sensors Article

SERS Taper-Fiber Nanoprobe Modified by Gold Nanoparticles Wrapped with Ultrathin Alumina Film by Atomic Layer Deposition Wenjie Xu, Zhenyi Chen, Na Chen *, Heng Zhang, Shupeng Liu, Xinmao Hu, Jianxiang Wen and Tingyun Wang Key Laboratory of Specialty Fiber Optics and Optical Access Networks, School of Communication and Information Engineering, Shanghai University, 333 Nanchen Road, Shanghai 200444, China; [email protected] (W.X.); [email protected] (Z.C.); [email protected] (H.Z.); [email protected] (S.L.); [email protected] (X.H.); [email protected] (J.W.); [email protected] (T.W.) * Correspondence: [email protected] Academic Editor: W. Rudolf Seitz Received: 6 January 2017; Accepted: 15 February 2017; Published: 25 February 2017

Abstract: A taper-fiber SERS nanoprobe modified by gold nanoparticles (Au-NPs) with ultrathin alumina layers was fabricated and its ability to perform remote Raman detection was demonstrated. The taper-fiber nanoprobe (TFNP) with a nanoscale tip size under 80 nm was made by heated pulling combined with the chemical etching method. The Au-NPs were deposited on the TFNP surface with the electrostatic self-assembly technology, and then the TFNP was wrapped with ultrathin alumina layers by the atomic layer deposition (ALD) technique. The results told us that with the increasing thickness of the alumina film, the Raman signals decreased. With approximately 1 nm alumina film, the remote detection limit for R6G aqueous solution reached 10−6 mol/L. Keywords: taper-fiber nanoprobe; gold nanoparticles; atomic layer deposition; surface enhancement Raman spectroscopy

1. Introduction Serving as an attractive analytical spectral technique, Raman spectroscopy provides a “fingerprint” of information of molecular structures [1,2]. Unfortunately, due to its relatively low Raman scattering cross-section (~10−30 cm2 per molecule) [3] and being hidden in fluorescence spectroscopy [4], the original Raman scattering is extremely weak and its potential applications were limited, until surface enhancement Raman spectroscopy (SERS), which paved the way for the extension of Raman spectroscopy to the realm of trace detection [5], was discovered in 1974 [6]. SERS can strongly increase the Raman signals of a molecule attached to nanoscale noble metallic structures, such as gold, silver and copper, etc. [7–9]. Among the surfaces of these noble metals, the localized surface plasma and the Raman signals will appear as resonance, and then the Raman signals will be extremely enhanced. In addition, the morphology of these noble metal nanostructures will significantly influence the SERS effect, leading to various morphological substrates investigated and novel nanostructures proposed [10–12]. A single-molecule detection was demonstrated by the SERS technique in different environments [13], even in single living cells [14]. Optical fibers, as sensors, have been used for SERS to detect molecules in remote locations and biomedical applications [15]. This group of optical fiber sensors, whose remote capabilities are often called “optrode” [5], can be classified into two categories by the Raman signals collected either in backward scattering through the same fiber or in forward scattering by the other separated fiber. There are types of optical fiber structures serving as SERS detectors: the liquid-core photonic crystal

