Sulfanilic Acid Pending on a Graphene Scaffold: Novel, Efficient

Subscriber access provided by UNIV OF DURHAM

Organic Electronic Devices

Sulfanilic Acid Pending on Graphene Scaffold: Novel, Efficient Synthesis and Much Enhanced Polymer Solar Cell Efficiency and Stability Using It as Hole Extraction Layer Fu-Gang Zhao, Cheng-Min Hu, Yu-Ting Kong, Bingyige Pan, Xiang Yao, Jian Chu, Zi-Wen Xu, Biao Zuo, and Wei-Shi Li ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.8b06562 • Publication Date (Web): 03 Jul 2018 Downloaded from http://pubs.acs.org on July 5, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Sulfanilic Acid Pending on Graphene Scaffold: Novel, Efficient Synthesis and Much Enhanced Polymer Solar Cell Efficiency and Stability Using It as Hole Extraction Layer Fu-Gang Zhao,*,† Cheng-Min Hu,† Yu-Ting Kong,† Bingyige Pan,† Xiang Yao,†, ‡ Jian Chu,† Zi-Wen Xu,†, ‡ Biao Zuo,† and Wei-Shi Li*,‡ †Department of Chemistry, Zhejiang Sci-Tech University, 928 Second Street, Hangzhou 310018, China ‡Key Laboratory of Synthetic and Self-Assembly Chemistry for Functional Molecules, Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, 345 Lingling Road, Shanghai 200032, China

ABSTRACT: In this contribution, we describe a novel, facile and scalable methodology for high-degree functionalization towards graphene by the reaction between bulk graphite fluoride and in-situ generated amine anion. Using this, the rationallydesigned sulfanilic acid pending on graphene scaffold (G−SO3H), a 2D -conjugated counterpart of poly(styrenesulfonate), is available. Combined reliable characterizations demonstrate that a very large quantity of sulfanilic blocks are linked to graphene through the foreseen substitution of carbon-fluorine units and an unexpected reductive defluorination simultaneously proceeds during the one-step reaction, endowing the resultant G−SO3H with splendid dispersity in various solvents and filmforming property via the former, and with recovered 2D -conjugation via the latter. Besides, the work function of G−SO3H lies at -4.8 ev, well-matched with P3HT donor. Awarded with these fantastic merits, G−SO3H behaves capable in hole collection and transport, indicated by the enhanced device efficiency and stability of polymer solar cells (PSC) based on intensively studied P3HT:PCBM blends as active layer. In particular comparison with conventional PEDOT:PSS and recently rising and shining GO, G−SO3H outperforms above 17% and 24%, respectively, in efficiency. More impressively, when these three unencapsulated devices are placed in a N2-filled glovebox at around 25 ºC for 7 weeks, or subject to thermal treatment at 150 ºC for 6 hours also in N2 atmosphere, or even rudely exposed to indoor air, G−SO3H-based PSCs exhibits the best stability. These findings enable G−SO3H to be a strongly competitive alternative of the existing hole extraction materials for PSC real-life applications. KEYWORDS: hole extraction material; functionalized graphene; polymer solar cell; sulfonation; graphite fluoride; defluorination

INTRODUCTION Solution-processable bulk-heterojunction (BHJ) polymer solar cells (PSCs), as a promising green solar power generation technology, have been attracting constant attention due to the integrated advantages of being flexible, lightweight, cost-efficient, semi-transparent, adaptable to highthroughput industrial manufacture technology, etc.1-11 During past few decades, a large number of high-performance electron donor materials12-18 as well as acceptors19-28 have been ingeniously invented, and substantially improved the record of light-to-electric power conversion efficiency (PCE) exceeding 10%.29-31 Apart from the active layer consisting of a blend of donor/acceptor, in view of the layered architecture of BHJ PSCs, other components are also equally important to enlarge PCE, such as hole extraction layer and electron extraction layer (HEL and EEL).32-34 With regard to HEL, its insertion between the interfaces of active layer and commonly used indium tin oxide (ITO) anode largely facilitates holes to transport towards and be collected on this electrode, meanwhile, selectively blocks electrons to diminish recombination of carriers, thus maximize device efficiency.34 The mostly employed HEL at present is poly(3,4ethylenedioxythiophene) doped with poly(styrenesulfonate) (PEDOT:PSS) having merits of being commercially available, water soluble and efficient collection of holes.35,36 But unfortunately, the PEDOT:PSS-based devices generally suffer from performance decay due to the extensively

