Ion-sensitive photonic-crystal nanolaser sensors

In general, biochemical sensors based on photonic cavities are used to detect changes in the refractive index of the environment. In this study, however, a GaInAsP semiconductor photonic-crystal nanolaser sensor that we recently developed was found to detect not only the environmental refractive index but also the surface charge. In contrast to the pH sensitivity we reported previously, this is an ultra-sensitive detection mechanism capable of identifying proteins and deoxyribonucleic acids (DNA) at a femtomolar-order or lower concentrations. When the device is exposed to plasma or DNA solutions, the laser wavelength simultaneously changes with the zeta potential and the flat-band potential of the semiconductor surface. This indicates that the charged functional groups on the surface, which are formed by these treatments, modify the Schottky barrier near the semiconductor surface, trap the excited carriers in the barrier, and change the refractive index of the semiconductor via the carrier effects. These findings also suggest that some other photonic sensors may also exhibit similar electrochemical and optoelectronic effects. © 2017 Optical Society of America OCIS codes: (280.4788) Optical sensing and sensors; (140.5960) Semiconductor lasers. References and links 1. W. C. Law, K. T. Yong, A. Baev, R. Hu, and P. N. Prasad, “Nanoparticle enhanced surface plasmon resonance biosensing: application of gold nanorods,” Opt. Express 17(21), 19041–19046 (2009). 2. A. M. Armani, R. P. Kulkarni, S. E. Fraser, R. C. Flagan, and K. J. Vahala, “Label-free, single-molecule detection with optical microcavities,” Science 317(5839), 783–787 (2007). 3. M. Iqbal, M. A. Gleeson, B. Spaugh, F. Tybor, W. G. Gunn, M. Hochberg, T. Baehr-Jones, R. C. Bailey, and L. C. Gunn, “Label-free biosensor arrays based on silicon ring resonators and high-speed optical scanning instrumentation,” IEEE J. Sel. Top. Quantum Electron. 16(3), 654–661 (2010). 4. H. Zhu, I. M. White, J. D. Suter, P. S. Dale, and X. Fan, “Analysis of biomolecule detection with optofluidic ring resonator sensors,” Opt. Express 15(15), 9139–9146 (2007). 5. S. Zlatanovic, L. W. Mirkarimi, M. M. Sigalas, M. A. Bynum, E. Chow, K. M. Robotti, G. W. Burr, S. Esener, and A. Grot, “Photonic crystal microcavity sensor for ultracompact monitoring of reaction kinetics and protein concentration,” Sens. Actuators B Chem. 141(1), 13–19 (2009). 6. M. Lončar, A. Scherer, and Y. Qiu, “Photonic crystal laser sources for chemical detection,” Appl. Phys. Lett. 82(26), 4648–4650 (2003). 7. S. Kita, S. Hachuda, S. Otsuka, T. Endo, Y. Imai, Y. Nishijima, H. Misawa, and T. Baba, “Super-sensitivity in label-free protein sensing using a nanoslot nanolaser,” Opt. Express 19(18), 17683–17690 (2011). 8. T. Baba, “Biosensing using photonic crystal nanolasers,” MRS Commun. 5(4), 555–564 (2015). 9. S. Hachuda, T. Watanabe, D. Takahashi, and T. Baba, “Sensitive and selective detection of prostate-specific antigen using a photonic crystal nanolaser,” Opt. Express 24(12), 12886–12892 (2016). 10. I. M. White and X. Fan, “On the performance quantification of resonant refractive index sensors,” Opt. Express 16(2), 1020–1028 (2008). 11. F. Vollmer, S. Arnold, D. Braun, I. Teraoka, and A. Libchaber, “Multiplexed DNA quantification by spectroscopic shift of two microsphere cavities,” Biophys. J. 85(3), 1974–1979 (2003). 12. F. Vollmer, D. Braun, A. Libchaber, M. Khoshsima, I. Teraoka, and S. Arnold, “Protein detection by optical shift of a resonant microcavity,” Appl. Phys. Lett. 80(21), 4057–4059 (2002). 13. X. Zhang, L. Liu, and L. Xu, “Ultralow sensing limit in optofluidic micro-bottle resonator biosensor by selfreferenced differential-mode detection scheme,” Appl. Phys. Lett. 104(3), 033703 (2014). 14. O. Scheler, J. T. Kindt, A. J. Qavi, L. Kaplinski, B. Glynn, T. Barry, A. Kurg, and R. C. Bailey, “Label-free, multiplexed detection of bacterial tmRNA using silicon photonic microring resonators,” Biosens. Bioelectron. 36(1), 56–61 (2012). Vol. 25, No. 20 | 2 Oct 2017 | OPTICS EXPRESS 24469 #303546 https://doi.org/10.1364/OE.25.024469 Journal © 2017 Received 28 Jul 2017; revised 22 Sep 2017; accepted 22 Sep 2017; published 26 Sep 2017 15. K. De Vos, J. Girones, T. Claes, Y. De Koninck, S. Popelka, E. Schacht, R. Baets, and P. Bienstman, “Multiplexed Antibody Detection With an Array of Silicon-on-Insulator Microring Resonators,” IEEE Photonics J. 1(4), 225–235 (2009). 16. A. L. Washburn, M. S. Luchansky, A. L. Bowman, and R. C. Bailey, “Quantitative, label-free detection of five protein biomarkers using multiplexed arrays of silicon photonic microring resonators,” Anal. Chem. 82(1), 69– 72 (2010). 