Metal Ion Coordination, Conditional Stability Constants, and Solution

Apr 4, 2014 - ... *E-mail: [email protected]., *E-mail: [email protected]. ... Molecular Organization of an Adsorbed Layer: A Zwitterionic, ...
0 downloads 0 Views 727KB Size
Article pubs.acs.org/Langmuir

Metal Ion Coordination, Conditional Stability Constants, and Solution Behavior of Chelating Surfactant Metal Complexes Ida Svanedal,*,† Susanne Boija,† Ann Almesåker,† Gerd Persson,† Fredrik Andersson,† Erik Hedenström,† Dan Bylund,‡ Magnus Norgren,*,† and Håkan Edlund*,† †

Fibre Science and Communication Network (FSCN), Mid Sweden University, SE-851 70 Sundsvall, Sweden Department of Natural Sciences, Mid Sweden University, SE-851 70 Sundsvall, Sweden



S Supporting Information *

ABSTRACT: Coordination complexes of some divalent metal ions with the DTPA (diethylenetriaminepentaacetic acid)based chelating surfactant 2-dodecyldiethylenetriaminepentaacetic acid (4-C12-DTPA) have been examined in terms of chelation and solution behavior. The headgroup of 4-C12DTPA contains eight donor atoms that can participate in the coordination of a metal ion. Conditional stability constants for five transition metal complexes with 4-C12-DTPA were determined by competition measurements between 4-C12DTPA and DTPA, using electrospray ionization mass spectrometry (ESI-MS). Small differences in the relative strength between the coordination complexes of DTPA and 4-C12-DTPA indicated that the hydrocarbon tail only affected the chelating ability of the headgroup to a limited extent. The coordination of Cu2+ ions was investigated in particular, using UV−visible spectroscopy. By constructing Job’s plots, it was found that 4-C12-DTPA could coordinate up to two Cu2+ ions. Surface tension measurements and NMR diffusometry showed that the coordination of metal ions affected the solution behavior of 4-C12-DTPA, but there were no specific trends between the studied divalent metal complexes. Generally, the effects of the metal ion coordination could be linked to the neutralization of the headgroup charge of 4-C12-DTPA, and the resulting reduced electrostatic repulsions between adjacent surfactants in micelles and monolayers. The pH vs concentration plots, on the other hand, showed a distinct difference between 4-C12-DTPA complexes of the alkaline earth metals and the transition metals. This was explained by the difference in coordination between the two groups of metal ions, as predicted by the hard and soft acid and base (HSAB) theory.

1. INTRODUCTION Chelating agents are polydentate ligands capable of forming strong coordination complexes with divalent and trivalent metal ions. When the chelating ability is paired with surface activity, a chelating surfactant is created. In the past few decades, chelating surfactants and metallosurfactants have gained quite a lot of attention, both from a scientific and an industrial perspective.1−4 Practical applications are found within magnetic resonance imaging,5−7 templating of mesoporous materials,8 and catalysis of carboxylic and phosphate ester hydrolysis.9 The amphiphilic character offers the possibility of separation, and there have been some studies regarding the use of chelating surfactants to remove metal ions from aqueous solutions by foam flotation,10,11 or micellar-enhanced ultrafiltration.12 The chelating surfactant 4-C12-DTPA, described in this paper, was designed with the purpose to form strong coordination complexes with divalent metal ions, and to be separable from water. We have shown in an earlier study that 4C12-DTPA has promising applications in paper pulp bleaching.13 Moreover, 4-C12-DTPA has proved to be efficient for recovery of Cu2+ and Mn2+ ions by foam flotation in the presence of a foaming cosurfactant.11,14,15 The headgroup of 4© 2014 American Chemical Society

C12-DTPA contains eight potential donor atoms, in the form of three amine nitrogen atoms and five carboxylate oxygen atoms, and is thus an amphoteric surfactant. The negatively charged donor atoms bind to monovalent cations such as Na+ via ionic bonds, just like conventional anionic surfactants do. However, when it comes to interactions with divalent or trivalent metal ions, the headgroup of the chelating surfactant differs from conventional surfactants due to the possibility of forming coordination complexes of high stability, where all eight donor atoms may contribute to the coordination of the metal ion. Divalent counterions are therefore very tightly bound to the chelating surfactant, and not exchanged regularly as counterions normally are. Properties of the coordinated ion, such as ionic radius and crystal field stabilization energy, affect the strength of the complex and the strength varies therefore substantially even between ions of the same valence. In two recent papers, we presented the results regarding the solution behavior of the chelating surfactant 4-C12-DTPA,16 as Received: January 20, 2014 Revised: March 12, 2014 Published: April 4, 2014 4605

