IN
DEGREE PROJECT ENGINEERING PHYSICS, SECOND CYCLE, 30 CREDITS
,
STOCKHOLM SWEDEN 2017
Effect of X-ray Irradiation on the
Blinking of CdSe/ZnS Nanocrystals
Contents
1 Introduction 4
2 Theory 6
2.1 X-Ray Interaction with Matter . . . 6
2.2 Core/ Shell CdSe/ZnS QD’s . . . 7
2.3 Blinking in Nanocrystals . . . 9
2.4 Types of Blinking . . . 10
3 Experimental Techniques and Instrumentation 13 3.1 PL-Setup . . . 13
3.2 X-Ray Irradiation Setup . . . 15
3.3 Nanoparticles Sample Preparation . . . 17
4 Results 19 4.1 PL Spectrum of Ensemble and Single Dot . . . 19
4.2 Time Evolution of Single Dot Blinking Trace . . . 21
4.3 Analysis of X-ray effect on Single Dot Emission . . . 24
4.3.1 Radiation Hardness: Number of Luminescent Dots . . 24
4.3.2 Radiation Hardness: Integrated Mean Intensity . . . . 30 4.3.3 Blinking: Distribution of power exponents Monand Mof f 32
5 Conclusions 38
List of Figures 39
Abstract
Different semiconductor nanocrystals exhibit size dependent properties due to confinement effect. Light emission from these nanocrystals may turn ON and OFF seemingly at random, an effect known as blinking. In this work blinking studies have been done to monitor the effect of X-ray exposure and to investigate the radiation hardness of CdSe/ZnS QD’s. Correct parameters to dilute and spcoat the obtained sample were found to get access to in-dividual single dots. Blinking of these dots was analyzed using Image J and MATLAB plug-in, where ON and OFF-times distribution power exponents Mon and Mof f have been extracted to see the change in emission intermit-tency after a total cumulative dose of ∼1026 Gy (absorbed by SiO2) in steps. It was observed that blinking was quenched and consequently the QD’s went permanently to off state as a result of X-ray exposure.
Acknowledgments
This thesis summarizes my work as a Master Student at the Department of Materials and Nano Physics at KTH Royal Institute of Technology.During my work i have received help and support from many people which I would like to acknowledge. I would like to thank my examiner Ilya Sychugov for his guidance and giving me the opportunity to complete this project work.His advice was always encouraging and thoughtful. I would also like to thank Federico Pevere who guided me through every step of this work. He has always welcomed to discuss and ask questions if I find any difficulty in com-pleting a specific task. His knowledge and way of thinking have helped me to reach the required goals. I am grateful to all my office colleagues Miao Zhang and Karin Th¨orne that were really kind and made office atomsphere pleasant.
Finally I would like to thank my family for supporting and encouraging me to never give up. I would dedicate this work to my late father whose advice made me a better hardworking person in life.
1
Introduction
The study of the materials in the nano-scale has opened up a whole new portal to explore the underlying physics behind different phenomenon. Nan-otechnology refers to the study of the technical details of objects that ranges between 1 and 100 nm in size. Semi-conductor nanocrystals have electronic properties that are intermediate between those of bulk and molecules. Un-like their bulk counterparts the electron energy levels are not continuous but rather discrete. This is due to the quantum confinement effect (i.e. sizes are closer to the exciton Bohr radius). This property allows a shift of the absorp-tion band edges to higher energies by decreasing the size and vice versa[1]. This change of the optical property can be used in broad range of applications which includes as active material in lasers, solar cells[2], optical sensors[3] as well as in biological labeling[4] and high speed signal-processing filters. The shape and size of II-VI group colloidal QD’s, such as CdSe QD’s, can be variably controlled, which makes them suitable for bio-medical imaging by having a wide range of tunable narrow emission[5].These dots have a unique property of randomly switching between bright (ON-state) and going to dark state (OFF-state), which is known as blinking. This unwanted phenomenon lowers the quantum yield of the CdSe QD’s[6].
The underlying mechanism is still an enigma to be solved, although differ-ent studies have showed that surface chemistry plays a vital role in controlling the charging effects, which leads to blinking[7]. In order to suppress the blink-ing behavior a thorough understandblink-ing of the phenomenon is required. CdSe has its size tunable emission spectra, which lie in visible wavelength region (450-650 nm). In applications these particles must be stable under different environments. Bare CdSe QD’s when interact with water (e.g.in bio sens-ing) experience formation of new non-radiative recombination centers[4]. To prevent this phenomenon, they are usually coated with an inorganic entity with a wider bandgap, like ZnS. It is a good shell material due to its wider bandgap, which gives a good exciton confinement, and also features lattice matching, which prevents interface defect formation.
The fluorescence intermittency in single semiconductor shows a power law distribution which is independent of temperature[8]. Studies have also been done to suppress this unwanted behavior.Yuan et. al showed that the blink-ing behavior can be suppressed by couplblink-ing QD’s with silver nanoprisms[9]. Similarly, a high level of blinking suppression was obtained by spin - coating QD’s on TiO2 thin film[10]. These statistical power law distributions reveal
that very long OFF-times are possible, and closing this non-radiative chan-nels can substantially enhance the luminescence intensity[11]. That is why a thorough study of this phenomenon is required, which can then be used in different photonic applications.
