Plasmonics

, Volume 6, Issue 1, pp 29–34

Enhanced Sensitivity of Surface Plasmon Resonance Phase-Interrogation Biosensor by Using Silver Nanoparticles

Authors

  • Tzu-Cheih Peng
    • Institute of Optoelectronic SciencesNational Taiwan Ocean University
  • Wen-Chi Lin
    • Institute of Optoelectronic SciencesNational Taiwan Ocean University
  • Chih-Wei Chen
    • Institute of Optoelectronic SciencesNational Taiwan Ocean University
  • Din Ping Tsai
    • Department of PhysicsNational Taiwan University
    • Instrument Technology Research CenterNational Applied Research Laboratories
    • Institute of Optoelectronic SciencesNational Taiwan Ocean University
    • Institute of PhysicsAcademia Sinica
    • Instrument Technology Research CenterNational Applied Research Laboratories
Article

DOI: 10.1007/s11468-010-9165-4

Cite this article as:
Peng, T., Lin, W., Chen, C. et al. Plasmonics (2011) 6: 29. doi:10.1007/s11468-010-9165-4

Abstract

It is demonstrated that the sensitivity of surface plasmon resonance phase-interrogation biosensor can be enhanced by using silver nanoparticles. Silver nanoparticles were fabricated on silver films by using thermal evaporation. Sizes of silver nanoparticles on silver thin film can be tuned by controlling the deposition parameters of thermal evaporation. By using surface plasmon resonance heterodyne interferometey to measure the phase difference between the p and s polarization of incident light, we have demonstrated that sensitivity of glucose detection down to the order of 10−8 refractive index units can be obtained.

Keywords

Surface plasmonsPhase measurementSilver nanoparticlesBiosensor

Introduction

Since the accomplishment of various optical methods in the excitation of the surface plasmon resonance (SPR) at a metal–dielectric interface [1], it has been well known that such an excitation can be utilized to achieve sensing of various interfacial phenomena with ultrahigh sensitivity. These include, for example, chemical and biological sensing [29], film-thickness sensing [10], temperature sensing [11], and angular measurement [12]. Recently, it has been demonstrated that this SPR monitoring technique for chemical and biological sensing can achieve very high sensitivity, down to the order of 10−8 [8, 9] refractive index units (RIU).

For the sensing application by using SPR, it is straightforward to adopt the attenuated total reflection method by coupling incident light to excite surface plasmon resonance at the metal-dielectric interface, leading to a dip in the reflection spectrum which can be monitored to follow any property changes that take place in the proximity [1]. There are at least four kinds of parameters that can be monitored in the SPR sensing process: (a) the change of the resonant angle, (b) the change of reflectance at fixed incident angle, (c) the change of resonant wavelength at fixed incident angle, and (d) the phase difference between p- and s-polarized light in the reflection spectrum [2]. The detection methods that correspond to these four kinds of parameters are angular interrogation, intensity interrogation, wavelength interrogation, and phase interrogation, respectively.

Among these various monitoring techniques, it is well known that the “phase interrogation” technique is by far the more sensitive one in many applications [3, 59, 11, 12]. Recently, we have found that the sensitivity of this technique has strong dependence on the wavelength of incident light in our study of SPR temperature sensing [11]. In other experiment, we have also observed that high-resolution angular measurement and biological sensing can be achieved by SPR phase interrogation at optimized incident wavelength [9, 12]. However, the value of the optimal wavelength will change when the film thickness varies. It is thus possible to reach optimal sensitivity by tuning the thickness of metallic film at fixed wavelength of incident light. Furthermore, modification of the surface roughness of the metallic film will also change the sensitivity of SPR sensor [1, 13, 14].

It was theoretically reported in the literature that metallic nanoparticles dispersed over metallic film could enhance the sensitivity of SPR sensor [13]. This theoretical prediction has been experimentally demonstrated by immobilizing colloidal Au nanoparticles onto a thermally evaporated Au film. It was found that large changes in SPR reflectivity could be reached using this wet-chemistry approach [14]. However, it is time consuming in the film production process where colloidal Au nanoparticles had to be first prepared via citrate reduction of HAuCl4 [14]. If one can prepare the nanoparticle film in only one deposition process, this kind of SPR sensing chip will be more promising in practical applications.

