Publications
Publications
Publications
www.elsevier.com/locate/cplett
Abstract
Ensembles of Y2O2S:Eu3+ widegap semiconductor nano-crystals exhibit ON–OFF fluorescence blinking phenomenon, which mimic
II–VI semiconductor quantum-structures. Attempt has been made to explain this phenomenon under the ambit of diffusion controlled
electron transfer (DCET) model. Insights derived using opto-impedance technique in conjunction with the DCET model clearly establish
the relaxation of surface state(s) through a diffusion process. Ensembles of Y2O3:Eu3+ nano-crystals, the insulating counterpart inves-
tigated for comparison does not show such fluorescence blinking.
2007 Elsevier B.V. All rights reserved.
0009-2614/$ - see front matter 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.cplett.2007.01.009
156 A. Nakkiran et al. / Chemical Physics Letters 436 (2007) 155–161
same as described before [9–13]. YOS nano-samples used in regions of YOS and YOX samples as mounted in the fluo-
this study had a crystallite size of 7 nm while YOX nano- rescence spectrophotometer.
samples had a size of 5 nm. Transmission electron micro-
scope image(s) and the corresponding selected area electron 3. Results and discussion
diffraction (SAED) pattern consistent with the target phase
under investigation are given in Fig. 1 inset. For recording 3.1. ON–OFF fluorescence intermittence in oxysulfide
‘ON–OFF’ blinking events, a digital camera (Olympus quantum-dot ensembles
Version 1.1) equipped with a CCD detector operated in
multiple flash modes was used. The blinking images were Similar Stark splitting patterns of photo-luminescence
recorded while the samples being continuously excited emission spectra for the bulk and nano-YOS samples con-
using UV radiation with wavelength(s) corresponding to firm identical chemical surroundings for the Eu3+ center.
respective fundamental band-edge, charge transfer band Whilst the diffuse reflectance spectra (Fig. 1) corresponding
Fig. 1. Diffuse reflectance spectra of bulk (A) and nano-crystalline (B) Y2O2S:Eu3+ samples. Inset: Transmission electron microscope image of yttrium
oxysulfide nano-crystals (Y2O2S:Eu3+) and the corresponding SAED pattern.
A. Nakkiran et al. / Chemical Physics Letters 436 (2007) 155–161 157
to the fundamental absorption edge (260 nm) of the host- [15]. The moderate dynamic range observed may be attrib-
matrix shows a significant blue shift of 0.3 eV. This indi- uted to a low time resolution of the experimental set-up
cates a Bohr exciton size of 5.3 nm much comparable to the used. Also the large collection of quantum-dots in the sam-
crystallite size in suggesting a strong quantum confinement ple configuration may lead to poor ON–OFF time averag-
[11] while for yttrium oxide:Eu3+ nano-crystals these shifts/ ing moments. Whilst YOX nano-crystals do not show any
features are not observable or appreciable. such fluorescence blinking (Fig. 3a).
For the YOS nano-crystalline ensembles when being The relative intensities of ON–OFF events determined
excited continuously using UV radiation the fluorescence with respect to time and number of events/occurrences,
intensity goes ON–OFF (Fig. 2a) resembling luminescence especially the latter follow a Gaussian pattern(s) (Fig. 2
blinking reported in other II–VI compounds [4,6,14]. The insets b–c). Because photo-luminescence efficiencies of
luminescence intensity variation between the dark and YOS and YOX systems differ considerably, there is a signif-
bright events (referred as dynamic range) is very moderate icant difference in the luminescent intensity scales for fixing
limited to about 2–3 folds. However, this convincing differ- position(s) or limit(s) of the ON–OFF switching pattern(s).
ence in intensity scales between bright and dark periods can The blinking phenomenon observed in the YOS quan-
be ascribed to luminescence blinking mechanism in a simi- tum-structure can be explained using a four level scheme
lar fashion as observed in InGaAS/GaAs quantum-dots known as diffusion controlled electron transfer (DCET)
b c
Fig. 2. (a) Photograph images of successive frames showing ON–OFF blinking in Y2O2S:Eu3+ nano-crystals, (b) corresponding fluorescence intensity
versus time plot. (c) histogram showing the PL intensity versus number of ON–OFF occurrences fitted under single Gaussian. Using PL intensity scales
‘ON’ and ‘OFF’ states are identified as marked. Also (d) opto-impedance data corresponding to Warburg impedance part (diffusion process) on
Y2O2S:Eu3+ nano-crystals showing variation in slope (m) upon UV shining in a comparable time scale as the blinking is indicated.
