Huaiyu
Chen
ab,
Jinyong
Zhou
ab and
Jianping
Deng
*ab
aState Key Laboratory of Chemical Resource Engineering, Beijing University of Chemical Technology, Beijing 100029, China. E-mail: dengjp@mail.buct.edu.cn
bCollege of Materials Science and Engineering, Beijing University of Chemical Technology, Beijing 100029, China
First published on 14th October 2015
Hybrid materials combining chirality and magneticity are stimulating much interest in diverse research areas. This article reports the preparation of a new type of optically active, magnetic Fe3O4 NP@polyacetylene core/shell microspheres (Fe3O4@PA MPs) consisting of magnetic Fe3O4 NPs as the core and helical polyacetylene as the shell. The Fe3O4@PA MPs integrate two significant concepts, “macromolecular helicity-derived chirality” and “magneticity” in one single microsphere entity. The composite MPs were prepared by emulsion polymerization approach and characterized by TEM, XRD, FT-IR, VSM, TGA, circular dichroism and UV-vis absorption spectroscopy techniques. They simultaneously showed fascinating optical activity and considerable magneticity. The MPs were further used as a chiral additive to induce enantioselective crystallization of racemic threonine. L-Threonine was preferentially induced to form rectangular-shaped crystals with an enantiomeric excess up to 90% (after enantioselective crystallization twice). The microspheres can be recycled conveniently with the assistance of an external magnetic field, demonstrating the MPs’ significant potential applications in chiral fields.
Synthetic helical polymers have received a great deal of attention owing to their distinctive helical structures and the corresponding properties that cannot be found in ordinary polymers. Helical polymers have visually elegant structures and possess fascinating optical activity,23 chiral resolution ability,24etc. Over the past two decades, a series of helical polymers were synthesized,25–40 among which helical polyacetylenes are the most intensively explored. Since the pioneering work of Heeger, MacDiarmid, and Shirakawa41–43 on polyacetylenes, considerable efforts have been devoted to the polymer and its derivatives, in particular the helical substituted polyacetylenes in which the conjugated polymer main chains adopt helical conformations while the pendant groups endow the desired functions.44–47 Although helical polyacetylenes possess fascinating properties, they frequently suffer from disadvantages like insufficient solubility, thermal instability and/or unsatisfactory processability.48 Therefore it is of great significance to find effective ways to circumvent the disadvantages. Design and synthesis of acetylene-based nanoparticles (NPs) seem to be an efficient and facile approach to overcome these limitations. The Mecking group49 has made breakthroughs in preparing acetylene-based polymer NPs. We also have prepared optically active helical substituted polyacetylene particles via emulsion polymerization,50 precipitation polymerization,51 and helix-sense-selective polymerizations of achiral monomers.52,53 Particularly, we synthesized core/shell polyacetylene NPs composed of optically active helical polyacetylene cores and different shells.1,54 The helical polyacetylene NPs promisingly find remarkable applications in enantioselective crystallization,54 chirally-controlled release,55 chiral catalysis,56 among the other significant applications. Unfortunately, the recovery of the chiral NPs after use remains challenging, a big issue in terms of practical applications. Endowing the optically active core/shell particles with magnetic properties will undoubtedly help improve their recoverability.
In the present study, we synthesized optically active, magnetic core/shell structured Fe3O4@polyacetylene microspheres (defined as Fe3O4@PA MPs) by taking magnetic Fe3O4 NPs as the core and helical substituted polyacetylene as the shell. The composite MPs integrated the magneticity of the Fe3O4 core and the intriguing optical activity of the helical polymer shell. Accordingly, the core/shell MPs in theory shall possess both rapid magnetic responsivity and optical activity. Excitingly, this hypothesis was validated in the present work. We further employed the resulting MPs for performing enantioselective crystallization by taking rac-threonine as a model compound. L-Threonine was found to be preferentially induced to crystallize. It is worth emphasizing that the composite MPs can be easily recycled with the help of a magnetic field after the enantioselective crystallization process.
Scheme 1 Schematic illustration for preparing the optically active, magnetic Fe3O4@PA core/shell microspheres (Fe3O4@PA MPs). |
After enantioselective crystallization, threonine crystals filtered from the residual solution were re-dissolved in water. The Fe3O4@PA MPs originally coated by the resulting crystals were separated by a magnet, washed with water three times, and dried at 40 °C for 24 h. The collected Fe3O4@PA MPs were reused again for conducting a second cycle of enantioselective crystallization. In the second enantioselective crystallization, a predetermined amount of L-threonine crystals obtained in the first cycle was dissolved in water to form a threonine supersaturated aqueous solution while the other experimental conditions remained the same as in the first cycle. For comparison, achiral alkynyl-Fe3O4 NPs were also utilized as additives instead of chiral Fe3O4@PA MPs for crystallization of racemic threonine in the same manner.
