Open Access
Add to BibSonomy
Abstract
We conceptualized and numerically investigated a photonic crystal fiber (PCF)-based surface plasmon resonance (SPR) sensor for rapid detection and quantification of novel coronavirus. The plasmonic gold-based optical sensor permits three different ways to quantify the virus concentrations inside patient’s body based on different ligand-analyte conjugate pairs. This photonic biosensor demonstrates viable detections of SARS-CoV-2 spike receptor-binding-domain (RBD), mutated viral single-stranded ribonucleic acid (RNA) and human monoclonal antibody immunoglobulin G (IgG). A marquise-shaped core is introduced to facilitate efficient light-tailoring. Analytes are dissolved in sterile phosphate buffered saline (PBS) and surfaced on the plasmonic metal layer for realizing detection. The 1-pyrene butyric acid n-hydroxy-succinimide ester is numerically used to immobilize the analytes on the sensing interface. Using the finite element method (FEM), the proposed sensor is studied critically and optimized for the refractive index (RI) range from 1.3348-1.3576, since the target analytes RIs fluctuate within this range depending on the severity of the viral infection. The polarization-dependent sensor exhibits dominant sensing attributes for x-polarized mode, where it shows the average wavelength sensitivities of 2,009 nm/RIU, 2,745 nm/RIU and 1,984 nm/RIU for analytes: spike RBD, extracted coronavirus RNA and antibody IgG, respectively. The corresponding median amplitude sensitivities are 135 RIU-1, 196 RIU-1 and 140 RIU-1, respectively. The maximum sensor resolution and figure of merit are found 2.53 × 10−5 RIU and 101 RIU-1, respectively for viral RNA detection. Also, a significant limit of detection (LOD) of 6.42 × 10−9 RIU2/nm is obtained. Considering modern bioassays, the proposed compact photonic sensor will be well-suited for rapid point-of-care COVID testing.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Considering the existing and continuously developing label-free biosensing assays, the emergence of fast mutating Severe Acute Respiratory Syndrome Coronavirus-2 (SARS-CoV-2) necessitated more compact, inexpensive, scalable, and portable solutions for rapid viral pathogen detection and associated integrative systems with point-of-care (PoC) compatibility. Having highly contagious nature, as of May 27, 2022, the number of worldwide confirmed COVID-19 cases has been over 525 million with reported casualties of over 6.2 million since its exposure in the city of Wuhan, China in 2019 [1]. In 2020, the World Health Organization (WHO) acknowledged this unprecedented outbreak as a global pandemic posing a high risk to vulnerable countries with underdeveloped healthcare systems [2]. SARS-CoV-2 belongs to the positive-sense single-stranded RNA virus group which manifests quick respiratory illness than its former strains, namely, SARS-CoV and MERS-CoV (Middle Eastern Respiratory Syndrome Coronavirus) [3]. In addition to predominant abnormal respiratory symptoms, infected patients with acute cardiac arrests, kidney damages and liver function irregularities have been largely documented indicating myocardial, renal, and hepatic injury due to the pathogenic activity of this virus [4]. The capacity to jump over cross-species barrier facilitates this virus to mutate at a large scale which is apparent by its frequently introducing variants e.g. B.1.1.7 (Alpha), B1.351 (Beta), and B.1.617.2 (Delta) [5]. Regardless of the variants, this virus consists of four biologically significant proteins. Spike glycoproteins act as the viral RBD, whereas the envelope proteins and the membrane proteins are scattered on the lipid bilayer membrane. Nucleocapsid protein or the single-stranded RNA is encapsulated inside this external frame. To get access to host cells, after infiltrating the prospective body from contaminated surfaces or aerosols, a subunit of spike RBD binds with angiotensin-converting enzyme 2 (ACE2) which is available in most of the vital human organs [6]. Therefore, this subunit has been at the peak of researchers’ attentions, as preventing surface receptor binding of the virus could be a potential treatment for this disease [7]. On the other hand, extraction of the viral RNA has also been highlighted for studying reverse genetically engineered antiviral platforms [8]. Moreover, the RNA proteins and spike RBD proteins are being constantly used in laboratory-based coronavirus screening.
To date, real-time reverse-transcriptase polymerase chain reaction (RT-PCR) is regarded as the gold standard for RNA detection. However, it is restricted to well-equipped healthcare facilities with experienced users and hence, not suitable for in situ monitoring of COVID-19 patients. To avoid procurement delays, loop-mediated isothermal amplification (LAMP) has been adopted by several researchers. Lately, Tang et al. [9] presented a LAMP assay coupled with a glass nanopore-based digital amplicon counter which is only enabled for qualitative analysis of the virus. The lateral flow immunoassays (LFIAs) are very cost-efficient paper-strips-based testing kits, commonly used in medicine for PoC medical diagnostics. Despite being convenient for rapid low-cost detection of SARS-CoV-2 in resource-poor areas, this simplified assay can only give binary (yes/no) outputs disabling further quantitative assessment [10]. The new chemiluminescent assay (CLIA) possesses limited scalability for its costly operation [11]. Nouri et al. [12] presented a clustered regularly interspaced short palindromic repeats (CRISPR)-assisted nanopore biosensor which suffers from sensitivity constraints for a limited turnaround period.
Viruses need to intrude appropriate host cells for replicating and propagating genetic materials. The human immune system produces monoclonal antibodies (mAbs) i.e. different immunoglobulins to terminate infected cells via phagocytosis and among them, IgG usually becomes available within 3-6 days following infection [13]. For SARS-CoV-2, the measurement of immune response to this antigenic invasion by detecting and quantifying mAbs plays an important role both in rapid detection of the virus and antibody therapeutics. For home use, 47 antibody PoC diagnostic test kits have been approved by The Board Decision on Additional Support for Country Responses to COVID-19 [14]. As their accuracies were not up to the mark, some lab-based detection schemes are also presented by biomedical engineers. For example, Tré-Hardy et al. [15] depicted the use of an enzyme-linked immunosorbent assay (ELISA) for detecting anti-spike protein antibodies.
