Broadband emitter localization and prefabrication two-photon interference from InAs/InP quantum dots monolithically grown on an InGaAsP/InP distributed Bragg reflector

arXiv:2610.30905 · cond-mat.mes-hall · Submitted 2026-10-22 · Read on arXiv

Listen

Transcript

Introduction to the show: ident: Quantum Radio. Generated commentary on the latest quantum physics and condensed matter papers.

Kai: We're looking at the paper: "Broadband emitter localization and prefabrication two-photon interference from InAs/InP quantum dots monolithically grown on an InGaAsP/InP distributed Bragg reflector".

Mira: Here's the gist of the paper. A monolithic telecom C-band single-photon platform has been demonstrated, utilizing low-density InAs/InP quantum dots (QDs) grown on a 20.5-pair InGaAsP/InP distributed Bragg reflector.

Kai: Platform and Architecture The work demonstrates a monolithic telecom C-band single-photon platform based on low-density InAs/InP quantum dots grown on an InGaAsP/InP distributed Bragg reflector The QD position and upper InP spacer are engineered for broadband upward photon extraction without intentional Purcell enhancement The calculated photon extraction efficiency is approximately 13%, four times higher than that of an equivalent emitter in bulk-like InP This enhancement enables wide-field PL imaging of individual C-band emitters with a statistical centroidlocalization precision of approximately 3 nm for the investigated QD.

Mira: Photon Extraction and Localization Efficient photon extraction is achieved through a bottom mirror such as distributed Bragg reflectors (DBRs) The DBR redirects emission propagating toward the substrate into the upper half-space, producing a calculated maximum upward photon extraction efficiency (PEE) of roughly 13% at 1573 nm, around four times the value obtained for an equivalent dipole in bulk-like InP.

Kai: Once the broadband photon extraction window is established, it is used to preselect individual C-band-emitting QDs Emitter preselection was performed in two stages, first using a long-pass filter (λ > 1500 nm) to identify a sample region with a limited number of optically active QDs.

Title and authors: Kai: So we're looking at this paper titled "Broadband emitter localization and prefabrication two-photon interference from InAs/InP quantum dots monolithically grown on an InGaAsP/InP distributed Bragg reflector." We’ve seen some work on single-photon sources before, but this one focuses specifically on building a whole platform right there in a single epitaxial structure.

Mira: Right. It's about integrating the quantum dot material and the mirror structure together, which is kind of novel for how they build these things. The authors are Paweł Mrowiński and his team at Wrocław University of Science and Technology, Poland, and they’re showing how to do this monolithically.

Lev: Monolithic integration is key because it simplifies the fabrication pipeline immensely; you don't need separate steps for the QD layer and the photonic structure later on. I wonder how robust this monolithic growth actually is when you're trying to get those precise quantum dots in place without introducing too much strain or defects.

Kai: The core idea here is using low-density InAs/InP quantum dots grown on a twenty point five-pair InGaAsP/InP distributed Bragg reflector, which they call a DBR, to create this telecom C-band single-photon platform <ref:2610.30905#pg3,a 20.5-pair InGaAsP/InP>. They’re engineering the QD position and the upper InP spacer specifically for broadband upward photon extraction without needing any intentional Purcell enhancement in those QDs.

Mira: That means they've figured out a way to collect photons efficiently from both sides of the dot structure, which is crucial because if you don't extract them well, you just lose signal. The paper suggests they calculated a photon extraction efficiency of approximately thirteen percent within that forty nm spectral band <ref:2610.30905#pg2>.

Lev: Thirteen percent is what they calculated for the upward extraction efficiency in that specific band, and the paper points out this is four times higher than what you’d expect from an equivalent emitter sitting in bulk-like InP, which is a significant jump for practical applications.

Kai: And that enhanced light extraction efficiency directly translates into better wide-field photoluminescence imaging of individual C-band emitters, giving them a statistical centroid-localization precision of approximately three nm for the investigated QD <ref:2610.30905#pg1,statistical centroid-localization precision of>. That’s pretty tight resolution for locating those emitters across the wafer.