dx.doi.org/10.1021/la5002336 | Langmuir 2014, 30, 4605−4612

Langmuir

Article

where [DTPA]tot and [4-C12-DTPA]tot are the total concentrations of the chelating agents, and the KM‑DTPA values are known from the literature.19 The assumptions thereby made are that no significant side reactions occur for the chelating agents and that any potential ESI suppression effects due to the extra counterions in the sample solutions will affect both chelating agents equally. 2.4. Cu2+ Coordination. A Varian CARY 100 Bio UV−visible spectrophotometer was used to measure UV−visible spectra. The ratio of Cu2+ ions coordinated by the chelating surfactant was determined using Job’s plots and the continuous variation method.20 A range of aqueous solutions with varying concentration of copper chloride and the chelating surfactant, with a total concentration of 10 mM, were prepared, and pH was adjusted (using a combined glass pH electrode) with dilute solutions of hydrochloric acid and tetra-tert-butylammonium hydroxide, before measuring the UV−visible spectra and constructing the Job’s plots. 2.5. Sample Preparation for pH, Surface Tension, and NMR Diffusometry Measurements and Phase Behavior at Higher Concentrations. 1:1 metal complex solutions were prepared by adding appropriate amounts of the respective metal chloride to 400 mM 4-C12-DTPA solutions. The pH was adjusted with sodium hydroxide to approximately 5 or 7, and then the solutions were diluted to 200 mM to obtain the stock solutions. The Ni2+−4-C12-DTPA complex at pH 5 was prepared at 100 mM due to the formation of a highly viscous liquid at higher concentrations. The labeling of solutions as pH 5 or pH 7 refers to the pH of the adjusted stock solutions. Samples for pH measurements (using a Mettler Inlab micro pH electrode), surface tension, and NMR diffusometry were prepared by diluting the stock solutions of the metal complexes with pH adjusted Milli-Q water. Portions of the stock solutions of the metal complexes were freezedried, and the resulting solids were used without further purification for the investigation of the phase behavior of the metal complexes at higher concentrations. Samples were prepared by mixing appropriate amounts of the specific freeze-dried metal complex with Milli-Q water in 8 mm test tubes, which then were flame-sealed and equilibrated at room temperature. The samples were inspected visually using crossed polarizers. 2.6. Surface Tension Measurements. The surface tension was measured with a Krüss K6 tensiometer and a platinum du Noüy ring at a temperature of approximately 22 °C. The surface tension was measured during dilution of the stock solutions. Each sample was measured once right after sample preparation, and then the sample was left for 5 min before it was measured four times consecutively. Mean values are reported. C20 and γcmc are determined from surface tension vs concentration plots. For the time dependence study of surface tension, each sample was first measured right after pouring of the sample into the measuring cup, after which the samples were covered and left for 24 h before they were measured again. 2.7. NMR Diffusometry. The NMR diffusometry measurements were carried out using the stimulated echo pulsed field gradient (STEPFG) NMR sequence. This technique is well established, and a detailed description can be found elsewhere.21−23 The experiments were performed on a Bruker Avance DPX 250 MHz NMR spectrometer equipped with a Bruker self-diffusion probe capable of providing magnetic field strengths up to 1200 G/cm. The temperature was kept at 22 °C unless otherwise stated. The attenuation of the echo signal was fitted using eq 2 to extract the self-diffusion coefficient D:

well as the interactions in micellar systems of 4-C12-DTPA and ionic surfactants.17 In the present paper we extend our investigations to include metal complexes formed by this surfactant. To begin with, conditional stability constants are determined for five transition metal complexes with 4-C12DTPA and the coordination of Cu2+ ions is investigated in particular. These results are compared to the conventional chelating agent DTPA. Furthermore, the time dependence of monomer adsorption at the air−water interface is described. In the last section, we discuss the solution behavior of the metal complexes in terms of cmc and surface tension reduction. We also examine the pH vs concentration plots for the metal complexes in correlation to the formation of micelles.

2. EXPERIMENTAL SECTION 2.1. Materials. The chelating surfactant 2-dodecyldiethylenetriaminepentaacetic acid (4-C12-DTPA) was delivered by Syntagon AB. The synthesis and analyses have been reported earlier.16,18 Water for preparation of samples was of Milli-Q grade. Diethylenetriaminepentaacetic acid (DTPA), manganese chloride, magnesium chloride, cobalt chloride, copper chloride, calcium chloride, zinc chloride, nickel chloride, ammonium hydroxide, triethylamine, acetic acid, methanol, hydrochloric acid, tetra-tert-butylammonium hydroxide, and sodium hydroxide were of analytical grade and used without further purification. 2.2. Sample Preparation for ESI-MS Measurements. 10 mM stock solutions with 4-C12-DTPA and DTPA, respectively, were prepared by dissolving weighted substance material in small amounts of diluted ammonium hydroxide and further diluted with Milli-Q water. For each metal ion, 10 mM stock solutions were prepared, where all metal salts were dissolved and diluted in Milli-Q water except for zinc chloride, which was dissolved and diluted in 5 mM triethylamine acetic acid at pH 5. The exact metal ion concentration in each stock solution was determined with inductively coupled plasma mass spectrometry (Agilent 7700 series ICP-MS). Calibration solutions were prepared with a 1:1 molar relationship between 4C12-DTPA and DTPA at concentrations of 1, 5, and 10 μm. Sample solutions with each metal ion were prepared at a concentration relationship of 8:8:6 μm for 4-C12-DTPA:DTPA:metal ion. Calibration and sample solutions were diluted with 20 vol % methanol/80 vol % 5 mM triethylamine acetic acid that had been pH-adjusted to either pH 5 or 7, using a combined glass pH electrode. 2.3. Conditional Stability Constants. Determinations of conditional stability constants for the different 4-C12-DTPA metal complexes were performed by electrospray ionization mass spectrometry (ESI-MS) measurements on an API 3000 instrument (AB SCIEX) with the software Analyst 1.6.1. Instrument parameters were optimized in negative ion mode at the ionization source voltage −2.9 kV. Using an Agilent 1100 autosampler, 25 μL of sample was injected into a 100 μL/min flow of mobile phase (20 vol % methanol/80 vol % water) delivered with a Shimadzu LC-10AD pump. Signals for DTPA and 4C12-DTPA were recorded in Q1Multiple ion mode at m/z 392 and m/ z 560, respectively, with a dwell time of 500 ms for each ion, and a total acquisition time of 7 min. The system was washed between sample injections with repeated injections of 50 vol % methanol and 50 vol % water. The integrated peak areas for the calibration solutions were used to construct linear calibration curves (R2 > 0.991), which was then used to determine the free, uncomplexed concentrations of chelating agents ([DTPA]u and [4-C12-DTPA]u) in the sample solutions. Conditional stability constants for 4-C12-DTPA were finally determined through the following expression:

2

(2)

where I denotes the observed echo intensity, I0 is the echo intensity in the absence of field gradient pulses, γ is the magnetogyric ratio, G is the field gradient strength, δ is the duration of the gradient pulse, and Δ is the time between the leading edges of the gradient pulses. In all experiments δ was set to 1.5 ms and Δ to 40 ms while the gradient G was varied linearly. For the water self-diffusion measurements we used a repetition time of 10 000 ms, gradient starting value of 5 G/cm, and end value 75 G/cm, and 21 data points were measured, while for the surfactant measurements we used a repetition time of 3000 ms,

⎛ log KM ‐ 4 ‐ C12 ‐ DTPA = log⎜KM‐DTPA ⎝ ×

4

I = I0 e−D(2πγGδ) [Δ− (δ /3)] × 10

[DTPA]u ([4‐C12‐DTPA]tot − [4‐C12‐DTPA]u ) ⎞ ⎟ [4‐C12‐DTPA]u ([DTPA]tot − [DTPA]u ) ⎠

(1) 4606

dx.doi.org/10.1021/la5002336 | Langmuir 2014, 30, 4605−4612

Langmuir

Article

gradient starting value of 75 G/cm to suppress the water signal as much as possible without compromising the surfactant signals, end value adjusted to the expected self-diffusion coefficient for the surfactant, and at least 31 data points. The cmc’s are determined from the self-diffusion coefficient vs concentration plots.