The QD’s used in this work were fabricated by the thermal decomposi-tion reacdecomposi-tions by injecting organic precursors complexes into the hot solvent by another research group[2, 12]. These QD’s were used as received to do blinking experiments. This work depicts the study of these luminescence properties (blinking) under the influence of X-ray irradiation. The motiva-tion behind this experiment was to control defect formamotiva-tion in the QD core and shell layers, and to monitor its effects on optical performance. To study the behavior of these dots it was important to dilute the sample further with additional toluene to attain conditions for single dot blinking measurements. The work in this regard consisted of diluting the sample with reasonable amount of solvent, performing luminescence measurements, and later ana-lyzing the blinking sequence by extracting the data in ImageJ and further analyzing ON-/ OFF time distributions by using enhanced MATLAB plugin. In this case two set of samples were prepared for comparison. One sample was given a total cumulative dose of 3078 Gy in steps, while the other non-irradiated sample was taken as the reference measurement. The results show that after a total cumulative dose of 3078 Gy the luminescence intensity of the dots was drastically decreased and blinking behavior was substantially modified already after a few hundreds of Gy. Characteristic blinking distri-bution power exponents Mon and Mof f for both the samples were extracted, which shows that after a large radiation dose the duty cycle was decreased and the dots had shorter ON-time and relatively longer OFF-times. These observations are consistent with the formation of non-radiative defects in the QD core and shell layers. In addition to the reported blinking dependence on the radiation dose, results of this thesis quantify radiation hardness of such fluorophores.
2
Theory
2.1
X-Ray Interaction with Matter
Depending upon the photon energy of the incident X-rays and atomic num-ber (Z) of the material which is being irradiated, two types of effects has been seen mostly responsible for the ionization events. For low electron energies (<100keV) photoelectric effect is dominating, while for higher electron ener-gies ( 1MeV- 4MeV) compton effect is more prevalent as shown in Figure.1. In this work electron energies up to 100keV have been used to study the irradiation effect on CdSe/ZnS QD’s for blinking studies. The interaction of X-rays with the material is a point interaction where secondary particles are generated.
For a quantum dot this ejection of an electron will result in the ionization of the core, leading to non-radiative Auger process. As a result the QD will go to the dark OFF state[13]. In addition, such charging of an atom in the shell/interface of QD’s may lead to the formation of trap sites and surface defect states for excited electrons effecting blinking behavior. If the energy of the X-ray source is high enough it can also introduce some additional lattice displacement damages[14]. Irradiation studies on CdSe/ZnS for high energy gamma-rays suggest a blue shift in the optical absorption spectra which re-sults in poor radiation hardness[15]. The blinking studies done in this work also suggest that at relatively high irradiation doses OFF time distribution power exponent Mof f is decreased which results in longer OFF-times.
Figure 1: Two types of collision when X-rays interact with matter.(a) Pho-toelectric effect (b) Compton scattering[16].
2.2
Core/ Shell CdSe/ZnS QD’s
In colloidal nanocrystals optical excitation creates electron holes pair and the average separation distance between these pairs is known as Bohr exciton radius. Different materials have different Bohr radius depending upon the energy structure, dielectric constant and separation distance of excitons of different materials. When the size of the quantum dot is decreased and its size becomes comparable to the Bohr radius, the energy level becomes discrete. This effect is known as quantum confinement effect. The effective bandgap of QD’s is always greater than its bulk counterpart.
In Type-I quantum dot the bandgap of the core material is smaller than the shell material. The band edges of the core lie within the shell which makes electron and hole to confine in the core. CdSe has a bandgap of 1.74eV and ZnS 3.54eV, which provides a good exciton confinement. This also means that spectrum of the CdSe nanocrystal lie within the visible range as opposed to the bulk counterpart which lies in deep red/Infrared region. The optical properties of CdSe is slightly modified when it is covered with a shell material. Usually the shell material provides a good passivation, promoting high quantum yield. The fluorescence studies demonstrates that as the thickness of ZnS is increased the trap states on the surface are also increased. The shape of such QD’s can be controlled by the fabrication parameters[17]. These shape variations can also be seen in the Figure.2 on the sample which have been exported from other research group to do blinking measurements. Different temperature dependent studies have also been made on Core-Shell CdSe/ZnS QD’s which shows a red shift of the peak wavelength as the temperature is increased which can be associated with band gap shrinkage[18].
Figure 2: STEM image of the Core/Shell CdSe/ZnS QD’s sample.The shell grows anisotropically on the dot preferring growth on certain facets rather than others [provided by University of Amsterdam research group].
In Type-II configuration the valence and conduction band edges are ei-ther higher or lower than the band edges of the shell. As a result, one carrier is confined in the core and the other is confined in the shell. The properties of Type-II are fundamentally different from Type-I due to the confinement of only one charge particle in the core and shell. This quantum dot het-erostructure can be engineered in such a way that a high degree of control
for the charge separation can be used in, e.g. photovoltaic cells[19]. On the other hand, the long radiative life times of these dots can give rise to more non-radiative decay which results in lower QY. Thus CdTe/CdSe can emit energy which are smaller than CdTe and CdSe individual material.This is shown in Figure.3. This band gap engineering can be controlled by changing the thickness of the core as well as the shell material[20].
Figure 3: Two types of Core/Shell QD’s. (a) Type-I CdSe/ZnS the carri-ers are confined in the core (b) Type-II CdTe/CdSe the charge carricarri-ers are segregated in core and shell[21].
2.3
Blinking in Nanocrystals
Fluorescence intermittency or blinking is observed in single colloidal nanocrys-tals which shows a distinct switching between a bright ON state and a dark OFF state. These QD’s are well suitable for blinking studies because they prove to be more stable under different excitation conditions. It has been observed that this phenomenon is dependent on temperature, the chemical composition and the excitation intensity which makes it harder to study[22]. Photoluminescence is the most dominant technique used today to observe the blinking behaviour by time resolved PL measurements.
