Oscillatory behavior in the electrochemical oxidation of formic acid on Pt(100) rotation and...

download Oscillatory behavior in the electrochemical oxidation of formic acid on Pt(100)  rotation and temperature effects.pdf

of 8

Transcript of Oscillatory behavior in the electrochemical oxidation of formic acid on Pt(100) rotation and...

  • 7/21/2019 Oscillatory behavior in the electrochemical oxidation of formic acid on Pt(100) rotation and temperature effects.pdf

    1/8

    Journal of Electroanalytical Chemistry 500 (2001) 3643www.elsevier.nl/locate/jelechem

    Oscillatory behavior in the electrochemical oxidation of formicacid on Pt(100): rotation and temperature effects

    T.J. Schmidt *, B.N. Grgur 1, N.M. Markovic, P.N. Ross, Jr.

    Materials Science Diision, Lawrence Berkeley National Laboratory, Uniersity of California, Berkeley, CA 94720, USA

    Received 8 May 2000; received in revised form 27 June 2000; accepted 1 August 2000

    Dedicated to Professor R. Parsons on the occasion of his retirement from the position of the Editor in Chief of the Journal of

    Electroanalytical Chemistry and in recognition of many contributions to electrochemistry

    Abstract

    We investigated the oscillatory behavior in the kinetics of formic acid electrooxidation on Pt(100) in 1 mM HClO4 solution. We

    studied the effect of different experimental parameters on the oscillatory behavior, viz. defined HCOOH mass-transport to the

    electrode surface by using the rotating disk electrode technique, the temperature of the supporting electrolyte, and the nature of

    anions. We suggest that the interdependence of the reaction steps during HCOOH oxidation, the adsorption of anions and the

    competition for adsorption sites among the reaction partners and intermediates lead to complex non-linear kinetics. It was evident

    that once the individual reactions in the dual path mechanism reach steady state the oscillations vanish. These conditions can be

    reached either by enhanced formic acid reaction rates induced by electrode rotation or by increased temperature. Under specific

    conditions of anion and formic acid concentration, relaxational oscillations can be transformed into mixed-mode oscillations. 2001 Elsevier Science B.V. All rights reserved.

    Keywords: Oscillatory electrochemical reaction; Formic acid; Pt(100) electrode

    1. Introduction

    Oscillation of current or potential in electrochemical

    reactions is a well known phenomenon, e.g. in metal

    dissolution [1,2], peroxide reduction [3], the hydrogen

    oxidation reaction [1,4], or in the oxidation of small

    organic molecules like ethylene [5], formaldehyde [6,7],methanol [8] or formate/formic acid [5,9,10]. In the past

    decade oscillatory instabilities during formic acid oxida-

    tion have been extensively studied on low-index Pt(hkl)

    electrodes [11 15]. Although general agreement exists

    that current oscillations during formic acid oxidation

    arise from oscillating surface coverages by adsorbed

    formic acid, adsorbed poisoning intermediates, oxygen

    containing species and anions from the supporting elec-

    trolytes, the true nature of the oscillatory behavior is

    still controversial. For example, one explanation for

    sustained oscillations under potentiostatic control was

    the local variation of the pH close to the electrode

    during HCOOH oxidation on Pt(100), resulting in a

    sudden change of the surface coverage by electroactive

    oxygenated species (OHad), therefore changing the reac-

    tion rate for removing the poisoning intermediate (e.g.adsorbed carbon monoxide, COad) which is formed by

    HCOOH decomposition [11,14]. Additionally to the pH

    effect, Tripkovic et al. proposed in their study that the

    number of surface sites occupied by bridge-bonded

    COad plays an important role in the oscillatory behav-

    ior of formic acid oxidation on stepped Pt(hkl) elec-

    trodes [12]. In contrast to the pH hypothesis, Markovic

    and Ross proposed that the periodicity, the amplitude,

    and the absolute magnitude of the current oscillations

    during HCOOH oxidation on Pt(100) depend on at

    least three parameters, viz. electrode potential and po-tential history, formic acid concentration, and the con-

    centration of the anion of the supporting electrolyte

    [13]. Closely following a concept described by Degn [1],

    * Corresponding author. Tel. +1-510-4864793; fax: +1-510-

    4865530.

    E-mail address: [email protected] (T.J. Schmidt).1 Current address: Faculty of Technology and Metallurgy, Univer-

    sity of Belgrade, Karnegijeva 4, 11000 Belgrade, Yugoslavia.

    0022-0728/01/$ - see front matter 2001 Elsevier Science B.V. All rights reserved.

