Multiple adsorbate interactions between reduced CO2 adsorbates and either allyl alcohol or propargyl alcohol residues on platinum in 0.5 M sulphuric acid

Abstract The interaction of propargyl alcohol and allyl alcohol with a platinum (Pt) electrode modified by the presence of reduced CO 2 adsorbates (r-CO 2 ) is studied through voltammetry and potentiostatic current transients by using a flow cell technique. The interaction between each alcohol and r-CO 2 is promoted at three constant potentials corresponding to different degrees of Pt surface coverage by H-atoms. For both alcohols the anodic stripping peaks and the charge balance are interpreted in terms of the formation of composed adsorbates. Based upon the probable r-CO 2 structure and H-bonding interactions between r-CO 2 and the alcohol molecules or their adsorbed residues, average alcohol residues/r-CO 2 molecular ratios are estimated.


INTRODUCTION
The formation of reduced carbon dioxide on platinum electrodes takes place when COr dissolved in an acid electrolyte interacts with a Pt electrode covered by Hatoms [ 1,2]. Depending on the adsorption conditions, the presence of two types of reduced COr (r-COr) species can be inferred from the electrooxidation voltammograms. Both species can be electrodesorbed within a potential range which lies close to the threshold potential of the O-adatom electroadsorption reaction[3-51. Similar conclusions concerning the number of r-CO, species can be derived from IRRAS and PTIRRAS data [6,7].
As there is a relatively wide potential range where stable reduced C02-adsorbates can be formed on Pt, the possibility of promoting the formation of composed adsorbates through the addition of a foreign substance to the solution can be explored. The formation of composed adsorbates has been recently investigated on polycrystalline Pt electrode saturated with CO adsorbates interacting with unsaturated alcohol residues in acid solutions [8]. This type of process has been reported for CO-modified Pt electrodes with other simple organic molecules, and the composed adsorbates formed under these conditions have been described in terms of mixed adlattices involving lateral modifications and adsorbate reorganisations to approach detlned structural situations [9].
The present paper refers to new interactions at constant potential in acid solution between reduced CO1 adsorbates on Pt, and residues which result from the electroadsorption of either propargyl alcohol (PA) or ally1 alcohol (AA). These results allow us to *Visiting Professor. Permanent addressz Consejo National de Investigaciones Cientiticas y T&r&as, Argentina. advance limiting average stoichiometries for possible coadsorbates produced at different potentials.

EXPERIMENTAL
Runs were made in 0.5 M H,SO, by using a 50 ml capacity flowing electrolyte electrochemical cell technique [lO] with the usual arrangement of three electrodes. A polycrystalline (PC) Pt working electrode (0.23 cm2 real surface area) was encased in a glass holder. The counter electrode was a Pt electrode (ca 1 cm2 geometric area) mounted in a separate compartment. A Hg/Hg2S0,(s)/K2S0,(sat.)/0.S M H2S0, reference electrode (mse) (E/V (nhe) = 0.659) was employed. Potentials in the text are referred to the mse scale. The working electrode pretreatment consisted of a potential cycling in the -0.650 to 0.800 V range at 0.2 V s-' in 0.5 M H2S04.
The base electrolyte was prepared from 98 % sulfuric acid (Merck p.a.) and Millipore-MilliQ*-water. The solutions of the different alcohols (PA and AA) were prepared from twice distilled p.a. chemicals. CO2 was prepared from the reaction between 50% H2S0, and NaHCO,, and bubbled through the electrolyte solution under saturation. Runs were made at 25°C under either Ar or CO2 atmosphere, depending on the stage of the experiment.
Each run consisted of the following steps. Firstly, the voltammetric behaviour of the H-atoms electrosorption reactions were followed under potential cycling at 0.2V s-l. The stability of the cyclovoltammograms was taken as a purity test of the system following the criteria reported in the literature [ll]. Then, the potential was set to Ead(C02), the adsorption potential for C02. As the null current at this potential was reached, the solution was saturated with CO2 to form the reduced CO2 adsorbates. Subsequently, the solution was replaced by 0.5 M HzS04

