Article pubs.acs.org/accounts
Folding of ortho-Phenylenes C. Scott Hartley* Department of Chemistry & Biochemistry, Miami University, Oxford, Ohio 45056, United States S Supporting Information *
CONSPECTUS: In nature, the folding of oligomers and polymers is used to generate complex three-dimensional structures, yielding macromolecules with diverse functions in catalysis, recognition, transport, and charge- and energy-transfer. Over the past 20−30 years, chemists have sought to replicate this strategy by developing new foldamers: oligomers that fold into well-defined secondary structures in solution. A wide array of abiotic foldamers have been developed, ranging from non-natural peptides to aromatics. The ortho-phenylenes represent a recent addition to the family of aromatic foldamers. Despite their structural simplicity (chains of benzenes connected at the ortho positions), it was not until 2010 that systematic studies of ophenylenes showed that they reliably fold into helices in solution (and in the solid state). This conformational behavior is of fundamental interest: o-Arylene and o-heteroarylene structures are found embedded within many other systems, part of an emerging interest in sterically congested polyphenylenes. Further, o-phenylenes are increasingly straightforward to synthesize because of continuing developments in arene−arene coupling, the Asao−Yamamoto benzannulation, and benzyne polymerization. In this Account, we discuss the folding of o-phenylenes with emphasis on features that make them unique among aromatic foldamers. Interconversion between their different backbone conformers is slow on the NMR time scale around room temperature. The 1H NMR spectra of oligomers can therefore be deconvoluted to give sets of chemical shifts for different folding states. The chemical shifts are both highly sensitive to conformation and readily predicted using ab initio methods, affording critical information about the conformational distribution. The picture that emerges is that o-phenylenes fold into helices with offset stacking between every third repeat unit. In general, misfolding occurs primarily at the oligomer termini (i.e., “frayed ends”). Because of their structural simplicity, the folding can be described by straightforward models. The overall population can be divided into two enantiomeric pools, with racemization and misfolding as two distinct processes. Examination of substituent effects on folding reveals that the determinant of the relative stability of different conformers is (offset) aromatic stacking interactions parallel to the helical axis. That is, the folding of ophenylenes is analogous to that of α-helices, with aromatic stacking in place of hydrogen bonding. The folding propensity can be tuned using well-known substituent effects on aromatic stacking, with moderate electron-withdrawing substituents giving nearly perfect folding. The combination of a simple folding mechanism and readily characterized conformational populations makes ophenylenes attractive structural motifs for incorporation into more-complex architectures, an important part of the next phase of foldamer research.
■
INTRODUCTION Polyphenylenes, compounds composed primarily of directly connected arenes, are now well-established in materials chemistry, valued for their inherent conjugation, chemical stability, and straightforward syntheses.1 The simplest polyphenylenes are the three linear (unbranched) architectures derived from the para, meta, and ortho connectivity of a disubstituted arene. Of the three, para-phenylenes are by far the most thoroughly investigated, especially as conjugated polymers.2 meta-Phenylenes are less well studied, but have been used as helical polymers (and oligomers)3 and as cross-conjugated systems.4 In contrast, only a few examples of ortho-phenylenes were known until recently,5−11 presumably because they were comparatively difficult to synthesize and unlikely to be of interest as conjugated materials. In 1998, Simpkins pointed out that o-phenylenes could potentially fold into helices.12 This hypothesis was supported by © XXXX American Chemical Society
a crystal structure of an o-phenylene hexamer that adopts a helical conformation in the solid state. The o-phenylenes therefore had potential as a class of foldamers: oligomers that fold into welldefined secondary structures because of noncovalent interactions between repeat units.13−15 Foldamers have long been of interest because of their relationship to folded biological molecules, especially peptides. It is hoped that we will achieve the activity and specificity of functional biomacromolecules, ultimately complementing these systems with equally sophisticated but chemically orthogonal structures. The folding of o-phenylenes is of particular interest given their structural simplicity and the possibility that they could be used to better understand the conformational behavior of polyphenylenes in general. However, for 10 years following Simpkins’s Received: January 20, 2016
A
DOI: 10.1021/acs.accounts.6b00038 Acc. Chem. Res. XXXX, XXX, XXX−XXX
Article
Accounts of Chemical Research
oligomer length (reaching Φf = 0.18 for oP8(H)8).21 Unusually, the fluorescence spectra of o-phenylenes give a hypsochromic shift as n increases (Figure 1). An interesting consequence of this is that the fluorescence behavior of the isomeric linear polyphenylenes is quite different: with increasing length, pphenylenes exhibit a pronounced shift to the red (as expected for conjugated oligomers), m-phenylenes exhibit no shift at all (because they are cross-conjugated), and o-phenylenes exhibit a shift to the blue.24 The origin of this effect is discussed further below.
original report the field lay largely dormant.16 This Account focuses on recent investigations of the solution-phase folding of o-phenylenes, focusing mostly on the oPn(X)n series of [n]-mers shown in Scheme 1. They do indeed fold into helical structures, Scheme 1. o-Phenylene Oligomers Discussed in this Account
■
o-PHENYLENES AS FOLDAMERS
Conformational Analysis of Coupled 2,2′-Biaryls
The backbone conformation of an o-phenylene [n]-mer will be dictated by the n−3 internal torsional angles φi, as shown in Figure 2. Much of the folding of o-phenylenes can be understood
but beyond this simple fact they exhibit a rich conformational behavior that allows much to be understood about their folding and their potential for use in complex foldamer architectures.
