Slow Anion-Exchange Reaction of Cesium Lead Halide Perovskite Nanocrystals in Supramolecular Gel Networks.

Cesium lead halide perovskite nanocrystals are widely studied as among the most attractive emissive nanomaterials because of their high photoluminescence quantum yield and tunable emission wavelengths over the whole visible-light region by the halide ion-exchange reaction. However, the reactions were often observed in solution and generally very fast, which interferes with the fine-tuning capability of the emission properties. Here, we report a novel nanocrystal-organogel hybrid soft material in which the perovskite nanocrystals in a supramolecular gel exhibit extremely slow and inhomogeneous anion-exchange reactions that are different from those in solution. Furthermore, the inhomogeneous emission in the gel became homogeneous over several days due to a slow diffusion.

Introduction

Semiconductor nanocrystals (NCs), known as quantum dots, are among the most attractive photoluminescent materials because of their size-dependent emission wavelengths, narrow emission line width, and high photodurability.[1,2] Among the semiconductor NCs, cesium lead halide (CsPbX3, X = Cl, Br, I) ternary perovskite NCs exhibit high photoluminescence quantum yield and halide ion-exchange reactions that enable tuning of the emission wavelengths over the whole visible-light region.[3−5] Since Kovalenko and co-workers developed a method to synthesize NCs,[3] the optical properties that are tunable by the anion-exchange reaction have been actively investigated at ensemble levels in solution.[6−12] Recently, an in situ observation of emission behaviors during the anion-exchange reaction at the single NC level has been revealed.[13] However, the anion-exchange reaction is generally very fast, which interferes with the fine-tuning capability of the emission color. The fast reaction can be controlled by increasing the viscosity, for example, via gelation. Although gels containing semiconductor NCs such as CdSe and CdSe/ZnS NCs have been studied to obtain soft materials,[14−18] the anion-exchange reaction of perovskite NCs in gels has not been revealed so far. On the basis of this hypothesis, we demonstrate a novel nanocrystal-supramolecular gel hybrid system in which CsPbBr3 perovskite NCs in a supramolecular gel exhibit extremely slow and inhomogeneous anion-exchange reactions that are substantially different from those in solution. Moreover, the inhomogeneous emission area in the gel became homogeneous over several days due to the slow diffusion of ligands and NCs.

Results and Discussion

CsPbBr3 perovskite NCs (Br-NCs) covered with the surface-protecting ligands of oleic acid and oleylamine were synthesized according to the reported method (see the Experimental Section).[3] The UV/vis absorption spectrum of a diluted cyclohexane solution (SolBr-NC, c = 0.01 μM) showed a first peak at 500 nm (Figure S3a), by which the size of Br-NC was estimated to be approximately 7.4 nm.[19] The size is consistent with the transmission electron microscopic (TEM) analysis of the Br-NCs (7.3 ± 0.8 nm, Figure S3b). The photoluminescence (PL) spectrum under 405 nm excitation showed a blue–green emission at 510 nm with a narrow full width at half-maximum (FWHM) of 25 nm (Figure S3a).

