Electrochemical Polymerization of Furfural on a Platinwn Electrode in Aqueous Solutions of Potassiwn Biphthalate

Three different electrochemical methods confirm the growth processes of polyfurfural on platinum electrodes in aqueous solutions. The electrochemical oxidative polymerization of furfural occurs only with 0.1 O moi L-1 potassium biphthalate as the supporting electrolyte. Electrochemical and spectroscopic methods are employed to characterize the polymeric film produced. Based on spectroscopic data, a polymeric structure involving furfural and biphthalate anions is discussed.

Conducting polymers are regarded as potential materiais for the electronic industry.The possibility of combining the properties of organic polymers and the electronic properties of semiconductors in these new materiais has been the driving force for various applications.Conducting polymers can be prepared via chernical 1 • 3 or electrochemical polymerization•.Electronically conducting polymer films generated by electrochemical polymerization are usually deposited onto a supporting electrode surface by various electrochernical techniques, including potentiostatic ( constant-potential), galvanostatic (constant current) and potentiodynarnic (cyclic voltammetry) methods.Conjugated polymers have attracted considerable scientific interest for their potential use dueto their unusual electrical, electrochernical and optical properties 5 .A new generation of electroactive polymers based o n polyheterocycles (polypyrrole, polythiophene) has been the object of investigation" 10 Attention so far has focused primarily on the chernical polymerization of conjugated polymers bearing furan rings, while little auention has been dedicated to their electrochemical polymerization 11 • 13 This paper deals with a potential candidate for new applications developed from laboratory studies based on furfural (2-furancarboxyaldehyde).Furfural is produced from many sources because most agricultura!wastes and woods contain sufficient quantities of pentosans, thus justifying this research.The worldwide production offurfural in 1985 was about 200,000 tons, originating mostly from comcobs, cottonseed, rice hulls, sugarcane bagasse, wood, and invol ved various wood technologies and organic solvent plants 14 .Despi te the poor chernical stability offurfural (2-furancarboxaldehyde), the literature contains evidence confirrniog the electroactivity ofthis compound in differentelectrodes 1 s. 16.However, much less attention has so far focused on the electropolymerization of polyfuraos 17 , probably because of their supposedly less well-defioed structures and poor stability.Furfuryl alcohol polymerizatioo is a well known homogeneous acíd catalysis process 1 .Furfural, which has been characterized as an electrophore, can be reduced into furfuryl alcohol or oxidized into furoic acíd 15 The maio purpose of this study is directed to obtain a stable polymeric film fromelectrooxidation of furfural in order to offer a organic barrier against corrosion processes.The next step should apply these procedures on non-noble metallic surfaces such as steel, zinc and others metais.
•e-mail: reinaldo@iq.ufrgs.brln a previous investigation in o ur laboratory, the presence of an organic polyfurfural film was proposed to explain the changes in the electrochemical behavior of a platinum platinized electrode 16 Current versus potential curves and electrochemical irnpedance measurements conclusively characterized the presence of the organic film on the electrode surface.The polymer film was electrogrown by keeping the curreot density constant at 34 j.tA cm 1 (galvanostatic method) aod applyiog a constant potential of around 0.30 V (SCE) for varying periods of time (potentiostatic method) i o aceto nitrile.
The present work describes lhe electrochemical polymerization of furfural io ao aqueous solution formed oo platioum by three distinct electrochemical methods.
The film was characterized by scanniog electron microscopy (SEM), differential scanning calorimetry (DSC) and UV spectroscopy.Electrochemical experiments were also conducted to confirm the presence of the organic film on the electrode surface.This verification was essentially based on cyclic voltammetry.

Reagents and materiais
Pure furfural was obtained from Merck; however, periodical distillation was required before use to maintain its levei of purity.An aqueous solution containing 80. O mmol L-1 of furfural was prepared by dissolving this compound with pure conductive water.Potassium biphthalate (p.a.) from Merck was used as the supportiog electrolyte.Only 0.1 O moi L-1 supportiog electrolyte aqueous solutions were tested.

Cell and electrode m ateriais
Electrochemical polymerization was carried out in a conventional three-electrode electrochernical cell.The working electrodes with distinct active surfaces were made of pure platinum wire and the auxiliary electrode also consisted of a platinum wire, while the reference was a saturated calomel electrode.The working electrodes were subjected to conventional polishing with alurnioa (50-200 mícron), degreased with acetone and dried with hot air, prior to the electropolymerization step, thus avoiding contamination and superficial irregularities.

