In-Parallel Polar Monitoring of Chemiluminescence Emission Anisotropy at the Solid–Liquid Interface by an Optical Fiber Radial Array

: Chemiluminescence (CL) detection is widely employed in biosensors and miniaturized analytical devices since it o ﬀ ers high detectability and ﬂexible device design (there are no geometry requirements for the measurement cell, except the ability to collect the largest fraction of emitted photons). Although the emission anisotropy phenomenon for an emitting dipole bound to the interface between two media with di ﬀ erent refractive index is well known for ﬂuorescence, it is still poorly investigated for CL reactions, in which the excited-state reaction products can di ﬀ use in solution before the photon emission event. In this paper, we propose a simple method for the real-time evaluation of the CL emission anisotropy based on a radial array of optical ﬁbers, embedded in a poly(methyl methacrylate) semicylinder and coupled with a Charge-Coupled Device (CCD) camera through a suitable interface. The polar-time evolutions of the CL emission have been studied for catalyzing enzymes immobilized onto a solid surface (heterogeneous conﬁguration) or free in solution (homogeneous conﬁguration). Evidence of the anisotropy phenomenon is observed, indicating that the lifetime of the excited-state products of the enzyme-catalyzed reactions is shorter than the time required for their di ﬀ usion in solution at a distance at which the CL can be considered isotropic. These results open new perspectives in the development of CL-based miniaturized analytical devices.


Introduction
Chemiluminescence (CL, i.e., the light emission originating from a strongly exergonic chemical reaction yielding a product in its electronically excited state) is an optical detection method which, due to its intrinsic characteristics, is gaining a growing interest in the development of high-performance biosensing devices [1][2][3]. In fact, when compared to other optical detection principles, such as absorption or fluorescence, CL offers considerable advantages including a high detectability even in low volumes, due to the possibility of producing photons "in the dark", thus minimizing any other nonspecific signal contribution coming from excitation photons and maximizing the resulting signal-to-noise ratio. As the CL light signal is the result of a chemical reaction, no external source, together with the connected filtering apparatus, is requested. All these features can lead to the development of  Figure 1 sketches the experimental setup adopted for collecting the CL signal and studying its behavior at the interface in terms of emission anisotropy (i.e., polar distribution of the emission). All transparent elements were made of PMMA, with a mean refractive index n PMMA = 1.50 in the 420-480 nm wavelength interval (the HRP-and ALP-catalyzed CL emissions have maxima at about 425 and 470 nm, respectively).

Optical Setup and Methods
Chemosensors 2020, 8 Figure 1 sketches the experimental setup adopted for collecting the CL signal and studying its behavior at the interface in terms of emission anisotropy (i.e., polar distribution of the emission). All transparent elements were made of PMMA, with a mean refractive index nPMMA = 1.50 in the 420-480 nm wavelength interval (the HRP-and ALP-catalyzed CL emissions have maxima at about 425 and 470 nm, respectively). Multimode PMMA optical fibers Broadcom HFBR-RUS100Z (Broadcom, San Jose, CA, USA) were used, characterized by a diameter of 1 mm and a numerical aperture (NA) of 0.47 and cut in pieces 60 cm in length. The ends of each fiber were accurately polished by means of a lapping machine, using aluminum oxide and silicon carbide abrasive disks with different grind size from 12 μm down to 0.3 μm (Buehler FibrMet Disc, PSA Backed, 4''). All Buehler products were obtained from Buehler (Lake Bluff, IL, USA).

