Escherichia coli Fiber Sensors using Concentrated Dielectrophoretic

May 17, 2018 - A sensitive tapered optical fiber tip combined with dielectrophoretic (DEP) trapping was used to rapid and label-free detection of bact...
0 downloads 0 Views 825KB Size
Download PDF
Subscriber access provided by Kaohsiung Medical University

Article

Escherichia coli Fiber Sensors using Concentrated Dielectrophoretic Force with Optical Defocusing Method Yi-Hsin Tai, Chia-Wei Lee, Dao-Ming Chang, Yu-Sheng Lai, Ding-Wei Huang, and Pei-Kuen Wei ACS Sens., Just Accepted Manuscript • DOI: 10.1021/acssensors.8b00258 • Publication Date (Web): 17 May 2018 Downloaded from http://pubs.acs.org on May 17, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 10 1 2 3 4 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 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

Escherichia coli Fiber Sensors using Concentrated Dielectrophoretic Force with Optical Defocusing Method †

†

‡

§

∥

Yi-Hsin Tai , Chia-Wei Lee , Dao-Ming Chang , Yu-Sheng Lai , Ding-Wei Huang and Pei-Kuen Wei†,⊥,#,* †

Research Center for Applied Sciences, Academia Sinica, Taipei, 11529, Taiwan

‡

Department of Biochemical Science and Technology, National Taiwan University, Taipei 10617, Taiwan

§

Department of Biochemistry and Molecular Biology, College of Medicine National Taiwan University, Taipei, 10051, Taiwan ∥

Graduate Institute of Photonics and Optoelectronics, National Taiwan University, Taipei 10617, Taiwan

⊥

Institute of Biophotonics, National Yang-Ming University, Taipei 11221, Taiwan

#

Institute of Optoelectronic Sciences, National Taiwan Ocean University, Keelung 20224, Taiwan

KEYWORDS: Fiber tip, Defocused image, Refractive index detection, Dielectrophoretic force, E. coli

ABSTRACT: A sensitive tapered optical fiber tip combined with dielectrophoretic (DEP) trapping was used to rapid and label-free detection of bacteria in water. The angular spectrum of optical field at the fiber tip was changed with surrounding refractive index (RI). By measuring far-field intensity change at the defocus plane, the intensity sensitivity was up to 95,200 %/RIU (RI unit) and the detection limit was 5.2 × 10-6 RIU at 0.5% intensity stability. By applying an AC voltage to a Ti/Al coated fiber tip and an indium–tin-oxide glass, the DEP force effectively trapped the Escherichia coli (E. coli) near the fiber tip. Those bacteria can be directly measured from optical intensity change due to the increase of surrounding RI. By immobilizing antibody on the Ti/Al fiber tip, the tests for specific K12 bacteria and nonspecific BL21 bacteria verified the specificity. The antibody-immobilized Ti/Al coated fiber tip with DEP trapping can detect bacteria at a concentration about 100 CFU/mL.

Optical fiber sensors take advantages of flexible, light, small-volume, rapid and remote detections and are easily combined with microfluidic devices.1-6 By immobilizing the specific antibody on the surface of the sensing area, optical sensors can identify specific targets for biosensing applications.7,8 The detection sensitivity can be increased by immobilizing metal nanoparticles on the targets9,10 or optimizing the structures and coating materials on the fibers.11-13 In addition, the targets at low concentration can be enriched on the sensing area by the electrical or magnetic force.14-16 Most optical fiber detections are based on the fluorescence labeling or absorption of targets. Labelfree detections, such as surface plasmon resonance (SPR)17,18 and optical interference19 have been applied for fiber sensors. Flat surfaces are often required for fiberSPR excitation or forming the optical interference. The flat surface cannot generate large gradient of the electric field for trapping small particles. In this study, we developed a label-free, highly sensitive fiber tip sensor combined with large gradient of the electric field for rapid detection of specific bacteria in water. The detection is based on measuring the far-field intensity of a tapered fiber tip. The tip sensor takes advantages

of ultra-small sample volume, easy fabrication and detection. In our previous work,20 we have demonstrated that the optical intensity at the fiber tip is sensitive to the environmental refractive index (RI) change. The detection using a lower numerical aperture (NA) lens had a higher RI sensitivity. In this work, we proposed a simple way to further enhance the sensitivity by using defocused image method. With the defocused images, the measured RI sensitivity can be greatly enhanced up to 95,200 %/RIU (RI unit), which was about 10 times higher than our previous work using photo-detectors. This new approach is applied to detect specific bacteria in water. The bacteria near the fiber will increase the surrounding RI so does the optical distribution is changed. By measuring the transmission intensity change, the bacterial concentration can be detected without any labeling. Moreover, using a thin Ti/Al coating on the fiber tip and an indium–tin-oxide (ITO) glass substrate as a counter electrode, a large gradient of electric field can be formed near the fiber tip. Such gradient field provides a dielectrophoretic (DEP) force for trapping bacteria and enriching the target concentration.21 When specific antibody was immobilized on the

ACS Paragon Plus Environment

ACS Sensors 1 2 3 4 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 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

coated fiber tip, the modified fiber tip can detect specific bacteria at a low concentration in a short time. It is noticed that other fiber optical sensors have been demonstrated for bacteria detections, such as prior works using fiber gratings.6 Their approach used the resonant wavelength shift as the signal. The wavelength interrogation is more robust than the intensity interrogation. It still needs a spectrometer and gives no information of the optical image of the detection region. Besides, the DEP is hard to be included in the fiber grating setup. In our work, by using the fiber tip, an ITO electrode and a simple optical microscope, we demonstrated an alternative method for sensitive bacterial detection with the functionality of DEP enhancement and optical images.

