arxiv: 2604.23259 · v1 · submitted 2026-04-25 · ❄️ cond-mat.mtrl-sci

Wafer-scale hybrid molecular beam epitaxy of BaTiO3 and SrTiO3 on silicon

Xiaodong Tian , Yan Lin , Hanbin Gao , Han Yu , Yunpeng Ma , Ruiqi Liang , Changfu Chen , Wei Li

show 3 more authors

Chenguang Deng Qiang Zheng Qian Li

This is my paper

Pith reviewed 2026-05-08 08:00 UTC · model grok-4.3

classification ❄️ cond-mat.mtrl-sci

keywords barium titanatehybrid molecular beam epitaxywafer-scale growthelectro-optic coefficientsilicon integrationferroelectric thin filmsstrontium titanate buffer

0 comments

The pith

Hybrid molecular beam epitaxy produces barium titanate on silicon with a 248 pm/V electro-optic coefficient.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper shows that a hybrid molecular beam epitaxy process can grow uniform barium titanate films across entire 4-inch silicon wafers. It uses a titanium precursor to enable fast, self-regulated growth on a strontium titanate buffer layer while keeping the interface atomically sharp. Films made this way display higher crystal quality and a larger effective electro-optic coefficient than films grown by pulsed laser deposition on the same templates. The difference matters for building energy-efficient photonic circuits directly on silicon chips.

Core claim

We demonstrate the continuous, uniform wafer-scale growth of high-quality BTO films on SrTiO3-buffered 4-inch Si(001) wafers using a fully hybrid molecular beam epitaxy approach. By utilizing titanium tetraisopropoxide as a titanium precursor, we achieve a self-regulating, adsorption-controlled layer-by-layer growth at rates exceeding 75 nm/h, while maintaining an atomically sharp and structurally coherent BTO/STO interface. The optimized hMBE-grown BTO exhibits superior crystallinity and a larger effective EO coefficient of 248 pm/V, surpassing that of the PLD-grown films (220 pm/V).

What carries the argument

The fully hybrid molecular beam epitaxy process with titanium tetraisopropoxide precursor, which delivers self-regulating adsorption-controlled growth of BTO on STO/Si templates.

Load-bearing premise

Measurements of electro-optic response on the hybrid MBE and pulsed laser deposition films are directly comparable without hidden differences in thickness, domain structure, or test conditions that could account for the 28 pm/V gap.

What would settle it

Independent measurements on BTO films of matched thickness and domain configuration grown by both methods that yield identical electro-optic coefficients would show the reported advantage is not caused by the growth technique.

read the original abstract

The integration of epitaxial barium titanate (BTO) on silicon represents a highly promising pathway for next-generation, energy-efficient photonic integrated circuits due to BTO's exceptionally high Pockels coefficients. However, the scalable epitaxy of BTO on Si remains hindered by complex stoichiometric control and slow growth rates. In this work, we demonstrate the continuous, uniform wafer-scale growth of high-quality BTO films on SrTiO3 (STO)-buffered 4-inch Si(001) wafers using a fully hybrid molecular beam epitaxy (hMBE) approach. By utilizing titanium tetraisopropoxide as a titanium precursor, we achieve a self-regulating, adsorption-controlled layer-by-layer growth at rates exceeding 75 nm/h, while maintaining an atomically sharp and structurally coherent BTO/STO interface. We systematically compare the structural, ferroelectric, and electro-optic (EO) properties of fully hMBE-grown BTO with those deposited via pulsed laser deposition (PLD) on identical STO/Si templates. While both techniques yield high-quality c-domain dominated films, the optimized hMBE-grown BTO exhibits superior crystallinity and a larger effective EO coefficient of 248 pm/V, surpassing that of the PLD-grown films (220 pm/V). These results highlight the advantages of the fully hMBE approach as a scalable, deterministic, and high-performance materials platform for wafer-scale integrated ferroelectric photonics.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

hMBE with the titanium precursor scales BTO growth to 4-inch Si wafers at >75 nm/h and claims a modest EO edge over PLD, but the 28 pm/V difference needs matched-film controls to be convincing.

read the letter

This paper shows a practical hybrid MBE process for growing c-domain BTO on STO-buffered 4-inch silicon wafers. Using titanium tetraisopropoxide as the Ti source gives self-regulated layer-by-layer growth above 75 nm/h while keeping atomically sharp interfaces. They run the same templates through PLD for comparison and report better crystallinity plus an effective EO coefficient of 248 pm/V versus 220 pm/V for the PLD films.

