Fabrication of AO/LDH fluorescence composite and its detection of Hg2+ in water

Divalent mercury ion (Hg2+) is one of the most common pollutants in water with high toxicity and significant bioaccumulation, for which sensitive and selective detection methods are highly necessary to carry out its detection and quantification. Fluorescence detection by organic dyes is a simple and rapid method in pollutant analyses and is limited because of quenching caused by aggregation dye molecules. Hydrotalcite (LDH) is one of the most excellent carrier materials. In this study, an organic dye acridine orange (AO) was successfully loaded on the LDH layers, which significantly inhibited fluorescence quenching of AO. The composite AO/LDH reaches the highest fluorescence intensity when the AO initial concentration is 5 mg/L. With its enhanced fluorescent property, the composite powder was fabricated to fluorescence test papers. The maximal fluorescence intensity was achieved with a pulp to AO/LDH ratio of 1:5 which can be used to detect Hg2+ in water by naked eyes. Hg2+ in aqueous solution can be detected by instruments in the range of 0.5 to 150 mM. The novelty of this study lies on both the development of a new type of mineral-dye composite material, as well as its practical applications for fast detection.

Hydrotalcites (LDH), a family of two-dimensional anionic layered materials, have attracted extensive interests, because of their novel properties and extensive application for catalysts, molecular container, flame retardants, acid absorbents, drug delivery, etc [22][23][24][25] . LDH are minerals with a high anion exchange capacity in contrast to the well-known clay minerals 26 , which have cation exchange properties 27 . The structure of these minerals consists of brucite-like positively charged layers resulting from partial substitution of the original Mg 2+ by Al 3+ 28  In this study, we prepared a fluorescence test paper for the detection of Hg 2+ . Due to quenching effect, the photoactive molecule AO was loaded on the surface of LDH to reach maximal separation and minimal aggregation. Fluorescence quenching was assessed using variety of solutions. The AO/LDH fluorescence test paper resulted in a superior response to Hg 2+ detection.