Table 1. Configurations of Hydrated Metal Ions, Ionic Radii, and Conditional Stability Constants log K

Me2+

3. RESULTS AND DISCUSSION 3.1. Conditional Stability Constants and Coordination Behavior. The molecular structure of the chelating surfactant 4-C12-DTPA, with eight possible donor atoms that can coordinate metal ions, is shown in Figure 1. Conditional

2+

Figure 1. Molecular structure of 4-C12-DTPA at high pH, with eight potential donor atoms, i.e., the five carboxylate oxygen atoms and the three nitrogen atoms.

a

stability constants for five divalent metal complexes with 4-C12DTPA were determined by competition measurements where 4-C12-DTPA and DTPA competed for the metal ion. The molar ratios of uncomplexed chelating agents were determined using ESI-MS. Competition measurements utilizing ESI-MS have been reported before, for crown ethers and metal ions.24 Full spectra of sample solutions, as partly presented in Figure 2 with the chelating agents and Mn2+, shows a well resolved

configurationa

ionic radiusb (Å)

Mg

octahedron

0.86

Ca2+

square antiprism

1.14

Mn2+

octahedron

0.97

Zn2+

octahedron

0.88

Co2+

octahedron

0.88

Ni2+

octahedron

0.83

Cu2+

octahedron (JT)

0.87

pH 5 7 5 7 5 7 5 7 5 7 5 7 5 7

[DTPA]u/[4C12-DTPA]uc e

n/a n/ae n/ae n/ae 0.89 0.76 1.05 0.86 1.09 0.82 0.83 0.70 0.84 0.74

4-C12DTPA e

n/a n/ae n/ae n/ae 6.2 10.0 9.9 13.0 10.0 13.5 11.4 14.5 11.9 15.5

DTPAd 1.9 4.3 3 5.7 6.3 10.2 9.8 13.1 9.9 13.7 11.6 14.9 12.1 15.8

From ref 25. bFrom ref 26. c±0.02. dFrom ref 19. eNot determined.

pH. This is not surprising since the hydrocarbon tail supports neighboring donor atoms with electrons, leading to increased pKa values, i.e., reduced acidity, of those donor atoms. At pH 7, all ratios are close to each other, ranging from 0.70 to 0.86. At pH 5, the relative amounts of free chelating agents were more spread, from 0.83 to 1.09. This increased difference between the ratios indicates that the strength of the complexes is not solely dependent on the pKa values of the chelating agents. It might also be affected by small differences in 3-D structure of the respective chelating agent, probably affecting the property more when the chelating agents have the same degree of protonation. Despite small difference in strength, at the investigated pHs the conditional stability constants presented in Table 1 show the same selectivity order for both chelating agents: Mn2+ < Zn2+ < Co2+ < Ni2+ < Cu2+. Even though the differences in conditional stability constants between the chelating agents are very small, they can be determined in high precision by the analytical method applied. The concept of hard and soft acids and bases (HSAB), which implies that hard bases preferably coordinate hard acids and soft bases prefer soft acids, is useful when discussing coordination of metal ions.27 The relative strength between the coordination of different metal ions with a specific chelating agent generally correlates to the classification of the metal ions on the scale from hard to soft acids. A typical hard acid is a small metal ion with high charge density and low polarizability, whereas soft acids are larger, have lower charge density, and are more polarizable. Among the examined metal ions, the divalent alkaline earth metals are classified as harder acids than the divalent transition metals. Donor atoms are described as hard or soft bases in a similar manner. When it comes to the two different donor atoms in the headgroup of 4-C12-DTPA, the negatively charged carboxylate oxygen atoms are harder bases than the neutral nitrogen atoms. Generally, smaller ions are coordinated more strongly due to stronger electrostatic interactions. For the transition metal ions, there are additional contributions from the crystal field stabilization energy, as well as the Jahn−Teller distortion in the case of Cu2+, that increases the stability.25 The examined divalent alkaline earth metals, being hard acids, are consequently coordinated more strongly

Figure 2. ESI-MS full spectrum in negative ion mode of a sample solution containing DTPA, 4-C12-DTPA, and Mn2+. Peak for uncomplexed DTPA is at m/z 392 and at m/z 560 for uncomplexed 4-C12-DTPA, and the peaks at m/z 445 and m/z 613 come from Mn2+ complexes with DTPA and 4-C12-DTPA, respectively.

spectrum and strong signals for DTPA and 4-C12-DTPA, at m/ z 392 and m/z 560, respectively. These peaks are thus easily integrated at the concentration levels applied (1−10 μM). Also, Mn2+ coordinated by the chelating agents show peaks at m/z 445 and m/z 613, respectively. The other small peaks within the interval shown are due to small amounts of Na+, which are also present in the calibration solutions and should thus only have negligible effects on the result interpretations. The determined ratios of uncomplexed chelating agents in sample solutions are shown in Table 1. A value above 1 of the [DTPA]u/[4-C12-DTPA]u ratio, which is noted for Zn2+ and Co2+ complexes at pH 5, indicates that 4-C12-DTPA forms stronger complexes than DTPA, while the opposite is true for the values below 1. For all metal ion-chelating agent mixtures studied, the relative amount of uncomplexed DTPA is consistently lower at pH 7, indicating that DTPA is the stronger chelating agent for all transition metals explored at this 4607

