Facile construction of a hyperbranched poly(acrylamide) bearing tetraphenylethene units: a novel fluorescence probe with a highly selective and sensitive response to Zn2+

Thermo-responsive hyperbranched copoly(bis(N,N-ethyl acrylamide)/(N,N-methylene bisacrylamide)) (HPEAM-MBA) was synthesized by using reversible addition–fragmentation chain-transfer polymerization (RAFT). Interestingly, the zinc ion (Zn2+) was found to have a crucial influence on the lowest critical solution temperature (LCST) of the thermo-responsive polymer. The tetraphenylethylene (TPE) unit was then introduced onto the backbone of the as-prepared thermo-responsive polymer, which endows a Zn2+-responsive “turn-off” effect on the fluorescence properties. The TPE-bearing polymer shows a highly specific response over other metal ions and the “turn-off” response can even be tracked as the concentration of Zn2+ reduces to 2 × 10−5 M. The decrement of fluorescence intensity was linearly dependent on the concentration of Zn2+ in the range of 4–18 μmol L−1. The flexible, versatile and feasible approach, as well as the excellent detection performance, may generate a new type of Zn2+ probe without the tedious synthesis of the moiety bearing Zn2+ recognition units.


Introduction
Zinc is widespread in nature, such as in food (vegetables, wheat, fruit, and milk powder), beverages (beer, and wine), animals (cows, and chickens), soil, minerals and the human body. The detection of zinc has raised attention in elds including wastewater treatment, food analysis, environmental monitoring and medical treatment. [1][2][3] Zinc content in soil and irrigation water will have an effect on the quality of wheat. Zinc in milk powder has greatly affected the growth of infants. Zinc is one of the important nutritional attributes in many beverages. The zinc ion is the second most abundant transition metal ion in the body and plays an active role in various physiological processes including cell growth, adaptive immunity, gene transcription, apoptosis and neurotransmission. [4][5][6][7] An imbalance in zinc ions is closely associated with diseases, such as diabetes and neurodegenerative diseases. [8][9][10] Therefore, the accurate and sensitive detection of zinc ions would be signicant for practical application, such as exploring disease mechanisms, food analysis and environmental monitoring. A number of analytical techniques have been employed to detect and quantify zinc in biological systems, including inductively coupled plasma atomic emission spectroscopy, atomic absorption spectrophotometry and radioisotope detection. [11][12][13][14] Comparing with these methods, uorescent sensor with real-time monitoring, high sensitivity and noninvasive, shows attractive prospect in the detection of intracellular ion levels. [15][16][17] The past two decades have witnessed the development of a series of small moleculebased uorescent zinc ion probes. These synthetic Zn 2+ probes possessing diverse colors, response mechanisms, kinetics, and binding affinities, have proven to be valuable tools. [18][19][20][21][22][23][24][25][26][27][28][29][30][31][32] Despite the progress, for selectively detecting the zinc ions in live cells, a suitable group being of recognition in zinc ion has to be contained in the uorescent probes. Thus, there remains much to be done to develop novel probes, or seek other way for circumventing this problem.
Since zinc ion is of 3d 10 electron conguration, it is prone to chelate with other elements (N, S, O) and thereby inducing changes in the uorescence properties. Zinc ion can also bind with the polymers and result in the conformational changes or cross-linking structures. Zuckermann et al. 33 introduced a highaffinity zinc-binding function on a peptoid (N-substituted glycine polymer) two-helix bundle, where the conformational change in polymers was systematically investigated. Yano et al. 34 reported a Zn 2+ salt of ethylenemethacrylic acid ionomer which possesses pressure responsiveness. The response was caused by the change of coordination structure around zinc ion. The conformational change may also be able to trigger a uorescence change, and thereby endow a "turn on" or "turn off" response as the polymers binding with the Zn 2+ in uorescence polymers.
