Effect of [Zr( α -PW 11 O 39 ) 2 ] 10 − Polyoxometalate on the Self-Assembly of Surfactant Molecules in Water Studied by Fluorescence and DOSY NMR Spectroscopy

: The catalytic fragmentation of hydrophobic proteins by polyoxometalates (POMs) requires the presence of surfactants in order to increase the solubility of the protein. Depending on the nature of the surfactant, different effects on the kinetics of protein hydrolysis are observed. As the molecular interactions between the POMs and surfactants in solutions have been scarcely explored, in this study, the interaction between the catalytically active Keggin polyoxometalate [Zr( α -PW 11 O 39 ) 2 ] 10 − and four different surfactants—sodium dodecyl sulfate (SDS), dodecyldimethyl(3-sulfopropyl)ammonium (Zw3-12), dodecyldimethyl(3-sulfopropyl) ammonium (CHAPS), and polyethylene glycol tert -octylphenyl ether (TX-100)—have been studied in aqueous media. The effect of polyoxometalate on the self-assembly of surfactant molecules into micelles and on the critical micellar concentration (CMC) has been examined by ﬂuorescence spectroscopy and diffusion ordered NMR spectroscopy (DOSY).

The successful hydrolysis of proteins and other biological substrates is governed by a number of factors. One of the most important aspects is that both the protein substrate and the catalytic POMs are soluble in the same medium (i.e., water). In the case of the hydrolysis of hydrophobic proteins, a common approach to circumvent the low solubility (or often insolubility) of the proteins in aqueous solution is to use surfactants. However, as the surfactants exhibit polar or charged headgroups, they can also be involved in a variety of electrostatic or hydrogen bonding interactions with the POM [13]. This could hinder the binding of the POM to a protein, or it could inhibit their catalytic activity [14].
In order to address this issue, in this work, we seek to obtain molecular insights into the interactions between the [Zr(α-PW 11 O 39 ) 2 ] 10− polyoxoanion and commonly used dodecyldimethyl(3-sulfopropyl) ammonium (CHAPS), dodecyldimethyl(3-sulfopropyl)ammonium (Zw3-12), and surfactants-sodium dodecyl sulfate (SDS) (Figure 1), by using fluorescence spectroscopy and diffusion ordered NMR spectroscopy (DOSY). In the absence of POMs, water soluble surfactants at low concentrations populate the water/air interface. However, as the surfactant concentration gradually increases the interface becomes saturated and nanoscale supramolecular assemblies, such as micelles, are formed. The micelles form above a critical micellar concentration (CMC), which can be influenced by the following: (i) the charge of the hydrophilic headgroup, (ii) the length of the hydrophobic moiety, and (iii) the ionic strength of the solution [15]. Considering that POMs directly influence (i) and (iii), it is expected that the presence of POMs will influence the CMC. Furthermore, as POMs may be partially or fully encapsulated during the micelle formation, they may also affect the micellar size. The formation of surfactant encapsulated clusters (SECs) involving POMs is not unusual [16][17][18]. However, so far POM SECs have been mainly explored in non-aqueous media, although they have been attracting much attention in material science and catalysis [19,20].

The Interaction between Micelles and 1 Studied Using the Pyrene Fluorescence Method
In 1977, Kalyanasundaram and Thomas demonstrated that the fluorescence intensity of the vibronic fine structure of the pyrene monomer fluorescence is strongly solvent dependent [24]. Pyrene is characterized by five distinct vibronic peaks at 372 nm, 378 nm, 383 nm, 389 nm, and 392 nm, which are labeled I, II, III, IV, and V, respectively. The ratio of the vibronic peaks at 372 nm and 383 nm (typically referred to as the I:III pyrene ratio) is indicative of a micelle formation and has become a popular method for the determination of the CMC of surfactant solutions [25]. At surfactant concentrations below the CMC, the I:III ratio corresponds to a polar environment. When the surfactant concentration increases, the ratio decreases rapidly as the pyrene molecules sense the hydrophobic environment of the interior of the micelle. Above the CMC, the I:III ratio remains roughly constant when pyrene is incorporated into the micelle.