Different theories have been proposed to understand the blinking be-haviour in these nanocrystals. When an electron is excited by absorbing a photon, it is excited to a higher energy level where it leaves behind a hole and forms an exciton (electron-hole pairs). This photo-excited electron emits
light when it is radiatively relaxed rendering ON-state. Although if multiple electrons are excited in the vicinity it leads to a strong interaction between the excitons which may leave behind a charged quantum dot core by ejecting an electron (hole) to the interface/shell. In this case the electrons are relaxed by generally faster non-radiative recombination rate through Auger process[23]. This results in the reduction of overall light emission and consequently can be reflected in the fluorescence lifetime. This charging blinking model has been extensively used to explain the underlying blinking mechanism. Dif-ferent blinking behaviour have been observed in difDif-ferent nanocrystals. In CdSe/ZnS core–shell nanocrystals this ON-OFF distribution follows a power law and the duty cycle is decreased by increasing the excitation densities whereas mono-exponential statistics has been seen in Si QD’s[24]. However, sometimes it is hard to observe blinking behaviour at relatively higher exci-tation density because the exciton which has been photo excited can easily be further excited via intraband excitation, which prevents the exciton from relaxing to the ground state for radiative recombination giving out a pho-ton, which results in photoluminescence fluctuations known as flickering of the QD’s[25]. The surrounding matrix assembly also plays a vital role in understanding blinking of nanocrystals. Different surface related trap states have been found to effect the blinking behaviour. Time resolved PL measure-ments on CdSe/ZnS Core–Shell nanocrystals showed that these trap states influence the blinking behaviour by creating extra site for the electrons to relax non-radiatively[26]. Also some trap sites are likely to exist between the epitaxial layers of the core-shell nano-structure as well. While it is accepted that trap sites play central role in the blinking process, the exact mechanism of the blinking in nanocrystals is still not completely clear.
2.4
Types of Blinking
Different theories have been proposed in this regard to understand the be-haviour of the electrons and holes with existence of these trap sites. These trap states can be introduced between the core and shell interface where ex-tra site is available for the photo excited electron to reside in. Tunneling probability theory suggests that these photo excited electrons can be ejected to the shell material, resulting in the QD OFF-state.
To understand the behavior of the kinetics of the OFF-state , the charg-ing blinkcharg-ing model gives us a good approximation of the blinkcharg-ing
mecha-nism. A deeper analysis by Galland et al. found out that injecting inten-tionally an electron to the core-shell results in change in the fluorescence lifetimes. When these lifetimes were measured two distinct types of blinking were observed[27].
• The A-Type blinking mechanism is due to the charging and discharg-ing of the core which has been described in the conventional chargdischarg-ing blinking model. The addition of the electron in the core results in the increased number of excitons where radiative decay is suppressed by relatively faster Auger processes. This quenches the emission and hence fluorescence lifetime of the OFF- state is decreased as quantum dots goes to the dark-state[27]. This is shown in Figure.4b where the OFF-state contains an additional electron in the excited state.
• In B-type blinking the fluorescence lifetime remains the same which means that the transitions to the dark state is accompanied by an intermediate trap states. This can be explained by the excitation of the electrons which are well above the lowest conduction band edge or so called hot electrons[23]. These hot electrons can reside in those trap states which acts as non-radiative recombination centers for the Auger process. The B-Type blinking can be suppressed by closing these channels. This can be done by applying an external potential which can populate these trap sites and close the channel[27]. Figure.4c shows hot electrons which have been trapped by the trap surface states. These will relax non-radiatively with the hole after residing at the trap centers giving B- type blinking. We have observed that by applying high energy radiation to CdSe/ZnS QD’s more trap states were created where Auger process becomes dominant and hence shorter ON-time and longer OFF-times.
Figure 4: (a) Excitation of a charge carrier.(b) In an A-Type blinking event, the OFF state contains an additional electron in an excited state.(c) In a B-Type event, "hot electrons" are trapped by surface states immediately following photo excitation, and combine non-radiatively with the remaining hole[23].
3
Experimental Techniques and
Instrumenta-tion
3.1
PL-Setup
All the blinking sequences and the PL measurements were carried out in the PL lab in Kista at KTH. The setup consisted of a Zeiss Observer Z1m inverted optical microscope with an objective of 100x magnification and 0.9 numerical aperture (NA) lens. On the left side port of microscope, Andor iXon 888 EMCCD camera thermally cooled down to -100◦C is installed to capture the PL sequence. The EMCCD is set to cool down at -100◦C by default to reduce the noise associated with dark counts during the image acquisition. The prepared sample was put on X-Y-Z stage of the optical mi-croscope, which could be manually controlled by means of a stage controller. The sample was attached to a microscope glass slide with thermal tape and put inverted onto the stage. The sample was left 1-2 hours on the stage prior doing measurements to reduce the thermal drift and defocusing. For the PL measurements the excitation source used was Omicron PhoxX 405 nm diode laser, which was set to run in continuous wave mode with a power of 20mW and incident angle of 70◦ normal to the surface in a dark field configuration as shown in Figure.5. The fluorescence cube used for the PL measurements was composed of 395-440 nm Zeiss band pass filter. The bandpass filters in general are denoted by their central wavelength and FWHM. Apart from that it consists of 460 nm Zeiss short-pass dichroic mirror and a Semrock 442 nm ultra-steep long-pass edge filter. The long-pass edge filters have a very sharp slope and denoted by their cut on wavelength at 50% of peak trans-mission. For the acquisition of the blinking sequence the EMCCD camera was mounted directly on the left port of the microscope. For measurement of PL spectrum, it was connected to an Andor-SR500 imaging spectrometer with a diffraction grating having spectral resolution of 0.9 nm.