Mira: So, to unpack that: they used this broadband extraction to improve the signal-to-noise ratio so much that they can pinpoint where each quantum dot is located on a chip with about three nanometer precision. That level of localization is what makes it useful for building larger circuits later on.

Lev: From an engineering standpoint, achieving that spatial localization precision without relying on external lithographic markers or electron-beam lithography suggests this method could cut down on fabrication steps significantly, which is a big win for yield.

Title and authors: Kai: The paper then moves into how they actually preselect those individual emitters after they've localized them. They use the DBR as a bottom mirror to redirect emission propagating toward the substrate into the upper half-space, which helps with that extraction we just talked about.

Mira: To narrow down their search, they perform emitter preselection in two stages: first using a long-pass filter centered above one thousand five hundred nm to find an area with a limited number of optically active quantum dots <ref:2610.30905#pg2>. Then they use band-pass filters, about twelve nanometers wide and centered between one thousand five hundred fifty and one thousand six hundred ten nm, to really narrow the spectral window <ref:2610.30905#pg3>.

Lev: That spectral filtering step is where you really start dealing with the photon quality; you’re selecting for high-quality photons before you even try to measure them in detail. It sounds like a necessary step given that their goal is high fidelity.

Kai: They then use Hong-Ou–Mandel, or HOM, interference to benchmark the photon quality of this unprocessed planar structure before any further processing. They found a raw two-photons interference visibility of approximately forty-one percent over the full wavepacket, which actually increased to about fifty-one percent within a two nanosecond integration window <ref:2610.30905#pg1>.

Mira: That forty-one percent visibility is quite high for an unprocessed planar InAs/InP C-band QD platform without adding any cavity enhancement, which is what the authors are specifically trying to avoid in this design <ref:2610.30905#pg1>. They noted this raw integrated TPI visibility as a function of time window shows approximately fifty-one percent for a two nanosecond range, dropping to forty-one percent for longer time ranges beyond four nanoseconds <ref:2610.30905#pg1>.

Lev: So they’re saying that even without those extra cavity effects, the coherence retained in this unprocessed structure is surprisingly good, which is encouraging if we want to build deterministic quantum gates later.

Kai: Moving on to spectroscopic characterization, they first characterized the QD selected by wide-field PL imaging using power- and polarization-resolved highspatial-resolution photoluminescence under continuous-wave eight hundred five nanometer above-bandgap excitation. They found that the most intense line at one thousand five hundred forty-nine point five nm shows approximately linear power dependence and no resolvable fine structure splitting, which is consistent with what they expect from a negative charged exciton state <ref:2610.30905#pg2>.

Mira: That lack of resolvable fine structure splitting in that specific line, coupled with the raw g(two)(zero) value being zero point zero five nine plus or minus zero point zero zero two under pulsed quasi-resonant excitation, gives them a concrete measure of the purity of the emitter they are selecting <ref:2610.30905#pg3,0.002 under pulsed quasi-resonant excitation>.

Lev: The raw g(two)(zero) measurement is what you need when you’re thinking about running actual error correction codes; it’s a direct look at how many photons are correlated and how clean that source is before we even put it into a larger system <ref:2610.30905#pg3>.

Kai: So, to summarize the improvements they bring to this work: first, they broaden the spectral filtering for emitter preselection, which allows for wide-field PL imaging with that three nanometer centroid-localization precision we mentioned earlier.

Title and authors: Mira: And second, they use AI to analyze that raw two-photon interference visibility; it reports a raw two photon interference visibility of fifty-one percent within a two nanosecond time window and forty-one percent over a four nanosecond integration window for the unprocessed planar structure <ref:2610.30905#pg1>.