known drawbacks of high acidity and hygroscopicity, which could corrode ITO and induce water penetration into the active layer.34, 37-39 Several wide bandgap inorganic materials, such as V2O5, MoO3 and NiO, have been exploited as the alternatives of PEDOT:PSS.40, 41 But such metal oxide HELs are vacuum-deposited with a high manufacturing cost, obviously incompatible with solution-processable and printable technological advantages of organic electronics. As an old but previously silent material,42, 43 graphene oxide (GO) glows renewed vitality since the birth of graphene, and is found to be talented in versatile fields ranging from nanocomposites,44-46 to energy conversion and storage,47-51 environment,52, 53 and so many others. Recently, GO has been reported to serve as an efficient hole transporting material and a promising candidate of PEDOT:PSS alternative for PSCs.34, 37, 54-60 However, the chemical structure and composition of GO are notoriously complicated and still remain ambiguous to date, although some oxygenated functional groups (e.g. −OH, C=O, −COOH, epoxy and phenolic groups), often accompanied by some sulfur species (classical Hummers’ method) are widely accepted to be included.61-64 As a matter of fact, the final structure of GO largely depends on each chemical oxidation process used during the preparation.65 What is more, the oxygen-containing functionalities are labile, may causing performance degradation of the devices using GO as HEL. With these concerns, well-defined sulfanilic acid pending on graphene scaffold, inspired by two-dimensional (2D) structure of GO, extended

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

conjugation of PEDOT, together with sulfonic acid moiety of PSS, is rationally designed and readily synthesized via the graft strategy of sodium sulfanilate to commercially available bulk graphite fluoride (bGF). After final exchange of sodium with proton, sulfanilic acid functionalized graphene (designated as G−SO3H) is harvested, and investigated as HEL for the most popular BHJ PSC device based on poly(3hexylthiophene):[6,6]-phenyl-C61 butyric acid methyl ester (P3HT:PC61BM) blends as active layer. Owing to the combined virtues of G−SO3H with unique 2D structure, highly electrical conductivity, wonderful solution processability and suitable energy level alignment, this device exhibits the best characteristics (Table 1, vide infra) in comparison with other three control cells with state-of-the-art PEDOT:PSS, shining GO star as HEL, and without HEL. More impressively, the exploitation of G−SO3H as HEL endows model PSC devices with a much longer lifetime in stability experiments. These findings clearly demonstrate that G−SO3H, as a 2D conjugated counterpart of PSS, is a promising HEL material and a highly practical replacement for conventional PEDOT:PSS. RESULTS AND DISCUSSION Considering that bGF behaves very chemically inert, as illustrated in Scheme 1, sodium sulfanilate reagent is firstly deprotonated with stoichiometric NaNH2 in order to strengthen its reaction activity (Caution: NH3 release during the process). Afterwards, bGF powders are dispersed in N,N-dimethylformamide (DMF) with the assist of gentle ultrasonication, then the suspension is injected dropwisely into the pre-synthesized sodium sulfanilate anion solution. After careful quenching this reaction with methanol, sodium sulfanilate functionalized graphene (G−SO3Na) intermediate is available by PTFE membrane filtration of the mixture and subsequent purification for the filter cake. Since interfacial protonic doping towards donor polymer can achieve favourable Ohmic contacts and boost photovoltaic characteristics,43, 66 proton exchange for sodium is successfully carried out and finally affords the target sulfonated graphene (G−SO3H) products (See Experimental Section for more details on each procedure).

Scheme 1. Synthetic overview of sulfanilic acid covalently functionalized graphene (G−SO3H). To verify the effective graft of sulfanilate moiety onto graphene sheet, elemental analysis (EA), thermogravimetric analysis (TGA), X-ray photoelectron spectroscopy (XPS),