To conduct the anion-exchange reaction, we selected benzoyl halides as halide precursors,[20] and then benzoyl iodide in cyclohexane (SolI, c = 82 mM) and benzoyl chloride in cyclohexane (SolCl, c = 86 mM) were prepared. Upon adding 20 μL of SolI to 3 mL of SolBr-NC, the emission intensity decreased significantly without a redshift of the emission wavelength (Figure a, Figure S4a). The absorption spectra showed a slight redshift of a first peak with increasing absorption around 550 nm (Figure S4b), indicating the occurrence of the anion-exchange reaction but not decomposition. After several hours, weak-emissive film-like precipitates were formed at the bottom of the solution (Figure S4c). These unexpected PL decrease and precipitation are due to the agglomeration of NCs. In this process, benzoyl iodide reacted with the ligands on the NCs to release I–, thus forming ligand-free NCs, which give rise to defects. As a result, the NCs underwent agglomeration due to the low solubility in cyclohexane. However, the addition of oleic acid and oleylamine to the precipitates enabled the emergence of an orange emission with a maximum PL wavelength (λPL) at 594 nm, corresponding to an I-rich NC, that is, restoration of the NCs (Figure a,c). To hinder ligand removal from the surface of NCs, we prepared a Br-NC solution containing 2 μL of oleic acid and 2 μL of oleylamine (SolBr-NC*). The PL intensity increased 2-fold, indicating that the number of defects decreased due to the presence of ligands (Figure S5). As expected, the anion-exchange reaction in this condition occurred without precipitation (Figure b), and the changes were monitored by PL spectroscopic measurements. Upon adding 20 μL of SolI to 3 mL of SolBr-NC* ([Br]parent/[I]incoming = 1:8) and shaking, λPL redshifted to 652 nm within several minutes (Figure c). The PL decay curves showed that the average lifetime of Br-NC* at τ = 5.7 ns changed to 18 ns (Figure S6), suggesting the formation of the I-rich NC* by the anion-exchange reaction. On the other hand, the addition of SolCl to SolBr-NC* ([Br]parent/[Cl]incoming = 1:8) resulted in a blueshift of the PL wavelength to 414 nm (Figure c) with a decrease in the average lifetime to τ = 1–2 ns (Figure S6). The change clearly indicates the formation of the Cl-rich NC*.

Figure 1

Schematic of the cyclohexane solutions of (a) Br-NC and (b) Br-NC* upon the anion-exchange reaction by adding benzoyl iodide (I–). Here, the asterisk indicates the addition of oleic acid and oleylamine. The vials are placed under a UV lamp. (c) PL spectra of Br-NC and Br-NC* in cyclohexane upon the addition of benzoyl iodide and benzoyl chloride.

Although the emission color change of Br-NC* in solution by the anion-exchange reaction occurred quickly, we expected the color change in the gel to be slow due to the high viscosity. Here, we selected a supramolecular gel that is categorized as a physical gel and shows the reversible sol–gel transition as a function of temperature.[21] As an organogelator suitable for hydrophobic NCs, cholesterol-functionalized diamide molecule 1 (Figure a) was newly synthesized through three steps (Scheme S1, Figures S1 and S2, see the Experimental Section). In this molecular design, the diamide is introduced for double hydrogen-bonding, and the cholesterol moiety can act as an auxiliary aggregation moiety,[22−25] and the long alkyl chain is functionalized for the enhancement of solubility in organic solvents. We first attempted a gelation experiment in cyclohexane. Upon cooling the hot solution (c ≥ 3 mM, 2 g/L), a transparent gel was formed (Figure d and Figure S7). To observe the morphology, the gel was heated gently at 60 °C and then spin-coated onto a carbon-coated grid for TEM measurements. The TEM images displayed the formation of helical fibers and their intertwined network structure (Figure b,c and Figure S8). To prepare a supramolecular gel containing emissive Br-NC*s (GelBr-NC*), we added 9.0 mg of 1 to 1.5 mL of SolBr-NC* and subjected the mixture to heating and cooling to obtain emissive GelBr-NC* (Figure g). Owing to the characteristics of the supramolecular gel, the emissive gel showed a temperature-driven reversible sol–gel transition (Figure S9). The PL behaviors of GelBr-NC* showed almost the same properties as those of SolBr-NC*, which suggests that neither interactions between Br-NCs nor a deterioration of Br-NCs occurred in the gel (Figure S10). In addition, the PL of Br-NCs in the gel was stable for at least several months, probably due to the presence of sufficient ligands and/or being in the highly viscous state in which the motion of NCs becomes slower than that in solution. The TEM images of a sample that was spin-coated from gently heated GelBr-NC* also displayed Br-NCs dispersed in the supramolecular fiber network (Figure e,f). These NCs were not adsorbed to the fibers because 1 does not have a moiety that adheres to NCs. Therefore, the NCs in the gel can diffuse slowly.

