Photoluminescence Imaging of Polyfluorene Surface Structures on

Dec 6, 2016 - Tunable room-temperature single-photon emission at telecom wavelengths from sp3 defects in carbon nanotubes. Xiaowei He , Nicolai F...
0 downloads 0 Views 2MB Size
Download PDF
Photoluminescence Imaging of Polyfluorene Surface Structures on Semiconducting Carbon Nanotubes: Implications for Thin Film Exciton Transport Nicolai F. Hartmann,† Rajib Pramanik,† Anne-Marie Dowgiallo,‡ Rachelle Ihly,‡ Jeffrey L. Blackburn,*,‡ and Stephen K. Doorn*,† †

Center for Integrated Nanotechnologies, MPA-CINT, Los Alamos National Laboratory, Los Alamos, New Mexico 87545, United States ‡ National Renewable Energy Laboratory, Golden, Colorado 80401, United States S Supporting Information *

ABSTRACT: Single-walled carbon nanotubes (SWCNTs) have potential to act as light-harvesting elements in thin film photovoltaic devices, but performance is in part limited by the efficiency of exciton diffusion processes within the films. Factors contributing to exciton transport can include film morphology encompassing nanotube orientation, connectivity, and interaction geometry. Such factors are often defined by nanotube surface structures that are not yet well understood. Here, we present the results of a combined pump−probe and photoluminescence imaging study of polyfluorene (PFO)wrapped (6,5) and (7,5) SWCNTs that provide additional insight into the role played by polymer structures in defining exciton transport. Pump−probe measurements suggest exciton transport occurs over larger length scales in films composed of PFO-wrapped (7,5) SWCNTs, compared to those prepared from PFO-bpy-wrapped (6,5) SWCNTs. To explore the role the difference in polymer structure may play as a possible origin of differing transport behaviors, we performed a photoluminescence imaging study of individual polymer-wrapped (6,5) and (7,5) SWCNTs. The PFO-bpy-wrapped (6,5) SWCNTs showed more uniform intensity distributions along their lengths, in contrast to the PFO-wrapped (7,5) SWCNTs, which showed irregular, discontinuous intensity distributions. These differences likely originate from differences in surface coverage and suggest the PFO wrapping on (7,5) nanotubes produces a more open surface structure than is available with the PFO-bpy wrapping of (6,5) nanotubes. The open structure likely leads to improved intertube coupling that enhances exciton transport within the (7,5) films, consistent with the results of our pump−probe measurements. KEYWORDS: carbon nanotubes, energy harvesting, photovoltaics, exciton transport, surface structure

B

conduction bands) can also be readily tuned for optimal overlap with other device materials and for directed charge transport.11,12 Charge13,14 and exciton15,16 transport can be ultrafast in these materials, and exciton dissociation at acceptor interfaces17−19 has been demonstrated to be highly efficient. Despite these characteristics, demonstrated SWCNT-based PV device performance has been limited to power conversion efficiencies in the 1−2% range for single-chiral devices.3,5,20 A significant limitation inherent to such devices is that the narrow

roadening the range of efficient light-harvesting materials for solar photovoltaic (PV) applications remains an important challenge for meeting growing demands for renewable energy resources. Single-walled carbon nanotubes (SWCNTs) are drawing interest as light-harvesting materials, with potential to serve as the active light-harvesting elements in thin film PV devices or to enhance the response of primary elements in hybrid devices.1−8 Among the attractive characteristics that SWCNTs bring to this role is their strong absorptivity at visible and near-infrared (NIR) wavelengths.1,9 Because of the sensitivity of SWCNT electronic transitions to changes in nanotube structure,10 optical absorbance can be effectively tuned across the visible and NIR wavelength range. Likewise, redox properties (in particular, offsets for valence and © 2016 American Chemical Society

Received: October 24, 2016 Accepted: December 6, 2016 Published: December 6, 2016 11449

DOI: 10.1021/acsnano.6b07168 ACS Nano 2016, 10, 11449−11458

Article

www.acsnano.org

Article

ACS Nano

Figure 1. (a) Molecular structures of the polyfluorene polymers PFO-bpy (left) and PFO (right). (b) Absorbance of (6,5)/PFO-bpy sample, with PL spectrum (dotted trace) obtained for E22 excitation at 570 nm. (c) Absorbance of (7,5)/PFO sample, with PL spectrum (dotted trace) obtained for E22 excitation at 650 nm. (d) PLE map of (6,5)/PFO-bpy sample. (e) PLE map of (7,5)/PFO sample.

interactions will also be important.25,27 Such interactions are determined by direct SWCNT surface connectivity, which is in turn defined by the surface structures generated by the chemical agents (e.g., surfactants, polymers) used for preparing the SWCNT solution suspensions. In the case of surfactant-based suspensions, Kataura and coworkers have demonstrated that choice of surfactant can directly determine thin film morphology.28 For example, sodium dodecyl sulfate (SDS) suspensions result in relatively open networks of highly bundled SWCNTs. Sodium deoxycholate suspensions (DOC), however, yield dense random networks with high degrees of interconnectivity between individual nanotubes.28 Filtration conditions for generating such films can also be chosen such that highly aligned, densely packed morphologies result.29 Surfactant choice can also affect access to the SWCNT surface, which in turn helps to define communication between SWCNTs and external entities (e.g., other SWCNTs, redox species, etc.). SDS has been shown to create tunable surface structures that in general provide somewhat randomly oriented coverage and more open access to the nanotube surface.30,31 In contrast, DOC produces highly ordered close-packed surface structures that limit direct intertube interactions.31,32 While surfactant-based suspensions have shown some success as source material for PV devices,6 polymer-based suspensions using polyfluorene (PFO) wrappings such as PFO and PFObpy (poly[(9,9-dioctylfluorenyl-2,7-diyl) and poly[(9,9-dioctylfluorenyl-2,7-diyl)-alt-co-(6,6′-{2,2′-bipyridine})], respectively; see Figure 1) have the advantage of yielding high chiral purity semiconducting SWCNTs33,34 and, for single chiral devices,

SWCNT exciton absorbances limit the effective wavelength range of the device. Multichiral architectures can boost this efficiency by harvesting a greater wavelength range but have still been limited to performances below 4%.7 Significant opportunity also remains for optimization of cell design, including parameters such as chiral purity, SWCNT length, orientation and network morphology, effect and extent of bundling, active layer thickness, and thickness of other device materials.20−22 Many of these parameters ultimately have a direct influence on energy and charge transport through the devices, which are likely the primary limitations to current performance. While ideally cells would harness the exceptional axial transport characteristics of the SWCNT structure,21−23 current designs are not well-suited for accessing such transport. Instead, the random planar network architecture of current cells are limited to transverse intertube exciton and charge hopping as the primary transport mechanisms.23−25 Arnold and co-workers, for example, have shown that exciton transport occurs via such an intertube hopping, with observed average hopping distances of 10 nm,24 ultimately limiting film thickness and light-harvesting ability.20−24 As noted above, energy and charge transport are ultimately limited in these systems by the details of the SWCNT network interconnectedness, as defined by film morphology and nanotube interaction geometry.23−27 Film morphology determines the relative nanotube orientation and whether nanotubes are incorporated as ordered or random structures. Such parameters also determine the degree of close packing experienced within the film, all impacting performance via restrictions on transport. Ultimately, the strength of intertube 11450

