Threonine functionalized colloidal cadmium sulfide (CdS) quantum dots: the role of solvent and counterion in ligand induced chiroptical properties

Abstract

The functionalization of semiconductor nanocrystals, quantum dots (QDs), with small organic molecules has been studied extensively to gain better knowledge on how to tune the electronic, optical and chiroptical properties of QDs. Chiral QDs have progressively emerged as key materials in a vast range of applications including biosensing and biorecognition, imaging, asymmetric catalysis, optoelectronic devices, and spintronics. To engage the full potential of the unique properties of chiral nanomaterials and to be able to prepare them with tailorable chiroptical characteristics, it is essential to understand how chirality is rendered from chiral molecular ligands at the surface of nanocrystals to the electronic states of QDs. Using a series of polar protic and aprotic solvents together with ammonium (NH₄⁺), tetramethylammonium (TMA⁺), and tetrabutylammonium (TBA⁺) countercations in the preparation of threonine-functionalized cadmium sulfide (Thr-CdS) QDs by phase transfer ligand exchange approach, we demonstrated the significance of the role both the solvent and the countercations play in the transfer of chirality from chiral molecular ligand to achiral semiconductor QDs as apparent by the modulations of the signatures and anisotropy of the circular dichroism (CD) spectra. Moreover, we have utilized tetrabutylammonium countercation to successfully synthesize chiral QDs in nonpolar cyclohexane solvent for the first time. This study provides further insights into the origin of the ligand induced chirality of colloidal nanomaterials and facilitates the synthesis of tailormade chiral QDs.

Graphical Abstract

1. Introduction

The significance of small molecules to the surface science of nanomaterials is widely recognized with recent efforts focusing on molecular-level understanding and control of chemical and physical processes.[1–6] Colloidal semiconductor nanocrystals (quantum dots, QDs) represent an ideal system for studying the capping ligand effects on the optical band gap, absorption coefficients, fluorescence quantum yield, or chiroptical properties.[7–12] Chiral, optically active QDs and quantum rods (QRs) can be easily prepared as aqueous solutions by ligand exchange of achiral capping ligand to chiral capping ligands.[13–19] Chiral QDs functionalized with chiral ligands were successfully applied in sensing and biosensing, stereoselective catalysis, circularly polarized light (CPL) emitters, and spintronics.[20–33] Previous studies have shown that the structure of the capping ligand as well as the binding geometry directly affected the shape and intensity of the CD signal.[34–42] The origin of ligand induced chirality of QDs was proposed to be the hybridization of the (achiral) QD and (chiral) ligand electronic states,[14] or ligand induced chiral dislocations and defects on the QD surface,[43] nonetheless, a thorough understanding still remains elusive.[34, 44–46] The polarity of organic solvents has been shown to cause the shifts of the absorption spectra of colloidal cadmium selenide (CdSe) QDs,[47, 48] however, the effect of solvent or counterion on CD spectra and chiral absorption anisotropy factor (g_CD factor) have not been explored. Given that chiral QDs are typically prepared with polar capping ligands which make them soluble in water, the majority of chiral QDs were prepared and studied for their CD characteristics in water, with only a handful of studies done in methanol (MeOH), dimethylformamide (DMF), or chloroform (CHCl₃).[37, 38, 40] Since the dynamic equilibrium exists between the free and the surface-bound chiral ligands, it is reasonable to expect that both solvent and countercation will have a pronounced effect on the chirality of the QDs, as reflected by their CD spectra: phase transfer ligand exchange is affected by the countercation of the negatively-charged capping ligand, and chirality of the QDs is controlled by the ligand-QD interactions.

Herein we present a study of the chiroptical properties of threonine-functionalized CdS (Thr-CdS) QDs prepared by a ligand exchange procedure from oleic acid capped CdS (OA-CdS) QDs in cyclohexane into polar protic, polar aprotic and even nonpolar solvents using three structurally different ammonium countercations. Our results showed that both solvent and countercation had a strong effect on chiroptical properties of QDs. Moreover, by using a suitable solvent-counterion combination, chiral Thr-CdS QDs can be prepared in nearly any solvent, and chiroptical properties of Thr-CdS QDs can be easily tuned. Overall, our data provide additional evidence for the important role the solvent and counterion play in the ligand-surface interactions and therefore subsequently in the ligand induced chirality of QDs.