158 A. Nakkiran et al. / Chemical Physics Letters 436 (2007) 155–161
b c
Fig. 3. (a) Photograph images of successive frames showing the absence of any blinking phenomenon in Y2O3:Eu3+ nano-crystals. Based on emission
intensity of events/frames can be considered to represent ON events only (b) corresponding fluorescence intensity versus time plot (c) histogram showing
the PL intensity versus number of ON occurrences fitted under single Gaussian. Also (d) opto-impedance data corresponding to Warburg impedance part
on Y2O3:Eu3+ nano-crystals showing no variation in slope (m).
model proposed by Tang and Marcus [14]. In this model, a Fig. 4. The charge transfer state (CTS) of europium in
back and forth spectral diffusion process between two yttrium oxysulfide lying near the vicinity of the band-edge
tracks comprising four states viz., light (neutral-quantum- can be analogized to the situation of an exciton weakly
dot) and dark quantum-dots (charge separated states- coupled to divalent europium species. Upon size miniatur-
CSS) is considered. In a similar way, the presence of large ization these CTS states may get ruptured leading to a sit-
ensemble of two kinds of YOS quantum-dots viz., neutral uation similar to that of charge separated state of the
and dark dots are possible here. The latter may be attrib- DCET model. Upon optical pumping the exciton trapped
uted to the presence of innumerable optically induced sur- at the surface-states will get relaxed via localized f–f levels
face states especially in the vicinity of conduction band and of europium, in particular 5D0 ! 7FJ levels of europium
charge separated state. Optical pumping can hasten the thereby explaining the blinking observed. This means that
relaxation of surface states or the release of captured elec- photo-induced surface state relaxation may generate an
trons to feed the luminescence transitions via the excited electron transfer process between charged and neutral
CTS through a resonance process as schematized in quantum-dots. The electron transfer between the two states
A. Nakkiran et al. / Chemical Physics Letters 436 (2007) 155–161 159
Fig. 4. Schematic diagram to illustrate the blinking mechanism in Y2O2S:Eu3+ nano-crystals following diffusion assisted charge transfer between neutral
and charge separated states (CSS). EFS and ECTS are the Fermi level of surface states and Eu3+ – ligand charge transfer states in YOS nano-crystals.
has been established to follow a one-dimensional back and II, III in Fig. 5) undergoes significant modification.
forth diffusion process explained by the potential cross- The semi-circular region I in the high frequency side
over model [16]. As will be seen in the subsequent sections, (1 MHz–60 Hz) can be attributed to a space-charge build-
our results based on opto-impedance study corroborate the up across the dielectric medium by considering a parallel
presence of a diffusion regulated electron transfer process. resistor-capacitance combination [20]. Whilst the second
In the case of insulating Y2O3:Eu3+ nano-crystals, the semi-circular region marked II in the middle frequency
weak quantum confinement will facilitate the direct feeding (60–10 Hz) would correspond to relaxation of surface
of the CTS level which in turn can feed f–f luminescence states. Different models have been proposed for explaining
transitions of Eu3+ in the same fashion as its bulk counter- contribution from surface state(s) on the impedance spec-
part resulting in the absence of quantum blinking. This tra for several semi-conductor electrode systems [21–23].
insulating system having a larger band-gap will have little Of these, the most viable equivalent circuit fixed for this
scope for the ruptured CSS to relax radiatively via f–f level contribution is a resistor (R) and capacitor (C) components
and instead the exciton will follow an Auger process. The in parallel combination [23]. The third region (region
constant intensity level observed in this case can be marked III) comprising a linear portion having profound
explained in terms of direct photo-luminescence. dependence on excitation energy on the slope would corre-
spond to an electron/charge transfer process between core
3.2. Photo-chemical processes and surface state relaxation and surface state(s) in the vicinity of conduction band-edge
and subsequent relaxation through diffusion of charge car-
Impedance spectroscopy is an excellent tool for studying riers/defects. It should be noted that the photo-relaxation
charge exchange between core state(s) (conduction/valence kinetics of these processes differ in time-scale by several
band) and charges trapped at the surface state(s) [17]. orders.
