Conductive composites of polyaniline–polyacid complex and graphene nanostacks

Synthetic Metals 211 (2016) 89–98

Contents lists available at ScienceDirect

Synthetic Metals journal homepage: www.elsevier.com/locate/synmet

Conductive composites of polyaniline–polyacid complex and graphene nanostacks Olga D. Iakobsona , Oxana L. Gribkovaa,* , Alexey R. Tameeva , Valery V. Kravchenkob , Alexander V. Egorovc , Anatoly V. Vannikova a b c

A.N Frumkin Institute of Physical Chemistry and Electrochemistry of the Russian Academy of Sciences, 31, bld.4, Leninsky Prospect, Moscow 119071, Russia M.V Lomonosov Moscow State University of Fine Chemical Technologies, 86, Vernadskogo Prospect, Moscow 119571, Russia M.V Lomonosov Moscow State University, 1, Leninskie Gory, Moscow 119991, Russia

A R T I C L E I N F O

A B S T R A C T

Article history: Received 21 October 2015 Received in revised form 10 November 2015 Accepted 17 November 2015 Available online xxx

Nanocomposites based on graphene and polyaniline–polyacid complexes with tunable electrical conductivity are elaborated. An influence of graphene oxidation degree on conductivity of the nanocomposites is investigated. The change of optical and electrical properties after graphene introduction into polyaniline–polyacid complexes is explained by the formation of graphene nanostacks of different size and their different distribution in the film bulk. The role of (i) internal interactions between graphene sheets revealed by high-resolution TEM and AFM and (ii) external interactions between graphene and polyaniline or polyacid of different hydrophobicity elucidated by UV–vis, FTIRspectroscopies and pH-measurements is discussed. In case of uniform distributed graphene sheets having a low oxidation degree, the electrical conductivity of the nanocomposites based on polyaniline complexed with more hydrophilic polyacid increases up to 20 times in respect to initial polyaniline complex. ã 2015 Elsevier B.V. All rights reserved.

Keywords: Polyaniline Graphene Electrical conductivity UV–vis-spectroscopy FTIR-spectroscopy TEM AFM

1. Introduction Successful preparation of graphene/polymer nanocomposites is one of the topical issues of the modern material science aiming to improve both the processability of graphene and the physical properties of the host polymers [1]. Graphene is nearly transparent, cost-effective and relatively easy to prepare (“easy of access”) material possessing high electron mobility and stability [2]. It has been established already that small graphene sheets can provide polymer nanocomposites with new functions and applications in various electronics-related systems [3]. In turn, a conductive polymer is a preferable component of the graphene/polymer nanocomposites, since it is known to possess sufficient conductivity [4–6]. Such a combination promotes synergistic effect of the advantages of the both components [7]. Due to known environmental stability, low cost of monomer and simplicity of the polymer synthesis, polyaniline (PANI) seems to be one of the most promising conductive polymers. Moreover, the compatibility of PANI and graphene resulting from interactions between them is expected [8]. The different types of PANI–

* Corresponding author. E-mail address: [email protected] (O.L. Gribkova). http://dx.doi.org/10.1016/j.synthmet.2015.11.018 0379-6779/ ã 2015 Elsevier B.V. All rights reserved.

graphene nanocomposites and their potential applications are briefly reviewed in Refs. [6,9,10]. PANI–polyacid complex prepared by chemical oxidative polymerization of aniline in the presence of a polyacid seems to be particularly promising for nanocomposite creation. In addition to other advantages described thoroughly in refs. [11,12], the most attractive property of PANI–polyacid complex for practical application is its water-dispersancy. The introduction of graphene into PANI–polyacid complexes has been reported in several articles only [13–17]. Majority of these studies considers the introduction of graphene into the reaction medium before aniline polymerization (in situ chemical polymerization of aniline over graphene dispersion) [13–16]. In the preliminary studies [18,19] on one type of PANI–polyacid complex, we have already shown that the electrical conductivity of its thin films can be increased upon the introduction of a small amount of graphene. In the present study, we report a detailed investigation of the preparation of nanocomposites based on graphene of different oxidation degree and different PANI–polyacid complexes. Polyacids with various hydrophilicity of the side chain were employed that allowed us to regulate the interactions between nanocomposite components. Moreover, the nanocomposite formulation was optimized by using graphene of the different

90

O.D. Iakobson et al. / Synthetic Metals 211 (2016) 89–98

Chart 1. The monomer units of polyacids.

oxidation degree. Two different procedures for the preparation of PANI–polyacid/graphene nanocomposite were developed. Ultraviolet-visible and Fourier transform infrared-spectroscopes (UV– vis and FTIR-spectroscopes, respectively) as well as atomic-force microscopy (AFM) and high-resolution transmission electron microscopy (HR TEM) were used to elucidate the improvement of the electrical conductivity of the nanocomposite thin films. Such complex approach in the study of the graphene/conductive polymer nanocomposites was used for the first time. 2. Experimental The aniline polymerization was carried out in the presence of poly(2-acrylamido-2-methyl-1-propanesulfonic acid) (PAMPSA) or poly(4-styrenesulfonic acid) (PSSA) (Chart 1) with ammonium persulfate as an oxidizer. The details of the synthesis is described in [11]. Aniline was preliminary distilled; ammonium persulfate was used without additional purification. PAMPSA (Mw
2,000,000, 15 wt.% in H2O, Aldrich) and PSSA (Mw
75,000, 30 wt.% in H2O, Aldrich) were used as purchased. The molar ratio of aniline to oxidizer was 1:1 mol/mol, the ratio of aniline to a sulfonic group of polyacid was always kept at 1:2 mol/g-eq. sulfonic groups. Aniline concentration in all cases was equal to 0.003 M. pH-measurement was made by an OP-208/1 p=-meter (Radelkis), the accuracy of the measurement was 0.05.

