8792
Langmuir 2004, 20, 8792-8795
Effect of Structure on the Reduction Potentials of Films of Constitutional Isomers of Iron-Sulfur Cluster Core Dendrimers Tyson L. Chasse and Christopher B. Gorman* Department of Chemistry, North Carolina State University, Box 8204, Raleigh, North Carolina 27695-8204 Received May 21, 2004. In Final Form: July 19, 2004 The thermodynamic redox potentials of films of constitutional isomers of iron-sulfur cluster core dendrimers were measured and compared. It was determined that the primary structure of the dendrimer influences its reduction potential. Dendrimers containing so-called backfolded linkages were more difficult to reduce than their extended analogues. This behavior is rationalized by suggesting that the backfolded isomers pack more tightly around the iron-sulfur cluster, creating a more hydrophobic local microenvironment. Also, all of these molecules are easier to reduce in the film than in dimethyl formamide solution. The variation in redox potential between film and solution environment was compared to that of dendrimers of differing generations and correlated with the amount of hydrophobic dendron surrounding the cluster.
Electronic conduction through films is a key behavior in a variety of solid-state devices. These include batteries, solar cells (including the membrane-bound photosynthetic apparatuses), sensors, and emerging molecular electronics devices. This type of conduction generally appears to involve a series of discrete electron hopping events. Focusing specifically on electron hopping through molecular films, a number of investigations on polymer films,1,2 polymer melts,3-6 and dendrimers containing redox-active units7,8 have been undertaken to elucidate the parameters that govern rate and driving force through these films. We have recently shown that electroactive core dendrimers constitute a system in which the effects of molecular architecture on electron-transfer behavior can be studied.7 Dendrimers of varying generations changed the electrochemical potential of a redox-active, iron-sulfur core by ca. 500 mV when going from a second to a fourth generation dendrimer of the type shown in Figure 1. At the same time, the effective rate of electron transfer remained unchanged. More recently, we have shown that the rate and extent of electron transfer through films of iron-sulfur cluster core dendrimers is strongly influenced by the type of counterion available during reduction/ reoxidation. This behavior suggests that counterion permeation into these films is rate limiting.8 In this paper, we address the issue of how dendrimer architecture influences the redox potential of the ironsulfur cluster when in a film environment. Specifically, * To whom correspondence should be addressed. E-mail:
[email protected]. (1) Cameron, C. G.; MacLean, B. J.; Pickup, P. G. Macromol. Symp. 2003, 196, 165-171. (2) Pickup, P. G. J. Mater. Chem. 1999, 9, 1641-1653. (3) Sullivan, M. G.; Murray, R. W. J. Phys. Chem. 1994, 98, 43434351. (4) Sosnoff, C. S.; Sullivan, M.; Murray, R. W. J. Phys. Chem. 1994, 98, 13643-13650. (5) Long, J. W.; Kim, I. K.; Murray, R. W. J. Am. Chem. Soc. 1997, 119, 11511-11515. (6) Dickinson, E.; Williams, M. E.; Hendrickson, S. M.; Masui, H.; Murray, R. W. J. Am. Chem. Soc. 1999, 121, 613-616. (7) Gorman, C. B.; Smith, J. C. J. Am. Chem. Soc. 2000, 122, 93429343. (8) Chasse, T. L.; Smith, J. C.; Carroll, R. L.; Gorman, C. B. Langmuir 2004, 20, 3501-3503.
