Katika saketi jumuishi za fotoniki, sehemu kubwa ya swichi za optiki huelekeza mwanga tu katika ndege sambamba na uso wa chipu. Kadiri idadi ya ports inavyoongezeka, muundo huu wa pande mbili unaweza kusababisha njia za kuunganisha kuwa ndefu zaidi na eneo la chipu kuongezeka. Utafiti huu unapendekeza usanifu wa swichi ya optiki ya pande tatu unaotumia birefringence ya nematic liquid crystal inayoweza kubadilishwa chini ya electric field ili kuelekeza mwanga kwa usawa ndani ya safu hiyo hiyo ya fotoniki na kwa wima kati ya safu tofauti zilizopangwa juu ya nyingine. Katika design, refractive index ya polycarbonate (PC) core ni 1.583, huku refractive index ya PMMA cladding ikiwa 1.482. Extraordinary na ordinary refractive indices za 5PCH liquid crystal iliyotumika ni \(n_e=1.5918\) na \(n_o=1.4739\) mtawalia, na thamani hizi hutoa ulinganifu wa karibu na PC na PMMA. Bila voltage, LC hubaki katika hali ya low-index inayofanana na cladding, huku electric field ikizungusha molekuli, kuongeza effective refractive index na kuunda temporary optical core. Electrode geometry hutengeneza core hii inayodhibitiwa na voltage kwa umbo la horizontal au vertical na hivyo kuruhusu mwanga kuelekezwa upya kati ya channels katika three-layer 3×3 switch architecture. Chanzo kinaripoti transition time ya 1.1 ms kwa ON na 45 ms kwa OFF chini ya drive ya takribani 5 V; hata hivyo, kuweka moja kwa moja LC thickness na material constants zilizotolewa katika chanzo kwenye equations husika za time constant hakuzalishi tena thamani hizi na kuna inconsistency ya namba ndani ya chanzo. Optical mode simulations zinaonyesha kuwa mwanga unaweza kufungiwa katika channel iliyochaguliwa; lakini kwa kuwa scattering na geometric waveguide loss hazikujumuishwa katika hesabu, real insertion loss na crosstalk bado hazijathibitishwa kwa kiasi. Utengenezaji na upimaji wa physical prototype umeainishwa kama future work. Kwa hiyo, utafiti wa sasa unatoa zaidi device architecture iliyochanganuliwa kwa namba kwa quantum switching networks kuliko kifaa cha low-voltage three-dimensional LC photonic switch kilichokamilishwa kwa majaribio.

Kwa nini optical switching inahitaji kuhamishwa kwenye dimension ya tatu?

Katika traditional planar photonic integrated circuit architectures, optical paths huundwa kwa kiasi kikubwa juu ya uso huo huo. Ingawa Mach–Zehnder interferometer, directional coupler na multimode-interference based switches zinaweza kutoa performance ya juu, kadiri idadi ya ports inavyoongezeka horizontal routing complexity na required chip area vinaweza kuongezeka. Muundo unaopendekezwa unalenga kuongeza photonic interconnect density kwa kutumia pia stacked optical layers kama routing axis.

Liquid crystal inageukaje kuwa electrical waveguide?

Nematic liquid crystals ni anisotropic; refractive index inayoonekana na mwanga hutegemea angle kati ya molecules na optical axis. Katika utafiti, sifa hii inatumika ili eneo moja la LC lifanye kazi kama cladding au core kulingana na voltage state.

Voltage ikiwa imezimwa, molecules hujipanga kando ya alignment layer na optical region hutenda takribani kwa \(n_o\). Electric field inapowashwa, molecules huzunguka kuelekea field direction na kuongeza refractive index inayoonekana na TM mode kuelekea \(n_e\). Hivyo, ndani ya LC block ambayo kimwili ni stationary, waveguide segment inayoweza kuwashwa na kuzimwa kwa umeme huundwa.

3×3 architecture inayopendekezwa imepangwaje?