17. A. L. Washburn, L. C. Gunn, and R. C. Bailey, “Label-Free Quantitation of a cancer biomarker in complex media using silicon photonic microring resonators,” Anal. Chem. 81(22), 9499–9506 (2009). 18. D. Dorfner, T. Zabel, T. Hürlimann, N. Hauke, L. Frandsen, U. Rant, G. Abstreiter, and J. Finley, “Photonic crystal nanostructures for optical biosensing applications,” Biosens. Bioelectron. 24(12), 3688–3692 (2009). 19. Y. Zou, S. Chakravarty, D. N. Kwong, W.-C. Lai, X. Xu, X. Lin, A. Hosseini, and R. T. Chen, “Cavitywaveguide coupling engineered high sensitivity silicon Photonic crystal microcavity biosensors with high yield,” IEEE J. Sel. Top. Quantum Electron. 20(4), 1–10 (2014). 20. S. Chakravarty, A. Hosseini, X. Xu, L. Zhu, Y. Zou, and R. T. Chen, “Analysis of ultra-high sensitivity configuration in chip-integrated photonic crystal microcavity bio-sensors,” Appl. Phys. Lett. 104(19), 191109 (2014). 21. S. Chakravarty, Y. Zou, W.-C. Lai, and R. T. Chen, “Slow light engineering for high Q high sensitivity photonic crystal microcavity biosensors in silicon,” Biosens. Bioelectron. 38(1), 170–176 (2012). 22. D. Yang, S. Kita, F. Liang, C. Wang, H. Tian, Y. Ji, M. Lončar, and Q. Quan, “High sensitivity and high Qfactor nanoslotted parallel quadrabeam photonic crystal cavity for real-time and label-free sensing,” Appl. Phys. Lett. 105(6), 063118 (2014). 23. F. Liang, N. Clarke, P. Patel, M. Lončar, and Q. Quan, “Scalable photonic crystal chips for high sensitivity protein detection,” Opt. Express 21(26), 32306–32312 (2013). 24. V. Toccafondo, J. García-Rupérez, M. J. Bañuls, A. Griol, J. G. Castelló, S. Peransi-Llopis, and A. Maquieira, “Single-strand DNA detection using a planar photonic-crystal-waveguide-based sensor,” Opt. Lett. 35(21), 3673–3675 (2010). 25. M. G. Scullion, A. Di Falco, and T. F. Krauss, “Slotted photonic crystal cavities with integrated microfluidics for biosensing applications,” Biosens. Bioelectron. 27(1), 101–105 (2011). 26. S. Pal, E. Guillermain, R. Sriram, B. L. Miller, and P. M. Fauchet, “Silicon photonic crystal nanocavity-coupled waveguides for error-corrected optical biosensing,” Biosens. Bioelectron. 26(10), 4024–4031 (2011). 27. T. Watanabe, Y. Furuta, S. Hachuda, Y. Nishijima, and T. Baba, “Electrochemical sensing by photonic crystal nanolaser sensors,” presented at Biosensors (2016). 28. J. P. Kim, B. Y. Lee, S. Hong, and S. J. Sim, “Ultrasensitive carbon nanotube-based biosensors using antibodybinding fragments,” Anal. Biochem. 381(2), 193–198 (2008). 29. A. Kim, C. S. Ah, H. Y. Yu, J. H. Yang, I. B. Baek, C. G. Ahn, C. W. Park, M. S. Jun, and S. Lee, “Ultrasensitive, label-free, and real-time immunodetection using silicon field-effect transistors,” Appl. Phys. Lett. 91(10), 103901 (2007). 30. G. Zheng, F. Patolsky, Y. Cui, W. U. Wang, and C. M. Lieber, “Multiplexed electrical detection of cancer markers with nanowire sensor arrays,” Nat. Biotechnol. 23(10), 1294–1301 (2005). 31. H. A. Budz, M. M. Ali, Y. Li, and R. R. LaPierre, “Photoluminescence model for a hybrid aptamer-GaAs optical biosensor,” J. Appl. Phys. 107(10), 104702 (2010). 32. Y. S. Liu, Y. Sun, P. T. Vernier, C. H. Liang, S. Y. C. Chong, and M. A. Gundersen, “pH-sensitive photoluminescence of CdSe/ZnSe/ZnS quantum dots in human ovarian cancer cells,” J. Phys. Chem. C 111(7), 2872–2878 (2007). 33. K. Watanabe, Y. Kishi, S. Hachuda, T. Watanabe, M. Sakemoto, Y. Nishijima, and T. Baba, “Simultaneous detection of refractive index and surface charges in nanolaser biosensors,” Appl. Phys. Lett. 106(2), 021106 (2015). 34. M. Sakemoto, Y. Kishi, K. Watanabe, H. Abe, S. Ota, Y. Takemura, and T. Baba, “Cell imaging using GaInAsP semiconductor photoluminescence,” Opt. Express 24(10), 11232–11238 (2016). 35. S. Kita, K. Nozaki, S. Hachuda, H. Watanabe, Y. Saito, S. Otsuka, T. Nakada, Y. Arita, and T. Baba, “Photonic crystal point-shift nanolaser with and without nanoslots—design, fabrication, lasing and sensing characteristics,” IEEE J. Sel. Top. Quantum Electron. 17(6), 1632–1647 (2011). 36. M. Narimatsu, S. Kita, H. Abe, and T. Baba, “Enhancement of vertical emission in photonic crystal nanolasers,” Appl. Phys. Lett. 100(12), 121117 (2012). 37. N. Jacob, Israelachvili, Intermolecular and Surface Forces Third Edition (Academic, 2011). 38. B. Seger, T. Pedersen, A. B. Laursen, P. C. K. Vesborg, O. Hansen, and I. Chorkendorff, “Using TiO2 as a conductive protective layer for photocathodic H2 evolution,” J. Am. Chem. Soc. 135(3), 1057–1064 (2013). 39. S. L. Chuang, Physics of Photonic Devices, 2nd ed. (WILEY, 2008). 40. M. Krijn, “Heterojunction band offsets and effective masses in III-V quaternary alloys,” Semicond. Sci. Technol. 6(1), 27–31 (1991). 41. K. Nozaki, S. Kita, and T. Baba, “Room temperature continuous wave operation and controlled spontaneous emission in ultrasmall photonic crystal nanolaser,” Opt. Express 15(12), 7506–7514 (2007). 42. M. Fujita, A. Sakai, and T. Baba, “Ultrasmall and ultralow threshold GaInAsP-InP microdisk injection lasers: design, fabrication, lasing characteristics, and spontaneous emission factor,” IEEE J. Sel. Top. Quantum Electron. 5(3), 673–681 (1999). Vol. 25, No. 20 | 2 Oct 2017 | OPTICS EXPRESS 24470


