CHEMPHYSCHEM ARTICLES DOI: 10.1002/cphc.201402296
Isothermal Titration Calorimetry Study of a Bistable Supramolecular System: Reversible Complexation of Cryptand[2.2.2] with Potassium Ions Maria G. del Rosso, Artur Ciesielski, Silvia Colella, Jack M. Harrowfield, and Paolo Samor*[a] Isothermal titration calorimetry (ITC) is used to investigate the thermodynamics of the complexation of potassium ions by 1,10-diaza-4,7,13,16,21,24-hexaoxabicyclo[8.8.8]hexacosane (cryptand[2.2.2]) in aqueous solution. By changing the pH of
the solution it was possible to trigger the reversible complexation/decomplexation of the cryptand in consecutive in situ experiments and to assess for the first time the use of ITC to monitor the thermodynamics of a bistable system.
1. Introduction One of the challenges in supramolecular chemistry,[1] and more generally in the chemistry of complex multicomponent systems, is to quantify the contribution of non-covalent interactions to the free-energy landscape of a given molecule upon interaction with a binding partner. Such information is of paramount importance in biology and pharmacology for the design of drugs that bind to a given target in a reaction and give a product of optimal stability.[2] The ability to correlate enthalpy and entropy changes to the binding affinity of investigated molecules interacting with their binding partners can offer a clear picture of structure–energy relationships, which is particularly important for the design of new drugs and functional supramolecular systems.[3] The advent of highly sensitive titration microcalorimeters in the late 1980s opened the door to the successful extraction of thermodynamic parameters from simple experiments.[4] Isothermal titration calorimetry (ITC) is now an established and invaluable method for determining with a high degree of precision the thermodynamic parameters, association constants and stoichiometry of molecular interactions at equilibrium in solution.[5] Unlike other methods commonly employed for the characterization of equilibria (e.g. NMR, Raman, electronic absorption and fluorescence spectroscopy), ITC makes it possible to determine simultaneously and with a high degree of precision the thermodynamic parameters and stoichiometry of a ligand– receptor interaction in a single experiment by measuring the increments of heat occurring during stepwise titration experiments.[6] ITC enables one to avoid the time-consuming process of van’t Hoff analysis based on the temperature dependence of equilibrium constants that is otherwise needed to acquire binding constants as well as entropy and enthalpy values. [a] M. G. del Rosso, Dr. A. Ciesielski, Dr. S. Colella, Prof. J. M. Harrowfield, Prof. P. Samor ISIS & icFRC, Universit de Strasbourg & CNRS 8 alle Gaspard Monge, 67000 Strasbourg (France) E-mail: [email protected] Supporting Information for this article is available on the WWW under http://dx.doi.org/10.1002/cphc.201402296.
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Hence, ITC has been widely employed to explore biological systems[7] involving protein–ligand, protein–nucleic acid and protein–protein interactions as well as model systems for protein–metal ion interactions based on transition metal complexes of oligopeptides.[8] It has also been used to unveil supramolecular interactions involved in host–guest binding of a wide variety of synthetic receptors including crown ethers,[9] expanded[10] and protonated polyaza-[11] cryptands, cyclodextrins,[12] calixarenes and resorcin[4]arenes/cavitands,[13] cucurbiturils[14] and calixpyrroles,[15] as well as processes leading to oligomerisation or polymerisation through H-bonding[16] and metal–ligand interaction in the field of coordination chemistry.[17] However, much of this work has been carried out in nonaqueous solvents. Surprisingly, only a few calorimetric measurements of any kind have focussed on well-defined, simple aqueous systems involving assembly of complexes of alkali metal ions, although the application of ITC to such a case was considered a useful means to further evaluate the merit of the technique. Cryptands, in their simplest form,[18] are synthetic macrobicyclic multidentate ligands that can host, amongst others, alkali and alkaline earth metal cations, usually in the form of 1:1 inclusion complexes. This capacity has recently been used as a “remote control” to trigger the in situ reversible conformational interconversion of oligoheterocyclic strands consisting of alternating pyridine and pyrimidine subunits,[19] as well as the operation of a dynamic switch between different self-assembled forms of guanine.[20] Thus, a detailed understanding of the thermodynamics of the involved reversible complexation reactions is fundamental to the analysis of such systems. For these reasons, in the present work, we focused on the thermodynamics of the complexation of potassium ions by 1,10-diaza-4,7,13,16,21,24-hexaoxabicyclo[8.8.8]hexacosane (cryptand[2.2.2]) in aqueous media. To this end, we performed different in situ consecutive experiments, starting with the simple addition of an aqueous solution of potassium chloride to a solution of cryptand[2.2.2] , followed by the consecutive addition of an acid and a base to trigger the reversible comChemPhysChem 0000, 00, 1 – 7
&1&
These are not the final page numbers! ÞÞ
CHEMPHYSCHEM ARTICLES plexation/decomplexation process with the ultimate goal of monitoring quantitatively for the first time by ITC a bistable supramolecular system in action. Due to differences in the free energy of solvation of the cations in various solvents, the stability of the complexes of cryptand[2.2.2] with alkali and alkaline earth metal cations increases with decreasing solvent polarity.[18, 21] Nonetheless, because of its importance as a reference system, we decided to perform all experiments in water, which engendered some particular problems associated with the use of a protic solvent. A solution of cryptand[2.2.2] in water is alkaline, and the resulting hydrolysis reaction results in a dilute aqueous solution of cryptand[2.2.2] containing a significant fraction of the ligand in its monoprotonated form. Protonation is one means of inhibiting metal-cation binding by the cryptand. Thus, in the present work we determined the optimum conditions for determination of the complexation constant for a ligand that may undergo significant hydrolysis in water.
2. Results and Discussion Self-assembly of the complex K[2.2.2] + was studied initially as a direct reaction between the two components, that is, K + and cryptand[2.2.2], and also as a competition between potassium ions and protons for the cryptand. This was done by firstly measuring the heat changes in the direct reaction of K + with cryptand[2.2.2], treating the resulting solution with a strong acid (HCl) to destroy the complex, and finally treating this solution with a strong base (1,1,3,3-tetramethylguanidine, TMG, pKa = 13.9[22]) to restore the complex along the path portrayed in Scheme 1. The choice of HCl and TMG was aimed at avoiding further complications in the experiment, since neither chloride nor TMG or its conjugated acid can be included in the protonated cryptand.
Scheme 1. Stepwise in situ reversible assembly between cryptand[2.2.2] and cryptate K[2.2.2] + . The blue sphere represents K + .
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemphyschem.org All experiments were performed with a Microcal VP-ITC instrument and the data were analysed with the program HypDH.[23] Since cryptand[2.2.2] is a base, its reaction with water must be taken into account in addition to those with H3O + and K + . One of the long-known complications in ITC analysis is to know the exact nature of the reacting species in solution,[24] and thus to assign the observed heat changes to particular events. The complication of strong ion pairing in weakly polar solvents, for example, has even proven to render some ITC data uninterpretable[25] and this is not the only complication arising in non-aqueous solvents.[15b] In preliminary experiments (see Section 2.1 of the Supporting Information) 0.5 mm solutions of cryptand[2.2.2] were titrated with HCl in order to characterise the two protonation steps. The result of this experiment led to some confusion related to the appearance of ITC end points after addition of about 0.7 and 2.0 equivalents of acid (see Figure S1 in the Supporting Information). If the literature values of the acidity constants for cryptand[2.2.2][26] are used for a speciation calculation at this concentration, about 60 % of the added cryptand[2.2.2] would have been converted to its monoprotonated derivative along with an equivalent amount of hydroxide ions, so that the ITC results can be explained by the initial dominance of heat evolution due to the reaction H + + OH , which would of course reduce the accuracy of the determination of the enthalpy change for the first step of cryptand[2.2.2] protonation. Since experiments on the protonation of cryptand[2.2.2] at a concentration of 1 mm led to ITC end points close to one and two equivalents of acid (Figure 1 b), all further experiments
Figure 1. a) Representative ITC raw data for interaction of cryptand[2.2.2] and HCl, obtained by using a 1.0 mm solution of cryptand[2.2.2] and a 30.2 mm solution of HCl. b) Left y axis: experimental (&) and calculated (~) heats for each injection; right y axis: speciation curves relative to cryptand[2.2.2] (c free cryptand, c monoprotonated cryptand H[2.2.2] + , c diprotonated cryptand H2[2.2.2]2 + ).
