Introduction

Research related to the use of HCOOH (formic acid (FA)) has increased significantly in recent years due to its high-efficiency power conversion, easy transportation, and storage, and because it offers a clean and sustainable alternative [1, 2]. Thus, the FA electro-oxidation using different metals at macroscopic and nanometric level promoted the early research in this field [3]. As a result of its high activity, platinum is the most used electro-catalyst anode [4, 5], however, nowadays its use has decreased because, in addition to high production costs and low durability [4, 6], it is rapidly poisoned during the oxidation processes.

The FA electro-oxidation processes to obtain CO2 require the breaking of two FA molecule bonds, namely OH and CH. Breaking the OH fragment is relatively easy, owing to the acid-base equilibrium inherent to FA, but the second breakdown requires high affinity and specificity of the electrode used as anode, favoring hydrogenation and subsequent dehydration of the electrode, being platinum, as already mentioned, the one with the highest efficiency. By this route, when metal-based electrodes are used, the formation of CO(ads) at low potential can be observed; this interaction is very strong and difficult to break, leading ultimately to CO2 that requires the transfer of an OH group in the process. This path is known as the direct oxidation pathway whose reaction mechanism is not yet fully elucidated [7]:

$$ \begin{array}{c}\hfill \mathrm{HCOOH}\to {CO}_2+2{\mathrm{H}}^{+}+2\mathrm{e}-\hfill \\ {}\hfill \mathrm{HCOOH}\to {CO}_{\mathrm{ads}}+{\mathrm{H}}_2\to {CO}_2+2{\mathrm{H}}^{+}+2\mathrm{e}-\hfill \end{array} $$

As mentioned, one of the main disadvantages in using Pt or other metal-based electrodes is the poisoning as the reaction proceeds, losing its activity as is reused; therefore, it has been sought to conjugate the properties of other systems in such a way that the synergy of the combination of these materials with metal nanoparticles helps to enhance efficiency and reduce platinum poisoning.

Recently, Pt composites with conducting polymers (CPs) [8–11], such as polythiophene [12, 13], polyaniline [14], polypyrrole [15], among others, have been employed to study FA electro-catalysis, based on the advantages that CPs present, e.g., large surface area and high chemical and environmental stability, in addition to the synergistic effect that occurs between CPs and Pt particles [16–19]. New technologies and advances in this area, which allow nanostructuring-forming nanocomposites with templates [8–10], have considerably broadened the already known yield of bulky polymers [8, 11].

In the current paper, electrodes modified with polymeric nanostructures previously reported by us [12, 13] will be tested. The assembly consists of poly(1-amino-9,10-anthraquinone) P1AAQ conducting polymer nanowires (P1AAQnw), electro-synthesized on a poly(1-amino-9,10-anthraquinone-co-o-phenylenediamine) (COP) electrode using stainless steel (SS) as substrate. To this purpose, a SiO2 template (tp) that guides the growth of the nw was electro-deposited on the COP; the tp is subsequently chemically removed [14]. Next, through an electro-chemical method also developed and checked by our group [11], incorporation of dispersed platinum nanoparticles (Ptnp) on these modified electrodes (SS|COP-P1AAQnw-Ptnp) is performed (Fig. 1). The assembly was further evaluated respect to FA electro-oxidation. Hence, if the activity were favored, the costs associated with this type of anodes would be lowered, enabling an important breakthrough for this kind of fuel cells. Another important advantage of this system is the preparation of nanostructures (polymer and Pt) in situ, directly upon the electrode, which ensures reproducibility that is almost impossible to achieve if nanoparticles are synthesized ex situ and subsequently placed over the substrate.

