LSPR Review2

Localized Surface Plasmon Resonance Sensors:ReviewSi-yi Yang1, Xiao-xue Sun1, Cheng Wang1,Dong-hong Cai2, Li Zhang1, Guo-hua Liu1(1.College of Information Technology and Science, Nankai University, Tianjin 300071, China；2.Department of Physics, Kashi teachers college, Xinjiang 844000, China) Abstract: Recently, the research on the localized surface plasmon resonance (LSPR) sensors is one of the hot-spots in the filed of sensor technology. The LSPR sensors have obviously technology advantage in detecting and analyzing the properties of physics and chemistry, especially biology. This review introduces the technical principles and features of LSPR sensors and compares the current research of several LSPR sensor structures and production methods.Meanwhile, it discusses the future developments and commercial views of LSPR sensors.Key Words: Localized Surface Plasmon Resonance (LSPR); Biosensor; Nanoparticle 1.IntroductionIn recent years, scientists have high interests in the nanomaterials for the special characters in optics, electromagnetics and mechanics. Nanoparticles of the noble metals exhibit a strong UV-Vis absorption band that is not present in the spectrum of the bulk metal [1-8]. Scientists have reported that the characters of noble metal nanoparticle suspensions arise from their strong interaction with light and the advent of the field of nanoparticle optics has allowed for a deeper understanding of the relationship between material properties such as composition, size, shape, and local dielectric environment and the observed colour of a metal suspension. An understanding of the optical properties of noble metal nanoparticles holds both fundamental and practical significance. Fundamentally, it is important to systematically explore the nanoscale structural and local environmental factors that cause optical property variation, as well as provide access to regimes of predictablebehaviour. Practically, the tunable optical properties of nanostructures can be applied as materials for surface-enhanced spectroscopy [9-13], optical filters[14, 15], plasmonic devices [16-19]and sensors.It was realised that the sensor transduction mechanism of this LSPR-based nanosensor is analogous to that of SPR sensors. LSPR sensors can be seen as the continuation and expansion of SPR sensors. The former took place in the localized metal nanoparticles and the latter took place in the surface of metal films. However, the optical character of LSPR is different from that of SPR. LSPR has great potential in sensing field so that it has been widely studied [20-26]. It is known that nanoparticles, such as gold and silver, have strong absorption in the visible region, often considered as localized surface Plasmon absorption [27, 28].Such LSPR occurs when the incident photon frequencies match the collective oscillations of the conduction electrons of metal nanoparticles or metal islands. The particles in the nanoscale exhibit unique optical responses within the UV-Vis region [29, 30], where the absorbance shows an exponential decay with decreasing photon energy (the so-called Mie scattering), onto which an LSPR band, specific for the particle material, is superimposed. The surface plasmon energy and intensity have been found to be sensitive to a numberof factors, including particle conformation, immediate surrounding media, etc.[31-35] .The reactions between solution and probe on the molecular nanoparticles can cause change the thickness of the molecular layer, which led to the shift of LSPR absorption, so this method can be used in detecting dynamic real-time detection [36-37]. Take Ag nanoparticles fabricated using nanosphere lithography (NSL) as an example. Consistent wavelength shifts to the red with increasing density and thickness of adsorbate layers. The wavelength shift response is determined by the size and packing density of the molecules on the nanoparticle surface and the limitof detection of the system are determined by the surface-confined binding constant between the capture ligand on the surface and the target molecule in solution. As the systems reveal no nonspecific binding, the entire response can be attributed to the specific interactions between the molecules of choice. Optimization of the LSPR nanosensor has been realized by adjusting the size and shape of the nanoparticles[30, 38].For the nanomaterials have the same scale as biopolymers, proteins and nucleic acids in sizes, the researches and optimizations of LSPR-based sensing technology have been carried on in the biomedical field. The basic of applications in bioresearch, biosensing, cell labeling, fixed-point diagnosis, molecular dynamics research, and