Multiplexed Simultaneous High Sensitivity Sensors

sensors Article

Multiplexed Simultaneous High Sensitivity Sensors with High-Order Mode Based on the Integration of Photonic Crystal 1 ˆ 3 Beam Splitter and Three Different Single-Slot PCNCs Jian Zhou, Lijun Huang, Zhongyuan Fu, Fujun Sun and Huiping Tian * State Key Laboratory of Information Photonics and Optical Communications, School of Information and Communication Engineering, Beijing University of Posts and Telecommunications, Beijing 100876, China; [email protected] (J.Z.); [email protected] (L.H.); [email protected] (Z.F.); [email protected] (F.S.) * Correspondence: [email protected]; Tel.: +86-10-6228-2153 Academic Editors: Alexandre Francois and Al Meldrum Received: 1 April 2016; Accepted: 6 July 2016; Published: 7 July 2016

Abstract: We simulated an efficient method for the sensor array of high-sensitivity single-slot photonic crystal nanobeam cavities (PCNCs) on a silicon platform. With the combination of a well-designed photonic crystal waveguide (PhCW) filter and an elaborate single-slot PCNC, a specific high-order resonant mode was filtered for sensing. A 1 ˆ 3 beam splitter carefully established was implemented to split channels and integrate three sensors to realize microarrays. By applying the three-dimensional finite-difference-time-domain (3D-FDTD) method, the sensitivities calculated were S1 = 492 nm/RIU, S2 = 244 nm/RIU, and S3 = 552 nm/RIU, respectively. To the best of our knowledge, this is the first multiplexing design in which each sensor cite features such a high sensitivity simultaneously. Keywords: photonic crystal; resonators; filter; optical sensing and sensor; multiplexing

1. Introduction Label-free optical sensors have recently garnered increasing interest for high-performance quantitative measurements without labeling heterogeneity introduced by fluorescent techniques. In particular, photonic crystal (PhC) sensors, because of the strong capability to manipulate light propagation, have been an attractive candidate for available label-free systems [1–24]. Recently, optical microcavities have been studied extensively for sensing. The resonant wavelength simply shifts as the environmental refractive index nenv varies. Nanocavities in the one-dimensional (1D) PhC nanobeam slab confine light into an ultrasmall volume of the order of optical wavelength. In addition, a narrow slot introduced has increasingly improved the optical field localization contributing to the strong light–matter interaction [25–27]. The above-mentioned high-performance and miniaturized geometries enable multiple sensor microcavities to be integrated on a chip. The integration of PhC sensors requires that multiple PhC microcavities must be positioned on one chip. Previously, Mandal et al. [6] demonstrated a nanoscale opto-fluidic sensors array based on five one-dimensional (1D) photonic crystal micro-cavities side-coupled to a bus waveguide. Although the authors of [6] presented a sensor array based on five one-dimensional (1D) photonic crystal nanobeam, the maximum sensitivity was only 130 nm/RIU. Yang et al. [21,22] demonstrated the nanoscale PhC sensors array on monolithic substrates based on a side-coupled resonant cavity array on silicon, and the sensitivities were approximately 160 nm/RIU. Caër et al. [27] demonstrated a high sensitivity of 235 nm/RIU based on an infiltrated high-Q slot photonic crystal cavity, but this sensor possessed a large footprint so that it was not suited to large-scale integration. Zou et al. [28] presented a dense microarray of 64 microcavity-based sensor nodes with series and parallel connected PhC microcavity Sensors 2016, 16, 1050; doi:10.3390/s16071050

www.mdpi.com/journal/sensors

Sensors 2016, 16, 1050 Sensors 2016, 16, 1050

2 of 10 2 of 10

sensors. et al. [29] proposed the integration a scheme to multiplexreferences multi-resonance PhC cavity realize aYan multiplexed sensor array. However, of the above-mentioned demonstrated low sensors with an additional PhC waveguide (PhCW) bandpass filter to realize a multiplexed sensor sensitivities or a low-site sensor array, and the combination thereof will be a promising candidate for array. the above-mentioned references demonstrated low sensitivities or a low-site sensor futureHowever, sensor applications. array,In and the combination will be a promising candidate for future sensor applications. order to realize a thereof high-sensitivity sensor array, the parallel connection of single-slot PhC In order to realize a high-sensitivity sensor array, the parallel connection of single-slot nanobeam cavity (PCNC) sensors was simulated. The parallel connector was obtained by a 1 × 3PhC PhC nanobeam cavity (PCNC) The parallel connector was obtained bywhich a 1 ˆ 3makes PhC beam splitter. Because thesensors PCNCwas has simulated. several resonances in the transmission spectrum, beam the PCNC has several resonances in (PhCW) the transmission makes them splitter. difficultBecause to multiplex, an additional PhC waveguide bandpassspectrum, filter waswhich integrated on them an additional PhC waveguide was integrated each difficult channel to ofmultiplex, the multiplexed sensor array to select a (PhCW) specific bandpass resonance.filter Through adjusting on the each multiplexed array to select specificthree resonance. Throughhigh-order adjusting the taper taperchannel regionof ofthe PCNC and the sensor lattice constant (a) ofaPhCW, distinguished resonant region PCNC and the transmission lattice constant (a) of PhCW, three distinguished high-order resonant peaks peaks of appeared in the spectrum. In addition, the calculated high sensitivities of the appeared in the transmission spectrum. addition, calculated of the multiplexed multiplexed sensor array were S1 = 492In nm/RIU, S2the = 244 nm/RIU,high andsensitivities S3 = 552 nm/RIU, respectively. sensor array were S1 = 492 nm/RIU, S2 = 244 nm/RIU, and S3 = 552 nm/RIU, respectively. 2. The 1 × 3 PhC Beam Splitter Design 2. The 1 ˆ 3 PhC Beam Splitter Design As a key component in photonic integrated circuits (PICs), beam splitters are indispensable. To As a keyamount component in photonic integrated circuitsusing (PICs), beam splitters indispensable. date, a large of components have been designed elementary beam are splitters, including To date, a large amount of components have been designed using elementary beam splitters, direct splitting, Y-branch [30,31], and T-branch [32,33], directional coupling splitters [34,35], 1×4 including direct splitting, Y-branch and T-branch [32,33], coupling splitters [34,35], splitters [36], etc. In this section, we[30,31], describe our design of a 1 × 3directional beam splitter, and the principle was 1motivated ˆ 4 splitters [36], etc. In this section, we describe our design of a 1 ˆ 3 beam splitter, and the principle by [37]. A basic PhC slab-based beam splitter was formed by a triangular lattice air was motivated byin[37]. A basic PhCsilicon slab-based beam splitter was a triangular lattice air cylinders etched a 220-nm-thick layer (n Si = 3.46) lying on formed top of a by 2-μm buried silicon oxide cylinders etched in awhich 220-nm-thick silicon layer (nSi = 3.46) on top of athe 2-µm buried silicon oxide layer (nSiO2 = 1.45), is clearly shown in Figure 1. In lying this structure, lattice constant was 460 layer (n = 1.45), which is clearly shown in Figure 1. In this structure, the lattice constant was SiO2 nm, and the radius of the holes was 147 nm. The former of the beam splitter consisted of a Y branch 460 nm, and the radius of the was 147 nm. The former of the beam splitter consisted of a Y branch with a 120-degree angle andholes a waveguide connected at the junction. The latter is composed of two with a 120-degree angle and a waveguide connected at the junction. The latter is composed of two 120-degree low-loss waveguide bend structures. The transmission for three branches was measured 120-degree low-loss waveguide bend structures. The transmission for three branches was measured through three detectors (Dectectors 1, 2, and 3), respectively. The detailed parameters set in Regions through detectors (Dectectors 1, 2, and The detailed parameters in Regions (R1, R2,three and R3) are shown in the right inset3),inrespectively. Figure 1. The triangle polygons wereset introduced to (R1, R2, and R3) are shown in the right inset in Figure 1. The triangle polygons were introduced to reduce the reflection loss and guide the incoming wave into the branches effectively [37]. The key reduce the reflection and constant guide thethickness incomingh wave branches effectively [37]. key parameters are w, L,loss θ, and = 220into nm,the where w is the width of theThe triangle parameters w, L, θ, and constant thickness h =sharp 220 nm, whereand w isθ the width of theangle triangle polygon, polygon, Lare is the length from the bottom to the corner, is the rotation versus the xLaxis. is the length from the bottom to the sharp corner, and θ is the rotation angle versus the x-axis.