Figure 2

(a) Molecular structure of compound 1. (b) TEM image of the fibrous structures of the supramolecular gel of 1. (c) Magnified TEM image of helical fibers. (d) Photograph of the gel. (e) TEM image of fibrous structures containing Br-NCs. The inset shows magnified TEM image of Br-NCs. (f) Magnified TEM image of dispersed Br-NCs with a fiber. (g) Photograph of the emissive gel containing Br-NCs under UV light.

To perform an anion-exchange reaction in the gel, 10 μL of SolI was dropped on top of GelBr-NC* and then monitored for a change in emission colors under excitation at 365 nm. The method of the anion exchange was defined as Method 1. In sharp contrast to the fast and homogeneous anion-exchange reaction in solution, an inhomogeneous change in emission colors was observed at a local reaction area, and the area migrated downward with time (Figure a). In the local area, various emission colors from green (λmax = 521 nm) to red (λmax = 642 nm) depending on the degree of anion-exchange reaction were detected by spatially resolved PL spectroscopic measurements under excitation with a 405 nm continuous wave (CW) laser (Figure S11a). To our knowledge, this is the first demonstration of an anion-exchange reaction in a supramolecular gel exhibiting a unique emission behavior.

Figure 3

Photographs of the emissive gel containing Br-NCs under UV light after the addition of (a) SolI and (c) SolCl. The vials are placed under a UV lamp. Time-dependent PL spectral changes in areas A and B after the addition of (b) SolI or (d) SolCl (λex = 405 nm).

To obtain insight into the unique anion-exchange reaction in the gel, the spatially resolved time-dependent emission spectra at two local areas, A and B, which are located at approximately the top and middle, respectively, were measured (Figure a). At area A after dropping the SolI, the PL intensity of Br-NC at λmax = 510 nm gradually decreased with time; weak PL peaks at approximately 650 nm corresponding to I-rich NC emerged at t = 0–5 min, and then the intensity decreased after t = 5 min (Figure b). These results suggest that a large number of Br-NCs in area A were quenched without the preferential formation of I-rich NCs. The PL quenching at area A results from the formation of ligand-free NCs. In this area, the fraction of benzoyl iodide with respect to Br-NC is too large compared to that of the homogeneous solution (SolBr-NC*), leading to the formation of nonemissive NCs. In contrast, area B exhibited a steady redshift of the PL spectrum through broadening of the PL spectra from t = 205 min (Figure c). The broadened spectra indicate the formation of a mixture of mixed-halide CsPb(Br/I)3 NCs with various ratios of Br/I. At t = 1650 min, a sharp peak (FWHM = 42 nm) with λmax = 660 nm was detected, suggesting the preferential formation of I-rich NCs. The movement of the reaction area to the bottom was completed within 1 day, indicating that this time corresponds to the diffusion time of anions from top to bottom in the gel. After several days, notably, the quenched areas containing area A showed a restoration of red emission corresponding to I-rich NCs, and the inhomogeneous gel became homogeneous (Figure a and Figure S11b). The PL of I-rich NCs in the gel was maintained for at least several weeks. In the case of the anion-exchange reaction with benzoyl chloride in the gel, a similar slow and inhomogeneous PL color change with a blueshift was observed followed by the gel changing into a homogeneously emissive gel after 4 days (Figure c,d and Figure S11b). Considering that the lack of ligands on the NC surface could induce the PL quenching, as observed in solution, the observed restoration behavior in the gel is probably due to the adhesion of free ligands onto quenched NCs via long-term diffusion. In addition, NCs can move into the top area as well as the bottom area during this long-term diffusion. The slow diffusion of NCs in the gel was probably due to the suppression of the NC movement by the fiber networks. Accordingly, the restored homogeneous emission over a long time resulted from the slow diffusion of ligands and/or NCs in the gel.

