We present a polymer biochip with embedded optics which allows the detection of supercritical angle fluorescence (SAF) without losses due to total internal reflection within the substrate. The chip design comprises structured spherical and aspherical optical elements on the bottom, while the top is chemically functionalized for direct binding of biomolecules. Furthermore, this design facilitates integration in lab-on-a-chip systems with appropriate microfluidics. In the confocal optical setup an ellipsoidal mirror is used for collection of SAF light above the critical angle of the water-polymer interface, which is detected by a photon-counting detector. The work presented here represents a proof of concept for performing sensitive and rapid point-of-care testing, using this low-cost, robust and disposable optical biochip platform. The performance of the platform was validated using direct binding DNA and human IgG assays which yielded low limits of detection 10pM for DNA and 10pg/ml for human IgG.
© 2013 Optical Society of America
Planar substrates are widely used for fluorescence-based immuno- and DNA assays. Glass and polymer substrates are commonly reported in the development of optical biochip systems while various polymer materials are used in commercial fluorescence plate reader instrumentation [1,2]. The excitation and detection optics can be mounted either above or below the chip and light detection from the labeled biomolecules is implemented using photomultiplier tubes, diodes or CCD cameras. It has been established that, due to the anisotropic nature of the fluorescence emitted from molecules at the interface between two dielectric media, a large proportion of the radiation is emitted into the substrate above the critical angle and so does not reach a detector which is placed either directly above or below the chip. This light constitutes the Supercritical Angle Fluorescence or SAF which is not captured in conventional detection systems which are also often characterized by low numerical aperture optics. This light can be harnessed by a suitable re-design of the solid substrate . As well as enhanced light collection, another key advantage of SAF for bioassay applications is that SAF detection allows collection of fluorescence only from molecules adjacent to the chip surface, hence minimizing background fluorescence from the bulk solution . This is an important factor in designing highly sensitive assay platforms. Some SAF strategies used in the past by the authors and others include the use of a parabolic glass or polymer elements to collect the SAF light and re-direct to the detector [5,6] and array-based collection elements . A disadvantage of the paraboloid approach reported previously  is that optical contact between the chip and the collection optics has to be made using refractive index matching oils or gels which is inconvenient for point-of-care applications and adds to the complexity of the design. In this paper, we present an improved SAF chip design, based on an injection-moulded Zeonor element, which provides a planar top surface for assay immobilization and incorporates a light collection structure on the underside of the chip which collects the SAF light in conjunction with a separate elliptical mirror which directs the light to the detector below. This design while incorporating the two main SAF features of harnessing the light which is usually lost in the substrate and discriminating between surface and bulk fluorescence, also allows for immobilization of arrays of biomolecules on the top surface and associated microfluidics, and obviates the use of index matching fluids. The key features of the paper are (i) the use of a novel, compact and robust ring-lens biochip platform which delivers enhanced sensitivity by harnessing the SAF fluorescence and (ii) validation of the chip design and performance using a model DNA assay and model IgG immunoassay assay.
2. Ring-lens biochip design and optical interrogation
A schematic diagram of the instrument is shown in Fig. 1. The main light collection spherical element in Fig. 2(a), the ring-lens chip, was fabricated by injection moulding (Protomould, UK) using Zeonor, which is an optical cyclo olefin polymer. This is shown in the expanded circle on the right of Fig. 1 and shows the underside of the Zeonor chip which contains a spherical ring structure,which ensures the collection of virtually all the unrefracted supercritical angle fluorescence from molecules immobilized on the interrogation spot of diameter ~300 µm and within ~100 nm of the upper planar surface. This ring represents a spherical segment with a radius of 4.19 mm centered on top of the chip and central angles of 123° to 160.0°. The bottom radius of the ring is 3.68 mm.
This image also shows the aspheric lens which is located in the optical centre on the bottom of the chip which focuses the Gaussian laser beam on the interrogation spot with a diameter of 300 µm on the chip surface. The asphere has a focal length of 2.3 mm and is designed to produce a homogeneous light disk of ~300 μm diameter. We calculated the surface sag for a conical constant of −0.35 and a lens radius of 0.85 mm. Due to limitations in the thickness of the moulded chip (limited to 2 mm) because of the requirement to integrate it to the centrifugal microfluidics (see below) it is not possible to use the full potential of an aspheric lens surface on the bottom side. For example, a more optimal thickness of ~5 mm for a one-sided aspheric lens would facilitate correction of a lateral mismatch of the incident laser beam by several hundreds of micrometer . In this design we can only compensate for less than 100 micrometers. This fact makes it crucial to position the chip very accurately over the laser beam, which is achieved by using precise mounts for the chips.