    PII: S 0 0 2 2 - 0 7 2 8 ( 0 0 ) 0 0 3 4 2 - 9

  • 7/21/2019 Oscillatory behavior in the electrochemical oxidation of formic acid on Pt(100) rotation and temperature effects.pdf

    2/8

    T.J. Schmidt et al./ Journal of Electroanalytical Chemistry 500 (2001) 3643 37

    Strasser et al. pointed out that the presence of a suffi-

    ciently large ohmic resistance is crucial to develop

    oscillatory instabilities in formic acid oxidation on all

    three low-index Pt single crystals [15]. For details about

    the theory of electrochemical oscillators we refer to

    Refs. [16,17]. All results for formic acid oxidation on

    the Pt(hkl) electrodes were obtained on stationary elec-

    trodes, i.e. the oscillatory behavior was not investigatedunder well defined hydrodynamic conditions. In order

    to simulate enhanced mass transport conditions, the

    effect of electrolyte stirring on the oscillatory behavior

    during formic acid oxidation was studied and discussed

    in Refs. [11,1315]. The main findings from the applied

    turbulent mass-transport induced by magnetic stirring

    of the electrolyte is that under certain experimental

    conditions the enhanced mass transfer of HCOOH

    caused the oscillations either to cease [11,14] or to be

    sustained with different shapes and periodicity [13,15].

    It is important to note that Schell and coworkers [6,10]

    found that sustained current oscillations can be devel-

    oped during the galvanostatic oxidation of formic acid

    on a rotating polycrystalline platinum disk electrode. In

    addition, these studies showed that oscillatory behavior

    on polycrystalline Pt was not substantially affected by

    the temperature of the electrolyte [6,10]. To the best of

    our knowledge there are no similar studies of the

    oscillatory behavior on Pt(hkl) electrodes under (i) well

    defined mass-transport conditions and (ii) variable

    temperature.

    In this paper we present results on oscillations devel-

    oped during formic acid oxidation on Pt(100) in 1 mMHClO4 solution. Our aim is to demonstrate the influ-

    ence ofdefinedmass-transport to the electrode surface

    by utilizing the rotating disk electrode (RDE) tech-

    nique. Additionally, the effect of temperature was stud-

    ied under specific experimental conditions in order to

    isolate the effect of surface reaction kinetics on the

    oscillations. In the last part, we report on so-called

    mixed-mode oscillations obtained under specific condi-

    tions concerning anion and formic acid concentration.

    2. Experimental

    The results presented here were conducted using twodifferent Pt(100) single crystals (0.283 cm2, Figs. 14;

    0.572 cm2, Fig. 5). The single crystals were flame an-

    nealed in a hydrogen flame and cooled in a mild stream

    of 2% H2+Ar. One crystal (0.283 cm2) was mounted

    into the disk position of an insertable ring disk elec-

    trode assembly (Pine Instruments), fully described pre-

    viously [18]. Both electrodes were transferred into a

    thermostated standard three compartment electrochem-

    ical cell and immersed into the Ar-purged electrolyte

    under potential control at 0.2 V (Ar, Matheson

    Research Purity; 1 mM HClO4, EM Science suprapure;

    or 5 mM H2SO4, J.T. Baker ultrex, both prepared with

    triply pyrodistilled water). The reference electrode was

    a saturated calomel electrode (SCE) separated by a

    closed bridge from the working electrode compartment

    in order to avoid chloride contamination. A circulating

    constant temperature bath (Fischer Isotemp Circulator)

    maintained the temperature of the electrolyte within

    0.5C. The measurements were carried out either at

    298 K or at 318 K. All measurements were conducted

    non-isothermally, i.e. keeping the temperature of the

    reference electrode constant (298 K) while that of the

    working electrode was varied. After recording the basevoltammetry of the Pt(100) electrode in order to check

    the cleanliness of the electrode preparation, the elec-

    trode was held at 0.2 V and formic acid (Aldrich

    ACS Reagent) was injected to obtain a 12 or 35 mM

    HCOOH solution, respectively. All experiments were

    carried out on the unreconstructed Pt(100)(11) sur-

    face present in acid and alkaline electrolyte [19].

    3. Results

    3.1. Oscillation characteristics on stationary electrode

    The general shape of the polarization curve for for-

    mic acid oxidation on Pt(100) in 1 mM HClO4 (12 mM

    HCOOH, 10 mV s1, 298 K) between hydrogen ad-

    sorption (referred to as Hupd) and oxide formation

    (referred to as OHad) is shown in Fig. 1. This curve

    reproduces the behavior we reported previously [13].

    Since the oxidation currents for formic acid oxidation

    under potentiodynamic conditions overlap in the posi-

    tive and negative sweeps, respectively, the voltam-

    mograms, for clarity, are unfolded with respect to thepotential axis. Starting from the negative potential

    limit, the oxidation of HCOOH is suppressed up to ca.

    Fig. 1. (a) Unfolded potentiodynamic curve for formic acid oxidation

    on Pt(100) (10 mV s1). Sweep reversed at 0.7 V. (b) Potentiostatic

    transient at 0.63 V in a sweep-and-hold experiment after sweepingfrom 0.7 V negatively. 1 mM HClO4+12 mM HCOOH, 0 rpm, 298

    K.

