Analyst

Published on 20 January 2014. Downloaded by Cornell University on 22/04/2014 13:58:44.

PAPER

View Article Online View Journal | View Issue

Cite this: Analyst, 2014, 139, 2515

Alternating polymer micelle nanospheres for optical sensing† Anna Kisiel, Katarzyna Kłucin´ska, Zuzanna Głe ˛bicka, Marianna Gniadek, Krzysztof Maksymiuk and Agata Michalska* A novel concept of nanosized fluorimetric sensors is proposed, using alternating polymers as self assembling micelles that can be crosslinked resulting in stable polymeric nanoparticles. The thus obtained nanospheres have sizes close to 250 nm or 130 nm, depending on the preparation procedure and the negative surface charge, due to the presence of carboxyl groups on the surface. By a simple procedure, the nanospheres can be effectively loaded with compounds of choice, e.g. ionophores and ion-exchangers previously used to induce ionic sensitivity in polyacrylate or poly(vinyl chloride) microand nanospheres (miniature optrodes), thus allowing for optical or fluorimetric quantification of analytes. As a proof of concept, H+ sensitive colorimetric and fluorimetric sensors and K+ fluorimetric sensors using classical optrode approach were prepared and tested. The obtained sensors were characterized by high sensitivity, fast and reversible responses. Both K+ and H+ sensors were characterized by a broad response range resulting from the significant effect of processes occurring on the surface of the

Received 18th December 2013 Accepted 20th January 2014

nanospheres. Due to this effect, the fluorimetric responses of the obtained spheres are significantly different from those typically observed for miniature optrode systems, and were linear within a range of at least 5 logarithmic units of analyte concentration. As shown, the surface groups of the herein

DOI: 10.1039/c3an02344c

proposed nanospheres can be used for the covalent linking of fluorophores that can be used as markers

www.rsc.org/analyst

(if applied alone) or as reference dyes for fluorescent ion-sensitive nanospheres.

Introduction Polymeric micro- and nanospheres are attractive systems for imaging, detection or drug delivery. There are many potential analytical benets of the application of polymeric micro- or nanospheres; they include, among others, faster response times (compared to lms) or the possibility of direct intracellular detection of analytes. Micro- and nanoparticles are especially attractive for optical detection based analytical techniques, with an emphasis on uorimetry.1 Fluorimetric sensing employing polymeric micro- or nanoparticles can be achieved upon incorporation of a uorescent dye (uorescent receptor) within the matrix – for this purpose, latex, silica or polyacrylamide are oen used, e.g. ref. 2–4. An alternative analytical approach benets from the concomitant presence in the polymeric micro- or nanospheres matrix of both a (uorescent silent) receptor and a uorescent active transducer. This approach allows the quantication of analytes that cannot easily be determined otherwise, e.g. taking advantage of experience and reagents originally developed for ion-selective

Department of Chemistry, Warsaw University, Pasteura 1, 02-093 Warsaw, Poland. E-mail: [email protected] † Electronic supplementary 10.1039/c3an02344c

information

(ESI)

This journal is © The Royal Society of Chemistry 2014

available.

See

DOI:

electrodes, i.e. micro- or nanospheres, can be considered as miniaturized optrodes. Naturally, for these type of miniature sensors, polymer matrices have mostly been used to prepare ion-selective electrodes or optrodes – plasticized poly(vinyl chloride)5–10 or polyacrylates, e.g. poly(n-butyl acrylate),11 poly(dodecyl methacrylate)12,13 or others.14 Despite the well documented benets of the application of these polymers for the preparation of classical ion-selective electrodes or optrodes,15 this approach has some disadvantages. For example, preparation of microspheres from poly(vinyl chloride) or solution processable polyacrylates requires a specialized set up and experimental expertise. To prepare poly(vinyl chloride) microspheres of diameters in the mm (or tens of mm) range, a special particle casting electronic apparatus is required e.g. ref. 6, 7, 9 and 10; less oen, an alternative approach based on injection of membrane solution into the sonicated surfactant aqueous solution has been used, resulting in beads of mm sizes.5,8 Polyacrylate microspheres can also be prepared by microemulsion polymerization (performed e.g. at elevated temperatures4,12,13 or under UV radiation11). Most oen, the resulting beads are of micrometric size; preparation of nanometric size spheres is certainly more demanding when using poly(vinyl chloride) or polyacrylate as matrices. For most of the thus obtained polymeric microspheres leading to a change of emission/absorption spectra occur within the whole volume of the polymeric

Analyst, 2014, 139, 2515–2524 | 2515

View Article Online

Published on 20 January 2014. Downloaded by Cornell University on 22/04/2014 13:58:44.

