Double-Antibody Sandwich Immunoassay and Plasmonic Coupling Synergistically Improved Long-Range SPR Biosensor with Low Detection Limit

A long-range surface plasmonic resonance (LR-SPR) biosensor modified with double-antibody sandwich immunoassay and plasmonic coupling is demonstrated for human-immunoglobulin G detection with a low limit of detection (LOD). The double-antibody sandwich immunoassay dramatically changes the average refractive index of the medium layer on the sensor surface. The near-field electron coupling between the localized surface plasmon and the long-range surface plasmon leads to a significant perturbation of the evanescent field. The large penetration depth and the long propagation distance of the long-range surface plasmonic waves facilitate the LR-SPR sensor in the detection of biological macromolecules. The unique light absorption characteristic of the nanocomposite material in the sensor provides the in situ self-compensation for the disturbance. Therefore, besides the inherent advantages of optical fiber sensors, the developed biosensor can realize the detection of biomolecules with high sensitivity, low LOD and high accuracy and reliability. Experimental results demonstrate that the LOD of the biosensor is as low as 0.11 μg/mL in the detection of the phosphate-buffered saline sample, and the spike-and-repetition rate is 105.56% in the detection of the real serum sample, which partly shows the practicability of the biosensor. This indicates that the LR-SPR biosensor provides better response compared with existing similar sensors and can be regarded as a valuable method for biochemical analysis and disease detection.


Introduction
Surface plasmonic resonance (SPR) biosensors based on optical fiber have been widely studied in immunoassay, analytical chemistry and disease examination due to their outstanding properties of compact size, anti-electromagnetic interference, high sensitivity and label-free detection. However, traditional fiber based SPR sensors have limited ability in further reducing the limit of detection (LOD) because the relatively shallow penetration of the evanescent field prevents the effective detection of the refractive index variation of biomolecules. Furthermore, the large full width at half-maximum (FWHM) and the low figure of merit (FOM, the ratio between the sensitivity of the sensor and the FWHM) in the fiber SPR-sensing spectrum have seriously restricted the resolution and accuracy of the detection. Therefore, methods for improving the performance of SPR sensors have

Sensing Principle and Simulation Analysis
As shown in Figure 1a, the essential film structure of the LR-SPR sensor includes SPF, LML, metallic layer and target analyte. Since the metallic layer is sandwiched by two dielectric layers with similar refractive indices, SPWs are generated on both sides of the metallic layer. According to the coupling mode theory [17], the energy of transmission light is coupled into two rows of SPWs. The two SPWs will interfere with each other. The constructive interference forms long-range SPWs (LR-SPWs) and the destructive interference forms short-range SPWs (SR-SPWs). These effects excite LR-SPR and shortrange SPR [18][19][20], respectively. Therefore, the transmission spectrum of the LR-SPR sensor contains a long-range resonance curve and a short-range resonance curve, as shown in Figure 1b. Since the field distribution of the LR-SPWs [21] shown in Figure 1b is mainly concentrated in the dielectric layers, which indicates a weaker restriction on the vertical axis and a lower loss on the parallel axis, the penetration depth of LR-SPWs in the analyte is larger (in the order of micrometers) and the propagation distance of LR-SPWs on the interface between the metallic layer and the medium layer is longer. Therefore, the long-range resonance curve has narrower FWHM and higher FOM, and also has more significant advantages in detecting biological macromolecules compared with traditional SPR sensors [18,19]. The energy of SR-SPWs is mainly focused on the metallic layer surface, and thus it can be used for the monoatomic layer detection [22]. The short-range resonance curve may disappear due to the optimized design of the film thickness.
Nanomaterials 2021, 11,2137 3 of 12 in size, low in cost, easy to form the distributed sensing and can realize remote real−time online detection for H−IgG with low LOD and high accuracy and reliability.

