Sodium Alginate Cross-Linkable Planar 1D Photonic Crystals as a Promising Tool for Pb2+ Detection in Water

Due to its high toxicity, Pb2+ pollution is a serious threat for human health and environments. However, in situ real-time detection of Pb2+ pollution is difficult and laboratory instruments are usually required. Then, the possibility to monitor water quality without laboratory instruments could lead to the extensive assessment of polluted water sources, especially in rural environments and developing countries where large lead concentrations are often found in surface water. Consequently, new simple colorimetric sensors are highly interesting in the field. In this work we report for the first time disposable polymer planar 1D photonic crystals made of poly (N-vinylcarbazole) as high refractive index medium and sodium alginate as low refractive index and active medium for the detection of Pb2+ in water. The detection relies on the ionic exchange occurring into the alginate matrix. This process effectively induces a physical cross-linking phenomenon, which inhibits water solubilization of the polymer. In turn, this affects the spectral response of the planar 1D photonic crystals modifying its color.


Introduction
Lead is a toxic poisonous substance for humans and animals. In recent years, several Pb 2+ leaches from water piping to drinking water made it an important health threat, especially for elder people and infants [1,2]. Indeed, Pb 2+ exposure causes blood and nervous disorders [3], as it is easily complexed by biomolecules containing oxygen, nitrogen and sulfur. This complexation can affect protein structures and enzymatic processes [4]. In addition, lead is a great threat also for animal and plant life [5,6]. Nowadays, national regulations limit the content of heavy metals in drinking water. According to European and United States drinking water directive, the maximum amount of lead allowed in water for human consumption is 10 µg/L in the European Union [7] and 15 µg/L in the United States. In developing countries, these limits are hardly respected due to a wide series of industrial/social reasons. In this regard, several recent works showed that lead concentration in waste water [8,9], natural rivers [10,11] and even drinking water [12] reaches maximum values of milligrams per liter, well above the limits and not compatible with human and animal life. For example, Pb 2+ concentration in the Ganges river in India can approach 27 mg/L (10 −4 M), depending on the investigation site and season [13]. Consequently, the possibility to semi-quantitively detect lead with a fast, simple and cost-effective sensor represents an important milestone. Unfortunately, sampling is often not representative, especially for different timeframes. Moreover, costly and time demanding laboratory analyses makes real time monitoring of pollution nearly impossible [14,15]. These data makes it clear In this work we employ SA in spun-cast polymer DBRs for the detection of Pb 2+ ions in water in a concentration range commonly found in polluted surface water. In details, we demonstrate the use of SA together with poly(N-vinylcarbazole) as DBR building block for the first time. After the optical characterization of the new planar 1D photonic crystal, we test and analyze its dynamic behavior in water polluted with Pb 2+ with different concentrations. All the characterizations are repeated for two DBR structures with different number of periods to confirm the optical behavior of the system.

Materials and Methods
DBRs were fabricated by the alternating spin-coating of SA (Sigma Aldrich, MV = 400 kDa, M/G 1.5) solution in deionized water with a concentration of 15 mg/mL (containing Triton X-100 with a concentration of 0.1% v/v to reduce the solution surface tension and increase the processability) and of PVK (Acros Organics, Mw = 40,000 kDa) solution in toluene with a concentration of 30 mg/mL. In a typical experiment, 100 μL of each solution were alternatively casted on a glass substrate and dried for 120 s at a rotation speed of 140 round per second to obtain 6.5 or 10.5 bilayers; first and last layers are always made of PVK. Once fabricated, the samples were baked for 120 s at 60°C on a hotplate to remove any residual solvent. The samples were cut into portion and exposed to Pb 2+ solutions (PbBr2, Sigma Aldrich) in deionized water with concentration ranging from 1 × 10 −5 M to 1 × 10 −2 M in an aluminum cell connected to a reflectance fiber probe. The optical response of the DBRs was evaluated comparing the reflectance spectra collected after the first thermal annealing and after the desorption. Both the measurements were performed at room temperature. The optical response of the sensors was sampled at set intervals along the entire cation exposure using a halogen-deuterium DH-BAL-2000 light source and an CMOS spectrometer Avantes-2048-Evo for the detection. Steady state response was finally investigated after drying the sample in air at 60°C. The reflectance spectra of the DBR structure were modeled using a transfer matrix method formalism as described in Ref. [32] using the refractive index dispersion of SA and PVK, which were retrieved from literature as input (nPVK,500 nm = 1.68, nSA, 500 nm = 1.51, Ref. [87][88][89]) and the geometric thickness of the layers as fitting parameters. The layer thicknesses were retrieved for the sample made of 10.5 periods and employed for the modeling of the structure made of 6.5 bilayers fabricated using the same conditions. In this work we employ SA in spun-cast polymer DBRs for the detection of Pb 2+ ions in water in a concentration range commonly found in polluted surface water. In details, we demonstrate the use of SA together with poly(N-vinylcarbazole) as DBR building block for the first time. After the optical characterization of the new planar 1D photonic crystal, we test and analyze its dynamic behavior in water polluted with Pb 2+ with different concentrations. All the characterizations are repeated for two DBR structures with different number of periods to confirm the optical behavior of the system.