we compare the electronic behaviors of pairs of films consisting of constitutional isomers of iron-sulfur cluster core dendrimers (Figure 2). These molecules contain alternately 3,5-branched or 2,6-branched dibenzyloxy linkages which we will subsequently refer to as extended and backfolded, respectively. Previous studies9 of these isomer pairs in solution revealed that the backfolded isomers displayed a smaller, more compact architecture compared to their extended counterparts. This difference in molecular conformation led to slower electron-transfer kinetics and a larger negative reduction potential for these backfolded molecules compared to their extended isomeric analogues. Here, we probe how redox potential is biased by the dendrimer structure when in a film environment. To rationalize the redox potential of surface-confined iron-sulfur cluster dendrimers, the relative hydrophobicity of the microenvironment around the cluster appears to be the key matter at hand. In previous studies, the dianion [Fe4S4(SR)4]2- (R ) various aryl, alkyl, and dendritic groups) becomes more difficult to reduce (e.g., the reduction potential becomes more negative) when the environment (e.g., the R groups and/or the solvent) around it becomes less polar.10 This effect may be very important in tuning the redox potential of various ferredoxin proteins.11 In the types of dendrimer films studied here, the dendrons surrounding the iron-sulfur cluster are thought to be relatively nonpolar, particularly when compared to ionic and thus relatively polar electrolyte found within a thin film. Within this model, as the amount and/or density of packing of the dendrons around the core increases, the redox potential should decrease (e.g., become larger but more negative). Indeed, this is the general effect observed in these films. Films of each dendrimer on a platinum (Pt) electrode were prepared by drop coating a solution of each onto the (9) Chasse, T. L.; Sachdeva, R.; Li, Q.; Li, Z.; Petrie, R. J.; Gorman, C. B. J. Am. Chem. Soc. 2003, 125, 8250-8254. (10) Beinert, H.; Holm, R. H.; Mu¨nck, E. Science 1997, 277, 653659. (11) Bentrop, D.; Capozzi, F.; Luchinat, C. In Handbook of Metalloproteins; Bertini, I., Sigel, A., Sigel, H., Eds.; Marcel Dekker: New York, 2001; pp 357-460.
10.1021/la048733a CCC: $27.50 © 2004 American Chemical Society Published on Web 08/28/2004
Redox Potential of Films of Constitutional Isomers
Langmuir, Vol. 20, No. 20, 2004 8793
Figure 1. Structures of the iron-sulfur cluster core dendrimers considered in which the generation of the dendrimer was varied. Each dendrimer has the form (nBu4N)2[Fe4S4D4], where D indicates a dendron substituted with a focal aromatic thiol. For each molecule, four identical ligands are attached to the iron-sulfur core (denoted by a circled D).
surface followed by solvent evaporation as described in the Experimental Section. This protocol is similar to that employed previously by us7,8 as adapted from the work of Murray et al.5,6,12-14 Cyclic voltammetry was used to determine the redox potential for each dendrimer isomer. A mixture of 80%/20% v/v propylene glycol/propylene carbonate was used as an electrolyte solvent. This mixture was sufficiently polar to act as an electrolyte solvent. At the same time, this solvent mixture did not dissolve the dendrimer films, thus permitting this study. An increase in the volume fraction of propylene glycol in the solvent rendered the supporting electrolyte insoluble, while a decrease resulted in dissolution of the dendrimer film. Cyclic voltammograms (Figure 3) were obtained for each dendrimer isomer film. Their redox potentials are tabulated in Table 1 and compared with the redox potentials measured previously in dimethylformamide (DMF) solution.9 Keeping in mind the caveat associated with comparing electrochemical potentials in different media, two noticeable trends in redox potentials are noted. First, all of the dendrimer isomers were easier to reduce in the film form than in solution. To rationalize this behavior, these data were compared with those obtained previously for 4,4-bis(hydroxyphenyl)pentanol-based (Val) iron-sulfur core dendrimer solutions and films7 (Table 1). The molecule G2Val was also easier to reduce in a film than in solution.15 In contrast, films of G3Val were more difficult to reduce (e.g., they required a larger negative (12) Williams, M. E.; Masui, H.; Long, J. W.; Malik, J.; Murray, R. W. J. Am. Chem. Soc. 1997, 119, 1997-2005. (13) Poupart, M. W.; Velazquez, C. S.; Hassett, K.; Porat, Z.; Haas, O.; Terrill, R. H.; Murray, R. W. J. Am. Chem. Soc. 1994, 116, 11651166. (14) Vela´zquez, C. S.; Hutchison, J. E.; Murray, R. W. J. Am. Chem. Soc. 1993, 115, 7896-7897. (15) The solution redox potential of G2Val is lower than that of the dendrimer isomers, and this behavior is most probably due to the electron withdrawing para-amidophenyl linkage in the former compared to the electron donating para-alkoxyphenyl linkage in the latter. Here, we focus on the difference between film and solution rather than the difference in the electron donating or withdrawing nature of the group attached to the iron-sulfur cluster.
Table 1. Redox Potentials for the One-Electron Redox Couple [Fe4S4(S-Dend)4]2-/3- in Solution and in Films and the Corresponding Molecular Weights for Iron-Sulfur Core G2Val, G3Val,7 and Three Pairs of Constitutional Isomer Dendrimers molecule 2,2d 2B,2Bd 2,3d 2B,3d 3,2d 3,2Bd G2Vale G3Vale
redox potential (mV)a
molecular weight
filmb
solutionc
4229 4229 5090 5090 5518 5518 5995 11507
-1210 (2) -1256 (2) -1209 (1) -1241 (1) -1205 (1) -1238 (2) -1315 (13) -1515 (14)
-1465 (5) -1560 (1) -1460 (4) -1465 (7) -1440 (4) -1569 (7) -1366 (10) -1371 (19)
a Thermodynamic redox potential, E , observed vs ferrocene/ 1/2 ferrocinium external reference in the solvent systems specified below. Values in parentheses represent 90% confidence intervals of the average values found. b Film electrochemistry was carried out in 100 mM tetraethylammonium hexafluorophosphate in 80% propylene glycol/20% propylene carbonate for the constitutional isomers and 100% propylene carbonate for the valerate dendrimers. Values for the valerate dendrimers are from ref 7. c Solution electrochemistry was carried out in 100 mM tetraethylammonium tetrafluoroborate/dimethyl formamide. Values are from refs 9 and 7 and are adjusted when necessary so that all values in this table are reported against the same reference (Fc/Fc+). d Structure shown in Figure 2. e Structure shown in Figure 1.
potential) compared to G3Val in DMF solution. The structural difference between G2Val and G3Val is the greater number of more hydrophobic dendron repeat units around a charged (dianion) core. Thus, an increase in dendrimer generation makes the microenvironment within the dendrimer film progressively more nonpolar and the clusters within it progressively more difficult to reduce. Since the repeat units within the dendrimer isomers are different than those found in G2Val and G3Val, comparisons between them were made on the basis of molecular weight (Table 1). The dendrimer isomers have molecular weights similar to that of G2Val. Thus, we conclude that dendrimer isomer films contain a relative
8794
Langmuir, Vol. 20, No. 20, 2004
Chasse and Gorman
Figure 2. Structures of the iron-sulfur cluster core dendrimer isomer pairs. Each dendrimer has the form (nBu4N)2[Fe4S4D4], where D indicates a dendron substituted with a focal aromatic thiol. For each molecule, four identical ligands are attached to the iron-sulfur core (denoted by a circled D).
volume fraction of nonpolar dendron to polar cluster small enough to create a microenvironment within the film that is effectively more polar than DMF solution. The second notable correlation in these data comes if the redox potentials of constitutional isomer films are compared. In this case, each of the backfolded architectures displayed negative shifts in redox potential compared to their extended counterparts (Table 1). This trend was similar to that observed, although to a smaller degree, for each of the isomer pairs in solution.9 Previously, we determined that the backfolded architectures created a smaller, more compact dendrimer in solution. This bias in dendrimer conformation created a more effectively hydrophobic microenvironment surrounding the ironsulfur core. This difference in microenvironment would rationalize the relative reduction potentials between these isomer pairs. The negative shift in redox potential plausibly corresponds to some degree of “shape persistence” of the backfolded architectures compared to their extended counterparts when deposited as thin films. The observed molecular shape persistence is likely due to the greater steric congestion near the iron-sulfur cluster core in the backfolded architectures. Although it was of interest to compare the rate of charge transfer through films of dendrimer isomers, attempts to probe the kinetic behavior of the isomer thin films were unsuccessful and inappropriate for the following reasons. First, the amount of current that was passed decreased as the film was subjected to multiple reduction/reoxidation cycles. Attempts at determining the possible causes of
this behavior, such as altering film thicknesses and solvent concentrations or utilizing different electrochemical potential windows and scan rates, were unsuccessful. Thus it was deemed inappropriate to make any conclusions about the kinetics of charge transport from these films. In addition, it was realized that dendrimer isomers likely have different sizes when packed in films than they do in solution.9 Thus, it was concluded that any apparent kinetic differences between films could just as easily be due to uncertainty in the effective concentration of redox units within the films16 as to any real differences in rates of charge transport through the film. Experimental Section Dimethylformamide, propylene glycol, and propylene carbonate used in electrochemical experiments were anhydrous, purchased from Aldrich or Acros and stored under N2 atmosphere in a drybox. Electrolyte salts were recrystallized three times from ethanol and stored under N2 atmosphere in a drybox. The iron-sulfur core, constitutional isomer dendrimers ((nBu4N)2[Fe4S4(dend)4]) were prepared as described previously.9 (16) Assuming a cubic closest packing model of spheres, an effective concentration can be postulated and the Dahms-Ruff equation Dapp ) Dphys + kexδ2C/6 can be used to determine the intermolecular selfexchange rate, kex, through a film (where C is the concentration of redox units within the film, and δ is the distance between redox sites). However, given that films of dendrimer isomers likely have different packing/ organization, significant differences in C and δ between them would be expected. Thus, as Dapp does not distinguish between hopping rate and film organization, it is not appropriate to extract and compare values for kex for films composed of molecules with such different sizes.