dx.doi.org/10.1021/la5002336 | Langmuir 2014, 30, 4605−4612

Langmuir

Article

hydrocarbon tail of 4-C12-DTPA has little effect on the coordination behavior of the chelating headgroup. It was assumed that other metal ions would coordinate to 4-C12DTPA in similar ways as to DTPA, and this is assumed throughout the rest of the paper. 3.2. Determination of Cmc and Time Dependence of Surface Tension for Metal Complexes. The chelating surfactant has a complex solution behavior due to the numerous donor atoms (see Figure 1) with overlapping pKa values. For each functional group there is an equilibrium reaction between the protonated and the dissociated form, and many of these equilibria are overlapping, which has been thoroughly discussed elsewhere.16,17 At any intermediate pH, there will be a distribution of differently charged species, and the complexity depends on the number of overlapping pKa values at the specific pH. As micelles start to form, the distribution may change since the surfactant can accept or donate protons to the aqueous solution in order to screen the electrostatic repulsion between adjacent headgroups in micelles. At the examined pH levels (pH 5 and 7) the chelating surfactant is zwitterionic with a negative net charge, and surfactants in micelles will consequently become protonated to a higher degree than the free monomers to screen the repulsion between surfactants in micelles. In a recent study of the pure 4-C12-DTPA, we reported that the determination of the critical micelle concentration, cmc, from the surface tension plot could not be done in a conventional way.16 This was because the point usually interpreted as cmc did not match the cmc determined by NMR diffusometry, and the interpretation of data from NMR diffusometry was considered unambiguous. When increasing the concentration, we found that the surface tension declined to a point that was interpreted as surface saturation. This was followed by a region of constant surface tension up to the cmc, where the surface tension began to increase again, see Figure 3a.

by the carboxylate groups than by the amine groups. The divalent transition metal ions, on the other hand, are coordinated strongly by both types of donor atoms. This is consistent with the higher log K values of the transition metal ions seen in Table 1. Mn2+ is sometimes classified as hard and sometimes as intermediate, and its log K is indeed between those of the alkali earth metal ions and those of the other examined transition metal ions. To gain insight into the coordination behavior of the chelating headgroup, Cu2+ complexes with 4-C12-DTPA were studied in solution by UV−visible spectroscopy. Absorptions due to d−d transitions of Cu2+ complexes can be seen in the visible region and give some hints about the coordination environment.28 The coordination environment of Cu2+ ions with the headgroups of 4-C12-DTPA was investigated at pH 5, 6, 8, and 10, see Table 2. The coordination behavior of 4-C12Table 2. Absorption Maxima, λmax, and Extinction Coefficients, ε, of Cu2+ Complexes of DTPA and 4-C12DTPA at Different pHs λmax [nm] (ε [M−1 cm−1])

DTPA DTPA 4-C12DTPA 4-C12DTPA a

ratio to Cu2+

pH 5

pH 6

pH 8

pH 10

1:1 1:2 1:1

684 (64) 740 700 (66)

666 (76) 740 (118) 664 (85)

660 (81) 740 (117) 660 (91)

n/aa n/aa 660 (92)

1:2

725

721 (146)

721 (146)

n/ab

Not measured. bNot measured due to precipitation.

DTPA was believed to be similar to that of commonly used DTPA. Hence, the UV−visible spectra of Cu2+ complexes of 4C12-DTPA were compared to those of DTPA. DTPA is known to be able to coordinate one or two Cu2+ ions.29 DTPA first coordinates one Cu2+ ion, i.e., Cu2+−DTPA is formed, with an absorption maximum at 685, 666, and 660 nm at pH 5, 6, and 8, respectively. A change in the UV−visible absorption to longer wavelengths is seen for solutions containing more than one equivalent of Cu2+ ions. The (Cu2+)2−DTPA species had spectra showing a maximum at 740 nm at all the investigated pHs, Table 2. As expected, similar trends were seen for the Cu2+ complexes formed with 4-C12-DTPA, which had a maximum at 700, 664, 660, and 660 nm at pH 5, 6, 8, and 10, respectively for Cu2+−4-C12-DTPA, whereas the (Cu2+)2− 4-C12-DTPA had a maximum at 725 at pH 5 and at 721 at pH 6 and 8. A precipitate was formed in the 2:1 solution at pH 10, likely indicating the release of some Cu2+ ions and formation of Cu2+ hydroxides. The coordination ratios were also examined by Job’s plots, shown in the Supporting Information. From these results, it was clearly seen that a 1:1 complex is formed for solutions with less and exactly one equivalent of Cu2+ ions present and that 2:1 complexes were formed for both DTPA and 4-C12-DTPA. No more than two Cu2+ ions could be coordinated to the chelating agents in the studied solutions. This was also shown for the DTPA complexes in 1957 by Chaberek et al.29 It is worth noticing here that ethylenediaminetetraacetic acid (EDTA), which is one of the most commonly used chelating agents, is known to only coordinate one metal ion.30 The similar absorption spectra of the Cu2+ complexes of DTPA and 4-C12-DTPA clearly indicates that the

Figure 3. Correlations between NMR diffusometry (···■···), surface tension (···●···), and pH (···▲···) vs concentration for (a) 4-C12DTPA and (b) Ni2+−4-C12-DTPA. The determined cmc’s are highlighted by the gray zones. 4608

dx.doi.org/10.1021/la5002336 | Langmuir 2014, 30, 4605−4612

Langmuir

Article

energy of the phase boundary is however of great importance, and allowing the system to equilibrate resulted in reduced surface tension due to the increased adsorption. Both electrostatic interactions and hydrogen bonding between neighboring surfactants might reduce the headgroup area and increase the packing parameter. The surface tension plots at different states in the adsorption process are shown in Figure 4 for the Mg2+−4-C12-DTPA

The increasing surface tension was explained by the monomers favoring the micelles over the saturated surface. The most likely reason for this is the conical geometry of the chelating surfactant, i.e., its low critical packing parameter, cpp.17 The cpp is the ratio between the cross-sectional area of the hydrophobic and the hydrophilic parts. Because of its large headgroup, we can expect an extremely low cpp for the chelating surfactant, most likely significantly lower than the value of 1/3 that is often observed for ionic surfactants. This makes the chelating surfactant more favorably packed in spherical micelles than at planar surfaces. Comparisons between surface tension measurements and NMR diffusometry for the metal complexes of 4-C12-DTPA studied in the present paper show that this is the case also here; the Ni2+ complex is shown as an example in Figure 3b. The chelating surfactant contains eight donor atoms, and as seen in Table 1, all examined metal ions except Ca2+ are six-coordinating. This means that, in the metal complexes, the number of free donor atoms that can participate in protonation equilibria are considerably lower than for the pure chelating surfactant, since the majority of the donor atoms are engaged in the coordination of the metal ion. Although the number of dissociation states is markedly reduced, the anomalous surface tension behavior still persists. This further confirms our hypothesis that the main reason for the increasing surface tension at the cmc is the conical geometry of the surfactant. Although an increase in the cpp can be expected from coordination of a metal ion, from the similarities between the surface tension plots in Figure 3a (4C12-DTPA) and b (Ni2+−4-C12-DTPA) we cannot see any significant effects on the anomalous desorption of surfactants from the surface at the formation of micelles. It is however unclear to what extent we can expect the surface tension plot to change with the cpp. Another possible explanation to the shape of the surface tension plot could lie in the distribution of differently charged species, i.e., that the more hydrophobic species are leaving the surface in favor of the micelles after the cmc. Considering that the anomaly remains to the same extent despite the fact that the number of dissociation states is markedly reduced by the coordination of a metal ion, and thus the number of species, this explanation is less likely. Determinations of cmc’s are therefore based on these observations, as shown in Figure 3. The increase in pH at the micellization as a result of the higher degree of protonation in micelles compared to monomers, i.e., higher apparent pKa values in micelles, is also shown in the figure. The pH behavior of the Ni2+ complex is representative of the studied transition metal complexes, whereas complexes of the alkaline earth metal ions show decreasing pH after the formation of micelles as a result of the different coordination of these ions, as predicted by the HSAB theory. The pH behavior of the metal complexes is further discussed at the end of section 3.3. The time dependence of the 4-C12-DTPA monomer adsorption at the air−water interface16 was examined for the metal complexes. Generally, the adsorption of monomers at the surface reaches equilibrium within seconds, but in the case of 4C12-DTPA the surface tension continued to decrease for several hours and 24 h equilibration was allowed to ensure equilibrium surface tension. The slow equilibrium process of monomer adsorption was explained by the electrostatic surface potential created by the multiply charged chelating surfactants already adsorbed at the surface, which acts as a repulsive barrier that must be overcome for further adsorption. The driving force toward dense packing at the surface and lowering of the free

Figure 4. Surface tension vs concentration at different aging times of the air−water interface; t0 (−□−), subsequent dilution (−▲−), and t24h (−●−) for Mg2+−4-C12-DTPA. The determined cmc is highlighted by the gray zone.

complex, as an example. The surface tension was first measured during subsequent dilution of the stock solution. A portion of the sample was removed, and the remaining was diluted to obtain the next sample. Each removed portion was then measured twice: once directly after pouring the sample into the measuring cup (the t0 set) and then after 24 h equilibration (the t24h set). An important note here is that the surface is only partially destroyed and re-created during the subsequent dilution, in contrast to the t0 set, which is measured shortly after the surface was created. The subsequent dilution set is thus a state further into the adsorption process than the t0 set. The large difference between the surface tension of the t0 set and the subsequent dilution set is seen in Figure 4. Since the du Noüy ring method does not measure dynamic surface tension, the surface tension value at t0 is by no means equal to the initiation of the adsorption, but rather the surface tension at a state well into the adsorption process. This approach thus serves only to study slow equilibria, while fast processes would have reached equilibrium already at t0. As seen in Figure 4, the time dependence of the relaxation process is different at different concentrations. At low concentrations, and consequently low surface coverage, the effect emerges already on the short time scale from t0 to the subsequent dilution set, which in turn is rather close to the t24h set. As the surface becomes increasingly populated, when approaching the cmc, the relaxation time seems to increase and the effect becomes more pronounced on the longer time scale. At concentrations above the cmc, no significant effect is seen during the short time equilibration, but it emerges after the 24 h equilibration. Even though the overall effect of equilibration on the surface tension, in the studied time interval, is most pronounced at concentrations up to cmc, the relaxation is actually slower at concentrations above the cmc. As shown here, the procedure of establishing equilibrium surface tension plots for the investigated chelating surfactant and its metal complexes is very time-consuming. Therefore, the surface tension measurements during the subsequent dilution of the stock solutions are reported throughout the rest of the paper. 4609

dx.doi.org/10.1021/la5002336 | Langmuir 2014, 30, 4605−4612

Langmuir

Article

3.3. Effects of Metal Ion Coordination on the Solution Behavior. In the absence of metal ions, the numerous donor atoms in the headgroup of 4-C12-DTPA are titrating and the charge varies with pH, in theory from +3 at low pH to −5 at high pH (see Figure 1). At the two investigated pH levels (pH 5 and 7), the chelating surfactant is zwitterionic with a negative net charge. Furthermore, there is flexibility in the configuration of the headgroup, as well as in the allocation of protons within the headgroup, which is affected by the interactions between neighboring surfactants in adsorbed monolayers and micelles. The coordination of a metal ion markedly changes the headgroup of the chelating surfactant. The negative net charge is reduced by the coordination of the positively charged metal ions, there is less flexibility in the headgroup, and the number of dissociation states is reduced. The addition of the respective metal salt also increases the ionic strength, which may affect the results. This can however be expected to have a minor effect compared to the effect of metal ion coordination. For the six divalent metal ion complexes with 4-C12-DTPA in Table 3, the

system the intensity of the same peak decreased at higher concentrations. The presence of Ca2+ and Mg2+ ions only alters the line shape of the peaks from the protons in the headgroup, due to the coordination, leaving the main C12 methylene peak unaffected. Both the Co2+ and Ni2+ complexes show an increase in the 1H line width at higher concentrations. This is usually associated with an increase in the aggregate size due to micellar growth. However, we cannot exclude a line broadening due to relaxation caused by the cobalt and nickel ions. The solution behavior of the pure chelating surfactant has been described in a previous paper.16 The metal complexes are compared to the pure chelating surfactant to sort out the effects of the metal ion coordination on the solution behavior. When comparing the metal complexes with the pure chelating surfactant, referred to as Na+−4-C12-DTPA in Table 3, it is obvious that the chelation of metal ions affects the solution behavior. Because of the overlapping surface tension plots of the metal complexes, Cu2+−4-C12-DTPA and Ni2+−4-C12DTPA are shown as examples in Figure 5. At both investigated