Materials and Methods
DBRs were fabricated by the alternating spin-coating of SA (Sigma Aldrich, MV = 400 kDa, M/G 1.5) solution in deionized water with a concentration of 15 mg/mL (containing Triton X-100 with a concentration of 0.1% v/v to reduce the solution surface tension and increase the processability) and of PVK (Acros Organics, Mw = 40,000 kDa) solution in toluene with a concentration of 30 mg/mL. In a typical experiment, 100 µL of each solution were alternatively casted on a glass substrate and dried for 120 s at a rotation speed of 140 round per second to obtain 6.5 or 10.5 bilayers; first and last layers are always made of PVK. Once fabricated, the samples were baked for 120 s at 60 • C on a hotplate to remove any residual solvent. The samples were cut into portion and exposed to Pb 2+ solutions (PbBr 2 , Sigma Aldrich) in deionized water with concentration ranging from 1 × 10 −5 M to 1 × 10 −2 M in an aluminum cell connected to a reflectance fiber probe. The optical response of the DBRs was evaluated comparing the reflectance spectra collected after the first thermal annealing and after the desorption. Both the measurements were performed at room temperature. The optical response of the sensors was sampled at set intervals along the entire cation exposure using a halogen-deuterium DH-BAL-2000 light source and an CMOS spectrometer Avantes-2048-Evo for the detection. Steady state response was finally investigated after drying the sample in air at 60 • C. The reflectance spectra of the DBR structure were modeled using a transfer matrix method formalism as described in Ref. [32] using the refractive index dispersion of SA and PVK, which were retrieved from literature as input (n PVK,500 nm = 1.68, n SA, 500 nm = 1.51, Ref. [87][88][89]) and the geometric thickness of the layers as fitting parameters. The layer thicknesses were retrieved for the sample made of 10.5 periods and employed for the modeling of the structure made of 6.5 bilayers fabricated using the same conditions. Figure 2a schematizes the DBRs fabricated in this work. There, thin films of SA and PVK are alternated to form a lattice. Figure 2b displays the normal incidence reflectance spectra collected over 9 different spots of the sample surface for the DBR made of 10.5 bilayers (see next paragraph for the same characterization for the sample made of 6.5 bilayers). All the spectra show two main features: first, an intense reflectance peak at about 510 nm with maximum value above 80%. This signal is assigned to the DBR PBG. The second feature is an oscillating pattern in the background of all the spectra. This pattern takes the name of Fabry-Pérot pattern and arises from the interference of beams partially reflected at the top and bottom interfaces of the photonic structure [32]. The presence of these two characteristics demonstrates that the DBRs possess good optical quality [90]. Moreover, the superimposition of the nine spectra proves the rather good homogeneity of the DBR structure. To the best of our knowledge, so far this is the first planar 1D photonic crystals containing SA reported in the literature.