Modern plasmonic optical biosensors are anticipated to have great prospects for multiplexed PoC coronavirus sensing. In a work by Cady et al. [16], the emission of fluorescence in a multiplexed coronavirus sensor was amplified by surface plasmon polaritons (SPPs) to successfully boost up its sensitivity. Furthermore, SPP-assisted quantum dot-based aptasensor and microplate reader are depicted by other researchers [17,18]. Localized SPR sensors integrated with microfluidic chip and silver nanoarray are lately employed for detections of antiviral antibodies and spike glycoproteins, respectively [19,20]. Two Kretschmann configuration-based COVID-19 sensors are demonstrated in Ref. [21,22] which suffer from the disadvantages of having cumbersome rotary apparatus. To avoid such bulkiness, optical fiber-based SPR biosensors have been heavily appreciated in reported studies [23]. Cennamo et al. [24,25] designed two plastic optical fiber sensors with decent LODs for rapidly screening spike RBD of COVID-19 virions. The above discussion points out several drawbacks of the current assaying technologies for SARS-CoV-2 particle detections, such as (i) not all existing diagnostic assays and biosensors are PoC-supported; (ii) some required lengthy processing time with trained medical personnel; (iii) many assays are limited by either high cost and/or procurement time and/or portability and/or scalability; (iv) some showed high propensities of false-positive responses; (v) numerous present assays provide only binary (infected/not infected) results and thus, do not allow quantitative analysis; (vi) several coronavirus sensors have not proffered practically favorable performance; (vii) to enable additional sensing options, careful integration of multiple assaying setups is needed; (viii) A very few SARS-CoV-2 sensors are enabled for all 2 types of viral and 1 type of antiviral analytes (spike RBD, RNA, mAbs) detection, and therefore, identifying false-positive outcome becomes difficult.
Simultaneous multianalyte usability makes plasmonic sensors more compact and multi-operational. Because of the high degree of freedom in designing PCFs, they are considered one of the perfect candidates for researching multianalyte SPR biosensing [26]. In PCF sensors, sensitivities are enhanced by efficiently facilitating the SPP excitation by means of structural variations. For instance, wavelength sensitivities of a multianalyte sensor by Zheng et al. [27] could reach up to 1535 nm/RIU and 1550 nm/RIU for its different channels using a wagon-wheel-shaped design. Controlled single-mode light propagation and birefringence let enhanced superiority to the PCF-based plasmonic sensors over the regular optical fiber sensors. Moreover, microstructured PCFs can demonstrate desirable confinement parameters without requiring specific dopant inclusion for performing critical internal reflections [28]. In 2019, Kaur and Singh [29] showed an undoped silica-based PCF sensor for dual-channel sensing with channel 1 and 2 peak wavelength sensitivities of 1000 nm/RIU and 3750 nm/RIU, respectively. Prior to this, another dual-analyte PCF-SPR sensor also exhibited similar decent sensitivities [30]. External etching of conventional fiber cladding using different mechanical and chemical methods produces various unexpected experimental errors [31]. These problems are absent in PCF fabrication which allow it greater robustness for sensing applications. Yet, no PCF-based SPR sensors have been explored for the purpose of COVID screening so far.
In this work, we conceptualized a PoC compatible plasmonic biosensor for accurately quantifying COVID-19 viral specimens based on a propagation-controlled core PCF which is supported by the SPR phenomena. We did FEM-based numerical simulations and attained improved sensitivities and LODs for our proposed photonic sensor. The location of the analyte layer is selected to be outside the fiber for facilitating fast operations and to simplify the swapping procedure of analytes. Being a stable element, gold (Au) is employed externally to constitute the plasmonic sensing interface and thereby, the likelihood of oxidation is minimized. Two individual channels are introduced to enable real-time on-site multianalyte detections. It is shown that the proposed dual-channel sensor operates without the deficiencies categorized in the texts above. All structural parameters of the PCF (pitch and air hole sizes) and Au thickness are critically studied and numerically optimized. This being the first PCF sensor for early novel coronavirus detection and quantification, it will play an important role in our fight against COVID-19.
2. Sensor design & operation
2.1 Theoretical modeling
A biosensor can be viewed as a device including a transducer and signal processing unit. The sensing interface of the transducer, in association with bioreceptor(s), specifies certain analytes which are eligible for detection by a particular sensor. Our PCF-SPR sensor can promptly detect all three analytes which are investigated so far for SARS-CoV-2 screening. To activate the sensor, certain bioreceptor molecules (which show affinity to the target analyte) act as ligands and are immobilized onto the sensing surface to bind with the chosen proteins. The RI of the sensitive interface of the device then changes depending on the dynamics of the microprotein organisms. An SPR will be mandated when these ligand-analyte conjugates receive the highest energy from the transmitted light within the leaky fiber core. This peak energy transaction results in a sharp intensity drop (confinement loss) of the input signal at a definite wavelength (aka resonant wavelength) for a fixed analyte RI. Also, the optical frequency perfectly matches with the frequency of naturally oscillating SPPs under satisfied resonance conditions.
We designed a silica PCF that has three sequential rings of hexagonally patterned air holes. The core of the fiber is constructed by scaling down several air holes (RI of air = 1) from all of these rings so that light can propagate within the core by performing modified total internal reflections (mTIR). The electromagnetic wave will distribute throughout the holy core, as the smaller air holes are allowing some high RI regions (fused silica RI ∼1.45 for visible lights) surrounding them. The refractive properties of background silica can be obtained using the Sellmeier Eq. (1) as a nonlinear function of broadband wavelengths [32].
Here, Bi = 1, 2, 3 and Ci = 1, 2, 3 are defined as the Sellmeier coefficients (B1 = 0.696166300, B2 = 0.407942600, B3 = 0.897479400, C1 = 4.67914826.10−3µm2, C2 = 1.35120631.10−2µm2 and C3 = 97.9340025 µm2).
As depicted in Fig. 1(a), four, six and two air holes diameters are decreased in size from the first, second and third rings, respectively. Thus, the shape of the core resembles that of a marquise-cut diamond. The stacking of glass capillaries to obtain this kind of lattice configuration is displayed in Fig. 1(b) which allows easy fabrication and facilitates mass production. This arrangement of core (small) airholes deliberates 2 horizontal leakage paths through which evanescent electric fields will dissipate energy to the plasmonic Au layer for SPP excitation. Drude-Lorentz model is employed to functionalize the complex relative permittivity of the metal and the optical constants of Au can be accessed from Ref. [32].
