PaperlayerView exact revision on arXiv

Source-linked AI summary

Photon temporal modes: a complete framework for quantum information science

B. Brecht, Dileep V. Reddy, C. Silberhorn, M. G. Raymer

arXiv:1504.06251v4quant-ph

TL;DR

Photonic quantum information needs controlled resource-state generation, targeted manipulation, and efficient detection, while temporal-mode encoding had not yet been demonstrated as a viable basis. The paper develops a complete TM framework with arbitrary-dimensional state generation, tomography, quantum pulse-gate operations, and proposed QIS applications. It concludes that TM-based QIS is feasible with current technology, while practical performance remains bounded by mode-structure, memory, bandwidth, and gate-fidelity limitations.

  • Problem

    Temporal modes had not yet been demonstrated as a viable quantum-information basis, and photonic QIS requires controlled resource generation, quantum operations, and efficient detection.

  • Method

    The paper combines dispersion-engineered parametric down-conversion, shaped pump pulses, quantum pulse gates, TM tomography, and TM-based operations to construct a complete framework.

  • Results

    TM sorting achieves efficiency and selectivity exceeding 99.5 percent, while the framework supports arbitrary-dimensional entangled resources, tomography, and proposed photonic QIS applications.

  • Takeaways & Limitations

    Temporal modes provide a practical high-dimensional photonic encoding framework compatible with single-mode fiber networks and a range of quantum-information applications.

  • Takeaways & Limitations

    Two-stage quantum pulse gates reach about 95.4% maximum gate fidelity, and increasing selectivity trades against total internal gate losses.

Abstract

from arXiv · show

Field-orthogonal temporal modes of photonic quantum states provide a new framework for quantum information science (QIS). They intrinsically span a high-dimensional Hilbert space and lend themselves to integration into existing single-mode fiber communication networks. We show that the three main requirements to construct a valid framework for QIS -- the controlled generation of resource states, the targeted and highly efficient manipulation of temporal modes and their efficient detection -- can be fulfilled with current technology. We suggest implementations of diverse QIS applications based on this complete set of building blocks.

INTRODUCTION

The paper introduces temporal modes as a practical basis for photonic quantum information, addressing generation, manipulation, and detection requirements with current technology. TMs provide field-orthogonal wave-packet states that occupy a complete energy-domain basis and can integrate with single-mode fiber networks.

  • Framework requirements: The framework targets three requirements for photonic quantum networks: reliable resource-state generation, controlled quantum operations, and efficient detection.These requirements organize the paper’s proposed building blocks for photonic subsystems.
  • TM manipulation: Quantum pulse gates enable TM sorting with efficiency and selectivity exceeding 99.5 percent through dispersion-engineered, multistage frequency conversion.The conversion is driven by spectrally and temporally shaped laser control pulses.
  • Applications: The paper proposes flexible generation of arbitrary-dimensional entangled resource states, TM tomography, and implementations of photonic co-processors and quantum communication applications.The authors identify boson sampling as one application and suggest reduced switching losses for larger networks.
  • Temporal modes: Temporal modes are field-orthogonal broadband wave-packet states that form a complete basis for single-photon states in the energy degree of freedom.They overlap in time and frequency while remaining orthogonal under frequency or time integration.
  • Integration advantages: Temporal modes share one spatial field distribution, supporting waveguide devices and existing single-mode fiber networks while remaining insensitive to stationary or slowly varying medium perturbations.Their overlapping spectra underlie the stated robustness to linear dispersion.
  • Formal basis: TM states can be represented as coherent superpositions of monochromatic or creation-time components and expanded in a basis with complex coefficients.The paper illustrates the first three basis members using Hermite-Gaussian functions in frequency and time.

Quantum information encoding with TMs

Temporal modes offer a high-dimensional encoding basis for photonic quantum information. The paper defines TM qudits as coherent superpositions of TM states and relates TM qubits to orthogonal modes and their superpositions.

  • High-dimensional encoding: Temporal modes span an infinite-dimensional Hilbert space, potentially increasing information capacity per photon and security compared with two-dimensional encoding.Information carriers in a d-dimensional Hilbert space are called qudits.
  • TM qudits: A TM qudit is a coherent superposition of d temporal-mode states.This provides the stated d-dimensional encoding construction.
  • TM qubits: A TM qubit uses two orthogonal temporal modes as logical states, with coherent superpositions represented on a Bloch sphere.The paper illustrates the basis using zeroth- and first-order Hermite-Gaussian frequency modes.

Mutually unbiased bases

Mutually unbiased bases provide complementary measurement bases for TM qubits, while PDC decomposes photon-pair states into correlated temporal-mode pairs. Dispersion engineering can isolate a single TM pair, but that state is not an entangled resource and general PDC does not precisely control dimensionality.