vector treatment is the interaction between biomoleculars and nanomaterials.LSPR nanosensors are useful in detecting biological molecules [22, 39, 40].2. The principle of LSPR sensorsLSPR is collective oscillations of the conduction electrons confined to metallic nanoparticles and metallic nanostructures [41-45]. It occurs in metallic nanostructures, such as nanoparticles, nanotriangles [46], and nanoisland [42]. LSPR is observed when the frequencies of incident photon match the collective oscillations of the conductive electrons of metal nanoparticles. Excitation of localized surface plasmons by an electric field at an incident wavelength where resonance occurs results in a strong light scattering, in the appearance in intense surface plasmon absorption bands, and the enhancement of the local electromagnetic field. The nanoparticles exhibit unique optical responses within the UV-Vis region [28], where the absorbance shows an exponential decay with decreasing photon energy (the so-called Mie scattering) onto which an LSPR band.The frequency and intensity of the surface plasmon absorption bands highly depend on the type of the material (gold, silver, platinum), the size and the shape of nanostructures as well as on their surrounding environment [29, 30, 34, 47].LSPR occurs when the frequency of the incident photons resonates with the collective oscillations of free electrons in the metals.The simplest model of nanoparticle optical responses is Mie theory. It describes extinction amounts of the long-wavelength spherical metal particles. The form is as follow [35]:E(λ) is the extinction which is equal to the sum of absorption and scattering, N A isthe areal density of nanoparticles, a is the radius of the metallic nanosphere, εm is the dielectric constant of the medium surrounding the metallic nanosphere (assumed to be a positive, real number and wavelength independent), λis the wavelength of the absorbing radiation, εi is the imaginary portion of the metallic nanosphere’s dielectric function, and εr is the real portion of the metallic nanosphere’s dielectric function. It is easy to see that the LSPR condition is met when (εr+2εm)2 approaches zero.In this primitive model, it is clear that the LSPR spectrum of an isolated metallic nanosphere embedded in an external dielectric medium will depend as follows: the radius a, the nanoparticle material (εi and εr), and the nanoenvironment’s dielectric constant (εm). Furthermore, when the nanoparticles are not spherical, as is always the case in real samples, the extinction spectrum will depend on the nanoparticle’s in-plane diameter, out-of-plane height, and shape. In this case the resonance term from the denominator of (1) is replaced with:Where χ, a shape factor term [11],is a term that describes the nanoparticle’s aspect ratio. Thus, the LSPR will also depend on interparticle spacing and the substrate dielectric constant. It is apparent from Eq. 1 that the location of the extinction maximum of noble metal nanoparticles is highly dependent on the dielectric properties of the surrounding environment, and that wavelength shifts in the extinction maximum of nanoparticles can be used to detect molecule-induced changes surrounding the nanoparticle.3. The configuration of LSPR sensing systemLSPR-based device can be set up without using the specific configurations, for example, the attenuated total reflection (ATR) optical setup or waveguide coupling, it is possible to fabricate very small devices based by NSL or other technologies. These properties have prompted an intense interest in LSPR-based sensors and biosensors.Generally, an LSPR optical sensor comprises an optical system, a transducingmedium which interrelates the optical and biochemical domains, and an electronic system supporting the optoelectronic components of the sensor and allowing data processing. The transducing medium transformes changes in the quantity of interest into changes in the refractive index which may be determined by optically interrogating the LSPR. The optical part of the LSPR sensor contains a source of optical radiation and an optical structure in which surface plasmon wave is excited and interrogated. In the process of interrogating the LSPR, an electronic signal is generated and processed by the electronic system. Major properties of an LSPR sensor are determined by properties of the sensor’s su bsystems. The sensor sensitivity, stability, and resolution depend upon properties of both the optical system and the transducing medium. The selectivity and response time of the sensor are primarily determined by the properties of the transducing medium. Take the fiber-based biosensor system as an