  • 7/21/2019 Oscillatory behavior in the electrochemical oxidation of formic acid on Pt(100) rotation and temperature effects.pdf

    3/8

    T.J. Schmidt et al./ Journal of Electroanalytical Chemistry 500 (2001) 364338

    Fig. 2. Potentiodynamic curves for formic acid oxidation on Pt(100)

    (10 mV s1) at 0 rpm, 1600 rpm and 3600 rpm, respectively (only the

    negative sweep is shown). 1 mM HClO4+12 mM HCOOH, 298 K.

    static current oscillations are typical for so-called relax-

    ation oscillations, where the system remains in either

    one of two quasi-stationary states, followed by a sharp

    transition from one state to another. Hence, the time

    the system stays in the active and inactive state is

    defined as tA and tI, respectively (see insert in Fig. 1).

    The total charge passed in a single oscillation can be

    referred to as QA and QI flowing during tA and tI. It isnoteworthy that the sustained potentiostatic oscillations

    at 0.63 V (insert Fig. 1) are relatively stable in ampli-

    tude and frequency over a long period of time (i.e.

    15 min) before significant changes in the shape are

    observed.

    3.2. Influence of electrode rotation on oscillatory

    behaior

    3.2.1. Potentiodynamic conditions

    The effects of rotation rate on the oscillatory behav-

    ior during the first positive and consecutive negative

    sweeps, respectively, are shown in Fig. 2. There are two

    characteristics that represent the effect of rotation: (i) in

    the positive sweep the oscillations cease immediately

    under electrode rotation; and (ii) the number of spikes

    recorded in the negative sweep direction decreases on

    increasing the rotation rate. In the latter case, the

    number of oscillations is decreased from eight at the

    stationary electrode to four when rotating at 1600 rpm,

    and finally to two when the rotation rate is increased up

    to 3600 rpm. At rotation rates above 3600 rpm, we

    were not able to develop oscillations in these potentio-dynamic experiments. Fig. 2 also shows that with in-

    creasing rotation rate the system stays longer in both

    the deactivated (tI increases) and activated (tA in-

    creases) states. As a consequence the frequency of the

    oscillations goes down with increasing rotation rate.

    The effect of the rotation rate on the amplitude of the

    oscillations is not as pronounced, although a detailed

    inspection of Fig. 2 reveals a small decrease with in-

    creasing electrode rotation. Finally, we note that in the

    consecutive two following sweeps, the number of oscil-

    lations was always smaller than in the first sweep or no

    oscillations were observed anymore (especially at 3600

    rpm, the polarization curve presented was a first sweep

    phenomenon).

    3.2.2. Potentiostatic conditions

    Fig. 3 shows the influence of rotation on potentio-

    statically developed oscillations at 0.63 V. In these

    experiments the potential was swept (10 mV s1) from

    the negative potential limit to 0.7 V, and then after the

    reversal of the sweep direction the potential was held at

    0.63 V. At this potential we were recording the sus-

    tained oscillations shown on the left-hand side of Fig.3a, which qualitatively resemble the current transient

    from the insert in Fig. 1. We then started to rotate the

    0.2 V due to the presence of adsorbed CO containing

    poisoning intermediates formed by formic acid decom-

    position (for simplicity, hereafter denoted as COad,

    although other organic intermediates might be present

    [20,21]). The concomitant removal of COad and the

    beginning of formic acid oxidation starts at ca. 0.2 V,resulting in the increase in the current. Eight subse-

    quent anodic current oscillations are observed between

    ca. 0.55 and 0.7 V. After reversal of the potential at 0.7

    V, 12 additional oscillations appear up to 0.4 V, fol-

    lowed by a monotonic decrease of the current when

    approaching the initial potential. We note that when

    sweeping the electrode consecutively over a longer pe-

    riod of time, however, the number of oscillations in

    both positive and negative sweeps was reduced before

    the oscillations ceased completely, indicating that the

    oscillatory behavior seen in potentiodynamic experi-

    ments is of time-dependent nature. The potentiody-namic data in Fig. 1 are in general agreement with data

    presented in Refs. [13,14], which were recorded under

    similar conditions.

    The insert in Fig. 1 is representative for current

    oscillations recorded under potentiostatic conditions. In

    this experiment, the potential was swept positively from

    0.2 to 0.7 V and stopped at 0.63 V upon the reversal

    point (sweep-and-hold experiment). This procedure

    produced almost the same current transients compared

    to potential stepping experiments from 0.2 V to the

    region where current oscillations occur. We, therefore,focused in this study on sweep-and-hold experiments.