Techniques and instrumentation
The polymers wereelectrosynthesized using a TECfROL model TCHl000-005 de power source to grow the film galvanostatically and an AUTOLAB PGSTAT 30 potentiostat/galvanostat for potentiostatic and potential cycling film electrogrowth.Micrographs of the elect:rodes were taken using a JSM model 5800 microscope.A Perkin-Eimer DSC-4 calorimeter was employed for the differential scanning calorimetry experiments.UV studies were made using a Shimadzu UV-VIS model1601 PC spectrophotometer.

Galvanostatic experiments
ln this study, the polymer film was grown anodically by passing 10 mA of cunent through the working electrode immersed in an aqueous solution of0.10 moi L• 1 potassium biphthalate containing 80 mmol L 1 of furfural.Various polarization times were tested.Although the presence ofthe film on the electrode surface became visib te after 30 minutes of polarization, the electrode surface was totally covered only after90 minutes ofpolarization, as shown in Figure 1.This image, taken with a microscope and magnified 64 times, shows the electrode covered only partially by the polymer film.A point worth noting is that, by this method, the film grew continuously throughout the polarization time, suggesting that it may have conducted the electricity since the process was not intezmpted after the first layer ofthe organic polymer was fotmed.This may be considered the first piece of evidence that the film should support the charge transfer process on the polymeric structure.The film's motphology and the inftuence ofthe polarization time on the structure ofthe film fotmed on the electrode surface will be discussed below.
Despite the visib te modification ofthe electrode surface, confirmation ofthepresence ofthe film wasobtained by comparing the resulting voltammetric pro file when the modifiedelectrode was transferred to an aqueous solutionof0.10molL 1 potassium biphthalate without furfural.The I( E) profiles of the electrode without the presence of the film, which wez•e recorded at 0.10 Vs• 1 in the potential range of 1.00 V(SCE) to 1.75 V(SCE), clearly show electrooxidation ofthe electrode swface in this interval.A simple strategy was employed to confirm the presence ofthe film formed on the electrode sw•face as a function ofthe electrolysis time.Figw•e 2 shows the charge associated with this electrooxidation process (Qb ).A comparison ofthe I(E) profiles ofboth modified and platinum electrodes indicates that this charge decreases.This effect is dependent on the polarization time, suggesting that the film thickness increases propottionally.The coverage ofthe electrode (9)was calculated from the charges using the following equation: where Qb is the charge associated with the electrooxidation process of the unaltered electrode, and Q,. is the charge associated with the electrooxidation process of the electrode modified by the film formed during different polarization times.
Figure 3 shows theeffect ofthe coverage ofthe electrode surface as function of the polarization times on the electrooxidation process ofthe metal.As expected, the coverage ofthe electrode swface increases along with the polarization time.The electrode surface was saturated after 60 minutes of polarization.ln addition to this evidence, another unequivocal strategy was applied to confirm the modification ofthe electrode's electrochemical behavior as a function ofthe presenceofthe polymez'ic film.The electrode was transfened to another cell containing 0.10 moi L-1 of aqueous sulfuric acid.Voltammograms ofplatinum electrodes in this medium are well defined in the literature 18 • The changes in the I(E) profiles revealed through a comparison ofthe two electrodes may be attributed to the presence ofthe film.
The data obtained fi'om the galvanostatic experiments were characterized by comparing the I(E) curves ofthe platinum electrode in aqueous sulfuric acid solutions with and without the film on the electrode surface.Figure 4  ences the electrochemical behavior ofthe platinum electrode in an appropriate potential window.The choice of potential range served to confirm how the charge transfer processes can be affected in the shift from a double layer formation (O. O V /SCE) to a platinum oxide formation (1.2 V/SCE).The anodic currents relating tothe platinum electrode's electrooxidation process were shown to decrease in a comparison ofthe two electrodes.On the other hand, the cwrents observed during the cathodic potential sweep relating to the electroreduction oftheplatinum oxide practicallydisappeared.A possibleexplanation ofthis effect is that, after 90 minutes ofpolarization, the electrode was ahnost totally covered by the organic fihn, which effectively prevented the formation of platinum oxide.
The cun•ent ofthe modified electrode decreased in the more negative potential range in which cathodic currents relating to hydrogen adsotption/desorption processes prevailed.Both the anodic and the cathodic processes on the platinum sw•face in aqueous sulfuric acid solutions were clearly altered by the presence of the fihn on the electrode surface.It is important to point out the fihn's electrical conductivity, which was evidenced by the nonintetruption of the charge transfer processes on the electrode surface as the polymet• was formed.