Figure1.1. Schematic Schematic view view of 3 beam splitter lying on on toptop of aof2Figure of the the 2D 2D photonic photoniccrystal crystal(PhC) (PhC)slab slab1 1× ˆ 3 beam splitter lying buried silicon oxide layer, where a = 460 147rnm, andnm, h = 220 parameters aμm 2-µm buried silicon oxide layer, where a =nm, 460r =nm, = 147 andnm. h =The 220detailed nm. The detailed in Regions (R1, R2, and R3) are shown in the right inset, where w is the width of the triangle polygon, parameters in Regions (R1, R2, and R3) are shown in the right inset, where w is the width of the L is the length from the bottom to the sharp corner, and θ is the rotation angle versus the x-axis. triangle polygon, L is the length from the bottom to the sharp corner, and θ is the rotation angle versus the x-axis.

In this work, the three-dimensional finite-difference-time-domain (3D-FDTD) method with perfectly matched layer boundary conditions was utilized for the simulations (Lumerical solutions, Inc., Vancouver, BC, Canada) [38]. Many simulations are done by modifying the parameters L, w, and

Sensors 2016, 16, 1050

3 of 10

In this work, the three-dimensional finite-difference-time-domain (3D-FDTD) method with perfectly Sensors 2016,matched 16, 1050 layer boundary conditions was utilized for the simulations (Lumerical solutions, 3 of 10 Sensors 2016, 16, 1050 10 Inc., Vancouver, BC, Canada) [38]. Many simulations are done by modifying the parameters L, w,3 of and θ; θ; in this way, we obtain optimized results. The optimal parameters parameters set set were were as as follows: follows: LL ==782 782nm, nm, θ; in this way, we obtain optimized results. The optimal parameters set were as follows: L = 782 nm, ˝ (Region w θ =θ 300° 1); L1);= 1150 nm, w = 103.04 nm, and = 210° (upper part in Region w ==220.8 220.8nm, nm,and and = 300(Region L = 1150 nm, w = 103.04 nm,θ and θ= 210˝ (upper part in w = 220.8 nm, and θ = 300° (Region 1); L = 1150 nm, w = 103.04 nm, and θ = 210° (upper part in Region 2); L = 1150 = 103.04 θ = 150° in˝ Region L =in 782 nm, w 2); = 220.8 Region 2); nm, L = w1150 nm, nm, w =and 103.04 nm, (lower and θ part = 150 (lower 2); part Region L = nm, 782 and nm, 2); L = 1150 nm, w = 103.04 nm, and θ = 150° (lower part in Region 2); L = 782 nm, w = 220.8 nm, and θw==60° (Region 3). θ = 60˝ (Region 3). 220.8 nm, and θ = 60° (Region 3). The steady-stateelectric electric field profile is shown in 2, Figure excited resonant The steady-state field profile is shown in Figure where2,thewhere excitedthe resonant wavelength The steady-state electric field profile is shown in Figure 2, where the excited resonant wavelength is 1491.94 nm.seen As clearly seen from Figure 2, the optical horizontal field was by is 1491.94 nm. As clearly from Figure 2, the horizontal fieldoptical was limited by alimited photonic wavelength is 1491.94 nm. As clearly seen from Figure 2, the horizontal optical field was limited by aband photonic band and gap the (PBG), andoptical the vertical optical field wasby manipulated by thereflection total internal gap (PBG), vertical field was manipulated the total internal (TIR). a photonic band gap (PBG), and the vertical optical field was manipulated by the total internal reflection (TIR).the Figure shows theof field of the to the red dash-line regions Figure 3 shows field3amplitude theamplitude corresponding tocorresponding the red dash-line regions (Regions 1, 2, and reflection (TIR). Figure 3 shows the field amplitude of the corresponding to the red dash-line regions (Regions 1, 2, andclearly 3) in Figure 1. Figure As clearly shown infield Figure the electricthrough field was 3) in Figure 1. As shown in 3, the electric was3,continuous the continuous waveguide, (Regions 1, 2, and 3) in Figure 1. As clearly shown in Figure 3, the electric field was continuous through waveguide, while the electric field wasthe discontinuous through thediscontinuity triangle polygon. This while thethe electric field was discontinuous through triangle polygon. This is because through the waveguide, while the electric field was discontinuous through the triangle polygon. This discontinuity is because of the fact that the electric displacement field is continuous across the index of the fact that the electric displacement field is continuous across the index contrast interface. Light discontinuity is because of the fact that the electric displacement field is continuous across the index contrast interface. Lightindex localization the low index material could be realized byfield the localization in the low materialin region could be realized byregion the discontinuity of electric contrast interface. Light localization in the low index material region could be realized by the discontinuity of electric field perpendicularly to the index contrast interface. As shown in Figure 4, perpendicularly to the index contrast interface. As shown in Figure 4, the flat transmission spectrum discontinuity of electric field perpendicularly to the index contrast interface. As shown in Figure 4, the transmission spectrum in anm. range from nm to 1570 nm. Thisthe flatdemand band was wasflat obtained in a range from was 1470obtained nm to 1570 This flat1470 band was able to meet of the flat transmission spectrum was obtained in a range from 1470 nm to 1570 nm. This flat band was able to meet the demand parallel connectors for the sensor array. parallel connectors for theofsensor array. able to meet the demand of parallel connectors for the sensor array.

