A nanowaveguide platform for collective atom-light interaction
AA nanowaveguide platform for collective atom-light interaction
Y. Meng, a) J. Lee, a) M. Dagenais, b) and S. L. Rolston Department of Electrical and Computer Engineering, University of Maryland, College Park,Maryland 20742, USA Joint Quantum Institute, Department of Physics, University of Maryland and NIST, College Park,Maryland 20742, USA
We propose a nanowaveguide platform for collective atom-light interaction through evanescent field coupling.We have developed a 1 cm-long silicon nitride nanowaveguide can use evanescent fields to trap and probe anensemble of Rb atoms. The waveguide has a sub-micrometer square mode area and was designed with tapersfor high fiber-to-waveguide coupling efficiencies at near-infrared wavelengths (750 nm to 1100 nm). Inversetapers in the platform adiabatically transfer a weakly guided mode of fiber-coupled light into a stronglyguided mode with an evanescent field to trap atoms and then back to a weakly guided mode at the other endof the waveguide. The coupling loss is − ∼
80 % coupling efficiency) at 760 nm and 1064 nm,which is estimated by a propagation loss measurement with waveguides of different lengths. The proposedplatform has good thermal conductance and can guide high optical powers for trapping atoms in ultra-highvacuum. As an intermediate step, we have observed thermal atom absorption of the evanescent componentof a nanowaveguide, and have demonstrated the U-wire mirror magneto-optical trap that can transfer atomsto the proximity of the surface.Atom-light interactions can be harnessed for a num-ber of quantum-based applications, such as quantuminformation processing and atomic sensing . En-hancing the atom-light interaction with small mode ar-eas has been demonstrated in several platforms such ashollow-core fibers , hollow-core waveguides , ta-pered fibers and a nanowaveguide . An array ofa few micrometer-square free-space optical spots in atrench from an optical waveguide array were also recentlyused to interrogate ultracold atoms formed by a mirrormagneto-optical trap (MOT) above the trench . In con-trast to thermal alkali vapors or a free-space modefrom an optical waveguide array , laser-cooled ultracoldatoms trapped in the evanescent field of a waveguide canexhibit much stronger coupling between the atoms andphotons due to good overlap of the atoms with the smalloptical mode, and the lack of Doppler broadening. In ad-dition, cold trapped atoms have a longer residence timein the field and thus a longer coherence time. All-opticalswitching with optically trapped atoms has previouslybeen implemented in a hollow fiber , and nanofiber-based optical lattices for Cs and Rb atoms were alsodemonstrated . Although nanofiber atom traps cre-ate strong atom-light interactions, they are not easilyscalable to realize complex photonic circuits, while anintegrated optics approach with waveguides enjoys bothstrong interactions and the potential for scalability.Here we present a potentially scalable nanowaveguideplatform, which can operate as the atom-light inter-face between an evanescent field probe and cold neutralatoms. An open-window in the middle of the waveg-uide is used to trap and probe atoms through the evanes-cent field. Atoms loaded from an atom-chip mirror MOT a) Y. Meng and J. Lee contributed equally to this work. b) Electronic mail: [email protected]. will be optically-trapped with a two-color evanescent fieldatom trap provided by red-detuned counterpropagatingwaveguide fields and a blue-detuned traveling waveguidefield with van der Waals potential. For the atom trap,it is necessary to perform measurements characterizingthe optical properties of such waveguides, focusing ondesigns that can simultaneously support the two opticalfrequencies.Fig. 1 (a) shows the schematic image of a nanowaveg-uide platform, and Fig. 3 (a), (b) shows the image of thereal waveguide sample. Instead of using a nanofiber or a nanophotonic crystal waveguide , we consideran integrated design of a nanowaveguide and an atomchip mirror MOT which has the advantage of betterheat dissipation and good scalability. The transportingand loading procedure from a mirror MOT above thesurface to the free-space mode of an optical waveguidehaving a few micrometer square area (4 × µm ) in thetrench structure was previously demonstrated . In ourexperiment, nanowaveguide atom trapping with a sub-micrometer square evanescent field mode will be possibleand the two-color evanescent field atom trap minimumwill be around 150 nm above the waveguide surface (seeFig. 