ChemPhysChem 0000, 00, 1 – 7
&2&
These are not the final page numbers! ÞÞ
CHEMPHYSCHEM ARTICLES were performed at this concentration. However, this did not completely exclude the effects of ligand hydrolysis (at 103 m, ca. 20 % conversion to H[2.2.2] + is still expected and must be taken into account). Since acidity constants determined by ITC were not available, their measurement was made part of the present work to give a consistent data set for the treatment of K + complexation. All experiments were performed at 298 K in degassed solutions. The initial titrand was a 1.1 mm solution of cryptand[2.2.2], which was loaded into the sample cell, and the titrants for the consecutive experiments were in order a 22 mm solution of potassium chloride, a 11.5 mm solution of HCl and a 20 mm solution of TMG. The concentration of the titrant must be at least ten times higher than that of the titrand, because the volume of the calorimetric pipette that contains the titrant is smaller than that of the sample cell. Heats of dilution for all reagents were determined in separate experiments and used to correct the raw ITC data. In all cases, tip-diffusion errors required the first and sometimes the second titration points to be discarded. Full details of the procedure are given in the Supporting Information. Moreover, for the second and third titration the concentration of the components present in the cell must be calculated by taking into account that VP-ITC operates in overfilled mode and the concentrations of the components in the cell change after every injection from the pipette into the cell (see Section 3 of the Supporting Information). Fitting of ITC data with the program HypDH requires the development of a chemical model for the studied system. As mentioned above, the acid and base were chosen so as to avoid the complication of inclusion reactions other than that of the potassium ion. Whereas various crystal structure determinations on acid salts of cryptand[2.2.2] have shown that in the solid state the proton attached to a cryptand N atom may be directed exo or endo with respect to the cavity, in solution it is known that inversion at these centres (in the unprotonated species as well) is fast,[18, 26b] so that the assumption was made that cryptand[2.2.2], H[2.2.2] + and H2[2.2.2]2 + could each be treated as a single species for a “slow” technique such as ITC. Since monoprotonation is considered to involve only one of the possible coordination sites of cryptand[2.2.2], it is conceivable that the species KH[2.2.2]2 + could be formed, but 1 H NMR spectroscopic monitoring of the titration of a solution of K[2.2.2] + in D2O with DCl (see Section 4 of the Supporting Information) showed that all changes could be accounted for in terms of the spectra of K[2.2.2] + and the two protonated forms of cryptand[2.2.2]. Hence, Scheme 1 is considered to fully describe the chemistry studied in the present work. As the auto-ionisation of water is a rather well studied reaction, literature values for pKw (13.99)[27] and DHw (13.4 kcal mol1)[28] at zero ionic strength were incorporated as constants in the models for all systems, but small errors may have resulted from the fact that the ionic strength, although low in all cases, did vary (from ca. 0.0002 to 0.003 m) as a result of the way the titrations were conducted. Since cryptand protonation/deprotonation reactions were involved in all systems studied, the simple titration of cryptand 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemphyschem.org [2.2.2] with HCl was used to provide values of the thermodynamic parameters for the two protonation steps, under the assumption that any protonation of ether O sites could be ignored under the conditions used. The parameters describing the complexation of the potassium ion (logarithm of the stability constant lg KK and the thermodynamic parameters) were in fact calculated by maintaining constant all the parameters concerning protonation/deprotonation reactions (pKa1, pKa2, enthalpies of protonation). The elaboration of the raw ITC data presented in Figure 1 a for a protonation experiment in which a solution of cryptand[2.2.2] with concentration close to 103 m was titrated is shown in Figure 1 b. It shows, as noted above, two apparent end points at molar ratios close to 1.0 and 2.0, which indicate a negligible influence of hydrolysis in the initial solutions of the cryptand. The values obtained for the protonation-step parameters (Table 1) compare well with those obtained by other methods,[26] in which supporting electrolytes were employed to give an ionic strength of 0.1 m. After the determination of pKa1 and pKa2 and their related enthalpy changes, the reversible complexation/decomplexation of K + by cryptand[2.2.2] was studied by in situ consecutive experiments in which pH-modulated switching was monitored by ITC. First, the simple complexation of cryptand[2.2.2] with potassium ion in water was executed by titrating a 1.1 mm solution of cryptand[2.2.2] with a 20 mm solution of potassium chloride. The raw ITC data and the observed and calculated heat as function of the molar KCl/cryptand ratio are shown in Figure 2. They reveal again that any influence of cryptand hydrolysis is hidden within the effects arising from K + complexation. Here, the cryptand protonation parameters determined from the preceding experiment were used as fixed constants (along with Kw and DHw) in the fitting, in order to define the equilibrium constant and enthalpy changes for the potassium complex. Although the values obtained were again in good agreement with literature data (ionic strength 0.05 or 0.1 m),[21, 26a, 29] reversal of the initial ligand hydrolysis during the titration had a significant influence on the heat changes, and the value found for KK indicates that, at a concentration near 103 m, any dissociation of the complex would be slight. For these reasons, a better way to perform the titration could be to add HCl to a solution of the complex. This experiment corresponds to the second in situ consecutive titration, carried out by titrating a 11.5 mm solution of HCl into the sample cell that now contains all of the cryptand in its cryptate form. Protonation of the cryptand causes displacement of the potassium ion and thus affects in particular the first part of the titration, as it is evident by comparing Figures 1 and 3. Indeed, the results obtained from the deceptively simple raw ITC data curve for this reverse titration with all other parameters kept as fixed constants (Figure 3) provides values that are still within the error bar of those reported in the literature (see Table 1).[21] To reassemble the K + complex of the cryptand, in situ deprotonation was performed by titrating the solution of H2[2.2.2]2 + in the presence of potassium ions with a 20 mm solution of TMG. Figure 4 shows the raw ITC data and the observed and calculated step reaction heats. In this case the enthalpogram clearly shows two inflection points. The first inflecChemPhysChem 0000, 00, 1 – 7
&3&
These are not the final page numbers! ÞÞ
10.78 0.01 11.06 0.01 10.43 0.01 5.5 0.1 5.8 0.1 4.9 0.1
[a] The values calculated as output parameters are given in italics; the other values are those employed in each fitting as fixed parameters. The Gibbs energy (DG8 and entropy DS8 were calculated by using the DH DG classic thermodynamic equations: DG ¼ RT ln K and DS ¼ T . The errors were generated by the fitting process of the program HypDH. Refs. [21, 29a]: The stability constants were determined by pHmetric and potentiometric methods at an ionic strength of I = 0.05 m and T = 25 8C, and the enthalpies by calorimetric measurements. Refs. [26a, 30]: The stability constants were determined by a pH-metric method at I = 0.1 m and T = 25 8C, and the enthalpies by calorimetric measurements.