10.5 Period DBRs
Figure 2a schematizes the DBRs fabricated in this work. There, thin films of SA and PVK are alternated to form a lattice. Figure 2b displays the normal incidence reflectance spectra collected over 9 different spots of the sample surface for the DBR made of 10.5 bilayers (see next paragraph for the same characterization for the sample made of 6.5 bilayers). All the spectra show two main features: first, an intense reflectance peak at about 510 nm with maximum value above 80%. This signal is assigned to the DBR PBG. The second feature is an oscillating pattern in the background of all the spectra. This pattern takes the name of Fabry-Pérot pattern and arises from the interference of beams partially reflected at the top and bottom interfaces of the photonic structure [32]. The presence of these two characteristics demonstrates that the DBRs possess good optical quality [90]. Moreover, the superimposition of the nine spectra proves the rather good homogeneity of the DBR structure. To the best of our knowledge, so far this is the first planar 1D photonic crystals containing SA reported in the literature. To further confirm the PBG assignment, we also measured the angular resolved transmittance spectra of the photonic structure. The spectra are reported in Figure 2c for s-polarized light and in Figure 2d for p-polarized light (see Ref. [32] for further details regarding the measurement). The spectra are reported as contour-plot where the light wavelength is reported on the ordinate axis while the light incidence angle as the abscissa axis. The transmittance intensity is instead reported as a color scale such that low intensities are in blue shades and high intensity values in red shades. When illuminated with s-polarized light, the spectral position of the PBG, which was visible in blue tones in panel c of the Figure, moves to the blue wavelength side of the spectrum increasing the angle and its spectral width remains unchanged, although its relative intensity increases. Concerning p- To further confirm the PBG assignment, we also measured the angular resolved transmittance spectra of the photonic structure. The spectra are reported in Figure 2c for s-polarized light and in Figure 2d for p-polarized light (see Ref. [32] for further details regarding the measurement). The spectra are reported as contour-plot where the light wavelength is reported on the ordinate axis while the light incidence angle as the abscissa axis. The transmittance intensity is instead reported as a color scale such that low intensities are in blue shades and high intensity values in red shades. When illuminated with s-polarized light, the spectral position of the PBG, which was visible in blue tones in panel c of the Figure, moves to the blue wavelength side of the spectrum increasing the angle and its spectral width remains unchanged, although its relative intensity increases. Concerning p-polarization instead, the spectral position of the PBG follows the same pattern just observed. Conversely, its intensity and spectral width decreases with the angle till the feature almost disappears slightly above 60 • , where the incident light approaches the Brewster angle for the structure. This behavior was in full agreement with previous reports for planar 1D photonic crystals [91] and with the photonic band structure predicted from theory [32,90]. The collected spectra were indeed in agreement with the modeled one for a structure having PVK and SA geometric thickness d PVK = 75 nm and d SA = 81 nm, respectively.
The sensitivity of the system to Pb 2+ was tested immersing the DBR into water solution of Pb 2+ ion with concentration of 1 × 10 −2 M. The top panels of Figure 3a show the normal incidence reflectance spectrum of the DBR before (black line) and after (red line) the sample swelling. Before the exposure, the sample show a clear signature of the PBG at about 510 nm and the spectrum background displays an interference pattern (black line). After the exposure, these features were slightly visible, and the reflectance was rather flat in the range between 25% and 35% (red line). This effect can be attributed to the severe swelling of the SA layers in water. Such phenomenon can indeed worse the lattice periodicity that generates the PBG and the surface optical quality that permits the interference pattern [53]. Removing the sample from the water environment and drying the structure at 60 • C, the two spectral features were again visible (green line). Nevertheless, the spectral position of the PBG results blue-shifted to about 475 nm, about 39 nm lower with respect to its initial value. This phenomenon can be explained considering the shrinkage of SA layers after cross-linking; indeed, the physical "bridges" created by Pb 2+ ions between alginate chains force the formed 3D-network to assume a more compact structure, which reduces the thickness of the DBR layers [92][93][94][95].