The Interaction between Micelles and 1 Studied Using the Pyrene Fluorescence Method
In 1977, Kalyanasundaram and Thomas demonstrated that the fluorescence intensity of the vibronic fine structure of the pyrene monomer fluorescence is strongly solvent dependent [24]. Pyrene is characterized by five distinct vibronic peaks at 372 nm, 378 nm, 383 nm, 389 nm, and 392 nm, which are labeled I, II, III, IV, and V, respectively. The ratio of the vibronic peaks at 372 nm and 383 nm (typically referred to as the I:III pyrene ratio) is indicative of a micelle formation and has become a popular method for the determination of the CMC of surfactant solutions [25]. At surfactant concentrations below the CMC, the I:III ratio corresponds to a polar environment. When the surfactant concentration increases, the ratio decreases rapidly as the pyrene molecules sense the hydrophobic environment of the interior of the micelle. Above the CMC, the I:III ratio remains roughly constant when pyrene is incorporated into the micelle.
This method has been widely used to study surfactants [26][27][28][29], mixed-surfactant systems [30,31], block-co-polymers [32][33][34], and the additive and temperature effects on the micellar properties of surfactants [35,36]. The emission spectrum of a saturated aqueous solution of pyrene at pH 7.4 (10 mM sodium phosphate buffer) is shown in Figure 2 and Figure S1a. The ratio of the fluorescence intensity of peak I to III plotted in the function of the surfactant concentration follows a Boltzman-type sigmoidal decrease (see inset Figure 2), which can be fitted to Equation (1).
Where I:III stands for the intensity ratio of peak I to III; x stands for the surfactant concentration, A1 and A2 are the upper and lower limits of the sigmoid, respectively; x0 is the center of the sigmoid; and Δx is related to the surfactant concentration range wherein the abrupt decrease of I:III occurs. According to Aguiar et al., the CMC of a surfactant is determined according to Equation (2) or Equation (3) [37], as follows: The ratio of the fluorescence intensity of peak I to III plotted in the function of the surfactant concentration follows a Boltzman-type sigmoidal decrease (see inset Figure 2), which can be fitted to Equation (1).
Where I:III stands for the intensity ratio of peak I to III; x stands for the surfactant concentration, A 1 and A 2 are the upper and lower limits of the sigmoid, respectively; x 0 is the center of the sigmoid; and ∆x is related to the surfactant concentration range wherein the abrupt decrease of I:III occurs.
According to Aguiar et al., the CMC of a surfactant is determined according to Equation (2) or Equation (3) [37], as follows: Generally, Equation (2) is used for surfactants with a high CMC, while Equation (3) is employed for low CMC values, typically below 1 mM. This different criterion has been explained in terms of the pyrene partition between the micelles and bulk solution. When the volume of the hydrophobic phase is not sufficient, pyrene is divided between the micelles and the bulk water, thereby providing an average I:III value that corresponds to a more polar environment, consequently prolonging the transition from high to low I:III ratios [38]. The calculated CMC values are summarized in Table 1 according to Equation (2) or Equation (3). The CMC of the SDS in pure water is reported to be 8 mM [37]. In a 10 mM sodium phosphate buffer at pH 7.4, the CMC is reduced to 4.16 mM. This decreases further to 1.40 mM in the presence of 1% 1. However, in the presence of 16 mM Et 2 NH 2 Cl, which is the approximate concentration of counterions that accompany 1% of 1 in a solution, the same CMC was observed. This indicates that the Et 2 NH 2 + cation is responsible for lowering the CMC of SDS, probably because of the screening of the repulsive anionic-anionic interactions, and not the POM polyanion itself. The CMC of TX-100 was increased from the reported 0.15-0.23 mM to 0.41 mM at pH 7.4 [39][40][41][42][43]. In the presence of Et 2 NH 2 + , the CMC was reduced to 0.26 mM, which was similar to the CMC in the presence of 1% of 1 (0.23 mM), again demonstrating that the POM itself has no significant influence on the CMC of TX-100. The CMC values that were