    As already mentioned previously [11,13] the potentio-

  • 7/21/2019 Oscillatory behavior in the electrochemical oxidation of formic acid on Pt(100) rotation and temperature effects.pdf

    4/8

    T.J. Schmidt et al./ Journal of Electroanalytical Chemistry 500 (2001) 3643 39

    RDE at 400 rpm as indicated in Fig. 3a for ca. 90 s

    before stopping the electrode rotation again. Obviously,

    the oxidation current for formic acid oxidation in-

    creased when the rotation started, indicating higher

    reaction rates induced by the enhanced mass-transport

    to the electrode surface. However, as in the sweep

    experiments (Fig. 2), the amplitude decreases slightly

    once rotation is applied but stays relatively stable

    within the duration of the experiment. An inspection of

    Fig. 2 also shows that, at 400 rpm, the current peaks

    are sharper than under stationary conditions, whereas

    the induction period from the deactivated to the acti-

    vated state (tI) increases with time. As a consequence,

    the frequency of oscillations decreases continuously

    (Fig. 3a). However, when the rotation was stopped

    after 80 s, the reverse behavior was observed, and over

    an additional 40 s the shape of the oscillations, the

    frequency and amplitude slowly recover to the state

    recorded before rotation. Similar observations were

    made by Strasser et al. on Pt(100) using magnetic

    stirring of the electrolyte [15]. Fig. 3b shows that the

    increase of the rotation rate from 400 to 1600 rpm leads

    to an increase of the reaction rate, a continuous de-

    crease of the amplitude of the current spikes and the

    frequency of oscillations. Fig. 3b also demonstratesthat, in contrast to Fig. 3a, the current peaks broaden

    and the poisoning period increases, so after 250 s the

    oscillations are annihilated completely. It is noteworthy

    that we were unable to restore oscillations over the next

    10 min, as observed in Ref. [15], or when we stopped

    the rotation. We also found that the time required for

    the oscillations to cease completely decreases with in-

    creasing rotation rate due to the enhanced formic acid

    concentration on the surface at higher rotation rates.

    3.3. Influence of temperature on oscillatory behaior

    Another possibility to enhance the reaction rate for

    formic acid is the increase in temperature of the elec-

    trolyte (Fig. 4). Note that due to the inverse propor-

    tionality of the diffusivity, D, and viscosity of the

    electrolyte, with increasing temperature, T, (DT/

    ), mass transport to the electrode is not affected by the

    temperature change. Since we were not able to develop

    current oscillations when we started the experiment at

    temperatures above 303 K (in contrast to galvanostatic

    oscillations on polycrystalline Pt reported in Ref. [10]),

    we carried out the following experiment. We started the

    experiment at 298 K and recorded potentiostatic cur-

    rent oscillations at 0.63 V under sweep-and-hold condi-

    tions. The resulting current transients are shown on the

    left side in Fig. 4. In order to demonstrate the regular

    shape of the oscillations, we magnified the time scale

    for the first 4 s of the experiment. After an additional

    60 s we started to increase the temperature of the

    electrolyte to 318 K and recorded the response of the

    system. The temperature increase results instanta-

    neously in higher HCOOH oxidation rates and a slight

    decrease in the frequency. The oscillations themselves,however, stay in the same regular shape as shown in the

    time-magnified part of Fig. 4. Only after ca. 145 s,

    when the desired temperature of 318 K equilibrated in

    the electrolyte, do the current spikes start to lose their

    regular shape, resulting in a more or less chaotic

    development of current oscillations with irreproducible

    frequency and amplitude (note that for 145t330 s

    the current transients are shown on an expanded time

    scale). The last current spike is observed at ca. 310 s

    before the oscillations cease completely. We also note

    that we were unable to revive the oscillatory behaviorafter this time through rotation or through a change of

    the potential.

    Fig. 3. Influence of rotation rate on potentiostatic transients at 0.63

    V: (a) after switching from 0 rpm to 400 rpm and vice versa; (b) after

    switching from 0 rpm to 1600 rpm. 1 mM HClO4+12 mM HCOOH,

    298 K.

    Fig. 4. Influence of temperature on potentiostatic transients at 0.63 V.As indicated, the temperature was increased from 298 to 318 K. 1

    mM HClO4, 12 mM HCOOH, 0 rpm.

  • 7/21/2019 Oscillatory behavior in the electrochemical oxidation of formic acid on Pt(100) rotation and temperature effects.pdf

    5/8

    T.J. Schmidt et al./ Journal of Electroanalytical Chemistry 500 (2001) 364340

    Fig. 5. Evaluation of so-called mixed-mode oscillations on Pt(100):

    potentiostatic transient at 0.47 V. The insert shows the section

    betweent=100 s and 130 s. 5 mM H2SO4+35 mM HCOOH, 0 rpm,

    298 K.

    tors. For a discussion of this issue, we refer to Section

    4.1.

    4. Discussion

    In Section 3, we presented results which illustrate

    that in addition to bulk concentration of formic acid,electrode potential and potential history, the concentra-

    tion of specifically adsorbing anions [13], or the pres-

    ence of a high ohmic resistance [15], two other

    parameters can be added as important factors in deter-

    mining the oscillatory behavior of formic acid. We

    demonstrated the strong influence of (i) enhanced and

    defined mass-transport to the electrode surface induced

    by electrode rotation and (ii) the electrolyte tempera-

    ture on the oscillatory instabilities. In what follows, our

    aim is neither to propose any detailed mechanism for

    oscillations during formic acid oxidation nor to present

    any model which reproduces the experimental results,

    but rather to report how the enhanced rate of formic

    acid oxidation, induced by an increase of both mass-

    transport of formic acid and temperature of the elec-

    trolyte, may effect the interdependence of the

    experimental parameters responsible for complex non-

    linear kinetics.