Sensors 2017, 17, 467; doi:10.3390/s17030467

www.mdpi.com/journal/sensors

Sensors 2017, 17, 467 Sensors 2017, 17, 467

2 of 9 2 of 9

fiber probe (LCPCF) prepared to detect the 10−10 mol/L R6G solution mixed with nanoparticles (NPs) −10 remote [16], the nanopillar array on a fiber facet for sensing of toluene vapor [17], fiber fiber probe (LCPCF) prepared to detect thein10situ mol/L SERS R6G solution mixed with nanoparticles tips with polishingarray [18], on flatafiber ablated by aremote femtosecond laser [19], and the vapor taper-fiber (NPs) [16],angled the nanopillar fibertips facet for in situ SERS sensing of toluene [17], −7 probetips withwith silver nanoplates fabricated to detect the 10 mol/L [20]. The taper-fiber fiber angled polishing [18], flat fiber tips ablated by4-ATP a femtosecond laser [19],tips andwith the 7 mol/L an optimalprobe cone angle can achieve superior performance fiber SERS detection, due to taper-fiber with silver nanoplates fabricated to detectfor thean 10−optical 4-ATP [20]. The taper-fiber increasing SERS active surface transmitting evanescentfor field the surroundings with an tips with anthe optimal cone angle can[21], achieve superiorthe performance an to optical fiber SERS detection, increased magnitude good collection [23]. Nevertheless, the SERS may due to increasing the [22], SERSand active surface [21], efficiency transmitting the evanescent field to thesubstrates surroundings be easily damaged by exposure air good or samples, andefficiency will not remain stable for a the long time. with an increased magnitude [22],toand collection [23]. Nevertheless, SERS substrates proposal of shell-isolated Raman spectroscopy [24]a was may The be easily damaged by exposurenanoparticle-enhanced to air or samples, and will not remain stable for long generally time. accepted to be a milestone in Ramannanoparticle-enhanced spectroscopy. Noble metal NPs coated with [24] metal oxide layers The proposal of shell-isolated Raman spectroscopy was generally were reported be able to in suppress surface reactions withNPs air, and preserve their morphology accepted to be to a milestone Ramanmetal spectroscopy. Noble metal coated with metal oxide layers at high temperatures fortoan extended period of time without their enhancement capabilities were reported to be able suppress metal surface reactions withlosing air, and preserve their morphology at [25–28]. These coated structures be applied complex environment detectioncapabilities and survive with high temperatures for an extendedcan period of time to without losing their enhancement [25–28]. long-term stability [29,30]. indicated, in some literatures,detection that an and Ag island film coated with These coated structures can It bewas applied to complex environment survive with long-term protected[29,30]. alumina to establish a reusable andthat stable substrate and Ag protected nanorods stability It was able indicated, in some literatures, an SERS Ag island film [31], coated with wrappedwas with ultrathin alumina layersand were ableSERS to stabilize sensitivity air for more alumina able to establish a reusable stable substratetheir [31],SERS and Ag nanorodsinwrapped with than 50 days [30]. Besides, nanostructures with ultrathin (~1.5 nm) layers will not decrease the SERS ultrathin alumina layers were able to stabilize their SERS sensitivity in air for more than 50 days [30]. performance much. Hence, it is highly towill wrap SERSthe nanoprobe with ultrathin Besides, nanostructures with ultrathin (~1.5desirable nm) layers not our decrease SERS performance much. aluminaitlayers. Hence, is highly desirable to wrap our SERS nanoprobe with ultrathin alumina layers. thisliterature, literature,based based previous work we prepared a taper-fiber nanoprobe In this onon ourour previous work [32], [32], we prepared a taper-fiber nanoprobe (TFNP) (TFNP) modified by Au-NPs, and wrapped it with ultrathin film. The nanoscale taper-fiber modified by Au-NPs, and wrapped it with ultrathin aluminaalumina film. The nanoscale taper-fiber probe probe tip was made through the processes of heated pulling combined with chemical etching. tip was made through the processes of heated pulling combined with chemical etching. Electrostatic Electrostatic self-assembly adopted to on deposit tip and surface the self-assembly technology wastechnology adopted to was deposit Au-NPs the tip Au-NPs surface ofon thethe TFNP, thenof it was TFNP, and then it was wrapped ultrathin alumina film “layer via theby ALD technique by layer”. wrapped with ultrathin alumina with film via the ALD technique layer”. Direct “layer measurement to Direct measurement to R6G adsorbed on Au-NPs with different alumina layers on this probe surface R6G adsorbed on Au-NPs with different alumina layers on this probe surface was made. Raman spectra was made. Raman spectrawith of R6G aqueous solutions with were detected in of R6G aqueous solutions different concentrations weredifferent detectedconcentrations in the remote detection mode by the remote detection mode by this fiber SERS nanoprobe. this fiber SERS nanoprobe. 2. Experiment Procedures and Methods 2.1. Fabrication of of the the Nanoscale Nanoscale Taper-Fiber 2.1. Fabrication Taper-Fiber A diameter of of 50/125 50/125µµm was used used in in our our experiment. experiment. A multimode multimode fiber fiber with with aa core/cladding core/cladding diameter m was The nanoscale taper-fiber was obtained through the heated pulling method combined with The nanoscale taper-fiber was obtained through the heated pulling method combined with chemical chemical etching The specific specific process process is is the etching [33–35]. [33–35]. The the same same as as our our previous previous works works [32]. [32]. After After etching, etching, the the fiber fiber was rinsed with deionized water. The final result of this taper-fiber tip is shown in Figure 1. was rinsed with deionized water. The final result of this taper-fiber tip is shown in Figure 1.

Figure 1. Optical microscope photographs of taper-fiber nanoprobe with 10× objective. Figure 1. Optical microscope photographs of taper-fiber nanoprobe with 10× objective.