Results and Discussion
Preparation Characterization of AO/LDH. The AO adsorption data were fitted using the Langmuir and Freundlich adsorption models (Fig. 1a). The Langmuir 30 equation can be described as: where q e (mmol·g −1 ) and q max (mmol·g −1 ) are the amount of AO adsorbed at equilibrium time and the maximum quantity of AO per unit clay to form a complete monolayer on the surface, respectively. C e (mmol·L −1 ) is the equilibrium solute concentration. K L (L·mmol −1 ) is Langmuir affinity, which represents enthalpy of sorption and related to temperature 31 . The Freundlich 32 model has the form: where K F and n are the Freundlich constants related to the sorption capacity and sorption intensity, respectively 33 . The r 2 was 0.98 when the data were fitted to the Langmuir model. The AO adsorption capacity was 6 mg/g. In comparison, the Freundlich isotherm model which involves multilayer adsorption on adsorption surface fitted Figure 1. AO adsorption isotherm on LDH (a) and Zeta potential of LDH after AO adsorption from different initial concentrations (inset). Characterization of raw LDH by X-ray diffraction (b) and TEM (inset). X-ray diffraction patterns of LDH as affected by initial AO concentrations (c).
the experimental data poorly, with an r 2 value of 0.93, and therefore was not adopted in this study. The positive properties of the surface are decreased after adsorption, indicating that some molecules are loaded onto the surface of LDH (Fig. 1a (Fig. 1b), which are typical for crystallized Mg-Al LDH (CO 3 2− ) phase. AO was retained on the surface of LDH instead of intercalated into the interlayer of LDH (Fig. 1c).
The morphologies of raw LDH and LDH after AO adsorption were characterized by SEM (Fig. 2). The samples are quite similar in shape (tabular) and uniform in size (40-500 nm). No changes in crystal morphology, nor in crystal size was observed after AO uptake on the LDH, suggesting no particle disintegration after upload of AO on LDH (Fig. 2b).
The luminescence properties of AO/LDH and fluorescence test paper (FTP). Due to the lack of strong luminescence, AO powder was rarely used as a luminescent material. This very low luminescence intensity was attributed to concentration quenching 35 . After being loaded on the surface of LDH, the luminescence intensity improved significantly (Fig. 3a), perhaps as a result of separation of AO molecules on the surface of LDH. LDH is equal to a solid dispersant which can disperse AO successfully. Thereby the radiationless transition reduces. The raw LDH is not luminous. The fluorescence intensity increases gradually with the increase of AO when different concentrations of AO are dispersed on the layer of LDH. The luminescence intensity of AO/LDH increased as the initial AO concentration increased to 5 mg/L (or at the AO loading of 4 mg/g), at which the highest luminescence intensity of AO/LDH was obtained (Fig. 3b). However, AO fluorescence molecules continue to reunite as the concentration of AO increase resulting in luminescence intensity decreasing.
To fabricate test strips similar to a litmus paper, pulp needs to be added. The amount of pulp added affected the performance of AO/LDH (Fig. 4). Visualization of FTP prepared with different pulp to AO/LDH ratios under  Fig. 4a. The AO/LDH power can be further dispersed when combing with pulp. However, it is the best proportion when the ratio of pulp to AO/LDH is 1:5 to achieve the largest luminous intensity (Fig. 4a). Thus, the pulp to AO/LDH ratio of 1:5 was used for fabrication of the FTP to be used to detection of other solutes in aqueous solution. The luminescence property of the composite material changed significantly under different temperatures, suggesting that the thermal stability of the materials is a principal factor affecting its application. The temperature rise reduces the luminescence efficiency of the material, which is called temperature quenching. Quenching means that there is no radiative transition between the electron levels. The rate of radiationless transition increases with temperature, causing the temperature quenching. Despite the luminous intensity of FTP weakened as temperature increased, it did not disappear until about 120 °C (Fig. 4a inset). AO, loaded on the layer of LDH, is under the protection of LDH layer, making it better adaptive to changing ambient temperature.
Screening and quantification of Hg 2+ by FTP. The FTP was tested to screen its response to a variety of solutes at a concentration of 0.1 M (Fig. 5). Among the tested solutes the fluorescence response to Hg 2+ was quite remarkable, while quenching by other solutes did not occur, indicating a high selectivity of the FTP for Hg 2+ .
The photoluminescence (PL) quenching of FTP in the presence of Hg 2+ in aqueous solution was further investigated as a function of initial Hg 2+ concentration. Over the concentration range of 0.5-200 mM, an exponential decrease in fluorescence intensity was observed (Fig. 6a). A full PL quenching can be observed with naked eyes at an Hg 2+ concentration of 50 mM. The detection limit of Hg 2+ using FTP by naked eyes was 5 mM. The images of FTP with addition of Hg 2+ under UV light showed a consistent tendency (Fig. 6b).
Mechanism of fluorescence enhancement. Minute adsorption of AO from water by LDH was observed at an input concentration of 0.5-80 mg/L, resulting in an AO adsorption capacity of 6 mg/g that could be reached at an initial concentration of 40 mg/L. The LDH is an anion exchanger with positively charged surfaces while the  AO is positively charged when the solution pH is lower than its pKa value of 10.4 24 . Thus, the influence of equilibrium solution pH on AO uptake was further assessed. As the pH increased from 2 to 10, the AO uptake increased progressively (Fig. 7). The total charge of a mineral is made of permanent charge, which is attributed to isomorphic substitution in crystal lattice (Mg 2+ substitution by Al 3+ in case of LDH 36 ), and pH dependent charge, due to protonation/deprotonation 37 . The metal ions of LDH will dissolve under the condition of strong acidity (pH < 4) and its structure changes. LDH can be dissolved affecting its adsorption effect under the condition of strong acidity 38 . The phenomenon of LDH dissolution is not observed under other pH conditions. The adsorption of AO from pH 2 to 4 becomes better because of the less dissolution of LDH. The points of zero charge (PZC) refers the pH value at which the total charges was 0 under a certain temperature, pressure and medium. PZC can be further divided into point of zero net proton charge (PZNPC) and point of zero net charge (PZNC) 39 . The research found that hydrotalcite exists PZC 40 . The adsorption proton charge on the surface of hydrotalcite would change with the pH value of the medium. However, charges on the surface of LDH could change with the pH of system. The basal surfaces (perpendicular to crystallographic c) and prismatic surfaces (parallel to c) may show different affinity for AO molecules under different pH conditions due to electrostatic effect. The prismatic surface may bear negative charges under high pH due to deprotonation, which may result in an increased AO uptake (Fig. 7).
To support the electrostatic interactions between AO and LDH surfaces, molecular dynamics simulation studies were performed under low and high pH conditions for the interactions between the basal or prismatic surfaces. The positive charge of AO is located at the dimethylammonium group. Under both low and high pH conditions the interactions between AO and the basal surface of LDH were all minimal (Fig. 8), suggesting that  the basal surfaces were dominated by permanent surface charges, resulting in a net repulsion between the positively charged AO and the LDH basal surface. A similar result was found for the interactions between AO and prismatic surfaces of LDH under low pH condition (Fig. 9a). However, under a high pH condition, it seemed that the dimethylammonium group of AO moved towards the prismatic surfaces of LDH (Fig. 9b). Thus, it is very likely that the deprotonation of the prismatic surface of LDH played a crucial role in AO uptake.