-
holding the value of &(C02). Afterwards, the potential was changed to E&X), the adsorption potential for the unsaturated alcohol (X = PA, AA), and 2 ml of either 0.1 M PA or 0.1 M AA solution was added to produce the composed adsorbate. Simultaneously the current transient of the second electroadsorption process was recorded during 3 min, the time necessary to reach an adsorption-time-invariant anodic stripping voltammogram for the adsorbed residues. Occasionally, the adsorption of both substances was made at $ I I I I the same potential. From the current transients the E total electroadsorption charge resulting from the addig 32 _ tion of X was determined. Finally, after replacing the 3 solution again by 0.5 M H,SO,, the stripping voltammogram of the composed adsorbates was obtained "0 .at 0.2V s-l as indicated elsewhere [l2]. The follow-o ing adsorption potentials were chosen: Ead(CO1) = -0.650 V and -0.500 V; E&X) = -0.650, -0.500 and -0.250 V. At these potentials f?n, the degree of surface coverage by H-atoms on Pt, is close to 1, 0.5 and 0, respectively.

Potential/V
Results from the blank are depicted in  ( Table 1). The electroadsorption current transients 1. PA electroadrorption on Pt modified by r-CO, and the anodic stripping voltammograms of the adsorbates produced on Pt from PA ( Fig. 1 b) and Firstly CO2 was adsorbed at E,(C02) = -0.650 V AA ( Fig. lc) at different E,(X) coincide with data (0, = l), for about 10 min to reach the r-CO2 recently published[ 131.
saturation coverage. The electroadsorption of PA  at E,,(PA) = -0.650 V produces an electroreduction current transient (Fig. 2a). The charge density for this process after 3 min becomes considerably smaller than that found for the blank (Fig. lb). In contrast to Pt modified by CO adsorbates [8], the electroreduction of PA in the presence of r-CO1 is not inhibited because H-atoms are available on Pt. The anodic stripping voltammogram resulting from r-CO, and PA residues differs from that depicted for r-CO2 and PA residues (compare Figs la, lb, and 2a). It shows firstly a small anodic current which is probably due to the electrodesorption of residual H-atoms and traces of molecular hydrogen in solution, followed by two partially overlapping anodic peaks (1 and 2) in -0.100 to 0.800 V range, peak 1 being higher than peak 2. When PA is adsorbed at E&PA) = -0.25OV (Fig. 2b), the electroadsorption current transient becomes definitely anodic comprising a limiting charge density value of ca 0.04 mC cmm2. The anodic stripping voltammogram of the adsorbed residues presents also two anodic current peaks qualitatively -0.1 0.3 Potential /V similar to those displayed in Fig. 2a, but in this case the initiation of the anodic processes takes place at potentials which are close to the 0-electroadsorption threshold potential.
Results obtained by setting Ead(C02) = -0.650 V and E&PA) = -0.500 V, differ also from the blank, as for the latter only an anodic PA electroadsorption current transient is observed. In this case the contribution of peak 1 is enhanced, its peak potential being intermediate between the peak potential values described previously.
When the r-CO, adsorbates are produced at E,(CO,) = -0.500 V, ie below the r-CO2 saturation coverage, and the PA electroadsorbed residues are formed at E&PA) = -0.650 V, the resulting electrochemical features are qualitatively comparable to those depicted in Fig. 2a.
On the other hand, when E&PA) = -0.250 V (Fig. 3a), only an anodic electroadsorption current transient is observed, and the anodic stripping voltammogram shows that peak 2 results higher than peak 1. Finally, in runs made by setting &(COz) = ,?$#A) = -0SOOV (Fig. 3b) the electroadsorption current transients exhibit firstly a very small cathodic spike followed by an anodic current which reaches a peaked value, then decreases, becomes cathodic and tends slowly to zero. In this case the anodic stripping voltammogram becomes comparable to that resulting for EJCOz) = -0.650 V and E&PA) = -0.500 V, except that the peak 1 to peak 2 height ratio approaches 1.
In all these runs only a few cyc1ovoltammograms are required for cleaning the Pt electrode and recovering the voltammogram of the Pt10.5 M H,SO, system.

