Anal. Chem.
2008, 80, 573-582
Articles
Poly(styrene-block-vinylpyrrolidone) Beads as a
Versatile Material for Simple Fabrication of
Optical Nanosensors
Sergey M. Borisov,* Torsten Mayr, and Ingo Klimant
Institute of Analytical Chemistry and Radiochemistry, University of Technology Graz, Stremayrgasse 16, 8010 Graz, Austria
A versatile platform for designing optical nanosensors is
proposed. The “sensing chemistries” are entrapped into
the poly(styrene-block-vinylpyrrolidone) nanobeads hav-
ing the average size of 245 nm in aqueous media.
Addressable staining into the core or the shell of the beads
results in nanosensors for essential analytes such as
dissolved oxygen, temperature, pH, chloride, and copper
ions. Two immobilization procedures are developed:
staining in the polystyrene core is performed from a
tetrahydrofuran/water mixture (50:50 v/v) and staining
in the poly(vinylpyrrolidone) shell is achieved by using
the ethanol/water mixture (70:30 v/v). The oxygen and
temperature indicators should be preferably immobilized
into the core, whereas nanosensors for ions are manu-
factured by staining into the shell. In the case of the
lipophilic pH indicators both procedures result in similar
pKa values. The unique properties of the beads make them
promising for sensing and imaging even in very complex
media, multianalyte sensing, and monitoring of very fast
processes.
Optical chemical sensing has become very widespread in the
last 2 decades. It is mostly based on the use of smart materials
with optical properties that respond to a chemical parameter.1,2
Most sensor materials consist of an indicator dye dissolved in a
polymeric matrix which also acts as a solid support and as a
permeation-selective membrane. Such materials are used in
numerous formats including planar sensor spots,3,4 fiber-optic
systems,5 two-dimensional foils,6,7 and pressure-sensitive paints.8,9
* Corresponding author. Phone: +43 316 873 4326. Fax: +43 316 873 4329.
E-mail: sergey.borisov@tugraz.at.
(1) Wolfbeis, O. S. Anal. Chem. 2006, 78, 3859-3874.
(2) Wolfbeis, O. S. J. Mater. Chem. 2005, 15, 2657-2669.
(3) Hartmann, P.; Trettnak, W. Anal. Chem. 1996, 68, 2615-2620.
(4) Mills, A.; Thomas, M. Analyst 1997, 122, 63-68.
(5) Rosenzweig, Z.; Kopelman, R. Anal. Chem. 1995, 67, 2650-2654.
(6) Koese, M. E.; Carrol, B. F.; Schanze, K. S. Langmuir 2005, 21, 9121-
9129.
(7) Liebsch, G.; Klimant, I.; Frank, B.; Holst, G.; Wolfbeis, O. S. Appl. Spectrosc.
2000, 54, 548-559.
(8) Zelelow, B.; Khalil, G.; Phelan, G.; Carlson, B.; Gouterman, M.; Callis, J. B.;
Dalton, L. R. Sens. Actuators, B 2003, 96, 304-314.
(9) Hradil, J.; Davis, C.; Mongey, K.; McDonagh, C.; MacCraith, B. D. Meas.
Sci. Technol. 2002, 13, 1552-1557.
10.1021/ac071374e CCC: $40.75 © 2008 American Chemical Society
Published on Web 01/04/2008
Imaging, in particular, becomes increasingly popular due to the
possibility of real time analyte monitoring over a large area or in
volume. Imaging of oxygen,10,11 pH,12,13 CO2,14,15 ions,16 and
biomolecules such as glucose17 is of wide interest in various fields
of science and technology including marine research,10,13,14,18
clinical medicine,19 and biotechnology15 to mention only a few.
Imaging of the total pressure of air (via oxygen pressure) on the
surface of aircrafts and cars20,21 is performed by means of pressure-
sensitive paints. Planar sensor foils and pressure-sensitive paints
which have been mostly used so far for imaging purposes are,
however, not universal tools for several reasons. These include
(a) inability of monitoring of very fast processes such as enzymatic
reactions since significant thickness of the sensor layer (usually
several micrometers) and the presence of a thick virtually gas-
impermeable support results in relatively long response times.
This can be handled by using (sub)micrometer-sized fiber-optic
sensors5,22 which respond virtually in real time; however, mechan-
ical stability of those is often poor; (b) unsuitability for imaging
in a variety of flow-through cells and bioreactors, where only small
spots of the planar material can be positioned; (c) certain
difficulties for multianalyte sensing, which can be compromised
by inhomogeneities of a sensor layer and reduced photostability.23
The above limitations can be overcome by making use of analyte-
(10) Koenig, B.; Kohls, O.; Holst, G.; Glud, R. N.; Kuehl, M. Mar. Chem. 2005,
97, 262-276.
(11) Schroeder, C. R.; Polerecky, L.; Klimant, I. Anal. Chem. 2007, 79, 60-70.
(12) Liebsch, G.; Klimant, I.; Krause, Ch.; Wolfbeis, O. S. Anal. Chem. 2001,
73, 4354-4363.
(13) Stahl, H.; Glud, A.; Schroder, C. R.; Klimant, I.; Tengberg, A.; Glud, R. N.
Limnol. Oceanogr.: Methods 2006, 4, 336-345.
(14) Zhu, Q.; Aller, R. C.; Fan, Y. Mar. Chem. 2006, 101, 40-53.
(15) Borisov, S. M.; Krause, Ch.; Arain, S.; Wolfbeis, O. S. Adv. Mater. 2006,
18, 1511-1516.
(16) Mayr, T.; Liebsch, G.; Klimant, I.; Wolfbeis, O. S. Analyst 2002, 127, 201-
203.
(17) Schaeferling, M.; Wu, M.; Wolfbeis, O. S. J. Fluoresc. 2004, 14, 561-568.
(18) Holst, G.; Grunwald, B. Sens. Actuators, B 2001, 74, 78-90.
(19) Babilas, P.; Liebsch, G.; Schacht, V.; Klimant, I.; Wolfbeis, O. S.; Szeimies,
R. M.; Abels, C. Microcirculation 2005, 12, 477-487.
(20) Demas, J. N.; DeGraff, B. A.; Coleman, P. B. Anal. Chem. 1999, 71, 793A-
800A.
(21) Gouterman, M.; Callis, J.; Dalton, L.; Khalil, G.; Mebarski, Y.; Cooper, K.
R.; Greiner, M. Meas. Sci. Technol. 2004, 15, 1986-1994.
(22) Tan, W.; Shi, Z.-Y.; Kopelman, R. Anal. Chem. 1992, 64, 2985-2990.
(23) Borisov, S. M.; Neurauter, G.; Schroeder, C.; Klimant, I.; Wolfbeis, O. S.
Appl. Spectrosc. 2006, 60, 1167-1173.
Analytical Chemistry, Vol. 80, No. 3, February 1, 2008 573
Figure 1. Chemical structures of the indicators used in nanosensors.
sensitive nanobeads which will also allow for three-dimensional
(3D) imaging. Moreover, since nanobeads can be dispersed in
small volume, very low absolute limits of detection are achieved.