Analyst

material, leading to sigmoidal shape responses with a linear response range usually covering about 3 orders of magnitude e.g. ref. 11. However, as was shown recently, limitation of the accessible area or slow equilibration within the sphere volume results in uorimetric signal formation in the top surface layer of the microsphere, ultimately leading to pronounced changes in the obtained response pattern and linear dependence covering as many as 7 orders of magnitude.16 From the materials point of view, some limitations in the practical application of polymeric microparticles can result from leakage of the plasticizer from the poly(vinyl chloride) matrix or relatively slow diffusion of ions within polyacrylate polymers.17–19 Thus, the application of novel types of polymeric materials offering easier preparation of nanometric size particles seems to be an interesting alternative. In recent years there has been signicant interest in nanometric size polymeric particles, especially in the context of drug delivery, e.g. ref. 20–22; in addition, other materials such as lipid bilayers have been applied to prepare nanocapsules.23 Lately, signicant emphasis in this context has been placed on block copolymers with amphiphilic character. The clear advantage of these materials, apart from commercial availability and characteristics established by the producers, is a tendency to self-assemble into spherical micelles, e.g. ref. 24 and 25. The formed structures – polymeric micelles – are usually sub-100 nm in size, e.g. ref. 26 and 27. A variety of materials has been investigated with respect to drug delivery and target release; their ability to cross biological barriers and cellular membranes are important properties. Moreover, contrary to surfactant micelles, polymeric amphiphilic micelles usually have low CMC values, and they are reported to be stable both in solutions and in vivo, e.g. ref. 24. To the best of our knowledge, polymeric micelles for sensing purposes, especially uorimetric sensor construction, have not been extensively applied. A ratiometric Fe(III) sensor was prepared using poly(ethylene oxide)-b-polystyrene micelles of nanometric size.28 However, uorescent dyes, e.g. pyrene or perylene, have been encapsulated within polymeric micelles, e.g. to study membrane pores.29 A particular example of amphiphilic copolymers, useful in the preparation of polymeric micelles, are alternating polymers. Especially interesting are alternating polymers in which every second unit is maleic anhydride, e.g. poly(maleic anhydride-alt1-tetradecene) or poly(maleic anhydride-alt-1-octadecene), as upon spontaneous hydrolysis anhydride groups are transformed into carboxyl groups, resulting in stability of the thus obtained particles (micelles) in hydrophilic (e.g. aqueous) environments.30–33 Additionally, crosslinking of the obtained nanoparticles can be achieved using relatively cheap and commercially available materials in the course of (spontaneous) reaction with amine compounds, e.g. bis(6-aminohexyl)amine.30 The usefulness of polymers of this type for the preparation of nanostructures has been proven by their application to solubilize different nanocrystals in water, including quantum dots, resulting in stable structures of high uorescence intensity.30,34,35 However, to the best of our knowledge these structures have not been applied to prepare uorimetric sensors.

2516 | Analyst, 2014, 139, 2515–2524

Paper

In our opinion, these polymers, without nanoparticles encapsulated inside, can be of interest as a novel type of matrix for the preparation of optical sensors. Aliphatic moieties of easy to prepare nanostructures – micelles formed from alternating polymers such as poly(maleic anhydride-alt-1-octadecene) (PMAO) – are attractive alternatives to the majority of the proposed polymeric spheres applied as miniature optrodes. In this work, we explore for the rst time a novel concept of preparation of optical and uorescent nanospheres embedding a model receptor – an optically silent ionophore, valinomycin – and an optical transducer, chromoionophore, within the crosslinked micelles prepared from PMAO to yield optical nanosensors for potassium ions. Alternatively, incorporation of chromoionophore (without valinomycin) results in pH-sensitive uorescent nanospheres. It is also shown that the presence of amine groups on the surface of the micelle can be used to covalently attach a reference uorescent dye allowing for ratiometric sensing.

Experimental Reagents Poly(maleic anhydride-alt-1-octadecene) (average molar mass 30 000–50 000 g mol
1) (PMAO), bis(hexamethylene)triamine (BHMTA), potassium selective ionophore – valinomycin, chromoionophore I (N-octadecanoyl-Nile blue), potassium tetrakis(4-chlorophenyl)borate (KTChP), poly(vinyl alcohol) (PVA), tris(hydroxymethyl)aminomethane (Tris), tetrahydrofuran (THF), chloroform, 1-pyrenebutyric acid and N-(3-dimethylaminopropyl)-N0 -ethylcarbodiimide (EDC) were from Aldrich (Germany). All salts used, as well as hydrochloric acid and sodium hydroxide, were of analytical grade and were obtained from POCh (Gliwice, Poland). Doubly distilled and freshly deionised water (resistance 18.2 MU cm, Milli-Qplus, Millipore, Austria) was used throughout this work. The following pH buffers were used: 10
3 M Tris (adjusted with HCl) buffer of pH ¼ 7.3, and universal buffer for pH measurements (mixture of 0.109 M citric acid, 0.1 M Tris, 0.088 M NaH2PO4 and 0.1 M NaCl adjusted with HCl or NaOH to the desired pH values). Apparatus Fluorimetric experiments were performed using a Cary Eclipse spectrouorimeter (Varian). Aer exposure at an excitation wavelength of 580 nm, emission intensity was observed within the range of 600 to 750 nm. Unless otherwise stated the slits used were 10 nm for both excitation and emission, while the detector voltage was maintained at 900 V. Pyrene spectra (aer covalent attachment of 1-pyrenebutyric acid to the nanospheres) were recorded using an excitation wavelength of 340 nm, and emission was observed within the range of 360 nm to 600 nm, using slits of 5 nm, while the detector voltage was maintained at 900 V. The UV/vis experiments were performed using a LAMBDA 650 UV/vis spectrophotometer (Perkin Elmer).