DOI: 10.1021/acsnano.6b07168 ACS Nano 2016, 10, 11449−11458

Article

ACS Nano

Figure 2. Transient absorption spectra and kinetics recorded at the position of the trion-induced absorption (X1) for SWCNT thin films: (a) Transient absorption spectra obtained at 300 ps delay time for SWCNT/C60 bilayer films containing PFO-bpy-wrapped (6,5) (top, purple) and PFO-wrapped (7,5) (bottom, blue) nanotubes. Bleach of the respective E11 absorption is labeled as E11, while the trion-induced absorption is labeled as X1. (b) (6,5)/PFO-bpy, pumped at 1000 nm (E11) and probed at 1180 nm and (c) (7,5)/PFO SWCNT, pumped at 1050 nm (E11) and probed at 1230 nm. In each plot, the black trace is for a neat film with no C60, where the signal originates from the shoulder of a small peak (blue-shifted from the trion peak) that has been assigned to a biexciton-induced absorption19,39 (since there are no charges generated in the neat films, there are no trions generated by the probe pulse). The red traces in each plot are for the bilayer films where charges are generated by exciton dissociation at the SWCNT/C60 interface. The green dot-dashed line in each plot is a fit to the rise time of the trion-induced absorption signal intensity. In panel (b), the kinetics are fit well with a single-exponential rise time that is instrument-response-limted (∼120 fs). In panel (c), in addition to the instrument-response-limited rise time, there is a second rise time with time constant τ = 2.6 ± 1.9 ps. We note the biexciton absorbance is absent in the SWCNT/C60 bilayer films19 and thus does not contribute to the kinetic traces of the bilayers (red).

RESULTS AND DISCUSSION Toluene suspensions of either PFO-bpy-wrapped (6,5) or PFO-wrapped (7,5) nanotubes served as the primary samples for all studies in this work. Figure 1 shows the molecular structures of the two polyfluorene polymers used as the structurally selective SWCNT suspension agents. The resultant absorbance and room-temperature ensemble fluorescence spectra for each sample type are also shown in Figure 1. The primary spectral features seen in the figure correspond to the first and second excitonic transitions (E11 and E22), occurring at 570 and 1000 nm for (6,5) and 650 and 1050 nm for (7,5) nanotubes, respectively. Femtosecond Pump−Probe Studies. Pump−probe measurements were performed on SWCNT/C60 bilayer thin films incorporating electron donor films of either PFO-bpywrapped (6,5) nanotubes or PFO-wrapped (7,5) nanotubes. Excitation of both samples was tuned to the respective E11 resonance. As found previously,19 excitation of the (6,5) films leads to a bleach of the E11 absorbance at 1000 nm and rapid growth of an induced absorbance at 1174 nm (see Figure 2a, top). The induced absorbance is attributed to formation of a positively charged trion of the (6,5) E11 exciton by the probe pulse,19 where the extra hole of the trion results from ultrafast exciton dissociation and excited-state interfacial electron transfer from the nanotube to C60. While other groups have attributed this induced absorbance to the excited-state absorbance of triplet excitons,35 we only see this peak under conditions where an appreciable number of charges are generated on the SWCNTs19,36−38 and further note that in the absence of C60 acting as an electron acceptor (i.e., in neat SWCNT films) this peak is absent. Our data thus are not consistent with this peak arising from triplets. As detailed extensively in our previous studies, we intentionally utilize photon fluences below 3 × 1012 photons/cm2/pulse to ensure that the only trion peak that we observe arises from interfacial photoinduced charge transfer (electron transfer from SWCNTs to C60 in this case). As originally demonstrated by Yuma et al.,39 at sufficiently high fluences, exciton Auger ionization can give rise to a trion-induced absorbance. This Auger-induced trion

have shown the best performance and device reproducibility to date.2−5,20 Although polyfluorene surface structures on SWCNTs are likely a major determining factor in how film morphology and intertube interactions translate to exciton and charge transport, little is known about these polymer surface structures. As an example, recent femtosecond pump−probe studies of exciton dissociation dynamics within PFO-bpywrapped (6,5) SWCNT/C60 heterojunction bilayers showed efficient exciton dissociation at SWCNT/C60 interfaces.19 The lack of any signatures of exciton transport from deeper within the SWCNT film, however, suggests that film morphology and polymer−SWCNT surface structures may inhibit efficient exciton transport to the type-II exciton dissociation interface. A better understanding of the relevant polymer−SWCNT surface structures and how they translate to PV device performance is required to further advance capability for optimizing SWCNT-based light harvesting. In this work, we evaluate polyfluorene−SWCNT surface structures and their impact on exciton transport within thin film C60 heterojunction devices. Significantly different photophysical behaviors are found for PFO-bpy-wrapped (6,5) versus PFOwrapped (7,5) chiralities. In contrast to the PFO-bpy/(6,5) system, pump−probe results on PFO-wrapped (7,5) bilayers with C60 electron-accepting layers show two sources of charge generation at C60 interfaces: one from the dissociation of excitons generated immediately at the interface and a second source originating from excitons generated deeper within the (7,5) film. Photoluminescence imaging of the two systems suggests that PFO produces a more open surface structure on (7,5) chiralities than is found for PFO-bpy surface structures on (6,5) nanotubes. Combined, the results indicate very different surface structures for the two polymer/SWCNT systems. More open structures (as for (7,5)) likely promote intertube interactions that directly yield improved exciton transport for (7,5)-based devices, in comparison to those formed from (6,5). The results provide insight into how polymer surface structures may be tuned toward further optimization of SWCNT-based PV devices. 11451

DOI: 10.1021/acsnano.6b07168 ACS Nano 2016, 10, 11449−11458

Article

ACS Nano

Figure 3. Representative PL images of a (6,5) (a) and a (7,5) SWCNT (d). The scale bar represents 1 μm in both images. (b,e) Positiondependent normalized intensity distributions along the length of the tubes in (a) and (d), as defined by the blue dashed arrows. Histograms of the intensity values found along the length of the tubes are depicted in (c,f) and depict visually the significant difference in intensity distribution for (6,5) PFO-bpy-wrapped (c) and (7,5) PFO-wrapped (f) SWCNTs.