Page 2 of 12

Fourier transform infrared spectroscopy (FT-IR) and Raman characterizations are performed on the resultant products. Firstly, EA provides the precise H, N and S contents of 1.88%, 3.54% and 7.57%, respectively, for G−SO3H sample. Its F content sharply diminishes to 1.08% and C content rises to 72.98% as opposed to commercial bGF starting material, whose C and F contents are 49.18% and 49.84%. Apparently, the presence of H, N and S directly demonstrates the attachment of sulfanilate. More visually, we can observe the new peaks of N, O and S at 400.8, 533.8 and 169.2 eV, respectively, in the survey XPS spectra of G−SO3H (Figure 1a), which undoubtedly originates from para-aminobenzene sulfonic acid mother unit. What’s more, the intensity of F signal at 689.5 eV is much weaker than that for bGF, reflecting little F remaining in G−SO3H. These observations strongly agree with EA data. In addition, the appearance of a band at 285.9 eV attributed to C−N bond in deconvolved C 1s spectra for G−SO3H,67 which is absent in bGF C 1s spectra, confirms the sulfanilic functionalization towards graphene through C−N linker (Figure 1b and 1c). It is also seen that the peak intensities of the C−Fx components at binding energies of 287.5-292.9 eV in C 1s spectra of G−SO3H are tremendously suppressed with respect to that of bGF, uncovering the occurrence of substantial substitution of C−Fx species by sulfanilate. In N 1s spectra of G−SO3H (Figure 1e), one component at 400.8 eV associated with N−benzenesulfonic acid and the other at 402.3 eV assignable to N−graphene scaffold are reasonably fitted, while the N 1s spectra for sodium sulfanilate displays a single peak centred at 400.5 eV (Figure 1d). These results prove, once again, the valid coupling of sulfanilate block with graphene sheet via covalent C−N bond. Likewise, O 1s spectra for sodium sulfanilate appears a single and symmetric peak (Figure 1f), yet that for G−SO3H deconvolves another new minor component at 530.9 eV corresponding to C−O species (Figure 1g), in addition to the major and definite S−O3 moieties. The invasion of unexpected C−O units may be in connection with a little amount of NaOH impurities generated from as-supplied NaNH2 hydrolysis. S 1s and F 1s spectra of G−SO3H are presented in Supporting Figures. Due to the homology, the shapes of S 1s spectra for G−SO3H and sodium sulfanilate appear almost identical (Figure S1). The comparison for F 1s spectra between G−SO3H and bGF determines the carbon−fluorine units left in G−SO3H to be CF2 ones (Figure S2). Covalent loading of sulfanilate functionalities on bGF is further consolidated by FT-IR according to the presence of new bands, not present in bGF. For instance, the featured stretching vibrations at 1390 cm−1, and in the ranges of 3500-3050 cm−1 as well as 1300-830 cm−1 is attributed to C−N linker, active N−H/O−H species and sulfanilic group, respectively (Figure 1h).67 All these findings clearly verify the successful graft of the desired sulfanilic functionality onto the 2D graphene scaffold. For the topic regarding the covalent functionalization to graphene, the degree of functionalization, namely, the grafting amount of sulfanilic group in this contribution, is a vitally important factor. In fact, it can be exactly calculated using previous EA data (see Supporting Information for the calculation details). Referring to S content of 7.57%, the grafting amount of sulfanilic block per gram G−SO3H adducts is estimated to be  2.36 mmol ( 0.406 g). In comparison with other


Page 3 of 12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


strategies for sulfonation towards graphene (Figure S3), our proposal achieves an exceptionally high degree of sulfonic functionalization. Moreover, TGA is also employed to measure the grafting amount under N 2 atmosphere. First of all, the thermal behavior of bGF precursor is recorded, and we are aware that bGF is fairly stable below 450 °C and exhibits weight loss of 60% in the temperature range 450−650 °C (Figure 1i). The TGA profile unravels that the decomposition of G−SO3H adducts is triggered above 200 °C, and ultimately it loses roughly 41% of the weight when the temperature rises

1000 800

410

405

600

400

400

200

395

100

545

close to 460 °C. The weight loss during this temperature interval just corresponds to detachment/decomposition of the sulfanilate groups. Thus, the contribution from sulfanilate units in G−SO3H product suggested by TGA is about 41%, in perfect agreement with the value calculated from EA results. Incidentally, the thermal behavior of GO is also studied, and its TGA curve tells a weight loss of about 20% at 150 °C. These data clearly disclose the intrinsic lability of oxygen species within GO, and inferior stability to G−SO3H.

