Covalent Synthesis and Optical Characterization of Double-Walled

Apr 26, 2010 - Department of Chemistry, State University of New York at Stony Brook, Stony Brook, New York 11794-3400, and Condensed Matter Physics ...
0 downloads 0 Views 3MB Size
Download PDF
8766

J. Phys. Chem. C 2010, 114, 8766–8773

Covalent Synthesis and Optical Characterization of Double-Walled Carbon Nanotube-Nanocrystal Heterostructures Xiaohui Peng,† Matthew Y. Sfeir,‡ Fen Zhang,† James A. Misewich,‡ and Stanislaus S. Wong*,†,‡ Department of Chemistry, State UniVersity of New York at Stony Brook, Stony Brook, New York 11794-3400, and Condensed Matter Physics and Materials Sciences Department, BrookhaVen National Laboratory, Building 480, Upton, New York 11973 ReceiVed: January 20, 2010; ReVised Manuscript ReceiVed: April 1, 2010

The unique electronic structure and optical properties of double-walled carbon nanotubes (DWNTs) have made them a key focus material of research in recent years. However, the incorporation of DWNTs with quantum dots (QDs) into nanocomposites via a covalent chemical approach as well as the optical properties of the composites have rarely been explored. In particular, we have been interested in this model system to investigate whether nanomaterial heterostructures can provide efficient pathways for charge separation relative to loss mechanisms such as recombination. In this specific work, the synthesis of DWNT-CdSe QD heterostructures obtained by using a conventional covalent protocol has been demonstrated. CdSe QDs with terminal amino groups have been conjugated onto the surfaces of oxidized DWNTs by the formation of amide bonds. The observed trap emission of CdSe is thought to arise from the presence of 2-aminoethanethiol capping ligands and is effectively quenched upon conjugation with the DWNT surface because of the charge transfer from CdSe to DWNTs. Introduction In recent years, double-walled carbon nanotubes (DWNTs), consisting of two coaxial tubules, have attracted significant attention because of their unique properties.1-3 Indeed, because the outer and inner layers in a DWNT retain the basic electronic properties of each constituent graphene monolayer tubule, even though the detailed energy dispersion relations are affected by the interlayer interaction,4,5 the properties of DWNTs are predicted to be as promising as or superior to those of singlewalled carbon nanotubes (SWNTs) and multiwalled carbon nanotubes (MWNTs).6 For instance, with respect to fieldemission properties, it has been found that DWNTs not only maintain higher current densities but also combine the low threshold voltage for electron emission normally characteristic of SWNTs with the efficient emission stability typically associated with MWNTs.7,8 Moreover, the different permutations of interwall interactions between the inner and outer tubes of DWNTs also result in their electronic structure being somewhat different from that of SWNTs, thereby leading to their unique spectral features.4,9-12 By comparison with SWNTs shown to enhance electron transport,13 DWNTs with a high charge mobility have been utilized to construct efficient solar cells by facilitating hole transport.14,15 Moreover, the advanced structural stability,3 decent thermal conductivity, and high mechanical stiffness8 make DWNTs promising candidates for numerous applications.16,17 Most recently, chemical functionalization studies of DWNTs, including sidewall fluorination, have been carried out.18 Raman and optical absorption spectra results have indicated that fluorination can be used to suppress the optical properties of * To whom correspondence should be addressed. Phone: 631-632-1703 and 631-344-3178. Email: [email protected] and sswong@ bnl.gov. † State University of New York at Stony Brook. ‡ Brookhaven National Laboratory.

CNTs without interfering with the properties of inner tubes.19 That is, it is possible to suppress only the Raman radial breathing modes and absorption contributions from the outer tubes of DWNTs while keeping intact the intrinsic electronic structure of the inner tubes, as demonstrated by photoluminescence maps and optical absorption spectra.19 The synthesis of carbon nanotube-nanocrystal (CNT-NC) heterostructures has been of interest in research in recent years because of their attractive performance in a wide range of applications spanning catalysis, sensing, optoelectronics, and biological imaging.20-23 Different strategies have therefore been put forward for the fabrication of CNT-NC nanocomposites. Nonetheless, these fall into two basic classes. One main approach involves the prior synthesis of nanoparticles that are then subsequently connected to functionalized CNTs via either covalent (i.e., organic20,24,25 or biomolecular linkers26,27) or noncovalent28 interactions. The second key approach involves direct deposition of nanoparticles onto the surface of CNTs, either through formation of nanoparticles in situ,29,30 a reduction reaction,23,31 or an electrodeposition process using CNTs as templates.32,33 In the specific context of DWNT-NC heterostructures, fluorine atoms on the outer tubes of DWNTs have been found to be more effective at nucleating and growing 5-7 nm CdSe nanoparticles in situ than oxygen-containing functional groups.34 Nonetheless, a key disadvantage of direct formation of nanoparticles onto CNT surfaces has been the difficulty in defining and controlling the spatial coverage and distribution of NCs. In fact, the in situ chemical synthesis of semiconductor nanoparticles directly onto the surfaces of CNTs usually leads to the generation of structures which may be highly polydisperse and inhomogeneous in shape and size.35 The point therefore is to create well-defined heterostructure junctions that possess predictable chemistry and morphology. Conversely, covalently modifying the CNT surface by using functional groups containing desirable, pendant moieties has