Sensing Principle and Simulation Analysis
As shown in Figure 1a, the essential film structure of the LR−SPR sensor includes SPF, LML, metallic layer and target analyte. Since the metallic layer is sandwiched by two dielectric layers with similar refractive indices, SPWs are generated on both sides of the metallic layer. According to the coupling mode theory [17], the energy of transmission light is coupled into two rows of SPWs. The two SPWs will interfere with each other. The constructive interference forms long−range SPWs (LR−SPWs) and the destructive interference forms short−range SPWs (SR−SPWs). These effects excite LR−SPR and short−range SPR [18][19][20], respectively. Therefore, the transmission spectrum of the LR−SPR sensor contains a long−range resonance curve and a short−range resonance curve, as shown in Figure  1b. Since the field distribution of the LR−SPWs [21] shown in Figure 1b is mainly concentrated in the dielectric layers, which indicates a weaker restriction on the vertical axis and a lower loss on the parallel axis, the penetration depth of LR−SPWs in the analyte is larger (in the order of micrometers) and the propagation distance of LR−SPWs on the interface between the metallic layer and the medium layer is longer. Therefore, the long−range resonance curve has narrower FWHM and higher FOM, and also has more significant advantages in detecting biological macromolecules compared with traditional SPR sensors [18,19]. The energy of SR−SPWs is mainly focused on the metallic layer surface, and thus it can be used for the monoatomic layer detection [22]. The short−range resonance curve may disappear due to the optimized design of the film thickness.  (I: characteristic absorption line, II: long−range resonance curve, III: short−range resonance curve). Inset: the field/current distributions of three modes with the lowest loss in the insulator−metal−insulator (IMI) model (IV-VI: three modes, red line: electric field distribution, black line: magnetic field distribution, orange arrow: current conduction). Reproduced with permission from reference [21] by John Wiley and Sons.
Modifying the nanometal to the gold layer surface in the sensor can further enhance the sensitivity, as shown in Figure 1a. The finite element analysis shown in Figure 2 suggests that the electric field intensities at the top and the bottom of the gold nanospheres in the gold nanosphere−modified LR−SPR sensor are, respectively, 1.63 times (81.09-132.28 V/m) and 8.66 times (81.09-701.85 V/m) higher than those of the bare LR−SPR sensor surface, respectively. This is because the near−field electron coupling between the localized surface plasmon, originating from the nanometal, and the long−range surface plasmon, arising from the metallic layer, leads to a stronger evanescent field perturbation [6]. The sensitivity is related to the overlap integral of the electrical intensity in the analyte region [23], thus the gold nanosphere−modified LR−SPR sensor possesses higher sensitivity. The specific antigen can be detected by modifying the capture antibody (CAb) on the nonmetal 300 400 500 600 700 800 900 10001100   (I: characteristic absorption line, II: long-range resonance curve, III: short-range resonance curve). Inset: the field/current distributions of three modes with the lowest loss in the insulator-metal-insulator (IMI) model (IV-VI: three modes, red line: electric field distribution, black line: magnetic field distribution, orange arrow: current conduction). Reproduced with permission from reference [21] by John Wiley and Sons.
Modifying the nanometal to the gold layer surface in the sensor can further enhance the sensitivity, as shown in Figure 1a. The finite element analysis shown in Figure 2 suggests that the electric field intensities at the top and the bottom of the gold nanospheres in the gold nanosphere-modified LR-SPR sensor are, respectively, 1.63 times (81.09-132.28 V/m) and 8.66 times (81.09-701.85 V/m) higher than those of the bare LR-SPR sensor surface, respectively. This is because the near-field electron coupling between the localized surface plasmon, originating from the nanometal, and the long-range surface plasmon, arising from the metallic layer, leads to a stronger evanescent field perturbation [6]. The sensitivity is related to the overlap integral of the electrical intensity in the analyte region [23], thus the gold nanosphere-modified LR-SPR sensor possesses higher sensitivity. The specific antigen can be detected by modifying the capture antibody (CAb) on the nonmetal surface. The parameter that the sensor can detect is the average refractive index of all substances on its surface. In order to enhance the detection sensitivity, detection antibody (DAb) can be added after the antigen, to build the double-antibody sandwich immunoassay (i.e., CAb-antigen-DAb structure, shown in Figure 1a) [16,24]. This can increase the change of the refractive index in the medium layer on the sensor surface. surface. The parameter that the sensor can detect is the average refractive index of all substances on its surface. In order to enhance the detection sensitivity, detection antibody (DAb) can be added after the antigen, to build the double−antibody sandwich immunoassay (i.e., CAb−antigen−DAb structure, shown in Figure 1a) [16,24]. This can increase the change of the refractive index in the medium layer on the sensor surface. The LML used in this paper can absorb the light field near a wavelength of 350 nm. This produces a characteristic absorption line in the transmission spectrum of the sensor, as shown in Figure 1b. When the sensor is disturbed, the whole transmission spectrum could shift, but the wavelength difference between the resonance curve and the absorption line remains unchanged. Detection results can be obtained based on the stable wavelength difference rather than the single resonance curve. Therefore, the sensor possesses the in situ self−compensation function which allows corrections for errors of spectral drifts induced by the mechanical vibration and other disturbances [25]. Polydopamine (PDA), as a polymer material, can realize the controllable self−polymerization and possesses excellent biocompatibility and hydrophilicity. The strong adhesion (similar to the mussel mucus) and the strong hydrolysis resistance of the PDA make it more suitable as a biosensing matrix than the 11−mercaptoundecanoic acid. Therefore, PDA has been used as an antibody coupling agent, and the specific principle is shown in Figure 3. The ortho−diphenol functional group of dopamine will be converted into quinone group under the weakly basic conditions. Quinones can be covalently coupled with amino− or thiol−terminated biomolecules by Schiff base and Michael addition reaction [16].