Fig. 1
figure 1

Scheme to obtaining SS|COP-P1AAnw-Ptnw

Full size image

Materials and methods

SS electrode modification with COP (SS|COP) followed by template deposition (SS|COP-Tp)

All solutions were prepared with fresh Milli-Q grade water obtained from a Heal Force (Smart Series) deionizer. The experiments were conducted at room temperature (20 °C) under a high purity argon atmosphere in a three-compartment, three-electrode anchor-type electro-chemical cell. A 0.07-cm2 geometric area steel (AISI 316) disk (SS) was used as working electrode, and a platinum wire coil of large geometric area was the counter electrode. Ag|AgCl in a solution of tetramethylammonium chloride, whose potential matched that of a saturated calomel electrode (SCE), was the reference electrode [15]. The monomers used in the electrosynthesis were 1AAQ (98 %, Aldrich) and o-phenylenediamine (OPD) (99.5 %, Aldrich), which were prepared following the protocols reported elsewhere [16]. The template was prepared in an ethanol (98 %)∕Milli-Q water (50 % v/v) mixture containing 0.05 mol L−1 of potassium nitrate (KNO3, Aldrich) as supporting electrolyte, 0.0034 mol L−1 tetraethyl orthosilicate (TEOS, Aldrich) as precursor of silicon oxide (SiO2), and 0.115 mol L−1 hexadecyltrimethylammonium bromide (CTAB, Aldrich) as surfactant, by applying −1.400 V for 5 s, and then dried 12 h at 130 °C.

P1AAQnw electrosynthesis on SS|COP-Tp [17]

The P1AAQnw-modified electrodes (SS|COP-1AAQnw) have been obtained on SS|COP-Tp-modified electrodes, applying 1100 mV during 40 s to the SS|COP-Tp electrode immersed into 0.01 mol L−1 1AAQ solution; subsequently, they were dried during 12 h at room temperature. Removal of SiO2 template from SS|COP-1AAQnw electrodes was accomplished using 0.1 mol L−1 NaOH (99.0 %, Merck) solution; to this purpose, 10 μL was deposited on each electrode and allowed to stand 10 min. The solution was then neutralized and the electrode washed with 5 % w/w NaHCO3 (99 %, Merck). This extraction procedure is repeated twice for each device.

Ptnp dispersion over the SS|COP-1AAQnw-modified electrode

Electrosynthesis of Ptnp dispersed on P1AAQnw follows the designed and previously reported methodology. Therefore, a 0.001 mol L−1 of chloroplatinic acid (H2PtCl6) + 0.1 mol L−1 potassium hexafluorophosphate (KPF6) supporting electrolyte aqueous solutions was utilized. The dispersion is conducted using what we called 1 platinization cycle (PC) and consists of a methodology based on the p-doping process of conducting polymers: a fixed potential of 0.800 V is applied during 360 s to produce the p-doping from the anion provided by the acid, i.e., PtCl6 2−, and immediately the potential is set at −0.500 V for 10 s to reduce Pt(IV) that will become incorporated into the polymer matrix as Ptnp. Thus, the SS|COP-nw-Pt-modified electrode is obtained.

To assess FA electro-oxidation, solutions in the concentration range 0.1–8.0 mol L−1 FA (99.0 %, Merck) + 0.1 mol L−1 KPF6 (supporting electrolyte) were utilized.

Instruments

All electro-chemical measurements were carried out on a CH Instruments Electro-chemical Workstation. Images from the electrode surface (scanning electron microscope (SEM)) were obtained on an LEO VP1400 SEM and on a Hitachi HF2000-FEG transmission electron microscope (TEM), using carbon-coated nickel grids.

Results and discussion

Optimization of template electro-obtaining on the copolymer

Template preparation on SS|COP by cyclic voltammetry and chronoamperometry

The template electro-chemical preparation was conducted on a COP previously modified SS electrode by adapting methods reported elsewhere [10, 16]. Cyclic voltammograms (CV) (Fig. 2a) were recorded within a 0.300 and −1.600 V window. After drying (130 °C, 12 h), a very homogeneous and temperature resistant white coating was obtained. The successful obtainment of SiO2 upon the copolymer was predictable from AFM morphological characterization [16], wherein the formation of a homogeneous, very uniform, low roughness, and good conductivity deposit was verified.