AA electroadrorption on Pt modified by r-CO2
When E,(CO,) = E&AA) = -0.650 V (Fig. 4a), the AA electroadsorption produces a large cathodic current transient involving a charge density which is one order of magnitude greater than that of plain AA. The corresponding anodic stripping charge remains, however, equal to that one in the blank (Table l), although the potential range of the complex anodic stripping peak moves positively. It should be noted that the anodic stripping peak in the blank is somewhat distorted at its descending branch, whereas after the addition of AA a distortion can be seen at the ascending branch [compare Figs la and 4a].
Runs made at &,(C02) = -0.650 V and E&AA) = -0.250 V show an anodic electroadsorption current involving a limiting charge density. The corresponding anodic stripping voltammogram shows firstly a sharp anodic peak at 0.170 V (peak 1) and another one at 0.37OV (peak 2) which partially overlaps peak 1.
Finally, when &,(C02) = -0.650 V and E,(AA) = -0.5OOV (Fig. 4b), the AA electroadsorption produces a cathodic current transient, in contrast to the anodic ones found for both AA and PA in the absence of r-CO,. In this case the anodic stripping voltammogram exhibits peak 1 at 0.150 V with a small distortion at its ascending branch, and peak 2 at 0.37OV, the former being higher than the latter.
Runs made at &(COJ = -0.500 V and E&AA) = -0.65OV show a cathodic current transient related to AA electroadsorption, the corresponding charge density (Table 1) and the stripping voltammogram approaching those of the blanks. Conversely, when E,(AA) = -0.250 V (Fig. 5a) an anodic AA ekxtroadsorption current transient is observed, and the subsequent anodic stripping voltammogram shows 'two peaks of nearly the same height located at 0.180 V (peak 1) and 0.390 V (peak 2), respectively. Finally, for E&IA) = -0.500 V (Fig. 5b) the AA electroadsorption current transient is only cathodic, and the charge involved is smaller than that found for Ed(COz) = -0.650 V. The corresponding anodic stripping voltammogram is rather similar to that seen in Fig. 4b.
In all these experiments only a few cyclovoltammograms are also required to return to the voltammogram of the Pt/0.5 M H,SO, system.

Summary of results
The electroadsorption current transient and the anodic stripping voltammograms of the adsorbates produced from unsaturated alcohols (AA and PA) on a Pt electrode modified by the presence of r-CO, adsorbates are considerably dependent on Ed(X). These results suggest specific interactions between r-CO2 and the products formed from the unsaturated alcohol electroadsorption, and between the proper alcohol molecules with residual H-atoms on Pt.
Results from the different experiments are assembled in Tables 1 and 2. Table 1  density, qt(Y) Ty = r-CO, +X] resulting from the current transients at &(C02) and E&X); and the anodic stripping charge density, qox(Y), determined at u = 0.2 V s-'. These magnitudes can he derived from the experimental data. The rest of the symbols are explained further on in the discussion. The anodic stripping voltammetric peak potentials (peaks 1 and 2) are given in Table 2.

DISCUSSION
Results previously reported for Pt electrodes modified by CO adsorbates interacting with PA and AA electroadsorbed residues [8], show that the anodic stripping charge density resulting from the composed adsorbed residues was practically the sum of the anodic stripping charge densities of the blanks. This simple charge addition rule is not directly applicable to the present results. Then, to explain the present data it is convenient to consider firstly the electrochemical behaviour of r-CO, on Pt, and to continue with the composed systems resulting from the interactions between r-CO, and the unsaturated alcohols.
This procedure makes possible to estimate average stoichiometries and probable structures of the composed adsorbates.

Behaviour of reduced CO, arisorbates
Radiochemical [  producing finite groups as dimers carboxylic acids[l7]. However, in contrast to bulk compounds, the structure of r-CO, on Pt likely implies an asymmetric potential dependent, probably tlickering structure. This type of structure can explain the dependence of the electrochemical behaviour of r-CO, on the electrolyte composition [4,5], and allows us to understand the formation of a number of electroreduction products The multiplicity of voltammetric peaks related to the electrooxidation of r-CO2 on Pt in acid has been interpreted through a complex reaction mechanism which includes an interconversion of two types of r-CO, specie@].
The r-CO2 saturation electroadsorption charge density at Epd = -0.650 V (f+., = 1) is only 20% greater than that obtained at & = -0.500 V (qt., = 0.5). This is consistent with the fact that the initial CO, electroadsorption involves the participation of strongly bound H-atoms [3]. However, the fact that after reaching about 20% saturation coverage, the amount of r-CO, further increases, would indicate a redistribution of the electronic charge between Pt(H) and r-CO, leading to potential and environmental interaction dependent adsorbate structure@]. This fact also reflects on the redistribution of the residual H-atom electrodesorption peaks in the anodic stripping voltanunogram (Fig. la). These changes can explain why in the presence of r-CO, the H-atom electrosorption processes are not completely inhibited as it occurs in the presence of CO.