Materials and Reagents
The multimode fiber with core/cladding diameters of 105/125 μm and a numerical aperture of 0. The LML used in this paper can absorb the light field near a wavelength of 350 nm. This produces a characteristic absorption line in the transmission spectrum of the sensor, as shown in Figure 1b. When the sensor is disturbed, the whole transmission spectrum could shift, but the wavelength difference between the resonance curve and the absorption line remains unchanged. Detection results can be obtained based on the stable wavelength difference rather than the single resonance curve. Therefore, the sensor possesses the in situ self-compensation function which allows corrections for errors of spectral drifts induced by the mechanical vibration and other disturbances [25]. Polydopamine (PDA), as a polymer material, can realize the controllable self-polymerization and possesses excellent biocompatibility and hydrophilicity. The strong adhesion (similar to the mussel mucus) and the strong hydrolysis resistance of the PDA make it more suitable as a biosensing matrix than the 11-mercaptoundecanoic acid. Therefore, PDA has been used as an antibody coupling agent, and the specific principle is shown in Figure 3. The ortho-diphenol functional group of dopamine will be converted into quinone group under the weakly basic conditions. Quinones can be covalently coupled with amino-or thiol-terminated biomolecules by Schiff base and Michael addition reaction [16]. surface. The parameter that the sensor can detect is the average refractive index of all substances on its surface. In order to enhance the detection sensitivity, detection antibody (DAb) can be added after the antigen, to build the double−antibody sandwich immunoassay (i.e., CAb−antigen−DAb structure, shown in Figure 1a) [16,24]. This can increase the change of the refractive index in the medium layer on the sensor surface. The LML used in this paper can absorb the light field near a wavelength of 350 nm. This produces a characteristic absorption line in the transmission spectrum of the sensor, as shown in Figure 1b. When the sensor is disturbed, the whole transmission spectrum could shift, but the wavelength difference between the resonance curve and the absorption line remains unchanged. Detection results can be obtained based on the stable wavelength difference rather than the single resonance curve. Therefore, the sensor possesses the in situ self−compensation function which allows corrections for errors of spectral drifts induced by the mechanical vibration and other disturbances [25]. Polydopamine (PDA), as a polymer material, can realize the controllable self−polymerization and possesses excellent biocompatibility and hydrophilicity. The strong adhesion (similar to the mussel mucus) and the strong hydrolysis resistance of the PDA make it more suitable as a biosensing matrix than the 11−mercaptoundecanoic acid. Therefore, PDA has been used as an antibody coupling agent, and the specific principle is shown in Figure 3. The ortho−diphenol functional group of dopamine will be converted into quinone group under the weakly basic conditions. Quinones can be covalently coupled with amino− or thiol−terminated biomolecules by Schiff base and Michael addition reaction [16].

Materials and Reagents
The multimode fiber with core/cladding diameters of 105/125 μm and a numerical aperture of 0.

Materials and Reagents
The multimode fiber with core/cladding diameters of 105/125 µm and a numerical aperture of 0.