Fig. 2
figure 2

a Voltammetric profiles during template electro-obtainment on COP; b j/t transient during template electrodeposition on SS-modified copolymer (SS|COP-Tp). Interfaces: SS|0.0034 mol L−1 TEOS, +0.115 mol L−1 CTAB, +0.05 mol L−1 KNO3 in ethanol/Milli-Q water (50 % v/v) mixture. ν = 50 mV s−1

Full size image

Chronoamperometric (potentiostatic method (PM)) study of template electro-preparation on the COP was performed applying potentials from −1.500 to −1.200 V for 10 s. In this case, a homogeneous coating over the copolymer-modified electrode was achieved only at −1.400 V; therefore, this potential was chosen as the most suitable for template electrosynthesis. Thus, a continuous current increase was observed as the potential was applied.

The early signs concerning the optimal time for obtaining a permeable mesoporous film of template, which permits subsequent electrodeposition of P1AAQ in the confined spaces of their pores, were inferred from the recorded transient (Fig. 2b) that showed a slope variation at about 4 s of electrolysis. This current variation can be ascribed to surface area change, since at the early seconds of electro-obtaining, SiO2 growth would have been generated into the free space left by the surfactant that at negative potentials become compacted as micelle. After polymer growth has reached a height that exceeds the micelle height, the polymer starts growing on it and then grows upon itself, generating the observed slope change during the electro-preparation.

Following the proper potential optimization for electro-preparation of suitable SiO2 template using PM, the right time to completely coat the electrode and in addition to have a permeable film was surveyed. Assays were conducted during 4, 5, 7, and 10 s, observing that as time increases, it is visible at a glance that the coating upon the electrode surface is increasingly covering more area over the modified electrode but at the same time becomes less homogeneous. SEM images (Fig. 3) allowed corroborating that at longer times (Figs. 3c, d) the small heterogeneous spheres grow in volume and uniformity over the electrode, until the surfactant guided mesopores are covered, leading to obtaining an insulator electrode.

Fig. 3
figure 3

SEM images of template prepared upon COP by PM at −1.400 V during a 4, b 5, c 7, and d 10 s

Full size image

For 4 s of electrolysis, a great amount of heterogeneous spheres is not observed by SEM, and the electrode coating reproducibility is no good, hence this time was disregarded for template preparation. Instead, 5 s (Fig. 3b) of electro-preparation was selected to work, because the coating over the modified electrode is total, despite what it could be inferred from SEM images.

Permeability, diffusion coefficient, and effective area of the working electrode with the template upon the COP

The permeable film of SiO2 obtained was evaluated by CV of the modified electrodes using the well-defined ferrocene/ferrocenium redox couple. At the same time, this response was utilized to assess the effective area of the working electrode, considering that when nw are electro-synthesized within the template, they remain attached to the copolymer, leaving an effective area that must be recalculated. To this purpose, the Randles-Sevcik Eq. (1) [17, 18] was used:

$$ {I}_{\mathrm{p}}=K\cdotp {n}^{3/2}\cdotp A\cdotp \surd \left(D\cdotp \upnu \right)\cdotp C $$
(1)

where I p is the oxidation peak current, K a constant equal to 2.69 × 105 C mol−1 ν−1/2, n number of electrons involved in the redox reaction, A electrode area (cm2), D diffusion coefficient of the analyte (cm2 s−1), C concentration (mol cm−3), and ν scan rate (V s−1).

A ν study between 5 and 100 mV s−1 (Fig. 4a) revealed that the slope (m) of the plot in Fig. 4b enables recalculating the effective area when the template is present, taking ferrocene diffusion coefficient = 2.60 × 10−9 cm2 s−1 [19]. The calculated electrode area was 3.41 × 10−3 cm2 that corresponds to a decrease equal to one order of magnitude with respect to the geometric area of the SS substrate.

Fig. 4
figure 4

a Rate study for obtaining the effective area of the SS|COP-Tp; b diffusion coefficient determination. Interfaces: SS|COP-Tp 0.01 mol L−1 ferrocene, +0.1 mol L−1 TBAPF6 in CH3CN

Full size image

Electro-preparation of P1AAQnw and Ptnp dispersion

Electrosynthesis of P1AAQnw was accomplished by PM directly on the already electro-synthesized template [14]. Figure 5a illustrates a continuous current growth of the respective j/t transient that, after 25 s, starts declining; this could mean that the polymer is filling the mesopores. Polymer behavior in KPF6 after template removal is seen in Fig. 5b, from where the proper oxidation potential (p-doping process, Fig. 5c) can be also worked out, being possible to deduce that 0.800 V is the most appropriate potential to dope the obtained polymer.