determined at pH 7.4 for CHAPS and Zw3-12 (5.8 and 2.28 mM, respectively) fall within the range of the reported values for CHAPS (3 to 11 mM) [43][44][45][46][47][48] and Zw3-12 (2 to 4 mM) [49][50][51]. Both of the surfactants were unaffected by the presence of Et 2 NH 2 Cl, but in the presence of 1% 1, the CMC decreased to 2.0 mM and 1.10 mM for CHAPS and Zw3-12, respectively. This indicates that 1 screens the repulsive forces between the zwitterionic groups, allowing them to form micelles more easily. The structural information of the micelles can be derived from the I:III pyrene ratio as it is sensitive to its local environment. When the hydrophobicity of the local environment increases, the I:III ratio decreases. If additional water molecules permeate the micelle surface, then the ratio increases. Hence, the ratio can be used as a measure for the compactness of the head group packing at the micelle surface [24,25]. The I:III ratios of the pyrene populating the interior of the micelles in different conditions are summarized in Table 2.
The I:III ratio of pyrene dissolved within an SDS micelle is unaffected by the addition of Et 2 NH 2 Cl or 1, indicating that the packing of the polar, charged head groups remains the same. The surface of a TX-100 micelle becomes more compact when 1% of the POM is added, despite the unaffected CMC value. It is known that POMs can bind with polyether functions, and a similar binding could explain the closer packing of the polyethylene glycol head groups of TX-100. The I:III ratio of CHAPS was relatively unaffected by the addition of Et 2 NH 2 Cl, but in the presence of 1, the ratio decreased slightly. The POM probably shields the repulsive forces between the ammonium groups, but the vicinity of the anionic sulfonate causes repulsive interactions, avoiding a close packing of the head groups. Unfortunately, because of the overlapping peaks of the pyrene emission spectrum in the presence of Zw3-12, a I:III ratio could not be determined for these micelles (see Figure S1b in Supplementary Materials). However, because of the similar trend in the CMC values of Zw3-12 and CHAPS, one can assume that a similar trend would be observed for the packing of the micelle surface of Zw3-12.  (1)) was taken as the I:III ratio of pyrene populating the interior of the micelle; b All of the samples were buffered at pH 7.4 by a 10 mM sodium phosphate buffer; c Without 1 or its counterion (Et 2 NH 2 + ) present; d No I:III ratio could be calculated because of the overlap of peak II and III.

The Interaction between Micelles and 1 Studied by DOSY
DOSY is a 2D NMR technique that spectroscopically differentiates between particles in solution based on their translational self-diffusion coefficient (D). A DOSY spectrum consists of a standard 1D spectrum on the F2 (horizontal) axis and the diffusion coefficients in the F1 (vertical) dimension (see Figure 3 and Supplementary Materials), thus correlating the chemical shifts with the diffusion coefficients of the different components present in the solution. The I:III ratio of pyrene dissolved within an SDS micelle is unaffected by the addition of Et2NH2Cl or 1, indicating that the packing of the polar, charged head groups remains the same. The surface of a TX-100 micelle becomes more compact when 1% of the POM is added, despite the unaffected CMC value. It is known that POMs can bind with polyether functions, and a similar binding could explain the closer packing of the polyethylene glycol head groups of TX-100. The I:III ratio of CHAPS was relatively unaffected by the addition of Et2NH2Cl, but in the presence of 1, the ratio decreased slightly. The POM probably shields the repulsive forces between the ammonium groups, but the vicinity of the anionic sulfonate causes repulsive interactions, avoiding a close packing of the head groups. Unfortunately, because of the overlapping peaks of the pyrene emission spectrum in the presence of Zw3-12, a I:III ratio could not be determined for these micelles (see Figure  S1b in Supplementary Materials). However, because of the similar trend in the CMC values of Zw3-12 and CHAPS, one can assume that a similar trend would be observed for the packing of the micelle surface of Zw3-12.