296

292

288

284

296

540

535

530

525 545

540

292

535

288

530

284

525 4000

410

405

3000

400

2000

395

1000

bGF

80 60 40

GO G-SO3H 200

400

600

800

Figure 1. (a) XPS survey spectra of sodium sulfanilate, bGF and G−SO3H samples; (b, c) C 1s XPS spectra for bGF (b) and G−SO3H (c); (d, e) N 1s XPS spectra for sodium sulfanilate (d) and G−SO3H (e); (f, g) O 1s XPS spectra for sodium sulfanilate (f) and G−SO3H (g); (h) FT-IR spectra of sodium sulfanilate, bGF and G−SO3H; (i) TGA profiles of bGF, GO and G−SO3H; (j) electrically conductive properties of bGF, GO, PEDOT:PSS and G−SO3H; (k, l) Raman spectra of bGF (k) and G−SO3H (l). Since G−SO3H is aimed to be a carrier transport material, it should be electrically conductive. But it is well known that bGF and its building block of fluorographene, reckoned as the 2D counterpart of Teflon,68 are extremely electrically insulating. It is intriguing for the conductive behavior of graphene derivatives prepared from nonconductive bGF. Fourpoint probe measurement discloses that the drop-casted G−SO3H film possesses an electrical conductivity of  1.2 S m-1, not only approximately 12 and 4 orders of magnitude higher than bGF precursor ( 10-12 S m-1) and popular GO film (5.7×10-4 S m-1), respectively, even also 7.5 times more conductive than PEDOT:PSS (Clevios™) (Figure 1j). The

inversion in conductivity implies a -conjugated structure within G−SO3H, yet -conjugation in bGF is almost totally collapsed. In this regard, EA and XPS data also corroborate the recovery of conjugated structure in the course of the reaction between bGF and sodium sulfanilate anion as following deduction. For easy explanation, firstly, the carbon−fluorine bondings of bGF are idealized to be C−F singlet state, i.e. regardless of CF2 and CF3 units. Now back to the reaction again, the minimum loading weight of sulfanilic group is supposed to be exceeding 0.9 g per gram G−SO3H adducts in the case that C−F bonds are completely replaced by sulfanilic block. (See Supporting Information for calculation



details). Notably, the ideal value hugely deviates from the real ( 0.406). Another significant fact deserves notice that fluorines left in G−SO3H product are nearly negligible, indicated by F content of only 1.08% in EA together with almost entire vanish of C−Fx signals in C 1s and survey XPS spectra. These combined results proof the occurrence of accompanied reductive defluorination with the foreseen substitution reaction, which induces the rehabilitation of -conjugated system. As a more straightforward and powerful tool to characterize sp2 carbon conjugated structure, Raman spectra is performed on G−SO3H and bGF as a reference. Of expect, bGF presents a much intense D band around 1290 cm-1 which is induced by disordered sp3 carbons, and a medium G band around 1545 cm-1 corresponding to in-plane bond stretching of sp2 carbons.69 The intensity ratio of both bands (ID/IG) reaches 2.46, indicative of intrinsic high-density disordered C(sp3)–Fx domains within bGF. In contrast, the G band of G-SO3H gets a little stronger in its intensity, accordingly, the ID/IG value decreases to 1.49, suggesting that inplane conjugated C=C bonds largely grow during the reaction. That is, nucleophilic substitution of C–Fx and reductive defluorination concurrently proceed in the perspective of reaction mechanism,70 realizing the desired functionalization through the former, and also the repair of 2D -conjugation through the latter as an appealing bonus. However, we can also observe that D band of G–SO3H still remains very strong, and quantitatively, its intensity is 1.49 folds larger than G band. This fact validates, on the other hand, the attachment of a large amounts of sulfanilic blocks on graphene skeleton. To examine the morphology of bGF before and after introduction of sulfanilic blocks, field-emission scanning electron microscopy (FE-SEM) is carried out. As depicted in Figure 2a, bGF starting material, consisting of numerous tightly and orderedly stacked fluorographene monolayer, appears thick pizza-shaped monoliths. In contrast, G–SO3H product seems much fluffier than bGF, and the compact monolithic plates observed in bGF open up and separate into thin flakes. This observation demonstrates that sodium sulfanilate molecules behave as organic wedges to exfoliate bGF agglomerates during reactive functionalization. Furthermore, energy dispersive spectroscopy elemental mapping is performed on G–SO3H sample, and one can see that the elements of carbon, nitrogen, sulfur and oxygen are readily identified and distribute rather homogeneously throughout the bulk G– SO3H powders, suggesting that sulfanilic functionalities are really anchored on graphene surface again. In addition, the micromorphology and crystalline structure of the exfoliated G–SO3H nanosheets from its dilute DMF dispersion are observed by transmission electron microscopy (TEM) and the corresponding selected area electron diffraction (SAED). As shown in Figure S4, G–SO3H flakes appear almost transparent and wrinkled. Its SAED pattern exhibits well-defined six-fold symmetry dot arrays, which also suggests a well recovery of 2D crystal domains in G–SO3H sheets during the reaction.