For nanocomposite preparation the hydrophobic unoxidized graphene, G, and hydrophilic partially oxidized graphene, poG, were used. Graphenes of both types were obtained by mechanochemical procedure as described in Ref. [20] and Ref. [21], respectively. The detailed material characterization was made before [20,21]. The graphene content in nanocomposite was varied from 0.1 to 30 wt.% based on PANI–polyacid complex. Two procedures were applied for PANI–graphene nanocomposite preparation. The procedure I included graphene dispersion addition into the polymerization medium before PANI synthesis; procedure II consisted of the mixing obtained PANI–polyacid and graphene dispersions (Scheme 1). PoG was used in form of aqueous dispersion, G was dispersed in ethanol. Totally, 8 types of PANI– polyacid/graphene nanocomposites were obtained using two methods of nanocomposite preparation. After polymerization the obtained PANI–polyacid or PANI– polyacid/graphene dispersions were dialyzed against water (cellulose membrane ZelluTrans (Roth), MWCO 8000–10000). The electron spectra recording during PANI–polyacid complex synthesis without or with graphene and the final spectra of the nanocomposites were registered by a diode scanning spectrophotometer “AvaSpec 2048”. The thin nanocomposite films were obtained by drop-casting on pre-prepared glass support with following drying on air. The thickness of the films, t, was determined by KLA-Tencor D100 Profiler. The measured value was in the range of 50–60 nm. The DC-conductivity was measured by four-probe technique as described earlier [18,19]. Measurements of the electrical conductivity of the nanocomposites were carried out on different samples several times. The conductivity measurement error did not exceed 5%. The films of graphene–polyacid mixtures on silicon substrates for FTIR-studies were dried at a room temperature. FTIR-spectra were registered on an EQUINOX 55 FTIR spectrometer (Bruker) using the OPUS/IR program. The surface morphology of the nanocomposite films (AFM) was recorded using an Enviroscope scanning probe microscope (Bruker). HR TEM images of G and poG stacks in the nanocomposites with PANI–polyacid complexes were obtained using a Cs-corrected JEM2100 F transmission electron microscope operating at 200 kV. All the experiments were reproduced twice at least with reliable results obtained.

Scheme 1. Schematic presentation of applied procedures for PANI–polyacid/graphene nanocomposite preparation.

O.D. Iakobson et al. / Synthetic Metals 211 (2016) 89–98

D (a.u.)

1.0

G dispersion poG dispersion

271 nm 261 nm

0.8

0.6

0.8 0.6

0.4

0.4

D (a.u.)

1.2

0.2 0.0 200

250

300

350

0.2 400

λ (nm)

91

its hydrophobic properties it can be dispersed in other types of solvents as NMP, DMF, and EtOH [20]. The graphene used is characterized by different maximum position of the UV-absorption spectra that is clearly seen in Fig. 1. While G has the maximum at about 271 nm attributed to the overlapped graphene p–p* transition [22], the maximum of poG is shifted toward shorter wavelengths and locates at around 261 nm. These results prove that G is unoxidized, whereas poG is slightly oxidized in comparison with graphene oxide that possesses characteristic peak located at 230 nm [20,21]. The above described differences between G and poG properties are expected to influence the physico-chemical properties of its nanocomposites with conductive polymers. 3.2. Polymeric acids

Fig. 1. UV-absorption spectra of graphene dispersions: unoxidized graphene (G) in ethanol (0.03 mg mL1) and partially oxidized graphene (poG) in water (0.02 mg mL1).

3. Results and discussion 3.1. Nanostacks of graphene A novel mechanochemical procedure without using aggressive concentrated acids and multiple complicated steps has been used for graphene preparation. By careful choosing the reaction condition in this procedure it is possible to obtain unoxidized graphene (G) and graphene with a low oxidation degree [20,21]. As was shown earlier [21], the partially oxidized graphene (poG) has C¼C, C O, C¼O bonds corresponding to different functional groups such as carbonyl, hydroxyl and epoxy groups. The variation of graphene oxidation degree has been performed with the aim of providing new properties to graphene (for example, waterdispersancy) and modifying the existent ones [20,21]. It is the presence of functional groups that defines the differences of the properties between both graphene used. For instance, poG easily forms stable aqueous dispersions in contrast to hydrophobic G. It should be noted that actual nature of poG is likely to be amphiphilic because oxygen-containing groups are situated mostly on the edges of graphene oxide [7] or at least limited by isolated regions [1] along with the presence of large hydrophobic basal planes. The functional groups of graphene oxide are likely to prevent the stacking of nanolayers due to their electrostatic repulsive forces [7,9]. On the contrary, graphene without functionality forms large aggregates due to strong hydrophobic and p–p interactions during storage [9]. However, for the dispersions used in this study the initial characteristic sizes of both graphenes are
1 nm thick and
200 nm long [20,21] that is they form nanostacks of 2– 3 layers. It is necessary to note, that poG still possesses electrical conductivity that is an advantage over isolating graphene oxide in the application for conductive nanocomposite preparation. The high electrical conductivity is characteristic of G also, however, for