A laser diode (Hitachi HL6335G, 635 nm, 5mW, Thorlabs, USA) with a wavelength of 635nm is used to excite the fluorophores on the chip through an interference filter (FF01-625/26-25, Semrock Inc., USA) from below. The position of the laser is controlled by two mirrors.
The minimum size of the spherical ring structure is defined by the maximum collectable angle of the emitted supercritical angle fluorescence and is limited by the required minimal thickness of the chip as explained above and hence these geometrical and optical restrictions of the ring lens and the focal distance of the embedded lens result in a chip thickness of 2.0 mm.
The outcoupled SAF light is re-directed to the detector using a separate elliptical mirror. It was decided to use a separate mirror for this purpose rather than using an integrated injection-moulded element such as a paraboloid  due to limitations of the injection moulding process as the shape of a parabolic element does not allow an easily feasible demoulding of the chip.
The distance between chip and the rest of the optical setup needs to be adjusted as the first elliptical focus of the mirror needs to be slightly above the surface of the chip. Supercritical angle fluorescence emitted between 61.5° and 80.0° is collimated by the elliptical mirror surface and focussed through a detection aperture onto the photocathode of a photomultiplier (H8259-02 PMT module, Hamamatsu, Japan), which is run in a pre-set counting mode.
The elliptical metal mirror was fabricated by electroforming (Optiforms, USA) and is coated with silver. It has a diameter of 72.2 mm and a focal length of 4.85 mm measured from the top centre of the mirror. The distance between the two focal points is 110 mm.
A plano-convex lens (focal length = 150 mm, diameter = 50 mm) is used as a mount for a small prism mirror which is used to direct the laser beam to the centre of the ring lens chip.
This lens causes a shift of the second elliptic focus and therefore the aperture needs to be adjusted accordingly. A combination of an interference filter (FF01-676/29-25, Semrock Inc.) and detection aperture (diameter≈1.5 mm) is used to block off stray light. Both the restrictions of the ring lens and the geometry of the elliptical mirror limit the maximum collection angle to 80.0°, whereby the critical angle of water-Zeonor interface equals ~61.5 degrees. Angles below 61.5 degrees are not collected due to geometric restrictions of the chip design and the confocal assembly of the optical setup. Finally, four of the microfluidic-ringlens modules were attached to a standard CD as shown in Fig. 3(b) and mounted on a rotation stage to allow for centrifugal microfluidics for sample delivery as described below.
3. Microfluidics and assay design
3.1 Microfluidic chamber
We used centrifugal microfluidics [8–10] to flow the sample and washing solutions over the assay region of the ring lens chip. The design of the microfluidics on each chip is shown in Fig. 3(a) and comprises five chambers, two elliptical chambers for sample loading and washing using volumes from 10 to 100 µl, a central small, shallow chamber where the assay takes place and finally two flower-shaped waste chambers with ‘banana-shaped’ lateral protrusions which are designed to prevent flow back of fluids which sometimes could be observed during the spinning of chips. All chambers are interconnected with 2 mm wide channels. The waste from the sample chamber is collected in the flower-shaped chamber diagonally opposite in Fig. 3(a) while the wash chamber waste is collected in the other diagonally positioned flower-shaped chamber. This design prevents contamination of the sample region from waste fluids and also overcomes potential problems with bubbles and airlocks which sometimes are formed during the spinning of the chips. For the fabrication of the microfluidic chambers and channels we used laser-cut poly-methyl-methacrylate (PMMA) (Radionics, Ireland) and knife-cut pressure-sensitive adhesive (PSA) (Adhesives Research, Ireland) which we designed using CAD-software.
3.2 Functionalization of the Zeonor chip
The surface of the Zeonor chip was functionalized with APTES (3-amino-propyl-triethoxy-silane) using liquid phase-based coating. For liquid phase silanization with APTES, the Zeonor substrates were ultrasonically cleaned in surfactant Micro 90 (International Product Corporation, USA) and distilled water, and after drying with nitrogen, activated by oxygen plasma (40 kHz, 100 W, 0.2-1 mbar) for 10 minutes. The silianization was carried out by immersion in a 3% (v/v) solution of APTES in 95% ethanol for 2 h. The substrates were rinsed with ethanol and water, dried under nitrogen flow, and cured for 1h in an oven at 80°C. For immobilization of biomolecules, we used aldehyde-dextran as a linker which was prepared according the protocol in an earlier paper [5,11].