Gold and silver island films on glass substrate were recently reported to be reproducible by precise control of the deposition parameters of thermal evaporation with tunable SPR wavelengths [15]. Specific combinations of substrate temperature, deposition rate, and film thickness could produce island films with SPR wavelength tuning from visible to infrared. Based on this technique, one can employ thermal evaporation to produce metallic nano-thin film on the glass substrate first, and then control the deposition parameters to evaporate metallic nanoparticles over metallic film without opening the vacuum chamber. The size of nanoparticles can be controlled by deposition parameters, and therefore, the SPR wavelength of nanoparticle film could be tuned. As mentioned above, the wavelength of incident light is crucial to the sensitivity of SPR sensor based on phase-interrogation technique [9, 11, 12]. It is the purpose of our present work to extend SPR phase interrogation to refractive index measurement of glucose based on silver nanoparticle film produced by thermal evaporation. By controlling the deposition parameters of manufacture process, the size of silver nanoparticle can be changed, and, therefore, it is possible to achieve optimal sensitivity at fixed wavelength of incident light.

Experiment

SPR sensing chips based on silver nanoparticles over silver thin film were fabricated by using thermal evaporation. The main parameters of the thermal evaporation include the substrate temperature (Ts), deposition rate (Rd), and deposition thickness of the thin film (Tf). Tf is determined from the quartz oscillator of thermal evaporator. A thin silver film was first evaporated on the glass substrate, and metallic nanoparticles were then evaporated over metallic film by controlling the deposition parameters. The thermal evaporation parameters for silver nanoparticles are Ts from 50 to 200 °C, Rd from 0.3 to 1.2 Å/s, and Tf from 10 to 100 Å, respectively. SPR chips were fabricated under different combination of these manufacture parameters. The sensitivity of SPR phase detection for glucose sensing based on these silver nanoparticle films were compared, and optimal fabrication condition was determined.

The substrates (Matsunami cover glass, 22 mm × 22 mm, 0.12–0.17 mm thickness) were first cleaned using acetone and methanol in an ultrasonic bath for 30 min. Following this, the glass substrate was put into the evaporation chamber and air was pumped out of the evaporation chamber until 5 × 10−6 Torr of air pressure was achieved. The glass substrate was rotated during the evaporation process, and the deposition rate was set at 0.4 Å/s and the deposition thickness at 50 nm. After depositing a silver thin film on the glass substrate, silver nanoparticles were then deposited. By heating the glass substrate to a specific temperature and varying the deposition thickness and deposition rate, the size of the silver nanoparticles over silver thin film was controlled. Based on the above deposition process, different sizes of silver nanoparticles over silver thin film could be fabricated by varying the deposition parameters of thermal evaporation (see Table 1).
Table 1

Fabrication parameters of silver nanoparticle films and detected SPR phase sensitivity by using these films

Sample

1

2

3

4

5

6

7

8

9

10

Substrate temperature (°C)

50

75

100

125

150

100

100

100

100

100

Deposition thickness of nanoparticles layer (nm)

10

10

10

10

10

1

3

5

7

10

Average size of nanoparticles (nm)

58

58

44

46

47

52

40

43

44

Size deviation of nanoparticles (nm)

15.7

15.7

16.4

15

16.8

14.6

17.3

16.2

16.4

Sensitivity of SPR phase interrogation (RIU)

1.4 × 10−6

1.1 × 10−6

4.6 × 10−7

5.1 × 10−7

5.8 × 10−7

8.2 × 10−7

5.3 × 10−7

6.3 × 10−8

3.8 × 10−7

4.6 × 10−7

The morphology of the nanostructures was analyzed using atomic force microscope (AFM, Nanosurf Mobile S). All of the AFM images were collected in the contact mode with an applied force of 18 nN. The nanoprobe tips were made of silicon with 10 nm in diameter, and the scanning range was 3 μm × 3 μm.

