Skip to main content
NCBI home page
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2024 Oct 4;146(41):28580–28588. doi: 10.1021/jacs.4c11869

Solvation-Enhanced Salt Bridges

Ben Iddon 1, Christopher A Hunter 1,*
PMCID: PMC11487562  PMID: 39364787

Abstract

graphic file with name ja4c11869_0012.jpg

Salt bridges formed by amidines and carboxylic acids represent an important class of noncovalent interaction in biomolecular and supramolecular systems. Isothermal titration calorimetry was used to study the relationships between the strength of the interaction, the chemical structures of the components, and the nature of the solvent. The stability of the 1:1 complex formed in chloroform changed by 2 orders of magnitude depending on the basicity of the amidine and the acidity of the acid, which is consistent with proton transfer in the complex. Polar solvents reduce the stabilities of salt bridges formed with N,N’-dialkylamidines by up to 3 orders of magnitude, but this dependence on solvent polarity can be eliminated if the alkyl groups are replaced by protons in the parent amidine. The enhanced stability of the complex formed by benzamidine is due to solvation of the NH sites not directly involved in salt bridge formation, which become significantly more polar when proton transfer takes place, leading to more favorable interactions with polar solvents in the bound state. Calculation of H-bond parameters using density functional theory was used to predict solvent effects on the stabilities of salt bridges to within 1 kJ mol–1. While H-bonding interactions are strong in nonpolar solvents, and solvophobic interactions are strong in polar protic solvents, these interactions are weak in polar aprotic solvents. In contrast, amidinium–carboxylate salt bridges are stable in both polar and nonpolar aprotic solvents, which is attractive for the design of supramolecular systems that operate in different solvent environments.

Introduction

Salt bridges represent an important class of noncovalent interactions that involve both H-bonding and ion-pairing when a cationic H-bond donor interacts with an anionic H-bond acceptor. These interactions play a pivotal role in biological systems, particularly in protein folding, protein-nucleic acid recognition, and medicinal chemistry.1−4 The amidinium–carboxylate salt bridge has been widely used in synthetic supramolecular systems due to the large association constants found in nonpolar solvents and the well-defined geometry dictated by two cooperative H-bonds (Figure 1). Applications include sensing,5,6 crystal engineering,7,8 catalysis,9 H-bonded organic frameworks,10 polymer chemistry,11,12 self-replicating systems,13 self-assembly of duplexes and capsules,14−17 and template synthesis.18,19 In contrast to other noncovalent interactions that have been the subject of quantitative systematic studies,20−23 salt bridges have received relatively little attention. Reliable implementation of noncovalent chemistry in molecular design requires an understanding of the relationships between the strength of the interaction, the chemical structures of the components, and the nature of the solvent.24 Here, we use the formation of salt bridges between a series of carboxylic acid and benzamidine derivatives in a range of organic solvents to establish these principles.

Figure 1.

Figure 1

Open in a new tab

Salt bridge interaction between a benzoic acid and a benzamidine derivative. AH and B represent an acid and base, respectively, and X, Y, and R are substituents.

The combination of H-bonding and ion-pairing involved in the formation of a salt bridge means that multiple equilibria may be involved, as illustrated in Figure 1. Proton transfer may take place to different extents before and after the formation of the H-bonds in the salt bridge, so there is a complex interplay of acid–base chemistry and solvation, as well as the effects of charge on the strength of the H-bonds.25 Here, we show that, starting from the neutral species (top left in Figure 1), the stability of the salt bridge interaction in chloroform can be modulated by 2 orders of magnitude depending on the X and Y substituents. Polar solvents reduce the stabilities of salt bridges formed with N,N’-dialkylamidines by up to 3 orders of magnitude, but this dependence on solvent polarity can be practically eliminated if the alkyl groups are replaced by protons in the parent amidine (R = H in Figure 1). We show that this unusual property of salt bridges formed by amidines and carboxylic acids is due to changes in the solvation shell associated with proton transfer that takes place within the salt bridge.

Results and Discussion

Synthesis

All of the benzoic acids investigated were commercially available (X = H, NMe2, OMe, CF3, and NO2 in Figure 1). The synthesis of 4-substituted benzamidines was carried out by the routes shown in Scheme 1. 4-Hydroxybenzonitrile was first alkylated with 2-ethylhexyl bromide, and subsequent treatment with acetyl chloride in methanol followed by methanolic ammonia solution gave 4-alkoxybenzamidine 1. 4-Mercaptobenzonitrile was similarly alkylated and then oxidized with 3-chloroperbenzoic acid (mCPBA) to give the 4-sulfonylbenzonitrile 2. Conversion of 2 to the corresponding amidine gave 4-sulfonylbenzamidine 3.

Scheme 1. Synthesis of 4-Substituted Benzamidines (R’ = 2-Ethylhexyl).

Scheme 1

Open in a new tab

Conditions: (a) 1. 2-ethylhexyl bromide; 2. AcCl, MeOH then NH3; (b) 1. 2-ethylhexyl bromide; 2. mCPBA; (c) AcCl, MeOH then NH3.

N,N′-Dialkylbenzamidines 4–7 were synthesized by the routes shown in Scheme 2. The N,N′-dimethyl and N,N′-diethyl derivatives 4 and 5 were prepared by alkylation of benzonitrile with the relevant alkyl triflate, followed by reaction with the corresponding primary amine. The N,N′-di-i-propyl and N,N′-di-t-butyl derivatives 6 and 7 were obtained by reaction of the corresponding carbodiimide with phenyl magnesium bromide.

