Home
Search
Collections
Journals
About
Contact us
My IOPscience
Caution for monitoring the surface modification of dually emitted ZnSe quantum dots by timeresolved photoluminescence
This content has been downloaded from IOPscience. Please scroll down to see the full text. 2015 Nanotechnology 26 125703 (http://iopscience.iop.org/0957-4484/26/12/125703) View the table of contents for this issue, or go to the journal homepage for more
Download details: IP Address: 142.132.1.147 This content was downloaded on 11/08/2015 at 03:09
Please note that terms and conditions apply.
Nanotechnology Nanotechnology 26 (2015) 125703 (6pp)
doi:10.1088/0957-4484/26/12/125703
Caution for monitoring the surface modification of dually emitted ZnSe quantum dots by time-resolved photoluminescence Chunlei Wang, Zhiyang Hu, Shuhong Xu, Shujie Zhou, Zhuyuan Wang and Yiping Cui Advanced Photonics Center, School of Electronic Science and Engineering, Southeast University, Nanjing 210096, People’s Republic of China E-mail: [email protected] Received 12 December 2014, revised 24 January 2015 Accepted for publication 29 January 2015 Published 4 March 2015 Abstract
This work wants to give a caution for monitoring the surface modification of dually emitted ZnSe quantum dots (QDs) by using time-resolved photoluminescence (PL). Aqueous ZnSe QDs have two emission bands: namely ZnSe band gap emission centered at 395 nm and ZnSe trap emission centered at 470 nm. By fitting the measured PL spectra by two peaks, serious overlapping of two emission bands can be found in the range of 360–430 nm. As a result, the measured PL lifetimes at 395 nm (the peak position of ZnSe band gap emission) is just an apparent value, composing of both ZnSe band emission (contribution proportion about 80%) and ZnSe trap emission (contribution proportion about 20%). Due to the much smaller PL lifetime of ZnSe band gap emission (less than 20 ns) than that of ZnSe trap emission (about 50–70 ns), the elevated contribution proportion of ZnSe band gap emission at improved QD surface modification will lead to the decreased average PL lifetime at 395 nm. This result is completely opposite to the traditional result where improved QD surface modification leads to increased PL lifetimes on the basis of single emitted QDs. Hence, when time-resolved PL is used for monitoring the surface modification of dually emitted QDs, the emission bands overlapping should be taken into consideration with caution. Keywords: surface modification, ZnSe QDs, PL spectra (Some figures may appear in colour only in the online journal) 1. Introduction
used for monitoring the surface modification of QDs based on the monotone relationship between PL lifetimes, PL intensity and QD surface modification [4, 5]. Unlike elemental analysis measurements which require purified QD solution or QD powders, time-resolved PL can in situ monitor QD surface modification in solution. Therefore, purification-induced alteration of QD surface modification can be avoided. This advantage also makes time-resolved PL available for monitoring the dynamics of QDs surface modification at different environment conditions [6]. Dually emitted QDs, which indicate two emission bands from one single QD, are a new type of luminescent QD. Just by tuning the PL intensity ratio of the two emission peaks, the emission color of QDs can be tuned [7]. In this way, dually
Highly luminescent quantum dots (QDs) have received great attention due to their broad applications in bio-imaging, solar cells, LED, and so on [1–3]. The large surface-to-volume ratio and the disrupted crystal lattice on the surface lead to the presence of defects on the QD surface. For CdSe and CdTe QDs, which has been the workhorse of the luminescent QD family in the past two decades, surface defects mainly act as non-irradiative centers. As a result, improved QD surface modification leads to decreased non-irradiative recombination and hence increased photoluminescence (PL) lifetime and increased PL intensity of QDs [4, 5]. Time-resolved PL, which can measure the PL lifetime of QDs, is now widely 0957-4484/15/125703+06$33.00
1
© 2015 IOP Publishing Ltd Printed in the UK
Nanotechnology 26 (2015) 125703
C Wang et al
emitted doped ZnSe QDs with white emission were prepared and used for white light LED [8]. In addition, dually emitted QDs also provide the opportunity for applications in ion sensing or temperature sensing via the ratiometric fluorescence measurement [9, 10]. In comparison to traditional QD sensing based on the fluorescence intensity measurement, the ratiometric fluorescence measurement has less instrumental drift, high sensitivity, and better accuracy. In this work, we investigated the surface modification of dually emitted ZnSe QDs by using time-resolved PL. Unlike the monotone increasing between PL intensity and PL lifetime for single emitted CdTe or CdSe QDs, we observed an ‘abnormal’ phenomenon in the case of dually emitted ZnSe QDs: namely the increased ZnSe band gap emission corresponding to the decreased PL lifetimes. The possible reason for this ‘abnormal’ observation was investigated in detail.
To obtain CdTe QDs with different surface modification, a centrifugation method was used [12]. Freshly prepared CdTe QDs were precipitated from solution by centrifugation after the addition of an equal volume of isopropanol, and followed by re-dispersing into water. After repeating the precipitation and re-dispersing manipulation again, as-prepared purified CdTe QDs show worse surface modification than QDs before centrifugation. 2.5. Characterization
UV–vis absorption spectra (UV) were recorded with a Shimadzu 3600 UV–vis near-infrared spectrophotometer. PL measurements were performed with a Shimadzu RF-5301 PC spectrofluorimeter. The excitation wavelength was 350 nm for ZnSe QDs and 400 nm for CdTe QDs. Time-resolved PL spectra were measured by an Edinburgh FLS 920 spectrofluorimeter. All optical measurements were performed at room temperature under ambient conditions. Transmission electron microscopy (TEM) and high-resolution TEM were recorded by a Tecnai F20 electron microscope with an acceleration voltage of 200 kV. X-ray photoelectron spectroscopy (XPS) was investigated using a PHI550 spectrometer with an Mg Kα(1253.6 eV). QDs powder was used in XPS measurements. QD powder was obtained by precipitating QD from solution with isopropanol and followed by drying in vacuum.