2. Materials and methods

Starting materials:

L-threonine (L-Thr, 99%) was purchased from RPI and D-threonine (D-Thr, ≥99.0%) was purchased from Chem-Impex International. Methanol (MeOH, ≥99.9%), N,N-dimethylformamide (DMF, ≥99.8%), acetonitrile (MeCN, ≥99.9%), formamide (FA, 99%), and ammonium hydroxide (NH₄OH, 28.0-30.0% aqueous solution) were purchased from Fisher Scientific, N-methylformamide (NMF, >99%) was purchased from TCI. Tetramethylammonium hydroxide (TMAH, ≥97%), tetrabutylammonium hydroxide 30-hydrate (TBAH, ≥98.0%), dimethyl sulfoxide (DMSO, >99.9%), and cyclohexane (Cx, ≥99.8%) were purchased from Sigma-Aldrich. All commercial chemicals were used as received. D.I. water (DIW) was obtained from a Milli-Q system with a resistivity of 18.2 MΩ•cm. The diameter of the QDs was determined using Peng’s equation from the absorption spectra.[49]

Synthesis of OA-CdS DQs.

Oleic acid capped CdS (OA-CdS) QDs were synthesized using a slight modification of a previously reported procedure.[50] Briefly: Mixture of sulfur in octadecene (3.2 mg/ml) was heated to 150 °C in 10 min using heating mantle and then was allowed to cool to room temperature (RT) to yield transparent yellowish solution. Cadmium oxide (130 mg) was added at RT to octadecene (17.5 ml) and oleic acid (2.8 ml), and the mixture was heated to 160 °C using heating mantle. After 15 min at 160 °C, the temperature was increased to 275 °C, at which point the flask was removed from heating mantle. When reaction temperature dropped to 260 °C, prepared sulfur solution was quickly injected into cadmium precursor solution. Temperature of reaction mixture dropped to 230 °C, and the reaction mixture was kept at 230 °C for 2 min. Then, reaction mixture was swiftly cooled to RT. Synthesized OA-CdS QDs solution (22 ml) was mixed with isopropanol (88 ml) and the resulting solution was centrifuged (8000×g) for 10 min. The precipitate was resuspended in cyclohexane. Precipitation and centrifugation process was repeated twice to obtain pure OA-CdS QDs.

Phase transfer ligand exchange.

Phase transfer ligand exchange was performed using previously reported procedure.[40] L- or D-threonine and bases (TMAH or TBAH or NH₄OH) were added in polar solvent (3.0 mL) with stirring followed by the addition of cyclohexane (4.8 mL). The reaction mixture was then deoxygenated under vigorous stirring using two vacuum-N₂ cycles. A cyclohexane solution of oleic acid capped CdS QDs (OA-CdS, 0.2 mL) was added via a syringe to the deoxygenated mixture. The resulting reaction mixture was vigorously stirred at RT under N₂ in the absence of light for 30 min. The reaction mixture was then left to stand for 30 min to allow the phases to fully separate after which the layer containing CdS QDs was removed with a syringe, and the L- or D-threonine functionalized CdS QDs were characterized.

Spectroscopic characterization:

CD spectra were collected at 22 °C using a Jasco J-815 spectropolarimeter equipped with a single position Peltier temperature control system. Conditions were as follows: scanning speed 100 nm/min, data pitch 0.5 nm, DIT 1 s, and bandwidth 4.0 nm. A quartz cuvette with a 1.0 cm path length was used for all CD experiments. Each CD spectrum was an average of 8 scans. UV-vis absorption spectra were collected at 22 °C using a Jasco V-600 UV-vis double beam spectrophotometer equipped with a single position Peltier temperature control system. A quartz cuvette with a 1.0 cm path length was used for all UV-vis experiments. Photoluminescence data were collected at 22 °C using a Jasco FP-8300 fluorescence spectrophotometer equipped with a single position Peltier temperature control system. Conditions were as follows: scan rate of 600 nm/min, low sensitivity and 10/10 nm excitation and emission slits. A quartz cuvette with a 1.0 cm path length was used. ¹H NMR spectra were collected on a Bruker Avance NEO 500 MHz NMR spectrometer equipped with a solution state 5-mm broadband iProbe and using a standard Bruker 1D pulse sequence with a 30° pulse. The recycle delay was 1.0 s.