Optical Setup and Methods
A PMMA semicylinder was obtained by cutting a PMMA rod (35 mm in length, 40 mm in diameter) along a plane parallel to the cylinder axis and 1 mm far from it (see Figure 2). The obtained rectangular flat top surface (35 mm × 40 mm) was manually polished by using a nonabrasive cloth (Buehler microcloth PSA 2-7/8'') and colloidal diamond suspension (Buehler MetaDi TM Supreme-Polycrystalline Diamond Suspension) with a grit of 0.25 μm. The semicylinder block was then drilled in correspondence with its circular outer surface in order to realize eleven radial cylindrical holes (1.1 mm in diameter, 4 mm in depth) with a constant azimuthal angular spacing of 15°, necessary to house the input ends of the optical fibers used for CL signal collection. The fibers were glued inside the holes with the NOA 68 optical glue (Norland Products Inc., Cranbury, NJ, USA).
A PMMA slide (microfluidic ChipShop, Jena, Germany), with a thickness of 1 mm, was set on the top of the PMMA semicylinder and used as support for the CL reaction. With reference to Figure 2, considering that the height of the semicylinder is 19 mm (i.e., 1 mm less than the cylinder radius), the upper surface of the 1-mm-thick slide will be coincident with the cylinder axial plane. During the experiments, a refractive index matching oil (pure glycerol from Sigma-Aldrich, St. Louis, MO, USA, with a refractive index nOIL = 1.48) was used to improve the optical coupling condition of the PMMA slide with the underlying PMMA semicylinder. Multimode PMMA optical fibers Broadcom HFBR-RUS100Z (Broadcom, San Jose, CA, USA) were used, characterized by a diameter of 1 mm and a numerical aperture (NA) of 0.47 and cut in pieces 60 cm in length. The ends of each fiber were accurately polished by means of a lapping machine, using aluminum oxide and silicon carbide abrasive disks with different grind size from 12 µm down to 0.3 µm (Buehler FibrMet Disc, PSA Backed, 4"). All Buehler products were obtained from Buehler (Lake Bluff, IL, USA).
A PMMA semicylinder was obtained by cutting a PMMA rod (35 mm in length, 40 mm in diameter) along a plane parallel to the cylinder axis and 1 mm far from it (see Figure 2). The obtained rectangular flat top surface (35 mm × 40 mm) was manually polished by using a nonabrasive cloth (Buehler microcloth PSA 2-7/8") and colloidal diamond suspension (Buehler MetaDi TM Supreme-Polycrystalline Diamond Suspension) with a grit of 0.25 µm. The semicylinder block was then drilled in correspondence with its circular outer surface in order to realize eleven radial cylindrical holes (1.1 mm in diameter, 4 mm in depth) with a constant azimuthal angular spacing of 15 • , necessary to house the input ends of the optical fibers used for CL signal collection. The fibers were glued inside the holes with the NOA 68 optical glue (Norland Products Inc., Cranbury, NJ, USA).
A PMMA slide (microfluidic ChipShop, Jena, Germany), with a thickness of 1 mm, was set on the top of the PMMA semicylinder and used as support for the CL reaction. With reference to Figure 2, considering that the height of the semicylinder is 19 mm (i.e., 1 mm less than the cylinder radius), the upper surface of the 1-mm-thick slide will be coincident with the cylinder axial plane. During the experiments, a refractive index matching oil (pure glycerol from Sigma-Aldrich, St. Louis, MO, USA, The optical configuration allowed the maximum efficiency of light collection by optical fibers. Indeed, if r = 0.5 mm is the radius of the optical fiber and D = 16 mm is the distance between the fiber face and the CL drop, the acceptance angle, or angular resolution, for every fiber, assuming a point emitter placed on the semicylinder axis, can be calculated as: This value of the acceptance angle is under the numerical aperture of the fiber, so all the light reaching the fiber input end will be conveyed to the CCD camera. The eleven optical fiber output ends were fixed in the holes of a support obtained by drilling a parallelepiped of black polyvinylchloride (PVC) polymer (17 mm × 13 mm × 10 mm, see Figure 1) to obtain a 4 × 3 holes matrix (1.1 mm diameter, 5 mm spacing). Such support acted as an interface between the collecting system (PMMA semicylinder and plastic fibers) and the imaging detector (cooled MZ-2PRO CCD camera, MagZero, Pordenone, Italy) equipped for lensless contact imaging through a round fiber optic taper (25 to 11 mm, Edmund Optics, Barrington, NJ, USA) placed in contact with the CCD sensor [33]. Since each position in the 4 × 3 holes matrix of the support corresponded to a specific optical fiber, the CL emissions at different angles can be acquired independently and simultaneously (the twelfth position of the matrix was used to measure the background signal of the CCD camera). The measurements were performed inside a black box to avoid any spurious light contribution from the environment that could interfere with the measurement.