2. Experimental section 2.1. Materials
All materials used in this work were of analytical reagents. Zn (NO3)2, CdCl2, and isopropanol were purchased from Beijing Chemical Factory. NaBH4 was purchased from Guangdong Chemical Reagent Engineering Technological Research and Development Center. Se powder, Te powder, and 1-thioglycerol (TG) were purchased from Aldrich. NaHSe or NaHTe solution was prepared by using NaBH4 with Se or Te in aqueous solution [11].
3. Results and discussion CdTe (or ZnSe) QDs were prepared by a popular aqueous method. In brief, the reaction of NaHTe (or NaHSe) with Cd2+ (or Zn2+) in the presence of 1-thioglycerol (TG) ligand gives rise to the nucleation of CdTe (or ZnSe). After refluxing, luminescent CdTe (ZnSe) QDs were prepared. After post-synthesis treatments, QDs with different surface modification were obtained and used for time-resolved PL measurements.
2.2. Preparation of ZnSe QDs
Freshly prepared NaHSe solution was injected into a pH 12.0 aqueous solution under a N2 atmosphere, followed by heating the solution to 100 °C. Then the mixture of Zn(NO3)2 and TG at pH 12.0 was injected into the aforementioned solution under stirring. Further refluxing for 6 h, freshly prepared ZnSe QDs were obtained. The total concentration of Se in solution is 8.1 × 10−4 mol L−1, and the molar ratio of Se/Zn/TG is 1.0/ 4.0/8.0.
3.1. Fitting methods for time-resolved PL
The measured time-resolved PL required to be fitted by exponential equations [4, 5]:
2.3. Preparation of CdTe QDs
The mixture of CdCl2 and TG was adjusted to pH 12.5 with saturated NaOH solution. After aerating with N2 for 30 min, NaHTe solution was injected into the mixture under stirring. Further refluxing for 6 h, CdTe QDs were prepared. The total concentration of Te in solution is 6.5 × 10−3 mol L−1, and the molar ratio of CdCl2/TG/NaHTe is 1.0/2.4/0.5.
I=
∑Ai e−t /τ , i
(1)
where I, A, and τ are the normalized PL intensity, the amplitude, and the PL lifetime, respectively. The subscripts i indicate the component numbers of the exponential equation. In this work, just by using the single exponential equation and the bi-exponential equation, the goodness of fitting (symbolized by R2) can reach to 0.99, suggesting excellent matching for the fitted results to the measured PL decay curves. For the single exponential equation, i is 1. For the bi-exponential equation, i is 1 and 2. For single exponential equation fitting, there is only one fitted PL lifetime. For bi-exponential fitting, there are two fitted PL lifetimes. Thus, the average PL lifetime (τ ) is used
2.4. Post-synthesis treatments of QDs
To obtain QDs with the same size but different surface modification, post-synthesis treatments are used. Freshly prepared ZnSe QDs were irradiated by an UV lamp in the open air. The irradiation wavelength is 365 nm, and the irradiation time is 2 h. After UV irradiation, the emission of ZnSe QDs increased. 2
Nanotechnology 26 (2015) 125703
C Wang et al
Figure 1. UV–vis absorption spectra, PL spectra (a), and time-resolved PL spectra (b) for CdTe QDs before centrifugation (red dash or red cycle) and after centrifugation (black solid or square).
by considering the fractional contribution (f) of each component [13]: fi =
Ai τi , ∑ Ai τi
τ = fi τi .
aqueous ZnSe QDs have two emission bands [15]: namely ZnSe band gap emission around 395 nm (peak 1) and ZnSe trap emission around 470 nm (peak 2). By post-synthesis treatment of UV irradiation, ZnSe band gap emission at 395 nm increases dramatically, whereas ZnSe trap emission at 470 nm increases a little. The increased PL intensity of ZnSe QDs during UV irradiation implies the improved QD surface modification. In TEM images of figure 3, the sizes of ZnSe QDs increase from 4.6 nm before UV irradiation to 4.7 nm after UV irradiation. The slight size alteration means no serious photo-degradation or QD growth after a short time UV irradiation (2 h) in the current conditions. The most possible alteration during UV irradiation is the changed QD surface modification. In XPS results in figure 4, an obvious increased content of S and Zn can be seen after UV irradiation. Since QD surface is capped by complexes of Zn and thiol ligand [4, 5], the increased S and Zn content in XPS confirms the improved QD surface modification by UV irradiation. Note that irradiation-induced PL enhancement has been previously reported for thiol ligand-capped aqueous ZnSe and CdTe QDs [16, 17]. The irradiation-induced decomposition of thiol ligands followed by the incorporation of S2− on QDs are believed to be able to improve QD surface modification and reduce the non-irradiative defects on the QD surface. According to equations (4) and (5), the decreased non-irradiative recombination rate gives rise to improved emission. For ZnSe trap emission (peak 2) in figure 2(c), the PL decay curve for irradiated QDs is on the top, suggesting much longer PL lifetime. The fitted PL lifetime using the single exponential equation in table 1 is 70.9 ns for irradiated QDs, longer than that of freshly prepared QDs (56.8 ns). Obviously, the relationship between PL lifetime and the PL intensity of ZnSe trap emission (peak 2) is the same as CdTe QDs in figure 1. Interestingly, when ZnSe band gap emission (peak 1) is used, an ‘abnormal’ phenomenon appears. We can see the PL decay curve for irradiated QDs is on the bottom in figure 2(b), implying much shorter PL lifetime of ZnSe band gap emission (peak 1) for irradiated ZnSe QDs with improved surface modification. In table 1, only the bi-exponential equation is
(2) (3)
3.2. Time-resolved PL for single emitted CdTe QDs
Figure 1 shows the absorption, PL intensity, and PL lifetimes of aqueous CdTe QDs. As can be seen, there is no obvious shift in the QD first exciton peak in the absorption spectra after post-synthesis treatments. It indicates no serious alteration in QD size. From the PL spectra, CdTe QDs before centrifugation have a PL QY of 12.9%. The fitted PL lifetime is 9.2 ns by using single exponential equation fitting. After centrifugation, the PL QY of CdTe QDs decreases to 5.5% with PL lifetime of 7.3 ns. Obviously, experimental results suggest decreased PL intensity and decreased PL lifetime at worsened QD surface modification. Theoretically, the PL lifetime of CdTe band gap emission can be expressed as [14]:
(
τ = 1/ k f +
∑K ) ,
QY = k f × τ ,
(4) (5)
where τ, kf, ΣK, and QY represent the PL lifetime, the irradiative recombination rate, the non-irradiative recombination rate, and the PL QY. Worsened surface modification leads to increased non-irradiative recombination rate. As a result, the PL lifetime and the PL QY decrease. The direct proportion between PL intensity and PL lifetime has been widely used for monitoring QD surface modification in different environment conditions according to the reference work [6]. 3.3. Time-resolved PL for dually emitted ZnSe QDs
Figure 2 shows the absorption, PL intensity, and PL lifetimes of aqueous ZnSe QDs. Unlike single emitted CdTe QDs, 3
Nanotechnology 26 (2015) 125703
C Wang et al
Figure 2. UV–vis absorption spectra, PL spectra for ZnSe QDs (a). The measured (symbols) or the fitted (line) time-resolved PL spectra for ZnSe QDs measured at peak 1 of 395 nm (c) and peak 2 of 470 nm (d). Freshly prepared ZnSe QDs (black solid or cycle) and ZnSe QDs after 2 h irradiation by a 365 nm UV lamp (red dash or square) were used.