To better explain the phenomena occurring in the sensor, we monitored the entire spectral response during the swelling process reported in Figure 3b as a contour plot having the wavelength on the y-scale and the time on the x-scale. In figure we see that the PBG shifts to more than 100 nm within 5 min (300 s). At this time, notwithstanding the PBG was not detectable in the spectrum anymore, its higher order replicas were visible at about one half (marked with * in Figure 3b) and one third (marked with *' in Figure 3b) of its wavelength. These higher diffraction orders continue red-shifting until up to the sixth order of diffraction was detected in the spectral range investigated at about 410 nm. This effect owes to the severe swelling of the structure that, on the other hand, maintains the periodicity. We can then infer that this large structure swelling was provided by a great water intake, while SA cross-linking reaction prevents the dissolution of the structure. Under this assumption we can estimate an increase of optical thickness (n times d) for the SA layers much larger than 100% with respect to the initial value. Indeed, the swelling degree of SA hydrogels can be larger than 1000%. Then, we can estimate a refractive index variation for the fully swollen structure up to ∆n SA = −11%. The refractive index variation for swelling processes was indeed often considered negligible with respect to the thickness variation [32,53,96].
The cross-linking was confirmed by the spectral behavior of the DBR during the drying process ( Figure 3c, top panel) where the PBG was not detectable as soon as the water environment was removed. It becomes instead increasing evident from 80 s of drying, where the feature becomes clear in red shades. At this time, the spectral feature shifts to the blue side of the spectrum until it reaches about 475 nm-well below its initial position. As mentioned above, the smaller wavelength of the PBG testifies a reduction of the light optical path (refractive index times the thickness) within the structure. In details, the spectral position of the peak results blue-shifted by~40 nm. It is known that the cross-linking process provides a volume reduction in SA up to 100% with respect to its initial value while refractive index variations can approach 4%-5% [88]. Then, in first approximation and in agreement with the literature, the refractive index variation can be again neglected with respect to thickness variation [97]. The spectral shift can then be related to a shrinking of the SA layers by −20% with respect to its initial value. Furthermore, the restoration of both PBG and interference pattern testifies that alginate did not dissolve under the experiment condition.  The cross-linking was confirmed by the spectral behavior of the DBR during the drying process ( Figure 3c, top panel) where the PBG was not detectable as soon as the water environment was removed. It becomes instead increasing evident from 80 s of drying, where the feature becomes clear in red shades. At this time, the spectral feature shifts to the blue side of the spectrum until it reaches about 475 nm-well below its initial position. As mentioned above, the smaller wavelength of the Notice that the large hysteresis in the sorption-desorption process, together with the non-reversible response that characterizes cross-linked polymers makes these systems unsuitable for repeated used. The hysteresis can be assigned to the very diverse conditions for swelling and desorption processes. Indeed, the first occurs in water environment while the second through water evaporation.
As mentioned above, the lack of signal attributed to the PBG during the first stages of drying arises from different swelling degree of the DBR components and has been widely described in literature [54,97,98]. As previously reported [32], polymer multilayer sensors can be as small as few squared millimeters. The reduced size, together with the possibility to fabricate these structures on the very large area by industrial processes [59][60][61]99] makes them interesting low-cost disposable sensor.
The samples were then exposed to different concentrations of Pb 2+ ions to test whether the optical response can be used to retrieve this parameter and the lower detection limit of the system. The bottom panels of Figure 3 report the full optical response of the DBR to concentrations of Pb 2+ ranging between 5 × 10 −3 to 1 × 10 −5 M (the concentration value is reported at the left side of the y-axis). Again, panels a show the steady state response, while panels b and c show the full dynamic spectral evolution of the DBRs during exposure and desorption, respectively. As observed before, the PBG undergoes a blue-shift and a decrease in intensity upon exposure and drying (compare green and black lines in panels a). The responses appear rather similar to the one just observed, but with faster kinetics for decreasing concentrations (panels b and c). Moreover, the PBG intensity and spectral position collected after the drying decrease upon decreasing Pb 2+ concentration. Indeed, low lead concentrations induce low cross-linking effect and then facile solubilization of the alginate layers in the swollen structure. The solubilization of the swollen alginate layers affects the periodicity of the dried samples that result thinner. With decreasing lead concentration, the PBG blue-shift increases and its profile were increasingly inhomogeneously broadened (compare green lines in Figure 3a). These were clear symptoms of thinner layers in the lattice and of disorder, respectively.