Kuna channels tisa kwa jumla katika optical layers tatu. Katika layer iliyo karibu zaidi na silicon substrate ziko waveguides 1, 4 na 7; katika layer ya pili 2, 5 na 8; na katika layer ya tatu 3, 6 na 9.

Horizontal switching inafanywaje?

Kulingana na source text na Figure 2, Area A ni horizontal switching region. Stepped electrode structure huunda electric field profile na high-index region ya LC huchukua umbo la optical path. Mifano ya routing ni 3→6 na 3→9. Transition angle ya takribani 15° hutumiwa kupunguza mabadiliko makali ya optical direction.

Vertical switching inafanywaje?

Katika Area B, electrodes hupangwa juu na chini ya liquid crystal. Kutumia electric field kwa mwelekeo wa mteremko hubadilisha LC director profile kuwa core inayofanya smooth transition kati ya layers. Chanzo kinatoa routing examples za 3→2 na 3→1 na hutumia vertical slope ya takribani 11°.

Katika Figure 1 caption, herufi A/B za horizontal na vertical regions zimeandikwa kinyume na main method text. Kwa kuwa Figure 2 na maelezo yanayofuata yanaunga mkono A kuwa horizontal na B kuwa vertical, Verianla inatumia ulinganifu huo.

LC cell imeundwa na layers gani?

Muundo huu unatoa total thickness ya takribani 2 µm, huku katika switching equations LC layer thickness ikitumika kama \(d=1.7\ \mu m\).

Physical properties za 5PCH

Ni polarization gani inapaswa kutumika?

Design imefanywa kwa TM mode. Ili effective waveguide iundwe chini ya voltage, mwanga lazima u-couple na extraordinary refractive index \(n_e\). Kwa hiyo kifaa si polarization-independent katika hali yake ya asili. Kwa unpolarized au mixed-polarization inputs, photonic components za ziada kama polarization splitter–rotator zinahitajika.

Mbinu na Matokeo

Uhusiano wa voltage–refractive index

Figure 3 inaonyesha usambazaji wa LC director angle kati ya electrodes. Wakati 5 V inatumika, molecular angle katika central region hufikia takribani 90°. Katika Figure 4, average refractive index iliyohesabiwa kwa 1550 nm huongezeka pamoja na voltage.

Source note kuhusu average refractive index equation

Chanzo kinafafanua \(\langle n_{LC}\rangle\) katika Equation (1) kwa kutumia integral kupitia cell thickness. Katika equation iliyochapishwa, expected \(1/d\) normalization factor kwa average haionekani. Figure 4 hutumia matokeo kama dimensionless refractive index. Kwa hiyo kuna uwezekano wa normalization omission katika typesetting ya Equation (1); Verianla haisahihishi formula hii kimya kimya nje ya chanzo.

Switching time model

Chanzo kinafafanua OFF time constant kama:

\[ \tau_{OFF} = \frac{\gamma d^2} {K_{11}\pi^2} \tag{2} \]

na ON time constant kama:

\[ \tau_{ON} = \frac{\tau_{OFF}} {\left(V/V_{th}\right)^2-1} \tag{3} \]

.

Temporal response ya molecular orientation inamodeliwa na:

\[ \theta_{ON}(t) = \frac{\pi}{2} \left(1-e^{-t/\tau_{ON}}\right) \tag{4} \]

na:

\[ \theta_{OFF}(t) = \frac{\pi}{2}e^{-t/\tau_{OFF}} \tag{5} \]

.