Dissolved indicators can also be used, but such systems are prone
to various interferences, such as water, ions, and other species,
and they can interact with other components of the medium such
as proteins. Moreover, many of the indicators possess extremely
low solubility in aqueous solutions. Dendrimeric indicators are
promising,24 but sophisticated synthesis is necessary.
The group of Kopelman has developed a variety of analyte-
sensitive nanobeads, so-called “PEBBLEs”,25 which are mostly
based on polyacrylamide and poly(decyl methacrylate) and have
typical size from 50 to 300 µm. Nanobeads for sensing and imaging
oxygen,26,27 pH,28 Ca2+,28,29 Mg2+,30 K+,31 Fe3+,32 Zn2+,33 and Cl- 34
were reported. The group also developed nanobiosensors for
(24) Wilson, D. F.; Lee, W. M. F.; Makonnen, S.; Finikova, O.; Apreleva, S.;
Vinogradov, S. A. J. Appl. Physiol. 2006, 101, 1648-1656.
(25) Clark, H. A.; Hoyer, M.; Philbert, M. A.; Kopelman, R. Anal. Chem. 1999,
71, 4831-4836.
(26) Xu, H.; Aylott, J. W.; Kopelman, R.; Miller, T. J.; Philbert, M. A. Anal. Chem.
2001, 73, 4124-4133.
(27) Cao, Y.; Lee Koo, Y.-E.; Kopelman, R. Analyst, 2004, 129, 745-750.
(28) Clark, H. A.; Hoyer, M.; Parus, S.; Philbert, M. A.; Kopelman, R. Mikrochim.
Acta 1999, 131, 121-128.
(29) Clark, H. A.; Kopelman, R.; Tjalkens, R.; Philbert, M. A. Anal. Chem. 1999,
71, 4837-4843.
(30) Park, E. J.; Brasuel, M.; Behrend, C.; Philbert, M. A.; Kopelman, R. Anal.
Chem. 2003, 75, 3784-3791.
(31) Brasuel, M.; Kopelman, R.; Miller, T. J.; Tjalkens, R.; Philbert, M. A. Anal.
Chem. 2001, 73, 2221-2228.
(32) Sumner, J. P.; Kopelman, R. Analyst 2005, 130, 528-533.
(33) Sumner, J. P.; Aylott, J. W.; Monson, E.; Kopelman, R. Analyst 2002, 127,
11-16.
(34) Brasuel, M. G.; Miller, T. J.; Kopelman, R.; Philbert, M. A. Analyst 2003,
128, 1262-1267.
574 Analytical Chemistry, Vol. 80, No. 3, February 1, 2008
Cu2+ 35 and glucose.36 Other groups reported dyed polystyrene
nanoparticles for sensing oxygen,37 quantum dots for temperature
sensing (however, suitable for measurements only in some
nonaqueous solvents),38,39 and beads based on plasticized poly-
(vinyl chloride) for sensing chloride ion.40 Oxygen-sensitive
liposomes (∼70 nm in diameter) were shown to degrade con-
stantly with time.41
In this work we report poly(styrene-block-vinylpyrrolidone) as
a novel matrix for designing optical nanosensors. This versatile
matrix allows for “sensing chemistry” to be incorporated either
into the core or the shell of a bead which for some nanosensors
can slightly tune the sensitivity and helps to minimize interfer-
ences. Examples include nanosensors for oxygen, temperature,
pH, chloride, and copper ions.
EXPERIMENTAL SECTION
Materials and Suppliers. Platinum(II) 5,10,15,20-tetrakis-
(2,3,4,5,6-pentafluorophenyl)-porphyrin (PtTFPP) was obtained
from Frontier Scientific (www.frontiersci.com), poly(styrene-block-
vinylpyrrolidone) emulsion in water (38% w/w emulsion in water)
was purchased from Aldrich (www.sigmaaldrich.com). Nile red,
lucigenin, lucifer yellow-CH, sodium dodecyl sulfate (SDS),
hexadecyltrimethylammonium chloride (HDTMA), sodium hy-
(35) Sumner, J. P.; Westerberg, N. M.; Stoddard, A. K.; Fierke, C. A.; Kopelman,
R. Sens. Actuators, B 2006, 113, 760-767.
(36) Xu, H.; Aylott, J. W.; Kopelman, R. Analyst 2002, 127, 1471-1477.
(37) Im, S. H.; Khalil, G. E.; Callis, J.; Ahnb, B. H.; Gouterman, M.; Xia, Y. Talanta
2005, 67, 492-497.
(38) Jorge, P. A.; Mayeh, M.; Benrashid, R.; Caldas, P.; Santos, J. L.; Farahi, F.
Meas. Sci. Technol. 2006, 17, 1032-1038.
(39) Wang, S.; Westcott, S.; Chen, W. J. Phys. Chem. B 2002, 106, 11203-11209.
(40) Ceresa, A.; Qin, Y.; Peper, S.; Bakker, E. Anal. Chem. 2003, 75, 133-140.
(41) McNamara, K. P.; Rosenzweig, Z. Anal. Chem. 1998, 70, 4853-4859.
Table 1. Composition of the Nanosensors
content,
code analyte dye % w/w procedure
N1 oxygen PtTFPP 1.5 EtOH
N2 oxygen PtTFPP 1.5 THF
N3 T Eu(tta)3L 1.5 THF
N4 pH CHFOE 0.25 EtOH
N5 pH CHFOE 0.25 THF
drogen phosphate, sodium dihydrogen phosphate, and sodium
chloride were from Fluka (www.sigmaaldrich.com). Tetrahydro-
furan (THF) and ethanol were obtained from Roth (www.carl-
roth.de). Nitrogen and synthetic air (all of 99.999% purity) were
obtained from Air Liquide (www.airliquide.at). All the components
of the cultivation media were obtained from Becton, Dickinson
and Company (www.bd.com) and Roth.
Synthesis of 2′-chloro-7′-hexylfluorescein octadecylester (CH-
FOE) is reported elsewhere.42 The pH indicator 1-hydroxypyrene-
3,6,8-tris-octylsulfonamide (HPTS(OA)3) was prepared according
to the procedure of Mohr et al.43 The temperature probe europium-
(III) tris(thenoyltrifluoroacetonate) dipyrazoltriazine complex (Eu-
(tta)3L) was prepared according to the literature procedure.44 The
chemical structures of the indicators are presented in Figure 1.
Preparation of the Oxygen-Sensitive Nanobeads N1. Five
hundred twenty-six milligrams of the polymer emulsion (contain-
ing 200 mg of the polymer beads) was diluted with the mixture
of 80 mg of ethanol and 40 mg of water. Then, 3 mg of PtTFPP
was dissolved in 20 mL of ethanol, and the solution was added
dropwise under vigorous stirring into emulsion of the polymer.
The emulsion was concentrated under reduced pressure until all
ethanol was removed. It was then diluted with water up to 20 mL
overall volume.
Preparation of the Oxygen-Sensitive Nanobeads N2. Five
hundred twenty-six milligrams of the polymer emulsion was
diluted with the mixture of 50 mL of water and 30 mL of THF.
Three milligrams of PtTFPP was dissolved in 20 mL of THF, and
the solution was added dropwise under vigorous stirring into
emulsion of the polymer. THF was removed under reduced
pressure, and the suspension was diluted with water up to 20 mL
overall volume.
Preparation of the Temperature-Sensitive Nanobeads N3.
It was performed according to the previous protocol; however, 3
mg of Eu(tta)3L was used instead of the same amount of PtTFPP.
Preparation of the pH-Sensitive Nanobeads N4. One
thousand and fifty-two milligrams of the polymer emulsion was
diluted with the mixture of 80 mg of ethanol and 40 mg of water.
Then, 1 mg of CHFOE was dissolved in 20 mL of ethanol, and
the solution was added dropwise under vigorous stirring into the
emulsion of the polymer. The emulsion was concentrated under
reduced pressure until all ethanol was removed. It was then diluted
with water up to 20 mL overall volume. Preparation of the pH-
sensitive nanobeads N6 was performed in the similar manner;
however, 6 mg of the dye (HPTS(OA)3) was used. Other
(42) Weidgans, B. M.; Krause, Ch.; Klimant, I.; Wolfbeis, O. S. Analyst 2004,
129, 645-650.
(43) Mohr, G. J.; Werner, T.; Wolfbeis, O. S. J. Fluoresc. 1995, 5, 135-138.
(44) Borisov, S. M.; Wolfbeis, O. S. Anal. Chem. 2006, 78, 5094-5101.
content,
code analyte dye % w/w procedure
N6 pH HPTS(OA)3 1.5 EtOH
N7 pH HPTS(OA)3 1.5 THF
N8 Cl- lucigenin 0.23 EtOH
N9 Cu2+ lucifer 0.175 EtOH
yellow-CH
pH-sensitive nanobeads (N5 and N7) were prepared according to
the procedure for the oxygen-sensitive nanobeads N1, and 0.5 mg
of CHFOE and 1.5 mg of HPTS(OA)3 were used for 526 mg of
polymer emulsion, for preparation of N5 and N7, respectively.
Preparation of the Chloride-Sensitive Nanobeads N8.
Three hundred milligrams of polymer emulsion was diluted with
2.7 mL of water. Then, 200 µL of an aqueous lucigenin stock
solution (1.3 mg/mL) and 200 µL of an aqueous SDS stock
solution (30 mg/mL) were added to 2 mL of ethanol. This solution
was added dropwise under vigorous stirring into 1 mL of the
polymer emulsion. To this emulsion 9 mL of water was added.
The emulsion was dialyzed against water. The reservoir was
changed four times every 3 days until it was colorless.
Preparation of the Copper(II)-Sensitive Nanobeads N9.
Three hundred milligrams of polymer emulsion was diluted with
2.7 mL of water. Then, 200 µL of aqueous lucifer yellow-CH stock
solution (1 mg/mL) and 64 µL of aqueous HDTMA stock solution
(38 mg/mL) were added to 2 mL of ethanol. This solution was
added dropwise under vigorous stirring into 1 mL of the polymer
emulsion. To this emulsion 9 mL of water was added. The
emulsion was dialyzed against water. The reservoir was changed
four times every 3 days until it was colorless.
The composition of the sensor materials is summarized in
Table 1.
Measurements. The size of the beads was determined with
a particle size analyzer Zetasizer Nano ZS (www.malvern.de).
Luminescence phase shifts were measured with a two-phase
lock-in amplifier (SR830, Stanford Research Inc., www.thinksr-
s.com). The emulsion of the beads in water (∼1 mg/mL)
contained in a glass vial was excited with the light of a violet LED
(λmax 405 nm, www.roithner-laser.com) which was sinusoidally
modulated at 5 kHz or at 700 Hz, respectively, in the case of the
oxygen-sensitive and temperature-sensitive beads. A bifurcated
fiber bundle was used to guide the excitation light to the vial and
to guide back the luminescence after passing through the OG
590 (Schott) glass filter. The luminescence was detected with a
photomultiplier tube (H5701-02, Hamamatsu, www.sales.hamamat-
su.com). Temperature was controlled by a cryostat ThermoHaake
DC50. In the case of the oxygen-sensitive beads the temperature
was kept constant at 1, 25, and 50 °C. Gas calibration mixtures
were obtained using a gas mixing device (MKS, www.mksinst-
.com). Three independent measurements were performed to
obtain a calibration curve.
Fluorescence measurements for pH-sensitive beads N4 and
N5 were performed on a Hitachi F-7000 fluorescence spectrometer
(www.inula.at). The pH was adjusted to the desired value using
phosphate buffer. The pH of the buffer solutions was controlled
Analytical Chemistry, Vol. 80, No. 3, February 1, 2008 575
by a digital pH meter (InoLab pH/ion, WTW GmbH & Co. KG,
www.wtw.com) calibrated at 20 ( 2 °C with standard buffers of
pH 7.0 and 4.0 (WTW GmbH & Co. KG). The buffers were
adjusted to constant ionic strength using sodium chloride as the
background electrolyte. Three independent measurements were
performed to obtain a calibration curve. Fluorescence measure-
ments of the nanobeads of types N6 and N7 were performed in
96-well microplates using a BMG Labtech Fluorstar Optima
microplate reader (www.bmg-labtechnologies.com) equipped with
485/20 and 560/20 band-pass filters for excitation and emission,
respectively. Eight independent measurements were performed
in parallel to obtain a calibration curve.
Fluorescence measurements of ion-sensitive nanobeads were
performed in 96-well microplates using a BMG Labtech Fluorstar
Optima microplate reader with 420/20 and 540/20 band-pass filters
for excitation and emission, respectively. The pH of the copper-
(II)-containing solutions was adjusted to pH 5 using a 10 mM
acetate buffer. The pH of the chloride sample solutions was
adjusted to 7.1 with a phosphate buffer (C(H2PO4- + HPO42-) )
13 mM) and to constant ionic strength of 230 mM using sodium
fluoride as the background electrolyte. Eight independent mea-
surements were performed in parallel to obtain a calibration curve.
Leaching Tests. Two milliliters of the emulsion in water (10
mg beads/mL) was placed in a syringe and pressed through a
fine filter (220 nm pore size, Rotilabor, Roth, www.carl-roth.de).
The first 0.5 mL of the slightly turbid filtrate was discarded, and
the main transparent fraction was collected and investigated
spectroscopically. Leaching in the standard Luria broth () LB)
medium (10 g/L tryptone, 5 g/L yeast extract, 5 g/L NaCl) was
investigated in the same manner. Prior to filtration, the mixture
of 2 mL of the bead emulsion and 2 mL of LB medium was stirred
at room temperature for 24 h.
