- Email: [email protected]
Specific ion adsorption on alkyl carboxylate surfactant layers
Accepted Manuscript Title: Specific Ion Adsorption on Alkyl Carboxylate Surfactant Layers Author: Susanne Dengler Gordon J.T. Tiddy Lydia Zahnweh Werner Kunz PII: DOI: Reference:
S0927-7757(14)00539-1 http://dx.doi.org/doi:10.1016/j.colsurfa.2014.06.001 COLSUA 19280
To appear in:
Colloids and Surfaces A: Physicochem. Eng. Aspects
Received date: Revised date: Accepted date:
18-3-2014 31-5-2014 2-6-2014
Please cite this article as: S. Dengler, G.J.T. Tiddy, L. Zahnweh, W. Kunz, Specific Ion Adsorption on Alkyl Carboxylate Surfactant Layers, Colloids and Surfaces A: Physicochemical and Engineering Aspects (2014), http://dx.doi.org/10.1016/j.colsurfa.2014.06.001 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Colloids and Surfaces A: Physiochem. Eng. Aspects xx (2014) xx-xx
Specific Ion Adsorption on Alkyl Carboxylate Surfactant Layers Susanne Dengler1, Gordon J. T. Tiddy2, Lydia Zahnweh1, Werner Kunz*1 Institut für Physikalische und Theoretische Chemie, Universität Regensburg, Regensburg, Germany
2
School of Chemical Engineering and Analytical Science, University of Manchester, Oxford Road, Manchester, M13 9PL, UK
HIGHLIGHTS •
GRAPHICAL ABSTRACT
us
•
an
•
cr
Rubidium and caesium cations prefer bb-binding site. Affinity of rubidium and caesium to carboxylate is very weak and there is no detectable ion specificity. A sufficient interaction between anion and cation is a precondition for the determination of ion specificities by NMR.
ip t
1
ABSTRACT
Article history: Received Received in revised from Accepted Available online
Mixtures of dodecyl carboxylate, CsOH, RbOH and octanol in D2O were investigated to determine the relative ion specificity of Cs+ and Rb+ in their interactions with dodecyl carboxylate in a liquid crystalline phase. Octanol as cosurfactant was necessary to obtain the lamellar liquid crystalline phase. The technique used was the measurement of the NMR quadrupole splitting of 137Cs and 87Rb. The ratio of Rb to Cs was varied as well as the concentration of surfactant and octanol, whereas the ratio of surfactant to octanol was kept constant at 1:3. It turned out that there is no significant difference between both ions in their affinity towards carboxylate head groups. © 2014 Elsevier B.V. All rights reserved
Specific-ion effects
Ac ce p
NMR quadrupole splitting
d
Hofmeister Series
te
Keywords:
M
ARTICLE INFO
Carboxylate
1.
Introduction
1.1. Ion specificity
The specific interaction of charged headgroups with different counter ions matters in many biological as well as industrial systems [1, 2]. An example is the transition of micellar structures to vesicles [3]. To cite a more applied case, the Krafft temperature of surfactants used in washing processes increases with increasing ion binding. These behaviours can often be explained by Collins’ concept of matching water affinities, together with a Hofmeister-like classification of both ions and charged headgroups (Fig. 1) [49]. In Hofmeister series anions and cations are listed in separate series, sorted from salting out to salting in of proteins.
Figure 1: A typical Hofmeister series.
In this concept, it is supposed that two ions with high charge densities (“hard” ions or headgroup) of opposite *Corresponding author: Tel.: +49 941 9434296 E-mail address: [email protected]
charge have a very strong electrostatic interaction which is greater than the interaction between the ion and the water shell around them resulting in the formation of a contact ion pair. Some of the hydration water between the ions is released. Regarding “soft” ions (low charge density), the interaction between water and these ions is weak and hydration water molecules between the ions can easily be released, which enables also the formation of a contact ion http://dx.doi.org/10.1016/j.colsurfa.2014.xx.xxxx 0927-7757/© 2014 Elsevier B.V. All rights reserved.
Page 1 of 8
2
S. Dengler et al. / Colloids and Surfaces A: Physiochem. Eng. Aspects xx (2014) xx-xx
The core requirements for detecting quadrupole splittings of counter ions are the existence of an electric quadrupole moment and an anisotropic environment of the ion. Hence, only ions with a spin quantum number I > 1/2 can be detected. The anisotropy causes an electric field gradient with some orientations favoured over others. The splitting of the NMR resonance is the result of the interaction between the electric field gradient and the electric quadrupole moment at the nucleus. In an isotropic solution the orientation dependent quadrupole splitting averages to zero. For an anisotropic solution the resonance splits up into 2I peaks [4, 5, 15]. The frequency between the central peak and the first neighbouring peak is defined as the quadrupole splitting Δ. The magnitude of Δ is given by equation 2 for a powder sample (one with no macroscopic alignment of the liquid crystal phase) with Si, the order parameter describing the orientation of the fraction of molecules at site i (pi), given by Si = <3cos2θDMi-1>/2. θDMi is the angle between the liquid crystal axis (the director) and the electric field gradient. 4νQi is equal to the quadrupole coupling constant. Δ is a weighted average of the different values for the i sites. The quadrupole coupling constant is not known. Hence, absolute values of Δ can`t be determined. But the relative change of Δ can be observed and delivers an insight in the binding situation of the ions [4, 6, 14, 16].