Scheme 2. Synthesis of N,N’-Dialkylbenzamidines.

Scheme 2

Open in a new tab

Conditions: (a) 1. ROTf; 2. RNH2; (b) phenyl magnesium bromide.

Effect of Aromatic Substituents

The interaction of pairwise combinations of benzoic acid and benzamidine derivatives was investigated in chloroform solution using isothermal titration calorimetry (ITC). In each case, the titration data fit well to a 1:1 binding isotherm (see Supporting Information for details), and the resulting thermodynamic parameters are summarized in Table 1.

Table 1. Substituent Effects on the Thermodynamic Parameters for Salt Bridge Formation between Benzoic Acid and Benzamidine Derivatives Determined by ITC in CHCl3 at 298 Ka.

X Yb R Log (K/M–1) ΔG°/kJ mol–1 ΔH°c/kJ mol–1 ΔS°c/J K–1 mol–1 N
H H H 6.01 ± 0.05 –34.3 ± 0.3 –70.0 ± 3.0 –120 ± 10 1.07 ± 0.06d
NMe2 H H 5.10 ± 0.10 –29.1 ± 0.6 –60.0 ± 0.8 –104 ± 5 1.08 ± 0.06d
OMe H H 5.69 ± 0.03 –32.5 ± 0.2 –66.0 ± 3.0 –110 ± 10 1.10 ± 0.01d
CF3 H H 6.51 ± 0.02 –37.1 ± 0.1 –76.0 ± 1.0 –130 ± 4 0.80 ± 0.01d
NO2 H H 6.84 ± 0.09 –39.0 ± 0.5 –75.4 ± 0.6 –122 ± 4 0.90 ± 0.10d
H OR’ H 6.55 ± 0.09 –37.4 ± 0.5 –70.0 ± 10.0 –110 ± 30 1.20 ± 0.20d
H SO2R’ H 4.92 ± 0.03 –28.1 ± 0.2 –61.0 ± 1.0 –111 ± 4 0.90 ± 0.01e
H H Me 6.37 ± 0.05 –36.3 ± 0.3 –73.0 ± 9.0 –120 ± 30 1.20 ± 0.1e
H H Et 6.52 ± 0.09 –37.2 ± 0.5 –74.0 ± 3.0 –120 ± 10 1.20 ± 0.1d
H H iPr 6.40 ± 0.20 –37.0 ± 1.0 –75.0 ± 1.0 –128 ± 7 1.00 ± 0.1e
H H tBu 6.36 ± 0.08 –36.3 ± 0.5 –80.0 ± 3.0 –150 ± 10 1.17 ± 0.03e
Open in a new tab
a

Errors are twice the standard deviation of at least two repeat measurements.

b

R’ = 2-ethylhexyl.

c

Errors in ΔH° and ΔS° do not take into account the uncertainty in N.

d

N is the number of amidine molecules bound to one carboxylic acid in the complex.

e

N is the number of carboxylic acid molecules bound to one amidine in the complex.

The nature of the aromatic substituents X and Y has a large impact on the stability of the 1:1 complex formed in chloroform, and the association constants span almost 2 orders of magnitude. The stability of the complex increases for electron-withdrawing groups on the carboxylic acid and electron-donating groups on the amidine. Figure 2 shows that the association constants correlate well with the Hammett substituent parameter σ, which measures the effect of substituents on the acidity of the corresponding benzoic acid. The results show that the strength of the interaction between a neutral carboxylic acid and a neutral amidine can be directly and predictably tuned by changing the acidity and/or basicity of the interacting partners.

Figure 2.

Figure 2

Open in a new tab

Hammett plots showing the relationship between the association constant for salt bridge formation measured in chloroform at 298 K and substituents on the benzoic acid (X, green) or the benzamidine (Y, orange). KHH is the association constant for X = Y = H. The lines of best fit were fixed to pass through the origin and correspond to Log(KX/KHH) = 1.05 σX and Log(KY/KHH) = −1.57 σY.

The slopes of the Hammett plots in Figure 2, ρ, quantify the complexation-induced changes in charge on the carboxylic acid and amidine groups. The values of ρ are large and positive for X and large and negative for Y, which indicate a substantial change in charge for both partners when the salt bridge is formed. These observations suggest that a proton is transferred from the carboxylic acid to the amidine in the salt bridge, which exists in the zwitterionic form even in nonpolar solvents like chloroform. These results might be interpreted as a simple proton transfer reaction between the acid and amidine, generating the ionized species without forming an intermolecular complex. However, 1H NMR DOSY experiments in acetonitrile, a more polar solvent that would stabilize the separated ions better than chloroform, show that the complex is fully assembled in a 1:1 mixture at millimolar concentrations (see Supporting Information for details).