Page 4 of 12

Figure 2. FE-SEM images of (a) bGF and (b) G–SO3H sample; (c-f) elemental mapping images of G–SO3H sample. To be compatible with low-cost and high-throughout roll-to-roll production of PSCs, splendid dispersity and filmforming capabilities are fundamental requirements for HEL materials. Owing to intrinsic strong solubilization of sulfanilate functionality, G–SO3H wonderfully disperses in water as well as varieties of routine polar organic solvents over a prolonged period for more than 2 months (Figure S5), guaranteeing the solution processing feasible. Macroscopically visual and flexible G–SO3H films can be fabricated by drop-casting its DMF dispersion ( 2 mg mL-1) on the commonly used substrate (Figure S6), and Figure 3a shows a representative free-standing G–SO3H film with wrinkles and inward shrinks when stripped off ITO slide after natural evaporation of most DMF. Microcosmically, we clearly observe the characteristic flaky morphology of graphene family using atomic force microscopy (AFM) with mica as a support (Figure 3b). The thickness of G–SO3H sheets is about 1.4 nm, slightly surpassing that of GO plates ( 1 nm) (Figure S7). Like this, G–SO3H thin layers are also spincoated on ITO. Figure 3c and 3d show the AFM height images for a bare ITO surface before and after spincoated with a 0.8 mg mL-1 G–SO3H DMF dispersion. As the thickness of a singleor few-layer G-SO3H nanosheets is less than the root mean square (rms) roughness of ITO surface (rms roughness of 4.15 nm), it is no surprise that G–SO3H sheets are invisible on rough ITO (Figure 3d), and it is very difficult to accurately measure the thickness of G–SO3H thin layer. Despite we cannot catch sight of G–SO3H film, the deposition of GSO3H solution really causes ITO surface much smoother, which is also demonstrated by a smaller rms roughness (2.64 nm) than bare ITO. This change indicates the existence of a thin G–SO3H layer. To address the difficulty in determining the thickness of G-SO3H layer on ITO, G–SO3H DMF dispersion is spincoated on atomically flat Si


Page 5 of 12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


substrates under the same deposition conditions in devices. Figure 3e shows a representative spincoated G–SO3H film with a thickness of 2 nm on Si wafer. By finely tuning the concentration of G–SO3H dispersion and spin-coating rounds, uniform thin films with thicknesses ranging from 2 to 6 nm are obtained (see Experimental Section for details on film fabrication, and Figure S8). The observed fantastic film-forming properties of G–SO3H is significantly advantageous for its application as HEL in PSCs.

structure, wonderful dispersive behavior in common solvents, excellent film-forming property and weak absorption in visible range, promise that it should be an attractive hole extraction material.

2

0 0 0

300

-4

600

none GO PEDOT:PSS G-SO3H

-8 -12 0.0

0.2

0.4

0.6

60 40

none GO PEDOT:PSS G-SO3H

20

100 80

0 300

60

40

bare ITO 3 nm 5 nm

20

2 nm 4 nm 6 nm

400

600

800

Figure 3. (a) A digital photograph of free-standing dropcasted G–SO3H film peeling off ITO substrate; (b-e) AFM height images of G–SO3H sheets on mica (b), bare ITO surface (c), ITO surface spincoated with G–SO3H film (d) and G– SO3H film deposited on Si wafer (e); (f) Optical transmission spectra of bare ITO and ITO deposited with different thick G–SO3H films. How the deposition of G–SO3H film onto bare ITO surface impacts the optical transparency deserves investigation since high sunlight transmission of anode is the prerequisite for photovoltaic device. Figure 3f describes the transmission spectra of pristine ITO before and after deposited with various thicknesses of G–SO3H layers. The transmittance decreases slightly when the thickness of G–SO3H films grow from 2 to 4 nm. But a little more loss in transparency is found as G–SO3H films are thickened to 5 to 6 nm. Overall, it scarcely alters the high transparency of ITO anode in visible range. All the prominent features of G–SO3H described above, particularly for 2D -conjugated and conductive

600

700

0

2 nm 4 nm 6 nm

-3 400

500

3 nm 5 nm

-6 -9 0.0

0.2

0.4

0.6

Figure 4. (a) Schematic energy level diagrams of the bottom electrode ITO, HEL materials (G–SO3H, GO and PEDOT:PSS), P3HT (donor), PC61BM (acceptor), LiF EEL and the top electrode Al; (b) Schematic illustration of the PSC device structure with G–SO3H as HEL; (c, d) J–V characteristics (c) and EQE curves of the devices with no HEL, with PEDOT:PSS film (30 nm), GO film (2 nm) and G–SO3H film (2 nm) as HEL; (e) J–V curves of the PSCs with G–SO3H films as HEL with different thicknesses. Besides, the paramount property is whether the energy level alignment of G–SO3H material match properly the highest occupied molecular orbital (HOMO) level of the donor to minimize energy barriers for efficient hole extraction, should be determined prior to its use as HEL. The work function of G–SO3H