Structural characterization:

The powder diffraction data were measured for the samples at room temperature on a Bruker D8 Venture diffractometer equipped with a CMOS IμS Mo source and PHOTON detector. The data were measured using the Mo Kα radiation sources. The powder samples were rolled into ca. 0.33 mm wide balls and mounted to a MiTeGen micromount. The samples were centered, and the diffraction data measured in the 2θ range, 3–70 °. Imaging was performed on a FEI Tecnai G2 F20 scanning transmission electron microscope (STEM) operating at 200 kV. Samples for TEM were prepared by ultrasonic dispersion of the crude solution of L-threonine functionalized CdS QDs. The suspensions were then drop-cast onto carbon-coated copper grids and dried in air.

3. Results and Discussion

3.1. Solvent and countercation (base) effect on the phase transfer ligand exchange synthesis of Thr-CdS QDs

We started our studies by exploring the effect of the solvent’s structure and polarity on the phase transfer ligand exchange synthesis and circular dichroism of Thr-functionalized CdS QDs from achiral OA-capped CdS QDs (λ_max = 405.0 nm, ø_CdS = 3.55 nm). We used tetramethylammonium hydroxide (TMAH) as a base to deprotonate the Thr and to provide Thr with an amphiphilic methylammonium countercation to increase its solubility in the cyclohexane layer (see ESI for experimental details). Thr has been selected as the ligand, as we expected that solvent and countercation will alter its binding interaction with CdS QD due to the presence of several polar functional groups in its structure (carboxylic, amino, and hydroxy groups). We selected seven solvents that are not miscible with cyclohexane (see Chart 1 and Table 1 for their structures, abbreviations and relative permittivities). Four solvents were polar protic (DIW, MeOH, NMF, FA), i.e., they are H-bond donors and acceptors and can form H-bonds with another molecule of the same species, and three were polar aprotic (DMSO, DMF, MeCN), i.e., these solvents do not serve as H-bond donors but can still participate in H-bonds with molecules that contain H-bond donor groups. It is important to note that although these solvents are immiscible with cyclohexane, they are, to a very small extend, soluble in cyclohexane. Using ¹H NMR we have determined that the molar percentages of polar solvents in cyclohexane in a cyclohexane-polar solvent+base+Thr reaction mixtures varied between 0.8% and 5.4% (see Table S1).

Chart 1.

Structures and abbreviations of threonine, tetramethylammonium hydroxide, tetrabutylammonium hydroxide (top row), polar protic solvents (middle row), and polar aprotic solvents (bottom row). H-bond donors and acceptors in blue and red, respectively.

Table 1.

Percent yields of L-Thr-CdS QDs prepared from OA-CdS QDs by phase transfer ligand exchange from cyclohexane into various polar solvents using different ammonium countercations.^a

solvent	ε b	TMA+ (%)	TBA+ (%)	NH4+ (%)
DIW	80.1	no transfer	no transfer d	no transfer
MeOH	33.0	85 ±2	85 ±2	65 ±3
NMF	189.0	85 ±2	70 ±2	80 ±2
FA	111.0	55 ±3	75 ±3	40 ±3
DMSO	47.2	crash out c	85 ±2	35 ±4
DMF	38.3	crash out c	90 ±2	45 ±3
MeCN	36.6	crash out c	90 ±2	no transfer d

^aYield determined by UV-vis absorption spectroscopy.

^bRelative permittivity.[51]

^cQDs crashed out of the solution.

^dPartial ligand exchange without phase transfer; CdS QDs remained in the cyclohexane layer. Each experiment was reproduced at least 3 times.

Table 1 shows that TMA countercation gave very good yields of Thr-CdS QDs by OA→Thr phase transfer ligand exchange in MeOH and NMF, as determined by the UV-vis absorption spectroscopy of L-Thr-TMA⁺-CdS QDs. However, a moderate yield of Thr-CdS was obtained in FA and no phase transfer was observed into DIW, DMSO, DMF and MeCN (the CdS QDs either crashed out of the solution or remained in the cyclohexane layer). The ¹H NMR data revealed that L-Thr as well as L-Thr⁻-TMA ion pair were insoluble in DMSO (Table S2). TMA⁺ cation thus promoted the ligand exchange and phase transfer in protic solvents (with the notable exception of DIW, as previously reported)[35], whereas it failed to do so in aprotic solvents due to low stability/solubility of L-Thr-TMA⁺-CdS QDs.

