Role of Core–Shell Formation in Exciton Confinement Relaxation in

Letter pubs.acs.org/JPCL

Role of Core−Shell Formation in Exciton Confinement Relaxation in Dithiocarbamate-Capped CdSe QDs Sreejith Kaniyankandy* and Sandeep Verma Radiation & Photochemistry Division, Bhabha Atomic Research Centre, Trombay, Mumbai 400085, India S Supporting Information *

ABSTRACT: The possibility of exciton delocalization in alkyldithiocarbamate (ATC)capped CdSe has been investigated for several alkyl groups and compared with phenyldithiocarbamates (PTCs). We find a bathochromic shift for ATC similar to the one obtained for PTC. Our computational studies show reduction in HOMO−LUMO gaps in both PTC and ATC, albeit with a lower shift. However, TDDFT studies revealed that ATC-capped CdSe is more of a localized HOMO state as compared with partly delocalized HOMO in PTC-capped CdSe, hinting at a difference in electronic interaction between the two binding groups. We hypothesized the formation of sulfide layer over the CdSe QDs as the possible reason for the observed bathochromic shift, as verified by absorption spectra of S2− ligand exchange samples. The formation of CdS shell leads to substantial electron delocalization because CdSe CB is in close resonance with CdS, which is exactly the opposite of what was previously concluded in the literature.

S

ligands into pulling charge carriers out of the QDs by careful alignment of ligand energetic alignment at the interface.10−12 Recently, this was exactly what was claimed by Weiss and coworkers. In their work the native ligands were replaced by phenyldithiocarbamate (PTC) to achieve the aforementioned objective.13−15 Their optical absorption before and after the ligand exchange showed a substantial red shift in absorption spectra, which was implied as carrier (hole) confinement relaxation due to energetics of the highest occupied molecular orbital (HOMO) level with respect to valence band (VB). In the case of PTC, it was observed that presence of close resonance between the valence band (VB: −6.1 to −6.3 eV with respect to vacuum) and HOMO (−6.1 with respect to vacuum) of the ligand leads to confinement relaxation of holes, which directly leads to red shift in absorption spectra, as shown for CdSe, CdS, and PbS.13,14 Further studies by the same group reported that changing the ligand energetics by tuning the functional groups on the ligand lead to varying shift in delocalization of excitons.15 These observations led to the conclusion that ligand energetics have a substantial role to play in the bathochromic shift of the ground-state exciton transition in the presence of surface ligands. In general, ligand interaction with a given surface has been discussed based on information obtained from absorption spectra. With that in mind, it is necessary to ask if this is the only mechanism to explain the observation. To understand such electronic interactions, ab initio computational studies on clusters have been found to be remarkably useful.16,17 In terms of interactions of ligands with

emiconductor quantum dots (QDs) have emerged as the material of choice over the past several years owing to their increasing technological impact as active material in several optoelectronic devices.1−4 The typical feature of quantum dots that enable such versatility is the formation of size-tunable energetics of the bound state between electron and hole or, in other words, referred to as an exciton. Such a bound state is confined within the nanocrystal by balance of energetics. In most cases the wave function is buried well within the nanocrystal due to confining potential at the surface arising from surface ligands.5 The confining potential at surface prohibits the wave function from leaking far from the surface. Apart from purely electronic perspective, these ligands also additionally provide colloidal stability in a solvent.5,6 While this is beneficial from the point of view of colloidal stability and prevention of traps states at the surface, in the case of devices based on QDs, it is a vexing problem. This problem has a tremendous impact on most practical applications like optoelectronics, photovoltaics, and so on, where we expect the carriers to have wave functions that are exposed to the surrounding quantum dots, which enables an efficient transport.1,3−5,7 In the absence of such a “contact” the carriers are bound to undergo substantial scattering at the interface, leading to energy loss that is detrimental for most practical applications. Therefore, in most cases one employs postsynthesis processing procedures to quantum dots to alleviate these issues. One way of achieving the above objective is by linking of QDs or alternatively by replacing long-chain fatty acids at the surface with short-chain ligands.1,8,9 In this strategy we rely only on the role of distance in the transport, and ligand seldom plays a role other than as a role of a “ruler” by maintaining the distance. So, a natural question that arises is what if one can coerce the © XXXX American Chemical Society