  • Mutually unbiased bases: The three MUBs for a TM qubit use zeroth- and first-order Hermite-Gaussian temporal-mode shapes, with spectral amplitudes and phases shown separately.The qubit is encoded in the leftmost basis.
  • Mutually unbiased bases: MUBs have equal cross-basis overlaps, so measuring an eigenstate in another MUB yields uniformly random results.They support applications including quantum key distribution and quantum state tomography.
  • TM structure of photon pair states: A PDC joint spectral amplitude is decomposed into two uniquely defined signal and idler TM bases with pairwise correlations and weights √λk.The JSA combines the pump-envelope and phase-matching functions, with normalized weights satisfying Σk λk = 1.
  • TM structure of photon pair states: Dispersion-engineered PDC can excite one TM pair, yielding a known pure heralded single-photon TM state.The JSA has no signal-idler correlations and only one nonzero weighting coefficient.
  • TM structure of photon pair states: General PDC lacks precise control over the number of temporal modes, while a single-TM state is not an entangled resource for QIS applications.Thus, neither case alone supplies the required resource-state structure.

single-photon states

Quantum pulse gates coherently select, convert, and reshape temporal modes, enabling mode-selective manipulation of single-photon states. A two-stage interferometric design addresses the single-stage selectivity limit caused by time ordering.

  • single-photon states: A QPG uses a shaped pump pulse to select a targeted TM and convert it to an output state, while non-target modes remain transmitted.The conversion efficiency is η = sin^2(θi), and superpositions can also be selected with appropriate pulse shaping.
  • single-photon states: QPGs can convert arbitrary targeted TMs into one common output state or reshape that output state into arbitrary TMs.The common output enables interference between formerly orthogonal modes, while the output band can serve as a processing space.
  • single-photon states: Temporal mode selectivity measures the selected mode’s squared conversion efficiency relative to the conversion efficiencies of all modes.A selectivity of 1 denotes perfect single-TM operation, whereas 0 denotes no modal selectivity.
  • single-photon states: A single-stage QPG is limited to S = 0.85 because time ordering causes temporal multimode behavior above 90% conversion efficiency.A two-stage Mach-Zehnder/Ramsey-like setup with 50% efficiency per stage overcomes this limitation.
  • single-photon states: An experiment demonstrated 80% TM selectivity at η = 87% conversion efficiency for a single-stage QPG operated at the single-photon level.Alternative approaches may be simpler experimentally but cannot generally reach high mode selectivities.

COMPLETING THE TOOL KIT FOR A TM QIS FRAMEWORK

The completed TM toolkit combines controlled resource-state generation with temporal-mode operations for photonic QIS. Shaped dispersion-engineered PDC produces Bell states and higher-dimensional states with user-controlled dimensionality, while ideal QPGs support linear-optical operations.

  • COMPLETING THE TOOL KIT FOR A TM QIS FRAMEWORK: The framework adds arbitrary-dimension TM-state generation and single-photon and photon-pair TM tomography as missing components of the toolkit.These components are presented as enabling a complete TM framework for QIS.
  • COMPLETING THE TOOL KIT FOR A TM QIS FRAMEWORK: Ideal QPGs can implement linear-optics operations on single photons and photon pairs.This extends the toolkit from state generation and verification to quantum operations.
  • COMPLETING THE TOOL KIT FOR A TM QIS FRAMEWORK: A first-order Hermite-Gaussian pump drives dispersion-engineered PDC to produce exactly two equally weighted TM pairs, forming a TM Bell state.The Bell state is identified as a fundamental resource for QIS applications.
  • COMPLETING THE TOOL KIT FOR A TM QIS FRAMEWORK: A second-order Hermite-Gaussian pump produces exactly three TM pairs with unequal relative weights but well-defined dimensionality.Increasing the pump’s Hermite-Gaussian order successively adds further TM pairs.
  • COMPLETING THE TOOL KIT FOR A TM QIS FRAMEWORK: Higher-order shaped pumps generate high-dimensional photonic states with unprecedented control while keeping all temporal modes in one transverse spatial waveguide mode.The shared spatial mode supports experimental simplicity and robustness.

Photon TM-state tomography

TM tomography uses programmable QPG measurements and detected transmitted or converted outputs to recover density-matrix elements for single-photon and biphoton states.