Figure 5: PL schematics showing an excitation laser source light incident on the sample. Emitted light passes through an objective with 100x magnifica-tion and detected by the EMCCD camera.
The whole PL setup which was used for experiment is shown in Figure.6. Different PL time resolved blinking studies have been conducted on various QD’s which showed that CdSe/ZnS QD’s exhibit faster blinking as compared with the Si QD’s[28]. Photo bleaching can be a problem in these type of QD’s as it can affect the blinking behaviour by reducing the core size[29]. To avoid this type of artefact it was important to set the acquisition parameters, which would effectively give the best output with high signal to noise ratio. The acquisition parameters, which was set to capture the blinking sequence as well as the room temperature time resolved PL spectra, will be discussed in detail in Chapter 4.
Figure 6: PL Setup schematics, showing inverted optical microscope with X100 magnification in a light-proof black box with X-Y-Z stage, which is motorized. A spectrometer with a cooled CCD camera are attached for photoluminescence measurements with blue laser diode used for excitation of NCs.
3.2
X-Ray Irradiation Setup
One set of the samples were irradiated by using Hamamatsu L9181-04 X-ray source, having Tungsten as the source target material. This is shown in Figure.7, where the source is kept inside the chamber surrounded by lead bricks to absorb any stray irradiation. The source was warmed up for ap-proximately 20 minutes before operation. The rating of the X-ray source was set at a maximum power of 130 kV and 300 µA in continuous wave mode. This was to ensure that the emission spectrum of the source has a peak at 10 keV with a tail till 100 keV. The calculated absorbed dose rate was found out to be 17 Gy/min. The total cumulative dose of 3072 Gy was given to observe irradiation effect on the photoluminescence properties of CdSe/ZnS QD’s. For this purpose the sample was mounted on a lead brick placed 5 cm away from the focal spot. In addition to this experiment another test was performed at different doses i.e 256, 513, 769, 1026 Gy respectively to observe
the change in blinking parameters after each step. The setup is illustrated in Figure.8.
Figure 7: Hamamatsu L9181-04 X-ray source kept inside a lead-shielded chamber.
Figure 8: X-ray radiation setup showing the source emitting soft X-rays (130kV) on CdSe/ZnS sample taped to the lead brick at a distance of 5cm from the source focal spot[28].
3.3
Nanoparticles Sample Preparation
The stock solution of CdSe/ZnS dots immersed in toluene was imported from another research group and used to do the blinking experiments. This solution was stored in the dark to avoid any photo bleaching. The diameter of these dots as measured with TEM was approximately 5 nm with 35.2% quantum yield. In order to do single dot blinking experiments the dots needed to be diluted with additional toluene. The initial concentration of these colloidal NCs was 2.12E-4 M. The sample, which we received, had already been diluted 500 times with respect to the initial concentration.
This stock solution was further diluted 100 times to have isolated dots by using eq.1
M1V1 = M2V2 (1)
M1 and V1 is the initial concentration and volume. M2 and V2 is the final concentration and volume.
obtained solution was put in the middle of the ultrasonic bath for 30 minutes. Vortex mixer haven’t been used as it vigorously mixes the solution at a high rpm forming aggregates. Si chips were used as a substrate for the prepared colloidal QD’s solution. These Si chips were cleaned by dispersing them in isopropanol for 10 minutes in ultrasonic bath and afterwards in distilled water. Air gun was used to dry these Si chips and blow out any remaining fluid. Afterwards, spin coater was used to smear out these dispersed QD’s on the Si substrate. For this purpose, static dispense method was used (solution is placed on substrate while it is stationary) and spin coater was started as quickly as possible to avoid any evaporation of CdSe-NCs colloidal solutions. Pipette tip was angled at 90◦ to the substrate. The spin coating parameters used was two-step spin profile 3000 rpm for 10 seconds and 7000 rpm for 20 seconds, acceleration of 400 rpm/s and 10 µl volume of solution to avoid aggregation.
For this purpose, two set of samples were made following the same proce-dure. One was named as the reference sample or the control sample, which was kept non-irradiated throughout the experiment. The other sample un-derwent different irradiation doses at specific intervals of time and then com-pared with the reference sample to observe the effect of time.
4
Results
4.1
PL Spectrum of Ensemble and Single Dot
Due to the quantum confinement effect the bandgap of CdSe/ZnS will in-crease by decreasing the size and the discrete energy levels will arise at the band edges for both the conduction and valence bands. This confinement region of the exciton must be smaller than the Bohr radius to have discrete energy levels. For the bulk CdSe the Bohr exciton radius is around 5.4 nm. In our case the prepared QD’s had a diameter of 5 nm which corresponds to the radius of 2.5 nm which is in the strong confinement regime. The quan-tum yield of the sample was 35.2% and FWHM of 32 nm according to the data provided by the research group in the University of Amsterdam, who provided the starting material.