11.00 0.48 10.56 0.49 12.55 0.48
11[26a] 18.3[26a] 14.94 3.11
9.60 0.05[21] 10.00 0.01[30] 9.71[26a] 9.1 0.4 9.1 9.1 9.1 literature[21, 26a, 30]
[2.2.2] + HCl [2.2.2] + KCl K[2.2.2] + + HCl K + + (H2[2.2.2])2 + + TMG
H + + [2.2.2]!(H[2.2.2]) + pKa2 DH8 [kcal mol1]
10.8[26a] 11.02 0.01 11.02 11.02 11.02
DG8 [kcal mol1]
13.25[26a] 12.41 0.54
8.2[26a] 4.67 1.78
DS8 [cal mol1 K1]
7.28 0.05[21] 7.53 0.04[30] 7.31[26a] 6.4 0.7 6.4 6.4 6.4
4.5[26a] 4.27 0.1 4.27 4.27 4.27
9.95[26a] 8.74 0.95
5.58[26a]
7.61[26a]
7.50 0.14 7.92 0.14 6.68 0.14
11.5[26a]
14.1 2[29a] 11.4 0.2[29a]
DS8 [cal mol1 K1] H + + (H[2.2.2]) + !(H2[2.2.2])2 + pKa1 DG8 DH8 [kcal mol1] [kcal mol1]
5.4 0.4[21]
www.chemphyschem.org
Titration
Table 1. Selected thermodynamic parameters of the in situ reversible H + -dependent assembly between cryptand[2.2.2] and K + .[a]
[2.2.2] + KCl!K[2.2.2] + log KK DH8 [kcal mol1]
DG8 [kcal mol1]
DS8 [cal mol1 K1]
CHEMPHYSCHEM ARTICLES
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 2. a) Representative ITC raw data for the interaction of KCl and cryptand[2.2.2] showing 34 consecutive injections of 3 mL. b) Left y axis: experimental (&) and calculated (~) heats for each injection; right y axis: speciation curves relative to cryptand[2.2.2] (c cryptate K[2.2.2] + , c free cryptand, c monoprotonated cryptand H[2.2.2] + , c diprotonated cryptand H2[2.2.2]2 + ).
tion point, at a molar base/cryptand ratio of about 0.3, corresponds to the reaction of the excess acid from the previous titration (Figure 3) on addition of the base. In fact, if we perform the second experiment (Figure 3) and stop the titration exactly at a molar ratio equivalent to two (to avoid the excess of acid), the first part of the curve due to neutralisation of the acid with the base is not observed (see Section 2.4 in the Supporting Information). The second inflection point at a molar ratio of 2.27 is characteristic of the completion of the process of deprotonation of the second nitrogen atom and inclusion of the potassium ion (see speciation curves in Figure 4 b). Also in this case the curve is not a simple bisigmoidal curve, because parallel inclusion of the potassium ion is occurring. Although hydrolysis of a solution of cryptand[2.2.2] dihydrochloride may be less extensive than that of the free base, titration with alkali rather than acid can involve greater problems of reagent purity. The use of a base such as TMG, chosen both to avoid cryptand inclusion reactions and because it is more easily obtained in a pure form than guanidine itself, illustrates another problem, in that additional thermodynamic parameters are required to describe the complete system. For TMG, an approximate pKa value is available,[22] but its heat of neutralisation is not known, although it is expected to be similar to that of guanidine and thus close to that of OH .[31] Indeed, the value of the association constant of the potassium ion (lg KK), determined by keeping all the parameters constant except the enthalpy of protonation of TMG and the enthalpy of formation of the cryptate complex, is less accurate than that found in previous titrations. Moreover the value obtained for ChemPhysChem 0000, 00, 1 – 7
&4&
These are not the final page numbers! ÞÞ
CHEMPHYSCHEM ARTICLES
www.chemphyschem.org nificance in the absence of a precise pKa value for aqueous media.
3. Conclusions
Figure 3. a) Representative ITC raw data for the interaction of cryptate K[2.2.2] + and HCl. b) Left y axis: experimental (&) and calculated (~) heats for each injection; right y axis: speciation curves relative to cryptand[2.2.2] (c cryptate K[2.2.2] + , c free cryptand, c monoprotonated cryptand H[2.2.2] + , c diprotonated cryptand H2[2.2.2]2 + .
Figure 4. a) Representative ITC raw data for the deprotonation of the cryptand[2.2.2] in the presence of the potassium cation. b) Left y axis: experimental (&) and calculated (~) heats for each injection; right y axis: speciation curves relative to cryptand[2.2.2] (c free cryptand, c monoprotonated cryptand H[2.2.2] + , c diprotonated cryptand H2[2.2.2]2 + , c cryptate K[2.2.2] + ).
the heat of neutralisation of TMG is very high (13.91 0.01 kcal mol1). We can conclude that the last result is not of great sig 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
We have shown for the first time that ITC can be used to monitor the reversible process of complexation/decomplexation in a genuinely bistable supramolecular system in which the pH change is used as a remote control. As a model system. we chose the complexation of potassium ions by cryptand[2.2.2] . Its decomplexation followed by recomplexation was triggered by subsequent addition of a strong acid and a strong base, respectively. In all subsequent titrations, it was possible to determine the association constant and the thermodynamic parameters relative to the process of complexation of potassium ion by cryptand[2.2.2]. We emphasize that the protic nature of water as a medium engenders complications in the analysis of equilibria involving the binding of acidic and basic substrates, and hence in the present case a cryptand concentration of at least 1 mm is necessary to obtain optimum precision of the derived thermodynamic parameters. The results obtained by ITC for the protonation and complexation of K + by cryptand[2.2.2] are in agreement with those reported in the literature, based on studies carried out with other techniques. Even though it is known that the precision reached by other methods in the calculation of the stability constant (i.e. potentiometric methods) can be higher, the major advantage of ITC is the possibility to determine all the thermodynamic parameters in a single set of measurements. The separation of enthalpic and entropic contributions from the reaction free energy, which is readily obtained by ITC measurements, is instrumental for assessing the factors that can influence host–guest recognition. Unfortunately, if only free energies (equilibrium constants) are known, there is a tendency to emphasize influences such as bonding and solvation considered as contributions to an enthalpy change, whereas in many instances the entropy change dominates the enthalpy change as a contribution to the overall change in Gibbs free energy.[32] Recognition of this can provide a rather different perspective on the nature of the equilibrium involved. A comparison of the binding of one proton or one potassium ion by cryptand[2.2.2] is interesting in that, even though the guestbinding interactions are rather different, the enthalpy changes in the reactions are similar, although the entropy change is positive for protonation and negative for K + binding. The entropy decrease on K + binding has been ascribed to enhanced solvent structuring about the cation masked by the organic sheath of the cryptand ligand,[18] an argument which implies that K + binding might be simply enhanced by functionalising the exterior of cryptand[2.2.2] with hydrophilic units.