Received: May 20, 2017 Accepted: June 29, 2017 Published: June 29, 2017 3228

DOI: 10.1021/acs.jpclett.7b01259 J. Phys. Chem. Lett. 2017, 8, 3228−3233

Letter

The Journal of Physical Chemistry Letters

we only get spherical nanocrystals. Therefore, one can make a direct comparison between the exciton peak position obtained and the radii using the size scaling as obtained from Peng and coworkers.23 On the basis of the sizing curve, we obtain the sizes ranging from 1.9 to 4 nm for the QDs. The size region covers the region that Weiss and coworkers13,15 have taken. This enables a direct comparison of any size-dependent shift in ligand-exchanged samples unambiguously. These samples were further used to perform ligand exchange with dithiocarbamates. In all cases, as mentioned in the Supporting Information, the ligand exchange was performed to completeness as arrived from the saturation of the bathochromic shift. The optical absorption spectra of the ligand-exchanged samples dissolved in toluene are shown in the Figure 1b,c. Observations made on these samples reveal that there is a red shift in exciton transitions in all quantum dots. Additionally, the quantum of shift seems to be largely influenced by the size of nanocrystal. In other words, bathochromic shift is found to be higher for smaller quantum dots as compared with larger quantum dots. Furthermore, the shift does not depend on the type of alkyl group used in all ATC used; the shift is invariant with respect to the alkyl group. Moreover, the shift clearly matches well with the one obtained for PTC for samples of all sizes. To elucidate this point we used similar method as used in ref 13 to plot the change in the confinement energy with respect to that of different sizes of quantum dot (Figure 1d). The trend, as seen from the Figure 1d, indicates that the shift is clearly radii-dependent and is in accordance with that observed in CdSe QDs.13 Apart from the change with respect to radii of QDs, one can get information on the variation in confinement by comparing the change in the confinement radii with respect to the size, as seen in Figure 1d. The pattern obtained clearly indicated that change in both energy and confinement radii with respect to PTC clearly tallies well in the present case. This possibly indicates a similar origin for the shift in each of the cases. However, the origin of shift remains controversial, and there are several factors that could contribute to the shift in absorption spectra: (i) local field factors, (ii) change in the QD size during exchange, and (iii) possible change in exciton delocalization brought out by change in the ligand electronic structure. In the present case, we can rule out local field factors because this may not give rise to a large size-dependent variation. While exciton delocalization is a plausible scenario in the case of PTC, we hypothesize that if this is the case then in the present case the alky chain presents a HOMO level well within the valence band. Therefore, in such a scenario the possibility of delocalization remains remote at least in the present case. However, the shift obtained in the present case (ATC) is very similar to the ones obtained for PTC, which possibly indicates similarity in origin of bathochromic shift. In our studies, we also observed that the PL peak also shows a similar shift in both cases of dithiocarbamate (ATC and PTC). To gain a greater insight into the origin of the shift, we conducted computational studies on CdSe QD with different types of ligand. For this purpose, we chose a CdSe magic cluster Cd56Se50 as representative of the QD. The relevant ligands for the present cases are carboxylate (COOH), phenyl dithiocarbamate (PTC), and methyl dithiocarbamate (MTC). The carboxylates are the native ligands on the surface of quantum dot (SI). The starting base structure for molecular dynamics (MD) is provided in SI 2 (Supporting Figure 2, computational details in the SI). Figure 3a (SI 4) shows the isosurfaces of frontier

