Single-walled carbon nanotubes (SWCNTs) are attractive materials for next-generation energy-harvesting technologies, including thermoelectric generators, due to their tunable opto-electronic properties and high charge carrier mobilities. Controlling the Fermi level within these unique 1D nanomaterials is often afforded by charge transfer interactions between SWCNTs and electron or hole accepting species. Conventional methods to dope SWCNT networks typically involve the diffusion of molecular redox dopant species into solid-state thin films, but solution-phase doping could potentially provide routes and/or benefits for charge carrier transport, scalability, and stability. Here, we develop a methodology for solution-phase doping of polymer-wrapped, highly enriched semiconducting SWCNTs using a p-type charge transfer dopant, F4TCNQ. This allows doped SWCNT inks to be cast into thin films without the need for additional post-deposition doping treatments. We demonstrate that the introduction of the dopant at varying stages of the SWCNT dispersion process impacts the ultimate thermoelectric performance and observe that the dopant alters the polymer selectivity for semiconducting vs metallic SWCNTs. In contrast to dense semiconducting polymer films, where solution-phase doping typically leads to disrupted morphologies and poorer TE performance than solid-state doping, thin films of solution-doped s-SWCNTs perform similarly to their solid-state doped counterparts. Interestingly, our results also suggest that solution-phase F4TCNQ doping leads to fully ionized and dimerized F4TCNQ anions in solid-state films that are not observed in films doped with F4TCNQ after deposition. Our results provide a framework for the application of solution-phase doping to a broad array of high-performance SWCNT-based thermoelectric materials and devices that may require high-throughput deposition techniques.

Thermoelectric (TE) energy harvesting is an attractive technological solution for converting waste heat into electricity.1 The dimensionless TE figure of merit, zT, depends on electrical conductivity (σ), thermopower (α), and thermal conductivity (κ),

zT=σα2κT,
(1)

while TE materials and devices have incorporated inorganic semiconductors for many decades, substantial research over the past decade has explored the TE energy harvesting potential of organic semiconductors2—including small molecules, oligomers, and polymers—especially for low-grade waste heat recovery. Single-walled carbon nanotubes (SWCNTs) represent an interesting materials platform for energy harvesting technologies,3,4 with many properties that are akin to both crystalline inorganic materials and disordered or semi-crystalline organic materials. SWCNT-based TE materials and devices have shown significant improvements and promise over the past decade4–7 as researchers have focused on the roles of SWCNT electronic structure,8–10 precision carrier density/Fermi level tuning,11–14 morphology-dependent transport,15–17 and behavior within composites.18,19 As a result, several SWCNT-based TE materials platforms have realized high TE power factors (PF=σα2) and respectable zT values that continue to climb.

Highly enriched semiconducting SWCNTs (s-SWCNTs) have large intrinsic thermopower8,10,20 and large electrical conductivities that can be tuned via charge carrier doping. One effective SWCNT doping strategy involves electron or hole injection by molecular redox dopants.8,11,15,21,22 Conventionally, such redox doping has been performed once in solid-state SWCNT films. While this approach has several advantages, it requires additional, sometimes iterative, processing steps to tune the doping level for individual films and may not be scalable for ultimate device fabrication and integration. For solution-processed semiconductors, films can be deposited from “inks” containing the semiconductor of choice and strategically chosen additives that enhance the resulting films and devices. The semiconducting polymer community has widely explored the incorporation of “solution-phase” molecular redox dopants for tuning the polymer doping level prior to film deposition.23,24 From a technological standpoint, solution-phase doping has potential scaling advantages, since large volumes of ink can be processed at high throughput to yield huge batches of films and devices with nominally identical properties, without the need for extensive post-processing.25 From a scientific standpoint, solution-phase and solid-state doping can lead to very different film morphologies, inter-molecular interactions, and doping efficiencies,26–29 so it is important to compare these doping strategies to understand the true potential of redox doping for optimizing TE performance.

While solution-phase doping has been explored for SWCNTs containing both metallic and semiconducting SWCNTs, these approaches have frequently utilized very strong oxidants or reductants (e.g., chlorosulfonic acid30 or alkali metals31) There is a need for less aggressive approaches amenable to high-throughput solution-processing, especially for highly enriched s-SWCNT inks and films. Although a number of dopants have been used to produce p- and n-type SWCNTs in solid-state films, an additional challenge for solution-phase doping is finding a dopant that is compatible with the particular solvent and deposition method of choice. Since highly enriched s-SWCNTs can be produced by selective polymer extraction in organic solvents, such as toluene, one attractive option is 2,3,5,6-tetrafluoro-7,7,8,8-tetracyanoquinodimethane (F4TCNQ), a dopant that has been studied extensively for solid-state SWCNT doping32–34 and recently applied to solution-phase doping of s-SWCNTs for transistor applications.35 

Here, we demonstrate that F4TCNQ can tune nanotube carrier concentration in highly enriched s-SWCNT inks and that films resulting from these solution-doped inks have large TE PFs (138 μW m−1 K−2) that compare well to solid-state doped films of similar composition. Spectroscopic analysis allows us to identify significant differences between inks and films resulting from the incorporation of F4TCNQ at different stages of the dispersion process. Unique to films prepared from solution-doped s-SWCNTs, F4TCNQ molecules appear to form fully ionized (TCNQ)2 radical anion dimers. Thus, our data indicate that solution-phase doping may provide a strategy for manipulating molecular interactions in solid-state s-SWCNT networks. Our results demonstrate a scalable and generalizable strategy for controlling carrier density and TE performance in SWCNT-based TE materials and devices.