10.1021/jp100580h  2010 American Chemical Society Published on Web 04/26/2010

Synthesis and Characterization of DWNT-NC Heterostructures been an important and popular method for creating CNT-based hybrid nanostructures because it involves a relatively robust, straightforward, and facile protocol.28,36 There are several advantages to this particular approach. First, the shape and size of individual nanoparticles or NCs can be easily tailored by sophisticated synthesis methods prior to combination with CNTs, thereby mitigating the influence of the CNTs themselves on the nucleation and growth processes of either nanoparticles or quantum dots. Second, covalent bonds can rigidly connect the linker molecules and CNTs in a reliable and robust manner, such that the nanoparticles will not become easily dislodged even after either sonication or extensive washing. Third, the physical integrity and hence corresponding optoelectronic properties of the DWNT structure need not be necessarily compromised. In this work, we present the covalent, linkermediated synthesis of a DWNT-NC heterostructure, prepared by anchoring derivatized CdSe quantum dots (QDs) onto the surface of complementarily functionalized DWNTs. The investigation of optoelectronic properties of CNT-QD heterostructures is crucial for developing a fundamental understanding of exciton dissociation via a charge/energy transfer process, thereby providing critical insights into the development of photovoltaic devices. A significant prior observation20 from our lab is the quenching of luminescence emission from the QD when in close proximity to SWNTs because of the charge transfer (CT) process, which could potentially occur either between the excited states of QDs and CNTs or between the trap states of QDs and CNTs.37 It has been demonstrated that the quenching of excitonic luminescence seems to depend only on the relative separation of the NT and QD as it has been observed with chemical linkers20,38 and polymer spacers.39 However, there are few if any studies reporting the influence of surface ligands of QDs on the CT process between QDs and CNTs, thereby potentially revealing insight into the interfacial interaction between these two. In fact, the coupled interaction between QDs and CNTs, even at the interfacial level, as well as the subsequent charge separation of electron-hole pairs, photogenerated in QDs, inevitably influence the nature of the QD photoluminescence behavior. Additionally, it is well known that the capping ligands on the surface of QDs can play a key role in influencing the luminescence behavior of QDs either through improved passivation40-42 or by additional surface trapping due to the alteration of the surface states.43-45 Therefore, by choosing the appropriate capping ligand, it is reasonable to conceive that we are able to probe the CT process via a surfacerelated phenomenon such as trap emission. Therefore, in this work, the role of 2-aminoethanethiol (AET) in modifying the surface of CdSe QDs not only allowed for covalent conjugation between QDs and DWNTs but also provided for the possibility of probing the surface states of CdSe by inducing a characteristic trap emission. As such, the optical properties of heterostructures composed of DWNTs and CdSe QDs were studied by using steady-state fluorescence measurements. The observed trap emission ascribed to AET was consequently quenched by CNT upon conjugation. To this end, for the first time, we have demonstrated that, by controlling the outer surface chemistry of CdSe QDs, it is possible to manipulate charge separation and transfer between trap states of CdSe QDs and CNTs, thereby paving the way for probing interfacial interactions between CdSe and CNTs. Experimental Section 1. Synthesis. Reasonably purified DWNTs have been obtained by purification of commercial DWNTs (Helix Material

J. Phys. Chem. C, Vol. 114, No. 19, 2010 8767 Solutions, Inc.) so as to remove metal catalysts and carbonaceous impurities, including SWNTs and amorphous carbon. Specifically, 50 mg of DWNTs were dispersed in 8 M HNO3 by sonication and then heated to 60 °C for 30 h. The analogous use46 of 18% HCl at 100 °C for 10 h followed by air oxidation at 500 °C for 30 min has been reported as an alternative technique for DWNT buckypaper purification. The resulting, purified DWNTs (pDWNTs) were filtered through a 0.2 µm polycarbonate membrane (Millipore), thoroughly washed by excess water, and ultimately dried, prior to TGA measurements. It has been noted that nitric acid treatment not only purified but also oxidized our DWNTs, as shown by the infrared spectra reported herein. We should state though that we could not totally remove MWNT content from the commercial sample used in our experiments. CdSe QDs have been prepared according to a well-known protocol described in the prior literature.47-49 Upon further processing, oleic acid capping agents of as-prepared CdSe QDs were replaced by AET via a ligand exchange reaction, resulting in water-soluble, amino-terminated CdSe QDs (AET-CdSe QDs). In a typical experiment, a solution of AET in methanol was added to a suspension of as-prepared CdSe QDs. QDs were then immediately precipitated upon ligand exchange. These QDs were then collected by centrifugation and subsequently washed with ethanol and methanol. The resulting QDs were finally redispersed in dimethyl sulfoxide (DMSO). AET-CdSe QDs were attached to pDWNTs by formation of amide bonds between amino terminal groups on the surfaces of AET-CdSe QDs and carboxylic groups on the pDWNTs, via a standard coupling reaction.50 In a typical experiment, 1 mg of pDWNTs was dispersed in 10 mL DMSO by sonication. A total of 80 mg of EDC (N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide)/NHS (N-hydroxysuccinimide) and an aliquot of MES (2-(N-morpholino) ethanesulfonic acid) buffer were subsequently added, thereby allowing for the carboxylic groups of pDWNTs to be activated, followed by addition of 1 mL of AETCdSe QDs. The resulting solution was stirred in the dark for 24 h. Upon completion of the reaction, the solution was vacuumfiltered by using a 0.2 µm polycarbonate membrane (Millipore), extensively washed with DMSO and distilled water, and finally oven-dried for subsequent measurements. 2. Characterization. Low-magnification TEM images were taken at an accelerating voltage of 80 kV on a FEI Tecnai12 BioTwinG2 instrument, equipped with an AMT XR-60 CCD digital camera system. High-resolution images were obtained on a JEOL 2010F instrument, equipped with an INCA EDS system, at accelerating voltages of 200 kV. Specimens for all of these TEM experiments were prepared by dispersing the asprepared product in ethanol, sonicating for 2 min to ensure adequate dispersion of the nanostructures, and dipping one drop of solution onto a 300 mesh Cu grid, coated with a lacey carbon film. Crystallographic and purity information on NT-NC heterostructures were obtained by using powder XRD. To analyze these materials, as-prepared samples were subsequently sonicated and dispersed in ethanol for about 2 min and finally airdried upon deposition onto glass slides. Diffraction patterns of these materials were collected by using a Scintag diffractometer, operating in the Bragg configuration by using Cu KR radiation (λ ) 1.54 Å) from 20 to 80° at a scanning rate of 0.25° per minute. FT-mid-IR data were obtained on a Nexus 670 (Thermo Nicolet) equipped with a single-reflectance zinc selenide (ZnSe) ATR accessory, a KBr beam splitter, and a DTGS KBr detector.