It is important to recall, that a successful phase transfer ligand exchange synthesis depends on several steps: (i) transfer of the deprotonated and negatively charged Thr molecules in an ion pair with positively charged alkylammonium countercation from the polar solvent layer to the phase boundary or to the nonpolar cyclohexane layer, (ii) substitution of OA by Thr on the CdS QD surface, and finally (iii) the reverse transfer of the Thr-CdS QDs into the polar solvent layer. Since the amphiphilicity of the negatively charged Thr can be altered by its positively charged counterion (i.e., the cation of the hydroxide base used for the deprotonation of the Thr), we decided to test other surfactant bases, namely a more hydrophobic tetrabutylammonium hydroxide (TBAH) and a more hydrophilic ammonium hydroxide (NH₄OH). Importantly, TBA⁺ and NH₄⁺ cations are chemically inert to the QDs and do not functionalize QDs’ surface. TBAH have been used previously for the ligand exchange synthesis of achiral water soluble QDs.^[52–55] As can be seen from Table 1, TBA⁺ cation was more effective in phase transfer ligand exchange than TMA⁺ cation and promoted OA→Thr ligand exchange with all solvents except DIW, with yields ≥70% for protic and ≥85 for aprotic solvents. ¹H NMR spectra have been used to confirm successful OA→Thr ligand exchange on CdS QDs (Figure S16) whereas the ATR FTIR spectra have been used to characterize the L-Thr-CdS QDs (Figure S17). NH₄⁺ cation gave lower yields than TBA⁺ (40-80% for protic and 35-45% for aprotic solvents) and failed to promote the ligand exchange with two solvents, DIW and MeCN. Powder XRD and TEM data of the L-Thr-CdS and OA-CdS QDs (Figure 1) confirmed that the structure of the QD’s core has not been changed and that the size distribution has not been altered (Figure S18). Our ¹H NMR data showed that the increased hydrophobicity of the Thr⁻-TBA⁺ ion pair led to its increased partition into the cyclohexane layer in comparison to the more hydrophilic Thr⁻-TMA⁺ (Table S2). It is reasonable to assume that higher concentration of Thr in cyclohexane layer shifted the equilibrium towards substitution of the OA ligand by the Thr ligand.

Figure 1.

TEM images of (a) L-Thr-TMA⁺-CdS QDs prepared in NMF and (b) L-Thr-TBA⁺-CdS QDs prepared in DMF. (c) Powder XRD patterns of OA-CdS QDs (dashed orange line), L-Thr-TMA⁺-CdS QDs prepared in NMF (green line), L-Thr-TBA⁺-CdS QDs prepared in DMF (red line), CdS (cubic) reported in the ICDD powder diffraction files (No. 01-089-0440, blue line).

3.2. Solvent and countercation (base) effect on the induced CD and CD anisotropy

While the achiral OA-CdS QDs did not display any CD signal, functionalization of CdS QDs with chiral L-Thr ligands made them CD active. As expected, the solvent and the countercation (base) used in the phase transfer ligand exchange had a strong effect on the circular dichroism (CD) spectra of L-Thr-CdS QDs. Both the shape of the CD spectra and the CD anisotropy factor (g_CD) displayed large alteration depending on the solvent and countercation used in the ligand exchange (Figure 2 and Table 2). The anisotropy factor g is defined as g_CD = Δε/ε = (A_L - A_R)/A where A represents the conventional absorbance of non-polarized light and A_L and A_R are the absorptions of left and right circularly polarized light, respectively. The g_CD is concentration independent if the CD and absorption spectra are measured on the same sample. The g_CD maxima of L-Thr-CdS QDs prepared by phase transfer ligand exchange into different polar protic and aprotic solvents using TMA⁺, TBA⁺ and NH₄⁺ countercations are summarized in Table 2 (see Figures S4–S6 for the g_CD plots).

Figure 2.

(a) CD spectra of L-Thr-TBA⁺-CdS QDs prepared by phase transfer ligand exchange from OA-CdS QDs in cyclohexane into polar solvents using TBAH as a base, (b) Comparison of the CD spectra of L-Thr-TBA⁺-CdS QDs in MeCN and FA. Each ligand exchange experiment has been repeated at least 3 times followed by CD spectra measurements and representative spectra are shown.