We employ polyfluorene polymer-wrapped SWCNTs as a model system for solution-phase doping. SWCNTs are prepared in-house by laser vaporization (LV), and we utilize poly[(9,9-dihexylfluorenyl-2,7-diyl)-co(9,10-anthracene)] (PFH-A) as the SWCNT dispersant. We previously demonstrated that this polymer disperses only s-SWCNTs, with high selectivity and throughput, for this LV SWCNT raw material.36 We have also adapted this solution-phase doping process to s-SWCNTs dispersed with several other fluorene-based polymers, and Salazar-Rios et al.35 have used a similar process on thiophene-based polymers. We study all films in air with no encapsulation.

We first probe the degree to which the incorporation of F4TCNQ into the SWCNT dispersion process impacts the resulting SWCNT inks. Figure 1 shows a schematic of the process flow for creating both undoped SWCNT inks and inks that are doped at various stages of the dispersion process. We study four different variations of SWCNT inks within this study: (1) “undoped,” (2) “pre-dispersion doped,” (3) “post-dispersion/before-PR,” and (4) “post-dispersion/after-PR” inks. “Undoped inks” are prepared by our standard process where SWCNT raw soot is tip-sonicated with PFH-A and centrifuged to provide a s-SWCNT enriched ink that contains excess polymer.36 Films prepared from undoped inks are still lightly p-type due to adventitious doping in air. “Pre-dispersion” inks are prepared by incorporating varying amounts of F4TCNQ into the initial dispersion mixture with the raw SWCNT soot and PFH-A, all of which then undergoes tip sonication and centrifugation to produce the ink. Before any inks are sprayed into films, they undergo a second polymer removal (PR) ultracentrifugation step, after which they can be sprayed into electronically coupled thin films.8 Post-dispersion inks are prepared by adding varying amounts of F4TCNQ to undoped, enriched s-SWCNT inks, either before or after the PR step to produce post-dispersion/before-PR or post-dispersion/after-PR inks, respectively.

FIG. 1.

Schematic illustrating the process flow used to generate the four inks, and the subsequent solid-state films.

FIG. 1.

Schematic illustrating the process flow used to generate the four inks, and the subsequent solid-state films.

Close modal

Absorbance spectroscopy can be used to assess the electronic purity of inks and thin films and follow the impact of doping on their optical properties.37Figure 2(a) shows visible/near-infrared (NIR) absorbance spectra for representative undoped, pre-dispersion doped, and post-dispersion/after PR doped inks. The undoped ink shows three distinct multi-peak envelopes that correspond to the first (S11, ca. 1400–2000), second (S22, ca. 800–1150), and third (S33, ca. 450–550 nm) excitonic transitions of s-SWCNTs. For both doped inks, F4TCNQ (250 μg/ml) causes nearly complete bleaching of the S11 excitonic transitions and partial bleaching of the S22 transitions, clearly demonstrating successful solution-phase doping. An obvious difference in pre- and post-doped dispersions is the presence of another envelope of optical transitions in the 600–750 nm range for the pre-dispersion doped ink that corresponds to the excitonic transitions of metallic SWCNTs. The presence of these peaks suggests that incorporation of F4TCNQ during the tip-sonication dispersion process causes the PFH-A polymer to lose its selectivity for s-SWCNTs, likely due to the similar electronic polarizability imparted by the hole density injected by F4TCNQ into both s- and m-SWCNTs.38,39

FIG. 2.

(a) Absorbance spectra of undoped, pre-dispersion doped, and post-dispersion/after PR doped inks. (b) Absorbance spectra of films sprayed from pre-dispersion ink, post-dispersion/after PR doped ink, and an undoped ink. The film prepared from the undoped ink was doped post-deposition by F4TCNQ. (c) Photographs of (left) an undoped ink, along with a conductive film prepared from this ink that was doped post-deposition by F4TCNQ and (right) a post-dispersion/after PR ink, along with the film prepared from this ink. (d) The impact of the F4TCNQ concentration on the bleach of the S11 peak envelope for post-dispersion doped inks.

FIG. 2.

(a) Absorbance spectra of undoped, pre-dispersion doped, and post-dispersion/after PR doped inks. (b) Absorbance spectra of films sprayed from pre-dispersion ink, post-dispersion/after PR doped ink, and an undoped ink. The film prepared from the undoped ink was doped post-deposition by F4TCNQ. (c) Photographs of (left) an undoped ink, along with a conductive film prepared from this ink that was doped post-deposition by F4TCNQ and (right) a post-dispersion/after PR ink, along with the film prepared from this ink. (d) The impact of the F4TCNQ concentration on the bleach of the S11 peak envelope for post-dispersion doped inks.

