Primary potentiometric measurement using CPE incorporating Sodium tetrakis(1-imidazolyl)borate ion-exchanger (SIB, Fig. 1) and (4-nitrobenzo-15-crown-5 and dibenzo-18-crown-6 ionophores, Fig. 1) was performed for the anions and heavy metal ions including Ag+. The primary potentiometric results confirmed the Ag+ selectivity using 4-nitrobenzo-15-crown-5 only.Fig. 1Chemical structure of dibenzo-18-crown-6 ionophore (A), 4-nitrobenzo-15-crown-5 ionophore (B), ETH 5294 chromoionophore (C), and Sodium tetrakis(1-imidazolyl)borate ion-exchanger (D).Effect of compositionThe optode composition was varied, as in Table 1, by changing type and amount of ionophore, type of plasticizer and amount of ion-exchanger, while the amounts of PVC, plasticizer and ETH 5294 were fixed. The existence of an ion-exchanger is mandatory for the sensors that operates via neutral carrier mechanism11.Table 1 The composition of the different optodes with its concentration range and detection limit.In optode 1, low amounts of ionophore and SIB of (ionophore/SIB) molar ratio ≈ 2, were dissolved in THF with the other components; it exhibited good response in limited concentration range of 10−5–10−3 M with a detection limit of 1.0 × 10−5 M. In optode 2, higher amounts of ionophore and ion-exchanger with higher molar ratio of 2.5 exhibited better concentration range (10−7–10−3 M) and lower detection limit of 9.7 × 10−8 M (Table 1). The ionophore/SIB molar ratio of 1.6 in optode 3 with higher amounts of ionophore and ion-exchanger, exhibited better response in the concentration range of (10−11–10−8), and the lowest observed detection limit of 8.8 pico-Molar (pM) (Fig. 2). Also, the highest signal intensity (sensitivity) was observed for optode 3 (compared to optode 1 and 2), with two maxima at 550 and 665 nm (Fig. 2). The lowest detection limit and good concentration range of optode 3 can be interpreted by the suitable amount of the ion-exchanger SIB that facilitated Ag+ transfer from the aqueous phase to the membrane phase11, shifting the concentration range and detection limit to the lower values. Also, SIB was added to compensate for the charge imbalance that may exist in the membrane21. The same amount of the ion-exchanger SIB was tested without an ionophore in optode 4; the concentration range was decreased, shifted to higher concentrations and higher detection limit of (1.0 × 10−9 M); the response of optode 4 is due to the interaction between Ag+ and SIB, as confirmed in the literature by formation of white precipitate22; the higher detection limit and small concentration range of optode 4 are due to the absence of the ionophore, that is responsible for the selective interaction with the analyte Ag+11. Also, optode 5 was prepared for comparison using Sodium tetrakis 3,5-bis(trifluoromethyl)phenyl borate ion-exchanger instead of SIB; the absorbance change was very small (˂ 0.01) in the concentration range of 10−8–10−5 M, with higher detection limit of 10−8 M (Table 1); the spectrum of optode 5 exhibited the normal shape of spectra shown by ETH5294; two maxima at 620 and 630 nm (Supplementary Fig. S1). For confirming role of plasticizer, TCP was tested instead of NPOE in sensor 6, with the same other constituents of optode 3; it exhibited diminished response in small concentration range with high detection limit (Table 1). In sensor 7, the 4-nitrobenzo-15-crown-5 ionophore is replaced by dibenzo-18-crown-6 as an ionophore, with the same other constituents of optode 3; it exhibited no response to Ag+. These results confirm the role of the ion-exchanger SIB and the 4-nitrobenzo-15-crown-5 ionophore. Accordingly, optode 3 was the sensor of choice during this study.Fig. 2Visible spectrum of optode 3 at different concentrations (10–11–10–8 M) (A); spectrum spotted in the maximum region around 550 nm (B), and the calibration curve for optode 3 (closed circle) and optode 4 (open circle) at λmax = 550 nm (C).Optode response mechanism and characterizationThe interaction mechanism was elucidated by performing FTIR for the ionophore, and ion-exchanger before and after interaction with Ag+. In ionophore spectrum, bands at 2993, 2908, and 1651 cm−1 in the ionophore spectrum was assigned to the aromatic C–H, aliphatic C–H, and aromatic C–C vibrations, respectively23,24; after interaction with Ag+, these bands were shifted to 2985, 2909, and 1658 cm−1, respectively, which confirmed the ionophore–Ag+ interaction (Fig. 3). However, the most important and noticeable peak shift was observed in the fingerprint region; ionophore peaks at 697, 1044 and 1319 cm−1 assigned to –CH2, C–O–C, C–C–O and –CH were shifted to 693, 1041 and 1317 cm−1, after interaction with Ag+25,26. It can be observed that C–O–C band shift in Fig. 3 and Table 2 confirmed the complexation via oxygen atoms of the crown ether to form penta-coordinated complex, as confirmed in the literature23,24.Fig. 3FTIR spectra of optode 3 before (red) and after (blue) interaction