The Interaction between Micelles and 1 Studied by DOSY
DOSY is a 2D NMR technique that spectroscopically differentiates between particles in solution based on their translational self-diffusion coefficient (D). A DOSY spectrum consists of a standard 1D spectrum on the F2 (horizontal) axis and the diffusion coefficients in the F1 (vertical) dimension (see Figure 3 and Supplementary Materials), thus correlating the chemical shifts with the diffusion coefficients of the different components present in the solution.
The diffusion dimension in the DOSY spectra is normally given in a logarithmic scale (as log D values in [log(m 2 ·s −1 )] units), which is a more convenient presentation when there is a large difference in the diffusion coefficients of the different components of the studied mixture. The diffusion coefficient of a compound is determined by its size, weight, and shape. If the compounds interact, their diffusion coefficient will decrease because of the larger size and molecular  The diffusion dimension in the DOSY spectra is normally given in a logarithmic scale (as log D values in [log(m 2 ·s −1 )] units), which is a more convenient presentation when there is a large difference in the diffusion coefficients of the different components of the studied mixture.
The diffusion coefficient of a compound is determined by its size, weight, and shape. If the compounds interact, their diffusion coefficient will decrease because of the larger size and molecular weight of the formed complex [52].The diffusion coefficients of the different surfactants, the counter ion (Et 2 NH 2 + ), and the POM cluster measured in their individual solutions and in the mixtures with a given composition are summarized in Table 3. With the aim to independently asses the interaction of the surfactant with the counterion Et 2 NH 2 + and with the POM polyanion, we first measured a DOSY spectrum of a mixture of each surfactant and Et 2 NH 2 Cl.
The results demonstrate that the interaction between the CHAPS micelles and Et 2 NH 2 + is negligible, as only a minor decrease in their diffusion coefficients was detected (see Table 3). The slight decrease of the D value of Et 2 NH 2 + from 70.5 × 10 −11 m 2 ·s −1 in its neat solution to 63.9 × 10 −11 m 2 ·s −1 in the mixture with CHAPS indicates that the counterion is likely to be involved in a fast equilibrium between its free state and its micelle bound state, which is strongly shifted toward the free state. These observations are in agreement with the pyrene fluorescence studies showing that the presence of the Et 2 NH 2 + counterion resulted in a minor decrease of the CMC of CHAPS (Table 1). In the following step, the mixture of CHAPS and 1 was investigated. Figure 3 shows the DOSY spectrum of a mixture of 50 mM CHAPS in the absence and in the presence of 9 mM of 1 or 90 mM of Et 2 NH 2 Cl. The spectrum demonstrates that in the surfactant/POM mixture, the diffusion coefficient of CHAPS (8.78 × 10 −11 m 2 ·s −1 ) is much lower (by a factor of 1.7) compared with the value measured in both the individual CHAPS solution and in the CHAPS/Et 2 NH 2 Cl mixture ( Table 3). The same effect was observed by 31 P DOSY for the diffusion coefficient of the POM cluster, which was 2.6 times lower in the mixture. The relatively close values of the diffusion coefficients of the CHAPS (8.78 × 10 −11 m 2 ·s −1 ) and 1 (8.21 × 10 −11 m 2 ·s −1 ) suggest the formation of a relatively stable micelle/POM complex because of the electrostatic interactions between the negatively charged POM and the positively charged quaternary ammonium group of CHAPS. The diffusion coefficient of the counterion Et 2 NH 2 + in the CHAPS/1 mixture (43.2 × 10 −11 m 2 ·s −1 ) was lower by a factor of 1.2 and 1.5, respectively, compared with its value in the individual solution of 1 (51.1 × 10 −11 m 2 ·s −1 ) and in the CHAPS/Et 2 NH 2 Cl solution. This trend could be explained by the weak binding of the counterion to the larger CHAPS/POM complex, probably through its interaction with the negatively charged POM and, to a lesser extent, with the sulfonate group of the CHAPS. This proposal is corroborated by the higher D value of the counterion measured in the CHAPS/Et 2 NH 2 Cl solution (63.9 × 10 −11 m 2 ·s −1 ), compared with its diffusion coefficient in the individual solution of 1 (51.1 × 10 −11 m 2 ·s −1 ), showing that the counterions interact more strongly with the POM cluster than with the negatively charged sulfonate groups of the surfactant micelles. The higher diffusion coefficient of the counterion compared with the diffusion coefficients of both the surfactant and the POM indicates that it is involved in a dynamic equilibrium between the bulk solution and the CHAPS/POM/ Et 2 NH 2 + complex.