Reproducibility problem katika switching-time numbers

Thamani za \(\gamma=0.1507\ {\rm Pa\,s}\), \(d=1.7\ \mu m\) na \(K_{11}=9.6\times10^{-12}\ {\rm N}\) zilizotolewa katika chanzo zikitumika moja kwa moja kwenye Equation (2), \(\tau_{OFF}\) hutoka takribani 4.60 ms; kwa 5 V na \(V_{th}=1\) V, Equation (3) hutoa takribani 0.192 ms. Thamani hizi hazilingani na 25.45 ms na 0.89 ms katika Figure 5. Thamani iliyoripotiwa ya 25.45 ms inaweza kupatikana katika equation hiyo hiyo kwa kutumia thickness ya takribani 4 µm. Kwa hiyo matokeo ya Figure 5 yanahifadhiwa kama simulation values zilizoripotiwa katika chanzo, lakini hayachukuliwi kuwa yamethibitishwa upya kwa kujitegemea kutoka equations.

Je, mwanga unaweza kufungiwa ndani ya waveguide?

Kwa kutumia Equivalent Refractive Index approach, beam profile imehesabiwa kwenye input, LC transition regions na output. Katika Figure 6, optical intensity inaonyeshwa ikiwa imefungiwa ndani ya channel core iliyochaguliwa.

Basic period katika planar direction ni takribani:

4 µm core + 4 µm cladding

na vertical layer thickness ni takribani:

2 µm

.

Hata hivyo, Figure 6 caption inaeleza wazi kuwa LC scattering na waveguide-shape losses hazikujumuishwa. Kwa hiyo simulation inaunga mkono mode confinement lakini haibainishi real insertion loss value.

Je, 5 V inawezekana kwa upande wa dielectric?

Kuna PMMA ya takribani 0.1 µm kati ya LC cell electrodes zilizo karibu. Kwa 5 V, chanzo kinakokotoa:

\[ E\approx50\ {\rm kV/mm} = 0.5\ {\rm MV/cm} \]

.

Ikilinganishwa na 2–5 MV/cm intrinsic breakdown levels katika thin-polymer literature inayotumiwa na chanzo, thamani hii ni ya chini. Hata hivyo, kwa kuwa real breakdown voltage ni nyeti kwa film defects, solvent residues, cross-linking na fabrication conditions, chanzo hakijathibitisha tathmini hii kwa reliability test iliyofanywa kwenye physical device.

Ulinganisho na teknolojia nyingine za optical switch

Maneno kama “Excellent”, “Good” na mengine ya aina hiyo si standardized experimental benchmark scores; ni tathmini za waandishi katika qualitative comparison table ya makala.

10×10 scalability

LC electrode pixel pitch ya sasa ni takribani 8 µm. Waandishi wanakadiria kuwa minimum pitch ya 4–5 µm inaweza kuwa inawezekana baadaye na matrix ya ports 10×10 inaweza kuundwa katika eneo la takribani 1 mm × 1 mm. Hili si physical prototype wala demonstration ya experimental switch yenye ports 100.

Matokeo yanayoungwa mkono na utafiti

• Birefringent behavior ya liquid crystal inaweza kutumika kwa voltage-controlled optical core/cladding transition.
• Electrode geometries tofauti zinaweza kubuniwa kwa horizontal na vertical routing kwenye LC platform hiyo hiyo.
• 3×3 architecture inayopendekezwa inaunganisha photonic layers tatu katika switch structure moja.
• Kwa kuwa thamani za \(n_o\) na \(n_e\) za 5PCH ziko karibu na indices za PMMA na PC, voltage-controlled index matching approach inaweza kutumika.
• Source simulations zinaonyesha channel mode confinement katika 1550 nm.
• Source switching simulation inaripoti takribani 1.1 ms ON na 45 ms OFF kwa 5 V.
• LC switching mechanism haihitaji mechanical moving parts.
• Three-dimensional stacking inatoa architectural approach inayoweza kuongeza port density baadaye.