Sensors 2017, 17, 467 Sensors 2017, 17, 467

3 of 9 3 of 9

taper-fiberhas hasa avery very sharp its angle is about 30°. It is illustrated that thelength cone The taper-fiber sharp tiptip andand its angle is about 30◦ . It is illustrated that the cone length tip is measured at around 330Generally, μm. Generally, taper-fiber of this of tipthis is measured at around 330 µm. taper-fiber probeprobe with with largerlarger tapertaper angleangle and and shorter cone length more to efficient collect light. scattering light. inHowever, in the shorter taper taper cone length is more isefficient collect to scattering However, the applications applications of puncture measurements, angle required, so with considerations, the integrative of puncture measurements, smaller taper smaller angle is taper required, soiswith the integrative considerations, the taper should a trade-off the taper angle should beangle a trade-off inbe some cases. in some cases. 2.2. Preparation of of Gold Gold NPs 2.2. Preparation NPs Analogy our previous previous works works [32], [32], Gold Gold colloids colloids were were still still prepared prepared with the classic classic Frenz Frenz [36] [36] Analogy to to our with the method, for its itsbeing beingrapid rapidand anduniform. uniform. The preparation procedures were repeated as follows: method, for The preparation procedures were repeated as follows: 100 100 0.01% chloroauric acid (HAuCl prepared beaker,and andthen thenheated heatedto to boiling; boiling; 4 ) solution mL mL 0.01% (w)(w) chloroauric acid (HAuCl 4) solution prepared inin a abeaker, 0.5 O77··2H 2H22O) O)solution, solution,which whichmakes makes Au-NPs Au-NPs 55 55 nm nm in in average average 0.5 mL mL 1% 1% (w) (w) trisodium trisodium citrate citrate(Na (Na33C C66H H5O diameter, added to to the the boiling boiling HAuCl HAuCl44solution solutionand andkept keptthe themixture mixturesolution solutionboiling boilingstate statefor for diameter, was was added 5 5 min. That Au-NPs with diameters between 50 nm and 60 nm have the optimal enhancement effect min. That Au-NPs with diameters between 50 nm and 60 nm have the optimal enhancement effect for for the the R6G’s R6G’s Raman Raman spectra spectra detection detection has has been been demonstrated demonstrated [37]. [37]. 2.3. Modified Tapered Tapered Fiber 2.3. Modified Fiber Nanoprobe Nanoprobe with with Gold Gold NPs NPs Although it is is time-consuming, time-consuming, electrostatic electrostatic self-assembly self-assembly technology, technology, due Au-NPs Although it due to to it it making making Au-NPs uniform on the taper-fiber tips, is still adopted to form the SERS-active substrate, as our previous uniform on the taper-fiber tips, is still adopted to form the SERS-active substrate, as our previous works works [32]. [32]. Firstly, the taper-fiber taper-fiber tip tipshould shouldbebetotally totally cleaned immersing a piranha solution Firstly, the cleaned byby immersing intointo a piranha solution (3:1 (3:1 mixture of 96% concentrated sulfuric acid and 30% hydrogen peroxide) for 30 min, and then mixture of 96% concentrated sulfuric acid and 30% hydrogen peroxide) for 30 min, and then rinsing rinsing in deionized water and ethanol respectively. Secondly, the cleaned fiber was soaked twice intwice deionized water and ethanol respectively. Secondly, the cleaned fiber was soaked in the in the mixture solution with 5% v/v deionized water, 5% v/v (3-Aminopropyl) trimethoxysilane mixture solution with 5% v/v deionized water, 5% v/v (3-Aminopropyl) trimethoxysilane (APTMS (APTMS 90% ethanol 30 Again, min. Again, it was rinsed twice deionizedwater waterand and ethanol ethanol 97%) and97%) 90%and ethanol for 30 for min. it was rinsed twice in in deionized ◦ C for 30 min. respectively to remove the residual APTMS. Thirdly, it was kept in the incubator at 90 respectively to remove the residual APTMS. Thirdly, it was kept in the incubator at 90 °C for 30 min. Finally, Finally, this this fiber fiber was was cleaned cleaned in in the the ethanol ethanol and and then thenimmersed immersedinto intothe thegold goldcolloids colloidsfor for48 48h. h. The photographs of of our our TFNP TFNP modified modified by by Au-NPs The morphology morphology photographs Au-NPs were were obtained obtained by by scanning scanning electron As expected, the Au-NPs on the electron microscopy microscopy (SEM) (SEM) (Figure (Figure 2). 2). As expected, the Au-NPs were were uniformly uniformly distributed distributed on the surface of TFNP, and the nanoprobe has a tip dimension under 80 nm (Figure 3b) which have great surface of TFNP, and the nanoprobe has a tip dimension under 80 nm (Figure 3b) which have great potential potential in in puncture puncture measurements. measurements.

Figure 2. SEM photographs of the distribution of Au-NPs on the TFNP surface and nanoprobe tip Figure 2. SEM photographs of the distribution of Au-NPs on the TFNP surface and nanoprobe tip inset. inset.

2.4. Wrap the Modified Nanoprobe with Alumina Layers The taper-fiber nanoprobes prepared above, in puncture measurement, were used to detect the intracellular Raman signals but the corresponding biologic Raman signals were very weak and even sometimes no biologic Raman signal was detected. Therefore, we assumed that due to the friction

Sensors 2017, 17, 467 Sensors 2017, 17, 467

4 of 9 5 of 9

Figure 3. (a) Profile illustration of Au-NPs modified TFNP with ultrathin alumina layers; (b) SEM Figure 3. (a) Profile illustration of Au-NPs modified TFNP with ultrathin alumina layers; photograph of Au-NPs modified TFNP with ultrathin alumina layers (with 10 cycles); (c) SEM (b) SEM photograph of Au-NPs modified TFNP with ultrathin alumina layers (with 10 cycles); photograph of fiber on which deposited alumina film by ALD technique with 5000 cycles (first 2000 (c) SEM photograph of fiber on which deposited alumina film by ALD technique with 5000 cycles cycles and then 3000 cycles). (first 2000 cycles and then 3000 cycles).