the surface, carboxylates and dithiocarbamates can be regarded as donors of electron, which are possibly spread over both oxygen and sulfur, and unligated Cd ion on the surface, which bears a positive charge.5 This interaction is widely regarded as a case where the ligand plays a surface charge compensation role. Interaction between PTC and CdSe has been shown to be substantial, and possible formation of hybridized HOMO state does exist.18 In a previous computational study on complexation of CdSe QDs with SCH3, it was shown that the HOMOlowest unoccupied molecular orbital (LUMO) difference was reduced by ∼0.2 eV on the CdSe surface.19 However, on careful analysis of carrier localization, it was observed that the thiolcapped surface introduces a trap at the surface and the HOMO level has a fairly localized character in the thiolate group rather than a hybridized delocalized one. Observations that thiolate binding on CdSe creates a new level in the mid gap region are not new and have been observed from several experimental studies where the emission spectra are completely quenched on thiolate addition and highly red-shifted trap-state emission.20−22 Therefore, it is necessary to ask if in the case of PTC such an interaction takes place. But apart from this scenario that has been mentioned in their paper, are there other effects taking place in tandem? Therefore, we asked the question, “What if one were to have an alkyldithiocarbamate (ATC) in place of PTC?” In ATC, it is natural to assume that the HOMO level of ATC is not in resonance with CdSe VB. Therefore, we should not obtain a bathochromic shift in the case of ATC. In the present study, we arrive at a different conclusion for the bathochromic shift observed in the case of dithiocarbamates based on our experiments and computational studies. Figure 1 shows the absorption spectra of as-prepared oleicacid-capped QDs. The lowest exciton peak position ranged from 470 to 580 nm. In the synthesis methodology employed,

Figure 1. Steady-state absorption spectra of CdSe QDs of different sizes (in nm) (i) 2.1, (ii) 2.2, (iii) 2.4, (iv) 2.6, (v) 3, (vi) 3.4, and (vii) 3.6 capped with (a) oleic acid, (b) PTC, (c) OTC, and (d) ΔEg and ΔR versus R plot (ΔEg (Red), decrease in bandgap energy observed from shift in absorption peak shift after dithiocarbamate ligand exchange; ΔR (Blue), corresponding change in exciton radius after dithiocarbamate ligand exchange). (e) TEM image of ATC-CdSe QDs (size 3.6 nm). 3229

DOI: 10.1021/acs.jpclett.7b01259 J. Phys. Chem. Lett. 2017, 8, 3228−3233

Letter

The Journal of Physical Chemistry Letters

Figure 2. Projected density of states (PDOS) of (a) COOH-Capped CdSe and (b) PTC-capped CdSe (Note: the energy levels are with respect to VB).

Figure 3. Isosurfaces of frontier orbitals of ATC-capped CdSe. Isosurface values were set to 0.003 for all orbitals.

orbitals of CdSe with COOH. The first frontier orbital or HOMO level clearly indicates that the wave function is delocalized over the entire nanocrystal with negligible contribution on ligands. The isosurface of LUMO shows a spherical symmetry clearing, indicating the lowest LUMO of exciton to have S-symmetry and the next three frontier orbitals to have P-type symmetry. This indicates that the ligands “do not disturb” the exciton wave function in ligated QDs and only have the role of the confining group. This observation is also ascertained by our projected density of states (PDOS) computations (Figure 2), which show HOMO−LUMO energy separation of ∼2.0 eV. This value is lower than experimental value by almost 1 eV but can be attributed to the inability of generalized gradient approximation (GGA) to correctly account for the exchange correlation interaction.24 Nevertheless, the DOS captures most of the features as expected for the electronic structure of CdSe. The PDOS for the VB predominantly arises from Se 4p and CB arises from 4s orbital Cd. As evidenced from the isosurface, none of the frontier orbital has localization on the ligands and additionally there are no localized states originating at the surface, which clearly indicates the robustness of capping in the case of COOH. While frontier orbitals remain undisturbed, we find that well within the individual bands there is a large overlap between the ligand and QDs. Contributions to levels ∼1 eV below the VB edge in the case of COOH come from oxygen. These computational results are in accordance with previous studies by Azpiroz and De Angelis,18 and Voznyy et al.25,26 on similar ligands. It must be mentioned that surface reconstruction does not lead to the presence of localized mid-gap states, clearly indicating that these orbitals have been pushed deep within the bands.

