DOI: 10.1002/chem.201406507
Full Paper
& Self-Assembly and Separation
Controllable Assembly and Separation of Colloidal Nanoparticles through a One-Tube Synthesis Based on Density Gradient Centrifugation Xiaohan Qi,[a] Minglin Li,[a] Yun Kuang,[a] Cheng Wang,[a] Zhao Cai,[a] Jin Zhang,[a] Shusen You,[a, b] Meizhen Yin,[a, b] Pengbo Wan,[a] Liang Luo,*[a] and Xiaoming Sun*[a]
field to overcome the random Brownian motion of NPs and benefit the assembly effect. Such a facile “one-tube synthesis” approach couples assembly and separation in one centrifuge tube by centrifuging once. The method can be tuned by changing the concentration of interference salt layer, encapsulation layer, and centrifugation rate. Furthermore, positively charged fluorescent polymers such as perylenediimide-poly(N,N-diethylaminoethyl methacrylate) could encapsulate the assemblies to give tunable fluorescence properties.
Abstract: Self-assembly of gold nanoparticles into one-dimensional (1D) nanostructures with finite primary units was achieved by introducing a thin salt (NaCl) solution layer into density gradient before centrifugation. The electrostatic interactions between Au nanoparticles would be affected and cause 1D assembly upon passing through the salt layer. A negatively charged polymer such as poly(acrylic acid) was used as an encapsulation/stabilization layer to help the formation of 1D Au assemblies, which were subsequently sorted according to unit numbers at succeeding separation zones. A centrifugal field was introduced as the external
Introduction
assembly degree, orientation, and spacing between repeat units. However, a significant challenge in the self-assembly is the precisely control of ordered structures that adopts specific interparticle orientations, which prevents colloidal jamming. With the aim to achieve synthetic efficiency, a variety of successful strategies have been developed through anisotropic chemical modification of the nanoparticle surface, including the use of templates such as DNA linkers,[8, 9] block copolymers,[1, 10] and patchy particles,[11, 12] or by changing environmental conditions, such as pH value,[13] temperature,[14] and electrostatic interactions.[15] However, the 1D assembly of colloids is regarded the result of dynamic balance between various interactions.[16] For the obtained assembled structures based on current fabrication methods, it is hard to realize ideal monodispersity as well as the corresponding unique optical properties, and the related study remains rather limited. Very recently, the density gradient centrifugation separation method has emerged as a powerful tool to sort nanoparticles with different compositions, sizes, and morphologies in a short time.[18, 19] Moreover, in our preliminary studies, this method also provided the opportunity to isolate/capture intermediate product for the observation and understanding of chemical reaction, growth process, or phase transition.[17, 20–22] Herein, such a method was further developed to capture the “aggregation intermediate”, which means only the initial state of aggregation when several colloids collide with each other in salt solution. Since the aggregation degree can be well-controlled and terminated in certain degree, it is also called as-
Self-assembly of nanoparticles as the building blocks into welldefined nanostructures is of pivotal significance for the design and fabrication of next-generation functional nanodevices. Linear polymer-like one-dimensional (1D) assembled nanostructures in particular are rapidly gaining attention as model systems and a fundamental step owing to their unique physical and chemical properties.[1] In particular, 1D assembled nanostructures based on noble metal nanoparticles (NPs), termed plasmonic polymers, exhibit strong coupling of localized surface plasmon resonance (LSPR) between adjacent NPs,[2] which lead to a wide range of promising applications such as chemical/bio sensing (for example SERS), data storage, nanocircuits, nanoantennas, and optoelectronics.[3–7] The plasmon coupling can be controlled and tuned by changing the
[a] X. Qi, M. Li, Y. Kuang, C. Wang, Z. Cai, J. Zhang, S. You, Prof. M. Yin, Dr. P. Wan, Dr. L. Luo, Prof. X. Sun State Key Laboratory of Chemical Resource Engineering P.O. Box 98, Beijing University of Chemical Technology Beijing 100029 (P. R. China) E-mail: [email protected] [email protected] [b] S. You, Prof. M. Yin Beijing Laboratory of Biomedical Materials Beijing University of Chemical Technology Beijing 100029 (P. R. China) Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/chem.201406507. Chem. Eur. J. 2015, 21, 1 – 7
These are not the final page numbers! ÞÞ
1
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
Full Paper semblies were seriously aggregated after moving down to the bottom of the tube. There were partially chain-like assemblies for f1 and f3, but from f5 afterward, the fractions tended to aggregate into almost bulk randomly rather than nanochains. Digital camera images of centrifugation tube before and after separation (Supporting Information, Figure S1) showed that the color of the fractions in the separation zone changed from violet (up) to light blue (down), indicating the serious aggregation. UV/Vis spectra of the corresponding fractions (Figure 1 B) further confirmed the trend. Compared with isolated Au NPs, there was significant reduced absorbance and slight red-shift for the separated fractions from 520 to 540 nm as the intrinsic absorption peak of Au NPs, indicating the aggregation of Au NPs. For the longitudinal mode, the absorption peak appeared at about 702 nm for f3, which indicated 1D oligomer formation. Blue-shift was observed from f5 afterward, owing to the reduced aspect ratio (thicker diameter) of the random aggregated nanostructures. Thus, even if exposed for a very short time and separated quickly, the obtained intermediate oligomers were still reactive to consume each other by random combination during the separation process. In this context, a polymer (here negatively charged poly(acrylic acid) (PAA)), which was demonstrated as the very effective stabilizers to end reactions and/or isolate intermediates by Chen and co-workers,[16, 23] was introduced to encapsulate the assemblies and avoid further aggregation as the isolation layer. The assembly and separation process with PAA layer (1 mL solution with the concentration of 1 mg mL1) which was placed blow the interference layer, is represented in the Supporting Information, Figure S2 A. After the centrifugation, the color of the fractions changed to red and violet, indicating the limited assembly degree. Compared with isolated Au NPs, there was an obvious red-shift about 20 nm for the extinction peak (Supporting Information, Figure S2B) of the four typical fractions (f1, f3, f5, and f7), owing to the increase of surface refractive index, indicating the existence of PAA shell for each fraction. However, the very concentrated distribution of fractions and the irregular trend of right tails of absorption peaks of fractions indicated the separation process need further optimization. Theoretically, according to the classical sedimentation theory,[22, 24] the sedimentation velocity v of colloidal particles in a given medium can be described as Equation (1):
sembly. Such a facile “one-tube synthesis” approach can realize the self-assembly of Au NPs into 1D nanostructures through density gradient ultracentrifugation rate separation. NaCl solution was introduced as the interference layer to change the electrostatic interactions between Au NPs and induce the assembly behavior. Centrifugal field was introduced as the external field to provide tunable gravity to overcome the random Brownian motion of NPs. On the basis of assemblies isolated by encapsulation of polyacrylic acid (PAA) as the stabilization layer, the obtained various sized 1D Au nanochains were effectively separated in the following density gradient layers through centrifugation. All the processes were realized in one centrifuge tube at one time. Furthermore, the fluorescent species such as perylenediimide–poly(N, N-diethylaminoethyl methacrylate) (PDI-PDMAEMA) could also be introduced onto the surface of the assemblies to give fluorescence properties which could be controlled by tuning the longitudinal surface plasmon resonance.