    4.1. Formic acid oxidation mechanism and oscillating

    surface coerages

    It is now well established by various electrochemicaland spectroscopical studies that HCOOH electrooxida-

    tion on Pt proceeds via the so-called dual-path mecha-

    nism proposed by Capon and Parsons [22] which is

    summarized in the following reaction scheme:

    According to this mechanism the direct oxidation

    path can be ascribed to dehydrogenation of HCOOH to

    CO2 via an active intermediate (often referred to as

    COOHad, reaction 1) [22]. The CO2 production in the

    parallel path can be assigned to a (non-faradaic) dehy-

    dration of HCOOH to adsorbed CO (representing one

    of several possible products; reaction 2) in a first step

    which is followed by the COadoxidation reaction (reac-tion 4). OH adsorption, which is generally assumed to

    be necessary for the CO-oxidation reaction, occurs in

    3.4. Influence of bisulfate anion concentration on

    oscillatory behaior

    Fig. 5 shows some potentiostatic current transients

    recorded in 35 mM HCOOH containing 5 mM H2SO4electrolyte at 0.47 V (sweep-and-hold experiment in the

    negative sweep, 0 rpm). In our previous publication (see

    figure 11 in Ref. [13]) no oscillations were observed

    with the same amount of sulfuric acid but at slightly

    higher potentials (0.53 V) and formic acid concentra-

    tion (50 mM HCOOH), indicating the strong influence

    of both parameters on the oscillatory behavior. Under

    our present experimental conditions, we developed ini-

    tially periodic oscillations with very low frequency (ca.

    0.1 s1) and a long deactivation period (Fig. 5). In the

    third oscillatory wave, however, first instabilities on the

    deactivation branch occurred, followed by a sudden

    frequency increase in the next current waves. Addition-

    ally, the aforementioned current instabilities in the de-

    activation period developed to periodically returning

    current spikes. The overall current transient behavior

    illustrated in Fig. 5 represents so-called mixed-modeoscillations. These are characterized by the appearance

    of an oscillation with a large amplitude which is fol-

    lowed periodically by a small current spike on the

    deactivating branch of the main oscillatory wave (see

    insert in Fig. 5). Similar sequences of mixed-mode

    oscillations were found for galvanostatic HCOOH oxi-

    dation on polycrystalline Pt (2 M HCOONa+0.5 M

    H2SO4) [6] and in the potentiostatic oxidation of

    HCOOH on Pt(100) (1 M HCOONa+0.5 M H2SO4,

    applying a large external ohmic resistance) [15]. In the

    latter work the mixed-mode oscillations were explainedby the presence of a chemical species as a second

    feed-back variable giving rise to two interacting oscilla-

  • 7/21/2019 Oscillatory behavior in the electrochemical oxidation of formic acid on Pt(100) rotation and temperature effects.pdf

    6/8

    T.J. Schmidt et al./ Journal of Electroanalytical Chemistry 500 (2001) 3643 41

    acid electrolyte by decomposition of water (reaction

    3). In what follows, we propose that the interdepen-

    dence of the reaction steps in reactions 1 4 and the

    competition for adsorption sites among the reaction

    partners and intermediates lead to complex non-linear

    kinetics. For example, if the reaction rate kpkox,

    then COad accumulates on the surface and poisons

    reactions 1, 3 and 4. For these conditions, therefore,the rate of oxidation of formic acid is mostly deter-

    mined by the rate of both the formation of OHad [23]

    and the oxidative removal of COad through a Lang-

    muirHinshelwood (L-H) type reaction [24]. It is im-

    portant to note that the rate of the L-H reaction is

    strongly affected by the delicate balance between the

    coverage of COad, OHad and anions from the support-

    ing electrolytes (reaction 5, kA). While anions have

    negligible effects on the interaction of COad with the

    surface atoms, they indeed have a strong effect on

    OHad adsorption [25] and adsorption of formic acid

    [13]. Hence, in order to develop a more realistic pic-

    ture about the competition of reactants, intermediates

    and products for bare Pt sites the anion adsorption/

    desorption from the supporting electrolyte must be

    incorporated in the reaction scheme proposed above

    for the oxidation of HCOOH. In our previous study,

    we emphasized that anions play a very important role

    in the development of the oscillatory behavior during

    the oxidation of HCOOH on the Pt(100) surface [13].

    The most important observation from this work is

    that by adding small amounts of sulfuric acid (up to 1

    mM HSO4

    ) to the base electrolyte (1 mM HClO4) weobserved an increase in the induction time, tI, and a

    decrease of the time in the activated state, tA. AtcH2SO41 mM, no oscillation could be observed any-

    more. The same effects were observed by adding small

    amounts of the more strongly adsorbing Cl, but at

    concentrations three orders of magnitude smaller. The

    effects of anions on tI can be interpreted by the afore-

    mentioned competition between anions and OHad for

    the same adsorption sites (initially starting at defect

    sites of the crystal), which in terms of our reaction

    scheme implies a decrease of kOH. The time in the

    activated state, tA, decreases for the same reasons,

    namely the competition of formic acid and anions for

    the same Pt surface sites. At this point one may em-

    phasize that in order to produce sustained current

    oscillations it is necessary to create surface conditions

    which allow the system to reach the specific balance of

    the individual reaction rates presented in our scheme.