The morphology and the size of the Fe3O4 NPs and the Fe3O4@PA MPs were characterized by TEM (Fig. 1) and SEM (Fig. S1 in the ESI†). Fig. 1a displays the typical TEM images of Fe3O4 NPs. The average diameter is ca. 280 nm. The morphology and size of the Fe3O4@PA MPs are shown in Fig. 1b, in which the average diameter of the composite microspheres is approximately 1050 nm. More noticeably, the core/shell structures can be observed. To make the core/shell structure more clear, phosphotungstic acid was used to stain the polyacetylene shell, so the PA shells seem to be much darker than the Fe3O4 cores. In contrast, the average diameter of the pure polyacetylene nanoparticles (PA NPs), which were prepared under similar conditions and taken as a control sample, was only about 680 nm (Fig. 1c). Similar results were obtained in SEM images (Fig. S1†). A comparison of the TEM and SEM images definitely demonstrates that in the Fe3O4@PA core/shell MPs, the substituted polyacetylene was successfully attached onto the surface of the Fe3O4 NPs.
Fig. 2 shows the XRD patterns of (a) Fe3O4 NPs, (b) optically active, magnetic Fe3O4@PA core/shell MPs, and (c) pure polyacetylene NPs as a reference. Diffraction peaks (111), (200), (311), (222), (400), (422), (511), and (440) shown in Fig. 2a, and can be indexed as face centered cubic Fe3O4 (the Joint Committee on Powder Diffraction Standards (JCPDS) reference (no. 19-0629)). Fig. 2b shows the XRD pattern of Fe3O4@PA MPs. All the diffraction peaks of Fe3O4 NPs shown in Fig. 2a can be found in Fig. 2b. The XRD pattern of the pure PA NPs has a wider peak at about 2θ = 19.20 in Fig. 2c. Accordingly, the wider peak at 2θ = 19.26 shown in Fig. 2b originates from amorphous substituted polyacetylene shells. The observations further show us that Fe3O4@PA MPs were successfully prepared.
Fig. 2 XRD patterns of (a) Fe3O4 NPs, (b) optically active, magnetic Fe3O4@PA core/shell MPs, and (c) pure substituted polyacetylene NPs. |
Next, the products were subjected to FT-IR spectra measurements, as shown in Fig. 3. The vibrational absorption peak at 588 cm−1 (Fig. 3a) can be assigned to the characteristic of the Fe–O stretching vibration.63 The absorption peaks corresponding to Si–O (1128 cm−1) and CC (2103 cm−1) can be found in the spectrum of alkynyl-Fe3O4 NPs (Fig. 3b), indicating that the Fe3O4 NPs were successfully functionalized by alkynyl moieties. The FT-IR spectrum of the Fe3O4@PA core/shell MPs is presented in Fig. 3c, in which the new peaks at 1525, 1665, and 1790 cm−1 are attributed to the amide I band (CO stretching), amide II band (N–H bending), and CO stretching of the lactone group (in the M1 units). The peak at 1740 cm−1 corresponds to the ester group in both M1 units and the cross-linking agent (M2). The peaks provide evidence for the presence of M1 units and cross-linking agent units in the composite microspheres.57 The FT-IR spectra further prove that magnetic Fe3O4 NPs and helical substituted polyacetylene chains are both present in the composite microspheres. For this conclusion, CD and UV-vis spectra will offer more information, and will be discussed later.
Fig. 3 FT-IR spectra of (a) Fe3O4 NPs, (b) alkynyl-Fe3O4 NPs, (c) optically active, magnetic Fe3O4@PA core/shell MPs, and (d) pure polyacetylene NPs (KBr tablet). |
The thermal properties of the Fe3O4@PA MPs and the PA NPs were measured by the TGA technique, as illustrated in Fig. 4. Compared to the original Fe3O4 NPs, the TGA curve of the alkynyl-Fe3O4 NPs showed a weight loss of 9.6% (Fig. 4b). The weight loss is due to OPNTU silane, which further confirms that the Fe3O4 NPs were successfully functionalized by alkynyl moieties. Fe3O4@PA MPs showed a weight loss of 85% when the temperature was increased from 100 to 700 °C (Fig. 4c). So the residual amount of Fe3O4@PA MPs was about 15%. This value is reasonably much higher than that of the pure PA NPs (Fig. 4d).