A similar trend was observed for the diffusion coefficients of the other zwitterionic surfactant Zw3-12, the counterion Et 2 NH 2 + , and the POM cluster in both the Zw3-12/Et 2 NH 2 Cl and Zw3-12/1 mixtures (Table 3 and Figures S1-S3). In the Zw3-12/1 mixture, the diffusion coefficients of the Zw3-12 and the counterion Et 2 NH 2 + were significantly lower compared with their values measured in the Zw3-12/Et 2 NH 2 Cl solution, indicating the formation of the Zw3-12/1/Et 2 NH 2 + complex. However, the diffusion coefficient of the POM cluster (9.24 × 10 −11 m 2 ·s −1 ) measured by 31 P DOSY in the Zw3-12/1 mixture was larger than the D value of the surfactant (6.12 × 10 −11 m 2 ·s −1 ). The 1D 31 P spectrum of the Zw3-12/1 mixture shows only one signal at −14.69 ppm, contrary to the 31 P spectrum of pure 1, which displays two resonances at −14.57 and −14.65 ppm, which is characteristic of the dimeric mono-Zr substituted Keggin POM because of the presence of two units with different bonding modes (bond lengths and bond angles), as determined by X-ray crystallography [53,54]. The observed changes in the 31 P spectrum indicate that in the presence of Zw3-12 micelles, 1 undergoes structural transformations towards a monomeric kegginoidal polyoxoanion [53][54][55], with a lower overall negative charge than the parent dimeric [Zr(α-PW 11 O 39 ) 2 ] 10− cluster. A probable cause of this is the weaker electrostatic interactions between the positively charged ammonium group of the Zw3-12 and the POM cluster. The lower stability of the Zw3-12/POM complex corroborates the observed differences in the micellar and POM diffusion coefficients, as in this scenario, the contribution of the free micelles and that of the free POM species to the measured diffusion coefficients are expected to be more pronounced.
These results indicate that the POM/micelle/Et 2 NH 2 + complex formed in the Zw3-12/1 solution is different from the complex in CHAPS/1 mixture (see above), which also corroborates the differences in the pyrene spectra (see Figure S1b). The minor decrease in the diffusion coefficient of the SDS/1 mixture measured by the 31 P DOSY (from 20.9 × 10 −11 m 2 ·s −1 in its pure solution to 19.5 × 10 −11 m 2 ·s −1 in the SDS/1 mixture) corroborates that the anionic SDS micelles do not interact with 1. However, the diffusion coefficient of the SDS micelles decrease from 9.49 × 10 −11 m 2 ·s −1 in the pure SDS solution to 6.76 × 10 −11 m 2 ·s −1 in the presence of 9 mM 1, because of the interaction of the negatively charged sulfate group of the surfactant with the cationic Et 2 NH 2 + counterion of the POM. This interaction is further confirmed by the observed decrease in the diffusion coefficients of both the surfactant and the Et 2 NH 2 + counterion in the SDS/Et 2 NH 2 Cl mixture ( Table 3). The higher diffusion coefficient of Et 2 NH 2 + measured in the SDS/1 mixture (39.9 × 10 −11 m 2 ·s −1 ) compared with its value in the SDS/Et 2 NH 2 Cl mixture (26.6 × 10 −11 m 2 ·s −1 ) suggests that in the SDS/1 mixture, the counterion is most probably involved in a complex dynamic equilibrium because of its competitive interactions with both the SDS micelles and the polyoxoanion. The final surfactant, the neutral TX-100 (D pure = 4.55 × 10 −11 m 2 ·s −1 ), was largely unaffected by the presence of 1 (D 1 = 4.46 × 10 −11 m 2 ·s −1 ) or Et 2 NH 2 Cl (D count = 46.3 × 10 −11 m 2 ·s −1 ), as only minor changes of its diffusion coefficient were detected in both the TX100/1 and TX100/Et 2 NH 2 Cl mixtures. As expected, because of the lack of any charge, TX-100 does not actually interact with 1 or the counter ion.