■
BASIC PROPERTIES OF o-PHENYLENES Studies of systematic series of o-phenylenes began to appear in 2010 and 2011, with reports from us17 and Fukushima and Aida.18 Not surprisingly, synthesis of o-phenylene oligomers has principally been done using aryl−aryl coupling methods, especially those that tolerate sterically hindered substrates. Our work has primarily made use of Buchwald’s active Suzuki− Miyaura coupling catalysts,19 whereas Fukushima and Aida have used copper-mediated coupling of aryl lithiums. In general, ophenylenes are easy to work with; for example, they exhibit much better solubility than analogous p-phenylenes.20 Representative UV−vis and fluorescence spectra for the parent series oPn(H)n are shown in Figure 1.21 Overall shifts in the UV−
Figure 2. Key torsional angles defining the backbone conformation of an o-phenylene.
beginning with consideration of the behavior of 2,2′disubstituted biaryls, shown in the inset to Figure 3. A 2,2′disubstituted biaryl will have four distinct conformers defined by φ. There are two sets of enantiomers: in one, the two substituents are syn, with φ = ± α, and in one they are anti, with φ = ± β. Of course, the number of possible conformers increases exponentially when many biaryl units are coupled together to give an o-phenylene. For the disubstituted o-terphenyl system, shown in Figure 3, there are 16 possible conformations resulting from the different combinations of ±α and ±β.25 However, the two biaryl bonds in an o-terphenyl system are conformationally coupled, which greatly simplifies the system. For typical ophenylenes, the values of α and β are found to be approximately 55° and 130°, respectively (see below). Consider then the behavior of oP5(H)5, shown in Figure 4. If one dihedral is held fixed at φfix = −α (−55°), minima on the conformational energy surface are only obtained for the other dihedral when φvar = −α or +β. Similarly, if φfix = +β (+130°), minima are also only obtained if φvar = −α or +β. That is, the configuration of φfix dictates the allowed values of φvar, with only (−α/+β)/(−α/+β) or (+α/−β)/(+α/−β) pairings allowed. The stable conformers correspond to the conformers with quasi-parallel rings (Figure 3). This relationship has important consequences when extrapolated along an o-phenylene of arbitrary length. For a single molecule of an idealized o-phenylene, we expect all of the key torsional angles to be restricted to two values, either −α/+β or +α/−β. Thus, the folding of an o-phenylene backbone can be described using a simple binary notation. We typically define A = −α and B = +β (and analogously A′ = +α and B′ = −β). Thus, for an o-phenylene octamer, the AAAAA (A5) conformer is a compact left-handed helix with (offset) stacking between every fourth repeat unit, as shown in Figure 5. The BBBBB (B5) conformer is an extended right-handed helix. There will, of course, be a large number of misfolded conformers that must also be considered (e.g., AAAAB).
Figure 1. UV−vis (solid) and normalized fluorescence (dashed) spectra of oPn(H)n (cyclohexane). Adapted with permission from ref 21. Copyright 2011 American Chemical Society.
vis spectra of o-phenylene oligomers of increasing length are small. The effective conjugation length in the parent series is short (necl = 4), although it can be much longer for simple substituted o-phenylenes (e.g., necl = 8 for oPn(OMe)n).17 In systems with terminal dimethylamino−nitro-substitution (push−pull), the charge-transfer interaction is absent for n > 3.22 Taken together, these observations are consistent with the expectation that delocalization in o-phenylenes is attenuated by twisting.23 The parent o-phenylenes (except for the trimer) are moderately fluorescent, with quantum yields increasing with B
DOI: 10.1021/acs.accounts.6b00038 Acc. Chem. Res. XXXX, XXX, XXX−XXX
Article
Accounts of Chemical Research
Figure 3. Pseudo-Newman projections for rotation around the biaryl bonds in a substituted o-terphenyl. Inset: Basic conformational behavior of 2,2′disubstituted biphenyl.
Figure 5. Representative conformers of octa(o-phenylene). Adapted with permission from ref 32. Copyright 2014 American Chemical Society.
Figure 4. Coupled rotation in oP5(H)5. Energies were calculated at the B97-D/TZV(2d,2p) level (see the Supporting Information).
The overall conformational population of an o-arylene can therefore be divided into two enantiomeric pools, A/B and A′/ B′. We can consider two separate dynamic processes for these structures: misfolding (AA···A ⇌ AA···B ⇌ etc. ⇌ BB···B) and racemization (AA···A/AA···B/etc. ⇌ A′A′···A′/A′A′···B′/etc.). Inspection of Figure 4 suggests that the energetic difference between the A and B states should be relatively small in simple ophenylenes and related systems. Further, ideal o-phenylenes will not naturally incorporate defects leading to racemization. Racemization will require either a concerted motion of the full oligomer or nonideal behavior (i.e., not conforming to the analysis above), such as the propagation of a defect that allows the coexistence of B/B′ (or A/A′) states.26,27 The rate of racemization should therefore be strongly length-dependent, and
should be slow, in general, relative to misfolding, although the precise mechanism remains an open question. Experimental Determination of o-Phenylene Folding
When synthesizing a series of o-phenylene [n]-mers, it quickly becomes apparent that the different backbone conformers are in slow exchange on the NMR time scale for n ≥ 5 (around room temperature).21 The 1H NMR spectrum of hexa(o-phenylene) oP6(H)6 is shown in Figure 6 (top) as a representative example. With fast exchange, the 2-fold-symmetric oP6(H)6 would exhibit 11 total signals; instead, the spectrum comprises many overlapping signals from three separate conformers. Two are symmetric; one is asymmetric. C
DOI: 10.1021/acs.accounts.6b00038 Acc. Chem. Res. XXXX, XXX, XXX−XXX
Article
Accounts of Chemical Research
For oP6(H)6 in Figure 6, the observed conformers can be assigned to the AAA (major), AAB, and BAB geometries. Once the 1H NMR signals have been assigned to specific geometries, the conformational population can be quantified by simple integration. Simple o-phenylenes do indeed preferentially fold into compact, stacked An−3 helices in solution. For the oPn(H)n and oPn(OMe)n series, these perfectly folded conformers account for about half of the populations depending on the length.21,25 Defects (B states) tend to occur at the ends of the chains: that is, the An−4B conformers are the next-most populated; hence, o-phenylenes can be described as helical oligomers with frayed ends.
■
WHY DO o-PHENYLENES FOLD?
Substituent Effects on the Folding of Hexa(o-phenylene)s
In principle, the folding behavior of an o-phenylene must be governed by a combination of inherent torsional preferences about φi, steric strain, and arene−arene interactions of various types; intersubstituent interactions could also play a role in some cases. Solvent effects are also likely to be significant although they have to this point received only limited study,31,32 and none at all in protic media. In order to study the folding mechanism, we examined the series of terminally substituted o-phenylene hexamers oP6(X)2, shown in Scheme 2.33 The substituents
Figure 6. 1H NMR and EXSY spectra of oP6(H)6 (500 MHz, CDCl3, − 5 °C). The sets of circled cross-peaks track one proton through the three observed conformers. Adapted with permission from ref 21. Copyright 2011 American Chemical Society.
Scheme 2. Terminally Substituted o-Phenylenes
The complex NMR spectra of o-phenylenes are very useful for analysis of the folding behavior. EXSY28 (i.e., NOESY) spectra allow the overlapping signals in the 1D 1H NMR spectra to be deconvoluted: As shown in Figure 6, clear cross-peaks connect the signals associated with a specific proton as it exchanges between each of the three conformers (there are four total signals because the asymmetric conformer has two distinct environments). In combination with standard COSY, HMQC, and HMBC spectra, it is straightforward to determine the chemical shift assignments for the observable backbone conformers. For hexamers, it is usually possible to fully account for the complete population. For longer oligomers, it may be necessary to focus on just the most heavily populated conformers. Because the different conformers vary greatly in the positioning of hydrogen atoms with respect to the shielding zones of nearby aromatic rings, the 1H chemical shifts δexp are highly sensitive to the particular folding state of the oligomer. Variations of up to 2 ppm are observed in some cases (e.g., the labeled signals in Figure 6). The NMR data can then be directly connected to specific molecular geometries through ab initio predictions of isotropic shieldings, with the conformational analysis outlined above providing the exact backbone geometries that must be considered. The method is not particularly sensitive to the quality of the input geometry: standard DFT methods such as (gas-phase) B3LYP/6-31G(d) work well. In our work, NMR shieldings δcalc have been calculated and referenced as recommended by Bally and Rablen29 using the inexpensive PCM/WP04/6-31G(d) method.30 One then simply compares δcalc for each candidate geometry with δexp for each conformer. In general, there is a clear distinction between good matches, for which the root-mean-squared (rms) difference is less than 0.25 ppm (and usually less than 0.15 ppm), and bad matches, for which the rms difference is typically above 0.3 ppm. To this point, we have not observed any conformers by NMR that could not be assigned to one of the expected A/B-type backbone geometries.
were found to exert a strong effect on the conformational behavior. Focusing on the AAB ⇌ AAA equilibrium (where AAA represents perfect folding), there is a good correlation between the free energy difference between the two conformers ΔG°AAA (extracted from NMR spectra) and the Hammett constants σm for the substituents, as shown in Figure 7. Note that in the absence of any stabilizing interactions the relative concentrations of AAB:AAA would be 2:1 simply because of the lower symmetry of the AAB conformer; this corresponds to ΔG°AAA = +0.38 kcal/
Figure 7. ΔG°AAA plotted against substituent σm for oP6(X)2 (277 K, CDCl3). Adapted with permission from ref 33. Copyright 2013 American Chemical Society. D
DOI: 10.1021/acs.accounts.6b00038 Acc. Chem. Res. XXXX, XXX, XXX−XXX
Article
Accounts of Chemical Research
on the number of “B” states describing the conformation, as shown in Figure 9b. Each φi in the B state will eliminate exactly one stacking relationship between arenes; thus, the good linear dependence strongly suggests that the factor most responsible for the folding preferences of o-phenylenes is the aromatic stacking interactions between every fourth repeat unit. The same relationship holds for the conformers of oP8(H)8 (Figure 9b). The slopes in the two cases are parallel at 2 kcal/mol per B state, which is comparable to the binding energy of the offset-stacked benzene dimer (2.75 kcal/mol at the B97-D/TZV(2d,2p) level).38
mol (at 277 K). Hence, for all oP6(X2) that were considered the equilibrium is biased toward the AAA conformer. This experimental result provides the first evidence that the difference in stability between the AAA and AAB conformers is determined by a single aromatic stacking interaction, as shown in Figure 8. The overall trend, enhanced stability of the AAA
A Simple Model for o-Phenylene Folding
The evidence above suggests that the essential features of ophenylene folding can be described by a model with two simple rules: 1. “ABA” sequences of torsional angles are forbidden because they introduce too much strain. 2. Conformer stability is otherwise determined by offset aromatic stacking interactions parallel to the helical axis, easily determined from the number of torsional angles in the A state. These two rules can be combined to give a semiquantitative model of o-phenylene folding that explicitly considers every possible folding state for an idealized oligomer.33 We define an oligomer as being “well-folded” if it is either perfectly folded (An−3) or has defects only at the ends (An−4B or BAn−5B). The fraction of the total population that should be well-folded as a function of oligomer length and interaction energy is shown in Figure 10. In the absence of stabilizing interactions, the expected
Figure 8. Equilibrium between the AAB and AAA conformers of oP6(X)2, with the key aromatic stacking interaction indicated.
conformer with increasing σm, is broadly consistent with classic results for aromatic stacking.34,35 Further, the simple model in Figure 8, if true, suggests that the AAB ⇌ AAA equilibrium is directly analogous to molecular balances and related systems used to quantify aromatic stacking.36 In this context, the measured energetic difference between the two conformers (0.3−1.1 kcal/mol, accounting for symmetry) is in good agreement with the strengths of aromatic stacking interactions measured for comparable systems, as is the overall substituent effect of ΔΔG° ≈ 1 kcal/mol in going from X = Me to CN.37 Further insight into the folding of o-phenylenes comes from DFT calculations.33 Although common DFT functionals fail to predict the relative stability of o-phenylene conformers (e.g., B3LYP), the dispersion-corrected B97-D/TZV(2d,2p) method,38 often used to study aromatic stacking,39 correctly predicts the stabilities of observed oP6(H)6 conformers (AAA > AAB > BAB > all others), as shown in Figure 9a. The method also does a reasonably good job of reproducing the substituent effects in Figure 7. Examining the oP6(H)6 energies in Figure 9a, it is clear that the ABA conformer is a special case, less stable than the other five by >10 kcal/mol. The instability arises from steric strain as the two terminal rings are forced into each other. Curiously, the energies of the remaining five conformers show a good linear dependence
Figure 10. Predicted fraction of well-folded oligomers as a function of length and stacking interaction energy. Adapted with permission from ref 33. Copyright 2013 American Chemical Society.
fraction of well-folded oligomers falls off quickly as the number of possible conformers increases exponentially. For an interaction energy of 0.5 kcal/mol, as is expected for the unsubstituted ophenylene (in chloroform), the folding is improved but still not particularly good. Thus, the parent poly(o-phenylene) should not fold particularly well. However, the folding propensity should be very sensitive to the strength of the stabilizing interaction over the range 0−2 kcal/mol.
■
Figure 9. Relative B97-D/TZV(2d,2p) conformer energies for oP6(H)6 and oP8(H)8. (a) Energies of all six possible conformers of oP6(H)6. (b) Energies vs number of B states for oP6(H)6 and oP8(H)8; “ABA” conformers have been excluded. The energies of the oP 6(H)6 conformers have been offset for clarity. Adapted with permission from ref 33. Copyright 2013 American Chemical Society.
TUNING THE FOLDING OF o-PHENYLENES Tuning aromatic stacking interactions between 0−2 kcal/mol should be possible using simple substituent effects. Thus, we prepared the series of fully substituted oligomers oPn(X)n (X = H, OMe, OAc, OTf, and CN, Scheme 1).32 Indeed, electronE
DOI: 10.1021/acs.accounts.6b00038 Acc. Chem. Res. XXXX, XXX, XXX−XXX
Article
Accounts of Chemical Research withdrawing substituents greatly decrease the proportion of imperfectly folded conformers, to the point that they are undetectable by 1H NMR for oP6(CN)6, as shown in Figure 11.
NMR assignments can be made, this qualitative approach is probably the easiest and most direct way to assess the folding state of an o-phenylene backbone.32 It is likely that some future applications of folded o-phenylenes will rely on interactions with the surface of the helix (i.e., by analogy with α-helices). Unfortunately, the substitution pattern in the oPn(X)n series arranges the substituents in a zigzag pattern along each stack of arenes, as shown in Figure 13. With this in
Figure 11. 1H NMR spectra (500 MHz, CDCl3, 268 K) of oP6(OMe)6A and oP6(CN)6A (aromatic regions). The peaks labeled with an asterisk (*) are the isotopic side bands for the chloroform. Adapted with permission from ref 32. Copyright 2014 American Chemical Society.
Figure 13. Substituent (X) group alignment in oPn(X)n and oPn(X)niso. Adapted with permission from ref 32. Copyright 2014 American Chemical Society.
mind, we also investigated the folding of the isomeric series oPn(X)niso.32 Given that the strength of stacking interactions between substituted aromatics depends strongly on the relative positioning of the substituents,39 it is not surprising that the folding propensities of this new series are distinct. Good folding is observed in all cases, but no correlation with σm values is observed, likely because direct substituent−substituent interactions are of greater importance in this series and not accounted for by Hammett constants. DFT calculations on simple models do a fair job of predicting the folding behavior, however. Interestingly, Fukushima and Aida have reported a series of ophenylenes that demonstrate a surprising effect of terminal substituents on folding.31 In most solvents (CDCl3, toluene-d8, DMSO-d6, and DMF-d7), oligomers oP10(OMe)20(X)2, shown in Scheme 3, are not particularly well-folded (as judged by 1H
The folding can be quantified according to the equilibrium [all misfolded conformers] ⇌ An−3 with associated free energy change ΔG°fold. For the oP6(X)6 series, a good correlation is found with σm for the substituents. Notably, excellent folding (>90% of the perfectly folded conformer) is observed even for the only moderately electron-withdrawing acetoxy substituent. These same substituent effects carry forward to oligomers at least as long as [10]-mers. The 1H NMR spectrum of oP10(OAc)10 is shown in Figure 12. While there are some small signals resulting from misfolded states, more than 90% of the population is perfectly folded into the A7 conformer.
Scheme 3. o-Phenylenes Reported by Fukushima and Aida
Figure 12. 1H NMR spectrum of oP10(OAc)10 (500 MHz, CDCl3, rt). The model corresponds to the solid-state structure with the acetoxy groups omitted for clarity. The hydrogen atoms have been colored according to their chemical shifts. Adapted with permission from ref 32. Copyright 2014 American Chemical Society.
NMR). However, in acetonitrile-d3, the folding improves dramatically for all of the substituted oligomers (X ≠ H). This appears to be due to a steric effect; however, to the best of our knowledge, its precise origin, and the special role played by the solvent, is as yet unexplained.
With the availability of data for a range of different oligomers, it was possible to draw some general guidelines for the rapid analysis of o-phenylene conformation. In Figure 12, the A7 geometry of oP10(OAc)10 is shown with the aromatic hydrogen atoms colored according to their chemical shifts. In the folded state, the stacking orients specific hydrogen atoms in the shielding zones of neighboring aromatic rings. This positioning alone can shield the protons by ≥1 ppm. For example, the circled protons are in nearly identical chemical environments in terms of through-bond substituent effects, but, as indicated in the spectrum, their chemical shifts differ by 1.2 ppm. So long as 1H
■
PROPERTIES DERIVED FROM o-PHENYLENE FOLDING Unlike the isomeric m- and p-phenylenes, it is the conformational behavior of o-phenylenes that determines most of their properties of interest. Take, for example, the anomalous hypsochromic shift in the fluorescence of o-phenylene oligomers with increasing length (Figure 1). It is well-known that the Stokes F
DOI: 10.1021/acs.accounts.6b00038 Acc. Chem. Res. XXXX, XXX, XXX−XXX
Accounts of Chemical Research shift in biaryls results from excited-state planarization.40 For ophenylenes, planarization is clearly not possible because of simple sterics; however, reductions in φi will cause the favored An−3 conformer to compress like a spring. As the length of the oligomer increases, this compression is less well accommodated and so the fluorescence shifts to the blue.24,41 As this phenomenon is more directly tied to the oligomer geometry than it is to its electronic structure, the shift in emission is observed over a longer length scale than that of absorption. As shown by Fukushima and Aida, the folding of o-phenylenes also has a strong effect on their electrochemical properties.18 Oxidation of octamer oP8(OMe)16(NO2)2 (Scheme 3) yields a radical cation with the hole delocalized over the complete oligomer, as determined by crystallography and DFT calculations. Further work with oligomers such as oP10(OMe)20(X)2 has demonstrated better electrochemical reversibility under conditions of near-perfect folding.31 Our work on detecting and controlling folding described above focused on racemic o-phenylene oligomers. However, Fukushima and Aida have looked at racemization rates for some of their oligomers. Compound oP8(OMe)16(NO2)2 crystallizes as a conglomerate, and therefore can be resolved through physical separation.18 Although the helix is configurationally stable in the solid state, it racemizes quickly in solution (t1/2 = 5.9 min, MeCN, −10 °C); interestingly, the (delocalized) radical cation racemizes much more slowly (t1/2 = 44 h). Variabletemperature NMR shows that the racemization rate of (neutral) veratrole-based o-phenylenes slows with increasing length.31 Oligomers longer than the [10]-mer can therefore be resolved by chiral-phase HPLC; the [16]- and [24]-mers have half-lives for racemization on the order of hours. Interestingly, the free energy of activation for racemization of these oligomers increases as ΔΔG‡ ≈ 0.4 kcal/mol per additional repeat unit, consistent with the breaking of an additional arene-stacking interaction.
Article
■
CONCLUSION AND OUTLOOK
■
ASSOCIATED CONTENT
o-Phenylenes exist as compact, stacked helices in solution (and in the solid state17,18,31−33). They are therefore a new class of foldamers, with secondary structures stabilized by attractive interactions parallel to the helical axis. In this way their folding is analogous to the helical folding of peptides, except with aromatic stacking in place of hydrogen bonding. Given the many examples of foldamers, why study ophenylenes? First, because of modern aryl−aryl coupling methods, they are now relatively easy to synthesize. New ophenylene-based architectures are being enabled by the appearance of new, complementary methods, such as the Asao−Yamamoto benzannulation52 and benzyne polymerization.53,54 Second, they have simple structures, arguably the simplest possible structures. Consequently, their folding is now well understood and predictable using straightforward models, as described above. Third, their simple arene-rich but minimally conjugated structure yields unique properties. For example, the structural parallel between the folded helix and a stack of triphenylenes allows them to act as alignment layers for discotic liquid crystals,55 and they have been used as unconjugated moieties in high triplet energy materials.56 Fourth, it is possible to determine their precise folding state in solution. For many foldamer architectures, solution phase behavior is inferred from bulk spectroscopic measurements that give only a general view of the population. Alternatively, they are characterized by crystallography, which gives very precise information about the solid-state but leaves open the question of how well folded the oligomers are in solution. As discussed above, for o-phenylenes it is straightforward to derive exact solution-phase folding populations. Of course, the dependence on NMR spectroscopy and ab initio calculations means that this strategy is limited to oligomers of short to moderate length (n ≲ 12); however, studies on these shorter systems have allowed general guidelines to be derived which should be very useful in assessing larger systems. These properties make o-phenylenes attractive structures for the next phase of foldamer research, as the field moves away from the study of local folding to the higher-order positioning of folded subunits relative to each other.15 They are also part of a recent interest in the study of various sterically congested polyphenylenes, including polyphenylene polymers and oligomers with mixed connectivity57,58 and more complex twodimensional architectures.52 Conformational behavior is among the defining characteristics of all of these systems.
■
HETEROCYCLIC ANALOGUES OF o-PHENYLENES The study of polymers and oligomers based on 1,2-linked aromatic heterocycles predates much of the work on ophenylenes. The best examples are the poly(quinoxaline-2,3diyl)s studied by Ito and Suginome.42,43 These compounds are well-established as an important class of helical polymers, exhibiting remarkable chiroptical properties44 and being used as ligands for asymmetric catalysis.45 The conformational analysis of this system closely parallels that of o-phenylenes.46 However, in contrast to the o-phenylenes, these systems favor the fully extended helical Bn−3 conformers (Figure 5). This behavior appears to be inherent to the pyrazine-2,3-diyl repeat unit, as indicated by DFT calculations on simple model compounds (see the Supporting Information). Recently, Tokoro reported a series of pyridine-containing ophenylene hexamers.47 Folding into the AAA conformer directly stacks two pyridine rings and is very effective, with no evidence for minor conformational states. This behavior is consistent with the enhanced aromatic stacking interactions expected for pyridine dimers (of certain geometries).48 Several analogues based on five-membered aromatic heterocycles have also been reported. In general, good folding may be observed in the solid state49,50 but solution-phase behavior is complex in the absence of secondary interactions between appended units.51 Unlike o-phenylenes, conformational exchange in oligomers of five-membered heterocycles appears to be fast on the NMR time scale.
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.accounts.6b00038.
■
Cartesian coordinates for AAA and BBB conformers of hexa(pyrazine-2,3-diyl) (TXT) Details of computational results (PDF)
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Notes
The authors declare no competing financial interest. G
DOI: 10.1021/acs.accounts.6b00038 Acc. Chem. Res. XXXX, XXX, XXX−XXX
Article
Accounts of Chemical Research Biography
oligomers with a surprisingly long effective conjugation length. J. Am. Chem. Soc. 2010, 132, 13848−13857. (18) Ohta, E.; Sato, H.; Ando, S.; Kosaka, A.; Fukushima, T.; Hashizume, D.; Yamasaki, M.; Hasegawa, K.; Muraoka, A.; Ushiyama, H.; Yamashita, K.; Aida, T. Redox-responsive molecular helices with highly condensed π-clouds. Nat. Chem. 2011, 3, 68−73. (19) Walker, S. D.; Barder, T. E.; Martinelli, J. R.; Buchwald, S. L. A rationally designed universal catalyst for Suzuki−Miyaura coupling processes. Angew. Chem., Int. Ed. 2004, 43, 1871−1876. (20) Banerjee, M.; Shukla, R.; Rathore, R. Synthesis, optical, and electronic properties of soluble poly-p-phenylene oligomers as models for molecular wires. J. Am. Chem. Soc. 2009, 131, 1780−1786. (21) Mathew, S. M.; Hartley, C. S. Parent o-phenylene oligomers: synthesis, conformational behavior, and characterization. Macromolecules 2011, 44, 8425−8432. (22) He, J.; Mathew, S. M.; Cornett, S. D.; Grundy, S. C.; Hartley, C. S. ortho-Phenylene oligomers with terminal push-pull substitution. Org. Biomol. Chem. 2012, 10, 3398−3405. (23) Tour, J. M. Soluble oligo- and polyphenylenes. Adv. Mater. 1994, 6, 190−198. (24) Guiglion, P.; Zwijnenburg, M. A. Contrasting the optical properties of the different isomers of oligophenylene. Phys. Chem. Chem. Phys. 2015, 17, 17854−17863. (25) Hartley, C. S.; He, J. Conformational analysis of o-phenylenes: helical oligomers with frayed ends. J. Org. Chem. 2010, 75, 8627−8636. (26) Delsuc, N.; Kawanami, T.; Lefeuvre, J.; Shundo, A.; Ihara, H.; Takafuji, M.; Huc, I. Kinetics of helix-handedness inversion: folding and unfolding in aromatic amide oligomers. ChemPhysChem 2008, 9, 1882− 1890. (27) Abramyan, A. M.; Liu, Z.; Pophristic, V. Helix handedness inversion in arylamide foldamers: elucidation and free energy profile of a hopping mechanism. Chem. Commun. 2016, 52, 669−672. (28) Perrin, C. L.; Dwyer, T. J. Application of two-dimensional NMR to kinetics of chemical exchange. Chem. Rev. 1990, 90, 935−967. (29) Jain, R.; Bally, T.; Rablen, P. R. Calculating accurate proton chemical shifts of organic molecules with density functional methods and modest basis sets. J. Org. Chem. 2009, 74, 4017−4023. (30) Wiitala, K. W.; Hoye, T. R.; Cramer, C. J. Hybrid density functional methods empirically optimized for the computation of 13C and 1H chemical shifts in chloroform solution. J. Chem. Theory Comput. 2006, 2, 1085−1092. (31) Ando, S.; Ohta, E.; Kosaka, A.; Hashizume, D.; Koshino, H.; Fukushima, T.; Aida, T. Remarkable effects of terminal groups and solvents on helical folding of o-phenylene oligomers. J. Am. Chem. Soc. 2012, 134, 11084−11087. (32) Mathew, S.; Crandall, L. A.; Ziegler, C. J.; Hartley, C. S. Enhanced helical folding of ortho-phenylenes through the control of aromatic stacking interactions. J. Am. Chem. Soc. 2014, 136, 16666−16675. (33) Mathew, S. M.; Engle, J. T.; Ziegler, C. J.; Hartley, C. S. The role of arene−arene interactions in the folding of ortho-phenylenes. J. Am. Chem. Soc. 2013, 135, 6714−6722. (34) Hunter, C. A.; Sanders, J. K. M. The nature of π-π interactions. J. Am. Chem. Soc. 1990, 112, 5525−5534. (35) Cozzi, F.; Cinquini, M.; Annunziata, R.; Dwyer, T.; Siegel, J. S. Polar/π interactions between stacked aryls in 1,8-diarylnaphthalenes. J. Am. Chem. Soc. 1992, 114, 5729−5733. (36) Mati, I. K.; Cockroft, S. L. Molecular balances for quantifying noncovalent interactions. Chem. Soc. Rev. 2010, 39, 4195−4205. (37) Gung, B. W.; Xue, X.; Reich, H. J. The strength of paralleldisplaced arene−arene interactions in chloroform. J. Org. Chem. 2005, 70, 3641−3644. (38) Grimme, S. Semiempirical GGA-type density functional constructed with a long-range dispersion correction. J. Comput. Chem. 2006, 27, 1787−1799. (39) Wheeler, S. E. Local nature of substituent effects in stacking interactions. J. Am. Chem. Soc. 2011, 133, 10262−10274. (40) Momicchioli, F.; Bruni, M. C.; Baraldi, I. Fluorescence and absorption spectra of polyphenyls. Theoretical study on the band shape. J. Phys. Chem. 1972, 76, 3983−3990.