Manufacturing of Sensing Probe
The manufacturing process of the sensing probe is described as follows. 1.18 g of Tb (III) acetate is firstly dissolved in 20 mL of deionized water at 0 • C. Then, 1.50 mL of hexafluoroacetylacetone is drop wise added to the above solution. The mixture will produce white-green precipitations after stirring for 180 min. Finally, precipitations are filtered and recrystallized from distilled water to produce acicular crystals. The LML solution is obtained by dissolving the crystals in the absolute ethanol. A 100 nm LML with a refractive index of approximately 1.365-1.400 [26,27]  and recrystallized from distilled water to produce acicular crystals. The LML solution is obtained by dissolving the crystals in the absolute ethanol. A 100 nm LML with a refractive index of approximately 1.365-1.400 [26,27] is coated on the surface of the polishing area of the SPF (polishing depth: 50 μm, polishing length: 20 mm) by a dip−coater (SYDC−200, Shanghai SanYan Technology Corp., Ltd. China, Shanghai, China). A 40 nm gold layer is then coated using a magnetron sputtering machine (MSP−3200, Beijing Chuangshiweina Technology Corp., Ltd. China, Beijing, China) to complete Step 1 in Figure 4a.
The sensor in Step 1 is covalently coupled with a layer of gold nanospheres through additional gold−sulfur bonds to complete Step 2. The sensor in Step 2 is immersed in dopamine solution (2 mg/mL, pH = 8.8) and constantly shaken to form a self−polymerized dopamine layer. The resonance spectrum of the sensor in Figure 4b is monitored during the self−polymerization process, and the resonance curve shows significant redshift and broadening, indicating that the sensitivity is enhanced but the detection accuracy is decreased. Experimental results show that the sensor performs better when the self−polymerization time is 30 min. The scanning electron microscope (SEM) image of the cross−section of the sensor is shown in Figure 4b. The sensor was dried and was then immersed in 200 μg/mL rabbit anti H−IgG solution. The whole component was incubated overnight at 4 °C to fully immobilize the antibody. In order to investigate whether the antibody is immobilized to the sensor surface, the resonance spectrum of the sensor is monitored as shown in Figure 4b when the sensor is immersed in the antibody solution. It is found that the resonance curve appears obvious redshift which indicates that the antibody is effectively immobilized. Additionally, BSA has been used as blocking agent in surface passivation.

Experimental Setup
The sensor is used as a sensing probe to connect to the experimental setup shown in Figure 5  The sensor in Step 1 is covalently coupled with a layer of gold nanospheres through additional gold-sulfur bonds to complete Step 2. The sensor in Step 2 is immersed in dopamine solution (2 mg/mL, pH = 8.8) and constantly shaken to form a self-polymerized dopamine layer. The resonance spectrum of the sensor in Figure 4b is monitored during the self-polymerization process, and the resonance curve shows significant redshift and broadening, indicating that the sensitivity is enhanced but the detection accuracy is decreased. Experimental results show that the sensor performs better when the self-polymerization time is 30 min. The scanning electron microscope (SEM) image of the cross-section of the sensor is shown in Figure 4b. The sensor was dried and was then immersed in 200 µg/mL rabbit anti H-IgG solution. The whole component was incubated overnight at 4 • C to fully immobilize the antibody. In order to investigate whether the antibody is immobilized to the sensor surface, the resonance spectrum of the sensor is monitored as shown in Figure 4b when the sensor is immersed in the antibody solution. It is found that the resonance curve appears obvious redshift which indicates that the antibody is effectively immobilized. Additionally, BSA has been used as blocking agent in surface passivation.

Experimental Setup
The sensor is used as a sensing probe to connect to the experimental setup shown in Figure 5. The light emitted by the wide-spectrum lamp (DH-2000-BAL, Ocean Insight, Inc., Orlando, FL, USA) passes through the unpolished fiber into the sensing probe area located in the customized glass tube. The transmitted light is received by the spectrometer (Maya2000 Pro, Ocean Insight, Inc., Orlando, FL, USA), and the transmission spectrum is displayed on the computer. The H-IgG solution with specific concentration is injected into the glass tube via a micro-injection pumper (LSP01-3A, Longer Precision Pump Corp., Ltd. Baoding, China) through the inlet to complete Step 3, and the excess solution flows out to the waste cylinder through the outlet. The reaction lasted for 40 min after the injection of H-IgG solution. Then, the mouse anti H-IgG solution with a concentration of 40 µg/mL is injected to complete Step 4 with a reaction lasting 15 min. To detect other concentrations of H-IgG, the sensor can be rinsed several times with PBS and the above steps will be repeated. Note that the above immunoreaction has been carried out in a room with a constant temperature of 25 • C. To detect other concentrations of H−IgG, the sensor can be rinsed several times with PBS and the above steps will be repeated. Note that the above immunoreaction has been carried out in a room with a constant temperature of 25 °C.