1 (a)(i)).Using the simulation software, FimmProp (see Fig. 2),we have designed a nanowaveguide (800 nm ×
300 nm, n Si N = 2 and n SiO = 1.46) that guides 760 nm blue-detuned trapping light, 1064 nm red-detuned trappinglight, and 780 nm probe light for Rb atoms . Aninverse-tapered structure that expands the waveguidemode is widely used as a mode converter, and can of-fer good mode matching and coupling efficiency betweena nanowaveguide and an optical fiber. For instance,some previous work on the use of tapers have demon-strated good performance at telecom wavelengths , butits advantages and applicability have seldom been re-ported at near-infrared wavelengths in situations where1 a r X i v : . [ phy s i c s . a t o m - ph ] A ug a) Inverse Taper SiSiO Si N (b) (c) y xz xz MOT atomsSiO y xz FiberNanowaveguideSiO zzSi Au(i) (ii) SiSiO Si N xz Si N Au (0.1µm) Au (0.1µm) U z [ µ K ] Distance [µm] (i)
Atoms
FIG. 1. The concept of a nanowaveguide platform forcollective atom-light interaction. (a) Integration of ananowaveguide and a mirror MOT (3D-plot, not-to-scale)(i) The total potential along the vertical direction ofthe waveguide for Rb atoms ( λ Rb = 780 nm) is U tot = πc ω (cid:16) Γ∆ blue I blue ( r ) + Γ∆ red I red ( r ) (cid:17) − C vdW z , where opticalpowers of a blue-detuned traveling waveguide field (760 nm)and one of red-detuned counterpropagating fields (1064 nm)are P blue (cid:39)
12 mW and P red (cid:39)
10 mW. (b) Silicon nitride(Si N ) core and inverse tapers of the nanowaveguide (3D-plot, not-to-scale). (c) 2-D geometry of the nanowaveguide(i) Side view of the nanowaveguide input (xz plane, not-to-scale). (ii) Cross-sectional view of the nanowaveguide at theopen-window (xz plane, not-to-scale); the width of the open-window is (cid:39) µ m. many wavelengths are used between 750 nm and 1100 nm,which is the typical situation used in atomic physics (ourwaveguide design is also compatible with trapping Csatoms as well) and in nonlinear optics. Fig. 2 (a) showsthe simulation, which demonstrates that our taper struc-ture can achieve high coupling efficiency for a wide near-infrared wavelength range.According to the simulation, the horizontal and verti-cal TE mode field sizes (1 /e in power) of the waveguidewithout inverse tapers are (703 nm ,
377 nm) at 760 nmand (765 nm ,
474 nm) at 1064 nm. For a vacuum com-patible platform, we used a fiber-to-waveguide couplingapproach based on gluing the fiber to the waveguide chipwith UV-epoxy (Epotek OG 116-31). We used FibercoreSM750 (core diameter = 3.82 µ m, cladding diameter =125 µ m, NA = 0.14) fiber, with mode-field diameters of4.32 µ m at 760 nm ( n core = 1.46077, n clad = 1.45405) and5.4 µ m at 1064 nm ( n core = 1.45635, n clad = 1.44963).The mode mismatch between the fiber and the waveg-uide would lead to a coupling efficiency of only around3 %, which is too low to provide sufficient intensity totrap atoms with reasonable laser powers. We thus de-signed an inverse taper to improve the mode matchingand increased the measured coupling efficiency. Simula-tions show that the coupling efficiencies of ∼
80 % and the