This journal is © The Royal Society of Chemistry 2014

View Article Online

Published on 20 January 2014. Downloaded by Cornell University on 22/04/2014 13:58:44.

Paper

Unless otherwise stated, emission and absorption spectra were recorded aer a 5 minute contact time of the nanospheres with the sample. Longer response times (of up to 1 h) were also tested, conrming that responses are stable in this timescale. Emulsions were prepared using a homogenizer from Hielscher, model UP 200S. The obtained nanospheres were separated from the solutions using a MPW-251 centrifuge (MPW Med. Instruments), the details of the parameter settings are stated below. To obtain TEM images of the prepared nanospheres, a Zeiss LIBRA 120 TEM (HT ¼ 120 kV, LaB6 cathode) apparatus was used, and the obtained nanoparticles were also characterized using a Malvern Zetasizer Nano ZS (scattering angle 173 degrees). Synthesis of nanospheres 500 ml of a solution of PMAO in chloroform (20 mg ml
1) was slowly added, dropwise, to 5 ml of a 1% (w/w) PVA aqueous solution containing 8.5 mg of dissolved BHMTA under sonication (cycle 0.5, power 70%). Sonication was continued for another 20 min (procedure A) or for another 5 min (procedure B). Then the mixture was placed on a stirring plate; aer 20 min (procedure A) or 10 min (procedure B) of stirring, 400 ml of 0.18 M H2SO4 aqueous solution was added and the mixture was le overnight on the stirring plate. The resulting mixture was centrifuged (5800 rpm) to separate bulky particles, the obtained supernatant was again centrifuged (18 000 rpm) (procedure A) or the mixture was directly centrifuged (18 000 rpm) (procedure B) to separate nanospheres from the polymerization mixture. The nanospheres obtained were dispersed in deionized water, again centrifuged and nally dispersed in 1% w/w PVA. Thus obtained spheres were used to prepare nanosphere sensors. Unless stated otherwise, nanospheres obtained according to procedure A were used in the experiments. Preparation of nanosphere sensors H+-sensitive nanospheres. 500 ml of a suspension of nanospheres in PVA were mixed with 100 ml of THF solution containing, in 1 ml, 0.8 mg of chromoionophore I. The mixture was le for 20 min in an open vial. Then the mixture was centrifuged (18 000 rpm) and the deposit obtained was dispersed in 0.5 ml of deionized water, centrifuged again (1800 rpm) and nally dispersed in 0.5 ml of PVA. K+-sensitive nanospheres. The procedure was the same as that described above, however, THF solution was used, containing, in 1 ml: 0.8 mg of chromoionophore I, 1 mg of KTChP and 4 mg of valinomycin. 1-Pyrenebutyric acid functionalization of K+-sensitive nanospheres for ratiometric sensors A saturated solution of 1-pyrenebutyric acid was prepared by adding 1 mg of 1-pyrenebutyric acid to 10 ml of 0.1 M acetic buffer at pH 5 and placing the mixture in an ultrasonic bath (50  C, 30 min); the resulting mixture was centrifuged at 18 000 rpm. To 2 ml of the thus obtained supernatant, 1 ml of the as prepared nanosphere suspension was added, then to this

This journal is © The Royal Society of Chemistry 2014

Analyst

mixture 15 ml of EDC was added and the mixture was placed in the ultrasonic bath for 2 h. Then the nanospheres were separated by centrifugation, followed by washing. Finally, the nanospheres were resuspended in 1 ml of PVA – this mixture was either used directly in measurements, or alternatively pyrene labeled nanospheres were loaded with chromoionophore I, valinomycin and the ion-exchanger – as described above – and used as a potassium uorimetric sensor. Optical measurements For measurements, 50 ml of nanospheres were added to 2 or 3 ml of sample for UV/vis or uorescence measurements, respectively. Potassium sensitive nanospheres were tested in Tris buffer at pH 7.3. Samples for dynamic light scattering sizing and zeta potential measurements were prepared by mixing 3 ml of 10
3 M Tris buffer at pH 7.3 with 50 ml of nanospheres suspension, in the case of the sample of nanospheres studied in the presence of KCl the above mixture was additionally spiked with KCl solution to reach a potassium ion concentration of 10
1 M. For the purpose of TEM imaging, samples of the as prepared nanospheres suspensions were diluted 100 times with deionized water and applied onto a Formvar grid mesh 400, and were le to dry at room temperature.