2.5. SERS Measurements 2.4. Wrap the Modified Nanoprobe with Alumina Layers Rhodamine 6G (R6G) aqueous solution was used as the standard sample, to ascertain the SERS The taper-fiber nanoprobes prepared above, in puncture measurement, were used to detect the performances of the TFNP with ultrathin alumina layers. The Raman spectra of R6G were detected intracellular Raman signals but the corresponding biologic Raman signals were very weak and even by the above prepared fiber SERS nanoprobe with a commercial confocal micro-Raman spectrometer sometimes no biologic Raman signal was detected. Therefore, we assumed that due to the friction and (Renishaw inVia plus) by direct measurement and in optrode configuration. The experimental collision between nanoparticles and cell membrane, the dropping and surface morphology changing setups illustrate in Figure 4. of the AuNPs might cause the signals weak and even disappearing. In order to gain better potential for puncture biosensing such as intracellular, we wrap the modified nanoprobe Raman spectrometerwith alumina layers, Raman spectrometer which can sacrifice itself to protecting the Raman enhancement substrates from being damaged or dropping in the process of piercing the cell or other analytes.Objective The alumina nano-layers were deposited Objective on the gold modified TFNP with the ALD equipment (TFS 200, Beneq, Espoo, Finland). There is a disc shape substrate with diameter 22 cm and the height of 3 cm in its reaction chamber. Excitation light Raman scatteringUsing signal high Excitation lightas the carrier Raman scattering signal purity nitrogen and purge gas, two precursors, trimethyl aluminum (TMA) and water Sample vapors, were valve controlled and carried into the heated chamber (about 210 ◦ C) by the way of pulse alternatively. Quintessentially, one reaction cycle consisted of four steps: (1) TMA reactant adsorbed Sample Optical fiber tip Optical fiber on the surface of sample; (2) nitrogen purging; (3) water vapor exposure in the chamber and reaction (a) (b) with TMA; (4) nitrogen purging. One reaction cycle completed, the monolayer Al2 O3 was deposited on the sample-surface. Through repeating the deposition cycle, Al2 O3 film grows thicker layer by layerFigure and its4.thickness can setup be precisely controlled by altering thedetection numberinofoptrode cycles.configuration. Experimental for (a) direct measurement and (b) As to our ALD equipment, one technique corresponds to one thickness of the film in a deposition cycle.ByIndirect ordermeasurement, to confirm thethe thickness of the film per in ourstage alumina deposition, we have tip of TFNP was put oncycle an object (Figure 4a), adjusting the deposited 300 ALD cycles on the clean silicon wafer, and the total thickness of the arbitrary six points objective to optimum spot for measurement. Using 633 nm He-Ne laser as exciting light, whose of the is coated film was ahead measured by ellipsometer, and the results showntipinsurface Table 1. It was power around 2 mW objective, the SERS enhancement effectare of TFNP including demonstrated that the thickness was uniform, the refractive index of alumina film was 1.658, and the Au-NPs wrapped with different alumina layers was investigated. The microscope objective used 50× deposition rate is approximate 0.12 nm/cycle. We also have deposited 5000 ALD cycles on the clean for high efficient collecting SERS signal from the tip surface of TFNP. R6G aqueous solution with a −6 mol/L surface of fiberof by10 two steps (first 2000 cycles then 3000 cycles), and its SEM photograph is shown in concentration was used as the and detection sample. Figure 3c. The SEM photograph told us that an average deposition rate iscut approximate 0.1112 In optrode configuration, the fiber SERS nanoprobe was usually into segments of nm/cycle. 15–20 cm Considering above data, finally regarded 0.12 nm/cycle as the deposition By this approach, with a flat end surface forwe laser injection. Through the 10× microscope objectiverate. for matching the NA our gold633 modified nanoprobe wasthe coated with eight ALD cycles nm). The diagram representing of fiber, nm He-Ne laser with power of approximate 2 mW(~1 was injected into the fiber from its profile this after-wrapped nanoprobe illustrates in R6G Figure 3a. flat endofsurface. The backward SERS signal of the molecules on the nanoprobe surface was collected and transmitted by the same nanoprobe, coupled to the same “injection” objective, finally entered the Raman spectrometer (Figure 4b). The exposure time of Raman spectrometer was set to 10 s for each measurement. All the experimental conditions were the same in the whole measurement, to guarantee the results believable.

Sensors 2017, 17, 467

5 of 9

Table 1. Thickness of the film on the silicon wafer in 6 different points. Position

Thickness (nm)

N (633 nm)

#1 36.618 1.658 #2 37.083 1.658 #3of Au-NPs modified 37.822 TFNP with1.658 Figure 3. (a) Profile illustration ultrathin alumina layers; (b) SEM #4 TFNP with36.672 1.658 photograph of Au-NPs modified ultrathin alumina layers (with 10 cycles); (c) SEM #5 37.021 1.658 photograph of fiber on which deposited alumina film by ALD technique with 5000 cycles (first 2000 #6 37.706 1.658 cycles and then 3000 cycles).

2.5. 2.5. SERS SERS Measurements Measurements Rhodamine Rhodamine 6G 6G (R6G) (R6G) aqueous aqueous solution solution was was used used as as the the standard standard sample, sample, to to ascertain ascertain the the SERS SERS performances of the TFNP with ultrathin alumina layers. The Raman spectra of R6G were performances of the TFNP with ultrathin alumina layers. The Raman spectra of R6G were detected detected by by the the above above prepared prepared fiber fiber SERS SERS nanoprobe nanoprobe with with aa commercial commercial confocal confocal micro-Raman micro-Raman spectrometer spectrometer (Renishaw inVia plus) by direct measurement and in optrode configuration. The experimental setups (Renishaw inVia plus) by direct measurement and in optrode configuration. The experimental illustrate in Figure 4. setups illustrate in Figure 4. Raman spectrometer

Raman spectrometer Objective

Objective

Excitation light

Excitation light

Raman scattering signal

Raman scattering signal Sample

Sample Optical fiber tip

(a)

Optical fiber

(b)

Figure 4. Experimental setup for (a) direct measurement and (b) detection in optrode configuration. Figure 4. Experimental setup for (a) direct measurement and (b) detection in optrode configuration.

By By direct direct measurement, measurement, the the tip tip of of TFNP TFNP was was put put on on an an object object stage stage (Figure (Figure 4a), 4a), adjusting adjusting the the objective objective to to optimum optimum spot spot for for measurement. measurement. Using Using 633 633 nm nm He-Ne He-Ne laser laser as as exciting exciting light, light, whose whose power power is is around around 22 mW mW ahead ahead objective, objective, the the SERS SERS enhancement enhancement effect effect of of TFNP TFNP tip tip surface surface including including Au-NPs wrapped with different alumina layers was investigated. The microscope objective × Au-NPs wrapped with different alumina layers was investigated. The microscope objectiveused used50 50× for SERS signal signal from from the thetip tipsurface surfaceofofTFNP. TFNP.R6G R6Gaqueous aqueoussolution solutionwith witha for high high efficient efficient collecting collecting SERS −6 mol/L was used as the detection sample. aconcentration concentrationofof1010 −6 mol/L was used as the detection sample. In In optrode optrode configuration, configuration, the the fiber fiber SERS SERS nanoprobe nanoprobe was was usually usually cut cut into into segments segments of of 15–20 15–20 cm cm with a flat end surface for laser injection. Through the 10 × microscope objective for matching the NA of with a flat end surface for laser injection. Through the 10× microscope objective for matching the NA fiber, 633 nm He-Ne laser with the power of approximate 2 mW was injected into the fiber from its flat of fiber, 633 nm He-Ne laser with the power of approximate 2 mW was injected into the fiber from its end The backward SERS signal of the R6G molecules on the nanoprobe surface was collected flat surface. end surface. The backward SERS signal of the R6G molecules on the nanoprobe surface was and transmitted by the same nanoprobe, coupled to the same “injection” objective, finally entered the collected and transmitted by the same nanoprobe, coupled to the same “injection” objective, finally Raman spectrometer (Figure 4b). The exposure time of Raman spectrometer was set to 10 s for each entered the Raman spectrometer (Figure 4b). The exposure time of Raman spectrometer was set to 10 measurement. All the experimental conditions were the same in the measurement, to guarantee s for each measurement. All the experimental conditions were the whole same in the whole measurement, the results believable. to guarantee the results believable. 3. Results and Discussions The direct measurement results are shown in Figure 5, with the baseline subtracted and the Savitzky-Golay filter applied for clarity and to increase the signal-to-noise ratio without greatly distorting the signal [38], respectively. For each spectrum the baseline is the broken line composed of a series of segments, which is obtained by connecting the two lowest points of each Raman peak [32].