Redox Potential of Films of Constitutional Isomers
Figure 3. Cyclic voltammograms of each constitutional isomer as a thin film (dotted) and in solution (solid). Given the uncertainty in film thickness and charge transport dynamics through it, the magnitude of the current in film measurements is not meaningful. All thin films were drop-coated onto a Pt button by a procedure analogous to that described previously by Murray et al.5 Thin film deposition was sensitive to small fluctuations in dendrimer concentration as well as droplet size. As a result, determining an ideal deposition protocol was required to create a dendrimer film that maximizes current passed while achieving the greatest stability. Films that were excessively thin displayed a maximum initial current followed by a rapid current decay during the next few cyclic voltammetry scans. Conversely, films that were excessively thick passed insufficient current for analysis. Ideal film thicknesses were determined empirically by comparing the
Langmuir, Vol. 20, No. 20, 2004 8795 cyclic voltammograms of multiple depositions using one, two, or five 5 µL drops of concentration 0.5, 1.0, or 2.0 mM dendrimer in THF. The optimal conditions were determined to be coating with one 5 µL drop of 2 mM dendrimer/THF solution. Following deposition, the films were allowed to dry for approximately 10 min. Pt button working electrodes were polished, conditioned, and referenced to a dendrimer standard, previously studied within our group,17 prior to and following thin film deposition. This procedure was performed to ensure that the electrode was not passivated by any undesired components during the thin film electrochemical experiments. Electrochemical experiments were carried out on a Bioanalytical Systems CV-50W voltammetric analyzer. Cyclic voltammetry measurements were carried out in a three-electrode cell consisting of a platinum disk working electrode with a geometric area of 0.0201 cm2, a Pt auxiliary electrode, and a homemade nonaqueous Ag/AgNO3 reference electrode (Ag wire contacting a MeCN solution of 0.01 M AgNO3 and 0.1 M supporting electrolyte, TEAPF6). All redox potentials were referenced to the ferrocene/ferrocinium redox couple by adding ferrocene to the solution after the experiment was complete. All electrochemical experiments were carried out in a nitrogen-filled Vacuum Atmospheres drybox at room temperature. The Pt working electrode coated with the chosen dendrimer sample was placed into a three-electrode cell containing a solution of TEAPF6 (100 mM) in 80% propylene glycol/20% propylene carbonate. Each experiment was repeated six times for each dendrimer sample at scan rates of 5, 10, 20, 50, and 100 mV/s on two preconditioned Pt electrodes. In each study, voltammetric precycling of approximately 5-10 scans was required to obtain a stable response. Precycling was done for each thin film until a stable response was reached whereupon data were collected. No dissolution of the films was visibly evident due to the molecules being highly colored and obvious when solvated.
Acknowledgment. We thank Jennifer Smith for advice on film preparation and Drew Wassel and Ryan Fuierer for a critical reading of the manuscript. This work was supported in part by the National Science Foundation (CHE-9900072 and CHE-0315311) and a Graduate Assistance in Areas of National Need (GAANN) Electronic Materials fellowship. LA048733A (17) Gorman, C. B.; Smith, J. C.; Hager, M. W.; Parkhurst, B. L.; Sierzputowska-Gracz, H.; Haney, C. A. J. Am. Chem. Soc. 1999, 121, 9958-9966.