Effect of N-Alkyl Amidine Substituents

N,N′-Dialkylamidines have been commonly used in supramolecular systems because they can easily be made from N,N’-dialkylcarbodiimides and show increased solubility in organic solvents compared with the parent amidines.26Figure 3 shows how the stability of the salt bridge formed with benzoic acid is affected by N-alkyl substituents (R) of increasing steric bulk in chloroform. Alkylation of the amidine increases the stability of the complex relative to the parent amidine (R = H), but the association constants measured for all four N,N’-dialkylamidines are the same within experimental error. The steric bulk of the alkyl groups does not play a role in determining the salt bridge stability. By analogy with the results for substituent effects described above, the increase in stability observed for the N,N’-dialkylamidines is most likely due to the higher basicity of more substituted nitrogen atoms.

Figure 3.

Figure 3

Open in a new tab

Effect of benzamidine N,N’-dialkyl substituents (R) on the association constant for salt bridge formation with benzoic acid measured in chloroform at 298 K.

Solvent Effects

Although association constants for amidinium–carboxylate salt bridges have previously been measured in different solvents and in solvent mixtures,16,27−31 no systematic study has been attempted. To investigate the role of the solvent, ITC titrations were used to measure the interaction of formic acid with benzamidine (Y = H, R = H) and with N,N′-dimethylbenzamidine (Y = H, R = Me) in eight organic solvents with a wide range of polarities. Polar protic solvents were not included in these experiments because of the increased probability of proton transfer between solute and solvent, which would change the nature of the free species on the left-hand side of the equilibrium (for details, see eq 7 and the associated discussion later in the text). Table 2 summarizes the thermodynamic parameters obtained from the ITC experiments, and Figure 4 compares the stabilities of the complexes formed by the two different amidines. The association constant decreases with increasing solvent polarity for both benzamidine (red) and N,N′-dimethylbenzamidine (blue), but the behavior of the two systems is quite different. The benzamidinium–formate complex is much less sensitive to the solvent polarity than the N,N’-dimethylbenzamidinium–formate complex. Benzamidine forms a slightly less stable salt bridge than N,N’-dimethylbenzamidine in chloroform, but N-alkylation leads to a decrease in the association constant by 2 orders of magnitude in THF and DMF compared with the parent benzamidine.

Table 2. Solvent Effects on the Thermodynamic Parameters for Salt Bridge Formation between Formic Acid and Benzamidine Derivatives Determined by ITC at 298 Ka.

Solvent R Y Log (K/M–1) ΔG°/kJ mol–1 ΔH°b/kJ mol–1 ΔS°b J/K–1 mol–1 Nc
CHCl3 H H 6.26 ± 0.05 –35.7 ± 0.3 –58.0 ± 1.0 –75 ± 4 1.01 ± 0.01
PhOMe H H 5.90 ± 0.20 –34.0 ± 1.0 –67.8 ± 0.1 –113 ± 4 0.80 ± 0.30
MeCN H H 5.73 ± 0.06 –32.7 ± 0.3 –63.0 ± 5.0 –102 ± 18 0.90 ± 0.20
Acetone H H 5.50 ± 0.10 –31.4 ± 0.6 –64.0 ± 2.0 –109 ± 9 0.89 ± 0.03
EtOAc H H 5.37 ± 0.01 –30.6 ± 0.1 –65.0 ± 4.0 –120 ± 10 0.89 ± 0.03
DME H H 5.34 ± 0.03 –30.5 ± 0.2 –68.0 ± 2.0 –126 ± 7 0.96 ± 0.04
THF H H 4.96 ± 0.07 –28.3 ± 0.4 –61.0 ± 0.4 –110 ± 3 0.79 ± 0.06
DMF H H 4.94 ± 0.05 –28.2 ± 0.3 –54.9 ± 0.3 –90 ± 2 0.97 ± 0.01
CHCl3 Me H 6.60 ± 0.05 –37.7 ± 0.3 –60.0 ± 10 –75 ± 30 0.93 ± 0.03
PhOMe Me H 5.40 ± 0.01 –30.8 ± 0.1 –51.0 ± 0.1 –68 ± 1 0.88 ± 0.01
MeCN Me H 4.50 ± 0.07 –25.7 ± 0.4 –53.0 ± 5.0 –90 ± 20 0.95 ± 0.07
Acetone Me H 4.14 ± 0.06 –23.6 ± 0.3 –37.0 ± 2.0 –45 ± 8 1.03 ± 0.01
EtOAc Me H 4.04 ± 0.05 –23.0 ± 0.3 –44.0 ± 0.7 –70 ± 3 0.93 ± 0.04
DME Me H 3.60 ± 0.02 –20.5 ± 0.1 –46.0 ± 4.0 –90 ± 10 1.00d
THF Me H 2.93 ± 0.05 –16.7 ± 0.3 –23.0 ± 2.0 –21 ± 8 1.00d
DMF Me H 3.34 ± 0.07 –19.1 ± 0.4 –50.0 ± 10.0 –100 ± 30 1.00d
Open in a new tab
a

Errors are twice the standard deviation of at least two repeat measurements.

b

Errors do not take into account the uncertainty in N.

c

N is the number of carboxylic acid molecules bound to one amidine in the complex.

d

Titrations carried out in the low c-value regime were analyzed with N fixed at 1.00.32.

Figure 4.

Figure 4

Open in a new tab

Solvent effects on the association constant for salt bridge formation between formic acid and benzamidine (red) or N,N’-dimethylbenzamidine (blue).