Introduction
Biosensors detect biomolecules such as proteins and deoxyribonucleic acids (DNAs) in a solution through a change in some physical quantity of the sensor.In this decade, various photonic biosensors based on optical resonance, which do not require labels such as fluorescent molecules or colorimetric substances, have been studied and developed [1][2][3][4][5].These sensors detect the target through a shift in the resonance wavelength or resonance angle of light when the target molecules are adsorbed on the sensor.Usually, this shift is considered to arise from an increase in the environmental refractive index near the sensor's surface, leading to a change in the equivalent index of the optical mode [6].Thus, photonic sensors have been recognized as environmental-index sensors.This classical assumption may be valid when the concentration of the target molecules in the solution is of the order of micromolar (μM) or higher, for which significant adsorption is expected.However, when the concentration is of the order of picomolar (pM) or lower, the shift in the resonance wavelength or resonance angle must be explained differently because the index change produced by the adsorption would be too small.To the best of our knowledge, however, no studies have been conducted to directly measure the number of molecules adsorbed on the surface by other means, to estimate the environmental and modal index changes, and to confirm the resonance shift that arises from these index changes.We have previously developed GaInAsP photonic-crystal (PC) nanolaser sensors that can be used to detect certain proteins at ultra-low concentrations, i.e., in the range of subattomolar (sub-aM) to pM, based on a wavelength shift Δλ [7][8][9].However, the order of magnitude of the observed Δλ was much larger than that expected for such low concentrations [10].The relation between Δλ and the concentration of biomolecules has been reported for several photonic sensors, including our nanolasers [2][3][4][5][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25][26][27], as summarized in Fig. 1.If the sensor responds to a change in the environmental index, the corresponding Δλ would also occur with the concentration (or the number of target molecules adsorbed).Silica cavities and Si micro-rings exhibit such a trend (represented by the green-shaded band).On the other hand, PC nanolasers and PC passive cavities exhibit Δλ of approximately the same order of magnitude even at concentrations lower than the order of pM (red-shaded band).However, this observation has not been addressed in these previous reports.In our previous experiment for the detection of prostate specific antigen (PSA), we doubted the dilution process used for the sample solutions and double-checked the concentration using an enzyme-linked immunosorbent assay (ELISA).However, we did not find any significant problems at sub-pM and higher concentrations, which is the dynamic range of the ELISA [9].Meanwhile, ion-sensitive field-effect transistors (IS-FETs) are well-known electronic biosensors that detect the electric charge of biomolecules and have a limit of detection between aM and femtomolar (fM) orders [28][29][30].The photoluminescence (PL) from photonic semiconductors, such as III-V compounds, is known to be influenced by the pH of the solution in which the semiconductors are immersed [31,32].This phenomenon is generally attributed to nonradiative surface recombination, which is affected by the pHdependent surface charge of the semiconductors.We have also observed that the PL intensity of GaInAsP/InP semiconductor wafers and the laser-emission intensity of GaInAsP nanolasers are influenced by the pH and adsorption of some charged polyelectrolytes [33,34].We applied these phenomena to sensing high concentrations of negatively charged DNA and to imaging living cells with locally modified pH through the change in the emission intensity while disregarding Δλ due to the change in the environmental index.Conversely, in this study, we demonstrate the possibility that the wavelength of a nanolaser is also very sensitive to the surface charge of the adsorbed biomolecules, even at ultra-low biomolecule concentrations.