Figure 2. Steady-state Steady-stateelectric electricfield fielddistribution distribution fundamental TE-like (Transverse Electric) Figure 2. forfor thethe fundamental TE-like (Transverse Electric) mode Figure 2. Steady-state electric field distribution for the fundamental TE-like (Transverse Electric) mode propagating through the optimized PhCsplitter slab splitter in (a) x-y and plane y-z plane. propagating through the optimized PhC slab in (a) the x-ythe plane (b)and the(b) y-zthe plane. mode propagating through the optimized PhC slab splitter in (a) the x-y plane and (b) the y-z plane.

Figure 3. Field amplitude distributions for the optimized model corresponding to the three regions Figure 3. Field amplitude distributions for the optimized model corresponding to the three regions Figure 3. 1, Field amplitude for the optimized model corresponding to the three regions (Regions 2, and 3). (a) distributions Field amplitude corresponding to the Region 2; (b) Field amplitude (Regions 1, 2, and 3). (a) Field amplitude corresponding to the Region 2; (b) Field amplitude (Regions 1, 2, to and (a) Field amplitude corresponding to the Region 2; 3. (b) Field amplitude corresponding the3). Region 1; (c) Field amplitude corresponding to the Region corresponding to the the Region Region 3. 3. corresponding to to the the Region Region 1; 1; (c) (c) Field Field amplitude amplitude corresponding corresponding to

Sensors 2016, 16, 1050 Sensors Sensors 2016, 2016, 16, 16, 1050 1050

4 of 10 44 of of 10 10

Figure 4. (blue Figure 4. 4. Three-dimensional Three-dimensional finite-difference-time-domain finite-difference-time-domain (3D-FDTD) (3D-FDTD) transmission transmission spectra spectra (blue (blue Figure Three-dimensional finite-difference-time-domain (3D-FDTD) transmission spectra and black curves) for TE-polarized light at the Detectors 11 and 3, and Detector 22 at the output of the and black curves) for TE-polarized light at the Detectors and 3, and Detector at the output of the and black curves) for TE-polarized light at the Detectors 1 and 3, and Detector 2 at the output of the optimum 11 ×× 33 beam splitter. The flat-band range is to the benefit of outstanding parallel connector. optimum beam splitter. The flat-band range is to the benefit of outstanding parallel connector. optimum 1 ˆ 3 beam splitter. The flat-band range is to the benefit of outstanding parallel connector.

3. PCNC Design 3. Single-slot Single-slot 3. Single-slotPCNC PCNCDesign Design The combination PCNC the air-slot waveguide in the strong light–matter The combination combination of of the the PCNC PCNC and and the the air-slot air-slot waveguide waveguide results results in in the the strong strong light–matter light–matter The of the and results interaction [23,26], which contributes to high sensitivity. We demonstrate here that the structure interaction [23,26], which contributes to high sensitivity. We demonstrate here that the interaction [23,26], which contributes to high sensitivity. We demonstrate here that the structurestructure consists consists of two parallel suspended nanobeams separated by aa small air gap. The schematic of consists of two parallel suspended nanobeams separated by small air gap. The schematic of the the of two parallel suspended nanobeams separated by a small air gap. The schematic of the model is model is shown in Figure 5. The refractive index of the silicon is n si = 3.46. The lattice constant a = 430 model is shown in Figure 5. The refractive index of the silicon is n si = 3.46. The lattice constant a = 430 shown in Figure 5. The refractive index of the silicon is nsi = 3.46. The lattice constant a = 430 nm, width nm, w == 360 thickness tt == 220 nm, and w nm. nm,=width width wnbnbwaveguide 360 nm, nm, waveguide waveguide andwslot slot=width width wslot slot = = 100 100 nm. The The design design w 360 nm, thickness t =thickness 220 nm, and220 slotnm, width 100 nm. The design principle is nb slot principle is referred to the deterministic cavity optimized method [39]. To create the Bragg mirror, principletoisthe referred to the cavity deterministic cavity optimized [39]. To mirror, create the mirror, referred deterministic optimized method [39]. Tomethod create the Bragg the Bragg air-holes radii the air-holes radii are tapered from rrcenter 120 to rend == 93 nm, i.e., the parabolic air-holestapered radii from are parabolic center = 120“nm nm nm, i.e.,, q{k2max are rparabolic = 120 tapered nm to rendfrom = 93 nm, i.e.,= r pnq rcenterto` kr2endprend ´93rcenter center 22 22 , (k increases from 0 to k max ). Therefore, a Gaussian type confinement r n  r  k r  r k , (k increases from 0 to k max ). Therefore, a Gaussian type confinement     r increases n  rcenter k rend center from end center max (k 0 torcenter kmax ).kmax Therefore, a Gaussian type confinement will be realized. In order to will realized. In to three sensing, we three different realize parallel multiplexed sensing, we chosemultiplexed three different numbers of tapered but will be bethree realized. In order order to realize realize three parallel parallel multiplexed sensing, we chose chose threeholes different numbers of holes region. To the studied numbersthe of tapered tapered holes but but without the mirror mirror region. To reduce reduce the simulation simulation time, we studied without mirror region. Towithout reduce the simulation time, we studied the highertime, orderwe modes of the higher of cavity possessing moderate Q-factor, the higher order order modes modescavity of aa waveguide-coupled waveguide-coupled cavity possessing moderate Q-factor, and the total total athe waveguide-coupled possessing a moderate Q-factor, and aathe total number of kand mirror pair number kk mirror was 16, 20, Figure 66 shows the number of ofwas mirror pair segments was 13, 13, 16, and and 20, respectively. respectively. Figure shows the transmission transmission segments 13, 16,pair and segments 20, respectively. Figure 6 shows the transmission spectra obtained by exciting spectra obtained by with aa broadband waveguide mode The spectra obtained by exciting exciting the input input waveguide waveguide with broadband waveguide mode source. source. The the input waveguide with a the broadband waveguide mode source. The three separated high-order three separated high-order resonances will be chosen for parallel multiplexed sensing. Figures 7a–c three separated high-order resonances will be chosen for parallel multiplexed sensing. Figures 7a–c resonances will be chosen for parallel multiplexed sensing. Figure 7a–c show the major field-component show field-component distribution of three high-order resonant modes. As it show the the major major field-component distribution of the the As three high-order resonant modes. As expected, expected, it distribution of the three high-order resonant modes. expected, it can be clearly seen that the electric can clearly that the field profile confined within the of can be be clearly seen that confined the electric electric field the profile is strongly confined within which the slot slotcontributes of the the low-index low-index field profile is seen strongly within slotis ofstrongly the low-index material, to the material, which to interaction between analytes material, which contributes contributes to the the strong strong interaction betweenmode. analytes and and the the cavity cavity resonant resonant mode. mode. strong interaction between analytes and the cavity resonant

