. [31]tal and Population Balance Study, Scientific Reports 15, 6301 (2025). [31] D. A. Hardy, J. F. Robinson, T. G. Hilditch, et al., Accurate Measurements and Simulations of the Evaporation and Trajectories of Individual Solution Droplets, The Journal of Physical Chemistry B 127, 3416 (2023).DOI
  32. [32]L. Gao, X. Luo, J.-L. Sun, et al., Room-Temperature Solvent Evaporation Induced Crystallization: A General Strategy for Growth of Halide Perovskite Single Crystals by Applying the Le Chatelier’s Principle, Small 19, 2303687 (2023).DOI
  33. [33]V. L. Pool, B. Dou, D. G. Sun Campen, et al., Thermal Engineering of FAPbI3 Perovskite Material via Radiative Thermal Annealing and in Situ XRD, Nature Communications 8, 1407 (2017).DOI
  34. [34]M. Z. Bazant, Theory of Chemical Kinetics and Charge Transfer based on Nonequilibrium Thermodynamics, Accounts of Chemical Research 46, 1144 (2013).
  35. [35]C. Vallance, An Introduction to Chemical Kinetics (Morgan & Claypool Publishers, 2017).DOI
  36. [36]A. H. Persad and C. A. Ward, Expressions for the Evaporation and Condensation Coefficients in the Hertz-Knudsen Relation, Chemical Reviews 116, 7727 (2016).DOI
  37. [37]O. J. J. Ronsin, D. Jang, H.-J. Egelhaaf, et al., Phase-Field Simulation of Liquid–Vapor Equilibrium and Evaporation of Fluid Mixtures, ACS Applied Materials & Interfaces 13, 55988 (2021).arxiv.org/abs/2106.14460
  38. [38]M. Bond and H. Struchtrup, Mean Evaporation and Condensation Coefficients Based on Energy Dependent Condensation Probability, Physical Review E 70, 061605 (2004).pubmed.ncbi.nlm.nih.gov/15697379
  39. [39]H. Schlichting and K. Gersten, Boundary-Layer Theory (Springer, Berlin, Heidelberg, 2017).
  40. [40]D. E. Bornside, R. A. Brown, P. W. Ackmann, et al., The Effects of Gas Phase Convection on Mass Transfer in Spin Coating, Journal of Applied Physics 73, 585 (1993).DOI
  41. [41]D. Meyerhofer, Characteristics of Resist Films Produced by Spinning, Journal of Applied Physics 49, 3993 (2008).
  42. [42]J. J. van Franeker, D. Westhoff, M. Turbiez, et al., Controlling the Dominant Length Scale of Liquid–Liquid Phase Separation in Spin-Coated Organic Semiconductor Films, Advanced Functional Materials 25, 855 (2015).DOI
  43. [43]V. I. Kalikmanov, Nucleation Theory, Lecture Notes in Physics, Vol. 860 (Springer, Heidelberg, 2013).
  44. [44]O. J. J. Ronsin, D. Jang, H.-J. Egelhaaf, et al., A Phase-Field Model for the Evaporation of Thin Film Mixtures, Physical Chemistry Chemical Physics 22, 6638 (2020).DOI
  45. [45]C. D. Van Siclen, Random Nucleation and Growth Kinetics, Physical Review B 54, 11845 (1996).
  46. [46]J. S. Blázquez, F. J. Romero, C. F. Conde, and A. Conde, A Review of Different Models Derived from Classical Kolmogorov, Johnson and Mehl, and Avrami (KJMA) Theory to Recover Physical Meaning in Solid-State Transformations, physica status solidi (b) 259, 2100524 (2022).DOI
  47. [47]C. B. Whitehead, S. Özkar, and R. G. Finke, Lamer’s 1950 Model of Particle Formation: A Review and Critical Analysis of Its Classical Nucleation and Fluctuation Theory Basis, of Competing Models and Mechanisms for Phase-Changes and Particle Formation, and Then of Its Application to Silver Halide, Semiconductor, Metal, and Metal-Oxide Nanoparticles, Materials Advances 2, 186 (2021).DOI
  48. [48]N. T. K. Thanh, N. Maclean, and S. Mahiddine, Mechanisms of Nucleation and Growth of Nanoparticles in Solution, Chemical Reviews 114, 7610 (2014).DOI
  49. [49]K. H. Stone, A. Gold-Parker, V. L. Pool, et al., Transformation from Crystalline Precursor to Perovskite in PbCl2-Derived MAPbI3, Nature Communications 9, 3458 (2018).DOI
  50. [50]R. Szostak, S. Sanchez, P. E. Marchezi, et al., Revealing the Perovskite Film Formation Using the Gas Quenching Method by In Situ GIWAXS: Morphology, Properties, and Device Performance, Advanced Functional Materials 30, 2007473 (2020).DOI
  51. [51]T.-B. Song, Z. Yuan, M. Mori, et al., Revealing the Dynamics of Hybrid Metal Halide Perovskite Formation via Multimodal In Situ Probes, Advanced Functional Materials 30, 1908337 (2020).DOI
  52. [52]E. Kneschaurek, A. Hinderhofer, B. Hofferberth, et al., Compact Sample Environment for In Situ X-ray Scattering during Spin-Coating, Review of Scientific Instruments 94, 063901 (2023).DOI
  53. [53]L. Merten, T. Eberle, E. Kneschaurek, et al., Halide Segregated Crystallization of Mixed-Halide Perovskites Revealed by In Situ GIWAXS, ACS Applied Materials & Interfaces 16, 8913 (2024).DOI
  54. [54]M. Majewski, S. Qiu, O. Ronsin, et al., Simulation of Perovskite Thin Layer Crystallization with Varying Evaporation Rates, Materials Horizons 10.1039/D4MH00957F (2024).arxiv.org/abs/2404.09739

Paper details

vlnc:W_2fwwdsec
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents