Resonance energy transfer in conjugates
of semiconductor nanocrystals and
organic dye molecules
Mikhail Artemyev
Downloaded From: http://nanophotonics.spiedigitallibrary.org/ on 11/06/2012 Terms of Use: http://spiedl.org/terms
Resonance energy transfer in conjugates of
semiconductor nanocrystals and organic dye molecules
Mikhail Artemyeva,b
a
Belarusian State University, Institute for Physico-Chemical Problems, 220030 Minsk, Belarus
[email protected]
b
Moscow Engineering Physics Institute, Laboratory of Nano-Bioengineering,
31 Kashirskoe shosse, 115409 Moscow, Russian Federation
Abstract. I analyze the efficiency of Förster resonance energy transfer (FRET) in luminescent
donor-acceptor complexes based on conjugates of CdSe/ZnS quantum dots and nanorods and
the luminescent dyes. Semiconductor nanocrystals serve either as FRET donors or acceptors.
Experimentally observed reduced FRET efficiency in complexes of nanorods and dye molecules
as compared to quantum dots are found to be attributable to a distance-limited energy transfer
rate in case of point-like dye dipoles and extended nanorod dipole. © 2012 Society of Photo-Optical
Instrumentation Engineers (SPIE). [DOI: 10.1117/1.JNP.6.061705]
Keywords: quantum dots; excitons; fluorescence.
Paper 12085SS received Jun. 30, 2012; revised manuscript received Sep. 1, 2012; accepted for
publication Sep. 6, 2012; published online Oct. 5, 2012.
1 Introduction
Förster resonance energy transfer (FRET) in a system of semiconductor quantum dots (QDs) and
organic dye molecules is important for practical applications in biology, medicine, sensing,
photovoltaic devices, etc.1–7 Conjugates of QDs and organic molecules serve as pH and redox
sensors1–5 and multiplexed markers.6–9 Most experiments are done with CdSe or CdTe based
core-shell QDs serving as FRET donors and various luminescent dye molecules serving as
acceptors.9–12 Quantum dots are almost ideal candidates for FRET donors or acceptors, due
to their extended optical absorption, tunable and narrow emission band, and relatively high luminescence quantum yield. Semiconductor nanorods (NRs) attract considerably less attention,
although they possess unique optical properties.13 NRs absorb and emit strongly polarized light,
due to the optical dipole oriented and extended along the NR long axis.14 FRET in the system
of oriented dipoles is an important tool for studying the relative orientation and dynamics of
chromophores. Additionally, an extended optical dipole in NRs causes a much larger molar
absorption coefficient, which is important for achieving efficient FRET.
Usually, FRET is studied between different organic molecules, inorganic ions, or small nanoparticles of more or less similar size. In this case, the classic Förster theory describing interaction
between point-like optical dipoles is valid. In case of semiconductor NRs, their length is about two
orders of magnitude larger than the size of an organic dye molecule. Taking into account that
the optical dipole in a semiconductor NR is extended over the whole NR length, the FRET problem
in case of an NR-dye conjugate can no longer follow the Förster rules. Since I am interested mainly
in FRET efficiency in dye-nanoparticle conjugates for practical applications, the nanoparticle
geometry and dimension can play a very important role. In this paper, I make a comparative study
of the FRET efficiency in conjugates of fluorescent dye molecules and CdSe/ZnS core-shell QDs
and NRs where both QDs and NRs serve as either the FRET donors or acceptors.
2 Experiment and Materials
Hydrophobic CdSe-ZnS core-shell QDs (4 nm in diameter) and NRs (4 nm in diameter and
23 nm in length) were synthesized according to previously published procedures based on
0091-3286/2012/$25.00 © 2012 SPIE
Journal of Nanophotonics
061705-1
Downloaded From: http://nanophotonics.spiedigitallibrary.org/ on 11/06/2012 Terms of Use: http://spiedl.org/terms
Vol. 6, 2012
Artemyev: Resonance energy transfer in conjugates of semiconductor nanocrystals : : :
high-temperature reaction between organometallic precursors in coordinating solvents.15 The
nanocrystals were solubilized in water with D,L-cysteine (Cys), or a mixture of aminoethanethiol and dimethylaminoethanethiol. Briefly, dry nanocrystals (ca. 5 mg) were dissolved in
chloroform (2 ml), and methanol (3 ml) was then added. The solid phase was centrifuged,
washed with methanol, and dissolved in chloroform (3 ml). Next, an equimolar mixture of aminoethanethiol hydrochloride (Aldrich, 0.3 ml) and dimethylaminoethanethiol hydrochloride
(Aldrich, 0.3 ml) in methanol (0.1 mmol∕mL each) was added to the nanocrystals in chloroform,
and the mixture was stirred for 20 min to complete the binding of thiols to the surface of the
nanocrystals.
In another option, 0.5 mL of D,L-cysteine hydrochloride (Cys) in methanol (5 mg∕mL) was
added to nanocrystals in chloroform, and the mixture was stirred for 10 to 15 min to complete the
binding of Cys molecules to the surface of nanocrystals.
The solid phase containing the nanocrystals with the monolayer of aminothiols on their surfaces was centrifuged, washed two times with methanol, and dissolved in anhydrous dimethylsulfoxide (DMSO, 3 ml). Alexafluor®647 carboxyl acid succinimidyl ether (Invitrogen) was
dissolved in anhydrous DMSO at the concentration of 0.1 mM, and 100 μL of this solution
was added to the nanocrystals in DMSO solubilized with the aminoethanethiol/dimethylaminoethanethiol mixture. The mixture was stirred in the dark at room temperature for 2 h
in order to complete the reaction between the succininimidyl ether groups of Alexafluor®647
and amino groups of aminoethanethiol on the surface of the nanoparticles. The conjugates of
nanocrystals with dye molecules where dialyzed against bidistilled water. During the dialysis,
the conjugates formed the aggregates, which were removed from the solution by centrifugation
(5000 rpm, 3 min). The solid phase was washed several times by bidistilled water, in order to
remove unbound dye molecules, and redispersed in a fresh portion of DMSO (3 ml).
For Cys-stabilized nanocrystals, fluorescein isothiocyanate (FITC, Sigma) was conjugated
with NRs and QDs according to the published protocol.16 Briefly, 1 mg of FITC was dissolved in
5 ml of anhydrous DMSO and stored at 4°C. Then two 20-μL aliquots of FITC solution were
added to 2-mL aliquots of aqueous solutions of QDs and NRs at room temperature and stirred for
1 h in the dark. Then both solutions were purified by dialysis against 1∶100 diluted phosphate
buffer in order to remove unbound FITC. QD-FITC and NR-FITC conjugates were embedded
into the polymeric matrices and studied in a solid form.
To prepare polymeric films containing QD-FITC and NR-FITC conjugates, a polymer polyvinylpyrrolidone (PVPR, Aldrich, MW ¼ 1300000) was added to each solution to the concentration of 5 mg∕mL. Additionally, three reference aqueous solutions with the same concentration
of PVPR and pH were prepared with Cys-solubilized QDs, Cys-soluibilized NRs, and pure
FITC. Finally, all five solutions were drop-cast on to round quartz substrates 2.5 cm in diameter
and dried at room temperature in the dark to obtain polymeric films. The average thickness of all
polymeric films was about 10 μm, and the optical density of all samples at the emission range
of nanocrystals (ca. 600 to 650 nm) was kept below 0.3 in order to avoid the emission’s
reabsorption.
The photoluminescence (PL) spectra of all samples were recorded at room temperature with
the Jobin-Yvon Fluoromax-2 spectrofluorimeter, and the optical absorption spectra were
recorded with the Ocean Optics HR 2000 spectrophotometer.
3 Results and Discussion
Semiconductor QDs and NRs possess extended optical absorption above their band edge. This
causes efficient excitation of QD and NR PL at different wavelengths and makes the studied
CdSe QDs and NRs ideal candidates for FRET donors. Since dye molecules serving as FRET
acceptors have a relatively narrow absorption band, one may easily achieve selective QD or NR
optical excitation behind the optical absorption of dye molecule. Figure 1 shows the room
temperature optical absorption and PL spectra of CdSe/ZnS core-shell QDs and NRs and their
conjugates with AlexaFluor®647 dye.
The FRET efficiency in QD-dye and NR-dye conjugates may be determined according to the
formula:
Journal of Nanophotonics
061705-2
Downloaded From: http://nanophotonics.spiedigitallibrary.org/ on 11/06/2012 Terms of Use: http://spiedl.org/terms
Vol. 6, 2012
Artemyev: Resonance energy transfer in conjugates of semiconductor nanocrystals : : :
Fig. 1 Room temperature absorption (a) and PL (b) and (c) spectra of CdSe/ZnS QDs (red
curves), NRs (green curves), and their conjugates with AlexaFluor®647 dye in DMSO. The
blue and cyan curves in (a) correspond to the absorption and PL spectra of free AlexaFluor®647
dye in DMSO. The blue curves in (b) and (c) correspond to the PL spectra of AlexaFluor®647 dye
conjugated with QDs and NRs, respectively. The PL excitation wavelength λex ¼ 450 nm.
E ¼ 1 − FDA ∕FD ;
(1)
where FDA and FD are the fluorescence intensities of FRET donors with and without the acceptor, respectively, normalized to the nanoparticles concentration.17 Based on the data of Fig. 1, the
FRET efficiencies (E) for QD-dye and NR-dye conjugates were determined to be E ¼ 0.96 and
E ¼ 0.87, respectively. However, the different concentration of dye molecules per single QD and
NR must be taken into account. Since the surface area of QDs is much smaller than that of NRs,
the number of attached dye molecules is different. In order to estimate the number of dye molecules per single nanocrystal, the concentrations of QDs, NRs, and dye were determined from the
magnitude of absorption peak (first excitonic transition for semiconductor nanocrystals) and the
Journal of Nanophotonics
061705-3
Downloaded From: http://nanophotonics.spiedigitallibrary.org/ on 11/06/2012 Terms of Use: http://spiedl.org/terms
Vol. 6, 2012
Artemyev: Resonance energy transfer in conjugates of semiconductor nanocrystals : : :
reference data on the molar absorption coefficient for AlexaFluor®647, CdSe/ZnS QDs, and
NRs18 (the molar absorption coefficient for CdSe NRs demonstrates simple volume-dependent
behavior for fixed diameter19).
The average number of dye molecules per nanocrystal for each conjugate is 2.7 for QDs and
12 for NRs, respectively. Therefore, the FRET efficiency is much lower in NR-dye conjugates
than it is in QD-dye conjugates, taking into account the larger amount (roughly five times) of
acceptor molecules per donor in the NR-dye complex.
In the case of multiple acceptors, one should take into account possible nonlinear effects, like
FRET saturation. To study the saturation effect in QD-dye and NR-dye conjugates, the series of
conjugates were prepared with the average number of AlexaFluor®647 molecules varying from
1.5 to 27 per NR and from 1.5 to 3.4 per QD. The protocol for the preparation of conjugates with
different dye/nanoparticle ratios was the same as that described above, and only the amount of
dye used for conjugation was varied. Using Eq. (1), the FRET efficiency was determined within
the prepared conjugates, and these results are presented in Fig. 2 in the form of FRET efficiency
E plotted versus the average number of dye molecules per NR or QD.
Figure 2 clearly shows the presence of the saturation effect in FRET efficiency with an
increased number of dye molecules per CdSe NR. It must be pointed out that the saturation
effect is inherent to QDs, NRs, and other types of FRET pairs. However, the FRET efficiency
in QD-dye conjugate appears to be already above about three dye molecules per QD. The origin
of slower FRET saturation for NR-dye conjugates in Fig. 2 is not clear yet. It’s possible that the
dye–dye interaction growing with the increased amount of dye molecules bound to the surface of
each NR causes the reorientation of optical dipoles of dye molecules toward less favorable FRET
conditions.
Another important conclusion is that the E values for QD-dye and NR-dye conjugates
prepared at 1∶1 dye/nanoparticle ratios and derived from the corresponding fitting curves
point to the FRET efficiency from a single NR to a single dye (E ¼ 0.18) being less than
half that of a QD-dye pair (E ¼ 0.42). The reason for the different FRET efficiencies in
QD-dye and NR-dye complexes lies in the specific dipole–dipole interaction between the
extended optical dipole of a single CdSe NR and the point-like dipole of a single dye molecule
(in the case of a QD-dye complex, the classic FRET between two point-like dipoles is achieved).
To study this effect in a more complex way, the reverse situation was realized experimentally,
where CdSe QDs and NRs served as FRET acceptors, while FITC molecules served as FRET
donors. FITC molecules were chemically bound to the surface of Cys-solubilized nanocrystals,
and the resultant conjugates were embedded in solid polymeric film. Figure 3 shows the experimental PL and photoluminescence excitation (PLE) spectra of QD-FITC and NR-FITC
conjugates in PVPR.
Fig. 2 FRET efficiency E versus the number of AlexaFluor®647 dye molecules per CdSe/ZnS QD
(red triangles) or NR (green squares). The dotted red and green curves are the first-order exponential decay fits for QDs and NRs dye-conjugates, respectively. The arrows correspond to the
interpolated values of FRET efficiency for a single dye molecule per QD (red) or NR (black).
Journal of Nanophotonics
061705-4
Downloaded From: http://nanophotonics.spiedigitallibrary.org/ on 11/06/2012 Terms of Use: http://spiedl.org/terms
Vol. 6, 2012
Artemyev: Resonance energy transfer in conjugates of semiconductor nanocrystals : : :
Fig. 3 The PL (blue curves, λexc ¼ 400 nm) and PLE (red and black curves) spectra of free FITC
(a), QD-FITC (b), and NR-FITC (c) in PVPR films. For the red curves, λdetect ¼ 535 nm, and for
the black curves, λdetect ¼ 620 nm.
Both types of CdSe nanocrystals exhibit PL peaking at around λ ¼ 620 nm, while the FITC
PL band is located at around λ ¼ 540 nm. The PLE spectra, shown by black curves in Fig. 3(b)
and 3(c), show that the excitonic emission both from QDs and NRs in their conjugates with FITC
can be efficiently excited in the region of the absorption band of FITC molecules. This fact
demonstrates the presence of very efficient FRET from FITC molecules to nanocrystals. The
Journal of Nanophotonics
061705-5
Downloaded From: http://nanophotonics.spiedigitallibrary.org/ on 11/06/2012 Terms of Use: http://spiedl.org/terms
Vol. 6, 2012
Artemyev: Resonance energy transfer in conjugates of semiconductor nanocrystals : : :
Fig. 4 A scheme of the NR-dye and QD-dye complexes involved in FRET. The blue bold arrows
show FRET between NR or QD and dye molecules bound to the surface of the corresponding
nanocrystals. Unlike with the QD, only the strongly distance-limited part of the NR may ensure
efficient FRET.
FRET efficiency can be determined with the same Eq. (1) using the values of the PL signal from
free FITC molecules and FITC bound to the acceptors.
Again, the estimated FRET efficiency E derived from the data of Fig. 3 is higher for the QDFITC complex (E ¼ 0.8) than it is for NR-FITC (E ¼ 0.6). These values were calculated taking
into account the relative concentration of FITC molecules in all studied samples.
The number of FITC molecules per QD is estimated at around 6, while it is around 75 for
NRs.20 Such large discrepancies in the number of dye molecules per nanocrystal in the case of
FITC and AlexaFluor®647 dye molecules is due to the much smaller size of FITC (molecular
weight of 389, compared with 1300 for AlexaFluor®647) and the more efficient conjugation of
FITC molecules with amino groups of Cys molecules on the surface of CdSe/ZnS nanocrystals.
Taking into account the much larger concentration of FITC molecules per NR, one may conclude
that the FRET efficiency from dye to NR in our NR-dye conjugates is much smaller than that
from dye to QD. The physical form of the studied conjugates (a solid sample in the case of FITC
and a colloidal solution in the case of AlexaFluor®647) does not affect significantly the general
conclusion that semiconductor NRs are less efficient FRET donors or acceptors than QDs. The
reason of such weak efficiency lies in the nature of dipole–dipole interaction between dye molecules and nanocrystals. The length of the effective region in the NR to which the energy from/to
dye molecule may be transferred is of the order of R—the distance between dye and the NR
center (Fig. 4). It is about 3 to 4 nm and close to the average diameter of our NRs.
The conventional point-dipole FRET theory is not valid when the FRET-based configuration
involving extended oriented dipoles is concerned. A number of sophisticated FRET theories and
models have been developed for the systems of interacting extended and point-like optical
dipoles.21–23 The fact that the energy transfer rate is distance-limited in the case of the energy
transfer between extended and point-like dipoles reduces the NR region from which the efficient
FRET may occur to only a few nanometers.21 For example, the one-dimensional origin of the
optical dipole in an NR-based FRET system may cause the slower spatial decay of energy
transfer rate γ ∼ 1∕R5 , compared with γ ∼ 1∕R6 for conventional dipole–dipole interaction.21
However, a much larger number of dye molecules per NR compared with QDs equalizes
the situation and brings the total FRET efficiency in an NR-dye complex close to 100%.
4 Conclusion
I compared the efficiency of resonance energy transfer in the conjugates of CdSe/ZnS quantum
dots and nanorods, and different dye molecules served either as FRET donors or acceptors. A
sufficiently smaller FRET efficiency for nanorod-dye complexes was observed and explained by
the distance-limited energy transfer rate in the case of extended and point-like dipoles. However,
Journal of Nanophotonics
061705-6
Downloaded From: http://nanophotonics.spiedigitallibrary.org/ on 11/06/2012 Terms of Use: http://spiedl.org/terms
Vol. 6, 2012
Artemyev: Resonance energy transfer in conjugates of semiconductor nanocrystals : : :
the fact that the molar absorption coefficient of CdSe nanorods is almost an order of magnitude
larger than that of quantum dots, together with the bigger surface area that allows more dye
molecules to be attached to their surface, make nanorods-based FRET complexes very efficient
and perspective as fluorescent sensors.
Acknowledgments
This work was supported in part by the X11MC-27 grant of the Foundation for Basic Investigations (Belarus) and by the Ministry of Higher Education and Science of the Russian Federation
(Grants No. 11.G34.31.0050).
References
1. E. R. Goldman et al., “A hybrid quantum dot–antibody fragment fluorescence resonance
energy transfer-based TNT sensor,” J. Am. Chem. Soc. 127(18), 6744–6751 (2005), http://
dx.doi.org/10.1021/ja043677l.
2. M. Tomasulo, I. Yildiz, and F. M. Raymo, “pH-sensitive quantum dots,” J. Phys. Chem. B
110(9), 3853–3855 (2006), http://dx.doi.org/10.1021/jp060185h.
3. M. Suzuki et al., “Quantum dot FRET biosensors that respond to pH, to proteolytic or
nucleolytic cleavage, to DNA synthesis, or to a multiplexing combination,” J. Am.
Chem. Soc. 130(17), 5720–5725 (2008), http://dx.doi.org/10.1021/ja710870e.
4. I. L. Medintz et al., “Interactions between redox complexes and semiconductor quantum
dots coupled via a peptide bridge,” J. Am. Chem. Soc. 130(49), 16745–16756 (2008), http://
dx.doi.org/10.1021/ja805456x.
5. D. M. Willard et al., “Directing energy flow through quantum dots: towards nanoscale sensing,” Anal. Bioanal. Chem. 384(3), 564–571 (2006), http://dx.doi.org/10.1007/s00216
-005-0250-z.
6. A. R. Clapp et al., “Quantum dot-based multiplexed fluorescence resonance energy transfer,” J. Am. Chem. Soc. 127(51), 18212–18221 (2005), http://dx.doi.org/10.1021/ja054630i.
7. T. Pons et al., “Solution-phase single quantum dot fluorescence resonance energy transfer,”
J. Am. Chem. Soc. 128(47), 15324–15331 (2006), http://dx.doi.org/10.1021/ja0657253.
8. A. Sukhanova et al., “Nanocrystal-encoded fluorescent microbeads for proteomics: antibody profiling and diagnostics of autoimmune diseases,” NanoLett. 7(8), 2322–2327
(2007), http://dx.doi.org/10.1021/nl070966+.
9. I. Nabiev et al., Colloidal Nanoparticles in Biotechnology, A.Elaissari, Ed., pp. 133–168,
Wiley Interscience, New York (2008).
10. A. R. Clapp et al., “Fluorescence resonance energy transfer between quantum dot donors
and dye-labeled protein acceptors,” J. Am. Chem. Soc. 126(1), 301–310 (2004), http://dx
.doi.org/10.1021/ja037088b.
11. Q. Zhang et al., “Highly efficient resonant coupling of optical excitations in hybrid organic/
inorganic semiconductor nanostructure,” Nat. Nanotechnol. 2(9), 555–559 (2007), http://dx
.doi.org/10.1038/nnano.2007.253.
12. H. Lu et al., “Self-assembled donor comprising quantum dots and fluorescent proteins for
long-range fluorescence resonance energy transfer,” J. Am. Chem. Soc. 130(14), 4815–4827
(2008), http://dx.doi.org/10.1021/ja078243f.
13. S. Halivni et al., “Effect of nanoparticle dimensionality on fluorescence resonance energy
transfer in nanoparticle_dye conjugated systems,” ACS Nano 6(6), 2758–2765 (2012),
http://dx.doi.org/10.1021/nn300216v.
14. J. Hu et al., “Linearly polarized emission from colloidal semiconductor quantum rods,”
Science 292(5524), 2060–2063 (2002), http://dx.doi.org/10.1126/science.1060810.
15. Z. A. Peng and X. J. Peng, “Mechanisms of the shape evolution of CdSe nanocrystals,”
J. Am. Chem. Soc. 123(7), 1389–1395 (2001), http://dx.doi.org/10.1021/ja0027766.
16. T. G. Hermanson, Bioconjugate Technique, Academic Press, San Diego (1998).
17. B. W. van der Meer, G. Coker, and S.-Y. Chen, Resonance EnergyTransfer: Theory and
Data, Wiley-VCH, New York (1994).
Journal of Nanophotonics
061705-7
Downloaded From: http://nanophotonics.spiedigitallibrary.org/ on 11/06/2012 Terms of Use: http://spiedl.org/terms
Vol. 6, 2012
Artemyev: Resonance energy transfer in conjugates of semiconductor nanocrystals : : :
18. M. Hardzei et al., “Comparative efficiency of energy transfer from CdSe–ZnS quantum dots
or nanorods to organic dye molecules,” Chem. Phys. Chem. 13(1), 330–335 (2012), http://
dx.doi.org/10.1002/cphc.201100552;
19. E. Shaviv, A. Salant, and U. Banin, “Size dependence of molar absorption coefficients of
CdSe semiconductor quantum rods,” Chem. Phys. Chem. 10(7), 1028–1031 (2009), http://
dx.doi.org/10.1002/cphc.v10:7.
20. M. Artemyev, E. Ustinovich, and I. Nabiev, “Efficiency of energy transfer from organic
dye molecules to CdSe-ZnS nanocrystals: nanorods versus nanodots,” J. Am. Chem. Soc.
131(23), 8061–8065 (2009), http://dx.doi.org/10.1021/ja809839h.
21. P. L. Hernandez-Martinez and A. O. Govorov, “Exciton energy transfer between nanoparticles and nanowires,” Phys. Rev. B 78(3), 035314 (2008), http://dx.doi.org/10.1103/
PhysRevB.78.035314.
22. B. P. Krueger, G. D. Scholes, and G. R. Fleming, “Calculation of couplings and energytransfer pathways between the pigments of LH2 by the ab initio transition density cube
method,” J. Phys. Chem. B 102(27), 5378–5386 (1998), http://dx.doi.org/10.1021/
jp9811171.
23. S. Sadhu, M. Tachiya, and A. Patra, “A stochastic model for energy transfer from CdS
quantum dots/rods (donors) to Nile red dye (acceptors),” J. Phys. Chem. C 113(45),
19488–19482 (2009), http://dx.doi.org/10.1021/jp906160z.
Mikhail Artemyev received his PhD in physical chemistry in 1999 from
Belarusian State University. From 1999 to 2000, he was a fellow in the
chemistry department of Massachusetts Institute of Technology. In 2009,
he received a doctor of science degree in physical chemistry from BSU.
He is currently the head of the Laboratory of Nanochemistry in the Institute
for Physico-Chemical Problems at BSU and a nanochemistry group leader
in the Laboratory of Nano-Bioengineering at the Moscow Engineering
Physics Institute (Russia). His scientific interests relate to the synthesis
of semiconductor and metal nanocrystals and nanostructures, applications
in optics, optoelectronics, and biomedicine. He published more than 100 scientific papers in
high-ranked international journals.
Journal of Nanophotonics
061705-8
Downloaded From: http://nanophotonics.spiedigitallibrary.org/ on 11/06/2012 Terms of Use: http://spiedl.org/terms
Vol. 6, 2012