METHODS AND MATERIALS Optical setup. Figure 1a illustrates the measurement setup for bacteria detection using a Ti/Al-coated fiber tip.

Figure 1. (a) The optical setup for the fiber tip with DEP measurement. An AC voltage was applied to the Ti/Al-coated fiber tip and the ITO electrode to generate the DEP force. The optical image was taken by a CCD through an objective lens. The inset illustrates the focusing effect near the fiber tip (RI = n1) for difference RI environment (RI = n2). (b) Simultaneous RI measurement and bacteria trapping using DEP effect near the fiber tip. Optical images of the fiber tip with various RI (n = 1.332, 1.358, and 1.401) at the focal plane (c)– (e), and the defocused plane (f)–(h).

The tapered fiber tip was made from a single-mode fiber (Thorlabs SM600) by a modified wet etching method.22 The 632-nm diode laser was coupled into the fiber, and the optical image near the tip of the tapered fiber was taken by an objective lens and a charge-coupled device (CCD). The average intensity of the CCD image was used as the optical signal. The fiber tip was fixed on a precision stage with 1 µm accuracy for tuning the focusing distance. To obtain the focused position, the smallest light spot on the tip was first determined through the CCD image. Then, the defocusing distance was decided by tuning the position of objective lens. We have checked the average

Page 2 of 10

intensity variations at same defocusing distance with different fibers. The variations were below 5% for repeated measurements. An alternating current (AC) voltage was applied to the Ti/Al coated fiber tip and the indium–tinoxide (ITO) glass. A steady water flow is applied to the sample holder for removing the untrapped bacteria. Figure 1b shows the magnification of the sensing area. The gradient of the electric field generates the DEP force, which attracts bacteria toward the fiber tip. With the DEP trapping force, bacteria were gathered near the fiber tip and the surrounding RI was increased accordingly. Figures 1c–1e show the CCD images of optical fields when the fiber tip was surrounded with different RI environment. The optical intensity and spot size were increased with the environmental RI. The change of spot size is owing to the change of the focusing effect near the tapered tip. As illustrated in figure 1a, refractive index difference between the glass tip and its environment forms refracted beams focusing at the fiber tip. When environmental refractive index increases, the refraction angle is reduced and the spot size is increased. After adjusting the distance of 0.91 mm below the focal plane, the optical fields for various surrounding RIs are shown in Figures 1f–1h. Compared with the changes in the focal plane, the defocused images show a larger sensitivity to the environmental RI. A suitable defocused distance can greatly enhance the optical detection sensitivity. Theoretical calculations. To verify the RI sensitivity of the fiber tip at different focal plane, we employed the finite-difference time-domain (FDTD) method (Fullwave, Rsoft) to calculate the optical near-field distribution and far-field angular spectrum The optical field near the tapered fiber tip is shown in Figure 2a. The parameters for the simulations were ∆x = ∆y = ∆z = 20 nm, time step (c∆t) = 10 nm, input wavelength = 632 nm. The boundary condition was set as a perfectly matched layer (PML) condition. The refractive index difference between the glass tip and its environment forms a focused spot near the fiber tip. When environmental refractive index increases, the refraction angle is reduced and the spot size is increased. From Fourier optics theory, such larger spot has a narrower angular spectrum. The collected intensity is higher than the small focused spot when a low numerical angle (NA) lens is used.20 The far-field angular spectrum can be calculated by taking the Fourier transform of optical spot at the fiber tip. Figure 2b shows the angular spectrum of the fiber tip for various surrounding RI. With the increase of surrounding RI, the intensity is decreased at a larger diffraction angle and increased at a smaller angle. The optical intensity with a small diffraction angle has a larger intensity change with the environmental RI. It is known the maximum collection angle is related to numerical aperture (NA). A lower NA lens has a lower collection angle. Figure 2c shows the transmission intensities in various surrounding media with different RIs and NAs. The transmission intensity was increased with the RI of the surrounding medium. The intensity sensitivity was greatly

2 ACS Paragon Plus Environment

Page 3 of 10 1 2 3 4 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 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors the AC voltage. The accumulation of targets increases the environmental RI of the tip. It thus enhances the detection of the biological particles on the Ti/Al coated fiber tip and reduces the detection time.