Results and Discussion Figure 1 A shows a representation of the assembly and separation procedure through centrifugation. First, the positively charged quasispherical Au NPs with a size of (20 2) nm were synthesized in cetyltrimethylammonium chloride (CTAC) solution as the target assembly units. Their size uniformity and
Figure 1. A) Illustration of the assembly and separation of Au NPs in a density gradient centrifugation system. B) UV/Vis spectra of Au NPs and five typical fractions.
sphericity minimized the shape effect on the self-assembly of the NPs. A buffer layer including 0.1 mm CTAC solution was placed below Au NPs to prevent spontaneous assembly of Au NPs in ethylene glycol (EG). An interference layer including 400 uL 0.05 mm NaCl solution was introduced to break the surface charge balance of Au NPs and thereby induce the assembly. A multilayer density gradient (40 %, 50 %, 60 %, 70 %, 80 %, 90 %, and 100 %) was made as the separation zone by sequentially layering aqueous solutions with different concentrations of EG in a centrifuge tube. The centrifugal field provided a centrifugal force (9645 g, 7500 rpm) to drive Au NPs through the buffer layer and into the separation zone. After centrifugation for 30 min, Au NPs were driven from the interference layer into the subsequent separation zone. However, as shown in the Supporting Information, Figure S1, transmission electron microscope (TEM) images of five typical fractions (f1, f3, f5, f7, and f9) indicated that the separated as&
&
Chem. Eur. J. 2015, 21, 1 – 7
www.chemeurj.org
v ¼ 2ð1p 1s Þ r 2 g0 =ð9hs Þ
ð1Þ
where v is the sedimentation velocity of colloidal nanoparticle, 1p and 1s are the densities of nanoparticle and the surrounding fluid, r is the radius of nanoparticle, g’ is the centrifugal acceleration, and hs is the viscosity of the surrounding fluid. Hence, the density and the radius of nanoparticle was constant for the assemblies after passing through PAA layer, and the separation effect was determined by the centrifugation rate and the solvent. For the separation process with EG/H2O as the solvent, the relative large viscosity (EG: 22.1 MPa s at 20 8C) of surrounding fluid resulted in long centrifugation time and espe2
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!
Full Paper cially more energy needed for passing through the gradient phase interfaces. The separation zone was thus applied by using 50 %, 60 %, 70 %, 80 %, 90 %, and 100 % (by volume) solutions of deuteroxide in water as the gradient solvents, and PAA solution was applied as the encapsulation layer, as shown in Figure 2 A. The digital camera images before and after centrifugation with 7000 rpm for 15 min (Figure 2 B) showed that the fractions distributed separately, and the color changed slightly from light pink to violet, indicating the limited assembly degree of Au NPs and ideal separation effect. As the salt layer and encapsulation layer were both very thin, and thereby the time for each assembled fraction reaching to the top layer of density gradient was almost the same with the biggest time interval no more than 1 min. Besides the obvious red-shift about 20 nm for the extinction peak of the all fractions (Figure 2 C) compared to isolated Au NPs, there was also a regular red-shift trend (from 530 to 550 nm for f1–f7) for the extinction peak and increased absorbance of right tail for each fraction. Different with the fractions obtained by the separation without PAA, there was no strong longitudinal peaks as PAA was involved,
Figure 3. A) TEM images of four typical fractions: f1, f3, f5, and f7 (scale bars: 100 nm), HRTEM image of nanostructure (scale bar: 5 nm), and representation of the obtained nanostructure. B) Size distribution of the four typical fractions.
Furthermore, the dynamic light scattering (DLS; Supporting Information, Figures S3 and S4) results also exhibited the increase trend of overall size distribution for the fractions from top to down, confirming the corresponding dominant constitution. To obtain deeper insight into the one-tube synthesis process, control experiments by changing Figure 2. A) Illustration of the assembly and separation of Au NPs in a density gradient centrifugation system with the concentration of interference a PAA layer. B) Digital camera images of centrifugation tube before and after centrifugation. C) UV/Vis spectra of Au NPs and the four fractions. layer (NaCl) and encapsulation layer (PAA) and the centrifugation rate were carried out to investigate the assembly behavior which was due to the limited assembly degree based on relaof Au NPs. For the study of the effect of interference layer, tive small size of Au nanoparticles (20 nm) as the assembly three different concentrations of 400 uL NaCl solution, namely unit. Figure 3 A shows the TEM images of four typical fractions 0, 0.05 and 0.1 m, were compared (Figure 4 A). As the concen(f1, f3, f5, and f7), which clearly exhibited the effective assemtration of NaCl was 0 m, there was no obvious aggregation bly and separation of linear nanostructures as monomers, after centrifugation, and the color of each fraction was still the dimers, tri/tetramers, and oligomers. High-resolution TEM characteristic red of Au NPs. The corresponding UV/Vis spectra (HRTEM) images confirmed the Au assemblies@PAA core–shell of assemblies and isolated Au NPs were almost the same structure, and the mean shell thickness were 1.9 nm. After we except the red-shift for the assemblies induced by the coating carefully rechecked the TEM images of f1, we confirmed that of PAA (Supporting Information, Figure S5 A). As the concentrathey consisted of single nanoparticles. The coating of PAA was tion of NaCl was 0.05 m, the regular assembly was realized, the carried out in the centrifuge tube, so it is unlikely that it will UV-Vis spectra of various fractions (Supporting Information, occur with vigorous stirring. Thus, we cannot guarantee that Figure S5B) was the same as Figure 2 C. At a higher concentrathe shell thickness is the same between assembled nanopartition of NaCl as 0.1 m, the color of the dominant fractions was cles. The varied shell thicknesses can be observed under TEM bluish violet, which indicates over-aggregation during the asobservation and confirmed by broadened absorption peak (f1, sembly process, and the broadened absorption peaks in UV/ Figure 2 C) even for the single nanoparticles (f1, as confirmed Vis spectra especially the bottom fractions such as f7 (Supportby TEM images). The size distribution (Figure 3 B) of the four ing Information, Figure S5C) also validated the result. Further typical fractions based on TEM characterization showed that evidence of the effect of NaCl concentration is more clearly the mean size of f1, f3, f5, and f7 were about (23 3) nm, seen from the comparison of corresponding UV/Vis spectra of (51 17) nm, (110 30) nm, and (142 32) nm, respectively. Chem. Eur. J. 2015, 21, 1 – 7
www.chemeurj.org
These are not the final page numbers! ÞÞ
3
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
Full Paper Furthermore, as the basic parameter to control separation effect, three centrifugation rates as 3000, 7000, and 11 000 rpm were applied to monitor the assembly process of Au nanoparticles, respectively. For 3000 rpm, the centrifugal force was not enough, and the NPs moved slowly. There was much more time for Au NPs passing through NaCl layer and separation under 3000 rpm, thereby the NPs aggregated seriously and randomly as a result of Brownian motion. The UV/Vis spectra of various fractions (Supporting Information, Figure S7 A) confirmed the seriously aggregation, especially the over-broadened and red-shifted absorption peak of f5, and the relatively blue-shifted peak of f7 owing to the reduced aspect ratio by over-aggregation (Supporting Information, Figure S8). When the centrifugation rate was 7000 rpm, the time for the Au NPs staying in the assembly layer was long enough to form Au chains with less than 10 NPs. For 11 000 rpm, the relative high centrifugation rate induced relative high g force, which would provide a strong directional driving force to facilitate the linear aggregation for the unstable intermediates just after passing through NaCl and PAA layers during centrifugation. The UV/Vis spectra (Supporting Information, Figure S7C) showed that there was a regular red-shift trend for the fractions from the upper layer to below, but also an obvious increase absorbance trend for the right tail of absorption peaks, which indicated the longer linear aggregation (Supporting Information, Figure S8). The comparison of corresponding absorption spectra of f5 based on the three parallel experiments was also summarized in Figure 4 F. Owing to the assembly process was carried out during centrifugation, besides the electrostatic force, the assembly process was mainly due to the competition between Brownian motion and external field. According to Stokes–Einstein equation[25] [Eq. (2)], diffusion constant D of nanoparticles could be calculated as:
Figure 4. Digital camera images of tubes before and after centrifugation with the change of A) NaCl concentration from 0 to 0.1 m, C) PAA concentration from 0.5 to 2 mg mL1, E) centrifugation rate from 3000 to 11 000 rpm. B), D, F) UV/Vis spectra of Au NPs and the typical fraction f5 based on the three parallel experiments.