The majority of works is devoted to the investigation of nanocomposites based on “ordinary” PANI and graphene that is bicomponent systems are considered. In this work we use PANI– polyacid complex for nanocomposite preparation with graphene aiming to have additional constituent part, namely, polymeric acid, for regulation of nanocomposite properties. Two different polyacids distinguished in hydrophilicity of their side chain has been chosen for the research (Chart 1). Despite electrical structure of PANI obtained in the presence of these polyacids demonstrates similarity [11], the different types of interaction between polymeric acid and small molecules or ions is possible due to different the hydrophobic–hydrophilic nature of the polyacid [23]. It can be also applicable to different particles including graphene with various oxidation degrees. It is seen that PAMPSA is more hydrophilic due to the presence of the amide groups in its structure compared to PSSA containing hydrophobic phenylene groups. Since graphene used also possesses different hydrophobic/hydrophilic nature, it is highly necessary to compare the PANI complexes with these two polyacids with the aim of elucidation of the interaction between the nanocomposite components. 3.3. Electrical conductivity of nanocomposite thin films In the study, two procedures are employed for the preparation of the PANI–polyacid/graphene nanocomposites. The procedure I consists of the graphene dispersion addition into polymerization medium before PANI synthesis (in-situ aniline polymerization in the presence of polyacid with graphene dispersion addition). This procedure has been shown to be useful for tuning the properties of nanocomposites [9,24] and it has prevailed among other procedures for the preparation of the nanocomposites based on graphene and PANI–polyacid complexes [13–16]. The procedure II, mixing individual components of PANI/ graphene nanocomposites, is not so widely used for “ordinary” PANI due to its insolubility in the majority of solvents [9].

Table 1 The electrical conductivity of PANI–polyacid/graphene nanocomposite films. Nanocomposite preparation

Conductivity [S cm1] PANI–PAMPSA

Procedure I

Procedure II

Before dialysis After dialysis

PANI–PSSA

Initial

+G

+poG

Initial

+G

+poG

2.0  101 1.5  102

2.0  101 3.0  102

5.0  101 2.0  102

9.0  104 3.0  104

5.0  104 3.0  104

8.0  104 4.0  104

1.5  102

3.0  102

2.5  101

3.0  104

3.0  104

4.0  104

92

O.D. Iakobson et al. / Synthetic Metals 211 (2016) 89–98

This procedure is applicable to water-dispersible systems (i.e., PANI–polyacid complexes) and, therefore, provides more technological approach. The procedure II is also used in the present work. Firstly, 1 wt.% of graphene content in the nanocomposite with PANI–polyacid complex has been used for comparison with [18,19]. Electrical conductivity of thin films of all nanocomposites prepared by the both procedures is shown in Table 1. For the nanocomposites prepared by procedure I, the electrical conductivity decreases after the purification by dialysis and the final conductivity demonstrates slight (if any) increase compared to conductivity of the original PANI–polyacid complex free from graphene. However, co-products present in unpurified product (the traces of unreacted monomer, oxidizer, low-molecular-weight by-products, etc.) influence negatively the stability of nanocomposite properties. The procedure II yields nanocomposite films with enhanced conductivity compared to initial PANI–polyacid complex (Table 1). Moreover, the nanocomposite of poG with PANI–PAMPSA possesses the conductivity that is only slightly lower than that in the nanocomposite obtained by procedure I before dialysis, yet, it is completely free from undesirable co-products. The graphene nanocomposites with PANI–PSSA demonstrate equal values of conductivity for procedure I (after dialysis) and procedure II (Table 1). This fact may result from nonoptimal graphene content; therefore, we have resumed the research of the dependence of electrical properties on the content of graphene for nanocomposites obtained by procedure II (Fig. 2a and b).

It is seen clearly that the dependence of electrical conductivity on the graphene concentration is significantly different for PANI– PAMPSA nanocomposites with G or poG (Fig. 2a). Even small addition of poG (
0.1 wt.%) into PANI–PAMPSA complex leads already to 9-times increase in the conductivity. The further gain in conductivity with increase of poG concentration (up to 1 wt.%) is followed by the small decay to the value which remains constant up to 20 wt.% of the graphene concentration (not shown). On the other hand, in the case of G addition into PANI–PAMPSA complex the increase of conductivity does not exceed 3-times in a wide range of graphene content. Thus, the degree of conductivity increase of PANI–PAMPSA/graphene nanocomposite film depends on graphene oxidation degree exceeding more than one order (about 20 times) in case of poG. An additional advantage of the poG is that it forms an aqueous dispersion; this allows one to avoid utilization of harmful and volatile organic solvents and meet the requirements of environmental safety. The influence of the oxidation degree of graphene on the electrical conductivity of the PANI–PSSA/graphene nanocomposites differs markedly from that of the PANI–PAMPSA nanocomposites (Fig. 2b). Significant increase of the conductivity (16-times maximum) is seen in the case of PANI–PSSA/G nanocomposite, whereas PANI–PSSA/poG nanocomposite demonstrates 2.5-times increase only. Moreover, in both cases this influence is seen only at comparatively high content of graphene (more than 10 wt.%). In [25] an additional doping of PANI by carboxylic acid groups of oxidized graphene was mentioned among the reasons of the increase of conductivity of PANI/graphene oxide composite. In such

Fig. 2. The influence of graphene content in nanocomposite film on the electrical conductivity (s , S/cm) and the shift of localized polarons absorbance maxima position of visspectra (D l, nm) for PANI–PAMPSA (a and c) or PANI–PSSA (b and d) complexes.