The anion-exchange reaction can also occur by mixing perovskite NCs with different halogens.[6,7,12] Thus, we synthesized pure CsPbI3 perovskite NCs (I-NCs) exhibiting red emission at 674 nm (see the Experimental Section).[3] To observe the anion-exchange reaction between NCs in gels, we prepared supramolecular gels containing emissive Br-NCs (GelBr-NC) and I-NCs (GelI-NC). Upon contacting the gels (defined as Method 2), a change in the emission color at the contacted area was observed within several minutes (Figure a). This result suggests an anion-exchange reaction between Br-NCs and I-NCs in the gel. As shown from the results, the anion exchange by Method 2 occurred from the middle area located at the interface between the gels, while the anion exchange by Method 1 occurred from the top of the gel. In contrast to Method 1, free halide ions do not exist in the gel, and thus, the anion exchange by Method 2 requires an approach between Br-NC and I-NC, in which it caused the shuttling of halide ions between approached NCs. Due to the slow diffusion of NCs in the gel, a slow and inhomogeneous change in emission color was observed over several days, as confirmed by the spatially resolved time-dependent emission spectra at the three local areas A, B, and C, which are located at approximately the top, middle, and bottom, respectively (Figure b). Unlike the case of anion exchange in gels triggered by adding halide precursors (Method 1), this anion exchange between NCs did not induce the formation of a nonemissive gel area, indicating that the surface condition of NCs was maintained. During the reaction, in area B, relatively fast spectral changes were observed, but area A containing I-NCs underwent a slow blueshift by 155 nm for several days as a result of an increase in the Br fraction. On the other hand, area C containing Br-NCs showed a slight redshift by 10 nm (Figure b). After 5 days, a homogeneous gel showing green emission at approximately 520 nm was obtained (Figure a,b). These results indicate the formation of Br-rich CsPb(Br/I)3 NCs in the gel via the anion-exchange reaction induced by the long-term diffusion of NCs. As the timescale to become homogeneous in this case is similar to the case of anion exchange in gels triggered by adding halide precursors (Method 1), we concluded that the anion-exchange reaction in gels depends on the diffusion speed of the NCs.

Figure 4

(a) Photographs of the emissive gels containing Br-NCs and I-NCs during the anion-exchange reaction. The vials are placed under a UV lamp. The upper blue light is a scattered light. (b) Time-dependent PL spectral changes in areas A, B, and C after contacting GelI-NC and GelBr-NC (λex = 405 nm).

Conclusions

In summary, we demonstrated the very slow change in the emission color of CsPbBr3 perovskite NCs embedded in a supramolecular gel in comparison to that in solution. In a cyclohexane solution, the emission color change by the anion-exchange reaction using benzoyl halides occurred at a quite fast speed within several seconds. The anion-exchange reaction induced PL quenching of NCs due to the decrease in the number of ligands on the NCs, but the addition of ligands to the solution can allow the NCs to maintain the PL intensity. In contrast, the NCs in supramolecular gel networks underwent slow and inhomogeneous emission color changes via the formation of PL-quenching areas and shifted emission areas. After several days, the inhomogeneous emission gels became homogeneous as a result of the slow diffusion of NCs. Slow anion exchange was also observed by making contact between different emissive gels containing CsPbBr3 NCs and CsPbI3 NCs. These results suggest that slow anion exchange can be realized by gelation. Furthermore, our results suggest the importance of the gelation of NCs, which can lead to the creation of new hybrid soft materials exhibiting unique behaviors depending on the viscosity.