In Figure.9 we can see the absorption and emission optical spectra of the ensemble for the used CdSe/ZnS QD’s. This suggests us that the emission maxima is at 626 nm which corresponds to orange-red colour in the visible light spectrum. The effective bandgap in this case will be 1.97eV which is greater than bulk CdSe counterpart 1.74eV. To obtain a good emission spectrum the absorption wavelength should be in between 350 nm to 450 nm. In this case we are using an excitation source of 405 nm. There are many factors which affect the emission and absorption spectrum in addition to the energy structure of the material itself. Most notably the type of solvent plays an important role in this regard. Few studies have been made on CdSe/ZnS QD’s in different polar and non-polar solvents which showed two emission peak[30]. This shows that if there are more layers of ZnS shell It can lead to an interaction between the solvent and shell. This effect can then consequently reduce the quantum yield.
Figure 9: Optical Spectra for both CdSe core and CdSe/ZnS QD’s. The first exciton peak for CdSe/ZnS is at 613 nm and emission peak at 626 nm [provided by University of Amsterdam research group].
The variation in size and shape of the ensemble can lead to inhomoge-neous broadening of emission spectra and this can be avoided by measuring luminescence spectra for single dot. To have single dots for the blinking experiment it was imperative to do PL measurements first. In this regard excitation source of 405 nm was used in dark-field configuration with power density of ∼4 W/cm2and 20mW laser power. This light passes through 100x Zeiss lens with a dichroic mirror at 405 nm and long pass filter (LPF) at 442 nm. The PL spectrum of single dots is shown in the Figure.10. In this case we can clearly see that FWHM is less than that of the ensemble, i.e. 32 nm, which gives a good approximation that the dots used for blinking are single ones. Also the variation of peak position from dot-to-dot supports this explanation.
Figure 10: PL spectra for different single QD’s. Highlighted lines show FWHM which is less than the ensemble.
4.2
Time Evolution of Single Dot Blinking Trace
The florescence intermittency in single nanocrystal is an intrinsic property, which shows transition from ON-OFF and OFF-ON state. In case of
Cd-Se/ZnS the blinking is measured in time intervals of 0.1 seconds. The two step binary blinking gives another strong evidence that the obtained spec-tra originate from a single nanocrystal. In case of an ensemble the blinking behaviour is averaged over the nano-particles and we don’t observe two dis-tinct levels. The acquisition parameters which have been used to extract the blinking trace is listed below.
• Total Acquisition time for 20 000 frames: 35 minutes • Number of frames: 20 000 frames
• Read-out time: 0.1 seconds • Pre-amplifier gain: x5.2 • EM gain :20
• Pixel read-out rate :5MHz
In order to get the best signal to noise ratio it was necessary to have acquisition time of 0.1 seconds for blinking measurements. Although with lower time one can decrease the probability of the missed blinking event, the trade-off is a poor signal-to-noise (S/N) ratio. The blinking data were extracted by determining the on and off distributions from the time traces of individual blinking dots. The threshold value was set above the back-ground and at a specific intensity to discriminate between the two states. The threshold value for each image is given by the red dotted line. The two level blinking intensity can be clearly seen to decrease with increased X-ray irradiation dose.
Figure 11: Typical intensity time trace for CdSe(ZnS) dot 5 blinking at different X-ray irradiation doses absorbed by SiO2 with the same time scales. The on time is defined as the time interval when the signal is above selected threshold value.
4.3
Analysis of X-ray effect on Single Dot Emission
The analyzed sample was given different doses with the time interval of 3 days between each successive dose session. As stated earlier the X-ray was operated in continuous wave mode. Different areas on the sample were se-lected to choose different single dots. The purpose of selecting these dots on different areas was to average out the effect of irradiation as the X-rays beam striking at the center of the sample is maximum and intensity decreases as it moves towards the edge of the sample. In other words the X-ray intensity distribution is not uniform and can have different effect, which leads slightly varying value of absorbed dose.
4.3.1 Radiation Hardness: Number of Luminescent Dots
One set of the sample was given different irradiation doses and PL image of the same area (520x520 pixels) was taken to see the change in the number of the dots. With increased irradiation dose the QD’s goes more to the off state with complete loss of intensity at irradiation dose higher than 3078 Gy. This is shown in Figure.12. The corresponding measurement made on the reference sample for the same time duration reveals that the number of dots remained almost constant as shown in Figure.13.
Figure 12: Different PL images for the same sample area (520x520 pixels) at different irradiation doses with 10 sec of acquisition time.
Figure 13: Different PL images for the same sample area (520x520 pixels) for control sample at different time intervals with 10 sec of acquisition time. One should note that a continuous laser excitation on the dots for several hours can cause the loss of their fluorescence intensity. So it was important not to have too long acquisition time as intensity of the blinking dots could be quenched by laser and not by X-ray irradiation. This give us a benchmark for the total acquisition time, which was less than 35 minutes for 20,000 frames. This seems to be in the reasonable limit to see the effect of blinking with time as well with increased irradiation dose. In Figure.14 it can be observed that after 3 hours of continuous excitation 70% of the dots were gone to the off-state, while within 35 minutes the change is reasonable.
Figure 14: Quenching of the single CdSe/ZnS QD’s. Upon continuous laser excitation, 70 % of the number of luminous dots are fully quenched with time.
This radiation hardness experiment was performed on the X-ray irradi-ated sample and the same area (520x520 pixels) was taken and analyzed in ImageJ after each successive irradiation step to see the number of lumines-cent dots. It was observed that the absolute number of single dots decreased from 60 to less than 10 after 3078 Gy of absorbed dose. This is shown in Figure.15.
Figure 15: Absolute number of luminescent dots for the X-ray irradiated sample as a function of dose.
Same acquisition parameters were set for to observe the effect in control sample. The absolute number of dots only decreased to 12% of the initial number as shown in Figure.16.
Figure 16: Absolute number of luminescent dots for the control sample (with-out continuous excitation or X-ray irradiation).