In PTC, close resonance between HOMO and VB is expected to change the QD electronic structure significantly. To gain greater insight into the nature of the interaction we compare the frontier orbitals of COOH and PTC (SI 4a,b). We find that isosurfaces for the VB (HOMO) are not spread uniformly over all of the ligands but only a few ligands. In the case of PTC, while CB (LUMO) maintains the spherical symmetry, the higher CB orbitals (P) clearly show disruption of P-symmetry, indicating that there is a substantial interaction between PTC and CdSe QDs. We also find that not all ligands contribute to the HOMO uniformly, and the delocalization seems to be aided by N, C (connected to S) of the carbamate and S, clearly indicating that atoms of the linker group, which in this case is a dithiocarbamate, contribute to HOMO orbitals. This clearly indicates that lone pair on the nitrogen may have a role in possible delocalization of wave function. While isosurfaces does show a delocalization of the exciton wave function enabled primarily by linker groups, it does not tell much about the magnitude of this shift. To ascertain the role of the ligand to the bathochromic shift we compare the PDOS for each of the cases. The PDOS of PTC samples is given in Figure 2b. The first observation that we can make from the Figure is a reduced gap by ∼0.2 eV as compared with COOH. In PDOS of COOH we find that oxygen of ligands has only negligible contribution to frontier orbitals both HOMO and LUMO. On the contrary, in the case of PTC, S and N atoms are clearly seen to contribute toward HOMO. This clearly indicates that the wave function leaks into the ligand shell comprising dithiocarbamate and phenyl groups. Wave function delocalization to the ligand shell clearly helps us explain the reduction in band gap in the case of PTC, which in the present case is ∼0.2 3230

DOI: 10.1021/acs.jpclett.7b01259 J. Phys. Chem. Lett. 2017, 8, 3228−3233

Letter

The Journal of Physical Chemistry Letters

absorption. Therefore, we carried out TDDFT computations on all clusters. The results of TDDFT (SI 5) clearly indicate that there is a fundamental difference between interaction between QDs, PTC, and ATC. The reference for this study is COOH-QD, as this is known to give a clean electronic structure without traps,18 and we find that in the case of PTC the highest absorption comes from the HOMO−1−LUMO transition at 1.94 eV. The HOMO−LUMO transition has a very low extinction coefficient. On the contrary for ATC, the highest extinction coefficient appears for transition at 2.05 eV, and there are several transitions from 1.94 to 2 eV, albeit with lower extinction coefficient. In both of the cases it is likely that the low extinction coefficient states are traps, as also observed from the isosurfaces. Therefore, even in the case of PTC it is clear that the shift that is seen from TDDFT for the state with transitions with highest extinction coefficients is only 0.1 eV, and ATC does not exhibit any shift based on the transition with the highest extinction value. In the present study, all computations studies have been conducted on a CdSe QD of ∼1.7 nm size, and the shift in comparison with experiments by Sardar and coworkers27 is far less and cannot be clearly accommodated to just a simple ligand exchange scenario. In light of these observations from both theory and experiments, it is important to ask if the interaction between QD and dithiocarbamates is a trivial case of ligand exchange. This answer was partly answered by Kilina et al. recently by considering the reaction chemistry at the surface of CdSe and PTC.28 The result as reported by Kilina et al. clearly demonstrates that there is a degradation of phenyldithiocarbamate at the surface of CdSe. Decomposition of PTC in the presence of an acidic counterion like either triethylamine ion (TEA+) or NH4+ leads to the formation of the corresponding amines, carbon disulfide, thiuram sulfides. Previous literature studies by Green et al.29 demonstrated that in the case of mercury-based dithiocarbamates, decomposition of the metal dithiocarbamate complex occurs even at room temperature, leading to the formation of HgS quantum dots after several hours. These studies also demonstrate that the formation of sulfides from a starting metal dithiocarbamate complex is not unusual and clearly hints at the role of possible dynamic nature of ligand QD surface reaction chemistry. Metal dithiocarbamates also have been used in the synthesis of QDs and sulfide shell formation by Pradhan et al.,30 even in case of cadmium, zinc-based dithiocarbamates via thermal decomposition. In the case of CdS they in fact established the role of additional reactants in decomposition. They concluded that the presence of trioctylphosphine (TOP) in solution showed a reduction in decomposition temperature for the formation of CdS to room temperature. Consolidating our observation in light of several previous experimental and computations reports, we can speculate that in our case alkyl dithiocarbamate may lead to the formation of a shell sulfide, which, in turn, leads to a core− shell system with a CdS-like shell. Consequently, this picture might help explain relaxation of exciton confinement. The surface reaction in our case is expected to be self-limiting, assuming that a very low amount of Cd is present either in the reaction solution or especially on the surface of QD. Such monolayer deposition may not be readily visible in a normal TEM with a single monolayer change of only ∼0.16 nm in thickness coming from the deposition of a layer of S (or ∼0.32 nm with deposition of one monolayer of CdS). We believe that further studies of decomposition products released may help reveal the exact decomposition species and henceforth the