C. Scott Hartley was born in Ottawa, Ontario, Canada in 1977. He received his B.Sc.H. (2000) and Ph.D. (2005) degrees from Queen’s University under the supervision of Robert P. Lemieux. He was then a postdoctoral researcher at the University of Illinois at Urbana− Champaign with Jeffrey S. Moore. He joined the faculty at Miami University in 2007.
■
ACKNOWLEDGMENTS Many thanks to the current and former co-workers who contributed to this project, which was supported by National Science Foundation (CHE-0910477 and CHE-1306437) and the donors of the American Chemical Society Petroleum Research Fund (47926-G7).
■
REFERENCES
(1) Hammer, B. A. G.; Müllen, K. Dimensional evolution of polyphenylenes: expanding in all directions. Chem. Rev. 2016, 116, 2103−2140. (2) Grimsdale, A. C.; Leok Chan, K.; Martin, R. E.; Jokisz, P. G.; Holmes, A. B. Synthesis of light-emitting conjugated polymers for applications in electroluminescent devices. Chem. Rev. 2009, 109, 897− 1091. (3) Ben, T.; Goto, H.; Miwa, K.; Goto, H.; Morino, K.; Furusho, Y.; Yashima, E. Synthesis and helix formation of poly(m-phenylene)s bearing optically active oligo(ethylene oxide) side chains in protic media. Macromolecules 2008, 41, 4506−4509. (4) Patel, D. G. D.; Ohnishi, Y.-Y.; Yang, Y.; Eom, S.-H.; Farley, R. T.; Graham, K. R.; Xue, J.; Hirata, S.; Schanze, K. S.; Reynolds, J. R. Conjugated polymers for pure UV light emission: poly(meta-phenylenes). J. Polym. Sci., Part B: Polym. Phys. 2011, 49, 557−565. (5) Kovacic, P.; Uchic, J. T.; Hsu, L.-C. Polymerization of aromatic nuclei. XII. Oligomerization of halobenzenes by aluminum chloride− cupric chloride. J. Polym. Sci., Part A-1: Polym. Chem. 1967, 5, 945−964. (6) Kovacic, P.; Ramsey, J. S. Polymerization of aromatic nuclei. XIV. Polymerization of toluene with aluminum chloride−cupric chloride. J. Polym. Sci., Part A-1: Polym. Chem. 1969, 7, 111−125. (7) Hsing, C.-F.; Jones, M. B.; Kovacic, P. Polymerization of aromatic nuclei. XXVII. ESR studies on polymers and oligomers from reaction of benzenoid and naphthalene monomers with Lewis acid catalyst− oxidant systems. J. Polym. Sci., Polym. Chem. Ed. 1981, 19, 973−984. (8) Wittig, G.; Lehmann, G. Ü ber die reaktionsweise von 2.2′dilithium-diphenyl gegenüber metallchloriden; gleichzeitig ein beitrag zur synthese von poly-o-phenylenen. Chem. Ber. 1957, 90, 875−892. (9) Ibuki, E.; Ozasa, S.; Murai, K. Studies of polyphenyls and polyphenylenes. I. The syntheses and infrared and electronic-spectra of several sexiphenyls. Bull. Chem. Soc. Jpn. 1975, 48, 1868−1874. (10) Ibuki, E.; Ozasa, S.; Fujioka, Y.; Kitamura, H. Syntheses and physical properties of several symmetrical sexiphenyls. Chem. Pharm. Bull. 1980, 28, 1468−1476. (11) Ozasa, S.; Fujioka, Y.; Fujiwara, M.; Ibuki, E. Syntheses and physical properties of several octiphenyls and a septiphenyl. Chem. Pharm. Bull. 1980, 28, 3210−3222. (12) Blake, A. J.; Cooke, P. A.; Doyle, K. J.; Gair, S.; Simpkins, N. S. Poly-orthophenylenes: synthesis by Suzuki coupling and solid state helical structures. Tetrahedron Lett. 1998, 39, 9093−9096. (13) Hill, D. J.; Mio, M. J.; Prince, R. B.; Hughes, T. S.; Moore, J. S. A field guide to foldamers. Chem. Rev. 2001, 101, 3893−4011. (14) Gellman, S. H. Foldamers: a manifesto. Acc. Chem. Res. 1998, 31, 173−180. (15) Guichard, G.; Huc, I. Synthetic foldamers. Chem. Commun. 2011, 47, 5933−5941. (16) Voisin, E.; Williams, V. E. Do catechol derivatives electropolymerize? Macromolecules 2008, 41, 2994−2997. (17) He, J.; Crase, J. L.; Wadumethrige, S. H.; Thakur, K.; Dai, L.; Zou, S.; Rathore, R.; Hartley, C. S. ortho-Phenylenes: unusual conjugated H
