Low voltage and high-bandwidth thin-film lithium tantalate modulator on a silicon dioxide substrate
Pith reviewed 2026-05-10 10:43 UTC · model grok-4.3
The pith
Thin-film lithium tantalate modulators on fused silica reach 64 GHz electro-optic bandwidth with 1.53 V half-wave voltage.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
By employing a slow-wave electrode design to achieve velocity matching between microwave and optical signals, the demonstrated modulator achieves a 3-dB electro-optic bandwidth of 64 GHz with a low half-wave voltage of 1.53 V, with potential to operate at the measured 100 GHz electrical bandwidth if the employed spectral biasing is removed. The modulator moreover exhibits low bias drift, with a constant switching voltage down to 10 mHz. Using the fabricated devices, a net single lane data rate of 440.6 Gbps is achieved using PAM8 signaling.
What carries the argument
Slow-wave electrode design to achieve velocity matching between microwave and optical signals on a fused-silica substrate
If this is right
- High-speed single-lane data transmission reaches 440.6 Gbps using PAM8 modulation.
- The device can potentially operate at the full 100 GHz electrical bandwidth once spectral biasing is removed.
- Low bias drift maintains constant switching voltage at frequencies as low as 10 mHz.
- Performance matches state-of-the-art lithium niobate modulators fabricated on quartz substrates.
Where Pith is reading between the lines
- Reduced birefringence in lithium tantalate may enable denser and more complex photonic integrated circuits than lithium niobate allows.
- Higher optical power handling could extend use to applications requiring elevated optical intensities without material damage.
- Fabrication on 4-inch fused silica wafers opens a path to lower-cost, higher-volume production of high-performance modulators.
Load-bearing premise
The slow-wave electrode design achieves true velocity matching on the fused-silica substrate without introducing unaccounted losses or fabrication defects that limit the measured bandwidth and voltage performance.
What would settle it
A direct measurement revealing significant velocity mismatch between microwave and optical signals, or an observed 3-dB electro-optic bandwidth well below 64 GHz under the same drive conditions, would show the claimed matching has not been realized.
Figures
read the original abstract
Modern communication networks demand ever-increasing transmission bandwidth, placing stringent requirements on low-cost, high-performance electro-optic modulators. Substantial advances have been made in integrated photonics employing lithium niobate on insulator. In contrast, photonic integrated circuits based on lithium tantalate -- a material already commercially adopted for wireless filters -- have been developed, offering reduced DC drift, higher optical power handling, and lower birefringence. These advantages enable more complex and dense photonic integrated circuits, and make lithium tantalate a promising material platform for next-generation integrated electro-optic modulators. However, in contrast to the extensively studied thin-film lithium niobate platform, thin-film lithium tantalate modulators have only been explored on silicon substrates. Here, we report the first fabrication and characterization of thin-film lithium tantalate electro-optic modulators manufactured on a 4-inch (100 mm) fused-silica substrate for adapting a low-loss slow-wave microwave electrode to improve the electro-optic bandwidth. By employing a slow-wave electrode design to achieve velocity matching between microwave and optical signals, the demonstrated modulator achieves a 3-dB electro-optic bandwidth of 64 GHz with a low half-wave voltage of 1.53 V, with potential to operate at the measured 100 GHz electrical bandwidth, if the employed spectral biasing is removed. The modulator moreover exhibits low bias drift, with a constant switching voltage down to 10 mHz. This performance enables high-speed data transmission comparable to state-of-the-art lithium niobate modulators fabricated on quartz substrates. Using the fabricated devices, a net single lane data rate of 440.6 Gbps is achieved using PAM8 signaling.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The manuscript reports the first fabrication of thin-film lithium tantalate electro-optic modulators on 4-inch fused-silica substrates. A slow-wave electrode design is used to achieve velocity matching between microwave and optical signals, yielding a measured 3-dB electro-optic bandwidth of 64 GHz at a half-wave voltage of 1.53 V. Additional results include low bias drift (constant switching voltage to 10 mHz), an electrical bandwidth of 100 GHz, and a net single-lane data rate of 440.6 Gbps using PAM8 signaling. The work positions LT on quartz as competitive with lithium niobate platforms for high-speed, low-voltage modulation.
Significance. If the velocity-matching claim and headline metrics are substantiated, the result would be significant for integrated photonics. It establishes a new substrate platform for LT modulators that exploits the material's lower birefringence, reduced DC drift, and higher optical power handling relative to LN, while demonstrating data rates comparable to state-of-the-art LN devices on quartz. The experimental demonstration on 4-inch wafers also supports scalability.
major comments (3)
- [Abstract / Results] Abstract and bandwidth-measurement section: the central claim of 64 GHz 3-dB EO bandwidth at Vπ = 1.53 V is presented without error bars, uncertainty estimates, or a description of the S21 extraction method (including how spectral biasing was accounted for). This makes it impossible to judge whether the quoted numbers are robust or whether the bandwidth is truly limited by velocity matching rather than electrode loss.
