Skip to main content
NCBI home page
UKPMC Funders Author Manuscripts logoLink to UKPMC Funders Author Manuscripts
. Author manuscript; available in PMC: 2022 Apr 13.
Published in final edited form as: J Am Chem Soc. 2021 Sep 30;143(40):16682–16692. doi: 10.1021/jacs.1c07778

Conformationally-controlled linear and helical hydrocarbons bearing extended side-chains

Lin Guo ∫, Oliver J Dutton ∫, Murat Kucukdisli ∫, Matthew Davy ∫, Olivier Wagnières ∫, Craig P Butts ∫,*, Eddie L Myers †,*, Varinder K Aggarwal ∫,*
PMCID: PMC7612001  EMSID: EMS138312  PMID: 34590479

Abstract

Conformationally controlled flexible molecules are ideal for applications in medicine and materials, where shape matters but an ability to adapt to multiple and changing environments is often required. The conformation of flexible hydrocarbon chains bearing contiguous methyl substituents is controlled through the avoidance of syn-pentane interactions: alternating syn–anti isomers adopt a linear conformation whilst all-syn isomers adopt a helical conformation. From a simple diamond lattice analysis, larger substituents, which would be required for most potential applications, result in significant and un-avoidable syn-pentane interactions, suggesting substantially reduced conformational control. Through a combination of computation, synthesis and NMR analysis, we have identified a selection of substitution patterns that allow large groups to be incorporated on conformationally controlled linear and helical hydrocarbon chains. Surprisingly, when the methyl substituents of alternating syn–anti hydrocarbons are replaced with acetoxyethyl groups, the main chain of almost 95% of the population of molecules adopt a linear conformation. Here, the side-chains adopt non-ideal eclipsed conformations with the main chain, thus minimizing syn-pentane interactions. In the case of all-syn hydrocarbons, concurrent removal of some methyl groups on the main-chain adjacent to the large substituents is required to maintain a high population of molecules adopting a helical conformation. This information can now be used to design flexible hydrocarbon chains displaying functional groups in a defined relative orientation for multivalent binding or cooperative reactivity, for example, in targeting the interfaces defined by disease-relevant protein–protein interactions.

Introduction

Hydrocarbon chains bearing substituents with a 1,3-relationship adopt conformations that minimise destabilizing syn-pentane interactions,1–7 an interaction that imposes a ~14 kJ/mol penalty (Figure 1a).8 Therefore minimizing steric interactions, while maximizing stereoelectronically favourable arrangements,9–12 can allow flexible acyclic hydrocarbon molecules to adopt well-defined conformations with a level of control that rival cyclic systems. Indeed, it has been proposed that Nature evolved polypropionate metabolism to create flexible hydrocarbon chains punctuated at every other carbon atom with a methyl substituent, which act like levers to control conformation.8, 13–16 The flexibility enables the molecule to change shape as required, for example, for transport across membranes, and the conformational control provides increased binding to its biological target. Hoffmann, who investigated the conformational properties of acyclic systems, provided experimental data to support the conformational control that is conferred by a 1,3-positional pattern of methyl substituents. Notably, even greater control could be achieved by placing a t-Bu group at one end,15,16 but the degree of control rapidly diminishes with increasing chain length owing to the enthalpic gain in minimizing syn-pentane interactions being out-weighed by the entropic cost of greater rigidity (Figure 1b).

Figure 1.

Figure 1

Open in a new tab

(a) Energy penalty incurred by a syn-pentane conformation; (b) Population of helical conformations in isotactic oligopropylene, a shape induced predominantly by the avoidance of syn-pentane interactions; (c) and (d) Diamond-lattice overlay of helical and linear conformations of a section of all-syn hydrocarbon 1 and alternating syn–anti hydrocarbon 2, respectively.

We showed that high levels of conformational control could be maintained in longer chains by doubling the number of interactions, specifically by placing a methyl substituent on every carbon atom of the main chain, thereby doubling the enthalpic gain in avoiding syn-pentane interactions without increasing the number of rotatable bonds.17 The effect of contiguous methyl substitution can be easily appreciated by viewing the acyclic structure on a diamond lattice (Figure 1c/d). We also showed that the configurational pattern of methyl substitution controls the shape of the major conformation, a tactic reminiscent of the use of L- to D-amino acid mutations to control peptide conformation and ultimately aggregation.18 Specifically, the alternating syn-anti isomer 2 adopts a single zig–zag linear conformation representing 95% of the population; the all-syn isomer 1 adopts a helical conformation representing 74% of the population (ignoring end-group rotamers).17 In a fascinating twist, we subsequently discovered that the helical propensity of the all-syn isomer 1 alternated with increasing chain length: chains with an even number of carbon atoms showing higher and more enduring levels of helical control.19