2c corresponds to a biexponential rise of the induced absorption intensity, convoluted with the instrument response (τIRF = 120 fs). For this fit, the faster of the two exponential rise times is fixed at 120 fs, as determined in our previous study. The slower of the two exponential rise times and the amplitudes of both components are allowed to float to find the best fit to the data. The resulting fit suggests that the slower rise time, τslow = 2.5 ± 1.9 ps, contributes approximately 18% of the total signal intensity for the trion-induced absorbance. In this case, we assign the instrument-limited rise to exciton dissociation occurring promptly for excitons generated in the immediate vicinity of the SWCNT/C60 interface, while the longer-time component likely reflects a diffusion-limited process in which excitons are first transported to the C60 interface on a picosecond time scale, followed by dissociation via charge transfer. Although the noise associated with the transient places a significant degree of uncertainty on the value of this longer-time component, this value is consistent with the time scales expected for diffusion-limited intertube exciton transport in PFO-wrapped (7,5) films.25 The pump−probe results indicate that significantly different exciton-to-free carrier conversion processes occur for the two types of films studied. While the (7,5) film appears to be capable of supporting exciton transport from locations throughout the film thickness, such processes appear to be significantly suppressed for the (6,5) film. While the films differ in composition by SWCNT chirality, the differing behaviors also likely depend on the very different molecular structures of the wrapping polymers and their interactions at the SWCNT surface. Both differences in the polymer molecular structure and differences in how the polymer interacts with specific chiralities may lead to different surface structures. The ability for specific polymers to selectively suspend specific SWCNT chiralities certainly provides precedence for such an expectation. The result could translate to differences in intertube interactions that would yield very different exciton transport behaviors occurring in the respective assemblies. Specifically, the pump−probe results lead us to speculate that PFO−(7,5) SWCNT interactions may result in more open surface structures that promote improved intertube interactions in comparison to those for PFO-bpy−(6,5) SWCNT and thus

peak, however, can only be observed when exciting neat (6,5) SWCNT samples if the photon fluence exceeds ∼6 × 1012 photons/cm2/pulse.39 Our own studies on neat (6,5) and (7,5) SWCNT thin films confirm this finding19 and indicate no observable trion peak within neat SWCNT films at the low fluences used here. Our previous studies also demonstrate that a weak induced absorbance to the blue of the trion peak (at 1130 nm) can be seen in the photoexcited neat film.19 This peak was assigned previously to the creation of a biexciton,39 which rapidly decays in ∼5 ps (see Figure 2b). This peak is absent in the photoexcited bilayer19 and does not influence our analysis detailed below. The long decay time of the trion (340 ps to 16 ns, see ref 19) indicates a persistent lifetime for the charge-separated state. Of primary interest here, however, is the rise time of the trion feature, which will reflect the rate of interfacial exciton dissociation processes plus any exciton diffusion that might occur prior to this final step. The early time transient response of the trion absorbance is shown in Figure 2b. Notably, the rise time of this feature is instrument-limited and indicates that exciton dissociation and charge transfer occur within the 120 fs excitation pulse width. No additional contributions to the rise time are resolved, suggesting that additional diffusive transport processes to the interface may be minimal for the (6,5) PFObpy system. Similar processes may be probed in PFO-based (7,5) thin films. Of note is the difference in molecular structures of the wrapping polymers used for isolating (6,5) versus (7,5) SWCNT material (see Figure 1), which may lead to differing photophysical behaviors for films derived from the two chiralities. For the (7,5) system, excitation is performed at 1050 nm. Again, a bleach of the E11 absorbance is observed upon excitation, accompanied by an induced trion absorbance occurring now at 1240 nm (Figure 2a, bottom). While the E11 bleach recovery and the decay of the trion absorbance occur on similar time scales for (7,5) as was found for the (6,5) film, we now observe significantly different kinetic behavior for the rise of the (7,5) trion feature. As seen in Figure 2c, the induced trion absorbance in the (7,5) case also shows an instrumentlimited rise time, but superimposed upon this ultrafast rise is a second, longer-time component. The dashed fit line in Figure 11452

DOI: 10.1021/acsnano.6b07168 ACS Nano 2016, 10, 11449−11458

Article

ACS Nano

occurrences of intensities near the normalized maximum than is found for the (7,5) example (Figure 3f). Furthermore, the intensity is found to spread over a broad range of variability for the (7,5) nanotube, while the intensities for the (6,5) example show a much narrower range. These observations are quantified as the average Irel and σ of each histogram (0.89 and 0.08 for (6,5) and 0.63 and 0.18 for the (7,5) examples, respectively). We obtained average Irel and σ values for 50 individual, randomly chosen, (6,5) and (7,5) nanotubes and plotted the results as histograms in Figure 4. The average Irel and σ values

enhanced exciton transport/diffusion. To test this hypothesis, we performed photoluminescence (PL) imaging measurements on individual PFO-bpy and PFO-wrapped (6,5) and (7,5) nanotubes. Direct Photoluminescence Imaging of Polyfluorene− SWCNT Surface Structures. Because the SWCNT exciton PL emission is highly sensitive to the surface environment, PL spectra and imaging can be useful probes of surface structure. PL intensity and spectral changes have been shown to provide a useful measure of surfactant reorganization at the nanotube surface in aqueous suspensions.30−32 Additionally, whole-tube PL imaging techniques can directly resolve variations in surface chemistry.40,41 Here, we use PL imaging at the single-nanotube level to probe polymer surface structures on (6,5) and (7,5) SWCNTs. Representative examples of PL images of individual (6,5) and (7,5) nanotubes are shown in Figure 3a,d, respectively (additional PL images are shown in the Supporting Information). The images of the two structures are quite different. PL intensity along the length of the (6,5) nanotube shows a generally uniform intensity distribution. In contrast, the (7,5) image shows more spatial intensity variability and even apparent gaps in emission in certain regions of the SWCNT. This contrasting behavior is found to be a general result across many tubes (see Supporting Information). Variability in the surface environment can directly modulate PL intensity along the tube length.30−32 We note that in our experiments we obtain PL images from SWCNTs that remain in a toluene solution environment (see Methods), rather than obtaining images from nanotubes that have been drop-dried on a substrate. This approach allows us to minimize possible substrate interactions that could also modulate PL intensity along a tube and possibly also perturb the solution-phase polymer structures we wish to probe. The differences in intensity distributions observed for the (6,5) versus (7,5) structures may thus originate from different regions of a given tube interacting variably with the polymer or solvent environment. The uniform intensity distribution along the (6,5) tube indicates a stable environment along its length, while the changing intensity along the (7,5) tube indicates a more variable environment. The observed differences in intensity distribution therefore suggest that the (6,5) SWCNTs experience a more continuous polymer coverage along their surface, while coverage along the (7,5) nanotubes may be discontinuous. This visual impression, however, is qualitative and somewhat subjective. We therefore also perform a more quantitative comparison of the two behaviors. For quantitative comparisons, the intensity distribution along the length (blue dashed lines in Figure 3a,d) of all sampled nanotubes is extracted using ImageJ software. The extracted intensities are normalized with respect to the brightest intensity observed. The resulting intensity distributions for the two examples shown in Figure 3a,d are plotted in Figure 3b,e. Shown in Figure 3c,f are histograms of the normalized intensities obtained at each pixel along the tube lengths. We note that in order to not count the effects of end quenching on PL intensities,15 the response from the first 500 nm in from either tube end was not included in the histograms (500 nm is around a factor of 2 larger than typical intratube SWCNT exciton diffusion lengths15). The plotted intensity variations thus represent primarily the impact of the local surface environment on emission intensities. For the (6,5) example (Figure 3c), the histogram is found to have many more

Figure 4. Global PL intensity behavior. Histograms showing the average σ (a) and mean relative intensity (Irel) (b) of the normalized intensity distributions for 50 individual (6,5) (top row) nanotubes and for 50 different (7,5) (bottom row) nanotubes (c,d).

