Increased valence theory and the four electron reduction of
O2 to H2O
Citation for published version (APA):
Putten, van der, A. M. T. P., Elzing, A., Visscher, W., Barendrecht, E., & Harcourt, R. D. (1988). Increased valence theory and the four electron reduction of O2 to H2O. Journal of Molecular Structure: THEOCHEM, 49, 309-318. https://doi.org/10.1016/0166-1280(88)80097-6
DOI:
10.1016/0166-1280(88)80097-6
Document status and date: Published: 01/01/1988
Document Version:
Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers)
Please check the document version of this publication:
• A submitted manuscript is the version of the article upon submission and before peer-review. There can be important differences between the submitted version and the official published version of record. People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher's website.
• The final author version and the galley proof are versions of the publication after peer review.
• The final published version features the final layout of the paper including the volume, issue and page numbers.
Link to publication
General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
• Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain
• You may freely distribute the URL identifying the publication in the public portal.
If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, please follow below link for the End User Agreement:
www.tue.nl/taverne Take down policy
If you believe that this document breaches copyright please contact us at: openaccess@tue.nl
providing details and we will investigate your claim.
Journal of Molecular Structure (Theochem), 186 (1988) 309-318
Elsevier Science Publishers B.V., Amsterdam - Printed in The Netherlands
309
INCREASED VALENCE THEORY AND THE FOUR ELECTRON REDUCTION OF O2 TO Hz0
A. VAN DER PUTTEN, A. ELZING, W. VISSCHER and E. BARENDRECHT
Laboratory for Electrochemistry, Department of Chemical Technology, Eindhoven University of Technology, P.O. Box 513,560O MB Eindhoven (The Netherlands)
R.D. HARCOURT
Department of Physical Chemistry, University of Melbourne, Parkville, Victoria 3052 (Australia)
(Received 23 November 1987)
ABSTRACT
The four electron reduction of oxygen to water, without the intermediate formation of hydrogen peroxide has been reported on dicofacial dicobalt porphyrins and planar dicobaltchelates, con- taining two CON, units in the plane of the molecules. These two classes of compounds display “reversed selectivity” with respect to the solution pH. Dicofacial dicobalt porphyrins directly reduce O2 to Hz0 in acid solution, and to hydrogen peroxide in alkaline solution. Planar dico- baltchelates show four electron reduction of 0, in alkaline solution, but yield exclusively hydrogen peroxide at low pH.
By using the increased valence theory, an attempt is made to provide a valence bond repreeen- tation for the electronic reorganization that could be associated with each of these phenomena. For the results obtained in acid solution, this theory indeed offers an excellent description. In alkaline solution, however, application of the increasedvalence theory alone does not fully account for the experimental observations, unless the following condition is also satisfied. Besides bridge adsorption of oxygen, appropriate values for the redox potentials of the Co centres also seem to be a prerequisite for the occurrence of direct reduction of oxygen to water.
INTRODUCTION
The electrochemical reduction of 0, has been extensively studied [ 11, not only for a better understanding of the way nature reduces 02, but also for ap- plication in oxygen electrodes for fuel cells and metal-air batteries. These studies have shown that O2 can be reduced in two ways:
(i) 4e -reduction of 0, to Hz0 without hydrogen peroxide as an interme- diate, the so-called direct reduction.
(ii) 2e -reduction of O2 to H,Oz which can be the stable end product or can be subsequently reduced to H,O.
Discrimination between the two pathways is possible with the use of the
a
C
b
d
.2 “20
Fig. 1. Molecular structure of some dicobalt chelates: (a) face-to-face Co-Co-4 porphyrin; (b) 1,8- anthryl dicobalt porphyrin; (c) CozTAPH(N03),; (d) Co,(dpt)Cl,.
rotating ring-disc electrode technique [ 21. At the disc the reduction of 0, is monitored as a function of the applied potential and the rotation frequency of the electrode, while the production of hydrogen peroxide is detected at the ring electrode. Wroblowa et al. [ 31 developed a criterion to determine whether or not direct reduction is taking place. Studies of the O2 reduction on noble metals like Au and Pt have shown that this selectivity is determined by the structure of the 0, adsorption on the catalyst [ 41. If the O2 molecule interacts with only one metal atom (end-on or side-on adsorption), then (at least initially) H,O, is produced (Au); if the O2 molecule is adsorbed at two metal atoms (bridge adsorption), direct 4e - reduction to H,O becomes possible (Pt). This model is corroborated by recent observations, such as the role of interatomic Pt-Pt spacing [5] and the effect of underpotential deposited layers (UPD) of Pb
[6], Tl [7] or Bi [8] on Au. In the first case, by alloying pure Pt with foreign metals, an optimum Pt-Pt distance for bridge adsorption of the O2 was ob- tained; in the second case, the presence of an adsorbed foreign metal on an Au single crystal surface resulted in the direct reduction of 0, to H,O, contrary to the pure Au electrode, which yields H,O, exclusively. This change in selectivity is explained with a bimetal bridge adsorption, i.e., the formation of Me-O-O- Au adducts.