Although these studies showed that the conformation of carbon chains bearing simple methyl groups can be controlled, it was unclear how the conformational landscape would be affected by larger substituents. Confirmation of, and the ability to predict, the level and sense of conformational control for such systems would be needed for their use as scaffolds in materials and medicine. For example, chains presenting hydrophobic and hydrophilic substituents could be designed as multivalent inhibitors of protein–protein interactions, which frequently involve large featureless interfaces. However, a diamond-lattice analysis suggested that high levels of conformational control might be difficult to achieve for such systems. For example, for linear molecule 2, replacement of a single methyl group with an ethyl group gives hydrocarbon 3 (Figure 2a). Placing hydrocarbon 3 on a diamond lattice reveals that C2″ on the ethyl side-chain cannot avoid syn-pentane interactions, thereby suggesting that the strong preference for a linear conformation will be eroded. As drawn in Figure 2a, C2″ of the ethyl side-chain has a syn-pentane interaction with the C3′ methyl group. In principle, replacement of the C3′ methyl group for a hydrogen atom, to give methylene 4, will eliminate this syn-pentane interaction (Figure 2b), thus allowing the ethyl side-chain to occupy one low-energy conformation that is devoid of syn-pentane interactions. However, as we have shown previously, the small number of discrete dihedral angles defined by the diamond lattice can be too restrictive, leading to incorrect conclusions.19 Herein, we use a synergistic combination of computation, synthesis, and NMR analysis to define the accessible conformational space and identify rules on how the conformation of hydrocarbon chains bearing larger substituents can be controlled.

Figure 2.

Figure 2

Open in a new tab

Diamond lattice overlay of linear conformers of (a) syn–anti hydrocarbon 3 containing an ethyl side-chain (red) and exhibiting syn-pentane interactions in each of the three rotamers; (b) syn–anti hydrocarbon 4 where the syn-methyl has been replaced with a hydrogen atom; (c) 1,3-diethyl syn–anti hydrocarbon 5 where the second ethyl side-chain (red) has syn-pentane interactions in each of the three rotamers; (d) 1,3-diethyl syn–anti hydrocarbon 6 where the configuration at C5 and C7 has been inverted to remove syn-pentane interactions; (e) Lowest energy conformer 6b1, with the backbone adopting a linear conformation and both side-chains in a staggered conformation; second lowest energy conformer 6b2, with the backbone adopting a linear conformation and one side-chain adopting an eclipsed conformation. Newman projections looking down the C2-C2′ bond.

Results and Discussion

In assessing the effect of groups larger than methyl, we focussed on 1,2 (gauche) and 1,3 (syn-pentane) interactions, which dominate the conformations of hydrocarbons like 1 and 2. We used acetoxyethyl groups as analogues of longer alkyl substituents for synthesis and computation because the polar acetoxy group increased chemical shift dispersion of NMR signals, thus facilitating conformational analysis; notably, comparison with the computational data of simple ethyl analogues showed that the acetoxy dipole moments did not disrupt the conformational ensembles.

Linear hydrocarbons with 1,3-related large substituents

Hydrocarbons can be conferred with a low-energy linear conformation through imposing an alternating syn–anti substitution pattern. As indicated above, when one methyl group in a contiguously methyl-substituted hydrocarbon is replaced for an ethyl group, syn-pentane interactions can no longer be avoided, with C2″ of the ethyl side-chain having one such interaction with the C3′ methyl group. In principle, replacement of the C3′ methyl group with a hydrogen atom to give methylene 4 eliminates this syn-pentane interaction (Figure 2b), thus allowing the ethyl side-chain to occupy one low-energy conformation devoid of syn-pentane interactions while the main chain maintains the desired zig–zag linear conformation. Another methyl → ethyl substitution at C4 (to give 5), leads to further syn-pentane interactions (Figure 2c), which can be avoided by either replacing the C5 methyl group in 4 with a hydrogen atom or by inversion of configuration at C5 (and also C7 to give 6) thus maintaining the necessary alternating syn–anti substitution pattern on the main chain flanking the methylene group. However, replacing a methyl group with a hydrogen atom in the main chain would further erode conformational control owing to an increased number of accessible conformations devoid of syn-pentane interactions. Therefore, this diamond-lattice analysis suggests that hydrocarbon 6, which exhibits a 1,3-diethyl substituted 3-carbon fragment embedded within an alternating syn–anti contiguously methyl-substituted chain and has only one conformation devoid of syn-pentane interactions, should have good conformational control.

