Are Keggin's POMs Charged Nanocolloids or Multicharged Anions

Dec 26, 2017 - Indeed, the investigation of the micelles by SAXS and the variation in the cloud points of a polyethoxy surfactant with POMs have shown...
3 downloads 12 Views 1MB Size
Subscriber access provided by NORTH CAROLINA STATE UNIV

Article

Are Keggin’s POMs charged nano-colloids or multicharged anions? Alla Malinenko, Alban Jonchere, Luc Girard, Sandra Parrès-Maynadié, Olivier Diat, and Pierre Bauduin Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.7b03640 • Publication Date (Web): 26 Dec 2017 Downloaded from http://pubs.acs.org on December 27, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Langmuir is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Are Keggin’s POMs charged nano-colloids or multicharged anions? Alla Malinenko, Alban Jonchère, Luc Girard, Sandra Maynadié-Parres, Olivier Diat and Pierre Bauduin* Institut de Chimie Séparative de Marcoule (ICSM), UMR 5257 (CEA, CNRS, UM, ENSCM) CEA Marcoule, BP 17171, 30207 Bagnols-sur-Cèze, France.

ABSTRACT. Owing to their multiple charges and their nanometric size polyoxometalates (POMs) are at the frontier between ions and charged colloids. We investigated here the effect of POM-POM electrostatics repulsions on their self-diffusion in water by varying POM and supporting salt concentrations. The self-diffusion coefficients of two Keggin’s POMs: silicotungstate (SiW12O404-) and phosphotungstate (PW12O403-) were determined by dynamic light scattering (DLS) and DOSY 1H/31P NMR whereas POM-POM electrostatic repulsions were investigated by the determination of the static structure factors using small angle X-ray scattering (SAXS). The self-diffusion coefficients for the two POMs and for different POM/background salt concentrations were collected in a master curve by comparing the averaged POM-POM distance in solution to the Debye length. As for “classical” charged colloids, we show that the POM’s counterions should not be considered in the calculation of the ionic strength that governs POM-

ACS Paragon Plus Environment

1

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 43

POM electrostatic repulsions. This result was confirmed by fitting the POM-POM structure factor by considering a pair potential of spherical charged particles using the well-known Hayter mean spherical approximation (MSA). These Keggin’s POMs also behave as (super-)chaotropic anions, i.e. they have a strong propensity to adsorb on (neutral polar) surfaces, was also investigated, here on the surface of octyl-beta-glucoside (C8G1) micelles. The variations of (i) the chemical shift of 1

H/31P NMR signals and of (ii) the self-diffusion coefficients obtained by DOSY 1H/31P NMR of

PW3- and of C8G1 were in good agreement, confirming the strong adsorption of POMs on the micelle polar surface on a static and a dynamic point of view. We concluded that Keggin’s POMs behave (i) as anions, because they adsorb on surfaces as chaotropic anions do, as well (ii) as colloids because they can be described by a classical colloidal approach by dynamic and static scattering techniques, i.e. by the investigation of their inter-particle electrostatic structure factor and self-diffusion without considering the POM’s counterions in the ionic strength calculation.

This work highlights the dynamic properties of POMs at soft interfaces, compared to bulk aqueous solution, which is essential in the understanding of functional properties of POMs, such as (photo)-catalysis or in the rational design of POM-based hybrid nano-materials from soft templating routes i.e. in aqueous solutions at room temperature.

INTRODUCTION The general interest in polyoxometalate (POM) nanometric ions has spawned from their wide variety of compositions and structural versatility.1-3 The presence of transition metal atoms in high oxidation state confer to POMs an extensive range of chemical and physical properties such as catalytic,4-6 redox,2, 7 magnetic,8 dielectric,9 photoluminescence and photochromic.10-12 These many features make POMs used in a wide range of applications such as chemical synthesis,