example. The configuration of fiber-based biosensor system is shown in Fig.1. The system consists of a laser, a chopper, a fiber coupler, a sensing fiber, a liquid cell, a photo receiver, and a lock-in amplifier [48].Fig.1 The schematic representation of the LSPR biosensor system4. The fabrication and technology of LSPR sensorMetal nanostructures are the key to fabricate the LSPR biosensors. They mainly include nanowire array,nanoparticles, nanoislands and other structures, then we willintroduce the fabrication and technology of LSPR sensor with the three Metal Nanostructures in details.4.1 The fabrication of gold nano-wire array chipsA 1.25% solution of polymethyl methacrylate (PMMA) is spin coated on the silicon substrate to prepare the film with the thickness of 50nm. Roast the film under 150°C for half an hour to remove the aqueous vapor and set the arrangements of PMMA. Nanometer imprint is carried by using atomic force microscope (AFM) (Smena-HV, NT-MDT, Russia) and the diameter of the probe needlepoint is about 20nm (NSC15, MikroMasch, Russia). The main process is vapor plating the system with the electron gun after the nanogap is scratched in the PMMA film on the Si substrate. First, a Ti layer with thickness of 1nm is vapor plated, and then, a Au layer with thickness of20nm is plated. At last, a gold nano-wire array with the sub-100nm-linewidth can be got. The whole process is shown in the Fig.2.At room temperature，the produced gold nano-wire array chips are immersed into an ethanol solution of 10-3M octadecanethiol (ODT), and after a period of specific time, take out the chips and rinse the chips with an ethanol solution immediately, and then, the chips are dried with N2. At last, the chips are roasted on the heating plate for about 10min in the air, in order to removal all the moisture stayed on the chips, and then, ODT-modified nano-wire arrays can be prepared [49].Fig.2 Schematic representation of producing the gold nano-wire array4.2 The fabrication of Ag nanoparticles based on NSL technologyNanosphere lithography (NSL) is one of nanometer machining and producing technology, and we can use it to control the shape, size and structure of nanoparticles’surface. In the experimental study, we can compose Ag nanometer triangle particles with the width of 100nm and the height of 500nm based on NSL technology, and we can measure the optical properties of particles by using LSPR spectroscopic measurement.For preparing LSPR biological nanometer sensor, we need make Ag nanometer triangle particles have adsorption function by self-assembled monolayer, then biotin will be covalently linked to carboxy by zero-length coupled reagent.The whole process is shown in the Fig.3. The fabrication steps of Ag nanoparticles are as follows: (1) clean the substrate; (2) drip polystyrene nanoparticles on the substrate and cover it; (3) dry the monolayer film filled closely of nanometer sphere and form its mask; (4) Ag evaporates and precipitates on samples; (5) remove the mask of nanometer sphere in ethanol solution by ultrasonic degradation; (6) finish Ag nanometer particles samples [50].Fig. 3 The processing operation of Ag nanoparticles based on NSL technology4.3 Fabrication of gold NI well chips and detection of recombinant protein using the gold NI well chipsA 200μm-thick layer of SU-8 100 (Microchem, USA) was formed on the gold NI chip by spin-coating (500 rpm for 5 s, followed by 1500 rpm for 30 s) and soft-baked on a hot plate at 95 ◦C for 100 min. An 8×8 array of Ø800μm-wells was patterned by UV exposure at a wavelength of 365 nm and a dose of 630 mJ/cm2. After exposure,the post-exposure baking was performed at 95 ◦C for 30 min, and non-cross-linked SU-8 resins in the regions of wells were then removed by dipping in an SU-8 developer for 3 min with weak sonication, leading to exposure of the gold NI surfaces in a portion of the wells.Functionalization of the gold NI surfaces with GSH was carried out using the same procedure as described previously. 