4 Assay design
4.1 DNA assay design
10µM amino-functionalized capture probe DNA (5′-NH2-TTCAAAATTGCGAAGTTGGG-3′) was mixed with oxidized dextran in 3xSSC (saline-sodium citrate buffer (pH 7.0, Sigma Aldrich) and incubated for 30 min at 30°C. This solution was spotted on the APTES coated chip using a sciFLEXARRAYER S3 (Scienion, Germany), a piezo driven non-contact dispensing system, forming a dense rectangular array of spots with array size 2x2 millimeter. This was left overnight in a humidity chamber at RT. Non bound capture probes were removed by washing with 2xSSC + SDS, dried and finally the chip. As a result of the reaction of the aldehyde-dextran/amino-DNA with amino-reactive APTES, surface imine bonds are formed.10µM DNA of complimentary DNA (Eurofins MWG Operon Germany) labelled with cy5 (5′-cy5-CCCAACTTCGCAATTTTGAA-3′) was used as the fluorescent target probe. Solutions of varying concentrations of target probe in 6xSSC, 0.1% SDS (sodium dodecyl sulfate, Sigma Aldritch) were heated to 45°C until the precipitated SDS was dissolved and 100µl were added to each of the sample chambers of the chips. These chips were mounted on a circular holder and this disc illustrated in Fig. 3(b) was spun while the flow of the liquids was monitored on a screen and the speed varied accordingly (120-450RPM). The centrifugal driven flow of the sample solution reached the detection area of the chip and the DNA strands hybridised with the surface bound capture probes. The spinning of the chips was continued until the whole solution reached the waste chamber. For washing, the second channel on the chip was used by turning the chip 45 degrees anti-clockwise. Subsequently we washed the chips by spinning with firstly 2xSSC, 0.2% SDS, secondly with 2xSSC and finally with 0.2xSSC. After these washing steps, the fluorescence signal of the chips was measured on the SAF instrument.
4.2 Human-IgG assay design
The IgG-antibody (10 µg/ml) (Sigma Aldrich), was mixed with 2 vol% oxidized dextran, 1 vol% trehalose (Sigma Aldrich) in 50 mM NaPO4 buffer (pH 7.5, Sigma Aldrich) and incubated for 30 minutes at 40°C. An array of IgG-antibodies/aldehyde-dextran was printed on the APTES functionalized chip using a protocol similar to that described for the DNA assay in the previous section. This was left in a humidity chamber 1h at 37°C. Afterwards, we blocked the surface with BSA (1 w/w%) for 1hr at 37°C, washed with PBS/Tween and dried the chip under a nitrogen stream. In the next step, we attached the ring lens chip to the microfluidic chip. We added different concentrations of antigen human IgG (Sigma Aldrich) in 1w/w% BSA and PBS/Tween to the sample chambers. The chips were mounted on the circular holder shown in Fig. 3(b) and spun while the flow of the liquids was monitored on a screen and the speed varied as described in the previous section. For detection, we added 200nM IgG-antibody labeled with cy5 and spun again until the fluid has reached the waste chamber. As above, for washing, the second channel on the chip was used by turning the chip 45 degrees anti-clockwise. Subsequently we washed the chips twice with PBS/Tween and once with PBS. At this stage, the fluorescence was measured on the SAF instrument.
The graphs in Fig. 4 show a limit of detection (LOD), based on 3σ standard deviation for the IgG assay was measured as 10 pg/mL while that of the DNA assay was 10 pM. It has already been established that, with the advantages of greater light collection efficiency and surface selectivity, SAF assays have enhanced performance compared to planar assays of which Elisa assays represent the state-of-the-art. Currently, we are among a limited number of groups developing SAF-based assay platforms. The group of Ruckstuhl has published on the performance of a number of SAF biochip platforms for example using a glass parabolic element  where analyte concentrations down 28.5 pg/ml were detected but the platform involved the use of a glass SAF element and used refractive index matching oil which represents a less convenient and less versatile approach than that presented here. This group has recently published a novel assay platform which uses polymer test tubes. The bottom of these tubes is modified to incorporate an embedded moulded SAF element. They have achieved LODs of 0.27 pM (4.5 pg/mL for an Interleukin 2 assay and 1.9 pM (30 pg/ml) for an Interferon gamma assay . This high performance SAF platform is quite different to the one presented here but both approaches complement each other, our platform allowing for a planar platform but with embedded SAF while facilitating conventional microfluidic approach and achieving similar LOD performance.