f5 as the typical representative fraction based on the three parallel experiments, shown in Figure 4 B. An appropriate concentration of NaCl was pivotal for the control of assembly degree. The concentration of PAA as the encapsulation layer also played an important role in isolating assemblies and the subsequent separation. PAA solution (1 mL) with low concentration as 0.5 mg mL1 could not provide effective isolation for various assemblies. After centrifugation, the color of the dominant assemblies was bluish violet (Figure 4 C), indicating over-aggregation such as in the situation without PAA (Supporting Information, Figure S1). The corresponding UV/Vis spectra (Supporting Information, Figure S6 A) showed that the absorption peaks of upper fractions were almost the same as isolated Au NPs (above f5), and the lower fractions (below f7) were over-aggregated owing to the significant broadened and red-shifted absorption peaks. As the concentration of PAA was 1 mg mL1, it was the appropriate value and most PAA was consumed without causing side effect (Figure 2 B). With increasing concentration of PAA to 2 mg mL1 could cause a significant side effect: overcoating the unstable intermediate assembles by PAA could induce overloading of surface charge, which separated the nanoparticles into smaller fractions and even individual nanoparticles. The color of fractions in the centrifugation tube remained the characteristic red of isolated Au NPs, indicating the very limited assembly degree that was confirmed by the redshift and the inadequate broadening for all the fractions according to UV/Vis spectra (Supporting Information, Figure S6 C). As revealed by the comparison of corresponding UV/ Vis spectra of f5 based on the three parallel experiments (Figure 4 D), the concentration of PAA was further found to be important to influence the assembly/disassembly behavior. &
&
Chem. Eur. J. 2015, 21, 1 – 7
www.chemeurj.org
D ¼ kB T=ð6 prhs Þ
ð2Þ
where kB is Boltzmann’s constant and T is the absolute temperature. As long as the external environment (T and hs) was constant, the Brownian motion was constant as 6.64 mm2 s1 according to the calculation (we took trimer in pure D2O as the calculation model), which means a trimer particle is displaced about 1.84 mm in one second. For the effect of centrifugation rate (external field), the sedimentation velocity (equation (1)) was calculated to be 33.17 mm s1 at 3000 rpm, 206.14 mm s1 at 7000 rpm and 443.38 mm s1 at 11 000 rpm, respectively (Table 1). Thus, the higher centrifugation rate benefited the assembly in one direction, rather than random aggregation induced by Brownian motion. Moreover, the average translational friction coefficient of the falling assemblies was determined by the orientation of assembled nanostructures, which was the same as the direction of external field. A high centrifugation rate was also helpful to obtain relative regular 1D nanostructures during the separation process. Considering of the electrostatic attraction, the surface coating could be further realized by adding other positively charged functional polymers, such as fluorescent polymer per4
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!
Full Paper excitation rate of PDI-PDMAEMA based on LSPR coupling. Similarly, f1 showed presented a little better LSPR match than f5. After diluting the fractions according to the difference value of absorbance of fluorescent and non-fluorescent fractions to the same level, the fluorescence spectra of normalized free PDIPDMAEMA, f1 and f5 were detected, as shown in Figure 5 E. The fluorescence intensity of f1 and f5 clearly verified better enhancement than free PDI-PDMAEMA, and f5 exhibited more fluorescence enhancement than that of f1. The evidence was further complemented with time-resolved fluorescence measurements (Figure 5 F). The fluorescence lifetime of free PDIPDMAEMA, f1 and f5 were almost the same, namely 6.9, 5.6, and 6.8 ns, respectively. Therefore, the fluorescence enhancement was mainly induced by the energy transfer from Au assemblies to the PDI-PDMAEMA shell owing to the match of LSPR and excitation band of PDI-PDMAEMA,[26] and the match degree was easily controlled by the assembly degree from the one-tube synthesis. Compared to other approaches, the proposed one-tube synthesis could realize convenient assembly and separation in one tube at one time, and more importantly, allow the isolation of the initial assembled 1D nanostructures from salt layer duly in very short time, which was the guarantee of the formation of 1D nanostructures rather than overjammed aggregates, as demonstrated in control experiments (Supporting Information, Figure S10). At the same time, using such technique, quick chemical environments switching was realized, salt and PAA did not exist in one solution. The subsequent coating with PAA or other polymers just after assembly provided convenient protection and functionalization.
Table 1. Calculation of Brownian motion and sedimentation rate of assemblies in D2O/H2O based on three different centrifugation rates. Rate [rpm] 3
1P [kg m ] 1s [kg m3] R [nm] g’ [m s2] hs [Pa s] D [mm2s1] v [mm s1]
3000
7000 4
1.30 10 1.10 103 30 1.52 104 1.095 103 6.65 33.17
12 000 4
1.30 10 1.10 103 30 9.45 104 1.095 103 6.65 206.14
1.30 104 1.10 103 30 2.03 105 1.095 103 6.65 43.38
ylenediimide-poly(N, N-diethylaminoethyl methacrylate) (PDIPDMAEMA; Figure 5 A; see details of the synthesis in the Supporting Information). Purple PDI-PDMAEMA was added as the second encapsulation layer (Figure 5 B). The sum of shell thickness of obtained assemblies was confirmed as being 5 nm from HRTEM image (Figure 5 C), and the 2 nm-thick inner shell of PAA acted as the spacer between fluorescent PDI-PDMAEMA and Au NPs to avoid the quenching induced by directly contact. After the centrifugation, the assemblies were separated in various layers, and exhibited red fluorescence under UV lamp. We chose f1 and f5 as the model fractions to compare the fluorescence intensity. Figure 5 D shows the absorption spectra of free PDI-PDMAEMA, Au NPs and bishelled fractions: f1 and f5. There were three characteristic absorption peaks for PDIPDMAEMA: 447, 541, and 581 nm, and the shell induced redshifted LSPR of f1 and f5 exhibited better match with the peak at 541 nm of PDI-PDMAEMA than Au NPs, indicating the better
Conclusion Controllable assembly of Au nanoparticles into one-dimensional nanostructures and consequent separation was achieved by a one-tube synthesis that requires centrifugation only once. NaCl solution was introduced as the interference layer to change the electrostatic interactions between Au NPs and induce the assembly. Charged polymers such as polyacrylic acid (PAA) or even fluorescent species such as perylenediimide-poly(N,N-diethylaminoethyl methacrylate) (PDI-PDMAEMA) could be introduced as encapsulation/stabilization layer and functionalization layer. Subsequently, the obtained variously sized 1D Au nanochains were effectively separated in the following density gradient layers through centrifugation. The process could be easily controlled by changing the concentration of interference layer, encapsulation layer, and centrifugation rate. This method provided a very facile and controllable way to obtain separated assembled colloidal nanoparticles.
Experimental Section Figure 5. A) Chemical structural formula of PDI-PDMAEMA. B) Digital camera images of centrifugation tube before and after centrifugation and under UV light, with PAA and PDI-PDMAEMA as the two-step encapsulation layers. C) HRTEM image of Au assemblies@PDI-PDMAEMA nanostructure. D) UV/Vis spectra of free PDI-PDMAEMA, Au NPs, and two fractions: f1 and f5. E) Fluorescence spectra and F) transient fluorescence spectra (logarithmic scale on y axis) of free PDI-PDMAEMA and two fractions: f1 and f5. Chem. Eur. J. 2015, 21, 1 – 7
www.chemeurj.org
These are not the final page numbers! ÞÞ
Materials: D2O was purchased from J&K Chemicals Company, and other chemicals were purchased from Beijing Chemical Reagent Co., Inc. All the reagents were A.R. grade and used as received without further purification. Synthesis of Au NPs: Gold nanoparticles were synthesized on two steps by seed-mediated method using cetyltrimethylammonium
5
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
Full Paper chloride (CTAC) as surfactant. Seed solution was prepared by vigorous mixing of 10 mL of HAuCl4 (2.5 104 m) aqueous CTAC (0.1 m) solution with 0.45 mL NaBH4 (0.02 m). The growth solution was prepared by adding 5 mL of HAuCl4 (5 mm), 50 mL of NaBr (0.01 m) and 500 mL of ascorbic acid (0.1 m) to 100 mL of aqueous CTAC (0.1 m) solution. The seed solution was aged for 60 min and diluted 10 times before adding it to the growth solution. To this colorless growth solution, depending on the size of particles required, 2 mL of diluted seed solution was added under vigorous agitation and left undisturbed one week. Density gradient preparation: All separation experiments were performed using a Beckman Optima L-100XP ultracentrifuge. Typically the separation layer step gradient was made using 50 %, 60 %, 70 %, 80 %, 90 %, and 100 % (by volume) solutions of D2O in H2O. For instance, a volume ratio of D2O/H2O = 8:2 was used to make the 80 % solution. A step gradient was created directly in Beckman centrifuge tubes by adding layers to the tube with decreasing density (that is, lower D2O concentration). To make a (50 % + 60 % + 70 % + 80 % + 90 % + 100 %) gradient, 1 mL of 100 % solutions of D2O in H2O was first added to the centrifuge tube, then 1 mL 90 % solutions of D2O in H2O was slowly layered above the 100 % layer. The subsequent layers were made following the same procedure and resulted in a density gradient along the centrifuge tube.