No correlation was found between the association constants and common descriptors of the solvent polarity. For example, Figure 5a shows the relationship between the association constants and the solvent dielectric constant, εr, which indicates that the ionizing power of the solvent is not an important factor governing the observed solvent effects. Figure 5b shows the relationship with solvent polarity parameter ET(30), which highlights the failure of bulk solvent descriptors to account for the observed solvent effects. Similar results were found for the Hansen solubility parameters (see Supporting Information). However, a correlation was found between the solvent hydrogen bond acceptor parameter, βS, and the difference between the free energy changes for the formation of the benzamidine and N,N’-dimethylbenzamidine complexes (ΔΔG°, Figure 5c). The enhanced stability of the benzamidine complex in polar solvents is therefore related to interactions between solvent H-bond acceptors and H-bond donor sites in the benzamidine complex that are not present in the N,N’-dimethyl complex. Since the difference between the two complexes is simply the replacement of two NH H-bond donor sites with nonpolar methyl groups, this observation suggests that a more explicit analysis of the details of the solvation shell may shed light on the effect of solvent on salt bridge stability.

Figure 5.

Figure 5

Open in a new tab

Comparison of the association constants for salt bridge formation between formic acid and benzamidine (red) or N,N’-dimethylbenzamidine (blue) with (a) solvent dielectric constant, εr, and (b) solvent polarity, ET(30).37 (c) Comparison of the difference between the free energy changes for the formation of the benzamidine and N,N’-dimethylbenzamidine complexes (ΔΔG°) with the solvent hydrogen bond acceptor parameter (βS); R2 = 0.81. (See the Supporting Information for details.)

Solution-phase complexation can be described by a solvent competition model that uses the parameters α and β to quantify the noncovalent interaction properties of solute and solvent functional groups.24,33−35 These parameters can be determined by experiment or calculated using density functional theory (DFT) molecular electrostatic potential surfaces in conjunction with a footprinting algorithm described previously.36 The free energy contribution due to an intermolecular interaction between two functional groups is given in eq 1.

graphic file with name ja4c11869_m001.jpg 1

This approach can be used to build up a quantitative picture of the solvent–solute interactions that govern the behavior of the salt bridges. Figure 6 compares the primary interactions in the solvation shells of the free and bound species involved in the formation of the two different salt bridges. The major difference between the two complexes relates to the solvation of the two peripheral NH sites that are not directly involved in the solute–solute H-bonding interactions (the solvation interactions are highlighted in blue in Figure 6). These NH groups become significantly more polar in the zwitterionic complex compared with the neutral free state, which will lead to stronger interactions with the solvent in the bound state. It is the change in solvation of these peripheral sites that accounts for the difference in behavior between benzamidine and N,N’-dimethylbenzamidine. Stronger solvation of the peripheral NH protons upon formation of the salt bridge enhances the stability of the benzamidine complex by almost 2 orders of magnitude in DMF compared with the N,N’-dimethylbenzamidine complex.

Figure 6.

Figure 6

Open in a new tab

Primary solute–solvent interactions in the solvation shells of the free and bound species involved in the formation of (a) the benzamidinium–formate salt bridge and (b) the N,N’-dimethylbenzamidinium–formate salt bridge. The major difference is due to the solvation interactions highlighted in blue.

The change in solvation free energy between free and bound species on formation of the salt bridge (ΔΔG°solv) can be calculated in terms of the H-bond parameters for the solvent and solute by summing the free energy contributions of each pairwise interaction shown in Figure 6 (eq 2).

graphic file with name ja4c11869_m002.jpg 2

where αf, βf, αb, and βb are the H-bond parameters of the sites on the free and bound solutes, and αS and βS are the H-bond parameters of the solvent.

H-bond parameters for all of the sites on the free and bound solutes were calculated or obtained from experimental data (Figure 7). The calculated H-bond parameters confirm that the peripheral NH protons and oxygen lone pairs that are not directly involved in the salt bridge H-bonds become significantly better hydrogen bond donors and acceptors, respectively, when the proton is transferred in the salt bridge. Using these H-bond parameters in eq 2, it is possible to estimate how differences in solvation energy affect the relative stability of the two different salt bridges (eqs 3 and 4).

graphic file with name ja4c11869_m003.jpg 3
graphic file with name ja4c11869_m004.jpg 4

Figure 7.

Figure 7

Open in a new tab

Free and bound solute H-bond parameters calculated using DFT (the carboxylic acid parameters labeled with an asterisk were obtained from experimental data).35,36

The coefficients of the solvent H-bond parameters in eqs 3 and 4 predict that the N,N’-dimethyl complex should be significantly more sensitive to solvent polarity than the complex formed with the parent benzamidine, particularly with respect to the H-bond acceptor properties of the solvent, which is consistent with the experimental observations described above. Figure 8 compares the change in solvation energy predicted by eqs 3 and 4 (see Supporting Information for solvent H-bond parameters) with the experimental values of the free energy change for the formation of the salt bridge in different solvents. There is an excellent correlation for both the benzamidine (R = H, red) and N,N’-dimethylbenzamidine (R = Me, blue) complexes, and the slope of the line of best fit is 1.0 in both cases. In other words, the experimentally observed solvent effects on the stability of the salt bridge interactions are almost perfectly described by the primary solvation model in Figure 6.

Figure 8.