The importance of bridge adsorption on transition metal complexes was shown by Collman et al. [9,10]. They synthesized dicofacial dicobalt porphyr- ins (Fig. 1 (a) ), two Co porphyrin units linked such that the planar macrocy- cles are held in a face-to-face orientation. By comparing complexes with
311
different interplanar distances, it was observed that direct reduction_ to Hz0 only occurred if the two Co centres were located at a distance of ca. 4 A, which apparently enables bridge adsorption of oxygen. Other dicofacial dicobalt por- phyrins were synthesized like the anthracene bridged dicobalt diporphyrin of Liu and co-workers [ 111 (Fig. 1 (b) ), giving identical results to the Collman
complexes. Direct reduction to H,O was observed in acid solution; in alkaline solution only H,Oz was formed.
This bridge adsorption should also be possible on a planar dicobalt chelate, containing two CON, centres in the plane of the molecule. Indeed Sarangapani
[ 121 and Yeager [ 131 reported the reduction of 0, to H,O on Co,TAPH (NO,),
(Fig. 1 (c) ). In alkaline solution a mixed production of HZ0 and HzOz was
obtained. Evaluation of their data [ 12,131 with the aid of the Wroblowa cri-
terion [ 31 showed that besides the peroxide pathway, direct reduction of O2 to
H,O is also occurring. We obtained similar results [ 151 with Cozdpt,C1, (Fig.
1 (d) ). In acid solution O2 is reduced to HzOz exclusively. The cofacial com- plexes yield virtually no H,O, in acid solution whereas the planar chelates give considerable amounts in alkaline solution; this is caused by the difference in activity of the corresponding monomeric 0, adducts in acid and alkaline so- lution [15]. If O2 is only attached to one Co centre, the reduction to H202 probably proceeds via 0; as an intermediate [ 16,171. The driving force for
this pathway is much greater in alkaline than in acid solution [ 151. In other
words, in alkaline solution the direct pathway and the reduction to H,O, pro- ceed at comparable rates with low overpotential; in acid solution only the direct reduction can proceed at low overpotential. For the reduction to H,O, a much higher overpotential is necessary.
Summarizing, one can conclude that dicofacial dicobalt porphyrins reduce O2 directly to H,O in acid solution and to H,O, in alkaline solution. Planar dicobalt chelates show 4e - reduction of O2 to H,O in alkaline solution and reduction to Hz02 in acid solution. This reversed behaviour with respect to the solution pH has not been recognized so far. One of the theoretical concepts used to explain the behaviour of transition metal macrocycles in the presence of oxygen, is the increased valence theory, developed by Harcourt [l&19,21-
23]. This theory has been able to explain the observation that ,u-0x0 adducts are formed in the case of water-soluble iron porphyrin but not with the corre- sponding Co compound [ 191. Moreover, it has provided a model for the O2
reduction on cytochrome c oxidase and the formation of a diamagnetic oxygen adduct in the case of haemoglobin [ 231. Therefore we decided to apply the
increased valence theory to the O2 reduction behaviour of the different dicobalt chelates in order to obtain a better understanding of the observed phenomena.
The basic concepts of the increased valence theory are given in refs. 18-23. Here we will only summarize the results for oxygen adducts.