  • Photon TM-state tomography: TM tomography retrieves complex density-matrix entries, addressing verification of arbitrary-dimensional TM state generation.The higher-dimensional TM space distinguishes this procedure from polarization-state tomography.
  • Photon TM-state tomography: A QPG selects a coherent superposition of two TMs, while transmitted and converted count rates provide measurement data.The selected superposition is programmed through the QPG pump-pulse shape.
  • Photon TM-state tomography: Setting ζ = 0 or ζ = 1 directly measures diagonal coefficients, while ζ = 1/√2 with φ = 0 and φ = π/2 retrieves complex off-diagonal coefficients.These settings allow complete density-matrix reconstruction or sampling of an experimentally feasible subset.
  • Photon TM-state tomography: Any chosen density-matrix portion can be directly measured without reconstructing the entire state when the QPG achieves unit selectivity.Without high selectivity, this direct-measurement advantage is not fully available.
  • Photon TM-state tomography: Biphoton tomography sends the two photons through separate QPGs and uses coincidence rates between transmitted and converted detector outputs.Cycling through the QPG parameter space reveals desired two-photon density-matrix coefficients.
  • Photon TM-state tomography: Non-mode-selective detection generally leaves a sibling photon in a mixed state, whereas QPG-based TM heralding can produce a pure state at reduced heralding rate.This illustrates the role of mode-selective detection in preparing and characterizing photonic states.

QIS APPLICATIONS

The paper combines its TM building blocks into applications spanning photon purification and reshaping, quantum communication, gate operations, and cluster-state generation.

  • QIS APPLICATIONS: The framework is applied to photon TM purification, TM reshaping, quantum communication, single-qubit gate operations, and cluster-state generation.These applications are presented to highlight the versatility of the TM framework.

Photon TM “purification”

QPG-based TM selection purifies heralded photons from general PDC states and enables coherent reshaping for interfaces between photons and quantum memories.

  • Photon TM “purification”: Spectrally narrow intensity filtering is commonly used to obtain approximately pure heralded photons from general PDC states, discarding much of the generated light.This setting arises when a source cannot be engineered to generate only a single TM pair.
  • Photon TM “purification”: A QPG can select a single TM in one photon and, upon converted-output detection, herald its partner in a pure state.The QPG acts as a complex spectral-amplitude shape filter with efficiency η.
  • Photon TM “purification”: The purification lowers the heralding rate by the factor λ_i, but produces a photon in a desired TM rather than merely a spectrally filtered photon.The trade-off is between state selectivity and heralding rate.
  • Photon TM “purification”: The same TM framework can support TM-selective photon subtraction from continuous-variable multimode states for entanglement distillation.This operation uses a QPG operated at intentionally low conversion efficiency.
  • Photon TM “purification”: Two QPGs can coherently reshape a photon, and tailoring the second QPG’s phasematching function enables arbitrary reshaping in principle.The approach can adapt interfaces between telecommunication-wavelength photons and specific quantum memories with a single QPG.

Quantum communication

Temporal modes support multiplexed and genuinely high-dimensional quantum communication, including generalized BB84, while QPGs provide channel add/drop and readout operations.

  • Quantum communication: TM multiplexing uses distinct orthogonal temporal modes as independent channels transmitted through one single-mode fiber.Alice adds channels with QPGs, while Bob demultiplexes them with cascaded QPGs for readout.
  • Quantum communication: TM communication can encode information directly in arbitrary superpositions of single-photon TMs, enabling genuinely high-dimensional quantum communication.A separate multiplexing approach uses TMs as channels while encoding information in another degree of freedom, such as polarization.
  • Quantum communication: Unlike conventional time- or frequency-based multiplexing, field-orthogonal TMs can provide in-principle zero crosstalk while densely packing channels in time-frequency phase space.Conventional schemes use separated pulses or narrow spectral windows and therefore lower packing density to maintain approximate orthogonality.
  • Quantum communication: Generalized BB84 in four dimensions uses 20 possible basis states across five mutually unbiased bases, with Bob requiring three QPGs to resolve each basis.More generally, resolving a basis of size d requires d − 1 QPGs.
  • Quantum communication: High-dimensional encoding increases information capacity per photon and can increase QKD security because interception introduces a larger error.These are identified as advantages of high-dimensional communication schemes.
  • Quantum communication: The paper also considers TM qubits in linear optical quantum computation and graph-state generation through cluster-state quantum computation.These routes use single- and two-qubit operations or fuse multiple TM qubits into tailored entanglement structures.

LOQC

Temporal-mode qubits support arbitrary single-qubit gates through sequential quantum pulse gates (QPGs), phase shifts, and mode-selective operations. The same framework extends to single-qudit operations by acting successively on two-dimensional subspaces.