eV; however, for smaller QDs the experimental bathochromic shift is close to 0.5 eV. This difference between experimental and theoretical value cannot be solely attributed to underestimation of band gap using DFT. In comparison with computational studies on thiolates as reported previously by Infante et al.,19 it is clear that PTC exhibits much stronger interaction with QDs, which helps to explain larger bathochromic shift in the absorption spectra; extending this argument one can envisage that alkyl dithiocarbamate may not exhibit a larger delocalization due to lack of resonance between VB and HOMO. Therefore, to ascertain the cause of bathochromic shift in ATC we conducted computational studies on alkyl dithiocarbamates with methyl dithiocarbamate as a model for computational studies for all ATC, and the isosurface obtained for alkyl dithiocarbamates is given in Figure 3. It is clear from the isosurfaces for HOMO that the HOMO level has been completely disrupted and there is a substantial change in the electronic structure for the top of the VB. On the contrary, LUMO orbitals remain unchanged in contrast with PTC even though the linker group has remained the same (see SI 4). Additionally, it is also seen that HOMO is more localized toward the ligand in contrast with PTC. It is clear from the PDOS represented in Figure 4 that the S and Se atoms contribute significantly toward the HOMO level.

Figure 4. PDOS of ATC-capped CdSe (Note: the energy levels are with respect to VB).

There seems to be some hybridization in HOMO for this particular case. The energy gap in this case is lower as compared with COOH, but the change is found to be lesser than what is obtained for the case of PTC. The PDOS plotted clearly hints at the point that the linker group atoms N and S contribute substantially toward the delocalization. In the case of ATC, PDOS clearly reveals that a large fraction of HOMO levels is contributed solely by the S atom, which, in turn, hints at the fact that holes are most likely localized on S. This clearly indicated that the holes are basically trapped on the sulfide binding groups of the ligands. Nevertheless, in comparison with both dithiocarbamates, we can conclude that just exciton delocalization cannot be completely attributed to wave function spread to phenyl groups alone. Furthermore, we find that comparison made from KS orbitals obtained from just a ground-state calculation cannot be used for directly comparing the bathochromic shift obtained in optical 3231

DOI: 10.1021/acs.jpclett.7b01259 J. Phys. Chem. Lett. 2017, 8, 3228−3233

Letter

The Journal of Physical Chemistry Letters

have been synthesized and compared with phenyldithiocarbamate, which has been shown to delocalize holes in CdSe. Our studies reveal that the PTC is not a unique ligand to show bathochromic shift, but alkyldithiocarbamates show similar behavior. Our computational studies clearly show that ligands have a disruptive influence of the electronic structure of QDs and collapse of frontier orbitals in comparison with PTC. For example, the higher LUMO orbitals have a substantially hybridized contribution from the S of the ligands. The studies reveal that in the case of ATC we found the contribution of S atoms directly on the HOMO level of the QD. However, there is a fundamental difference between the interaction of ATC and PTC. The ATC orbitals were found to not contribute to optical absorption based on these analyses; we believe that in the case of ATC it is a local surface trap rather than a delocalized one. We hypothesized that the discrepancy between theoretical and experimental studies could be resolved by considering the dynamic nature of ligands at the surface and role of decomposition of dithiocarbamates. On the basis of these studies we hypothesized that surface S2− deposition on the decomposition of PTC and the formation of CdS shell might possibly be the reason for exciton delocalization. We verified this claim by conducting ligand exchange with sulfide, and interestingly we observe that the bathochromic shift is similar for PTC, ATC, and S2− ligand-exchanged samples, indicating the possibility of the similarity of the origin.