Table 3. Effects of Metal Ion Coordination on pH, Cmc, γcmc, C20, and D0 for Men+−4-C12-DTPA

a

Men+−4C12-DTPA

pH1a

pH2b

Na+

n/ae

Mg2+

4.4

Ca2+

4.4

Mn2+ Co2+ Ni2+

3.0 2.3 2.1

5 7 5 7 5 7 5 5 5 7

Cu2+

2.6

5

cmcc (mM) 20 ± 3f 35 ± 10f 10 ± 2 14 ± 6 15 ± 5 14 ± 3 13 ± 3d 16 ± 4 14 ± 3 31 ± 7, 20 ± 5d 14 ± 3d

γcmcd (mN/m)

C20d (mM)

D0c (10−10 m2/s)

33.2f 35.5f 36.6 41.0 39.7 35.8 40.0 39.4 39.2 37.0

0.39f 2.3f 0.98 0.27 0.60 1.2 1.2 0.69 1.0 1.1

3.4 3.3 3.5 3.5g 3.5 3.3 n/ae 3.2 3.1 2.9

37.1

1.2

Figure 5. Surface tension vs concentration for 4-C12-DTPA at pH 5 (−●−), Cu2+−4-C12-DTPA at pH 5 (···■···), Ni2+−4-C12-DTPA at pH 7 (···Δ···), and 4-C12-DTPA at pH 7 (−▼−).

n/ae

pH levels, all metal complexes exhibit lower cmc than the pure 4-C12-DTPA. This is a natural consequence of the reduced electrostatic repulsions due to the reduced negative headgroup charge by the coordination of the positively charged metal ion. Generally, reduced charge will reduce the electrostatic repulsions at the air−water interface, which is often associated with reduced surface tension at cmc, γcmc, as a consequence of the increased adsorption of surfactants at the surface. The increased γcmc by the coordination of metal ions may therefore seem unexpected at first. This effect can however be explained when considering the pH dependence of γcmc for the pure 4C12-DTPA, where a minimum in γcmc was found at pH 5.16 This minimum indicates an optimum in attractive interactions between the headgroups at the surface, i.e., an optimal charge distribution for maximum surface tension reduction at this pH. With this in mind, it is not surprising that the γcmc increases since the interactions between surfactants at the surface in fact become less attractive by the chelation of metal ions, even though the net charge is lower. The concentration where the surface tension of the solvent is reduced by 20 mN/m, C20, is also affected by the coordination of metal ions. The effect is however different for the two investigated pH levels, when the metal complexes are compared to the pure chelating surfactant at the respective pH. In fact, the C20 of the metal complexes at both pH levels fall between the C20 of the pure 4-C12-DTPA at the two pH levels (see Figure 5). Therefore, when the metal complexes are compared to each other, there is no significant difference in C20 between the two pH levels.

b

pH of the solution after addition of the respective metal salt. pH of the adjusted stock solution prior to dilution. cDetermined by NMR diffusometry at 295 K unless otherwise stated. dDetermined by surface tension measurements. eCould not be obtained. fFrom ref 16. g Measured at 299 K.

solution phase extends up to the studied 50 wt %. The metal complexes thus have higher solubility than the pure 4-C12DTPA, which forms a liquid crystalline phase from about 30 wt %.16 The solution properties of the metal complexes were studied by tensiometry and NMR diffusometry (see Table 3). In accordance with the reasoning at the beginning of section 3.2, the surface tension plots were in good agreement with the cmc’s determined by NMR diffusometry, the only exception being the Ni2+−4-C12-DTPA complex at pH 7. At both studied pH levels, the monomer self-diffusion coefficient, D0, for the Ni2+−4-C12-DTPA complex was slightly lower than for the pure chelating surfactant. Because of the paramagnetic nature of the Ni2+ (and Co2+) ions, their presence changes the appearance of the NMR spectrum. According to Latos-Grazynski et al. the location of the proton peaks in DTPA is shifted over 100 ppm at low field due to the formation of Ni2+−DTPA complexes.31 We found that the peaks yielded by the DTPA headgroup in our surfactant disappeared when Ni2+ was added. In the Co2+− 4-C12-DTPA system the shape of the main C12 methylene peak changed with concentration, whereas in the Ni2+−4-C12-DTPA 4610