Potentiostatic experiments
fu this experiment, the potential of the working electrode was kept constant at 2. 65 V(SCE) during various polarization times.The organic film grew fi•om the sarne solution tested before.Deaerated solutions were tested to prevent parallel reactions on the electrode surface.The sarne strategy was applied to confirm the presence of the fihn on the electrode surface.The resulting voltarnmetric profile when the modified electrode was transferred to an aqueous solution ofO.lO moi L• 1 potassium biphthalate without fwfural was used for this pwpose.Figure 5 shows the degree of coverage ofthe sw"face at different polarization times.The sarne conclusions can be drawn by analyzing the data showed in this figw•e, i.e., coverage of the electrode swface increased with polarization time.
However, the main evidence of the electrochemical oxidation of furfural on the electrode surface, which resulted in v isib le covet•age, is shown in Figure 6.The current versus time curves obtained when the electrode was polarized at 2.65 V(SCE) in the absence and in the presence of furfural are depicted.A comparison of the cun•ent densities in the presence and in the absence of furfw•al reveals an unequivocal increase in the cw1•ent density ofthe electrode with this organic compound.The anodic cwrent was likely related with the electrooxidation process of furfural (1), giving the furfural dication (2), according to the Scheme 1.
The onset ofpolymer growth was likely related with the fotmation of furfural dication (2).The stability of the dication using both positions2 and 5 on the furan ring was theoretically determined 19 and the sarne mechanism was proposed to explain the polymet• formed from furfuryl alcohol;n.

.1.3. Potentiodynamic experiments
Potentiodynamic experiments were also carried out in thís medi um to check the possibílíty ofthe anodíc polyrnerization offurfural in the potential range of 2.00 V to 2.65 V(SCE) at 50 mV s• 1 .The film grew after 400 cycles.When the number of cycles increased, a spongy whíte film appeared on the electrode surface.The sarne approach was applíed to characterize the presence of the organic film formed on the electrode surface.The resulting voltarnrnetríc profiles when the modífied electrode was transferred to an aqueous solu tion of 0.1 O moi L-1 potassium bíphthalate without the presence of furfural were used for this purpose.Figure 7 shows the degree of surface coverage in dífferent sweep cycles.The sarne conclusions can be drawn by analyzing the curve shown in this figure, i.e., coverage of the electrode surface increased with the number of cycles.
The sarne strategy was applíed before transferring the electrode modífied by the film to another electrochemícal cell containing aqueous sulfuric acíd solution.Figure 8 shows the voltarnrnograrns of the modífied electrode in deaerated 0.50 moi L-1 sulfuric acíd solution, as well as the I(E) curve of the bare electrode.The anodíc current increased during the anodíc potential sweep.The sarne effect was observed during the cathodíc potential sweep when comparing the cathodíc peak with the bare electrode.Both oxídation/reduction electrochemícal processes apparently increased in response to the enhancement of the electrode's active surface.These contríbutions may be assocíated with the redox charge-díscharge processes of the polyfurfural, as well as the formation of a Ptü monolayer and the oxídation of the film.

Comparison among the three ele ctrochemical m ethods of growing
The structure of the film obtained by the galvanostatic method as well as the polymer's adherence was dependent on the polyrnerization time and current applíed.The current was controlled by the equipment; however, the electrode potential was shífted without control to more positive potential value.This variable should affect the crystallíne structure ofthepolymer as well as theroughness.Studies involving others electrochemícal methods the electrode potential was kept under control.The selected values were not so positive, inducíng a polyrneríc structure with a granular morphology.The gal vanostatic electrogrow should be suítable if the adherence of the film on the electrode surface is used as a criteríon to choíce between the three methods.

.1.5. Surface morphology of the film
Having clearly characterized the presence of the organic film by the electrochemícal methods approach, the next step was to determine the morphology and structure of the newly formed polymer.
The surface morphology ofthe organic film was studíed by SEM photomícrographs.Figure 9 shows the SEM photomícrographs ofthe electrodes after the polyfurfural film was grown using three electrochemícal techniques.A surface examination revealed considerable dífferences in the film morphology accordíng to the electrochemícal synthesis method employed.The coatings on the electrode surface formed by the potentiodynamic and potentiostatic methods showed a granular morphology.The film showed líttle adherence when few cycles or short polarization times were applíed.On the other hand, the structure of the film obtained by the galvanostatic method and the polyrner's adherence were also dependent on the polyrnerization time and current applíed.Fast film growth led to poor adherence of the polymer on the substrate, whíle longer polyrnerization times resulted in more adherent films with a more compact structure and higher rou ghness.An analysis of the film synthesized during short polymerization times revealed a laminar structure.I tis a well known fact that compact layers are resistant to degradation whíle laminar structures hydrolyze more easíly 21 .Figure 1 O shows how the polarization time affects the morphology of the film forrne d on the electrode surface produced by the galvanostatic method.
A laminar structure is visible with short polarization times, whíle 90 minutes of galvanostatic electrogrowth lead to a more compact morphology.