The final comparison was made by taking the two individual graph and plotting it with the same corresponding time scale normalizing the number of initial dots to unity The strong effect of X-ray irradiation on QD performance is obvious from Figure.17.
Figure 17: No. of luminescent Dots for irradiated and control sample. The luminescent dots are decreasing with increased irradiation dose.
4.3.2 Radiation Hardness: Integrated Mean Intensity
As an alternative figure of merit for the radiation hardness the integrated mean intensity value was evaluated and analyzed. First, the captured PL im-ages were analyzed using ImageJ. An area of 5x5 pixels was selected around the dot to calculate the integrated intensity which is the sum of each indi-vidual pixel intensities within that area. The background intensity has been subtracted which leaves with only the sum of integrated intensity /area of the particular dot. Integrated mean intensity has been calculated for all the num-ber of luminescent dots in this case. The integrated intensity distribution of these dots has been shown by the error bars in Figure.18 and Figure.19. We can see that the trend for Integrated Mean Intensity of the dots (Figure.18 and Figure.19) looks similar with Figure.17 which means the mechanism of
quenching luminescence is the same as in the previous section. The error bars in the graph shows the intensity distribution within the probed ensemble of luminescent dots.
Figure 19: Mean Integrated Intensity vs time for the control sample. 4.3.3 Blinking: Distribution of power exponents Mon and Mof f The duty cycle is degrading over time with increased dose of X-ray irra-diation, which means that the dots have shorter ON times and longer OFF times.The probability distribution P(t) is given by the on and off events with the length of the respective intervals t and shows the number of observed in-tervals for a certain length.For the case of CdSe/ZnS this distribution follows a pure power law, also typically observed in literature[31]. Power exponents Mon and Mof f are obtained by analyzing individual dots from the on-off time histograms probability distribution. Fitting curves for each of these selected dots in the MATLAB plugin were used. In this representation we plot the on and off time bins vs on and off probability distribution for the respective dots and extract these power exponents. The Integration time of 0.1 sec-onds provides good signal to noise ratio which gives good power law blinking
statistics.
P (t) = At−M on P (t) = At−M of f
Where Mon and Mof f are the blinking power exponents. The best signal-to-noise individual blinking dots were selected to be used for this analysis.
For the control sample average Mon and Mof f remains linearly constant with time. This indicates that the blinking statistics for on and off times doesn’t vary too much for the control sample. This can be seen in Figure.22 and Figure.20.
For the irradiation sample the situation is clearly different. Mon is in-creasing with increased absorbed dose. That implies that the duty cycle is degrading, reducing on-time intervals. At the same time Mof f is decreasing, which means that the dots are staying in the dark off state longer. This is illustrated in Figure.21 and Figure.23. The observed behaviour can be ex-plained by formation of more charge traps as a result of X-ray irradiation. Charged nanocrystals experience efficient non-irradiated Auger emission. We have checked the irradiation sample after 3 weeks and observed that these dots once gone to the off state remained permanently quenched.
At higher irradiation these dots turn darki.e. don’t show any blinking behaviour, so it was impossible to extract any usable data. One can even use lower acquisition time with good S/N ratio to achieve higher resolution but the limitation of the PL equipment in our case doesn’t allow it. Nevertheless this result is consistent with previous studies and gives a good estimation for the radiation hardness and blinking changes of these QD’s.
Figure 21: Blinking distribution power exponent Mof f for the irradiated sample.
Figure 23: Blinking distribution power exponent Monfor the irradiated sam-ple.
5
Conclusions
In this work we have made time resolved photoluminescence measurements on single core shell CdSe/ZnS QD’s at room temperature and observed blinking behaviour. The QD’s sample and relevant parameters were obtained from University of Amsterdam research group. This sample was diluted further with toluene to do single dot measurement. Blinking time evolution of single dot trace shows that the florescence intensity is decreased as more charge states appear with increased irradiation doses. This result in more B-Type blinking where the electron reside in the defect states (hot electrons) before combining non-radiately with the hole i.e transition back to the charged core. This results in faster Auger process and the dots go permanently to the OFF state. Different PL images of the irradiated and control sample was taken at relatively specific time intervals to see the change in the number of the luminescent dots. The results also shows a significantly decrease in the luminescent intensity and the number of blinking dots for the irradiated sample and no significant change was observed in control sample. These dots were almost fully quenched at an absorbed dose of ∼3078 Gy. Mean integrated Intensity distribution of these dots versus absorbed dose shows the same quenching behavior. This result provided us an estimation of the value for the maximum absorbed dose which was later used for blinking experiment. The blinking behavior of the dot was analyzed using enhanced MATLAB plug-in and distribution power exponent Mon and Mof f was extracted for the best dots. The results give us a power law behaviour for both ON-OFF and OFF-ON time intervals. The duty cycle was degrading over time and after a cumulative dose of ∼1026 Gy , the blinking was quenched and the dots remained permanently OFF. At any higher dose it was not possible to extract blinking trace. Blinking distribution power exponents Mon and Mof f shows that as the irradiation doses are increased, the value of Mof f was decreased which shows longer OFF-times and shorter ON-times. One can also try to do the same experiment with core only CdSe QD’s and compare the effect with passivated shell. Overall this shows us that these dots are susceptible to radiation defects and not as stable as Si QD’s which showed no significant change even after ∼300,000 Gy[28].