cause of the shift in the case of alkyl dithiocarbamate and, in general, for all thiocarbamates. However, on the basis of the above assessment we can verify the claim of sulfide deposition based on S2− ligand exchange on the surface. Therefore, to test our hypothesis, we exchange the native carboxylate ligand on surface of QD to S2−. Figure 5 shows the

Figure 5. Steady-state absorption spectra of CdSe QDs of different sizes after S2− exchange: (i) 2.1, (ii) 2.4, (iii) 2.6, and (iv) 3 nm. Inset: ΔEg and ΔR versus R plot (ΔEg, decrease in bandgap energy observed from shift in absorption peak shift after dithiocarbamate ligand exchange; ΔR, corresponding change in exciton radius after dithiocarbamate ligand exchange; (a) butyldithiocarbamate and (b) S2−).

■

comparison to PTC and S2−. The shift, as seen from delocalization radii and energy shift, is very similar in all of the cases. This ascertains the possibility that exciton delocalization is most likely due to S2− shell formation on CdSe surface. Previously, such formation of S2− shell over CdSe has been shown to give a highly red-shifted first exciton state by Talapin and coworkers.31 In the case of nanoplatelets it was clear that for smaller nanoplatelets the shift is considerably larger than for larger platelets, clearly demonstrating the role of possible sulfide formation at the surface. In the present case, such an explanation needs to be invoked to explain the red shift in our case, as our TDDFT studies indicated that bathochromic shift cannot be attributed to adsorbed ATC. Therefore, we reason that the shift in the case may be assigned to the possible formation of core and shell sulfide, leading to a confinement relaxation. Additionally, the reason for the bathochromic shift on S2− shell formation is completely different from that invoked in the case of PTC. In the case of S2− the delocalization is most likely an electron delocalization rather than hole delocalization ascribed to PTC because the outer sulfide with an outer Cd layer coating essentially tries to mimic a CdS shell. The band alignment at the interface is the one of CdSe core and CdS shell. The band alignments for the bulk CdSe and CdS are plotted in SI 7, which clearly indicates a close resonance between the CB of CdSe and CdS on the contrary VB energy offsets are significant. In the presence of confinement, CB levels of CdSe are expected to shift upward, forming a a quasi-Type II alignment32 at the interface. To delineate this point we calculated the electronic structure for a smaller cluster and coated it with CdS shell (SI 9). The figure represented in SI 9 clearly shows a large reduction in band gap for CdSe covered with CdS shell with a value of ∼0.6 to 0.7 eV, which clearly accounts for the shift for the smallest CdSe samples measured.27 To summarize, we state that the role of alkyldithiocarbamate in wave function delocalization has been investigated by both experimental and computational studies. Alkyl dithiocarbamates

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpclett.7b01259. Details of CdSe synthesis, spectroscopic studies, and additional computational studies on ligated CdSe QDs. (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Tel: 022-25593627. Fax: (+) 9122-25505331/25505151. ORCID

Sreejith Kaniyankandy: 0000-0003-3097-604X Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank Dr. H. Pal and Dr. P. D. Naik for their constant support and encouragement. We thank Dr. Suman Neogy, Mechanical Metallurgy Division, BARC for TEM measurements. S.K. is grateful to staff members of the Supercomputing Center, BARC for their kind support.