In Section 2, we describe some interesting behaviors of the nanolaser wavelength in the presence of ultra-low concentrations of biomolecules that cannot be explained by the change in environmental index.In Section 3, we focus on the condition in which the effect of the change in surface charge is dominant over that of the change in the environmental index.We compare the observed temporal Δλ with the surface charge and the internal Schottky barrier of the semiconductor via measurement of the zeta potential and flat band potential, respectively, and confirm that the wavelength behavior correlates well with the change in these electrical properties.In Section 4, we discuss a mechanism that carrier effects such as optoelectronic effects, which are induced by these electrical properties, modify the modal equivalent index and shift the wavelength of the nanolaser.

Nanolaser and biomolecule sensing
The detailed fabrication process for the PC nanolaser, as shown in Fig. 2(a), is described in [35].The PC slab comprises a weak n-type GaInAsP air-bridge membrane that is approximately 185 nm thick with a triangular lattice of holes with a diameter of 250 nm and a pitch of 500 nm.The membrane comprises a single-quantum-well (SQW) of 4-5 nm thick and optical confinement layers (OCLs) with bandgap wavelengths of 1.20, 1.15, and 1.10 μm (band gap energy E g of 1.03, 1.08, and 1.13 eV, respectively).The PL peak wavelength is 1.55 μm.The H0-type laser cavity is formed by slightly shifting two or four adjacent airholes to the outside (we set s x = 120 nm and s y = 60 nm in this study).A nanoslot with a width of 40-60 nm is also formed at the center of the cavity to suppress the thermal effects, reduce the spectral width, and enhance the sensitivity.Moreover, a double-periodic modulation of the hole diameters around the cavity is applied, which induces asymmetry in the cavity structure, breaks unwanted destructive interference of the modal electric field, and forms a vertical output beam detectable by a 50 × objective lens [36].The entire exposed surface of the device is coated with an approximately 3-nm thick ZrO 2 layer by atomic layer deposition (ALD) (Ultratech/Cambridge Nanotech, Savannah) for chemical stabilization in a solution, in which the nanolaser is immersed for sensing.A fabricated nanolaser chip is placed between the glass slides filled with the solution and photopumped by a pulsed laser light at a wavelength of 0.98 μm through the objective lens.The laser emission is coupled with an optical fiber through the same lens, and its wavelength is measured using an optical spectrum analyzer with a wavelength resolution of 0.2 nm.In the biosensing experiments, we usually perform appropriate functionalization on a device chip of typically 1 mm 2 size, adsorb biomolecules in a solution in a microtube for 1 h, rinse by pure water (deionized water whose electrical resistance is 18.3 MΩ) in another microtube, remove and set between two panes of glass, inject pure water, as shown in Fig. 2(b), and measure the wavelength of 10-20 nanolasers on the chip using an automatic system.The nanolaser with the biomolecules adsorbed usually exhibits a red shift in wavelength, as shown in Fig. 2(c).Then, we remove it from within the two glass panes and place it into another microtube including the biomolecule solution of different concentration.We repeated this procedure to obtain the Δλ for various concentrations.The wavelength fluctuation was measured to be less than ± 0.1 nm during such a procedure when all the solutions are just pure water.From Fig. 1, we discussed an unexpectedly large Δλ in the PC nanolasers and PC cavities in the presence of ultra-low concentrations of proteins.Here we present additional details regarding the unique behaviors in detecting antibody-antigen interactions of proteins using the nanolaser.The functionalization procedure we employed was as follows.Immediately following the ALD of ZrO 2 , a 3-aminopropyltriethoxysilane (APTES) monolayer was selfassembled on the nanolaser surface in the same vacuum chamber as the silane-coupling treatment, which chemically binds inorganic ZrO 2 and the following organic material, i.e., glutaraldehyde in this experiment.After a liquid-phase treatment application of glutaraldehyde as the linker, an antibody was immobilized on the surface in pure water in a microtube at 