ACS Paragon Plus Environment

2

Page 3 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

material science, electrochemistry or medicine.1, 13-16 As surface effects are essential for most of the POM applications, see for example in reference7, numerous investigations have focused on their adsorption on soft or solid support or via (hybrid-) self-assembling.17-24 Nevertheless, in contrast to the extensive interest in these unique characteristics, the investigation of the dynamic properties of POMs in aqueous solution has gained much less attention25-33 whereas they are essential to control all transport or reaction phenomena as well as their selfassembling or building of complex 2D or 3D architectures. Among dynamic properties, the selfdiffusion coefficient (D) is of particular significance to better understand molecular processes and interactions in solutions. The self-diffusion coefficients of the most common Keggin’s POMs obtained from previous works are gathered in Table 1 and are discussed in the following paragraphs. Some of the pioneer works on diffusion of POMs in aqueous solution were performed by Pope and co-workers.25-26 Their first study on the subject focused on the investigation of the D of two isomorph pairs of POMs, molybdosilicate (SiMo12O404-) and tungstosilicate (SiW12O404-) as well as molybdocobaltiate (CoMo6O213-) and molybdochromiate (CrMo6O213-). Using a method based on a specific radioactive marker of the POMs solute the authors demonstrated that POM selfdiffusion was independent of their molecular weight and for the large species it is rather influenced by the solvent surrounding the POM. Although their second work26 was focused on the reduction properties of POMs, the diffusion coefficients of tungstophosphate (PW12O403-) and tungstosilicate (SiW12O404-) were deduced from electrochemistry measurements. Different diffusion coefficients were obtained for these two POMs and the authors suggested that this difference could originate from the different viscosities of the supporting electrolytes.

ACS Paragon Plus Environment

3

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 43

In 2000, Grigoriev V. A. et al. have investigated ion pair formation between POM and different counterions in various media using chronoamperometry.30 In this study they determined the diffusion coefficients of POMs in order to estimate their apparent radii and they did not observe ion pairs in water. In contrast, ion pairs between POMs and cations were observed in water/alcohol mixtures, as expected from the lower dielectric constant of these media compared to pure water.30 Furthermore, using a molecular dynamics (MD) approach, Leroy et al. have analyzed the self-diffusion of Keggin-type POMs with different counterions in water29 and their results have shown some consistency with data published by Grigoriev V. A. et al. One of the conclusions drawn from this simulation work was that the diffusion coefficient dependence on the charge of POM as well as on the size of the counterion is not trivial. It depends on the subtle interplay between direct POM-cations electrostatic interactions and also on the stability of the solvation shell of the ions. Meanwhile, Olynyk et al. studied the diffusion coefficients of 12-tungstosilicate by cyclic voltammetry in the presence of an excess of supporting electrolyte to suppress the migration of the electroactive species.33 They obtained the diffusion coefficient at infinite dilution by extrapolation of the experimental values from the mean spherical approximation (MSA) transport calculations in agreement with conductivity results. This work suggests that there is no ion association (ion pairs) in water,34 this is in agreement with previous works.29-30 Ion-pairing between Keggin’s POMs monovalent cations was so far only observed in the special case of niobate based POMs, such as [SiNb12O40]16− and [GeNb12O40]16−, due to their extremely high charge density.35-36 In 2003, diffusion coefficients of 12-tungstosilicate in acid form at low POM concentrations and at various temperatures were calculated by Horky et al. via the ion mobility, obtained from

ACS Paragon Plus Environment

4

Page 5 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

conductivity measurements.32 At room temperature the D value obtained in this latter study was around 20% lower than the one obtained by Olynyk et al .33 More recently Poulos with co-workers investigated the diffusion of 12-tungstophosphate in the confined hydrophilic region of lyotropic lamellar phases by pulsed gradient spin echo (PGSE) NMR and they reported a strong reduction of the diffusion coefficient compared to those determined in bulk water at the same concentration. The authors explained this effect by the adsorption of POMs onto the non-ionic surfactant bilayers.27 A recent MD study on 12tungstophosphate showed that discrepancies in the diffusion coefficient values exist depending on the chosen charge model.28 Table 1. Type of POM, POM anion molecular weight (IW), charge of POM (z), POM concentration ([POM]), medium, viscosity of the medium (η), temperature (T), diffusion coefficient (D), hydrodynamic radius (Rh), ionic strength (I) and technique reported in the referred publications.

POM

[CoMo6O21]3[CrMo6O21]3[SiMo12O40]4[SiW12O40]4-

[PW12O40]3-

D IW η [POM] T Rh g/m z Medium mPa m2/s ⋅ mM ˚C nm 10 ol ·s 10 Baker L. C. W. and Pope M. T., JACS, 1960, 4176-417925 7.4 ± 971 3 0.2* 8 – 21 7.4 ± 964 3 0.2* H2O + NaClO4 30 6.2 ± 1820 4 IS = 1 0.2* 3 – 11 6.1 ± 2875 4 0.2* Pope M. T. and Varga G. M., Inorg. Chem., 1966, 1249–125426 H2O + 2877 3 0.05 – 5 1.2 25 3.36 ± 0.56 1M