The semicircular dual-channels can be used to simultaneously run two independent COVID-19 tests without any interference between them. And there are also no prerequisite conditions for doing so i.e. both tests can either be of the same analyte of different RIs or of different SAR-CoV-2-associated samples. The analyte, dissolved in Sterile Phosphate Buffered Saline (PBS), will be streamed over the sensor surface for detection. With a pH of ∼7.4, uncontaminated PBS depicts an RI of 1.3348. Virus transport media (VTM) are used to limit the chance of contamination while carrying or experimenting with any viral sample. The use of PBS is encouraged by the Food and Drug Administration owing to the shortage of conventional COVID-19 VTM [33]. Moreover, as the analyte-PBS solution yields an increasing RI with any denser viral concentration which is slightly greater than clean deionized (DI) water (RI = 1.33), we leveraged this to complete our desired SPR functionalization for target molecules. To visualize the multianalyte sensing feasibility, in Fig. 1(c), we showed the x-polarized mode confinement loss response of our proposed sensor when both channels are filled with the same analyte and when they are filled with different analytes. One can see that the loss spectrum consists of only one resonant peak for the given PBS RI 1.3348 filled in both of the channels. For unknown analytes of RI 1.33 and RI 1.35 in opposite channels, two resonant peaks are observed within a single loss curve.
Fig. 1. (a) Cross-sectional 2D view of the PCF sensor. (b) Stacked capillaries for PCF drawing, and (c) Verification of multianalyte operation feasibility.
2.2 Structure & sensing operation
Fabrication of our proposed PCF biosensor is very convenient in all respects. This type of hexagonal lattice with uniformly spaced air holes does not require micro-drilling, since the structure can be achieved by simple stack and draw method involving capillaries of different wall thicknesses. Cladding layer assists in controlling the desirable levels of light confinement and the field leakage by means of mTIR. Therefore, our FEM-based investigation involves optimization of the core-clad dimensions to confirm our expected waveguiding operation. We finalized the optimum values for regular air hole diameters (d) and small air hole diameters (dc) which are 1.075 µm and 0.33 µm, respectively. A 1µm analyte layer (AL) is found to be adequate for the sensor to function sensitively. The proposed PCF has constant center-to-center spacing (pitch size, $\wedge$) of two air holes situated next to each other. Upon verifying the holy fiber for several pitches, we selected the optimal pitch to be 1.65 µm. In terms of performance, we showed that an Au nanolayer with a width (t) of 40 nm exhibited improved sensitivity. We recommend atomic layer deposition or nanoparticle layer deposition for placing the Au layer over the round surface of the fiber to confirm greater uniformity and no contamination [34,35].
Figure 2 displays the schematic setup of the presented sensing system. As our sensor is practically a single-mode fiber coated with an Au annular nanofilm, a broadband light source transmits a range of optical frequencies through the fiber after the light is polarized by a transverse magnetic or p-polarizer. Our sensor operates within the suitable visible wavelengths due to the plasma frequency characterized by Au. The COVID-19 samples can be collected from affected patient’s nasal cavity using nasopharyngeal swabs and are inactivated through heat or gamma ray [36], whereas the IgG proteins are only available through patient’s blood sample. Due to fine compatibility of our sensing kit with most of the RNA extraction procedures, user conserves the freedom to choose either magnetic nanoparticles (MNP) or RT-PCR for segregating the virus RNA [37]. After the samples are dissolved in PBS and floated over the sensor surface, the change of molar concentration varies the activity of the target analytes significantly and results in equivalent RI fluctuation in a linear manner [38]. So, the resonant peaks are relocated by the different molar concentrations of analytes in the sensing channels. 1-pyrene butyric acid n-hydroxy-succinimide ester (PBSE) is responsible for immobilizing the viral and antiviral particles onto the sensor surface. The immobilized bioparticles work as the acting ligand and attach to the target analytes. Besides PBSE, modified ACE2 can be alternatively utilized as spike protein immobilizers [39]. For the range of RI variation for SARS-CoV-2 detection (RI 1.3348-1.3576), a digital system is trained beforehand with the experimental loss curves. The polarized light will show a maximum loss in intensity based on the unknown RI flowing over the sensor. A charged-coupled device or an optical spectra analyzer, probed to the aforesaid computational system, is deployed to collect those modal loss curve readouts. After necessary interpolation, the system retrieves the exact losses and compares them with the in-memory SPR responses to display the final result on a monitor.
Fig. 2. Schematic setup for sample preparation and the protocol for SARS-CoV-2 quantification by the proposed sensor.
It is understandable that our PCF-SPR sensor is basically a refractometer that can precisely measure the RI of the aqueous solution flown over its active surface. To make this optical device viable for SARS-CoV-2 detection, we translated the molar concentrations of COVID-19 RNA, spike receptor-binding subunit and mAb IgG particles in terms of their equivalent RIs. Mimicking the actual interaction mechanism between SARS-CoV-2 antigens and human ACE2, Forssén et al. [40] and Lan et al. [41] established a rate constant distribution algorithm with the help of adaptive interaction distribution algorithm (AIDA) which is used to approximate the concerned RIs initially in our study. Later, we finalized the RIs by evaluating linear regressions following articles [42–44]. The equivalence is listed in Table 1 where the COVID positive/negative threshold is defined in Ref. [45].
Table 1. RIs of isolated coronavirus RNA, spike RBD and human mAbs IgG at different molar concentrations
3. Simulation results and discussions
3.1 Performance analysis
To perform our wave optics modal analysis based on FEM, we sketched, modeled and simulated the sensor in the commercial software COMSOL v5.5. We selected a large mesh size to increase the modal data accuracy. The simulation was conducted by choosing a predefined cylindrical perfectly matched layer with default boundary conditions which is liable for absorbing the scattered evanescent field.