Figure 3. TEM, HRTEM (inset), and size distribution for ZnSe QDs before UV irradiation (a) and after UV irradiation (b). The scale bars are
10 nm for TEM images and 5 nm for HRTEM images.
4
Nanotechnology 26 (2015) 125703
C Wang et al
3.5. Mechanism for abnormal relationship between PL intensity and PL lifetime of dually emitted ZnSe QDs
By simultaneous consideration of the fitted PL decay curves (table 1) and the overlapping of emission peaks at 395 nm (table 2), we can deduce the possible reason for the abnormal PL decay curves in figure 2(b). As can be seen in table 1, biexponential fitting is required for the PL decay curves of ZnSe band gap emission. The fitted long PL lifetime component (τ2) at peak 1 (66.2 and 69.1 ns) is close to the measured PL lifetime of ZnSe trap emission (56.8 and 70.9 ns of peak 2). In comparison, the short PL lifetime component (τ1) is only about 10 ns, far less than that of ZnSe trap emission. It should be assigned to the ZnSe band gap emission. This short PL lifetime is also close to the PL lifetimes of band gap emission of CdTe or CdSe QDs as reported in [4, 5]. Note that single emitted CdTe or CdSe QDs may also be fitted by the bi-exponential equation. The short lifetime (2–10 ns) assigns to the intrinsic recombination of populated core state [4, 5]. The long lifetime (15–30 ns) component assigns to QD band gap emission which can be affected by the surface modification [4, 5]. In other words, these two PL lifetime components for CdSe and CdTe are both assigning to the band gap emission of QDs. Obviously, the bi-exponential fitting of the current dually emitted ZnSe QDs is quite different from that of single emitted CdTe or CdSe QDs. The long lifetime (about 60 ns) of ZnSe QDs is much longer than the reported surface-related band gap emission (15–30 ns) of CdTe or CdSe QDs. In comparison to PL lifetime of ZnSe trap emission (56.8 and 70.9 ns of peak 2), the long lifetime (about 60 ns) component is definitely assigning to the trap emission rather than the band gap emission. Comparing table 1 with table 2, the following conclusions can be deduced. (i) The average PL lifetime (τ ) at peak 1 is just a combination of ZnSe band gap emission and ZnSe trap emission with a certain manner. The average PL lifetime cannot represent the real PL lifetime of ZnSe band gap emission. (ii) In the whole emission windows of ZnSe band gap emission from 360 to 430 nm, serious emission bands overlapping can be found in figure 5. Therefore, no matter which wavelength is used, the measured average PL lifetime for ZnSe band gap emission is an apparent value. (iii) UV irradiation promotes the elevated proportion of the real ZnSe band emission in the measured apparent one. Due to much shorter PL lifetime of ZnSe band gap emission, the elevated proportion of the real ZnSe band emission leads to the decreased average PL lifetimes. (iv) By using the real ZnSe band gap PL lifetimes (τ1), a monotone relationship between PL lifetimes and PL intensity can be found. It accords with the observations in CdTe QDs. (iv) Since no emission bands overlap at 470 nm, the measured PL lifetime at 470 nm can represent the real PL lifetimes of ZnSe trap emission. This should be the reason why ZnSe trap emission do not show abnormal PL lifetimes in figure 2(c).
Figure 4. XPS results of Se, S, and Zn elements for freshly prepared ZnSe QDs (black line at the top panel) and ZnSe QDs after 2 h irradiation by a 365 nm UV lamp (red line at the bottom panel).
used, the fitted results can show a high goodness of fitting (0.99). The fitted average PL lifetime is 58.1 ns for the band gap emission (peak 1) of irradiated ZnSe QDs, shorter than that of freshly prepared ZnSe QDs (61.4 ns).