Figure 2. (a) Calculated optical fields near the fiber tip. (b) The angular spectrum of the fiber tip for different surrounding refractive indices. The NA indicates the numerical aperture of the lens. (c) The far-field optical intensity change as a function of refractive index for different NA value. (d) The equivalent NA can be reduced by using the defocusing method.

enhanced by using a low NA. Usually, the NA is small by using a low magnified lens. Herein, we show an alternative approach without changing the high-NA lens. Figure 2d shows the defocused image method. The idea is to adjust the lens-sample distance in the optical setup. When defocused distance is increased, the collecting angle of lens becomes small. It is equivalent to the decrease of the NA. Dielectrophoretic (DEP) force. By creating a nonuniform electric field, the DEP force can be exerted on dielectric particles in small volumes of liquids for the manipulation or the concentration.23,24 The general equation for the DEP force is shown as bellow.   

  

 |  |

  

(1)

Where   is the electric field by applying the AC voltage.  and  are the frequency-dependent complex permittivities of the particle and the medium. is the radius of the particle. By adjusting the frequency  of the AC voltage, the     can be changed to be plus or minus sign. Therefore, the DEP force can attract or repulse the dielectric particle in the medium by controlling the frequency. By using a tip electrode, the high gradient difference of the squared electric field can create strong DEP force without applying a high AC voltage. In our setup, the high gradient difference occurred at the Ti/Al coated fiber tip. The tip can concentrate or repel targets near its environment by tuning the frequency of

Antibody immobilization procedures. AntiEscherichia coli antibodies were adhered to the aluminum surface of the fiber tip by following surface modification. First, the Ti/Al fiber tip was treated with O2 plasma (500 W for 10 min) to create a solid OH group on the surface. The fiber tip was immersed in 10% (v/v) 3ammineopropyl-triethoxysilane (APTES) (Sigma-Aldrich A3648) in methanol for 10 min25 to form amino group on the surface. After immersion, the fiber tip was washed with methanol and dried for 1 h at 120 °C. The fiber tip was then incubated for 1 h with 10 mM solution of NHSLC-Biotin (Thermo 21336) in 150 μL of dimethyl sulfoxide and 850 μL of phosphate buffer (pH 8.0). 26-28 The Biotin will be conjugated on the tip surface. Next, the fiber tip was washed with phosphate buffer for 5 min, and the tip was immersed in 10 μg/mL streptavidin (Sigma-Aldrich S6402) phosphate-buffered saline (PBS) with 0.1% Tween20 for 1 h.29-31 The tip was washed in PBS with 0.1% Tween20 for 5 min and PBS alone for 5 min. Next, the tip was immersed in 4 mg/mL of anti-E.coli antibody (Biotin) (Abcam ab20640) with PBS for 1 h and washed in PBS for 5 min.32,33 The fiber tip was soaked in PBS with 3% BSA overnight at 4 °C to prevent nonspecific binding. Finally, the antibody-immobilized fiber tip was washed in PBS for 5 min and used in the experiments. All the coating and washing steps were performed with a shaker at 2 Hz to improve the uniformity and cleanliness on the aluminum surface of the tip. Bacterial preparation. Nonspecific BL21 and specific K12 bacteria were used in the antibody experiment. BL21 was produced from ECOS™ competent cells [BL21(DE3)], and K12 was prepared using a commercial kit (pGLO™ Bacterial Transformation Kit #1660003). The bacteria were processed with a plasmid to transfer the green fluorescence protein for staining in microscopy. Deionized water (DI) was used as a buffer medium for bacteria in the antibody experiments.

RESULTS AND DISCUSSION Optimized fabrication condition of the fiber tip sensor. The tapered fiber tip was fabricated through the wet etching method. To obtain the optimal fiber tip for the RI measurement, the wet-etching temperatures were set at 10, 20, 30, 40, and 50 °C in hydrofluoric (HF) acid, respectively. For testing the RI sensitivity, medium with different RI was prepared by mixing glycerin and DI water. The RI difference is the difference between pure water and the medium. We used a refractometer to measure RI of glycerol solutions at different concentrations. The change in the transmission intensity was determined for different RIs of the surrounding media. All the measurements were fixed at the same focal plane. After recording

3 ACS Paragon Plus Environment

ACS Sensors 1 2 3 4 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 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

changes in the intensity distribution, the RI sensitivity was calculated. The etching temperature for fiber tips was optimized according to the RI sensitivity. SEM images of the tip and tip angle versus etching temperature are presented in Figure 3a-b. The angle of the etched tip in

Figure 3. (a) SEM images for various fiber tip angles at different etching temperatures. (b) Histograms of the tip angles at different temperature (N = 5). (c) Sensitivities of the etched fiber tips in media with various refractive indices (N = 5).

creased with the increase of the temperature. The sensitivity of the fiber tip at different etching temperature was measured as seen Figure 3c. The average intensity sensitivity at etching temperatures between 20°C and 30°C was higher than those at other etching temperatures. The highest intensity sensitivity was about 3,500 %/RIU. It is noted that due to the wet etching process, the tip shape cannot be well controlled. Some tips may look not as regular as symmetrical cones. In Fig. 3, we can see variations for the optically RI measurements under the same fabrication condition. The variation is larger for larger tip’s angle. The shape of the fiber tip would also affect the DEP trapping force. Because the gradient of the electric field is also dependent on the sharpness of the fiber tip. Since the tip shape cannot be well controlled under current fabrication process, it is necessary to do the calibration curves of the RI for each fiber tip before the detections for the analytical samples Defocusing technique for enhancing sensitivity. The fiber tip was placed in media with various refractive indices, and the 10× objective lens was defocused at 0, 0.72, 1.45, and 2.15 mm to measure the relation between the defocal distance and sensitivity of the RI. The average intensity in the CCD was then calculated and recorded. The CCD images and the sensitivities of the RI at different focal distances are shown in Figure 4a-b. Compared with the focal distance (d), the intensity sensitivity at defocused planes 0.72, 1.45 and 2.15 mm show significant im-