To confirm the presence of the PBG after exposure, Figure 4 shows the angular dispersion of the PBG after exposure and drying for the samples exposed to concentration down to 10 −4 M. It was not possible to measure the spectra for the lower concentration (10 −5 M) as the PBG spectral position ranges between 350 and 525 nm, close to the absorption inset of the glass substrate. Then, we can only infer than the broad feature detected in the bottom panel of Figure 3a was assigned to a PBG. Nevertheless, its intensity and spectral position follows the observed trend for ion concentration observed for the other samples. For concentration ranging from 10 −2 to 10 −3 the PBG dispersions reported as contour-plot fully resemble the data observed in Figure 2 for the unexposed structure. For the concentration 10 −4 M, the low intensity of the PBG makes it hardly detectable in the contour plot then we also reported the pristine spectra collected for the sample. The bottom panels of Figure 4 display that for both polarizations the spectral position of the broad peak moves to the short wavelength side of the spectrum confirming its nature.

6.5 Period DBRs
To confirm the behavior observed, we repeated the characterization for a DBR made of 6.5 bilayers. Figure 5a displays the normal incidence reflectance spectra collected over 9 different spots of the sample surface. All the spectra show an intense reflectance peak assigned to the PBG at about 510 nm with reflectance value above 60% and a Fabry-Pérot pattern. The superimposition of the nine spectra, demonstrates the rather good homogeneity of the DBR. The angular resolved transmittance spectra are reported in Figure 5b,c for p-polarized and s-polarized light, respectively. When illuminated with p-polarized light, the spectral position of the PBG, which was visible in blue tones in panel b, moves to the short wavelengths increasing the angle, and its spectral width decreases approaching the Brewster angle. Concerning s-polarization, the PBG position follows the same trend, while its intensity slightly changes. This behavior agrees with the data collected for the 10.5 bilayer sample. The lower intensity of the peaks assigned to PBG owes to the lower number of periods composing the structure. The collected spectra were indeed well-superimposed on the modeled one (black dotted line) using the geometric thickness retrieved from the fitting of the spectra of the sample made of 10.5 bilayers.

6.5 Period DBRs
To confirm the behavior observed, we repeated the characterization for a DBR made of 6.5 bilayers. Figure 5a displays the normal incidence reflectance spectra collected over 9 different spots of the sample surface. All the spectra show an intense reflectance peak assigned to the PBG at about 510 nm with reflectance value above 60% and a Fabry-Pérot pattern. The superimposition of the nine spectra, demonstrates the rather good homogeneity of the DBR. The angular resolved transmittance spectra are reported in Figure 5b,c for p-polarized and s-polarized light, respectively. When approaching the Brewster angle. Concerning s-polarization, the PBG position follows the same trend, while its intensity slightly changes. This behavior agrees with the data collected for the 10.5 bilayer sample. The lower intensity of the peaks assigned to PBG owes to the lower number of periods composing the structure. The collected spectra were indeed well-superimposed on the modeled one (black dotted line) using the geometric thickness retrieved from the fitting of the spectra of the sample made of 10.5 bilayers. The DBR was then portioned and exposed to Pb 2+ solutions with concentration ranging from 1 × 10 −2 to 1 × 10 −4 M. The low number of periods allows faster sorption kinetics and makes improbable the possibility to detect concentrations lower than 10 −4 M. Figure 6 reports the exposure data collected for Pb 2+ concentrations of 1 × 10 −2 , 5 × 10 −3 , 1 × 10 −3 and 1 × 10 −4 M, from top to bottom. In full agreement with the behavior observed in Figure 3, after the exposure the PBG blue-shifts and its intensity decreases for decreasing concentration (green lines in Figure 6a). Regarding the swelling kinetics it was faster than for the DBR with larger number of layers (Figure 6b). In