    To see this balance in action, we created the sustained

    mixed-mode oscillations under specific conditions of

    anion and formic acid concentrations (5 mM H2SO4,

    35 mM HCOOH), Fig. 5. For similar types of oscilla-

    tions on Pt(100) Strasser et al. [15] suggested that theyare triggered and controlled by an additional negative

    feed back variable, which gives rise to two interacting

    oscillators and consequently to mixed-mode oscilla-

    tions. Most importantly, however, they ascribed the

    additional negative feed back variable as the result of

    the presence of some (unknown) chemical species.

    From our experiments, it turned out that these chemi-

    cal species may be specifically adsorbing anions, which

    can control the adsorption of both OHad and formic

    acid molecules.At this point, it is appropriate to note that recently

    Strasser et al. pointed out that the presence of a large

    ohmic resistance, R, is a crucial parameter to develop

    oscillations during formic acid oxidation [15,26]. The

    mechanism of action of a large ohmic resistance is

    rather simple. The potential drop from the output of

    the potentiostat/galvanostat to a point in the bulk

    electrolyte in front of the reference electrode, E, is

    split into the potential drop across the double-layer

    (double-layer potential, DL) and the ohmic potential

    drop (E=DL+IR). The latter includes all ohmicdrops, including those from resistors added in series in

    the external circuit [27]. Under current flow, at con-

    stant E, DL changes when the current changes. In

    terms of the dual path mechanism for HCOOH oxida-

    tion, the changing double layer potential leads to the

    following scenario under potentiostatic conditions

    (E=const., R=const). When the current is high as in

    the activated state (direct path, reaction 1) the poten-

    tial drop across the double-layer is much lower than

    E, leading to slow OHad formation rates (reaction 3)

    and, as a consequence to COad accumulation on thesurface (reaction 2). Due to the poisoning of the sur-

    face by COad the direct path becomes inhibited and

    the current decreases. Lower currents, however, in-

    crease DL and, consequently, the OHad formation

    rate is increased, which, in turn, leads to freeing of

    surface sites for the direct path, and a concomitant

    rise in the current to complete the cycle. In a simple

    picture, the formation of OHad can be switched on

    and off periodically like the light in a room by inser-

    tion of a large ohmic resistance. It is, however, not

    clear whether the presence of an ohmic resistance issufficient to create the oscillatory behavior during for-

    mic acid oxidation. Our experiment clearly showed

    that simply by varying the concentration of anions one

    can control the mode of oscillation. Unfortunately, in

    the model proposed by Strasser et al. anion adsorption

    is not taken into account [26]. In addition, this model

    also cannot explain the oscillatory behavior found un-

    der conditions of rather high electrolyte/anion concen-

    tration [5,6,9,12,28], i.e. under conditions where the IR

    drop seems to be negligible. However, closely follow-

    ing Koper and Sluyters [16], a minimum value of theIR drop is necessary to observe oscillations, but this

    parameter does not seem to represent a sufficient one.

  • 7/21/2019 Oscillatory behavior in the electrochemical oxidation of formic acid on Pt(100) rotation and temperature effects.pdf

    7/8

    T.J. Schmidt et al./ Journal of Electroanalytical Chemistry 500 (2001) 364342

    4.2. Current oscillations at defined mass -transport

    As mentioned in Section 1, all previous results for the

    oscillatory behavior during formic acid oxidation on

    platinum single crystal electrodes were obtained on a

    stationary electrode. A common shortcoming of this

    electrode configuration is that the mass transport of

    reacting species is not well controlled, and for lowconcentrations of electroactive species one usually mea-

    sures pure transient currents. Transient behavior during

    the oxidation of formic acid on stationary electrodes

    arises from a slow diffusion of HCOOH through a

    relatively thick diffusion layer (200500 m). The best

    way to overcome this limitation is to use the RDE

    method, where a known pattern of hydrodynamic flow

    is imposed on the solution through the Nernst diffusion

    layer with thickness of typically 50 m. Conse-

    quently, the RDE offers (pseudo) steady state currents

    for the oxidation of formic acid even when small con-

    centrations of HCOOH are used and relatively fast

    sweep rates are applied (Fig. 2). With a more effective

    supply of HCOOH to the Pt(100) electrode in RDE

    measurements, we can nicely control the surface con-

    centration of HCOOH, c0. Thus, RDE experiments

    may create very similar conditions to those we achieved

    by monitoring the effects of the formic acid concentra-

    tion on the oscillatory behavior but on a stationary

    electrode [13]. For the latter system, we found that with

    increasing formic acid concentration (i) tI increases, (ii)

    tAdecreases and (iii) the amplitude of the current spikes

    rises. Due to the higher HCOOH bulk concentration,cb, the HCOOH surface concentration, c0, is increased.