To test the conclusions of the above analysis, we subjected acetoxyethyl analogue 6b (X=OAc, R=Me, Figure 2d,e) to molecular mechanics conformational search with subsequent DFT energy calculations (see the SI for details). The lowest energy structure found adopts a linear backbone conformation, but with a relatively low population (27.5%, 6b1). Only one other linear conformer was predicted to have a significant population (20%, 6b2), indicating that hydrocarbon 6 would have poor conformational control compared to parent hydrocarbon 2, which adopted a linear conformation with a population of 95%.20 Interestingly, although both ethyl side-chains in conformer 6b1 adopted a staggered arrangement with respect to the main chain, one ethyl side-chain in conformer 6b2 adopted an off-diamond lattice eclipsed arrangement (H2–C2–C2′–C2″ dihedral angle: 5.3°). Furthermore, conformer 6b2 was only 0.8 kJ/mol above conformer 6b1, indicating that the penalty for adopting the partially-eclipsed conformation was surprisingly small. Such off-lattice conformations were not observed for the main chain, suggesting that side-chains can adopt off-lattice conformations to avoid syn-pentane interactions, without perturbation of the main chain structure. Because it was clear that side-chains could adopt an off-lattice eclipsed conformation without a significant energetic penalty, we decided to reinstall the methyl substituent at C3, leading to structure 7 (Figure 3), which we hoped would exhibit a level of main-chain conformational control similar to that of parent hydrocarbon 2.

Figure 3.

Figure 3

Open in a new tab

(a) Alternating syn–anti hydrocarbon 7 with 1,3-related large substituents; (b) Idealised diamond lattice conformation of hydrocarbon 7 with syn-pentane interactions highlighted; (c) Illustrative off-diamond lattice conformation of hydrocarbon 7, highlighting the eclipsed conformations of the extended side-chains, as adopted to avoid syn pentane interactions. Newman projections looking down the C4−C4′ bond and C2−C2′ bonds.

Subjecting hydrocarbon 7b (X=OAc, R=Me, Figure 3) to a molecular mechanics conformational search with subsequent DFT calculation revealed seven low-energy linear conformers (>97% population) differing only by rotation around the bonds of the acetoxyethyl side-chains. As expected, both side-chains at C2 and C4 adopt partially-eclipsed conformations with respect to the main chain, albeit with slightly larger dihedral angles (~20–23°) compared to hydrocarbon 6b; these eclipsing interactions avoid the higher-energy syn-pentane interactions of an on-diamond lattice conformation. With a good theoretical basis established, hydrocarbon 7b, and 6b, which would be used as a control, were taken forward for synthesis and conformational analysis.

Linear hydrocarbons with 1,2-related large substituents

Linear (alternating syn–anti) hydrocarbons with 1,2-related large substituents can be configured 1,2-syn- or 1,2-anti to give structures 8 and 9, respectively (Figure 4). On a diamond lattice, linear conformers of both structures have at least one syn-pentane interaction. However, based on the calculated off-lattice side-chain conformations for the 1,3-related large substituents, we investigated whether hydrocarbons 8 and 9 would still retain good conformational control of the main chain. Indeed, subjecting hydrocarbon 8b (X=OAc, R=Me, Figure 4a) to a molecular mechanics conformational search with subsequent DFT calculation revealed a predominance of low-energy linear conformers (>95% population). As anticipated, this is allowed by eclipsing of the side-chains and in the dominant conformer (71%) both side-chains adopt an eclipsed conformation with respect to the main chain (Figure 4(a)), thus facilitating the avoidance of on-lattice high-energy syn-pentane interactions. In contrast, for the 1,2-anti hydrocarbon (9b, X=OAc, R=Me) only ~50% of the conformers are linear with eclipsed conformations of the side-chains (Figure 4(b)), the remaining population adopting non-linear conformations exhibiting eclipsing interactions within the main chain (C1–C2 or C4–C5 bonds). Owing to the predicted higher level of linear conformational control in C 2-symmetric 1,2-syn hydrocarbon 8b, which can be prepared by using an efficient bidirectional synthetic strategy, and the more labour-intensive monodirectional synthesis that would be required for accessing non-symmetric 1,2-anti hydrocarbon 9b, we elected to take hydrocarbon 8b forward for synthesis and conformational analysis.

Figure 4.

Figure 4

Open in a new tab

Alternating syn–anti hydrocarbons with (a) 1,2-syn (8) and (b) 1,2-anti (9) ethyl substituents. Left - idealised diamond-lattice conformation with syn-pentane interactions; Right – lowest-energy conformer as determined by DFT computation with the extended side-chains showing off-diamond lattice eclipsed conformations, thus avoiding syn-pentane interactions.