Referee Report

2 major / 2 minor

Summary. The manuscript reports the development of a hybrid molecular beam epitaxy (hMBE) process for wafer-scale growth of BaTiO3 (BTO) and SrTiO3 (STO) on 4-inch Si(001) wafers using titanium tetraisopropoxide as a Ti precursor. It achieves self-regulated layer-by-layer growth at rates >75 nm/h with atomically sharp BTO/STO interfaces, and compares structural, ferroelectric, and electro-optic properties of optimized hMBE BTO films against PLD-grown films on identical STO/Si templates, claiming superior crystallinity and an effective Pockels coefficient of 248 pm/V versus 220 pm/V.

Significance. If the reported advantages in growth rate, interface coherence, and electro-optic performance are substantiated with matched samples and statistics, this work would represent a meaningful advance in scalable epitaxial integration of high-Pockels-coefficient ferroelectrics on silicon, directly relevant to energy-efficient silicon photonics platforms.

major comments (2)

[Abstract and EO characterization section] Abstract and electro-optic results: The central claim that hMBE-grown BTO exhibits a larger effective EO coefficient (248 pm/V) than PLD-grown films (220 pm/V) is load-bearing for the superiority assertion. However, effective r_eff extraction depends on film thickness (via optical-mode overlap), c-domain fraction, residual strain, and electrode/wavelength geometry. The manuscript states both films are 'c-domain dominated' on 'identical STO/Si templates' but does not report matched thicknesses, quantitative domain fractions, or identical measurement protocols; without these, the 28 pm/V difference cannot be unambiguously attributed to growth-method differences rather than geometric or structural variations.
[Results on structural and ferroelectric properties] Structural and ferroelectric characterization: Claims of 'superior crystallinity' for hMBE films and 'atomically sharp' interfaces require quantitative metrics (e.g., XRD rocking-curve FWHM values, TEM-measured interface roughness, or defect densities) with sample statistics and error bars. These data are needed to establish that the structural improvements are statistically significant and directly responsible for the EO advantage, rather than post-growth selection effects.

minor comments (2)

[Abstract and Methods] The abstract and methods should include error bars, number of samples measured, and explicit measurement protocols (wavelength, electrode configuration, fitting assumptions) for the EO coefficients and growth rates to allow independent assessment of the reported numerical outcomes.
[Figure captions and experimental details] Figure captions and text should clarify whether the PLD and hMBE films were grown to identical nominal thicknesses and whether any post-growth annealing or electrode deposition steps differed between the two sets.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We are grateful to the referee for their thorough review and valuable suggestions, which have helped us improve the clarity and rigor of our manuscript. We address each major comment below.

read point-by-point responses

Referee: [Abstract and EO characterization section] Abstract and electro-optic results: The central claim that hMBE-grown BTO exhibits a larger effective EO coefficient (248 pm/V) than PLD-grown films (220 pm/V) is load-bearing for the superiority assertion. However, effective r_eff extraction depends on film thickness (via optical-mode overlap), c-domain fraction, residual strain, and electrode/wavelength geometry. The manuscript states both films are 'c-domain dominated' on 'identical STO/Si templates' but does not report matched thicknesses, quantitative domain fractions, or identical measurement protocols; without these, the 28 pm/V difference cannot be unambiguously attributed to growth-method differences rather than geometric or structural variations.

Authors: We appreciate the referee pointing this out. The films were deposited on identical STO/Si templates with similar thicknesses, and both are c-domain dominated. The EO characterization was performed with identical protocols. However, to fully address the concern and allow unambiguous attribution, we will revise the manuscript to explicitly report the film thicknesses, quantitative domain fractions from XRD, and include error bars on the EO coefficients from multiple measurements. This will clarify that the difference is due to the growth method's impact on crystallinity. revision: yes
Referee: [Results on structural and ferroelectric properties] Structural and ferroelectric characterization: Claims of 'superior crystallinity' for hMBE films and 'atomically sharp' interfaces require quantitative metrics (e.g., XRD rocking-curve FWHM values, TEM-measured interface roughness, or defect densities) with sample statistics and error bars. These data are needed to establish that the structural improvements are statistically significant and directly responsible for the EO advantage, rather than post-growth selection effects.