3.4. Influence of emission overlapping on the PL lifetimes of ZnSe
To investigate the possible reason for the abnormal PL lifetime of ZnSe band gap emission in figure 2(b), we fitted the emission peak of dually emitted ZnSe QDs in figure 5. As can be seen, ZnSe trap emission is partly overlapping with ZnSe band gap emission at the peak position of 395 nm. Obviously, only the fitted peak 1 can represent the real emission from ZnSe band gap emission. The measured peak 1 before fitting is just an apparent value, which is composed of both ZnSe band gap emission and ZnSe trap emission. The possible reason for the overlapping of ZnSe band gap emission and ZnSe trap emission is (i) the presence of size distribution in a QD ensemble (usually around 15% for aqueous QDs [11]), and (ii) the broad wavelength of trap emission from multiple trap energy levels. To quantitative evaluation the overlapping of two emission peaks, we defined the contribution proportion as the ratio between the fitted peak intensity and the measured peak intensity at the same peak position. For ZnSe trap emission at 470 nm, the contribution proportion of fitted peak 2 is 100% in table 2. It means no overlapping at 470 nm. For ZnSe band gap emission at 395 nm, the real ZnSe band gap emission (namely fitted peak 1) only contributes a proportion of 75.4% in the apparent ZnSe band gap emission (namely the measured peak 1). It is reasonable due to the serious overlapping of the two emission bands at 395 nm according to figure 5. After UV irradiation, this contribution proportion is increased from 75.4% to 81.3%. It suggested UV irradiation is more effective to improve ZnSe band gap emission. This result coincides with PL results in figure 2(a) where UV irradiation leads to dramatically increased ZnSe band gap emission and slightly increased ZnSe trap emission.
4. Conclusion When time-resolved PL is used for monitoring the surface modification of dually emitted ZnSe QDs, the overlapping of 5
Nanotechnology 26 (2015) 125703
C Wang et al
Table 1. The fitting parameters of freshly prepared ZnSe QDs and ZnSe QDs after 2 h irradiation fitted at peak 1 of 395 nm and peak 2 of
470 nm. Sample Fresh (peak 1) Irradiation (peak 1) Fresh (peak 2) Irradiation (peak 2) a
A1
τ1
A2
τ2
τ
R2a
0.37 ± 0.02 0.55 ± 0.03 …. ….
9.2 ± 1.0 12.3 ± 0.9 …. ….
0.56 ± 0.02 0.42 ± 0.03 0.84 ± 0.008 0.81 ± 0.01
66.2 ± 5.3 69.1 ± 8.8 56.8 ± 1.5 70.9 ± 2.7
61.4 58.1 …. ….
0.99 0.99 0.99 0.98
R2 indicates the goodness of fitting.
used. We think the current work provides an in time and significant foundation for correctly monitoring surface modification of dually emitted QDs by using time-resolved PL.
Acknowledgment This work is supported by the National Key Basic Research Program of China (Grant No. 2015CB352002), National Natural Science Foundation of China (Grant Nos. 61475034, 21403034, 61177033), the Fundamental Research Funds for the Central Universities (No. 2242014R30006), the natural science foundation of Jiangsu Province Youth Fund (No. BK20140650, BK20140635).
References
Figure 5. The measured (cycle) and the fitted PL spectra of ZnSe
QDs after 2 h irradiation by a 365 nm UV lamp was used. [1] Du X, Chen Z, Li Z, Hao H, Zeng Q, Dong C and Yang B 2014 Adv. Energy Mater. 4 1400135 [2] Zhou D, Lin M, Chen Z, Sun H, Zhang H, Sun H and Yang B 2011 Chem. Mater. 23 4857–62 [3] Chen Z, Zhang H, Du X, Cheng X, Chen X, Jiang J and Yang B 2013 Energy Environ. Sci. 6 1597–03 [4] Wang X, Qu L, Zhang J, Peng X and Xiao M 2003 Nano Lett. 3 1103–6 [5] Zhao K, Li J, Wang H, Zhuang J and Yang W 2007 J. Phys. Chem. C 111 5618–21 [6] Xu S, Wang C, Zhang H, Wang Z, Yang B and Cui Y 2011 Nanotechnology 22 315703 [7] Battaglia D, Blackman B and Peng X 2005 J. Am. Chem. Soc. 127 10889–97 [8] Wang C, Xu S, Wang Y, Wang Z and Cui Y 2014 J. Mater. Chem. C 2 660–66 [9] McLaurin E J, Vlaskin V A and Gamelin D R 2011 J. Am. Chem. Soc. 133 14978–80 [10] Wang C, Zhou S, Xu S, Wang Z and Cui Y 2014 Nanotechnology 25 375602 [11] Zhang H, Zhou Z, Yang B and Gao M 2003 J. Phys. Chem. B 107 8–13 [12] Tang Z, Podsiadlo P, Shim B S, Lee J and Kotov N A 2008 Adv. Funct. Mater. 18 3801–08 [13] Li R, Xu S, Wang C, Shao H, Xu Q and Cui Y 2010 ChemPhysChem. 11 2582–88 [14] Yang Y, Chen Q, Angerhofer A and Cao Y C 2009 Chem. Eur. J. 15 3186–97 [15] Han J, Zhang H, Tang Y, Liu Y, Yao X and Yang B 2009 J. Phys. Chem. C 113 7503–10 [16] Shavel A, Gaponik N and Eychmüller A 2004 J. Phys. Chem. B 108 5905–08 [17] Bao H, Gong Y, Li Z and Gao M 2004 Chem. Mater. 16 3853–59
Table 2. The peak position, peak intensity, and calculated contribution proportion of the measured or fitted peaks.
Sample Fresh Fresh Fresh Fresh Irradiation Irradiation Irradiation Irradiation
Peak type
Position
Intensity
Proportion
Measured peak 1 Fitted peak 1 Measured peak 2 Fitted peak 2 Measured peak 1 Fitted peak 1 Measured peak 2 Fitted peak 2
395 nm 395 nm 470 nm 470 nm 395 nm 395 nm 470 nm 470 nm
120.2 90.6 114.7 114.7 239.4 194.7 126.9 126.9
…. 75.4% …. 100% …. 81.3% …. 100%
dual emission peaks should be taken into consideration. Due to serious emission overlapping at the position of ZnSe band gap emission, the measured average PL lifetimes are actually an apparent value, composed by both the PL lifetimes of ZnSe band gap emission and ZnSe trap emission. In this case, the improved QD surface modification leads to increased proportion of the real ZnSe band gap emission with much shorter PL lifetimes, and hence decreased average PL lifetimes of QDs. For correctly monitoring surface modification of dually emitted QDs by time-resolved PL, the PL peak without emission bands overlapping is recommended. If the PL peak with serious emission bands overlapping is used, the fitted real PL lifetime at the corresponding emission band should be 6
Search
Collections
Journals
About
Contact us
My IOPscience
Caution for monitoring the surface modification of dually emitted ZnSe quantum dots by timeresolved photoluminescence
This content has been downloaded from IOPscience. Please scroll down to see the full text. 2015 Nanotechnology 26 125703 (http://iopscience.iop.org/0957-4484/26/12/125703) View the table of contents for this issue, or go to the journal homepage for more
Download details: IP Address: 142.132.1.147 This content was downloaded on 11/08/2015 at 03:09
Please note that terms and conditions apply.