Techniques

Open capillary method

Polarographic

ACS Paragon Plus Environment

5

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 43

H2SO4 [SiW12O40]4-

[PVW11O40]3-

[SiVW11O40]4-

[AlVW11O40]5-

[PW12O40]3[SiW12O40]4[AlW12O40]5-

[SiW12O40]4-

[SiW12O40]4-

2875 4

H2O + 0.9 M Na2SO4

0.15* 1.5

25

2.56 ± 0.15*

method 0.56

Grigoriev V. A. et al, J. Am. Chem. Soc. 2000, 3544-354530 0.62 H2O:TBA 2.38 – 60 – a (2:3 v/v) 2.78 2744 3 1 0.72d H2Ob 0.89 25 0.57 0.68 H2O:TBA 2.08 – 60 – (2:3 v/v)a 2.56 2741 4 1 0.83d H2Ob 0.89 25 0.56 0.96 1.64 – H2O: TBA 60 – (2:3 v/v)a 1.79 2740 5 1 1.03d H2Ob 0.89 25 0.59 29 Leroy F. et al, J. Phys. Chem. B 2008, 8591–8599 1.8 – 2877 3 2.0 1.1 – 2875 4 54 H2O 0.89 25 1.6 1.1 – 2873 5 1.7 Olynyk T. et al, J. Phys. Chem. B 2001, 7394-739833 4.5 – 4.9c H2O + 2875 4 0.5 25 HClO4 5.1* Horky A. et al, Horky, J. Electrochem. Soc., 200332 1.30 10 2.4* 6 1.05 18 3.2* 3 at H2O 2875 4 infinite 0.89 25 3.9* dilution 0.79 30 4.8* 7 0.54 50 7.0* 7 Poulos A. S. et al. , J. Phys. Chem. B, 2010, 220-22727

singlepotential-step chronoampero metry

MD

square wave voltammetry Voltammetry + MSAtransport theory

Conductivity measurements

ACS Paragon Plus Environment

6

Page 7 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

[PW12O40]3-

2877 3

~ 94

H2O

4.6

0.48

Lopes X. et al, J. Phys. Chem. A, 2005, 1216-122228 2.4* [PW12O40]32877 3 H2O 0.89 3.2* * - diffusion coefficient at infinite dilution

PGSE-NMR

MD (2 models)

TBA - tert-butyl alcohol a

- addition of XCl, where X = Li, Na, K. The concentrations of added salt depend on X and POM and vary in the range from 85 to 202 mM b

- addition of XCl, where X = Li, Na, K. The exact concentrations of added salts were not specified c

- varies with the concentration of added HClO4 (0.11 M < [HClO4] < 1M)

d

- effective radius

Table 2. Experimental results from the present work: type of POM, POM anion molecular weight (IW), charge of POM (z), POM concentration ([POM]), medium, viscosity of the medium (η), temperature (T), diffusion coefficient (D), hydrodynamic radius (Rh), ionic strength (I) and technique.

POM

[SiW12O40]4-

IW z g/mol

2875

[POM] mM

Medium

η mPa·s

10 15 20 25 30 35

H2O

0.89

4 10

20

H2O + 50 mM NaCl H2O + 75 mM NaCl H2O + 100 mM NaCl H2O + 250 mM NaCl H2O + 50 mM NaCl H2O + 100

T ˚C

D m2/s ⋅ 1010 9.0 11.1 12.1 11.7 12.8 11.2

Rh nm

Technique

5.93 25 0.90± 0.01

DLS

5.21 4.87

0.50

4.69

0.52

7.19 5.85

ACS Paragon Plus Environment

7

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

mM NaCl H2O + 300 [PW12O40]3- 2877 3 30 1.30a 25 mM NaCl a - viscosity of 30 mM of H3PW in presence of 300 mM NaCl