Authors: We concur that quantitative metrics with statistics are necessary. We will update the results section to include XRD rocking-curve FWHM values with error bars from multiple samples, TEM-measured interface roughness, and any available defect density estimates. These revisions will establish the statistical significance of the structural improvements and their relation to the EO properties. revision: yes

Circularity Check

0 steps flagged

No circularity: purely experimental comparison with no derivation chain

full rationale

The manuscript is an experimental materials report comparing hMBE and PLD growth of BTO films on Si. It presents measured structural, ferroelectric, and electro-optic data (including the 248 pm/V vs 220 pm/V EO coefficients) without any equations, fitted models, predictions, or self-referential derivations. No load-bearing step reduces by construction to an input defined inside the paper; the central claim rests on direct experimental outputs rather than any mathematical or parametric loop.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on experimental growth and characterization results rather than theoretical derivations; no free parameters, new entities, or non-standard axioms are introduced beyond routine assumptions of epitaxial growth kinetics.

axioms (1)

domain assumption Adsorption-controlled layer-by-layer growth occurs under the chosen hybrid MBE conditions with titanium tetraisopropoxide.
Invoked to explain self-regulating stoichiometry and high growth rate.

pith-pipeline@v0.9.0 · 5590 in / 1300 out tokens · 25348 ms · 2026-05-08T08:00:38.339883+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

5 extracted references

[1]

Despite its dominance of large-scale integration, silicon is a centrosymmetric crystal structurally devoid of a linear electro-optic (Pockels) response

INTRODUCTION The relentless push for higher bandwidth and minimal energy dissipation in next-generation optical interconnects has exposed the intrinsic limitations of current photonic platforms [1–3]. Despite its dominance of large-scale integration, silicon is a centrosymmetric crystal structurally devoid of a linear electro-optic (Pockels) response. Con...
[2]

STEM was performed on a double-aberration-corrected transmission electron microscope (Spectra 300, Thermo Fisher Scientific) at an accelerating voltage of 300 kV

zone axis using focused ion beam (FIB) in a dual-beam microscope (Zeiss Crossbeam). STEM was performed on a double-aberration-corrected transmission electron microscope (Spectra 300, Thermo Fisher Scientific) at an accelerating voltage of 300 kV. High-angle annular dark-field (HAADF) images were collected under a probe convergence angle of 25 mrad and wit...

2000
[3]

Lithium niobate photonics: Unlocking the electromagnetic spectrum

Boes A, Chang L, Langrock C, et al. Lithium niobate photonics: Unlocking the electromagnetic spectrum. Science 2023, 379: eabj4396. [3] Wan Y, He W, Jaussi J, et al. Integrating silicon photonics with complementary metal–oxide–semiconductor technologies. Nat Rev Electr Eng 2025, 3: 15–31. [4] Reed GT, Mashanovich G, Gardes FY, et al. Silicon optical modul...

2023
[4]

The interface of epitaxial SrTiO3 on silicon: in situ and ex situ studies

Hu X, Li H, Liang Y, et al. The interface of epitaxial SrTiO3 on silicon: in situ and ex situ studies. Appl Phys Lett 2003, 82: 203–5. [20] Baiutti F, Wrobel F, Christiani G, et al. Oxide molecular beam epitaxy of complex oxide heterointerfaces. In: Metal Oxide-Based Thin Film Structures, Elsevier; 2018: 53–78. [21] Jalan B, Engel-Herbert R, Wright NJ, et...

2003
[5]

Delineating complex ferroelectric domain structures via second harmonic generation spectral imaging

Li W, Ma Y, Feng T, et al. Delineating complex ferroelectric domain structures via second harmonic generation spectral imaging. Journal of Materiomics 2023, 9: 395–402. [37] Li W, Wang S, Peng P, et al. Terahertz excitation of collective dynamics of polar skyrmions over a broad temperature range. Nat Phys 2025, 21: 1965–72. [38] Hsu MHM, Thourhout DV, Pan...

2023