η = 1 + Ac 1 / 2 + Bc η0
Ac ce p
te
d
The coefficient can be positive or negative. Strongly hydrated ions have positive B coefficients and weakly hydrated ions have negative B coefficients. The point of change in JonesDole coefficient sign represents the ideal behaviour at which the interaction water-ion is the same as water-water interaction. In table 1 the B coefficients of some ions are given [12, 13]. Consequently, Rb+ as well as Cs+ are weakly hydrated whereas the carboxylate head group of fatty acid is strongly hydrated. So, according to Collins concept the interactions between the head group and the counterions should be very small and of comparable weakness. Admittedly, this is a very indirect conclusion, because it is not evident that viscosities of simple electrolyte solutions reflect partition coefficients as they are measured here in liquid crystalline systems. Nevertheless, we could show in a former work that the different ion bindings of lithium and sodium ions to carboxylate and sulphate headgroups in such lamellar systems reflect surprisingly well the expected trend as predicted by Collins. As in the present study, the used technique was NMR quadrupole splitting [14]. Here we check now if the expected (non-)binding of caesium and rubidium to carboxylate headgroups can also be detected by this technique and we further show how nicely the results fit with viscosity B coefficients.
Cations
B
Anions
B
Ca2+
0.285
CH3COO-
0.250
Li+
0.150
SO42-
0.208
Na+
0.086
Cl-
-0.007
Rb+
-0.030
NO3-
-0.046
Cs+
-0.045
I-
-0.068
Table 1: Jones-Doyle viscosity B coefficients
Δ=
∑pv
i Qi
Si
(2)
1.3. Ion interaction within lamellar phase
Within the investigated concentration range and in all compositions considered here, lamellar phases exist (Fig. 3) [14]. The ionic strength of the samples in water without considering the ion-binding is given in table 2.
M
(1)
us
Ion-water interactions are reflected in many properties of electrolyte solutions, for example in the viscosity of aqueous solutions. Especially, the Jones-Dole viscosity B coefficient is believed to be a good parameter to estimate ion hydration. It can be calculated from salt solution viscosity η by equation 1 with η0 the viscosity of pure water, A, an electrostatic term, being 1 for moderate salt concentrations and c the salt concentration. B is a direct measure of the strength of ionwater interactions normalized to the strength of water-water interactions.
an
Figure 2: Depending on the ion size direct ion pairs are formed.
cr
ip t
pair of two soft oppositely charged ions. By contrast, for the combination of hard and soft ions the formation of contact ion pair is not possible. The soft ion cannot penetrate the water shell of the hard ion. Hence, the general rule of this concept is “like seeks like” (Fig. 2) [10, 11].
Table 2: Ionic strength of the samples for all compositions and all concentrations.
The counter ions can be free in the bulk phase, bound between the surfactant head groups (bb-site) or on the surface of the head groups (bs-site). Depending on the binding situation of the cation, the corresponding quadrupole splitting Δ changes. For free ions the contribution to Δ is zero. Bound ions have an anisotropic environment, giving a certain Δ value. The Δ value can be positive or negative depending on the angle between the liquid crystal axis and the electric field gradient. Ions bound in the bs-site give a positive value, whereas ions bound at the bb-site give a negative quadrupole splitting value.
Figure 3: Schematic Drawing of counter ion binding at the lamellar surface. Three possible binding situations are shown: (a) Ion moves freely in the bulk phase. (b) The ion is located perpendicular to the surfactant head group in the lamellar layer. (c) The ion is bound between amphiphile head groups.
2.
Materials and Methods
2.1. Materials
1.2. NMR quadrupole splitting
Page 2 of 8
3
S. Dengler et al. / Colloids and Surfaces A: Physiochem. Eng. Aspects xx (2014) xx-xx
Dodecanoic acid (COOH, 99.6%) was received from Alfa Aesar (USA). Caesium hydroxide (CsOH, 50 wt% in H2O, 99.9%), rubidium hydroxide (RbOH, 50 wt% in H2O, 99.9%) and octanol were obtained from Sigma Aldrich (Germany). Deuterated water was purchased from Deutero (Germany). 2.2. Methods
Results
ip t
cr us
87
Figure 5: Rb-NMR spectrum of a 85 wt% sample with a composition of CsDC/RbDC = 0:1 at 300 K. The relative intensity of the signals is plotted against the frequency in [Hz].
3.2. Influence of surfactant concentration on the quadrupole splitting
The variations of the caesium and rubidium splitting for the single ion systems, with increasing surfactant and octanol concentration are shown in figure 6 and figure 7. The caesium splitting decreases until a concentration of 75 wt%, whereas the rubidium splitting decreases initially and is roughly constant above a concentration of 55 wt% and decreases again for the highest concentration. This nonuniform behaviour might be due to the different binding sites at the lamellar interface as can be seen above in figure 3. However, the decrease of rubidium splitting at 85 wt% surfactant after increase at 75 wt% is unlikely and can not be explained by different binding sites. It might result from only partially dissolved Rb or from the removal of inner-sphere hydration.
M
3.
Figure 4: 137Cs-NMR spectrum of a 85 wt% sample with a composition of CsDC/RbDC = 1:0 at 300 K. The relative intensity of the signals is plotted against the frequency in [Hz].
an
137 Cs- and 87Rb-NMR were performed using a Bruker Avance 400 spectrometer operating at 52.48 MHz in the case of caesium and operating at 130.93 MHz in the case of rubidium. For 137Cs resonance a 30° pulse length of 10.90 μs and for 87Rb resonance a 30° pulse length of 7.00 μs was used. The spectra were collected with a number of scans of 1024. The measurements were performed at 300 K. Each sample (3g) was prepared in a tube and mixed with a magnetic stirrer until the sample was homogeneous. Then about 700 mg was transferred in a NMR tube (5 mm o.d.). Each sample contains D2O, octanol, surfactant and H2O from the alkali metal hydroxid. The surfactants are either caesium dodecanoate (CsDC) or rubidium dodecanoate (RbDC). The surfactants were prepared in situ from dodecanoic acid and the corresponding alkali metal hydroxide. The molar ratio of CsDC and RbDC varies from 0 to 100 % caesium surfactant in 20 % steps. In each sample the molar ratio of surfactant to octanol was 1:3. The concentrations of surfactant + octanol in D2O are 45 wt% to 85 wt% in 10 wt% steps. In order to prove the formation of a lamellar phase, the samples were observed visually via crossed polar filters. All samples appeared as single phases with typical lamellar textures.