Table 2.

Anisotropy g-factors (g_CD) of L-Thr-CdS QDs prepared by phase transfer ligand exchange from OA-CdS QDs using various polar solvents and ammonium countercations.

	gCD (× 10−4) /λ (nm)
	L-Thr-TMA+	L-Thr-TBA+	L-Thr-NH4+
DIW	no transfer a	no transfer b	no transfer a
MeOH	3.5 (0.03) c / 400.5	2.9 (0.03) / 401.0	6.8 (0.04) / 387.0
NMF	1.0 (0.05) / 399.5	0.5 (0.08) / 399.0	3.1 (0.03) / 387.5
FA	−0.3 (0.09) / 371.0	−1.1 (0.05) / 373.5	1.8 (0.05) / 389.5
DMSO	-- d	−0.2 (0.10) / 402.0	2.3 (0.02) / 387.0
DMF	-- d	0.3 (0.10) / 405.0	1.1 (0.04) / 393.5
MeCN	-- d	0.7 (0.08) /377.0	no transfer b

^aNo phase transfer.

^bPartial ligand exchange without phase transfer; chiral CdS QDs in the cyclohexane layer.

^cNumbers in parentheses designate standard deviation (n≥3).

^dQDs crashed out of the solution.

The data showed that larger CD anisotropies g_CD have been induced in polar protic solvents (MeOH, NMF, FA) than their aprotic counterparts (DMSO, DMF, MeCN) regardless of the countercation/base used. From all the solvents explored, the largest g_CD of L-Thr-CdS QDs have been detected for MeOH for all three countercations, TMA⁺ (3.5 × 10⁻⁴), TBA⁺ (2.9 × 10⁻⁴), and NH₄⁺ (6.8 × 10⁻⁴). Weaker chiral induction was observed for L-Thr-CdS QDs prepared in the three aprotic solvents, with largest g_CD calculated for L-Thr-NH₄⁺-CdS QDs in DMSO (2.3 × 10⁻⁴) and DMF (1.1 × 10⁻⁴). No direct correlation was observed between polarity of the protic or aprotic solvents and CD anisotropy, further underlying the importance of the chemical structure of the solvent and the solvent-ligand-countercation intermolecular interactions during the ligand exchange.

The countercation/base used in the ligand exchange also had a very strong impact on CD anisotropy, as can be illustrated on the example of the L-Thr-CdS QDs in NMF (Table 2) where TMA⁺ gave g_CD = 1.0 × 10⁻⁴ (399.5 nm), TBA⁺ gave g_CD = 0.5 × 10⁻⁴ (399.0 nm), and NH₄⁺ gave g_CD = 3.1 × 10⁻⁴ (387.5 nm); i.e., a six-fold difference in anisotropy was observed as a direct effect of the countercation. Overall, while the largest and most non-polar TBA⁺ cation was the most versatile and enabled the synthesis of L-Thr-functionalized CdS QDs in diverse solvents with a wide range of polarity, from MeOH (ε= 33.0) to NMF (ε= 189.0), it was the NH₄⁺ cation, the smallest and most polar, that led to the formation of QDs with the highest CD anisotropy g_CD. In summary, L-Thr-functionalized CdS QDs with the highest calculated anisotropy factor (g_CD 6.8 × 10⁻⁴) were synthesized in MeOH, a polar protic solvent, and in the presence of the NH₄⁺ countercation.

It is also important to emphasize that the solvent and countercation used for the ligand exchange did not just influence the intensity of the CD signal but had an equally strong effect on the shape of the CD spectra (Figure 2a), i.e., the position of positive and negative CD bands (Cotton effects).

The striking effect of the solvent on the chirality of QDs can be illustrated by comparing the CD spectra of L-Thr-TBA⁺-CdS QDs in protic FA and aprotic MeCN that displayed mirror-image looking CD spectra (Figure 2b) indicating mirror-image-like binding geometries of L-Thr capping ligands on the surface of CdS QDs. Similarly pronounced impact on the shape of the CD spectra was observed by varying the countercation (Figures S7–S11) as can be illustrated on the example of the CD spectra and g factor plots of L-Thr-CdS QDs in NMF prepared with TMA⁺, TBA⁺, and NH₄⁺ cations where a partial inversion of the CD signal was observed (Figure 3a). Despite the dynamic nature of the binding (equilibrium between bound and free ligands) and weak interaction interactions between the carboxyl group of Thr ligands and CdS QDs, the Thr-induced CD anisotropy of CdS QDs was comparable to the CD anisotropy induced by tightly bound chiral thiol ligands. Therefore, we anticipate that the chirality of CdS QDs and the observed CD signal originated from the hybridization of the (achiral) QD and (chiral) ligand electronic states. However, due to the complexity of the problem and lack of consensus of the field on the origin of ligand induced chirality in QDs, it is not possible to provide a more in-depth analysis at the present time.