Fig. 5
figure 5

a j/t transients obtained during nw electrosynthesis on SS|COP-Tp and TEM images of SS|COP-P1AAQnw; b P1AAQnw electro-chemical response in 1 × 10−4 mol L−1 KPF6, ν = 100 mV s−1; c P1AAQnw p-doping in 10−1 mol L−1 KPF6, ν = 100 mV s−1

Full size image

Thus, when 0.800 V is applied, thanks to the p-doping process of the nanostructured polymer, PtCl6 −2 inclusion inside the P1AAQnw is favored. Then, all what is needed is to apply a reduction potential for a short period of time to achieve reduction and homogenous dispersion of Ptnp into the polymeric matrix [12]. The corresponding voltammetric response is depicted in Fig. 5.

If the P1AAQnw electro-chemical responses are compared before and after the addition of Ptnp (Fig. 6), two oxidation peaks are observed at 0.120 and 0.500 V when Ptnp was incorporated, as already corroborated by polymers having very similar structures and Ptnp dispersed over its surface [20]. This, together with TEM micrographs (Fig.7), would demonstrate that the proposed modification has been successfully accomplished.

Fig. 6
figure 6

SS|COP-P1AAQnw and SS|COP-P1AAQnw-Ptnp voltammetric profile in 0.1 mol L−1 H2SO4 + 10−1 mol L−1 KPF6, ν = 100 mV s−1

Full size image
Fig. 7
figure 7

TEM micrographs of Ptnp dispersed on the SS|COP-P1AAQnw electrode

Full size image

Figure 7 shows a representative image and nanoparticles size of Pt. It can be seen that Ptnp are dispersed over SS|COP-P1AAQnw with average particle size of 20 nm (dTEM; Table 1). Deposited Pt mass was calculated by integration of the electrical charge consumed during the deposition process Q (C cm−2), assuming that the only contributing reaction was the faradaic process [21]:

$$ {\mathrm{PtCl}}_6^{-2}+4{\mathrm{e}}^{-1}\to Pt+6Cl $$
Table 1 Physical characteristics of SS|COP-P1AAQnw-Ptnp
Full size table

Then, the quantity of platinum on the electrode (w) is calculated from:

$$ w=Q\cdotp \mathrm{M}/\left(z\cdotp F\right) $$
(2)

where Μ is Pt molar mass (195.09 g mol−1), z number of exchanged electrons, and F the Faraday constant (96,485.3 C mol−1). Then, the Pt deposited on P1AAQ nanowires was 5.7 mg cm−2.

Formic acid electro-oxidation on the modified electrodes

In Fig. 8a, the electro-oxidation of FA during the last two stages of electrode manufacturing is observed; hence it was demonstrated that although the inclusion of P1AAQnw allows obtaining an effective electro-catalysis, Ptnp inclusion improves even more the electro-catalytic process, since a significant current increase is evidenced.

Fig. 8
figure 8

FA oxidation voltammograms on (a) SS|COP-P1AAQnw and SS|COP-P1AAQnw-Ptnp and (b) SS|COP, SS|COP-Ptnp, and SS|COP-(Tp)-Ptnp in 0.1 mol L−1 formic acid solution. ν = 50 mV s−1

Full size image

In the voltammetric profiles shown in Fig. 8a, it is also interesting to note just one oxidation peak (∼1000 mV), which would account for a very high reaction kinetics for the electro-oxidation process on SS|COP-P1AAQnw-Ptnp, or else, that due to the large increase in effective area, given by the P1AAQ nano-structuring and Ptnp inclusion, oxidation of the entire FA molecules on the electrode surface would be possible, without resulting in electro-chemically detectable CO adsorption generated during this catalytic process [22, 23].

