Wednesday, October 26, 2016

Towards High Quality Semiconducting Nanocrystals and Surface Ligand Exchange for Solar Cell Application

(Taye Zewdu)

I) Introduction
As ´there is plenty of room at the bottom´ prediction, exploiting the very tiny building blocks of unique mechanical, optical and electrical properties for various applications have been a global trend and an active research arena of huge investment for more than a decade. [1] Nanotechnology is a trillion dollar worth market being explored for applications in materials, pharmaceuticals, electronics, chemical manufacturing, aerospace science and others. Of all other hurdles, energy has come the number one global challenge of 14 TW a year as we approached the peak oil (Hubbert´s peak) and the end of the oil age on top of which sustainability and environmental issues urge alternative and renewable power sources to take stage. [2]
Promising enough, the clean solar- energy of the future, is showing enormous breakthroughs a century after the demonstration of the first 1% solar cell. [3] Depending on the materials and configurations, they can be generally classified as inorganic, organic or hybrid under which a list of state-of-the art technologies are systematically grouped as first, second, third and next generation solar cells.

Efficiency, cost and stability determine the niche (competitiveness) of the devices converting light into electricity in the global market. Theoretical predictions showed that semiconducting nanocrystals could be promising in terms of efficiency that significantly exceeds the schockley-casler limit () and the ease of solution processing. Consequently, quantum dots, nanorods, nanowires, tetrapods and nanocrystals of different heterostructures has been incorporated in Schottky, dye-sensitized and bulk heterojunction solar cell configurations (Fig. 1). However, these types of solar cells are not yet efficient enough to show up commercially, best efficiency reported being 6%, 5.4% and 3.2% for architectures shown in the schemes below, respectively. [4] A major limitation is the electron mobility impeded by organic passivating ligands on the nanocrystals, which insulate the charge transport in the interpenetrating network. This report covers the syntheses of high quality CdSe quantum dots and tetrapods and their respective treatments to enhance electron mobility. Later is discussed some aspect of the optical and electrical device characterization with respect to J-V properties, IPCE, CE and TPV and TAS results.

II)      Syntheses and Characterizations

Optimized routes for CdSe quantum dot and tetrapod preparation were studied at different growth times and temperatures. Optical, TEM and XRD characterizations of typical samples were conducted. New synthesis methods for hollow nanoparticles and ODE-Se mediated long tetrapods were tried on the basis of collaboration.

CdSe quantum dots

The well-known rapid and hot injection synthesis was followed in all cases with a modified recipe. Briefly,  1 mM of 99.99% CdO (128.4 mg) was mixed with 6 mM of oleic acid (2 mL) and 20 mL 1-octadecene in a four-necked flask. The mixture was degased for few minutes and heated up with a reflux until the solution turns clear (usually around 240 oC). The Se precursor was prepared by weighing out 118 mg of Se powder in a glove bag and dissolving in 1 mL TOP (90%) under sonication for 5 minutes in a small vial. As soon as the Cd-oleate turns clear, the temperature was adjusted for injection and the argón flow and heat mantle set so as to control the growth temperature which in most cases was few degrees lower than the injection temperature.  In a typical synthesis where 0.34 mL of TOPSe was injected at 250 oC, the particles were grown around 245 oC for two minutes after which the heat mantle was quickly removed and argón flow increased to cool the sulution fast. The cleaning was done with acetone three times at 4000 rpm for 1hr, 30 and 10 minutes. The resulting precipitate was dried and redispersed in chloroform for characterization.
Fig. 2 Synthesis set up