Majumuisho yasiyoungwa mkono na utafiti

• Physical 3D LC switch prototype haikutengenezwa wala kupimwa katika makala hii.
• Hakuna numerical experimental result iliyotolewa kwa real insertion loss.
• Crosstalk na isolation hazikupimwa kimwili.
• Figure 6 haijumuishi scattering na geometric propagation loss.
• Kifaa cha ports 10×10 bado hakijatekelezwa.
• 5 V dielectric reliability haijathibitishwa kwa real lifetime test.
• Real QKD key rate, quantum bit error rate au entanglement fidelity havijapimwa.
• Design si polarization-independent na inahitaji TM input.
• 1.1 ms / 45 ms switching values haziwezi kuzalishwa tena moja kwa moja kwa parameters zilizotolewa katika source equations.
• Nematic LC switch haifai kwa classical high-speed packet-level telecom.

Maelezo ya Chanzo na Mbinu

Utafiti asilia:Three-Dimensional Liquid Crystal Optical Switch for Quantum Optical Communication.

Waandishi: Takao Tomono na Rumiko Yamaguchi.

Taasisi: Graduate School of Science and Technology, Keio University; Research and Development Directorate, Japan Aerospace Exploration Agency (JAXA); Graduate School of Engineering Science, Akita University.

Mwandishi wa mawasiliano: Takao Tomono.

Jarida:Entropic and Disordered Matter, MDPI.

Bibliographic record: 2026, 1(1), Article 2.

DOI: 10.3390/edm1010002.

Publication process: Received 23 April 2026; Revised 11 June 2026; Accepted 25 June 2026; Published 9 July 2026.

Publication status: Published Version of Record iliyopitia peer review.

License: Creative Commons Attribution, CC BY. Copyright inabaki kwa waandishi.

Aina ya utafiti: Development ya three-dimensional photonic switch architecture, nematic LC electro-optic modeling, effective-index calculation, switching-time analysis na optical mode-confinement simulation. Physical switch prototype haikutengenezwa.

Operating wavelength: 1550 nm telecom wavelength katika refractive-index na optical routing analyses.

LC material: Pure nematic 5PCH; polymer-dispersed LC haikutumika.

Polarization: TM. Chanzo kinataja hitaji la polarization splitter–rotator kwa mixed au unpolarized inputs.

Area A/B note ndani ya chanzo: Main method text na Figure 2 zinafafanua Area A kama horizontal na Area B kama vertical switching; katika Figure 1 caption herufi A/B zimegeuzwa.

Equation (1) note: Katika thickness integral iliyotolewa kwa “Average refractive index”, \(1/d\) normalization factor haionekani katika equation iliyochapishwa.

Switching-time QA note: Chanzo kinatoa \(\tau_{ON}=0.89\) ms na \(\tau_{OFF}=25.45\) ms katika Figure 5. Hata hivyo, thamani za \(d=1.7\ \mu m\), \(\gamma=0.1507\) Pa·s, \(K_{11}=9.6\times10^{-12}\) N, \(V=5\) V na \(V_{th}=1\) V zikitekelezwa kwenye Equation (2)–(3), time constants za takribani 4.60 ms OFF na 0.192 ms ON hupatikana mtawalia. Kwa hiyo reproducibility ya calculation ndani ya chanzo ni ndogo.

Loss-model limit: Figure 6 inaeleza wazi kuwa LC scattering na waveguide-shape losses hazikujumuishwa. Chanzo hakina real insertion loss, crosstalk au isolation value.

Editorial residue: Sentensi “I corrected the numbers and the order.” katika Discussion imebaki katika Version of Record kama editing statement nje ya scientific context.

Ufadhili: Center of Innovations for Sustainable Quantum AI (JST), Grant JPMJPF2221.

Upatikanaji wa data: Imeelezwa kuwa data si public lakini inaweza kupatikana kutoka kwa waandishi kwa reasonable request.

Mgongano wa maslahi: Waandishi wanatangaza kuwa hakuna conflict of interest.

Matumizi ya picha: Ingawa makala ina leseni ya CC BY, kwa publishing line ya Verianla ni bora kuchora upya kwa asili 3D photonic layers, LC electrodes na routed light paths badala ya kutumia Figure 1–6 za chanzo moja kwa moja.