Taper−end’s width (nm) C oup li ng ( % ) Wavelength (nm)(a)(b)(c) C oup li ng ( % ) C oup li ng ( % ) C oup li ng Lo ss ( d B ) C oup li ng Lo ss ( d B ) C oup li ng Lo ss ( d B ) -10-2-3-6-2-2 FIG. 2. The simulation of the fiber-to-waveguide couplingwith the TE mode (a) The wavelength vs. the coupling ef-ficiency with 500 µ m long and 70 nm wide tapers. (b) Therectangle length at the end of inverse tapers vs. the couplingefficiency for 760 nm (blue) and 1064 nm (red). (c) The widthof inverse taper-ends vs. the coupling efficiency for 760 nm(blue) and 1064 nm (red) with an extra 50 µ m-long rectangle. coupling losses of − µ m taper length that satisfiesthe adiabaticity condition; Fig. 2 (c) shows the simula-tion. A longer taper, such as 1 mm in length, can increasethe coupling efficiency by a few percent, but a longer lin-ear taper also induces more propagation loss and morestitching errors during electron beam lithography (EBL)in practice. Therefore, we chose to work with a 500 µ mtaper. In addition, the coupling efficiency is very sen-sitive to the taper-end width especially for 760 nm asshown in Fig. 2 (c), which makes cleaving another impor-tant issue. Our cleaving technique offers a near perfectmirror-like facet but may have a ± µ m error in cleavingposition. If we cleave 40 µ m inside the designed taper, thecoupling efficiency for 760 nm will drop to around 25 %from 79 %. We thus added a rectangular structure at thetaper-end to compensate for a potential cleaving error.As we simulated in Fig. 2(b), an additional 50 µ m-longrectangle only decreases the coupling efficiency by about3 %. In this way, we can reliably achieve a high couplingefficiency.The general fabrication process of our nanowaveg-uide is described in Fig. 3 (c). A one-centimeter longSi N waveguide with an inverse taper at both ends isfabricated on a 5 µ m-thick thermal SiO layer (Fig. 1(b)). The 300 nm Si N was deposited by LPCVD, andEBL was used to pattern the nanowaveguide and cre-ate an inverse-tapered structure. Due to our limitedEBL writing-field size of order one hundred microme-2 iSi SiO Ebeam Resist Si N SiER ERSiER ERChromiumSi N DepositionE-beam LithographyCr Deposition SiICP Etching Cr Etching PECVD SiO
Deposition SiSi SiSiPRSiPRPhoto ResistSihoto ResDevelop Photo ResistPhoto Lithography Au DepositionBOE EtchingUVSiLift-off SiLift-off Si N SiO SiO SiO SiO SiO SiO SiO SiO SiO SiO SiO Si N Si N Si N Si N Si N Si N Si N Si N Si N Si N (a) Open windowGold MirrorInverse Taper Nanowaveguide (b)
Inverse Taper End (c) (d) (e)
Waveguide Length (cm) O p t i c a l Lo ss ( d B ) FIG. 3. Nanowaveguide fabrication and propagation loss mea-surement (a) The real waveguide sample; taper length 500 µ m, waveguide length = 1cm. (b) The SEM image of an in-verse taper-end. (c) Fabrication Process. (d) Propagation lossmeasurement patterns with multiple waveguides in differentlengths; inset is the image of a waveguide pattern with a bend-ing. (e) The optical loss vs. the waveguide length. The esti-mated propagation losses from fitting lines are − . − . − .
11 dB at 760 nm and − .
044 dB at 1064 nm. The coupling losses are − .
055 dB perfacet (78.4 % coupling efficiency) at 760 nm and − .
022 dB perfacet (79 % coupling efficiency) at 1064 nm. ters, a one-centimeter long waveguide will cross 100 writ-ing fields (each field being 100 µ m by 100 µ m) leading torandom stitching errors at the writing field boundaries.In order to get a continuous, long waveguide, a FixedBeam Moving Stage (FBMS) and writing-field overlap-ping technique were used to minimize stitching errors.Since we have a 5- µ m-thick SiO insulating substrate,another conducting polymer was deposited on top of theebeam resist (PMMA) to reduce charging effects. Aftere-beam lithography, inductively coupled plasma (ICP)fluorine etching was used to etch the Si N . We usedPECVD to deposit another 4 µ m-thick SiO above the Si N core. Then with a photolithography pattern andgold deposition, an open-window is made by bufferedoxide etch (BOE) in the middle of the sample on thewaveguide. The window opening in the center of thewaveguide exposes the waveguides to vacuum and thusatoms for trapping. Finally, the sample is cleaved forfiber coupling. Since the atom trap needs to operate in ahigh vacuum environment, the optical alignment betweenthe fiber