Acknowledgements We thank Thomas M. Hermans for enlightening discussions. This work was financially supported by EC through the Marie-Curie ITN GENIUS (PITN-GA-2010-264694), the ERC project SUPRAFUNCChemPhysChem 0000, 00, 1 – 7
&5&
These are not the final page numbers! ÞÞ
CHEMPHYSCHEM ARTICLES
www.chemphyschem.org
TION (GA-257305), the Agence Nationale de la Recherche through the LabEx project Chemistry of Complex Systems (ANR-10-LABX0026_CSC) and the International Center for Frontier Research in Chemistry (icFRC). Keywords: calorimetry · cryptands · isothermal titration calorimetry · potassium · supramolecular chemistry [1] a) J. M. Lehn, Science 1993, 260, 1762 – 1763; b) T. F. A. de Greef, E. W. Meijer, Nature 2008, 453, 171 – 173; c) G. R. Desiraju, Acc. Chem. Res. 2002, 35, 565 – 573. [2] E. Freire, A. J. Ruben, Y. Kiso, Chem. Biol. Drug Des. 2006, 67, 2 – 4. [3] a) K. E. Riley, P. Hobza, Acc. Chem. Res. 2013, 46, 927 – 936; b) J. L. Asensio, A. Ard, F. J. CaÇada, J. Jimnez-Barbero, Acc. Chem. Res. 2013, 46, 946 – 954. [4] a) T. Wiseman, S. Williston, J. F. Brandts, L. N. Lin, Anal. Biochem. 1989, 179, 131 – 137; b) J. F. Brandts, L. N. Lin, T. Wiseman, S. Williston, C. P. Yang, Am. Lab. 1990, 22, 30 – 41; c) I. Wadso, Chem. Soc. Rev. 1997, 26, 79 – 86. [5] a) R. J. Falconer, B. M. Collins, J. Mol. Recognit. 2011, 24, 1 – 16; b) P. Arranz-Mascaros, C. Bazzicalupi, A. Bianchi, C. Giorgi, M. L. GodinoSalido, M. D. Gutierrez-Valero, R. Lopez-Garzon, M. Savastano, J. Am. Chem. Soc. 2013, 135, 102 – 105. [6] “Isothermal Titration Calorimetry in Supramolecular Chemistry,” F. P. Schmidtchen in Analytical Methods in Supramolecular Chemistry (Ed.: C. A. Schalley), Wiley-VCH, Weinheim, 2007, pp. 55 – 78. [7] a) V. Ball, C. Maechling, Int. J. Mol. Sci. 2009, 10, 3283 – 3315; b) J. Ladbury, I. Haq, J. Mol. Recognit. 2000, 13, 188 – 197. [8] a) N. Grossoehme, A. Spuches, D. Wilcox, J. Biol. Inorg. Chem. 2010, 15, 1183 – 1191; b) L. Q. Hatcher, L. Hong, W. D. Bush, T. Carducci, J. D. Simon, J. Phys. Chem. B 2008, 112, 8160 – 8164; c) L. A. Hong, T. M. Carducci, W. D. Bush, C. G. Dudzik, G. L. Millhauser, J. D. Simon, J. Phys. Chem. B 2010, 114, 11261 – 11271. [9] a) W. B. Turnbull, A. H. Daranas, J. Am. Chem. Soc. 2003, 125, 14859 – 14866; b) N. Rawat, R. B. Gujar, M. S. Murali, B. S. Tomar, V. K. Manchanda, Thermochim. Acta 2009, 488, 21 – 26; c) R. M. Izatt, J. S. Bradshaw, S. A. Nielsen, J. D. Lamb, J. J. Christensen, Chem. Rev. 1985, 85, 271 – 339. [10] a) A. M. P. Pederson, E. M. Ward, D. V. Schoonover, C. Slebodnick, H. W. Gibson, J. Org. Chem. 2008, 73, 9094 – 9101; b) A. M. P. Pederson, R. C. Vetor, M. A. Rouser, F. Huang, C. Slebodnick, D. V. Schoonover, H. W. Gibson, J. Org. Chem. 2008, 73, 5570 – 5573. [11] a) V. Amendola, G. Alberti, G. Bergamaschi, R. Biesuz, M. Boiocchi, S. Ferrito, F. P. Schmidtchen, Eur. J. Inorg. Chem. 2012, 3410 – 3417; b) P. Mateus, N. Bernier, R. Delgado, Coord. Chem. Rev. 2010, 254, 1726 – 1747. [12] a) Y. Inoue, M. V. Rekharsky, Chem. Rev. 1998, 98, 1875 – 1917; b) K. Bouchemal, S. Mazzaferro, Drug Discovery Today 2012, 17, 623 – 629. [13] a) A. F. Danil de Namor, R. M. Cleverley, M. L. Zapata-Ormachea, Chem. Rev. 1998, 98, 2495 – 2526; b) J. Hornung, D. Fankhauser, L. D. Shirtcliff,
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
[14]
[15]
[16] [17] [18] [19] [20]
[21] [22] [23] [24] [25] [26] [27] [28] [29]
[30] [31] [32]
A. Praetorius, W. B. Schweizer, F. Diederich, Chem. Eur. J. 2011, 17, 12362 – 12371; c) F. Corbellini, L. Di Costanzo, M. Crego-Calama, S. Geremia, D. N. Reinhoudt, J. Am. Chem. Soc. 2003, 125, 9946 – 9947; d) C. Sgarlata, C. Bonaccorso, F. G. Gulino, V. Zito, G. Arena, D. Sciotto, New J. Chem. 2009, 33, 991 – 997; e) M. Dionisio, G. Oliviero, D. Menozzi, S. Federici, R. M. Yebeutchou, F. P. Schmidtchen, E. Dalcanale, P. Bergese, J. Am. Chem. Soc. 2012, 134, 2392 – 2398. a) F. Biedermann, V. D. Uzunova, O. A. Scherman, W. M. Nau, A. De Simone, J. Am. Chem. Soc. 2012, 134, 15318 – 15323; b) F. Biedermann, M. Vendruscolo, O. A. Scherman, A. De Simone, W. M. Nau, J. Am. Chem. Soc. 2013, 135, 14879 – 14888. a) J. L. Sessler, D. E. Gross, W. S. Cho, V. M. Lynch, F. P. Schmidtchen, G. W. Bates, M. E. Light, P. A. Gale, J. Am. Chem. Soc. 2006, 128, 12281 – 12288; b) M. Valik, V. Kral, E. Herdtweck, F. P. Schmidtchen, New J. Chem. 2007, 31, 703 – 710. T. F. A. De Greef, M. M. J. Smulders, M. Wolffs, A. P. H. J. Schenning, R. P. Sijbesma, E. W. Meijer, Chem. Rev. 2009, 109, 5687 – 5754. M. Ostermeier, M. A. Berlin, R. M. Meudtner, S. Demeshko, F. Meyer, C. Limberg, S. Hecht, Chem. Eur. J. 2010, 16, 10202 – 10213. J.-M. Lehn, Acc. Chem. Res. 1978, 11, 49 – 57. M. Barboiu, J.-M. Lehn, Proc. Natl. Acad. Sci. USA 2002, 99, 5201 – 5206. a) G. P. Spada, S. Pieraccini, S. Masiero, O. Pandoli, P. Samori, Org. Lett. 2006, 8, 3125 – 3128; b) A. Ciesielski, S. Lena, S. Masiero, G. P. Spada, P. Samori, Angew. Chem. 2010, 122, 2007 – 2010; Angew. Chem. Int. Ed. 2010, 49, 1963 – 1966. J.-M. Lehn, J. P. Sauvage, J. Am. Chem. Soc. 1975, 97, 6700 – 6707. R. W. Taft, H. B. Yang, J. Am. Chem. Soc. 1971, 93, 1310 – 1311. P. Gans, A. Sabatini, A. Vacca, J. Solution Chem. 2008, 37, 467 – 476. Y. Zhang, S. Akilesh, D. E. Wilcox, Inorg. Chem. 2000, 39, 3057 – 3064. D. E. Gross, F. P. Schmidtchen, W. Antonius, P. A. Gale, V. M. Lynch, J. L. Sessler, Chem. Eur. J. 2008, 14, 7822 – 7827. a) G. Anderegg, Helv. Chim. Acta 1975, 58, 1218 – 1225; b) R. Pizer, J. Am. Chem. Soc. 1978, 100, 4239 – 4241. A. V. Bandura, S. N. Lvov, J. Phys. Chem. Ref. Data 2006, 35, 15 – 30. I. Grenthe, H. Ots, M. Ginstrupp, Acta Chem. Scand. 1970, 24, 1067 – 1080. a) E. Kauffmann, J.-M. Lehn, J. P. Sauvage, Helv. Chim. Acta 1976, 59, 1099 – 1111; b) J. Gutknecht, H. Schneider, J. Stroka, Inorg. Chem. 1978, 17, 3326 – 3329. F. Arnaud-Neu, B. Spiess, M.-J. Schwing-Weill, Helv. Chim. Acta 1977, 60, 2633 – 2643. L. Fabbrizzi, M. Micheloni, P. Paoletti, G. Schwarzenbach, J. Am. Chem. Soc. 1977, 99, 5574 – 5576. a) F. P. Schmidtchen, Coord. Chem. Rev. 2006, 250, 2918 – 2928; b) F. P. Schmidtchen, Chem. Soc. Rev. 2010, 39, 3916 – 3935; c) A. Ursu, F. P. Schmidtchen, Angew. Chem. 2012, 124, 246 – 250; Angew. Chem. Int. Ed. 2012, 51, 242 – 246.
Received: May 3, 2014 Published online on && &&, 2014
ChemPhysChem 0000, 00, 1 – 7
&6&
These are not the final page numbers! ÞÞ
ARTICLES pH-modulated complexation and release of metal ions based on reversible assembly of the K + complex of cryptand[2.2.2] in aqueous solution is explored by means of isothermal titration calorimetry. The thermodynamic parameters associated with the interconversion between cryptand and cryptate are elucidated by sequential addition of KCl, HCl and 1,1,3,3-tetramethylguanidine.