Helical hydrocarbons with 1,3-related large substituents

These types of hydrocarbons can be conferred with a low-energy helical conformation through an all-syn substitution pattern. A diamond-lattice analysis of 1,3-diethyl-substituted hydrocarbon 10 (all other main-chain carbon atoms bearing a methyl substituent) shows that all accessible helical conformers exhibit two syn-pentane interactions, including one conformer where both ethyl side-chains have a syn-pentane interactions with the same C2 methyl group (Figure 5a). This analysis was consistent with conformational search and subsequent DFT calculations on hydrocarbon 10b (X=OAc, R=Me, Figure 5a), which showed helical control reduced to ~50%. Therefore, unlike alternating syn–anti linear hydrocarbon 7b, off-lattice conformations of the side-chains in all-syn helical systems, like 10b, do not mitigate the deleterious impact of these high-energy interactions on conformational control.

Figure 5.

Figure 5

Open in a new tab

Diamond lattice analyses of (a) all-syn hydrocarbon with 1,3-related ethyl substituents (10), which has two syn-pentane interactions in every helical conformer; (b) all-syn hydrocarbon with 1,3-related ethyl substituents flanking a methylene unit (11), which is devoid of syn-pentane interactions; (c) all-syn hydrocarbon with 1,3-related ethyl substituents flanking a methylene unit (12b), with inversion of configuration at one terminus to reinforce one screw-sense; (d) all-syn hydrocarbon with 1,2-related ethyl substituents (13), which has two syn-pentane interactions in every helical conformer; (e) all-syn hydrocarbon with 1,2-related ethyl substituents sandwiched between two methylene units (14), which is devoid of syn-pentane interactions; (f) hydrocarbons 12b and 15b taken forward for synthesis

The significant energy penalty imposed by the syn-pentane interactions in hydrocarbon 10 can be eliminated by replacing the C3 methyl group with a hydrogen atom, leading to methylene 11 (Figure 5b). As we have described previously,19 the preferred screw-sense (P or M) of an all-syn configured hydrocarbon is dictated by the configuration of the end-groups. An all-syn configured hydrocarbon with an even number of main-chain carbon atoms would have end-groups exhibiting the same configuration (R or S), thus inducing the same sense of helicity (P or M), a reinforcement that will engender a higher degree of helical control. On the other hand, a hydrocarbon with an odd number of main-chain carbon atoms will have end-groups with opposite configurations, which will induce opposing P and M screw-senses, thus creating internally defective (kinked) helical conformers and overall poor conformational control. Methylene 11, which has an odd number of main-chain carbon atoms with a mirror plane containing all three atoms of the CH2 group, would therefore be predicted to have poor conformational control.

Helical conformational control could be increased while retaining the odd number of main-chain carbon atoms by changing the configuration of one of the stereogenic centers at the termini (Figure 5c). For demonstration, we considered truncated structure 12b, which exhibits different configurations at the termini. Molecular mechanics conformational search with subsequent DFT calculation predicts that 85% of hydrocarbon 12b molecules would adopt helical main-chain conformations. with acetoxyethyl side-chains adopting a perfectly staggered conformation with respect to the main chain. This analysis gave us cause to take hydrocarbon 12b forward for synthesis. Notably, inversion of C5 could also have been avoided by designing an analogue with an even number of carbon atoms, and this strategy is demonstrated in the 1,2-disubstituted helical case, which follows.

Helical hydrocarbons with 1,2-related large substituents

On a diamond lattice, helical conformers of 1,2-diethyl hydrocarbon 13 exhibit syn-pentane interactions between the ethyl side-chains and the methyl groups on adjacent main-chain carbon atoms (Figure 5d). Replacement of these methyl groups for hydrogen atoms gives hydrocarbon 14, which should adopt a helical conformation devoid of syn-pentane-interactions (Figure 5e). Subjecting truncated 1,2-disubstituted acetoxyethyl hydrocarbon 15b (X=OAc) to molecular mechanics conformational search with subsequent DFT calculation predicts that 85% of hydrocarbon 15b molecules will adopt helical main-chain conformations comprising a 4:1 ratio of P/M helices. Like 1,3-substituted hydrocarbon 12b, the acetoxyethyl side chains of hydrocarbon 15b also adopt a perfectly staggered conformation with respect to the main chain.