Page 4 of 10

provement. When the RI was changed from 1.332 to 1.431,

Figure 4. (a) Optical image was collected at defocusing distances of 0 mm, 0.72 mm, 1.45 mm, and 2.15 mm below the focal plane. (b) Average intensity was measured in media with various refractive indices from 1.332 to 1.431 at defocusing distances of 0 mm, 0.72 mm, 1.45 mm, and 2.15 mm below the focal plane. (c) The stability was measured in solutions

with various refractive indices. the average difference in the intensity at d = 1.45 mm was 65 times higher than that at the focal plane. The optimal defocused case had an averaged intensity sensitivity up to 95,200 %/RIU. In our experiments, the tip fiber, lens and the CCD detector were fixed during the measurement. Only the medium was changed. It assured that the intensity signals coming from the environmental refractive index change near the fiber tip. Figure 4c shows the stability test of the system. With 0.5% intensity noise in our current system, the defocusing method can detect ~ 5.2 × 10-6 RIU change. This resolution is comparable with SPR sensing system using an optical prism and precisely angular detection system.34 It is noted that the distance between then tip and the ITO surface was kept at tens of micron meter. There was no change in the distance for the DEP trapping. For the optical measurement, the defocused length was tuned by changing the position of the objective lens. The best length was 1.45 mm from the focal position of the objective. The optimal condition for the defocused length occurs when the optical pattern fits with the size of the CCD sensor. For a smaller defocused length (d = 0, 0.72 mm), the NA is high. Therefore, the sensitivity is low. For a longer defocused length (d = 2.15 mm), the optical pattern is larger than the CCD size. Some information about of intensity change is lost. It results in a decrease of sensitivity. Defocusing and DEP techniques for bacterial measurement. Figure 5a presents the results for detecting bacterial in the medium by using the defocusing method

4 ACS Paragon Plus Environment

Page 5 of 10 1 2 3 4 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 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

and the DEP force. For the frequency-dependent force

to surrounding RI change.20 If the optical signals come from the scattering, the decrease of NA will reduce the collection of scattering photons. However, in the measurement, the optical intensity had an obvious increase with the decrease of NA value as seen in Figure 5a. Therefore, the surrounding RI change due to the bacteria adhesion plays the major role for the optical signals. The time-dependent trapping can be fitted by the exponential growth equation.    

Figure 5. (a) The response curve for different focal lengths, 0 4 mm and 0.1 mm. The bacterial concentration was 5.5 × 10 CFU/mL. The bacteria were repelled and attracted at AC frequencies of 1 KHz and 1 MHz, respectively. (b) The measured intensity changes at different defocusing distances after trapping bacteria. (c) The response curve for different applied voltages, 1.5 and 1 Vrms. Note that at 300 seconds, a steady flow is applied to remove the unattached bacteria. At 1500 seconds, the DEP is off. (d) Statistic results of bacteria trapping under different conditions: non-trapping (black bar), trapping with different voltages (red bar), and trapping with flow (blue bar).

under the condition of water   and bacteria  , it was reported that 1 KHz AC frequency generated the repulsive force and 1 MHz produced the trapping force.35 In this experiment, 1 KHz AC frequency was first applied to repel environmental particles. After 100 seconds, the AC frequency was switched to 1 MHz to attract the bacteria. We used a 40× objective lens instead of 10× lens for magnifying the tip area in the captured image. By using the 40× objective lens, the average intensity for the bacteria detection was much higher at 0.1-mm defocused distance than at the focal plane. To find the optimal defocused length for trapping bacterium on the tip, the change of average intensity was measured at different defocusing distances: 0 mm, 0.1 mm, 0.2 mm, 0.3mm, 0.4 mm and 0.5 mm. The resultent was shown in Figure 5b. With 40× objective lens, the most change in intensity was occured at 0.1-mm defocusing distance. There were fluctuations in the signals because some bacteria were not attached on the tip. They moved around the tip region and confined by the DEP force. Due to the random motion, the surrounding RI had a variation with time. It is noted that the signals from the fiber tip can be classified into two parts. One is the scattering of optical near-field by microparticles.36 The other is the transmission intensity change due

!"