Figure 8

Open in a new tab

Comparison of the experimental free energy changes for the formation of salt bridges in different solvents with the associated change in solvation free energy calculated using eqs 3 and 4 (ΔΔG°solv). The lines of best fit are ΔG°expt = 38.2 + 1.00 ΔΔG°solv (R = H, red, RMSE = 1 kJ mol–1) and ΔG°expt = 42.3 + 0.95 ΔΔG°solv (R = Me, blue, RMSE = 1 kJ mol–1).

The intercepts on the ΔG°expt axis in Figure 8 represent the intrinsic stabilities of the salt bridges in a completely nonpolar solvent (i.e., ΔΔG°solv = 0) and give −38 kJ mol–1 for the benzamidinium–formate complex and −44 kJ mol–1 for the N,N’-dimethylbenzamidinium–formate complex. Using the change in solvation energy from eqs 3 and 4 together with these values allows prediction of the stability of the salt bridge relative to the neutral carboxylic acid and amidine in any solvent for which the H-bond parameters are available (eqs 5 and 6). The only caveat is that the solutes should not be significantly ionized in the free state; otherwise, the competing equilibria shown in Figure 1 would complicate the analysis. Figure 9 compares the association constants calculated using eqs 5 and 6 with those obtained experimentally.

graphic file with name ja4c11869_m005.jpg 5
graphic file with name ja4c11869_m006.jpg 6

Figure 9.

Figure 9

Open in a new tab

Comparison of experimentally determined association constants for the formation of salt bridges with formic acid in different solvents (Kexpt) with the corresponding values calculated using eqs 5 and 6 (Kcalc): benzamidine complex in red, and N,N’-dimethylbenzamidine complex in blue. The dashed line is y = x (RMSE = 0.25).

Equation 5 predicts that the association constant for the benzamidinium–formate salt bridge should be 6 × 107 M–1 in water (αS = 2.8, βS = 4.5). However, experimental measurements for similar systems, e.g., the guanidinium–acetate salt bridge, show that salt bridges are much less stable in water (K < 1 M–1).38,39 The discrepancy comes from differences in the nature of the free species. The experiments described here were all carried out in organic solvents in which ionization of the free carboxylic acid and free amidine is negligible. In water, the ionized species are substantially populated in the free state (Figure 1), and these competing equilibria must be considered in the estimation of the overall association constant for the formation of the salt bridge. The association constant for the formation of a salt bridge (K) can be expressed in terms of the equilibrium constants for protonation of the amidine (KA), deprotonation of the acid (KC), and formation of the salt bridge starting from the neutral species (KN) (eq 7).

graphic file with name ja4c11869_m007.jpg 7

The equilibrium constants for ionization of the free species in water can be calculated from the acidity constants of benzamidinium (pKa = 11.6)40 and formic acid (pKa = 3.8)41 giving KA = 4.0 × 104 and KC = 1.6 × 103 at pH 7. Using these values in eq 7 together with the association constant predicted by using eq 5 (KN) gives an association constant of K = 0.9 M–1 for the benzamidinium–formate salt bridge in water at pH 7, which is consistent with the experimental observations. The difference between the behavior in water and in organic solvents lies in the ability of water to strongly solvate both anions and cations in the free state, whereas polar aprotic solvents solvate anions poorly.

Analogous behavior is observed in aprotic solvents if salts of the acid and amidine are used as the starting materials instead of the neutral species. For example, the association constant for the salt bridge formed on mixing tetrabutylammonium benzoate and benzamidinium chloride in dimethyl sulfoxide (DMSO) is 2,500 M–1.7 Although we have not measured association constants in DMSO, it is possible to predict the value with eq 5 using αS= 1.4 and βS = 8.6. The calculated association constant for the formation of the benzamidinium–formate complex from the neutral species is 3 × 105 M–1. The experimental value reported above is 2 orders of magnitude lower due to the two competing ion-pair equilibria in the free state, analogous to the competing ionization equilibria described by eq 7.

Conclusions

A systematic investigation of factors affecting the strength of the amidinium–carboxylate salt bridge was carried out by measuring association constants for 27 different systems using isothermal titration calorimetry. Aromatic substituent effects show that electron-rich amidines and electron-poor carboxylic acids form the most stable complexes, suggesting that there is extensive proton transfer on the salt bridge formation. The steric size of the alkyl substituents on the nitrogen atoms of the amidine does not affect the stability of the salt bridge.

The results show that the solvent plays an important role in determining the stability of salt bridges, and the effects cannot be explained with bulk solvent descriptors. The complex of formic acid with N,N’-dimethylbenzamidine shows surprisingly different behavior compared to to the corresponding complex formed with benzamidine. In chloroform, the presence of methyl groups increases the stability of the salt bridge slightly. In more polar solvents, there is a decrease of 3 orders of magnitude in the stability of the N,N’-dimethylbenzamidine complex, whereas the association constant measured for the formation of the benzamidinium–formate salt bridge is between 105 and 106 M–1 in eight different solvents, ranging in polarity from chloroform to dimethylformamide.

The increase in stability of the benzamidine complex relative to the N,N’-dimethyl analogue correlates with the solvent H-bond acceptor parameter βS, which indicates that the stabilization is due to interactions with the two additional NH H-bond donor sites that are present in the benzamidine complex. There is a substantial increase in the polarity of these two peripheral NH groups when the proton transfer takes place in the benzamidine salt bridge, and the associated increase in free energy contributions due to H-bonding interactions with polar solvents stabilizes the complex. In very polar solvents (THF and DMF), these solvation effects enhance the stability of the benzamidine–formate complex by 2 orders of magnitude.