Figure 3.

(a) CD spectra and (b) g_CD plots of L-Thr-TMA⁺-CdS QDs, L-Thr-TBA⁺-CdS QDs, and L-Thr-NH₄⁺-CdS QDs prepared by phase transfer ligand exchange from OA-CdS QDs into NMF using different countercations. Each ligand exchange experiment has been repeated at least 3 times followed by CD spectra measurements and representative spectra are shown.

3.3. Solvent and countercation effects on the absorption spectra and band gaps

The solvent and the countercation used in the phase transfer ligand exchange (all other experimental parameters have been kept constant) also had a pronounced effect on the absorption spectra of L-Thr-CdS QDs and the position of the first exciton peaks (Figures 4, S1–S3, and Table 3). L-Thr-TBA⁺-CdS QDs displayed bathochromic shift of the 1^st exciton peak with polar aprotic solvents inducing larger shifts (from +6.4 nm/+47.6 meV to +11.2 nm/+82.4 meV) than their polar protic counterparts (from +0.2 nm/+1.5 meV to +6.2 nm/+46.2 meV). For the polar aprotic solvents, the increased chemical polarity (i.e., relative permittivity, see Table 1) of the solvent correlated with larger red shift of Thr-TBA⁺-CdS QDs. However, no noticeable correlation was evident for protic solvents even though the largest shift was again induced by the most polar protic solvent, NMF. Overall, it is apparent that the ability of the solvent to donate/accept H-bond has a more pronounced impact on the band gap of CdS QDs (as determined by the shift of the 1^st exciton peak) than the polarity of the solvent. As can be seen from the Table 3 and Figures S1–S3, L-Thr-TMA⁺-CdS QDs and L-Thr-NH₄+-CdS QDs either displayed smaller bathochromic shifts than L-Thr-TBA⁺-CdS QDs or even a hypsochromic shifts of the 1^st exciton peak. Only hypsochromic shifts have been detected for L-Thr-NH₄⁺-CdS QDs in polar aprotic solvents. As was the case with TBA⁺ cation, no apparent correlation was observed for TMA⁺ and NH₄⁺ cations for protic or aprotic solvents.

Figure 4.

Normalized absorption spectra of the lowest energy exciton peaks of L-Thr-TBA⁺-CdS QDs prepared by ligand exchange into various polar solvents in comparison to OA-CdS QDs in cyclohexane (dashed orange curve). Each ligand exchange experiment has been repeated at least 3 times followed by UV-vis absorption spectra measurements.

Table 3.

Wavelength (nm) and energy (meV) shifts of the lowest energy exciton peak of L-Thr-CdS QDs relative to the OA-CdS QDs starting material.

solvent	TMA+	TBA+	NH4+
	Δ nm (meV)	Δ nm (meV)	Δ nm (meV)
DIW	no transfer a	no transfer b	no transfer a
MeOH	+1.6 (+12.1)	+1.6 (+12.1)	−0.6 (−4.5)
NMF	+5.4 (+40.3)	+6.2 (+46.2)	0 (0)
FA	−1.4 (−10.6)	+0.2 (+1.5)	+0.8 (+6.0)
DMSO	-- c	+11.2 (+82.4)	−2.2 (−16.7)
DMF	-- c	+7.6 (+56.4)	−1.0 (−7.6)
MeCN	-- c	+6.4 (+47.6)	no transfer b

^aNo phase transfer.

^bPartial ligand exchange without phase transfer, CdS QDs remained in the cyclohexane layer.

^cQDs crashed out of the solution.