Sensors 2017, 17, 467

6 of 9

3. Results and Discussions The direct measurement results are shown in Figure 5, with the baseline subtracted and the Savitzky-Golay filter applied for clarity and to increase the signal-to-noise ratio without greatly Sensors 2017, 17, 467 6 of 9 distorting the signal [38], respectively. For each spectrum the baseline is the broken line composed of a series of segments, which is obtained by connecting the two lowest points of each Raman peak [32]. Obviously, all coated Au-NPs exhibited good SERS sensitivity, and the Raman peaks of the R6G molecule were marked marked in in the the two two figures, with the peaks’ position difference from the data process molecule were anan alumina layer, in Figure 5a, errors. Compared with with the the pink pinkline, line,which whichrepresents representsAu-NPs Au-NPswithout without alumina layer, in Figure the the Raman peakpeak intensity of theofAu-NPs wrapped with alumina layers layers (~1.2 nm) was lower around 5a, Raman intensity the Au-NPs wrapped with alumina (~1.2 nm) wasby lower by half. As half. a main to the enhancements of electromagnetics, Au-NPs were imprisoned by around Ascontributor a main contributor to the enhancements of electromagnetics, Au-NPs were alumina shells theirshells performance ofperformance enhancementofwas limited. In was addition, theIn Raman spectra imprisoned by and alumina and their enhancement limited. addition, the of Au-NPs with thicknesses alumina layers are more clearly compared Figure 5b. Raman spectra of different Au-NPs with differentofthicknesses of alumina layers are more clearly in compared in As expected, Raman signal intensity decreased withdecreased the increase of the Figure 5b. Asthe expected, the Raman signal intensity with the alumina increasethickness. of the alumina thickness.

1000

1652

800

1186

800

1312

600

600

10 cycles 20 cycles 30 cycles

1652.02

Intensity(a.u.)

1200

1364.26

1400

1000

1509.89

10 cycles 20 cycles 30 cycles none

1510

1186.65

1364

1600

Intensity(a.u.)

1800

400

200

400 200

0 0 1000

1100

1200

1300

1400

1500 -1

Raman Shift(cm )

(a)

1600

1700

1800

1000

1200

1400

1600

1800

-1

Raman Shift(cm )

(b)

Figure 5. (a) Comparison of Raman spectra of R6G (10−6 mol/L) adsorbed on gold NPs coated with Figure 5. (a) Comparison of Raman spectra of R6G (10−6 mol/L) adsorbed on gold NPs coated with and without alumina layers; (b) Raman spectra of R6G (10−6 mol/L) adsorbed on gold NPs coated and without alumina layers; (b) Raman spectra of R6G (10−6 mol/L) adsorbed on gold NPs coated with different alumina layers. with different alumina layers.

In order order to to ascertain ascertain the the remote remote sensing sensing potential potential of of the the TFNP TFNP with with ultrathin ultrathin alumina alumina protective protective In film, Raman spectra of the fiber itself and R6G aqueous solutions with four different concentrations film, Raman spectra of the fiber itself and R6G aqueous solutions with four different concentrations −4 mol/L, 10−5 mol/L, 10−6 mol/L, 10−7 mol/L) were measured in optrode configuration, shown (10−4 (10 mol/L, 10−5 mol/L, 10−6 mol/L, 10−7 mol/L) were measured in optrode configuration, shown in in Figure Raman spectra, except those thefiber fiberitself itselfand andthe thelowest lowestconcentration concentration of of the the R6G R6G Figure 6a.6a. Raman spectra, except those ofofthe −7 mol/L) aqueous solution due to them being too low to detect, from 1000 cm−1 to 1800 cm−1 with (10 −7 −1 −1 (10 mol/L) aqueous solution due to them being too low to detect, from 1000 cm to 1800 cm with the baselines baselines subtracted subtracted and and the the Savitzky-Golay Savitzky-Golay filter filter applied, applied, are are shown shown in in Figure Figure 6b. 6b. It It is is obviously obviously the demonstrated that our nanoprobes possess preferable remote sensing potential and Raman intensity demonstrated that our nanoprobes possess preferable remote sensing potential and Raman intensity in each peak position which became weaker and weaker with the decrease of the R6G concentration. in each peak position which became weaker and weaker with the decrease of the R6G concentration. −1 1363 cm−1 , 1510 cm−1 , 1651 cm−1 ) can still Furthermore, four Furthermore, four Raman Raman characteristic characteristicpeaks peaks(1186 (1186cm cm−1,, 1363 cm−1, 1510 cm−1, 1651 cm−1) can still be − 6 be recognized at the minimum concentration of 10 mol/L. Due beingwrapped wrappedwith withaa layer layer of of −6 recognized at the minimum concentration of 10 mol/L. Due to to being −6 mol/L in the ultrathin alumina film, the detection limit for R6G aqueous solution only reached 10 ultrathin alumina film, the detection limit for R6G aqueous solution only reached 10−6 mol/L in the optrode configuration, configuration,two twoorders ordersof ofmagnitude magnitudelower lowerthan thanininour ourprevious previous work [32] (with a limit optrode work [32] (with a limit of −8 mol/L). It is obviously shown, from these remote sensing results, that this kind of fiber SERS of 10 −8 10 mol/L). It is obviously shown, from these remote sensing results, that this kind of fiber SERS nanoprobe with with ultrathin ultrathin alumina alumina layers layers (~1 (~1 nm) nm) still still possesses possesses aa good good enhancement enhancement effect effect and and the the nanoprobe NPs would be protected by the alumina film for reusability and puncture measurements. Besides, NPs would be protected by the alumina film for reusability and puncture measurements. Besides, the fiber fiber background decreasing thethe peak of the background in in our ourexperiment experimentisisstill stillstrong strong(Figure (Figure6a), 6a),and andtherefore therefore decreasing peak thethe background will be be oneone of the future research aspects. of background will of the future research aspects.