Nanotechnology Nanotechnology 26 (2015) 125703 (6pp)
doi:10.1088/0957-4484/26/12/125703
Caution for monitoring the surface modification of dually emitted ZnSe quantum dots by time-resolved photoluminescence Chunlei Wang, Zhiyang Hu, Shuhong Xu, Shujie Zhou, Zhuyuan Wang and Yiping Cui Advanced Photonics Center, School of Electronic Science and Engineering, Southeast University, Nanjing 210096, People’s Republic of China E-mail: [email protected] Received 12 December 2014, revised 24 January 2015 Accepted for publication 29 January 2015 Published 4 March 2015 Abstract
This work wants to give a caution for monitoring the surface modification of dually emitted ZnSe quantum dots (QDs) by using time-resolved photoluminescence (PL). Aqueous ZnSe QDs have two emission bands: namely ZnSe band gap emission centered at 395 nm and ZnSe trap emission centered at 470 nm. By fitting the measured PL spectra by two peaks, serious overlapping of two emission bands can be found in the range of 360–430 nm. As a result, the measured PL lifetimes at 395 nm (the peak position of ZnSe band gap emission) is just an apparent value, composing of both ZnSe band emission (contribution proportion about 80%) and ZnSe trap emission (contribution proportion about 20%). Due to the much smaller PL lifetime of ZnSe band gap emission (less than 20 ns) than that of ZnSe trap emission (about 50–70 ns), the elevated contribution proportion of ZnSe band gap emission at improved QD surface modification will lead to the decreased average PL lifetime at 395 nm. This result is completely opposite to the traditional result where improved QD surface modification leads to increased PL lifetimes on the basis of single emitted QDs. Hence, when time-resolved PL is used for monitoring the surface modification of dually emitted QDs, the emission bands overlapping should be taken into consideration with caution. Keywords: surface modification, ZnSe QDs, PL spectra (Some figures may appear in colour only in the online journal) 1. Introduction
used for monitoring the surface modification of QDs based on the monotone relationship between PL lifetimes, PL intensity and QD surface modification [4, 5]. Unlike elemental analysis measurements which require purified QD solution or QD powders, time-resolved PL can in situ monitor QD surface modification in solution. Therefore, purification-induced alteration of QD surface modification can be avoided. This advantage also makes time-resolved PL available for monitoring the dynamics of QDs surface modification at different environment conditions [6]. Dually emitted QDs, which indicate two emission bands from one single QD, are a new type of luminescent QD. Just by tuning the PL intensity ratio of the two emission peaks, the emission color of QDs can be tuned [7]. In this way, dually
Highly luminescent quantum dots (QDs) have received great attention due to their broad applications in bio-imaging, solar cells, LED, and so on [1–3]. The large surface-to-volume ratio and the disrupted crystal lattice on the surface lead to the presence of defects on the QD surface. For CdSe and CdTe QDs, which has been the workhorse of the luminescent QD family in the past two decades, surface defects mainly act as non-irradiative centers. As a result, improved QD surface modification leads to decreased non-irradiative recombination and hence increased photoluminescence (PL) lifetime and increased PL intensity of QDs [4, 5]. Time-resolved PL, which can measure the PL lifetime of QDs, is now widely 0957-4484/15/125703+06$33.00
1
© 2015 IOP Publishing Ltd Printed in the UK
Nanotechnology 26 (2015) 125703
C Wang et al
emitted doped ZnSe QDs with white emission were prepared and used for white light LED [8]. In addition, dually emitted QDs also provide the opportunity for applications in ion sensing or temperature sensing via the ratiometric fluorescence measurement [9, 10]. In comparison to traditional QD sensing based on the fluorescence intensity measurement, the ratiometric fluorescence measurement has less instrumental drift, high sensitivity, and better accuracy. In this work, we investigated the surface modification of dually emitted ZnSe QDs by using time-resolved PL. Unlike the monotone increasing between PL intensity and PL lifetime for single emitted CdTe or CdSe QDs, we observed an ‘abnormal’ phenomenon in the case of dually emitted ZnSe QDs: namely the increased ZnSe band gap emission corresponding to the decreased PL lifetimes. The possible reason for this ‘abnormal’ observation was investigated in detail.
To obtain CdTe QDs with different surface modification, a centrifugation method was used [12]. Freshly prepared CdTe QDs were precipitated from solution by centrifugation after the addition of an equal volume of isopropanol, and followed by re-dispersing into water. After repeating the precipitation and re-dispersing manipulation again, as-prepared purified CdTe QDs show worse surface modification than QDs before centrifugation. 2.5. Characterization
UV–vis absorption spectra (UV) were recorded with a Shimadzu 3600 UV–vis near-infrared spectrophotometer. PL measurements were performed with a Shimadzu RF-5301 PC spectrofluorimeter. The excitation wavelength was 350 nm for ZnSe QDs and 400 nm for CdTe QDs. Time-resolved PL spectra were measured by an Edinburgh FLS 920 spectrofluorimeter. All optical measurements were performed at room temperature under ambient conditions. Transmission electron microscopy (TEM) and high-resolution TEM were recorded by a Tecnai F20 electron microscope with an acceleration voltage of 200 kV. X-ray photoelectron spectroscopy (XPS) was investigated using a PHI550 spectrometer with an Mg Kα(1253.6 eV). QDs powder was used in XPS measurements. QD powder was obtained by precipitating QD from solution with isopropanol and followed by drying in vacuum.