Reagents and Reaction Protocols
The enzymes HRP (type VI-A, from horseradish) and ALP (from bovine intestinal mucosa) were purchased from Sigma-Aldrich. Enzyme solutions were prepared in 0.1 M phosphate-buffered saline (PBS), pH 7.7. The Super Signal ELISA Femto CL cocktail for HRP was purchased from Thermo Fisher Scientific (Rockford, IL, USA), while the Lumiphos Plus CL cocktail for ALP was from Lumigen (Southfield, MI, USA). Figure 3 illustrates the heterogeneous and homogeneous configurations adopted in the experiments for measuring the polar distribution of the emission from enzyme-catalyzed CL reactions. For enzyme immobilization (heterogeneous configuration), the PMMA slides were first functionalized by one-minute dip-coating in a 2 mM Eudragit L100 copolymer (Evonik Degussa GmbH, Düsseldorf, Germany) ethanol solution, then air-dried to form a polymeric thin film on the surface. Eudragit L100 is an anionic copolymer made of methacrylic acid and methyl methacrylate used to provide carboxylic (-COOH) functional groups for the binding of biospecific molecules. The slide surfaces were activated with 2 mM 1-ethyl-3-[3-dimethylaminopropyl] carbodiimide hydrochloride (EDC) and 5 mM N-hydroxysuccinimide (NHS) (Thermo Fisher Scientific). Then, 1 μL of 1 μg/mL enzyme solution (HRP or ALP) was deposited on the activated PMMA slide surface and incubated for 1 h in a humid chamber at room temperature. The slides were then thoroughly washed with PBS and dried under a nitrogen stream. The optical configuration allowed the maximum efficiency of light collection by optical fibers. Indeed, if r = 0.5 mm is the radius of the optical fiber and D = 16 mm is the distance between the fiber face and the CL drop, the acceptance angle, or angular resolution, for every fiber, assuming a point emitter placed on the semicylinder axis, can be calculated as: This value of the acceptance angle is under the numerical aperture of the fiber, so all the light reaching the fiber input end will be conveyed to the CCD camera.
The eleven optical fiber output ends were fixed in the holes of a support obtained by drilling a parallelepiped of black polyvinylchloride (PVC) polymer (17 mm × 13 mm × 10 mm, see Figure 1) to obtain a 4 × 3 holes matrix (1.1 mm diameter, 5 mm spacing). Such support acted as an interface between the collecting system (PMMA semicylinder and plastic fibers) and the imaging detector (cooled MZ-2PRO CCD camera, MagZero, Pordenone, Italy) equipped for lensless contact imaging through a round fiber optic taper (25 to 11 mm, Edmund Optics, Barrington, NJ, USA) placed in contact with the CCD sensor [33]. Since each position in the 4 × 3 holes matrix of the support corresponded to a specific optical fiber, the CL emissions at different angles can be acquired independently and simultaneously (the twelfth position of the matrix was used to measure the background signal of the CCD camera). The measurements were performed inside a black box to avoid any spurious light contribution from the environment that could interfere with the measurement.

Reagents and Reaction Protocols
The enzymes HRP (type VI-A, from horseradish) and ALP (from bovine intestinal mucosa) were purchased from Sigma-Aldrich. Enzyme solutions were prepared in 0.1 M phosphate-buffered saline (PBS), pH 7.7. The Super Signal ELISA Femto CL cocktail for HRP was purchased from Thermo Fisher Scientific (Rockford, IL, USA), while the Lumiphos Plus CL cocktail for ALP was from Lumigen (Southfield, MI, USA). Figure 3 illustrates the heterogeneous and homogeneous configurations adopted in the experiments for measuring the polar distribution of the emission from enzyme-catalyzed CL reactions. For enzyme immobilization (heterogeneous configuration), the PMMA slides were first functionalized by one-minute dip-coating in a 2 mM Eudragit L100 copolymer (Evonik Degussa GmbH, Düsseldorf, Germany) ethanol solution, then air-dried to form a polymeric thin film on the surface. Eudragit L100 is an anionic copolymer made of methacrylic acid and methyl methacrylate used to provide carboxylic (-COOH) functional groups for the binding of biospecific molecules. The slide For the measurement, the slide was positioned on the PMMA semicylinder with the spot of immobilized enzyme placed in correspondence with the focus of the optical collecting system, univocally identified by the axis of the PMMA semicylinder and the plane of the fibers. Then, 10 µ L of CL cocktail (Super Signal ELISA Femto or Lumiphos Plus for HRP or ALP, respectively) were added and CL images were acquired.
A similar procedure was adopted for the measurements in the homogeneous configuration, except for the fact that pristine PMMA slides were used. For each measurement, 1 μL of 10 ng/mL HRP enzyme solution and 9 μL of Super Signal ELISA Femto CL cocktail were dispensed on the slide in correspondence with the focus of the optical collecting system and CL images were acquired.