O.D. Iakobson et al. / Synthetic Metals 211 (2016) 89–98 Table 2 pH of the reaction medium before PANI synthesis. Medium

Aniline + PAMPSA solution

Aniline + PSSA solution

Free from graphene G addition poG addition

3.0 3.0 3.7

3.1 3.1 3.1

a case we would observe the similar influence of graphene oxidation degree for both PANI–PAMPSA and PANI–PSSA. Thus, the mentioned reason [25] does not account for our data. From Fig. 2 a clear correlation between the increasing conductivity and bathochromic shift of the peak corresponding to localized polarons absorption (
750 nm) [26] in comparison to position of this peak in the PANI complex free from graphene is revealed. Similar to the conductivity change, the most significant bathochromic shift is exhibited by PANI–PAMPSA/poG and PANI– PSSA/G nanocomposites. On the other hand, only minor shift of the peak position, if any, is observed for the other two nanocomposites. At the same time, no increase of the absorption in near IR-range (>900 nm) for nanocomposite films (attributed to absorption of delocalized polarons [26]) is found. The observed shift of localized polarons absorption may be caused by partial charge delocalization over PANI macromolecules and graphene nanostacks. The delocalization reflects the formation of new energy levels within band gap of PANI that facilitates interchain charge transfer. These energy states serve as additional hole transport sites in nanocomposite enhancing the conductivity. 3.4. Reasons for the conductivity change 3.4.1. Aniline, polyacid, and graphene interactions (pH-measurement and FTIR-spectroscopy) The reason of influence of graphene with various degree of oxidation on optical and electrical properties of the nanocomposite may originate from the interaction between graphene and PANI– polyacid complex [18,19]. Indeed, despite that the procedure I of

(a)

93

the nanocomposite preparation yields unstable product, it allows us to elucidate the nature of the interactions between nanocomposite components because a component addition proceeds consistently. First of all, before aniline polymerization the pH of reaction medium (aniline, polyacid with or without graphene) has been measured. The data are presented in Table 2. It is seen that there is no influence on the pH of the solution containing aniline and PAMPSA after G addition. The same situation is observed for poG (or G) and aniline + PSSA medium. It means that graphene has no effect on the sites of localization of aniline in the vicinity of polyacid. On the contrary, in the case of poG added into aniline + PAMPSA containing solution the pH rises significantly. It may result from the competition between poG and aniline for the interaction with the polyacid that causes aniline to leave the sites of the polyacid. The accepted conception supposes that aniline after its protonation gains hydrophilic properties and can be easily dissolved in aqueous solution. Therefore, it shows that poG tends to interact with hydrophilic parts of polyacid (sulfonic acid groups or amide groups). However, the absence of competition between poG and aniline in PSSA solution suggests that amide groups seem to interact with poG. In order to confirm the hypothesis about the interaction between graphene of different oxidation degree and polyacid of different nature, we have performed the analysis of FTIR-spectra of polyacid before and after addition of graphene in the region of absorption of functional groups (Fig. 3). In Fig. 3a, one can see that after introduction of graphene of both types in PAMPSA there are some shifts in band position and changes in the absorption band intensity compared to pristine PAMPSA. Both in the case of G and poG addition, one can see broadening the hydrogen-bonded NH stretching band centered at around 3300 cm1 with respect to its height that can be a sign of decrease of the concentration of amide–amide hydrogen bonds in PAMPSA [27]. However, this process is not accompanied by the appearance of “free” NHgroup at around 3450 cm1. After the

(b)

Fig. 3. FTIR spectra of PAMPSA (a) or PSSA (b) without/with G or poG added.

94

O.D. Iakobson et al. / Synthetic Metals 211 (2016) 89–98

(a)

(b)

2.5

PANI-PAMPSA PANI-PAMPSA/G PANI-PAMPSA/poG

2

1.5

D (a.u).

D (a.u.)

2

1 0.5 0

0

150

300

450

600

PANI-PSSA PANI-PSSA/G PANI-PSSA/poG

2.5

750

900

t (s)

1.5 1 0.5 0

0

150

300

450

600

750

900

t (s)

Fig. 4. Kinetics of PANI absorption raise at 750 nm during the course of the PANI synthesis in the presence of PAMPSA (a) or PSSA (b). The wavelength corresponds to localized polarons absorption band [24].