  • Single-qubit gates: 100% and 50% conversion-efficiency QPGs, combined with channel-dependent phase shifts, implement arbitrary single-qubit operations on the {|A0⟩, |A1⟩} space.The phase shift acts only in the internally used green processing space, while the input and output remain red temporal modes.
  • Single-qubit gates: Pauli-X and Pauli-Y gates each have two experimental implementations differing in the order of addressing the two logical red temporal modes.The implementations use the pump shapes associated with the logical “0” and “1” modes.
  • Implementation: The phase-shift gate can be simplified by imprinting the phase (φ + π) onto one pump pulse, eliminating the channel-dependent phase shift.This retains the sequential QPG-based implementation while reducing its required operation set.
  • Implementation: Internal green-channel processing reduces phase-maintenance demands between different red-channel frequency bands and supports compact monolithic implementations.The sequential components are described as suitable building blocks for integrated networks.
  • Single-qudit operations: Any single-qudit operation can be realized by concatenating single-qubit operations across successive two-dimensional subspaces.The pump shapes must be selected so that each relevant qubit subspace is addressed in turn.

Cluster state quantum computation

Temporal-mode QPGs provide the operations needed to build cluster states, including local Hadamard transformations, projective measurements, and linear-cluster fusion. A QPG-based fusion circuit heralds success through single-photon detection and returns a qubit state that can be combined into one spatial mode.

  • Cluster-state construction: Cluster-state construction from Bell pairs requires local Hadamard transformations, projective measurements, and operations for merging linear clusters into two-dimensional clusters.The paper identifies the first two operations as already implementable with temporal modes and emphasizes linear-cluster generation as the remaining requirement.
  • Qubit fusion: Two spatially separated temporal-mode qubits are sent to QPGs that selectively convert chosen red modes into green modes.The QPGs apply complementary mode-selective operations to beams a and b before interference.
  • Qubit fusion: The converted green channels interfere at a balanced beamsplitter, erasing distinguishing information before detection at two output ports.The beamsplitter and detectors form the heralding stage of the fusion operation.
  • Qubit fusion: A single detected green photon heralds successful qubit fusion, with the detector port determining the sign in the resulting Kraus operator.The postselected state is again a qubit state.
  • Qubit fusion: The two parts of the fused qubit can be deterministically combined into one spatial mode using the QPG’s add/drop functionality.This provides a mode-combination step after successful fusion.

CHALLENGES

Temporal-mode QIS is feasible but faces limits from timing, bandwidth, memory multimode capacity, optical loss, accessible mode number, switching speed, and gate fidelity. The paper also identifies integration and detection requirements that remain technically demanding.

  • Bandwidth and interfaces: 9.7 GHz phasematching bandwidth is calculated for a 90 mm lithium-niobate waveguide, close to the 9.2 GHz bandwidth of a cited Raman memory.The QPG phasematching bandwidth determines the minimal bandwidth of the reshaped temporal mode.
  • Quantum-memory interfaces: A Raman memory cited in the paper stores only a single temporal mode, although concatenated Raman memories have been shown to store high-dimensional temporal-mode states.Multimode memory capability is therefore a complication for flying-to-stationary qudit interfaces rather than an established universal capability.
  • Loss budget: Coupling a fiber to a 90 mm QPG and back is estimated to incur roughly 1.0 dB insertion loss, with fiber couplings as the main loss source.Waveguide-to-fiber coupling above 92% and propagation losses as low as 0.016 dB/cm are cited, but current total losses remain prohibitively high.
  • Accessible dimensionality: 10 temporal modes can be addressed above 95% selectivity in one modeled realization, increasing to 20 after simple bandwidth and waveguide-length optimization.The bound arises from increasing mode complexity and spectral extent, plus QPG bandwidth and pulse-shaper resolution constraints.
  • Pulse shaping: Commercial 4096-pixel spatial light modulators can shape temporal modes of order 100 with fidelity above 99.9%, but switching speeds of a few tens of kHz limit application speed.The pulse-shaper constraint is therefore accompanied by an electronic switching bottleneck.
  • Gate fidelity: 98.46% maximum selectivity in a two-stage QPG corresponds to around 95.4% maximum gate fidelity when each LOQC gate uses three QPGs.The paper states that this does not meet fault-tolerant LOQC requirements but may support small co-processing operations.
  • Synchronization: Temporal-mode orthogonality degrades under timing jitter, making synchronization of active components more demanding than in some other encoding bases.The paper proposes weak coherent pilot pulses as a strategy for long-distance timing synchronization.
  • System-level challenges: Temporal modes offer single-mode-fiber compatibility and robustness to linear dispersion, but practical QIS still requires integrated fabrication, timing electronics, efficient detection, and low-loss programmable routing.The paper presents these as shared challenges for emerging optical QIS frameworks, alongside the need to exploit networkability and higher dimensionality.
Loading 1504.06251v4…