37°C for 1 h.Then, the corresponding antigen protein was adsorbed in pure water in another microtube at 37°C for 1 h through the antibody-antigen interaction.After each process, the device chip was rinsed with pure water.Subsequently, the laser wavelength was measured in water, as mentioned above.This procedure was repeated for a range of concentrations from low to high.In these experiments, immunoglobulin G (IgG) and PSA were tested as targets.Figure 3(a) depicts the results obtained for IgG sensing.When the anti-IgG antibody (Sigma-Aldrich, M8642, 150 kDa) was immobilized at a concentration of 6.7 nanomolars (nM), Δλ was as small as 0.1 nm.However, this small shift can be reasonably attributed to the change in the environmental index due to the presence of the antibody.When the IgG antigen (Sigma-Aldrich, I5681, 150 kDa) was applied at a concentration of 10 fM, a shift of 0.4 nm was observed.Despite the molecular weights of the antibody and antigen being similar, meaning they provide similar impacts to the environmental index, and the antigen being at a 10 5 -fold lower concentration, the adsorption of the antigen exhibited a 4fold larger shift.A similar behavior was observed in the detection of PSA, as shown in Fig. 3(b).The shift due to 30 aM PSA (Sigma-Aldrich, P338, 33 kDa) was 3-fold larger than that due to the immobilization of 6.7 nM anti-PSA antibody (JQ7, JCL, 150 kDa).Thus, the shifts observed here do not appear to originate from changes in the environmental index.We also detected 12-mer DNA (Hokkaido System Science Co., Ltd., 5′-TGC ACA GAC TAG-3′, 3.6 kDa).We employed the single strand DNA rather than the double strand DNA although the latter might be more interesting in terms of biosensing, because the single strand one has a smaller molecular weight, meaning a smaller index effect.In this experiment, we particularly aimed at a much smaller index effect and much stronger negative-charge effect than those of proteins.To simplify the sensing, the DNA was physisorbed on the device in pure water with no functionalization after the deposition of ZrO 2 .As depicted in Fig. 3(c), a shift of 0.8 nm was observed in the presence of 10-fM single-stranded DNA, even though the molecular weight of the DNA is 1-2 orders smaller that of the IgG and PSA.Similar DNA has been previously detected using a silica cavity, but the behaviors reported were completely different from those observed in this experiment; in the previous study, the concentration was as high as 1 μM and the measured Δλ was as small as 0.03 nm [11].This comparison further indicates that Δλ of the nanolaser is not due to a change in the environmental index.

Response to changes in the surface charge
For the unique shifts shown in Section 2 and also for those observed previously in nanolasers (and probably in PC nanocavities), as shown in Fig. 1, we consider the charge effect, similar to that in IS-FETs, as a possible cause.In the antibody-antigen interactions, the antibody and antigen having similar steric structures and opposite electrical charges bind, and the charges are locally cancelled out [37], which may strongly influence the surface charge of the nanolaser.In the adsorption of DNA, the charge effect is stronger than that for proteins because DNA is negatively charged in pure water due to the ionization of the phosphoric acid groups.To verify this hypothesis, we directly examined the charge effect on the nanolaser.
To induce an electrostatic charge in the nanolaser chip, we expose the nanolaser chip to air plasma (SAKIGAKE-Semiconductor, YHS-R), as shown in Fig. 4(a).The duration of the exposure was 1 min because no additional effect was observed for longer treatment times.
The chip was immersed into pure water immediately after plasma exposure and the laser wavelength was measured.As shown in Fig. 4(b), the wavelength was found to be drastically blue-shifted immediately after the exposure (time t = 0 in the figure) and then gradually redshifted, approaching a wavelength approximately 0.3 nm shorter than that before the exposure at t > 30 min.This shorter wavelength is thought to be due to the surface oxidation through plasma exposure.The large blue shift and gradual red shift cannot be attributed to the change in the environmental index because no materials can be adsorbed on the nanolaser surface while submerged in pure water.