Close modal

Figure 2(b) displays absorbance spectra of F4TCNQ-doped films prepared from the inks shown in Fig. 2(a). Since the SWCNTs in both doped inks are already heavily p-type, no post-deposition F4TCNQ doping is needed. The film prepared from the undoped ink was treated in the solid-state, after deposition, in a 250 μg/ml solution of F4TCNQ. Interestingly, an entirely new peak is observed at 630 nm in the film [asterisk, blue trace Fig. 2(b)] deposited from the post-dispersion ink that is not observed in the ink itself [blue trace Fig. 2(a)]. This strong visible transition in the red region of the spectrum is not observed for either of the other films that are either doped pre-dispersion or doped post-deposition in the solid-state. This peak also gives the post-dispersion solution-phase doped film a distinct green color [Fig. 2(c)], despite the ink itself being vibrant red due to the strong absorption of excess neutral F4TCNQ molecules in the green region of the spectrum. While not as obvious, due to the similar position of the s-SWCNT S22 transitions, the thin film deposited from the post-dispersion doped ink also contains a strong broad peak at 920 nm [double asterisk, blue trace Fig. 2(b)] that is not present in any of the inks or the other two films in Fig. 2(b). These two peaks at 630 and 920 nm provide an insight into the unique molecular configurations adopted by F4TCNQ anions in these films, and we discuss this in more detail below.

Figure 2(d) compares the extent of doping-induced S11 bleaching for two post-dispersion F4TCNQ-doped inks. While both inks undergo a PR, the first ink is doped before this PR (post-dispersion/before PR) and the second is doped after the PR (post-dispersion/after PR). Figure 2(d) demonstrates that it is most beneficial to dope the s-SWCNTs after the PR step, since the extent of S11 bleaching is stronger across all F4TCNQ doping concentrations when doped post-PR. This difference arises partly from reduced competition of F4TCNQ interaction with SWCNTs as compared to mixtures containing both excess polymer and dopant and also partly because the PR step removes some amount of physically adsorbed F4TCNQ molecules from s-SWCNT surfaces. Based on the collective insights from Fig. 2, we determined that the best methodology for precisely and reproducibly tuning the doping level of the ink and resulting thin film was to dope the SWCNT ink post-dispersion (to avoid incorporation of m-SWCNTs) and after PR (to avoid irreproducible changes to the adsorbed F4TCNQ and hole density).

Figure 3 summarizes the TE properties achieved by solution-phase F4TCNQ doping of LV SWCNTs. Based on the extent of S11 quenching induced by different concentrations of F4TCNQ doping and our previous TE studies,8,11,15,21 we determined that an F4TCNQ concentration of ca. 150 μg/ml would allow us to achieve a peak TE power factor. As such, all samples in Fig. 3(a) are doped with 150 μg/ml of F4TCNQ, but at different stages of the dispersion process. The film deposited from the pre-dispersion doped ink performs the worst by far, with low thermopower (31 μV K−1), moderate conductivity (27 355 S m−1), and the lowest PF (26 μW m−1 K−2). The low thermopower results from the incorporation of m-SWCNTs into this sample, since our previous studies and those of others have demonstrated that m-SWCNTs have inherently low thermopower, relative to s-SWCNTs. The film deposited from the post-dispersion/before PR doped ink has a relatively high thermopower (95 μV K−1), but at the expense of a very low conductivity (3845 S m−1). The 35 μW m−1 K−2 PF for this sample is just slightly above the film prepared from the pre-dispersion ink. These mediocre metrics indicate a relatively low hole concentration, consistent with a sizable reduction in the quantity of adsorbed F4TCNQ molecules after the PR step (reduced S11 quenching observed in the absorbance spectrum after the PR, data not shown). Finally, the film deposited from the post-dispersion/after PR doped ink performs the best, achieving high conductivity (57 517 S m−1), moderate thermopower (49 μV K−1), and high PF (138 μW m−1 K−2).

FIG. 3.

(a) Thermopower, electrical conductivity, and power factor for SWCNT thin films prepared from solution-phase F4TCNQ-doped inks ([F4TCNQ] = 150 μg/ml) where the F4TCNQ is added at different steps of the dispersion process. All samples are measured after a PR step was performed to remove equivalent amounts of PFH-A wrapping polymer for all samples. (b) Thermopower and power factor for the film deposited from the post-dispersion/after PR doped ink for a variety of F4TCNQ concentrations. (c) Comparison of the conductivity-dependent power factors for previously published polymer:LV thin films doped post-deposition (solid-state)8,15 to a film deposited from a solution-phase doped ink (post-dispersion/after PR).

FIG. 3.