Figure Schematics of of the the proposed single-slot photonic crystal nanobeam cavity Figure 5. (PCNC). The Figure 5. 5. Schematics Schematics of the proposed proposed single-slot single-slot photonic photonic crystal crystal nanobeam nanobeam cavity cavity (PCNC). (PCNC). The The structure is symmetric with respect to its center (blue dashed line). The photonic structure pitch == 430 structure is is symmetric symmetric with with respect respect to to its its center center (blue (blue dashed dashed line). line). The The photonic photonic mirror mirrormirror pitch aa pitch 430 anm, = the 430thickness nm, thett ==thickness t =width 220 w nm, the width = nm, and the slot separation nb slot nm, 220 and separation w nm. The the thickness 220 nm, nm, the the width wnbnb == 360 360 nm, nm, andwthe the slot360 separation wslot slot = = 100 100 nm. The hole hole w = nm. tapered The hole radii are= 120 parabolic tapered r=center = on 120 nm in the center to slot are nm center to 93 side. radii parabolic from rrcenter center = 120 nm in in the the centerfrom to rrend end = 93 nm nm on either either side. radii are100 parabolic tapered from rend = 93 nm on either side.

Sensors 2016, 16, 1050 Sensors 2016, 2016, 16, 16, 1050 1050 Sensors

5 of 10 of 10 10 55 of

spectra of of single-slot PCNC sensor 3D-FDTD simulation. simulation. The Figure 6. 6. Transmission Transmission spectra single-slot PCNC sensor from from 3D-FDTD The optimized optimized Figure structure with k (13, 16, 20) mirror segments in the taper region and no additional mirrors. The taper (13, (13, 16, 16, 20) 20) mirror mirror segments segments in in the the taper region and additional The structure with taper taper is set set as as RI applied background refractive refractive index index is RI == 1.000. 1.000. The The three three high-order high-order resonant resonant peaks peaks will will be be applied background multiplexed sensing. sensing. for parallel parallel multiplexed for

Figure 7. 3D-FDTD 3D-FDTD simulation of the major field distribution profile in the single-slot PCNC PCNC Figure Figure 7. 7. 3D-FDTDsimulation simulationof of the the major major field field distribution distribution profile profile in in the the single-slot single-slot PCNC corresponding to the three resonant modes in Figure 6. corresponding to the three resonant modes in Figure 6. corresponding to the three resonant modes in Figure 6. (a) λ = 1535.1 nm; (b) λ = 1544.25 nm; (c) λ = 1548.55 nm.

4. Bandpass Bandpass Filter Filter Design Design 4. 4. Bandpass Design Based on onFilter the above above designed sensors, sensors, in in order order to to select select three three specific specific resonant resonant peaks peaks to to realize realize aa Based the designed multiplexed sensor array, an additional additional PhCW bandpass filter was investigated. Sometoon-chip on-chip multiplexed array, an bandpass investigated. Some Based onsensor the above designed sensors,PhCW in order to select filter three was specific resonant peaks realize photonic bandpass filters, such as arrayed waveguide gratings (AWGs) and waveguide Bragg photonic bandpass filters, such as arrayed waveguide gratings (AWGs) and waveguide Bragg a multiplexed sensor array, an additional PhCW bandpass filter was investigated. Some on-chip gratings (WBGs), were considered for the filters. However, WBGs are commonly long and need to gratings (WBGs), were considered for the filters. However, WBGs are commonly long and need to photonic bandpass filters, such as arrayed waveguide gratings (AWGs) and waveguide Bragg gratings operate in in reflection [40].for The size of AWGs AWGs is usually usually too commonly large [41] [41] and and not not suitable for dense dense operate reflection [40]. The size of is too large for (WBGs), were considered the filters. However, WBGs are long andsuitable need to operate in integration. In particular, the PhCW filter is a natural choice due to the sensor chip composed by PhC, integration. In particular, the PhCW filter is a natural choice due to the sensor chip composed by PhC, reflection [40]. The size of AWGs is usually too large [41] and not suitable for dense integration. In leading to to the the easier passband alignment. leading alignment. particular, the easier PhCWpassband filter is a natural choice due to the sensor chip composed by PhC, leading to the easier passband alignment.