Since this behavior strongly suggests that the wavelength shift is related to the variation of the surface charge, we next investigated two electrical parameters, i.e., the zeta potential [37] and the flat band potential [38].It is well-known that the surface charge attracts counter ions in the solution and the Helmholtz layer is formed, at which the counter ions are strongly fixed face-to-face with the surface charge.The electric potential of this plane against the bulk solution region is defined as the Helmholtz potential V H .The counter ions located slightly further from the surface are also bound weakly.The interface of such ions is called the slipping plane and its potential is known as the zeta potential V Zeta .In biochemical experiments, V Zeta is usually measured as a measure for V H and the surface charge.In this experiment, we measured V Zeta for the original epiwafer of the nanolaser as a function of t, as shown in Fig. 4(c), using a zeta potential analyzer (Otsuka Electronics Co. Ltd., ELSZ-2000Z), which evaluates the electroosmotic flow near the wafer surface based on the laser Doppler shift.Figure 4(c) depicts the change in the surface charge from positive to negative or vice versa during the period after the plasma exposure.The polarity characteristics evaluated based on the electroosmotic flow might be influenced by the complicated electrostatic conditions in the analyzer chamber.However, if we focus only on the variation in V Zeta , it converged at t > 30 min.This observation indicates that a certain functional group such as OH 2+ was modified on the nanolaser surface by the plasma exposure and slowly neutralized by reactions with OH -ions in pure water.
Such a change in the surface charge usually modifies the electronic potential of a semiconductor.Under these conditions, the electrons, which are the majority carriers in the ntype semiconductor used here, leak into the solution through the thin ZrO 2 film so that the Fermi level E F of the semiconductor, which is almost equal to the conduction band, and the redox potential E Red in the solution are balanced, causing a Schottky barrier to form near the surface [38].The barrier height ΔU is approximated as E Red -U fb where U fb is the flat-band potential (the surface energy level of the conduction band).Since we did not use any pH solution in the experiments of this study, E Red was kept constant, and ΔU is determined only by U fb .Therefore, we measured U fb with time t by measuring the capacitance of the space charge layer of the original epiwafer (Mott-Schottky plot analysis), as shown in Fig. 4(c).Initially, U fb shifted to a negative value immediately after the plasma exposure, indicating an increase in ΔU, but it slowly returned to its initial potential and equilibrated at t > 30 min.Thus, the surface charge and barrier height behave in a manner similar to Δλ.
To confirm the relationship between the nanolaser's wavelength and the charge effect in the presence of low concentrations of biomolecules, V Zeta and U fb were also measured following the physical adsorption of the DNA, as shown in Fig. 5.Both V Zeta and U fb began to change at a concentration around 1 fM, which corresponds well with the Δλ shown in Fig. 3(c).To explain these behaviors, we consider the possibility that a limited number of charged DNA molecules change the functional groups on the entire surface of the device.The surface charge represented by V Zeta shifted to more negative values with increasing DNA concentrations, which is a reasonable result considering the negative charge of DNA.In contrast, it might be reasonable to consider that the negative surface charge modifies the flat band potential to the negative direction, but as shown in Fig. 5, U fb shifted to more positive values.From this result, we consider that the adsorption of a small amount of DNA changes the OH groups on most ZrO 2 surface without DNA to OH 2+ , leading to the flat band potential being shifted positively due to the positive surface charge and ΔU decreased.The negative shift of V Zeta and positive shift of U fb seen in Fig. 5 and the redshift of the wavelength in Fig. 3(c) are all similar to those observed following the plasma exposure.

Mechanism of wavelength shift
In this section, we discuss the mechanism by which ΔU influences the index change Δn sem in the semiconductor, leading to a modal equivalent index change Δn eq and a wavelength shift Δλ. Figure 6 summarizes the electrochemical and optoelectronic phenomena that occur inside and outside of the semiconductor.The effect of ΔU can occur on the front and rear surfaces of the PC slab where a U-shaped potential is formed in the OCLs sandwiching the SQW.The ΔU can also occur on the sidewalls of the holes, where a complicated potential distribution is formed from multiple holes.In the following, we mainly discuss the effect of the U-shaped vertical potential.Regarding the lateral ΔU, we finally discuss our expectation that its effect on Δλ is small.The wavelength shift Δλ is related to Δn eq as Δλ/λ = Δn eq /n eq .If we assume λ = 1,550 nm and n eq = 2.6 as typical and temporary values, Δn eq required for Δλ > 1 nm (seen in Figs. 3  and 4) is calculated as 1.7 × 10 −3 .In the U-shaped vertical potential, we can consider an optoelectronic phenomenon that modifies n sem in the SQW and in the OCLs.If we neglect the existence of hole arrays in the PC slab, the optical confinement factor in the SQW, Γ SQW , is calculated using the transfer matrix method as 2.1% when we approximate n sem = 3.4 and n env = 1.35 for pure water [35].With the holes filled with pure water, this value should be smaller because of the mode penetration into the holes.To create the above Δn eq with such a small Γ SQW , a large Δn sem of approximately 0.1 is necessary in the SQW, but such a large Δn sem is unlikely to arise from any optoelectronic effects.Therefore, we focus on a phenomenon in the OCLs.We calculated the laser mode for the nanolaser structure described in Section 2 using a three-dimensional finite-difference time-domain method.In this calculation, we neglected the existence of the SQW because the OCLs occupy more than 97% of the semiconductor slab.We calculated the index sensitivity of the laser wavelength, Δλ/Δn sem , as 370 nm/RIU (RIU denotes the refractive index unit).For Δλ > 1 nm, Δn sem > 2.7 × 10 −3 is required on an average in the OCLs, which is likely to occur.