INCREASED VALENCE THEORY AND OXYGEN ADDUCTS
For the description of the chemical binding between two atoms, the concept introduced by Lewis in 1916 is often used [ 201. The binding occurs via spin
pairing into a so-called “shared pair of electrons”. The Lewis concept does not mean that only electron pair binding can occur; one-electron bonds are also possible as in the He,+ ion [ 211. In some cases (for instance the 0, molecule)
two one-electron bonds are energetically even more favourable than an elec- tron pair bond, i.e. the valence bond structure for the ground state of 0, is [ 201
:p+: rather than $=c$
In molecular orbital (MO) theory this can be represented as
The one-electron bond is also essential for the increased valence theory: in a system consisting of three atoms and four electrons (according to the Lewis concept, a shared pair of electrons and a lone pair), delocalizations are possi- ble, leading to a structure with a fractional two-electron bond and a one-elec- tron bond [18,22,23]. Often, a driving force for this delocalization is the reduction of the formal charges on the atoms; the result is a structure with a higher number of bonding electrons, thereby increasing the valence of the in- volved atoms. The increased valence theory originated from studies of N,H, and N204 [ 22 1, explaining that the N-N distance is much longer in N204 than
in N,H,; moreover, that the N-O distance is even somewhat smaller in Nz04 than in HNO which contains four bonding electrons. According to standard Lewis valence bond formulae, the N-O bond in N,O, should contain an average of three bonding electrons.
The increased valence theory has already been used to explain experimental observations regarding O2 adducts of haemoglobin [ 231, a number of dissolved
Fe and Co porphyrins and phthalocyanines [ 191, and the reduction of 0, on
cytochrome c oxidase [ 231. In haemoglobin, each Fe2+ is surrounded by five
nitrogen atoms, four of the porphyrin itself and one of a histidine group which axially coordinates to the Fe2+. In the ground state the Fe2+ has four unpaired electrons (S = 2 ) and therefore the molecule is paramagnetic. The O2 adduct
however is diamagnetic (S = 0). To account for this, the Fe (II) (S = 2 ) is pro- moted to the intermediate spin state Fe (II) (S = 1) which reacts via spin-pair- ing with ground state O2 to afford increased valence structure (2 ) .
313
(For simplicity, the organic ligands in the increased valence structures ( 11) - ( 14)) are not drawn: they lie in the plane that is perpendicular to the plane of the paper.)
Another interesting reaction is the formation of p-ox0 complexes of Fe (II) porphyrins in Oz saturated solutions [19]. These flu-0x0 complexes have not been observed with Co(I1) porphyrins. Increased valence structure (2) reacts with a second intermediate spin Fe( II) (S= 1) to generate structure (3). In (3 ), two O-O bonding electrons delocalize, together with four Fe and O-O
non-bonding electrons, to afford (4), in which the O-O bond number is even less than unity. O-O bond breaking occurs and the Fe (II)0 species formed reacts with Fe(I1) (S=l) to givep-0x0 species (5).
l 'O ;;p:’ :;a. .g. fe: Fe2Z i+ '*\
l
0** I l O* 2+ .b' . . l y:- ;0/,:' K 4 Fe . .'Fe': %e* . I . . . . .;;. . . *Fe: (51 ;3; ;&-IThe corresponding MO scheme for the Fe (II)0 intermediate can be repre- sented as
Fe11110 .t-:
The similarity between the MO scheme of Fe(II)O and O2 is striking: the Fe (II) 0 species therefore will be relatively stable. For 0, bridged dimeric co- balt complexes a similar reaction scheme can be set up leading to structure (6). The same delocalizations as in structure (3 ) however cannot take place. The only way of loosening the O-O bond is by the construction of (7)
j ‘i;l: I.+.[;: .r+~O.: _ :o -.A
..;b:t-;,
203
+.O I- . :co.l.fl . . . . I61 I71the number of binding electrons does not increase. Therefore, the electronic reorganization indicated in (6) should proceed only to a small extent, and the O-O bond will not split; indeed no p-0x0 Co dimers have been observed experi- mentally in solution.
The reduction of Oz as performed in nature by cytochrome c oxidase has also been described [23]. It has a high resemblance to the O-O bond breaking at the Fe-O-O-Fe adduct except that one Fe (II) is replaced by a Cu (I)
(01 (91
After the bond breaking both species finally react with two protons and one electron to produce two HzO, Fe (III) and Cu( II). The latter two species are regenerated with two electrons to the initial Fe (II) and Cu (I).
APPLICATION OF THE INCREASED VALENCE THEORY TO O2 REDUCTION ON DI- MERIC 0, ADDUCTS
For the above described increased valence formulae, the solution pH is not taken into consideration. These structures, however, are not protonated and can therefore be used to explain the 0, reduction results, obtained in alkaline solution. Since structure (6) does not allow splitting of the O-O bond, it is not surprising that the dicofacial dicobalt porphyrins yield only H,O, in alkaline solution. In acid solution, the 0, adduct must be protonated according to the mechanism of Collman et al. [lo]. The relatively increased electronegativity of an 0 atom in
(10)
allows the delocalization of a cobalt non-bonding elec- tron into a Co-O bonding orbital1101 (11)
In species (
11))
the O-O bond is weakened. This bond is now able to split, giving reduction to water according to the schemeH>,,.+ ,.,+
Hv,@ :o e- - < ‘I. - :[o.l.fl 1 -:Co* . . *Hz0 :co* I4 . . . .315
The Co (III), at least initially formed as intermediate spin S= 1, can be re- generated to the initial Co (II) (S= f ); perhaps even diprotonation is possible, leading to ( 12 ) . The bond number for the O-O a-bond will be smaller than in structure (
11)
Q:
.