Sensors 2016, 16, 1050 Sensors 2016, 16, 1050

6 of 10 6 of 10

A line defect in a 2D PhC lattice on a 220-nm-thick slab established a PhCW, as shown in Figure 8. In terms of the bandpass filter, light with a longer wavelength above the passband located within the bandgap of the PhC was forbidden, and shorter wavelengths outside the passband were above the light line so that the light cannot be well-confined in the waveguide. The resonant filter was created by connecting the PhCW and the sensor (Figure 5) in series. The lattice constant of the PhCW could be engineered so that a passband formed in the transmission spectrum. The PhCW filter selected the specific high-order resonance of the sensor. The filtered resonances are used for sensing when analytes are filled into the sensor. As shown in Figure 9, passbands are observed in the transmission spectra with regard to three different lattice constants (a = 547 nm, 552 nm, and 553 nm). The three high-order Figure 8. The schematic of proposed photonic crystal waveguide (PhCW) bandpss filter. The6 of9.10 resonant modes Sensors 2016, 16, 1050are located within the passbands and easily filtered, as clearly observed in Figure thickness is 220 nm.

A line defect in a 2D PhC lattice on a 220-nm-thick slab established a PhCW, as shown in Figure 8. In terms of the bandpass filter, light with a longer wavelength above the passband located within the bandgap of the PhC was forbidden, and shorter wavelengths outside the passband were above the light line so that the light cannot be well-confined in the waveguide. The resonant filter was created by connecting the PhCW and the sensor (Figure 5) in series. The lattice constant of the PhCW could be engineered so that a passband formed in the transmission spectrum. The PhCW filter selected the specific high-order resonance of the sensor. The filtered resonances are used for sensing when analytes are filled into the sensor. As shown in Figure 9, passbands are observed in the transmission spectra with regard to three different lattice constants (a = 547 nm, 552 nm, and 553 nm). The three high-order resonant modes are located within the passbands and easily filtered, as clearly Theschematic schematic of proposed photonic waveguide The Figure 8.8.The of proposed photonic crystalcrystal waveguide (PhCW)(PhCW) bandpssbandpss filter. Thefilter. thickness observed in Figure 9. thickness is 220 nm.is 220 nm.

A line defect in a 2D PhC lattice on a 220-nm-thick slab established a PhCW, as shown in Figure 8. In terms of the bandpass filter, light with a longer wavelength above the passband located within the bandgap of the PhC was forbidden, and shorter wavelengths outside the passband were above the light line so that the light cannot be well-confined in the waveguide. The resonant filter was created by connecting the PhCW and the sensor (Figure 5) in series. The lattice constant of the PhCW could be engineered so that a passband formed in the transmission spectrum. The PhCW filter selected the specific high-order resonance of the sensor. The filtered resonances are used for sensing when analytes are filled into the sensor. As shown in Figure 9, passbands are observed in the transmission spectra with regard to three different lattice constants (a = 547 nm, 552 nm, and 553 nm). The three high-order resonant modes are located within the passbands and easily filtered, as clearly observed in Figure 9.

Figure Transmission spectra Figure 9. 9. Transmission spectra of of PhCW PhCW filters filters with with different different lattice latticeconstants. constants.

5. MultiplexingSensor Sensor Array Array Design Design 5. Multiplexing Multiplexing Multiplexing is is realized realized by by placing placing arbitrary arbitrary resonant resonant cavities cavities butt-coupled butt-coupled to to each each beam beam splitter splitter branch; branch; in in this this way, way,arbitrary arbitrary resonant resonant peaks peaks can can be beobtained. obtained. The The multiple multiple resonant resonant cavity cavity sensors sensors can device. For proof-of-concept demonstration of can be be interrogated interrogatedsimultaneously simultaneouslyon onmonolithic monolithicPhCs PhCs device. For proof-of-concept demonstration the proposed PhCW filter and multiplexing, a 3-channel PhC sensor array was designed. A 1 ˆ 3 PhC of the proposed PhCW filter and multiplexing, a 3-channel PhC sensor array was designed. A 1 × 3 beam splitter was used split waveguides. Through properproper engineering of the of passband, three PhC beam splitter was to used tothe split the waveguides. Through engineering the passband, high-sensitivity multi-mode single-slot PCNCPCNC sensors couldcould be multiplexed into ainto single-input and three high-sensitivity multi-mode single-slot sensors be multiplexed a single-input three-output system as shown in Figure 10. In each channel, the former features as a wavelength filter and three-output system as shown in Figure 10. In each channel, the former features as a wavelength and latter a sensing site with cascaded transmission band. In order realize the sensor filterthe and the acts latterasacts as a sensing sitea with a cascaded transmission band. Intoorder to realize the Figure 9. Transmission spectra of PhCW filters with different lattice constants.

5. Multiplexing Sensor Array Design

Sensors Sensors 2016, 2016, 16, 16, 1050 1050 Sensors 2016, 16, 1050

77 of of 10 10 7 of 10

sensor sensor array, array, the the PhCW PhCW filters filters were were carefully carefully employed employed to to clearly clearly distinguish distinguish designate designate high-order high-order resonant peaks. Multiple sensors could then be interrogated simultaneously from the eventual resonant peaks. Multiple sensors could then be interrogated simultaneously from the eventual array, the PhCW filters were carefully employed to clearly distinguish designate high-order resonant transmission spectra. These sensors allowed different analytes to be filled into each of them. Based transmission spectra. These sensors allowed different analytes to be filled into each of them. Based peaks. Multiple sensors could then be interrogated simultaneously from the eventual transmission on the integrated structure in Figure 10, by applying the 3D-FDTD method, the output transmission on the integrated structure in Figure 10,analytes by applying 3D-FDTD output transmission spectra. These sensors allowed different to be the filled into eachmethod, of them.the Based on the integrated spectra are calculated in Figure 11. As clearly shown in Figure 11, three resonant peaks appear in spectra are calculated in Figure 11. As clearly shown in Figure 11, three resonant peaks appear in the the structure in Figure 10, by applying the 3D-FDTD method, the output transmission spectra are calculated transmission spectra. transmission spectra. in Figure 11. As clearly shown in Figure 11, three resonant peaks appear in the transmission spectra.