dx.doi.org/10.1021/la5002336 | Langmuir 2014, 30, 4605−4612

Langmuir

Article

4. CONCLUSIONS When compared with the conventional chelating agent DTPA, it was clear that the hydrocarbon tail of 4-C12-DTPA only affected the coordination chemistry of the headgroup to a limited extent. Competition measurements utilizing ESI-MS, where 4-C12-DTPA and DTPA competed for the metal ion, proved to be very useful for accurate determinations of conditional stability constants. By this approach, small differences in the conditional stability constants between the two chelating agents were detectable. The order of selectivity between the five examined transition metal ions was however not affected by the addition of the hydrocarbon tail to the structure of DTPA. Neither was the ability of the headgroup to coordinate up to two Cu2+ ions in the presence of excess Cu2+ ions. From our measurements, 4-C12-DTPA seems to be comparable to DTPA in terms of chelating ability of divalent transition metal ions. As expected, significant effects on the solution behavior of the chelating surfactant were observed at the addition of metal salts. The solution phase for the metal complexes extended over the studied concentration interval up to 50 wt %, unlike the pure 4-C12-DTPA that forms a liquid crystalline phase from around 30 wt %. Generally, the investigated divalent metal ions affected the solution behavior of 4-C12-DTPA similarly, and the effects could be linked to the neutralization of the negative net charge of the headgroup as a result of the coordination of the positively charged metal ion. This finding is valuable in practical applications for chelating surfactants, such as foam flotation, when prediction of the solution properties of the metal complexes is necessary. The addition of a metal ion to the chelating headgroup increased the ability of the chelating surfactant to form aggregate in solution. When the donor atoms are engaged in coordinating a metal ion, the titrating ability is reduced and there is less flexibility in the configuration and allocation of protons in the headgroup. Because of this, the strong pH dependence in cmc, γcmc, and C20 that was observed for the pure 4-C12-DTPA was markedly reduced by the coordination of a metal ion. The investigated metal complexes of 4-C12-DTPA showed the same anomalous surface tension behavior as the pure chelating surfactant. This implies that it is indeed the conical geometry of the surfactant, i.e., the low cpp, that causes the anomalous desorption of surfactants from the surface at the formation of micelles. By simply measuring the pH decrease at the addition of the respective metal salt, a difference between the coordination of the alkaline earth metals and the transition metals was observed, which was in agreement with the general HSAB theory. Coordination of transition metal ions, by both types of donor atoms, caused more release of protons than coordination of alkaline earth metal ions, which are coordinated primarily by the negatively charged carboxylate oxygen atoms. This was confirmed by the pH vs concentration plots of the metal complexes, where the change in titration behavior of the complexes at the formation of micelles was studied. For the transition metal complexes, increasing pH of the bulk solution at the formation of micelles due to increased protonation of complexes as a result of repulsion between negative charges was observed. The alkaline earth metal complexes, on the other hand, showed decreasing pH of the micellar bulk solution due to release of protons from the complex as a result of repulsion between protonated positively charged amine groups not engaged in coordination of the metal ion. Because of the

During preparation of the metal complexes, addition of the respective metal salt to the stock solution of 4-C12-DTPA at pH 6.6 caused release of protons due to the coordination of the metal ion. As shown in the second column of Table 3, the pH of the solution decreased to varying degrees depending on the added metal salt. The difference in the release of protons is in agreement with the difference in the coordination of the metal ions, as discussed in section 3.1. Coordination of the transition metals, by both types of donor atoms, caused more release of protons than coordination of the alkaline earth metals, which are coordinated primarily by the negatively charged carboxylate oxygen. The difference between the pH vs concentration plots for the metal complexes provides additional support for different coordination chemistry of the two groups of metal ions, represented by the Mg2+ (alkaline earth metals) and Co2+ (transition metals) complexes shown in Figure 6. The pH behavior of the pure 4-C12-DTPA is also shown in the figure.

Figure 6. pH vs concentration for 4-C12-DTPA (−●−), Co2+−4-C12DTPA (−Δ−), and Mg2+−4-C12-DTPA (−■−). The determined cmc’s are indicated by the arrows.

Because of its amphoteric nature, 4-C12-DTPA can accept or donate protons to the aqueous solution in order to adjust its electrical charge.17 At the investigated pH, 4-C12-DTPA has a negative net charge. In order to screen the electrostatic repulsions between negatively charged headgroups, surfactants in micelles will become protonated to a higher degree than the free monomers.16 The pH increase for the pure 4-C12-DTPA as a result of the increased protonation at the formation of micelles is seen in Figure 6. For the Co2+ complex, the pH behavior is similar as for the pure chelating surfactant. The plot indicates that there are still repulsions between negative charges at the formation of micelles, just as for the pure 4-C12-DTPA, but of smaller magnitude. The increase in pH is less pronounced due to reduced requirements to screen the electrostatic repulsions in micelles of cobalt complexes compared to micelles of the pure 4-C12-DTPA. For the Mg2+ complex, the pH is practically constant at the formation of micelles but starts to decrease at a slightly increased concentration. From this, a different coordination of Mg2+ compared to Co2+ can be assumed. The decreasing pH as a result of release of protons indicates that there are repulsions between positive charges in the micellar solution of Mg2+ complexes, i.e., repulsions between protonated amine groups. This is in line with the discussions in section 3.1, that alkaline earth metal ions are coordinated preferably by the carboxylate groups, leaving the amine groups protonated to a higher degree. 4611