DOI: 10.1021/acs.accounts.6b00038 Acc. Chem. Res. XXXX, XXX, XXX−XXX
Article
Accounts of Chemical Research (41) Hartley, C. S. Excited-state behavior of ortho-phenylenes. J. Org. Chem. 2011, 76, 9188−9191. (42) Ito, Y.; Ihara, E.; Murakami, M. Enantioselective polymerization of 1,2-diisocyanoarenessynthesis of optically active, helical poly(quinoxaline-2,3-diyl)s. Angew. Chem., Int. Ed. Engl. 1992, 31, 1509− 1510. (43) Suginome, M.; Ito, Y. Transition metal-mediated polymerization of isocyanides. Adv. Polym. Sci. 2004, 171, 77−136. (44) Nagata, Y.; Nishikawa, T.; Suginome, M. Chirality-switchable circularly polarized luminescence in solution based on the solventdependent helix inversion of poly(quinoxaline-2,3-diyl)s. Chem. Commun. 2014, 50, 9951−9953. (45) Yamamoto, T.; Akai, Y.; Suginome, M. Chiral palladacycle catalysts generated on a single-handed helical polymer skeleton for asymmetric arylative ring opening of 1,4-epoxy-1,4-dihydronaphthalene. Angew. Chem., Int. Ed. 2014, 53, 12785−12788. (46) Ito, Y.; Ihara, E.; Murakami, M.; Sisido, M. Studies on the conformation of helical poly(2,3-quinoxaline)s. Empirical energy calculation and theoretical circular dichroism. Macromolecules 1992, 25, 6810−6813. (47) Tokoro, Y.; Ohtsuka, N.; Toh, A.; Fukuzawa, S.-i. Synthesis and structure of N-hetero-ortho-phenylene hexamers containing 2,3substituted pyridine moieties. Eur. J. Org. Chem. 2015, 2015, 7666− 7669. (48) Hohenstein, E. G.; Sherrill, C. D. Effects of heteroatoms on aromatic π−π interactions: benzene−pyridine and pyridine dimer. J. Phys. Chem. A 2009, 113, 878−886. (49) Marsella, M. J.; Yoon, K.; Almutairi, A.; Butt, S. K.; Tham, F. S. Helical sexithiophenes: an experimental and theoretical study implicating the alternating 2,2′:3,3′ regioisomer as a reliable helical motif. J. Am. Chem. Soc. 2003, 125, 13928−13929. (50) Fallon, T.; Willis, A. C.; Rae, A. D.; Paddon-Row, M. N.; Sherburn, M. S. β-Oligofurans. Chem. Sci. 2012, 3, 2133−2137. (51) Nakashima, T.; Imamura, K.; Yamamoto, K.; Kimura, Y.; Katao, S.; Hashimoto, Y.; Kawai, T. Synthesis, structure, and properties of α,βlinked oligothiazoles with controlled sequence. Chem.Eur. J. 2014, 20, 13722−13729. (52) Hein, S. J.; Arslan, H.; Keresztes, I.; Dichtel, W. R. Rapid synthesis of crowded aromatic architectures from silyl acetylenes. Org. Lett. 2014, 16, 4416−4419. (53) Ito, S.; Takahashi, K.; Nozaki, K. Formal aryne polymerization: use of [2.2.1]oxabicyclic alkenes as aryne equivalents. J. Am. Chem. Soc. 2014, 136, 7547−7550. (54) Mizukoshi, Y.; Mikami, K.; Uchiyama, M. Aryne polymerization enabling straightforward synthesis of elusive poly(ortho-arylene)s. J. Am. Chem. Soc. 2015, 137, 74−77. (55) Kajitani, T.; Suna, Y.; Kosaka, A.; Osawa, T.; Fujikawa, S.; Takata, M.; Fukushima, T.; Aida, T. o-Phenylene octamers as surface modifiers for homeotropic columnar ordering of discotic liquid crystals. J. Am. Chem. Soc. 2013, 135, 14564−14567. (56) Lee, D. R.; Lee, C. W.; Lee, J. Y. High triplet energy host materials for blue phosphorescent organic light-emitting diodes derived from carbazole modified orthophenylene. J. Mater. Chem. C 2014, 2, 7256− 7263. (57) Arslan, H.; Saathoff, J. D.; Bunck, D. N.; Clancy, P.; Dichtel, W. R. Highly efficient benzannulation of poly(phenylene ethynylene)s. Angew. Chem., Int. Ed. 2012, 51, 12051−12054. (58) Manabe, K.; Kimura, T. Repetitive synthetic method for o,o,poligophenylenes using C−H arylation. Org. Lett. 2013, 15, 374−377.
I
DOI: 10.1021/acs.accounts.6b00038 Acc. Chem. Res. XXXX, XXX, XXX−XXX