with Ag+ in the range 400–2000 cm−1.Table 2 FTIR peak shifts in cm−1 for optode 3 and optode 4.In addition to the peak shift, peak broadening was observed at 693, 1658 and 1590 cm−1 (Fig. 3), which confirmed the role of the ionophore for Ag+ complexation. Also, the interaction between the ion-exchanger SIB with Ag+ exhibited peak broadening and peak shifts at same previous positions (Table 2), and as was reported in the literature22, where a whiter precipitate between Ag+ and SIB is formed. Table 2 shows that peak shift in case of ion-exchanger and ionophore existence (optode 3) is greater than that in case of existence of ion-exchanger only (optode 4); this confirmed the role and importance of ionophore in the response mechanism.According to previous data, and according to literature, the response mechanism depends on Ag+ transfer from the aqueous analyte solution phase to the organic membrane optode phase, which is facilitated with the ion-exchanger SIB (\({R}_{m}^{-}\)) via low reversibility reaction, Eq. (1); this is followed by the crown ionophore complexation for Ag+ to form Lm–Ag+ complex, Eq. (2), which is followed by the deprotonation of the chromoionphore ETH 5294 (\({CH}_{m}^{+}\)) to get the deprotonated form of the chromoionphore (Cm), Eq. (3); deprotonation leads to decrease of the absorbance at 550 or 665 nm (Fig. 2)11,21. This mechanism was reported earlier for Ag-selective optodes using the same chromoionophore18,19.$${R}_{m}^{-} + {Ag}_{a}^{+}\to {\text{RAg}}_{\text{m}}$$
(1)
$${\text{RAg}}_{\text{m}}+ {\text{L}}_{\text{m}}\to {R}_{m}^{-}+{LAg}_{m}^{+}$$
(2)
$${CH}_{m}^{+}\to {\text{C}}_{\text{m}}+ {H}_{a}^{+}$$
(3)
$${R}_{m}^{-} + {\text{L}}_{\text{m}}+ {CH}_{m}^{+} +{Ag}_{a}^{+} \to {R}_{m}^{-}+{LAg}_{m}^{+}+ {\text{C}}_{\text{m}}+ {H}_{a}^{+}$$
(4)
The interaction between SIB and Ag+, Eq. (1), was confirmed by the FTIR (Supplementary Figs. S2 and S3). Also, the low reversibility of SIB-Ag+ interaction of Eq. (1) is responsible for the shift of the isosbestic point to the UV-region, with change of maxima from 665 to 550 nm (Fig. 2); both the absence of isosbestic point in visible region with its existence in UV-region, and maximum wavelength change in spectrum were reported for other optodes27.Optode 3 was imaged using AFM before and after measurement; it is shown in Fig. 4C that, the membrane surface contains agglomeration for particles with sizes 1–5 µm due to the low solubility of the ionophore in THF, resulting in heterogeneity, roughness and high surface area of membrane optodes; this roughness caused retention of hydrated ions and water flux, increased sensitivity to pico-Molar concentration, facilitate diffusion of membrane components to aqueous solution (short lifetime) and aqueous ions to membrane surface28,29,30,31. Membrane retention of water flux and diffusion of membrane components to aqueous solution decreased the measured surface area from 9.51 to 8.91 µm2, and roughness decreased from 1.63 to 1.39 before (Fig. 4A) and after (Fig. 4B) insertion in measurement solution.Fig. 43D-AFM images of optode 3 before soaking (A), and after soaking in 10–8 M Ag+ (B); size distribution of aggregations (C).Effect of pHIn optodes, the measurement is transduced to the optical signal by the chromoionphore, depending on chromoionophore protonation-deprotonation equilibrium11,32. Optode 3 was studied at different pH values of 4.2, 4.5 and 5 using acetate buffer; it gave almost the same concentration range with very small shift in the detection limit to lower values at pH 5.0 (Supplementary Table S1). Accordingly, and as reported in the literature18, pH 5.0 was selected for performing measurements.SelectivitySelectivity defines the sensor ability for differentiating the analyte ion from the interfering species; this is important, especially in real samples. The separate solution method was applied, where calibration curves for the Ag+ and the interfering species were performed using optode 3 (Fig. 5).Fig. 5Response of optode 3 towards different interfering cations and anions at 550 nm.Optode 3 exhibited high Ag+ selectivity over other interfering heavy metal cations, chloride and sulphate anions (Fig. 5). As recommended in literature11, the calculated selectivity coefficients by separate solution method at α = 0.5 \({logK}_{Ag+, cation}^{opt}\) was found to be − 4.3 for Cu2+, − 5.6 for Ni2+ and − 5.0 for Cd2+; the other interfering species (Hg2+, Mn2+ and Pb2+, \({NH}_{4}^{+}\), K+, Mg2+, Cl− and \({SO}_{4}^{2-}\)) exhibited lower or minimum interference with neglected response (Supplementary Fig. S4).Mixed solution method was performed to confirm selectivity observed by separate solution method (Supplementary Table S2); change in absorbance was minimum and can be neglected for all interfering specie, with minimum interference from cadmium ion. This ensures that, 4-nitrobenzo-15-crown-5 ionophore with the ion-exchanger SIB is highly selective for Ag+, even