Sensors 2017, 17, 467 Sensors 2017, 17, 467

7 of 9 7 of 9

1510.84

1649.43

30000

1363.39

1510.84

2000

1185.75

40000

1185.75

Intensity(a.u.)

50000

2500

Intensity(a.u.)

60000

1500

-4

10 mol/L R6G -5 10 mol/L R6G -6 10 mol/L R6G

1651.19

Fiber Background -4 10 mol/L R6G -5 10 mol/L R6G -6 10 mol/L R6G -7 10 mol/L R6G

1363.39

3000

70000

1000

20000

500

10000 200

400

600

800

1000

1200 -1

Raman Shift(cm )

(a)

1400

1600

1800

0 1000

1100

1200

1300

1400

1500

1600

1700

1800

-1

Raman Shift(cm )

(b)

Figure 6. (a) SERS spectra of R6G with different concentrations in remote mode; (b) Baseline subtracted Figure 6. (a) SERS spectra of R6G with different concentrations in remote mode; (b) Baseline from (a). subtracted from (a).

4. Conclusions 4. Conclusions We proposed a nanoscale taper-fiber SERS probe structure modified by Au-NPs wrapped with We proposed a nanoscale taper-fiber SERS probe structure modified by Au-NPs wrapped with alumina ultrathin film and presented its fabrication method. Its SERS performance in the optrode alumina ultrathin film and presented its fabrication method. Its SERS performance in the optrode configuration was demonstrated in experiments. The taper-fiber probe with a tip size under 80 nm configuration was demonstrated in experiments. The taper-fiber probe with a tip size under 80 nm was realized through heated pulling and chemical etching methods, with Au-NPs deposited on was realized through heated pulling and chemical etching methods, with Au-NPs deposited on the the nanoprobe surface by electrostatic self-assembly technology; the nanoprobe was wrapped with nanoprobe surface by electrostatic self-assembly technology; the nanoprobe was wrapped with ultrathin alumina layers using the ALD technique. It is indicated that in the same concentration of ultrathin alumina layers using the ALD technique. It is indicated that in the same concentration of R6G aqueous solution, the alumina layers were thicker and the Raman spectra enhancement effect was R6G aqueous solution, the alumina layers were thicker and the Raman spectra enhancement effect weaker due to the Au-NPs imprisoned by the thick alumina layers. Also, with approximately 1 nm was weaker due to the Au-NPs imprisoned by the thick alumina layers. Also, with approximately 1 alumina layers, the detection limit of the R6G aqueous solution reached 10−6−6mol/L in the optrode nm alumina layers, the detection limit of the R6G aqueous solution reached 10 mol/L in the optrode configuration. The nanoscale, remote sensing SERS-active nanoprobe, wrapped with ultrathin alumina configuration. The nanoscale, remote sensing SERS-active nanoprobe, wrapped with ultrathin layers for protecting the AuNPs, may have great potential for puncture biosensing such as intracellular alumina layers for protecting the AuNPs, may have great potential for puncture biosensing such as applications. An a great deal of effort should be devoted to suppressing the background Raman intracellular applications. An a great deal of effort should be devoted to suppressing the background scattering of the fiber itself to get the higher signal-to-noise ratio and long-distance sensing capabilities. Raman scattering of the fiber itself to get the higher signal-to-noise ratio and long-distance sensing capabilities. Acknowledgments: This work was supported by the Natural Science Foundation of China (NSFC) (61475095, 61520106014, 61422507), and the authors thank the support of the Key Laboratory of Specialty Fiber Optics and Optical Access Networks (SKLSFO2015-06). Acknowledgments: This work was supported by the Natural Science Foundation of China (NSFC) (61475095, 61520106014, 61422507), Wenjie and theXu authors thank support Key Laboratory Specialty Fiber Optics and Author Contributions: processed thethe data, wrote of thethe manuscript and wasofresponsible for interpretation Optical Access (SKLSFO2015-06). of the data andNetworks partial viewpoint. Zhenyi Chen and Na Chen proposed the idea and guided the experiments, Heng Zhang and Xinmao Hu performed the experiments, Shupeng Liu guided the writing, Jianxiang Wen and Author Xu the processed the equipment. data, wrote the manuscript and was responsible for TingyunContributions: Wang provided Wenjie and guided use of ALD interpretation of the data and partial viewpoint. Zhenyi Chen and Na Chen proposed the idea and guided the Conflicts of Interest: The authors declare no conflict of interest. experiments, Heng Zhang and Xinmao Hu performed the experiments, Shupeng Liu guided the writing, Jianxiang Wen and Tingyun Wang provided and guided the use of ALD equipment.