and the waveguide should be permanently fixed.Consequently, we glued the fiber onto the waveguide facetwith UV epoxy.To evaluate the coupling efficiency, we measured thepropagation loss of the waveguide. The cut-back tech-nique has been widely used for waveguide propagationloss measurements, but the accuracy of this measure-ment is highly dependent on the cleave quality. Smalldifferences in the cleaving may lead to large changes inthe coupling loss, which then make the propagation lossmeasurement unreliable. Here we use another method tomeasure the propagation loss without cleaving the chipmore than twice. Figure 3 (d) shows the pattern usedfor this measurement. Each waveguide contains 2 half-circle bends which are exactly the same. They have a400 µ m diameter, which is used to minimize the bendingloss. Each waveguide is 4 mm different in length withthe neighboring waveguide. After cleaving just once, wecan assume that the coupling loss for each waveguideis the same so it offers better accuracy than the tradi-tional cut-back technique. We estimated the couplingefficiencies at 760 nm and 1064 nm by the propagationloss measurement with waveguides of different lengths.Assuming both facets have the same coupling efficiency,the coupling losses per facet are − ∼
80 % for 760 nm and 1064 nmas shown in Figure 3 (e). The measured coupling efficien-cies are well matched with the simulation results shownin Fig. 2.By exposing the waveguide to a thermal vapor of Rbatoms, we were able to observe atomic absorption of theevanescent field of the waveguide (see Fig. 4 (a)), sim-ilar to that observed in a reference . The absorptionspectrum is significantly broadened, both due to Dopplerbroadening and transit-time broadening through the sub-micrometer scale waveguide mode. To make this mea-surement we used a L-shaped nanowaveguide (shown inFig. 4 (a) inset) to suppress light diverging through theSiO layer entering the output port, by more than 60 dB.We also produced the U-wire mirror MOT atoms (U-MOT) around the surface of a centimeter square goldmirror as shown in Fig. 4 (b), which is similar to an atomchip mirror MOT . The U-MOT atoms are createdby four cooling beams with the external U-wire (Fig. 4(b)(ii)) and the Helmholtz bias coil; we chose the ex-ternal U-wire instead of the microfabricated U-wire forsimplicity. We transported the steady state U-MOT’scenter position toward the mirror surface by adjustingthe magnetic field zero (see Fig. 4 (b)(v)), but found aU-MOT (cid:46) µ m significantly decreased the atom num-3 OTxyz (b) (c)(a) Rb: F=3 to F’ Rb: F=2 to F’Frequency (GHz) -2 -1 1 T r an s m i ss i on A side-cut nanowaveguide sample
AlN substrateend-capGold mirror
An integrated nanowaveguide atom-chip (iii) y x yz x y (ii)(iV) (V) z yz y (i) FIG. 4. (a) Absorption profile of thermal atoms with a L-shaped nanowaveguide. The profile shows Rb and Rb’sD2 transitions. The data (red line) of thermal atoms wasmeasured at (cid:104) T (cid:105) (cid:39) ◦ C with a nanowaveguide probe ( P (cid:39) µ W). The measured data has the total broadening of2 π ·
765 MHz (HFHM). The black line is the calculated ab-sorption profile only with Doppler-broadening in a similar wayof a reference , and the saturation power is P sat (cid:39)
25 nW,where
I/I sat = P/P sat (cid:39)
80; The interaction length ofthe L-shaped waveguide is l int = 3 mm; the energy of a780 nm waveguide probe coupled to thermal atoms is 3 . × × , P (cid:39) × − mbar); the fiber-couplednanowaveguide (white dotted line) is positioned at the edgeof the side-cut nanowaveguide sample. (i) A compact glasschamber with a 500 µ m thin AlN substrate end-cap. (ii)U-wire and Z-wire below the AlN substrate end-cap. (iii)Top view of a fiber-coupled side-cut nanowaveguide sampleabove the AlN substrate end-cap assembled with a gold de-posited silicon substrate mirror (xy plane). (iv) Side viewof U-MOT atoms (yz plane, no waveguide) (v) Side view oftransported U-MOT atoms (yz plane, no waveguide); the U-MOT atoms are centered at ∼ µ m from the surface, andthe edge of U-MOT atoms is contiguous with the surface. Thetotal number of U-MOT atoms ( N (cid:39) × ) decreases asthe atom center moves closer to the surface. (c) A compactglass chamber with a fiber-coupled, integrated nanowaveguideatom-chip (1 . × , see Fig. 1 (a)) for atom trapping,where nanowaveguides are fabricated along y-axis and MOTatoms are transported along z-axis. ber. Thus atom transport to the evanescent field atomtrap a few hundred nm from the waveguide surface willbe necessary. Microfabricated wires on the surface maybe useful to bring atoms closer to the surface to allowloading efficiency into the evanescent field atom trap.The detection of atomic absorption from the U-MOTwas unsuccessful with the fiber-coupled nanowaveguideof the side-cut sample, as the shadow cast by residualsilicon substrate nearby the nanowaveguide limited the cold atom density directly above the waveguide. Thiscan be improved by using an integrated waveguide (Fig. 4(c)), reducing the opening in the gold mirror to reducethe shadow, using some small angle beams to providecooling at the waveguide surface, or dropping the atomsand switching on the waveguide field, or transferring theatoms to a purely magnetic trap.In conclusion, we have designed a silicon nitrideevanescent-field nanowaveguide platform for trapping Rbatoms with wavelengths of 760 nm ( λ blue ) and 1064 nm( λ red ) and probing Rb atoms with the wavelength of780 nm ( λ Rb ). In a UHV environment, this nanowaveg-uide platform is compatible with the atom chip U-MOTconfiguration that will laser cool and confine atoms nearthe waveguide. The simulation of the nanowaveguidewith inverse tapers shows high coupling efficiencies overa wide range of near-infrared wavelengths from 750 nmto 1100 nm. A propagation loss measurement based ona modified “cut-back” technique was used to confirmthe high coupling efficiency per facet measured exper-imentally for the two-color atom-trap wavelengths. Acentimeter-long nanowaveguide with a two-color evanes-cent field atom trap is expected to confine many atomsloaded from a mirror MOT, creating a high opticaldepth atomic sample with a strong atom-light interac-tion. As a step toward the realization of an evanescentfield atom-light interaction, we measured an absorptionprofile of thermal atoms with a L-shaped nanowaveg-uide, and the absorption spectrum was strongly broad-ened due to Doppler broadening and transit-time broad-ening. In addition, we demonstrated U-MOT atoms ina compact UHV glass chamber, and the created U-MOTatoms are transportable to the surface. Such nanowaveg-uide atom traps may operate as an optically guided atominterferometer , and can be used for inertial atomic sen-sors or atomic magnetometry. In general, the evanescentfield nanowaveguide platform can also be used for bio-molecule sensing , gas detection, and chemical solutionsensing, which may be enhanced with dual color opera-tion. Moreover, because of its good heat dissipation capa-bility through its substrate, this scalable nanowaveguideplatform has the potential for use in the implementationof collective-atom-based quantum networks. ACKNOWLEDGMENTS
This work is funded by ARO Atomtronics MURIproject.
REFERENCES J. V. Porto, S. Rolston, B. Laburthe-Tolra, C. J.Williams, and W. D. Phillips,“Quantum informationwith neutral atoms as qubits,” Philos. T. Roy. Soc. A , 1808, 14171427 (2003).4