Page 6 of 12

material is measured to be −4.8 eV by scanning Kelvin nm-thick G–SO3H thin film. Notably, a big advance of probe microscopy. Such value matches well with HOMO 17.5% in device efficiency by the utilization of G–SO3H level of P3HT (Figure 4a), and facilitates the formation as HEL is mainly ascribed to an increase of FF by 13.6%, of an Ohmic contact between the P3HT:PCBM active in comparison to PEDOT:PSS. In a good consistency layer and the ITO electrode, which is well established in with the J-V curves, the external quantum efficiency pioneering works.43, 66 To assess the potentiality of G– (EQE) spectra illustrated in Figure 4d also reveals that SO3H as HEL in PSCs, finally, we construct the extenthe hole-transporting capability of G–SO3H lies at the sively-studied bulk heterojunction devices based on top and slightly surpasses that of the conventional P3HT:PCBM as active layer with the configuration of PEDOT:PSS. We attribute so exceptional hole-extracITO/G–SO3H(2 nm)/P3HT:PC61BM (200 nm)/LiF(1 tion performance to the integrated merits with the 2D nm)/Al(100 nm) (Figure 4b). For comparison, one ITOplanar conductive structure, suitable energy level only device consisting of ITO/P3HT:PC 61BM/LiF/Al, alignment and splendid film-forming property of and another two incorporating 30 nm PEDOT:PSS and G−SO3H. 2 nm GO as HEL, respectively, are fabricated. The phoFurthermore, we investigate the influence of tovoltaic characteristics of these devices are acquired G−SO3H film thickness on device characteristics. The under simulated AM1.5G illumination at 100 mW cm-2. different thicknesses of G−SO3H layer (26 nm) are Figure 4c shows the current density-voltage (J−V) achieved by carefully modulating the concentration of curves together with the numerical performance data DMF dispersion and spin-coating rounds. It is found listed in Table 1. As can be seen, the HEL-free device afthat the device parameters nearly remain unchanged as fords a short-circuit current density (Jsc) of 7.90 mA cmthe thickness increases from 2 to 3 nm, and slightly di2, open-circuit voltage (V oc) of 0.47 V, fill factor (FF) of minish while the thickness further improves to 4 nm 0.49, and very limited power conversion efficiency (Figure 4e and Table 1). However, further thickening (PCE) of 1.81%. The insertion of a 2 nm-thick GO thin G−SO3H film exerts larger detrimental effects upon defilm between ITO and active layer considerably provice performance. Specifically, the values of JSC, FF and motes photovoltaic efficiency up to 3.08%, accompanyPCE drop down to 6.56 and 6.38 mA cm –2, 0.57 and 0.51, ing with a VOC of 0.60 V, JSC of 8.86 mA cm–2 and FF of 2.28 and 1.98%, respectively, when the thickness 0.58. Such photovoltaic performance is roughly parallel grows to 5 and 6 nm. It is not surprising because sucto that of the cell based on state-of-the-art PEDOT:PSS cessive thickening G−SO3H layer leads to a growth in (30 nm) as interfacial layer, which delivers a VOC of 0.61 the series resistance (Rs) on the one hand (Table 1), and V, JSC of 9.08 mA cm–2, FF of 0.59, and PCE of 3.26%. also gradually lessens ITO transparency on the other Among these, that device exhibits the best performance hand (Figure 3f). Both are responsible for the reduced Jsc, of a VOC of 0.61 V, JSC of 9.38 mA cm–2, FF of 0.67, and PCE FF and the resultant PCE values. up to 3.83% when the counterpart is changed to be 2 Table 1. The performance of PSC devices without HEL material and with GO, PEDOT:PSS and G–SO3H (2–6 nm) as HEL. HEL material

aPCE

thickness

Jsc

Voc

(nm)

(mA cm-2)

(V)

none

/

7.9

0.47

PEDOT:PSS

30

9.08

GO

2

G–SO3H

FF

PCEa

Rs

(%)

( cm2)

0.49

1.81 (1.72)

10.4

0.61

0.59

3.26 (3.15)

2.5

8.86

0.60

0.58

3.08 (2.83)

3.6

2

9.38

0.61

0.67

3.83 (3.69)

1.8

G–SO3H

3

9.36

0.61

0.66

3.78 (3.56)

2.0

G–SO3H

4

9.02

0.61

0.62

3.42 (3.22)

2.7

G–SO3H

5

6.56

0.61

0.57

2.28 (2.10)

4.9

G–SO3H

6

6.38

0.61

0.51

1.98 (1.77)

8.1

value of the best device and data in parentheses are average PCE.