(2011CBA00503), and Program for Changjiang Scholars and Innovative Research Team in University. Keywords: centrifugation · core–shell nanostructures · density gradients · self-assembly · separation methods
[1] H. Wang, L. Chen, X. Shen, L. Zhu, J. He, H. Chen, Angew. Chem. Int. Ed. 2012, 51, 8021 – 8025; Angew. Chem. 2012, 124, 8145 – 8149. [2] T. Chen, M. Pourmand, A. Feizpour, B. Cushman, B. M. Reinhard, J. Phys. Chem. Lett. 2013, 4, 2147 – 2152. [3] O. L. Muskens, V. Giannini, J. A. Snchez-Gil, J. Gmez Rivas, Nano. Lett. 2007, 7, 2871 – 2875. [4] A. Kinkhabwala, Z. Yu, S. Fan, Y. Avlasevich, K. Mullen, W. E. Moerner, Nat. Photonics 2009, 3, 654 – 657. [5] A. Lee, G. F. S. Andrade, A. Ahmed, M. L. Souza, N. Coombs, E. Tumarkin, K. Liu, R. Gordon, A. G. Brolo, E. Kumacheva, J. Am. Chem. Soc. 2011, 133, 7563 – 7570. [6] J. Du, B. Zhu, X. Chen, Small 2013, 9, 4104 – 4111. [7] H.-E. Lee, H. K. Lee, H. Chang, H.-Y. Ahn, N. Erdene, H.-Y. Lee, Y.-S. Lee, D. H. Jeong, J. Chung, K. T. Nam, Small 2014, 10, 3007 – 3011. [8] J.-H. Lee, J.-M. Nam, K.-S. Jeon, D.-K. Lim, H. Kim, S. Kwon, H. Lee, Y. D. Suh, ACS Nano 2012, 6, 9574 – 9584. [9] P. Yuan, R. Ma, Z. Guan, N. Gao, Q.-H. Xu, ACS Appl. Mater. Interfaces 2014, 6, 13149 – 13156. [10] Y. Wang, M. Becker, L. Wang, J. Liu, R. Scholz, J. Peng, U. Gçsele, S. Christiansen, D. H. Kim, M. Steinhart, Nano Lett. 2009, 9, 2384 – 2389. [11] B. Gao, G. Arya, A. R. Tao, Nat. Nanotechnol. 2012, 7, 433 – 437. [12] Y. Wang, A. D. Hollingsworth, S. K. Yang, S. Patel, D. J. Pine, M. Weck, J. Am. Chem. Soc. 2013, 135, 14064 – 14067. [13] H. Xia, G. Su, D. Wang, Angew. Chem. Int. Ed. 2013, 52, 3726 – 3730; Angew. Chem. 2013, 125, 3814 – 3818. [14] Y. Liu, X. Han, L. He, Y. Yin, Angew. Chem. Int. Ed. 2012, 51, 6373 – 6377; Angew. Chem. 2012, 124, 6479 – 6483. [15] H. Zhang, D. Wang, Angew. Chem. Int. Ed. 2008, 47, 3984 – 3987; Angew. Chem. 2008, 120, 4048 – 4051. [16] M. Yang, G. Chen, Y. Zhao, G. Silber, Y. Wang, S. Xing, Y. Han, H. Chen, Phys. Chem. Chem. Phys. 2010, 12, 11850 – 11860. [17] M. S. Arnold, A. A. Green, J. F. Hulvat, S. I. Stupp, M. C. Hersam, Nat. Nanotechnol. 2006, 1, 60 – 65. [18] L. Bai, X. Ma, J. Liu, X. Sun, D. Zhao, D. G. Evans, J. Am. Chem. Soc. 2010, 132, 2333 – 2337. [19] M. L. Mastronardi, F. Hennrich, E. J. Henderson, F. Maier-Flaig, C. Blum, J. Reichenbach, U. Lemmer, C. Kbel, D. Wang, M. M. Kappes, G. A. Ozin, J. Am. Chem. Soc. 2011, 133, 11928 – 11931. [20] X. Ma, Y. Kuang, L. Bai, Z. Chang, F. Wang, X. Sun, D. G. Evans, ACS Nano 2011, 5, 3242 – 3249. [21] C. Zhang, L. Luo, J. Luo, D. G. Evans, X. Sun, Chem. Commun. 2012, 48, 7241 – 7243. [22] S. Song, Y. Kuang, L. Luo, X. Sun, Dalton Trans. 2014, 43, 5994 – 5997. [23] M. Sindoro, Y. Feng, S. Xing, H. Li, J. Xu, H. Hu, C. Liu, Y. Wang, H. Zhang, Z. Shen, H. Chen, Angew. Chem. Int. Ed. 2011, 50, 9898 – 9902; Angew. Chem. 2011, 123, 10072 – 10076. [24] Y. Kuang, J. Liu, X. Sun, J. Phys. Chem. C 2012, 116, 24770 – 24776. [25] A. Einstein, Ann. Phys. 1905, 322, 549 – 560. [26] C. Wang, Y. Kuang, L. Luo, X. Sun, J. Mater. Chem. C 2013, 1, 4146 – 4152.
Typical centrifugation of PAA coating: The pre-concentrated 20 nm Au nanoparticle solution (1 mL) was put on top of the gradient and the buffer solution contained 0.1 mm CTAC were in the 10 % D2O solution layer. The assembly (0.05 m NaCl) and coating (1 mg mL1 PAA) solution was in the 30 % and 40 % layer, respectively. The buffer and assembly layer were 400 mL, the coating layer was 1 mL. Then the tubes were centrifuged at 7 000 rpm for 15 min (a photograph of the ultracentrifuge tube after separation is shown in Figure 2 B). Fractions (200 mL each) were obtained by orderly manual extraction along the centrifuge tube after ultracentrifugation. Typical centrifugation of PDI-PDMAEMA coating: The coating layer contains three parts: the PAA solution (1 mg mL1), the PDIPDMAEMA solution (1 mg mL1) and a buffer layer made by 35 % solution of D2O in water was put between the two solutions. Then the tubes were centrifuged at 7 000 rpm for 15 min (a photograph of the ultracentrifuge tube after separation is shown in Figure 5 B). Fractions (200 mL each) were obtained by orderly manual extraction along the centrifuge tube after centrifugation. Characterization: The size and morphologies of samples were characterized by transmission electron microscopy (TEM; FEI G220) and high-resolution transmission electron microscopy (HRTEM; JEOL, JEM-2100, 200 kV). Fractions obtained by gradient separation were directly dried on carbon film supported on copper grids. The optical properties of samples were characterized by UV/vis absorbance spectroscopy (UV-2501PC, Shimadzu). The fluorescence spectra were taken using a fluorescence spectrophotometer (HITACHI, F-7000). The size distribution of samples was characterized by dynamic light scattering (DLS, Malvern, ZS-90).
Acknowledgements
Received: December 16, 2014 Revised: January 26, 2015 Published online on && &&, 0000
This work has been supported by the Natural Science Foundation of China, the National Basic Research Program 973
&
&
Chem. Eur. J. 2015, 21, 1 – 7
www.chemeurj.org
6
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!
Full Paper
FULL PAPER & Self-Assembly and Separation
A one-tube synthesis approach couples the assembly and separation of colloidal nanoparticles in one centrifuge tube. This method was introduced to construct one-dimensional nanostructures with finite primary units of gold nanoparticles by centrifuging once. Tuning was achieved by changing the concentration of interference salt layer, encapsulation layer, and centrifugation rate.