These conclusions are supported by density functional theory calculations of the H-bond parameters for all of the H-bonding sites in the free and bound species. The H-bond parameters were used to calculate the free energy contributions due to solvent–solute interactions in the primary solvation shell, and the calculations quantitatively predict the experimentally observed solvent effects to within 1 kJ mol–1. The approach provides a simple method that accurately predicts the stability of amidinium–carboxylate salt bridges in any solvent. The model also provides a quantitative explanation for the low stability of salt bridges in water if the equilibria between neutral and ionized species in the free state are taken into account. The solvent effects observed here represent an extreme example of what might be expected to be a more general phenomenon: if formation of a complex increases the polarity of peripheral functional groups exposed to the solvent, then more favorable interactions in the solvation shell will give rise to an unexpected stabilization of the complex in polar solvents.42,43

While H-bonding interactions are strong in nonpolar solvents, such as chloroform, and solvophobic interactions are strong in polar protic solvents, such as water, both types of interactions are weak in polar aprotic solvents, such as DMF. The exceptionally large association constants reported here in polar aprotic solvents highlight the unique properties of the amidinium–carboxylate salt bridge, making this interaction a very attractive option for the design of supramolecular systems that operate in a largely unexplored area of solvent space.

Acknowledgments

We thank the Herchel Smith Fund for financial support.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/jacs.4c11869.

  • Materials and methods, detailed synthetic procedures, characterization including 1H and 13C NMR spectra of all compounds, ITC titration data, DOSY NMR experiments, and solvent parameters (PDF)

Author Contributions

The manuscript was written through contributions of all authors.

The authors declare no competing financial interest.

Supplementary Material

ja4c11869_si_001.pdf (2.7MB, pdf)