3.4. Ligand exchange without phase transfer: chiral Thr-(OA)-CdS QDs in a highly nonpolar solvent

Further examination of cyclohexane layer from the failed ligand exchange experiments (see Table 1) where CdS QDs remained in the cyclohexane layer (as was confirmed by UV-vis and emission spectra, Figures S12–S15) revealed that cyclohexane layer from the MeCN-NH₄OH and DIW-TBAH experiments displayed CD signal in the 320-440 nm spectral region (Figure 5), whereas the cyclohexane from the DIW-NH₄OH and DIW-TMAH experiment were CD silent.

Figure 5.

CD spectra of L- and D-Thr-(OA)-CdS QDs in cyclohexane layer for (a) DIW-TBAH and (b) MeCN-NH₄OH ligand exchange experiments. Each ligand exchange experiment has been repeated at least 3 times followed by CD spectra measurements

The presence of the CD signal together with the fact that the CdS QDs were soluble in cyclohexane suggested partial substitution of achiral OA ligands on the surface of CdS by chiral L-Thr ligands, and we labelled them as L-Thr-(OA)-NH₄⁺-CdS QDs and L-Thr-(OA)-TBA⁺-CdS QDs. To confirm that the CD spectra originated from the ligand-QD interactions, we have also performed the ligand exchange experiments with enantiomeric D-Thr and obtained mirror-image CD spectra (Figure 5a,b). The observation of mirror-image CD spectra provided additional evidence of the direct bonding of chiral Thr ligands to the OA-CdS nanocrystal surface.

The comparison of emission spectra of L-Thr-(OA)-NH₄⁺-CdS QDs in cyclohexane (from cyclohexane-MeCN mixture) and OA-CdS QDs in cyclohexane revealed that while the fluorescence spectrum of OA-CdS QDs in cyclohexane (λ_exc = 400 nm) displayed a sharp near band-edge emission peak at 442 nm and a deep-trap emission band between 500 and 750 nm (Figures 6, magenta curve) and emitted mauve light under the UV lamp irradiation (Figure 6 inset, sample a), the emission spectrum of L-Thr-(OA)-NH₄⁺-CdS QDs in cyclohexane (cyclohexane layer from cyclohexane-MeCN reaction mixture) showed a decreased band-edge emission (Figure 6, orange curve) and emitted an orange glow under UV light (Figure 6 inset, sample c). To complete the comparison, a fluorescence spectrum of L-Thr-TBA⁺-CdS QDs in MeCN (MeCN layer from cyclohexane-MeCN reaction mixture) has also been measured and revealed almost no fluorescence signal (Figure 6, dark cyan line) and no emission was detected in the MeCN layer under the UV light (Figure 6 inset, sample b). Fluorescence data thus presented additional evidence for the partial functionalization of the CdS QD surface by the L-Thr capping ligands in L-Thr-NH₄⁺-CdS QDs in cyclohexane.

Figure 6.

Emission spectra (λ_exc = 400 nm) of (a) OA-CdS QDs (in cyclohexane; magenta curve), (b) L-Thr-TBA⁺-CdS QDs (in MeCN layer from cyclohexane-MeCN reaction mixture; dark cyan curve), and (c) L-Thr-(OA)-NH4⁺-CdS QDs (in cyclohexane layer from cyclohexane-MeCN reaction mixture; orange curve) taken at identical concentrations. Inset: photographs taken under ambient light (left) and UV light (365 nm; right) of (a) OA-CdS QDs in cyclohexane layer over MeCN layer before the ligand exchange reaction, (b) L-Thr-TBA⁺-CdS QDs in MeCN underneath the cyclohexane layer after the ligand exchange reaction, and (c) L-Thr-(OA)-NH4⁺-CdS QDs in cyclohexane layer over the MeCN layer after the ligand exchange reaction. Each ligand exchange experiment has been repeated at least 3 times followed by emission spectra measurements.

The optical and chiroptical characteristics of Thr-(OA)-CdS QDs in cyclohexane layer are summarized in Table 4 and showed stronger anisotropy and a smaller shift of the 1^st exciton peak for L-Thr-(OA)-NH₄⁺-CdS QDs (from cyclohexane-MeCN) than L-Thr-(OA)-TBA⁺-CdS QDs (from cyclohexane-DIW). The low CD anisotropy of L-Thr-(OA)-TBA⁺-CdS QDs likely originated from a smaller partition of Thr from DIW into cyclohexane. The data show that (i) a complete exchange of a nonpolar ligand for a polar one is not necessary for the successful induction of chirality in QDs, (ii) combination of chiral and achiral ligands can be used to induce chirality in QDs, and (iii) chiral QDs can be prepared in highly non-polar solvents using a mixture of polar and non-polar capping ligands. These are important observations because most commercially available chiral capping ligands are polar organic compounds

Table 4.