Toxicity and Penetration Tests. A preculture of Pichia
pastoris (strain: X-33 pGAPZB-PaHNL5R (BT2869)) was grown
in 50 mL YPD 1% (w/v) medium (1% yeast extract, 2% peptone,
1% dextrose) in 500 mL baffled shake flasks to OD600 ) 2.0. The
main culture was incubated to OD600 ) 0.26 (7.5 mL of the
preculture was added to 50 mL of the main culture) in BMD 1%
(w/v) medium (1.34% yeast nitrogen base w/o amino acids, 1%
dextrose, 0.02% D-biotin, 2% 10× PPB (30.13 g/L K2HPO4‚3H2O,
118.13 g/L KH2PO4)) in 500 mL buffeled shake flasks. Growth
conditions for both cultures were 28 °C and 120 rpm (incubator:
Certomat BS-1, www.sartorius-stedim.com). Two milliliters of the
emulsion of the nanobeads (N3, 10 mg/mL) was added to the
main culture, and the cells were monitored microscopically after
6, 12, and 24 h of cultivation.
A preculture of Escherichia coli (strain: BL 21 gold) was grown
in 25 mL of LB medium to OD600 ) 4.0 in 100 mL shake flasks.
The main culture was incubated to OD600 ) 0.52 (7.5 mL of the
preculture was added to 50 mL of the main culture) in LB medium
in 600 mL shake flasks. Growth conditions for both cultures were
30 °C and 160 rpm (incubator: Certomat BS-1). Two milliliters of
the emulsion of the nanobeads (N3, 10 mg/mL) was added to
the main culture, and the cells were monitored microscopically
after 3 and 6 h of cultivation.
The images were acquired on an Axiolab microscope (Zeiss,
www.zeiss.de) equipped with a ProgRes C14 digital camera
(Zeiss).
576 Analytical Chemistry, Vol. 80, No. 3, February 1, 2008
Figure 2. Size and distribution of the beads in different solvents
and solvent mixtures.
Fitting was performed using Origin version 7.5 (www.origin-
lab.com) software.
RESULTS AND DISCUSSION
Properties and Staining of the Beads. As specified by the
manufacturer, the poly(styrene-block-vinylpyrrolidone) nanobeads
consist of 64% w/w of styrene and 36% w/w of vinylpyrrolidone
and are <500 nm in size. In aqueous medium, the beads of this
block copolymer have core-shell structure in which nonpolar
polystyrene blocks form the core, whereas solvent-compatible
poly(vinylpyrrolidone) blocks form the interfacial region.45 Such
structure makes it possible for the beads to exist in water as a
very stable emulsion but also act as efficient solubilizers of
aromatic compounds.45,46
Measurements of the ζ-potential of the nanobeads water
emulsion revealed a potential of -12.1 mV, which indicates a
slightly negatively charged surface. We have found that if the
emulsion is dialyzed against water for 1 week the ζ-potential
decreases to -7.5 mV. We assume that the negative charge of
the beads is determined by residue of the ionic species such as
ascorbic acid, used as a stabilizer, whereas the beads are virtually
not charged. For comparison, the ζ-potential of 200 nm polystyrene
(PS) nanobeads with carboxylated surface (FluoSpheres, Molec-
ular Probes, www.probes.com) was measured to be -47 mV. The
small charge of the beads, the absence of reactive groups, and
the presence of the hydrophilic poly(vinylpyrrolidone) shell are
responsible for the unique properties of the material. The beads
are perfectly dispersible in water and do not aggregate even at
very high concentrations. Also, the addition of electrolytes (such
as, for example, saturated aqueous solution of NaCl) does not
result in aggregation. The beads can be used even in very complex
media such as those used during fermentation processes. In fact,
if dispersed in water, the size of the beads (as found by a particle
size analyzer) varies from 100 to 500 nm (Figure 2). The average
diameter (dav) was found to be 245 nm (peak width w is 72 nm).
The beads evidently do not aggregate in standard LB medium
(concentration of the beads was as high as 10 mg/mL), since the
size dav was measured to be 249 nm (w is 57 nm). In addition,
when freeze-dried, the beads can again be easily dispersed in
(45) Nagarajan, R.; Barry, M.; Ruckensteinf, E. Langmuir 1986, 2, 210-215.
(46) Haulbrook, W. R.; Feerer, J. L.; Hatton, T. L.; Tester, J. W. Environ. Sci.
Technol. 1993, 27, 2783-2788.
water. Water uptake by the dried particles was estimated to be
0.13 g per 1 g of the beads.
Addressed staining into PS core or poly(vinylpyrrolidone)
(PVP) shell can be performed easily using practically any lipophilic
dye. Note that staining of the other nanoparticles is usually rather
a sophisticated procedure.47 The swelling properties of the beads
were studied, and two different procedures for addressable
staining were established. It was found that, in ethanol, the
nanobeads behave similarly to water and evidently do not swell
(dav is 243 nm, w is 81 nm). Generally, lipophilic indicators
dissolved in ethanol/water mixture (70:30 v/v) can be forced into
the nanobeads by slowly elevating water content by evaporating
ethanol under reduced pressure. Addition of such solvents as
acetone, THF, dichloromethane, etc. to the emulsion of the beads
in water or ethanol results in swelling of the beads. Such solvents
as acetone and THF are of primary interest, since most lipophilic
indicators can be dissolved there, whereas the solvents could be
removed easily from their mixtures with water. Nanobeads were
found to increase in size significantly in acetone/water and THF/
water mixtures (Figure 2). Coagulation of the beads was observed
in pure acetone and THF, and their mixtures with water as well,
in which the solvent content exceeded 60-65%. Therefore,
mixtures of 50:50 v/v were used. As can be seen, the highest
degree of swelling is observed in the THF/water mixture (dav 780
nm, w 240 nm), whereas in the acetone/water mixture it is less
significant (dav 460 nm, w 98 nm). Immobilization in the lipophilic
core of the beads was achieved by adding the solution of a
lipophilic dye in THF dropwise to the emulsion of the beads in
the THF/water mixture so that ratio of THF/water 50:50 v/v was
achieved. Tetrahydrofuran was then removed under reduced
pressure. In both procedures indicator molecules are forced into
the bead (either core or shell) by slowly elevating water content.
In contrast to most bead staining procedures, dyeing of PS-PVP
particles is unsophisticated. Moreover, the process can be up-
scaled easily. Since PS and PVP differ significantly in polarity, we
used the polarity-sensitive dye Nile red to confirm localization of
the indicators. The fluorescence band of the probe is known to
shift bathochromically with increased solvent polarity.48 When the
dye was embedded from the THF/water mixture, the emission
maximum was found to be 574 nm (Figure 3). When incorporated
from the ethanol/water mixture the emission maximum was 623
nm; this indicates much higher polarity of the medium and
consequently localization in the hydrophilic shell of a nanobead.