##$ ⁄%

(2)

Where '( is the average intensity. *+ , (+ , ,- and . are the fitting parameters for background, initial time, amplitude and time constant, respectively. In Figure 5a, two fitting curves at different defocusing distances have a similar . ~ 72 seconds. ,- is 1.46 at 0.1-mm defocusing distance. This value is three times larger than the ,- at focal distance due to the enhanced detection sensitivity. In order to exclude the variation and unattached bacteria near the tip, we used a syringe pump to provide a steady flow to remove the unattached bacteria. The AC voltage was maintained below the redox potential of the aluminum (1.66 V) to avoid the damage of Ti/Al coating layer. Figure 5c shows the DEP trapping with different voltages. For 1 Vrms trapping voltage, there was a large fluctuation of signals due to the small trapping force. At 300 seconds trapping, a steady flow with 200 μL/min rate was applied. The signals suddenly dropped because unattached bacterium were taken away. Figure 5d shows the statistical result for DEP trapping with and without flow under different applied voltage and bacterial concentration. For 1 Vrms DEP force, there was no bacteria on the tip under the flow condition. For 1.5 Vrms DEP force, a stable bacteria signal can be measured. More than 4% intensity change was measured for the bacterial concentration of 6.8 × 103 CFU/mL. Antibody experiment. Using the DEP effect, bacteria can be trapped near the tip. Once the DEP force is released, the bacteria will be taken away by the fluidic force. However, when the tip surface is modified with the specific antibody, the binding force can capture the bacteria without the DEP force. It can be applied for rapid detection of specific bacteria without labeling. The antibody immobilization procedures are shown in Figure 6a. After the biotin conjugation, we used additional fluorescent streptavidin-Cy3 to verify the surface modification. The fluorescent distribution of streptavidin-Cy3 on the tip was shown in Figure 6b. Uniform distribution of streptavidin can be seen on the fiber tip surface. It is noted that antibody concentration is a key issue for the optimization of the specific bacterial detection. To check the optimal condition, different concentrations of biotinylated antibodies were connected to the streptavidin surface. The concentrations were 0.1, 0.4, 4 mg/mL. The adhesion of antibody was checked by the fluorescence of IgG antibody

5 ACS Paragon Plus Environment

ACS Sensors 1 2 3 4 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 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

DyLight 550 (Bethyl A120-100D3). Figure 6c shows the fluorescent images of IgG antibody. Obvious binding of

Page 6 of 10

Through the same antibody immobilization process, the fiber tip can be used for another detection. To study the detection limit and linearity, K12 bacteria at concentrations of 4 × 102, 4 × 103, 4 × 104, 4 × 105, 4 × 106 and 4 × 107 CFU/mL were tested. They were measured sequentially using the antibody-immobilized fiber tip. The experimental results were shown in Figure 7a. The

Figure 6. (a) The antibody immobilization procedures are shown below. (1) Aluminum surface on the fiber tip was modified with APTES. (2) NHS-LC-Biotin was bound with APTES. (3) NHS-LC-Biotin used biotin to connect streptavidin. (4) Biotinylated antibodies were connected to the streptavidin surface by using the biotin part. (5) BSA medium was used to block the surface. (6) E. coli was bound by the antibody on the fiber surface. (b) Fluorescent image of Streptavidin-cy3 on the fiber tip. (c) The fluorescent images of IgG

antibody on the fiber tip for modification with different antibody concentrations: 0.1, 0.4, 4 mg/ml, respectively. (d) SEM image of the adhered K12 bacteria on the tip. Upperright image: the fluorescence of adhered K12 bacteria on the tip. Lower-right image: the overlapping image of fluorescence of adhered K12 and optical spot at the fiber tip.

IgG antibody was found. Antibody concentration at 4 mg/mL made evenly adhesion. Therefore, in our experiment, we used 4 mg/mL for producing enough antibodies on the tip surface. After immobilization of specific antibody, the modified fiber tip was used to trap bacteria in water with the DEP force and flow. The image on the fiber tip after the bacteria capture was examined using a scanning electron microscope. Figure 6d shows the SEM image. It identified the bacterial adhesion on the tip area. To further check the bacteria capture, we used green fluorescence protein (GFP) transferred bacteria for optical images. The inset of Figure 6d shows the green fluorescence of bacteria at the tip. The GFP fluorescence was excited by 490-nm-wavelength light. The GFP image was taken using a dichroic mirror and the 40× objective lens. The fluorescence image can be compared with the image obtained at 632-nm-wavelength light emitted from the tip. This red light was used for the surrounding RI sensing. It is obvious that fluorescence of K12 bacteria and the red spot were overlapped. The overlapping image also indicated the bacteria at the tip region. The antibody immobilized fiber tip cannot be refreshed again after trapping bacterium due to the strong binding between the antibody and bacterial surface. But the fiber tip could be reused after removing the aluminum layer by chemical etching.