Page 8 of 43

3.25

0.52

DOSY NMR

Many factors such as concentration, nature and concentration of the supporting electrolyte, viscosity of the solution and temperature influence the self-diffusion of charged solutes. The POM/electrolyte concentrations are the most influent parameters on the POM self-diffusion through the change in the POM-POM electrostatic interactions. POM nanometric anions are much larger than standard anions, such as chloride for example, and can therefore be seen as small (nano-)colloids or as big (macro-)ions.37 They have many delocalized charges but their overall charge densities are low, and their polarizabilities are therefore expected to be high. As a consequence it is likely that POMs interact through dispersion forces (induced dipole-induced dipole) in addition to classical electrostatics effects.38 In a previous work we highlighted that Keggin’s POMs have a strong tendency to adsorb, through non-covalent interactions, on soft hydrated surfaces in water media.22 Indeed it was shown that PW12O403- and SiW12O404- adsorb strongly on non-ionic surfactant micelles and at the water/air interface covered by nonionic surfactants, i.e. polyethoxy- or sugar-based surfactants. It was proposed that the process of POM adsorption originates mainly from an entropically-driven process through the partial dehydration of the POMs and of the interface. The ability of POMs to adsorb on hydrated surfaces was further demonstrated on hydrophilic oligomers (polyethylene glycol, PEG) in bulk water.21 Indeed, it was shown that POMs builds soluble POM-PEG nanoassemblies in water. POMs were assigned within the anion classification of Hofmeister to the term “super-chaotropic anions” because of their adsorption properties and their ability to increase strongly the cloud point of non-ionic surfactants.22 Chaotropic anions are indeed usually referred to as large and polarizable (low charge density), which makes Keggin’s POMs archetypes of

ACS Paragon Plus Environment

8

Page 9 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

super-choatropic anions. The general super-chaotropic behavior of POMs, and more specifically their adsorption on hydrated polar interfaces, appears to be essential for the design of hybrid materials using soft methods, which is a field of growing interest.12, 21, 39-41 Moreover it seems that the super-chaotropic behavior is not exclusive to POMs but it is a more general property of nanoions with delocalized charges, such as boron clusters: dodecaborate,42 which was recently described as a super-chaotrope, or metallacarboranes.43-45 Considering the IUPAC’s definition of a “colloidal system”,46 which is based on a unique criterion of size, i.e. “…particles dispersed in a medium have at least in one direction a dimension roughly between 1nm and 1µm”, Keggin POMs (≈ 1 nm) are borderline cases to be called “colloids”. However it is commonly accepted that Van der Waals (VdW) forces are determinant in colloidal systems whereas ions are mostly controlled by pure electrostatics. Indeed VdW forces are at least in the same order of magnitude as electrostatics for (large) charged colloids, whose equilibrium of forces is well picked up by the classical DLVO theory.47 However earlier evidences have shown that “ions” of 2 – 3 nm in size, such as gold nanoparticles covered by SAM (self-assembled monolayers) ended with charged carboxylate groups,48 or Keplerate POMs49-51, do not behave like colloids because VdW forces are typically two orders of magnitude lower than electrostatic interactions for these systems. Recent simulations have shown for Kelplerate’s POMs (2 – 3 nm) that the increase in the electrical charge could produce the condensation of counterions, which subsequently leads to the Keplerate’s POMs self-assembly.49 Counterions condensation, induced by increasing surface charge density, is a common (pure electrostatic) effect in colloids and surface science, for example it is observed with charged surfactant monolayers by increasing the surface concentration,52 or in polyelectrolytes (known as Manning condensation). For Keggin’s POMs (≈ 1 nm) the situation seems to be very much different. Indeed a recent work by

ACS Paragon Plus Environment

9

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 43

Bera et al. has investigated Keggin’s POMs with 3, 4 and 5 charges, namely PW12O403-, SiW12O404- and AlW12O405-, by SAXS experiments using synchrotron radiation.53 They concluded that Keggin’s POMs with higher charges show stronger electrostatic repulsions but also that the POM of lower charge density (PW12O403-) form nano-aggregates of POMs. As a consequence, there is an apparent contradiction on the effect of the electrical charge on the self-assembly (aggregation) of POMs, with Keplerate’s POMs on one side and Keggin’s POMs on the other side, suggesting a different origin for the POM aggregation. Keplerate’s POMs show a classical electrostatic effect which induces counterion condensation (and self-assembly) presumably due to their high charge density whereas the aggregation of low charge density Keggin’s POMs, e.g. PW12O403-, may be related to their super-chaotropic behavior. This is supported by a previous investigation showing that the super-chaotropic behavior of PW12O403- was more pronounced than the one of SiW12O404-.22 The apparent dual behavior of POMs between salt and colloid was carefully discussed by Tianbo Liu in a recent contribution to this journal.17 On one hand it was pointed out that POMs, because of their delocalized charges, cannot be described by classical electrostatic theory, such as DebyeHückel which assumes point charges. On the other hand POMs are soluble species which makes them different from colloidal particles stabilized by the combination of repulsive and attractive interactions as described in the DLVO theory. The previous contribution and discussion by T. Liu focused on much larger POMs of Keplerate’s type, i.e. POMs which have a wheel structure with typically above 100 of metal centers and at least ten charges. As stated above, Keplerate’s POMs have the general tendency to self-assemble in large blackberry vesicles mediated by counterions condensation17, 54-56 while on the contrary Keggin’s POMs are fully dissociated from their (monocharged) counterions and generally show no self-assembly in water.33