agreement with theory, the diffusivity of small molecules into a slab, was indeed a function of its thickness for thin films [100]. For the largest Pb 2+ concentration (10 −2 M), all the features assigned to the PBGs fade at about 400 s of exposure. Then, the lower number of layers appears to facilitate the introduction of disorder in the structure. The behavior during desorption was analogous to that observed before, as no major differences were expected at the working temperature for thin films of comparable thickness (~1 μm for the 6.5 period DBR and ~1.6 μm for the 10.5 period). Overall, the very similar behavior observed for the two DBRS, together with the faster swelling and larger disorder of the 6.5 period one confirms the behavior of the sensor, whose response was ruled by antagonist cross-linking and dissolution phenomena. The DBR was then portioned and exposed to Pb 2+ solutions with concentration ranging from 1 × 10 −2 to 1 × 10 −4 M. The low number of periods allows faster sorption kinetics and makes improbable the possibility to detect concentrations lower than 10 −4 M. Figure 6 reports the exposure data collected for Pb 2+ concentrations of 1 × 10 −2 , 5 × 10 −3 , 1 × 10 −3 and 1 × 10 −4 M, from top to bottom. In full agreement with the behavior observed in Figure 3, after the exposure the PBG blue-shifts and its intensity decreases for decreasing concentration (green lines in Figure 6a). Regarding the swelling kinetics it was faster than for the DBR with larger number of layers (Figure 6b). In agreement with theory, the diffusivity of small molecules into a slab, was indeed a function of its thickness for thin films [100]. For the largest Pb 2+ concentration (10 −2 M), all the features assigned to the PBGs fade at about 400 s of exposure. Then, the lower number of layers appears to facilitate the introduction of disorder in the structure. The behavior during desorption was analogous to that observed before, as no major differences were expected at the working temperature for thin films of comparable thickness (~1 µm for the 6.5 period DBR and~1.6 µm for the 10.5 period). Overall, the very similar behavior observed for the two DBRS, together with the faster swelling and larger disorder of the 6.5 period one confirms the behavior of the sensor, whose response was ruled by antagonist cross-linking and dissolution phenomena.
To confirm that dissolution plays a fundamental role in the DBR response, we report its spectral behavior in deionized water (Figure 7). In this case, the SA layers easily dissolve and the PBG cannot be restored anymore. In Figure 7a, the DBR initially shows a PBG at about 490 nm. During the swelling all the structures fades in about 400 s (Figure 7b). Conversely to what observed for Pb 2+ solutions, the spectral features cannot be restored anymore by the drying process ( Figure 7c) and the resulting spectrum was a flat line (green line in Figure 7a). Figure 6. Full spectral response for the 6.5 periods DBR to Pb 2+ solutions with concentrations of (from top to bottom) 1 × 10 −2 , 5 × 10 −3 , 1 × 10 −3 and 1 × 10 −4 M. (a) Normal incidence reflectance spectra before (black line) and after (red line) exposure to Pb 2+ water solution and spectra after drying. Full dynamic spectral response of the swelling (b) and drying (c) processes.
To confirm that dissolution plays a fundamental role in the DBR response, we report its spectral behavior in deionized water (Figure 7). In this case, the SA layers easily dissolve and the PBG cannot be restored anymore. In Figure 7a, the DBR initially shows a PBG at about 490 nm. During the swelling all the structures fades in about 400 s (Figure 7b). Conversely to what observed for Pb 2+ Figure 6. Full spectral response for the 6.5 periods DBR to Pb 2+ solutions with concentrations of (from top to bottom) 1 × 10 −2 , 5 × 10 −3 , 1 × 10 −3 and 1 × 10 −4 M. (a) Normal incidence reflectance spectra before (black line) and after (red line) exposure to Pb 2+ water solution and spectra after drying. Full dynamic spectral response of the swelling (b) and drying (c) processes. solutions, the spectral features cannot be restored anymore by the drying process (Figure 7c) and the resulting spectrum was a flat line (green line in Figure 7a).