For comparison, in a PS film fluorescence maximum of Nile red
is 577 nm, whereas in a dry PVP film it is located at 635 nm. It
should be considered that the proposed immobilization procedures
can favor but not ensure the addressable staining into the core
or the shell of the beads. In fact, the staining also is influenced
by other factors, such as, e.g., the polarities of the indicators and
their solubilities in PS and PVP. Only staining from EtOH/H2O
can ensure that a dye is localized in the shell, whereas staining
from THF/H2O will result in a certain distribution of an indicator
(47) Borisov, S. M.; Mayr, T.; Karasyov, A. A.; Klimant, I.; Chojnacki, P.; Moser,
C.; Nagl, S.; Schaeferling, M.; Stich, M. I.; Vasylevska, G. S.; Wolfbeis, O.
S. New Plastic Microparticles and Nanoparticles for Fluorescent Sensing
and Encoding. In Springer Series of Fluorescence; Berberan-Santos, N., Ed.;
Vol. 4: Fluorescence of Supermolecules, Polymers, and Nanosystems, 2008;
pp 431-463. DOI: 10.1007/4243_013.
(48) Boldrini, B.; Cavalli, E.; Painelli, A.; Terenziani, F. J. Phys. Chem. A 2002,
106, 6286-6294.
Figure 3. Fluorescence spectra of Nile Red when immobilized into
the core and the shell of the beads (λexc ) 515 nm).
between the core and the shell. In the latter case, even for
nonpolar indicators excellently soluble in PS some amount of them
can be located in the PVP shell or in the intermediate region.
Nanobeads for Sensing Oxygen. The platinum(II) complex
of 5,10,15,20-tetrakis-(2,3,4,5,6-pentafluoro-phenyl)-porphyrin (Pt-
TFPP) was chosen as an oxygen-sensitive probe because of its
excellent photostability49,50 and good brightness.51 The complex
was incorporated into the PS core using the THF/water procedure.
If the beads with the immobilized complex are then dispersed in
ethanol, and the dispersion is passed through a fine filter (220
nm pore size, Rotilabor), the transparent filtrate remains practically
colorless. This indicates little leaching into ethanol and again
confirms localization of the indicator in the core of a nanoparticle.
The oxygen indicator also was immobilized into the shell of the
beads via the EtOH procedure. The calibration curves for the
nanobeads of both types are shown in Figure 4. As can be seen,
the Stern-Volmer plots (τ0/τ vs oxygen concentration) are not
linear. Notably, a similar degree of nonlinearity also is observed
for the same indicator if contained in a PS film.50,52 The depen-
dence was fitted using the common equation from the “two-site
model” which assumes the dye to be located in two regions of
different microenvironment:53
I τ f1 f2
) ) + (1)
I0 τ0 1 + K [O ] 1 + K2 [O ]
1
SV 2 SV 2
where f1 and f2 are the fractions of the total emission for each
component, respectively (with f1 + f2 being 1), and K1SV and K2SV
are the Stern-Volmer constants for each component. A fit nicely
matches experimental findings (r2 > 0.9997). Since f1 values are
rather uniform for all the temperatures investigated, medium
values were used for calculation of the Stern-Volmer constants.
For the beads obtained from ethanol the fitting gives K1SV of
0.201, 0.329, and 0.500 kPa-1 for 1, 25, and 50 °C, respectively.
The K2SV was found to be 0, and f1 was 0.836 ( 0.01. For the
beads stained from THF the fitting gave K1SV of 0.151, 0.271, and
(49) Lee, S.; Okura, I. Anal. Commun. 1997, 34, 185-188.
(50) Amao, Y.; Miyashita, T.; Okura, I. J. Fluorine Chem. 2001, 107, 101-106.
(51) Spellane, P. J.; Gouterman, M.; Antipas, A.; Kim, S.; Liu, Y. C. Inorg. Chem.
1980, 19, 386-391.
(52) Borisov, S. M.; Klimant, I. Anal. Chem. 2007, 79, 7501-7509.
(53) Sacksteder, L.; Demas, J. N.; DeGraff, B. A.; Bacon, J. R. Anal. Chem. 1993,
65, 3480-3483.
Analytical Chemistry, Vol. 80, No. 3, February 1, 2008 577
Figure 4. Calibration curves for the oxygen-sensitive beads based on PtTFPP: (a and b) beads stained from EtOH/H2O (N1); (c and d) beads
stained from THF/H2O (N2). Fitting of the Stern-Volmer plots (b and d) is performed using eq 1.

0.357 kPa-1, K2SV of 0.022, 0.012, and 0.017 kPa-1 for 1, 25, and 50
°C, respectively (f1 was 0.767 ( 0.14). It is evident that staining
from ethanol results in more sensitive nanobeads. Fine-tuning of
sensitivity to oxygen is therefore possible. Notably, quenching
efficiency in the case of the beads stained in the core is slightly
higher than that for the PS film (τ0/τ at 21.3 kPa was 2.9, 3.4, and
3.9, respectively, for 1, 25, and 50 °C in the case of the beads,
and τ0/τ was 2.9, 3.2, and 3.3, respectively, for 1, 25, and 50 °C in
the case of the sensor films). This can be explained by the fact
that the decay times in the absence of oxygen are higher in the
beads (65.1, 61.1, and 56.9 µs at 1, 25, and 50 °C, respectively)
than in the film (59.2, 55.4, and 51.0 µs at 1, 25, and 50 °C,
respectively).
As expected, the nanosensors exhibit higher sensitivity to
oxygen at elevated temperatures (Figure 4). Such behavior is
common for all optical oxygen sensors, and the temperature
effects need to be compensated for if measurements are performed
at varying temperatures.
Evidently, the sensitivity of the nanosensors can be varied
significantly if other oxygen indicators are used instead of the
platinum(II) porphyrin. For example, the palladium(II) complex
of 5,10,15,20-tetrakis-(2,3,4,5,6-pentafluoro-phenyl)-porphyrin (PdT-
FPP) is an ideal candidate if measurements of trace oxygen in
practically anaerobic solutions are performed, since its decay time
of ∼1 ms will enable high oxygen sensitivity of the beads. On the
other side, if measurements need to be carried out under pO2
exceeding 21 kPa, a widely used ruthenium(II)-tris-4,7-diphenyl-
1,10-phenanthroline probe (Ru-dpp, τ0 ∼ 6 µs) can be im-
mobilized. In fact, we have found that PdTFPP and Ru-dpp, as
well as phosphorescent iridium(III) coumarin complexes52 are
easily immobilized into the beads using the same protocols.54 For
example, in the case of Ru-dpp τ0/τ was found to be 1.45 at 25
°C and 21.3 kPa.54
Figure 5. Calibration curves for the temperature-sensitive beads
of type N3. Fitting is performed using eq 2.