We have described a low-cost, supercritical angle fluorescence chip reader for rapid biochip analysis which is suitable for point of care testing. This design has advantages over previously published parabolic element-based platforms in that it allows for efficient SAF collection underneath the chip while facilitating simple microfluidics on the top surface on which arrays of biomolecules can be printed. In particular, the combination of the high collection efficiency and the surface selectivity of the SAF detection technique yield a generic LOD of 10 pg/ml for human IgG and 10 pM for DNA. These results are comparable to those achieved on commercial bench-top instrumentation and also compare well with previous assays using SAF detection. The results demonstrate the potential of the SAF-based, compact and portable biochip platform for point-of-care biomedical applications. The presented microfluidic chip only shows one possible method to carry out an assay using a chip with ring structures. An alternative platform has been shown recently showing a serial siphon-enabled microfluidic disc for automated sequential reagent delivery . Future work will explore the feasibility of incorporating the ring lens structures and microfluidic channels in an integrated, moulded centrifugal platform.
This work was supported by the Science Foundation Ireland. (Grant Nos. 05/CE3/B754, 05/CE3/B754S6 and 10/CE/B1821).
References and links
1. T. I. Koshy and K. F. Buechler, “The Triage® System,” in The Immunoassay Handbook, D. Wild, eds. (Elsevier Science, 2013), pp. 541–544.
2. R. Bingisser, C. Cairns, M. Christ, P. Hausfater, B. Lindahl, J. Mair, M. Panteghini, Ch. Price, and P. Venge, “Cardiac troponin: a critical review of the case for point-of-care testing in the ED,” Am. J. Emerg. Med. 30(8), 1639–1649 (2012). [CrossRef] [PubMed]
3. J. S. Yuk, M. Trnavsky, C. McDonagh, and B. D. MacCraith, “Surface plasmon-coupled emission (SPCE)-based immunoassay using a novel paraboloid array biochip,” Biosens. Bioelectron. 25(6), 1344–1349 (2010). [CrossRef] [PubMed]
5. D. Kurzbuch, J. Bakker, J. Melin, Ch. Jonsson, T. Ruckstuhl, and B. MacCraith, “A biochip reader using super critical angle fluorescence,” Sens. Actuators B Chem. 137(1), 1–6 (2009). [CrossRef]
8. R. Gorkin 3rd, C. E. Nwankire, J. Gaughran, X. Zhang, G. G. Donohoe, M. Rook, R. O’Kennedy, and J. Ducrée, “Centrifugo-pneumatic valving utilizing dissolvable films,” Lab Chip 12(16), 2894–2902 (2012). [CrossRef] [PubMed]
9. J. Ducrée, S. Haeberle, S. Lutz, S. Pausch, F. Stetten, and R. Zengerle, “The centrifugal microfluidic Bio-Disk platform,” J. Micromech. Microeng. 17(7), S103–S115 (2007). [CrossRef]
10. J. Ducrée, “Next-generation microfluidic lab-on-a-chip platforms for point-of-care diagnostics and systems biology,” Procedia Chem. 1(1), 517–520 (2009). [CrossRef]
11. C. Jönsson, M. Aronsson, G. Rundström, C. Pettersson, I. Mendel-Hartvig, J. Bakker, E. Martinsson, B. Liedberg, B. MacCraith, O. Ohman, and J. Melin, “Silane-dextran chemistry on lateral flow polymer chips for immunoassays,” Lab Chip 8(7), 1191–1197 (2008). [CrossRef] [PubMed]
12. C. E. Nwankire, G. G. Donohoe, X. Zhang, J. Siegrist, M. Somers, D. Kurzbuch, R. Monaghan, M. Kitsara, R. Burger, S. Hearty, J. Murrell, Ch. Martin, M. Rook, L. Barrett, S. Daniels, C. McDonagh, R. O’Kennedy, and J. Ducrée, “At-Line bioprocess monitoring by immunoassay with rotationally controlled serial siphoning and integrated supercritical angle fluorescence optics,” Anal. Chim. Acta 781, 54–62 (2013). [CrossRef] [PubMed]