Quantitative DBR Response
Due to small inhomogeneity of the DBR spectra (see Figures 2 and 5), it was hard to quantify the spectral variations from the absolute PBG spectral positions and intensities measured before and after exposure and drying. Then, Figure 8 summarized the data relative to the initial spectral position and reflectance values for each sample for both the 6.5 (panel a-c) and 10.5 (panel a'-c') period DBRs. The spectra of Figure 8a and a' show that both PBG reflectance intensity and spectral position decrease with decreasing Pb 2+ concentration for both the DBRs. Figure 8b, b' c and c' show the PBG spectral position and reflectance variation in percent, measured after sample swelling and drying. Both the variations increase for decreasing Pb 2+ concentration. As discussed previously, this behavior can be explained considering competitive effects of SA dissolution and cross-linking. At large concentrations the alginate cross-links and shrinks. The reticulation was indeed well known to make the alginate insoluble in water solutions. At lower concentrations of divalent ions, the cross-linking was less efficient implying a partial solubilization of the polymer. This behavior was in full agreement with the data reported in Figures 3, 6 and 7. The partial solubilization of the alginate layers can indeed lead to its thinning and thus to the blue-shift of the PBG spectral feature. Concerning the reflectance value, such partial solubilization can introduce disorder in the DBR, reducing the reflectance intensity and introducing inhomogeneous broadening of the PBG spectral signature. Indeed, for decreasing lead concentration the PBG intensity decreases and its line-shape becomes asymmetrical [97]. Panels b and b' show an exponential dependence between the reflectance reduction and the lead ion concentration in solution. In both cases it was possible to distinguish the investigated concentration values. Concerning the spectral shift instead, there was not a clear correlation for the 6.5 bilayers when exposed to the highest lead concentrations, while for smaller value the variation increases. For the 10.5-period instead, there was a clear exponential dependence between concentration and reflectance variation in the entire range (Figure 8b'). Concerning the spectral position variation, it decreases exponentially for concentrations from 1 × 10 −3 to 1 × 10 −5 M, while for larger values the variations reach a plateau. Then, we can reasonably assume that increasing the number of layers allows better sensitivity, especially for low lead concentrations and lower detection limit. This was probably due to the larger resistance of the structure having 10.5 bilayers with respect to the other one, as observed in Figures 3 and 6.

Quantitative DBR Response
Due to small inhomogeneity of the DBR spectra (see Figures 2 and 5), it was hard to quantify the spectral variations from the absolute PBG spectral positions and intensities measured before and after exposure and drying. Then, Figure 8 summarized the data relative to the initial spectral position and reflectance values for each sample for both the 6.5 (panel a-c) and 10.5 (panel a'-c') period DBRs. The spectra of Figure 8a,a' show that both PBG reflectance intensity and spectral position decrease with decreasing Pb 2+ concentration for both the DBRs. Figure 8b,b',c,c' show the PBG spectral position and reflectance variation in percent, measured after sample swelling and drying. Both the variations increase for decreasing Pb 2+ concentration. As discussed previously, this behavior can be explained considering competitive effects of SA dissolution and cross-linking. At large concentrations the alginate cross-links and shrinks. The reticulation was indeed well known to make the alginate insoluble in water solutions. At lower concentrations of divalent ions, the cross-linking was less efficient implying a partial solubilization of the polymer. This behavior was in full agreement with the data reported in Figures 3, 6 and 7. The partial solubilization of the alginate layers can indeed lead to its thinning and thus to the blue-shift of the PBG spectral feature. Concerning the reflectance value, such partial solubilization can introduce disorder in the DBR, reducing the reflectance intensity and introducing inhomogeneous broadening of the PBG spectral signature. Indeed, for decreasing lead concentration the PBG intensity decreases and its line-shape becomes asymmetrical [97]. Panels b and b' show an exponential dependence between the reflectance reduction and the lead ion concentration in solution. In both cases it was possible to distinguish the investigated concentration values. Concerning the spectral shift instead, there was not a clear correlation for the 6.5 bilayers when exposed to the highest lead concentrations, while for smaller value the variation increases. For the 10.5-period instead, there was a clear exponential dependence between concentration and reflectance variation in the entire range (Figure 8b'). Concerning the spectral position variation, it decreases exponentially for concentrations from 1 × 10 −3 to 1 × 10 −5 M, while for larger values the variations reach a plateau. Then, we can reasonably assume that increasing the number of layers allows better sensitivity, especially for low lead concentrations and lower detection limit. This was probably due to the larger resistance of the structure having 10.5 bilayers with respect to the other one, as observed in Figures 3 and 6.