In light of real-life application concerns, a long-term operation stability of PSCs is highly desirable because it probably lie above all the performance evaluation indexes for commercialization. Disappointedly, the devices based on classic PEDOT:PSS as HEL usually suffer from inferior prolonged stability as a result of the previously mentioned adversely acidic and hygroscopic feature.34, 37-39 In order to wonder the stability of G–SO3H-based devices, the performance is plotted versus storage period in a N2-filled glovebox with no encapsulation. As a control, the PCE evolution

of PEDOT:PSS-, GO-based devices along with time is also recorded under the same conditions. As plotted in Figure 5a, despite the PCE values of three kinds of devices all gradually decay, PEDOT:PSS-, and GO-based devices are subjected to higher PCE losses than G–SO3H-based counterpart after storage duration for 7 weeks (sequential PCE retention rates of 74% and 87% versus 90% of the initial value). More impressively, when these three devices are subject to rude exposure to indoor air conditions, the PCE value of PEDOT:PSS-based cell rapidly degrades to nearly zero after


Page 7 of 12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


9 hours, and GO-based device sacrifices a half of original PCE (Figure 5b). In sharp contrast, G–SO3H-based counterpart can reserve above 60% of the initial PCE in this case. One more attractive merit is that G–SO3H-based PSC is highly comparable to PEDOT:PSS-based, but outperforms GO-based counterpart in tolerance to thermal treatment. As shown in Figure 5c, upon thermal annealing at 150 ºC for 6h in N2 atmosphere, the devices based on PEDOT:PSS and G– SO3H keep 91% and 90% of the original PCE, respectively, while such retention value is 78% for the device based on GO. We deduce that the disparities may originate from the thermal decomposition of labile oxygen-containing groups amongst GO and excellent thermal stability of G–SO3H material. All the results suggest that G–SO3H is a more stable and practical material for hole extraction than conventional PEDOT:PSS as well as the so hot GO of today, in upgrading PSC PCE and longevity.

1.0

0.8

G-SO3H GO PEDOT:PSS 0.6 0

14

28

42

1.0

G-SO3H

0.5

GO PEDOT:PSS 0.0 0

3

6

9

Figure 5. PCE decay of unencapsulated PSC devices based on G–SO3H, PEDOT:PSS and GO as the HEL along with storage time in (a) a N2-filled glovebox; (b) indoor air; (c) after thermal annealing at 150 ºC for 6h in a N2-filled glovebox.

N,N-dimethyl formamide was obtained by refluxing with CaH2 under vacuo for 12h and freshly distilled prior to use. All other chemicals were used as received, without any further purification. 1.2 Synthesis of sodium sulfanilate functionalized graphene (G−SO3Na) intermediate (a) In-situ generation of active sodium sulfanilate anion: to a 500 mL two-necked flask, 7.7435g (39.48 mmol) white sodium sulfanilate powders and 1.5397 g (39.48 mmol) NaNH2 plates were added. After vacuum pumping and N2 charging for three rounds, 200 mL freshly distilled DMF was injected into the flask using a syringe. The mixture was slowly warmed to 80 °C and vigorously stirred for 24h. Careful attention should be paid to the increasing pressure of the system due to NH3 release as the reaction is proceeding. After that, the mixture was cooled to room temperature, affording sodium sulfanilate anion solution; (b) Nucleophilic substitution reaction of sodium sulfanilate anion with bulk graphite fluoride (bGF): to another 200 mL flask, 1.0122 g bGF (F%: 50%, 26.68 mmol F) and 100 mL dry DMF were added. The mixture was subject to ultrasonic treatment for 2h (40 KHz, 50W). Afterwards, the suspension was dropwisely injected into the flask charged with sodium sulfanilate anion under ice/water bath using a syringe while vigorously stirring in 2h. After taking the ice/water bath away, the mixture was slowly warmed to 120 °C and vigorously stirred for 72h. Finally, this reaction was naturally cooled down to room temperature and carefully quenched with 50 mL methanol. The mixture was filtered over a 0.22 μm PTFE membrane, and the filter cake was subjected to thorough washing with ethanol (500 mL*3), tetrahydrofuran (500 mL*3), acetone (500 mL*3), dichloromethane (500 mL*3) and deionized water (500 mL*3) for removing any impurities, affording a crude product. To harvest fine product, the crude product was re-dispersed in 1000 mL DMF with the assist of ultrasonication for 30 min. The dispersion was subject to centrifugation (8000 rpm, 20 min), collecting upper suspension and abandoning the precipitates. After vacuum evaporation of DMF of all dispersion, we obtain the brightly black products, assigned name of G– SO3Na (0.6684 g). Elemental analysis (%), S: 6.98; C: 66.86; H: 1.44; N: 3.11; F: 0.96; Na:4.87%. 1.3 Synthesis of G–SO3H by exchange of Na+ with proton: 0.3439 g G–SO3Na powders were dispersed in 300 mL 0.2 mol L–1 HCl aqueous solution with the assist of ultrasonication for 30 min. Then, the dispersion was filtered over a 0.22 µm Nylon membrane under vacuo, and the filter cake was washed with 50 mL 0.5 mol L–1 HCl solution for three times again. Finally, the filter cake was washed with a large amount of deionized water till the pH value of filtrate was about 7, affording final G–SO3H sample (0.3087 g). Elemental analysis (%), S: 7.57; C: 72.98; H: 1.88; N: 3.54; F: 1.08; Na: 0.003. 2. Characterizations.