over all (6,5) tubes (Figure 4a,b) are found to be 0.81 and 0.12, respectively. The corresponding values for the (7,5) data set (Figure 4c,d) are 0.70 and 0.18. A comparison of the average σ values corroborates our initial visual impression of the images shown in Figure 3a,d: there is a larger variability present in the observed emission intensities for the (7,5) nanotubes, in comparison to that found for the (6,5) nanotubes. Additionally, the (7,5) intensities show deeper minima in the observed intensities than found for the (6,5) nanotubes (as reflected in the smaller Irel values found for the (7,5) examples). It is difficult to conclusively capture these trends on viewing the single images shown in Figure 3. However, an alternative way of visualizing and analyzing these intensity distribution results can make the differences between the (6,5) and (7,5) behaviors more apparent. In this analysis, shown in Figure 5, the intensity distributions (analogous to Figure 3b,e) for several tubes are linked together as one long “synthetic” tube. It is clear in this representation of the data that the intensity losses along the nanotube lengths are consistently much deeper and broader for the (7,5) nanotubes than are observed for the (6,5) structures. The body of data in Figures 3−5 thus indicate that PFO-wrapped (7,5) nanotubes show significantly more intensity variability along their lengths than do the PFO-bpywrapped (6,5) nanotubes. The histograms of Figure 4 and the global intensity distributions represented by the synthetic “long” nanotubes of Figure 5 confirm another qualitative visual impression of our imaging data. While we find a number of instances of (6,5) nanotubes showing discontinuous intensity distributions along their lengths (similar to that generally found for all (7,5) nanotubes), we most frequently observe a continuous and bright emission along the (6,5) nanotubes. In contrast, all (7,5) nanotubes displayed some degree of intensity discontinuity and dimming along their lengths, while we never observe a case of a 11453

DOI: 10.1021/acsnano.6b07168 ACS Nano 2016, 10, 11449−11458

Article

ACS Nano

along the nanotube surface than is provided by the PFO-bpy wrapping of the (6,5), thus creating regions that are free to more directly interact with the surrounding solvent environment. Given the more continuously bright intensity distribution observed for the (6,5), we assign the bright areas to regions more completely coated by the polymer, while dimmer areas are those that are more open to the surrounding environment. The observation of brighter emission from the polymercoated regions, in some respects, parallels the behavior found for aqueous-based surfactant-suspended SWCNTs. In the latter, surfactant surface structures are able to insulate excitons from environmental interactions that quench emission.31,32 Specifically, surfactants serve to minimize SWCNT interaction with water, whose highly polar nature, as well as its ability to participate in redox doping of SWCNTs via the H2O/O2 redox couple, allows it to act as a quenching center.11,32,42 Such quenching mechanisms, however, are unlikely for interaction with the nonpolar solvent toluene, used for suspending the PFO-wrapped SWCNTs. As an alternative mechanism for the dimming of the SWCNT surface in contact with toluene, we hypothesize that excitons formed in these regions lie at a slightly higher potential energy than those in the polymercoated regions. Such an energy difference will induce a directed diffusion of excitons out of the solvent-contacted area, thus dimming these regions. This is reasonable in that the PL imaging shows the dimmer regions to be typically at most a few diffusion lengths (ld ∼ 200 nm)15 in size. This proposed mechanism is further supported by spectroscopic imaging results (Figure 6) that are directly correlated to the PL intensity distributions found on both (6,5) and (7,5) examples. In these experiments, PL intensity maps of individual nanotubes and PL spectroscopic images of the exact same nanotubes are generated in a confocal spectral imaging configuration (see Methods). When the intensity images are compared to the spectroscopic images, we find that for both the (6,5) (Figure 6a,b) and (7,5) (Figure 6c,d) SWCNTs, regions of high intensity correlate to a red-shifting of the PL spectrum, while regions of low intensity correlate to blue-shifting of the spectrum. (For additional examples, see Supporting Information Figures S3 and S4.) This finding directly supports the idea that the dimmer regions will be at higher energy than the brighter regions. We suggest that π−π stacking of the polymer with the underlying SWCNT lattice may serve to stabilize excitons in these regions. The local dielectric environment associated with the polymer might also provide additional

Figure 5. Intensity fluctuations along the nanotube axis for intensities of several tubes stitched together for (6,5) (a) and (7,5) (b) examples, with the total length of each “synthetic” stitched-together tube being equal (60 μm). The right-hand side panels show the distribution of counts over relative intensities and depict visually the difference in fluctuation along a connected “synthetic” tube of equal length and same number of data points for (6,5) and (7,5) SWCNTs.

(7,5) nanotube that shows a continuous and bright intensity distribution along its length. This observation of mixed behavior of the (6,5) intensity distributions is reflected in the appearance of two populations in both the σ and mean Irel plots of Figure 4a,b and also by continuous stretches of high-intensity data in the synthetic nanotube plot of Figure 5a, with occasional interruption by spikes of discontinuity. As noted above, the intensity variability along the nanotube length is likely a direct reflection of variability in surface structure and coverage provided by the different polymer wrappings. The results of Figures 3−5 suggest that the PFO wrapping of the (7,5) nanotube provides a more open structure

Figure 6. Confocal PL intensity image (a) and spectroscopic image (b) of PFO-bpy-wrapped (6,5) SWCNT. Confocal PL intensity image (c) and spectroscopic image (d) of PFO-wrapped (7,5) SWCNT. Spectral image indicates peak emission energy for the E11 emission line of the respective SWCNT (centered at ∼1.235 and 1.185 eV for the (6,5) and (7,5) examples, respectively). Note that regions of bright intensity correlate to regions of red-shifted PL in both the (6,5) and (7,5) cases. The scale bar represents 0.5 μm (a,b) and 1.0 μm (c,d). Additional examples of single-tube E11 emission spectra and intensity distributions correlated to spectroscopic distributions are given in the Supporting Information. 11454

DOI: 10.1021/acsnano.6b07168 ACS Nano 2016, 10, 11449−11458

Article

ACS Nano

transport of excitons to the C60 interface in such thin films. In addition to probing variations in surface coverage, PL imaging allows an evaluation of SWCNT length in the (6,5) and (7,5) source material. As seen in Figure 7, we find the PFO−(7,5)