References

Conflicts of Interest: The authors declare no conflict of interest. 1. Tamor, M.A.; Vassell, W.C. Raman “fingerprinting” of amorphous carbon films. J. Appl. Phys. 1994, 76, 3823–3830. [CrossRef] References 2. Cao, Y.W.C.; Jin, R.; Mirkin, C.A. Nanoparticles with Raman spectroscopic fingerprints for DNA and RNA 1. Tamor, M.A.; Vassell, Raman ‘‘fingerprinting’’ of amorphous carbon films. J. App. Phys. 1994, 76, detection. Science 2002,W.C. 297, 1536–1540. [CrossRef] [PubMed] 3823–3830. 3. Nie, S.; Emory, S.R. Probing single molecules and single nanoparticles by surface-enhanced Raman scattering. 2. Cao, Y.W.C.; Mirkin, C.A. Nanoparticles with Raman spectroscopic fingerprints for DNA and RNA Science 1997, Jin, 275,R.; 1102–1106. [CrossRef] [PubMed] Science 2002, 297, 1536–1540. 4. detection. Gessner, R.; Rösch, P.; Petry, R.; Schmitt, M.; Strehle, M.A.; Kiefer, W.; Popp, J. The application of a SERS 3. Nie, Emory, S.R. Probing single molecules and single nanoparticles by 129, surface-enhanced Raman fiberS.; probe for the investigation of sensitive biological samples. Analyst 2004, 1193–1199. [CrossRef] scattering. Science 1997, 275, 1102–1106. [PubMed] 4. Gessner, R.; Rösch, P.; Petry, R.; Schmitt, M.; Strehle, M.A.; Kiefer, W.; Popp, J. The application of a SERS fiber probe for the investigation of sensitive biological samples. Analyst 2004, 129, 1193–1199.

Sensors 2017, 17, 467

5. 6. 7.

8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29.

8 of 9

Andrade, G.F.S.; Fan, M.K.; Brolo, A.G. Multilayer silver nanoparticles-modified optical fiber tip for high performance SERS remote sensing. Biosens. Bioelectron. 2010, 25, 2270–2275. [CrossRef] [PubMed] Fleischmann, M.; Hendra, P.J.; McQuillan, A. Raman spectra of pyridine adsorbed at a silver electrode. Chem. Phys. Lett. 1974, 26, 163–166. [CrossRef] Nguyen, C.T.; Nguyen, J.T.; Rutledge, S.; Zhang, J.; Wang, C.; Walker, G.C. Detection of chronic lymphocytic leukemia cell surface markers using surface enhanced Raman scattering gold nanoparticles. Cancer Lett. 2010, 292, 91–97. [CrossRef] [PubMed] Mulvihill, M.J.; Ling, X.Y.; Henzie, J.; Yang, P. Anisotropic etching of silver nanoparticles for plasmonic structures capable of single-particle SERS. J. Am. Chem. Soc. 2009, 132, 268–274. [CrossRef] [PubMed] Shao, Q.; Que, R.; Shao, M.; Cheng, L.; Lee, S.T. Copper Nanoparticles Grafted on a Silicon Wafer and Their Excellent Surface-Enhanced Raman Scattering. Adv. Funct. Mater. 2012, 22, 2067–2070. [CrossRef] Lin, X.M.; Cui, Y.; Xu, Y.H.; Ren, B.; Tian, Z.Q. Surface-enhanced Raman spectroscopy: substrate-related issues. Anal. Bioanal. Chem. 2009, 394, 1729–1745. [CrossRef] [PubMed] Betz, J.F.; Wei, W.Y.; Cheng, Y.; White, I.M.; Rubloff, G.W. Simple SERS substrates: Powerful, portable, and full of potential. Phys. Chem. Chem. Phys. 2014, 16, 2224–2239. [CrossRef] [PubMed] Schlücker, S. Surface-Enhanced Raman spectroscopy: Concepts and chemical applications. Angew. Chem. Int. Ed. 2014, 53, 4756–4795. [CrossRef] [PubMed] Pieczonka, N.P.W.; Aroca, R.F. Single molecule analysis by surfaced-enhanced Raman scattering. Chem. Soc. Rev. 2008, 37, 946–954. [CrossRef] [PubMed] Scaffidi, J.P.; Gregas, M.K.; Seewaldt, V.; Vo-Dinh, T. SERS-based plasmonic nanobiosensing in single living cells. Anal. Bioanal. Chem. 2009, 393, 1135–1141. [CrossRef] [PubMed] Stoddart, P.R.; White, D.J. Optical fibre SERS sensors. Anal. Bioanal. Chem. 2009, 394, 1761–1774. [CrossRef] [PubMed] Zhang, Y.; Shi, C.; Gu, C.; Seballos, L.; Zhang, J.Z. Liquid core photonic crystal fiber sensor based on surface enhanced Raman scattering. Appl. Phys. Lett. 2007, 90, 193504. [CrossRef] Yang, X.; Ileri, N.; Larson, C.C.; Carlson, T.C.; Britten, J.A.; Chang, A.S.; Bond, T.C. Nanopillar array on a fiber facet for highly sensitive surface-enhanced Raman scattering. Opt. Express 2012, 20, 24819–24826. [CrossRef] [PubMed] Viets, C.; Hill, W. Single-fibre surface-enhanced Raman sensors with angled tips. J. Raman Spectrosc. 2000, 31, 625–631. [CrossRef] Lan, X.; Han, Y.; Wei, T.; Zhang, Y.; Jiang, L.; Tsai, H.L.; Xiao, H. Surface-enhanced Raman-scattering fiber probe fabricated by femtosecond laser. Opt. Lett. 2009, 34, 2285–2287. [CrossRef] [PubMed] Cao, J.; Zhao, D.; Lei, X.; Liu, Y.; Mao, Q. One-pot hydrothermal synthesis of silver nanoplates on optical fiber tip for surface-enhanced Raman scattering. Appl. Phys. Lett. 2014, 104, 201906. [CrossRef] Lucotti, A.; Zerbi, G. Fiber-optic SERS sensor with optimized geometry. Sens. Actuators B Chem. 2007, 121, 356–364. [CrossRef] Sharma, A.K.; Jha, R.; Gupta, B.D. Fiber-optic sensors based on surface plasmon resonance: A comprehensive review. IEEE Sens. J. 2007, 7, 1118–1129. [CrossRef] Jayawardhana, S.; Mazzolini, A.P.; Stoddart, P.R. Collection efficiency of scattered light in single-ended optical fiber sensors. Opt. Lett. 2012, 37, 2142–2144. [CrossRef] [PubMed] Li, J.F.; Huang, Y.F.; Ding, Y.; Yang, Z.L.; Li, S.B.; Zhou, X.S.; Fan, F.R.; Zhang, W.; Zhong, Z.Y.; Wu, D.Y.; et al. Shell-isolated nanoparticle-enhanced Raman spectroscopy. Nature 2010, 464, 392–395. [CrossRef] [PubMed] Whitney, A.V.; Elam, J.W.; Stair, P.C.; Van Duyne, R.P. Toward a thermally robust operando surface-enhanced Raman spectroscopy substrate. J. Phys. Chem. C 2007, 111, 16827–16832. [CrossRef] Formo, E.V.; Mahurin, S.M.; Dai, S. Robust SERS substrates generated by coupling a bottom-up approach and atomic layer deposition. ACS Appl. Mater. Interfaces 2010, 2, 1987–1991. [CrossRef] John, J.F.; Mahurin, S.; Dai, S.; Sepaniak, M.J. Use of atomic layer deposition to improve the stability of silver substrates for in situ, high-temperature SERS measurements. J. Raman Spectrosc. 2010, 41, 4–11. [CrossRef] Bachenheimer, L.; Elliott, P.; Stagon, S.; Huang, H. Enhanced thermal stability of Ag nanorods through capping. Appl. Phys. Lett. 2014, 105, 213104. [CrossRef] Zhang, W.; Dong, J.C.; Li, C.Y.; Chen, S.; Zhan, C.; Panneerselvam, R.; Yang, Z.L.; Li, J.F.; Zhou, Y.L. Large scale synthesis of pinhole-free shell-isolated nanoparticles (SHINs) using improved atomic layer deposition (ALD) method for practical applications. J. Raman Spectrosc. 2015, 46, 1200–1204. [CrossRef]