0.2μL of the cell lysates containing various amounts of GST-hIL6 were dropped into the wells on the gold NI chips, and the well chips were then incubated for 3 h at 25 ◦C in a sealed petridish to prevent the evaporation of sample solutions. The incubated chips were extensively rinsed with PBS and DI water, followed by drying with N2 [51].4.4 Surface Processing technology of NanoparticlesIn Figure 4, the nanoimprint patterning process is schematically illustrated. First, Au layer of 40 nm in thickness was deposited on a glass substrate using an e-beam evaporator. A thin (2 nm) chromium under-layer was pre-deposited to enhance adhesion of the Au layer to the substrate. Then, polystyrene (PS) solutions at various concentrations (1.5, 0.8, and 0.5 wt %, M w of PS: 45 730) were spin-coated onto the thin Au films. The polymeric film was annealed at a temperature above the glass transition temperature T g and imprinted with a pre-designed composite PDMS mold, which was fabricated by following a standard procedure. This imprint process took 30 min to 1 h under a low pressure of 1.60 kPa in a vacuum at 150 °C. In scheme A of Figure 4 for high PS concentrations, the PS film was thick enough to fill the cylindrical holes between the PDMS mold and the substrate. Meanwhile, at low PS concentrations below 0.8 wt %, the PS film was thin and partially filled the gap as illustrated in scheme B of Figure 4. In this case, the PS film climbed the PDMS walls under the action of capillary forces and formed the circular undulations around the walls. The PS patterns were treated with reactive ion etching (RIE) of CF4/O2 to remove the residual PS layers on the substrate. The RIE modified slightly the feature sizes of the PS patterns. After being cooled to room temperature, the PS patterns were used as mask for Ar ion milling for 2 min, which left behind the patterned Auarrays. The DC bias for the ion milling was 400 V, and the Ar pressure was kept below 5*10-4 Torr. In particular, the thin circular PS dots with undulated edges produced eventually the patterned Au rings as shown in scheme B of Figure 4. Finally, the residual PS mask was removed by sonicating the substrate in toluene for 15 min.To modify the surfaces of Au islands, we used self-assembled monolayer (SAM) of aminoundecanethiol (AUT) as shown in scheme C of Figure 4. To ensure a well organized SAM of AUT onto the Au islands, the sample was incubated in the AUT solution for 24h. After careful rinsing and thorough drying with pure N2 gas, the patterned Au array was activated by incubating in 10 mM of bis-[sulfosuccinimidyl] suberate (BS3) in phosphate buffered saline (PBS) solution. Subsequently, the substrate was incubated overnight in 10wt % methanolic solution of G4 dendrimers of amine terminated poly (amidoamine) (henceforth, PAMAM). Then, the anchored PAMAM was biotinylated by attaching sulfo-NHS-LC-biotin covalently via amid bond formation. Finally, the biotinylated Au dot and ring arrays were exposed to 20 μM streptavidin solution to induce the binding of streptavidin to the biotinylated Au surface. LSPR was measured on UV-visible near-IR spectroscopy (JASCO: V-570) of the double-beam system with a monochromator. The scanning wavelength ranged from 300 to 2500 nm. The light source was an unpolarized halogen lamp (200~2500 nm). AFM (Seiko Instruments: SPA400) was used to obtain topographic images of the nanopattern arrays. The images were taken with contact mode under ambient conditions [52].Fig. 4 Schematic of nanopatterning of Au islands fabricated by imprint lithography and ion milling.(A and B) Dot and ring pattern formation using thick and thin PS films, respectively. (C) Procedure for immobilizing SA protein on the Au patterns using the SAM of amino-terminated PAMAMdendrimers as an intermediate coupling layer. Sulfo-NHS-LC-biotin was used to bind the SA protein.5. Application of LSPR sensorsPhysical phenomena occurring in various optical transducing materials have beenalso exploited for the development of LSPR-sensing devices including a humiditysensor utilizing humidity-induced refractive index changes of porous thin layers andpolymers and a temperature sensor based on the thermooptic effect inhydrogenated amorphous silicon.While in specific systems of limited complexity variations in the concentration ofanalyte may be determined by directly measuring refractive index using an LSPRsensor (e.g. monitoring distillation processes ), most chemical LSPR sensors arebased on the measurement of LSPR variations due to adsorption or a chemicalreaction of an analyte with a transducing medium which results in changes in itsoptical properties. It can be applied to detect the reaction of molecule by the changeof refractive index of conversion layer.LSPR sensors have also been widely used in biological field. The detection ofbiospecific interaction was developed by also some other groups. In LSPR biosensors,the analyte quantification was carried out by direct detection of the binding reaction.Earlier works were focusing mainly on antigen-antibody interactions, thestreptatividin–biotin reaction, and some IgG examinations, especially to test newalgorithms in biospecific molecular interaction analysis, to characterize newlydeveloped LSPR set-ups, and to improve surface chemistry. Current researchincludes far more advanced systems. One of the new areas