screening with respect to the regions that are open to toluene. These effects likely combine to ultimately modulate surface potentials along the polymer-wrapped SWCNTs. We note that we have also considered that polymer-induced strain may create similar effects. Strain has been shown to lower E11 transition energies for mod 1 SWCNTs like (6,5) structures.43 However, strain will have the opposite effect on the mod 2 (7,5) structure, so we rule out strain as a possible origin for the intensity modulation. The conclusions we can draw from the PL imaging regarding polymer surface structures provide a physical basis for understanding the differences between the (6,5) and (7,5) pump−probe data of Figure 2. The relatively more open surface structure found on the (7,5) tubes can promote more intimate intertube interactions in the disordered network films through direct contact between tubes at the gaps in surface coverage. Such direct interactions between tubes may be inhibited in (6,5) films due to the more continuous polymer coverage found for these. The resultant differences can have a direct impact on exciton transport processes in the respective thin films. Enhanced intertube connectivity enabled by gaps in the polymer surface structure for the PFO−(7,5) structures can in turn provide a more efficient route for exciton hopping between tubes, thus ultimately enabling effective exciton transport from deeper within the films to the SWCNT/C60 interface. Such a process is suggested by the multicomponent rise time found in the (7,5) pump−probe measurements (Figure 2c). In contrast, reduced intertube interactions resulting from the more continuous surface coverage found in the PFO-bpy−(6,5) structures prevents effective exciton transport to the C60 interface from within the SWCNT layer. Excitons contributing to charge generation are therefore effectively limited to originating at just the interface layer. The difference in the (6,5) versus (7,5) exciton transport behaviors in PV devices thus ultimately originates in the gaps occurring in the surface structure of the PFO wrappings that are absent with PFO-bpy wrapping. This interpretation of the combined pump−probe and PL imaging results is further supported by the findings of Grechko et al.25 In their work, femtosecond transient absorption spectroscopy was used to study exciton migration within networked SWCNT films of varying morphology. In agreement with our findings, modeling of the migration dynamics indicated that exciton transport between tubes and tube bundles was dominated by hopping at points of direct contact between tubes. Through-space energy transfer or exciton hopping was found to be inefficient, with even a single-polymer layer being sufficient to inhibit transport between tubes.25 Grechko et al. also find that an important aspect of their exciton transport mechanism is that intertube hopping is paired with intratube transport to the tube/tube contact or hopping points via exciton diffusion along the tube.25 Such intratube transport is expected to add a length dependence to the efficiency of this process. Hertel et al. have shown that the SWCNT ends will act to quench excitons, with the likelihood of end quenching increasing as the tubes become shorter.44 Longer tubes, in contrast, are expected to lead to a higher probability of each exciton encountering a tube/tube junction before being quenched at a SWCNT end. It is therefore expected that longer tubes will be more effective than short tubes at carrying excitons through networked films.21,22 Our PL imaging results also provide evidence that intratube exciton diffusion can be an additional contributing factor to effective

Figure 7. Histograms of nanotube lengths evaluated from PL imaging for (a) PFO-bpy−(6,5) and (b) PFO−(7,5) SWCNT source material. The solid black curves represent log-normal distribution fits to the histogram, resulting in μ = 1.92 ± 0.02 μm and 2σ = 0.25 ± 0.02 for the (6,5) and μ = 2.52 ± 0.17 μm and 2σ = 0.49 ± 0.06 for the (7,5) sample.

SWCNTs to be on average longer than the PFO-bpy−(6,5) SWCNTs (2.5 vs 1.9 μm average length, respectively). The (7,5) tubes also have a much broader length distribution, with significant numbers of tubes observed with lengths in the 3−7 μm range. These findings from the PL imaging are corroborated by AFM measurement of tube lengths in the two source materials, as shown in the Supporting Information. The longer tube lengths present in the (7,5) sample may be an additional contributing factor (by enabling more effective harnessing of intratube exciton transport) in the observed enhanced exciton transport efficiency evident from our pump− probe results. We speculate that the observed differences in tube lengths likely originate in the SWCNT processing steps. Significant ultrasonication of SWCNT source material is required to obtain high-quality dispersions of the polymer-wrapped tubes (see Methods). Sonication is known to effectively cut nanotubes to shorter segments; a process that is more effective for smaller diameter nanotubes (i.e., the (6,5): d = 0.757 nm) due to the higher level of curvature-induced strain present in comparison to the larger diameter (7,5) (d = 0.829 nm).

CONCLUSIONS In summary, our pump−probe results show a clear difference in exciton transport and dissociation kinetics on comparing PFObpy (6,5) films to those composed of PFO-wrapped (7,5) SWCNTs. Excitons generated within the (7,5) film are found to contribute to photogenerated charge at the SWCNT/C60 interface, while those generated within deeper layers of the (6,5) film cannot. Exciton transport from within the (6,5) film is therefore inhibited in comparison to the behavior of the (7,5) film. We speculate that this contrast in behaviors is due to changes in SWCNT connectivity within the film that ultimately originate in the polymer structures occurring at the nanotube surface. This conclusion is based on PL imaging results that suggest open surface structures generated on the PFO-wrapped 11455

DOI: 10.1021/acsnano.6b07168 ACS Nano 2016, 10, 11449−11458

Article

ACS Nano

laser pumps an optical parametric amplifier to generate the pump pulse, which was tuned to the E11 of each SWCNT sample (1000 nm for (6,5) and 1050 nm for (7,5)). The pump pulse was chopped at a rate of 500 Hz, and the pump fluence was kept in the range of ∼3 × 1012 photons pulse−1 cm−2 to avoid trion formation demonstrated to occur via Auger recombination processes as a consequence of excitation at higher fluences.19,24 The near-infrared probe pulse (800 nm < λprobe < 1700 nm) was generated by passing a portion of the amplified 800 nm light through a sapphire plate, and the pump−probe delay was tuned with a motorized translation stage. For the TA measurements, all thin film samples were sealed in a nitrogen glovebox using a separate glass slide and a polymer film (Surlyn, Solaronix) sandwiched between the two slides on the outer edges. This precaution was taken to avoid the possibility of photoinduced degradation requiring both photons and exposure to oxygen and/or water. Direct Photoluminescence Imaging. After suitable dilution of the initial suspension of polymer-wrapped SWCNTs with toluene, a 2 μL droplet of SWCNT suspension was deposited on a 25 × 25 mm microscope cover glass. The droplet was sandwiched with an 18 × 18 mm microscope cover glass on top, and the seam was sealed with white lithium grease. After a waiting time of at least 4−6 h, the SWCNTs’ diffusive motion is slowed sufficiently that they can be imaged on a home-built PL microscope setup based on a commercially available Olympus IX71 body. For (6,5) excitation, a 568 nm laser (Coherent Sapphire) was used, and for (7,5), a 660 nm laser (Coherent Cube) was used. After reflection at a beamsplitter (Semrock, Di02-R594-25x36 for 568 nm, and Semrock, FF705-Di0125x36 for 660 nm excitation), the laser was focused onto the back aperture of a NA = 1.49 oil immersion objective (Olympus UApoN 100×) for wide-field illumination. In the case of (6,5), the wide-field image was detected via an EMCCD (Princeton Instruments, ProEM 512 × 512 pixel) camera after optical filtering for the (6,5) PL (Thorlabs, FB990-10). The (7,5) PL was imaged through a 2D InGaAs detector (Princeton Instruments, OMA-V 320, 320 × 356 pixel) after optical filtering for the (7,5) PL (Thorlabs, FEL1000). Since the tube lens of the microscope body was removed, the widefield images were focused onto the two cameras with a f = 750 mm lens (Thorlabs, LA) for increased magnification. Image analysis and extraction of PL profiles were done in ImageJ.46 The representation of the “synthetic” tube (Figure 5) required interpolation of the (6,5) data points since the EMCCD camera is equipped with pixels smaller (16 μm) than those of the InGaAs array (30 μm). A meaningful comparison can only be obtained for two “synthetic” tubes of equal length and equal amount of data points per unit length. Confocal Spectroscopic Imaging. The same microscope is also equipped with a piezo-xy stage (Mad City Laboratories, Nano-T 200 × 200 μm) for confocal sample scanning. Excitation occurs here by sending a collimated laser beam into the objective in order to obtain a confocal excitation spot. Instead of a single-point detector, the signal was sent to a monochromator (Princeton Instruments, Acton 2500i) and the spectra were recorded with a linear InGaAs detector (Princeton Instruments, OMA V 1024 linear InGaAs). The scan movement and the recording of spectra are synchronized in order to capture a full spectrum at every image pixel. Individual spectra were fitted (Python with LMFIT)47 using three Voigt peaks (E11, RBM phonon48 and K-momentum phonon sideband;49 see also Figures S3a and S4a) and a global linear background resulting in a full parameter set (center position, amplitude, width) for each peak at each pixel.