dx.doi.org/10.1021/la5002336 | Langmuir 2014, 30, 4605−4612

Langmuir

Article

(12) Pramauro, E.; Bianco, A.; Barni, E.; Viscardi, G.; Hinze, W. L. Preconcentration and Removal of Iron(III) from Aqueous-Media Using Micellar-Enhanced Ultrafiltration. Colloids Surf. 1992, 63, 291− 300. (13) Högberg, I.; Zasadowski, D.; Karlsson, A.; Wikman, B.; Andersson, F.; Hedenström, E.; Edlund, H.; Norgren, M. Brightness Development of a Hydrogen Peroxide Bleached Spruce TMP. Comparisons of Pre-Treatments with DTPA and a Separable Chelating Surfactant. Nord. Pulp Pap. Res. J. 2012, 27, 50−55. (14) Zasadowski, D.; Hedenström, E.; Edlund, H.; Norgren, M. Removal of Lipophilic Extractives and Manganese Ions from Spruce TMP Waters in a Customized Flotation Cell. Bioresources 2012, 7, 2376−2392. (15) Zasadowski, D.; Hedenström, E.; Edlund, H.; Norgren, M. Use of a Voith Flotation Cell for Removal of Lipophilic Extractives and Mn Ions from Spruce Thermomechanical Pulping Process Waters. Bioresources 2012, 7, 2784−2798. (16) Svanedal, I.; Persson, G.; Norgren, M.; Edlund, H. Anomalies in Solution Behavior of an Alkyl Aminopolycarboxylic Chelating Surfactant. Langmuir 2013, 29, 13708−13716. (17) Svanedal, I.; Persson, G.; Norgren, M.; Edlund, H. Interactions in Mixed Micellar Systems of an Amphoteric Chelating Surfactant and Ionic Surfactants. Langmuir 2014, 30, 1250−1256. (18) Högberg, I.; Andersson, F.; Hedenström, E.; Norgren, M.; Edlund, H. The Interaction Parameter in Binary Surfactant Mixtures of a Chelating Surfactant and a Foaming Agent. Prog. Colloid Polym. Sci. 2011, 138, 17−20. (19) Keys to Chelation; Form No. 298-717-80; The Dow Chemical Company: Midland, MI, 1974. (20) Vosburgh, W. C.; Cooper, G. R. Complex Ions. I. The Identification of Complex Ions in Solution by Spectrophotometric Measurements. J. Am. Chem. Soc. 1941, 63, 437−442. (21) Stilbs, P. Fourier Transform Pulsed-Gradient Spin-Echo Studies of Molecular Diffusion. Prog. Nucl. Magn. Reson. Spectrosc. 1987, 19, 1−45. (22) Lindblom, G.; Oradd, G. NMR-Studies of Translational Diffusion in Lyotropic Liquid-Crystals and Lipid-Membranes. Prog. Nucl. Magn. Reson. Spectrosc. 1994, 26, 483−515. (23) Söderman, O.; Stilbs, P. NMR-Studies of Complex Surfactant Systems. Prog. Nucl. Magn. Reson. Spectrosc. 1994, 26, 445−482. (24) Kempen, E. C.; Brodbelt, J. S. A Method for the Determination of Binding Constants by Electrospray Ionization Mass Spectrometry. Anal. Chem. 2000, 72, 5411−5416. (25) Persson, I. Hydrated Metal Ions in Aqueous Solution: How Regular Are Their Structures? Pure Appl. Chem. 2010, 82, 1901−1917. (26) Wulfsberg, G. Principles of Descriptive Inorganic Chemistry; University Science Books: Sausalito, 1991. (27) Pearson, R. G. Hard and Soft Acids and Bases. J. Am. Chem. Soc. 1963, 85, 3533−3539. (28) Lever, A. B. P. Inorganic Electronic Spectroscopy, 2nd ed.; Elsevier Science: 1985. (29) Chaberek, S.; Frost, A. E.; Doran, M. A.; Bicknell, N. J. Interaction of Some Divalent Metal Ions with Diethylenetriaminepentaacetic Acid. J. Inorg. Nucl. Chem. 1959, 11, 184−196. (30) Kovacs, A.; Nemcsok, D. S.; Kocsis, T. Bonding Interactions in EDTA Complexes. J. Mol. Struct.: THEOCHEM 2010, 950, 93−97. (31) Latos-Grazynski, L.; Jezowska-Trzebiatowska, B. Nuclear Magnetic-Resonance Studies of the Configuration and Ligand Rearrangement in Complexes of Nickel(II) and Cobalt(II) with Diethylenetriaminepentaacetic Acid and Triethylenetetraaminehexaacetic Acid. J. Coord. Chem. 1980, 10, 159−169.

ability of amphiphilic chelating agents to self-associate, measurements of pH in correlation to the formation of micelles offers a unique way of studying the metal complexes.



ASSOCIATED CONTENT

* Supporting Information S

Job’s plots for the Cu2+ complexes formed with 4-C12-DTPA and DTPA. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. *E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was financed by the European Union European Regional Development Fund and the County Administrative Board of Västernorrland.



REFERENCES

(1) Scrimin, P.; Tecilla, P.; Tonellato, U.; Vendrame, T. Aggregate Structure and Ligand Location Strongly Influence Cu-2+ Binding Ability of Cationic Metallosurfactants. J. Org. Chem. 1989, 54, 5988− 5991. (2) Griffiths, P. C.; Fallis, I. A.; Willock, D. J.; Paul, A.; Barrie, C. L.; Griffiths, P. M.; Williams, G. M.; King, S. M.; Heenan, R. K.; Gorgl, R. The Structure of Metallomicelles. Chem.Eur. J. 2004, 10, 2022− 2028. (3) Bordes, R.; Holmberg, K. Physical Chemical Characteristics of Dicarboxylic Amino Acid-Based Surfactants. Colloids Surf., A 2011, 391, 32−41. (4) El-Sukkary, M. M. A.; Soliman, E. A.; Ismail, D. A.; El Rayes, S. M.; Saad, M. A. Synthesis and Properties of Some N-Acylethylenediamine Triacetic Acid Chelating Surfactants. Tenside, Surfactants, Deterg. 2011, 48, 82−86. (5) Andre, J. P.; Toth, E.; Fischer, H.; Seelig, A.; Macke, H. R.; Merbach, A. E. High Relaxivity for Monomeric Gd(DOTA)-Based MRI Contrast Agents, Thanks to Micellar Self-Organization. Chem. Eur. J. 1999, 5, 2977−2983. (6) Moghaddam, M. J.; de Campo, L.; Waddington, L. J.; Weerawardena, A.; Kirby, N.; Drummond, C. J. Chelating OleylEDTA Amphiphiles: Self-Assembly, Colloidal Particles, Complexation with Paramagnetic Metal Ions and Promise as Magnetic Resonance Imaging Contrast Agents. Soft Matter 2011, 7, 10994−11005. (7) Moghaddam, M. J.; de Campo, L.; Kirby, N.; Drummond, C. J. Chelating DTPA Amphiphiles: Ion-Tunable Self-Assembly Structures and Gadolinium Complexes. Phys. Chem. Chem. Phys. 2012, 14, 12854−12862. (8) Lee, H. S.; Kim, W. H.; Lee, J. H.; Choi, D. J.; Jeong, Y. K.; Chang, J. H. Transition Metal-Chelating Surfactant Micelle Templates for Facile Synthesis of Mesoporous Silica Nanoparticles. J. Solid State Chem. 2012, 185, 89−94. (9) Menger, F. M.; Gan, L. H.; Johnson, E.; Durst, D. H. Phosphate Ester Hydrolysis Catalyzed by Metallomicelles. J. Am. Chem. Soc. 1987, 109, 2800−2803. (10) Allen, W. D.; Jones, M. M.; Mitchell, W. C.; Wilson, D. J. Adsorbing Colloid Flotation of Cu(II) with a Chelating Surfactant. Sep. Sci. Technol. 1979, 14, 769−776. (11) Norgren, M.; Hedenström, E.; Edlund, H.; Andersson, F.; Högberg, I. Chemical and Method for Chelating Metal Ions Included in Water and Separating/Recovering of Formed Chelate. Patent WO 2009151366, 2009. 4612

dx.doi.org/10.1021/la5002336 | Langmuir 2014, 30, 4605−4612