Nanobeads for Sensing Temperature (N3). We have previ-
ously shown44 that the europium(III) tris(thenoyltrifluoroaceto-
nate) complex with coordinated antenna chromophore (“Eu-
(tta)3L”) can serve as an excellent temperature probe for
luminescence sensing and imaging. This complex is excitable by
visible light (λmax 402 nm in toluene) and benefits from very strong
brightness (Bs, defined as the product of quantum yield and the
molar absorbance at the excitation wavelength) which is ∼28 000
at 25 °C and ∼48 000 at 1 °C.44 In this work, the complex was
immobilized into the core of the nanobeads using the THF/water
procedure. It should be noted that it is not possible to immobilize
the europium(III) complex via the ethanol procedure due to
extremely poor stability of the complex in ethanol. When im-
mobilized from THF, the dye in the beads was found to remain
stable even at pH ∼1, while the indicator in solution is already
destroyed at higher pH. Elevating the temperature from 1 to 51
°C results in ∼2-fold decrease of the luminescence decay time
(Figure 5). A fit (with a correlation coefficient r2 > 0.9998) is
performed using an Arrhenius-type equation:55,56

(54) Borisov, S. M.; Klimant, I. Unpublished results, 2007. (55) Coyle, L. M.; Gouterman, M. Sens. Actuators, B 1999, 61, 92-99.

578 Analytical Chemistry, Vol. 80, No. 3, February 1, 2008

 (
τ ) k0 + k1 exp - (
∆E
R(T + 273) )) -1
(2)

where k0 is the temperature-independent decay rate for the

excited-state deactivation, k1 is the pre-exponential factor, ∆E is
the energy gap between emitting level and higher excited-state
level, and R is the gas constant. The temperature dependence of
the decay time (in the absence of oxygen) can be fitted with the
following parameters: k0 ) 1491 s-1, k1 ) 1.98 × 1010 s-1, and ∆E
) 43.7 kJ‚mol-1. Since cross-sensitivity to oxygen is rather
pronounced, the temperature measurements can be compromised
when the oxygen concentration differs significantly. In this case,
oxygen concentration should be known, which can be achieved Figure 6. Emission spectra of the pH-sensitive nanobeads N4
by dispersing oxygen-sensitive nanobeads along with the temper- (CHFOE from EtOH) at different pH (IS 0.1 M, λexc ) 475 nm).
ature-sensitive ones. Both can be excited, for example, with a
bright 405 nm LED, with emissions separated spectrally or by
decay time. As was previously demonstrated for planar sensor
foils,44 several iterations are enough to obtain true values of pO2
and temperature.
Nanobeads for Sensing pH. The nanosensors for pH make
use of the lipophilic indicators CHFOE and HPTS(OA)3. CHFOE
incorporated into a polyurethane hydrogel membrane was shown
previously42 to be suitable for optical pH sensing at physiological
conditions (pKa 6.96). Moreover, ratiometric measurements are
possible, since both protonated and deprotonated forms of the
pH indicator emit light at different wavelengths (Figure 6). Figure
7a shows dependence of the ratio of fluorescence intensities at
522 and 566 nm (λexc ) 475 nm) on pH at different ionic strengths
(IS). Minor dependence is observed at low IS (pKa is 6.2 and 5.9 Figure 7. Calibration curves for the pH-sensitive beads: (a) CHFOE
at IS 0.02 and 0.05 M, respectively); however, there is practically from EtOH (N4); (b) HPTS(OA)3 from THF (N7).
no dependence at higher IS (pKa 5.9 at IS 0.4 M) which is relevant
for physiological applications (IS ∼ 0.1 M) and marine systems collected using a 560/20 band-pass filter (e.g., in the microplate
(IS < 0.6 M). In contrast to many charged polymer matrixes, the reader). The pKa values obtained in both methods are essentially
PS-PVP is highly promising for designing pH nanosensors, since the same and are ∼0.4 units higher than those obtained by
it makes it possible to significantly reduce cross-sensitivity to IS. ratiometric measurements.
It should be mentioned that, unlike other indicator loaded beads, The calibration curves for the sensor beads N7 (HPTS(OA)3
both immobilization methods result in virtually identical results. immobilized via the THF procedure) are presented in Figure 7b.
In fact, when using the THF/H2O procedure, pKa of the beads In contrast to the widely used 1-hydroxypyrene-3,6,8-trisulfonate,
was measured to be 6.0 at IS 0.05 M, compared to the value of the sulfonamide derivative bears little negative charge so that the
5.9 for the EtOH/H2O procedure. It is likely that the amphiphilic sensor beads exhibit virtually no cross-sensitivity to IS. In fact,
nature of the indicator (which is composed of a hydrophilic the pKa values were found to be 6.53, 6.51, 6.37, and 6.37 for IS
chromophore and a long lipophilic tail) favors its localization in 0.02, 0.05, 0.2, and 0.5 M, respectively. The same trend is observed
the region of intermediate polarity between the core and shell. for the sensor beads N6, which are stained from EtOH, and the
It should be considered that the pKa values obtained by pKa values are 6.69, 6.57, 6.59, and 6.48 for the same IS. Notably,
ratiometric measurements are only apparent ones because of the staining from EtOH/H2O results in brighter beads than staining
fluorescence energy resonance transfer (FRET) which complicates from THF/H2O. Nevertheless, as in the case of CHFOE both
the situation. In fact, the emission spectrum of the acidic form procedures result in rather similar materials in respect to pKa
overlaps with the excitation spectrum of the basic form so that values.
FRET becomes increasingly efficient with increasing pH and the Although the sensing materials based on CHFOE and HPTS-
fluorescence intensity of the acidic form drops faster. The problem (OA)3 are suitable for measurements in physiological condition
can be of course tackled by preparing beads with much lower and application in biotechnology, they are not fully adequate for
amount of the dye (C , 0.6% in the PVP shell), but this will result some applications (e.g., for those of marine science). Different
in significantly lower brightness. More accurate determination of substituents were shown to shift the pKa of fluorescein derivatives
the acidity constant is, however, possible by measuring the to higher or lower values.42 These lipophilic indicators can be
fluorescence intensity in the emission maximum of the basic form. incorporated into the PS-PVP beads in the same manner as was
Alternatively, the part of the emission from the basic form is shown for CHFOE.54 Such beads can be used for pH monitoring
in neutral condition, slightly acidic, or slightly basic media. For
(56) Liebsch, G.; Klimant, I.; Wolfbeis, O. S. Adv. Mater. 1999, 11, 1296-1299. example, at IS ) 0.2 M the pKa values for the beads stained with
Analytical Chemistry, Vol. 80, No. 3, February 1, 2008 579
Figure 8. Calibration curves for the nanosensors for Cl- (N8) and
Cu2+ (N9). Figure 9. Course of the enzymatic oxidation of glucose monitored
with help of nanobeads N1 (2 mg/mL) and of a standard sensor foil
(PtTFPP in PS, 1.5% w/w, ∼6 µm thick). The concentration of glucose
2′,7′-didodecylfluorescein (from THF) and 2′,7′-dichlorofluorescein is 0.25 M. An amount of 100 µL of glucose oxidase solution (C ) 10
mg/mL) is added to the test solution (V ) 5 mL). Excitation source:
octadecylester (from THF) were found to be 7.7 and 5.7,
405 nm LED. The excitation and emission filters are BG 12 and RG
respectively.54 Being based on different fluorescein derivatives, 630, respectively. The modulation frequency is 5 kHz. The experiment
all the beads exhibit virtually identical spectral properties (excita- was performed at room temperature.