Notably, the above analysis reveals that only an alternating syn–anti configurational pattern allows a contiguously substituted hydrocarbon bearing extended side-chains to retain the well-defined main-chain conformation exhibited by the all-methyl substituted parent (in this case, linear). This retained control is facilitated by the extended side-chains adopting off-diamond-lattice eclipsing interactions with the main chain, thus avoiding the high-energy syn-pentane interactions that would exist in a fully on-diamond-lattice conformation. The equivalent eclipsing interactions in a related all-syn contiguously substituted hydrocarbon imposes a higher energetic penalty because the methyl substituents on adjacent main-chain carbon atoms are in closer proximity to the extended side-chains, thus leading to a collection of near-degenerate helical and non-helical conformations and ultimately poor conformational control (e.g. 10 cf 7). Conversely, the analysis also reveals that only the all-syn configurational pattern can allow the conformation of hydrocarbons bearing extended side-chains to be controlled through the judicious removal of methyl substituents. Here, the main-chain helical conformation is the only conformation devoid of syn-pentane interactions. The equivalent pattern of extended side-chains and methylene units in the alternating syn–anti hydrocarbons leads to a collection of near degenerate linear and non-linear conformations devoid of syn-pentane interactions, thus leading to poor conformational control (e.g. 12 cf 6).

Having established design principles for linear and helical control of hydrocarbon motifs with either 1,2 and 1,3-related large substituents, hydrocarbons 8b, 7b, 15b, and 12b, which are all predicted to have strong preferences for either linear or helical conformations, were synthesised and analysed by NMR spectroscopy. Hydrocarbon 6b, which was predicted to have lower levels of (linear) conformational control was also synthesised as a control system.

Synthesis

The designed hydrocarbons were synthesized by using our lithiation–borylation methodology, which allows carbon chains to be grown, one carbon at a time, with complete control of configuration. 21 The substituted lithium carbenoid building blocks, (S)-18, (R)-18, were generated in situ through stereospecific Sn–Li exchange of highly enantioenriched α-stannyl 2,4,6-triisopropylbenzoate ester precursors, whilst (S)-26 and (R)-26 were generated in situ through stereospecific sulfoxide–Li exchange of the corresponding diastereo- and enantiopure sulfoxide precursors.22 Hydrocarbons 6b, 8b and 15b, which have C 2 symmetry, were synthesized by using a step-economic bidirectional strategy, where carbon chains were grown in both directions concurrently (Figure 6). Hydrocarbons 7b and 12b, which are non-symmetric, were synthesized by using the standard uni-directional strategy (Figure 7).

Figure 6. Bidirectional synthesis of C 2-symmetric hydrocarbons: (a) 8b; (b) 15b; (c) 6b.

Figure 6

Open in a new tab

Figure 7. Unidirectional synthesis of non-symmetric hydrocarbons: (a) 7b; (b) 12b.

Figure 7

Open in a new tab

For C 2-symmetric hydrocarbons 8b and 15b, we used Morken’s rhodium(I)-catalysed enantioselective diboration of alkenes to initiate the synthesis (Figure 6a/b).23 Thus, catalytic diboration of TBS-protected trans-3-hexene-1,6-diol 16 with B2(cat)2 gave the desired bis(boronic ester) 17 in 71% yield with high enantioselectivity (> 99:1 e.r.). Bidirectional iterative homologation24 of bis(boronic ester) 17 with in-situ generated lithium carbenoids (S)-18, (R)-18 and (R)-18 gave bis(boronic ester) 19 in 34% yield (three steps, 6 transformations) with > 95:5 d.r..17 Subsequent Zweifel olefination25 with 2-propenyl lithium followed by removal of the tert-butyldimethylsilyl protecting groups with TBAF gave diol 21 in 79% yield (two steps). Finally, hydrogenation and acylation gave desired hydrocarbon 8b, which exhibits 1,2-related large substituents within an anti–syn (linear) configurational pattern of substituents. .

Similarly, all-syn hydrocarbon 15b was prepared in high d.r. (>95:5) through iterative homologation of bis(boronic ester) 17 with lithium carbenoids 22 and (S)-18, followed by Zweifel olefination with 2-propenyl lithium, TBAF-mediated removal of the silyl groups, hydrogenation and acetylation (Figure 6b). The structure and relative configuration of intermediate diol 24 was confirmed by X-ray analysis. Notably, diol 24 adopted the expected helical conformation in the solid state.

Starting with diborylmethane 25, C 2-symmetric hydrocarbon 6b was prepared through four consecutive bidirectional homologation reactions with lithium carbenoids (R)-26 (×1), (R)-18 (×1) and (S)-18 (×2), without purification of intermediates between steps (40% yield), followed by Zweifel olefination with 2-propenyl lithium, TBAF-mediated removal of the silyl groups, acetylation and hydrogenation (47% yield, four steps, Figure 6c).