3.1. Phase structure and NMR spectra
Ac ce p
te
d
A Cs splitting was observed for all samples, whereas we had problems to detect the rubidium splitting in the case of 85 wt% samples. The splitting of some compositions of the 85 wt% samples was too broad to be detected within the chosen frequency range. In figure 4 and 5 typical 137Cs- and 87Rb-NMR spectra can be seen. The number of peaks is given by 2I. Accordingly, in the 137Cs-NMR spectrum (Cs: I = 7/2) seven peaks occur and in the 87Rb-NMR spectrum (Rb: I =3/2) three peaks can be observed. Due to the small quadrupole moment (Q = -4 * 10-31 m2) of Cs, the signals are narrow [17]. By contrast, the quadrupole moment of Rb is Q = 0.14 * 10-28m2, resulting in broader signals [18]. The quadrupole splitting (Δ) is defined as half the distance between the two first neighbouring peaks to the central peak. As can be seen in the figures below, the caesium splitting is clearly smaller than the rubidium splitting. As it is well known for liquid crystalline samples, asymmetric central peaks occur for both ions. This effect is known as chemical shift anisotropy [17]. Due to the lower splitting of caesium, the anisotropy is more pronounced in the case of 137Cs-NMR. Note that there are three little peaks or shoulders at ca.965 Hz, 403 Hz and -498 Hz. This is a hint at a possible different lamellar structure in equilibrium with the main one. Its appearance is not a consequence of an insufficient homogenisation of the samples, because we took care to thoroughly mix them. Further, samples were prepared twice and independently to check this point. These small extra peaks occur for all samples.
Figure 6: Caesium splitting with increasing concentration of surfactant and octanol at 300 K.
Figure 7: Rubidium splitting with increasing concentration of surfactant and octanol at 300 K.
Page 3 of 8
S. Dengler et al. / Colloids and Surfaces A: Physiochem. Eng. Aspects xx (2014) xx-xx
our results. The single ion splitting of both cations decreases with increasing concentration of surfactant and octanol up to a certain concentration and increases for higher concentrations. For rubidium splitting an unlikely decrease after increase is observed. Both ions prefer the bb-binding site (negative Δ values) resulting in a decreasing ion splitting. For a very high surfactant concentration the surfactant head groups get very close. Consequently ions are moved to the surfactant head group surface (bs-site; positive Δ values) resulting in an increased splitting. Hence we can speculate that the kink results from the closer packing of the surfactants. 4.2. Ion specifity of carboxylate
The caesium and rubidium splitting remain nearly constant for all concentrations and fractions, except of 85 wt%. This is due to the low affinity of carboxylate to caesium and rubidium. The kink observed for all concentrations except 85 wt% samples points to a special lattice structure of bound ions at 1:1 ratio Cs/Rb. But ions are only loosely bound. For 85 wt% the course of splitting is different. The caesium splitting decreases and the rubidium splitting shows a discontinuous trend. Probably rubidium surfactant was not completely dissolved for this concentration. Nevertheless, visually these samples appear as homogenous as all other samples. In summary, the affinity of rubidium and caesium to carboxylate is very weak in both cases and there is no detectable ion specificity, unlike the ion binding of lithium and sodium studied in a previous paper [14], where preferential binding of lithium was observed. But sodium and lithium having also positive B coefficients like carboxylate are supposed to interact much more strongly with carboxylate. Accordingly, a precondition for the determination of ion specificities by NMR seems to be a sufficient interaction between anion and cation.
M
an
us
In figures 8 and 9 the caesium and rubidium splitting of the 45 wt% samples with an increasing fraction of caesium and rubidium respectively is shown. Caesium splitting remains nearly constant for all composition and rubidium shows only a slight decrease. Hence carboxylate does not show any ion specificity towards caesium or rubidium for this composition. Similarly, for the 55, 65 and 75 wt% samples (Fig. 8 and 9), the Δ values of caesium remain constant. The rubidium splitting is nearly constant for 55 and 75 wt% and shows only small increase for 65 wt%. Consequently, for these compositions carboxylate also does not show any ion specificity towards caesium and rubidium. Except of 85 wt% samples a kink at ratio 1:1 Cs/Rb can be observed for both splittings. This non-monotonic behaviour might point to a special lattice structure for this composition. Finally, for the 85 wt% samples (Figs. 8 and 9) the Δ values of caesium decreases with increasing amount of caesium. In contrast, the rubidium splitting increases up to an amount of rubidium of 65 wt% with a slight decline for 55 wt% rubidium and decreases significantly for higher amount of Rb.
ip t
3.3. Ion specifity of carboxylate
cr
4
Acknowledgments
Ac ce p
te
d
Figure 8: Caesium splitting with increasing amount of caesium for: () 45 wt%, (z) 55 wt%. (S) 65 wt%, (T) 75 wt% and () 85 wt% CsDC/RbDC/octanol in D2O
Figure 9: Rubidium splitting with increasing amount of rubidium for: () 45 wt%, (z) 55 wt%. (S) 65 wt%, (T) 75 wt% and () 85 wt% CsDC/RbDC/octanol in D2O.
4.