- [Electrode design / Microwave measurements] Electrode-design and microwave-characterization section: the slow-wave geometry is asserted to produce velocity matching on fused silica, yet no measured microwave effective index, group-velocity comparison, or control device without the slow-wave features is reported. Without this datum, the causal link between the electrode design and the observed 64 GHz bandwidth cannot be verified; the bandwidth could instead be set by dielectric dispersion or conductor losses.
- [High-speed data transmission] Data-transmission section: the 440.6 Gbps PAM8 result is stated as a 'net' rate, but no BER curves, DSP details, or overhead accounting are supplied. This prevents assessment of whether the rate is limited by the modulator or by the measurement setup.
minor comments (2)
- [Abstract] The statement that the device 'has potential to operate at the measured 100 GHz electrical bandwidth, if the employed spectral biasing is removed' requires a brief explanation of the biasing technique and its bandwidth-limiting mechanism.
- [Figures] Figure captions and axis labels should explicitly state the substrate (fused silica) and electrode geometry parameters to allow direct comparison with prior LN-on-quartz work.
Simulated Author's Rebuttal
We thank the referee for the careful and constructive review of our manuscript. The comments highlight important aspects that will improve the clarity and rigor of the presentation. We address each major comment point by point below and will incorporate revisions where appropriate to strengthen the paper.
read point-by-point responses
-
Referee: [Abstract / Results] Abstract and bandwidth-measurement section: the central claim of 64 GHz 3-dB EO bandwidth at Vπ = 1.53 V is presented without error bars, uncertainty estimates, or a description of the S21 extraction method (including how spectral biasing was accounted for). This makes it impossible to judge whether the quoted numbers are robust or whether the bandwidth is truly limited by velocity matching rather than electrode loss.
Authors: We agree that additional details on measurement uncertainty and the S21 extraction procedure are needed. In the revised manuscript, we will add error bars based on repeated measurements across multiple devices and a step-by-step description of the S21 extraction, including normalization to the low-frequency response and the quadrature biasing condition used to mitigate spectral biasing effects. These additions will confirm that the 64 GHz 3-dB point reflects the velocity-matched performance rather than electrode losses alone. revision: yes
-
Referee: [Electrode design / Microwave measurements] Electrode-design and microwave-characterization section: the slow-wave geometry is asserted to produce velocity matching on fused silica, yet no measured microwave effective index, group-velocity comparison, or control device without the slow-wave features is reported. Without this datum, the causal link between the electrode design and the observed 64 GHz bandwidth cannot be verified; the bandwidth could instead be set by dielectric dispersion or conductor losses.
Authors: The referee correctly notes the absence of direct microwave index measurements or control devices. Our design relied on electromagnetic simulations to achieve velocity matching between the microwave and optical modes on the fused-silica substrate, and the resulting 64 GHz bandwidth with low Vπ is consistent with this matching. In revision we will expand the electrode-design section with the simulated microwave effective index, group-velocity dispersion curves, and a quantitative comparison to the optical group index. We did not fabricate control devices without slow-wave features in this initial study, as the focus was on demonstrating the first LT-on-quartz platform; however, the performance metrics support the design choice over unmatched electrodes, which typically exhibit earlier roll-off. revision: partial
-
Referee: [High-speed data transmission] Data-transmission section: the 440.6 Gbps PAM8 result is stated as a 'net' rate, but no BER curves, DSP details, or overhead accounting are supplied. This prevents assessment of whether the rate is limited by the modulator or by the measurement setup.
Authors: We will revise the data-transmission section to include the measured BER curves versus received power, a description of the DSP chain (including linear equalization, nonlinear compensation, and FEC decoding), and explicit overhead accounting that yields the net 440.6 Gbps rate. The BER remained below the post-FEC error-free threshold, confirming that the demonstrated rate is supported by the modulator's 64 GHz bandwidth and 1.53 V drive voltage rather than being limited by the test equipment. revision: yes
Circularity Check
No circularity: purely experimental report with measured results only
full rationale
The paper is an experimental fabrication and characterization study of thin-film lithium tantalate modulators on fused-silica substrates. All headline metrics (64 GHz EO bandwidth, 1.53 V Vπ, 440.6 Gbps PAM8 data rate, low bias drift down to 10 mHz) are stated as direct measurement outcomes with no equations, derivations, fitted parameters, or model-based predictions. The slow-wave electrode is described as employed to achieve velocity matching, but this is presented as a design choice whose outcome is verified by measured bandwidth rather than by any self-referential calculation or self-citation chain that reduces the result to its inputs. No load-bearing steps exist that could be circular by the enumerated patterns.
Axiom & Free-Parameter Ledger
Reference graph
Works this paper leans on
-
[1]
Optical layer lithography
-
[2]
Electrode lithography
-
[3]
Metal lift-off and capping LT SiO2 Si FS DLC PR Au
-
[4]
1:Low voltage and high bandwidth thin film lithium tantalate modulators on a fused silica substrate
Wafer expansion Activated surfaces FIG. 1:Low voltage and high bandwidth thin film lithium tantalate modulators on a fused silica substrate. (a)Microscope picture of a chip comprising three modulators from a 4-inch wafer(b).(c)Fabrication process flow for the fused-silica-based lithium tantalate modu- lator, including the substrate preparation, device fab...