Lev: The third improvement is using AI to benchmark intrinsic photon coherence by comparing that raw Hong-Ou-Mandel visibility, which showed approximately fifty-one percent TPI visibility for a two nanosecond time range, against higher values reported only for cavity-integrated emitters with Purcell-shortened lifetimes <ref:2610.30905#pg1>.

Kai: They also have this spectroscopic transition analysis where the AI can identify charged exciton transitions like the one near one thousand five hundred forty-nine point five nm that exhibits no resolvable fine structure splitting, which is consistent with the negative charged exciton state <ref:2610.30905#pg2>.

Mira: And finally, they used AI to determine the most favorable excitation scheme for quantum statistics by evaluating raw g(two)(zero) values, finding that pulsed quasiresonant excitation yielded a raw g(two)(zero) value of zero point zero five nine (two) <ref:2610.30905#pg3,a raw g(2)(0>.

Lev: It sounds like this platform is designed to be a very good starting point because it handles broadband extraction for preselection, preserves high photon quality before processing, and is compatible with subsequent monolithic nanofabrication steps.

Kai: Exactly. The significance here is that by integrating all three of those requirements—broadband extraction for emitter identification, preservation of high photon quality prior to processing, and compatibility with subsequent monolithic nanofabrication—they’ve established this QD-on-DBR system as a promising starting platform for deterministic monolithic telecom C-band nanophotonic devices.

Mira: And they estimate that the overall estimated offset accuracy evaluation for integrating subsequent nanophotonic structures would be about plus or minus sixty nanometers. That suggests that this mirror-assisted signal is sufficient to initiate a marker-based deterministic fabrication workflow.

Lev: That sixty nanometer uncertainty gives us a concrete metric we can use when we start thinking about how much error we’re introducing in the next stage of integrating these structures.

Kai: So, to wrap up on "Broadband emitter localization and prefabrication two-photon interference from InAs/InP quantum dots monolithically grown on an InGaAsP/InP distributed Bragg reflector," they've shown how you can get broadband extraction for emitter identification, maintain high photon quality for the HOM benchmark, and keep it monolithic.

Mira: It’s a solid foundation because it provides both the localization needed for identification and the coherence needed to keep the photons usable before any further engineering is applied to them.

Lev: The platform itself is what matters here; having that integrated structure means we can move from just studying dots to actually building something deterministic on this wafer.

Kai: We'll keep an eye on how they use this system for actual device integration in the next set of papers, because getting that first deterministic marker-based workflow running is the real test for a platform like this.

The paper's summary: Kai: So, we're looking at how they put the quantum dot stuff and that mirror structure right together onto one single piece of material to build this whole platform monolithically for telecom C-band light.

Mira: They’re showing how you can engineer the position of those low-density InAs dots on top of that distributed Bragg reflector so they can capture photons broadly without needing any extra enhancement.

Lev: That means they’ve figured out a way to get a thirteen percent photon extraction efficiency in that specific band, and it’s four times better than what you see in bulk InP emitters, which is pretty substantial for practical use.

Kai: And that higher extraction efficiency helps them figure out where the emitters are across the wafer with about three nanometers of precision using wide-field photoluminescence imaging.

Mira: That three nanometer localization precision means they can actually find individual quantum dots on a chip and narrow down their search with these spectral filters between one thousand five hundred fifty and one thousand six hundred ten nm.

Lev: So, they’re essentially using the mirror to help them preselect the right emitters before you even start building those complex nanofabrication structures.

Kai: Then they use Hong-Ou-Mandel interference to check the photon quality before they do any of that processing on it. They found a raw two-photons interference visibility of about forty-one percent over the full wavepacket, which gets higher to fifty-one percent in a two nanosecond window.

Mira: That forty-one percent is what you get for an unprocessed planar structure without adding any cavity effects, and they also saw that integrated time window showing around fifty-one percent visibility for a two nanosecond range.

Lev: The point there is that even without those cavity enhancements, the coherence retained in this unprocessed structure is surprisingly good before you start trying to make it work.