Results and discussion Poly(maleic anhydride-alt-1-octadecene) has been successfully used to solubilize different nanostructures in water.27,28,34,35 To prepare optical nanosensors, a similar procedure was used, however, instead of wrapping the polymer around the nanotemplate, a spontaneous micelle formation was applied to obtain polymeric micelles according to Scheme 1. The synthetic procedure was adapted from previous reports27,28,34,35 with some modication, and comprises four steps: A – dissolution of the polymer in organic solvent, B – introduction of this solution to water and spontaneous micellization, followed by C – reaction of maleic groups with bis(hexamethylene)triamine to stabilize the formed structures, and nally D – spontaneous hydrolysis of unreacted maleic groups. This procedure proved to be effective in generating nanostructures of sizes (according to dynamic light scattering measurements, DLS) close to 240 nm (Fig. S1, ESI†) prepared according to procedure A and close to 135 nm for spheres prepared according to procedure B. Fig. 1 presents TEM images of the as obtained nanospheres – the size of the spheres obtained under TEM measurement conditions is close to 250 nm for spheres prepared according to procedure A, and close to 130 nm for spheres prepared according to procedure B. Following the introduction of the potassium ionophore, chromoionophore and ion-exchanger to the nanospheres, the size of the spheres prepared by procedure A, according to DLS measurements, was close to 190 nm, and the size distribution was also narrower (Fig. S1, ESI†). The contraction of the size of the PMAO nanospheres upon introduction of the ionophore may result from the improved removal of chloroform residues, used to introduce PMAO to the solution (by a mixture

Analyst, 2014, 139, 2515–2524 | 2517

View Article Online

Published on 20 January 2014. Downloaded by Cornell University on 22/04/2014 13:58:44.

Analyst

Scheme 1

Paper

Schematic representation of the synthetic steps in the preparation of polymeric nanospheres.

of THF–water used to introduce the potassium ionophore, chromoionophore and ion-exchanger to the nanospheres). This hypothesis is in line with the difference in nanosphere sizes obtained by TEM and DLS discussed above. On the other hand, for the as prepared nanospheres prepared according to procedure B, the size determined under conditions of DLS measurement was similar to that obtained from TEM measurement, and was close to 135 nm. This suggests that shortening the time of the initial synthesis steps ultimately leads to lower incorporation of the solvent into the nanospheres. Introduction of the ionophore into these nanospheres resulted in an increase in the size of the spheres to about 170 nm. This suggests that reagents introduced to the nanospheres that are virtually free from solvent interact with each other or with water molecules leading to expansion of the sphere. Introduction of potassium ions (0.1 M KCl) to the K+-sensitive nanospheres (prepared according to procedure A) suspension resulted in a further decrease in the size of the nanospheres determined under DLS conditions to ca. 150 nm (Fig. S1†). This can result from interactions of carboxyl groups on the surface of the micelle with potassium ions in a relatively concentrated electrolyte solution (0.1 M KCl), leading to a decrease in electrostatic repulsion between these groups and ultimately resulting in shrinkage of the nanosphere. This conclusion is supported by the results of zeta potential measurements. For the as obtained nanospheres and K+sensitive nanospheres, the determined zeta potential value was close to
15 mV. This result clearly shows that, despite crosslinking with amine groups of BHMTA, some unbound carboxyl groups remain on the surface of the nanospheres. Following

2518 | Analyst, 2014, 139, 2515–2524

contact of the K+-sensitive nanospheres with potassium ions, the zeta potential was close to +7 mV, thus, most probably, the obtained nanospheres are able to react with the target analyte ions not only due to potassium ions interacting with the valinomycin ionophore present inside, but also using the carboxyl groups (not involved in crosslinking with BHMTA) located on the surface, actively participating in ion-exchange between the lipophilic moiety of the micelle and the sample. This effect is highly promising for analytical applications of the obtained nanospheres. On the other hand, for nanospheres prepared according to procedure B, contact with relatively concentrated electrolyte solution (0.1 M KCl) resulted in a small increase in size to about 220 nm. This effect suggests increased repulsion within the nanosphere, which can result from incorporation of the analyte (K+) cations into the inside of the sphere. However, it should be stressed that, contrary to nanospheres prepared according to procedure A, for these nanospheres prepared according to procedure B, interaction with 0.1 M KCl did not result in a change of sign of the zeta potential, nevertheless an increase of potential was observed, similarly to that described above. The potential change for as prepared K+-sensitive nanospheres of type B was close to
14 mV, i.e. was comparable with the value obtained for nanospheres of type A. Upon contact of the nanospheres of type B with 0.1 M KCl the potential changed to
8.5 mV, thus although an increase in potential was observed, the overall potential of the nanospheres of type B remained negative. This indicates that even in a relatively concentrated sample (0.1 M KCl), nanospheres prepared according to procedure B will retain some negative charge on

This journal is © The Royal Society of Chemistry 2014

View Article Online

Published on 20 January 2014. Downloaded by Cornell University on 22/04/2014 13:58:44.