(a) Thermopower, electrical conductivity, and power factor for SWCNT thin films prepared from solution-phase F4TCNQ-doped inks ([F4TCNQ] = 150 μg/ml) where the F4TCNQ is added at different steps of the dispersion process. All samples are measured after a PR step was performed to remove equivalent amounts of PFH-A wrapping polymer for all samples. (b) Thermopower and power factor for the film deposited from the post-dispersion/after PR doped ink for a variety of F4TCNQ concentrations. (c) Comparison of the conductivity-dependent power factors for previously published polymer:LV thin films doped post-deposition (solid-state)8,15 to a film deposited from a solution-phase doped ink (post-dispersion/after PR).

Close modal

Figure 3(b) shows how TE metrics can be tuned in films deposited from these optimized solution-phase F4TCNQ-doped inks. Tuning F4TCNQ concentration in the s-SWCNT dispersion after PR affords concomitant tuning of the conductivity, thermopower, and PF over several orders of magnitude. Figure 3(c) demonstrates that this solution-phase doping strategy allows us to achieve optimized PFs in line with the maximum PFs we have achieved for s-SWCNTs utilizing LV SWCNTs and non-cleavable polymers, such as PFH-A or PFO-BPy.8,15 While this LV SWCNT/PFH-A polymer combination does not produce the highest PF that we have produced to date, it is a well-tested model system that provides informed insights for future extension of solution-phase doping to other SWCNT/dispersant combinations. Interestingly, the similar performance of solution-doped s-SWCNT thin films to their solid-state doped counterparts contrasts strongly to many reports from the semiconducting polymer community. In those cases, solution-phase doping of polymer inks has been shown to lead to poor film quality26–28 and low dopant ionization efficiency,29 relative to films doped in the solid-state by either dopant solutions or vapors.

We now return to the observation of unique peaks in absorbance spectra of thin films prepared from solution-doped inks, specifically for the case of the post-deposition/after PR sample [Fig. 2(b)—asterisks, blue trace]. To investigate this phenomenon in more detail, we show the absorbance spectra of a series of inks prepared in this manner with varying F4TCNQ concentrations [Fig. 4(a)] and corresponding films prepared from these inks [Fig. 4(b)]. Ink spectra show the typical behavior expected for p-type doping, where the S11 and S22 peaks are progressively bleached as the concentration of F4TCNQ in the ink is increased. A new strong absorption occurs below 600 nm, due the presence of neutral F4TCNQ molecules and F4TCNQ anions.40,41

FIG. 4.

(a) Absorbance spectra of a series of inks, doped in solution (post-dispersion/after PR) with varying amounts of F4TCNQ. (b) Thin films prepared from those same inks. (c) Difference spectra of the thin films shown in panel (b), where the spectrum of the film prepared from the undoped ink (0 μg/ml) is subtracted from the spectrum for a film prepared from a given doped ink. (d) Comparison of the 150 μg/ml difference spectrum to literature spectra of monomer TCNQ anions and40 dimer (TCNQ)2 anions40,41 (both in solution) and monomer F4TCNQ anions42 (within a solid-state polythiophene film). Literature spectra are digitized from plots in the respective citations.

FIG. 4.

(a) Absorbance spectra of a series of inks, doped in solution (post-dispersion/after PR) with varying amounts of F4TCNQ. (b) Thin films prepared from those same inks. (c) Difference spectra of the thin films shown in panel (b), where the spectrum of the film prepared from the undoped ink (0 μg/ml) is subtracted from the spectrum for a film prepared from a given doped ink. (d) Comparison of the 150 μg/ml difference spectrum to literature spectra of monomer TCNQ anions and40 dimer (TCNQ)2 anions40,41 (both in solution) and monomer F4TCNQ anions42 (within a solid-state polythiophene film). Literature spectra are digitized from plots in the respective citations.

Close modal

Film spectra elicit quite different features, relative to the corresponding ink spectra. A short post-deposition toluene soak removes unbound neutral F4TCNQ molecules, allowing for the clean discernment of new spectral features that are specific to bound F4TCNQ anions. While S11 transitions are progressively bleached with the increasing F4TCNQ concentration, there exists a clear and pronounced increase in absorbance at all wavelengths below ca. 1400 nm that is correlated with the F4TCNQ concentration. The spectral shape of this absorbance can be more clearly discerned by taking difference spectra [Fig. 4(c)]. At low F4TCNQ concentrations, distinct features appear between 360 and 500 nm, along with a broad feature near 800 nm. The 360 nm absorption and the broad ca. 800 nm feature are in line with optical transitions widely observed for dimeric and monomeric TCNQ anions—(TCNQ)2 and TCNQ, respectively—that have been observed in a large number of molecular charge transfer complexes.40,41 At the highest F4TCNQ concentration, the spectrum is dominated by several strong absorbance features at 360, 630, and 920 nm, along with a broad background and a small cluster of peaks centered at 500 nm.