Refractive Index Sensing Experiment
To investigate the influence of plasmonic coupling on the sensor performance, the refractive index sensing performance of the bare LR−SPR sensor and the gold nanosphere−modified LR−SPR sensor have been explored. Experimental results are shown in Figure 6 and Table 1. The modification of gold nanospheres increases the surface molecular weight in the sensor and strengthens the intensity of localized electromagnetic field. Therefore, the resonance curve position of the gold nanosphere−modified LR−SPR sensor exhibits redshift, and the average sensitivity shows an enhancement of 838.28 nm/RIU (2946.44-3784.72 nm/RIU) compared with that of the bare LR−SPR sensor. Due to the light scattering characteristics of gold nanospheres, the average FWHM of the gold nanosphere−modified LR−SPR sensor is increased by 37.51 nm (75.63-113.14 nm) compared with that of the bare LR−SPR sensor. For both types of sensors, the resonance curves will be gradually broadened with the increase in the refractive index. This is because the number of modes, which excite the SPR, increases. In other words, the phase−matching conditions for exciting SPR become less stringent.

Refractive Index Sensing Experiment
To investigate the influence of plasmonic coupling on the sensor performance, the refractive index sensing performance of the bare LR-SPR sensor and the gold nanospheremodified LR-SPR sensor have been explored. Experimental results are shown in Figure 6 and Table 1. The modification of gold nanospheres increases the surface molecular weight in the sensor and strengthens the intensity of localized electromagnetic field. Therefore, the resonance curve position of the gold nanosphere-modified LR-SPR sensor exhibits redshift, and the average sensitivity shows an enhancement of 838.28 nm/RIU (2946.44-3784.72 nm/RIU) compared with that of the bare LR-SPR sensor. Due to the light scattering characteristics of gold nanospheres, the average FWHM of the gold nanospheremodified LR-SPR sensor is increased by 37.51 nm (75.63-113.14 nm) compared with that of the bare LR-SPR sensor. For both types of sensors, the resonance curves will be gradually broadened with the increase in the refractive index. This is because the number of modes, which excite the SPR, increases. In other words, the phase-matching conditions for exciting SPR become less stringent.
Gold nanosphere modified LR-SPR sensor Figure 6. The resonance spectra of the (a) LR-SPR sensor and (b) gold nanosphere-modified LR-SPR sensor. Inset: red curve represents the binomial fitting of refractive index and resonance wavelength, and the tangent slope at each point represents sensitivity, and blue curve represents the FWHM variation with different refractive index. The developed sensor has been used to detect PBS-prepared H-IgG solutions with different concentrations, and detection results are shown in Figure 7. In the process of combining the two antibodies and antigens to form the macromolecular complex, the average refractive index of the sensor surface increases continuously, and the resonance curve exhibits redshift, as shown in Figure 7a. In addition, the self-compensation function of the sensor improves the detection reliability. Applying the Langmuir fitting to the low concentration part of the curve in Figure 7b, the tangent slope at the first point (1 µg/mL) is considered as the sensor's sensitivity S nλ and its value is 2.20 nm/(µg/mL). The wavelength resolution of the experimental system is 0.24 nm according to the 2ρ principle [28]. The LOD of the sensor is 0.109 µg/mL according to Equation (1) [29,30].
where ρ is the standard deviation (S.D.) obtained from the repeated measurement of the resonance wavelength in the sensor corresponding to a specific refractive index every 30 s, according to Equation (2).
where λ i and λ are the resonance wavelength value measured each time and the average value of resonance wavelength, respectively. i = 1, 2, 3, . . . , 100, n = 100. The LOD of 0.109 μg/mL is obviously lower than the expected range of H−IgG concentration in the blood plasma or serum (4.07 × 10 3 -2.17 × 10 4 μg/mL [31]). This can effectively reduce the influence of measurement errors on detection results. In addition, the principle of the developed sensor works based on the immune reaction between the antigen and the antibody. Other types of antigens can also be detected when corresponding antibodies are employed. Therefore, biomarkers with lower concentrations in the blood plasma or serum such as prostate specific antigen [32] can be detected by utilizing the SPR The LOD of 0.109 µg/mL is obviously lower than the expected range of H-IgG concentration in the blood plasma or serum (4.07 × 10 3 -2.17 × 10 4 µg/mL [31]). This can effectively reduce the influence of measurement errors on detection results. In addition, the principle of the developed sensor works based on the immune reaction between the antigen and the antibody. Other types of antigens can also be detected when corresponding antibodies are employed. Therefore, biomarkers with lower concentrations in the blood plasma or serum such as prostate specific antigen [32] can be detected by utilizing the SPR sensor with a lower LOD.
In fact, the LOD of the sensor is related to the wavelength resolution of the spectrometer, the detectable concentration range, the data processing and many other factors. IgG detection results using the electromagnetic resonance sensor have also been listed for similar methods and analytes, as shown in Table 2. It indicates that results in this paper have achieved good agreements with the reported research works. Different protein solutions with the same concentration of 10 µg/mL are also detected using this developed sensor. H-IgG solutions with the same concentration of 10 µg/mL have been detected multiple times. As Figure 8a shows, the resonance curve exhibits obvious redshift in the detection of H-IgG because the specific binding only occurs between the antibody and H-IgG. As shown in Figure 8b, the resonance wavelength redshift does not change much and gets slightly smaller after five cycles because the amount of antibody decreases during the dissociation and the rinse of the sensor. Above-experimental results indicate that the sensor has good specificity and repeatability. Meanwhile, the experimental setup exhibits good reproducibility. Nanomaterials 2021, 11, 2137 9 of 12 results indicate that the sensor has good specificity and repeatability. Meanwhile, the experimental setup exhibits good reproducibility.