2

3

Figure 7. (a) K12 bacteria at concentrations of 4 × 10 , 4 × 10 , 4 5 6 7 4 × 10 , 4 × 10 , 4 × 10 and 4 × 10 CFU/mL. They were measured sequentially using the antibody-immobilized fiber tip. (b) The measured intensities for various samples: control, BL21 and K12 in DI water. The concentration for K12 and BL21 6 bacteria were about 10 CFU/mL. (c) The measured intensities for various samples: control, BL21 and K12 in lake water.

flow rate of water was maintained at 200 μL/min. The referenced optical signal was set at AC voltage = 0 Vrms. After t = 200 seconds, K12 bacteria (4 × 102 CFU/mL) was injected into the channel and a 1.5-Vrms 1-MHz AC power was applied for the DEP trapping. The bacteria were trapped on the tip for 400 seconds. Then, the voltage was turned off. These measurement was repeated for different bacterial concentrations and operated at room temperature. The result shows a good linearity when bacteria number is between 102 to 106 CFU/mL. For 4 × 104 CFU/mL bacteria, the optical intensity was two times higher than the initial intensity. For 4 × 106 CFU/mL bacteria, increasing the bacterial concentration increased the optical intensity when DEP was on. However, the intensity remained at the similar level as 4 × 105 CFU/mL when DEP was off. This is attributed to the limited area for the bacteria binding on the tip. When the tip surface was occupied with bacteria, excessive bacteria would not cause the increase of signals. If the concentration is too high, the scattering of bacteria would further decrease the transmission intensity as observed in the case of 4 × 107 CFU/mL. For the proposed fiber tip sensor, the highest detectable concentration for bacteria was around 106 CFU/mL. For the detection limit, the noise level is the key

6 ACS Paragon Plus Environment

Page 7 of 10 1 2 3 4 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 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

issue. Under current system, the mean noise level is about 0.5%, the detection limit is about 102 CFU/mL. This detection limit is useful for the fast evaluation of bacteria in water and drink. Specificity experiment. For the specificity test, BL21 and K12 bacteria were examined sequentially using the antibody-immobilized fiber tip. Two types of bacterial media were prepared by collecting one colony of the bacteria on a dish and then dissolving it in 1-mL DI water. First, the tip was immersed in DI water. After 100 s, the BL21 medium flowed into the sensing area at a flow rate of 200 μL/min, and 1.5-Vrms AC voltage was applied. After 500 seconds, the DEP was off and DI water flowed into the fiber tip to remove the unbounded bacteria. The measurement steps were repeated by replacing the BL21 with the K12 bacteria. For the control experiment, the fiber tip without antibody modification was tested for K12 bacteria with the same concentration. The results of the specific binding test were presented in Figure 7b. The control experiment and non-specific BL21 bacteria had similar measured intensities with the initial level. It indicates that BL21 was not adhered to the tip surface. The K12 showed significant increase of the optical signal which was 3.6 times higher than the control and BL21. In Fig. 7b, the concentration for BL21 bacteria was about 106 CFU/mL. As compared with the result shown in Fig. 7a, the K12 has a averaged intensity value higher than BL21 ( 106 CFU/mL) even when its concentration is lower than 4 × 102 CFU/mL. There is about three orders of magnitudes for specific and non-specific bacterial concentration. The result demonstrated the specificity of the antibodyimmobilized fiber tip. The antibody coated fiber tip was also tested for bacteria in lake water. The lake water was collected from Dahu park in Taipei city. The lake water was left to stand for 24 hours in order to separate from silt. The nonspecific bacterial BL21, has a similar response to the control condition. The K12 shows better binding signals on the tip surface. However, the intensity difference between specific and non-specific bacteria is only 1.6 times for lake water. It is only 0.45 times of the intensity signals in pure water. We deduce that K12 binding is affected by the large particles in lake water. As indicated in Eq. (1), the DEP force is proportional to cube of the particle’s diameter. Those larger particles will cover the tip surface during DEP trapping and hinder the binding of K12 bacteria. Therefore, the number of immobized bactria after DEP and washing flow is much smaller than that in pure water. The pretreament of sample is necessay for using the fiber tip-DEP method for real sample measurement. For water samples with larger particles, such as lake water and juice, it is suggested to use membranes with a pore size of 10 µm to remove large particles. For biological samples, the specific antibody was used to recognize K12

bacteria. Since the DEP can only trap and enhance the micron biological particle numbers near the fiber tip, the effect of small biomolecules can be reduced by applying DEP effect. But for complicated biological samples, such as human blood, due to high concentration of biological particles (such as red cells), the sample pretreatment is necessary. A possible way to reduce the substrate interference is to filter the sample with micron size pore filters. Another way is to dilute the sample with pure buffer. Under the dilute condition, the bacterial number is proportionally reduced. The DEP active time may need to be increased for attracting more bacteria to the fiber tip.

CONCLUSIONS An optical image method for antibody modified fiber tip was proposed for measuring bacteria with good specificity and sensitivity. Unlike the surface plasmon resonance sensors or gratings-based fiber sensors, the proposed fiber tip sensor has no obvious frequency change. It can only be applied for intensity interrogation measurement. In the current system, the intensity sensitivity was 95,200 %/RIU. With 0.5% noise level, the detection resolution for RIU is up to 5.2 × 10-6, which is comparable to conventional prism-based SPR system. The fiber tip sensor provides the advantages of small detection volume, fast and label-free detection. The measurement setup is also simple and non-expensive. By coating a thin Ti/Al film on the fiber tip, a large DEP force is produced near the tip region. With the modification of specific antibody on tip surface, the sensor can be used for detecting specific bacteria in water. Currently, the detectable range for bacteria is about 102 to 106 CFU/mL. Substantial reduction of bacteria immobilization would happen when samples contain large particles. A suitable sample pretreatment such as micro filtration using membranes with a pore size larger than the bacterial size can help reduce the substrate interference. With the pretreatment, this simple and cheap detection method could be applied for fast evaluation of bacteria in drinks and lake waters.