Figure 4(d) plots the 150 μg/ml difference spectrum along with the solution-phase absorbance spectra of monomeric TCNQ and F4TCNQ anions and dimeric (TCNQ)2 anions.40–43 Since we have not found a spectrum of F4TCNQ anion dimers in the literature, we primarily refer here to the extensive literature on TCNQ anions, especially very clean solution-phase spectra, dating back to the 1960s.40,41 The heavily doped film's difference spectrum shows distinct features that can easily be assigned to the optical transitions of monomer and dimer anions. By comparison with the (TCNQ)2 dimer spectrum, we attribute features at 360 and 630 nm in the difference spectrum to intra-molecular optical transitions in (F4TCNQ)2 dimers and the peak at 920 nm to the inter-molecular charge transfer (CT) transition for (F4TCNQ)2 dimers. The shoulder at 420 nm and the broad absorbance background between the 630 and 920 nm in the difference spectrum can be assigned to the intra-molecular optical transitions in F4TCNQ monomer.

Spectral identification in Fig. 4(d) has a number of important and interesting implications. First, clear observation of F4TCNQ and (F4TCNQ)2 anion features in the F4TCNQ-doped film (prepared by solution-phase doping) implies that F4TCNQ molecules are fully ionized and have undergone integer charge transfer with the s-SWCNTs. Second, the predominance of dimer-related absorbance features in the heavily doped film, but not in the solution-phase doped ink from which this film originates, suggests that dimeric inter-molecular F4TCNQ interactions occur specifically in solid-state. These interactions are absent in films that are doped “post-deposition.” We hypothesize that formation of these strongly interacting (F4TCNQ)2 dimers occurs within s-SWCNT bundles during film formation, where F4TCNQ molecules on individual s-SWCNTs dimerize with other surface-bound F4TCNQ molecules on adjacent nanotubes as the nanotubes aggregate into bundles. These bundle-intercalated (F4TCNQ)2 dimers would not be present in the dispersion of isolated s-SWCNTs and would also not be able to form in pre-formed s-SWCNT networks containing tightly aggregated bundles.

In summary, we demonstrate that solution-phase redox doping of SWCNT inks enables control over charge-carrier density in the resulting films. The incorporation of F4TCNQ prior to dispersion reduces the selectivity of the conjugated polymer for s-SWCNTs, with unwanted m-SWCNTs limiting TE performance. The addition of F4TCNQ to the s-SWCNT ink after removal of excess polymer enables improved charge-carrier injection into nanotubes and solid-state TE performance that matches films prepared from an undoped ink and subsequently doped by immersion into a dopant solution. Our results illustrate that solution-phase doping could provide routes and/or benefits for charge carrier transport, process scalability, and device stability. This study also points to modified dopant-nanotube and dopant-dopant interactions within films that can be attributed to the way the F4TCNQ molecules are incorporated into the SWCNT bundles during network formation from the doped ink. Our results provide a framework to employ solution-phase doping for the fabrication of high-performance SWCNT-based thermoelectric materials and devices.

See the supplementary material for experimental methods.

This work was authored by the National Renewable Energy Laboratory (NREL), operated by Alliance for Sustainable Energy, LLC, for the U.S. Department of Energy (DOE) under Contract No. DE-AC36-08GO28308. The investigation of the thermoelectric properties of the SWCNT networks was supported by the Laboratory Directed Research and Development (LDRD) Program at NREL. The development of the solution-phase doping and s-SWCNT separations at NREL was funded by the Solar Photochemistry Program, Division of Chemical Sciences, Geosciences, and Biosciences, Office of Basic Energy Sciences, U.S. DOE. B.N.-B. and N.J. S. received support from the U.S. DOE, Office of Science, Office of Workforce Development for Teachers and Scientists (WDTS) under the Science Undergraduate Laboratory Internships (SULI) Program. The views expressed in the article do not necessarily represent the views of the DOE or the U.S. Government.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