In this work, we synthesized a backbone temperaturesensitive hyperbranched polyamide in which the temperature sensitivity is imparted by temperature-responsive backbone. 35 The conformation of polymer chains will be transformed from the extended state into aggregation as the temperature exceeding lower critical solution temperature (LCST). More importantly, we found that the LCST i.e. conformational change in polymers can also be affected by the ions due to the existence of N element. Fortunately, in comparison with other cations, Zn 2+ exhibited a critical inuence on LCST. Aggregationinduced emission (AIE) uorogens rstly presented by Benzhong Tang 36-39 have attracted considerable interest in the uorescence probes eld for its unique uorescence emission mechanism. The AIE uorescent probes can be conferred with "responsive or recognition" feature by introducing a special group that affects the molecular aggregation aer reacting with other compound. Taking advantage of AIE feature, Benzhong Tang's group 40 synthesized a new probe (DEVD-TPE) that might monitor the cell apoptosis. Bo Tang's group 41 prepared a rapidresponse uorescent probe for hydrogen peroxide. Herein, the introduction of tetraphenylethene into the backbone of the thermo-responsive poly(acryl amide) may create a response of Zn 2+ to the uorescence change. A hyperbranched polymer bearing tetraphenylethene units was synthesized via RAFT method. The zinc ion stimulates the conformational change of polymer chains and thereby prohibit the intramolecular rotations of AIE units. The TPE-bearing polymer can monitor the zinc ion via the decrease in uorescence intensity, which can be called as a "turn off" response. The development of the AIEbased uorescent probe for detection of zinc ion without resorting to traditional Zn 2+ recognition units. The as-prepared polymer can be used as an efficient uorescent probe for detecting the Zn 2+ on trace levels.
Synthesis hyperbranched copoly(bis(N,N-ethyl acryl amide)/ (N,N-methylene bisacrylamide)) (HPEAM-MBA) Polymerization was carried out in a glass tube that sealed with a rubber cap under nitrogen atmosphere. EBA (336 mg, 2 mmol), MBA (46.2 mg, 0.3 mmol), BDAAT (32.43 mg, 0.115 mmol), APS (52.44 mg, 0.23 mmol) and DMF (8 mL) were added to a Schlenk tube. Oxygen was removed by repeated vacuumnitrogen cycles. Then, the polymerization was conducted at 70 C in an oil bath for 24 h. Aerwards, the obtained polymer was precipitated by dropping the solution into a large excess of acetone to remove the excessive monomers and impurities. The precipitated polymer was separated by centrifugation, and then dissolved in water. The nal product HPEAM-MBA was given aer lyophilization (Scheme 1). Oxygen was removed by repeated vacuum-nitrogen cycles. Then, the polymerization was conducted at 70 C in an oil bath for 24 h. Aerwards, the obtained polymer was precipitated by dropping the solution into a large excess of acetone to remove the excessive monomers and impurities. The precipitated polymer was separated by centrifugation, and then dissolved in water. The nal product HPEAM-TPEAH was given aer lyophilization (Scheme 2).

Measurement of lower critical solution temperature (LCST)
LCSTs were dened as the temperature corresponding to 50% transmittance of aqueous solution during the heating process.
UV-visible 2550 Spectrophotometer (Shimadzu, Japan) at 500 nm with the heating rate as 0.1 C min À1 , and the UV-vis spectra of polymer solution was also obtained through same spectrophotometer from 250 nm to 400 nm. Equivalent amounts of polymers were added to 4.0 mL of aqueous solution of different salts. The mixture was cooled at 4 C for 10 min to ensure that the polymer was dissolved completely.

Cytotoxicity study
The cytotoxicity of HPEAM-TPEAH on Hela cells was examined as follows: Hela cells (5 Â 103 cells) in 100 mL of DMEM containing 10% FCS were plated in a 96-well plate and incubated for 24 h in a humidied atmosphere of 5% CO 2 in air at 37 C (Sanyo, Model MCO-18AIC, Japan) Then 100 mL of a HPEAM-TPEAH in DMEM containing 10% FCS and 2% DMSO was added to each well. Incubation was carried out for 24 h at a HPEAM-TPEAH concentration of 0.5 mM (nal DMSO content was 1% in all cases). Cells without HPEAM-TPEAH treatment were used as control group. Aer 24 h of treatment, MTT dye solution (20 mL, 5 mg mL À1 ) was added to each cell. Cells were incubated for another 4 h and then analyzed using a microplate spectrophotometer (BioRad Model 3550, USA) at 570 nm. The percentage cell survival was calculated by normalization with respect to the value for no HPEAM-TPEAH treatment.