    Assuming positive order reaction kinetics for formic

    acid dehydration (reaction 2), this results in higher

    COad formation rates and at constant kOH in slower

    rates for the oxidative removal of COad, kox (increasedtI), in agreement with the negative reaction order kinet-

    ics for oxidation of solution phase CO [24]. The same is

    valid for decreased tA: the higher poisoning rate forces

    the system to leave the activated state faster due to a

    faster inhibition of reaction 1. The higher amplitude

    can be explained by the higher HCOOH surface con-

    centration, c0, assuming positive order kinetics for for-

    mic acid oxidation (see e.g. Ref. [29] and references

    therein). Note, at formic acid concentration above 100

    mM, no oscillations could be found anymore [13],

    consistent with our supposition that if the rate of

    reaction 4 (oxidative removal of COad) becomes con-

    stant and is in equilibrium with kp and kOH, the oscilla-

    tions will cease and the overall reaction will reach its

    steady state.

    In RDE measurements, the formation of OHad is a

    rotation rate independent process, and consequently the

    major effect of the rotation on the oscillatory behaviorshould be discussed in the light of the variation of the

    surface concentration of formic acid. Figs. 2 and 3

    clearly reveal that the rate of formic acid oxidation and

    the activated/deactivated time is increased by increasing

    the mass transport of HCOOH (rotation of electrode)

    to the electrode. On the other hand, the amplitude and

    the frequency are decreased by increasing the rotation

    rate, and at 1600 rpm the oscillations cease after ca. 250

    s. Interestingly, the steady state current observed after

    the oscillatory behavior vanishes represents the deacti-vated state, indicating that the formic acid dehydration

    step (reaction 2) is becoming the dominant process with

    constant reaction rates while kox, kp and kOH are in

    equilibrium. We recall that the same happens on the

    stationary electrode, but for a much higher bulk con-

    centration of formic acid. Additionally, these experi-

    ments may shed some light on the influence of an

    ohmic resistance on the oscillatory behavior. Since the

    magnitude of an eventually present IR drop in the

    electrolyte should not be influenced by electrode rota-

    tion, we conclude that the presence of a large IR drop

    is not a sufficient parameter to induce the development

    of sustained current oscillations during formic acid

    oxidation. However, we want to note that the theoreti-

    cal model presented in the work by Strasser et al. [26]

    does fit their experimental results under their experi-

    mental conditions [15], which, in turn, are different

    from ours.

    4.3. Temperature effects on oscillatory instabilities

    The effects of temperature on the oscillatory behav-

    ior in the electrochemical oxidation of formic acid onPt(100) are in nature very similar to the rotation effects

    discussed above. Most notably, the reaction steps 14

    were affected differently by the temperature variation.

    As a consequence, the periodicity, the amplitude and

    the absolute magnitude of the current oscillations

    changed continuously with an increase of temperature.

    Recently, we found that the energetics of OH formation

    on a platinum single crystal are only slightly affected

    over the temperature range of 278 333 K [25]. This

    indicates that the temperature has only a minor effect

    on the reaction of OH formation (reaction 3 in our

    scheme). On the other hand, we showed that the rates

    of both the direct and the poison oxidation pathways

    of formic acid are enhanced at higher temperatures.

    The latter process increases the formation of COadwhich, in turn, can effectively block the sufficient ad-

    sorption of OHad required to trigger the transition of

    the system from an inactive to an active state. It

    appears, therefore, that the temperature has an indirect

    effect on the surface concentration of OHad. Due to a

    lack of OHad at higher temperature, smaller current

    amplitudes are obtained at 318 K than at 298 K, and

    the system stays longer in the deactivated, i.e. COadpoisoned state. It is interesting that at 318 K after ca.

    150 s the stable relaxational oscillations were slowly

  • 7/21/2019 Oscillatory behavior in the electrochemical oxidation of formic acid on Pt(100) rotation and temperature effects.pdf

    8/8

    T.J. Schmidt et al./ Journal of Electroanalytical Chemistry 500 (2001) 3643 43

    transformed into a chaotic-like type of oscillation

    (Fig. 5). Although the information is extremely limited,

    it is reasonable to propose that the development of the

    chaotic current spikes after ca. 145 s may tentatively

    be ascribed to arise due to some fluctuations of the

    thickness of the diffusion layer by thermal motion of

    the electrolyte. Nevertheless, regardless of the true na-

    ture of the chaotic behavior, our experiments showunambiguously that the main effect of the temperature

    on the oscillatory behavior is to convert the complex

    non-linear kinetics to steady state oxidation of formic

    acid. Fig. 5 shows that at 318 K after ca 300 s the

    oscillations ceased completely, indicating that the over-

    all rate of reaction is under steady state conditions.