Non-symmetric hydrocarbon 7b was prepared through nine consecutive homologation reactions starting with commercially available iPrBpin 28 and using lithium carbenoids (S)-18, (R)-18, (S)-26 and (R)-26 in the appropriate sequence (~10% yield over 9 steps, Figure 7a). The desired functionality at the end-group and side-chains was established through Zweifel olefination with 2-propenyl lithium, TBAF-mediated removal of the silyl groups, hydrogenation and acetylation (30% yield over four steps).

Finally, non-symmetric hydrocarbon 12b was also prepared by using a unidirectional homologation strategy (Figure 7b). Firstly, (+)-sparteine-mediated enantioselective deprotonation of isobutyl 2,4,6-triisopropylbenzoate ester 33 gave a lithium carbenoid that was used to homologate MeBpin 32 to give boronic ester 34 in 74% yield and 98:2 e.r.. Four consecutive monodirectional homologations by using (R)-26, 22, (R)-26, and (R)-18, in that order, followed by Zweifel olefination with 2-propenyl lithium, TBAF-mediated removal of the silyl groups, acetylation and hydrogenation, gave desired hydrocarbon 12b.

NMR Analysis

With target linear and helical hydrocarbons 7b, 8b, 12b and 15b in hand, their conformational behaviour in solution was investigated by NMR spectroscopy. The experimental 1H chemical shifts for these compounds were extracted from 1H spectra or, where multiplets were heavily overlapped, from homonuclear decoupled PSYCHE and pure-shift HSQC spectra, which allowed discrimination of 1H chemical shifts in overlapping multiplets to <0.01 ppm. Assignment of diastereotopic protons was achieved by matching the corresponding NOE-distances (determined from 1D-NOESY spectra) for each proton with that predicted by DFT. 13C chemical shifts were obtained from 13C{1H} spectra and assigned using combinations of COSY, pure-shift HSQC and HMBC spectra. 1H–1H couplings constants were extracted from resolved multiplets by hand, and (with the precise chemical shifts from above) used to inform iterative full spin-simulations of overlapped multiplet regions of the spectrum. Examples of these simulations and the quality of fit can be found in the Supporting Information.1H–13C scalar coupling constants were extracted from IPAP–HSQMBC spectra and interproton NOE-distances were measured from 1D-NOESY spectra on resolved multiplets, using CSSF-chemical shift selection to maximise the number of multiplets that could be studied. Full details of the NMR spectroscopic analysis and images of the spectra taken can be found in the Supporting Information.

DFT-calculated 1H–1H and 1H–13C scalar coupling constants, and interproton NOE-distances, were population averaged according to the DFT-calculated energies of conformers in order to generate experimentally comparable values for each NMR parameter (see SI for details). These predicted population-averaged NMR parameters were then compared to the corresponding experimentally measured values (Figure 8). For all four hydrocarbons, the goodness of fit was in line with that observed by us previously in conformational analyses of structurally complex hydrocarbons of this class17,19 (1H–1H coupling: mean absolute deviation (MAD) = 0.4–0.7 Hz, standard deviation (s.d.) = 0.5–0.9 Hz; 1H–13C coupling: MAD = 0.3–0.6 Hz, s.d. = 0.4–0.7 Hz; NOE distances: MAD = 2–5%, s.d. = 3–6%). This analysis confirms that the conformational population of hydrocarbons 8b, 7b, 15b and 12b in solution is in line with that predicted by DFT. Notably, we are not claiming that the precise DFT-predicted energies (and thus populations) of each conformer are exact; however, the quality-of-fit values demonstrates that the DFT-predicted conformational space of each molecule is broadly in line with the experiment. Thus, the design principles on which these molecules were based is validated. Conducting the same DFT/NMR analysis on control molecule 6b, where linear conformers were predicted to be <50% of the population, provided comparable MADs for all NMR parameters. Notably, the standard deviations for 1H–1H and 1H–13C scalar coupling constants (1.12 Hz and 0.72 Hz, respectively) for 6b were slightly higher than those determined for the more controlled molecules. These larger deviations are consistent with a larger number of significantly populated conformers and thus a greater contribution from errors in DFT-estimated energies and population.

Figure 8. Comparison of DFT-predicted and experimental NMR parameters (J-coupling and NOE-distances) for hydrocarbons 7b, 8b, 12b, and 15b.