AUTHOR INFORMATION Corresponding Author * E-mail: [email protected] (886) 2-27873146.

Notes The authors declare no competing financial interest.

ACKNOWLEDGMENT The authors would like to express their sincere appreciation to laboratory members of Dr. Chau-Hwang Lee and Dr. JiYen Cheng in the Research Center for Applied Sciences of the Academia Sinica for helpful discussions on biological studies and the cultivated samples. This work was supported by Ministry of Science and Technology, Taiwan, R.O.C. (Contract No. 105-2627-B-001 -001 -) and the Thematic Project of Academia Sinica, Taiwan, R.O.C.

7 ACS Paragon Plus Environment

ACS Sensors 1 2 3 4 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 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

REFERENCES (1). Sadeghi, J.; Patabadige, D. E. W.; Culbertson, A. H.; Latifi, H.; Culbertson, C. T., Out-of-plane integration of a multimode optical fiber for single particle/cell detection at multiple points on a microfluidic device with applications to particle/cell counting, velocimetry, size discrimination and the analysis of single cell lysate injections. Lab on a Chip 2017, 17 (1), 145-155. (2). Sadeghi, J.; Ghasemi, A. H.; Latifi, H., A label-free infrared opto-fluidic method for real-time determination of flow rate and concentration with temperature cross-sensitivity compensation. Lab Chip 2016, 16 (20), 3957-3968. (3). Patabadige, D. E.; Sadeghi, J.; Kalubowilage, M.; Bossmann, S. H.; Culbertson, A. H.; Latifi, H.; Culbertson, C. T., Integrating Optical Fiber Bridges in Microfluidic Devices to Create Multiple Excitation/Detection Points for Single Cell Analysis. Anal Chem 2016, 88 (20), 9920-9925. (4). Cheri, M. S.; Shahraki, H.; Sadeghi, J.; Moghaddam, M. S.; Latifi, H., Measurement and control of pressure driven flows in microfluidic devices using an optofluidic flow sensor. Biomicrofluidics 2014, 8 (5). (5). Cheri, M. S.; Latifi, H.; Sadeghi, J.; Moghaddam, M. S.; Shahraki, H.; Hajghassem, H., Real-time measurement of flow rate in microfluidic devices using a cantilever-based optofluidic sensor. Analyst 2014, 139 (2), 431-8. (6) Dandapat, K.; Tripathi, S. M.; Chinifooroshan, Y.; Bock, W. J.; Mikulic, P., Compact and cost-effective temperatureinsensitive bio-sensor based on long-period fiber gratings for accurate detection of E. coli bacteria in water. Opt Lett 2016, 41 (18), 4198-201 (7). Kim, D.; Herr, A. E., Protein immobilization techniques for microfluidic assays. Biomicrofluidics 2013, 7 (4), 47. (8). Ricciardi, A.; Crescitelli, A.; Vaiano, P.; Quero, G.; Consales, M.; Pisco, M.; Esposito, E.; Cusano, A., Lab-on-fiber technology: a new vision for chemical and biological sensing. Analyst 2015, 140 (24), 8068-79. (9). Guo, L.; Jackman, J. A.; Yang, H.-H.; Chen, P.; Cho, N.-J.; Kim, D.-H., Strategies for enhancing the sensitivity of plasmonic nanosensors. Nano Today 2015, 10 (2), 213-239. (10). Quesada-Gonzalez, D.; Merkoci, A., Nanoparticle-based lateral flow biosensors. Biosens Bioelectron 2015, 73, 47-63. (11). Shukla, S.; Sharma, N. K.; Sajal, V., Sensitivity enhancement of a surface plasmon resonance based fiber optic sensor using ZnO thin film: a theoretical study. Sensors and Actuators B: Chemical 2015, 206, 463-470. (12). Hu, X.; Pun, C. F.; Tam, H. Y.; Megret, P.; Caucheteur, C., Highly reflective Bragg gratings in slightly etched step-index polymer optical fiber. Opt Express 2014, 22 (15), 18807-17. (13). Urrutia, A.; Goicoechea, J.; Arregui, F. J., Optical Fiber Sensors Based on Nanoparticle-Embedded Coatings. Journal of Sensors 2015, 2015, 1-18.