Infiuence of r-CO2 on PA and AA electroreduction
The electroreduction of PA and AA in acids yields a number of species with the predominance of different saturated hydrocarbons[l9] and minor amounts of the corresponding saturated alcohols. The latter can be voltammetrically detected particularly at the initial stages of the electroreduction process[l3].
In the presence of r-COr the rate of electroreduction of both alcohols at -0.650 V is considerably diminished, as seen from the corresponding cathodic current transients. For PA this effect becomes relatively greater than for AA. Otherwise, in contrast to AA (Table 1) the anodic stripping charge for PA exceeds that of the blank. However, in both cases the potential of the anodic current peaks are positively shifted as compared to the blank, and in the case of AA the peaks are slightly distorted. The preceding description depends somewhat on the amount of r-CO2 .
The results obtained for PA and AA electroadsorption at & = -0.650 V on Pt under r-CO, saturation coverage, suggest that the partial inhibition of PA and AA electroreduction reactions may be due partly to a decrease of en, and partly to the formation of new adsorbates acting progressively as poisons for the electroreduction reactions themselves. It appears that the first effect predominates for AA, whereas the second one becomes the most relevant for PA.

Formation of composed aakorbates
When the r-CO, covered Pt interacts with either PA or AA diluted solution there is a certain Ed value for each alcohol which is related to the appearance of an excess of anodic stripping charge with respect to the &Or blank. Under these circumstances, the anodic electroadsorption current transients for PA at Ead > -0.500 V, becomes comparatively smaller than that expected from the blanks. These results suggest that the interaction between PA and r-CO, likely involves adsorbed PA molecules rather than electroadsorbed residues.
The situation appears to be somewhat different for AA, since the addition of the latter at Ed = -0.500 V, in the presence of r-COr produces a cathodic current transient in contrast to the anodic current transient which results in the absence of &Or. This difference becomes even more remarkable as the amount of r-CO2 on Pt is increased. In this case, unlike the situation already described in sub-section 2, there is an apparent enhancement of the AA electroreduction process. On the other hand, the fact that at &= -0.25OV, the AA electroadsorption charge density on r-CO, covered Pt is about one half the value resulting for the blank, points out that the presence of r-CO, favours a molecular adsorption of AA rather than the formation of electroadsorbed residues.
The preceding analysis allows us to estimate at least the limiting average PA/r-CO, and AA/r-CO2 molecular ratios for the composed adsorbates produced between r-CO* and PA, and r-CO, and AA. For this purpose let us consider that under r-CO, saturation coverage the charge involved in the proper electrooxidation of r-CO2 approaches the H-atom monolayer charge. Then, the number of electrons per r-CO2 species should depend on both the stoichiometry of the r-CO2 adsorbates and the overall electrooxidation reaction. In this case, let us put forward the reaction in the following way [3,51: Pt,[H,CO,H,O] by taking x = 3 as the most probable value. On the other hand, for the composed adsorbate Aq,,,(X), the fraction of the anodic stripping charge associated with X, is given by AqJX) = qoxol) -q&-CO& (2) In these cases, one considers that COz and H+ are the only products of the anodic stripping reactions[l, 31. Analogously, the number of adsorbed residues related to each X molecule can be obtained by dividing AqOJX) by n(X), the number of electrons entering the electrooxidation of each X-adsorbate. Then, N(X), the number of adsorbed residues resulting from X per r-CO1 adsorbate can be defined by the ratio