Signal Acquisition and Processing
Images (grayscale images, 16-bit depth in the FITS file format) were acquired at regular time intervals employing a 5-s integration time and analyzed with the freely available open source ImageJ software [34]. The overall signal of each spot (corresponding to a fiber output end and, therefore, proportional to the CL emission intensity at the polar angle of the fiber) was evaluated by integration over the whole spot area. An example of a frame is reported in Figure 4. For the measurement, the slide was positioned on the PMMA semicylinder with the spot of immobilized enzyme placed in correspondence with the focus of the optical collecting system, univocally identified by the axis of the PMMA semicylinder and the plane of the fibers. Then, 10 µL of CL cocktail (Super Signal ELISA Femto or Lumiphos Plus for HRP or ALP, respectively) were added and CL images were acquired.
A similar procedure was adopted for the measurements in the homogeneous configuration, except for the fact that pristine PMMA slides were used. For each measurement, 1 µL of 10 ng/mL HRP enzyme solution and 9 µL of Super Signal ELISA Femto CL cocktail were dispensed on the slide in correspondence with the focus of the optical collecting system and CL images were acquired.

Signal Acquisition and Processing
Images (grayscale images, 16-bit depth in the FITS file format) were acquired at regular time intervals employing a 5-s integration time and analyzed with the freely available open source ImageJ software [34]. The overall signal of each spot (corresponding to a fiber output end and, therefore, proportional to the CL emission intensity at the polar angle of the fiber) was evaluated by integration over the whole spot area. An example of a frame is reported in Figure 4.
The twelfth position of the matrix, indicated in Figure 4 as BG-σ r , was used to measure the background signal of the CCD camera and for the estimation of the statistical relative error σ r of the measurements, calculated as the standard deviation-average signal ratio of said point.

Signal Acquisition and Processing
Images (grayscale images, 16-bit depth in the FITS file format) were acquired at regular time intervals employing a 5-s integration time and analyzed with the freely available open source ImageJ software [34]. The overall signal of each spot (corresponding to a fiber output end and, therefore, proportional to the CL emission intensity at the polar angle of the fiber) was evaluated by integration over the whole spot area. An example of a frame is reported in Figure 4.

Chemiluminescence Systems
Chemiluminescent measurements were performed by using the CL reactions catalyzed by HRP or ALP enzymes, which are widely used in bioanalytical applications, in particular in enzyme-based assays and in binding assays such as gene probe hybridization and immunoassays.
Calf intestinal ALP is a 160 kDa enzyme often detected by means of 1,2-dioxetane derivatives as enzymatic substrates. In this work, the Lumiphos Plus CL cocktail for ALP containing 4-methoxy-4-(3-phosphatephenyl)spiro[1,2-dioxetane-3,2 -adamantane] disodium salt (AMPPD) and an enhancer (which increases the duration and intensity of the emission) was employed. The catalytic action of the enzyme on AMPPD results in dephosphorylation of the aryl phosphate moiety of the dioxetane with subsequent cleavage of the 1,2-dioxetane ring and production of a phenolate moiety in its excited state [35].
Horseradish peroxidase, which is largely employed in bioanalytical assays owing to its rather small dimension (44 kDa) and high turnover rate, catalyzes the oxidation reaction of luminol, which produces 3-aminophthalate in its excited state. In this work, a commercial CL cocktail containing an N-alkylated phenothiazine enhancer and 4-morpholinopyridine as the acylation catalyst was employed. These molecules are added to the system to increase the enzyme turnover number and the equilibrium concentration of the key intermediate luminol radical anion, providing more intense and prolonged light emission [36].