Our structure shows very high polarization dependence which results in dominant plasmonic responses in one polarized mode in comparison to the other optical mode. This is engineered by introducing two energy transfer pathways in the lateral direction using our proposed core while blocking other possible paths in the vertical directions setting higher air-contrasted cladding layers. Figure 3(a) displays the leaky electric field dispersions of the fundamental core-guided mode and the SPP mode in the highly responsive x-polarized mode for the RI of uncontaminated PBS. The loss of light intensity due to this core to plasmon energy transference is viewed as the incapacity of confinement of light by the fiber. This is termed confinement loss which can be computed in dB/cm by the following equation where k0 is the wave number and Im(neff) is the imaginary modal effective index,
Fig. 3. (a) Electric field profiles of the core-guided mode and SPP mode at analyte RI 1.3348. (b) Satisfied phase-matching condition, and (c) Birefringent behavior.
The resonant wavelengths change towards the higher or lower spectral region, since they are contingent upon the concentration of microbes at the sensor plane. This happens, because the variation of SPP mode real RIs depending on the activeness of the microproteins defines a newer phase matched SPR point where the loss peak has been updated. Highly concentrated analytes on the sensitive layer produces a large number of ligand-analyte complexes and vice versa. In our PCF sensor, the peak energy transfer points move to the longer wavelengths (red-shifts) because of enhanced viral activity and return to shorter wavelengths for poor number of active analyte bindings. To demonstrate our claim, we numerically inspected all three approaches to detect and quantify SARS-CoV-2 infection severity and immune responses within patient’s body in terms of RIs (correspondent molar densities) of the pertinent analytes.
As discussed above, many of the extraction methods for isolating nucleocapsid protein are time-killing and not PoC usable. Therefore, a faster (∼30 minutes) MNP-based RNA isolation strategy has been proposed by a group of scientists that has the capacity to replace the conventional RT-PCR-based nucleic acid extraction in near future [45]. We recommend this novel technique for expediting this RNA extraction process from bulk viral samples. To ready the viral nucleic acid particles for detection, it is mixed with PBS and glided through both channels of the sensor for molar concentration measurement. This is done for RI 1.3462 to RI 1.3576 as mentioned in column 2 of Table 1 and the loss spectra are presented in Fig. 4(a), displaying the red-shifting SPR peaks. We can see the escalations of loss depths for increased applied RIs inside sensor channels. Because, while the test RI is increasing, the declining RI contrast between the fundamental core mode and the plasmon mode results in upgraded modal loss peaks [26]. The peak propagation loss increases from 62 dB/cm to 101 dB/cm when the uncontaminated PBS is exploited with 300 nM viral RNA, equaling an RI of 1.3576. In this experiment, the SPR spectral shifts are 24.3 nm, 2 nm, 3 nm, 6.5 nm, 6.5 nm, 7 nm and 9 nm for the test saline concentrations of 0 nM, 150 nM, 165 nM, 180 nM, 210 nM, 240 nM, 270 nM and 300 nM, respectively. The resonant shift caused by a unit change in RI (na) is defined as the wavelength sensitivity (4).
We sited a maximum sensitivity of 3948 nm/RIU at RNA concentration 270 nM or RI 1.35532 in the wavelength interrogation method, whereas the mean sensitivity is 2745 nm/RIU. Besides, for the positive threshold concentration, this sensitivity is 2851 nm/RIU. The sensing parameters are tabulated in Table 2. We intentionally took non-uniformly different molar concentrations of coronavirus RNA for the analysis as our highest priority was to achieve extremely exact operation around the positive threshold RI value. As a result, the corresponding RIs varied in accordance with molar densities and so do the sensitivities. The initial sensitivity decrease is proffered by the large RI gap form pure PBS to 150 nM RNA solution. Afterward, the sensitivities showed a consistent increase with our expected estimations. With an approximation of minimal spectral resolution (Δλmin) as 0.1 nm, a sensor resolution of 2.53 × 10−5 is calculated for the detection of this COVID-19 genetic material. In amplitude interrogation, a peak sensitivity of 240 RIU-1 is outputted for the initial RI with an average sensitivity of 196 RIU-1 (please see Fig. 4(b)). The threshold RI amplitude sensitivity is estimated to be 185 RIU-1. Amplitude sensitivities are calculated using the Eq. (5) below.Fig. 4. Viral RNA concentrations detection. (a) Fiber confinement loss spectra, (b) Pertinent amplitude sensitivities, (c) Polynomial fitting of resonant wavelengths.
Table 2. Performance of the prescribed sensor in the detection of single-stranded COVID RNAa
Figure of merit (FOM) can be attained up to 101 RIU-1 while RNA samples are being tested by this biosensor. Since FOM is the arithmetic value designated by the ratio of wavelength sensitivity and the full-width-half-maxima (FWHM) at the same RI, specifically for our sensor, FOMs show high proportionality to the sensitivities. Because the FWHMs are found significantly constant (37∼39 nm) for our designed structure. Including these abovementioned performance indicators, LODs (=Resolution/ Sλ) for RNA sensing are also contained in Table 2. A very small LOD of 6.42 × 10−9 RIU2/nm is offered by the proposed SPR sensor at applied RI of 1.35532. Figure 4(c) shows that the R2 is almost approximating unity (0.9992) when a second-order polynomial curve is fitted to interpolate the resonant wavelengths.
The antigenic spike RBD sensing and quantification involve the same series of tasks. In incorporation with PBS, the solution will be pumped into the sensing channels for PoC detection. The loss behavior and sensitivities are checked by applying the RIs from 1.3348 to 1.3468 who are representing concentrations from 0 to 62.5 nM (as listed in column 4 Table 1). The right-shifted resonant maxima are portrayed in Fig. 5(a) as they proffer wavelength sensitivities from 1867 nm/RIU to 2284 nm/RIU for spike protein RIs according to their order in the Tables 1, 3. The threshold sensitivities and the exact spectral positions for each SPR are also available in Table 3. The average amplitude sensitivity and FOM are 135 RIU-1 and 53 RIU-2, respectively and the individual sensitivity spectra are visualized in Fig. 5(b). Minimal wavelength resolution and LOD are estimated to be 4.38 × 10−5 RIU and 1.92 × 10−8 RIU2/nm while investigating spike glycoproteins for experimented molar concentrations. Here, the R2 equals unity and we depicted the fitted characterization curve in Fig. 5(c).