Discussion
4.1. Influence of surfactant concentration on the quadrupole splitting
The concept of the Hofmeister series predicts a higher affinity of rubidium to carboxylate in comparison to caesium [8], but the series does not give any information, how string this difference might be. Jones-Dole viscosity B coefficients suggest only a very small interaction of both cations with the carboxylate head group. This is confirmed by
We thank Fritz Kastner, Annette Schramm and Georgine Stühler, the staff of the NMR department of the University of Regensburg for performing the NMR experiments.
References
[1] P. Koelsch, P. Viswanath, H. Motschmann, V.L. Shapovalov, G. Brezesinski, H. Moehwald, D. Horinek, R.R. Netz, K. Giewekemeyer, T. Salditt, H. Schollmeyer, R. von Klitzing, J. Daillant, P. Guenoun, Specific ion effects in physicochemical and biological systems: Simulations, theory and experiments, Colloids and Surfaces A, 303 (2007) 110-136. [2] E.R.A. Lima, E.C. Boström, J. Biscaia, F.W. Tavares, W. Kunz, Ion Specific Forces between Charged SelfAssembled Monolayers Explained by Modified DLVO Theory, Colloids and Surfaces A, 346 (2009) 11-15. [3] N. Vlachy, M. Drechsler, J.-M. Verbavatz, D. Touraud, W. Kunz, Role
Page 4 of 8
S. Dengler et al. / Colloids and Surfaces A: Physiochem. Eng. Aspects xx (2014) xx-xx
ip t
[12] K.D. Collins, Ions from the Hofmeister series and osmolytes: effects on proteins in solution and in the crystallization process. , Methods, 34 (2004) 300-311. [13] J.B. Robinson, J.M. Strottmann, E. Stellwagen, Prediction of neutral salt elution profiles for affinity chromatography Proceedings of the National Academy of Sciences, 78 (1981) 2287-2291. [14] S. Dengler, A. Klaus, G.J.T. Tiddy, W. Kunz, How specific are ion specifities? A pilot NMR study., Faraday Discussions, 160 (2012) 1-13.
M
[5] I.D. Leigh, M.P. McDonald, R.M. Wood, G.J.T. Tiddy, M.H. Trevethan, Structure of Liquid-crystalline Phases formed by Sodium Dodecyl Sulphate and Water as determined by Optical Microscopy, X-ray Diffraction and Nuclear Magnetic Resonance Spectroscopy, Journal of the Chemical Society, Faraday Transaction 1: Physical Chemistry in Condensed Phases, 77 (1981) 2867.
[11] P. Jungwirth, D.J. Tobias, Specific ion effects at the air/water interface., Chemical Reviews, 106 (2006) 12591281.
cr
[4] G. Lindblom, B. Lindman, G.J.T. Tiddy, Ion binding studied using quadrupole splittings of sodium-23(1+) ions in lyotropic liquid rystals. The dependence on surfactant type., Journal of American Chemical Society, 100 (1978) 2299.
Surfactants, and Deep Eutectic Solvents, Doctoral Thesis, University of Regensburg, Regensburg, 2013.
us
of the surfactant headgroup on the counterion specificity in the micelle-tovesicle transition through salt addition, Journal of Colloid and Interface Science, 319 (2008) 542-548.
an
5
Ac ce p
te
d
[6] N. Schwierz, D. Horinek, R.R. Netz, Reversed Anionic Hofmeister Series: The Interplay of Surface Charge and Surface Polarity., Langmuir, 26 (2010) 7370. [7] W. Kunz, Specific Ion Effects, 2010, World Scientific Publishing, Singapore.
[8] W. Kunz, Specific ion effects in colloidal and biological systems, Current Opinion in Colloid & Interface Science, 15 (2010) 34-39. [9] N. Vlachy, B. Jagoda-Cwiklik, R. Vacha, D. Touraud, P. Jungwirth, W. Kunz, Hofmeister series and specific interactions of charged headgroups with aqueous ions., Advances in Colloid and Interface Science, 146 (2009) 42. [10] D. Rengstl, Choline as a Cation for the Design of Low-toxic and Biocompatible Ionic Liquids,
[15] H. Wennerstrom, G. Lindblom, B. Lindman, Theoretical aspects on the NMR of quadrupolar ionic nuclei in micellar solutions and amphiphilic liquid crystals., Chemica Scripta, 6 (1974) 97. [16] P. Jungwirth, D.J. Tobias, Specific Ion Effects at the Air/Water Interface, Chemical Reviews, 106 (2006) 1259. [17] H. Wennerstrom, N.O. Persson, G. Lindblom, B. Lindman, Chemical Shift Anisotropies of 133Cs+ Counterions in Lyotropic Liquid Crystals., Journal of Magnetic Resonance, 30 (1978) 133. [18] U. Meyer-Berkhout, Betimmung der elektrischen Quadrupolmomente der Kerne Rb85 und Rb87 durch Messung der Hochfrequenzübergänge im angeregten 6P3/2-Term des Rb-Atoms.,
Page 5 of 8
6
S. Dengler et al. / Colloids and Surfaces A: Physiochem. Eng. Aspects xx (2014) xx-xx
Ac ce p
te
d
M
an
us
cr
ip t
Zeitschrift für Physik, 141 (1955) 185187.