Characterization
Infrared spectroscopy was measured on a Varian 640-IR Fourier infrared spectrometer (Varian, Palo Alto, CA, USA) by using KBr pellets and solution in D 2 O. CaF 2 transmission windows and 0.05 mm Mylar spacers were used. For each spectrum, 256 interferograms of 2 cm À1 resolution were co-added. From the spectrum of each sample, a corresponding D 2 O spectrum was subtracted. All the spectra were baseline-corrected and normalized. All data processing was performed with Resolution soware. The samples were investigated at mass concentration of 0.5 wt% in D 2 O. 1 H NMR spectra were recorded on a Bruker AVANCE-600 600 MHz spectrometer (Bruker, Germany). The chemical shis are given in parts per million (ppm). Fluorescence spectrum analysis was measured on the RF-5301-PC (Shimadzu, Japan) uorescence spectrometer lighttransmittance of the polymer solution was measured on a temperature-controlled. Molecular weight, polydispersity, Mark-Houwink index (a) and intrinsic viscosity (IV) were determined by Viscotek 270-MDSEC (Malvern Instruments Ltd., Malvern, US) equipped with the differential refractive index (RI), viscometer, and two-angle light scattering (LS, Malvern Instruments Ltd., Malvern, US) triplet detectors. The eluent was 0.1 mol L À1 NaNO 3 aqueous solution at a ow rate of 1 mL min À1 at 15 C. For MDSEC, narrow dispersion polyethylene oxide std-PE022K was used to calibrate the instrument. TEM studies of copolymers were performed on a Tecnai G2 F20 S-TWIN (FEI Company, Hillsboro, USA) dedicated to cryogenic temperatures, operated at an acceleration voltage of 200 kV at low electron-dose conditions. A small drop of clear and cloudy copolymers solution (1 mg mL À1 ) containing 1 wt% phosphotungstic acid (PTA) was placed on the Cu grid with holey carbon lm held by tweezers of the vitrication system Vitrobot (FEI Company, Hillsboro, USA). DLS measurements were performed on a Brookhaven BI-200 goniometer (Brookhaven, New York, USA) with vertically polarized incident light of wavelength l ¼ 532 nm supplied by a helium-neon laser operated at 75 mW and a Brookhaven BI-4700 AT digital autocorrelator (Brookhaven, New York, USA).

Results and discussions
In a previous study, hyperbranched poly(bis(N,N-propyl acryl amide)) 44 and poly(bis(N,N-ethyl acrylamide))s 45 had been successfully synthesized with high yield by reversible additionfragmentation chain transfer polymerization (RAFT) method. In addition, the lower critical solution temperature (LCST) oen ranges from 19 to 28 C, and 6 to 17 C, respectively. In this study, the synthesis hyperbranched copoly(bis(N,N-ethyl acryl amide)/(N,N-methylene bisacrylamide)) (HPEAM-MBA) was conducted in order to achieve a desirable LCST near human body temperature. Fortunately, the LCST of HPEAM-MBA is determined to be 36.4 C (details were shown in the ESI †).