8768

J. Phys. Chem. C, Vol. 114, No. 19, 2010

Peng et al.

Figure 1. TEM images of (A) raw DWNTs and (B) the corresponding oxidized, pDWNTs (B). TGA data of (C) raw DWNTs and (D) the corresponding oxidized pDWNTs.

Solid samples were placed onto a ZnSe crystal. Measurements were obtained in absorbance mode by using the Smart Performer module. UV-visible spectra were collected at high resolution by using a Thermospectronics UV1 with quartz cells maintaining a 10 mm path length. Data were corrected to account for the solvent background. Samples for photoluminescence (PL) spectra were initially dispersed in ethanol by sonication. Steady-state fluorescence data were obtained at room temperature on a home-built confocal microscope coupled to a high-repetition rate (250 kHz) ultrafast Ti: sapphire-based amplified laser system. An optical parametric amplifier was used to generate visible laser pulses (∼200 fs) tuned to an absorption spectrum of the QDs. Before photoexcitation, the laser pulses were temporally broadened and spectrally narrowed to a bandwidth of 2 nm by using a set of interference filters. Emission from the QD solutions was collected into a single-grating spectrometer equipped with both a liquid nitrogen cooled CCD for acquiring data and an avalanche photodiode for dynamical measurements. Fluorescence images were taken by using a confocal fluorescence microscope (Leica, TCS SP5) equipped with an argon ion laser. Samples were excited under 488 nm irradiation, and images were separately collected via two different output channels, which were monitored in the range of 590-620 nm and 750-800 nm, respectively. Results and Discussion 1. Preparation of Oxidized pDWNTs. Commercial, assynthesized DWNTs typically were contaminated with metal particles, amorphous carbon, and SWNTs, as illustrated in Figure 1A. On the basis of the TGA data, as shown in Figure 1 C, upon heating as-synthesized DWNTs to 800 °C under air, the remaining weight % of sample residue was ∼10%, indicating a carbon yield of around 90%. As illustrated in Figure 1C, the data suggested that the bulk of the sample was destroyed at

∼560 °C, which is considered to be a temperature slightly more elevated than that expected for pure SWNTs because of the higher resistance of DWNTs to oxidation as a result of their coaxial structure.46 The shoulder at around 476 °C in the curve of derivative weight versus temperature can be attributed to the presence of amorphous carbon and SWNTs in the sample, likely due to the higher reactivity of amorphous carbon, graphitic carbon, and SWNTs as compared with DWNTs.6,51,52 It is worth noting that, in our study, metal particles, possessing a size similar to that of QDs, can interfere with the observation of DWNT-QD heterostructures under TEM. In addition, the presence of SWNTs may affect the final properties and ultimate performance of DWNT-QD heterostructures. Therefore, it was extremely critical to remove all those extraneous impurities in order to obtain highly pDWNTs, prior to reactive coupling. It is known that nitric acid can not only destroy the metal catalyst but also remove amorphous carbon and chemically active SWNTs. A comparison between TEM images of DWNTs before and after nitric acid treatment, especially under higher magnification (Figure 1A,B), provides a visual confirmation of a much higher sample quality associated with our pDWNTs. TGA measurements were also used to monitor the degree of purity of DWNTs. The results demonstrated that less than 1% of metal catalyst was contained in pDWNTs, and the disappearance of the shoulder, previously observed at 476 °C, indicated that our pDWNTs were, to a certain extent, free from amorphous carbon and SWNTs (Figure 1D) as well, although we have additional data showing that we were unable to completely eliminate MWNTs. 2. Characterization of DWNT-CdSe QD Composites. The oxidation of DWNTs was confirmed by IR spectra as illustrated in Figure 2. The obvious peak for oxidized DWNTs (Figure 2b) at 1700 cm-1 is associated with the asymmetric stretch of carboxylic acid groups, which had been conspicuously absent in the spectrum of the as-synthesized sample (Figure 2a), and

Synthesis and Characterization of DWNT-NC Heterostructures

Figure 2. Infrared spectra of (a) raw DWNTs, (b) oxidized pDWNTs, and (c) DWNT-CdSe QD heterostructures.