    5. Conclusions

    We investigated the oscillatory behavior in the kinet-

    ics of formic acid electrooxidation on Pt(100) in 1 mM

    HClO4 solution. We studied the effect of different

    parameters on the oscillatory behavior, viz. defined

    HCOOH mass-transport to the electrode surface by

    using the rotating disk electrode technique, the temper-

    ature of the supporting electrolyte, and the nature of

    anions. We suggest that the interdependence of the

    reaction steps during HCOOH oxidation, the adsorp-

    tion of anions and the competition for adsorption sites

    among the reaction partners and intermediates lead to

    complex non-linear kinetics. It was evident that once

    the individual reactions in the dual path mechanismreach steady state, the oscillations vanish. These condi-

    tions can be reached either by enhanced formic acid

    reaction rates induced by electrode rotation or by in-

    creased temperature. Under specific conditions of anion

    and formic acid concentration, relaxational oscillations

    can be transformed into mixed-mode oscillations.

    Acknowledgements

    We want to acknowledge H.A. Gasteiger for helpful

    discussions. This work was supported by the Assistant

    Secretary for Conservation and Renewable Energy,

    Office of Transportation Technologies, Electric and

    Hybrid Propulsion Division of the U.S. Department of

    Energy under Contract No. DE-AC03-76SF00098.

    References

    [1] H. Degn, J. Chem. Soc., Faraday Trans. 64 (1968) 1348.

    [2] P. Russel, J.S. Newman, J. Electrochem. Soc. 133 (1986)

    2093.

    [3] T.G.J. Van Venrooij, M.T.M. Koper, Electrochim. Acta 40

    (1995) 1689.

    [4] K. Krischer, M. Luebke, W. Wolf, M. Eiswirth, G. Ertl, Elec-

    trochim. Acta 40 (1995) 69.[5] J. Wojtowicz, N. Marincic, B.E. Conway, J. Chem. Phys. 48

    (1968) 4333.

    [6] M. Schell, F.N. Albahadily, J. Safar, Y. Xu, J. Phys. Chem.

    93 (1989) 4806.

    [7] M. Hachkar, M. Choy de Martinez, A. Rakotondrainibe, B.

    Beden, C. Lamy, J. Electroanal. Chem. 302 (1991) 173.

    [8] M. Krausa, W. Vielstich, J. Electroanal. Chem. 399 (1995) 7.

    [9] N.A. Anastasijevic, H. Baltruschat, J. Heitbaum, J. Elec-

    troanal. Chem. 272 (1989) 89.

    [10] F.N. Albahadily, M. Schell, J. Electroanal. Chem. 308 (1991)

    151.

    [11] F. Raspel, R.J. Nichols, D.M. Kolb, J. Electroanal. Chem.

    286 (1990) 279.[12] A.V. Tripkovic, K. Popovic, R.R. Adzic, J. Chim. Phys. 88

    (1991) 1635.

    [13] N. Markovic, P.N. Ross, Jr., J. Phys. Chem. 97 (1993) 9771.

    [14] F. Raspel, M. Eiswirth, J. Phys. Chem. 98 (1994) 7613.

    [15] P. Strasser, M. Lubke, F. Raspel, M. Eiswirth, G. Ertl, J.

    Chem. Phys. 107 (1997) 979.

    [16] M.T.M. Koper, J.H. Sluyters, J. Electroanal. Chem. 371 (1994)

    149.

    [17] P. Strasser, M. Eiswirth, M.T.M. Koper, J. Electroanal. Chem.

    478 (1999) 66.

    [18] N.M. Markovic, H.A. Gasteiger, P.N. Ross, J. Phys. Chem.

    99 (1995) 3411.

    [19] I.M. Tidswell, N.M. Markovic, P.N. Ross, Phys. Rev. Lett. 71

    (1993) 1601.

    [20] O. Wolter, J. Willsau, J. Heitbaum, J. Electrochem. Soc. 132

    (1985) 1635.

    [21] S.G. Sun, J. Clavilier, A. Bewick, J. Electroanal. Chem. 240

    (1988) 147.

    [22] A. Capon, R. Parsons, J. Electroanal. Chem. 45 (1973) 205.

    [23] T.J. Schmidt, R.J. Behm, B.N. Grgur, N.M. Markovic, P.N.

    Ross Jr., Langmuir 16 (2000) 8169.

    [24] N.M. Markovic, B.N. Grgur, C.A. Lucas, P.N. Ross, J. Phys.

    Chem. B 103 (1999) 487.

    [25] N.M. Markovic, T.J. Schmidt, B.N. Grgur, H.A. Gasteiger,

    P.N. Ross, Jr., R.J. Behm, J. Phys. Chem. B 103 (1999) 8568.

    [26] P. Strasser, M. Eiswirth, G. Ertl, J. Chem. Phys. 107 (1997)

    991.[27] A.J. Bard, L.R. Faulkner, Electrochemical Methods, Wiley,

    New York, 1980.

    [28] H. Okamoto, Electrochim. Acta 37 (1992) 37.

    [29] V.S. Bagotzky, Y.B. Vassilyev, Electrochim. Acta 12 (1967)

    1323.

    .