Experimental Section

General Methods

1H and 13C NMR spectra were recorded on a JEOL JNM ECX-500II NMR spectrometer, and the chemical shifts are reported in ppm (δ) with the signal of TMS (δ = 0) as the internal standard. High-resolution mass spectra were obtained with a Bruker Daltonics microTOF-Q spectrometer (ESI). UV/vis absorption and photoluminescence (PL) spectra were recorded on a Shimadzu UV-3600 spectrophotometer and a Hitachi F-7000 spectrofluorometer, respectively. Transmission electron microscopic (TEM) observation was carried out on a Tecnai G2 F20 (FEI) at acceleration voltage at 200 kV. The samples were prepared by spin-coating solutions onto a carbon-coated STEM Cu grid (SHR-C075) and dried under vacuum for 1 h. The TEM observations were conducted without staining.

PL Decay Curve Measurements[26]

The PL decay curves of the solution and gel were measured with picosecond-pulsed laser excitation at 405 nm (10.0 MHz, 90 ps full width at half-maximum) under an inverted confocal microscope (Olympus, IX-71). The beam was reflected by a dichroic mirror (Semrock, Di01-R405) and then focused to the sample in a quartz cuvette on a sample stage by an objective lens (Olympus, 20×, NA 0.4), and the PL photons emitted from the sample were collected by the same objective lens and passed through a confocal pinhole (100 μm) and a longpass filter (Semrock, LP02-442RU) to remove the excitation laser. The photons were detected by an avalanche photodiode (APD) single-photon counting module (PerkinElmer, SPCM-AQR-14). The signal from the APD was connected to a time-correlated single-photon counting PC board (Becker & Hickl, SPC630) to determine the PL lifetime. The time resolution of the lifetime measurement (IRF) was approximately 0.3 ns. All measurements were performed at room temperature under ambient conditions.

Spatially Resolved PL Spectroscopic Measurements

The PL spectra of the solution and gel during the anion-exchange reaction were measured by a combination of laser excitation (405 nm) and fiber optic spectrometer (Ocean Optics, USB400). A 405 nm CW laser in a diameter of 3 mm was used to irradiate the sample in a quartz cuvette, and the PL from the sample was detected through the optical fiber. PL in different areas was detected by moving the cuvette.

Anion-Exchange Reaction

The anion-exchange reaction in solution was performed by the following procedure. First, we estimate the number of Br atoms in a single CsPbBr3 NC (Br-NC). The length of one side in our cubic-shaped Br-NC is estimated to be 7.3 nm from the TEM image. The lattice constant of the NC was reported to be 5.8 Å.[19] Thus, a single Br-NC includes approximately 6000 Br atoms. Therefore, when a 3 mL solution of Br-NC ([Br-NC] = 0.012 μM) is used, the number of Br moles is estimated to be 2.2 × 10–7 mol. Accordingly, upon the addition of a 20 μL benzoyl iodide cyclohexane solution (c = 82 mM), in which 1.6 × 10–6 mol of I– is released into the 3 mL Br-NC solution, the [Br]parent/[I]incoming ratio is estimated to be approximately 1:8. For the addition of a 20 μL benzoyl chloride cyclohexane solution (c = 86 mM) into the 3 mL Br-NC solution, the [Br]parent/[Cl]incoming ratio is estimated to be approximately 1:8. For the anion-exchange reaction in the gel, we calculated the ratio based on the method mentioned above.

General Materials

Column chromatography separations were performed using silica gel 60 N (spherical, neutral, particle size 63–210 μm; Kanto Chemical). All commercially available reagents and solvents were of reagent grade and were used without purification. The solvents for measurements were all spectral grade and were used without purification.

Synthesis and Characterization of CsPbBr3 and CsPbI3 NCs

Chemicals

Cesium carbonate (Cs2CO3, reagent Plus, 99%), lead(II) iodide (PbI2, 99.999% trace metals basis), toluene (anhydrous, 99.5%), octadecene (ODE, technical grade, 90%), oleylamine (OLA, 70%), and oleic acid (OA, 90%) were purchased from Sigma-Aldrich, and lead(II) bromide (PbBr2, for the perovskite precursor) was purchased from Tokyo Chemical Industry. All chemicals were used without any further purification.