Page 8 of 10

(14). Sun, X. C.; Lei, C.; Guo, L.; Zhou, Y., Separable detecting of Escherichia coli O157H:H7 by a giant magneto-resistance-based bio-sensing system. Sens. Actuator B-Chem. 2016, 234, 485-492. (15). Yang, Z.; Sun, X. C.; Wang, T.; Lei, C.; Liu, Y.; Zhou, Y.; Lei, J., A giant magnetoimpedance-based biosensor for sensitive detection of Escherichia coli O157:H7. Biomedical Microdevices 2015, 17 (1).: 5. doi: 10.1007/s10544-014-9925-6. (16). Lu, L.; Jun, S., Evaluation of a microwire sensor functionalized to detect Escherichia coli bacterial cells. Biosensors & Bioelectronics 2012, 36 (1), 257-261. (17). Sharma, A. K.; Jha, R.; Gupta, B. D., Fiber-optic sensors based on surface plasmon resonance: A comprehensive review. Ieee Sensors Journal 2007, 7 (7-8), 1118-1129. (18). Caucheteur, C.; Guo, T.; Albert, J., Review of plasmonic fiber optic biochemical sensors: improving the limit of detection. Anal. Bioanal. Chem. 2015, 407 (14), 3883-3897. (19). Baldini, F.; Brenci, M.; Chiavaioli, F.; Giannetti, A.; Trono, C., Optical fibre gratings as tools for chemical and biochemical sensing. Anal. Bioanal. Chem. 2012, 402 (1), 109-116. (20). Tai, Y. H.; Wei, P. K., Sensitive liquid refractive index sensors using tapered optical fiber tips. Optics Letters 2010, 35 (7), 944-946. (21). Tai, Y. H.; Pan, M. Y.; Lin, E. H.; Huang, D. W.; Wei, P. K., Quality Detection of Alcoholic Beverages Using Optical Fiber Tips. Ieee Sensors Journal 2016, 16 (14), 5626-5631. (22). Wei, P. K.; Chen, Y. C.; Kuo, H. L., Systematic variation of polymer jacketed fibres and the effects on tip etching dynamics. Journal of Microscopy-Oxford 2003, 210, 334-339. (23). Pethig, R., Review Article-Dielectrophoresis: Status of the theory, technology, and applications. Biomicrofluidics 2010, 4 (2), 35. (24). Voldman, J., Electrical forces for microscale cell manipulation. In Annual Review of Biomedical Engineering, Annual Reviews: Palo Alto, 2006; Vol. 8, pp 425-454. (25). Chai, C.; Lee, J.; Park, J.; Takhistov, P., Antibody immobilization on a nanoporous aluminum surface for immunosensor development. Appl. Surf. Sci. 2012, 263, 195-201. (26). Sang, C. H.; Chou, S. J.; Pan, F. M.; Sheu, J. T., Fluorescence enhancement and multiple protein detection in ZnO nanostructure microfluidic devices. Biosensors & Bioelectronics 2016, 75, 285-292. (27). Liu, H. H.; Lin, T. H.; Sheu, J. T., Enhancement of detection by selective modification of silicon nanobelt fieldeffect transistors via localized Joule heating. Sens. Actuator BChem. 2014, 192, 111-116. (28). Negishi, A.; Chen, J. H.; McCarty, D. M.; Samulski, R. J.; Liu, J.; Superfine, R., Analysis of the interaction between adenoassociated virus and heparan sulfate using atomic force microscopy. Glycobiology 2004, 14 (11), 969-977.

8 ACS Paragon Plus Environment

Page 9 of 10 1 2 3 4 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 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

(29). Takmakov, P.; Vlassiouk, I.; Smirnov, S., Application of anodized aluminum in fluorescence detection of biological species. Anal. Bioanal. Chem. 2006, 385 (5), 954-958. (30). Berquand, A.; Mazeran, P. E.; Pantigny, J.; ProuxDelrouyre, V.; Laval, J. M.; Bourdillon, C., Two-step formation of streptavidin-supported lipid bilayers by PEG-triggered vesicle fusion. Fluorescence and atomic force microscopy characterization. Langmuir 2003, 19 (5), 1700-1707. (31). Abel, A. P.; Weller, M. G.; Duveneck, G. L.; Ehrat, M.; Widmer, H. M., Fiber-optic evanescent wave biosensor for the detection of oligonucleotides. Anal. Chem. 1996, 68 (17), 29052912. (32). Wang, Y. X.; Ping, J. F.; Ye, Z. Z.; Wu, J.; Ying, Y. B., Impedimetric immunosensor based on gold nanoparticles modified graphene paper for label-free detection of Escherichia coli O157:H7. Biosensors & Bioelectronics 2013, 49, 492-498. (33). Huang, T. T.; Sturgis, J.; Gomez, R.; Geng, T.; Bashir, R.; Bhunia, A. K.; Robinson, J. P.; Ladisch, M. R., Composite surface for blocking bacterial adsorption on protein biochips. Biotechnol. Bioeng. 2003, 81 (5), 618-624. (34). Homola, J.; Yee, S. S.; Gauglitz, G., Surface plasmon resonance sensors: review. Sens. Actuator B-Chem. 1999, 54 (1-2), 3-15. (35). Lewpiriyawong, N.; Kandaswamy, K.; Yang, C.; Ivanov, V.; Stocker, R., Microfluidic characterization and continuous separation of cells and particles using conducting poly(dimethyl siloxane) electrode induced alternating currentdielectrophoresis. Anal Chem 2011, 83 (24), 9579-85. (36). Tai, Y. H.; Chang, D. M.; Pan, M. Y.; Huang, D. W.; Wei, P. K., Sensitive Detection of Small Particles in Fluids Using Optical Fiber Tip with Dielectrophoresis. Sensors 2016, 16 (3), 9.

9 ACS Paragon Plus Environment

ACS Sensors 1 2 3 4 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 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 10

For TOC only

10 ACS Paragon Plus Environment