As a possible optoelectronic effect, we first considered the Franz-Keldysh effect [39] in the OCLs because the electric field in the U-shaped potential is mainly applied to these layers.Assuming a maximum ΔU of 0.8 V in Fig. 4(d) and a half thickness of the OCLs, i.e., 90 nm, the electric field induced is estimated to be as large as ~100 kV/cm.However, the bandgap wavelengths of the OCLs range from 1.1 to 1.2 μm, which are apart from the laser wavelength.Consequently, Δn sem induced by this effect is as small as 10 −5 .This effect cannot explain the experimental results.Next, we considered the carrier effects that electrons and holes produced by the photopumping that are trapped in the OCLs and n sem is modified.The maximum difference of E g in the OCLs is 0.1 eV.Since the valence band offset of the GaInAsP system is known to be 60% of E g [40], the maximum difference is only 0.06 eV for the holes, which is much smaller than the ΔU estimated above.This indicates that the holes produced in the OCLs by photopumping are easily pulled toward the front and rear surfaces of the slab by the U-shaped potential, and some electrons and holes are not relaxed into the SQW but instead are trapped in the OCLs.If such carriers exist over the OCLs with a reasonably high concentration, the carrier-plasma, band-filling, and band-shrinkage effects can produce a Δn sem of 10 −3 order.In this case, a large (or small) ΔU will produce a large (small) negative Δn sem and a blue (red) shift in the wavelength, which is consistent with the experimental results.The existence of such carrier trapping can be verified from the PL spectrum shown in Fig. 7(a).The nanolaser used in this measurement did not show the lasing but only showed the PL of the resonant mode due to its structural disordering.Here, we can observe a clear sub-peak at around 1.24 μm in addition to the main resonance peak around 1.6 μm.The sub-peak is considered to represent the PL from the OCLs (the sub-peak might extend further to shorter wavelengths since a long-pass filter with a cutoff at 1.15 μm was used to filter out the pump light).The high intensity of the sub-peak suggests that some fraction of the excited carriers was trapped in the OCLs without relaxing into the SQW.In addition, the intensity of the sub-peak decreased gradually after the plasma exposure, as shown by the open circles in Fig. 7(b).Such a trend has been reproduced in six devices, as shown in Fig. 7(c).The behavior differed slightly between the devices but the average intensity decreased within 30 min and converged thereafter.This behavior is consistent with those of Δλ, V Zeta , and ΔU in Fig. 4. In contrast, the intensity of the main peak rather increased with time, as shown by the closed circles in Fig. 7(b).This is attributed to the trapped carriers beginning to relax into the SQW with the decrease in the barrier height.
Let us estimate the carrier density in the OCLs and Δn sem due to the carrier effects.In Fig. 7(a), the PL intensity ratio of the sub-peak to main-peak is 0.35 at t = 0 min.The PL intensity is proportional to the carrier density N and layer thickness h, while inversely proportional to the carrier lifetime τ.Thus, we obtain the relation N OCL h OCL /τ OCL = 0.35N SQW h SQW /τ SQW .The carrier lifetime in the SQW is dominated by the cavity mode.It has been investigated theoretically and experimentally to be 7-10 times shorter than that of nonresonant modes due to the Purcell effect in PC nanolasers [41].Therefore, 1/τ SQW is enhanced by this factor, compared with 1/τ OCL .Using this relation with h OCL ≈180 nm and h SQW ≈4.5 nm, N OCL is estimated in the range of 0.062N-0.088NSQW .In the experiment in Fig. 7, we employed the same pumping condition as that for the laser operation in well-fabricated devices used for Figs.2-4.Even for the disordered non-lasing devices in Fig. 7, N SQW is comparable to or even higher than that in the lasing devices.Considering the typical carrier density in GaInAsP quantum wells under lasing [42], we set N SQW = 4 × 10 18 cm -3 , then obtain N OCL = 2.5 × 10 17 -3.5 × 10 17 cm −3 .Figure 7(c) further indicates that N OCL almost converges to 60% of the initial value at t = 0 min.From [43], which reports the index change due to the carrier-plasma, bandfilling, and bandgap-shrinkage effects for GaInAsP with E g = 1.03 eV, the N OCL is converted to Δn sem = 2.2 × 10 −3 -3.1 × 10 −3 .This well corresponds to the expected value of Δn sem for Δλ > 1 nm.For this discussion, one may suppose that N OCL , Δn sem , and Δλ are all increased by stronger pumping.However, the situation is not so simple.When N OCL is increased, more holes are pushed to the front and rear surfaces of the slab and separated from the electrons in space, which increases the internal electric field between them and cancels the U-shaped potential.This rather accelerates the relaxation of these carriers into the SQW, and compensates for the variation of Δn sem and Δλ.We observed that the change in the temporal behaviors of Δλ for the plasma exposure in Fig. 4(b) falls into the fluctuation level when the pump power is changed widely.