The experimental setup was identical to that used in Ref. [9]. Details are recaptured and summarized in the following. The light from the source was introduced through a polarizer into an electro-optic modulator (ConOptics) with fast axis in the horizontal direction, and then reflected from a 45–45–90° triangular SF11 prism (Casix) with refractive index \( n \simeq 1.78 \) at visible wavelengths. The prism was in contact with a glass slide coated with silver nanoparticles over thin silver film. To avoid refractive index mismatch, index-matching oil was added between the prism and the glass slide. The samples studied in this work were glucose solutions of different concentrations and were introduced into the flow cell that was clamped onto the prism, as shown in Fig. 1. The laser light was then passed through an analyzer ANt. Both the transmission axis of the polarizer and analyzer were at 45° relative to the horizontal direction. The light was then detected by a Si photodetector. The converted electric signal from the photodetector was phase-locked by a lock-in amplifier (Stanford Research SR830). A sawtooth signal was applied to the electro-optic modulator, and the sawtooth voltage was used as the phase-locking reference. By measuring the signal intensity and by assuming almost unity reflectivity for the s wave (since only p wave can excite SPR at a metal-dielectric surface), the relative phase difference between the p and the s waves arriving at the photodetector can finally be calculated using Jones calculus [9, 11, 12].
https://static-content.springer.com/image/art%3A10.1007%2Fs11468-010-9165-4/MediaObjects/11468_2010_9165_Fig1_HTML.gif
Fig. 1

Schematic picture of silver nanoparticles enhancing the SPR biosensor

Results and Discussion

In the fabrication process of SPR sensing chip, we first used thermal evaporation to fabricate a 50-nm-thick silver thin film with substrate temperature of 100 °C and evaporation rate of 0.4 Å/S. Following this, we fabricated silver nanoparticles on the silver thin film with thermal evaporation rate of 0.4 Å/S, film thickness of 10 nm, and different temperature setting for different samples, including 50, 75, 100, 125, and 150°C. These samples were labeled as sample 1, sample 2, sample 3, sample 4, and sample 5, respectively. We then fixed substrate temperature to 100°C and tried to find the optimal parameter in film thickness by varying film-thickness parameter for different sample. The film thickness were 1, 3, 5, 7, and 10 nm and were labeled as sample 6, sample 7, sample 8, sample 9, and sample 10, respectively.

After thermal evaporation fabrication process, the topography of silver nanoparticle films was scanned by AFM. Figure 2 shows the (a) 2D and (b) 3D AFM images of sample 8. From these images, the silver nanoparticles were needle like and distributed randomly on the silver surface. Figure 3 shows the size distribution of the silver nanoparticles in Fig. 2a analyzed by using Image J software. The diameters of the silver nanoparticle for samples 1~5 were found to be 58, 58, 44, 46, and 47 nm, respectively. From these five data shown in Table 1, we conclude that the average size was smallest in sample 3 when the substrate temperature was 100°C. We therefore fixed the substrate temperature at 100°C and varied the film thickness in the fabrication process of samples 6 to 10. Because the deposition thickness for sample 6 was set at 1 nm, no apparent nanoparticles could be observed by AFM. Therefore, it was not possible to analyze the size distribution for the nanoparticles. The average size of silver nanoparticle for samples 7~10 were found to be 52, 40, 43, and 44 nm, respectively. These experimental results showed that the deposition parameters of sample 8 would allow us to fabricate the smallest size of silver nanoparticles, as summarized in Table 1.
https://static-content.springer.com/image/art%3A10.1007%2Fs11468-010-9165-4/MediaObjects/11468_2010_9165_Fig2_HTML.gif
Fig. 2

a 2D and b 3D AFM images of sample 8

https://static-content.springer.com/image/art%3A10.1007%2Fs11468-010-9165-4/MediaObjects/11468_2010_9165_Fig3_HTML.gif
Fig. 3

Size distribution of the silver nanoparticles in Fig. 1 analyzed by using Image J software

Figure 4 shows the experimental results for the relative phase difference between the p and the s waves as a function of incident angle for various glucose solutions, measured at the wavelength of 632.8 nm for incident light. The SPR sensing chips employed in Fig. 4 were bare Ag film, sample 7, sample 8, and sample 9 for (a), (b), (c), and (d), respectively. The Ag film thickness in Fig. 4a is fixed at 50 nm, and the minimum increment of incident angle is fixed at 0.01° for the experiment. A series of solutions made of glucose powder (Sigma Inc.) dissolved in deionized water with concentrations ranging from 0% to 5% by weight was introduced into the flow cell. In Fig. 4a, b, the phases near-resonance angle changes from higher to lower values, while the reverse takes place in Fig. 4c, d. This “cross over behavior” has been observed and explained in the literature [9, 11, 12].
https://static-content.springer.com/image/art%3A10.1007%2Fs11468-010-9165-4/MediaObjects/11468_2010_9165_Fig4_HTML.gif
Fig. 4