destruction of these hydrogen bonds the amide groups of polyacid seem to take part in other hydrogen-bonding interaction, possibly with sulfonic groups. Such interaction may be the reason of the shift of position of asymmetric stretching of sulfogroups from 1224 cm1 (PAMPSA) to 1214 cm1 (PAMPSA and G or poG). The fact that the changes in position of sulfogroups are observed both upon addition of G (without functionality) and poG (contains functional groups on the surface) favors to the opinion that it is connected with the change in the inherent interaction between or inside PAMPSA chains due to interaction with graphene. Therefore, after the addition of graphene (both G and poG) the inside bonds of PAMPSA seem to rearrange from amide–amide interaction into amide-sulfonic groups interaction. After poG addition to polyacid, the position of nonbonded amide groups band (C¼O stretching (amide I) 1731 cm1) shifts to lower frequency (1720 cm1) suggesting the direct interaction between PAMPSA amide groups and poG. Therefore, the amide– amide bonds are destroyed. Released amide groups and amide groups with poG attached interact with sulfonic groups of polyacid, the latter interaction causes aniline to leave its initial sites. On the contrary, the position of the amide I band remains unchanged in the case of G addition due to none of interaction between them in the absence of functional groups on the surface of this graphene. Moreover, in this case, one can see the tendency to the hydrophobic interaction between G and more hydrophobic parts of PAMPSA that is proved by the shift of H3CC(CH2)CH3 (2938 cm1 and 2854 cm1 in pristine PAMPSA toward to 2917 cm1 and 2849 cm1 in nanocomposite with G) and C N  bands (1302 cm1). Taking place in the vicinity of amide groups this interaction seems to break the amide–amide bonds; as a result, however, it has no influence on the aniline location. FTIR-spectroscopy has been also used to elucidate the specific interactions between PSSA and graphene of different oxidation degree (Fig. 4b). No significant absorbance is found for PSSA above 1700 cm1 due to the absence of amide groups in its structure. In comparison to PAMPSA, sulfogroups of PSSA are more hydrated (symmetric  SO3 stretching at 1036 cm1 for PSSA vs. 1041 cm1 for PAMPSA) [28]. It may be a result of proposed in [29] spatial conformation where more hydrophobic phenylene groups in PSSA tend to cluster together and force charged sulfonic groups to stay on the coil shell. In such a case, sulfonic groups more easily undergo hydration, so PSSA obtains a significant hydrophobic core. Hence, hydrophobic G sheets tend to penetrate into the core

followed by aggregation as nanostacks. The interaction of graphene with phenylene groups is proved by the shift of the initial position of the band corresponding to aromatic C¼C stretching vibrations from 1452 cm1 to 1446 cm1. PoG can also localize inside PSSA core because of its amphiphilic nature. This is reflected in the similar changes of FTIR-spectra for both G and poG. Absence of the interactions and any other influence on sulfonic groups of PSSA (unchanged symmetric  SO3 stretching band position— 1036 cm1) explains unchanged pH after the G or poG addition to the reaction medium of aniline+PSSA. 3.4.2. PANI and graphene interactions (UV–vis-absorption spectroscopy) The rate of rise of the PANI absorption on the characteristic wavelength during aniline polymerization in the presence of any graphene slows down as compared to that of aniline polymerization in the absence of graphene (Fig. 4). For aniline polymerization in PAMPSA in the presence of poG, the PANI synthesis deceleration may result from the decrease of aniline local concentration in the vicinity of polyacid due to its displacement by graphene shown above (Table 2). This case is also accompanied by product yield decrease (based on adsorption decrease on the PANI characteristic wavelength). The deceleration also occurs in the aniline + PAMPSA reaction medium containing G in which, aniline state remains unchanged after the graphene addition according to pH measurement (Table 2). It might be explained by the interaction between G and growing PANI chains that hinders aniline polymerization. Actually, various interactions (including electrostatic, p–p stacking, and hydrogen bonding) between graphene sheets and growing PANI backbone are well-known [9]. Interactions are expected to proceed in the same way for PANI complexes obtained both in the presence of PAMPSA and PSSA due to similar electronic structure of PANIs [11]. It forces the PANI synthesis to slow down in the case of aniline + PSSA reaction medium containing G as well. Because of poG location inside hydrophobic PSSA core, poG has no influence on the PANI synthesis rate in PSSA presence. This fact is caused by both (i) unchanged aniline location and concentration and (ii) the weak interaction between poG and growing PANI chains. 3.4.3. Graphene–graphene interactions (AFM and HR TEM) The analysis of AFM- and HR TEM-images clearly shows that the revealed interactions between graphene of different oxidation

O.D. Iakobson et al. / Synthetic Metals 211 (2016) 89–98

95

Fig. 5. AFM-images and image profiles (insert) of the nanocomposite films: PANI–PAMPSA (a), PANI–PAMPSA with G (b) or poG (c); PANI–PSSA (d), PANI–PSSA with G (e) or poG (f). H = object height, s = object planar size. For the sake of comparison, the ordinate scales are the same for all image profiles.