Conclusions
In this study, an organic dye acridine orange (AO) was successfully loaded on LDH surfaces. The maximal AO separation resulted in significantly inhibition in fluorescence quenching. With its enhanced fluorescent property, the composite powder was mixed with pulp to fabricate a fluorescence test paper, which can detect Hg 2+ in water by naked eyes with a detection limit of 5 mM. Further optimization may find its practical application in environmental and wastewater treatment fields.

Experimental Section
Experimental materials. The Mg-Al-LDH was obtained from Ica Shanghai biological technology co., LTD (Russia) and was used as is without further purification. The AO (CAS # 65-61-2) used was in an HCl form purchased from Sigma-Aldrich. Its molecular formula is C 34 H 40 C l4 N 6 Zn and its formula weight is 739.94. It had a pKa value of 10.4, below which the molecule would be protonated to form AOH + . NaOH, HCl, distilled water was brought from Beijing chemical factory. Mercuric chloride was from best reagent.
Methods. For batch study, 0.1 g of LDH and 25 mL of AO solution at concentration of 0.5-80 mg/L was added to each 50 mL centrifuge tube. The mixtures were shaken at 300 rpm for 6 h and then were centrifuged at 7500 rpm for 2 min. The supernatants were then filtered through 0.22 μm syringe filters before being analyzed by a UV-vis spectrophotometer. The amount of AO adsorbed was determined from the differences between the initial and equilibrium concentrations. All experiments were done in duplicate. The solid product is denoted as AO/LDH. Similar experiments were done under different pH conditions.
TO fabricate the FTP, 1 g of pulp and varying amounts of AO/LDH at the pulp to AO/LDH ratios of 1:10, 1:5, 1:2, 1:1 and 2:3 were added together with 100 mL of distilled water to each 200 mL Erlenmeyer flask. The mixture was stirred for 10 h to increase the integration. The final product was named fluorescence test paper (FTP) with good flexibility, high luminescent property and controllability.
To assess the solute detection, the following solutions at a concentration of 0.1 M were initially screened for fluorescence quenching of AO: Al 3+ , Ba 2+ , Bi 2+ , Cd 2+ , Co 2+ , CTAB, Hg 2+ , Fe 3+ , Li + , dimethylimidazole, Mn 2+ , Pb 2+ , Ethyl Alcohol. They were dropped onto the FTP. Then, the fluorescence intensities were determined using a fluorescence spectrophotometer (HITACHI, F4600). To evaluate the Hg 2+ detection response, drops of Hg 2+ solutions of different concentrations were placed onto the FTP. The initial Hg 2+ concentrations were from 0.5-200 mM. The florescence intensity was measured by a fluorescence spectrophotometer (HITACHI, F4600).

Methods of analyses.
The absorbance values of equilibrium solutions were measured using a UV-vis spectrophotometer (T6 New Century) at wavelength of 490 nm.
Powder XRD analyses were performed on a Rigaku D/max-IIIa diffractometer with a Ni-filtered CuKa radiation at 30 kV and 20 mA. Orientated samples were scanned from 3° to 70° at 8°/min with a scan step of 0.02°/step.
Scanning electron micrography (SEM) was recorded using a JEOL-IT300 Scanning electron microscope at an accelerating voltage of 10 kV.
Photoluminescence emission (PL) spectra were acquired on a fluorescence spectrophotometer (HITACHI, F4600) over the range of 500−700 nm with a photomultiplier tube operated at 800 V. A 150 W xenon lamp was used as the excitation source, at an excitation wavelength of 488 nm. The scan speed was set at 240 nm/min.