Page 6 of 8
7
S. Dengler et al. / Colloids and Surfaces A: Physiochem. Eng. Aspects xx (2014) xx-xx
Highlights Rubidium and caesium cations prefer bb-binding site.
ip t
Affinity of rubidium and caesium to carboxylate is very weak and there is no
cr
detectable ion specificity. A sufficient interaction between anion
us
and cation is a precondition for the
an
determination of ion specificities by
Ac ce p
te
d
M
NMR
Page 7 of 8
Ac
ce
pt
ed
M
an
us
cr
i
*Graphical Abstract (for review)
Page 8 of 8
S0927-7757(14)00539-1 http://dx.doi.org/doi:10.1016/j.colsurfa.2014.06.001 COLSUA 19280
To appear in:
Colloids and Surfaces A: Physicochem. Eng. Aspects
Received date: Revised date: Accepted date:
18-3-2014 31-5-2014 2-6-2014
Please cite this article as: S. Dengler, G.J.T. Tiddy, L. Zahnweh, W. Kunz, Specific Ion Adsorption on Alkyl Carboxylate Surfactant Layers, Colloids and Surfaces A: Physicochemical and Engineering Aspects (2014), http://dx.doi.org/10.1016/j.colsurfa.2014.06.001 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Colloids and Surfaces A: Physiochem. Eng. Aspects xx (2014) xx-xx
Specific Ion Adsorption on Alkyl Carboxylate Surfactant Layers Susanne Dengler1, Gordon J. T. Tiddy2, Lydia Zahnweh1, Werner Kunz*1 Institut für Physikalische und Theoretische Chemie, Universität Regensburg, Regensburg, Germany
2
School of Chemical Engineering and Analytical Science, University of Manchester, Oxford Road, Manchester, M13 9PL, UK
HIGHLIGHTS •
GRAPHICAL ABSTRACT
us
•
an
•
cr
Rubidium and caesium cations prefer bb-binding site. Affinity of rubidium and caesium to carboxylate is very weak and there is no detectable ion specificity. A sufficient interaction between anion and cation is a precondition for the determination of ion specificities by NMR.
ip t
1
ABSTRACT
Article history: Received Received in revised from Accepted Available online
Mixtures of dodecyl carboxylate, CsOH, RbOH and octanol in D2O were investigated to determine the relative ion specificity of Cs+ and Rb+ in their interactions with dodecyl carboxylate in a liquid crystalline phase. Octanol as cosurfactant was necessary to obtain the lamellar liquid crystalline phase. The technique used was the measurement of the NMR quadrupole splitting of 137Cs and 87Rb. The ratio of Rb to Cs was varied as well as the concentration of surfactant and octanol, whereas the ratio of surfactant to octanol was kept constant at 1:3. It turned out that there is no significant difference between both ions in their affinity towards carboxylate head groups. © 2014 Elsevier B.V. All rights reserved
Specific-ion effects
Ac ce p
NMR quadrupole splitting
d
Hofmeister Series
te
Keywords:
M
ARTICLE INFO
Carboxylate
1.
Introduction
1.1. Ion specificity
The specific interaction of charged headgroups with different counter ions matters in many biological as well as industrial systems [1, 2]. An example is the transition of micellar structures to vesicles [3]. To cite a more applied case, the Krafft temperature of surfactants used in washing processes increases with increasing ion binding. These behaviours can often be explained by Collins’ concept of matching water affinities, together with a Hofmeister-like classification of both ions and charged headgroups (Fig. 1) [49]. In Hofmeister series anions and cations are listed in separate series, sorted from salting out to salting in of proteins.
Figure 1: A typical Hofmeister series.
In this concept, it is supposed that two ions with high charge densities (“hard” ions or headgroup) of opposite *Corresponding author: Tel.: +49 941 9434296 E-mail address: [email protected]
charge have a very strong electrostatic interaction which is greater than the interaction between the ion and the water shell around them resulting in the formation of a contact ion pair. Some of the hydration water between the ions is released. Regarding “soft” ions (low charge density), the interaction between water and these ions is weak and hydration water molecules between the ions can easily be released, which enables also the formation of a contact ion http://dx.doi.org/10.1016/j.colsurfa.2014.xx.xxxx 0927-7757/© 2014 Elsevier B.V. All rights reserved.
Page 1 of 8
2
S. Dengler et al. / Colloids and Surfaces A: Physiochem. Eng. Aspects xx (2014) xx-xx
The core requirements for detecting quadrupole splittings of counter ions are the existence of an electric quadrupole moment and an anisotropic environment of the ion. Hence, only ions with a spin quantum number I > 1/2 can be detected. The anisotropy causes an electric field gradient with some orientations favoured over others. The splitting of the NMR resonance is the result of the interaction between the electric field gradient and the electric quadrupole moment at the nucleus. In an isotropic solution the orientation dependent quadrupole splitting averages to zero. For an anisotropic solution the resonance splits up into 2I peaks [4, 5, 15]. The frequency between the central peak and the first neighbouring peak is defined as the quadrupole splitting Δ. The magnitude of Δ is given by equation 2 for a powder sample (one with no macroscopic alignment of the liquid crystal phase) with Si, the order parameter describing the orientation of the fraction of molecules at site i (pi), given by Si = <3cos2θDMi-1>/2. θDMi is the angle between the liquid crystal axis (the director) and the electric field gradient. 4νQi is equal to the quadrupole coupling constant. Δ is a weighted average of the different values for the i sites. The quadrupole coupling constant is not known. Hence, absolute values of Δ can`t be determined. But the relative change of Δ can be observed and delivers an insight in the binding situation of the ions [4, 6, 14, 16].