Figure 3. (A) TEM as well as (B and C) HRTEM images of DWNT-CdSe QD heterostructures.

provided evidence that pDWNTs had in fact been oxidized by nitric acid. Oxidized DWNTs also could be well dispersed in ethanol and even in water, because the presence of carboxylic groups on the surface of DWNTs enhances their solubility46 in polar solvents. During the process of attaching amine-derivatized QDs onto pDWNT templates, EDC/NHS was used to activate reactive carboxylic groups on the surfaces of the pDWNTs. The resultant formation of amide bonds was demonstrated by the appearance of peaks at 1656 and 1504 cm-1, corresponding to the CdO stretching mode (amide I band) and the CsNsH stretch-bending mode (amide II band), respectively,53,54 in the infrared spectra of the as-obtained DWNT-QD heterostructure (Figure 2c). The morphology of anchored CdSe QDs onto the oxidized DWNT template was evaluated by using TEM images. Specifically, QDs with an average diameter of 3.4 nm were clearly visible on pDWNT bundles, as illustrated in Figure 3A. It was noted that the coverage of QDs on pDWNTs was nonuniform, which could be attributed to either (i) a nonuniform distribution of defect sites, associated with reactive carboxylic acid groups, on the surface of DWNTs or (ii) to inefficient attachment of CdSe QDs onto the NTs themselves as a result of steric hindrance issues arising from possible aggregation of DWNTs. Nonetheless, this observation suggested the potential of covalently conjugating oxidized CNTs with QDs, as a means of minimizing nonspecific QD adsorption on the surfaces of DWNTs, because these QDs were not removed even after

J. Phys. Chem. C, Vol. 114, No. 19, 2010 8769

Figure 4. XRD patterns of (a) raw DWNTs and (b) DWNT-CdSe QD heterostructures. Relevant peaks have been indexed.

extended washing. Figure 3B,C shows typical high-resolution TEM images of different DWNT-CdSe QD heterostructures, wherein distinctive double walls of CNTs and highly crystalline CdSe NCs immobilized on the NT surface could be observed. It is worth noting that some of the tubes appeared to be wrapped in an amorphous coating, which likely originated from solvent residue, incurred during TEM sample preparation. Crystallographic information on commercial DWNTs and asprepared NT-NC heterostructures were obtained by using powder XRD (Figure 4). The peak at ca. 25.4° was associated with the presence of graphite (JCPDS # 75-1621) in CNTs. We noted that for the DWNT-QD heterostructures, peaks at 42.5° and 48.8°, indicated by arrows, though weak, could be reliably assigned to the reflections of hexagonal wu¨rtzite CdSe QDs (JCPDS # 02-0330), again consistent with the formation of DWNT-QD heterostructures. 3. Optical Characterization and Fluorescence Imaging. To test the optical properties of DWNT-CdSe QD heterostructures, both UV-visible and steady-state PL spectra were collected in Figure 5 and Figure 6. For absorption spectra, asprepared CdSe NCs (Figure 5a) exhibited a well-pronounced absorption peak at ∼580 nm, whereas a 5 nm red shift in peak position was noted for AET-CdSe (Figure 5b). Similarly, a physical mixture of oxidized DWNTs with the same concentration of AET-CdSe nanoparticles yielded an absorption spectrum (Figure 5c) identical to that of AET-CdSe QDs, as expected. However, the peak intensity was relatively weaker, an observation attributable to light scattering by DWNTs, causing less photons to be absorbed by the AET-CdSe QDs themselves. No obvious spectral feature was noted in the absorption spectrum (Figure 5d) of DWNT-CdSe QD heterostructures, presumably because of the decreased CdSe concentration in the final composites, which might have been below the detection limit of UV-visible spectroscopy. In addition, the collective absorption profile was affected not only by polydispersity issues in terms of the size and density of adsorbed QDs but also by a reasonably strong background scattering of DWNTs, which might have conceivably overwhelmed the signal generated by CdSe QDs alone. The steady-state luminescence spectrum of AET-CdSe QDs (Figure 6) consists of a high-energy peak, corresponding to excitonic luminescence (or band gap emission) at 599 nm, as well as a broad, low-energy emission band at 800 nm with

8770

J. Phys. Chem. C, Vol. 114, No. 19, 2010

Figure 5. UV-visible spectra of (a) as-prepared CdSe QDs, (b) AET-CdSe QDs, (c) mixture of DWNTs and AET-CdSe QDs, and (d) DWNT-CdSe QD heterostructures.

Figure 6. Steady-state PL spectra of (a) AET-CdSe QDs, (b) a physical mixture of pDWNTs and AET-CdSe QDs, and (c) DWNT-CdSe QD heterostructures.

significant intensity, originating from trap states because of either surface defect sites or internal defects.55-57 The asymmetric excitonic luminescence profile is suggestive not only of the broad size distribution of CdSe QDs but also of spectral fluctuations in their microenvironment. Moreover, the broad trap emission band centered at 800 nm indicated a wide energy distribution in the trap states, implying a variety of defects associated with light emission. By contrast with the AET-CdSe QD solution alone, as-prepared CdSe QDs exhibited a pronounced band gap emission at 582 nm with only a small shoulder centered at 790 nm, merging with the long tail of the band gap emission of CdSe (data not shown), implying that it