Experimental results for the relative phase difference between the p and the s waves as a function of incident angle for various glucose solutions, measured at the wavelength of 632.8 nm for incident light. The SPR sensing chips employed were bare Ag film, sample 7, sample 8, and sample 9 for a, b, c, and d, respectively

In order to compare the smallest refractive index resolvable among various SPR sensing chips in phase measurement (σn), we shall follow the definition of Nelson et al. [3] to consider the following quantity: \( {\sigma_n} = \frac{{\Delta n}}{{\Delta \varphi }}{\sigma_\varphi } \), where \( \frac{{\Delta n}}{{\Delta \varphi }} \)is the local slope of the curve for refractive index n vs. phase ϕ, and \( {\sigma_\varphi } \) is the finest resolution available with a value of 0.01°. We calculated the refractive index of different glucose concentrations by using Fresnel equations. The refractive indices of glucose solutions calculated from experimental results in Fig. 4a are 1.3333, 1.3347, 1.3363, 1.3378, 1.3392, and 1.3407 for weight concentrations from 0% to 5%, respectively. By following the method, we used in analyzing the sensitivity of glucose detection in Ref. [9], the smallest refractive indices resolvable σn were 6.4 × 10−7, 5.5 × 10−7, 6.3 × 10−8, and 3.8 × 10−7 for each of the figures (Fig. 4a–d, as shown in Table 1). From the experimental results, it is observed that the higher sensitivity can be achieved with smaller silver nanoparticles and the sensitivity is increased by one order of magnitude compared with the results of bare Ag film with that of sample 8. Figure 5 shows the relation between the sensitivity of SPR phase detection and average size of silver nanoparticles, with fixed thermal evaporation temperature of 100°C and deposition rate of 0.4 Å/S as shown in Table 1. It is obvious that sensitivity is higher when the average size of silver nanoparticles is smaller and sensitivity as small as 6.3 × 10−8 can be achieved for the fabrication parameters of sample 8 when the fabrication parameter for deposition thickness is at 5 nm corresponding to average nanoparticle size of 40 nm. Since the silver nanoparticles are needle like as observed from the 3D AFM images shown in Fig. 2, the interaction area within the laser spot is increased as the aspect ratio of needle-like silver nanoparticles increases. In other words, when the size of nanoparticles is getting smaller, it is much easier to detect the local change of reflective index and thus the sensitivity of SPR biosensing is enhanced.
https://static-content.springer.com/image/art%3A10.1007%2Fs11468-010-9165-4/MediaObjects/11468_2010_9165_Fig5_HTML.gif
Fig. 5

Relation between the sensitivity of SPR phase detection vs. average size of silver nanoparticles, with fixed thermal evaporation temperature 100°C and deposition rate 0.4 Å/S as shown in Table 1

Conclusion

We have successfully demonstrated that SPR sensing chip based on silver nanoparticles over silver thin films could be fabricated by controlling the fabrication parameters of thermal evaporation; and the detection sensitivity of SPR phase interrogation for glucose detection could be enhanced by using this kind of sensing chip. With substrate temperature 100°C, deposition rate 0.4 Å/S, and film thickness 5 nm, we could fabricate silver nanoparticle film with the size of nanoparticles about 40 nm in diameter over Ag film of 50 nm in thickness. We have also demonstrated that the sensitivity of SPR phase interrogation for glucose detection based on this silver nanoparticle film could be as low as 6.3 × 10−8 at the laser wavelength of 632.8 nm. Unlike the traditional approach for enhanced SPR sensing using nanoparticles, which is time consuming in the film production process, where colloidal Au nanoparticles were first prepared via citrate reduction of HAuCl4 [14], our silver nanoparticle film could be prepared in only one deposition process. We believe that this kind of SPR sensing chip combined with SPR phase-interrogation technique will provide more promising applications and simpler protocols.

Acknowledgments

H.-P. Chiang acknowledges the financial support from the Center for Marine Bioenvironment and Biotechnology, National Taiwan Ocean University, and the National Science Council of ROC under grant number NSC 97-2112-M-019-001-MY3.

Copyright information

© Springer Science+Business Media, LLC 2010