References

  1. Bosshard H. R.; Marti D. N.; Jelesarov I. Protein Stabilization by Salt Bridges: Concepts, Experimental Approaches and Clarification of Some Misunderstandings. J. Mol. Recognit. 2004, 17 (1), 1–16. 10.1002/jmr.657. [DOI] [PubMed] [Google Scholar]
  2. Levintov L.; Vashisth H. Role of Salt-Bridging Interactions in Recognition of Viral RNA by Arginine-Rich Peptides. Biophys. J. 2021, 120 (22), 5060–5073. 10.1016/j.bpj.2021.10.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Kurczab R.; Ṡliwa P.; Rataj K.; Kafel R.; Bojarski A. J. Salt Bridge in Ligand–Protein Complexes—Systematic Theoretical and Statistical Investigations. J. Chem. Inf. Model. 2018, 58 (11), 2224–2238. 10.1021/acs.jcim.8b00266. [DOI] [PubMed] [Google Scholar]
  4. Hauel N. H.; Nar H.; Priepke H.; Ries U.; Stassen J.-M.; Wienen W. Structure-Based Design of Novel Potent Nonpeptide Thrombin Inhibitors. J. Med. Chem. 2002, 45 (9), 1757–1766. 10.1021/jm0109513. [DOI] [PubMed] [Google Scholar]
  5. Kusukawa T.; Toyama K.; Takeshita S.; Tanaka S. Fluorescent Detection of Amidinium–Carboxylate and Amidinium Formation Using Anthracene-Based Diamidine: An Application for the Analysis of Dicarboxylic Acid Binding. Tetrahedron 2012, 68 (48), 9973–9981. 10.1016/j.tet.2012.09.072. [DOI] [Google Scholar]
  6. Kusukawa T.; Aramoto H.; Umeda T.; Kojima Y. Carboxylic Acid Recognition of Diamidine Having a Fluorescent 1,8-Diphenylanthracene Unit and Its Detection of Amidinium-Carboxylate and Amidinium Formation. Tetrahedron 2019, 75 (9), 1293–1305. 10.1016/j.tet.2019.01.040. [DOI] [Google Scholar]
  7. Papoutsakis D.; Kirby J. P.; Jackson J. E.; Nocera D. G. From Molecules to the Crystalline Solid: Secondary Hydrogen-Bonding Interactions of Salt Bridges and Their Role in Magnetic Exchange. Chem. – Eur. J. 1999, 5 (5), 1474–1480. . [DOI] [Google Scholar]
  8. Pop L.; Hadade N. D.; van der Lee A.; Barboiu M.; Grosu I.; Legrand Y.-M. Occurence of Charge-Assisted Hydrogen Bonding in Bis-Amidine Complexes Generating Macrocycles. Cryst. Growth Des. 2016, 16 (6), 3271–3278. 10.1021/acs.cgd.6b00246. [DOI] [Google Scholar]
  9. Wulff G.; Liu J. Design of Biomimetic Catalysts by Molecular Imprinting in Synthetic Polymers: The Role of Transition State Stabilization. Acc. Chem. Res. 2012, 45 (2), 239–247. 10.1021/ar200146m. [DOI] [PubMed] [Google Scholar]
  10. Yang W.; Li W.; Yu B.; Liu C.; Wang H. Fluorescence Charge-Assisted Hydrogen-Bonded Organic Frameworks Assembled from Tetraphenylethene Amidinium Cation. Inorg. Chem. Commun. 2022, 139, 109396. 10.1016/j.inoche.2022.109396. [DOI] [Google Scholar]
  11. Bardts M.; Ritter H. Novel Copolymers Showing Interactions of Amidinium-Carboxylate Groups in Water. J. Polym. Res. 2011, 18 (1), 151–155. 10.1007/s10965-010-9401-9. [DOI] [Google Scholar]
  12. Furusho Y.; Endo T.; Higaki K.; Kaetsu K.; Higaki Y.; Kojio K.; Takahara A. Supramolecular Network Polymers Formed from Polyamidine and Carboxy-Terminated Telechelic Poly(n-Butyl Acrylate) via Amidinium-Carboxylate Salt Bridges. J. Polym. Sci., Part A: polym. Chem. 2016, 54 (14), 2148–2155. 10.1002/pola.28082. [DOI] [Google Scholar]
  13. Terfort A.; von Kiedrowski G. Self-Replication by Condensation of 3-Aminobenzamidines and 2-Formylphenoxyacetic Acids. Angew. Chem., Int. Ed. 1992, 31 (5), 654–656. 10.1002/anie.199206541. [DOI] [Google Scholar]
  14. Katagiri H.; Tanaka Y.; Furusho Y.; Yashima E. Multicomponent Cylindrical Assemblies Driven by Amidinium–Carboxylate Salt-Bridge Formation. Angew. Chem., Int. Ed. 2007, 46 (14), 2435–2439. 10.1002/anie.200603921. [DOI] [PubMed] [Google Scholar]
  15. Yamada H.; Wu Z.-Q.; Furusho Y.; Yashima E. Thermodynamic and Kinetic Stabilities of Complementary Double Helices Utilizing Amidinium–Carboxylate Salt Bridges. J. Am. Chem. Soc. 2012, 134 (22), 9506–9520. 10.1021/ja303701d. [DOI] [PubMed] [Google Scholar]
  16. Yashima E.; Maeda K.; Furusho Y. Single- and Double-Stranded Helical Polymers: Synthesis, Structures, and Functions. Acc. Chem. Res. 2008, 41 (9), 1166–1180. 10.1021/ar800091w. [DOI] [PubMed] [Google Scholar]
  17. Thomas C. M.; Foyle É. M.; Walker S. E.; White N. G. An Investigation of Five Component [3 + 2] Self-Assembled Cage Formation Using Amidinium···Carboxylate Hydrogen Bonds*,†. Aust. J. Chem. 2021, 74 (11), 787–794. 10.1071/CH21101. [DOI] [Google Scholar]
  18. Nakatani Y.; Furusho Y.; Yashima E. Amidinium Carboxylate Salt Bridges as a Recognition Motif for Mechanically Interlocked Molecules: Synthesis of an Optically Active [2]Catenane and Control of Its Structure. Angew. Chem., Int. Ed. 2010, 49 (32), 5463–5467. 10.1002/anie.201002382. [DOI] [PubMed] [Google Scholar]
  19. Núñez-Villanueva D.; Hunter C. A. Replication of a Synthetic Oligomer Using Chameleon Base-Pairs. Chem. Commun. 2022, 58 (78), 11005–11008. 10.1039/D2CC04580J. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Cabot R.; Hunter C. A. Non-Covalent Interactions between Iodo-Perfluorocarbons and Hydrogen Bond Acceptors. Chem. Commun. 2009, (15), 2005–2007. 10.1039/b822284c. [DOI] [PubMed] [Google Scholar]
  21. Yang L.; Adam C.; Cockroft S. L. Quantifying Solvophobic Effects in Nonpolar Cohesive Interactions. J. Am. Chem. Soc. 2015, 137 (32), 10084–10087. 10.1021/jacs.5b05736. [DOI] [PubMed] [Google Scholar]