It is known that the monomer possessing a multi vinyl monomer (MVM) is usually employed to fabricate polymers with crosslinked network structures, even if only small amounts of MVM participates in the free radical polymerization. However, Sherrington 46 and Guan 47 respectively prepared hyperbranched polymers with MVM as branching units via a chain transfer controlled free radical polymerization. The crosslinking was effectively inhibited by using thiol or catalytic chain transfer catalyst species. In Sato's study, [48][49][50][51] the MVM was also used as the branching species in which the radical Scheme 2 Synthesis of HPEAM-TPEAH. polymerization was well controlled by a chain termination. Recently, controlled free radical polymerization was used in the MVM reaction system, which is able to reduce the reaction rate and avoid cross-linking in polymerization. In addition, the preparation of hyperbranched polymers can also be well controlled by copolymerization with other monomers via living polymerization. For an example, Wang and Howdle 52 synthesized hyperbranched polymers with highly branched structures via atom transfer radical polymerization (ATRP). In our study, RAFT polymerization was adopted for the synthesis of hyperbranched copoly(bis (N,N- The nitrogen in the backbone of as-prepared polymer is able to coordinate with metal ion to form M n+ -N x complex and result in the changes in hydration. Therefore, the LCST of HPEAM-MBAs may also be changed by the addition of metal ions. The inuence of metal ions, including Na + , K + , Mg 2+ , Ca 2+ , Mn 2+ , Fe 2+ , Zn 2+ , on the LCST was investigated. Zn 2+ was selected as a represent metal ion whose turbidity curves that reect the inuence of its concentration on the LCST values are shown in Fig. 1. As expected, the turbidity curves are shied to high temperature with the concentration of Zn 2+ increasing. The turbidity curves of other metal ions are displayed in the ESI † respectively. Aer extracted the LCST values, the plot of LCST values against concentration are illustrated in Fig. 2. It can be observed from the Fig. 2 that all of these metal ions exhibit the hydrotropy ability which can be owing to the salting-in effect. The sequence salting-in effect is Fe 2+ < Ca 2+ < Mg 2+ < Mn 2+ < K + z Na + ( Zn 2+ , which is consistent with the typical Hofmeister cation series. [55][56][57] It should be noted that the tendency of the plots of LCST vs. [Zn 2+ ] displays a signicant difference over than those of other metal ions that the LCST is dramatically increased even [Zn 2+ ] < 1 Â 10 À5 M. In order to compare the inuence of Zn 2+ on the LCST with other metal ions more quantitatively, the parameter S (named sensitivity) is introduced and dened as the reciprocal of N when the transmittance% of the solution is 50%. For getting the parameter S, the transmittance% of the solution under different concentration of metal ions was recorded and shown as Fig. 3. All of the curves display a similar tendency that the transmittance% of the solution increases as the N Zn 2+ ascends. Also, the temperature affects the amount of the Zn 2+ . Accompany with temperature of the solution rising up, the N Zn 2+ obviously increase. As pointed out in the aforementioned part, high temperature will destroy the hydrogen bonds between the water molecules and amide groups, which means the hydrophobicity of HPEAM-MBAs is proportional to the temperature of solution, therefore the more N Zn 2+ is needed as the temperature getting higher. The curve of N Zn 2+ with temperature is inserted into the lower right corner of Fig. 3. According with the denition of S, the N Zn 2+ decreases as the increase of S. The curves of transmittance% vs. N of other metals are in the ESI † and display the similar tendency to that in Fig. 3. Although the value of S is the highest as the temperature is closed to LCST, the S Zn 2+ / S metals is not the highest value which reveals that the HPEAM-MBAs solution does not display good selectivity towards Zn 2+ . Encouragingly, the value of the N Zn 2+ is far below the N of other metals when the temperature of solution is 20 degrees above the LCST, and the ratio of S Zn 2+/S metals can reach 1.28 Â 10 3 . To  graphically show the selectivity of HPEAM-MBAs solution towards Zn 2+ , the S Na + is chose as the reference because that the Na + is the most abundant ion in the human body. As shown in Fig. 4, the ratio of S Zn 2+ /S Na + is 5.75 Â 10 2 when the temperature is 57 C. These results indicate that HPEAM-MBAs solution is of high sensitivity in assessing Zn 2+ , and the HPEAM-MBAs solution can selectively recognize the Zn 2+ when the environment temperature is 20 degrees above the LCST.