A. M. Marino, R. C. Pooser, V. Boyer, and P. D.Lett, “Tunable delay of EinsteinPodolskyRosen entan-glement,” Nature , 859862 (2009). A. I. Lvovsky, B. C. Sanders, and W. Tittel, “Opticalquantum memory,” Nat. Photon. , 706714 (2009). G. K. Brennen, I. H. Deutsch, and P. S. Jessen, “Quan-tum logic gates in optical lattices,” Phys. Rev. Lett. ,1060(1999). H. J. Kimble, “The quantum internet,” Nature ,1023 (2008). K. S. Choi, H. Deng, J. Laurat, and H. J. Kimble,”Mapping photonic entanglement into and out of aquantum memory,” Nature , 67-71 (2008). S. M. Dickerson, J. M. Hogan, A. Sugarbaker, D. M. S.Johnson, and M. A. Kasevich, “Multiaxis inertial sens-ing with long-time point source atom interferometry,”Phys. Rev. Lett. , 083001 (2013). H. J. McGuinness, A. V. Rakholia, and G. W. Bieder-mann, “High data-rate atom interferometer for mea-suring acceleration,” Appl. Phys. Lett. , 011106(2012). I. K. Kominis, T. W. Kornack, J. C. Allred, and M. V.Romalis, “A subfemtotesla multichannel atomic mag-netometer,” Nature , 596599 (2003). A. D. Slepkov, A. R. Bhagwat, V. Venkataraman, P.Londero, and A. L. Gaeta, “Spectroscopy of Rb atomsin hollow-core fibers,” Phys. Rev. A , 053825 (2010). K. Saha, V. Venkataraman, P. Londero, and A. L.Gaeta, “Enhanced two-photon absorption in a hollow-core photonic-band-gap fiber,” Phys. Rev. A ,033833 (2011). W. Yang, D. B. Conkey, B. Wu, D. Yin, A. R. Hawkins,and H. Schmidt, “Atomic spectroscopy on a chip,” Nat.Photon. , 331335 (2007). B. Wu, J. F. Hulbert, E. J. Lunt, K. Hurd, A. R.Hawkins, and H. Schmidt, “Slow light on a chip viaatomic quantum state control,” Nat. Photon. , 776779(2010). S. M. Spillane, G. S. Pati, K. Salit, M. Hall, P. Kumar,R. G. Beausoleil, and M. S. Shahriar, “Observation ofnonlinear optical interactions of ultralow levels of lightin a tapered optical nanofiber embedded in a hot ru-bidium vapor,” Phys. Rev. Lett. , 233602 (2008). S. Hendrickson, M. Lai, T. Pittman, and J. Franson,“Observation of two-photon absorption at low powerlevels using tapered optical fibers in rubidium vapor,”Phys. Rev. Lett. , 173602 (2010). L. Stern, B. Desiatov, I. Goykhman, and U.Levy, “Nanoscale light-matter interactions in atomiccladding waveguides,” Nature Communications , Ar-ticle number: 1548 (2013). M. Kohnen, M. Succo, P. G. Petrov, R. A. Nyman, M. Trupke, and E. A. Hinds “An array of integrated atom-photon junctions,” Nature Photonics , 3538 (2011)doi:10.1038/nphoton.2010.255. M. Bajcsy, S. Hofferberth, V. Balic, T. Peyronel, M.Hafezi, A. S. Zibrov, V. Vuletic, and M. D. Lukin. “Ef-ficient all-optical switching using slow light within ahollow fiber,” Phys. Rev. Lett. , 203902 (2009). E. Vetsch, D. Reitz, G Sag´ue, R. Schmidt, S. T.Dawkins, and A. Rauschenbeutel, “Optical interfacecreated by laser-cooled atoms trapped in the evanes-cent field surrounding an optical nanofiber,” Phys. Rev.Lett. , 203603 (2010). J. Lee, J. A. Grover, J. E. Hoffman, L. A. Orozco, andS. L. Rolston, “Inhomogeneous broadening of opticaltransitions of Rb atoms in an optical nanofiber trap,”J. Phys. B: At. Mol. Opt. Phys. J. D. Thompson, T. G. Tiecke, N. P. de Leon, J. Feist,A. V. Akimov, M. Gullans, A. S. Zibrov, V. Vuleti´c,and M. D. Lukin, “Coupling a single trapped atom toa nanoscale optical cavity,” Science , no. 6137, pp.1202-1205 (2013). A. Goban, C.-L. Hung, S.-P. Yu, J. D. Hood, J. A. Mu-niz, J.H. Lee, M. J. Martin, A. C. McClung, K. S. Choi,D. E. Chang, O. Painter, and H. J. Kimble, “Atom-light interactions in photonic crystals,” Nat. Commun. :3808 doi: 10.1038/ncomms4808 (2014). J. Reichel, W. H¨ansel, and T. W. H¨ansch, “Atomicmicromanipulation with magnetic surface traps,” Phys.Rev. Lett. , 3398-3401 (1999). J. Lee, D H. Park, S. Mittal, M. Dagenais and S. L.Rolston, “Integrated optical dipole trap for cold neutralatoms with an optical waveguide coupler,” New Journalof Physics. , 043010 (2013). V. R. Almeida, R. R. Panepucci, and M. Lip-son,“Nanotaper for compact mode conversion,” Opt.Lett. , 1302 (2003). P. Siddons, C. S. Adams, C. Ge, and I. G. Hughes,“Absolute absorption on rubidium D lines: comparisonbetween theory and experiment,” J. Phys. B: At. Mol.Opt. Phys. G. D. McDonald, H. Keal, P. A. Altin, J. E. Debs, S.Bennetts, C. C. N. Kuhn, K. S. Hardman, M. T. Johns-son, J. D. Close, and N. P. Robins, “Optically guidedlinear Mach-Zehnder atom interferometer,” Phys. Rev.A , 013632 (2013). M. Weisser, G. Tovara, S. Mittler-Nehera, W. Knolla,F. Brosinger, H. Freimuthb, M. Lacherb, and W.Ehrfeldb, “Specific bio-recognition reactions observedwith an integrated MachZehnder interferometer, ”Biosensors and Bioelectronics14