is the examination ofprotein–protein or protein–DNA interactions. Take multiple label-free detection ofprotein in nanometer chip as an example. After the immobilization of the antibodieson the multiarray chip, different concentrations of antigen solutions (~0 to 100μg/mL) were introduced onto the 300 spots using the nanoliter dispensing systemand incubated for 30 min (Figure 5). Especially, the total sample volume wassignificantly reduced. Each spot was reacted with a sample solution of 100 nLcontaining the antigens at varying concentrations. After an incubation period of 30min, a stringent washing procedure with 1% (v/v) Tween-20 was applied to suppress the nonspecific adsorption. Subsequently, evaluation of the optical characteristics of the chip was carried out. All absorbance spectra were taken from a range of 400-800 nm on the UV-vis spectrometer at RT. White light emerging from the optical fiber bundle was incident onto the nanochip from the vertical direction. The reflected light was coupled into the detection fiber probe of the optical fiber bundle and analyzed by the UV-vis spectrometer. The evaluation of the results obtained from the nanochip surface revealed that the specific absorbance strength change was directly related to the applied antigen concentration on the spot [53].Fig. 5 Experimental setup of the multiarray LSPR-based nanochip6. Commercialization of LSPR biosensor technologyAs we know from above that the LSPR biosensor has many advantages. It can be easy to operate and can detect very small quantities of samples (such as the protein solution).After research of LSPR sensors we found that such sensors have great commercial potential. Take the gold nanorods LSPR sensor as an example. It can be produced without photolithography, so that the cost of the clinical diagnosis can be significantly reduced. And its measurement method is based on the spectra shift. If the quantity of non-specific binding is very small, the detection limit of LSPR sensors can reach pM concentrations.Medical reports show that milk allergy has been a great threat to human health.And after experiment we find that the gold nanoparticles LSPR sensors can detect casein allergens in the milk samples. So we can see that the LSPR biosensors will be widely used in the allergic diagnose area in the future, and also we can use LSPR sensors to develop highly integrated food safety control system.The LSPR sensor system has a high sensitivity and a very small detection limit, its fabrication is very simple, and it has a high integration and good manipulability. Based on the above characteristics, we can see that LSPR sensors will have a very broad commercial prospect in the medical diagnosis and real-time detection area. 7. Future trends in development of LSPR biosensorsThe nanomaterial has many special features including its small size effect surface effect and quantum size effect and so on, so that the LSPR sensors can have unique optical and electrical properties. Meanwhile gold nanoparticles have good biological compatibility. It has good hybridization with DNA molecule. The most important is that the cost of LSPR sensor is very low, so in the future we can produce these sensors in mass production. Based on these characters the LSPR sensor will be widely used in the molecular biology, clinical medicine and bio-chip. Using LSPR sensor, the DNA computer will be invented possibly. Therefore, LSPR biosensors will play an important role in life science and analytical chemistry research.Future trends in diagnostics will continue in miniaturization of biochip technology toward nanoscale. Multiarray LSPR-based nanochip provides a convenient, low-cost, and label-free method for specific and highly sensitive detection of the biomolecular interactions in a parallel format. We anticipate that this technology will extend the limits of current molecular diagnostics and enable point-of-care diagnosis as well as promote the development of personalized medicine. The multiparallel possibilities of biosensing applications have the potential to allow the optimization of biomarker research, cancer diagnosis, and also the detection of infectious microorganisms for biodefense. Especially, the cost for the fabrication of multiarray LSPR-based nanochip, including the optical apparatus, is significantly lower than that of a conventional SPR system. This LSPR method is “easy to operate” even by“nonspecialists”. The LSPR-based multiarray nanochip presents a highly versatile method that is readily applicable to the other kinds of bioassays, such as metabolomics and cellomics.8. ConclusionsIn the recent years, because of the special properties of LSPR