-
[5]
N. Harris,Lightmatter interconnect launch event at ofc 2025, URLhttps://lightmatter.co/resource/ lightmatter-interconnect-launch-event-at-ofc-2025/ #
work page 2025
-
[6]
C. Wang, M. Zhang, X. Chen, M. Bertrand, A. Shams-Ansari, S. Chandrasekhar, P. Winzer, and M. Lon ˇcar, Nature562, 101 (2018), ISSN 1476-4687, URLhttps://doi.org/10.1038/ s41586-018-0551-y
work page 2018
-
[7]
M. Xu, M. He, H. Zhang, J. Jian, Y . Pan, X. Liu, L. Chen, X. Meng, H. Chen, Z. Li, et al., Nat. Commun.11, 3911 (2020), ISSN 2041-1723, URLhttps://doi.org/10.1038/ s41467-020-17806-0
work page 2020
-
[8]
J. Holzgrafe, E. Puma, R. Cheng, H. Warner, A. Shams-Ansari, R. Shankar, and M. Lon ˇcar, Opt. Express32, 3619 (2024), URLhttps://opg.optica.org/oe/abstract.cfm?URI= oe-32-3-3619
work page 2024
-
[9]
Y . Xu, M. Shen, J. Lu, J. B. Surya, A. A. Sayem, and H. X. Tang, Optics Express29, 5497 (2021)
work page 2021
-
[10]
C. Wang, Z. Li, J. Riemensberger, G. Lihachev, M. Churaev, W. Kao, X. Ji, J. Zhang, T. Blesin, A. Davydova, et al., Nature 629, 784 (2024), ISSN 1476-4687, URLhttps://doi.org/ 10.1038/s41586-024-07369-1
-
[11]
C. Wang, D. Fang, J. Zhang, A. Kotz, G. Lihachev, M. Chu- raev, Z. Li, A. Schwarzenberger, X. Ou, C. Koos, et al., Optica 11, 1614 (2024), URLhttps://opg.optica.org/optica/ abstract.cfm?URI=optica-11-12-1614
work page 2024
-
[12]
soitec.com/en/capital-markets-day-2021
SOITEC,Capital markets day 2021(2021),https://www. soitec.com/en/capital-markets-day-2021
work page 2021
-
[13]
S. U. Hulyal, J. Hu, C. Wang, J. Cai, G. Lihachev, and T. J. Kip- penberg, Optica12, 978 (2025), URLhttps://opg.optica. org/optica/abstract.cfm?URI=optica-12-7-978
work page 2025
- [14]
- [15]
-
[16]
G. Chen, K. Chen, R. Gan, Z. Ruan, Z. Wang, P. Huang, C. Lu, A. P. T. Lau, D. Dai, C. Guo, et al., APL photonics7(2022)
work page 2022
-
[17]
E. Berikaa, M. S. Alam, W. Li, S. Bernal, B. Krueger, F. Pittal`a, and D. V . Plant, IEEE Photonics Technology Letters35, 850 (2023)
work page 2023
- [18]
-
[19]
Z. Li, R. N. Wang, G. Lihachev, J. Zhang, Z. Tan, M. Churaev, N. Kuznetsov, A. Siddharth, M. J. Bereyhi, J. Riemensberger, et al., Nat. Commun.14, 4856 (2023), ISSN 2041-1723, URL https://doi.org/10.1038/s41467-023-40502-8
- [20]
-
[21]
Nagarajan,Technique for measuring the Vpi-AC of a Mach-Zehnder modulator(Mar
R. Nagarajan,Technique for measuring the Vpi-AC of a Mach-Zehnder modulator(Mar. 20, 2001), U.S. Patent 6,204,954 B1, URLhttps://patents.google.com/ patent/US6204954B1/
work page 2001
-
[22]
S. H. Lin and S. Y . Wang,26, 1696 (1987), ISSN 1559-128X
work page 1987
- [23]
-
[24]
IEEE Std 802.3-2022 (Revision of IEEE Std 802.3-2018) pp. 1–7025 (2022)
work page 2022
-
[25]
A. Graell i Amat and L. Schmalen,F orward Error Correction for Optical Transponders(Springer International Publishing, Cham, 2020), pp. 177–257, ISBN 978-3-030-16250-4, URL https://doi.org/10.1007/978-3-030-16250-4_7
-
[26]
Recommendation G.975.1, Telecommunication standardization sector of International Telecommunication Union (2004), URL https://www.itu.int/rec/T-REC-G.975.1-200402-I/ en
work page 2004
- [27]
-
[28]
Q. Hu, R. Borkowski, Y . Lefevre, J. Cho, F. Buchali, R. Bonk, K. Schuh, E. De Leo, P. Habegger, M. Destraz, et al., J. Light. Technol.40, 3338 (2022)
work page 2022
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.