List of Figures
1 Two types of collision when X-rays interact with matter.(a) Photoelectric effect (b) Compton scattering[16]. . . 6 2 STEM image of the Core/Shell CdSe/ZnS QD’s sample.The
shell grows anisotropically on the dot preferring growth on certain facets rather than others [provided by University of Amsterdam research group]. . . 8 3 Two types of Core/Shell QD’s. (a) Type-I CdSe/ZnS the
carri-ers are confined in the core (b) Type-II CdTe/CdSe the charge carriers are segregated in core and shell[21]. . . 9 4 (a) Excitation of a charge carrier.(b) In an A-Type blinking
event, the OFF state contains an additional electron in an ex-cited state.(c) In a B-Type event, "hot electrons" are trapped by surface states immediately following photo excitation, and combine non-radiatively with the remaining hole[23]. . . 12 5 PL schematics showing an excitation laser source light incident
on the sample. Emitted light passes through an objective with 100x magnification and detected by the EMCCD camera. . . . 14 6 PL Setup schematics, showing inverted optical microscope with
X100 magnification in a light-proof black box with X-Y-Z stage, which is motorized. A spectrometer with a cooled CCD camera are attached for photoluminescence measurements with blue laser diode used for excitation of NCs. . . 15 7 Hamamatsu L9181-04 X-ray source kept inside a lead-shielded
chamber. . . 16 8 X-ray radiation setup showing the source emitting soft X-rays
(130kV) on CdSe/ZnS sample taped to the lead brick at a distance of 5cm from the source focal spot[28]. . . 17 9 Optical Spectra for both CdSe core and CdSe/ZnS QD’s. The
first exciton peak for CdSe/ZnS is at 613 nm and emission peak at 626 nm [provided by University of Amsterdam research group]. . . 20 10 PL spectra for different single QD’s. Highlighted lines show
11 Typical intensity time trace for CdSe(ZnS) dot 5 blinking at different X-ray irradiation doses absorbed by SiO2 with the same time scales. The on time is defined as the time interval when the signal is above selected threshold value. . . 23 12 Different PL images for the same sample area (520x520 pixels)
at different irradiation doses with 10 sec of acquisition time. . 25 13 Different PL images for the same sample area (520x520 pixels)
for control sample at different time intervals with 10 sec of acquisition time. . . 26 14 Quenching of the single CdSe/ZnS QD’s. Upon continuous
laser excitation, 70 % of the number of luminous dots are fully quenched with time. . . 27 15 Absolute number of luminescent dots for the X-ray irradiated
sample as a function of dose. . . 28 16 Absolute number of luminescent dots for the control sample
(without continuous excitation or X-ray irradiation). . . 29 17 No. of luminescent Dots for irradiated and control sample.
The luminescent dots are decreasing with increased irradiation dose. . . 30 18 Mean Integrated Intensity vs time for the irradiated sample. . 31 19 Mean Integrated Intensity vs time for the control sample. . . 32 20 Blinking distribution power exponent Mof f for the control
sample. . . 34 21 Blinking distribution power exponent Mof f for the irradiated
sample. . . 35 22 Blinking distribution power exponent Monfor the control
sam-ple. . . 36 23 Blinking distribution power exponent Mon for the irradiated
References
[1] J Drenth Moras, B Strandberg, D Suc, and K Wilson. Semiconductor clusters, nanocrystals, and quantum dots. science, 271:933, 1996. [2] Boris D Chernomordik, Ashley R Marshall, Gregory F Pach, Joseph M
Luther, and Matthew C Beard. Quantum dot solar cell fabrication pro-tocols. Chemistry of Materials, 29(1):189–198, 2016.
[3] Jos´e M Costa-Fern´andez, Rosario Pereiro, and Alfredo Sanz-Medel. The use of luminescent quantum dots for optical sensing. TrAC Trends in Analytical Chemistry, 25(3):207–218, 2006.
[4] Shan Jin, Yanxi Hu, Zhanjun Gu, Lei Liu, and Hai-Chen Wu. Applica-tion of quantum dots in biological imaging. Journal of Nanomaterials, 2011:13, 2011.
[5] Xiaogang Peng, Liberato Manna, Weidong Yang, Juanita Wickham, Erik Scher, Andreas Kadavanich, and A Paul Alivisatos. Shape con-trol of cdse nanocrystals. Nature, 404(6773):59–61, 2000.
[6] Siying Wang, Claudia Querner, Thomas Emmons, Marija Drndic, and Catherine H Crouch. Fluorescence blinking statistics from cdse core and core/shell nanorods. The Journal of Physical Chemistry B, 110(46):23221–23227, 2006.
[7] Steven D Quinn, Aidan Rafferty, Emma Dick, Michael J Morten, Fraser J Kettles, Chloe Knox, Mark Murrie, and Steven W Magennis. Surface charge control of quantum dot blinking. The Journal of Physical Chemistry C, 120(34):19487–19491, 2016.
[8] Jeffrey J Peterson and David J Nesbitt. Modified power law behavior in quantum dot blinking: a novel role for biexcitons and auger ionization. Nano letters, 9(1):338–345, 2008.
[9] CT Yuan, Pyng Yu, and Jau Tang. Blinking suppression of colloidal cd-se/zns quantum dots by coupling to silver nanoprisms. Applied Physics Letters, 94(24):243108, 2009.
[10] Hao Xu, Li Li, Otto Manneberg, Aman Russom, Kristinn B Gylfason, Hjalmar Brismar, and Ying Fu. Modulated fluorescence of colloidal
quantum dots embedded in a porous alumina membrane. The Journal of Physical Chemistry B, 117(45):14151–14156, 2013.