Figure 8

Open in a new tab

Conclusion

The conformation of hydrocarbon chains bearing methyl substituents can be well controlled: the alternating syn–anti isomer adopts a linear conformation and the all-syn isomer adopts a helical conformation. The preferred conformations arise from the avoidance of destabilizing syn-pentane interactions. However, any group larger than a methyl group on the backbone inevitably leads to syn-pentane interactions if the molecule adopts a fully staggered diamond-lattice conformation, thus implying reduced conformational control and limited applications. Therefore, an understanding of how conformational control can be maximized as larger substituents are incorporated is required for the use of these substituted hydrocarbons as scaffolds for applications in materials and medicine. We have identified a selection of design principles that allow larger groups (Et/acetoxyethyl) to be incorporated along the hydrocarbon chain whilst maintaining the linear/helical conformational control.

Figure 9 summarizes the outcomes for these systems. The well-controlled linear conformation of alternating syn–anti contiguously-substituted hydrocarbons is retained (>95% of the population adopt linear conformers) by replacing methyl substituents with larger acetoxyethyl substituents in a way that presents those larger substituents in a 1,2-syn or a 1,3-anti relationship. Here, syn-pentane interactions are reduced by the large substituents adopting an eclipsed conformation with respect to the main chain. The replacement of methyl substituents for hydrogen atoms in the vicinity of the large substituents, with the intention of facilitating perfectly staggered 1,2-relationships devoid of syn-pentane interactions, leads to a substantial drop in conformational control of the main chain. In the case of the all-syn substituted hydrocarbons, incorporating larger substituents in the available 1,2-syn and 1,3-syn positions requires the concurrent replacement of methyl groups for hydrogen atoms in the immediate vicinity (on either side of the 1,2-substituted unit or on the central carbon atom of 1,3-substituted units) so as to maintain high levels of helical conformational control (>85%). These results show that it is possible to design and synthesize linear and helical hydrocarbon chains bearing substituents larger than a methyl group. This information will enable the design of flexible carbon chains that adopt a particular low-energy conformation displaying substituents that, for example, have the functionality necessary for multivalent binding to a biological receptor or to hotspots on a large biological surface. Furthermore, investigation of the membrane permeability of substituted hydrocarbons endowed with polar and non-polar substituents and terminal groups may reveal interesting insights on contribution of flexibility and conformational control to this important biological transport process.

Figure 9. Structures and diamond lattice overlays of synthesized molecules 7b, 8b, 12b, and 15b, highlighting the structural features that lead to their conformation control.

Figure 9

Open in a new tab

Acknowledgment

We thank H2020 ERC (670668) for financial support. LG (GU 1935/1-1) and MK thank the DFG and OW thanks SNSF for research fellowships. We thank N. Pridmore and H. Sparkes (UoB) for assistance with the X-ray analysis and W. Gerrard (UoB) for his contribution to data processing. M.D. thanks C4X Discovery and the EPSRC National Productivity Investment Fund (NPIF) for Doctoral Studentship funding.

Footnotes

Author Contributions

The manuscript was written through contributions of all authors and all authors have given approval to the final version of the manuscript.

Supporting Information

Details of computational, synthetic and spectroscopic methodology and results.