Wavelength (nm) and energy (meV) shifts of the 1^st exciton peak (relative to the OA-CdS QDs) and the anisotropy factors g_CD of L-Thr-(OA)-NH₄⁺-CdS QDs (from cyclohexane-MeCN) than L-Thr-(OA)-TBA⁺-CdS QDs (from cyclohexane-DIW) in cyclohexane.

reaction mixture	Δ nm (meV)	gCD (× 10−4) /λ (nm)
cyclohexane / DIW–TBAH	+ 3.8 (+28.5)	−0.1 / 383.5
cyclohexane / MeCN–NH4OH	−0.2 (−1.5)	−1.8 / 375.5

4. Conclusions

We have shown that chiral L-threonine-functionalized cadmium sulfide quantum dots (L-Thr-CdS QDs) can be synthesized in polar protic as well as aprotic solvents by phase transfer ligand exchange from OA-CdS QDs using an appropriate ammonium countercation. A significance of the present study is that we have shown for the first time that the yield of the ligand exchange, the chiroptical properties (CD anisotropy), and optical properties (1^st excitonic band shift) of Thr-CdS QDs prepared by ligand exchange correlated with the ability of the solvents to H-bond (i.e., protic vs. aprotic solvent). At the present time, we could not draw a correlation with the polarity of the solvents and our observations. Tetrabutylammonium cation (TBA⁺) proved to be the most versatile base (almost ‘universal cation’) giving effective ligand exchange and high yields of chiral L-Thr-CdS QDs in all tested polar solvents except DIW. It is notable that TBA⁺ cation enabled the synthesis with a polar capping ligand in very diverse solvents: we prepared QDs in the most polar organic solvent, N-methylformamide (NMF, ε= 189.0), as well as in the least polar solvent, cyclohexane (ε= 2.0). On the other hand, it is the use of NH₄⁺, the smallest and most polar countercation during synthesis that yielded the highest calculated anisotropy factor g_CD in L-Thr-functionalized CdS QDs. L-Thr-TBA⁺-CdS QDs displayed bathochromic shift of the 1^st exciton peak for aprotic (up to 11.2 nm, 82.4 meV) as well as protic (up to 6.2 nm, 46.2 meV) solvents. Thr-CdS QDs displayed larger anisotropies g_CD in protic solvents than in aprotic solvents for all countercations, and the largest g_CD was calculated for Thr-CdS QDs in the protic solvent MeOH for all three countercations. Confirming our hypothesis that solvents and countercations would affect the ligand binding geometry as well as the induced chiroptical properties of CdS, we have also shown that different solvents and different ammonium countercations gave rise to different, in some cases even inverted, circular dichroism (CD) spectra. An examination of cyclohexane layer from the experiments where the CdS QDs did not transfer into the polar solvent revealed some of them to be chiral and optically active, thus outlining an approach for the synthesis of chiral QDs in non-polar solvents using a mixture of polar and non-polar capping ligands. Overall, we highlighted the critical role the solvent and the countercation played in (i) the interaction between chiral capping ligand L-Thr and the surface of the CdS QD, (ii) stability and solubility of Thr-CdS QDs, (iii) the energy of the band gap and the 1^st exciton peak, and (iv) the shape and anisotropy of the induced CD signal of L-Thr-CdS QDs. Our data demonstrated that chiral QDs can be synthesized in a wide variety of organic solvents not only via the use of chemically different weakly bound capping ligands but also by the choice of the ammonium countercations via a ‘mix and match’ approach. Given the weak electrostatic interaction between the thiol-free carboxylic ligand Thr and the surface of CdS QD together with the dynamic nature of the binding (equilibrium between bound and free ligands), we anticipate the chirality of Thr-functionalized CdS QDs and the observed CD signal likely originated from the hybridization of the (achiral) QD and (chiral) ligand electronic states, rather than chiral surface defects. Presented data further advance our understanding of induced chirality in semiconductor nanocrystals and outline prospects towards rational design of tailormade chiral materials.