Figure monolithic substrate. Figure 10. 10. Schematic Schematic of of the the PhCs PhCs parallel parallel integrated integrated sensor sensor array on the sensor array array on on the the monolithic monolithic substrate. substrate.

transmission spectra of three three set Figure Figure 11. 11. The The transmission transmission spectra of three parallel parallel branches branches observed observed where where three three sensors sensors are are set in in parallel connections. parallel connections.

In In order order to calculate sensor performance performance when when the the resonant resonant cavities cavities are are in in parallel-connections, parallel-connections, order to to calculate calculate sensor each sensor was independently influenced by the refractive index change. Figure 12a the each influenced by the indexindex change. FigureFigure 12a shows the output each sensor sensorwas wasindependently independently influenced by refractive the refractive change. 12a shows shows the output transmission spectra of the multiplexed sensor array when three sensors are filled with transmission spectra ofspectra the multiplexed sensor arraysensor when three filled with refractive output transmission of the multiplexed arraysensors when are three sensors are filledindex with refractive RI three resonant peaks are obviously shifted and the shift arrives RI = 1.005;index three resonant peaks are obviously shifted the resonant shift at 2.46 nm, nm, refractive index RI == 1.005; 1.005; three resonant peaks are and obviously shifted andarrives the resonant resonant shift1.22 arrives at 2.46 nm, 1.22 nm, and 2.76 nm, respectively. Through the definition of the sensitivity as S = Δλ/Δn, and 2.76 nm, respectively. the definition of the as the S =sensitivity ∆λ/∆n, we at 2.46 nm, 1.22 nm, and 2.76Through nm, respectively. Through thesensitivity definition of as calculated S = Δλ/Δn, we of single-slot PCNC SS11 == 492 SS22 nm/RIU, == 244 the sensitivitythe of sensitivity each single-slot sensor thatsensor is S1 =that 492is S2 = 244 and we calculated calculated the sensitivity of each eachPCNC single-slot PCNC sensor that isnm/RIU, 492 nm/RIU, nm/RIU, 244 nm/RIU, nm/RIU, and S 3 = 552 nm/RIU, respectively. When the refractive index of one sensor changed, the others Sand respectively. When When the refractive index ofindex one sensor the othersthe remained S3 =nm/RIU, 552 nm/RIU, respectively. the refractive of onechanged, sensor changed, others 3 = 552 remained fixed. As from 12b, resonant peak generates towards higher fixed. As seen from Figure 12b,Figure the resonant generates higher wavelength where the remained fixed. As seen seen from Figure 12b, the thepeak resonant peaktowards generates towards higher wavelength wavelength where positive effect This three sensors work To positive effect happens. This confirms that threethat sensors independently. To the best of best our where the the positive effect happens. happens. This confirms confirms that three work sensors work independently. independently. To the the best of our knowledge, this is the first geometry that features such high sensitivity simultaneously, it knowledge, this is this the first that features such high simultaneously, it is thus of our knowledge, is thegeometry first geometry that features such sensitivity high sensitivity simultaneously, it is is

Sensors 2016, 16, 1050 Sensors 2016, 16, 1050

8 of 10 8 of 10

thus potentially an ideal platform for larger-scale high-sensitivity multiplexing refractive indexpotentially an ideal platform for larger-scale high-sensitivity multiplexing refractive index-based based biochemical sensing. biochemical sensing.

Figure 12. spectra fromfrom (a) three wavelength shift when thewhen refractive 12.Measured Measuredtransmission transmission spectra (a) resonant three resonant wavelength shift the index of three sensors change from 1.000 to 1.005; (b) a resonant wavelength shift when the refractive refractive index of three sensors change from 1.000 to 1.005; (b) a resonant wavelength shift when the index of one sensor change from 1.000 to 1.005, thewhile refractive index of index the other two sensors refractive index of one sensor change from 1.000 while to 1.005, the refractive of the other two keeps fixed. sensors keeps fixed.

6. Conclusions Conclusions In summary, summary, aa method method of a multiplexed sensor array of high-sensitivity single-slot PCNC sensor was measurement between a single input and three obtained was simulated. simulated.The Thesimultaneous simultaneous measurement between a single input andoutputs three was outputs was by the optimized 1 ˆ 3 beam were combined with single-slot PCNC sensors, obtained by the optimized 1 ×splitter. 3 beam PhCW splitter.filters PhCW filters were combined with single-slot PCNC contributing to the filtering specific of high-order resonant peaks to realize a large-scale array. sensors, contributing to theoffiltering specific high-order resonant peaks to realize asensor large-scale By using the By 3D-FDTD simulation three high sensitivities were obtained. proposed sensor array. using the 3D-FDTDmethod, simulation method, three high sensitivities were The obtained. The geometry desirableisfor the performance of a highly parallel and parallel label-free detection system with proposed is geometry desirable for the performance of a highly and label-free detection multiplexing capability oncapability a monolithic it is more suitable for suitable monolithic system with multiplexing on asubstrate. monolithicFurthermore, substrate. Furthermore, it is more for ultralow concentration detection indetection actual biochemical applications.applications. monolithic ultralow concentration in actual biochemical Acknowledgments: This research research was was supported supportedby byNSFC NSFC(No. (No.61372038), 61372038), Fund State Key Laboratory Acknowledgments: This Fund of of State Key Laboratory of of Information Photonics and Optical Communications (Beijing University of Posts and Telecommunications) Information Photonics andPostgraduate Optical Communications (Beijing Posts and Telecommunications) IPOC2015ZC02, P.R. China, Innovation Fund of SICE,University BUPT, 2015ofand BUPT Excellent Ph.D. Students IPOC2015ZC02, P.R. China, Postgraduate Innovation Fund of SICE, BUPT, 2015 and BUPT Excellent Ph.D. Foundation (CX2015202). StudentsContributions: Foundation (CX2015202). Author Jian Zhou conceived and designed the strcture, contributed the FDTD simualtions, analyzed the data and wrote the paper; Lijun Huang contributed the FDTD simualtions and analyzed the data; Author Contributions: Jian Zhou conceived and designed the strcture, contributed the FDTD simualtions, Zhongyuan Fu and Fujun Sun contributed the FDTD simulations and the writing of the paper; Huiping Tian analyzed the data andarranged wrote thethe paper; Lijundesign Huangand contributed the the FDTD simualtions and analyzed data; conceived this study, strcture contributed wirting of the paper. All thethe authors commented Zhongyuan on Fu the andpaper. Fujun Sun contributed the FDTD simulations and the writing of the paper; Huiping Tian conceivedofthis study,The arranged strcture design and contributed the wirting of the paper. All the authors Conflicts Interest: authorsthe declare no conflict of interest. commented on the paper.