Chem. Eur. J. 2015, 21, 1 – 7
www.chemeurj.org
These are not the final page numbers! ÞÞ
X. Qi, M. Li, Y. Kuang, C. Wang, Z. Cai, J. Zhang, S. You, M. Yin, P. Wan, L. Luo,* X. Sun* && – && Controllable Assembly and Separation of Colloidal Nanoparticles through a One-Tube Synthesis Based on Density Gradient Centrifugation
7
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
Full Paper
& Self-Assembly and Separation
Controllable Assembly and Separation of Colloidal Nanoparticles through a One-Tube Synthesis Based on Density Gradient Centrifugation Xiaohan Qi,[a] Minglin Li,[a] Yun Kuang,[a] Cheng Wang,[a] Zhao Cai,[a] Jin Zhang,[a] Shusen You,[a, b] Meizhen Yin,[a, b] Pengbo Wan,[a] Liang Luo,*[a] and Xiaoming Sun*[a]
field to overcome the random Brownian motion of NPs and benefit the assembly effect. Such a facile “one-tube synthesis” approach couples assembly and separation in one centrifuge tube by centrifuging once. The method can be tuned by changing the concentration of interference salt layer, encapsulation layer, and centrifugation rate. Furthermore, positively charged fluorescent polymers such as perylenediimide-poly(N,N-diethylaminoethyl methacrylate) could encapsulate the assemblies to give tunable fluorescence properties.
Abstract: Self-assembly of gold nanoparticles into one-dimensional (1D) nanostructures with finite primary units was achieved by introducing a thin salt (NaCl) solution layer into density gradient before centrifugation. The electrostatic interactions between Au nanoparticles would be affected and cause 1D assembly upon passing through the salt layer. A negatively charged polymer such as poly(acrylic acid) was used as an encapsulation/stabilization layer to help the formation of 1D Au assemblies, which were subsequently sorted according to unit numbers at succeeding separation zones. A centrifugal field was introduced as the external
Introduction
assembly degree, orientation, and spacing between repeat units. However, a significant challenge in the self-assembly is the precisely control of ordered structures that adopts specific interparticle orientations, which prevents colloidal jamming. With the aim to achieve synthetic efficiency, a variety of successful strategies have been developed through anisotropic chemical modification of the nanoparticle surface, including the use of templates such as DNA linkers,[8, 9] block copolymers,[1, 10] and patchy particles,[11, 12] or by changing environmental conditions, such as pH value,[13] temperature,[14] and electrostatic interactions.[15] However, the 1D assembly of colloids is regarded the result of dynamic balance between various interactions.[16] For the obtained assembled structures based on current fabrication methods, it is hard to realize ideal monodispersity as well as the corresponding unique optical properties, and the related study remains rather limited. Very recently, the density gradient centrifugation separation method has emerged as a powerful tool to sort nanoparticles with different compositions, sizes, and morphologies in a short time.[18, 19] Moreover, in our preliminary studies, this method also provided the opportunity to isolate/capture intermediate product for the observation and understanding of chemical reaction, growth process, or phase transition.[17, 20–22] Herein, such a method was further developed to capture the “aggregation intermediate”, which means only the initial state of aggregation when several colloids collide with each other in salt solution. Since the aggregation degree can be well-controlled and terminated in certain degree, it is also called as-
Self-assembly of nanoparticles as the building blocks into welldefined nanostructures is of pivotal significance for the design and fabrication of next-generation functional nanodevices. Linear polymer-like one-dimensional (1D) assembled nanostructures in particular are rapidly gaining attention as model systems and a fundamental step owing to their unique physical and chemical properties.[1] In particular, 1D assembled nanostructures based on noble metal nanoparticles (NPs), termed plasmonic polymers, exhibit strong coupling of localized surface plasmon resonance (LSPR) between adjacent NPs,[2] which lead to a wide range of promising applications such as chemical/bio sensing (for example SERS), data storage, nanocircuits, nanoantennas, and optoelectronics.[3–7] The plasmon coupling can be controlled and tuned by changing the
[a] X. Qi, M. Li, Y. Kuang, C. Wang, Z. Cai, J. Zhang, S. You, Prof. M. Yin, Dr. P. Wan, Dr. L. Luo, Prof. X. Sun State Key Laboratory of Chemical Resource Engineering P.O. Box 98, Beijing University of Chemical Technology Beijing 100029 (P. R. China) E-mail: [email protected] [email protected] [b] S. You, Prof. M. Yin Beijing Laboratory of Biomedical Materials Beijing University of Chemical Technology Beijing 100029 (P. R. China) Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/chem.201406507. Chem. Eur. J. 2015, 21, 1 – 7
These are not the final page numbers! ÞÞ
1
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
Full Paper semblies were seriously aggregated after moving down to the bottom of the tube. There were partially chain-like assemblies for f1 and f3, but from f5 afterward, the fractions tended to aggregate into almost bulk randomly rather than nanochains. Digital camera images of centrifugation tube before and after separation (Supporting Information, Figure S1) showed that the color of the fractions in the separation zone changed from violet (up) to light blue (down), indicating the serious aggregation. UV/Vis spectra of the corresponding fractions (Figure 1 B) further confirmed the trend. Compared with isolated Au NPs, there was significant reduced absorbance and slight red-shift for the separated fractions from 520 to 540 nm as the intrinsic absorption peak of Au NPs, indicating the aggregation of Au NPs. For the longitudinal mode, the absorption peak appeared at about 702 nm for f3, which indicated 1D oligomer formation. Blue-shift was observed from f5 afterward, owing to the reduced aspect ratio (thicker diameter) of the random aggregated nanostructures. Thus, even if exposed for a very short time and separated quickly, the obtained intermediate oligomers were still reactive to consume each other by random combination during the separation process. In this context, a polymer (here negatively charged poly(acrylic acid) (PAA)), which was demonstrated as the very effective stabilizers to end reactions and/or isolate intermediates by Chen and co-workers,[16, 23] was introduced to encapsulate the assemblies and avoid further aggregation as the isolation layer. The assembly and separation process with PAA layer (1 mL solution with the concentration of 1 mg mL1) which was placed blow the interference layer, is represented in the Supporting Information, Figure S2 A. After the centrifugation, the color of the fractions changed to red and violet, indicating the limited assembly degree. Compared with isolated Au NPs, there was an obvious red-shift about 20 nm for the extinction peak (Supporting Information, Figure S2B) of the four typical fractions (f1, f3, f5, and f7), owing to the increase of surface refractive index, indicating the existence of PAA shell for each fraction. However, the very concentrated distribution of fractions and the irregular trend of right tails of absorption peaks of fractions indicated the separation process need further optimization. Theoretically, according to the classical sedimentation theory,[22, 24] the sedimentation velocity v of colloidal particles in a given medium can be described as Equation (1):
sembly. Such a facile “one-tube synthesis” approach can realize the self-assembly of Au NPs into 1D nanostructures through density gradient ultracentrifugation rate separation. NaCl solution was introduced as the interference layer to change the electrostatic interactions between Au NPs and induce the assembly behavior. Centrifugal field was introduced as the external field to provide tunable gravity to overcome the random Brownian motion of NPs. On the basis of assemblies isolated by encapsulation of polyacrylic acid (PAA) as the stabilization layer, the obtained various sized 1D Au nanochains were effectively separated in the following density gradient layers through centrifugation. All the processes were realized in one centrifuge tube at one time. Furthermore, the fluorescent species such as perylenediimide–poly(N, N-diethylaminoethyl methacrylate) (PDI-PDMAEMA) could also be introduced onto the surface of the assemblies to give fluorescence properties which could be controlled by tuning the longitudinal surface plasmon resonance.