Paper

Fig. 1 The TEM images of as obtained PMAO nanospheres according to (A) procedure A or (B) procedure B. The mean diameter of 9 spheres depicted in (B) is 130 nm  16 nm.

the surface – thus the surface effects leading to improved analytical responses of polymeric nanospheres are expected to be more pronounced for this type of nanosphere. It should also be stressed that when the supernatant from the nanospheres (prepared according to procedure A) containing chromoionophore I (H+-sensitive nanospheres) was separated from the nanospheres two weeks aer introducing this chromoionophore to the nanospheres, this solution was found to be optically silent under uorescence emission mode in alkaline solution. This result proves that leakage of the chromoionophore from the nanospheres is negligible. It should be stressed that just before separation of the nanospheres, this suspension was showing the usual uorescent responses to an increase of sample pH. Thus, it can be assumed that lipophilic ionophores introduced into the core of the micelle were effectively retained inside. H+-sensitive nanospheres Taking advantage of the presence of cation-exchanging carboxyl groups in the nanospheres, H+-sensitive spheres were prepared just by introducing chromoionophore I (alone) to the structures to obtain sensors for H+ ions. It was expected that upon the

This journal is © The Royal Society of Chemistry 2014

Analyst

change of pH, H+ ions can be exchanged between the nanosphere and solution, resulting ultimately in a change of visible absorption or uorescence emission (in the case of deprotonation) of the chromoionophore present in the nanosphere. Suspensions of the H+-sensitive nanospheres showed visible changes in the colour of the sample from blue to pink, similar to those observed for deprotonation of a chromoionophore present in polymeric beads,11 Fig. 2A. As can be seen in Fig. 2A, under acidic conditions (pH ¼ 2.1) a broad peak with two maxima at about 620 and 675 nm is present in the visible absorption spectra of chromoionophore I present in the nanospheres. Upon deprotonation of chromoionophore I (alkalization of solution) these peaks gradually decrease at the expense of formation of a broad peak with a maximum at about 530 nm, Fig. 2A. The obtained spectra, both in alkaline and acidic media, correspond well to literature data obtained for this compound in a PVC plasticized membrane, e.g. ref. 36. The change of the protonation state of chromoionophore I inside the polymeric nanostructure can be followed uorimetrically e.g. ref. 11, Fig. 2B. As can be seen in Fig. 2B, an increase of sample pH resulted in an increase in uorescence of the nanospheres, with the peak maxima recorded at about 680 nm. As can be seen in Fig. 2B, for pH < 6 only weak uorescence of the sample was recorded; however, increasing the sample pH resulted in an increase of uorescence intensity. Within the pH range of 8.2 to 12.2, a linear dependence of uorescence intensity upon pH was obtained with R2 ¼ 0.980. It should be stressed that in comparison to other nanospheres applied as pH uorimetric sensors, e.g. polyacrylic nanospheres11 containing the same uorophore, chromoionophore I, the herein obtained response range was much broader – it covered four orders of magnitude. Moreover, the response range was shied to lower pH, which could result from the presence of carboxyl groups. The observed sensitivity over a wide pH range could result from the properties of the nanosphere system. The inside of the nanosphere is a highly lipophilic environment, due to the presence of heteroatom-free hydrocarbon chains. Therefore the neutral deprotonated form of the chromoionophore within the spheres is more stable even over a wide pH range of the aqueous medium surrounding the nanosphere. This results in a signicant increase of the pH response range. The response time of the herein proposed nanospheres was tested during pH variation from 2.1 to 12.1 and from 12.1 to 2.1. The change in the maximum absorption in the visible spectra was achieved within 0.2 min, then the absorption remained constant (results not shown). Similarly, the complete change in the emission spectra, regardless of whether the sample was alkalinized or acidied, was observed within less than 1.5 min (this is the minimum time required to record the full spectra over the wavelengths of interest); no further changes in the spectra were observed (Fig. S2, ESI†). The response of nanospheres to pH variation from pH 2 to pH 12 followed by uorescence intensity measurements clearly shows that a stable signal is observed 2.5 s aer the change in pH of the sample, Fig. S3, ESI.† Smaller pH changes (close to 3 units) result in even faster responses, regardless of whether the

Analyst, 2014, 139, 2515–2524 | 2519

View Article Online

Paper

Published on 20 January 2014. Downloaded by Cornell University on 22/04/2014 13:58:44.

Analyst

(A) Changes in the visible absorption spectra of pH-sensitive nanospheres. The pH values of the samples of universal buffer (sodium salts were used) were equal to: 2.1 and 12.7. Inset: changes of colour accompanying the change of pH of the solution containing the nanospheres. (B) Changes in the fluorescence spectra of pH-sensitive nanospheres. The pH values of the samples of universal buffer (sodium salts were used) were equal to: 2.1; 3.1; 4.0; 5.1; 6.1; 7.2; 8.1; 9.1; 10.2; 11.2; 12.2 and 12.7. Inset: dependence of fluorescence intensity recorded at 680 nm on pH. Fig. 2

pH is increased or decreased. These effects are highly promising for the analytical application of the herein proposed nanospheres, and prove effective ion-exchange through the micelle “walls”. This effect, in our opinion, results from the properties of the applied material and the presence of cation-exchanging groups at the surface of nanospheres, facilitating ion-exchange between the lipophilic core of the structure and the sample. Although the response is slower than in the case of pH sensitive

2520 | Analyst, 2014, 139, 2515–2524

liposomes,23 it is much faster compared with nanospheres prepared from poly(n-butyl acrylate) of a similar size.11 Moreover, the full reversibility of pH responses of the herein proposed nanospheres was also proven. K+-sensitive nanospheres The above presented results clearly show that the synthesized nanospheres can not only be applied as pH sensors but also as

This journal is © The Royal Society of Chemistry 2014

View Article Online

Published on 20 January 2014. Downloaded by Cornell University on 22/04/2014 13:58:44.