Infrared absorption spectra was employed to investigate the interaction of HPEAM-MBA and Zn 2+ . Fig. 5(A) shows the infrared absorption spectrum of HPEAM-MBA (in KBr). The characteristic adsorptions at 1652 cm À1 represents the stretching vibration of C]O, the peak located at 1541 cm À1 corresponds to C-N stretching vibration in the trans-associated secondary amide, and the signal at 1439 cm À1 is the N-H bending vibration in cis-amide. Fig. 5(B) is the infrared absorption spectra of HPEAM-MBA measured in D 2 O, and measured with the presence of the metal ions in D 2 O, respectively. Due to the interaction between the D 2 O and HPEAM-MBA, the signals corresponding to C]O and N-H shi to the lower wavenumber region. For Na + , K + , Mg 2+ and Ca 2+ , the characteristic adsorptions of N-H shied to the lower wavenumber region, while the vibrational peak of C]O stretching unchanged, which indicates that the addition of Na + , K + , Mg 2+ and Ca 2+ promotes the dissociation of amide group. As the Zn 2+ , the characteristic adsorptions of C]O and N-H shied from 1634 cm À1 to 1651 cm À1 and 1450 cm À1 to 1472 cm À1 respectively. Evidently, the Zn 2+ binds with not only N atom, but also the O atom which can be due to its 3d 10 electron conguration. Therefore, Zn 2+ shows a much stronger salting-in effect as comparing to those of other metal ions.
Because that the HPEAM-MBA solution shows the feature of selectively recognizing the Zn 2+ as the environment temperature is 20 degrees higher than the LCST, the HPEAM possessing lower LCST was selected as the platform to copolymerize with Fig. 3 The curves of transmittance% vs. N Zn 2+ of HPEAM-MBA at different temperatures. Polymer concentration: 5 mg mL À1 . Fig. 4 The three-dimensional histogram of temperature vs. S metals /S Na + . TPEAH. The TPE-bearing polymer was synthesized by using reversible addition-fragmentation chain-transfer polymerization (RAFT). It shows a highly specic response over other metal ions. The mechanism of polymerization was shown in Scheme 2 and the detail was shown in ESI. † The resulting copolymer HPEAM-TPEAH(with yield as 57%.) was characterized by FTIR, and 1 H NMR, and the detailed assignment is shown in ESI. † The ratio of EAM\TPEAH is 2.9 on basis of the 1 H NMR results. The HPEAM-TPEAH is difficult to be dissolved in water, while the aqueous system containing HPEAM-TPEAH displays an obvious uorescence emission which can be owing to the aggregation-induced emission (AIE) effect of TPE units. Generally, the incorporation with metal ions would raise the solubility of the polymers in aqueous system and thereby result in the decreasing of uorescence intensity, which can be named as the "turn off" response. The "turn off" response of HPEAM-TPEAH on the metal ions including Zn 2+ , Mn 2+ , Na + , K + , Ca 2+ , Mg 2+ and Fe 2+ were investigated respectively. The results depicted on ESI † shows that the "turn off" response is hard to be detected for Mn 2+ , Na + , K + , Ca 2+ , Mg 2+ and Fe 2+ , while the addition of HPEAM-TPEAH into the Zn 2+ solution triggers a dramatically decrease in uorescence intensity. Furthermore, a simulate body uid (Na: 140 mmol L À1 ; K + : 5.5 mmol L À1 ; Ca 2+ : 1.33 mmol L À1 ; Mg 2+ : 5 mmol L À1 ; Fe 2+ : 3.0 mmol L À1 ) was prepared, where the "turn-off" cannot be detected with the addition of HPEAM-TPEAH into simulate body uid (shown in Fig. 6). While the uorescence intensity signicantly decreased as the addition of Zn 2+ into the above system. Results of the linear regression for I 0 -I and [Zn 2+ ] show that the linear correlation coefficient (R 2 ) is greater than 0.99 which indicates that the I 0 -I and [Zn 2+ ] shows a good linearity relationship in the range from 2 Â 10 À6 to 2 Â 10 À5 mol L À1 (details were shown in ESI †). The results show that HPEAM-TPEAH exhibits a highly specic "turn-off" response of Zn 2+ over other metal ions.