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
M. G. del Rosso, A. Ciesielski, S. Colella, J. M. Harrowfield, P. Samor* && – && Isothermal Titration Calorimetry Study of a Bistable Supramolecular System: Reversible Complexation of Cryptand[2.2.2] with Potassium Ions
ChemPhysChem 0000, 00, 1 – 7
&7&
These are not the final page numbers! ÞÞ
Isothermal Titration Calorimetry Study of a Bistable Supramolecular System: Reversible Complexation of Cryptand[2.2.2] with Potassium Ions Maria G. del Rosso, Artur Ciesielski, Silvia Colella, Jack M. Harrowfield, and Paolo Samor*[a] Isothermal titration calorimetry (ITC) is used to investigate the thermodynamics of the complexation of potassium ions by 1,10-diaza-4,7,13,16,21,24-hexaoxabicyclo[8.8.8]hexacosane (cryptand[2.2.2]) in aqueous solution. By changing the pH of
the solution it was possible to trigger the reversible complexation/decomplexation of the cryptand in consecutive in situ experiments and to assess for the first time the use of ITC to monitor the thermodynamics of a bistable system.
1. Introduction One of the challenges in supramolecular chemistry,[1] and more generally in the chemistry of complex multicomponent systems, is to quantify the contribution of non-covalent interactions to the free-energy landscape of a given molecule upon interaction with a binding partner. Such information is of paramount importance in biology and pharmacology for the design of drugs that bind to a given target in a reaction and give a product of optimal stability.[2] The ability to correlate enthalpy and entropy changes to the binding affinity of investigated molecules interacting with their binding partners can offer a clear picture of structure–energy relationships, which is particularly important for the design of new drugs and functional supramolecular systems.[3] The advent of highly sensitive titration microcalorimeters in the late 1980s opened the door to the successful extraction of thermodynamic parameters from simple experiments.[4] Isothermal titration calorimetry (ITC) is now an established and invaluable method for determining with a high degree of precision the thermodynamic parameters, association constants and stoichiometry of molecular interactions at equilibrium in solution.[5] Unlike other methods commonly employed for the characterization of equilibria (e.g. NMR, Raman, electronic absorption and fluorescence spectroscopy), ITC makes it possible to determine simultaneously and with a high degree of precision the thermodynamic parameters and stoichiometry of a ligand– receptor interaction in a single experiment by measuring the increments of heat occurring during stepwise titration experiments.[6] ITC enables one to avoid the time-consuming process of van’t Hoff analysis based on the temperature dependence of equilibrium constants that is otherwise needed to acquire binding constants as well as entropy and enthalpy values. [a] M. G. del Rosso, Dr. A. Ciesielski, Dr. S. Colella, Prof. J. M. Harrowfield, Prof. P. Samor ISIS & icFRC, Universit de Strasbourg & CNRS 8 alle Gaspard Monge, 67000 Strasbourg (France) E-mail: [email protected] Supporting Information for this article is available on the WWW under http://dx.doi.org/10.1002/cphc.201402296.
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Hence, ITC has been widely employed to explore biological systems[7] involving protein–ligand, protein–nucleic acid and protein–protein interactions as well as model systems for protein–metal ion interactions based on transition metal complexes of oligopeptides.[8] It has also been used to unveil supramolecular interactions involved in host–guest binding of a wide variety of synthetic receptors including crown ethers,[9] expanded[10] and protonated polyaza-[11] cryptands, cyclodextrins,[12] calixarenes and resorcin[4]arenes/cavitands,[13] cucurbiturils[14] and calixpyrroles,[15] as well as processes leading to oligomerisation or polymerisation through H-bonding[16] and metal–ligand interaction in the field of coordination chemistry.[17] However, much of this work has been carried out in nonaqueous solvents. Surprisingly, only a few calorimetric measurements of any kind have focussed on well-defined, simple aqueous systems involving assembly of complexes of alkali metal ions, although the application of ITC to such a case was considered a useful means to further evaluate the merit of the technique. Cryptands, in their simplest form,[18] are synthetic macrobicyclic multidentate ligands that can host, amongst others, alkali and alkaline earth metal cations, usually in the form of 1:1 inclusion complexes. This capacity has recently been used as a “remote control” to trigger the in situ reversible conformational interconversion of oligoheterocyclic strands consisting of alternating pyridine and pyrimidine subunits,[19] as well as the operation of a dynamic switch between different self-assembled forms of guanine.[20] Thus, a detailed understanding of the thermodynamics of the involved reversible complexation reactions is fundamental to the analysis of such systems. For these reasons, in the present work, we focused on the thermodynamics of the complexation of potassium ions by 1,10-diaza-4,7,13,16,21,24-hexaoxabicyclo[8.8.8]hexacosane (cryptand[2.2.2]) in aqueous media. To this end, we performed different in situ consecutive experiments, starting with the simple addition of an aqueous solution of potassium chloride to a solution of cryptand[2.2.2] , followed by the consecutive addition of an acid and a base to trigger the reversible comChemPhysChem 0000, 00, 1 – 7
&1&
These are not the final page numbers! ÞÞ
CHEMPHYSCHEM ARTICLES plexation/decomplexation process with the ultimate goal of monitoring quantitatively for the first time by ITC a bistable supramolecular system in action. Due to differences in the free energy of solvation of the cations in various solvents, the stability of the complexes of cryptand[2.2.2] with alkali and alkaline earth metal cations increases with decreasing solvent polarity.[18, 21] Nonetheless, because of its importance as a reference system, we decided to perform all experiments in water, which engendered some particular problems associated with the use of a protic solvent. A solution of cryptand[2.2.2] in water is alkaline, and the resulting hydrolysis reaction results in a dilute aqueous solution of cryptand[2.2.2] containing a significant fraction of the ligand in its monoprotonated form. Protonation is one means of inhibiting metal-cation binding by the cryptand. Thus, in the present work we determined the optimum conditions for determination of the complexation constant for a ligand that may undergo significant hydrolysis in water.
2. Results and Discussion Self-assembly of the complex K[2.2.2] + was studied initially as a direct reaction between the two components, that is, K + and cryptand[2.2.2], and also as a competition between potassium ions and protons for the cryptand. This was done by firstly measuring the heat changes in the direct reaction of K + with cryptand[2.2.2], treating the resulting solution with a strong acid (HCl) to destroy the complex, and finally treating this solution with a strong base (1,1,3,3-tetramethylguanidine, TMG, pKa = 13.9[22]) to restore the complex along the path portrayed in Scheme 1. The choice of HCl and TMG was aimed at avoiding further complications in the experiment, since neither chloride nor TMG or its conjugated acid can be included in the protonated cryptand.
Scheme 1. Stepwise in situ reversible assembly between cryptand[2.2.2] and cryptate K[2.2.2] + . The blue sphere represents K + .
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemphyschem.org All experiments were performed with a Microcal VP-ITC instrument and the data were analysed with the program HypDH.[23] Since cryptand[2.2.2] is a base, its reaction with water must be taken into account in addition to those with H3O + and K + . One of the long-known complications in ITC analysis is to know the exact nature of the reacting species in solution,[24] and thus to assign the observed heat changes to particular events. The complication of strong ion pairing in weakly polar solvents, for example, has even proven to render some ITC data uninterpretable[25] and this is not the only complication arising in non-aqueous solvents.[15b] In preliminary experiments (see Section 2.1 of the Supporting Information) 0.5 mm solutions of cryptand[2.2.2] were titrated with HCl in order to characterise the two protonation steps. The result of this experiment led to some confusion related to the appearance of ITC end points after addition of about 0.7 and 2.0 equivalents of acid (see Figure S1 in the Supporting Information). If the literature values of the acidity constants for cryptand[2.2.2][26] are used for a speciation calculation at this concentration, about 60 % of the added cryptand[2.2.2] would have been converted to its monoprotonated derivative along with an equivalent amount of hydroxide ions, so that the ITC results can be explained by the initial dominance of heat evolution due to the reaction H + + OH , which would of course reduce the accuracy of the determination of the enthalpy change for the first step of cryptand[2.2.2] protonation. Since experiments on the protonation of cryptand[2.2.2] at a concentration of 1 mm led to ITC end points close to one and two equivalents of acid (Figure 1 b), all further experiments
Figure 1. a) Representative ITC raw data for interaction of cryptand[2.2.2] and HCl, obtained by using a 1.0 mm solution of cryptand[2.2.2] and a 30.2 mm solution of HCl. b) Left y axis: experimental (&) and calculated (~) heats for each injection; right y axis: speciation curves relative to cryptand[2.2.2] (c free cryptand, c monoprotonated cryptand H[2.2.2] + , c diprotonated cryptand H2[2.2.2]2 + ).