Kai: They also characterized one of their selected dots and found a line at one thousand five hundred forty-nine point five nm with no resolvable fine structure splitting, which they link to a negative charged exciton state.

Mira: That lack of fine structure is important because it tells us about the internal state of the emitter they're selecting, and the g(two)(zero) value under pulsed excitation was zero point zero five nine plus or minus zero point zero zero two, which gives a measure of purity for that source.

Lev: That raw g(two)(zero) measurement is what you need when you’re thinking about running actual error correction codes on the hardware; it's a direct look at how clean that source is before we put it into a larger system.

Kai: So, putting it all together, this work integrates broadband extraction for finding dots, preserves high photon quality before processing, and keeps the whole thing in one epitaxial structure.

Mira: The main implication is that this QD-on-DBR system becomes a promising starting point for building deterministic monolithic telecom C-band nanophotonic devices.

Lev: They estimate that the overall offset accuracy evaluation for integrating subsequent structures would be about plus or minus sixty nanometers, meaning that mirror-assisted signal is enough to start a marker-based deterministic fabrication workflow.

Kai: That sixty nanometer uncertainty gives us a concrete number we can use when thinking about how much error we’re introducing in the next stage of building these devices.

Mira: It sets a benchmark for how well you can trust this initial signal before you move on to making more complex quantum components.

Lev: We need to see if that sixty nanometer offset is manageable when we start talking about coupling it to other elements in the circuit, like those transmon qubits we’ve been looking at.

The paper's improvements: Tom: So, we’re looking at what they suggest are the big improvements they made to this monolithic platform for locating dots and checking their quality before fabrication.

Kai: They’re saying that by using AI, they can analyze that raw two-photon interference visibility data and report a raw two photon interference visibility of fifty-one percent within a two nanosecond time window, which is pretty good for an unprocessed structure.

Mira: And the AI can benchmark the intrinsic photon coherence by comparing that raw Hong-Ou-Mandel visibility, which was about fifty-one percent for a two nanosecond range, against higher values only reported for cavity-integrated emitters with Purcell-shortened lifetimes.

Lev: That’s what I care about because it means we can actually use those measured coherence numbers to predict how well they’ll perform on real hardware where you don't have all these controlled conditions.

Kai: They also added a spectroscopic transition analysis where the AI can identify charged exciton transitions, like the one near one thousand five hundred forty-nine point five nm, that shows no resolvable fine structure splitting.

Mira: That lack of fine structure splitting is important because it confirms what they expected about the negative charged exciton state they were looking for in that specific emission line.

Lev: And then there’s optimizing the excitation scheme; the AI found that pulsed quasiresonant excitation gave a raw g(two)(zero) value of zero point zero five nine plus or minus zero point zero zero two, which is what we need to run error correction codes on.

Kai: So they’re using this AI not just for reporting numbers, but for actually optimizing the setup to get the cleanest possible quantum source from the start.

Mira: The implication of these improvements is that this platform isn't just a static measurement tool anymore; it’s becoming an active system that can self-optimize its own quality control.

Lev: If we can reliably use AI to benchmark coherence against cavity-enhanced systems, it opens up a whole new way to characterize the noise floor of these planar structures without needing those bulky cavities upfront.

Kai: It shows they’re thinking about how this raw signal is going to feed into larger quantum circuits where you need high fidelity and low error rates.

Mira: So, the takeaway is that this method moves beyond just measuring what's there; it’s about using computational analysis to understand the underlying physical assumptions of the emitters themselves.

Conclusion: Kai: So we’re wrapping up on "Broadband emitter localization and prefabrication two-photon interference from InAs/InP quantum dots monolithically grown on an InGaAsP/InP distributed Bragg reflector."

Mira: We established that this platform manages to integrate the extraction, the quality check, and the monolithic growth all into one epitaxial structure.

Lev: From my side, it’s a solid foundation because if we can get that first marker-based deterministic workflow running, we can actually start building something physical.