Synthesis of CsPbBr3 NCs

CsPbBr3 NCs were synthesized following the reported procedure by Protesescu et al. with some minor modification.[3] Briefly, Cs2CO3 (0.16 g, 0.50 mmol), OA (0.5 mL), and ODE (8 mL) were loaded into a three-necked round-bottom flask, dried/degassed for 1 h at 120 °C under vacuum, and then stirred under an argon atmosphere at 150 °C until all Cs2CO3 dissolved. This resulting clear solution was referred to as Cs-oleate solution. PbBr2 (0.069 g, 0.19 mmol) and ODE (5 mL) were loaded into another three-necked round-bottom flask, dried/degassed under vacuum at 120 °C for 1 h, and then heated under argon gas to 160 °C, and dried OA (0.5 mL) and dried OLA (0.5 mL) were injected. After complete dissolution of the PbBr2, the temperature was adjusted at 160 °C, and the Cs-oleate solution (0.4 mL) was swiftly injected. Immediately after the injection, the reacted solution was quickly cooled down to room temperature with an ice-water bath. The crude solution of the synthesized CsPbBr3 NC was first centrifuged at 5000 rpm for 5 min at 25 °C to remove the unreacted compounds. The precipitate was dispersed in cyclohexane and then centrifuged at 5000 rpm for 30 min at 25 °C. After centrifugation, the supernatant was collected and used for the measurement. The resulting solutions were stored in the dark at 25 °C. UV/vis in cyclohexane: λ (first peak) = 500 nm, DUV = 7.4 nm.[19] PL in cyclohexane: λmax = 510 nm, FWHM = 25 nm. TEM: DTEM = 7.3 ± 0.8 nm.

Synthesis of CsPbI3 NCs

CsPbI3 NCs were synthesized by almost the same synthetic method as the CsPbBr3 NC using PbI2 (0.087 g, 0.19 mmol) as the precursor. PL in cyclohexane: λmax = 674 nm, FWHM = 34 nm.

Synthesis and Characterization of 1

Synthesis and Characterization of 2

Compound 2 was synthesized following the reported procedure by Bunzen et al. with some minor modification.[27] First, ethylenediamine (8.9 mL, 130 mmol), triethylamine (0.9 mL, 6 mmol), and dry dichloromethane (90 mL) were loaded into a three-necked round-bottom flask and stirred at 0 °C under an argon atmosphere. To the clear solution, cholesteryl chloroformate (3.0 g, 6.7 mmol) and dry dichloromethane (90 mL) were injected, and then the reaction mixture was stirred at 25 °C under an argon atmosphere for 18 h. The formed precipitates were filtered off, and the filtrate was washed one time with water and four times with brine. The organic layer was dried over Na2SO4 and then evaporated to dryness under reduced pressure to obtain 2 as a white solid (2.9 g, 6.1 mmol, 91% yield). 1H NMR (500 MHz, CDCl3, TMS, 25 °C): δ (ppm) = 5.38–5.37 (m, 1H, C=CH), 4.97 (br, 1H, OCONH), 4.52–4.48 (m, 1H, COOCH), 3.24–3.20 (m, 2H), 2.81 (t, J = 5.9 Hz, 2H), 2.38–0.94 (m, 31H), 0.91 (d, J = 6.7 Hz, 3H), 0.87 (d, J = 6.5 Hz, 3H), 0.86 (d, J = 6.5 Hz, 3H), 0.68 (s, 3H).