■

REFERENCES

(1) Talapin, D. V.; Lee, J.-S.; Kovalenko, M. V.; Shevchenko, E. V. Prospects of Colloidal Nanocrystals for Electronic and Optoelectronic Applications. Chem. Rev. 2010, 110, 389−458. (2) Shirasaki, Y.; Supran, G. J.; Bawendi, M. G.; Bulović, V. Emergence of Colloidal Quantum-Dot Light-Emitting Technologies. Nat. Photonics 2012, 7, 13−23. (3) Pietryga, J. M.; Park, Y.-S.; Lim, J.; Fidler, A. F.; Bae, W. K.; Brovelli, S.; Klimov, V. I. Spectroscopic and Device Aspects of Nanocrystal Quantum Dots. Chem. Rev. 2016, 116, 10513−10622. 3232

DOI: 10.1021/acs.jpclett.7b01259 J. Phys. Chem. Lett. 2017, 8, 3228−3233

Letter

The Journal of Physical Chemistry Letters (4) Carey, G. H.; Abdelhady, A. L.; Ning, Z.; Thon, S. M.; Bakr, O. M.; Sargent, E. H. Colloidal Quantum Dot Solar Cells. Chem. Rev. 2015, 115, 12732−12763. (5) Boles, M. A.; Ling, D.; Hyeon, T.; Talapin, D. V. The Surface Science of Nanocrystals. Nat. Mater. 2016, 15, 141−153. (6) Green, M. The Nature of Quantum Dot Capping Ligands. J. Mater. Chem. 2010, 20, 5797−5809. (7) Kagan, C. R.; Murray, C. B. Charge Transport in Strongly Coupled Quantum Dot Solids. Nat. Nanotechnol. 2015, 10, 1013− 1026. (8) Williams, K. J.; Tisdale, W. A.; Leschkies, K. S.; Haugstad, G.; Norris, D. J.; Aydil, E. S.; Zhu, X.-Y. Strong Electronic Coupling in Two-Dimensional Assemblies of Colloidal PbSe Quantum Dots. ACS Nano 2009, 3, 1532−1538. (9) Liu, Y.; Gibbs, M.; Puthussery, J.; Gaik, S.; Ihly, R.; Hillhouse, H. W.; Law, M. Dependence of Carrier Mobility on Nano-crystal Size and Ligand Length in PbSe Nanocrystal Solids. Nano Lett. 2010, 10, 1960−1969. (10) Munro, A. M.; Zacher, B.; Graham, A.; Armstrong, N. R. Photoemission Spectroscopy of Tethered CdSe Nanocrystals: Shifts in Ionization Potential and Local Vacuum Level As a Function of Nanocrystal Capping Ligand. ACS Appl. Mater. Interfaces 2010, 2, 863−869. (11) Soreni-Harari, M.; Yaacobi-Gross, N.; Steiner, D.; Aharoni, A.; Banin, U.; Millo, O.; Tessler, N. Tuning Energetic Levels in Nanocrystal Quantum Dots through Surface Manipulations. Nano Lett. 2008, 8, 678−684. (12) Brown, P. R.; Kim, D.; Lunt, R. R.; Zhao, N.; Bawendi, M. G.; Grossman, J. C.; Bulovic, V. Energy Level Modification in Lead Sulfide Quantum Dot Thin Films through Ligand Exchange. ACS Nano 2014, 8, 5863−5872. (13) Frederick, M. T.; Weiss, E. A. CdSe Quantum Dots by Modification with a Conjugated Dithiocarbamate Ligand. ACS Nano 2010, 4, 3195−3200. (14) Frederick, M. T.; Amin, V. A.; Cass, L. C.; Weiss, E. A. A Molecule to Detect and Perturb the Confinement of Charge Carriers in Quantum Dots. Nano Lett. 2011, 11, 5455−5460. (15) Frederick, M. T.; Amin, V. A.; Swenson, N. K.; Ho, A. Y.; Weiss, E. A. Control of Exciton Confinement in Quantum Dot−Organic Complexes through Energetic Alignment of Interfacial Orbitals. Nano Lett. 2013, 13, 287−292. (16) Puzder, A.; Williamson, A. J.; Zaitseva, N.; Galli, G.; Manna, L.; Alivisatos, A. P. The Effect of Organic Ligand Binding on the Growth of CdSe Nanoparticles Probed by Ab Initio Calculations. Nano Lett. 2004, 4 (12), 2361−2365. (17) Kilina, S.; Ivanov, S.; Tretiak, S. Effect of Surface Ligands on Optical and Electronic Spectra of Semiconductor Nanoclusters. J. Am. Chem. Soc. 2009, 131, 7717−7726. (18) Azpiroz, J. M.; De Angelis, F. Ligand Induced Spectral Changes in CdSe Quantum Dots. ACS Appl. Mater. Interfaces 2015, 7, 19736− 19745. (19) Houtepen, A. J.; Hens, Z.; Owen, J. S.; Infante, I. On the Origin of Surface Traps in Colloidal II−VI Semiconductor Nanocrystals. Chem. Mater. 2017, 29, 752−761. (20) Wuister, S. F.; de Mello Donegá, C.; Meijerink, A. Influence of Thiol Capping on the Exciton Luminescence and Decay Kinetics of CdTe and CdSe Quantum Dots. J. Phys. Chem. B 2004, 108, 17393− 17397. (21) Kalyuzhny, G.; Murray, R. W. Ligand Effects on Optical Properties of CdSe Nanocrystals. J. Phys. Chem. B 2005, 109, 7012− 7021. (22) Baker, D. R.; Kamat, P. V. Tuning the Emission of CdSe Quantum Dots by Controlled Trap Enhancement. Langmuir 2010, 26, 11272−11276. (23) Yu, W. W.; Qu, L.; Guo, W.; Peng, X. Experimental Determination of the Extinction Coefficient of CdTe, CdSe, and CdS Nanocrystals. Chem. Mater. 2003, 15, 2854−2860.