  22. Yang L.; Adam C.; Nichol G. S.; Cockroft S. L. How Much Do van Der Waals Dispersion Forces Contribute to Molecular Recognition in Solution?. Nat. Chem. 2013, 5 (12), 1006–1010. 10.1038/nchem.1779. [DOI] [PubMed] [Google Scholar]
  23. Hunter C. A.; Sanders J. K. M. The Nature of.Pi.-.Pi. Interactions. J. Am. Chem. Soc. 1990, 112 (14), 5525–5534. 10.1021/ja00170a016. [DOI] [Google Scholar]
  24. Hunter C. A. Quantifying Intermolecular Interactions: Guidelines for the Molecular Recognition Toolbox. Angew. Chem., Int. Ed. 2004, 43 (40), 5310–5324. 10.1002/anie.200301739. [DOI] [PubMed] [Google Scholar]
  25. Young E. R.; Rosenthal J.; Nocera D. G. Spectral Observation of Conversion between Ionized vs. Non-Ionized Proton-Coupled Electron Transfer Interfaces. Chem. Commun. 2008, (20), 2322–2324. 10.1039/b717747j. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Kraft A.; Peters L.; Powell H. R. Hydrogen-Bonding between Benzoic Acid and an N,N′-Diethyl-Substituted Benzamidine. Tetrahedron 2002, 58 (18), 3499–3505. 10.1016/S0040-4020(02)00301-0. [DOI] [Google Scholar]
  27. Sánchez L.; Sierra M.; Martín N.; Myles A. J.; Dale T. J.; Rebek J. Jr.; Seitz W.; Guldi D. M. Exceptionally Strong Electronic Communication through Hydrogen Bonds in Porphyrin–C60 Pairs. Angew. Chem., Int. Ed. 2006, 45 (28), 4637–4641. 10.1002/anie.200601264. [DOI] [PubMed] [Google Scholar]
  28. López-Martín I.; Veiga-Herrero J.; Aparicio F.; González-Rodríguez D. A Modular and Convergent Synthetic Route to Supramolecular Cyclic Dimers Based on Amidinium-Carboxylate Interactions. Chem. – Eur. J. 2023, 29 (64), e202302279 10.1002/chem.202302279. [DOI] [PubMed] [Google Scholar]
  29. Sebo L.; Schweizer B.; Diederich F. Cleft-Type Diamidinium Receptors for Dicarboxylate Binding in Protic Solvents. Helv. Chim. Acta 2000, 83 (1), 80–92. . [DOI] [Google Scholar]
  30. Kirby J. P.; van Dantzig N. A.; Chang C. K.; Nocera D. G. Formation of Porphyrin Donor-Acceptor Complexes via an Amidinium-Carboxylate Salt Bridge. Tetrahedron Lett. 1995, 36 (20), 3477–3480. 10.1016/0040-4039(95)00569-X. [DOI] [Google Scholar]
  31. Wulff G.; Knorr K. Stoichiometric Noncovalent Interaction in Molecular Imprinting. Bioseparation 2001, 10 (6), 257–276. 10.1023/A:1021585518592. [DOI] [PubMed] [Google Scholar]
  32. Turnbull W. B.; Daranas A. H. On the Value of c: Can Low Affinity Systems Be Studied by Isothermal Titration Calorimetry?. J. Am. Chem. Soc. 2003, 125 (48), 14859–14866. 10.1021/ja036166s. [DOI] [PubMed] [Google Scholar]
  33. Cook J. L.; Hunter C. A.; Low C. M. R.; Perez-Velasco A.; Vinter J. G. Solvent Effects on Hydrogen Bonding. Angew. Chem., Int. Ed. 2007, 46 (20), 3706–3709. 10.1002/anie.200604966. [DOI] [PubMed] [Google Scholar]
  34. Driver M. D.; Williamson M. J.; Cook J. L.; Hunter C. A. Functional Group Interaction Profiles: A General Treatment of Solvent Effects on Non-Covalent Interactions. Chem. Sci. 2020, 11 (17), 4456–4466. 10.1039/D0SC01288B. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Storer M. C.; Hunter C. A. The Surface Site Interaction Point Approach to Non-Covalent Interactions. Chem. Soc. Rev. 2022, 51 (24), 10064–10082. 10.1039/D2CS00701K. [DOI] [PubMed] [Google Scholar]
  36. Storer M. C.; Zator K. J.; Reynolds D. P.; Hunter C. A. An Atomic Surface Site Interaction Point Description of Non-Covalent Interactions. Chem. Sci. 2023, 15 (1), 160–170. 10.1039/D3SC05690B. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Marcus Y.The Properties of Solvents; Solutions Chemistry; Wiley, 1998. [Google Scholar]
  38. Springs B.; Haake P. Equilibrium Constants for Association of Guanidinium and Ammonium Ions with Oxyanions: The Effect of Changing Basicity of the Oxyanion. Bioorganic Chem. 1977, 6 (2), 181–190. 10.1016/0045-2068(77)90019-0. [DOI] [Google Scholar]
  39. Kusukawa T.; Matsumoto K.; Nakamura H.; Iizuka W.; Toyama K.; Takeshita S. Four-Component Assembly in the Crystalline State Driven by Amidinium–Carboxylate Salt Bridge Formation from an Aqueous Solution. Org. Biomol. Chem. 2013, 11 (22), 3692–3698. 10.1039/c3ob40309b. [DOI] [PubMed] [Google Scholar]
  40. Albert A.; Goldacre R.; Phillips J. 455. The Strength of Heterocyclic Bases. J. Chem. Soc. 1948, 2240–2249. 10.1039/jr9480002240. [DOI] [Google Scholar]
  41. Braude E. A.; Nachod F. C.. Determination of Organic Structures by Physical Methods; Academic Press: New York, 1955, Vol. 1. [Google Scholar]
  42. Miriyala V. M.; Lo R.; Bouř P.; Wu T.; Nachtigallová D.; Hobza P. Unexpected Strengthening of the H-Bond Complexes in a Polar Solvent Due to a More Efficient Solvation of the Complex Compared to Isolated Monomers. J. Phys. Chem. A 2022, 126 (43), 7938–7943. 10.1021/acs.jpca.2c05992. [DOI] [PubMed] [Google Scholar]
  43. Manna D.; Lo R.; Vacek J.; Miriyala V. M.; Bouř P.; Wu T.; Osifová Z.; Nachtigallová D.; Dračinský M.; Hobza P. The Stability of Hydrogen-Bonded Ion-Pair Complex Unexpectedly Increases with Increasing Solvent Polarity. Angew. Chem., Int. Ed. 2024, 63 (20), e202403218 10.1002/anie.202403218. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ja4c11869_si_001.pdf (2.7MB, pdf)

Articles from Journal of the American Chemical Society are provided here courtesy of American Chemical Society

RESOURCES