Fig. 5. Antigenic spike glycoprotein concentrations detection. (a) Fiber confinement loss spectra, (b) Pertinent amplitude sensitivities, (c) Polynomial fitting of resonant wavelengths.
Table 3. Performance of the prescribed sensor in the detection of spike RBD glycoproteina
Separate SARS-CoV-2 viral specimens have distinct areas of diagnostic applications e.g. screening of this infection, genetic engineering and vaccinology. Serological tests or mAbs quantification is also of great interest in generating longitudinal data for qualitative studies on administered vaccines [47]. From the list in Table 1, RIs corresponding to IgG concentrations (RI 1.3348-1.3465) are studied and maximum and median spectral sensitivities are respectively obtained 2257 nm/RIU and 1984 nm/RIU, where the COVID positive threshold sensitivity is 1786 nm/RIU. The loss curves and representative amplitude sensitivity curves are depicted in Figs. 6(a) and 6(b), respectively. More numerical details regarding the detection of mAb IgG are summarized in Table 4. The highest FOM and smallest LOD are computed to be 59 RIU-1 and 1.96 × 10−8 RIU2/nm respectively for antibody sensing. The magnitude of R2 is again found to be 1 (see Fig. 6(c)) indicating simplified computation of arbitrary RI by the processor.
Fig. 6. Quantification of mAb IgG concentration. (a) Fiber confinement loss spectra, (b) Pertinent amplitude sensitivities, (c) Polynomial fitting of resonant wavelengths.
Table 4. Performance of the prescribed sensor in the detection of anti-spike monoclonal IgGa
3.2. Multianalyte operation
Figures 7(a-c) show real-time multianalyte sensing ability of the prescribed biosensor for practical coronavirus cases. Here, channel 1 is consistently infiltrated with unadulterated PBS when channel 2 analyte is sequentially varied by the highest concentrations of PBS-dissolved micro-analytes (300 nM RNA, 62.5 nM spike protein and 27.8 nM antiviral IgG). Their corresponding RIs can be located in Table 1. Within the singular loss spectrum, the resonant loss peaks in lower wavelengths are due to the pure PBS, whereas our analytes are causing the 2nd loss peaks at higher wavelengths in those three figures. Individual channels containing differently active micro-organic solutions are allowing this efficient multianalyte COVID-19 quantification by optically coupling with the same PCF core at separate SPR frequencies depending on the viral infectivity.
Fig. 7. Multianalyte detections of SARS-CoV-2 analytes. Loss spectrum (a) for pure PBS in Ch1 and high concentrated RNA in Ch2, (b) for pure PBS in Ch1 and high concentrated Spike RBD in Ch2 (c) for pure PBS in Ch1 and high concentrated human antibody IgG in Ch2. And loss curves for multiple RNA tests (d) for pairwise variation of RNA concentrations in Ch1, Ch2, (e) for pure PBS in Ch1 and varied RNA levels in Ch2, (f) Amplitude sensitivities.
To better understand this simultaneous multianalyte detection opportunity, we further elucidated the procedure while experimenting the viral RNA and evaluated the correspondent sensitivities. In Fig. 7(d), we can see similar double peak loss spectra when both channel 1 and channel 2 RIs are continuously being altered. For sample RI 1.3348 in Channel 1 and RI 1.35076 in channel 2, the peaks are sited at 615.7 nm and 655 nm respectively. Other two RI pairs representing RNA samples are (1.34730, 1.35532) and (1.34848, 1.35760) and the spectral locations (in nanometers) of their associated resonant loss peaks as ordered pairs are (646, 662) and (650, 674) respectively. On the other hand, Fig. 7(e) is portraying loss curves for analyte variations from RI 1.3462 to 1.3576 in channel 2 when channel 1 analyte is kept fixed at RI 1.3348. All of them depicted two distinguishable SPR peaks within a single curve. Therefore, it is also evident that the inductions of double SPRs are similarly distinct for constant RI in one channel. During multianalyte performance inspection, we found the highest amplitude sensitive response of 233 RIU-1 as shown in Fig. 7(f). Alongside real-time multipurpose detections, this feature gives the leverage of identifying false-positive and false-negative responses (if they occur) instantly unlike many other immunoassays. The erroneous results are detected when the user is aware that the channels are engaged with the same unknown SARS-CoV-2 sample of the same concentration and yet displaying more than one SPR peak in one loss spectrum. Afterward, the original sensing output can be obtained by simply recalibrating the system and repeating the test. This is very useful even though the proposed sensor is designed to avoid all false detections. Au-based plasmonic sensors usually show these fake responses because of losing the steepness of resonant curves in longer wavelengths [48]. The consistency of FWHM (38 nm as seen in Tables 3&4) for changing RIs warrants that this problem does not apply to our sensor.
3.3 Optimization & tolerance investigation
To maximize the sensor efficacy and estimate tolerable fabrication imperfections, we fixed all of the dimensional variables of the proposed PoC sensor by investigating the optimality of operation. The critical study is done in terms of modal loss variations for x-polarization when RI 1.3348 (Pure PBS) is infused in both channels. Because, sensor sensitivity is directly reliant upon the propagation loss proffered by the optical device [26]. The Thickness of the annular metal coating regulates the RI contrasts surrounding the analyte layer and directly modulates sensitivities as a consequence. Figure 8(a) depicts the confinement losses when the thickness of the Au deposit changes from 30 nm to 50 nm where that modal loss is suppressed by escalating widths of the Au layer. In addition to RI 1.3348, we used RI 1.3465 to evaluate the spectral and amplitude sensitivities. Resonant points moved in the forward direction by about 21 nm, 24 nm and 26 nm for t equaling 30 nm, 40 nm and 50 nm, respectively. This clearly implies better wavelength sensitivity for thickened Au layers. However, from the perspective of amplitude interrogation, the sensitivity at RI 1.3348 drops to 214 RIU-1 from 229 RIU-1 due to this increase (please see Fig. 8(b)). The changing trend of amplitude sensitivities is similar to the trend of loss depth variation. We selected an Au coating of 40 nm for our SPR sensor considering the trade-off of sensitivities in different interrogation methods so that we can receive decent performance regardless of the interrogative approach.