ChemPhysChem 0000, 00, 1 – 7
&2&
These are not the final page numbers! ÞÞ
CHEMPHYSCHEM ARTICLES were performed at this concentration. However, this did not completely exclude the effects of ligand hydrolysis (at 103 m, ca. 20 % conversion to H[2.2.2] + is still expected and must be taken into account). Since acidity constants determined by ITC were not available, their measurement was made part of the present work to give a consistent data set for the treatment of K + complexation. All experiments were performed at 298 K in degassed solutions. The initial titrand was a 1.1 mm solution of cryptand[2.2.2], which was loaded into the sample cell, and the titrants for the consecutive experiments were in order a 22 mm solution of potassium chloride, a 11.5 mm solution of HCl and a 20 mm solution of TMG. The concentration of the titrant must be at least ten times higher than that of the titrand, because the volume of the calorimetric pipette that contains the titrant is smaller than that of the sample cell. Heats of dilution for all reagents were determined in separate experiments and used to correct the raw ITC data. In all cases, tip-diffusion errors required the first and sometimes the second titration points to be discarded. Full details of the procedure are given in the Supporting Information. Moreover, for the second and third titration the concentration of the components present in the cell must be calculated by taking into account that VP-ITC operates in overfilled mode and the concentrations of the components in the cell change after every injection from the pipette into the cell (see Section 3 of the Supporting Information). Fitting of ITC data with the program HypDH requires the development of a chemical model for the studied system. As mentioned above, the acid and base were chosen so as to avoid the complication of inclusion reactions other than that of the potassium ion. Whereas various crystal structure determinations on acid salts of cryptand[2.2.2] have shown that in the solid state the proton attached to a cryptand N atom may be directed exo or endo with respect to the cavity, in solution it is known that inversion at these centres (in the unprotonated species as well) is fast,[18, 26b] so that the assumption was made that cryptand[2.2.2], H[2.2.2] + and H2[2.2.2]2 + could each be treated as a single species for a “slow” technique such as ITC. Since monoprotonation is considered to involve only one of the possible coordination sites of cryptand[2.2.2], it is conceivable that the species KH[2.2.2]2 + could be formed, but 1 H NMR spectroscopic monitoring of the titration of a solution of K[2.2.2] + in D2O with DCl (see Section 4 of the Supporting Information) showed that all changes could be accounted for in terms of the spectra of K[2.2.2] + and the two protonated forms of cryptand[2.2.2]. Hence, Scheme 1 is considered to fully describe the chemistry studied in the present work. As the auto-ionisation of water is a rather well studied reaction, literature values for pKw (13.99)[27] and DHw (13.4 kcal mol1)[28] at zero ionic strength were incorporated as constants in the models for all systems, but small errors may have resulted from the fact that the ionic strength, although low in all cases, did vary (from ca. 0.0002 to 0.003 m) as a result of the way the titrations were conducted. Since cryptand protonation/deprotonation reactions were involved in all systems studied, the simple titration of cryptand 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemphyschem.org [2.2.2] with HCl was used to provide values of the thermodynamic parameters for the two protonation steps, under the assumption that any protonation of ether O sites could be ignored under the conditions used. The parameters describing the complexation of the potassium ion (logarithm of the stability constant lg KK and the thermodynamic parameters) were in fact calculated by maintaining constant all the parameters concerning protonation/deprotonation reactions (pKa1, pKa2, enthalpies of protonation). The elaboration of the raw ITC data presented in Figure 1 a for a protonation experiment in which a solution of cryptand[2.2.2] with concentration close to 103 m was titrated is shown in Figure 1 b. It shows, as noted above, two apparent end points at molar ratios close to 1.0 and 2.0, which indicate a negligible influence of hydrolysis in the initial solutions of the cryptand. The values obtained for the protonation-step parameters (Table 1) compare well with those obtained by other methods,[26] in which supporting electrolytes were employed to give an ionic strength of 0.1 m. After the determination of pKa1 and pKa2 and their related enthalpy changes, the reversible complexation/decomplexation of K + by cryptand[2.2.2] was studied by in situ consecutive experiments in which pH-modulated switching was monitored by ITC. First, the simple complexation of cryptand[2.2.2] with potassium ion in water was executed by titrating a 1.1 mm solution of cryptand[2.2.2] with a 20 mm solution of potassium chloride. The raw ITC data and the observed and calculated heat as function of the molar KCl/cryptand ratio are shown in Figure 2. They reveal again that any influence of cryptand hydrolysis is hidden within the effects arising from K + complexation. Here, the cryptand protonation parameters determined from the preceding experiment were used as fixed constants (along with Kw and DHw) in the fitting, in order to define the equilibrium constant and enthalpy changes for the potassium complex. Although the values obtained were again in good agreement with literature data (ionic strength 0.05 or 0.1 m),[21, 26a, 29] reversal of the initial ligand hydrolysis during the titration had a significant influence on the heat changes, and the value found for KK indicates that, at a concentration near 103 m, any dissociation of the complex would be slight. For these reasons, a better way to perform the titration could be to add HCl to a solution of the complex. This experiment corresponds to the second in situ consecutive titration, carried out by titrating a 11.5 mm solution of HCl into the sample cell that now contains all of the cryptand in its cryptate form. Protonation of the cryptand causes displacement of the potassium ion and thus affects in particular the first part of the titration, as it is evident by comparing Figures 1 and 3. Indeed, the results obtained from the deceptively simple raw ITC data curve for this reverse titration with all other parameters kept as fixed constants (Figure 3) provides values that are still within the error bar of those reported in the literature (see Table 1).[21] To reassemble the K + complex of the cryptand, in situ deprotonation was performed by titrating the solution of H2[2.2.2]2 + in the presence of potassium ions with a 20 mm solution of TMG. Figure 4 shows the raw ITC data and the observed and calculated step reaction heats. In this case the enthalpogram clearly shows two inflection points. The first inflecChemPhysChem 0000, 00, 1 – 7
&3&
These are not the final page numbers! ÞÞ
10.78 0.01 11.06 0.01 10.43 0.01 5.5 0.1 5.8 0.1 4.9 0.1
[a] The values calculated as output parameters are given in italics; the other values are those employed in each fitting as fixed parameters. The Gibbs energy (DG8 and entropy DS8 were calculated by using the DH DG classic thermodynamic equations: DG ¼ RT ln K and DS ¼ T . The errors were generated by the fitting process of the program HypDH. Refs. [21, 29a]: The stability constants were determined by pHmetric and potentiometric methods at an ionic strength of I = 0.05 m and T = 25 8C, and the enthalpies by calorimetric measurements. Refs. [26a, 30]: The stability constants were determined by a pH-metric method at I = 0.1 m and T = 25 8C, and the enthalpies by calorimetric measurements.