Serum Matrix Sample Detection
Controlled experiment has also implemented to investigate the practicability of the developed sensor. 10 μg/mL H−IgG solutions are separately prepared by PBS and donkey serum. The sensor is used to detect the PBS matrix sample at first, and the resonance curve shift   is 16.38 nm. The sensor is then rinsed and is used to detect the serum matrix

Serum Matrix Sample Detection
Controlled experiment has also implemented to investigate the practicability of the developed sensor. 10 µg/mL H-IgG solutions are separately prepared by PBS and donkey serum. The sensor is used to detect the PBS matrix sample at first, and the resonance curve shift ∆λ PBS is 16.38 nm. The sensor is then rinsed and is used to detect the serum matrix sample, and the observed resonance curve shift ∆λ serum is 17.29 nm. It is worth noting that the initial positions of the resonance curve are different for the two test samples, because the refractive indices of the two samples are different. Corresponding experimental results are shown in Figure 9. Repetition rate is defined according to Equation (3) [24].

Serum Matrix Sample Detection
Controlled experiment has also implemented to investigate the practicability of the developed sensor. 10 μg/mL H−IgG solutions are separately prepared by PBS and donkey serum. The sensor is used to detect the PBS matrix sample at first, and the resonance curve shift P BS   is 16.38 nm. The sensor is then rinsed and is used to detect the serum matrix sample, and the observed resonance curve shift serum   is 17.29 nm. It is worth noting that the initial positions of the resonance curve are different for the two test samples, because the refractive indices of the two samples are different. Corresponding experimental results are shown in Figure 9. Repetition rate is defined according to Equation (3) [24]. The calculated repetition rate is 105.56%, which is close to 100%. This indicates that the nonspecific interaction induced by donkey serum matrix is acceptable, and it is feasible for the developed sensor to detect the real serum sample. For comparison, Table 3 lists the repetition rate obtained from our research and reported works using similar methods, which suggests the repetition rate achieved in this article is at a good level.  The calculated repetition rate is 105.56%, which is close to 100%. This indicates that the nonspecific interaction induced by donkey serum matrix is acceptable, and it is feasible for the developed sensor to detect the real serum sample. For comparison, Table 3 lists the repetition rate obtained from our research and reported works using similar methods, which suggests the repetition rate achieved in this article is at a good level.

Conclusions
A novel approach for introducing the double-antibody sandwich immunoassay and plasmonic coupling into the LR-SPR sensor has been reported and applied to the detection of H-IgG. The developed biosensor is used to detected the PBS matrix H-IgG sample with different concentrations, and the demonstrated sensitivity and LOD reach 2.20 nm/(µg/mL) and 0.11 µg/mL, respectively. The remarkable amplification of sensing response is attributed to: (1) the stronger local electromagnetic field obtained from the combination of LR-SPR and L-SPR, and (2) the more obvious refractive index change of the dielectric layer on the sensor surface obtained by the double-antibody sandwich immunoassay. The biosensor has also been employed to detect the PBS matrix sample and the serum matrix sample (with the same concentration), and the ratio of the resonance curve shift obtained in the detection is 105.56%. This indicates that the nonspecific binding induced by the serum matrix puts negligible effect on the sensing performance and the developed sensor possesses certain practicability. Furthermore, the self-compensation against the disturbance allows the developed biosensor to provide higher reliable detection results. In future research, antigen can be labeled on the surface of the gold nanosphere to bind with the CAb attached to the gold layer. Therefore, the antigen-antibody binding will occur in the gap between the gold nanosphere and the gold layer, and the enhancement of the localized electromagnetic field will be fully utilized. In order to further reduce the LOD of the sensor, two-dimensional nanomaterials with high refractive indices can also be introduced.