Synthesis and Characterization of 3

First, compound 2 (1.1 g, 2.3 mmol), 4-hydroxybenzoic acid (0.27 g, 2.0 mmol), 4-dimethylaminopyridine (DMAP, 24 mg, 0.2 mmol), and dry dichloromethane (40 mL) were loaded into a three-necked round-bottom flask and stirred at 25 °C under an argon atmosphere. To the clear solution, N,N′-dicyclohexylcarbodiimide (DCC, 0.40 g, 1.9 mmol), 1-hydroxy-7-azabenzotriazole (HOAt, 0.26 g, 1.9 mmol), and dry dichloromethane (25 mL) were injected, and then the reaction mixture was stirred at 25 °C under an Ar atmosphere for 72 h. The formed precipitates were filtered off, and the filtrate was washed one time with water and two times with brine. The organic layer was dried over Na2SO4 and then evaporated to dryness under reduced pressure to obtain 3 as a white solid (1.2 g, 2.0 mmol, 87% yield). 1H NMR (500 MHz, CDCl3, TMS, 25 °C): δ (ppm) = 7.72 (d, J = 8.6 Hz, 2H, Ar-H), 6.94 (br, 1H, CONH), 6.86 (d, J = 8.6 Hz, 2H, Ar-H), 5.34–5.33 (m, 1H, C=CH), 5.06 (br, 1H, OCONH), 4.50–4.49 (m, 1H, COOCH), 3.57–3.44 (m, 4H), 2.31–0.99 (m, 31H), 0.91 (d, J = 6.9 Hz, 3H), 0.87 (d, J = 6.5 Hz, 3H), 0.86 (d, J = 6.5 Hz, 3H), 0.67 (s, 3H). The 13C NMR spectrum was not obtained due to the low solubility in CDCl3. HRMS (ESI): m/z calcd for C37H56N2O4Na 615.4141 [M + Na]+, found 615.4132.

Synthesis and Characterization of 1

Compound 3 (0.5 g, 0.84 mmol), 18-bromo-1-octadecene (0.28 g, 0.84 mmol), K2CO3 (5.8 g, 42 mmol), and dry DMF (50 mL) were loaded into a three-necked round-bottom flask, and then the reaction mixture was stirred at 70 °C under an argon atmosphere for 26 h. The reacted solution was added into water to form precipitates, and they were filtered off, and the separated solid was extracted with chloroform and brine. The organic layer was dried over Na2SO4 and then evaporated to dryness under reduced pressure. The resulting solid was purified by column chromatography on silica gel (hexane/ethyl acetate = 1:1) to obtain 1 as a white solid (0.43 g, 0.51 mmol, 61% yield). 1H NMR (500 MHz, CDCl3, TMS, 25 °C): δ (ppm) = 7.76 (d, J = 8.6 Hz, 2H, Ar-H), 6.93 (br, 1H, CONH), 6.90 (d, J = 8.6 Hz, 2H, Ar-H), 5.85–5.77 (m, 1H, CH=CH2), 5.33–5.32 (m, 1H, CH=C), 5.10 (br, 1H, OCONH), 5.01–4.91 (m, 2H, CH=CH2), 4.50–4.49 (m, 1H, COOCH), 3.98 (t, J = 6.6 Hz, 2H, OCH), 3.61–3.39 (m, 4H), 2.30–0.96 (m, 61H), 0.91 (d, J = 6.9 Hz, 3H), 0.87 (d, J = 6.5 Hz, 3H), 0.86 (d, J = 6.5 Hz, 3H), 0.67 (s, 3H). 13C NMR (125 MHz, CDCl3, 25 °C): δ (ppm) = 167.59, 161.83, 157.62, 139.76, 139.30, 128.83, 126.15, 122.55, 114.21, 114.11, 77.31, 77.06, 76.80, 74.79, 68.20, 56.72, 56.16, 50.02, 42.33, 41.56, 40.56, 39.75, 39.54, 38.49, 36.97, 36.55, 36.21, 35.82, 33.98, 33.86, 31.92, 31.86, 29.72, 29.66, 29.63, 29.55, 29.45, 29.20, 28.98, 28.26, 28.04, 26.05, 25.64, 24.98, 24.31, 23.86, 22.86, 22.60, 21.05, 19.33, 18.74, 11.88. HRMS (ESI): m/z calcd for C37H56N2O4Na 865.6792 [M + Na]+, found 865.6793.