In Fig. 8b, the possible uses of each electrode obtained in the electro-oxidation, using SS|COP and SS|COP-Tp with and without Ptnp incorporation were analyzed during the manufacturing process, which allowed verifying that, since low currents were observed, none of these would be effective for the electro-catalysis. Anyway, SS|COP is the electrode exhibiting the best electro-catalytic response, possibly due to the low roughness of the copolymer that allows effective CO adsorption. On the other hand, the template provides a very small surface area (as inferred from potential scan rate studies), which prevents any electro-catalysis. Ptnp inclusion enables observing small currents for both electrodes but always lower than those already found for the nanostructured copolymer.

Thus, it is possible to corroborate that the incorporation of nw and Ptnp significantly improves the current response during electro-oxidation (twofold), showing better response than those reported in similar works under analogous working conditions [20, 21].

If the effect of concentration on the SS|COP-P1AAQnw-Ptnp optimized electrode is analyzed (Fig. 9a), it can be seen how, with increasing concentration, the electro-oxidation profile is maintained only until concentrations ≤6 mol L−1. At very high concentrations (6 and 8 mol L−1), while the electrode is not poisoned, the voltammetric profile starts disappearing. If the potential window for both concentrations (Fig. 9b) is expanded, the profile similarity with other previously reported by several authors [12, 20] can be revealed. They observed COads absorption proving that at high concentrations (≥6 mol L−1) oxidation to CO2, due to saturation of the system, would not be direct, preventing the swift release and conversion of FA into CO2, accumulating thus COads molecules on the Ptnp surface. Nevertheless, at lower concentrations (Fig. 9a), FA direct oxidation to CO2 indeed would take place.

Fig. 9
figure 9

a Voltammetric profile during FA electro-oxidation at different concentrations on SS|COP-nw-Pt; b enlargement of the working window for FA electro-oxidation at higher concentrations on SS|COP-nw-Pt. ν = 50 mV s−1

Full size image

From the results of FA electro-oxidation in Fig. 8a, it is also possible to conduct a more thorough analysis of the effective area of Ptnp on the SS|COP-P1AAQnw-Ptnp surface, based on the information reported by Gloaguen et al. [24]. Since increase of electro-catalytic activity seemed to be enhanced by the inclusion of Ptnp and this in turn depends on the number of available surface sites, the activity referred to the platinum surface area (A r), has a great physical significance. To obtain it, integration is first necessary to determine the area under the voltammetric profiles recorded on SS|COP-P1AAQnw and SS|COP-P1AAQnw-Ptnp; then, from its subtraction, is possible to calculate the actual charge (Q r), corresponding to what has been carried out just on the Ptnp, assuming that each platinum atom should interact with at least one COads. In this case, an actual Q of 3.18 mC cm−2 was found. Equation 3 enables establishing the amount of Ptnp atoms that have effectively participated in the electro-catalysis:

$$ {Q}_{\mathsf{r}}\left(\mathsf{m}\mathsf{C}\right)/\mathsf{real}\ \mathsf{c}{\mathsf{m}}^2=F\cdotp n/N $$
(3)

where N is the Avogadro’s constant (6.022 × 1023 atoms mol−1), which yields 1.99 × 106 platinum atoms cm2.

On the other hand, the Pt area (Ar, cm2) can be calculated through the relationship between charges Q r and the corresponding charge obtained from the working current (Q c), i.e., A r = Q c/Q r, thereby it is possible to determine the specific surface area (S) in m2 g−1 according to Eq. 4:

$$ S=100\ A\mathrm{r}/\left(W\cdotp {A}_{\mathrm{g}}\right) $$
(4)

This enables a macroscopic idea of the amount of Pt that effectively acts on the electro-catalysis, different from the actual amount, which is deposited on the electrode surface, disregarding an idea of the Ptnp morphology. Hence, applying a structural model to S, aided by the TEM images (Fig. 7), and assuming that Pt particles are highly homogeneous and with spherical shape, dispersed uniformly over the electrode surface, the electrode area A g, where Ptnp (3.41 × 10−3 C m2) have been electro-deposited, particle size (d) (nm) is then calculated from:

$$ d=6000/\left(\rho \mathit{\cdotp}S\right) $$
(5)

where ρ is the platinum density (21.04 g cm−3), obtaining a theoretical average particle size of ca. 16 nm, which is consistent with the ∼20 nm worked out from TEM (Fig. 6; Table 1) images.