η = 1 + Ac 1 / 2 + Bc η0
Ac ce p
te
d
The coefficient can be positive or negative. Strongly hydrated ions have positive B coefficients and weakly hydrated ions have negative B coefficients. The point of change in JonesDole coefficient sign represents the ideal behaviour at which the interaction water-ion is the same as water-water interaction. In table 1 the B coefficients of some ions are given [12, 13]. Consequently, Rb+ as well as Cs+ are weakly hydrated whereas the carboxylate head group of fatty acid is strongly hydrated. So, according to Collins concept the interactions between the head group and the counterions should be very small and of comparable weakness. Admittedly, this is a very indirect conclusion, because it is not evident that viscosities of simple electrolyte solutions reflect partition coefficients as they are measured here in liquid crystalline systems. Nevertheless, we could show in a former work that the different ion bindings of lithium and sodium ions to carboxylate and sulphate headgroups in such lamellar systems reflect surprisingly well the expected trend as predicted by Collins. As in the present study, the used technique was NMR quadrupole splitting [14]. Here we check now if the expected (non-)binding of caesium and rubidium to carboxylate headgroups can also be detected by this technique and we further show how nicely the results fit with viscosity B coefficients.
Cations
B
Anions
B
Ca2+
0.285
CH3COO-
0.250
Li+
0.150
SO42-
0.208
Na+
0.086
Cl-
-0.007
Rb+
-0.030
NO3-
-0.046
Cs+
-0.045
I-
-0.068
Table 1: Jones-Doyle viscosity B coefficients
Δ=
∑pv
i Qi
Si
(2)
1.3. Ion interaction within lamellar phase
Within the investigated concentration range and in all compositions considered here, lamellar phases exist (Fig. 3) [14]. The ionic strength of the samples in water without considering the ion-binding is given in table 2.
M
(1)
us
Ion-water interactions are reflected in many properties of electrolyte solutions, for example in the viscosity of aqueous solutions. Especially, the Jones-Dole viscosity B coefficient is believed to be a good parameter to estimate ion hydration. It can be calculated from salt solution viscosity η by equation 1 with η0 the viscosity of pure water, A, an electrostatic term, being 1 for moderate salt concentrations and c the salt concentration. B is a direct measure of the strength of ionwater interactions normalized to the strength of water-water interactions.
an
Figure 2: Depending on the ion size direct ion pairs are formed.
cr
ip t
pair of two soft oppositely charged ions. By contrast, for the combination of hard and soft ions the formation of contact ion pair is not possible. The soft ion cannot penetrate the water shell of the hard ion. Hence, the general rule of this concept is “like seeks like” (Fig. 2) [10, 11].
Table 2: Ionic strength of the samples for all compositions and all concentrations.
The counter ions can be free in the bulk phase, bound between the surfactant head groups (bb-site) or on the surface of the head groups (bs-site). Depending on the binding situation of the cation, the corresponding quadrupole splitting Δ changes. For free ions the contribution to Δ is zero. Bound ions have an anisotropic environment, giving a certain Δ value. The Δ value can be positive or negative depending on the angle between the liquid crystal axis and the electric field gradient. Ions bound in the bs-site give a positive value, whereas ions bound at the bb-site give a negative quadrupole splitting value.
Figure 3: Schematic Drawing of counter ion binding at the lamellar surface. Three possible binding situations are shown: (a) Ion moves freely in the bulk phase. (b) The ion is located perpendicular to the surfactant head group in the lamellar layer. (c) The ion is bound between amphiphile head groups.
2.
Materials and Methods
2.1. Materials
1.2. NMR quadrupole splitting
Page 2 of 8
3
S. Dengler et al. / Colloids and Surfaces A: Physiochem. Eng. Aspects xx (2014) xx-xx
Dodecanoic acid (COOH, 99.6%) was received from Alfa Aesar (USA). Caesium hydroxide (CsOH, 50 wt% in H2O, 99.9%), rubidium hydroxide (RbOH, 50 wt% in H2O, 99.9%) and octanol were obtained from Sigma Aldrich (Germany). Deuterated water was purchased from Deutero (Germany). 2.2. Methods
Results
ip t
cr us
87
Figure 5: Rb-NMR spectrum of a 85 wt% sample with a composition of CsDC/RbDC = 0:1 at 300 K. The relative intensity of the signals is plotted against the frequency in [Hz].
3.2. Influence of surfactant concentration on the quadrupole splitting
The variations of the caesium and rubidium splitting for the single ion systems, with increasing surfactant and octanol concentration are shown in figure 6 and figure 7. The caesium splitting decreases until a concentration of 75 wt%, whereas the rubidium splitting decreases initially and is roughly constant above a concentration of 55 wt% and decreases again for the highest concentration. This nonuniform behaviour might be due to the different binding sites at the lamellar interface as can be seen above in figure 3. However, the decrease of rubidium splitting at 85 wt% surfactant after increase at 75 wt% is unlikely and can not be explained by different binding sites. It might result from only partially dissolved Rb or from the removal of inner-sphere hydration.
M
3.
Figure 4: 137Cs-NMR spectrum of a 85 wt% sample with a composition of CsDC/RbDC = 1:0 at 300 K. The relative intensity of the signals is plotted against the frequency in [Hz].
an
137 Cs- and 87Rb-NMR were performed using a Bruker Avance 400 spectrometer operating at 52.48 MHz in the case of caesium and operating at 130.93 MHz in the case of rubidium. For 137Cs resonance a 30° pulse length of 10.90 μs and for 87Rb resonance a 30° pulse length of 7.00 μs was used. The spectra were collected with a number of scans of 1024. The measurements were performed at 300 K. Each sample (3g) was prepared in a tube and mixed with a magnetic stirrer until the sample was homogeneous. Then about 700 mg was transferred in a NMR tube (5 mm o.d.). Each sample contains D2O, octanol, surfactant and H2O from the alkali metal hydroxid. The surfactants are either caesium dodecanoate (CsDC) or rubidium dodecanoate (RbDC). The surfactants were prepared in situ from dodecanoic acid and the corresponding alkali metal hydroxide. The molar ratio of CsDC and RbDC varies from 0 to 100 % caesium surfactant in 20 % steps. In each sample the molar ratio of surfactant to octanol was 1:3. The concentrations of surfactant + octanol in D2O are 45 wt% to 85 wt% in 10 wt% steps. In order to prove the formation of a lamellar phase, the samples were observed visually via crossed polar filters. All samples appeared as single phases with typical lamellar textures.