(7,5) SWCNTs that are less prevalent for the PFO-bpywrapped (6,5) SWCNTs. Such open structures can promote nanotube interactions capable of facilitating the observed exciton transport. These findings are an excellent example of how molecular-scale changes in structure can translate to altered behaviors at the mesoscale across the dimensions of these films. As demonstrated here, direct whole-nanotube PL imaging will be an important probe of variations in long-range surface structures and their impacts on nanotube functionality. Our findings on potential length-dependent performance differences in the two film types also support modeling results that nanotube length must also be accounted for as an important design parameter, as well.21,22 Interestingly, such dramatic differences in film behaviors as observed here arise from a simple substitution of the bpy functionality into the PFO polymer structure. That this leads to relatively large-scale differences in surface structures (with alternating dark and bright areas extending 100s of nm on the (7,5) surface; see Figure 3) indicates that simple steric hindrance introduced by the bpy group is not likely the primary driver of these larger-scale structural differences. This presents an interesting challenge for further understanding of these surface behaviors via molecular dynamics simulations. Additionally, these results motivate further study of the additional roles wrapping polymers may play in promoting optimized SWCNT PV device performance. We note that, while specific polymers play an essential role in isolating the desired nanotube structures, the same polymer may not provide the ideal surface structures required for ultimately optimizing desired functionality. There is thus a need for also developing capability for substituting the most promising polymer wrappings postseparations. Emerging strategies for polymer exchange may be important toward this end.45

METHODS Polyfluorene-Wrapped SWCNTs. CoMoCat SG65i SWCNTs were purchased from Southwest Nanotechnologies, Inc., and PFO-bpy and PFO were purchased from American Dye Source. For PL imaging studies, highly enriched SWCNT dispersions were prepared by bath sonication. For (6,5) SWCNT dispersions, 20 mg of PFO-bpy was dissolved in 10 mL of toluene, after which 10 mg of CoMoCAT SG65i SWCNTs was placed into the PFO-bpy solution. The same procedure was used for (7,5), except that the polymer utilized was PFO. These mixtures were bath-sonicated for 60 min with no cooling, after which the samples were immediately centrifuged at 13 200 rpm (Beckman L100-XP ultracentrifuge, SW32Ti rotor) in thin polyallomer centrifuge tubes for 5 min at 20 °C. Following centrifugation, the supernatant was removed from the centrifuge tube, and this supernatant contained the enriched SWCNT dispersion that was used, without further processing, for PL imaging experiments. The same procedure was utilized for producing samples for transient absorbance measurements, except that the SWCNTs in this case were sonicated for 30 min with a tip-horn ultrasonic processor (Cole Parmer, 750 W) operating at 40% power with a half-inch tip. Thin films were prepared for TA measurements as previously described.19,36 Briefly, excess polymer was removed from the highly enriched SWCNT samples by successive overnight centrifugation runs at 24 100 rpm at 0 °C until a polymer/SWCNT mass ratio of approximately 1:1 was reached. The resulting inks were then deposited by ultrasonic spraying into ∼10 nm thick films on glass substrates (neat films). For bilayer films, ∼90 nm of C60 was deposited onto the neat films by thermal evaporation. Femtosecond Transient Absorbance Measurements. Femtosecond pump−probe TA experiments were performed, as described previously,19,36 on a 1 kHz regeneratively amplified Ti:sapphire laser system that produces 4 mJ laser pulses at 800 nm. The Ti:sapphire

ASSOCIATED CONTENT S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsnano.6b07168. Additional PL images of individual (6,5) and (7,5) SWCNTs showing the range of intensity distribution behavior along nanotube lengths; additional examples of single-tube spectroscopic imaging results; complemen11456

