Skip to content
Heliosync
Company
About What Heliosync is, and what it has built Leadership Who runs the company Careers Open applications Registration CIN, Udyam, DPIIT, registered office
Support
Cosmos Horizon

Research · Photonics and quantum communication

How to count single photons at room temperature, and how to keep a quantum key secret in open air.

Those are the two questions. This is a programme, not a body of results: we have reviewed the literature, written a test plan, and proposed an architecture. Nothing has been fabricated or measured yet, and we will publish results when there are some.

Literature reviews
3

internal, not peer-reviewed

Experiments planned
1

48 conditions, not yet run

Architectures proposed
1

source, circuit, detector

Results published
0

we will say when that changes

§ 01

Counting photons without cooling

Quantum communication sends information one photon at a time, mostly at telecom wavelengths where optical fibre loses least. The best single-photon detectors work there only when cooled. A detector is blind to light past its material's cutoff, so the choice of material decides what it can count at all.

Absorption cutoffs: CsPbBr3 near 539 nm, MAPbI3 near 775 nm, silicon near 1107 nm and germanium near 1879 nm, against the telecom O-band at 1260 to 1360 nm and C-band at 1530 to 1565 nm. Silicon cannot see either telecom band.O-bandC-bandCsPbBr₃ (perovskite)539 nm · 2.30 eVMAPbI₃ (perovskite)775 nm · 1.60 eVSilicon1,107 nm · 1.12 eVGermanium1,879 nm · 0.66 eV400600800100012001400160018002000wavelength, nm →
Fig. 1.1 What each material can see Literature A semiconductor absorbs only photons with more energy than its band gap, so it is blind beyond λc = hc / Eg ≈ 1240 nm·eV / Eg. Band gaps are approximate room-temperature values. Silicon stops near 1107 nm and cannot see either telecom band; germanium's line uses its indirect gap, and its absorption fades well before that (its direct gap, about 0.80 eV, sits near 1550 nm). The two perovskites absorb in the visible, which is why our proposal uses them as a photon source and pairs them with plasmonic structures to detect.
Two-slit interference: a bright central fringe and fringes falling away under a single-slit envelope, drawn as a curve and as the pattern seen on a screen.screen position x →intensity I(x), normalised
Fig. 1.2 Interference is the computation Literature I(x) ∝ cos²(πdx/λL) · sinc²(πax/λL), two-slit interference after Young (1807), computed for d ≈ 4.4a. On a photonic chip the same effect, inside a Mach–Zehnder interferometer, is what turns paths of light into logic.

What the reviews concluded

Our reviews looked at room-temperature options: silicon and germanium–silicon avalanche photodiodes, perovskite quantum dots, plasmonic enhancement, and two-dimensional materials.

One finding shaped everything after it: for graphene and twisted two-dimensional materials, single-photon detection has so far been shown only at cryogenic temperatures. Room temperature is the open part of the problem, and the reason for the experiment below.

§ 02

Keys made of light, in weather

Quantum key distribution lets two parties agree a secret key whose security rests on physics: measuring a photon disturbs it, so an eavesdropper leaves errors behind. Over a free-space link the enemy is also the sky. Fog, haze and rain scatter the photons the key is made of.

Ten photons of a BB84 exchange. In the six positions where Alice and Bob chose the same basis, Bob reads Alice’s bit and they keep it; the shared key is 0 0 1 0 1 0. The other four are random and discarded.0++0yes1×+0?—1+×1?—0++0yes1××1yes0×+1?—0++0yes1××1yes1+×0?—0××0yesAlice’s bitAlice’s basisPhoton sentBob’s basisBob readsKept?Shared key: 0 0 1 0 1 0? = random: the bases differed
Fig. 2.1 BB84, ten photons Conceptual The protocol of Bennett and Brassard (1984). Alice encodes each bit as a polarisation in one of two bases, + or ×; Bob measures in a basis of his own. Afterwards they compare bases, never bits, in public, and keep only the positions where they matched. They then check a sample for errors; too many, and they assume someone was listening and throw the key away. An example sequence, not a run.
Attenuation in decibels per kilometre against visibility, from the Kim model, for 1550 and 850 nanometres. In clear air it is under one decibel per kilometre; in dense fog it passes a hundred, and the two wavelengths converge.0.010.11101001,000foghazeclear0.050.10.51510501550 nm850 nmvisibility V, km (log scale) →dB/km
Fig. 2.2 What fog does to a free-space link Literature The Kim model (Kim, McArthur and Korevaar, 2001): α = (3.91 / V) · (λ / 550 nm)−q per km for visibility V, with q falling to zero in dense fog. Computed here for 1550 nm and 850 nm. In clear air a link loses well under 1 dB per kilometre; in dense fog it loses hundreds, and choosing a longer wavelength no longer helps. Simulations of BB84 over such channels have been published, for example Atieh, Raytchev and Buhari (2023); that work is theirs, not ours.

