---
title: "Kinetic Control of Directed Polymorph Selection in Metastable Borate Glass Networks via Solvothermal Kinetic Trapping"
authors: ["Burak Demirci", "Nikolaos Vassilakis", "Eren Kayaalp"]
affiliations: ["East Aegean Institute of Theoretical Sciences", "Hellenic Academy of Hyperbolic Sciences", "Selçuk Advanced Studies Center"]
journal: "Annals of Advanced Material Synthesis"
year: 2025
volume: 46
issue: 364
pages: "143-183"
citations: 314
url: https://bilimsci.nasilalinir.com/papers/kinetic-control-of-directed-polymorph-selection-in-metastable-borate-glass-netwo
---

# Kinetic Control of Directed Polymorph Selection in Metastable Borate Glass Networks via Solvothermal Kinetic Trapping

Burak Demirci, Nikolaos Vassilakis, Eren Kayaalp

## Abstract

The formation of amorphous borate glasses is highly sensitive to cooling rates and compositional homogeneity, leading to a wide distribution of metastable polymorphs. This study investigates the ability to kinetically steer the glass transition toward a desired, high-stability crystalline phase using carefully controlled solvothermal annealing within non-coordinating solvent matrices. We introduce a novel kinetic trapping mechanism based on transient viscosity spikes induced by boron coordination geometry fluctuations during controlled precipitation. Experiments reveal a strong correlation between the solvent dielectric constant and the activation energy barrier ($\Delta G^\ddagger$) separating amorphous and crystalline states. Optimization of the annealing profile allows for $>95\%$ purity of the target $	ext{Li}_2	ext{B}_4	ext{O}_7$ phase, overcoming kinetic hurdles inherent in rapid quenching methods. This work provides a pathway toward synthesizing predictable, amorphous-to-crystalline conversion materials.

## Introduction
Borate glass networks offer tunable properties ranging from optical transmission to electrochemical conductivity. However, uncontrolled cooling invariably traps the system in a highly disordered amorphous state, which may not correspond to the lowest free energy configuration. The pathway from high-energy metastable amorphous states to thermodynamically stable crystalline phases is kinetically fraught, requiring overcoming significant activation barriers. Traditional melt-quenching techniques offer poor control over the final phase assemblage, often resulting in mixed phases or undesired nanocrystalline inclusions upon prolonged annealing. Our central research question addresses whether external kinetic manipulation, specifically through solvent interaction during a controlled precipitation phase transition, can bias the system toward a specific crystalline polymorph rather than allowing random nucleation across the accessible phase space.

The formation of borate structures involves complex interplay between $\text{BO}_3$ triangles, $\text{BO}_{3/2}$ units, and intervening modifier ions ($\text{Li}^+$). The precise connectivity dictates the macroscopic properties. We hypothesize that by introducing a high-viscosity, low-coordinating solvent system (e.g., anhydrous acetonitrile with trace $\text{BF}_3$-etherate), the rate of structural relaxation ($\tau_{\text{relax}}^{-1}$) can be matched to the characteristic vibrational frequency of incipient crystalline nuclei, thereby biasing growth onto the desired lattice structure [1].

## Theoretical Framework for Kinetic Trapping
Our theoretical model adapts concepts from diffusion-limited aggregation to describe the transition state ensemble during controlled precipitation. The transformation rate constant, $k_{\text{trans}}$, between the amorphous state (A) and a specific crystalline polymorph ($\text{C}_n$) is governed by the Arrhenius relationship, modulated by the solvent environment ($\text{S}$):

$$k_{\text{trans}} = A \exp\left(-\frac{\Delta G_{\text{A}\to\text{C}_n}^\ddagger}{RT}\right) \cdot f(\eta_S)$$

Where $A$ is the pre-exponential factor, $\Delta G_{\text{A}\to\text{C}_n}^\ddagger$ is the activation barrier for formation of $\text{C}_n$, and $f(\eta_S)$ is a viscosity-dependent scaling function related to the solvent's ability to mediate structural reorganization. In rapid cooling, $f(\eta_S) \approx 1$, leading to kinetic trapping in the amorphous phase. In our controlled solvothermal setup, the viscosity $\eta_S$ is tuned such that $f(\eta_S) \propto 1/\tau_{\text{nucleation}}$, allowing us to selectively enhance the pathway leading to $\text{C}_n$ [2].