UV spectroscopic studies
To confirm the presence offurfural in the polyrner structure, several spectroscopíc measurements were taken in solutions containíng dífferent furfural concentrations in the 200 to 350 nm wavelength, as shown in Figure 11.The UV spectra of the samples were linear at 270 nm, where the maxímum absorption occurs, in the furfural concentration range of 2. 0 x l Q-5 moi L-1 to 8.0 x lQ-5 moi L-1 .The quantity of furfural consumed duríng the galvanostatic experiment 0.8 The difference in light absorption between furfural solutions before and after the electrolysis revealed that 2.80 x 10• 4 moles of furfural were consumed; indicating that 2. 69 x 10• 2 g of furfural was oxidized during this experiment.The fact that a ce~tain amount of furfural was consumed may be considered an indication that the compound participated effectively in the polymer's composition.This hypothesis was confumed by a calculation ofthe theoretical furfuralmass that should beconsumed when this cun•ent is applied for that pe~•iod oftime, assuming a cha~ge transfer process involving 2 F mot 1 of furfural, as proposed in Scheme 1.The value calculated theoretically is 2.68 x 10• 2 g, which is vety close to that dete~•mined from spectroscopic studies of the electrolytic solution.The coincidence ofthe numbers allows one to conclude that furfural effectively participated in the polyme~• structure.The UV spectroscopy of the fihn dissolved in absolute ethanol was conclusive with regard to the presence of furfw•al in the polymer stmcture, since a clearly identifiable peak at 270 nm was observed, as depicted in Figw•e 12.
Finally, the polymer formed on the electrode surface was studied in order to confirm the presence of furfural in the film stmcture.The organic fihn on the electrode surface was scratched and examined, based on an approach whereby, using Differential Scanning Calorimetry (DSC) and UV spectroscopic techniques, con•elations can be identified in the film's chemical structure.

Differential scanning calorimetty
Figure 13 shows the first scan of the resins obtained ti'om the galvanostatic electrogrowth of furfural according to the above described procedure.The highest tempe~•ature at which the polymer melting process occurred was found to be 202.01°C, which far exceeds the boiling point ofpure furfural (161.7 °C) and which can therefore not be associated with potassium biphthalate, since this compound decomposes before melting.These findings indicate that the polymer structure involved a furfural and potassium biphthalatebased resin.The second scan confitmed that the initial stmcture of the compound was destroyed, since the melting point was that of phthalate anhydride.
The electrochemical resinification of furfural is a more complex process than that proposed for chemically generated furfuraldehyde resins 21 • The stmcture of the polymer should agree with the one proposed for cationic polymerization, involving an initiation and propagation mechanism 22 • However, the spectroscopic data and other information suggest the mechanism presented in Scheme 2.

Conclusions
This paper demonstrated that a polyme~•ic furfural film was formed on a platinum surface in aqueous potassium biphthalate solutions.The process occw1•ed by electrochemical oxidation of furfural under current o r potential contra l.The fihn' s adherence, thickness and morpho logy proved to be dependent on polarization time.A complex polymeric structure involving the polyfurfural film and biphthalate anions was favored dueto the greate~• availability of anions.

_Figure 2 .Figure 3 .
Figure L Image of the platinum electrode surface covered with the organic film after 90 minutes of anodic growth with the application of 10 mA of current, taken with a digital camera with zoom.

Figure 4 .
Figure 4. Voltamogramms of platinum electrode at 25 oc in aqueous aerated solution of 0.50 moi L• 1 H2S04, recorded at v = 0.10 vs• 1 : a) bare electrode, and b) after galvanostatic electrogrowth of polyfurfural, applying 10 mA of current for 90 minutes in 0.10 moi L• 1 of potassium biphthalate and 80 mm o! L• 1 of furfural.

Figure 7 . 2 FigureS.Figure 9 .Figure 1
Figure 7. Effect of the number of cycles during CV experiments on the degree of c overage (8) of the electrode surface, measured during t he electrooxidation of the platinum electrode in an aqueous aerated solution of 0.10 mal L •' potassium biphthalate and recorded at v = 0.10 Vs•'-