Peng et al. is likely that the majority of defect-related emission is induced by the AET ligands. It has been reported that hole trapping on CdSe QD sites due to the presence of AET ligands was energetically favorable, because the thiol redox energy level is situated at an energy level noticeably higher than that of the valence band of CdSe.43 The quenching of the excitonic emission of QDs has been previously observed in SWNT-QD composites, as reported by our group and many others,24,35,58-60 especially when the NT and QD are in close spatial contact with one another. This phenomenon has been ascribed to the electron-acceptor behavior of SWNTs. That is, photoexcited electrons produced upon illumination of QDs can conceivably flow into SWNTs, nonradiatively, prior to their recombination with photogenerated holes in QDs. A similar explanation61-64 has been proposed for the PL quenching of CdSe and CdTe observed, associated with MWNT-QD composites. By contrast, it was interesting to note that, in our DWNT-CdSe QD heterostructures, the excitonic luminescence of QDs was still quite evident even after chemical conjugation with DWNTs. We have found that the exact structure and profile of the PL spectrum ascribed to the DWNT-CdSe heterostructure vary slightly from sample to sample, and that this particular observation is likely a consequence of sample inhomogeneity and a reduced fraction of emitters. However, as compared with the emission profile of AET-CdSe alone, it is clear that the average emission peak position for DWNT-CdSe QD composites blue shifted by 20 nm. Unlike the spectral shifts due to the ligand exchange process, where the surface of the QD is chemically modified, this blue shift is likely electronic in origin and has been previously observed in a similar system.62 Additionally, radiative trap recombination has completely disappeared from the spectrum of the DWNT-CdSe QD composite, suggesting a new competitive kinetic decay process for the trapped carriers. Analogous quenching of CdSe QD trap sites has been noted in the presence of gold spherical nanoparticles and nanorods, and this observation was attributed to a CT process.65 For comparison, control samples consisting of NTs alone were analyzed and noted to maintain featureless spectra. Another control sample consisting of a physical mixture of NTs and QDs, generated simply by adding DWNTs together with AET-CdSe QDs at the same concentration as that of the pure AET-CdSe QD solution, was also optically probed. It turns out that the observed peak position of the excitonic emission in the mixture was similar to that of AET-CdSe QDs alone; moreover, the trap emission state was still significant. Furthermore, the intensity ratio of the trap emission to excitonic emission (It/Ie) for the mixture sample was smaller than that of the AET-CdSe QD solution. Interestingly, when an excess of AET ligands was added to the mixture solution, an increased It/Ie ratio was observed, once again suggesting that the presence of AET ligands contributed to an increase in trap sites as well as to the observed trap emission. Hence, it is reasonable to infer that the equilibrium associated with the association and dissociation kinetics as well as the inherent ordering of AET ligands on the CdSe NC surface was affected by the presence of NTs in solution, resulting in the observed changes to It/Ie. We note that, although it is possible that the absolute value of the intensity difference of Ie between that of the AET-CdSe QD solution and the corresponding physically mixed sample could be due to factors such as scattering and reabsorption by the CNTs, short path length (1 and 2 mm) cells were used to minimize these effects.

Synthesis and Characterization of DWNT-NC Heterostructures

J. Phys. Chem. C, Vol. 114, No. 19, 2010 8771

Figure 7. Confocal scanning fluorescence microscopy images of AET-CdSe QDs collected at (A) ∼600 nm and (B) ∼800 nm. (C) Bright field image of DWNT-CdSe QD heterostructures. (D) Composite, merged image for the DWNT-CdSe QD heterostructure, including (C) and the corresponding fluoresecence images collected at both ∼600 and ∼800 nm. No detectable signal was observed in the near-infrared channel. Scale bars are 10 µm.

On the basis of all of these data, notably the obvious excitonic emission peak with shifted peak position and the quenching of the trap emission for DWNT-CdSe QD composite heterostructures, we can conclude that the conjugation of DWNTs with CdSe QDs plays a significant role in affecting the surface-state emission of the QDs and also profoundly alters the dominant relaxation pathway of photogenerated charge carriers within the QDs themselves. We should also note that all data were obtained reproducibly and were unlikely a consequence of a photoaging effect,66 that is, a worsening of the NC surface or passivation layer during the course of measurements. Figure 7 highlights confocal fluorescence images of both AET-CdSe QDs and DWNT-CdSe heterostructure composites. Images were collected at two different channels corresponding to excitonic emission and trap emission behavior of CdSe QDs, respectively. Not surprisingly, the AET CdSe QDs by themselves showed discernible fluorescence in images obtained at both ∼600 and ∼800 nm emission wavelengths (Figure 7A,B). For the heterostructures, the bright-field image (Figure 7C) along with the composite merged image (Figure 7D) included data collected from both channels and clearly correlated spectral overlap of CdSe excitonic fluorescence with the spatial localization of the NTs. We note that the distribution of fluorescence of CdSe along the NTs themselves was not necessarily uniform, an observation fully consistent with our TEM data. We also cannot rule out the possible existence of either small clusters or single CdSe NCs exhibiting weak fluorescence beyond the functional capabilities of our detector.

4. Insights into the Mechanism. To understand these experimental observations, several pathways for charge-carrier relaxation in the semiconducting QDs should be considered. In addition to the intrinsic QD relaxation processes (e.g., radiative recombination, electron/hole trapping, and Auger recombination), the DWNT-CdSe composite will have two additional quenching mechanisms: (a) CT of one charge carrier as well as (b) resonant energy transfer (RET) of both charge carriers through near-field dipolar coupling. We note that these processes mentioned above compete with each other on characteristic time scales. For example, the lifetime of excitonic luminescence can be on the order of nanoseconds or longer. Trapped charges possess lifetimes ranging from picoseconds to microseconds, depending on the nature of the trap states.57 The rates of the CT and RET are highly dependent upon the separation of donor and acceptor states67,68 and can even happen on time scales as short as hundreds of femtoseconds.69,70 In CdSe QDs, the exciton is highly localized in the interior of the QD and, as such, CT from this state is expected to proceed much more slowly than from surface trap states.71 This observation is due to the particularly steep distance scaling (about one order of magnitude per additional angstrom of separation) of the CT process.72 On the other hand, RET is a longer-range effect and should preferentially affect the exciton state. RET will reduce trap recombination only if the kinetics are so fast that they prevent trapping from occurring. Because trapping has been found to proceed on ultrafast time scales73,74 and because we still detect significant exciton luminescence in