[11] Xiaoyong Wang, Jiayu Zhang, Min Xiao, Lianhua Qu, and Xiaogang Peng. Surface-related emission in highly luminescent cdse quantum dots studied by time-decay photoluminescence measurements. In APS Meet-ing Abstracts, 2003.
[12] Olivier Carion, Benoˆıt Mahler, Thomas Pons, and Benoit Dubertret. Synthesis, encapsulation, purification and coupling of single quantum dots in phospholipid micelles for their use in cellular and in vivo imaging. Nature protocols, 2(10):2383–2390, 2007.
[13] AV Gomonnai, Yu M Azhniuk, Mladen Kranjˇcec, AM Solomon, VV Lo-pushansky, and IG Megela. Effect of x-ray irradiation on cds 1- x se x quantum dots optical absorption. Solid state communications, 119(7):447–451, 2001.
[14] Girija Gaur, Dmitry Koktysh, Daniel M Fleetwood, Robert A Reed, Robert A Weller, and Sharon M Weiss. Effects of x-ray and gamma-ray irradiation on the optical properties of quantum dots immobilized in porous silicon. In Proc. SPIE, volume 8725, page 87252D, 2013.
[15] Nathan J Withers, Krishnaprasad Sankar, Brian A Akins, Tosifa A Memon, Tingyi Gu, Jiangjiang Gu, Gennady A Smolyakov, Melisa R Greenberg, Timothy J Boyle, and Marek Osi´nski. Rapid degradation of cd se/ zn s colloidal quantum dots exposed to gamma irradiation. Applied Physics Letters, 93(17):173101, 2008.
[16] http://visielab.uantwerpen.be/. Accessed: 2017-11-30.
[17] S Mathew, Bishwajeet Singh Bhardwaj, Amit D Saran, P Radhakrish-nan, VPN Nampoori, CPG Vallabhan, and Jayesh R Bellare. Effect of zns shell on optical properties of cdse–zns core–shell quantum dots. Optical materials, 39:46–51, 2015.
[18] Cheng Cheng and Haizhen Yan. Bandgap of the core–shell cdse/zns nanocrystal within the temperature range 300–373k. Physica E: Low-dimensional Systems and Nanostructures, 41(5):828–832, 2009.
[19] Marion Dufour, Violette Steinmetz, Eva Izquierdo, Thomas Pons, Nico-las Lequeux, Emmanuel Lhuillier, Laurent Legrand, Maria Chamarro, Thierry Barisien, and Sandrine Ithurria. Engineering bicolor emis-sion in 2d core/crown cdse/cdse1–x te x nanoplatelet heterostruc-tures using band-offset tuning. The Journal of Physical Chemistry C, 121(44):24816–24823, 2017.
[20] Sungjee Kim, Brent Fisher, Hans-J¨urgen Eisler, and Moungi Bawendi. Type-ii quantum dots: Cdte/cdse (core/shell) and cdse/znte (core/shell) heterostructures. Journal of the American Chemical Society, 125(38):11466–11467, 2003.
[21] LA Kosyachenko. Solar cells-new approaches and reviews. Intech, Croa-tia, ISBN, pages 978–953, 2015.
[22] Pavel Frantsuzov, Masaru Kuno, Boldizsar Janko, and Rudolph A Mar-cus. Universal emission intermittency in quantum dots, nanorods and nanowires. Nature Physics, 4(7):519–522, 2008.
[23] Todd D Krauss and Jeffrey J Peterson. Quantum dots: A charge for blinking. Nature materials, 11:14–16, 2012.
[24] Benjamin Bruhn, Fatjon Qejvanaj, Ilya Sychugov, and Jan Linnros. Blinking statistics and excitation-dependent luminescence yield in si and cdse nanocrystals. The Journal of Physical Chemistry C, 118(4):2202– 2208, 2014.
[25] Li Li, Guangjun Tian, Yi Luo, Hjalmar Brismar, and Ying Fu. Blinking, flickering, and correlation in fluorescence of single colloidal cdse quantum dots with different shells under different excitations. The Journal of Physical Chemistry C, 117(9):4844–4851, 2013.
[26] Marcus Jones, Shun S Lo, and Gregory D Scholes. Quantitative mod-eling of the role of surface traps in cdse/cds/zns nanocrystal photolu-minescence decay dynamics. Proceedings of the National Academy of Sciences, 106(9):3011–3016, 2009.
[27] Christophe Galland, Yagnaseni Ghosh, Andrea Steinbr¨uck, Milan Sykora, Jennifer A Hollingsworth, Victor I Klimov, and Han Htoon. Two types of luminescence blinking revealed by spectroelectrochemistry of single quantum dots. Nature, 479(7372):203–207, 2011.
[28] Unaltered blinking in single silicon oxidized nanocrystals when x-ray irradiated. Master thesis, KTH.
[29] Wilfried G. J. H. M. van Sark, Patrick L. T. M. Frederix, Ageeth A. Bol, Hans C. Gerritsen, and Andries Meijerink. Blueing, bleaching, and blinking of single cdse/zns quantum dots. ChemPhysChem, 3(10):871– 879, 2002.
[30] K.H. Ibnaouf, Saradh Prasad, A. Hamdan, M. AlSalhi, A.S. Aldwayyan, M.B. Zaman, and V. Masilamani. Photoluminescence spectra of cd-se/zns quantum dots in solution. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 121(Supplement C):339 – 345, 2014. [31] Jeffrey J. Peterson and David J. Nesbitt. Modified power law behavior in
quantum dot blinking: A novel role for biexcitons and auger ionization. Nano Letters, 9(1):338–345, 2009. PMID: 19072721.