2. Experimental section 2.1. Materials
All materials used in this work were of analytical reagents. Zn (NO3)2, CdCl2, and isopropanol were purchased from Beijing Chemical Factory. NaBH4 was purchased from Guangdong Chemical Reagent Engineering Technological Research and Development Center. Se powder, Te powder, and 1-thioglycerol (TG) were purchased from Aldrich. NaHSe or NaHTe solution was prepared by using NaBH4 with Se or Te in aqueous solution [11].
3. Results and discussion CdTe (or ZnSe) QDs were prepared by a popular aqueous method. In brief, the reaction of NaHTe (or NaHSe) with Cd2+ (or Zn2+) in the presence of 1-thioglycerol (TG) ligand gives rise to the nucleation of CdTe (or ZnSe). After refluxing, luminescent CdTe (ZnSe) QDs were prepared. After post-synthesis treatments, QDs with different surface modification were obtained and used for time-resolved PL measurements.
2.2. Preparation of ZnSe QDs
Freshly prepared NaHSe solution was injected into a pH 12.0 aqueous solution under a N2 atmosphere, followed by heating the solution to 100 °C. Then the mixture of Zn(NO3)2 and TG at pH 12.0 was injected into the aforementioned solution under stirring. Further refluxing for 6 h, freshly prepared ZnSe QDs were obtained. The total concentration of Se in solution is 8.1 × 10−4 mol L−1, and the molar ratio of Se/Zn/TG is 1.0/ 4.0/8.0.
3.1. Fitting methods for time-resolved PL
The measured time-resolved PL required to be fitted by exponential equations [4, 5]:
2.3. Preparation of CdTe QDs
The mixture of CdCl2 and TG was adjusted to pH 12.5 with saturated NaOH solution. After aerating with N2 for 30 min, NaHTe solution was injected into the mixture under stirring. Further refluxing for 6 h, CdTe QDs were prepared. The total concentration of Te in solution is 6.5 × 10−3 mol L−1, and the molar ratio of CdCl2/TG/NaHTe is 1.0/2.4/0.5.
I=
∑Ai e−t /τ , i
(1)
where I, A, and τ are the normalized PL intensity, the amplitude, and the PL lifetime, respectively. The subscripts i indicate the component numbers of the exponential equation. In this work, just by using the single exponential equation and the bi-exponential equation, the goodness of fitting (symbolized by R2) can reach to 0.99, suggesting excellent matching for the fitted results to the measured PL decay curves. For the single exponential equation, i is 1. For the bi-exponential equation, i is 1 and 2. For single exponential equation fitting, there is only one fitted PL lifetime. For bi-exponential fitting, there are two fitted PL lifetimes. Thus, the average PL lifetime (τ ) is used
2.4. Post-synthesis treatments of QDs
To obtain QDs with the same size but different surface modification, post-synthesis treatments are used. Freshly prepared ZnSe QDs were irradiated by an UV lamp in the open air. The irradiation wavelength is 365 nm, and the irradiation time is 2 h. After UV irradiation, the emission of ZnSe QDs increased. 2
Nanotechnology 26 (2015) 125703
C Wang et al
Figure 1. UV–vis absorption spectra, PL spectra (a), and time-resolved PL spectra (b) for CdTe QDs before centrifugation (red dash or red cycle) and after centrifugation (black solid or square).
by considering the fractional contribution (f) of each component [13]: fi =
Ai τi , ∑ Ai τi
τ = fi τi .
aqueous ZnSe QDs have two emission bands [15]: namely ZnSe band gap emission around 395 nm (peak 1) and ZnSe trap emission around 470 nm (peak 2). By post-synthesis treatment of UV irradiation, ZnSe band gap emission at 395 nm increases dramatically, whereas ZnSe trap emission at 470 nm increases a little. The increased PL intensity of ZnSe QDs during UV irradiation implies the improved QD surface modification. In TEM images of figure 3, the sizes of ZnSe QDs increase from 4.6 nm before UV irradiation to 4.7 nm after UV irradiation. The slight size alteration means no serious photo-degradation or QD growth after a short time UV irradiation (2 h) in the current conditions. The most possible alteration during UV irradiation is the changed QD surface modification. In XPS results in figure 4, an obvious increased content of S and Zn can be seen after UV irradiation. Since QD surface is capped by complexes of Zn and thiol ligand [4, 5], the increased S and Zn content in XPS confirms the improved QD surface modification by UV irradiation. Note that irradiation-induced PL enhancement has been previously reported for thiol ligand-capped aqueous ZnSe and CdTe QDs [16, 17]. The irradiation-induced decomposition of thiol ligands followed by the incorporation of S2− on QDs are believed to be able to improve QD surface modification and reduce the non-irradiative defects on the QD surface. According to equations (4) and (5), the decreased non-irradiative recombination rate gives rise to improved emission. For ZnSe trap emission (peak 2) in figure 2(c), the PL decay curve for irradiated QDs is on the top, suggesting much longer PL lifetime. The fitted PL lifetime using the single exponential equation in table 1 is 70.9 ns for irradiated QDs, longer than that of freshly prepared QDs (56.8 ns). Obviously, the relationship between PL lifetime and the PL intensity of ZnSe trap emission (peak 2) is the same as CdTe QDs in figure 1. Interestingly, when ZnSe band gap emission (peak 1) is used, an ‘abnormal’ phenomenon appears. We can see the PL decay curve for irradiated QDs is on the bottom in figure 2(b), implying much shorter PL lifetime of ZnSe band gap emission (peak 1) for irradiated ZnSe QDs with improved surface modification. In table 1, only the bi-exponential equation is
(2) (3)
3.2. Time-resolved PL for single emitted CdTe QDs
Figure 1 shows the absorption, PL intensity, and PL lifetimes of aqueous CdTe QDs. As can be seen, there is no obvious shift in the QD first exciton peak in the absorption spectra after post-synthesis treatments. It indicates no serious alteration in QD size. From the PL spectra, CdTe QDs before centrifugation have a PL QY of 12.9%. The fitted PL lifetime is 9.2 ns by using single exponential equation fitting. After centrifugation, the PL QY of CdTe QDs decreases to 5.5% with PL lifetime of 7.3 ns. Obviously, experimental results suggest decreased PL intensity and decreased PL lifetime at worsened QD surface modification. Theoretically, the PL lifetime of CdTe band gap emission can be expressed as [14]:
(
τ = 1/ k f +
∑K ) ,
QY = k f × τ ,
(4) (5)
where τ, kf, ΣK, and QY represent the PL lifetime, the irradiative recombination rate, the non-irradiative recombination rate, and the PL QY. Worsened surface modification leads to increased non-irradiative recombination rate. As a result, the PL lifetime and the PL QY decrease. The direct proportion between PL intensity and PL lifetime has been widely used for monitoring QD surface modification in different environment conditions according to the reference work [6]. 3.3. Time-resolved PL for dually emitted ZnSe QDs
Figure 2 shows the absorption, PL intensity, and PL lifetimes of aqueous ZnSe QDs. Unlike single emitted CdTe QDs, 3
Nanotechnology 26 (2015) 125703
C Wang et al
Figure 2. UV–vis absorption spectra, PL spectra for ZnSe QDs (a). The measured (symbols) or the fitted (line) time-resolved PL spectra for ZnSe QDs measured at peak 1 of 395 nm (c) and peak 2 of 470 nm (d). Freshly prepared ZnSe QDs (black solid or cycle) and ZnSe QDs after 2 h irradiation by a 365 nm UV lamp (red dash or square) were used.