Paper

ion sensors using the known approach, in which the potassium sensitivity is driven by the concomitant presence within the nanosphere of a potassium ionophore – optically silent receptor – and a chromoionophore used as an optical transducer.11,15 In principle, for a constant amount of ion-exchange sites, and a constant pH of the sample solution, incorporation of potassium cations into the nanosphere results in deprotonation of the chromoionophore which, as shown above, can be followed both uorimetrically and in the visible absorption mode. As can be seen in Fig. 3A, an increase in the concentration of potassium ions in the sample resulted in a visible change of colour and a change in the visible spectra of the sample. In the absence of potassium ions, two peaks characteristic of the protonated form of chromoionophore I (optical transducer) were observed in the recorded spectra at about 620 and 675 nm. Thus, the introduction of potassium ions to the sample is apparently followed by incorporation of K+ into the

Analyst

nanospheres, resulting in a gradual decrease in intensity of these peaks and the formation of a broad peak characteristic of the deprotonated form of the optical transducer with a maximum at about 540 nm, similar to that observed previously.11 A similar experiment performed under uorimetric conditions, Fig. 3B, has shown that an increase of potassium ion concentration in the sample is accompanied by an increase of uorescence emission (maximum at 680 nm, from emission of the deprotonated form of chromoionophore I) as expected for the gradual deprotonation of chromoionophore I present within the nanospheres upon interaction of K+ ions with the surface of the spheres or/and incorporation of K+ ions into the nanostructures. For nanospheres prepared according to procedure A, the change of K+ concentration from 10
8 to 10
7 M resulted in a measurable increase in recorded uorescence; a similar increase was recorded for the concentration change

Fig. 3 (A) Visible spectra of K+-sensitive nanospheres recorded with the change of potassium ion concentration (from 10
8 to 0.1 M KCl) in solution. (B) Changes in the fluorescence spectra of K+-sensitive nanospheres. The concentration of K+ ions was varied from 10
8 to 0.1 M. Inset: dependence of fluorescence intensity recorded at 680 nm on the logarithm of the concentration of K+ ions. (C) Dependence of fluorescence intensity on the logarithm of the concentration of K+ ions, the mean signal recorded at 680 nm together with the standard deviation from three measurements for each concentration recorded for K+-sensitive nanospheres (prepared according to procedure B). (D) Dependence of fluorescence intensity on the logarithm of the concentration of K+ ions, the mean signal recorded at 680 nm recorded for K+-sensitive nanospheres (prepared according to procedure B) after ( ) 10 min and ( ) 28 h contact time with potassium ion-containing solutions.

This journal is © The Royal Society of Chemistry 2014

Analyst, 2014, 139, 2515–2524 | 2521

View Article Online

Published on 20 January 2014. Downloaded by Cornell University on 22/04/2014 13:58:44.

Analyst

from 10
7 to 10
6 M. Unlike previously described nanospheres11 for which sigmoidal type responses were reported, for the herein proposed nanospheres a high sensitivity was coupled with a broad response range – a linear dependence of uorescence intensities on the logarithm of the concentration of potassium ions in the sample was obtained over the range of 10
6 M to 10
2 M with R2 ¼ 0.991. The observed relatively broad response range of the herein proposed K+-sensor can be attributed to strong interactions of potassium ions with valinomycin, favoured by the presence of cation-exchanging groups not only inside the nanospheres but also on the outer shell. On the other hand, due to the highly lipophilic nature of the medium inside the nanospheres, electrically neutral forms would be preferred. Therefore, saturation may denote only partial involvement of the present valinomycin molecules in K+ complexation. Thus, further increase in the concentration of K+ ions in the sample, covering even a few orders of magnitude, results in only a small increase of K+ – valinomycin concentration in the spheres. However, as expected on the basis of the zeta potential determination discussed above, in 0.1 M KCl for nanospheres of type A, surface effects are diminished (positive potential), probably due to enhanced incorporation of K+ into the inside of the sphere. Taking into account the fact that the observed responses result from the interaction of K+ ions with the surfaces of the nanospheres, it can be expected that decreasing the size of the nanospheres, e.g. moving from nanospheres prepared according to procedure A to spheres prepared according to procedure B, will result in further enhancement of the linear responses of the nanospheres (higher contraction of volume of spheres compared to decrease in surface area). Indeed, as seen in Fig. 3C, nanospheres obtained according to procedure B showed linear responses within the range of 10
5 to 0.1 M KCl with R2 ¼ 0.9999, conrming the high linearity of the dependence of uorescence intensities on the logarithm of KCl concentration. To calculate the detection limit of the obtained responses, a method typically used for ion-selective electrodes can be applied – the cross-section point of the linear part of the uorimetric dependence and the part of the response for which there is only a small change in recorded emission with the change of KCl concentration results in a detection limit equal to 10
5.7 M. Fig. 3C also conrms that the obtained responses are highly reproducible, the standard deviation values of the mean signal of three measurements are less than 1% of the total uorescence intensity change observed for the tested concentration range, regardless of the concentration of KCl in the sample. High reproducibility was also observed for nanospheres (procedure B) prepared in different experimental runs, as shown in Fig. S4, ESI.† As can be seen in Fig. S4, ESI,† the standard deviation of the mean signal is less than 3% of the total uorescence intensity change observed for the tested concentration range, regardless of the KCl concentration in the sample. It seems justied to expect that, with a longer contact time between the nanospheres and the sample solution,