96

O.D. Iakobson et al. / Synthetic Metals 211 (2016) 89–98

degree and PANI or polyacid lead to different graphene distribution on the surface and in the bulk of the nanocomposite film and different size of stacked graphene particles. Indeed, the surface topography of PANI–polyacid complexes is rather smooth with uniform distribution of hills of the height of less than 2 nm (PANI–PSSA) or 5 nm (PANI–PAMPSA) (Fig. 5a and d). For the nanocomposite of PANI–PAMPSA with G, irregularly and rarely distributed large inclusions with a characteristic height of 40–70 nm are noticed on the surface of the film (Fig. 5b). These seem to be stacks of G. On the contrary, smaller objects with the size of less than 20 nm in height and about 100 nm in size are seen on the surface of the PANI–PAMPSA/poG nanocomposite layer (Fig. 5c). The image confirms the uniform distribution of poG in the PANI–PAMPSA film with the slight nanostack formation caused by the interaction between nanocomposite components. As for PANI/PSSA nanocomposite with both G and poG, large objects of 200–400 nm in size are observed (Fig. 5e). However, G and poG exhibit different height and distribution. In the PANI– PSSA/G film, the objects are result of p–p stacking of G. More or less uniform distribution of the G stacks results from their interaction with both PANI and PSSA backbones. Since poG demonstrates a minor tendency to stack formation, smaller objects are seen on the film of the PANI–PSSA/poG nanocomposite (Fig. 5f). It is worth noting additionally that PANI–polyacid/poG nanocomposite films (both for PANI–PAMPSA and PANI–PSSA) show smoother surface in comparison with those of the PANI–polyacid complex doped with G. As shown by HR TEM (Fig. 6), the number of the graphene sheet layers in a nanostack is different for G and poG which are introduced into either PANI–PAMPSA or PANI–PSSA matrix. In the PANI–PAMPSA/G nanocomposite, G forms large stacks consisting of more than 30 layers (Fig. 6a). Indeed, it is known that graphene without functionalities tends to p–p interaction with the formation of graphene stacks [9]. On the contrary, the average number of poG layers in a nanostack formed in PANI–PAMPSA/poG nanocomposite is much lower: 4–6 layers. Moreover, the poG nanostacks are distributed uniformly in the bulk of the film (Fig. 6b). It is worth emphasizing that both for G and poG the interlayer distance determined from the HR TEM-images is identical and equal to 3.4 Å, whereas for graphene oxide nanostacks the

interlayer distance was published to increase compared with neat graphene nanostacks [30]. In the graphenes under the study, the same distance is explained by low degree of oxidation of poG: large graphene basal planes in it neighbor with small areas of functional groups. Hence, poG nanostacks mainly are formed through regions without functionalities so no difference with G stacks structure is observed. However, the functional groups of poG are seen to prevent large stack formation even at their small content. In the bulk of the PANI–PSSA/graphene nanocomposite films, the uniform distributed large stacks consisting of about 10–15 layers of G are found, whereas the smaller nanostacks of 6–8 poG layers are localized as isolated aggregates (images are not presented). Hence, in the bulk of nanocomposite film the concentration and distribution of graphene nanostacks differ due to different capability to the nanostack formation. This capability depends on (i) the internal interactions between graphene planes revealed by HR TEM and AFM and (ii) the external interactions between graphene and PANI or polyacid of different hydrophobicity proved by UV–vis, FTIR-spectroscopies and pH-measurements. Entirely, all these factors underlay in the different influence of G and poG on electrical conductivity of the nanocomposite films. Due to the interaction between PANI chains and graphene additional charge-transport paths are formed, and, as a result, the electrical conductivity rises. The special measurements on the polyacid/graphene layer (without PANI) have been carried out and no conductivity for such systems has been found. This fact proves that the conductivity of PANI–polyacid/graphene nanocomposite rises due to increased number of elongated charge-transport paths formed by both PANI and graphene together but merely graphene. It means that percolation conductivity over graphene sheets is not responsible for the enhanced conductivity in PANI–polyacid/ graphene nanocomposite. Since in the PANI–PAMPSA matrix the G sheets form large stacks the number of new charge-transport paths obviously increases slightly and minor raise of conductivity is observed. On the contrary, in the PANI–PSSA matrix the G sheets interact with both PSSA and PANI so uniform distributed stacks of moderate size are formed. As a result, the electrical conductivity rises greater compared to PANI–PAMPSA/G nanocomposite. In contrast to G, functionalized poG exhibits even more uniform distribution in the PANI–PAMPSA matrix with the thinnest

Fig. 6. HR TEM-images of the nanocomposite films: PANI–PAMPSA/G (a), PANI–PAMPSA/poG (b).