(3)
In the electroadsorption of X the value of n(X) changes according to the stoichiometry of the adsorbed residue in the following way. When, the adsorbed residues is the X molecule itself, one has either or CH, = CH-CHrOH + 5H,O =3C02+16H++16e-.
Accordingly, n(X) = n(PA) = 14, and n(X) = n(AA) = 16. For a preset value of X and further assuming that there is no displacement of r-CO, caused through the electroadsorption of X, those figures provide limiting PA/r-CO, and AA/r-CO, molecular ratios for the coadsorbed species.
On the other hand, one can also assume that the electroadsorption of X as well as the corresponding stripping processes are represented by the same stoichiometric reactions already proposed for the blanks, either[ 131 12H+ + 12e- (6) or Pt(CH = CH) + Pt(CH0) + 5H,O =2Pt+3CO,+13H++13e-.
Then, n(X) = n(PA) = 12, and n(X) = n(AA) = 13. Therefore, in the absence of competitive adsorption these figures provide another limiting PA/r-CO, and AA/r-CO, molecular ratios for the coadsorbates.
Data assembled in Table 1 allow description of other features of the coadsorption phenomena involving r-CO, and products resulting from X electroadsorption.
Thus, if one assumes that the electrochemical reactions of the composed adsorbates are not much different to those taking place for the separate adsorbates, the value of n * which represents the X electroadsorption to coadsorbate electroadsorption charge ratio, permits to estimate qO,(th), the hypothetical anodic stripping charge of the composed adsorbate. Thus, for AA taking n* = 2, the values of q,,(th) always exceed the experimental ones, whereas for PA, taking 2.6 < n* c 4, the opposite result is obtained.
The fact that in the present cases the charge addition rule is not obeyed can be attributed to specific chemical interactions; the interactions between PA and r-CO2 being stronger than those involving AA and r-CO,. This conclusion is consistent with the chemical reactivity of these unsaturated alcohols. It is known that PA is able to produce dimers and polymers at metal surfaces [20]. Then, PA dimer formation can take place on the r-CO, covered Pt electrode, a fact which can explain the relatively large value of q&Y) resulting in this case.
Finally, in the presence of r-CO,, the reactivity of PA in the hydrogen evolution reaction (HER) potential range is considerably enhanced, and correspondingly, the electroadsorption charge is diminished as expected for a facile PA electroreduction promoted by H-atoms.

Possible interactions involving the complex adsorbates
The interaction of r-COr with unsaturated alcohols should involve a predominant participation of Hbonding. This type of interaction is well known for substances with -CH,OH functional groups either in the presence or in the absence of water[ 16, 11. Therefore, one should expect that the basic clathratetype structure of r-CO, largely remains for the composed adsorbates. In this case, the entire adsorbed layer can be thought of as consisting of two types of domains, namely r-CO, +X-residues and r-CO, domains, including lateral interactions among domains. This description is consistent with the fact that the anodic stripping voltammograms of the composed adsorbates exhibit a first anodic peak (peak 1) resembling that of r-CO*, and a second one (Peak 2) approaching the anodic stripping peak of the X-residues. Nevertheless, it should be noted that for PA + r-CO, composed adsorbates peaks 1 and 2 are displaced with respect to those ones in the blanks, the former positively and the latter negatively. These potential shifts depend on E,,(X) and operate in the direction of increasing peak overlapping, in agreement with a relatively strong interaction between the adsorbates domains.
Otherwise, for AA + r-CO, composed adsorbates peak 1 is also shifted positively, but the potential of peak 2 remains practically constant, although the difference between the potentials of peaks 1 and 2 decreases as &(AA) increases. Certainly, for AA the overall effect is considerably smaller than for PA, as one should expect from the different chemical reactivity of these unsaturated alcohols.
In conclusion, the presence of either PA or AA residues on the r-CO, covered Pt electrodes also delays the initiation of the voltammetric electrooxidation processes. From the structural standpoint the anodic stripping voltammograms are consistent with an adsorbate adlattice probably consisting of two domains, one in which the alcohol residues interact directly with r-CO, adsorbates, and a second domain where r-CO, adsorbates keep to some extent their own characteristics although the global properties of domains become interdependent. Then, in the present case, the composed adsorbates can be probably described as mixed adlattices involving lateral modifications and adsorbate ensemble reorganisation, approaching the structural situations which were advanced a few years ago for other adsorbates on Pt electrodes [9].

CONCLUSIONS
The adsorption of unsaturated alcohols (AA, PA) dissolved in 0.5 M H2S04 on a polycrystalline Pt electrode modified by the presence of r-CO, adsorbates produces a potentiostatic current transient, which is either anodic or cathodic depending on the adsorption potential. The anodic stripping charge of the anodic residues is greater than that expected from the adsorption/electroadsorption of the independent substances. The anodic stripping voltammograms are also different from those corresponding to the blanks.