Heterogeneous Configuration
In the heterogeneous configuration, the enzyme, either ALP or HRP, was immobilized onto the Eudragit functionalized PMMA slide. A drop of CL reaction cocktail was then deposited on the surface of the slide in correspondence with the axis of the optical collecting system (Figure 2).
A representative CL image acquired by the CCD camera is depicted in Figure 4. The elliptical shape of the spots is generated by a slight distortion introduced by the fiber taper adapter, considering also that the end of the taper is about 2 mm distant from the camera sensor. Figure 5a shows the kinetic profile of the emission obtained for ALP immobilized on the PMMA surface (i.e., the sum of the signals collected by the eleven fibers). The error bars were calculated considering the relative standard deviation σ r = 1% calculated on the 15 samples acquired in the BG-σ r point. The intensity of the CL emission increases with time in accordance to the typical kinetics of the CL reaction of 1,2-dioxetane substrates catalyzed by ALP. Indeed, such reactions usually show slow emission kinetics, and the CL signal often reaches a plateau 20-30 min after the start of the reaction [37]. A similar behavior can be observed in Figure 5b where the CL emissions at the various angles (i.e., the average of the signals measured for each couple of fibers at symmetrical angles) are reported as a function of time. In the case of immobilized HRP, the kinetic profile of the overall CL emission (Figure 6a) is quite different, since the maximum emission intensity is immediately reached, then the signal decreases with time. Again, the behavior is in line with the typical kinetics of the luminol/peroxide/enhancer CL system in the presence of HRP, characterized by a rapid onset of the CL emission followed by a gradual decrease due to reagent consumption and enzyme degradation caused by the CL radical reaction [37].  Figure 5b clearly shows that the intensity of the CL emission also depends on the polar angle. Indeed, the CL emission is strongly anisotropic (Figure 5c), with maxima at ±60 • angles that are very close to the critical angle of the PMMA-buffer interface (θ c,PMMA−bu f f er = sin −1 (1.34/1.50) 63 • ). This indicated that the CL emission occurs, on average, close to the PMMA surface, in particular within a distance from the surface less than or comparable with the CL emission wavelength (i.e., around 470 nm). Thus, the lifetime of the excited product of the enzyme-catalyzed reaction is significantly shorter than the time required for its diffusion far from the surface. As expected, the polar distribution remains practically unchanged over time, demonstrating that the CL emission occurs close to the PMMA surface for the entire duration of the reaction.
In the case of immobilized HRP, the kinetic profile of the overall CL emission (Figure 6a) is quite different, since the maximum emission intensity is immediately reached, then the signal decreases with time. Again, the behavior is in line with the typical kinetics of the luminol/peroxide/enhancer CL system in the presence of HRP, characterized by a rapid onset of the CL emission followed by a gradual decrease due to reagent consumption and enzyme degradation caused by the CL radical reaction [37]. The kinetic profiles of the CL signals measured at the two symmetrical angles are reported in Figure 6b, while the resulting polar distribution of the emission at different times is depicted in Figure 6c. The error bars were calculated considering the relative standard deviation σr = 1% calculated for the 18 samples acquired in the BG-σr point. As already observed for the immobilized ALP, the CL emission is markedly anisotropic and the maximum emission is at ±60° angles, again confirming the proximity of the CL emitters to the PMMA surface. The kinetic profiles of the CL signals measured at the two symmetrical angles are reported in Figure 6b, while the resulting polar distribution of the emission at different times is depicted in Figure 6c. The error bars were calculated considering the relative standard deviation σ r = 1% calculated for the 18 samples acquired in the BG-σ r point. As already observed for the immobilized ALP, the CL emission is markedly anisotropic and the maximum emission is at ±60 • angles, again confirming the proximity of the CL emitters to the PMMA surface. Figure 7 shows the comparison between the polar distributions of the CL emissions observed for immobilized ALP and HRP. It is evident that the emissions display the same anisotropy independently from the catalyzing enzyme and the absolute intensity of the emission.

Homogeneous Configuration
The CL emission in the homogeneous configuration was investigated to demonstrate that the employed measurement system was able to correctly differentiate between isotropic and anisotropic CL emission. In this configuration, the CL reaction cocktail containing the enzyme HRP was deposited onto pristine PMMA slide in the same position defined by the axis of the optical detection system.
As far as the HRP in solution is concerned, the kinetic profile of the CL emission is quite similar to that of the immobilized HRP ( Figure 8a). However, the kinetic profiles of the CL signals measured at the two symmetrical angles (Figure 8b) and the resulting polar distribution of the emission at different times ( Figure 8c) display a different behavior. The error bars were calculated considering the relative standard deviation σr = 1% calculated for the 18 samples acquired in the BG-σr point. Indeed, as expected, the CL emission is almost isotropic at short reaction times, according to the fact that the CL emitters produced by the chemical reaction are homogeneously distributed in the solution volume. However, with increasing time, the intensity of the CL emission decreases while its polar distribution evolves to anisotropy, becoming similar to that observed for HRP in the heterogeneous configuration. This can be reasonably due to the gradual absorption of HRP onto the PMMA surface, which moves the average reaction/emission volume closer to the surface. Indeed, it has been reported that protein adsorption on the surface of pristine PMMA, which occurs through interaction of the hydrophobic domains of the protein with the polymer surface, is quite fast, reaching a significant degree of surface coverage after only 10 min of incubation [38,39]. This observation demonstrated the possibility to use the developed system to monitor the absorption of the HRP enzyme on a surface in real time.