tion and emission) but different pKa values. Thus, a homogeneous
mixture of the beads can be prepared, which will possibly enable
mM copper(II) solution by a decrease of fluorescence intensity
pH monitoring in a wide range, compared to the dynamic range
of -15.1% and -76.6%. The dynamic range is from 0.01 to 100
of the pH electrode.
mM. Similar dynamic ranges were found for nanosensors prepared
Nanobeads For Sensing Ions. Sensing schemes for ion-
with various molar ratios of anionic dye and added lipophilic
sensitive nanobeads (Cl- and Cu2+) are based on the fluorescence
counterion. The dynamic range is 1000-fold higher than that found
quenching of a fluorescent indicator dye immobilized in the
for dye solutions and sensing membranes based on lucifer yellow
hydrophilic shell of the beads. Lucigenin and lucifer yellow-CH
immobilized on anion exchanger cellulose. This is attributed to a
are known to be efficiently quenched by chloride57 and copper-
repulsion of the analyte by an excess of immobilized lipophilic
(II),58 respectively. The dyes show good solubility in water due
cation.
to their positively charged amino (lucigenin) or negatively charged
It was also found that if the “sensing chemistries” were
sulfo groups (lucifer yellow-CH). The dyes were immobilized by
incorporated into the core of the beads by using the THF/water
exchanging the counterions with lipophilic counterions (dodecyl
procedure, the nanobeads did not exhibit any sensitivity for ions.
sulfate or hexadecyltrimethylamine). This yields ion pairs that are
Response. A very important feature of the nanosensors which
localized in the hydrophilic shell of the beads and do not leach
makes them advantageous over planar sensor foils and even
into the aqueous medium. In fact, the reservoir from dialysis was
microsensors includes very short response times. Since the
colorless after four cycles indicating that all indicator dye is located
nanobeads introduce virtually no diffusional barrier for the
in the beads, not in the solvent.
diffusion-limited quenching processes, real time monitoring of fast
Response of the chloride nanosensor (N8) is shown in Figure
processes becomes possible. Figure 9 shows the course of
8. The titration plot exhibits a typical sigmoidal shape. The
enzymatic reaction of glucose oxidation monitored with help of a
dynamic range of the nanosensors is 0.5-100 mM of chloride.
standard planar sensor foil (PtTFPP in PS, ∼6 µm thick) and using
The nanosensors respond to a 100 mM chloride solution by a
the oxygen-sensitive beads N1. The apparent response of the
decrease in fluorescence intensity of -80%, and the concentration
nanobeads is ∼4 s, which represents, however, the overall time
at which 50% of the fluorescence is quenched (c1/2) is 6.3 ( 0.3
needed for mixing the enzyme and the enzymatic reaction itself.
mM. The apparent c(1/2) is ∼3-fold higher than the value found in
In fact, if a concentrated emulsion of beads is added under
solution (c(1/2) ) 2.3 mM)59 and ∼4-fold lower than to the value
vigorous stirring to the anaerobe aqueous solution, the whole
reported for lucigenin immobilized to a hydrogel (c(1/2) ) 25 mM)
signal change occurs in less than 1 s. The response of the pH
composed of poly(acrylamide) and poly(acrylonitrile). The more
and ion nanosensors is also very fast and occurs in less than 1 s,
efficient quenching in solution can be attributed to the better
which is evidently the time needed for stirring. The nanosensor
accessibility of the quenchers to the fluorophores, whereas the
for temperature responds virtually instantaneously, and so are
lower c(1/2) compared to that of the hydrogel can be explained by
temperature-sensitive foils, since no diffusion of the analytes is
the more hydrophilic environment.
required.
The response of copper(II)-sensitive nanospheres (N9) is also
Dye Leaching. As was demonstrated above the nanobeads
shown in Figure 8. The nanosensors respond to a 0.1 and 100
can very easily be stained with the indicators. Since only covalent
(57) Huber, Ch.; Klimant, I.; Krause, Ch.; Werner, T.; Mayr, T.; Wolfbeis, O. S. coupling insures that no leaching occurs, it can be possible for
Fresenius’ J. Anal. Chem. 2000, 368, 196-202. the stained poly(styrene-block-vinylpyrrolidone) nanobeads. The
(58) Mayr, T.; Wencel, D.; Werner, T. Fresenius’ J. Anal. Chem. 2001, 371, 44- emulsion of the beads (10 mg/mL) was filtered through a fine
48.
(59) Huber, Ch.; Krause, Ch.; Werner, T.; Wolfbeis, O. S. Microchim. Acta 2003, filter (220 nm pore size), and absorption and emission spectra of
142, 245-253. the filtrate were recorded. It was found that no leaching was
580 Analytical Chemistry, Vol. 80, No. 3, February 1, 2008
observed for any of the nanosensors even after 2 months of
storage in aqueous solution. We also investigated leaching in the
complex (LB) medium, which contains proteins and amino acids
and an inorganic salt (NaCl). After 24 h of continuous stirring of
the emulsion in the LB medium (1:1 v/v) the filtration was
performed. In the case of the oxygen, temperature, and pH
nanosensors (N1-N7, stained both from EtOH/H2O and THF/
H2O) no leaching was detected. On the other side, lucifer yellow-
CH was leached completely out of the Cu2+ nanosensors (N9). It
is likely that leaching is promoted by the exchange of hexade-
cyltrimethylammonium cation in the ion pair by sodium cations.
The Cl- nanosensors can evidently not be used in LB medium,
and leaching there was therefore not investigated.
Interference Tests. Water-soluble indicators are rarely used
for sensing purposes in complex media since they tend to absorb
on the proteins and other components of the media which can
dramatically alter the sensing properties. Apart from those
interactions, quenching of luminescence by other components of
the medium can occur. These interferences are generally signifi-
cantly reduced, but not always removed completely, upon im-
mobilization. The nanosensors for oxygen, temperature, and pH
were thus investigated in LB medium. In the case of all the
nanosensors immobilization in the core via the THF/H2O proce-
dure results in more robust sensors. In fact, at 25 °C τ0 for the
N2 nanosensor (PtTFPP from THF/H2O) is 5% lower in the LB
medium compared to aqueous emulsion and τ0/τ is 2.7, which is
slightly lower than the value obtained in water (τ0/τ is 2.88). Some
decrease in decay time can be, however, due to the background
fluorescence of the LB medium which is not fully discarded using
the emission filters.
The LB medium definitely has pronounced effect on the decay
time of PtTFPP when the beads are stained from EtOH/H2O. In
fact, the decay time drops by 28% in LB medium. The sensitivity
also was much lower (τ0/τ was 2.5 and 3.72 for the beads in LB
medium and in water, respectively). The data indicate that PtTFPP
located in the shell is likely to interact with the components of
the medium.
In the case of the temperature-sensitive beads N3, the decay
times in LB medium are 6% and 11% lower (at 1 and 50 °C,
respectively) than in water. Since brightness of the beads
increases at lower temperatures, the influence of the LB medium
can again be partly cause by background fluorescence. From the
pH-sensitive nanobeads only those of type N5 (CHFOE from THF)
show identical behavior in aqueous solution and in LB medium.