11.00 0.48 10.56 0.49 12.55 0.48
11[26a] 18.3[26a] 14.94 3.11
9.60 0.05[21] 10.00 0.01[30] 9.71[26a] 9.1 0.4 9.1 9.1 9.1 literature[21, 26a, 30]
[2.2.2] + HCl [2.2.2] + KCl K[2.2.2] + + HCl K + + (H2[2.2.2])2 + + TMG
H + + [2.2.2]!(H[2.2.2]) + pKa2 DH8 [kcal mol1]
10.8[26a] 11.02 0.01 11.02 11.02 11.02
DG8 [kcal mol1]
13.25[26a] 12.41 0.54
8.2[26a] 4.67 1.78
DS8 [cal mol1 K1]
7.28 0.05[21] 7.53 0.04[30] 7.31[26a] 6.4 0.7 6.4 6.4 6.4
4.5[26a] 4.27 0.1 4.27 4.27 4.27
9.95[26a] 8.74 0.95
5.58[26a]
7.61[26a]
7.50 0.14 7.92 0.14 6.68 0.14
11.5[26a]
14.1 2[29a] 11.4 0.2[29a]
DS8 [cal mol1 K1] H + + (H[2.2.2]) + !(H2[2.2.2])2 + pKa1 DG8 DH8 [kcal mol1] [kcal mol1]
5.4 0.4[21]
www.chemphyschem.org
Titration
Table 1. Selected thermodynamic parameters of the in situ reversible H + -dependent assembly between cryptand[2.2.2] and K + .[a]
[2.2.2] + KCl!K[2.2.2] + log KK DH8 [kcal mol1]
DG8 [kcal mol1]
DS8 [cal mol1 K1]
CHEMPHYSCHEM ARTICLES
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 2. a) Representative ITC raw data for the interaction of KCl and cryptand[2.2.2] showing 34 consecutive injections of 3 mL. b) Left y axis: experimental (&) and calculated (~) heats for each injection; right y axis: speciation curves relative to cryptand[2.2.2] (c cryptate K[2.2.2] + , c free cryptand, c monoprotonated cryptand H[2.2.2] + , c diprotonated cryptand H2[2.2.2]2 + ).
tion point, at a molar base/cryptand ratio of about 0.3, corresponds to the reaction of the excess acid from the previous titration (Figure 3) on addition of the base. In fact, if we perform the second experiment (Figure 3) and stop the titration exactly at a molar ratio equivalent to two (to avoid the excess of acid), the first part of the curve due to neutralisation of the acid with the base is not observed (see Section 2.4 in the Supporting Information). The second inflection point at a molar ratio of 2.27 is characteristic of the completion of the process of deprotonation of the second nitrogen atom and inclusion of the potassium ion (see speciation curves in Figure 4 b). Also in this case the curve is not a simple bisigmoidal curve, because parallel inclusion of the potassium ion is occurring. Although hydrolysis of a solution of cryptand[2.2.2] dihydrochloride may be less extensive than that of the free base, titration with alkali rather than acid can involve greater problems of reagent purity. The use of a base such as TMG, chosen both to avoid cryptand inclusion reactions and because it is more easily obtained in a pure form than guanidine itself, illustrates another problem, in that additional thermodynamic parameters are required to describe the complete system. For TMG, an approximate pKa value is available,[22] but its heat of neutralisation is not known, although it is expected to be similar to that of guanidine and thus close to that of OH .[31] Indeed, the value of the association constant of the potassium ion (lg KK), determined by keeping all the parameters constant except the enthalpy of protonation of TMG and the enthalpy of formation of the cryptate complex, is less accurate than that found in previous titrations. Moreover the value obtained for ChemPhysChem 0000, 00, 1 – 7
&4&
These are not the final page numbers! ÞÞ
CHEMPHYSCHEM ARTICLES
www.chemphyschem.org nificance in the absence of a precise pKa value for aqueous media.
3. Conclusions
Figure 3. a) Representative ITC raw data for the interaction of cryptate K[2.2.2] + and HCl. b) Left y axis: experimental (&) and calculated (~) heats for each injection; right y axis: speciation curves relative to cryptand[2.2.2] (c cryptate K[2.2.2] + , c free cryptand, c monoprotonated cryptand H[2.2.2] + , c diprotonated cryptand H2[2.2.2]2 + .
Figure 4. a) Representative ITC raw data for the deprotonation of the cryptand[2.2.2] in the presence of the potassium cation. b) Left y axis: experimental (&) and calculated (~) heats for each injection; right y axis: speciation curves relative to cryptand[2.2.2] (c free cryptand, c monoprotonated cryptand H[2.2.2] + , c diprotonated cryptand H2[2.2.2]2 + , c cryptate K[2.2.2] + ).
the heat of neutralisation of TMG is very high (13.91 0.01 kcal mol1). We can conclude that the last result is not of great sig 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
We have shown for the first time that ITC can be used to monitor the reversible process of complexation/decomplexation in a genuinely bistable supramolecular system in which the pH change is used as a remote control. As a model system. we chose the complexation of potassium ions by cryptand[2.2.2] . Its decomplexation followed by recomplexation was triggered by subsequent addition of a strong acid and a strong base, respectively. In all subsequent titrations, it was possible to determine the association constant and the thermodynamic parameters relative to the process of complexation of potassium ion by cryptand[2.2.2]. We emphasize that the protic nature of water as a medium engenders complications in the analysis of equilibria involving the binding of acidic and basic substrates, and hence in the present case a cryptand concentration of at least 1 mm is necessary to obtain optimum precision of the derived thermodynamic parameters. The results obtained by ITC for the protonation and complexation of K + by cryptand[2.2.2] are in agreement with those reported in the literature, based on studies carried out with other techniques. Even though it is known that the precision reached by other methods in the calculation of the stability constant (i.e. potentiometric methods) can be higher, the major advantage of ITC is the possibility to determine all the thermodynamic parameters in a single set of measurements. The separation of enthalpic and entropic contributions from the reaction free energy, which is readily obtained by ITC measurements, is instrumental for assessing the factors that can influence host–guest recognition. Unfortunately, if only free energies (equilibrium constants) are known, there is a tendency to emphasize influences such as bonding and solvation considered as contributions to an enthalpy change, whereas in many instances the entropy change dominates the enthalpy change as a contribution to the overall change in Gibbs free energy.[32] Recognition of this can provide a rather different perspective on the nature of the equilibrium involved. A comparison of the binding of one proton or one potassium ion by cryptand[2.2.2] is interesting in that, even though the guestbinding interactions are rather different, the enthalpy changes in the reactions are similar, although the entropy change is positive for protonation and negative for K + binding. The entropy decrease on K + binding has been ascribed to enhanced solvent structuring about the cation masked by the organic sheath of the cryptand ligand,[18] an argument which implies that K + binding might be simply enhanced by functionalising the exterior of cryptand[2.2.2] with hydrophilic units.