DOI: 10.1021/acsnano.6b07168 ACS Nano 2016, 10, 11449−11458

Article

ACS Nano

(11) O’Connell, M. J.; Eibergen, E. E.; Doorn, S. K. Chiral Selectivity in the Charge-Transfer Bleaching of Single-Walled Carbon-Nanotube Spectra. Nat. Mater. 2005, 4, 412−418. (12) Hirana, Y.; Juhasz, G.; Miyauchi, Y.; Mouri, S.; Matsuda, K.; Nakashima, N. Empirical Prediction of Electronic Potentials of SingleWalled Carbon Nanotubes with a Specific Chirality (n,m). Sci. Rep. 2013, 3, 2959. (13) Javey, A.; Kim, H.; Brink, M.; Wang, Q.; Ural, A.; Guo, J.; McIntyre, P.; McEuen, P.; Lundstrom, M.; Dai, H. High-Kappa Dielectrics for Advanced Carbon Nanotube Transistors and Logic Gates. Nat. Mater. 2002, 1, 241−246. (14) Durkop, T.; Getty, S. A.; Cobas, E.; Fuhrer, M. S. Extraordinary Mobility in Semiconducting Carbon Nanotubes. Nano Lett. 2004, 4, 35−39. (15) Crochet, J. J.; Duque, J. G.; Werner, J. H.; Lounis, B.; Cognet, L.; Doorn, S. K. Disorder Limited Exciton Transport in Colloidal Single-Wall Carbon Nanotubes. Nano Lett. 2012, 12, 5091−5096. (16) Harrah, D. M.; Schneck, J. R.; Green, A. A.; Hersam, M. C.; Ziegler, L. D.; Swan, A. K. Intensity-Dependent Exciton Dynamics of (6,5) Single-Walled Carbon Nanotubes: Momentum Selection Rules, Diffusion, and Nonlinear Interactions. ACS Nano 2011, 5, 9898−9906. (17) Crochet, J. J.; Hoseinkhani, S.; Luer, L.; Hertel, T.; Doorn, S. K.; Lanzani, G. Free-Carrier Generation in Aggregates of Single-Wall Carbon Nanotubes by Photoexcitation in the Ultraviolet Regime. Phys. Rev. Lett. 2011, 107, 257402. (18) Bindl, D. J.; Ferguson, A. J.; Wu, M.-Y.; Kopidakis, N.; Blackburn, J. L.; Arnold, M. S. Free Carrier generation and Recombination in Polymer-Wrapped Semiconducting Carbon Nanotube Films and Heterojunctions. J. Phys. Chem. Lett. 2013, 4, 3550− 3559. (19) Dowgiallo, A.-M.; Mistry, K. S.; Johnson, J. C.; Blackburn, J. L. Ultrafast Spectroscopic Signature of Charge Transfer between SingleWalled Carbon Nanotubes and C60. ACS Nano 2014, 8, 8573−8581. (20) Guillot, S. L.; Mistry, K. S.; Avery, A. D.; Richard, J.; Dowgiallo, A.-M.; Ndione, P. F.; van de Lagemaat, J.; Reese, M. O.; Blackburn, J. L. Precision Printing and Optical Modeling of Ultrathin SWCNT/C60 Heterojunction Solar Cells. Nanoscale 2015, 7, 6556−6566. (21) Bellisario, D. O.; Jain, R. M.; Ulissi, Z.; Strano, M. S. Deterministic Modelling of Carbon Nanotube Near-Infrared Solar Cells. Energy Environ. Sci. 2014, 7, 3769−3781. (22) Bellisario, D. O.; Paulson, J. A.; Braatz, R. D.; Strano, M. S. An Analytical Solution for Exciton Generation, Reaction, and Diffusion in Nanotube and Nanowire-Based Solar Cells. J. Phys. Chem. Lett. 2016, 7, 2683−2688. (23) Sharma, A.; Singh, V.; Bougher, T. L.; Cola, B. A. A Carbon Nanotube Optical Rectenna. Nat. Nanotechnol. 2015, 10, 1027−1032. (24) Mehlenbacher, R. D.; Wu, M. Y.; Grechko, M.; Laaser, J. E.; Arnold, M. S.; Zanni, M. T. Photoexcitation Dynamics of Coupled Semiconducting Carbon Nanotube Thin Films. Nano Lett. 2013, 13, 1495−1501. (25) Grechko, M.; Ye, Y.; Mehlenbacher, R. D.; McDonough, T. J.; Wu, M.-Y.; Jacobberger, R. M.; Arnold, M. S.; Zanni, M. T. DiffusionAssisted Photoexcitation Transfer in Coupled Semiconducting Carbon Nanotube Thin Films. ACS Nano 2014, 8, 5383−5394. (26) Mehlenbacher, R. D.; Wang, J.; Kearns, N. M.; Shea, M. J.; Flach, J. T.; McDonough, T. J.; Wu, M.-Y.; Arnold, M. S.; Zanni, M. T. Ultrafast Exciton Hopping Observed in Bare Semiconducting Carbon Nanotube Thin Films with Two-Dimensional White-Light Spectroscopy. J. Phys. Chem. Lett. 2016, 7, 2024−2031. (27) Davoody, A. H.; Karimi, F.; Arnold, M. S.; Knezevic, I. Theory of Exciton Energy Transfer in Carbon Nanotube Composites. J. Phys. Chem. C 2016, 120, 16354−16366. (28) Shimizu, M.; Fujii, S.; Tanaka, T.; Kataura, H. Effects of Surfactants on the Electronic Transport Properties of Thin-Film Transistors of Single-Wall Carbon Nanotubes. J. Phys. Chem. C 2013, 117, 11744−11749. (29) He, X.; Gao, W.; Xie, L.; Zhang, Q.; Lei, S.; Li, B.; Robinson, J. M.; Headrick, R. J.; Haroz, E. H.; Doorn, S. K.; Pasquali, M.; Vajtai, R.; Ajayan, P. M.; Adams, W. W.; Hauge, R. H.; Kono, J. Wafer-Scale

tary AFM-determined length distributions for the two sample types (PDF)

AUTHOR INFORMATION Corresponding Authors

*E-mail: jeff[email protected]. *E-mail: [email protected]. ORCID

Stephen K. Doorn: 0000-0002-9535-2062 Notes

The authors declare no competing financial interest.

ACKNOWLEDGMENTS We thank G. Rumbles and S. Tretiak for helpful discussions. This work was supported in part by the Los Alamos National Laboratory LDRD program. Portions of this work were performed at the Center for Integrated Nanotechnology, a U.S. Department of Energy, Office of Science User Facility. NREL authors were supported by the Solar Photochemistry Program of the U.S. Department of Energy, Office of Science, Basic Energy Sciences, Division of Chemical Sciences, Geosciences, and Biosciences, under Contract No. DE-AC308GO28308 to NREL. REFERENCES (1) Arnold, M. S.; Blackburn, J. L.; Crochet, J. J.; Doorn, S. K.; Duque, J. G.; Mohite, A.; Telg, H. Recent Developments in the Photophysics of Single-Walled Carbon Nanotubes for Their Use as Active and Passive Material Elements in Thin Film Photovoltaics. Phys. Chem. Chem. Phys. 2013, 15, 14896−14918. (2) Bindl, D. J.; Wu, M.-Y.; Prehn, F. C.; Arnold, M. S. Efficiently Harvesting Excitons from Electronic Type-Controlled Semiconducting Carbon Nanotubes. Nano Lett. 2011, 11, 455−460. (3) Shea, M. J.; Arnold, M. S. 1% Solar Cells Derived from Ultrathin Carbon Nanotube Photoabsorbing Films. Appl. Phys. Lett. 2013, 102, 243101. (4) Ye, Y.; Bindl, D. J.; Jacobberger, R. M.; Wu, M.-Y.; Roy, S. S.; Arnold, M. S. Semiconducting Carbon Nanotube Aerogel Bulk Heterojunction Solar Cells. Small 2014, 10, 3299−3306. (5) Ferguson, A. J.; Dowgiallo, A.-M.; Bindl, D. J.; Mistry, K. S.; Reid, O. G.; Kopidakis, N.; Arnold, M. S.; Blackburn, J. L. Trap-Limited Carrier Recombination in Single-Walled Carbon Nanotube Heterojunctions with Fullerene Acceptor Layers. Phys. Rev. B: Condens. Matter Mater. Phys. 2015, 91, 245311. (6) Jain, R. M.; Howden, R.; Tvrdy, K.; Shimizu, S.; Hilmer, A. J.; McNicholas, T. P.; Gleason, K. K.; Strano, M. S. Polymer-Free NearInfrared Photovoltaics with Single Chirality (6,5) Semiconducting Carbon Nanotube Active Layers. Adv. Mater. 2012, 24, 4436−4439. (7) Gong, M.; Shastry, T. A.; Xie, Y.; Bernardi, M.; Jasion, D.; Luck, K. A.; Marks, T. J.; Grossman, J. C.; Ren, S.; Hersam, M. C. Polychiral Semiconducting Carbon Nanotube-Fullerene Solar Cells. Nano Lett. 2014, 14, 5308−5314. (8) Cui, N. H.; Anisimov, A. S.; Chiba, T.; Fujii, S.; Kataura, H.; Nasibulin, A. G.; Chiashi, S.; Kauppinen, E. I.; Maruyama, S. Air-Stable High Efficiency Solar Cells with Dry-Transferred Single-Walled Carbon Nanotube Films. J. Mater. Chem. A 2014, 2, 11311−11318. (9) Schoppler, F.; Mann, C.; Hain, T. C.; Neubauer, F. M.; Privitera, G.; Bonaccorso, F.; Chu, D.; Ferrari, A. C.; Hertel, T. Molar Extinction Coefficient of Single Wall Carbon Nanotubes. J. Phys. Chem. C 2011, 115, 14682−14686. (10) Bachilo, S. M.; Strano, M. S.; Kittrell, C.; Hauge, R. H.; Smalley, R. E.; Weisman, R. B. Structure-Assigned Optical Spectra of Single Walled Carbon Nanotubes. Science 2002, 298, 2361−2366. 11457