Kai: They showed that this mirror-assisted signal is sufficient to initiate a marker-based deterministic fabrication workflow with an offset accuracy evaluation of plus or minus sixty nanometers.

Mira: That means the mirror isn't just a passive part; it’s an active component that gives us the necessary spatial guidance for subsequent processing steps.

Lev: That sixty nanometer uncertainty is a concrete metric we can use when we start thinking about how much error we’re introducing in the next stage of integrating these structures, which is vital for error correction overhead.

Kai: It demonstrates that even without cavity-induced enhancement, this unprocessed planar structure retains enough coherence to be useful before you start adding more complexity to it.

Mira: Ultimately, this work sets up a promising starting platform for deterministic monolithic telecom C-band nanophotonic devices by meeting those sequential requirements in a single epitaxial layer.

Lev: It’s about proving that the material science and the optical engineering can actually coexist in one growth process without destroying the quantum properties.

Kai: So, this paper shows we can localize emitters with high precision and still keep them high quality for deterministic fabrication.

Mira: The implication is that we’re getting closer to a real, integrated source that doesn't require external markers for every single operation.

Lev: Next up, I want to talk about how this kind of deterministic localization can be applied when we start moving toward more complex photonic circuits where the error budget gets really tight.

Paweł Mrowiński, Jan Śmigiel, Mikołaj Badura, Beata Ściana, Wojciech Kijaszek, Emilia Zięba-Ostój, Paweł Holewa, Marcin Syperek

Department of Experimental Physics, Faculty of Fundamental Problems of Technology, Wrocław University of Science and Technology · Faculty of Electronics, Photonics and Microsystems, Wrocław University of Science and Technology

cond-mat.mes-hall

Submitted: 2026-10-22

Updated: 2026-10-22

License: http://creativecommons.org/licenses/by/4.0/

The gist: The gist: A monolithic telecom C-band single-photon platform has been demonstrated, utilizing low-density InAs/InP quantum dots (QDs) grown on a 20.5-pair InGaAsP/InP distributed Bragg reflector.

Key concepts

Distributed Bragg Reflector (DBR): A DBR is a layered structure of alternating InGaAsP and InP materials designed to reflect light efficiently. In this platform, the DBR acts as a bottom mirror, redirecting emitted photons toward the substrate to enhance upward photon collection, which is crucial for identifying individual quantum dots during preselection.
Photon Extraction Efficiency (PEE): PEE measures how effectively photons are collected from the quantum dot structure and directed out of the material. The platform achieved a calculated maximum PEE of about 13% at 1573 nm, which is four times higher than what is typically seen in bulk InP emitters, improving the efficiency of signal collection.
Hong-Ou-Mandel (HOM) Interference: HOM interference is a technique used to benchmark the quality and coherence of single photons. By sending two consecutive photons into a beam splitter, researchers measured their raw visibility. A high visibility (up to 51%) indicates that the photons are highly indistinguishable and maintain good quantum properties before further processing.
Statistical Centroid Localization: This refers to the ability to precisely locate where a single photon is emitted within a large array of emitters. Using broadband photoluminescence imaging, this platform achieved a statistical centroid-localization precision of approximately 3 nm for the investigated quantum dots, which is essential for accurate emitter identification.

Terminology

Localization — 1× in this paper · explained in 2 episodes

Summary

The gist: A monolithic telecom C-band single-photon platform has been demonstrated, utilizing low-density InAs/InP quantum dots (QDs) grown on a 20.5-pair InGaAsP/InP distributed Bragg reflector.

Platform and Architecture

The work demonstrates a monolithic telecom C-band single-photon platform based on low-density InAs/InP quantum dots grown on an InGaAsP/InP distributed Bragg reflector The QD position and upper InP spacer are engineered for broadband upward photon extraction without intentional Purcell enhancement The calculated photon extraction efficiency is approximately 13%, four times higher than that of an equivalent emitter in bulk-like InP This enhancement enables wide-field PL imaging of individual C-band emitters with a statistical centroidlocalization precision of approximately 3 nm for the investigated QD.