Results and Discussion Figure 1 A shows a representation of the assembly and separation procedure through centrifugation. First, the positively charged quasispherical Au NPs with a size of (20 2) nm were synthesized in cetyltrimethylammonium chloride (CTAC) solution as the target assembly units. Their size uniformity and
Figure 1. A) Illustration of the assembly and separation of Au NPs in a density gradient centrifugation system. B) UV/Vis spectra of Au NPs and five typical fractions.
sphericity minimized the shape effect on the self-assembly of the NPs. A buffer layer including 0.1 mm CTAC solution was placed below Au NPs to prevent spontaneous assembly of Au NPs in ethylene glycol (EG). An interference layer including 400 uL 0.05 mm NaCl solution was introduced to break the surface charge balance of Au NPs and thereby induce the assembly. A multilayer density gradient (40 %, 50 %, 60 %, 70 %, 80 %, 90 %, and 100 %) was made as the separation zone by sequentially layering aqueous solutions with different concentrations of EG in a centrifuge tube. The centrifugal field provided a centrifugal force (9645 g, 7500 rpm) to drive Au NPs through the buffer layer and into the separation zone. After centrifugation for 30 min, Au NPs were driven from the interference layer into the subsequent separation zone. However, as shown in the Supporting Information, Figure S1, transmission electron microscope (TEM) images of five typical fractions (f1, f3, f5, f7, and f9) indicated that the separated as&
&
Chem. Eur. J. 2015, 21, 1 – 7
www.chemeurj.org
v ¼ 2ð1p 1s Þ r 2 g0 =ð9hs Þ
ð1Þ
where v is the sedimentation velocity of colloidal nanoparticle, 1p and 1s are the densities of nanoparticle and the surrounding fluid, r is the radius of nanoparticle, g’ is the centrifugal acceleration, and hs is the viscosity of the surrounding fluid. Hence, the density and the radius of nanoparticle was constant for the assemblies after passing through PAA layer, and the separation effect was determined by the centrifugation rate and the solvent. For the separation process with EG/H2O as the solvent, the relative large viscosity (EG: 22.1 MPa s at 20 8C) of surrounding fluid resulted in long centrifugation time and espe2
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!
Full Paper cially more energy needed for passing through the gradient phase interfaces. The separation zone was thus applied by using 50 %, 60 %, 70 %, 80 %, 90 %, and 100 % (by volume) solutions of deuteroxide in water as the gradient solvents, and PAA solution was applied as the encapsulation layer, as shown in Figure 2 A. The digital camera images before and after centrifugation with 7000 rpm for 15 min (Figure 2 B) showed that the fractions distributed separately, and the color changed slightly from light pink to violet, indicating the limited assembly degree of Au NPs and ideal separation effect. As the salt layer and encapsulation layer were both very thin, and thereby the time for each assembled fraction reaching to the top layer of density gradient was almost the same with the biggest time interval no more than 1 min. Besides the obvious red-shift about 20 nm for the extinction peak of the all fractions (Figure 2 C) compared to isolated Au NPs, there was also a regular red-shift trend (from 530 to 550 nm for f1–f7) for the extinction peak and increased absorbance of right tail for each fraction. Different with the fractions obtained by the separation without PAA, there was no strong longitudinal peaks as PAA was involved,
Figure 3. A) TEM images of four typical fractions: f1, f3, f5, and f7 (scale bars: 100 nm), HRTEM image of nanostructure (scale bar: 5 nm), and representation of the obtained nanostructure. B) Size distribution of the four typical fractions.
Furthermore, the dynamic light scattering (DLS; Supporting Information, Figures S3 and S4) results also exhibited the increase trend of overall size distribution for the fractions from top to down, confirming the corresponding dominant constitution. To obtain deeper insight into the one-tube synthesis process, control experiments by changing Figure 2. A) Illustration of the assembly and separation of Au NPs in a density gradient centrifugation system with the concentration of interference a PAA layer. B) Digital camera images of centrifugation tube before and after centrifugation. C) UV/Vis spectra of Au NPs and the four fractions. layer (NaCl) and encapsulation layer (PAA) and the centrifugation rate were carried out to investigate the assembly behavior which was due to the limited assembly degree based on relaof Au NPs. For the study of the effect of interference layer, tive small size of Au nanoparticles (20 nm) as the assembly three different concentrations of 400 uL NaCl solution, namely unit. Figure 3 A shows the TEM images of four typical fractions 0, 0.05 and 0.1 m, were compared (Figure 4 A). As the concen(f1, f3, f5, and f7), which clearly exhibited the effective assemtration of NaCl was 0 m, there was no obvious aggregation bly and separation of linear nanostructures as monomers, after centrifugation, and the color of each fraction was still the dimers, tri/tetramers, and oligomers. High-resolution TEM characteristic red of Au NPs. The corresponding UV/Vis spectra (HRTEM) images confirmed the Au assemblies@PAA core–shell of assemblies and isolated Au NPs were almost the same structure, and the mean shell thickness were 1.9 nm. After we except the red-shift for the assemblies induced by the coating carefully rechecked the TEM images of f1, we confirmed that of PAA (Supporting Information, Figure S5 A). As the concentrathey consisted of single nanoparticles. The coating of PAA was tion of NaCl was 0.05 m, the regular assembly was realized, the carried out in the centrifuge tube, so it is unlikely that it will UV-Vis spectra of various fractions (Supporting Information, occur with vigorous stirring. Thus, we cannot guarantee that Figure S5B) was the same as Figure 2 C. At a higher concentrathe shell thickness is the same between assembled nanopartition of NaCl as 0.1 m, the color of the dominant fractions was cles. The varied shell thicknesses can be observed under TEM bluish violet, which indicates over-aggregation during the asobservation and confirmed by broadened absorption peak (f1, sembly process, and the broadened absorption peaks in UV/ Figure 2 C) even for the single nanoparticles (f1, as confirmed Vis spectra especially the bottom fractions such as f7 (Supportby TEM images). The size distribution (Figure 3 B) of the four ing Information, Figure S5C) also validated the result. Further typical fractions based on TEM characterization showed that evidence of the effect of NaCl concentration is more clearly the mean size of f1, f3, f5, and f7 were about (23 3) nm, seen from the comparison of corresponding UV/Vis spectra of (51 17) nm, (110 30) nm, and (142 32) nm, respectively. Chem. Eur. J. 2015, 21, 1 – 7
www.chemeurj.org
These are not the final page numbers! ÞÞ
3
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
Full Paper Furthermore, as the basic parameter to control separation effect, three centrifugation rates as 3000, 7000, and 11 000 rpm were applied to monitor the assembly process of Au nanoparticles, respectively. For 3000 rpm, the centrifugal force was not enough, and the NPs moved slowly. There was much more time for Au NPs passing through NaCl layer and separation under 3000 rpm, thereby the NPs aggregated seriously and randomly as a result of Brownian motion. The UV/Vis spectra of various fractions (Supporting Information, Figure S7 A) confirmed the seriously aggregation, especially the over-broadened and red-shifted absorption peak of f5, and the relatively blue-shifted peak of f7 owing to the reduced aspect ratio by over-aggregation (Supporting Information, Figure S8). When the centrifugation rate was 7000 rpm, the time for the Au NPs staying in the assembly layer was long enough to form Au chains with less than 10 NPs. For 11 000 rpm, the relative high centrifugation rate induced relative high g force, which would provide a strong directional driving force to facilitate the linear aggregation for the unstable intermediates just after passing through NaCl and PAA layers during centrifugation. The UV/Vis spectra (Supporting Information, Figure S7C) showed that there was a regular red-shift trend for the fractions from the upper layer to below, but also an obvious increase absorbance trend for the right tail of absorption peaks, which indicated the longer linear aggregation (Supporting Information, Figure S8). The comparison of corresponding absorption spectra of f5 based on the three parallel experiments was also summarized in Figure 4 F. Owing to the assembly process was carried out during centrifugation, besides the electrostatic force, the assembly process was mainly due to the competition between Brownian motion and external field. According to Stokes–Einstein equation[25] [Eq. (2)], diffusion constant D of nanoparticles could be calculated as:
Figure 4. Digital camera images of tubes before and after centrifugation with the change of A) NaCl concentration from 0 to 0.1 m, C) PAA concentration from 0.5 to 2 mg mL1, E) centrifugation rate from 3000 to 11 000 rpm. B), D, F) UV/Vis spectra of Au NPs and the typical fraction f5 based on the three parallel experiments.