DOI: 10.1021/acsnano.6b07168 ACS Nano 2016, 10, 11449−11458

Article

ACS Nano Monodomain Films of Spontaneously-Aligned Single-Wall Carbon Nanotubes. Nat. Nanotechnol. 2016, 11, 633−638. (30) Niyogi, S.; Densmore, C. G.; Doorn, S. K. Electrolyte Tuning of Surfactant Interfacial Behavior for Enhanced Density-Based Separations of Single-Walled Carbon Nanotubes. J. Am. Chem. Soc. 2009, 131, 1144−1153. (31) Duque, J. G.; Densmore, C. G.; Doorn, S. K. Saturation of Surfactant Structure at the Single-Walled Carbon Nanotube Surface. J. Am. Chem. Soc. 2010, 132, 16165−16175. (32) Duque, J. G.; Oudjedi, L.; Crochet, J. J.; Tretiak, S.; Lounis, B.; Doorn, S. K.; Cognet, L. Mechanism of Electrolyte Induced Brightening of Single-Wall Carbon Nanotubes. J. Am. Chem. Soc. 2013, 135, 3379−3382. (33) Nish, A.; Hwang, J.-Y; Doig, J.; Nicholas, R. J. Highly Selective Dispersion of Single-Walled Carbon Nanotubes Using Aromatic Polymers. Nat. Nanotechnol. 2007, 2, 640−646. (34) Ozawa, H.; Ide, N.; Fujigaya, T.; Niidome, Y.; Nakashima, N. One-Pot Separation of Highly Enriched (6,5) Single-Walled Carbon Nanotubes Using a Fluorene-Based Co-Polymer. Chem. Lett. 2011, 40, 239−241. (35) Park, J.; Deria, P.; Therien, M. J. Dynamics and Transient Absorption Spectral Signatures of the Single-Wall Carbon Nanotube Electronically Excited Triplet State. J. Am. Chem. Soc. 2011, 133, 17156−17159. (36) Dowgiallo, A.-M.; Mistry, K. S.; Johnson, J. C.; Reid, O. G.; Blackburn, J. L. Probing Exciton Diffusion and Dissociation in SingleWalled Carbon Nanotube-C60 Heterojunctions. J. Phys. Chem. Lett. 2016, 7, 1794−1799. (37) Ihly, R.; Dowgiallo, A.-M.; Yang, M.; Schulz, P.; Stanton, N. J.; Reid, O. G.; Ferguson, A. J.; Zhu, K.; Berry, J. J.; Blackburn, J. L. Efficient Charge Extraction and Slow Recombination in OrganicInorganic Perovskites Capped with Semiconducting Single-Walled Carbon Nanotubes. Energy Environ. Sci. 2016, 9, 1439. (38) Schulz, P.; Dowgiallo, A.-M.; Yang, M.; Zhu, K.; Blackburn, J. L.; Berry, J. J. Charge Transfer Dynamics between Carbon Nanotubes and Hybrid Organic Metal Halide Perovskite Films. J. Phys. Chem. Lett. 2016, 7, 418−425. (39) Yuma, B.; Berciaud, S.; Besbas, J.; Shaver, J.; Santos, S.; Ghosh, S.; Weisman, R. B.; Cognet, L.; Gallart, M.; Ziegler, M.; Honerlage, B.; Lounis, B.; Gilliot, P. Biexciton, Single Carrier, and Trion Generation Dynamics in Single-Walled Carbon Nanotubes. Phys. Rev. B: Condens. Matter Mater. Phys. 2013, 87, 205412. (40) Crochet, J. J.; Duque, J. G.; Werner, J. H.; Doorn, S. K. Photoluminescence Imaging of Localized and Delocalized Electronic Defects in Single-Wall Carbon Nanotubes. Nat. Nanotechnol. 2012, 7, 126−132. (41) Hartmann, N. F.; Yalcin, S. E.; Adamska, L.; Haroz, E. H.; Ma, X.; Tretiak, S.; Htoon, H.; Doorn, S. K. Photoluminescence Imaging of Solitary Dopant Sites in Covalently Doped Single-Wall Carbon Nanotubes. Nanoscale 2015, 7, 20521−20530. (42) Aguirre, C. M.; Levesque, P. L.; Paillet, M.; Lapointe, F.; StAntoine, B. C.; Desjardins, P.; Martel, R. The Role of the Oxygen/ Water Redox Couple in Suppressing Electron Conduction in FieldEffect Transistors. Adv. Mater. 2009, 21, 3087−3091. (43) Leeuw, T. K.; Tsyboulski, D. A.; Nikolaev, P. N.; Bachilo, S. M.; Arepalli, S.; Weisman, R. B. Strain Measurements on Individual SingleWalled Carbon Nanotubes: Structure-Dependent Spectral Shifts and Load Transfer. Nano Lett. 2008, 8, 826−831. (44) Hertel, T.; Himmelein, S.; Ackermann, T.; Stich, D.; Crochet, J. Diffusion Limited Photoluminescence Quantum Yields in 1-D Semiconductors: Single-Walled Carbon Nanotubes. ACS Nano 2010, 4, 7161−7168. (45) Joo, Y.; Brady, G. J.; Shea, M. J.; Oviedo, M. B.; Kanimozhi, C.; Schmitt, S. K.; Wong, B. M.; Arnold, M. S.; Gopalan, P. Isolation of Pristine Electronics Grade Semiconducting Carbon Nanotubes by Switching the Rigidity of the Wrapping Polymer Backbone on Demand. ACS Nano 2015, 9, 10203−10213. (46) Schneider, C. A.; Rasband, W. S.; Eliceiri, K. W. NIH to ImageJ: 25 Years of Image Analysis. Nat. Methods 2012, 9, 671−675.

(47) Newville, M.; Stensitzki, T.; Allen, D. B.; Zenodo, A. I.; LMFIT. Non-Linear Least-Square Minimization and Curve-Fitting for Python. Zenodo 2014, DOI: 10.5281/zenodo.11813. (48) Cambré, S.; Santos, S. M.; Wenseleers, W.; Nugraha, A. R. T.; Saito, R.; Cognet, L.; Lounis, B. Luminescence Properties of Individual Empty and Water-Filled Single-Walled Carbon Nanotubes. ACS Nano 2012, 6, 2649−2655. (49) Torrens, O. N.; Zheng, M.; Kikkawa, J. M. Energy of KMomentum Dark Excitons in Carbon Nanotubes by Optical Spectroscopy. Phys. Rev. Lett. 2008, 101, 157401.

11458

DOI: 10.1021/acsnano.6b07168 ACS Nano 2016, 10, 11449−11458