O.D. Iakobson et al. / Synthetic Metals 211 (2016) 89–98

nanostacks observed in the work. Thus, in PANI–PAMPSA/poG nanocomposite charge-transport paths create continuous chargetransport network which provides the highest conductivity of the material. As for PANI–PSSA matrix, poG is localized mostly in the hydrophobic PSSA core with no interaction with PANI so the conductivity increase actually is negligible. 4. Conclusions Summarizing, a procedure for the preparation of nanocomposites based on PANI–polyacid complex and graphene is developed. It enables to tune the electrical conductivity of the nanocomposite films. The conductivity is found to depend on graphene oxidation degree which, in turn, defines the interactions between both nanocomposite components (graphene, PANI, polyacid) and graphene sheets themselves. The interactions are responsible for the formation of graphene nanostacks, their size and distribution in PANI–polyacid matrix. In the PANI–PSSA complex, both unoxidized G and partially oxidized poG are likely to accumulate in hydrophobic core of the polyacid. Moreover, G interacts with PANI yielding the increased conductivity of the nanocomposite film. Yet, the increase is observed at the comparatively high content of G due to significant stacking of graphene. On the contrary, the lack of the interaction between poG and PANI results in minor increase of the electrical conductivity of the PANI–PSSA/poG nanocomposite. In the PANI–PAMPSA complex, the introduction of G also leads to increase of the conductivity. The interactions of G with both PANI and polyacid are responsible for more or less uniform distribution of G stacks similar to the G distribution in PANI–PSSA. Such distribution underlies in the formation of additional chargetransport paths in the nanocomposite providing somewhat increase in conductivity. Functional groups of poG ensure (i) less pronounced stacking of poG and (ii) interaction between all the components of the PANI– PAMPSA/poG nanocomposite. The both lead to the most uniform distribution of the thinnest graphene nanostacks in the bulk of PANI–PAMPSA/poG among all nanocomposites under study. This nanocomposite peculiarity is considered to promote the formation of the continuous charge-transport network which provides the conductivity increase of about 20 times compared to pristine PANI–PAMPSA. Acknowledgements The authors thank Dr. V.I. Zolotarevskii (Frumkin Institute of Physical Chemistry and Electrochemistry, Russian Academy of Sciences, Russia) for AFM measurements and discussing the obtained results and Dr. O.Yu. Posudievsky (Pisarzhevskii Institute of Physical Chemistry, National Academy of Sciences, Ukraine) for providing samples of graphene. The financial support from the Russian Science Foundation (grant No. 15-13-00170) for optical, electrical, morphological investigations and the Russian Foundation for Basic Research (grant No. 14-03-90413-Ukr_a) is also acknowledged. O.O. is grateful to the Grant Council of the President of the Russian Federation for financial support (SP-2994.2015.1). References [1] H. Kim, A.A. Abdala, C.W. Macosko, Graphene/polymer nanocomposites, Macromolecules 43 (2010) 6515–6530, doi:http://dx.doi.org/10.1021/ ma100572e. [2] A.K. Geim, K.S. Novoselov, The rise of graphene, Nat. Mater. 6 (2007) 183–191, doi:http://dx.doi.org/10.1038/nmat1849.