Homogeneous Configuration
The CL emission in the homogeneous configuration was investigated to demonstrate that the employed measurement system was able to correctly differentiate between isotropic and anisotropic CL emission. In this configuration, the CL reaction cocktail containing the enzyme HRP was deposited onto pristine PMMA slide in the same position defined by the axis of the optical detection system.
As far as the HRP in solution is concerned, the kinetic profile of the CL emission is quite similar to that of the immobilized HRP ( Figure 8a). However, the kinetic profiles of the CL signals measured at the two symmetrical angles (Figure 8b) and the resulting polar distribution of the emission at different times ( Figure 8c) display a different behavior. The error bars were calculated considering the relative standard deviation σ r = 1% calculated for the 18 samples acquired in the BG-σ r point. Indeed, as expected, the CL emission is almost isotropic at short reaction times, according to the fact that the CL emitters produced by the chemical reaction are homogeneously distributed in the solution volume. However, with increasing time, the intensity of the CL emission decreases while its polar distribution evolves to anisotropy, becoming similar to that observed for HRP in the heterogeneous configuration. This can be reasonably due to the gradual absorption of HRP onto the PMMA surface, which moves the average reaction/emission volume closer to the surface. Indeed, it has been reported that protein adsorption on the surface of pristine PMMA, which occurs through interaction of the hydrophobic domains of the protein with the polymer surface, is quite fast, reaching a significant degree of surface coverage after only 10 min of incubation [38,39]. This observation demonstrated the possibility to use the developed system to monitor the absorption of the HRP enzyme on a surface in real time.

Conclusions
In this manuscript, an optical detection methodology for the real-time polar monitoring of CL emission has been described. The optical device consisted of a PMMA semicylinder with a radial array of eleven collecting optical fibers. The output ends of the fibers were faced to a CCD camera, which allowed the single-shot acquisition of the optical signals from all fibers. In this way, it was possible to evaluate the polar distribution of the CL emission from a chemical reaction that took place at the liquid-solid boundary and its evolution with time. With this methodology, it is worth

Conclusions
In this manuscript, an optical detection methodology for the real-time polar monitoring of CL emission has been described. The optical device consisted of a PMMA semicylinder with a radial array of eleven collecting optical fibers. The output ends of the fibers were faced to a CCD camera, which allowed the single-shot acquisition of the optical signals from all fibers. In this way, it was possible to evaluate the polar distribution of the CL emission from a chemical reaction that took place at the liquid-solid boundary and its evolution with time. With this methodology, it is worth observing that the angular resolution can be easily improved by increasing the number of fibers and/or decreasing the core diameter of each fiber.
Two different configurations have been studied: heterogeneous (HRP or ALP enzymes bound on a PMMA surface) and homogeneous (HRP enzyme free in solution). Experimental results showed a markedly anisotropic emission in the heterogeneous configuration, confirming that the photon emission resulting from the decay of the excited-state products of the CL reaction catalyzed by immobilized enzymes occurred close to the PMMA surface. It is worth noting that this behavior might not be confirmed for other CL systems, being dependent on the lifetime of intermediate products of the CL reaction [30]. Additionally, the presence of enhancers in the CL substrate could also affect the polar behavior of the emission (e.g., by introducing energy acceptors that represent the actual emitting species of the CL reaction). For the homogeneous configuration, the CL emission was almost isotropic at short reaction times, while evolving towards an anisotropic emission over time. This suggested that absorption of the enzyme on the PMMA surface took place during the measurement, bringing the reaction volume closer and closer to the PMMA surface.
The observed anisotropic behavior of the CL emission occurring when the enzymes catalyzing the CL reaction are in close proximity to the boundary between two dielectric media with different refractive indexes opens new perspectives in the design of CL biosensors. This phenomenon must be taken into consideration when designing miniaturized CL-based analytical devices or even exploited to improve the signal collection efficiency. Indeed, in many biosensor configurations the detector is placed close to the bottom part of a transparent substrate on which the CL reaction takes place. With this setup, a large fraction of emitted photons remains undetected, thus greatly reducing assay detectability. On the contrary, anisotropic coupling of the CL emission to a transparent substrate could be used to guide the emitted photons towards the light detector, thus improving light collection efficiency without using dedicated optics.