technology, there are more and more people paying their attentions to this area. This paper reviews the recent development of LSPR technology and discusses the fabrication and application of this technology. Through the detection of small amount molecule or high-sensitive project in the physical, chemical and biologic area, we can have a conclusion that the LSPR biosensor will have great potential in the future. The LSPR biosensor technology will greatly promote the development of monitoring and diagnostics. The LSPR biosensor system has many advantages, such as simple structure, convenient carrying, low cost, high sensitivity. Therefore we can see that LSPR technology will have broad commercial prospect. It is necessary to complete the research of this technology.References[1] Haynes C. L., and Van Duyne R. P., J. Phys. Chem. B, 2001, 105, 5599~5611[2] Mulvaney P., MRS Bull., 2001, 26, 1009~1014[3] El-Sayed M. A., Accounts Chem. Res., 2001, 34, 257~264[4] Link S., and El-Sayed M. A., J. Phys. Chem. B, 1999, 103, 8410~8426[5] Kreibig U., Gartz M., Hilger A., et al., in Duncan, M.A., JAI Press Inc., Stamford, , 1998, Vol.4, 345~393[6] Mulvaney P., Langmuir, 1996, 12, 788~800[7] Kreibig U., in Hummel, R. E. and Wissmann, P. (Eds.) CRC Press, Boca Raton FL, 1997, Vol. II,145~190[8] Hulteen J. C., Treichel D. A., Smith M. T., et al., J. Phys. Chem. B, 1999, 103, 3854~3863[9] Freeman R. G., Grabar K. C., Allison K. J., et al., Science, 1995, 267, 1629~1632[10] Kahl M., Voges E., Kostrewa S., et al., Sens. Actuators B, Chem., 1998, 51, 285~291[11] Schatz G. C., and Van Duyne R. P., Wiley, New York, , 2002, Vol. 1[12] Haynes C. L., and Van Duyne R. P., J. Phys. Chem. B, 2003, 107, 7426~7433[13] Haynes C. L.,McFarland A. D.,Zhao L., et al., J. Phys. Chem. B, 2003, 107, 7337~7342[14] Dirix Y., Bastiaansen C., Caseri W., et al., Adv. Mater., 1999, 11, 223~227[15] Haynes C. L., and Van Duyne R. P., Nano Lett., 2003, 3, 939~943[16] Maier S. A., Brongersma M. L., Kik P. G., et al., Adv. Mater., 2001, 13, 1501~1505[17] Maier S. A., Kik P. G., Atwater H. A., et al., Nature Mater., 2003, 2, 229~232[18] Shelby R. A., Smith D. R., and Schultz S., Science, 2001, 292, 77~78[19] Andersen P. C., and Rowlen K. L., Appl. Spectrosc.7, 2002, 56, 124A~135A[20] Love J. C., Estroff L. A., Kriebel J. K., et al. Chem. Rev. 2005, 105, 1103~1169[21] Hutter E., Fendler J. H. Chem. Commun. 2002, 378~379[22] Haes, A. J., Van Duyne, R. P. J. Am. Chem. Soc. 2002, 124, 10596~10604[23] Frederix F., Friedt J. M., Choi, K. H., et al. Anal. Chem. 2003, 75, 6894~6900[24] Nath N., Chilkoti A. J. Fluoresc. 2004, 14, 377~389[25] Stuart D. A., Yonzon C. R., Zhang X., et al. Anal. Chem. 2005, 77, 4013~4019[26] Sönnichsen C., Reinhard B. M., Liphardt J., et al. Nat. Biotechnol. 2005, 23, 741~745[27] Jin R., Cao Y., Mirkin C. A., et al. Science 2001, 294, 1901~1903[38] Prodan E., Nordlander P., Halas N. J. Nano Lett. 2003, 3, 1411~1415[29] Haes A. J., Zou S., Schatz G. C., et al. J. Phys. Chem. B 2004, 108, 6961~6968[30] Haes A. J.; Zou S., Schatz G. C.; et al. J. Phys. Chem. B 2004, 108, 109~116[31] Yonzon C. R., Jeoung E., Zou S., et al. J. Am. Chem. Soc. 2004, 126, 12669~12676[32] Hong X., Kao F. J. Appl. Opt. 2004, 43, 2868~2873[33] Nath N., Chilkoti A. Anal.Chem 2004, 76, 5370~5378[34] Underwood S., Mulvaney P. Langmuir 1994, 10, 3427~3430[35] Stuart D. A., Haes A. J., Yonzon C. R., et a1. IEE Proc. Nanobiotechno1. ,2005, 152(1), 13~32[36] Haes A. J., Chang L., Klein W. L., et al. J. Am. Chem. Soc. 2005, 127, 2264~2271[37] Himmelhaus M., Takei H. Sens. Actuators, B 2000, 63, 24~30[38] Haes A. J., Zou S., Schatz G. C., et al. J. Phys. Chem. B 2004, in press.[39] Riboh J. C., Haes A. J., McFarland A. D., et al. (2003) J Phys Chem B 107, 1772~1780[40] Van Duyne R. P., Haes A. J., McFarland A. D. (2003) SPIE 5223, 197~207[41] Kreibig U., Vollmer M., Optical Properties of Metal Clusters, Springer-Verlag, Berlin,Heidelberg, 1995.[42] Hutter E., Fendler J. H., Adv. Mater. 16 (2004) 1685.[43] Riu J., Maroto A., Rius F. X., Talanta 69 (2006) 288.[44] Kreibig U., Genzel L., Surf. Sci. 156 (1985) 678.[45] Wagner F. E., Halsberck S., Stievano L., et al. Nature. 407 (2000) 691.[46] Malinsky M. D, Kelly K. L, Schatz G. C, et al. (2001) J Am Chem Soc 123, 1471~1482[47] Yonzon C. R., Jeoung E., Zou S., et al. J. Am. Chem. Soc. 126 (2004) 12669.[48] Tsao-Jen Lin, Kuang-Tse Huang, Chia-Yu Liu, Biosensors and Bioelectronics 22 (2006)513–518[49] 吴宜蓁廖健順林鶴南[50] 金友,现代应用光学,2007年1月[51] Yong-Beom Shin, Jeong-Min Lee, Mi-Ra Park, et al, Biosensors and Bioelectronics 22 (2007)2301–2307[52] Sarah Kim, et al, Langmuir 2006, 22, 7109-7112[53] Tatsuro Endo, et al, Anal. Chem. 2006, 78, 6465-6475。