1.
A.
Zevalkink
,
D. M.
Smiadak
,
J. L.
Blackburn
,
A. J.
Ferguson
,
M. L.
Chabinyc
,
O.
Delaire
,
J.
Wang
,
K.
Kovnir
,
J.
Martin
,
L. T.
Schelhas
,
T. D.
Sparks
,
S. D.
Kang
,
M. T.
Dylla
,
G. J.
Snyder
,
B. R.
Ortiz
, and
E. S. A.
Toberer
, “
Practical field guide to thermoelectrics: Fundamentals, synthesis, and characterization
,”
Appl. Phys. Rev.
5
,
021303
(
2018
).
2.
H.
Wang
and
C.
Yu
, “
Organic thermoelectrics: Materials preparation, performance optimization, and device integration
,”
Joule
3
,
53
80
(
2019
).
3.
J. L.
Blackburn
, “
Semiconducting single-walled carbon nanotubes in solar energy harvesting
,”
ACS Energy Lett.
2
,
1598
1613
(
2017
).
4.
J. L.
Blackburn
,
A. J.
Ferguson
,
C.
Cho
, and
J. C.
Grunlan
, “
Carbon-nanotube-based thermoelectric materials and devices
,”
Adv. Mater.
30
,
1704386
(
2018
).
5.
T.
Sun
,
B.
Zhou
,
Q.
Zheng
,
L.
Wang
,
W.
Jiang
, and
G. J.
Snyder
, “
Stretchable fabric generates electric power from woven thermoelectric fibers
,”
Nat. Commun.
11
,
572
(
2020
).
6.
J.
Choi
,
Y.
Jung
,
C.
Dun
,
K. T.
Park
,
M. P.
Gordon
,
K.
Haas
,
P.
Yuan
,
H.
Kim
,
C. R.
Park
, and
J. J.
Urban
, “
High-performance, wearable thermoelectric generator based on a highly aligned carbon nanotube sheet
,”
ACS Appl. Energy Mater.
3
,
1199
1206
(
2020
).
7.
J.
Choi
,
Y.
Jung
,
S. J.
Yang
,
J. Y.
Oh
,
J.
Oh
,
K.
Jo
,
J. G.
Son
,
S. E.
Moon
,
C. R.
Park
, and
H.
Kim
, “
Flexible and robust thermoelectric generators based on all-carbon nanotube yarn without metal electrodes
,”
ACS Nano
11
,
7608
7614
(
2017
).
8.
A. D.
Avery
,
B. H.
Zhou
,
J.
Lee
,
E. S.
Lee
,
E. M.
Miller
,
R.
Ihly
,
D.
Wesenberg
,
K. S.
Mistry
,
S. L.
Guillot
,
B. L.
Zink
,
Y. H.
Kim
,
J. L.
Blackburn
, and
A. J.
Ferguson
, “
Tailored semiconducting carbon nanotube networks with enhanced thermoelectric properties
,”
Nat. Energy
1
,
16033
(
2016
).
9.
Y.
Ichinose
,
A.
Yoshida
,
K.
Horiuchi
,
K.
Fukuhara
,
N.
Komatsu
,
W.
Gao
,
Y.
Yomogida
,
M.
Matsubara
,
T.
Yamamoto
,
J.
Kono
, and
K.
Yanagi
, “
Solving the thermoelectric trade-off problem with metallic carbon nanotubes
,”
Nano Lett.
19
,
7370
7376
(
2019
).
10.
Y.
Nakai
,
K.
Honda
,
K.
Yanagi
,
H.
Kataura
,
T.
Kato
,
T.
Yamamoto
, and
Y.
Maniwa
, “
Giant Seebeck coefficient in semiconducting single-wall carbon nanotube film
,”
Appl. Phys. Express
7
,
025103
(
2014
).
11.
B. A.
MacLeod
,
N. J.
Stanton
,
I. E.
Gould
,
D.
Wesenberg
,
R.
Ihly
,
Z. R.
Owczarczyk
,
K. E.
Hurst
,
C. S.
Fewox
,
C. N.
Folmar
,
K.
Holman Hughes
,
B. L.
Zink
,
J. L.
Blackburn
, and
A. J.
Ferguson
, “
Large n- and p-type thermoelectric power factors from doped semiconducting single-walled carbon nanotube thin films
,”
Energy Environ. Sci.
10
,
2168
2179
(
2017
).
12.
D.
Hayashi
,
Y.
Nakai
,
H.
Kyakuno
,
T.
Yamamoto
,
Y.
Miyata
,
K.
Yanagi
, and
Y.
Maniwa
, “
Improvement of thermoelectric performance of single-wall carbon nanotubes by heavy doping: Effect of one-dimensional band multiplicity
,”
Appl. Phys. Express
9
,
125103
(
2016
).
13.
Y.
Nonoguchi
,
K.
Ohashi
,
R.
Kanazawa
,
K.
Ashiba
,
K.
Hata
,
T.
Nakagawa
,
C.
Adachi
,
T.
Tanase
, and
T.
Kawai
, “
Systematic conversion of single walled carbon nanotubes into n-type thermoelectric materials by molecular dopants
,”
Sci. Rep.
3
,
3344
(
2013
).
14.
Y.
Nonoguchi
,
M.
Nakano
,
T.
Murayama
,
H.
Hagino
,
S.
Hama
,
K.
Miyazaki
,
R.
Matsubara
,
M.
Nakamura
, and
T.
Kawai
, “
Simple salt-coordinated n-type nanocarbon materials stable in air
,”
Adv. Funct. Mater.
26
,
3021
3028
(
2016
).
15.
B.
Norton-Baker
,
R.
Ihly
,
I. E.
Gould
,