8772

J. Phys. Chem. C, Vol. 114, No. 19, 2010

the steady-state emission, we conclude that CT is a significant relaxation pathway. This hypothesis would explain the quenching of trap emission in our steady-state luminescence measurements. A similar explanation has also been noted to explain optical behavior in ZnO-CNT hybrid systems, wherein a defectrelated emission band did not appear in steady-state luminescence results.75 However, it is still possible that RET contributes by directly quenching a fraction of the photogenerated carriers, thereby reducing the overall emission intensity of the exciton as well. Future time-resolved studies to measure these relative rates are planned. Conclusions In order to take advantage of the unique properties of DWNTs for diverse applications, it has been critical to explore the chemical modification of DWNTs with other functional moieties, such as QDs. In this study, we demonstrated that a conventional covalent approach can be applied to the reliable and reproducible synthesis of DWNT-CdSe QD heterostructures. Moreover, the presence of CT between CdSe QDs and DWNTs can account for the disappearance of trapped emission bands and the observation of excitonic luminescence, as illustrated in steady-state luminescence measurements indicating that the trapped charges were transferred to DWNTs in the DWNT-CdSe QD heterostructure. Hence, we can conclude that chemical conjugation of DWNTs with CdSe plays a significant role in affecting the surface-state emission of the QDs. Importantly, we should note that the interpretation of our results may have been complicated by the presence of MWNTs in our DWNT samples, although the extraneous presence of MWNTs should not influence the validity of our results, because the optical phenomena observed appear to be more fundamentally dependent upon and intrinsically coupled to not only the nature of the QD and chemical linkers themselves but also the presence of a NT motif itself, as opposed to the actual type of NT studied. Thus, although additional experimental and theoretical studies on the effect of chemical functionalization on the properties of DWNTs are still needed, the experimental evidence presented herein suggests that DWNT-QD heterostructures exhibit potential for incorporation into devices such as photovoltaic cells, especially considering the observation of effective charge separation between CdSe and DWNTs. Acknowledgment. We acknowledge the U.S. Department of Energy (DE-AC02-98CH10886) for the spectroscopy work and for personnel support. Moreover, research carried out (in whole or in part, such as a few spectroscopy studies) at the Center for Functional Nanomaterials, Brookhaven National Laboratory, is also supported by the U.S. Department of Energy, Office of Basic Energy Sciences, under Contract No. DE-AC0298CH10886. We also thank the National Science Foundation (CAREER Award DMR-0348239) and the Alfred P. Sloan Foundation for PI support and for the synthesis, diffraction, TGA, and electron microscopy studies. Moreover, we are grateful to D. Wang (Boston College) as well as to S. van Horn (SUNY Stony Brook) for additional assistance with electron microscopy. References and Notes (1) Tison, Y.; Giusca, C. E.; Stolojan, V.; Hayashi, Y.; Silva, S. R. P. AdV. Mater. 2008, 20, 189. (2) Hashimoto, A.; Suenaga, K.; Urita, K.; Shimada, T.; Sugai, T.; Bandow, S.; Shinohara, H.; Iijima, S. Phys. ReV. Lett. 2005, 94, 045504/1. (3) Saito, R.; Matsuo, R.; Kimura, T.; Dresselhaus, G.; Dresselhaus, M. S. Chem. Phys. Lett. 2001, 348, 187.