Fig. 4 TGA curves of (a) Fe3O4 NPs, (b) alkyne-Fe3O4 NPs, (c) Fe3O4@PA MPs, and (d) pure substituted polyacetylene NPs. All the TGA curves were measured at a scanning rate of 10 °C min−1 in air. |
Our previous investigations demonstrate that circular dichroism (CD) and UV-vis absorption spectroscopy are effective techniques for identifying the helical structures and optical activity of substituted polyacetylenes and the particles thereof.50–54,58 The optically active, magnetic Fe3O4@PA core/shell MPs and the pure PA NPs were thus subjected to CD and UV-vis spectrum measurements. To characterize the optical activity of the Fe3O4@PA MPs, we obtained pure microspheres by isolating them from the emulsions and washed them three times with H2O with the aid of a magnet. The Fe3O4@PA MPs were dispersed again in water and then subjected to measurements. The CD spectra of the Fe3O4@PA MPs dispersion in water are presented in Fig. 5a, while the corresponding UV-vis spectra are shown in Fig. 5b. The CD signals and UV-vis absorption of Fe3O4@PA MPs can be observed around at 340 nm, quite similar to the CD signal and UV-vis absorption of the pure PA NPs. This further identifies that the helical polyacetylene was successfully coated on Fe3O4 NPs. According to our earlier studies dealing with helical substituted polyacetylenes,22 we conclude that the Fe3O4@PA MPs were composed of helical polymers with predominantly one-handed screw sense and accordingly the MPs possessed considerable optical activity. We further infer that the presence of Fe3O4 cores did not affect the polymers in terms of forming helical structures of preferential helicity. It should be pointed out that in Fig. 5, both the CD signal and UV-vis absorption of the chiral composite NPs are weaker relative to the pure polymer NPs. This results from two aspects: the relatively low content of helical polymers in the composite MPs and the spectra were just qualitatively measured.
Fig. 5 CD and UV-vis absorption spectra of the optically active, magnetic Fe3O4@PA core/shell MPs and pure polyacetylene nanoparticles (PA NPs) in aqueous dispersions. |
The CD spectrum measurement clearly shows the optical activity of the composite MPs. Besides, the composite microspheres are also expected to show interesting magneticity. To make this point clear, the magnetic properties of the Fe3O4@PA MPs were examined by using a VSM at 300 K. The relevant results are illustrated in Fig. 6. Fig. 6 shows the hysteresis loops of the Fe3O4 NPs (a) and the Fe3O4@PA MPs (b), indicating that their saturation magnetization is 66.85 and 23.59 emu g−1, respectively. A certain difference was observed between the magnetization value measured and the corresponding theoretical value, most likely due to the fact that some Fe3O4 NPs were not completely coated by polyacetylene. The VSM results convincingly show that the Fe3O4@PA MPs possessed remarkable magnetic properties. The high saturation magnetization of the microspheres afforded them excellent magnetic responsivity. As the insets in Fig. 6 show, the time from the dispersion state (b) to the aggregation state (c) was achieved within 11 s under an external magnetic field. When the external magnet was removed, the state (b) can be completely recovered from the state (c) just by shaking. Therefore the optically active, magnetic core/shell MPs can be easily recycled with the help of a magnet. Herein it is important to note that photos of the pure Fe3O4 NPs (Fig. 6a inset) and the optically active, magnetic Fe3O4@PA MP (Fig. 6b) dispersions were taken. The color of the pure Fe3O4 NPs was black, while the Fe3O4@PA MPs exhibited a yellow-brown color because of the presence of yellow helical substituted polyacetylene. Helical substituted polyacetylenes always show a light yellow color (or yellowish colors) in both solution and particulate states.58
Scheme 2 A schematic for illustrating the two-cycle enantioselective crystallization of racemic threonine by using Fe3O4@PA MPs. |
In Scheme 2, in the first enantioselective crystallization cycle, Fe3O4@PA MPs were added into the racemic D,L-threonine supersaturated aqueous solution, as detailed in the Experimental section. L-Threonine crystals (ee approx. 25%) were more favorably induced in the first cycle of enantioselective crystallization. After magnetic separation, Fe3O4@PA MPs were restored and re-used for the second cycle of enantioselective crystallization, wherein the L-threonine crystals induced in the first cycle were used, aiming at further improving the ee. Excitingly, the ee of the L-threonine crystals after the second cycle of crystallization increased up to about 90%.