f5 as the typical representative fraction based on the three parallel experiments, shown in Figure 4 B. An appropriate concentration of NaCl was pivotal for the control of assembly degree. The concentration of PAA as the encapsulation layer also played an important role in isolating assemblies and the subsequent separation. PAA solution (1 mL) with low concentration as 0.5 mg mL1 could not provide effective isolation for various assemblies. After centrifugation, the color of the dominant assemblies was bluish violet (Figure 4 C), indicating over-aggregation such as in the situation without PAA (Supporting Information, Figure S1). The corresponding UV/Vis spectra (Supporting Information, Figure S6 A) showed that the absorption peaks of upper fractions were almost the same as isolated Au NPs (above f5), and the lower fractions (below f7) were over-aggregated owing to the significant broadened and red-shifted absorption peaks. As the concentration of PAA was 1 mg mL1, it was the appropriate value and most PAA was consumed without causing side effect (Figure 2 B). With increasing concentration of PAA to 2 mg mL1 could cause a significant side effect: overcoating the unstable intermediate assembles by PAA could induce overloading of surface charge, which separated the nanoparticles into smaller fractions and even individual nanoparticles. The color of fractions in the centrifugation tube remained the characteristic red of isolated Au NPs, indicating the very limited assembly degree that was confirmed by the redshift and the inadequate broadening for all the fractions according to UV/Vis spectra (Supporting Information, Figure S6 C). As revealed by the comparison of corresponding UV/ Vis spectra of f5 based on the three parallel experiments (Figure 4 D), the concentration of PAA was further found to be important to influence the assembly/disassembly behavior. &
&
Chem. Eur. J. 2015, 21, 1 – 7
www.chemeurj.org
D ¼ kB T=ð6 prhs Þ
ð2Þ
where kB is Boltzmann’s constant and T is the absolute temperature. As long as the external environment (T and hs) was constant, the Brownian motion was constant as 6.64 mm2 s1 according to the calculation (we took trimer in pure D2O as the calculation model), which means a trimer particle is displaced about 1.84 mm in one second. For the effect of centrifugation rate (external field), the sedimentation velocity (equation (1)) was calculated to be 33.17 mm s1 at 3000 rpm, 206.14 mm s1 at 7000 rpm and 443.38 mm s1 at 11 000 rpm, respectively (Table 1). Thus, the higher centrifugation rate benefited the assembly in one direction, rather than random aggregation induced by Brownian motion. Moreover, the average translational friction coefficient of the falling assemblies was determined by the orientation of assembled nanostructures, which was the same as the direction of external field. A high centrifugation rate was also helpful to obtain relative regular 1D nanostructures during the separation process. Considering of the electrostatic attraction, the surface coating could be further realized by adding other positively charged functional polymers, such as fluorescent polymer per4
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!
Full Paper excitation rate of PDI-PDMAEMA based on LSPR coupling. Similarly, f1 showed presented a little better LSPR match than f5. After diluting the fractions according to the difference value of absorbance of fluorescent and non-fluorescent fractions to the same level, the fluorescence spectra of normalized free PDIPDMAEMA, f1 and f5 were detected, as shown in Figure 5 E. The fluorescence intensity of f1 and f5 clearly verified better enhancement than free PDI-PDMAEMA, and f5 exhibited more fluorescence enhancement than that of f1. The evidence was further complemented with time-resolved fluorescence measurements (Figure 5 F). The fluorescence lifetime of free PDIPDMAEMA, f1 and f5 were almost the same, namely 6.9, 5.6, and 6.8 ns, respectively. Therefore, the fluorescence enhancement was mainly induced by the energy transfer from Au assemblies to the PDI-PDMAEMA shell owing to the match of LSPR and excitation band of PDI-PDMAEMA,[26] and the match degree was easily controlled by the assembly degree from the one-tube synthesis. Compared to other approaches, the proposed one-tube synthesis could realize convenient assembly and separation in one tube at one time, and more importantly, allow the isolation of the initial assembled 1D nanostructures from salt layer duly in very short time, which was the guarantee of the formation of 1D nanostructures rather than overjammed aggregates, as demonstrated in control experiments (Supporting Information, Figure S10). At the same time, using such technique, quick chemical environments switching was realized, salt and PAA did not exist in one solution. The subsequent coating with PAA or other polymers just after assembly provided convenient protection and functionalization.
Table 1. Calculation of Brownian motion and sedimentation rate of assemblies in D2O/H2O based on three different centrifugation rates. Rate [rpm] 3
1P [kg m ] 1s [kg m3] R [nm] g’ [m s2] hs [Pa s] D [mm2s1] v [mm s1]
3000
7000 4
1.30 10 1.10 103 30 1.52 104 1.095 103 6.65 33.17
12 000 4
1.30 10 1.10 103 30 9.45 104 1.095 103 6.65 206.14
1.30 104 1.10 103 30 2.03 105 1.095 103 6.65 43.38
ylenediimide-poly(N, N-diethylaminoethyl methacrylate) (PDIPDMAEMA; Figure 5 A; see details of the synthesis in the Supporting Information). Purple PDI-PDMAEMA was added as the second encapsulation layer (Figure 5 B). The sum of shell thickness of obtained assemblies was confirmed as being 5 nm from HRTEM image (Figure 5 C), and the 2 nm-thick inner shell of PAA acted as the spacer between fluorescent PDI-PDMAEMA and Au NPs to avoid the quenching induced by directly contact. After the centrifugation, the assemblies were separated in various layers, and exhibited red fluorescence under UV lamp. We chose f1 and f5 as the model fractions to compare the fluorescence intensity. Figure 5 D shows the absorption spectra of free PDI-PDMAEMA, Au NPs and bishelled fractions: f1 and f5. There were three characteristic absorption peaks for PDIPDMAEMA: 447, 541, and 581 nm, and the shell induced redshifted LSPR of f1 and f5 exhibited better match with the peak at 541 nm of PDI-PDMAEMA than Au NPs, indicating the better
Conclusion Controllable assembly of Au nanoparticles into one-dimensional nanostructures and consequent separation was achieved by a one-tube synthesis that requires centrifugation only once. NaCl solution was introduced as the interference layer to change the electrostatic interactions between Au NPs and induce the assembly. Charged polymers such as polyacrylic acid (PAA) or even fluorescent species such as perylenediimide-poly(N,N-diethylaminoethyl methacrylate) (PDI-PDMAEMA) could be introduced as encapsulation/stabilization layer and functionalization layer. Subsequently, the obtained variously sized 1D Au nanochains were effectively separated in the following density gradient layers through centrifugation. The process could be easily controlled by changing the concentration of interference layer, encapsulation layer, and centrifugation rate. This method provided a very facile and controllable way to obtain separated assembled colloidal nanoparticles.
Experimental Section Figure 5. A) Chemical structural formula of PDI-PDMAEMA. B) Digital camera images of centrifugation tube before and after centrifugation and under UV light, with PAA and PDI-PDMAEMA as the two-step encapsulation layers. C) HRTEM image of Au assemblies@PDI-PDMAEMA nanostructure. D) UV/Vis spectra of free PDI-PDMAEMA, Au NPs, and two fractions: f1 and f5. E) Fluorescence spectra and F) transient fluorescence spectra (logarithmic scale on y axis) of free PDI-PDMAEMA and two fractions: f1 and f5. Chem. Eur. J. 2015, 21, 1 – 7
www.chemeurj.org
These are not the final page numbers! ÞÞ
Materials: D2O was purchased from J&K Chemicals Company, and other chemicals were purchased from Beijing Chemical Reagent Co., Inc. All the reagents were A.R. grade and used as received without further purification. Synthesis of Au NPs: Gold nanoparticles were synthesized on two steps by seed-mediated method using cetyltrimethylammonium
5
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
Full Paper chloride (CTAC) as surfactant. Seed solution was prepared by vigorous mixing of 10 mL of HAuCl4 (2.5 104 m) aqueous CTAC (0.1 m) solution with 0.45 mL NaBH4 (0.02 m). The growth solution was prepared by adding 5 mL of HAuCl4 (5 mm), 50 mL of NaBr (0.01 m) and 500 mL of ascorbic acid (0.1 m) to 100 mL of aqueous CTAC (0.1 m) solution. The seed solution was aged for 60 min and diluted 10 times before adding it to the growth solution. To this colorless growth solution, depending on the size of particles required, 2 mL of diluted seed solution was added under vigorous agitation and left undisturbed one week. Density gradient preparation: All separation experiments were performed using a Beckman Optima L-100XP ultracentrifuge. Typically the separation layer step gradient was made using 50 %, 60 %, 70 %, 80 %, 90 %, and 100 % (by volume) solutions of D2O in H2O. For instance, a volume ratio of D2O/H2O = 8:2 was used to make the 80 % solution. A step gradient was created directly in Beckman centrifuge tubes by adding layers to the tube with decreasing density (that is, lower D2O concentration). To make a (50 % + 60 % + 70 % + 80 % + 90 % + 100 %) gradient, 1 mL of 100 % solutions of D2O in H2O was first added to the centrifuge tube, then 1 mL 90 % solutions of D2O in H2O was slowly layered above the 100 % layer. The subsequent layers were made following the same procedure and resulted in a density gradient along the centrifuge tube.