2522 | Analyst, 2014, 139, 2515–2524

Paper

incorporation of the analyte K+ ions into the volume of the sphere will occur, leading to a smaller effect of the surface related phenomena due to reaction occurring in the nanosphere volume. This is expected to result in a change of the response pattern from linear responses to a typical sigmoidal one. The results presented in Fig. 3D, obtained for nanospheres aer 10 min and 28 h contact with KCl solutions, respectively, clearly indicate that with extended reaction time of the nanospheres (prepared according to procedure B), the responses change from linear over a broad concentration range, to sigmoidal, which is typical for classical miniature optrode systems, e.g. ref. 11 and 37. On the other hand, taking into account the rapid formation of the uorescence signal and its stability within the timescale of a typical experiment, the change in the response pattern related to incorporation of the analyte into the inside of the sphere is unlikely to hinder the analytical applications of the herein proposed system. The selectivity of the proposed nanospheres was also tested in solutions of different concentrations of model interferent cations (Na+, NH4+, H+, Li+, Ca2+ or Mg2+) in the solution, resulting in only small changes in the recorded uorescence intensity; see Fig. S5, ESI.† These results prove the high selectivity of uorescent K+-sensitive nanospheres, as observed previously for sensors containing a valinomycin ionophore. Ratiometric K+-sensitive sensor The presence of carboxyl groups (not used for amine crosslinking) as well as amine groups introduced by the crosslinking agent on the nanospheres offers a unique possibility of the covalent attachment of various dyes, provided that they have any of the above mentioned functional groups. Thus, apart from just obtaining uorescent dye modied spheres, a dye intended to allow the ratiometric uorimetric sensing of target ions can be introduced to the system. In this work, as a model dye, 1pyrenebutyric acid was used, and it was covalently attached to the nanospheres using the amine groups of BHMTA through carbodiimide coupling. Indeed, for the modied nanospheres (ionophore-free nanospheres), characteristic spectra of the pyrene derivative dye were obtained, and as expected no sensitivity to the change in concentration of potassium ions in the sample was observed. Thus, pyrene derivatives linked to the surface of nanospheres can be used as a reference dye for K+sensitive uorescent nanospheres. Following the introduction of valinomycin, ion-exchanger and chromoionophore I to the nanospheres, i.e. preparation of K+-sensitive nanospheres, it was observed that pyrene spectra are not affected by change of the KCl concentration in the sample, as shown in Fig. 4, however, the uorescence emission at the wavelength characteristic of the deprotonated form of chromoionophore I was dependent on the change of potassium ion concentration in the sample, in similar way as was shown in Fig. 3B. Fig. 4 shows the dependence of the ratio of the uorimetric signal recorded at a wavelength characteristic for chromoionophore I to the signal of pyrene, recorded for individual concentrations of potassium ions in the sample within the range of 10
8 to 10
1 M. Similar to that for K+-sensitive

This journal is © The Royal Society of Chemistry 2014

View Article Online

Published on 20 January 2014. Downloaded by Cornell University on 22/04/2014 13:58:44.

Paper

Analyst

Fig. 4 Dependence of the ratio of fluorescence intensity recorded at 680 nm (chromoionophore I) to the signal recorded at 396 nm (pyrene) for the K+-sensitive nanospheres with covalently attached 1-pyrenebutyric acid on the logarithm of the concentration of K+ ions. Inset: pyrene spectra of K+-sensitive nanospheres with covalently attached dye recorded for different concentrations of potassium ions in the sample.

nanospheres not modied with the reference dye, within the KCl concentration range of 10
6 to 10
2 M, a linear dependence of the ratio of uorescence intensities on the logarithm of KCl concentration was obtained, with R2 ¼ 0.982. The results of the above described ratiometric experiment conrm the analytical usefulness of the herein proposed nanospheres.

Conclusions It was shown that alternating polymer, self assembled crosslinked micelles are highly attractive for optical sensing purposes. The obtained nanospheres can be loaded with dyes of choice, useful for both visible absorption sensing and uorimetric quantication. Due to the unique properties of the structure of these nanoparticles, the obtained analytical dependences are characterized by broad linear response ranges and fast response times.

Acknowledgements Financial support from the National Science Centre (NCN, Poland), project 2011/03/B/ST4/00747, in the years 2012–2015, is gratefully acknowledged. Michał Pieczykolan is kindly acknowledged for graphical assistance.

References 1 M.-J. Ruedas-Rama, J. D. Walters, A. Orte and E. A. H. Hall, Anal. Chim. Acta, 2012, 751, 1.