97

[3] K.K. Sadasivuni, D. Ponnamma, J. Kim, T. Sabu, Graphene-Based Polymer Nanocomposites in Electronics, Springer International Publishing, Cham, 2015, doi:http://dx.doi.org/10.1007/978-3-319-13875-6. [4] S.B. Lee, S.M. Lee, N. Il Park, S. Lee, D. Chung, Preparation and characterization of conducting polymer nanocomposite with partially reduced graphene oxide, Synth. Met. 201 (2015) 61–66, doi:http://dx.doi.org/10.1016/j. synthmet.2015.01.021. [5] S. Khamlich, F. Barzegar, Z.Y. Nuru, J.K. Dangbegnon, A. Bello, B.D. Ngom, et al., Polypyrrole/graphene nanocomposite: high conductivity and low percolation threshold, Synth. Met. 198 (2014) 101–106, doi:http://dx.doi.org/10.1016/j. synthmet.2014.10.004.   [6] G. Ciri c-Marjanovi c, Recent advances in polyaniline composites with metals, metalloids and nonmetals, Synth. Met. 170 (2013) 31–56, doi:http://dx.doi. org/10.1016/j.synthmet.2013.02.028. [7] M. Acik, Y.J. Chabal, Nature of graphene edges: a review, Jpn. J. Appl. Phys. 50 (2011) 070101, doi:http://dx.doi.org/10.1143/JJAP.50.070101. [8] B. Abad, I. Alda, P. Díaz-Chao, H. Kawakami, A. Almarza, D. Amantia, et al., Improved power factor of polyaniline nanocomposites with exfoliated graphene nanoplatelets (GNPs), J. Mater. Chem. A 1 (2013) 10450, doi:http:// dx.doi.org/10.1039/c3ta12105d. [9] Y. Sun, G. Shi, Graphene/polymer composites for energy applications, J. Polym. Sci. B Polym. Phys. 51 (2013) 231–253, doi:http://dx.doi.org/10.1002/ polb.23226. [10] Q. Wu, Y. Xu, Z. Yao, A. Liu, G. Shi, Supercapacitors based on flexible graphene/ polyaniline nanofiber composite films, ACS Nano 4 (2010) 1963–1970, doi: http://dx.doi.org/10.1021/nn1000035. [11] O.L. Gribkova, A.A. Nekrasov, M. Trchova, V.F. Ivanov, V.I. Sazikov, A.B. Razova, et al., Chemical synthesis of polyaniline in the presence of poly(amidosulfonic acids) with different rigidity of the polymer chain, Polymer (Guildford) 52 (2011) 2474–2484, doi:http://dx.doi.org/10.1016/j.polymer.2011.04.003. [12] V.F. Ivanov, O.L. Gribkova, K.V. Cheberyako, A.A. Nekrasov, V.A. Tverskoi, A.V. Vannikov, Template synthesis of polyaniline in the presence of poly-(2acrylamido-2-methyl-1-propanesulfonic acid), Russ. J. Electrochem. 40 (2004) 299–304 10.1023/B:RUEL.0000019668.68527.cc. [13] C. Basavaraja, G.T. Noh, D.S. Huh, Chemically modified polyaniline nanocomposites by poly(2-acrylamido-2-methyl-1-propanesulfonicacid)/ graphene nanoplatelet, Colloid Polym. Sci. 291 (2013) 2755–2763, doi:http:// dx.doi.org/10.1007/s00396-013-3016-8. [14] J. Luo, S. Jiang, Y. Wu, M. Chen, X. Liu, Synthesis of stable aqueous dispersion of graphene/polyaniline composite mediated by polystyrene sulfonic acid, J. Polym. Sci. A Polym. Chem. 50 (2012) 4888–4894, doi:http://dx.doi.org/ 10.1002/pola.26316. [15] J. Luo, S. Jiang, R. Liu, Y. Zhang, X. Liu, Synthesis of water dispersible polyaniline/poly(styrenesulfonic acid) modified graphene composite and its electrochemical properties, Electrochim. Acta 96 (2013) 103–109, doi:http:// dx.doi.org/10.1016/j.electacta.2013.02.072. [16] L. Wan, B. Wang, S. Wang, X. Wang, Z. Guo, H. Xiong, et al., Water-soluble polyaniline/graphene prepared by in situ polymerization in graphene dispersions and use as counter-electrode materials for dye-sensitized solar cells, React. Funct. Polym. 79 (2014) 47–53, doi:http://dx.doi.org/10.1016/j. reactfunctpolym.2014.03.012. [17] S. Cho, J.S. Lee, J. Jun, S.G. Kim, J. Jang, Fabrication of water-dispersible and highly conductive PSS-doped PANI/graphene nanocomposites using a highmolecular weight PSS dopant and their application in H2S detection, Nanoscale 6 (2014) 15181–15195, doi:http://dx.doi.org/10.1039/c4nr04413d. [18] O.D. Omelchenko, O.L. Gribkova, A.R. Tameev, S.V. Novikov, A.V. Vannikov, Thin nanocomposite layers based on a complex of polyaniline and graphene, Prot. Met. Phys. Chem. Surf. 50 (2014) 613–619, doi:http://dx.doi.org/10.1134/ s207020511405013x. [19] O.D. Omelchenko, O.L. Gribkova, A.R. Tameev, A.V. Vannikov, The effect of the degree of graphene oxidation on the electric conductivity of nanocomposites based on a polyaniline complex, Tech. Phys. Lett. 40 (2014) 807–809, doi: http://dx.doi.org/10.1134/s1063785014090260. [20] O.Y. Posudievsky, O. Khazieieva, V.V. Cherepanov, V.G. Koshechko, V.D. Pokhodenko, High yield of graphene by dispersant-free liquid exfoliation of mechanochemically delaminated graphite, J. Nanopart. Res. 15 (2013) 1–9, doi: http://dx.doi.org/10.1007/s11051-013-2046-y (2046). [21] O.Y. Posudievsky, O. Khazieieva, V.G. Koshechko, V.D. Pokhodenko, Preparation of graphene oxide by solvent-free mechanochemical oxidation of graphite, J. Mater. Chem. 22 (2012) 12465–12467, doi:http://dx.doi.org/10.1039/ c2jm16073k. [22] D. Wang, X. Wang, Self-assembled graphene/azo polyelectrolyte multilayer film and its application in electrochemical energy storage device, Langmuir 27 (2011) 2007–2013, doi:http://dx.doi.org/10.1021/la1044128. [23] J.H. Hwang, S.C. Yang, Morphological modification of polyaniline using polyelectrolyte template molecules, Synth. Met. 29 (1989) 271–276, doi:http:// dx.doi.org/10.1016/0379-6779(89)90306-8. [24] J. Xu, K. Wang, S.Z. Zu, B.H. Han, Z. Wei, Hierarchical nanocomposites of polyaniline nanowire arrays on graphene oxide sheets with synergistic effect for energy storage, ACS Nano 4 (2010) 5019–5026, doi:http://dx.doi.org/ 10.1021/nn1006539. [25] S. Bae, J.U. Lee, H. Park, E.H. Jung, J.W. Jung, W.H. Jo, Enhanced performance of polymer solar cells with PSSA-g-PANI/Graphene oxide composite as hole transport layer, Sol. Energy Mater. Sol. Cells 130 (2014) 599–604, doi:http://dx. doi.org/10.1016/j.solmat.2014.08.006.

98

O.D. Iakobson et al. / Synthetic Metals 211 (2016) 89–98

[26] A.G. MacDiarmid, A.J. Epstein, Secondary doping in polyaniline, Synth. Met. 69 (1995) 85–92, doi:http://dx.doi.org/10.1016/0379-6779(94)02374-8. [27] X. Lu, R.A. Weiss, Development of miscible blends of polyamide-6 and manganese sulfonated polystyrene using specific interactions, Macromolecules 24 (1991) 4381–4385, doi:http://dx.doi.org/10.1021/ ma00015a021. [28] X.Y. Lu, R.A. Weiss, Specific interactions and ionic aggregation in miscible blends of nylon-6 and zinc sulfonated polystyrene ionomer, Macromolecules 25 (1992) 6185–6189, doi:http://dx.doi.org/10.1021/ma00049a015.

[29] J.S. Tan, P.R. Marcus, Ion binding in sulfonate-containing polyelectrolytes, J. Polym. Sci. 14 (1976) 239–250. [30] X. Díez-Betriu, F.J. Mompeán, C. Munuera, J. Rubio-Zuazo, R. Menéndez, G.R. Castro, et al., Graphene-oxide stacking and defects in few-layer films: impact of thermal and chemical reduction, Carbon N. Y. 80 (2014) 40–49, doi:http:// dx.doi.org/10.1016/j.carbon.2014.08.016.