(2011CBA00503), and Program for Changjiang Scholars and Innovative Research Team in University. Keywords: centrifugation · core–shell nanostructures · density gradients · self-assembly · separation methods
[1] H. Wang, L. Chen, X. Shen, L. Zhu, J. He, H. Chen, Angew. Chem. Int. Ed. 2012, 51, 8021 – 8025; Angew. Chem. 2012, 124, 8145 – 8149. [2] T. Chen, M. Pourmand, A. Feizpour, B. Cushman, B. M. Reinhard, J. Phys. Chem. Lett. 2013, 4, 2147 – 2152. [3] O. L. Muskens, V. Giannini, J. A. Snchez-Gil, J. Gmez Rivas, Nano. Lett. 2007, 7, 2871 – 2875. [4] A. Kinkhabwala, Z. Yu, S. Fan, Y. Avlasevich, K. Mullen, W. E. Moerner, Nat. Photonics 2009, 3, 654 – 657. [5] A. Lee, G. F. S. Andrade, A. Ahmed, M. L. Souza, N. Coombs, E. Tumarkin, K. Liu, R. Gordon, A. G. Brolo, E. Kumacheva, J. Am. Chem. Soc. 2011, 133, 7563 – 7570. [6] J. Du, B. Zhu, X. Chen, Small 2013, 9, 4104 – 4111. [7] H.-E. Lee, H. K. Lee, H. Chang, H.-Y. Ahn, N. Erdene, H.-Y. Lee, Y.-S. Lee, D. H. Jeong, J. Chung, K. T. Nam, Small 2014, 10, 3007 – 3011. [8] J.-H. Lee, J.-M. Nam, K.-S. Jeon, D.-K. Lim, H. Kim, S. Kwon, H. Lee, Y. D. Suh, ACS Nano 2012, 6, 9574 – 9584. [9] P. Yuan, R. Ma, Z. Guan, N. Gao, Q.-H. Xu, ACS Appl. Mater. Interfaces 2014, 6, 13149 – 13156. [10] Y. Wang, M. Becker, L. Wang, J. Liu, R. Scholz, J. Peng, U. Gçsele, S. Christiansen, D. H. Kim, M. Steinhart, Nano Lett. 2009, 9, 2384 – 2389. [11] B. Gao, G. Arya, A. R. Tao, Nat. Nanotechnol. 2012, 7, 433 – 437. [12] Y. Wang, A. D. Hollingsworth, S. K. Yang, S. Patel, D. J. Pine, M. Weck, J. Am. Chem. Soc. 2013, 135, 14064 – 14067. [13] H. Xia, G. Su, D. Wang, Angew. Chem. Int. Ed. 2013, 52, 3726 – 3730; Angew. Chem. 2013, 125, 3814 – 3818. [14] Y. Liu, X. Han, L. He, Y. Yin, Angew. Chem. Int. Ed. 2012, 51, 6373 – 6377; Angew. Chem. 2012, 124, 6479 – 6483. [15] H. Zhang, D. Wang, Angew. Chem. Int. Ed. 2008, 47, 3984 – 3987; Angew. Chem. 2008, 120, 4048 – 4051. [16] M. Yang, G. Chen, Y. Zhao, G. Silber, Y. Wang, S. Xing, Y. Han, H. Chen, Phys. Chem. Chem. Phys. 2010, 12, 11850 – 11860. [17] M. S. Arnold, A. A. Green, J. F. Hulvat, S. I. Stupp, M. C. Hersam, Nat. Nanotechnol. 2006, 1, 60 – 65. [18] L. Bai, X. Ma, J. Liu, X. Sun, D. Zhao, D. G. Evans, J. Am. Chem. Soc. 2010, 132, 2333 – 2337. [19] M. L. Mastronardi, F. Hennrich, E. J. Henderson, F. Maier-Flaig, C. Blum, J. Reichenbach, U. Lemmer, C. Kbel, D. Wang, M. M. Kappes, G. A. Ozin, J. Am. Chem. Soc. 2011, 133, 11928 – 11931. [20] X. Ma, Y. Kuang, L. Bai, Z. Chang, F. Wang, X. Sun, D. G. Evans, ACS Nano 2011, 5, 3242 – 3249. [21] C. Zhang, L. Luo, J. Luo, D. G. Evans, X. Sun, Chem. Commun. 2012, 48, 7241 – 7243. [22] S. Song, Y. Kuang, L. Luo, X. Sun, Dalton Trans. 2014, 43, 5994 – 5997. [23] M. Sindoro, Y. Feng, S. Xing, H. Li, J. Xu, H. Hu, C. Liu, Y. Wang, H. Zhang, Z. Shen, H. Chen, Angew. Chem. Int. Ed. 2011, 50, 9898 – 9902; Angew. Chem. 2011, 123, 10072 – 10076. [24] Y. Kuang, J. Liu, X. Sun, J. Phys. Chem. C 2012, 116, 24770 – 24776. [25] A. Einstein, Ann. Phys. 1905, 322, 549 – 560. [26] C. Wang, Y. Kuang, L. Luo, X. Sun, J. Mater. Chem. C 2013, 1, 4146 – 4152.
Typical centrifugation of PAA coating: The pre-concentrated 20 nm Au nanoparticle solution (1 mL) was put on top of the gradient and the buffer solution contained 0.1 mm CTAC were in the 10 % D2O solution layer. The assembly (0.05 m NaCl) and coating (1 mg mL1 PAA) solution was in the 30 % and 40 % layer, respectively. The buffer and assembly layer were 400 mL, the coating layer was 1 mL. Then the tubes were centrifuged at 7 000 rpm for 15 min (a photograph of the ultracentrifuge tube after separation is shown in Figure 2 B). Fractions (200 mL each) were obtained by orderly manual extraction along the centrifuge tube after ultracentrifugation. Typical centrifugation of PDI-PDMAEMA coating: The coating layer contains three parts: the PAA solution (1 mg mL1), the PDIPDMAEMA solution (1 mg mL1) and a buffer layer made by 35 % solution of D2O in water was put between the two solutions. Then the tubes were centrifuged at 7 000 rpm for 15 min (a photograph of the ultracentrifuge tube after separation is shown in Figure 5 B). Fractions (200 mL each) were obtained by orderly manual extraction along the centrifuge tube after centrifugation. Characterization: The size and morphologies of samples were characterized by transmission electron microscopy (TEM; FEI G220) and high-resolution transmission electron microscopy (HRTEM; JEOL, JEM-2100, 200 kV). Fractions obtained by gradient separation were directly dried on carbon film supported on copper grids. The optical properties of samples were characterized by UV/vis absorbance spectroscopy (UV-2501PC, Shimadzu). The fluorescence spectra were taken using a fluorescence spectrophotometer (HITACHI, F-7000). The size distribution of samples was characterized by dynamic light scattering (DLS, Malvern, ZS-90).
Acknowledgements
Received: December 16, 2014 Revised: January 26, 2015 Published online on && &&, 0000
This work has been supported by the Natural Science Foundation of China, the National Basic Research Program 973
&
&
Chem. Eur. J. 2015, 21, 1 – 7
www.chemeurj.org
6
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!
Full Paper
FULL PAPER & Self-Assembly and Separation
A one-tube synthesis approach couples the assembly and separation of colloidal nanoparticles in one centrifuge tube. This method was introduced to construct one-dimensional nanostructures with finite primary units of gold nanoparticles by centrifuging once. Tuning was achieved by changing the concentration of interference salt layer, encapsulation layer, and centrifugation rate.
Chem. Eur. J. 2015, 21, 1 – 7
www.chemeurj.org
These are not the final page numbers! ÞÞ
X. Qi, M. Li, Y. Kuang, C. Wang, Z. Cai, J. Zhang, S. You, M. Yin, P. Wan, L. Luo,* X. Sun* && – && Controllable Assembly and Separation of Colloidal Nanoparticles through a One-Tube Synthesis Based on Density Gradient Centrifugation
7
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