.
.co:+
*I.
H I+) \d &,., - H$j%l~ H -K I> :$: :co.+ I* . . 1121 1131As one can see, the increased valence theory nicely describes how the O-O bond breaking is taking place on these complexes in acid solutions: the delo- calization of a Co non-bonding electron in a Co-O bonding orbital loosens the O-O bond. Moreover, the model explains why the Oz adduct has to be proton- ated to accomplish splitting of the O-O bond.
Since protonation is unlikely in alkaline solution, the 0, adduct of the planar dicobalt chelates in this electrolyte can be represented initially by resonance between the increased-valence structures (
14a)
and ( 14b), each of which has a fractional Co-O a-bond and a fractional Co-O x-bond. There would then be a strong tendency to form two Co-O a-bonds, as occurs in (15) with a strained peroxide linkage. This would proceed via the n-electron shifts that are indi- cated in (14a) and(14b),
and would be accompanied by orbital rehybridi- zation for both cobalt and oxygen1141) ll4bl 1151
Based on similar arguments as for structure (7)) this species should not be able to break the O-O bond in the solution phase: although the O-O bond in ( 15) is somewhat strained as compared to structure (7), the delocalization of a Co non-bonding electron into a Co-O bonding orbital still involves formal charge separation.
Moreover, if the destabilization of the O-O bond in (15) would already be enough to split the O-O bond in alkaline solution, one would certainly expect this to happen in acid solution via equivalent structures of ( 11) and ( 13 ). This, however, is not observed in practice [ 141.
These considerations show that application of the increased valence theory alone, does not lead to a consistent and complete explanation of all the ob- served experimental results. As already suggested in a previous paper [ 141, we think that an important precondition for the occurrence of direct reduction of O2 is that the redox potentials of the metal ions have appropriate values.
centres of their decofacial dicobalt porphyrins in acid solution are closely re- lated to the 0, reduction onset potential (ca. 0.9 V vs. a reversible hydrogen electrode). This indicates that the redox potentials have the appropriate val- ues in this case. Since Fe”‘/Fe” porphyrin redox potentials in general have lower values than Co”‘/Co”, the redox potentials of the dicofacial diiron che- lates are probably too low. In fact at low overpotential for 0, reduction, the iron centres are still in the Fe( III) state and will therefore have only low in- teraction with oxygen: no improved performance is obtained as compared to the monomeric Fe porphyrin.
If we now go to alkaline solution, the equilbirium potential for the O2 reduc- tion with respect to the reversible hydrogen reference electrode will not shift since both redox systems have the same pH dependence [ 151. If the potentials of the CO”‘/CO’~ redox couples are pH independent, as with CoPc [ 151, then they will be too high (ca. 1.5 V vs. RHE at pH 14) in alkaline solution. Of course, the same applies to the planar dicobalt chelates. Since the Co”‘/Co” redox couples therefore probably have values in alkaline solution which are too high, a possible explanation would be that in that case the Co”/Co’ couples are involved. The splitting of the O-O bond in the case of the planar dicobalt chelates can then be visualized as
In passing from the Con to the Co’ state the extra negative charge, imposed on the cobalt centres (indicated by the arrows), assists in stabilizing structures
( 16) and (17) by compensating the positive charge. This facilitates the de- localization of a cobalt non-bonding electron into a Co-O bonding orbital, thus leading to the splitting of the O-O bond.
CONCLUDING REMARKS
By comparing structures (6) and ( 15) it is clear that 0, adsorption in a cis configuration is advantageous for the destabilization of the O-O bond. The O2 valence state is
. . . .
:?-o: . rather than :‘d- j,: .
(18) 119)
In ( 18), the interelectronic repulsion will be higher than in (19), weaken- ing the O-O bond. Perhaps this phenomenon also explains the results of Liu et al. [ 241 who synthesized a dicofacial dicobalt porphyrin onto which the O2 had to be adsorbed in a cis configuration. These authors attributed the im-
317 proved activity of the latter molecule to a better accessibility of protons to the adsorbed O2 molecule.