in presence of other interfering ions in the same measurement solution.The selectivity data observed by separate solution and mixed solution methods can be interpreted in terms of lipophilicity-charge density relation, and size suitability for the host–guest interaction (Table 3). Charge of the metal cation divided by the ionic radius (charge density) is inversely related to water solubility; i.e. high charge density ions are water soluble and prefer hydrophilic media, and vice versa33,34,35. Ag+ exhibited the highest charge density, so high lipophilicity which accounts for Ag+ selectivity (Table 3). On the other hand, ionic radius of Ag+ is the most suitable for the cavity size of crown ionophore, which facilitate strong host-guest interaction, as confirmed in the literature36. These factors enhance the silver selectivity by crown-based optode incorporating imidazolyl borate ion-exchanger.Table 3 The ionic radius and charge density for Ag+ and different interfering ions34,35,37.Response time, lifetime and reversibilityResponse time (tR) is defined as the time taken to achieve about 90% of the sensor stable response11. Short tR is important issue for large number sample analysis, as in routine work. tR of optode 3 was 2–3 min, which is shorter than previous work17,18,38. This tR can be understood in terms of the time needed by reaction mechanism, which are three steps. The absorbance was recorded during 4 min till constant value, and it was plotted for different concentrations at the different response times, Fig. 6. It is clear in the figure that absorbance change is neglected and calibration curve is the same during 2–4 min, i.e tR is 2–3 min11,39.Fig. 6Effect of response time (tR) on Calibration curves of optode 3.Life time is related to consumption of membrane components by irreversible reaction or leaching to the measurement solutions, the stability of the different components within the membrane, and the stability of the membrane itself on the substrate11. The calibration curve for optode 3 was constructed several times at different time intervals to get the lifetime of the sensor (Table 4).Table 4 The concentration range and detection limit of optode 3 at different lifetimes.As in Table 4, small change in the concentration range and detection limit was observed until 10 days where both the concentration range and detection limit started to decrease. Indeed, the slope of the response curve decreased also during the 10 days (Supplementary Fig. S5). This is expected due to the leaching of the membrane components to the measurements solution, as confirmed by decrease in surface area and roughness of optodes in AFM images (Fig. 4)11,21.Reversibility of optode 3 was tested by measurement of different Ag+ concentrations from high to low concentrations and vice versa. Optode 3 was found to restore about 92% of its absorbance when performing calibration curve several times, but with two restrictions; Ag+ measurement should be in the ascending concentration direction (low to high concentrations) and not the reverse, and optode should be soaked for about 30–45 min in buffer OR soaking in 1 M HCl for 5 min followed by soaking in buffer for 10 min when calibration curve is to be reconstructed, as reported earlier39,40. These response characteristics of long restoration time can be explained by the memory effect due to the existence of some of analyte ion, complexed to ionophore within the membrane which need long time for de-complexation11; soaking in acid facilitates the protonation of the chromoionophore which facilitates analyte de-complexation to keep electro-neutrality within optode membrane. Also, the low reversibility of Ag+-SIB interaction interprets low reversibility of optode 3.ApplicationsOptode 3 was applied for measuring Ag+ amount in the pharmaceutical formulation PinkEye Relief® eye drop, and the results were compared to that of the reference inductively coupled plasma-mass spectrometry modified method41. Good and reliable recovery values (not exceeding 91%) in Table 5 confirm the accuracy of the method, and the low standard deviation values confirm the precision of measurement using the optode.Table 5 Determination of Ag+ in PinkEye Relief® eye drop using optode 3 and ICP-Ms reference method41.By comparing the results of optode 3 in this work to the previously mentioned ones in the literature, the response time decreased, the detection limit decreased to more sensitive values of pico-Molar, and concentration range is good (Table 6). The most important feature in this work is the improvement of selectivity, especially over heavy metal ions as can be observed in Table 6.Table 6 Comparison of the response characteristics of optode 3 to previously reported silver-selective optodes17,18,38.
PVC membrane bulk optode incorporating 4-nitrobenzo-15-crown-5 and sodium tetrakis(1-imidazolyl) borate for the pico-molar determination of silver ion in pharmaceutical formulation