3.1. Phase structure and NMR spectra
Ac ce p
te
d
A Cs splitting was observed for all samples, whereas we had problems to detect the rubidium splitting in the case of 85 wt% samples. The splitting of some compositions of the 85 wt% samples was too broad to be detected within the chosen frequency range. In figure 4 and 5 typical 137Cs- and 87Rb-NMR spectra can be seen. The number of peaks is given by 2I. Accordingly, in the 137Cs-NMR spectrum (Cs: I = 7/2) seven peaks occur and in the 87Rb-NMR spectrum (Rb: I =3/2) three peaks can be observed. Due to the small quadrupole moment (Q = -4 * 10-31 m2) of Cs, the signals are narrow [17]. By contrast, the quadrupole moment of Rb is Q = 0.14 * 10-28m2, resulting in broader signals [18]. The quadrupole splitting (Δ) is defined as half the distance between the two first neighbouring peaks to the central peak. As can be seen in the figures below, the caesium splitting is clearly smaller than the rubidium splitting. As it is well known for liquid crystalline samples, asymmetric central peaks occur for both ions. This effect is known as chemical shift anisotropy [17]. Due to the lower splitting of caesium, the anisotropy is more pronounced in the case of 137Cs-NMR. Note that there are three little peaks or shoulders at ca.965 Hz, 403 Hz and -498 Hz. This is a hint at a possible different lamellar structure in equilibrium with the main one. Its appearance is not a consequence of an insufficient homogenisation of the samples, because we took care to thoroughly mix them. Further, samples were prepared twice and independently to check this point. These small extra peaks occur for all samples.
Figure 6: Caesium splitting with increasing concentration of surfactant and octanol at 300 K.
Figure 7: Rubidium splitting with increasing concentration of surfactant and octanol at 300 K.
Page 3 of 8
S. Dengler et al. / Colloids and Surfaces A: Physiochem. Eng. Aspects xx (2014) xx-xx
our results. The single ion splitting of both cations decreases with increasing concentration of surfactant and octanol up to a certain concentration and increases for higher concentrations. For rubidium splitting an unlikely decrease after increase is observed. Both ions prefer the bb-binding site (negative Δ values) resulting in a decreasing ion splitting. For a very high surfactant concentration the surfactant head groups get very close. Consequently ions are moved to the surfactant head group surface (bs-site; positive Δ values) resulting in an increased splitting. Hence we can speculate that the kink results from the closer packing of the surfactants. 4.2. Ion specifity of carboxylate
The caesium and rubidium splitting remain nearly constant for all concentrations and fractions, except of 85 wt%. This is due to the low affinity of carboxylate to caesium and rubidium. The kink observed for all concentrations except 85 wt% samples points to a special lattice structure of bound ions at 1:1 ratio Cs/Rb. But ions are only loosely bound. For 85 wt% the course of splitting is different. The caesium splitting decreases and the rubidium splitting shows a discontinuous trend. Probably rubidium surfactant was not completely dissolved for this concentration. Nevertheless, visually these samples appear as homogenous as all other samples. In summary, the affinity of rubidium and caesium to carboxylate is very weak in both cases and there is no detectable ion specificity, unlike the ion binding of lithium and sodium studied in a previous paper [14], where preferential binding of lithium was observed. But sodium and lithium having also positive B coefficients like carboxylate are supposed to interact much more strongly with carboxylate. Accordingly, a precondition for the determination of ion specificities by NMR seems to be a sufficient interaction between anion and cation.
M
an
us
In figures 8 and 9 the caesium and rubidium splitting of the 45 wt% samples with an increasing fraction of caesium and rubidium respectively is shown. Caesium splitting remains nearly constant for all composition and rubidium shows only a slight decrease. Hence carboxylate does not show any ion specificity towards caesium or rubidium for this composition. Similarly, for the 55, 65 and 75 wt% samples (Fig. 8 and 9), the Δ values of caesium remain constant. The rubidium splitting is nearly constant for 55 and 75 wt% and shows only small increase for 65 wt%. Consequently, for these compositions carboxylate also does not show any ion specificity towards caesium and rubidium. Except of 85 wt% samples a kink at ratio 1:1 Cs/Rb can be observed for both splittings. This non-monotonic behaviour might point to a special lattice structure for this composition. Finally, for the 85 wt% samples (Figs. 8 and 9) the Δ values of caesium decreases with increasing amount of caesium. In contrast, the rubidium splitting increases up to an amount of rubidium of 65 wt% with a slight decline for 55 wt% rubidium and decreases significantly for higher amount of Rb.
ip t
3.3. Ion specifity of carboxylate
cr
4
Acknowledgments
Ac ce p
te
d
Figure 8: Caesium splitting with increasing amount of caesium for: () 45 wt%, (z) 55 wt%. (S) 65 wt%, (T) 75 wt% and () 85 wt% CsDC/RbDC/octanol in D2O
Figure 9: Rubidium splitting with increasing amount of rubidium for: () 45 wt%, (z) 55 wt%. (S) 65 wt%, (T) 75 wt% and () 85 wt% CsDC/RbDC/octanol in D2O.
4.