Regarding a possible use of the proposed sensor in the real environment, effects of parameters like the presence of other divalent cations and the effect of pH should also be considered. In these regards, several studies proved that alginate-based hydrogels show a considerable selectivity in the adsorption of divalent ions from aqueous solutions and that the largest affinity with Pb 2+ allow preferable cross-linking with the latter. For example, An et al. performed batch and column tests to evaluate the competitive sorption within alginate beads observing an affinity of the type of Pb 2+ >> Cu 2+ >> Ni 2+ > Ca 2+ , with the adsorption behavior of Pb not influenced by the medium pH [101]. Similarly, Papageorgiou et al. obtained that the maximum metal adsorption capacities of alginate hydrogels followed the order Pb 2+ > Cu 2+ > Cd 2+ [102], whereas Fatin-Rouge et al. showed that metal ions had the following affinity for Pb 2+ >> Cu 2+ > Zn 2+ > Ni 2+ with very good retentions obtained for Pb 2+ , Cu 2+ and Zn 2+ [103]. Yet, Sone et al. developed composite alginate-based foams for the selective elimination of Pb 2+ from contaminated water [104]. Moreover, the literature show that the investigation of DBR sensor response kinetics during analytes intercalation or multivariate analysis provide the sensor with selectivity even among very similar analytes such as benzene and toluene [32,53] based on different weak interactions and diffusivity of the specie within the DBRs. Nevertheless, the selectivity of the system and thus its response to other contaminants is still to be investigated. Regarding a possible use of the proposed sensor in the real environment, effects of parameters like the presence of other divalent cations and the effect of pH should also be considered. In these regards, several studies proved that alginate-based hydrogels show a considerable selectivity in the adsorption of divalent ions from aqueous solutions and that the largest affinity with Pb 2+ allow preferable cross-linking with the latter. For example, An et al. performed batch and column tests to evaluate the competitive sorption within alginate beads observing an affinity of the type of Pb 2+ >> Cu 2+ >> Ni 2+ > Ca 2+ , with the adsorption behavior of Pb not influenced by the medium pH [101]. Similarly, Papageorgiou et al. obtained that the maximum metal adsorption capacities of alginate hydrogels followed the order Pb 2+ > Cu 2+ > Cd 2+ [102], whereas Fatin-Rouge et al. showed that metal ions had the following affinity for Pb 2+ >> Cu 2+ > Zn 2+ > Ni 2+ with very good retentions obtained for Pb 2+ , Cu 2+ and Zn 2+ [103]. Yet, Sone et al. developed composite alginate-based foams for the selective elimination of Pb 2+ from contaminated water [104]. Moreover, the literature show that the investigation of DBR sensor response kinetics during analytes intercalation or multivariate analysis provide the sensor with selectivity even among very similar analytes such as benzene and toluene [32,53] based on different weak interactions and diffusivity of the specie within the DBRs. Nevertheless, the selectivity of the system and thus its response to other contaminants is still to be investigated.

Conclusions
In conclusion, we demonstrated new all-polymer planar 1D photonic crystals processable by spin-coating deposition made of sodium alginate and poly (N-vinylcarbazole). The new structures possess high optical quality and optical behavior in agreement with theory. The new photonic system shows large swelling in water and sensitivity to Pb 2+ ions owing to their ability to induce reticulation of alginate. The DBR demonstrated detection limit as low as 10 −5 M and is promising for the development of Pb 2+ sensing devices. Although, more analyses are needed to establish selectivity, enhance sensitivity and detection limit of the sensor, these results pave the way for the development of disposable chromatic sensors to be used in situ even by nonspecialized people.