Emission spectroscopy: All of the samples were buffered at pH 7.4 by a 10 mM sodium phosphate buffer and were saturated with pyrene (2-3 µM). The stock solutions were prepared using all of the surfactants and 1, using the buffered pyrene solution. A 10.0 mM quartz cuvette was used to record the emission spectra on an Edinburgh Instruments FLS-980 spectrometer (Edinburgh Instruments, Livingston, UK). The samples were excited at 337 nm and an emission spectrum was recorded from 350 to 410 nm. All of the fluorescence measurements were performed at ambient temperature.
The NMR spectra were recorded on a Bruker Avance II + 600 (600.01 MHz for 1 H, 242.89 MHz for 31 P) spectrometer (Bruker, Rheinstetten, Germany), equipped with a 5 mm direct detection dual broadband probe, with a gradient coil delivering maximum gradient strength of 53 G cm −1 . All of the experiments were performed at a temperature of 293 K. The DOSY spectra were measured using a convection compensating double-stimulated echo-based pulse sequence (dstegp3s), with monopolar square shaped gradient pulses. The spectra were acquired with 64k time domain data points in a t2 dimension, with 32 gradient strength increments, a diffusion delay of 100 ms, a gradient pulse length of 4 ms, 16 transients for each gradient step, and a relaxation delay of 2 s. The gradient strength was incremented from 4% to typically 70% of the maximum gradient output (from 1.92 to 33.7 G cm −1 ). For each sample, the maximum gradient strength was optimized to achieve optimal signal attenuation. The spectra were processed with an exponential window function (line broadening factor 0.5), 64k data points in F2, and 258 data points in the diffusion dimension. The evaluation of the diffusion coefficients was performed by fitting the diffusion profile (the normalized signal intensity as a function of the gradient strength G) at the chemical shift of selected signals in the DOSY spectrum with an exponential function using the variant of the Stejskal-Tanner equation adapted to the particular pulse sequence used.

Conclusions
In this work, we explored the effect of the Keggin polyoxometalate [Zr(α-PW 11 O 39 ) 2 ] 10− on the formation of surfactant-based micelles. Our experiments using fluorescent spectroscopy showed that the CMCs of the anionic SDS and the neutral TX-100 are unaffected by 1. DOSY NMR spectroscopy has further revealed that the anionic SDS does not bind with 1 but with the dissolved counter-cations. The neutral TX-100 is not affected by 1 or by the dissolved counteractions.
The CMCs of both CHAPS and Zw3-12 are significantly reduced under the influence of 1. 1 H and 31 P DOSY NMR spectroscopy indicated that both CHAPS and Zw3-12 interact with 1 and its counter ion, Et 2 NH 2 + . However, on the basis of 31 P NMR and fluorescence spectroscopy, it was shown that the encapsulated POM species preserves its dimeric structure in the CHAPS/1 complex, while the interaction with Zw3-12 resulted in structural transformations of the POM with the formation of monomeric Keggin type species with a lower overall negative charge. The overall work exhibits interesting aspects and offers an important contribution to the interactions of POMs and surfactants in aqueous media. The overall study is also useful for future theoretical studies focusing on molecular dynamics, which could provide further insights into the packing of individual POMs within the forming micelles [56].