Figure 3. TEM, HRTEM (inset), and size distribution for ZnSe QDs before UV irradiation (a) and after UV irradiation (b). The scale bars are
10 nm for TEM images and 5 nm for HRTEM images.
4
Nanotechnology 26 (2015) 125703
C Wang et al
3.5. Mechanism for abnormal relationship between PL intensity and PL lifetime of dually emitted ZnSe QDs
By simultaneous consideration of the fitted PL decay curves (table 1) and the overlapping of emission peaks at 395 nm (table 2), we can deduce the possible reason for the abnormal PL decay curves in figure 2(b). As can be seen in table 1, biexponential fitting is required for the PL decay curves of ZnSe band gap emission. The fitted long PL lifetime component (τ2) at peak 1 (66.2 and 69.1 ns) is close to the measured PL lifetime of ZnSe trap emission (56.8 and 70.9 ns of peak 2). In comparison, the short PL lifetime component (τ1) is only about 10 ns, far less than that of ZnSe trap emission. It should be assigned to the ZnSe band gap emission. This short PL lifetime is also close to the PL lifetimes of band gap emission of CdTe or CdSe QDs as reported in [4, 5]. Note that single emitted CdTe or CdSe QDs may also be fitted by the bi-exponential equation. The short lifetime (2–10 ns) assigns to the intrinsic recombination of populated core state [4, 5]. The long lifetime (15–30 ns) component assigns to QD band gap emission which can be affected by the surface modification [4, 5]. In other words, these two PL lifetime components for CdSe and CdTe are both assigning to the band gap emission of QDs. Obviously, the bi-exponential fitting of the current dually emitted ZnSe QDs is quite different from that of single emitted CdTe or CdSe QDs. The long lifetime (about 60 ns) of ZnSe QDs is much longer than the reported surface-related band gap emission (15–30 ns) of CdTe or CdSe QDs. In comparison to PL lifetime of ZnSe trap emission (56.8 and 70.9 ns of peak 2), the long lifetime (about 60 ns) component is definitely assigning to the trap emission rather than the band gap emission. Comparing table 1 with table 2, the following conclusions can be deduced. (i) The average PL lifetime (τ ) at peak 1 is just a combination of ZnSe band gap emission and ZnSe trap emission with a certain manner. The average PL lifetime cannot represent the real PL lifetime of ZnSe band gap emission. (ii) In the whole emission windows of ZnSe band gap emission from 360 to 430 nm, serious emission bands overlapping can be found in figure 5. Therefore, no matter which wavelength is used, the measured average PL lifetime for ZnSe band gap emission is an apparent value. (iii) UV irradiation promotes the elevated proportion of the real ZnSe band emission in the measured apparent one. Due to much shorter PL lifetime of ZnSe band gap emission, the elevated proportion of the real ZnSe band emission leads to the decreased average PL lifetimes. (iv) By using the real ZnSe band gap PL lifetimes (τ1), a monotone relationship between PL lifetimes and PL intensity can be found. It accords with the observations in CdTe QDs. (iv) Since no emission bands overlap at 470 nm, the measured PL lifetime at 470 nm can represent the real PL lifetimes of ZnSe trap emission. This should be the reason why ZnSe trap emission do not show abnormal PL lifetimes in figure 2(c).
Figure 4. XPS results of Se, S, and Zn elements for freshly prepared ZnSe QDs (black line at the top panel) and ZnSe QDs after 2 h irradiation by a 365 nm UV lamp (red line at the bottom panel).
used, the fitted results can show a high goodness of fitting (0.99). The fitted average PL lifetime is 58.1 ns for the band gap emission (peak 1) of irradiated ZnSe QDs, shorter than that of freshly prepared ZnSe QDs (61.4 ns).