Relaxation of surface states (SS) with the combination of Also the magnitude of charge carriers, (i.e. the total
electric-field resonance and optical excitation generating charge) associated with surface states as reflected in the
pairs of neutral and charged quantum-dots [18] and their opto-impedance data (having direct impact on the blinking
role in augmenting various photo-chemical properties such rate) can be obtained from the relation.
as luminescence, and solar cells have been well demon- Z V1
strated [19]. In semiconductor nano-crystals, there exists
eo N SS ¼ QSS ¼ C SS dV ð2Þ
two kind(s) of surface states viz., neutral SS (having no V0
net-charge) and charged SS+ (with charge carriers
released/captured by the intrinsic surface states). From Table 1 we have that optical excitation causes sub-
Total surface states density is given by the sum of the stantial increase in both RSS and CSS values corresponding
concentrations of these two states to the resistance and capacitance values of surface-states
suggesting significant increase in the photo-generated carri-
N SS ¼ ½SS þ ½SSþ ¼ nSS þ pSS ð1Þ
ers. This may stem from photo-induced relaxation of sur-
In a photo-irradiated quantum ensemble, there is consider- face sates, in particular SS+ (charged surface states). As a
able probability of finding both neutral and charged sur- consequence we can observe a twofold increase in capaci-
face states. tance, hence the charge associated QSS. Based on these re-
Upon UV shining on the YOS sample, the Nyquist type sults we can surmise a twofold increase in the total surface
complex impedance plot comprising three parts (marked I, state density NSS implying a pronounced increase in the
160 A. Nakkiran et al. / Chemical Physics Letters 436 (2007) 155–161
a b
Fig. 5. Nyquist impedance fitted plot for (a) Y2O2S:Eu3+ and (b) Y2O3:Eu3+ nano-crystals under dark and UV excitation conditions and the
corresponding electrical-impedance equivalent circuits. Insets: expanded portion of Warburg impedance pattern with variation in slope (m) upon UV
shining and absence of any variation in the respective cases.
Table 1
Fitted parameters corresponding to semi-circular portion of the impedance spectrum following photo-chemically induced space-charge region and surface
states relaxation and photo-chemically induced diffusion process of surface state relaxation
System Rsc (X) Csc (l) F RSS (X) CSS (lF) WsR · 106 (X cm2) WsT (s) WsP
NYOS (254 nm) 2.93E6 1.76E11 313700 1.58E8 2.48 23.06 0.4774
NYOS (no UV) 1.61E6 2.07E-11 285350 7.44E-9 – – –
Different Warburg parameters WsR (X), WsT (s), WsP indicate bulk (ohmic) resistance, Warburg resistance, Warburg time constant and order of the
process, respectively.
concentration of charge separated states (CSS). It is rea- impedance spectrum, a straight-line spectrum having a
sonable to expect that this significant increase will set-in slope m = 1 would indicate a diffusion mediated process
an electron transfer process between charged and neutral referred as Warburg impedance [24]. The value of slope
quantum-dots so as to achieve equalization of charge of the linear portion approaching unity indicates a clear
carriers. diffusion process attributed to the transport of charge car-
In the opto-impedance behavior of oxysulfide quantum- riers (electrons in the present case) through a diffusion pro-
ensembles investigated, the most important point that mer- cess. This can be represented by the equation
its a detailed explanation would be the variation in slope of
Z ¼ R tanhð½I T w ^ P Þ=ðI T wÞ ^ P : ð3Þ
the straight-line observed in the region III at the low fre-
quency side (0.1–10 Hz) indicated by m in Fig. 5a-inset From Table 1, it turns out that for the YOS nano-system
which can be fixed to motion of charge carriers. In an upon UV shining there is a substantial change in various
A. Nakkiran et al. / Chemical Physics Letters 436 (2007) 155–161 161
Warburg impedance parameters. Although there is only a processes of nearly comparable time-scales in a good agree-
moderate increase in Warburg Resistance (WsR), there is ment with the DCET model.
about twofold increase in the values of Warburg parame-
ters corresponding to kinetic processes such as WsT, Acknowledgements
WsP. It is pertinent to note that Warburg impedance is
known to have frequency dependence [25]. The emergence The authors sincere thank Dr. D.C. Trivedi and Profes-
of a clear Warburg regime with slope m = 1 upon optical sor. A.K. Shukla for the encouragement and support.
pumping on YOS quantum-ensembles is a convincing
proof for the verification of DCET model. It should also References
be mentioned that the DCET model has been proposed
on the basis of photo-induced electron transfer through a [1] A.N. Kapanidisa, S. Weiss, J. Chem. Phys. 117 (2002) 10953.
diffusion process from neutral to charged quantum-dots [2] B. Lounis, W.E. Moerner, Nature 407 (2000) 491.