The SEM technique was used to observe the L-threonine crystals induced in the first cycle as a function of crystallization time. Fig. 7 shows the relevant SEM images. In Fig. 7a, clusters of crystals commenced to form and could be clearly viewed (after crystallization for six hours). Twelve hours later (Fig. 7b), rectangular-shaped crystals were observed more clearly. In Fig. 7c and d, the crystals continuously grew larger but still well maintained the rectangular shape; moreover, the crystals became progressively more regular. The SEM images clearly show the appearance and the growth of the induced crystals. To acquire deeper insights into the growth of the crystals in the course of enantioselective crystallization in the presence of Fe3O4@PA MPs as chiral seeds, the yield of the crystals against crystallization time is presented in Fig. 8. At the early stage, no crystals were formed (within the first four hours). From then on, crystals began to form and the yield of the crystals drastically increased with prolonging crystallization time (4–12 h, the second stage). After that, the yield of the crystals increased slowly and then was nearly constant (the third stage). The maximum yield of the crystals could be as high as 40%. The observation is in agreement with the SEM images (Fig. 7). The L-threonine crystals in the second crystallization cycle showed the same crystal morphology as observed in Fig. S2 (ESI†). However, the crystals yield increased more rapidly at the second stage relative to that in the first cycle of enantioselective crystallization (Fig. 8).
Fig. 8 Yield of L-threonine crystals induced by Fe3O4@PA MPs as a function of the crystallization time. |
To further validate our conclusion, the crystals induced in the two cycles and the residual solutions were measured by CD spectra, as shown in Fig. 9. Just like pure L-threonine, the threonine crystals obtained in both the 1st and 2nd crystallization cycles showed a positive CD signal around 210 nm. The threonine crystals obtained in the 2nd crystallization cycle had a stronger CD signal than that from the 1st cycle. This observation is in accordance with the ee values of the induced L-threonine crystals (25% and 90% in the 1st and 2nd cycles, respectively). In sharp contrast to the induced L-threonine crystals, the corresponding residual solutions showed a negative CD signal at the same wavelength (210 nm). From the CD spectra, we know that L-threonine was preferentially induced to crystallize, while D-threonine predominantly remained in the residual solutions. To further justify the effects of Fe3O4@PA MPs in the enantioselective crystallization process, we measured the CD spectra of the threonine crystals induced by using achiral alkynyl-Fe3O4 NPs instead of chiral Fe3O4@PA MPs. The induced crystals assumed a similar morphology to that induced by using chiral composite MPs, but they did not show a CD signal in the wavelength range of interest (ee, about zero in this case). These results demonstrate that Fe3O4@PA MPs play essential roles in inducing the enantioselective crystallization of L-threonine.
In order to obtain a better understanding of the enantioselective crystallization, the ee of the threonine crystals (L-threonine in excess in the threonine crystals induced by Fe3O4@PA MPs, as discussed above) and the corresponding residual solutions were plotted against crystallization time. As Fig. 10a illustrates the ee of both the threonine crystals and the residual solutions increased as a function of the crystallization time. For the L-threonine crystals, a maximum ee of 25% was achieved around 18 h. From then on, ee decreased slightly, due to the fact that D-threonine began to crystallize faster than L-threonine. So the optimal crystallization time is determined to be 18 h. The changing trend in the ee of the residual solution is consistent with the above observation. Nonetheless, the ee value of the induced crystals is not high enough from the viewpoint of practical applications. To further improve the purity of the L-threonine crystals induced in the first enantioselective crystallization cycle, we performed the second cycle of the enantioselective crystallization by using the recycled Fe3O4@PA MPs after the first cycle use. As shown in Fig. 10b, after ca. 12 h, the maximum ee in the second crystallization cycle excitingly increased up to 90%, showing the significant potential applications of the Fe3O4@PA MPs as chiral additives. The Fe3O4@PA MPs recycled after the first two-cycle enantioselective crystallization were re-used for another “two-cycle” crystallization process in the same manner. The ee of the obtained crystals in this case was found to be up to 84%. This demonstrates that the Fe3O4@PA MPs were highly efficient in inducing enantioselective crystallization even after a “two-cycle” process.
Fig. 10 ee as a function of the crystallization time (L-threonine in excess) by using Fe3O4@PA MPs (a) in the 1st and (b) 2nd enantioselective crystallization cycles. |
XRD analyses were further performed on the L-threonine crystals obtained in the 2nd enantioselective crystallization cycle (Fig. 11). When compared to pure L-threonine, the XRD of L-threonine crystals induced by Fe3O4@PA MPs showed nearly the same XRD patterns, only with a slight difference in the diffraction intensity. The XRD measurement further demonstrates the high purity of the L-threonine crystals after the 2nd enantioselective crystallization cycle.
Fig. 11 Typical X-ray diffraction patterns of (a) pure L-threonine and (b) L-threonine crystals induced by Fe3O4@PA MPs after the 2nd crystallization. |
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5py01549a |
This journal is © The Royal Society of Chemistry 2016 |