So far, we only discussed the U-shaped vertical potential in Fig. 6.We will next focus on the effect of the lateral Schottky barrier around the holes.Since the SQW is exposed at the sidewalls of the holes, the nonradiative surface recombination of excited carriers occurs and it is modulated by the height of the Schottky barrier.Thus, the laser emission intensity is modified by the adsorption of charged materials, as already reported in [33].However, as it is a phenomenon in the SQW with the small modal confinement, its effect on Δλ should be small in any case.In addition, the lateral Schottky barrier does not work for trapping carriers in the OCLs because it shifts the energy levels of the SQW and the OCLs simultaneously and does not change the relation between them.

Conclusion
The ultra-sensitive changes in the emission wavelength of the GaInAsP PC nanolaser biosensor for detecting biomolecules can be explained by the electrochemical and optoelectronic effects rather than by the change in the environmental index.A large wavelength shift Δλ of over 1 nm appeared when the device was electrified by the plasma exposure and charged DNA was adsorbed.The wavelength varied consistently with the zeta potential reflecting the surface charge and with the flat-band potential, indicating that the Schottky barrier height in the semiconductor was increased or decreased.The observed Δλ can be explained by the carrier plasma, band-filling, and band-shrinkage effects of this semiconductor due to the photopumped carriers being partly trapped in the OCLs inside the Schottky barrier.The PC nanolaser is sensitive to the environmental index, but much more sensitive to the change of the surface charge.We consider such a phenomenon that the adsorption of a small number of charged biomolecules becomes a trigger that modifies the charged functional group covering the device.These phenomena are based on a mechanism which is different from that of the enhanced surface recombination that we reported previously for pH sensing based on the emission intensity of the same device [33].
This conclusion may raise another question about whether the ultra-high sensitivity of Si PC cavity sensors [20,22], also exhibiting the same trend as that for the nanolaser in Fig. 1, originating from a similar mechanism.Although Si cavities are passive devices, they can acquire their ion sensitivity when carriers are generated via defect-state absorption and/or two-photon absorption.

Fig. 1 .
Fig. 1.Relation between the wavelength shift and concentration of the target biomolecules for various label-free photonic biosensors that are operated based on optical resonance.

Fig. 2 .
Fig. 2. Photonic-crystal nanolasers and the measurement of their laser operation.(a) Scanning electron micrograph of the nanolaser array and a magnified view of an H0-type cavity with a nanoslot.The diameter of the holes inside the dashed lines are reduced by 30 nm from those outside to induce asymmetry in the cavity structure and form a vertical output beam.(b) The measurement setup used in the experiment.(c) Measured lasing spectra for increasing levels of the biomolecule.

Fig. 3 .
Fig. 3. Schematic of the nanolaser surface and the detection of the biomolecules: (a) IgG, (b) PSA, and (c) single-strand DNA.Each plot and error bar show average values and the scattering range of the measurements for 10-20 nanolasers, respectively.

Fig. 4 .
Fig. 4. Response to the plasma exposure.(a) Plasma exposure treatment.(b) Temporal changes of the nanolaser wavelength following immersion in pure water immediately after the plasma exposure.(c) Corresponding zeta potential and (d) flat-band potential of an original epiwafer treated similarly.

Fig. 5 .
Fig. 5. Zeta and flat-band potentials measured after the physical adsorption of single-strand DNA.

Fig. 6 .
Fig. 6.Schematic of the relation between the surface charge, Schottky barrier, carrier distribution, and light emission.The left lower figure schematically depicts the band lineup in both the SQW and the OCLs.

Fig. 7 .
Fig. 7. Temporal behaviors of the main spectral peak of the cavity mode and the sub peak of the PL from OCLs of the PC nanolaser which did not exhibit laser operation due to structural disordering.The device was exposed to the plasma and then operated in pure water.(a) Two spectra measured at 5 and 60 min after the plasma exposure are presented.The sharp sub peak is partly caused by a long-pass filter with a cutoff of 1.15 μm, which was used to filter out the pump light.(b) Variation of the main and sub peak intensities normalized to those at time zero.(c) Variation of the sub peak intensity.Gray and black plots show those for eight different devices and their averages, respectively.