Fig. 8. Fabrication tolerance investigation. (a) Propagation loss and (b) amplitude sensitivities while varying Au-layer thickness at RIs 1.3348 & 1.3465. Optimizations of (c) analyte layer width, (d) pitch size, (e) large air hole diameters, (f) small air hole diameters.
The elegance of the propagation-controlled design can be understood by the high structural tolerance it has demonstrated. The PCF showed high sustainability of its loss characteristics when scrutinized for structural changes up to ±10%. For instance, Fig. 8(c) is literally showing no fluctuations of SPRs in respects of spectral positions and depths of losses for varied analyte layers (AL) from our prescribed value. Furthermore, for any change in optimal pitch size ($\wedge$=1.65µm), the spectral shift is no more than 2 nm. Figure 8(d) displays that the confinement loss magnitudes increased by 16 dB/cm and decreased by 15 dB/cm for $\wedge$ of 1.82 µm and 1.48 µm, respectively. The cladding air holes changes depicted even lesser variations (see Fig. 8(d)). The change of losses in core-guided mode range from merely +9% to -9% for ±10% fluctuations from our suggested diameter size (d = 1.075). When moderating core air holes, peak losses of 47 dB/cm, 54 dB/cm, 71 dB/cm and 83 dB/cm were observed for successively varying the dc = 0.33µm by -10%, -5%, 5 and 10% as shown in in Fig. 8(f). Again, for both types of air holes, movements in SPR wavelengths are not heavily inflected showing maximum deviations of 2.5 nm due to these changes in diameters. The above discussion proves that the likely levels of fabrication defects will not have major impacts on the overall sensing functionality and detection performance.
4. Conclusions
We designed a PCF sensor for accurate multianalyte detection purposes of influential coronavirus specimens with PoC supportable features. The operating principle of this sensor is the SPR shifts of condiment loss tips resulting from varied analyte concentrations over the sensing interface. Molar concentrations of living micro-organic compounds are representative of their pathogenic activities inside the host’s body. Therefore, we interpreted equivalent RIs of those concerned concentrations and successfully showed detection feasibility by plasmonic PCF-based photonic sensing strategy. The proposed sensor offers highly sensitive detections of SARS-CoV-2 spike RBD and RNA proteins as well as serological quantification of antibody IgG. The sensor is designed to give COVID-positive results provided that at least the threshold concentrations were sensed and negative results for lower concentrations. Apart from only depicting positive or negative outcomes, it can appropriately measure the exact concentration of corona particles in a given viral sample. So, it can be used as a bioparticle counter which is in high demand in the field of antibody therapeutics and vaccinology. To further accelerate rapid diagnosis, we engineered this photonic sensor enabling the option of multiple operations using two channels for different COVID tests at the same time. In our proposed procedure, PBS solution and target groups of analytes act as solvent and solute respectively to realize the detection. To evaluate the performance, we adopted FEM-based numerical analysis and examined the spectral characteristics in terms of confinement loss curves. We found average wavelength sensitivities of 2,745 nm/RIU for RNA detection, 2,009 nm/RIU for spike protein detection and 1,984 nm/RIU for IgG quantification. Also, the corresponding amplitude sensitivities are 135 RIU-1, 140 RIU-1 and 196 RIU-1. The most intensified sensing responses are observed for viral RNA detection, and it exhibited a high FOM of 101 RIU-1 with a significantly small LOD of 6.42 × 10−9 RIU2/nm. The proposed sensor will be a very useful portable device in resource-limited medical facilities and will significantly surpass the limitations of existing viral assays.
Acknowledgments
Authors acknowledge the support from LightMode Solutions.
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. “WHO Coronavirus (COVID-19) Dashboard,” (May 27, 2022), retrieved May 30, 2022, https://covid19.who.int/.
2. C. Sohrabi, Z. Alsafi, N. O’neill, M. Khan, A. Kerwan, A. Al-Jabir, C. Iosifidis, and R. Agha, “World Health Organization declares global emergency: A review of the 2019 novel coronavirus (COVID-19),” Int. J. Surg. 76, 71–76 (2020). [CrossRef]
3. A. A. Rabaan, S. H. Al-Ahmed, S. Haque, R. Sah, R. Tiwari, Y. S. Malik, K. Dhama, M. I. Yatoo, D. K. Bonilla-Aldana, and A. J. Rodriguez-Morales, “SARS-CoV-2, SARS-CoV, and MERS-COV: a comparative overview,” Infez Med 28, 174–184 (2020).