A. D.
Avery
,
Z. R.
Owczarczyk
,
A. J.
Ferguson
, and
J. L.
Blackburn
, “
Polymer-free carbon nanotube thermoelectrics with improved charge carrier transport and power factor
,”
ACS Energy Lett.
1
,
1212
1220
(
2016
).
16.
W.
Gao
,
N.
Komatsu
,
L. W.
Taylor
,
G. V.
Naik
,
K.
Yanagi
,
M.
Pasquali
, and
J.
Kono
, “
Macroscopically aligned carbon nanotubes for flexible and high-temperature electronics, optoelectronics, and thermoelectrics
,”
J. Phys. D: Appl. Phys.
53
,
063001
(
2020
).
17.
Y.
Nonoguchi
,
A.
Takata
,
C.
Goto
,
T.
Kitano
, and
T.
Kawai
, “
Thickness-dependent thermoelectric power factor of polymer-functionalized semiconducting carbon nanotube thin films
,”
Sci. Technol. Adv. Mater.
19
,
581
587
(
2018
).
18.
C.
Bounioux
,
P.
Diaz-Chao
,
M.
Campoy-Quiles
,
M. S.
Martin-Gonzalez
,
A. R.
Goni
,
R.
Yerushalmi-Rozene
, and
C.
Muller
, “
Thermoelectric composites of poly(3-hexylthiophene) and carbon nanotubes with a large power factor
,”
Energy Environ. Sci.
6
,
918
925
(
2013
).
19.
C. K.
Mai
,
B.
Russ
,
S. L.
Fronk
,
N.
Hu
,
M. B.
Chan-Park
,
J. J.
Urban
,
R. A.
Segalman
,
M. L.
Chabinyc
, and
G. C.
Bazan
, “
Varying the ionic functionalities of conjugated polyelectrolytes leads to both p- and n-Type carbon nanotube composites for flexible thermoelectrics
,”
Energy Environ. Sci.
8
,
2341
2346
(
2015
).
20.
N. T.
Hung
,
A. R.
Nugraha
,
E. H.
Hasdeo
,
M. S.
Dresselhaus
, and
R.
Saito
, “
Diameter dependence of thermoelectric power of semiconducting carbon nanotubes
,”
Phys. Rev. B
92
,
165426
(
2015
).
21.
J. L.
Blackburn
,
S. D.
Kang
,
M. J.
Roos
,
B.
Norton-Baker
,
E. M.
Miller
, and
A. J.
Ferguson
, “
Intrinsic and extrinsically limited thermoelectric transport within semiconducting single-walled carbon nanotube networks
,”
Adv. Electron. Mater.
5
,
1800910
(
2019
).
22.
K. S.
Mistry
,
B. A.
Larsen
,
J. D.
Bergeson
,
T. M.
Barnes
,
G.
Teeter
,
C.
Engtrakul
, and
J. L.
Blackburn
, “
N-type transparent conducting films of small molecule and polymer amine doped single-walled carbon nanotubes
,”
ACS Nano
5
,
3714
3723
(
2011
).
23.
B.
Russ
,
A.
Glaudell
,
J. J.
Urban
,
M. L.
Chabinyc
, and
R. A.
Segalman
, “
Organic thermoelectric materials for energy harvesting and temperature control
,”
Nat. Rev. Mater.
1
,
16050
(
2016
).
24.
D.
Kiefer
,
L.
Yu
,
E.
Fransson
,
A.
Gómez
,
D.
Primetzhofer
,
A.
Amassian
,
M.
Campoy-Quiles
, and
C. A.
Müller
, “
Solution-doped polymer semiconductor: Insulator blend for thermoelectrics
,”
Adv. Sci.
4
,
1600203
(
2017
).
25.
Y.
Diao
,
L.
Shaw
,
Z.
Bao
, and
S. C. B.
Mannsfeld
, “
Morphology control strategies for solution-processed organic semiconductor thin films
,”
Energy Environ. Sci.
7
,
2145
2159
(
2014
).
26.
M. T.
Fontana
,
D. A.
Stanfield
,
D. T.
Scholes
,
K. J.
Winchell
,
S. H.
Tolbert
, and
B. J.
Schwartz
, “
Evaporation vs solution sequential doping of conjugated polymers: F4TCNQ doping of micrometer-thick P3HT films for thermoelectrics
,”
J. Phys. Chem. C
123
,
22711
22724
(
2019
).
27.
D. T.
Scholes
,
S. A.
Hawks
,
P. Y.
Yee
,
H.
Wu
,
J. R.
Lindemuth
,
S. H.
Tolbert
, and
B. J.
Schwartz
, “
Overcoming film quality issues for conjugated polymers doped with F4TCNQ by solution sequential processing: Hall effect, structural, and optical measurements
,”
J. Phys. Chem. Lett.
6
,
4786
4793
(
2015
).
28.
S. N.
Patel
,
A. M.
Glaudell
,
K. A.
Peterson
,
E. M.
Thomas
,
K. A.
O'Hara
,
E.
Lim
, and
M. L.
Chabinyc
, “
Morphology controls the thermoelectric power factor of a doped semiconducting polymer
,”
Sci. Adv.
3
,
e1700434
(
2017
).
29.
A.
Jha
,