Quantum dots of CdSe were synthesized at 190 oC and 250 oC injection temperatures at different growth times of 20 sec, 1, 2 and 5 minutes. Fig. 2 and 3 show the normalized absorbance and photoluminescence spectra of the as-synthesized CdSe nanocrystals at respective injection temperatures. At lower injection temperatures, the growth was relatively slower providing the smallest quantum dots as predicted in Table 1 by the following empirical relation.
where D is the diameter of the nanocrystal and λ is the first excitonic peak (nm). The quantum dots  show a stock´s shift about 24 nm.
Typical to these syntheses is the existence of numbers of other excitonic peaks which are favorable and an indication of the kind of crystallinity exibited. For example, the second excitonic peak for a zinc blend structure forms a shoulder which is characteristic to quantum dots. This could be evidenced from the XRD data shown in Fig. x later. These several band features which are found to be growth time and temperature dependent.
The TEM images taken with a high contrast electron microscope show a cluster of smaller nanocrystals of 2-3 nm in size.
The photoluminescence efficiency of the quantum dots were also measured by calculating the quantum yield as in equation 2 below. The fluorescence quantum yield (Φ) is the ratio of photons absorbed to photons emitted through fluorescence. In other words the quantum yield gives the probability of the excited state being deactivated by fluorescence rather than by another, non-radiative mechanism such as internal conversion and vibrational relaxation. This has been done following a guide to recording fluorescence quantum yields using the comparative method of Grabolle et al., which involves the use of well characterised standard samples with known ΦS values based on suitable procedures and achievable uncertainities. Rhodamine 6G was used as standard in these studies. The corresponding fluorescence quantum yield values are provided in Table 1. Compared to literature values (0.3-0.85), fluorescence is the determining deactivation mechanism in the nanocrystal.

Fig. 2 Optical characterization of CdSe QDs synthesized at a) 190  oC/170 oC, b) 250 oC/230 oC and C) 300 oC/280 oC injection and crystal growth temperatures

Table 1 Typical characteristics of CdSe coloidal quantum dots in 2 minute growth time
Temperature (oC)
1st exciton peak (nm)
PL max peak (nm)
FWHM (nm)
Size (nm)
Φ (%)



The nanocrystals show a broader PL peak at lower growth times indicating surface defect patterns in the dynamic nucleation and growth reactions. At the shortest growth time, blue nanocrystals of higher energy bandgap as much as xxeV were synthesized at 190/170 oC.   


Similar to the quantum dot synthesis, 1mM of 99.99% CdO (128.4 mg) was mixed with 6 mM of oleic acid (2 mL) and 20 mL 1-octadecene in a four-necked flask. The mixture was degased for few minutes and heated up with a reflux until the solution turns clear (usually around 250 oC). The Se precursor was prepared by weighing out 39 mg of Se powder in a glove bag and dissolving in 1.5 mL TOP (90%) under sonication for 5 minutes in a small vial. The TOPSe was mixed with 36 mg of CTAB in 2 mL of toluene. As soon as the Cd-oleate turns clear, the temperature was adjusted for injection.  In a typical synthesis where 3.5 mL of TOPSe was injected at 190 oC, the particles were grown at 170 oC for two minutes after which the heat mantle was quickly removed and argón flow increased to cool the sulution fast. Short arm tetrapods were synthesized using 103 mg of CdO and 32 mg of Se poder. The resulting solution was divided into four 40mL vials in which around 2 mL of chloroform and methanol mixture was added prior to centrifuging with approximately 20 mL of acetone for 20 minutes. Additional cleaning was done only with acetone two times at 4000rpm 10 minutes each. The resulting precipitate was dried and redispersed in chloroform or hexane for characterization.
As in the case of the CdSe quantum dots, a matrix of temperature and time controlled syntheses has been carried out to find the best experimental conditions that significantly affect the nanocrystal growth. Based on an optimized recipe, reproducibly shape-controlled tetrapods were prepared.
The absorption spectra of typical tetrapods is characterized by a sharp first excitonic peak followed by a shoulder a valley at lower wave length. Peak counts as much as eight were common in these syntheses while observing the second excitonic peak coresponding to a wurtzite structure of the tetrapods. As shown in Fig. 3, the peak broadens at higher growth temperature and time conditions which could be an indication of Ostwald ripening- the evolution of an inhomogenous structure over time.    
The flourescence of these semiconducting materials was uniquely sharper than its quantum dot counter parts. A characteristic PL is sharp, with minimum FWHM(about 20 nm), symetrical and with high emission. At lower growth times, the surface trap states are predominant because of a dynamic nucleation and growth reaction leaving incomplete covalent bonds. This can be seen from the lagging tail on the PL peak at higher wave length.
However, the choice of the experimental conditions depends on what arm length and thickness of the tetrapods is needed. In particular, long arm tetrapods are synthesized at 190 oC while the short arm tetrapods are prepared at 300 oC with a modified recipe.