4. M. Gheblawi, K. Wang, A. Viveiros, Q. Nguyen, J.-C. Zhong, A. J. Turner, M. K. Raizada, M. B. Grant, and G. Y. Oudit, “Angiotensin-converting enzyme 2: SARS-CoV-2 receptor and regulator of the renin-angiotensin system: celebrating the 20th anniversary of the discovery of ACE2,” Circ. Res. 126(10), 1456–1474 (2020). [CrossRef]
5. K. Dhama, S. K. Patel, K. Sharun, M. Pathak, R. Tiwari, M. I. Yatoo, Y. S. Malik, R. Sah, A. A. Rabaan, and P. K. Panwar, “SARS-CoV-2 jumping the species barrier: zoonotic lessons from SARS, MERS and recent advances to combat this pandemic virus,” Travel Med. Infect. Dis. 37, 101830 (2020). [CrossRef]
6. I. Astuti, “Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2): An overview of viral structure and host response,” Diabetes Metab. Syndr. 14(4), 407–412 (2020). [CrossRef]
7. F. Cocozza, N. Névo, E. Piovesana, X. Lahaye, J. Buchrieser, O. Schwartz, N. Manel, M. Tkach, C. Théry, and L. Martin-Jaular, “Extracellular vesicles containing ACE2 efficiently prevent infection by SARS-CoV-2 Spike protein-containing virus,” J. Extracell. Vesicles 10(2), e12050 (2020). [CrossRef]
8. A. A. Amarilla, J. D. Sng, R. Parry, J. M. Deerain, J. R. Potter, Y. X. Setoh, D. J. Rawle, T. T. Le, N. Modhiran, and X. Wang, “A versatile reverse genetics platform for SARS-CoV-2 and other positive-strand RNA viruses,” Nat. Commun. 12(1), 3431 (2021). [CrossRef]
9. Z. Tang, R. Nouri, M. Dong, J. Yang, W. Greene, Y. Zhu, M. Yon, M. S. Nair, S. V. Kuchipudi, and W. Guan, “Rapid detection of novel coronavirus SARS-CoV-2 by RT-LAMP coupled solid-state nanopores,” Biosens. Bioelectron. 197, 113759 (2022). [CrossRef]
10. Y.-W. Tang, J. E. Schmitz, D. H. Persing, C. W. Stratton, and A. J. McAdam, “Laboratory Diagnosis of COVID-19: Current Issues and Challenges,” J. Clin. Microbiol. 58(6), e00512 (2020). [CrossRef]
11. M. Infantino, V. Grossi, B. Lari, R. Bambi, A. Perri, M. Manneschi, G. Terenzi, I. Liotti, G. Ciotta, C. Taddei, M. Benucci, P. Casprini, F. Veneziani, S. Fabbri, A. Pompetti, and M. Manfredi, “Diagnostic accuracy of an automated chemiluminescent immunoassay for anti-SARS-CoV-2 IgM and IgG antibodies: an Italian experience,” J. Med. Virol. 92(9), 1671–1675 (2020). [CrossRef]
12. R. Nouri, Y. Jiang, Z. Tang, X. L. Lian, and W. Guan, “Detection of SARS-CoV-2 with Solid-State CRISPR-Cas12a-Assisted Nanopores,” Nano Lett. 21(19), 8393–8400 (2021). [CrossRef]
13. H. Hou, T. Wang, B. Zhang, Y. Luo, L. Mao, F. Wang, S. Wu, and Z. Sun, “Detection of IgM and IgG antibodies in patients with coronavirus disease 2019,” Clin. Transl. Immunol. 9, e1136 (2020). [CrossRef]
14. “List of SARS-CoV-2 Diagnostic test kits and equipments eligible for procurement according to Board Decision on Additional Support for Country Responses to COVID-19 (GF/B42/EDP11),” (The Global Fund, 2021), retrieved August 4, 2021, https://www.theglobalfund.org/media/9629/covid19_diagnosticproducts_list_en.pdf.
15. M. Tré-Hardy, A. Wilmet, I. Beukinga, J. Favresse, J. M. Dogné, J. Douxfils, and L. Blairon, “Analytical and clinical validation of an ELISA for specific SARS-CoV-2 IgG, IgA, and IgM antibodies,” J. Med. Virol. 93(2), 803–811 (2021). [CrossRef]
16. N. C. Cady, N. Tokranova, A. Minor, N. Nikvand, K. Strle, W. T. Lee, W. Page, E. Guignon, A. Pilar, and G. N. Gibson, “Multiplexed detection and quantification of human antibody response to COVID-19 infection using a plasmon enhanced biosensor platform,” Biosens. Bioelectron. 171, 112679 (2021). [CrossRef]
17. R. Chen, L. Kan, F. Duan, L. He, M. Wang, J. Cui, Z. Zhang, and Z. Zhang, “Surface plasmon resonance aptasensor based on niobium carbide MXene quantum dots for nucleocapsid of SARS-CoV-2 detection,” Microchim. Acta 188(1), 1–10 (2021). [CrossRef]
18. L. Huang, L. Ding, J. Zhou, S. Chen, F. Chen, C. Zhao, J. Xu, W. Hu, J. Ji, and H. Xu, “One-step rapid quantification of SARS-CoV-2 virus particles via low-cost nanoplasmonic sensors in generic microplate reader and point-of-care device,” Biosens. Bioelectron. 171, 112685 (2021). [CrossRef]
19. R. Funari, K.-Y. Chu, and A. Q. Shen, “Detection of antibodies against SARS-CoV-2 spike protein by gold nanospikes in an opto-microfluidic chip,” Biosens. Bioelectron. 169, 112578 (2020). [CrossRef]
20. Y. Yang, J. Murray, J. Haverstick, R. A. Tripp, and Y. Zhao, “Silver nanotriangle array based LSPR sensor for rapid coronavirus detection,” Sens. Actuators, B 359, 131604 (2022). [CrossRef]
21. X. Peng, Y. Zhou, K. Nie, F. Zhou, Y. Yuan, J. Song, and J. Qu, “Promising near-infrared plasmonic biosensor employed for specific detection of SARS-CoV-2 and its spike glycoprotein,” New J. Phys. 22(10), 103046 (2020). [CrossRef]
22. M. Moznuzzaman, I. Khan, and M. R. Islam, “Nano-layered surface plasmon resonance-based highly sensitive biosensor for virus detection: A theoretical approach to detect SARS-CoV-2,” AIP Adv. 11(6), 065023 (2021). [CrossRef]
23. A. A. Rifat, M. R. Hasan, R. Ahmed, and A. E. Miroshnichenko, “Microstructured optical fiber-based plasmonic sensors,” in Computational Photonic Sensors (Springer, 2019), pp. 203–232.