§ 03

The experiment we have planned

Perovskites detect light well and degrade quickly: heat, moisture and light itself break them down. Before any perovskite detector is worth building, we want to know which combination survives. The plan is written; none of it has been run.

A planned 48-condition experiment: three perovskites, each tested under four passivations and four encapsulations. All cells are empty because none has been run.MAPbI₃CsPbBr₃2D RPnonepyridinePCBMMABrencapsulation → bare · ALD Al₂O₃ · + UV epoxy · multilayerpassivation ↓
Fig. 3.1 48 conditions, all empty Planned A full factorial design: three perovskites, four surface passivations and four encapsulations. With six device and three glass replicates per condition, 432 samples. Results will be analysed by three-way ANOVA. Every cell is empty because none has been made.

How each sample would be stressed

  • Damp heat at 85 °C and 85% relative humidity
  • Light soaking at 1 sun and 5 suns
  • Thermal cycling between −40 °C and 85 °C
  • Up to 1,000 hours under test

What counts as passing

  • At least 70% of responsivity kept after 168 hours of damp heat
  • T80 beyond 1,000 hours at 1 sun and 65 °C
  • At least 85% kept after 200 thermal cycles
  • Dark current rising less than tenfold

Gates set in advance, so the result cannot choose its own bar.

§ 04

An architecture we have proposed

If the materials hold up, the long aim is one chip that makes single photons, processes them by interference, and counts them, all at room temperature. It is a design with targets, not a device.

Proposed architecture: a perovskite quantum-dot single-photon source, then a silicon-photonic interferometer circuit, then a perovskite-plasmonic detector.Perovskitequantum-dot sourcesingle photonsSilicon-photoniccircuit (MZI)interference as gatesPerovskite–plasmonicdetectorcounts at room temp.one chip, at room temperature · proposed, not built
Fig. 4.1 Source, circuit, detector Planned A perovskite quantum-dot source of single photons, a silicon-photonic circuit of Mach–Zehnder interferometers, and a perovskite–plasmonic detector. Proposed in our own strategy report; not fabricated.

Phase 1 Planned

Targets, not results · 1–2 years

  • A single-photon source with g(2)(0) below 0.1
  • Gate fidelity above 95%
  • Detector dark counts below 100 per second per mm²

Phase 2 Planned

Targets, not results · 3–5 years

  • A small processor of 4 to 8 photonic qubits

Elsewhere at Heliosync Research

Where this meets the digital human

  • A photonic reservoir to process timing-sensitive signals such as voice prosody, proposed for Shanaya

Reading

  1. Young, T. A Course of Lectures on Natural Philosophy and the Mechanical Arts. London, 1807.
  2. Bennett, C. H. and Brassard, G. Quantum cryptography: public key distribution and coin tossing. Proceedings of the IEEE International Conference on Computers, Systems and Signal Processing, Bangalore, 1984.
  3. Kim, I. I., McArthur, B. and Korevaar, E. Comparison of laser beam propagation at 785 nm and 1550 nm in fog and haze for optical wireless communications. Proc. SPIE 4214, 2001.
  4. Atieh, A., Raytchev, M. and Buhari, A. QKD over FSO under different weather conditions using OptiSystem software. Photonics North, IEEE, 2023. doi:10.1109/PN58661.2023.10223140

Our own reviews, test plan and proposal are internal working documents and are not published here.

Have a lab, a fab or a better idea?

The experiment needs one. We would like to hear from you.

Get in touch