References Conflicts of Interest: The authors declare no conflict of interest. 1.

Armani, A.M.; Vahala, K.J. Heavy water detection using ultra-high-Q microcavities. Opt. Lett. 2006, 31,

References 1896–1898. [CrossRef] [PubMed] 2. 1. 3. 2. 3. 4.

4.

Lal, S.; Link, Halas, K.J. N.J. Heavy Nano-optics sensing to waveguiding. Nat. Photonics 2007, 1, 2006, 641–648. Armani, A.M.;S.;Vahala, water from detection using ultra-high-Q microcavities. Opt. Lett. 31, [CrossRef] 1896–1898. Armani, A.M.; Fraser, S.E.; Flagan, R.C.; Vahala, K.J. Label-free, single-molecule detection Lal, S.; Link, S.;Kulkarni, Halas, N.J.R.P.; Nano-optics from sensing to waveguiding. Nat. Photon. 2007, 1, 641–648. with optical microcavities. Science 2007, 317, 783–787. [CrossRef] [PubMed] Armani, A.M.; Kulkarni, R.P.; Fraser, S.E.; Flagan, R.C.; Vahala, K.J. Label-free, single-molecule detection White, I.M.; Fan, X. On the performance quantification of resonant refractive index sensors. Opt. Express with optical microcavities. Science 2007, 317, 783–787. 2008, 16, 1020–1028. [CrossRef] [PubMed] White, I.M.; Fan, X. On the performance quantification of resonant refractive index sensors. Opt. Express 2008, 16, 1020–1028.

Sensors 2016, 16, 1050

5. 6. 7. 8. 9. 10. 11.

12. 13. 14. 15.

16. 17. 18. 19. 20.

21. 22. 23. 24.

25. 26. 27. 28.

9 of 10

Kita, S.; Nozaki, K.; Baba, T. Refractive index sensing utilizing a cw photonic crystal nanolaser and its array configuration. Opt. Express 2008, 16, 8174–8180. [CrossRef] [PubMed] Mandal, S.; Erickson, D. Nanoscale optofluidic sensor arrays. Opt. Express 2008, 16, 1623–1631. [CrossRef] [PubMed] Hanic, S.T.; Rahmani, A.; Steel, M.J.; Sterke, C.M. Comparison of the sensitivity of air and dielectric modes in photonic crystal slab sensors. Opt. Express 2009, 17, 14552–14557. [CrossRef] Kang, C.; Phare, C.T.; Vlasov, Y.A.; Assefa, S.; Weiss, S.M. Photonic crystal slab sensor with enhanced surface area. Opt. Express 2010, 18, 27930–27937. [CrossRef] [PubMed] Dahdah, J.; Courjal, N.; Baida, F.I. Analysis of a photonic crystal cavity based on absorbent layer for sensing applications. J. Opt. Soc. Am. B 2010, 27, 305–310. [CrossRef] Xu, T.; Zhu, N.; Xu, M.Y.-C.; Wosinski, L.; Aitchison, J.S.; Ruda, H.E. Pillar-array based optical sensor. Opt. Express 2010, 18, 5420–5425. [CrossRef] [PubMed] Caucheteur, C.; Shevchenko, Y.; Shao, L.; Wuilpart, M.; Albert, J. High resolution interrogation of tilted fiber grating SPR sensors from polarization properties measurement. Opt. Express 2011, 19, 1656–1664. [CrossRef] [PubMed] Yu, Z.F.; Fan, S.H. Extraordinarily high spectral sensitivity in refractive index sensors using multiple optical modes. Opt. Express 2011, 19, 10029–10040. [CrossRef] [PubMed] Huang, L.; Tian, H.; Zhou, J.; Liu, Q.; Zhang, P.; Ji, Y. Label-free optical sensor by designing a high-Q photonic crystal ring–slot structure. Opt. Commun. 2015, 335, 73–77. [CrossRef] Sevilla, G.A.C.; Finazzi, V.; Villatoro, J.; Pruneri, V. Photonic crystal fiber sensor array based on modes overlapping. Opt. Express 2011, 19, 7596–7602. [CrossRef] [PubMed] Grepstad, J.O.; Kaspar, P.; Solgaard, O.; Johansen, I.; Sudbo, A.S. Photonic-crystal membranes for optical detection of single nano-particles, designed for biosensor application. Opt. Express 2012, 20, 7954–7965. [CrossRef] [PubMed] Wang, Y.; Wang, H.; Xue, Q.; Zheng, W. Photonic crystal self-collimation sensor. Opt. Express 2012, 20, 12111–12118. [CrossRef] [PubMed] Lai, W.; Chakravarty, S.; Zou, Y.; Guo, Y.; Chen, R.T. Slow light enhanced sensitivity of resonance modes in photonic crystal biosensors. Appl. Phys. Lett. 2013, 102, 041111–041113. [CrossRef] [PubMed] Fan, F.; Gu, W.; Wang, X.; Chang, S. Real-time quantitative terahertz microfluidic sensing based on photonic crystal pillar array. Appl. Phys. Lett. 2013, 102, 121113–121116. [CrossRef] Huang, L.; Tian, H.; Zhou, J.; Ji, Y. Design Low Crosstalk Ring-Slot Array Structure for Label-Free Multiplexed Sensing. Sensors 2014, 14, 15658–15668. [CrossRef] [PubMed] Hachuda, S.; Otsuka, S.; Kita, S.; Isono, T.; Narimatsu, M.; Watanabe, K.; Goshima, Y.; Baba, T. Selective detection of sub-atto-molar Streptavidin in 1013 -fold impure sample using photonic crystal nanolaser sensors. Opt. Express 2013, 21, 12815–12821. [CrossRef] [PubMed] Yang, D.; Tian, H.; Ji, Y.; Quan, Q. Nanoscale photonic crystal sensor arrays on monolithic substrates using side-coupled resonant cavity arrays. Opt. Express 2011, 19, 20023–20034. [CrossRef] [PubMed] Yang, D.; Tian, H.; Ji, Y. Nanoscale Low Crosstalk Photonic Crystal Integrated Sensor Array. IEEE Photonics J. 2014, 6, 4200107–4200107. [CrossRef] Zhou, J.; Tian, H.; Yang, D.; Liu, Q.; Huang, L.; Ji, Y. Refractive index sensing utilizing parallel tapered nano-slotted photonic crystal nano-beam cavities. J. Opt. Soc. Am. B 2014, 31, 1746–1752. [CrossRef] Zhou, J.; Tian, H.; Yang, D.; Liu, Q.; Ji, Y. Integration of high transmittance photonic crystal H2 nanocavity and broadband W1 waveguide for biosensing applications based on Silicon-on-Insulator substrate. Opt. Commun. 2014, 330, 175–183. [CrossRef] Falco, A.D.; O’Faolain, L.; Krauss, T.F. Chemical sensing in slotted photonic crystal heterostructure cavities. Appl. Phys. Lett. 2009, 94, 063503–063505. [CrossRef] Jágerská, J.; Zhang, H.; Diao, Z.; Thomas, N.L.; Houdré, R. Refractive index sensing with an air-slot photonic crystal nanocavity. Opt. Lett. 2010, 35, 2523–2525. [CrossRef] [PubMed] Caër, C.; Serna-Otálvaro, S.F.; Zhang, W.; Roux, X.L.; Cassan, E. Liquid sensor based on high-Q slot photonic crystal cavity in silicon-on-insulator configuration. Opt. Lett. 2014, 39, 5792–5794. [CrossRef] [PubMed] Zou, Y.; Chakravarty, S.; Zhu, L.; Chen, R.T. The role of group index engineering in series-connected photonic crystal microcavities for high density sensor microarrays. Appl. Phys. Lett. 2014, 104, 141103. [CrossRef] [PubMed]