Although dicofacial diiron porphyrins did not give improved results in acid solution, this could change in alkaline solutions. In acid solution their redox potentials are too low: they should have appropriate values, i.e. in the vicinity of 0.9 V vs. RHE. Our own results with FePc [ 151 show that this molecule has a Fe”‘/Fe” redox peak at 0.85 V vs. RHE at pH 14. The behaviour of dicofacial diiron phthalocyanine would therefore be interesting, since this molecule could fulfil both preconditions for the occurrence of direct reduction in alkaline so- lution. Moreover, since this redox peak is pH dependent from pH 6-14, the catalyst could also be active in this entire pH range.
ACKNOWLEDGEMENTS
The present investigations were carried out with the support of the Neth- erlands Foundation for Chemical Research (S.O.N.) and with financial aid from the Netherlands Organization for the Advancement of Pure Research
(Z.W.O.).
REFERENCES
1 M. Tarasevich, A. Sadkowski and E. Yeager, Oxygen electrochemistry, in B. Conway, J. Bockris, E. Yeager, S. Khan and R. White (Eds.), Comprehensive Treatise of Electrochem- istry, Vol. 7, Kinetics and Mechanisms of Electrode Processes. Plenum Press, New York, 1983, Chapter 6, pp. 301-398.
2 Yu.V. Pleskov and V.Yu. Filinovskii, The Rotating Disc Electrode, Consultants Bureau, New York/London, 1976.
3 H.S. Wroblowa, Y.C. Plan and G. Razumney, J. Electronal. Chem., 69 (1976) 195. 4 P. Fischer and J. Heitbaum, J. Electronal. Chem., 112 (1980) 231.
5 V. Jalan and E.J. Taylor, J. Electrochem. Sot., 130 (1983) 2299.
6 K. Jiittner, Electrochim. Acta, 29 (1984) 1597.
7 R. Amadelli, N. Markovic, R. Adzic and E. Yeager, J. Electroanal. Chem., 159 (1983) 391.
8 SM. Sayed and K. Jiittner, Electrochim. Acta, 28 (1983) 1635.
9 J.P. Collman, M. Marocco, P. Denisevich, C. Koval and F.C. Anson, J. Electroanal. Chem., 101 (1979) 117.
10 J.P. Collman, P. Denisevich, Y. Konai, M. Marocco, C. Koval and F.C. Anson, J. Am. Chem. Sot., 102 (1980) 6027.
11 C.K. Chang, H.Y. Liu and I. Abdalmuhdi, J. Am. Chem. Sot., 106 (1984) 2725; H.Y. Liu, I. Abdalmuhdi, C.K. Chang and F.C. Anson, J. Phys. Chem., 89 (1985) 665.
12 S. Sarangapani, Ph.D. Thesis, Case Western Reserve University, Cleveland, 1983.
13 E. Yeager, Electrochim. Acta, 29 (1984) 1527.
14 A. van der Putten, A. Elzing, W. Visscher and E. Barendrecht. J. Chem. Sot. Chem. Com-
mun., (1986) 477.
15 A. van der Putten, A. Elzing, W. Visscher and E. Barendrecht, J. Electronal. Chem., 221 (1987) 95.
17 E. Yeager, Electrochim. Acta, 29 (1984) 1527.
18 R.D. Harcourt, J. Chem. Educ., 45 (1968) 779; 46 (1969) 856.
19 R.D. Harcourt, J. Inorg. Nucl. Chem., 39 (1977) 243 and references cited therein. 20 J.W. Linnett, The Electronic Structures of Molecules, Methuen, London, 1964. 21 R.D. Harcourt, J. Chem. Educ., 62 (1985) 99.
22 R.D. Harcourt, J. Mol. Struct., 9 (1971) 221; J. Am. Chem. Sot., 102 (1980) 5195,103 (1981) 5623; ref. 23, Chaps. 7,11 and 13.
23 R.D. Harcourt, Qualitative valence bond descriptions of electron-rich molecules: Pauling “3- electron bonds” and “increased valence” theory, in G. Berthier, M.J.S. Dewar, H. Fischer, F. Fukui, G.G. Hall, H. Hartmann, H.H. Jaffd, J. Jortner, W. Kutzelnigg, K. Ruedenberg and E. Scrocco (Eds. ) , Lecture Notes in Chemistry, Springer Verlag, Berlin, 1982.