Sensors 2017, 17, 467

30.

31.

32. 33. 34. 35. 36. 37.

38.

9 of 9

Ma, L.; Huang, Y.; Hou, M.; Xie, Z.; Zhang, Z.J. Silver nanorods wrapped with ultrathin Al2O3 layers exhibiting excellent SERS sensitivity and outstanding SERS stability. Sci. Rep. 2015, 5, 1–9. [CrossRef] [PubMed] Mahurin, S.M.; John, J.; Sepaniak, M.J.; Dai, S. A reusable surface-enhanced Raman scattering (SERS) substrate prepared by atomic layer deposition of alumina on a multi-layer gold and silver film. Appl. Spectrosc. 2011, 65, 417–422. [CrossRef] [PubMed] Chen, Z.Y.; Dai, Z.M.; Chen, N.; Liu, S.P.; Pang, F.F.; Lu, B.; Wang, T.Y. Gold nanoparticles-modified tapered fiber nanoprobe for remote SERS detection. IEEE Photonics Technol. Lett. 2014, 26, 777–780. [CrossRef] Lazarev, A.; Fang, N.; Luo, Q. Formation of fine near-field scanning optical microscopy tips. Part II. By laser heated pulling and bending. Rev. Sci. Instrum. 2003, 74, 3684–3688. [CrossRef] Lazarev, A.; Fang, N.; Luo, Q.; Zhang, X. Formation of fine near-field scanning optical microscopy tips. Part I. By static and dynamic chemical etching. Rev. Sci. Instrum. 2003, 74, 3679–3683. [CrossRef] Gu, N.; Li, C.; Sun, L.; Liu, Z.H.; Sun, Y.K.; Xu, L.N. Controllable fabrication of fiber nano-tips by dynamic chemical etching based on siphon principle. J. Vac. Sci. Technol. B 2004, 22, 2283–2285. [CrossRef] Frens, G. Controlled nucleation for the regulation of the particle size in monodisperse gold suspensions. Nature 1973, 241, 20–22. [CrossRef] Hu, L.; Chen, Z.Y.; Chen, N.; Zhang, W.; Zhu, H.; Liu, S.P.; Wang, T.Y. Preparation of gold colloid and its surface-enhanced Raman scattering properties. In Proceedings of the Communications and Photonics Conference and Exhibition, Shanghai, China, 13–16 November 2011. Savitzky, A.; Golay, M.J.E. Smoothing and differentiation of data by simplified least squares procedures. Anal. Chem. 1964, 36, 1627–1639. [CrossRef] © 2017 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/).