3.4. Influence of emission overlapping on the PL lifetimes of ZnSe
To investigate the possible reason for the abnormal PL lifetime of ZnSe band gap emission in figure 2(b), we fitted the emission peak of dually emitted ZnSe QDs in figure 5. As can be seen, ZnSe trap emission is partly overlapping with ZnSe band gap emission at the peak position of 395 nm. Obviously, only the fitted peak 1 can represent the real emission from ZnSe band gap emission. The measured peak 1 before fitting is just an apparent value, which is composed of both ZnSe band gap emission and ZnSe trap emission. The possible reason for the overlapping of ZnSe band gap emission and ZnSe trap emission is (i) the presence of size distribution in a QD ensemble (usually around 15% for aqueous QDs [11]), and (ii) the broad wavelength of trap emission from multiple trap energy levels. To quantitative evaluation the overlapping of two emission peaks, we defined the contribution proportion as the ratio between the fitted peak intensity and the measured peak intensity at the same peak position. For ZnSe trap emission at 470 nm, the contribution proportion of fitted peak 2 is 100% in table 2. It means no overlapping at 470 nm. For ZnSe band gap emission at 395 nm, the real ZnSe band gap emission (namely fitted peak 1) only contributes a proportion of 75.4% in the apparent ZnSe band gap emission (namely the measured peak 1). It is reasonable due to the serious overlapping of the two emission bands at 395 nm according to figure 5. After UV irradiation, this contribution proportion is increased from 75.4% to 81.3%. It suggested UV irradiation is more effective to improve ZnSe band gap emission. This result coincides with PL results in figure 2(a) where UV irradiation leads to dramatically increased ZnSe band gap emission and slightly increased ZnSe trap emission.
4. Conclusion When time-resolved PL is used for monitoring the surface modification of dually emitted ZnSe QDs, the overlapping of 5
Nanotechnology 26 (2015) 125703
C Wang et al
Table 1. The fitting parameters of freshly prepared ZnSe QDs and ZnSe QDs after 2 h irradiation fitted at peak 1 of 395 nm and peak 2 of
470 nm. Sample Fresh (peak 1) Irradiation (peak 1) Fresh (peak 2) Irradiation (peak 2) a
A1
τ1
A2
τ2
τ
R2a
0.37 ± 0.02 0.55 ± 0.03 …. ….
9.2 ± 1.0 12.3 ± 0.9 …. ….
0.56 ± 0.02 0.42 ± 0.03 0.84 ± 0.008 0.81 ± 0.01
66.2 ± 5.3 69.1 ± 8.8 56.8 ± 1.5 70.9 ± 2.7
61.4 58.1 …. ….
0.99 0.99 0.99 0.98
R2 indicates the goodness of fitting.
used. We think the current work provides an in time and significant foundation for correctly monitoring surface modification of dually emitted QDs by using time-resolved PL.
Acknowledgment This work is supported by the National Key Basic Research Program of China (Grant No. 2015CB352002), National Natural Science Foundation of China (Grant Nos. 61475034, 21403034, 61177033), the Fundamental Research Funds for the Central Universities (No. 2242014R30006), the natural science foundation of Jiangsu Province Youth Fund (No. BK20140650, BK20140635).
References
Figure 5. The measured (cycle) and the fitted PL spectra of ZnSe
QDs after 2 h irradiation by a 365 nm UV lamp was used. [1] Du X, Chen Z, Li Z, Hao H, Zeng Q, Dong C and Yang B 2014 Adv. Energy Mater. 4 1400135 [2] Zhou D, Lin M, Chen Z, Sun H, Zhang H, Sun H and Yang B 2011 Chem. Mater. 23 4857–62 [3] Chen Z, Zhang H, Du X, Cheng X, Chen X, Jiang J and Yang B 2013 Energy Environ. Sci. 6 1597–03 [4] Wang X, Qu L, Zhang J, Peng X and Xiao M 2003 Nano Lett. 3 1103–6 [5] Zhao K, Li J, Wang H, Zhuang J and Yang W 2007 J. Phys. Chem. C 111 5618–21 [6] Xu S, Wang C, Zhang H, Wang Z, Yang B and Cui Y 2011 Nanotechnology 22 315703 [7] Battaglia D, Blackman B and Peng X 2005 J. Am. Chem. Soc. 127 10889–97 [8] Wang C, Xu S, Wang Y, Wang Z and Cui Y 2014 J. Mater. Chem. C 2 660–66 [9] McLaurin E J, Vlaskin V A and Gamelin D R 2011 J. Am. Chem. Soc. 133 14978–80 [10] Wang C, Zhou S, Xu S, Wang Z and Cui Y 2014 Nanotechnology 25 375602 [11] Zhang H, Zhou Z, Yang B and Gao M 2003 J. Phys. Chem. B 107 8–13 [12] Tang Z, Podsiadlo P, Shim B S, Lee J and Kotov N A 2008 Adv. Funct. Mater. 18 3801–08 [13] Li R, Xu S, Wang C, Shao H, Xu Q and Cui Y 2010 ChemPhysChem. 11 2582–88 [14] Yang Y, Chen Q, Angerhofer A and Cao Y C 2009 Chem. Eur. J. 15 3186–97 [15] Han J, Zhang H, Tang Y, Liu Y, Yao X and Yang B 2009 J. Phys. Chem. C 113 7503–10 [16] Shavel A, Gaponik N and Eychmüller A 2004 J. Phys. Chem. B 108 5905–08 [17] Bao H, Gong Y, Li Z and Gao M 2004 Chem. Mater. 16 3853–59
Table 2. The peak position, peak intensity, and calculated contribution proportion of the measured or fitted peaks.
Sample Fresh Fresh Fresh Fresh Irradiation Irradiation Irradiation Irradiation
Peak type
Position
Intensity
Proportion
Measured peak 1 Fitted peak 1 Measured peak 2 Fitted peak 2 Measured peak 1 Fitted peak 1 Measured peak 2 Fitted peak 2
395 nm 395 nm 470 nm 470 nm 395 nm 395 nm 470 nm 470 nm
120.2 90.6 114.7 114.7 239.4 194.7 126.9 126.9
…. 75.4% …. 100% …. 81.3% …. 100%
dual emission peaks should be taken into consideration. Due to serious emission overlapping at the position of ZnSe band gap emission, the measured average PL lifetimes are actually an apparent value, composed by both the PL lifetimes of ZnSe band gap emission and ZnSe trap emission. In this case, the improved QD surface modification leads to increased proportion of the real ZnSe band gap emission with much shorter PL lifetimes, and hence decreased average PL lifetimes of QDs. For correctly monitoring surface modification of dually emitted QDs by time-resolved PL, the PL peak without emission bands overlapping is recommended. If the PL peak with serious emission bands overlapping is used, the fitted real PL lifetime at the corresponding emission band should be 6