As in the case of the oxygen indicator, immobilization of CHFOE
from THF/H2O helps to minimize the interferences. Notable
changes occur in the calibration curve for the beads stained from
EtOH/H2O (N4), and the pKa value was found to increase by 0.5
units. The calibration plots for the beads based on HPTS(OA)3
were both altered in the LB medium and lost their typical
sigmoidal form so that determination of acidity constants became
unreliable.
Toxicity and Uptake by Cells. We have demonstrated above
that the nanosensors for oxygen, temperature, and pH which are
obtained by staining from THF/H2O are rather robust to perform
adequately even in complex media such as those used in
bioreactors. It is, however, also important that the materials exhibit
Figure 10. Microscopic image of the cultivation media containing
E. coli (left) and P. pastoris (right) and the beads N3 stained with the
temperature indicator (0.5 mg/mL).
no cell toxicity and no uptake by the cell occurs over prolonged
time.
Therefore, cultivation of E. coli and P. pastoris was performed
in the presence of the nanobeads (N3). We have found that the
beads in a concentration of 0.5 mg/mL do not influence the growth
of the cells. We assume that the nanobeads are not toxic for these
robust cells; however, no experiment was performed with more
sensitive mammal cells. In addition, we have investigated the
possibility of the uptake of the beads by the cells. Figure 10 shows
the photographic images obtained with a fluorescence microscope
of the cultivation media after 3 and 6 h (in the case of E. coli and
P. pastoris, respectively). Evidently, no interaction is visible
between the cells and the beads which remain outside of the cell
membranes. Similar results were obtained after longer cultivation
times (6 and 24 h for E. coli and P. pastoris, respectively). The
results confirm that the beads can be successfully used for sensing
and imaging in cell cultures.
Storage Stability. All the beads can be stored as an aqueous
emulsion for at least 3 months (at 4 °C) without alteration of their
properties. Freeze-drying can enhance the storage durability. The
effect of freeze-drying was investigated for oxygen, pH, and
temperature nanosensors. The freeze-dried beads are dispersible
in water within 2 min under ultrasonication. The calibration curves
for the temperature nanosensor and the oxygen nanosensor N2
(PtTFPP stained from THF/H2O) are not affected by freeze-drying.
In the case of the oxygen-sensitive beads obtained from EtOH/
H2O an ∼25% decrease in the decay times (at 0 kPa O2 and at air
saturation) was observed after freeze-drying which can be caused
by aggregation of the indicator in the dry shell. Freeze-drying
has only a minor effect on the properties of the pH-sensitive
nanobeads except those based on HPTS(OA)3 stained from THF/
H2O (N7), the brightness of which was reduced significantly after
freeze-drying. For the other pH nanosensors, freeze-drying results
in the increase of pKa values. In fact, for the nanobeads N4, N5,
and N6 the pKa values increased by 0.1, 0.21, and 0.23 units,
respectively. The data indicate that all the beads can be freeze-
dried and redispersed without significant alteration of the sensing
properties, especially those stained from THF/H2O.
Multianalyte Sensing. Above we have shown that fluores-
cence intensity, fluorescence intensity ratio, and luminescence
Analytical Chemistry, Vol. 80, No. 3, February 1, 2008 581
decay time (via phase fluorometry) can serve as analytical
parameters for sensing with the help of the nanobeads. Ratiometric
measurements were demonstrated for the dually emissive pH-
sensitive probe. In general, ratiometric measurements can be
performed using a second (analyte-insensitive) dye; in this case
the reference dye and the indicator should possess overlapping
absorption spectra to enable their simultaneous excitation but
significantly different emission spectra. Note that ratiometric
measurements are quite common,25,30 especially when measure-
ments by means of confocal microscopy are performed. The
flexibility of the presented material can enable immobilization of
the indicator and the reference in different parts of the beads,
i.e., into the core and the shell. Alternatively, two types of beads
(i.e., sensor beads and reference beads) can be dispersed together. Figure 11. Emission spectra of the mixture of the oxygen-sensitive
In this case, the ratio of the components can be optimized easily. beads N2 (C ) 0.66 mg/mL) and the temperature-sensitive beads
N3 (C ) 0.167 mg/mL) in aqueous solution (λexc ) 405 nm). All the
Another exciting possibility includes multianalyte sensing and
solutions contained 0.1 M glucose. The deoxygenated solutions
imaging using the modified dual lifetime referencing method additionally contained 0.2 mg/mL glucose oxidase.
(mDLR).11,23,60 Particularly, nanosensors with long luminescence
decay time (for oxygen or temperature) can be mixed together
with fluorescent nanobeads (such as those for pH or ions). the beads by using different staining protocols. The practically
Multianalyte sensing with mDLR becomes possible if the two types uncharged PVP shell allows for the unique stability of the
of beads are excitable at a single wavelength and the emission of nanosensors which show no tendency to aggregate even when
both is detected simultaneously at different modulation frequen- used in high concentration in complex media. It is demonstrated
cies.23,60 Because of their small size, the beads result in truly
that the obtained materials can be used for optical sensing and
homogeneous emulsions.
imaging of important analytes, such as O2, Cl-, and Cu2+ but also
Since spectral properties of the nanosensors are very different
for determination of temperature and pH. Simultaneous sensing
and so are the optical windows required for excitation of the beads
of two analytes is also possible. It is shown that the nanosensors
and reading of their emission, multianalyte measurements can be
for oxygen, temperature, and pH can be used in complex LB
performed in another way. A typical example is a mixture of the
medium. The nanosensors can be used for sensing and imaging
oxygen-sensitive beads and the temperature-sensitive beads which
of even very fast processes in biotechnology, marine science,
is suitable for simultaneous determination of both parameters
environmental monitoring, etc.
(Figure 11). The emissions can be isolated with two sets of filters,
and the luminescence decay times can be detected independently
for each indicator. On the other side, no spectral separation of ACKNOWLEDGMENT
the signals is required using imaging in the time domain since This work was partly supported by the European Commission
the decay times of both indicators differ significantly and can be Fifth and Sixth Framework programs MAST (N 1645000012) and
resolved. CLINICIP (EU506965). We thank Dr. Jochen Gerlach from the
In summary, we have described a novel and versatile platform Research Center of Applied Biocatalysis (Petersgasse 14, 8010
for designing luminescent nanosensors. The sensing nanobeads Graz, Austria) for the help in performing the cell penetration tests.
can easily be prepared using commercially available poly(styrene-
block-vinylpyrrolidone) aqueous emulsion. The “sensor chemis-
tries” can be immobilized either into the core or into the shell of Received for review June 28, 2007. Accepted October 25,
2007.
(60) Vasilevska, A. S.; Borisov, S. M.; Krause, Ch.; Wolfbeis, O. S. Chem. Mater.
2006, 18, 4609-4616. AC071374E

582 Analytical Chemistry, Vol. 80, No. 3, February 1, 2008