X-ray powder diffraction is also crucial in studying the crystalinity of the tetrapods. This has been reported by several authors. Fig. 5 shows XRD of the as-synthesized quantum dots and tetrapods.
Fig. 4 Optical properties of tetrapods sybthesized a) 190 oC/170 oC c b) TEM, 190 oC/170 oC/2 min simple c)250 oC/230 oC d) 300 oC/280 oC

Table 2 Optical characteristics of CdSe TPs
Temperature (oC)
1st exciton peak (nm)
PL max peak (nm)
FWHM (nm)
Arm length (nm)
Φ (%)



Fig. 5 XRD of CdSe nanocrystals of QDs (green) tetrapods (red) sythesized at 250 oC for 2 min 

TOP-free tetrapods

Normally, tetrapods are synthesized by a rapid hot injection of TOP-Se in the presence of CTAB. Although TOP free strategies are available for colloidal quantum dots, it was barely applied for elongated nanocrystals. The preparation of these ´super-long tetrapods´is as follows: 1 mmol of CdO is added to 2 mL oleic acid , and heated with stirring under inert conditions until a clear cadmium oleate complex is formed. The solution is cooled to room temperature but 1 mL of toluene should be added when the temperature lowers below 100 oC. This helps the complex remain in liquid state. Similarly, equivalent amount of Se (to Cd), is added to 20 mL of ODE and heated to 280 oC (or another temperature depending on which the arm length and thickness is controlled). Adding 0.1 mmol of CTAB to the prepared Cd-oleate, and injecting into the ODE-Se adduct the synthesis is monitored at different growth times. The color development corresponding to the particle size is found gradual which benefits easy control of arm length and thickness.

Hollow nanoparticles

Hollow nanoparticles/nanocrystals are spherical nano-objects formed as a direct consequence of a classic mechanism in metallergy- the Kirkendall effect. The phenomenon, induced from the difference in diffusivities of atoms, is a generic route to diverse range of hollow nanostructures reported previously; such as ZnO, Co3S4, CoO, CoSe2, CoTe, Co2P and Ni2P nanoparticles, for example. Hollow nanocrystals and the method of making is recently patented by Alivisatos and co-workers. []The underlying mechanism of hollowing has been explained by the generation of small Kirkendall voids near interfaces as a result of different interdiffusion rates and surface diffusion of core material along the pore surface or a direct elapse of the material in the core.
Very recently, group of Wong at Rice University succeeded in developing a Kirkendell process to synthesize CdSe hollow nanoparticles of less than  20 nm in size. The HNPs thereof were ligand exchanged and applied in hybrid solar cell devices to study the charge transfer reactions.
In a typical synthesis, 2.47 g of cadmium nitrate tetrahydrate (Cd(NO3)2•4H2O, 8 mmol, purity = 99%, Sigma-Aldrich), 0.32 g of selenium powder (Se, 4 mmol, purity = 99.999%, Sigma-Aldrich), 20 mL of 1-octadecene ("ODE", purity = 90%, Sigma-Aldrich) and 0.15 g of hexadecyltrimethylammonium bromide ("CTAB", 0.4 mmol, purity >99%, Sigma-Aldrich) were added to a four-neck flask, which was then heated slowly to 190 °C at a rate of 11 °C/min. The Cd:Se:CTAB molar ratio was 2:1:0.1. The reaction flask was kept at this temperature for 5 min and cooled down to room temperature, resulting is a reddish-brown cloudy liquid. The resultant solution was centrifuged to remove the solids, resulting in a clear reddish-brown liquid. Ethanol was then added to the reddish brown liquid to crash out the HNPs. The HNP powder was washed several times with ethanol before being dispersed in chloroform. The synthesis method was also performed with no CTAB, and twice and triple the original amount of CTAB (i.e., 0 mmol, 0.8 mmol, and 1.2 mmol). 