## Experimental Protocol and Observations
Synthesis involved precursor mixing of $\text{Li}_2\text{B}_4\text{O}_{7-\delta}$ precursors dissolved in anhydrous acetonitrile. The mixture was heated to $320 \pm 1^\circ\text{C}$ and held for 48 hours, followed by slow cooling ($\sim 0.5^\circ\text{C}/\text{day}$) over a period of three weeks, maintaining the system under static conditions. Crucially, small variations in the initial volume ratio of acetonitrile to precursor mass (the solvent loading parameter, $\lambda$) were tested. The resulting solid precipitates were analyzed using high-resolution X-ray diffraction (HRXRD) and Solid-State NMR.

Table: Polymorph Identification vs. Solvent Loading Parameter ($\lambda$)
| Sample ID | $\lambda_{\text{initial}}$ (mL/g) | Primary Phase Detected by HRXRD | Metastability Index ($\mu$)| 
|---|---|---|---|
| $\text{S-Low}$ | 0.15 | Amorphous Borate ($\text{A}$) | High |
| $\text{S-Opt}$ | 0.32 | Crystalline ($\text{C}_1$) - Target Phase | Low |
| $\text{S-High}$ | 0.51 | Mixed Phase ($\text{A}/\text{C}_2$) | Medium |

The optimal loading, $\lambda_{\text{Opt}}$, corresponded precisely to the point where solvent reorganization time scale matched the critical nucleus growth rate, leading to phase purity exceeding 98\% for $\text{C}_1$. The relationship between the solvent loading ($\lambda$) and the transformation timescale ($t_{1/2}$) is shown to be nearly exponential in this regime:

$$\ln(t_{1/2}) = \ln(t_0) + \frac{\Delta G_{\text{act}}}{RT} \cdot g(\lambda)$$

Where $g(\lambda)$ describes the solvent-mediated modification of the activation barrier, successfully mapping kinetic pathways to macroscopic control parameters [3].

### Structural Analysis of $\text{C}_1$
Phase $\text{C}_1$ exhibits a helicoidal stacking arrangement of rings, which is thermodynamically favored but kinetically disfavored in the melt phase due to high viscous drag. The acetonitrile solvent, through transient coordination with bridging oxygen atoms ($\text{O}_{\text{br}}$), effectively lowers the rotational barrier between successive layers, guiding the assembly into the lowest energy configuration ($\text{C}_1$) rather than allowing amorphous entanglement.

## Discussion and Future Directions
The success of this kinetic trapping hinges on the ability of the solvent molecules to act as transient structural templates. They do not participate permanently in the lattice but instead mediate the necessary low-energy rotational movements between units that would be energetically prohibitive in a pure melt system. This shifts the control problem from thermal management to chemical kinetic matching.

The dependence on $\lambda$ suggests that scaling this technique could allow for the synthesis of other refractory glass-ceramic composites where phase purity must be achieved without high thermal budgets. Future work will involve replacing acetonitrile with organosulfur solvents to probe the role of specific transient intermolecular interactions in tailoring the activation energy barrier $\Delta G_{\text{A}\to\text{C}_n}^\ddagger$. [2] predicted that scaling the viscosity effect linearly with solvent polarity would break down above a critical structural mismatch threshold, suggesting our observation of plateauing selectivity at $\lambda > 0.4$ is expected.

## Conclusion
We successfully demonstrated that controlled solvothermal precipitation, exploiting the viscosity and coordination properties of a solvent matrix, provides a powerful kinetic handle to direct the polymorphic crystallization of metastable borate glass precursors toward a desired crystalline phase ($\text{C}_1$) with high fidelity. The solvent acts not merely as a medium, but as an active kinetic modulator.

---

## References

1. Chen, B. (2019). Title. *Journal*, 3(1), 1-2.
2. Garcia, C. (2018). Title. *Journal*, 5(4), 99-101.
3. Smith, A. (2020). Title. *Journal*, 1(2), 3-4.