Photon Extraction and Localization

Efficient photon extraction is achieved through a bottom mirror such as distributed Bragg reflectors (DBRs) The DBR redirects emission propagating toward the substrate into the upper half-space, producing a calculated maximum upward photon extraction efficiency (PEE) of roughly 13% at 1573 nm, around four times the value obtained for an equivalent dipole in bulk-like InP. Once the broadband photon extraction window is established, it is used to preselect individual C-band-emitting QDs Emitter preselection was performed in two stages, first using a long-pass filter (λ > 1500 nm) to identify a sample region with a limited number of optically active QDs. The spectral window was then narrowed using approximately 12 nm-wide band-pass filters centered between 1550 and 1610 nm.

Photon Coherence Benchmarking

Hong-Ou–Mandel (HOM) interference is employed to benchmark the photon quality retained by the unprocessed planar structure Consecutive photons display a raw two-photons interference visibility of approximately 41% over the full wavepacket, increasing to 51% within a 2 ns integration window. To the best of our knowledge, this result is among the highest raw full-wavepacket HOM visibilities reported for an unprocessed planar InAs/InP C-band QD platform without cavity-induced radiative rate enhancement. The raw integrated TPI visibility as a function of time window shows approximately 51% TPI visibility for a 2 ns time range, decreasing to 41% for time ranges longer than 4 ns.

Spectroscopic Characterization

The QD selected by wide-field PL imaging was first characterized by power- and polarization-resolved highspatial-resolution photoluminescence (µPL) under continuous-wave (CW) 805 nm above-bandgap excitation. The most intense line at 1549.5 nm exhibits approximately linear power dependence and no resolvable FSS. Under the (iii) excitation scheme, the resulting autocorrelation histogram yields a raw g(2)(0) = 0.059 ± 0.002. The raw two-photons interference visibility is approximately 41% over the entire detected wavepacket.

Conclusion and Significance

The significance of this platform resides in its integration of three sequential requirements within a single epitaxial structure: sufficient broadband extraction for emitter identification, preservation of high photon quality prior to processing, and compatibility with subsequent monolithic nanofabrication. The DBR enhances upward photon collection during emitter preselection while remaining an integral part of the wafer for subsequent deterministic nanofabrication. This work establishes our monolithic QD-on-DBR system as a promising starting platform for deterministic monolithic telecom C-band nanophotonic devices. The overall estimated offset accuracy evaluation for integration of subsequent nanophotonic structures would be ±60 nm. This demonstrates that the mirror-assisted signal is sufficient to initiate a marker-based deterministic fabrication workflow.

ACKNOWLEDGEMENTS

This work was supported by the National Science Centre, Poland, under Projects No. 2023/50/E/ST7/00581 and 2025/57/B/ST7/04916 (TSUNAMI) P.H. acknowledges support from the Polish National Agency for Academic Exchange under the Polish Returns programme The authors thank Paweł Wyborski for advice on polarization control in the two-photon-interference measurements, Elizaveta Semenova for support with epitaxial growth, and Jarosław Serafińczuk for X-ray diffraction measurements.

REFERENCES

[1] N. Somaschi et al., Near-optimal single-photon sources in the solid state, Nat. Photonics 10, 340 (2016)

[2] M. Schwartz, E. Schmidt, U. Rengstl, F. Hornung, S. Hepp, S. L. Portalupi, K. Llin, M. Jetter, M. Siegel, and P. Michler, Fully On-Chip Single-Photon Hanbury-Brown and Twiss Experiment on a Monolithic Semiconductor-Superconductor Platform, Nano Lett. 18, 6892 (2018)

[3] R. Uppu et al., On-chip deterministic operation of quantum dots in dual-mode waveguides for a plug-and-play single-photon source, Nat. Commun. 11, 3782 (2020)