This journal is © The Royal Society of Chemistry 2014

2 K. Michael, L. Taylor, S. Schulz and D. Walt, Anal. Chem., 1998, 70, 1242. 3 K. Albert and D. Walt, Anal. Chem., 2000, 72, 1947. 4 H. A. Clark, M. Hoyer, M. A. Philbert and R. Kopelman, Anal. Chem., 1999, 71, 4831. 5 I. Tsagkatakis, S. Peper and E. Bakker, Anal. Chem., 2001, 73, 315. 6 I. Tsagkatakis, S. Peper, R. Retter, M. Bell and E. Bakker, Anal. Chem., 2001, 73, 6083. 7 M. Telting-Diaz and E. Bakker, Anal. Chem., 2002, 74, 5251. 8 V. Bychkova and A. Shvarev, Anal. Chem., 2009, 81, 2325. 9 K. Wygladacz, A. Radu, Ch. Xu, Y. Qin and E. Bakker, Anal. Chem., 2005, 77, 4706. 10 Ch. Xu, K. Wygladacz, R. Retter, M. Bell and E. Bakker, Anal. Chem., 2007, 79, 9505. 11 M. J. Ruedas-Rama and E. A. H. Hall, Analyst, 2006, 131, 1282. 12 M. Brasuel, R. Kopelman, T. J. Miller, R. Tjalkens and M. A. Philbert, Anal. Chem., 2001, 73, 2221. 13 Y. Qin, S. Peper, A. Radu, A. Ceresa and E. Bakker, Anal. Chem., 2003, 75, 3038. 14 M. G. Brasuel, T. J. Miller, R. Kopelman and M. A. Philbert, Analyst, 2003, 128, 1262. 15 W. E. Morf, K. Seiler, B. Rusterholz and W. Simon, Anal. Chem., 1990, 62, 730. 16 E. Wo´znica, K. Maksymiuk and A. Michalska, Anal. Chem., 2014, 86, 411. 17 L. Y. Heng, K. Toth and E. A. H. Hall, Talanta, 2004, 63, 73.

Analyst, 2014, 139, 2515–2524 | 2523

View Article Online

Published on 20 January 2014. Downloaded by Cornell University on 22/04/2014 13:58:44.

Analyst

18 E. Wo´znica, J. Mieczkowski and A. Michalska, Analyst, 2011, 136, 4787. 19 A. Michalska, M. Wojciechowski, E. Bulska and K. Maksymiuk, Electroanalysis, 2012, 21, 1931. 20 E. Fleige, M. A. Quadir and R. Haag, Adv. Drug Delivery Rev., 2012, 64, 866. 21 K. Kataoka, A. Harada and Y. Nagasaki, Adv. Drug Delivery Rev., 2012, 64, 37. 22 Z. L. Tyrrell, Y. Shena and M. Radosz, Prog. Polym. Sci., 2010, 35, 1128. 23 S. A. Dergunov, B. Miksa, B. Ganus, E. Lindner and E. Pinkhassik, Chem. Commun., 2010, 46, 1485. 24 V. P. Torchilin, J. Controlled Release, 2001, 73, 137. 25 J. K. Kim, S. Y. Yang, Y. Lee and Y. Kim, Prog. Polym. Sci., 2010, 35, 1325. 26 Y. Bae and K. Kataoka, Adv. Drug Delivery Rev., 2009, 61, 768. 27 H. Otsuka, Y. Nagasaki and K. Kataoka, Adv. DrugDelivery Rev., 2012, 64, 246. 28 B. Ma, S. Wu, F. Zeng, Y. Lu, J. Zhao and Z. Tong, Nanotechnology, 2010, 21, 195501.

2524 | Analyst, 2014, 139, 2515–2524

Paper

29 Z. Yang, Z. Wang, X. Yao, X. Chen and Y. Wang, J. Membr. Sci., 2013, 441, 9. 30 T. Pellegrino, L. Manna, S. Kudera, T. Liedl, D. Koktysh, A. L. Rogach, S. Keller, J. R¨ adler, G. Natile and W. J. Parak, Nano Lett., 2004, 4, 703. 31 R. Di Corato, A. Quarta, P. Piacenza, A. Ragusa, A. Figuerola, R. Buonsanti, R. Cingolani, L. Manna and T. Pellegrino, J. Mater. Chem., 2008, 18, 1991. 32 W. W. Yu, E. Chang, J. Falkner, J. Zhang, A. M. Al-Somali, Ch. M. Sayes, J. Johns, R. Drezek and V. Colvin, J. Am. Chem. Soc., 2007, 129, 2871. 33 N. Tomczak, D. Ja´ nczewski, M. Hana and G. J. Vancso, Prog. Polym. Sci., 2009, 34, 393. 34 Ch. Zhoua, W. Xu, S. Lou, Z. Zhang, L. S. Li and J. Chen, Colloids Surf., A, 2010, 355, 139. 35 Ch. Zhou, H. Shen, Y. Guo, L. Xu, J. Niu, Z. Zhang, Z. Du, J. Chen and L. S. Li, J. Colloid Interface Sci., 2010, 344, 279. 36 E. Lindner, T. Zwickl, E. Bakker, B. Thi Thu Lan, K. Toth and E. Pretsch, Anal. Chem., 1998, 70, 1176. 37 J. D. Walters and E. A. H. Hall, Analyst, 2011, 136, 4718.

This journal is © The Royal Society of Chemistry 2014