Peng et al. (4) Pfeiffer, R.; Kramberger, C.; Simon, F.; Kuzmany, H.; Popov, V. N.; Kataura, H. Eur. Phys. J. B 2004, 42, 345. (5) Pfeiffer, R.; Kuzmany, H.; Kramberger, C.; Schaman, C.; Pichler, T.; Kataura, H.; Achiba, Y.; Kurti, J.; Zolyomi, V. Phys. ReV. Lett. 2003, 90, 225501/1. (6) Endo, M.; Muramatsu, H.; Hayashi, T.; Kim, Y. A.; Terrones, M.; Dresselhaus, N. S. Nature 2005, 433, 476. (7) Ha, B.; Shin, D. H.; Park, J.; Lee, C. J. J. Phys. Chem. C 2008, 112, 430. (8) Son, Y. W.; Oh, S.; Ihm, J.; Han, S. Nanotechnology 2005, 16, 125. (9) Ho, Y. H.; Chang, C. P.; Shyu, F. L.; Chen, R. B.; Chen, S. C.; Lin, M. F. Carbon 2004, 42, 3159. (10) Maeng, I.; Kang, C.; Oh, S. J.; Son, J. H.; An, K. H.; Lee, Y. H. Appl. Phys. Lett. 2007, 90, 051914/1. (11) Villalpando-Paez, F.; Son, H.; Nezich, D.; Hsieh, Y. P.; Kong, J.; Kim, Y. A.; Shimamoto, D.; Muramatsu, H.; Hayashi, T.; Endo, M.; Terrones, M.; Dresselhaus, M. S. Nano Lett. 2008, 8, 3879. (12) Ho, G. W.; Ho, Y. H.; Li, T. S.; Chang, C. P.; Lin, M. F. Carbon 2006, 44, 2323. (13) Khatri, I.; Adhikari, S.; Aryal, H. R.; Soga, T.; Jimbo, T.; Umeno, M. Appl. Phys. Lett. 2009, 94, 093509/1. (14) Somani, S. P.; Somani, P. R.; Flahaut, E.; Kalita, G.; Umeno, M. Jpn. J. Appl. Phys. 2008, 47, 1219. (15) Jia, Y.; Wei, J. Q.; Wang, K. L.; Cao, A. Y.; Shu, Q. K.; Gui, X. C.; Zhu, Y. Q.; Zhuang, D. M.; Zhang, G.; Ma, B. B.; Wang, L. D.; Liu, W. J.; Wang, Z. C.; Luo, J. B.; Wu, D. AdV. Mater. 2008, 20, 4594. (16) Jung, Y. C.; Shimamoto, D.; Muramatsu, H.; Kim, Y. A.; Hayashi, T.; Terrones, M.; Endo, M. AdV. Mater. 2008, 20, 4509. (17) Koizhaiganova, R. B.; Kim, H. J.; Vasudevan, T.; Lee, M. S. Synth. Met. 2009, 159, 2437. (18) Muramatsu, H.; Kim, Y. A.; Hayashi, T.; Endo, M.; Yonemoto, A.; Arikai, H.; Okino, F.; Touhara, H. Chem. Commun. 2005, 2002. (19) Hayashi, T.; Shimamoto, D.; Kim, Y. A.; Muramatsu, H.; Okino, F.; Touhara, H.; Shimada, T.; Miyauchi, Y.; Maruyama, S.; Terrones, M.; Dresselhaus, M. S.; Endo, M. ACS Nano 2008, 2, 485. (20) Banerjee, S.; Wong, S. S. Nano Lett. 2002, 2, 195. (21) Azamian, B. R.; Coleman, K. S.; Davis, J. J.; Hanson, N.; Green, M. L. H. Chem. Commun. 2002, 366. (22) Ellis, A. V.; Vjayamohanan, K.; Goswaimi, R.; Chakrapani, N.; Ramanathan, L. S.; Ajayan, P. M.; Ramanath, G. Nano Lett. 2003, 3, 279. (23) Choi, H. C.; Shim, M.; Bangsaruntip, S.; Dai, H. J. J. Am. Chem. Soc. 2002, 124, 9058. (24) Cho, N.; Choudhury, K. R.; Thapa, R. B.; Sahoo, Y.; Ohulchanskyy, T.; Cartwright, A. N.; Lee, K. S.; Prasad, P. N. AdV. Mater. 2007, 19, 232. (25) Coleman, K. S.; Bailey, S. R.; Fogden, S.; Green, M. L. H. J. Am. Chem. Soc. 2003, 125, 8722. (26) Bottini, M.; Cerignoli, F.; Dawson, M. I.; Magrini, A.; Rosato, N.; Mustelin, T. Biomacromolecules 2006, 7, 2259. (27) Moghaddam, M. J.; Taylor, S.; Gao, M.; Huang, S. M.; Dai, L. M.; McCall, M. J. Nano Lett. 2004, 4, 89. (28) Peng, X.; Chen, J.; Misewich, J. A.; Wong, S. S. Chem. Soc. ReV. 2009, 38, 1076. (29) Lee, H.; Yoon, S. W.; Kim, E. J.; Park, J. Nano Lett. 2007, 7, 778. (30) Banerjee, S.; Wong, S. S. J. Am. Chem. Soc. 2003, 125, 10342. (31) Wang, D.; Li, Z. C.; Chen, L. W. J. Am. Chem. Soc. 2006, 128, 15078. (32) Quinn, B. M.; Dekker, C.; Lemay, S. G. J. Am. Chem. Soc. 2005, 127, 6146. (33) Day, T. M.; Unwin, P. R.; Wilson, N. R.; Macpherson, J. V. J. Am. Chem. Soc. 2005, 127, 10639. (34) Kim, Y. A.; Muramatsu, H.; Park, K. C.; Shimamoto, D.; Jung, Y. C.; Kim, J. H.; Hayashi, T.; Saito, Y.; Endo, M.; Terrones, M.; Dresselhaus, M. S. Appl. Phys. Lett. 2008, 93, 051901/1. (35) Banerjee, S.; Wong, S. S. Chem. Commun. 2004, 1866. (36) Peng, X.; Wong, S. S. AdV. Mater. 2009, 21, 625. (37) Nikoobakht, B.; Burda, C.; Braun, M.; Hun, M.; El-Sayed, M. A. Photochem. Photobiol. 2002, 75, 591. (38) Haremza, J. M.; Hahn, M. A.; Krauss, T. D.; Chen, S.; Calcines, J. Nano Lett. 2002, 2, 1253. (39) Grzelczak, M.; Correa-Duarte, M. A.; Salgueirino-Maceira, V.; Giersig, M.; Diaz, R.; Liz-Marzan, L. M. AdV. Mater. 2006, 18, 415. (40) Hohng, S.; Ha, T. J. Am. Chem. Soc. 2004, 126, 1324. (41) Hammer, N. I.; Early, K. T.; Sill, K.; Odoi, M. Y.; Emrick, T.; Barnes, M. D. J. Phys. Chem. B 2006, 110, 14167. (42) Dannhauser, T.; O’Neil, M.; Johansson, K.; Whitten, D.; McLendon, G. J. Phys. Chem. 1986, 90, 6074. (43) Wuister, S. F.; Donega, C. D.; Meijerink, A. J. Phys. Chem. B 2004, 108, 17393. (44) Landes, C.; Burda, C.; Braun, M.; El-Sayed, M. A. J. Phys. Chem. B 2001, 105, 2981. (45) Cowdery-Corvan, J. R.; Whitten, D. G.; McLendon, G. L. Chem. Phys. 1993, 176, 377.