24. N. Cennamo, L. Pasquardini, F. Arcadio, L. Lunelli, L. Vanzetti, V. Carafa, L. Altucci, and L. Zeni, “SARS-CoV-2 spike protein detection through a plasmonic D-shaped plastic optical fiber aptasensor,” Talanta 233, 122532 (2021). [CrossRef]
25. N. Cennamo, G. D’Agostino, C. Perri, F. Arcadio, G. Chiaretti, E. M. Parisio, G. Camarlinghi, C. Vettori, F. Di Marzo, and R. Cennamo, “Proof of concept for a quick and highly sensitive on-site detection of SARS-CoV-2 by plasmonic optical fibers and molecularly imprinted polymers,” Sensors 21(5), 1681 (2021). [CrossRef]
26. A. Islam, F. Haider, R. A. Aoni, M. Hossen, F. Begum, and R. Ahmed, “U-grooved dual-channel plasmonic sensor for simultaneous multi-analyte detection,” J. Opt. Soc. Am. B 38(10), 3055–3063 (2021). [CrossRef]
27. Y. Zhang, C. Zhou, L. Xia, X. Yu, and D. Liu, “Wagon wheel fiber based multichannel plasmonic sensor,” Opt. Express 19(23), 22863–22873 (2011). [CrossRef]
28. I. Danlard and E. K. Akowuah, “Assaying with PCF-based SPR refractive index biosensors: From recent configurations to outstanding detection limits,” Opt. Fiber Technol. 54, 102083 (2020). [CrossRef]
29. V. Kaur and S. Singh, “A dual-channel surface plasmon resonance biosensor based on a photonic crystal fiber for multianalyte sensing,” J. Comput. Electron. 18(1), 319–328 (2019). [CrossRef]
30. A. Yasli, H. Ademgil, S. Haxha, and A. Aggoun, “Multi-channel photonic crystal fiber based surface plasmon resonance sensor for multi-analyte sensing,” IEEE Photonics J. 12(1), 1–15 (2020). [CrossRef]
31. Z. Samavati, A. Samavati, A. F. Ismail, N. Yahya, M. A. Rahman, and M. H. D. Othman, “Effect of acetone/methanol ratio as a hybrid solvent on fabrication of polymethylmethacrylate optical fiber sensor,” Opt. Laser Technol. 123, 105896 (2020). [CrossRef]
32. A. K. Shakya, A. Ramola, S. Singh, and V. Van, “Design of an ultra-sensitive bimetallic anisotropic PCF SPR biosensor for liquid analytes sensing,” Opt. Express 30(6), 9233–9255 (2022). [CrossRef]
33. K. G. Rodino, M. J. Espy, S. P. Buckwalter, R. C. Walchak, J. J. Germer, E. Fernholz, A. Boerger, A. N. Schuetz, J. D. Yao, M. J. Binnicker, and A. J. McAdam, “Evaluation of Saline, Phosphate-Buffered Saline, and Minimum Essential Medium as Potential Alternatives to Viral Transport Media for SARS-CoV-2 Testing,” J. Clin. Microbiol. 58(6), e00590 (2020). [CrossRef]
34. K. Cao, J. Cai, B. Shan, and R. Chen, “Surface functionalization on nanoparticles via atomic layer deposition,” Sci. Bull. 65(8), 678–688 (2020). [CrossRef]
35. A. Csaki, F. Jahn, I. Latka, T. Henkel, D. Malsch, T. Schneider, K. Schröder, K. Schuster, A. Schwuchow, and R. Spittel, “Nanoparticle layer deposition for plasmonic tuning of microstructured optical fibers,” Small 6(22), 2584–2589 (2010). [CrossRef]
36. R. Ahmed, C. F. Guimarães, J. Wang, F. Soto, A. H. Karim, Z. Zhang, R. L. Reis, D. Akin, R. Paulmurugan, and U. Demirci, “Large-Scale Functionalized Metasurface-Based SARS-CoV-2 Detection and Quantification,” ACS nano (2022).
37. G. Ramos-Mandujano, R. Salunke, S. Mfarrej, A. T. Rachmadi, S. Hala, J. Xu, F. S. Alofi, A. Khogeer, A. M. Hashem, and N. A. Almontashiri, “A robust, safe, and scalable magnetic nanoparticle workflow for RNA extraction of pathogens from clinical and wastewater samples,” Global Chall. 5(4), 2000068 (2021). [CrossRef]
38. F. Glover and J. Goulden, “Relationship between refractive index and concentration of solutions,” Nature 200(4912), 1165–1166 (1963). [CrossRef]
39. J.-H. Lee, Y. Lee, S. K. Lee, J. Kim, C.-S. Lee, N. H. Kim, and H. G. Kim, “Versatile role of ACE2-based biosensors for detection of SARS-CoV-2 variants and neutralizing antibodies,” Biosens. Bioelectron. 203, 114034 (2022). [CrossRef]
40. P. Forssén, J. R. Samuelsson, K. Lacki, and T. Fornstedt, “Advanced analysis of biosensor data for SARS-CoV-2 RBD and ACE2 interactions,” Anal. Chem. 92(17), 11520–11524 (2020). [CrossRef]
41. J. Lan, J. Ge, J. Yu, S. Shan, H. Zhou, S. Fan, Q. Zhang, X. Shi, Q. Wang, and L. Zhang, “Structure of the SARS-CoV-2 spike receptor-binding domain bound to the ACE2 receptor,” Nature 581(7807), 215–220 (2020). [CrossRef]
42. H. Zhao, P. H. Brown, and P. Schuck, “On the distribution of protein refractive index increments,” Biophys. J. 100(9), 2309–2317 (2011). [CrossRef]
43. C.-Y. Tan and Y.-X. Huang, “Dependence of refractive index on concentration and temperature in electrolyte solution, polar solution, nonpolar solution, and protein solution,” J. Chem. Eng. Data 60(10), 2827–2833 (2015). [CrossRef]
44. D. Fu, W. Choi, Y. Sung, Z. Yaqoob, R. R. Dasari, and M. Feld, “Quantitative dispersion microscopy,” Biomed. Opt. Express 1(2), 347–353 (2010). [CrossRef]
45. T. B. A. Akib, S. F. Mou, M. Rahman, M. Rana, M. Islam, I. M. Mehedi, M. Mahmud, and A. Z. Kouzani, “Design and numerical analysis of a graphene-coated SPR biosensor for rapid detection of the novel coronavirus,” Sensors 21(10), 3491 (2021). [CrossRef]
46. Q. Zhao, J. Liu, H. Yang, H. Liu, G. Zeng, and B. Huang, “High Birefringence D-Shaped Germanium-Doped Photonic Crystal Fiber Sensor,” Micromachines 13(6), 826 (2022). [CrossRef]
47. A. Petherick, “Developing antibody tests for SARS-CoV-2,” Lancet 395(10230), 1101–1102 (2020). [CrossRef]
48. H. Sarker, M. Faisal, and M. A. Mollah, “Slotted photonic crystal fiber-based plasmonic biosensor,” Appl. Opt. 60(2), 358–366 (2021). [CrossRef]