As aforementioned result shows that the N Zn 2+ or the concentration of Zn 2+ ([Zn 2+ ]) can affect the LCST of HPEAM-MBAs, which means [Zn 2+ ] can tune the conrmation of the HPEAM-MBAs. Therefore, the [Zn 2+ ] should be associated with the PL intensity of HPEAM-TPEAH. As shown in Fig. 7, along with the increase of [Zn 2+ ], the PL intensity gradually decrease. When the [Zn 2+ ] reaches 0.02 mmol L À1 , the PL intensity is very low and keeps unchanged even if the [Zn 2+ ] is higher. The relationship of PL intensity and concentration of other metals was also investigated and the corresponding curves are shown in ESI. † The PL intensity of HPEAM-TPEAH solution still unchanged, even if the [metals], including Na + , K + , Ca 2+ , Mg 2+ and Fe 2+ , reaching 100 mmol L À1 . The accuracy of the detection of Zn 2+ by using HPEAM-TPEAH is also evaluated. The plots of [Zn 2+ ] against I 0 -I are displayed in Fig. 8, where I 0 represents the initial PL intensity, and I is the PL intensity under the Fig. 6 Fluorescence intensity curves of HPEAM-TPEAH under simulate body fluid with (i) or without Zn 2+ (ii). The simulate body fluid contains 140 mmol L À1 Na + , 5.5 mmol L À1 K + , 1.33 mmol L À1 Ca 2+ , 5 mmol L À1 Mg 2+ , Zn 2+ and 3.0 mmol L À1 Fe 2+ . The concentration of Zn 2+ is 0.02 mmol L À1 .  show that the linear correlation coefficient (R 2 ) is greater than 0.995 which indicates that the I 0 -I and [Zn 2+ ] shows a good linearity relationship in the range from 4 Â 10 À6 to 1.8 Â 10 À5 mol L À1 . Traditionally, the synthesis of polymers bearing the detection capacity upon zinc ions were usually based on the introduction of Zn 2+ recognition units into the backbones. 15,24,25 For an example, the results given by the literature indicated that the detection specicity would be signicantly disturbed when the concentration of other metal ions, such as Na + , K + , Mg 2+ and Ca 2+ , reaches 50 mM. 24,25 In this study, HPEAM-TPEAH was prepared and used for the detection of Zn 2+ based on the interaction between Zn 2+ and N and O atoms accompanying with the aggregation-induced emission (AIE) effect. The results indicate that HPEAM-TPEAH holds a desirable specicity on the detection of Zn 2+ when the concentration of those metal ions reaches 100 mM, where the I 0 -I and [Zn 2+ ] shows a good linearity relationship in the range from 2 Â 10 À6 to 2 Â 10 À5 mol L À1 .
The cytotoxicity of HPEAM-TPEAH was tested by using Hela cells and the results are given in Fig. 9. HPEAM-TPEAH shows very low cytotoxicity (over 95% cell viability) within 24 h of incubation time, in which the concentration ranges from 1.0 Â 10 À6 to 5.0 Â 10 À5 mol L À1 .

Conclusions
In summary, a new thermo-responsive hyperbranched copoly(bis(N,N-ethyl acryl amide)/(N,N-methylene bisacrylamide)) (HPEAM-MBA) was synthesized by using RAFT. Zn 2+ was found to be able to bind with N atom and O atom in the backbone of HPEAM-MBA. When the environment temperature is 20 degrees above the LCST, HPEAM-MBAs solution exhibit high sensitivity in assessing Zn 2+ , which make it possible to gain a new selectively recognizing system. Furthermore, a new Zn 2+ specic uorescent probe (HPEAM-TPEAH) was developed by the copolymerization of N,N-ethyl acryl amide and TPEAH in this study. The introduction of TPE units gives a highly specic and sensitive Zn 2+ -sensitivity to HPEAM-TPEAH, as a result the uorescence exhibits a highly specic "turn-off" response for Zn 2+ .

Conflicts of interest
There are no conicts to declare.