Kinetic features of FA electro-oxidation on SS|COP-P1AAQnw-Ptnp electrode

Figure 10 shows a current density (j) vs. potential scan rate (v) plot for FA electro-oxidation on SS|COP-P1AAQnw-Ptnp, wherein it can be verified that the current density of the anodic peak is directly proportional to the square root of the scan rate, indicating a diffusion controlled process [25].

Fig. 10
figure 10

Anodic peak current vs. scan-rate square root plot of FA oxidation on SS|COP-P1AAQnw-Ptnp in 0.1 mol L−1 formic acid +0.1 mol L−1 KPF6 aqueous solution

Full size image

SS|COP-P1AAQnw-Ptnp stability depending on the concentration

SS|COP-P1AAQnw-Ptnp electrode stability was measured by comparing the anodic peak of FA electro-oxidation using 0.1 and 8 mol L−1 solutions, after 100 successive electro-oxidation cycles and can be seen in Figs. 11a, b, respectively.

Fig. 11
figure 11

Electro-chemical response and standard deviation of SS|COP-P1AAQnw-Ptnp in FA: (a) 0.1 mol L−1, (b) 8 mol L−1 FA. νν = 100 mV s−1

Full size image

These results using the same electrode indicate that the stability of the electrode during the continuous cycling of electro-catalysis presents lower standard deviation when working at lower FA concentrations (σ0.1M 1.03 × 10−6 and 1.08 × 10−4, respectively). The results also confirm what it has been already mentioned namely that at low concentrations, poisoning of the electrode would not occur, but the poisoning indeed takes place when greater FA concentrations are used. In this case, as noted, the CO adsorption on SS|COP-P1AAQnw-Ptnp surface is so high, which does not allow a reliable reuse of the electrode over 20 cycles, where the standard deviation is ≤10−5, showing a decrease of the catalytic activity, which is also evidenced by smaller currents.

It is noteworthy, finally, that the use of electro-chemical techniques for each of the steps of the electrode modification guarantee variables control and reproducibility, particularly when generating nanostructures in situ, phenomenon practically impossible when the modified electrodes are prepared adding nanostructures synthesized ex situ. In that case, it is very difficult to control that the surface be perfectly reproducible, as verified by previous studies concerning the characterization of the electrodes prepared and used in the current work [24].

Conclusions

Optimization of the SiO2 template electro-chemical obtaining on COP allowed the successful electrosynthesis of P1AAQnw on an SS-modified electrode. Template speed studies allowed the effective working area to be determined, this being 3.41 × 10−3 cm2.

The prepared P1AAQnw showed a p-doping response that enabled Ptnp, dispersion. Furthermore, voltammetric response and TEM micrographs of these electrodes corroborated the inclusion of the metallic nanoparticles (∼20 nm) over the P1AAQnw. It was also corroborated from the theoretical treatment of the charges, which yielded Ptnp sizes of 16 nm.

Evaluation of the effective FA electro-oxidation demonstrated the considerable improvement obtained when the polymer is nanostructured, becoming evident an increase of the anodic peak current (∼1000 mV) that increased even more, ca. twofold, when Ptnp was dispersed.

The systematic study of FA electro-oxidation at concentrations ≤6 mol L−1 does not possess the characteristic profile of CO adsorption due to rapid reaction kinetics on the electrode; however, at FA concentrations ≥6 mol L−1, typical profiles of CO adsorption is observed as a result of system saturation. These results were also consistent when the stability of the electrode was studied, showing that at low concentrations the variability of CO oxidation peak current is favored, which does not occur at high concentrations wherein reproducible FA electro-oxidation peak currents are not possible to be obtained.

Furthermore, the potential scan rate study showed that FA electro-oxidation on SS|COP-P1AAQnw-Ptnp is a diffusion-controlled process.

Thus, this method provides a highly reproducible nanostructured electrode that is prepared solely based on electro-chemical techniques and offers promising results for FA electro-catalytic oxidation.