References

  • (1).Tsuzuki S, Schafer L, Goto H, Jemmis ED, Hosoya H, Siam K, Tanabe K, Osawa E. Investigation of intramolecular interactions in n-alkanes. Cooperative energy increments associated with GG and GTG’ sequences. J Am Chem Soc. 1991;113:4665–4671. [Google Scholar]
  • (2).Scott RA, Scheraga HA. Conformational Analysis of Macromolecules. II. The Rotational Isomeric States of the Normal Hydrocarbons. J Chem Phys. 1966;44:3054–3069. [Google Scholar]
  • (3).Abe A, Jernigan RL, Flory PJ. Conformational Energies of n-Alkanes and the Random Configuration of Higher Homologs Including Polymethylene. J Am Chem Soc. 1966;88:631–639. [Google Scholar]
  • (4).Sykora S. Collect. Semiempirical calculation of conformational structure of 2,4-disubstituted pentanes. Czech Chem Commun. 1968;33:3514–3527. [Google Scholar]
  • (5).Wiberg KB, Murcko MA. Rotational barriers. 2. Energies of alkane rotamers. An examination of gauche interactions. J Am Chem Soc. 1988;110:8029–8038. [Google Scholar]
  • (6).Smith PW, Still WC. The effect of substitution and stereochemistry on ion binding in the polyether ionophore monensin. J Am Chem Soc. 1988;110:7917–7919. [Google Scholar]
  • (7).Goto H, Osawa E, Yamato M. How many conformers are there for small n-alkanes? Consequences of asymmetric deformation in GG’ segment. Tetrahedron. 1993;49:387–396. [Google Scholar]
  • (8).Hoffmann RW. Conformation Design of Open-Chain Compounds. Angew Chem Int Ed. 2000;39:2054–2070. [PubMed] [Google Scholar]
  • (9).Nilewski C, Geisser RW, Ebert M-O, Carreira EM. Conformational and Configurational Analysis in the Study and Synthesis of Chlorinated Natural Products. J Am Chem Soc. 2009;131:15866–15876. doi: 10.1021/ja906461h. [DOI] [PubMed] [Google Scholar]
  • (10).O'Hagan D. Polar Organofluorine Substituents: Multivicinal Fluorines on Alkyl Chains and Alicyclic Rings. Chem Eur J. 2020;26:7981–7997. doi: 10.1002/chem.202000178. [DOI] [PubMed] [Google Scholar]
  • (11).O'Hagan D. Organofluorine Chemistry: Synthesis and Conformation of Vicinal Fluoromethylene Motifs. J Org Chem. 2012;77:3689–3699. doi: 10.1021/jo300044q. [DOI] [PubMed] [Google Scholar]
  • (12).O'Hagan D. Understanding organofluorine chemistry. An introduction to the C–F bond. Chem Soc Rev. 2008;37:308–319. doi: 10.1039/b711844a. [DOI] [PubMed] [Google Scholar]
  • (13).Hoffmann RW. Flexible molecules with defined shape – conformational design. Angew Chem Int Ed. 1992;31:1124–1134. [Google Scholar]
  • (14).Hoffmann RW, Stahl M, Schopfer U, Frenking G. Conformation design of hydrocarbon backbones: a modular approach. Chem Eur J. 1998;4:559–566. [Google Scholar]
  • (15).Hoffmann RW, Göttlich R, Schopfer U. Conformation Induction Between Neighboring Dimethylpentane Segments. Eur J Org Chem. 2001:1865–1871. [Google Scholar]
  • (16).Alder RW, Maunder CM, Orpen AG. The conformational effects of quaternary centres. Tetrahedron Lett. 1990;31:6717–6720. [Google Scholar]
  • (17).Burns M, Essafi S, Bame J, Bull SP, Webster MP, Balieu S, Dale SW, Butts CP, Harvey JN, Aggarwal VK. Assembly-line synthesis of organic molecules with tailored shapes. Nature. 2014;513:183–188. doi: 10.1038/nature13711. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (18).Foley AR, Raskatov JA. Understanding and controlling amyloid aggregation with chirality. Curr Opin Chem Biol. 2021;64:1–9. doi: 10.1016/j.cbpa.2021.01.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (19).Pradeilles JA, Zhong S, Baglyas M, Tarczay G, Butts CP, Myers EL, Aggarwal VK. Odd–even alternations in helical propensity of a homologous series of hydrocarbons. Nat Chem. 2020;12:475–480. doi: 10.1038/s41557-020-0429-0. [DOI] [PubMed] [Google Scholar]
  • (20).For simplicity we define ‘linear’ structures as those where all the main chain dihedral angles are within 30° of anti arrangements and ‘helical’ structures as those where all the main chain dihedral angles are within 30° of repeating perfect anti/gauche(plus) or anti/gauche(minus) arrangements.
  • (21).Stymiest JL, Dutheuil G, Mahmood A, Aggarwal VK. Lithiated Carbamates: Chiral Carbenoids for Iterative Homologation of Boranes and Boronic Esters. Angew Chem Int Ed. 2007;46:7491–7494. doi: 10.1002/anie.200702146. [DOI] [PubMed] [Google Scholar]
  • (22).Casoni G, Kucukdisli M, Fordham JM, Burns M, Myers EL, Aggarwal VK. α-Sulfinyl Benzoates as Precursors to Li and Mg Carbenoids for the Stereoselective Iterative Homologation of Boronic Esters. J Am Chem Soc. 2017;139:11877–11886. doi: 10.1021/jacs.7b05457. [DOI] [PubMed] [Google Scholar]
  • (23).Morgan JB, Miller SP, Morken JP. Rhodium-Catalyzed Enantioselective Diboration of Simple Alkenes. J Am Chem Soc. 2003;125:8702–8703. doi: 10.1021/ja035851w. [DOI] [PubMed] [Google Scholar]
  • (24).Blair DJ, Tanini D, Bateman JM, Scott HK, Myers EL, Aggarwal VK. Selective Uni- and Bidirectional Homologation of Diborylmethane. Chem Sci. 2017;8:2898–2903. doi: 10.1039/c6sc05338f. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (25).Armstrong RJ, Aggarwal VK. 50 Years of Zweifel Olefination: A Transition-Metal-Free Coupling. Synthesis. 2017;49:3323–3336. [Google Scholar]

RESOURCES