Discussion
4.1. Influence of surfactant concentration on the quadrupole splitting
The concept of the Hofmeister series predicts a higher affinity of rubidium to carboxylate in comparison to caesium [8], but the series does not give any information, how string this difference might be. Jones-Dole viscosity B coefficients suggest only a very small interaction of both cations with the carboxylate head group. This is confirmed by
We thank Fritz Kastner, Annette Schramm and Georgine Stühler, the staff of the NMR department of the University of Regensburg for performing the NMR experiments.
References
[1] P. Koelsch, P. Viswanath, H. Motschmann, V.L. Shapovalov, G. Brezesinski, H. Moehwald, D. Horinek, R.R. Netz, K. Giewekemeyer, T. Salditt, H. Schollmeyer, R. von Klitzing, J. Daillant, P. Guenoun, Specific ion effects in physicochemical and biological systems: Simulations, theory and experiments, Colloids and Surfaces A, 303 (2007) 110-136. [2] E.R.A. Lima, E.C. Boström, J. Biscaia, F.W. Tavares, W. Kunz, Ion Specific Forces between Charged SelfAssembled Monolayers Explained by Modified DLVO Theory, Colloids and Surfaces A, 346 (2009) 11-15. [3] N. Vlachy, M. Drechsler, J.-M. Verbavatz, D. Touraud, W. Kunz, Role
Page 4 of 8
S. Dengler et al. / Colloids and Surfaces A: Physiochem. Eng. Aspects xx (2014) xx-xx
ip t
[12] K.D. Collins, Ions from the Hofmeister series and osmolytes: effects on proteins in solution and in the crystallization process. , Methods, 34 (2004) 300-311. [13] J.B. Robinson, J.M. Strottmann, E. Stellwagen, Prediction of neutral salt elution profiles for affinity chromatography Proceedings of the National Academy of Sciences, 78 (1981) 2287-2291. [14] S. Dengler, A. Klaus, G.J.T. Tiddy, W. Kunz, How specific are ion specifities? A pilot NMR study., Faraday Discussions, 160 (2012) 1-13.
M
[5] I.D. Leigh, M.P. McDonald, R.M. Wood, G.J.T. Tiddy, M.H. Trevethan, Structure of Liquid-crystalline Phases formed by Sodium Dodecyl Sulphate and Water as determined by Optical Microscopy, X-ray Diffraction and Nuclear Magnetic Resonance Spectroscopy, Journal of the Chemical Society, Faraday Transaction 1: Physical Chemistry in Condensed Phases, 77 (1981) 2867.
[11] P. Jungwirth, D.J. Tobias, Specific ion effects at the air/water interface., Chemical Reviews, 106 (2006) 12591281.
cr
[4] G. Lindblom, B. Lindman, G.J.T. Tiddy, Ion binding studied using quadrupole splittings of sodium-23(1+) ions in lyotropic liquid rystals. The dependence on surfactant type., Journal of American Chemical Society, 100 (1978) 2299.
Surfactants, and Deep Eutectic Solvents, Doctoral Thesis, University of Regensburg, Regensburg, 2013.
us
of the surfactant headgroup on the counterion specificity in the micelle-tovesicle transition through salt addition, Journal of Colloid and Interface Science, 319 (2008) 542-548.
an
5
Ac ce p
te
d
[6] N. Schwierz, D. Horinek, R.R. Netz, Reversed Anionic Hofmeister Series: The Interplay of Surface Charge and Surface Polarity., Langmuir, 26 (2010) 7370. [7] W. Kunz, Specific Ion Effects, 2010, World Scientific Publishing, Singapore.
[8] W. Kunz, Specific ion effects in colloidal and biological systems, Current Opinion in Colloid & Interface Science, 15 (2010) 34-39. [9] N. Vlachy, B. Jagoda-Cwiklik, R. Vacha, D. Touraud, P. Jungwirth, W. Kunz, Hofmeister series and specific interactions of charged headgroups with aqueous ions., Advances in Colloid and Interface Science, 146 (2009) 42. [10] D. Rengstl, Choline as a Cation for the Design of Low-toxic and Biocompatible Ionic Liquids,
[15] H. Wennerstrom, G. Lindblom, B. Lindman, Theoretical aspects on the NMR of quadrupolar ionic nuclei in micellar solutions and amphiphilic liquid crystals., Chemica Scripta, 6 (1974) 97. [16] P. Jungwirth, D.J. Tobias, Specific Ion Effects at the Air/Water Interface, Chemical Reviews, 106 (2006) 1259. [17] H. Wennerstrom, N.O. Persson, G. Lindblom, B. Lindman, Chemical Shift Anisotropies of 133Cs+ Counterions in Lyotropic Liquid Crystals., Journal of Magnetic Resonance, 30 (1978) 133. [18] U. Meyer-Berkhout, Betimmung der elektrischen Quadrupolmomente der Kerne Rb85 und Rb87 durch Messung der Hochfrequenzübergänge im angeregten 6P3/2-Term des Rb-Atoms.,
Page 5 of 8
6
S. Dengler et al. / Colloids and Surfaces A: Physiochem. Eng. Aspects xx (2014) xx-xx
Ac ce p
te
d
M
an
us
cr
ip t
Zeitschrift für Physik, 141 (1955) 185187.
Page 6 of 8
7
S. Dengler et al. / Colloids and Surfaces A: Physiochem. Eng. Aspects xx (2014) xx-xx
Highlights Rubidium and caesium cations prefer bb-binding site.
ip t
Affinity of rubidium and caesium to carboxylate is very weak and there is no
cr
detectable ion specificity. A sufficient interaction between anion
us
and cation is a precondition for the
an
determination of ion specificities by
Ac ce p
te
d
M
NMR
Page 7 of 8
Ac
ce
pt
ed
M
an
us
cr
i
*Graphical Abstract (for review)
Page 8 of 8