Sensors 2016, 16, 1050

29.

30.

31. 32. 33. 34. 35. 36.

37.

38. 39. 40.

41.

10 of 10

Yan, H.; Zou, Y.; Chakravarty, S.; Yang, C.; Wang, Z.; Tang, N.; Fan, D.; Chen, R.T. Silicon on-chip bandpass filters for the multiplexing of high sensitivity photonic crystal microcavity biosensors. Appl. Phys. Lett. 2015, 106, 121103. [CrossRef] [PubMed] Chen, J.H.; Huang, Y.T.; Yang, Y.L.; Lu, M.F. Design, Fabrication, and Characterization of Si-Based ARROW-B Photonic Crystal Sharp-Bend Waveguides and Power Splitters. J. Lightwave Technol. 2012, 30, 2345–2351. [CrossRef] Yang, D.; Tian, H.; Ji, Y. High-bandwidth and low-loss photonic crystal power-splitter with parallel output based on the integration of Y-junction and waveguide bends. Opt. Commun. 2012, 285, 3752–3757. [CrossRef] Jensen, J.S.; Sigmund, O. Topology optimization of photonic crystal structures: A high-bandwidth low-loss T-junction waveguide. J. Opt. Soc. Am. B 2005, 22, 1191–1196. [CrossRef] Fan, S.H.; Johnson, S.G.; Joannopoulos, J.D.; Manolatou, C.; Haus, H.A. Loss-induced on/off switching in a channel add/drop filter. Phys. Rev. B 2001, 64, 245302. [CrossRef] Zheng, W.; Xing, M.; Ren, G.; Johnson, S.G.; Zhou, W.; Chen, W.; Chen, L. Integration of a photonic crystal polarization beam splitter and waveguide bend. Opt. Express 2009, 17, 8657–8668. [CrossRef] [PubMed] Thorhauge, M.; Frandsen, L.H.; Borel, P.I. Efficient photonic crystal directional couplers. Opt. Lett. 2003, 28, 1525–1527. [CrossRef] [PubMed] Zhou, J.; Tian, H.; Yang, D.; Liu, Q.; Huang, L.; Ji, Y. Low-loss, efficient, wide-angle 1 ˆ 4 power splitter at ~1.55 µm wavelengths for four play applications built with a monolithic photonic crystal slab. Appl. Opt. 2014, 53, 8012–8019. [CrossRef] [PubMed] Tee, D.C.; Kambayashi, T.; Sandoghchi, S.R.; Tamchek, N.; Adikan, F.R.M. Efficient, Wide Angle, Structure Tuned 1 ˆ 3 Photonic Crystal Power Splitter at 1550 nm for Triple Play Applications. J. Lightwave Technol. 2012, 30, 2818–2823. Lumerical Solutions, Inc. Available online: http://www.lumerical.com (accessed on 12 June 2015). Quan, Q.; Lonˇcar, M. Deterministic design of wavelength scale, ultra-high Q photonic crystal nanobeam cavities. Opt. Express 2011, 19, 18529–18542. [CrossRef] [PubMed] Wang, X.; Shi, W.; Yun, H.; Grist, S.; Jaeger, N.A.F.; Chrostowski, L. Narrow-band waveguide Bragg gratings on SOI wafers with CMOS-compatible fabrication process. Opt. Express 2012, 20, 15547–15558. [CrossRef] [PubMed] Bogaerts, W.; Selvaraja, S.K.; Dumon, P.; Brouckaert, J.; de Vos, K.; van Thourhout, D.; Baets, R. Silicon-on-insulator spectral filters fabricated with CMOS technology. IEEE J. Sel. Top. Quantum Electron. 2010, 16, 33–44. [CrossRef] © 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).