A Miniaturized Therapeutic Chromophore for Multiple Metal Pollutant Sensing, Pathological Metal Diagnosis and Logical Computing

The efficacy of a miniaturized unimolecular analytic system is illustrated. The easily accessible therapeutic chromophore “temoporfin”, which responds differentially to bound metals at multiple wavelengths of Q-band absorption using chemometric analysis, expeditiously detects and discriminates a wide range of metals regarded as priority pollutants in water and hence may also be used for diagnosis of medically relevant metals in human urine. The molecule was further investigated as an electronic logic device, e.g. keypad lock device, to authorize multiple highly secure chemical passwords for information protection.

(Merck), Hg(II) acetate (Alfa Aesar), KCl (Sigma Aldrich), CuSO 4 .5H 2 O (Strem Chemicals) were all of highest analytical grade. Dulbecco's PBS buffer, pH=7.3, without calcium chloride and magnesium chloride used for metal ion detection was purchased from Sigma, Singapore. Millipore water was used for spectroscopy. Human urine was collected from a healthy person. Absorbance measurements were performed on a Perkin Elmer Envision 2104 Multilabel Reader. 96-well plates were used for absorbance measurements. Principal component analysis of the absorption spectra was performed using XLSTAT version 2015.2.02.17946. Figure S1. Effect of metallation on symmetry for porphyrin and chlorin.

S3
Upon metallation, porphyrin loses two protons and increases symmetry from D 2h to D 4h (Fig. S1). This increase in symmetry results in less molecular vibration along X-and Y-axis of the metalloporphyrin.
Hence, the intensity of Q-band absorption was reduced and only two peaks were observed. In case of chlorin which is an unsymmetrical, reduced form of porphyrin, the symmetry is reduced to C 2v . The unsymmetrical molecular vibration along the X-and Y-axis was increased in comparison to porphyrin, resulting in four Q-band absorption peaks. The symmetry of chlorin was further reduced by introducing a phenyl ring at the meso-position due to steric effect. Figure S2 illustrates porphyrin upon complexation with metals having different sizes, charges, binding constants resulted distortion at four pyrrole moieties of the porphyrin rings in different ways such as dome, saddle, ruffle, wave etc. which brings different molecular vibration along X-and Y-axis of the molecule and hence, generate different Q-band absorption spectra. Similar to porphyrin (Fig. S2), depending on charge and size of metal ions, several out-of-plane (Fig. 2, main text) and in-plane structural deformations can be observed. This further reduces symmetry and increases molecular vibration along the X-and Y-axis, and thus results in more Qband peaks.

3.b B-band absorbance response of sensor to different metal ions:
A solution of 1 (10 mM, 2 µL) in methanol was added to a solution of 196 µL PBS buffer /methanol (v/v=1/9) (pH= 7.3). To this solution containing 100 µM /198 µL of 1, a solution of a metal ion (100 mM, 2 µL) in water was added. The mixture was allowed to equilibrate for 5 min. Absorbance measurements were taken in 96-wall plates from 370-470 nm wavelengths for B-band spectra (Fig. S4). The spectra were recorded at steps of 5 nm.
The emission of the pure sensor (without metal ions) corresponds to an addition of water only.
All metal ions produced an identical B-band absorption spectrum at 410 nm, except for Cu 2+ and Fe 3+ . These latter metal ions exhibited a significant shift of the absorption peaks with an additional small shoulder-like peak. Hence, the B-band alone was not able to differentiate between many of the metal ions studied.  Each absorption spectrum represented is the average of five consecutive measurements (Fig. 3a, main text and Fig. S5). The emission of the pure sensor (without metal ions) corresponds to an addition of water only.  3.d Analysis of human urine samples: A solution of 1 (20 mM, 2 µL) in methanol was added to a solution of 196 µL Dulbecco's PBS buffer pH= 7.3: methanol (V:V=1:9) in 96-well plates. A solution of metal mixtures (2 µL) in human urine was added to above solution of 1 (200 µM, 198 µL). The mixture was allowed to equilibrate for 3 min. Absorbance measurements were taken at the wavelengths ranging from 270-800 nm. The spectra were recorded at steps of 5 nm as shown in figure S6. The emission of the pure sensor (without metal ions) corresponds to the addition of urine only. Each absorption spectrum represented is the average of six consecutive measurements (Fig. S6). The change in absorption intensities (ΔI abs = 1-I abs ) from a threshold intensity (I threshold = 1) at four different wavelengths (e.g. 515 nm, 545 nm, 595 nm, and 650 nm) is shown in Figure 5a (main text). Figure S6. Change in Q-band absorption intensities generated by 1 in Dulbecco's PBS buffer / methanol (1/9) (pH 7.3) upon addition of urine containing: A) chromium (0.53 mg/mL, dark red); B) chromium (0.26 mg/mL, green); C) cobalt (0.48 mg/mL, purple); D) cobalt ( 0.24 mg/mL, blue); E) mixture of chromium (0.53 mg/mL) and cobalt (0.48 mg/mL, orange); F) mixture of chromium (0.26 mg/mL) and cobalt (0.24 mg/mL, red); black spectrum represents addition of urine only.

3.e 2-Digit Keypad Lock:
A solution of 1 (20 mM, 2 µL) in methanol was added to a solution of 196 µL Dulbecco's PBS buffer/ methanol (1/9) (pH= 7.3) in 96-wall plates. A solution of a metal M0 (200 mM, 2 µL) in water as first input key was added to the above solution of 1 (200 µM, 198 µL). The mixture was allowed to equilibrate for 3 min. Subsequently, a solution of another metal M1 (200 mM, 2uL) in water was added as second input key and equilibrated for another 3 min. Absorbance measurement were taken at the wavelengths ranging from 270-800 nm. The spectra were recorded at steps of 5 nm.
Each absorption spectrum represented is the average of five consecutive measurements (Fig. 7a,  S7 text). The absorption of the pure sensor (without metal ions) corresponds to an addition of water only.
Few other metal combinations such Zn-Fe, Fe-Zn, Zn-Cu and Cu-Zn were tested in different sequence (Fig. S7). Zinc and iron metal in different sequence were producing same absorption signature whereas zinc and copper producing different absorption spectrum. It was observed that resultant absorption spectrum also depends on metal types along with concentration of metal and incubation time.

Principal Component Analysis (PCA).
The absorbance experiments were repeated for all the metal ions as Figure S3. A combination of 200 mM cobalt chloride and 200 mM cadmium chloride, or a combination of 200 mM iron chloride and 100 mM nickel chloride were used to study discrimination for mixture of metals. Both 200 mM or 100 mM solution of zinc chloride or iron chloride were used to study discrimination between different concentrations of metals. PCA was applied to distinguish between patterns generated by the absorbance intensities at six different wavelengths (e.g. 515 nm, 545 nm, 575 nm, 600 nm, 625 nm and 650 nm) in which maximal changes in intensities were observed. The PCA was able to discriminate all metal ions, different concentration of metal ions and combinations of toxic metal ions (Fig. 4, main text). The training set PCA was done taking these six wavelengths and color change further assisted to increase the accuracy (Table S1). Similarly, PCA differentiation between human urine containing cobalt and chromium metal