III) Ligand Exchange Strategies

Pyridine Exchange

Presynthesized and chloroform dispersed QDs (10 mg/mL) were added to a screw top vial with a septum cap. The content was dried with a flow of Ar until the solvent is completely evaporated or QDs are left a bit wet in some cases. The QDs were redispersed with 10 mL of pyridine with a glide syringe through the septum in the fume hood (under similar stirring and inert conditions). The ligand exchange for quantum dots is exceptionally slow and requires controlled heating and cooling steps. The pyridine exchange was comparatively more successful when heating the content to not more than 100 oC for 12 hours in an oil bath while redispersing to the same concentration so that all ligands are not evaporated through the vent needle. The same concentration of NCs was left for additional 12 hours at room temperature and with stirring. Thereafter, few drops of check-up sample were taken to assure the exchange has taken place and added to 5 mL hexane. A cloudy suspension means most of the pyridine exchange was performed in which case the parent solution was filled with at least 40 mL of hexane (until it turns cloudy) and centrifuged. The tendency of QDs capped in pyridine ligands is to aggregate. The hexane washed precipitate is later redissolved in chloroform to the desired concentration. This procedure was repeated whenever the surface passivation strategy was unsuccessful (QDs remain dissolved in the washing solvent).
Pyridine exchange for TPs and HNPs was relatively facile since room temperature stirring under inert conditions of 12 hours was effective with a final hexane wash and dispersion in pyridine:chloroform (1:9).

Butylamine Exchange

The short amine ligand exchange of quantum dots was done following a prior pyridine exchange; in which case 10 mL of butylamine is added to 100 mg of nanocrystals. The content was left stirring under argon for 24 hrs at 70 oC. Nanocrystals obtained were precipitated with acetone. The final volume redissolved in 1:9 butylamine: chloroform solution. Similarly, the TPs and HNPs were treated for coating exchanges overnight.

a)    Oleic acid     


b)   Pyridine      
c)    Butylamine

IV)        Photovoltaic performances

Rsh (MΩcm2)

IV)         Reproducing the syntheses

Few syntheses of QDs and TPs were conducted to see the reproducibility of the routes. For the experiments carried out it turns out to be far away from the expectations in terms of homogeneity. The full width at half máximum was almost twice the expected values. This is bad since there will be energy transfer in between polydispersed nanocrystals.

Experimental results

The fact that these experiments are susceptible to any changes in the syntheses conditions imposed a bottle neck in preparing nanocrystals of typical high quality characteristics. In the course of identifying the problem, the optical properties of the newly produced quantum dots and tetrapods were compared to common characteristic features.

Observation: Changing the volume of the flask, although a slight difference in injection temperatures, blue-shifted the absorption peak somehow. However, the synthesis was expected to produce QDs (abs. 540 nm) and TPs (abs. 570 nm) with FWHM values of less than 30 and 23 nm, respectively. It should be noted though a proper heating mantle was not used which resulted in a huge drop in temperature (more than 100 oC) upon room temperature injection of the TOP-Se. Unlike, previous syntheses, the Cd-oleate was a yellowish solution at injection temperature which showed some sort of impurities in the constituting reactants. As a result, all new reagents were purchased from Sigma Aldrich to the same impurity levels as in previous syntheses at Rice.

Experimental results

In this experiment, all new reagents were utilized. Temperature control was done in-situ using adjustable thermometers. The synthesis was better in that the Cd-oleate turned almost clear after degasing the ODE for 30 min, adding OA and degasing for 10 min and finally heating the reaction volume to 250 oC. The experimental conditions were well controlled and the right absorption bands were observed while homogeneity of the sample remains unresolved. This was evident from the fewer number of band features and the broadness of the FWHM. However, the PL peaks were so symmetric and no evidence of surface traps at higher growth times.

VI) Conclusion

So far, different shapes of particles were reported in literature, including, dots, tetrapods, tripods, tear drops and even hollow nanoparticles. Further optimization in getting highly crystalline nanocrystals is underway. Some of the important nanostructures were such as dots and tetrapods are being optimized to obtain the best results. However, in comparison to the the other shapes, the tetrapods outperformed in solar cell applications owing to their physical orientation that increases the electron mobility in the devices. No solar cells of HNPs in DSCs are reported in literature so far. We have achieved butylamine exchanged HNPs of higher efficiency compared to pyridine exchanged counterparts.

Institute of Chemical Research of Catalonia (ICIQ)

Universitat Rovira i Virgili

Rice University

Prof. Emilio Palomares

Prof. Michael S Wong

Suravani Gullipalli