(24) Perdew, J. P.; Levy, M. Physical Content of the Exact KohnSham Orbital Energies: Band Gaps and Derivative Discontinuities. Phys. Rev. Lett. 1983, 51, 1884−1887. (25) Voznyy, O. Mobile Surface Traps in CdSe Nanocrystals with Carboxylic Acid Ligands. J. Phys. Chem. C 2011, 115, 15927−15932. (26) Voznyy, O.; Mokkath, J. H.; Jain, A.; Sargent, E. H.; Schwingenschlögl, U. Computational Study of Magic-Size CdSe Clusters with Complementary Passivation by Carboxylic and Amine Ligands. J. Phys. Chem. C 2016, 120, 10015−10019. (27) Teunis, M. B.; Dolai, S.; Sardar, R. Effects of Surface-Passivating Ligands and Ultrasmall CdSe Nanocrystal Size on the Delocalization of Exciton Confinement. Langmuir 2014, 30, 7851−7858. (28) Munro, A. M.; Chandler, C.; Garling, M.; Chai, D.; Popovich, V.; Lystrom, L.; Kilina, S. Phenyldithiocarbamate Ligands Decompose During Nanocrystal Ligand Exchange. J. Phys. Chem. C 2016, 120, 29455−29462. (29) Green, M.; Prince, P.; Gardener, M.; Steed, J. Mercury (II) N,N′-Methyl-Phenylethyl-Dithiocarbamate and its Use as a Precursor for the Room-Temperature Solution Deposition of β-HgS Thin Film. Adv. Mater. 2004, 16, 994−996. (30) Pradhan, N.; Katz, B.; Efrima, S. Synthesis of High-Quality Metal Sulfide Nanoparticles from Alkyl Xanthate Single Precursors in Alkylamine Solvents. J. Phys. Chem. B 2003, 107, 13843−13854. (31) Ithurria, S.; Talapin, D. V. Colloidal Atomic Layer Deposition (c-ALD) using Self-Limiting Reactions at Nanocrystal Surface Coupled to Phase Transfer between Polar and Nonpolar Media. J. Am. Chem. Soc. 2012, 134, 18585−18590. (32) Zhu, H.; Song, N.; Rodríguez-Córdoba, W.; Lian, T. Wave Function Engineering for Efficient Extraction of up to Nineteen Electrons from One CdSe/CdS Quasi-Type II Quantum Dot. J. Am. Chem. Soc. 2012, 134, 4250−4257.

3233

DOI: 10.1021/acs.jpclett.7b01259 J. Phys. Chem. Lett. 2017, 8, 3228−3233