Synthesis and Characterization of DWNT-NC Heterostructures (46) Muramatsu, H.; Hayashi, T.; Kim, Y. A.; Shimamoto, D.; Kim, Y. J.; Tantrakarn, K.; Endo, M.; Terrones, M.; Dresselhaus, M. S. Chem. Phys. Lett. 2005, 414, 444. (47) Boatman, E. M.; Lisensky, G. C.; Nordell, K. J. J. Chem. Educ. 2005, 82, 1697. (48) Peng, Z. A.; Peng, X. G. J. Am. Chem. Soc. 2001, 123, 183. (49) Peng, Z. A.; Peng, X. G. J. Am. Chem. Soc. 2002, 124, 3343. (50) Nakajima, N.; Ikada, Y. Bioconj. Chem. 1995, 6, 123. (51) Murakami, T.; Matsumoto, K.; Kisoda, K.; Naito, R.; Nishio, K.; Isshiki, T.; Harima, H. J. Appl. Phys. 2008, 103, 114305/1. (52) Sugai, T.; Yoshida, H.; Shimada, T.; Okazaki, T.; Shinohara, H. Nano Lett. 2003, 3, 769. (53) Lin-Vien, D.; Colthup, N. B.; Fateley, W. G.; Grasselli, J. G. The Handbook of Infrared and Raman Characteristic Frequencies of Organic Molecules; Academic Press, Inc.: San Diego, 1991. (54) Miyazawa, T.; Shimanouchi, T.; Mizushima, S. I. J. Chem. Phys. 1958, 29, 611. (55) Kapitonov, A. M.; Stupak, A. P.; Gaponenko, S. V.; Petrov, E. P.; Rogach, A. L.; Eychmuller, A. J. Phys. Chem. B 1999, 103, 10109. (56) Wang, Y.; Suna, A.; McHugh, J.; Hilinski, E. F.; Lucas, P. A.; Johnson, R. D. J. Chem. Phys. 1990, 92, 6927. (57) Zhang, J. Z. J. Phys. Chem. B 2000, 104, 7239. (58) Banerjee, S.; Wong, S. S. AdV. Mater. 2004, 16, 34. (59) Robel, I.; Bunker, B. A.; Kamat, P. V. AdV. Mater. 2005, 17, 2458. (60) Biju, V.; Itoh, T.; Baba, Y.; Ishikawa, M. J. Phys. Chem. B 2006, 110, 26068. (61) Pan, B. F.; Cui, D. X.; Ozkan, C. S.; Ozkan, M.; Xu, P.; Huang, T.; Liu, F. T.; Chen, H.; Li, Q.; He, R.; Gao, F. J. Phys. Chem. C 2008, 112, 939.

J. Phys. Chem. C, Vol. 114, No. 19, 2010 8773 (62) Guldi, D. M.; Rahman, G. M. A.; Sgobba, V.; Kotov, N. A.; Bonifazi, D.; Prato, M. J. Am. Chem. Soc. 2006, 128, 2315. (63) Li, W.; Gao, C.; Qian, H.; Ren, J.; Yan, D. J. Mater. Chem. 2006, 16, 1852. (64) Pan, B.; Cui, D.; He, R.; Gao, F.; Zhang, Y. Chem. Phys. Lett. 2006, 417, 419. (65) Nikoobakht, B.; Burda, C.; Barun, M.; El Sayed, M. A. Photochem. Photobiol. 2002, 75, 591. (66) Babentsov, V.; Riegler, J.; Schneider, J.; Fiederle, M.; Nann, T. J. Phys. Chem. B 2005, 109, 15349. (67) Dibbell, R. S.; Watson, D. F. J. Phys. Chem. C 2009, 113, 3139. (68) Si, H. Y.; Liu, C. H.; Xu, H.; Wang, T. M.; Zhang, H. L. Nanoscale Research Lett. 2009, 4, 1146. (69) Mountrichas, G.; Sandanayaka, A. S. D.; Economopoulos, S. P.; Pispas, S.; Ito, O.; Hasobe, T.; Tagmatarchis, N. J. Mater. Chem. 2009, 19, 8990. (70) Vietmeyer, F.; Seger, B.; Kamat, P. V. AdV. Mater. 2007, 19, 2935. (71) Leatherdale, C. A.; Kagan, C. R.; Morgan, N. Y.; Empedocles, S. A.; Kastner, M. A.; Bawendi, M. G. Phys. ReV. B 2000, 62, 2669. (72) Noy, D.; Moser, C. C.; Dutton, P. L. Biochim. Biophys. Acta 2006, 1757, 90. (73) Guyot-Sionnest, P.; Shim, M.; Matranga, C.; Hines, M. Phys. ReV. B 1999, 60, R2181. (74) Kippeny, T. D.; Bowers, M. J., II; Dukes, A. D., III; McBride, J. R.; Orndorff, R. L.; Garrett, M. D.; Rosenthal, S. J. J. Chem. Phys. 2008, 128, 084713/1. (75) Li, F. S.; Cho, S. H.; Son, D. I.; Kim, T. W.; Lee, S. K.; Cho, Y. H.; Jin, S. H. Appl. Phys. Lett. 2009, 94, 111906/1.

JP100580H