EXPERIMENTAL SECTION 1. Material Synthesis. 1.1 Chemicals. Bulk graphite fluoride (bGF) was purchased from Shanghai CarFluor chemicals corporation. Sodium sulfanilate, sodium amide (NaNH2), N,N-dimethyl formamide (DMF) were purchased from Sigma-Aldrich. Anhydrous

Elemental analysis on C, H and N contents was carried out with an Elementar vario EL III elemental analyzer. The contents of F and S elements were measured by using a typical oxygen flask combustion method, in which the sample was totally burned in an O2 atmosphere, and the produced inorganic fluorine and sulfur compounds were individually



collected, and titrated with thorium nitrate and barium perchlorate, respectively. Sodium element was acquired by atomic absorption spectroscopy. Fourier transform infrared spectra were recorded on a Nicolet Avatar-360 Fourier transform infrared spectrophotometer using a potassium bromide pellet. Raman spectra were performed on a Renishaw inVia Reflex micro-Raman spectrometer using a 100fold objective lens and crystal laser excitation at 514.5 nm with a power of 0.1 mW. X-ray photoelectron spectroscopy was conducted on a PHI-5000 VersaProbe spectrometer under 10-7 Pa using a monochromatic Al Kα X-ray source operating at 100 W. The thermogravimetric analysis was carried out in N2 on a TA instrument with a heating rate of 10 °C min-1. The UV-vis transmission spectra were performed on a Hitachi U-3310 spectrophotometer at room temperature. Atomic force microscopy images were acquired under ambient conditions on a Veeco instrument Nanoscope IIIa Multimode apparatus operating in a non-contact mode with a silicon tip and cantilever operating at a frequency of 325 kHz and a scanning speed of 1 Hz. Samples were prepared by placing a drop of their dilute N,N-dimethyl formide dispersion on a fresh mica substrate and dried in a vacuum oven at room temperature. 3. Device Fabrication and Characterization. Prior to use, ITO glass substrates were cleaned for 10 min each sequentially with deionized water, acetone, and isopropanol, followed by drying with pressurized N2 flow and UV-ozone treatment for 15 min. A thin layer (30 nm) of filtered PEDOT:PSS (Heraeus Clevios P VP. Al 4083) was spincoated on the top of ITO at 4000 rpm for 60 s and baked at 120 °C for 15 min. GO thin film (2nm) was spincoated from its aqueous solution (1.0 mg mL-1) at 2000 rpm for 60 s, followed by heating at 110 °C for 15 min. G-SO3H films with thickness of 2, 3, 4, 5, 6 nm were spincoated from its DMF solution and achieved with the following conditions: 2 nm@ 0.6 mg mL-1, 2000 rpm, 60 s; 3 nm@ 1.0 mg mL-1, 2000 rpm, 60 s; 4 nm@ 0.6 mg mL-1, 2000 rpm, 60 s, repeated for two times; 5 nm@ 1.0 mg mL-1, 2000 rpm, 60 s firstly, then 0.6 mg mL-1, 2000 rpm, 60 s; 6 nm@ 1.0 mg mL-1, 2000 rpm, 60 s, repeated for two times. The G-SO3H-covered ITO substrates were heating at 110 °C for 15 min. Afterwards, the HEL-coated ITO substrates were transferred into a dry N2filled glovebox (O2

Sulfanilic Acid Pending on a Graphene Scaffold: Novel, Efficient

Recommend Documents