Acknowledgements We thank Thomas M. Hermans for enlightening discussions. This work was financially supported by EC through the Marie-Curie ITN GENIUS (PITN-GA-2010-264694), the ERC project SUPRAFUNCChemPhysChem 0000, 00, 1 – 7
&5&
These are not the final page numbers! ÞÞ
CHEMPHYSCHEM ARTICLES
www.chemphyschem.org
TION (GA-257305), the Agence Nationale de la Recherche through the LabEx project Chemistry of Complex Systems (ANR-10-LABX0026_CSC) and the International Center for Frontier Research in Chemistry (icFRC). Keywords: calorimetry · cryptands · isothermal titration calorimetry · potassium · supramolecular chemistry [1] a) J. M. Lehn, Science 1993, 260, 1762 – 1763; b) T. F. A. de Greef, E. W. Meijer, Nature 2008, 453, 171 – 173; c) G. R. Desiraju, Acc. Chem. Res. 2002, 35, 565 – 573. [2] E. Freire, A. J. Ruben, Y. Kiso, Chem. Biol. Drug Des. 2006, 67, 2 – 4. [3] a) K. E. Riley, P. Hobza, Acc. Chem. Res. 2013, 46, 927 – 936; b) J. L. Asensio, A. Ard, F. J. CaÇada, J. Jimnez-Barbero, Acc. Chem. Res. 2013, 46, 946 – 954. [4] a) T. Wiseman, S. Williston, J. F. Brandts, L. N. Lin, Anal. Biochem. 1989, 179, 131 – 137; b) J. F. Brandts, L. N. Lin, T. Wiseman, S. Williston, C. P. Yang, Am. Lab. 1990, 22, 30 – 41; c) I. Wadso, Chem. Soc. Rev. 1997, 26, 79 – 86. [5] a) R. J. Falconer, B. M. Collins, J. Mol. Recognit. 2011, 24, 1 – 16; b) P. Arranz-Mascaros, C. Bazzicalupi, A. Bianchi, C. Giorgi, M. L. GodinoSalido, M. D. Gutierrez-Valero, R. Lopez-Garzon, M. Savastano, J. Am. Chem. Soc. 2013, 135, 102 – 105. [6] “Isothermal Titration Calorimetry in Supramolecular Chemistry,” F. P. Schmidtchen in Analytical Methods in Supramolecular Chemistry (Ed.: C. A. Schalley), Wiley-VCH, Weinheim, 2007, pp. 55 – 78. [7] a) V. Ball, C. Maechling, Int. J. Mol. Sci. 2009, 10, 3283 – 3315; b) J. Ladbury, I. Haq, J. Mol. Recognit. 2000, 13, 188 – 197. [8] a) N. Grossoehme, A. Spuches, D. Wilcox, J. Biol. Inorg. Chem. 2010, 15, 1183 – 1191; b) L. Q. Hatcher, L. Hong, W. D. Bush, T. Carducci, J. D. Simon, J. Phys. Chem. B 2008, 112, 8160 – 8164; c) L. A. Hong, T. M. Carducci, W. D. Bush, C. G. Dudzik, G. L. Millhauser, J. D. Simon, J. Phys. Chem. B 2010, 114, 11261 – 11271. [9] a) W. B. Turnbull, A. H. Daranas, J. Am. Chem. Soc. 2003, 125, 14859 – 14866; b) N. Rawat, R. B. Gujar, M. S. Murali, B. S. Tomar, V. K. Manchanda, Thermochim. Acta 2009, 488, 21 – 26; c) R. M. Izatt, J. S. Bradshaw, S. A. Nielsen, J. D. Lamb, J. J. Christensen, Chem. Rev. 1985, 85, 271 – 339. [10] a) A. M. P. Pederson, E. M. Ward, D. V. Schoonover, C. Slebodnick, H. W. Gibson, J. Org. Chem. 2008, 73, 9094 – 9101; b) A. M. P. Pederson, R. C. Vetor, M. A. Rouser, F. Huang, C. Slebodnick, D. V. Schoonover, H. W. Gibson, J. Org. Chem. 2008, 73, 5570 – 5573. [11] a) V. Amendola, G. Alberti, G. Bergamaschi, R. Biesuz, M. Boiocchi, S. Ferrito, F. P. Schmidtchen, Eur. J. Inorg. Chem. 2012, 3410 – 3417; b) P. Mateus, N. Bernier, R. Delgado, Coord. Chem. Rev. 2010, 254, 1726 – 1747. [12] a) Y. Inoue, M. V. Rekharsky, Chem. Rev. 1998, 98, 1875 – 1917; b) K. Bouchemal, S. Mazzaferro, Drug Discovery Today 2012, 17, 623 – 629. [13] a) A. F. Danil de Namor, R. M. Cleverley, M. L. Zapata-Ormachea, Chem. Rev. 1998, 98, 2495 – 2526; b) J. Hornung, D. Fankhauser, L. D. Shirtcliff,
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
[14]
[15]
[16] [17] [18] [19] [20]
[21] [22] [23] [24] [25] [26] [27] [28] [29]
[30] [31] [32]
A. Praetorius, W. B. Schweizer, F. Diederich, Chem. Eur. J. 2011, 17, 12362 – 12371; c) F. Corbellini, L. Di Costanzo, M. Crego-Calama, S. Geremia, D. N. Reinhoudt, J. Am. Chem. Soc. 2003, 125, 9946 – 9947; d) C. Sgarlata, C. Bonaccorso, F. G. Gulino, V. Zito, G. Arena, D. Sciotto, New J. Chem. 2009, 33, 991 – 997; e) M. Dionisio, G. Oliviero, D. Menozzi, S. Federici, R. M. Yebeutchou, F. P. Schmidtchen, E. Dalcanale, P. Bergese, J. Am. Chem. Soc. 2012, 134, 2392 – 2398. a) F. Biedermann, V. D. Uzunova, O. A. Scherman, W. M. Nau, A. De Simone, J. Am. Chem. Soc. 2012, 134, 15318 – 15323; b) F. Biedermann, M. Vendruscolo, O. A. Scherman, A. De Simone, W. M. Nau, J. Am. Chem. Soc. 2013, 135, 14879 – 14888. a) J. L. Sessler, D. E. Gross, W. S. Cho, V. M. Lynch, F. P. Schmidtchen, G. W. Bates, M. E. Light, P. A. Gale, J. Am. Chem. Soc. 2006, 128, 12281 – 12288; b) M. Valik, V. Kral, E. Herdtweck, F. P. Schmidtchen, New J. Chem. 2007, 31, 703 – 710. T. F. A. De Greef, M. M. J. Smulders, M. Wolffs, A. P. H. J. Schenning, R. P. Sijbesma, E. W. Meijer, Chem. Rev. 2009, 109, 5687 – 5754. M. Ostermeier, M. A. Berlin, R. M. Meudtner, S. Demeshko, F. Meyer, C. Limberg, S. Hecht, Chem. Eur. J. 2010, 16, 10202 – 10213. J.-M. Lehn, Acc. Chem. Res. 1978, 11, 49 – 57. M. Barboiu, J.-M. Lehn, Proc. Natl. Acad. Sci. USA 2002, 99, 5201 – 5206. a) G. P. Spada, S. Pieraccini, S. Masiero, O. Pandoli, P. Samori, Org. Lett. 2006, 8, 3125 – 3128; b) A. Ciesielski, S. Lena, S. Masiero, G. P. Spada, P. Samori, Angew. Chem. 2010, 122, 2007 – 2010; Angew. Chem. Int. Ed. 2010, 49, 1963 – 1966. J.-M. Lehn, J. P. Sauvage, J. Am. Chem. Soc. 1975, 97, 6700 – 6707. R. W. Taft, H. B. Yang, J. Am. Chem. Soc. 1971, 93, 1310 – 1311. P. Gans, A. Sabatini, A. Vacca, J. Solution Chem. 2008, 37, 467 – 476. Y. Zhang, S. Akilesh, D. E. Wilcox, Inorg. Chem. 2000, 39, 3057 – 3064. D. E. Gross, F. P. Schmidtchen, W. Antonius, P. A. Gale, V. M. Lynch, J. L. Sessler, Chem. Eur. J. 2008, 14, 7822 – 7827. a) G. Anderegg, Helv. Chim. Acta 1975, 58, 1218 – 1225; b) R. Pizer, J. Am. Chem. Soc. 1978, 100, 4239 – 4241. A. V. Bandura, S. N. Lvov, J. Phys. Chem. Ref. Data 2006, 35, 15 – 30. I. Grenthe, H. Ots, M. Ginstrupp, Acta Chem. Scand. 1970, 24, 1067 – 1080. a) E. Kauffmann, J.-M. Lehn, J. P. Sauvage, Helv. Chim. Acta 1976, 59, 1099 – 1111; b) J. Gutknecht, H. Schneider, J. Stroka, Inorg. Chem. 1978, 17, 3326 – 3329. F. Arnaud-Neu, B. Spiess, M.-J. Schwing-Weill, Helv. Chim. Acta 1977, 60, 2633 – 2643. L. Fabbrizzi, M. Micheloni, P. Paoletti, G. Schwarzenbach, J. Am. Chem. Soc. 1977, 99, 5574 – 5576. a) F. P. Schmidtchen, Coord. Chem. Rev. 2006, 250, 2918 – 2928; b) F. P. Schmidtchen, Chem. Soc. Rev. 2010, 39, 3916 – 3935; c) A. Ursu, F. P. Schmidtchen, Angew. Chem. 2012, 124, 246 – 250; Angew. Chem. Int. Ed. 2012, 51, 242 – 246.
Received: May 3, 2014 Published online on && &&, 2014
ChemPhysChem 0000, 00, 1 – 7
&6&
These are not the final page numbers! ÞÞ
ARTICLES pH-modulated complexation and release of metal ions based on reversible assembly of the K + complex of cryptand[2.2.2] in aqueous solution is explored by means of isothermal titration calorimetry. The thermodynamic parameters associated with the interconversion between cryptand and cryptate are elucidated by sequential addition of KCl, HCl and 1,1,3,3-tetramethylguanidine.
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
M. G. del Rosso, A. Ciesielski, S. Colella, J. M. Harrowfield, P. Samor* && – && Isothermal Titration Calorimetry Study of a Bistable Supramolecular System: Reversible Complexation of Cryptand[2.2.2] with Potassium Ions
ChemPhysChem 0000, 00, 1 – 7
&7&
These are not the final page numbers! ÞÞ