[3] Y. Chang, J.M. Cargue, P.T. Snee, M.G. Bewandi, Appl. Phys. Lett.
[14]. Furthermore, we observe that there is a reasonable 85 (2004) 1460.
comparison between the time-scale corresponding to diffu- [4] F.D. Stefani, X. Zhong, W. Knoll, M. Han, M. Kreiter, New J. Phy. 7
sion of charge carriers reflected in the opto-impedance data (2005) 197.
and the blinking kinetics. [5] L. Jacak, P. Hawrylak, A. Wojs, Quantum Dots, Springer, Berlin,
Turning to explain the absence of luminescence blinking 1997.
[6] D.E. Gómez, J. van Embden, J. Jasieniak, T.A. Smith, P. Mulvaney,
in nano-YOX system: the respective impedance spectra Small 2 (2006) 204.
show relatively simple impedance spectrum comprising a [7] M. Nirmal, B.O. Dabbousi, M.G. Bawendi, J.J. Macklin, J.K.
semi-circle and a line without any spectacular change in Trautman, T.D. Harris, L.E. Brus, Nature 383 (1996) 802.
the opto-impedance pattern (Fig. 5b). However, there is a [8] I. Chung, M.G. Bawendi, Phys. Rev. B 70 (2004) 165304.
[9] J. Dhanaraj, R. Jagannathan, D.C. Trivedi, J. Mater. Chem. 13
convincing change in the impedance value (region I) repre-
(2003) 1778.
senting the space-charge build-up of the dielectric medium [10] J. Dhanaraj, M. Geethalakshmi, R. Jagannathan, T.R.N. Kutty,
of the material. This indicates the absence of any contribu- Chem. Phys. Lett. 387 (2004) 23.
tion from surface states (corresponding to the region II of [11] J. Thirumalai, R. Jagannathan, D.C. Trivedi, J. Lumin, in press.
YOS system). More important, there is not any obvious [12] P. Kalyani, R. Jagannathan, S. Gopukumar, C.H. Lu, J. Power
photo-induced change in the slope for the linear-portion Sources 109 (2002) 301.
[13] P. Kalyani et al., J. Phys. D-Appl. Phys. 38 (2005) 990.
(region III) consistent with the absence of any lumines- [14] J. Tang, R.A. Marcus, J. Chem. Phys. 123 (2005) 054704.
cence blinking. This lends support to our hypothesis that [15] X.Y. Wang, W.Q. Ma, J.Y. Zhang, G.J. Salamo, M. Xiao, C.K. Shih,
the regions II and III of the opto-impedance spectrum Nano Lett. 5 (2005) 1873.
for YOS nano-ensembles respectively represent surface [16] R.A. Marcus, N. Sutin, Biochim. Biophys. Acta 265 (1985) 881.
charges and their relaxation kinetics in the same time scale [17] J. Bisquert, Arie Zaban, P. Salvador, J. Phys. Chem. B 106 (2002)
8774.
as does the luminescence blinking. [18] E.A. Stinaff et al., Science 311 (2006) 636.
[19] S. Licht Encyclopedia of Electrochemistry, vol. 6, Wiley-VCH, 2002.
[20] K. Schwarzburg, F. Wiling, J. Phys. Chem. B 107 (2003) 3552.
4. Conclusions [21] A. Bewick, K. Kunimatsu, J.W. Russell, J. Electroanal. Chem. 276
(1987) 175.
The ON–OFF luminescence blinking observed in YOS [22] K. Chandrasekaran, J.Ó.M. Bockris, Electrochim. Acta 32 (1987)
nano-crystals follows a diffusion regulated electron transfer 1393.
[23] John Ó.M. Bockris, Shahed U.M. Khan, Surface Electrochemistry: A
process between the neutral and charge separated quan-
Molecular Level Approach, Plenum Press, New York, 1993.
tum-dots. This is revealed through the occurrence of fluo- [24] P. Jasinski, V. Petrovsky, T. Suzuki, H.U. Anderson, J. Electrochem.
rescence blinking and the Warburg opto-impedance Soc. 152 (4) (2005) J27.
profile corresponding to photo-induced electron diffusion [25] A. Compte, Phys. Rev. E 53 (1996) 4191.