H.-G.
Duan
,
V.
Tiwari
,
M.
Thorwart
, and
R. J. D.
Miller
, “
Origin of poor doping efficiency in solution processed organic semiconductors
,”
Chem. Sci.
9
,
4468
4476
(
2018
).
30.
V. A.
Davis
,
A. N. G.
Parra-Vasquez
,
M. J.
Green
,
P. K.
Rai
,
N.
Behabtu
,
V.
Prieto
,
R. D.
Booker
,
J.
Schmidt
,
E.
Kesselman
,
W.
Zhou
,
H.
Fan
,
W. W.
Adams
,
R. H.
Hauge
,
J. E.
Fischer
,
Y.
Cohen
,
Y.
Talmon
,
R. E.
Smalley
, and
M.
Pasquali
, “
True solutions of single-walled carbon nanotubes for assembly into macroscopic materials
,”
Nat. Nanotechnol.
4
,
830
834
(
2009
).
31.
A.
Pénicaud
,
P.
Poulin
,
A.
Derré
,
E.
Anglaret
, and
P.
Petit
, “
Spontaneous dissolution of a single-wall carbon nanotube salt
,”
J. Am. Chem. Soc.
127
,
8
9
(
2005
).
32.
J.
Tan
,
Z.
Chen
,
D.
Wang
,
S.
Qin
,
X.
Xiao
,
D.
Xie
,
D.
Liu
, and
L.
Wang
, “
Balancing the electrical conductivity and Seebeck coefficient by controlled interfacial doping towards high performance benzothienobenzothiophene-based organic thermoelectric materials
,”
J. Mater. Chem. A
7
,
24982
24991
(
2019
).
33.
S.
Kazaoui
,
Y.
Guo
,
W.
Zhu
,
Y.
Kim
, and
N.
Minami
, “
Optical absorption spectroscopy of single-wall carbon nanotubes doped with a TCNQ derivative
,”
Synth. Met.
135–136
,
753
754
(
2003
).
34.
J. G.
Jang
,
S. Y.
Woo
,
H.
Lee
,
E.
Lee
,
S. H.
Kim
, and
J.-I.
Hong
, “
Supramolecular functionalization for improving thermoelectric properties of single-walled carbon nanotubes–small organic molecule hybrids
,”
ACS Appl. Mater. Interfaces
12
,
51387
51396
(
2020
).
35.
J. M.
Salazar-Rios
,
A. A.
Sengrian
,
W.
Talsma
,
H.
Duim
,
M.
Abdu-Aguye
,
S.
Jung
,
N.
Fröhlich
,
S.
Allard
,
U.
Scherf
, and
M. A.
Loi
, “
Customizing the polarity of single-walled carbon-nanotube field-effect transistors using solution-based additives
,”
Adv. Electron. Mater.
6
,
1900789
(
2020
).
36.
K. S.
Mistry
,
B. A.
Larsen
, and
J. L.
Blackburn
, “
High-yield dispersions of large-diameter semiconducting single-walled carbon nanotubes with tunable narrow chirality distributions
,”
ACS Nano
7
,
2231
2239
(
2013
).
37.
J. L.
Blackburn
,
A. J.
Ferguson
, and
O. G.
Reid
, “
Spectroscopy of ground- and excited-state charge carriers in single-wall carbon nanotubes
,” in
Handbook of Carbon Nanomaterials
, edited by
R.
Bruce Weisman
and
J.
Kono
(
Rice University
,
2019
), pp.
237
296
.
38.
Z.
Li
,
J.
Ding
,
J.
Lefebvre
, and
P. R. L.
Malenfant
, “
Dopant-modulated conjugated polymer enrichment of semiconducting SWCNTs
,”
ACS Omega
3
,
3413
3419
(
2018
).
39.
H.
Wang
,
B.
Hsieh
,
G.
Jiménez-Osés
,
P.
Liu
,
C. J.
Tassone
,
Y.
Diao
,
T.
Lei
,
K. N.
Houk
, and
Z.
Bao
, “
Solvent effects on polymer sorting of carbon nanotubes with applications in printed electronics
,”
Small
11
,
126
133
(
2015
).
40.
R. H.
Boyd
and
W. D.
Phillips
, “
Solution dimerization of the tetracyanoquinodimethane ion radical
,”
J. Chem. Phys.
43
,
2927
2929
(
1965
).
41.
J. B.
Torrance
, “
The difference between metallic and insulating salts of tetracyanoquinodimethone (TCNQ): How to design an organic metal
,”
Acc. Chem. Res.
12
,
79
86
(
1979
).
42.
R.
Kroon
,
D.
Kiefer
,
D.
Stegerer
,
L.
Yu
,
M.
Sommer
, and
C.
Müller
, “
Polar side chains enhance processability, electrical conductivity, and thermal stability of a molecularly p-doped polythiophene
,”
Adv. Mater.
29
,
1700930
(
2017
).
43.
C.
Wang
,
D. T.
Duong
,
K.
Vandewal
,
J.
Rivnay
, and
A.
Salleo
, “
Optical measurement of doping efficiency in poly(3-hexylthiophene) solutions and thin films
,”
Phys. Rev. B
91
,
085205
(
2015
).

Supplementary Material