[4] A. Strittmatter et al., Site‐controlled quantum dot growth on buried oxide stressor layers, Physica Status Solidi (a) 209, 2411 (2012)

[5] M. Podhorský, M. Klonz, L. Böhmer, S. Kulig, C. C. Palekar, P. Klenovský, S. Rodt, and S. Reitzenstein, Buried Stressor Engineering for Position-Controlled InGaAs Quantum Dots with Local Density Variation for Integrated Quantum Photonics, ACS Photonics 13, 471 (2026)

[6] E. Schöll et al., Resonance Fluorescence of GaAs Quantum Dots with Near-Unity Photon Indistinguishability, Nano Lett. 19, 2404 (2019)

[7] J. Liu et al., A solid-state source of strongly entangled photon pairs with high brightness and indistinguishability, Nat. Nanotechnol. 14, 586 (2019)

[8] S. F. C. da Silva, G. Undeutsch, B. Lehner, S. Manna, T. Manna, Klonz, L., Böhmer, S., Kulig, C., Palekar, C., Klenovský P., Rodt S., and Reitzenstein S., GaAs quantum dots grown by droplet etching epitaxy as quantum light sources, Appl. Phys. Lett. 119, (2021)

[9] P. Holewa, S. Kadkhodazadeh, M. Gawełczyk, P. Baluta, A. J., Musiał V., G. Dubrovskii V., M. Syperek M., and E. Semenova E., Droplet epitaxy symmetric InAs/InP quantum dots for quantum emission in the third telecom window: morphology, optical and electronic properties, Nanophotonics 11, 1515 (2022)

[10] B. Fain et al.

Improvements for AI systems

Broadened spectral filtering for emitter preselection: The system can perform wide-field photoluminescence imaging with a statistical centroid-localization precision of approximately 3 nm due to improved signal-to-noise ratio achieved through broadband extraction.
Enhanced single photon characterization: The AI system can analyze raw two photon interference visibility, reporting a raw two photon interference visibility of 51% within a 2 ns time window and 41% over a 4 ns integration window for the unprocessed planar structure.
Coherence benchmarking capability: The AI can benchmark intrinsic photon coherence by comparing the raw Hong-Ou-Mandel visibility, which showed approximately 51% TPI visibility for a 2 ns time range, against higher values reported only for cavity-integrated emitters with Purcell-shortened lifetimes.
Spectral transition analysis: The system can characterize charged exciton transitions by identifying a line near 1549.5 nm exhibiting no resolvable fine structure splitting, consistent with the negative charged exciton (X-) state.
Excitation scheme optimization: The AI can determine the most favorable excitation scheme for quantum statistics by evaluating raw g(2)(0) values, finding that pulsed quasiresonant excitation yielded a raw g(2)(0) value of 0.059 (2).

Abstract

We demonstrate a monolithic telecom C-band single-photon platform based on low-density InAs/InP quantum dots grown on an InGaAsP/InP distributed Bragg reflector, with simulated upward light-extraction efficiency exceeding 10% within the 40 nm spectral band. Enhanced light extraction is crucial for wide-field photoluminescence imaging, enabling approximately 3 nm statistical centroid-localization precision of QD emitters due to improved signal-to-noise ratio. The identified charged exciton emitting at 1549.5 nm exhibits single-photon emission purity up to 94.7%, validated by a raw peak-area integration under pulsed quasi-resonant excitation. Under these conditions, photons emitted sequentially from the unprocessed planar structure reveal a raw two-photon interference visibility of 51% of the photon wave packet within the 2 ns time window and 41% over a 4 ns integration window, all without Purcell enhancement typical for cavity-embedded emitters. Combining broadband extraction, high photon coherence, and emitter preselection on a native III-V wafer, this architecture paves the way for marker-based deterministic fabrication of high-quality monolithic nanophotonic devices.