ACS Paragon Plus Environment

10

Page 11 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

In the present work we tackle a similar question as raised previously by T. Liu on the “ion” or “colloid” nature of POMs but we focus here on the general properties of Keggin’s POMs in water and at interfaces with a different approach based on the investigation of the dynamic and static properties of POMs. In order to understand deeper the behavior of POMs and to figure out if they behave as charged colloids or as salts, we followed an experimental approach that is classical to investigate charged colloids. The self-diffusion of POMs and the POM-POM structure factor, that both account for POM-POM electrostatic interactions, was investigated by using dynamic light scattering (DLS), diffusion ordered spectroscopy nuclear magnetic resonance (DOSY NMR) and small angle X-ray scattering (SAXS) techniques. Tungstophosphate (PW12O403-) and tungstosilicate (SiW12O404-) were chosen as representatives of the Keggin’s POMs because: (i) they were already well studied in the literature (see Table 1), (ii) they have an isomorph structure with a near spherical shape and (iii) they have different number of charges which enable to investigate specifically the effect of charge density on the electrostatic interactions. To the best of our knowledge the self-diffusion of free POMs in water was so far never investigated by dynamic light scattering. As POMs have shown the general property of a (super-)chaotropic anion, i.e. to with the propensity adsorb on (hydrated) surfaces, the dynamic and static properties of POMs were not only investigated in bulk water but also at the surface of non-ionic micelles, made of a sugar-based surfactant: n-Octyl-β-monoglucoside (C8G1). EXPERIMENTAL SECTION Materials Silicotungstic acid (H4SiW12O40xH2O, H4SiW, MW = 2878.17 g/mol, >99.9%), phosphotungstic acid (H3PW12O40xH2O, H3PW, MW = 2880.05 g/mol, >99.9%), n-Octyl−β-D-monoglucoside also

ACS Paragon Plus Environment

11

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 43

known as C8G1 (MW = 272.37 g/mol, purity > 98%, cmc = 25 mM) and sodium chloride (NaCl, >99.5%) were purchased from Sigma-Aldrich. The thermal gravimetric analysis (TGA) measurements showed that water content in POMs equals up to 10 and 5 % for H4SiW and H3PW respectively. The molar concentration of POM was converted to volume fraction using the density of the POMs, which is around 5.39 g/ml.57 All chemicals were used as received. The Milli-Q water (18.2 MΩ·cm at 25 °C) was used for sample preparation. Sample preparation All POM solutions were prepared by solubilisation of POM powders in Milli-Q water. For the DLS experiments all POM solutions were filtered just before the measurements using polytetrafluorethylene or cellulose acetate membrane with pore size of 0.4 µm. The POM solutions in the presence of NaCl were prepared by adding salt solutions at certain concentration to the POM powder. The solutions containing POM and C8G1 with and without NaCl were prepared by solubilisation of POM powders in a solution containing the surfactant/NaCl at the desired concentration. pH of the POM solutions was measured and was always below 4.0, i.e. in the range of chemical stability of the POMs in water.58-59 Methods SAXS measurements using Mo radiation (λ = 0.071 nm) were performed on a bench built by XENOCS. The collimation was applied using a 12:∝ multilayer Xenocs mirror (for Mo radiation) coupled to two sets of scatterless FORVIS slits providing a 0.8 × 0.8 mm X-ray beam at the sample position. The scattered beam was recorded using a large online scanner detector (diameter: 345 mm, from MAR Research). A large q-range (0.2 to 40 nm-1) was covered thanks to an offcenter detection. Pre analysis of data was performed using FIT2D software. The scattered

ACS Paragon Plus Environment

12

Page 13 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

intensities are expressed versus the magnitude of scattering vector  = [(4)⁄ ] sin (⁄2)), where λ is the wavelength of incident radiation and θ the scattering angle. 2 mm quartz capillaries were used as sample containers for the solutions. Usual corrections for background (empty cell and detector noise) subtractions and intensity normalization using high density polyethylene film as a standard were applied. Experimental resolution was ∆q/q = 0.05. For analyzing the SAXS spectra SASfit software was used.60 Diffusion coefficients of POMs were determined by dynamic light scattering using an ALV-CGS3 goniometer equipped with a 22 mW HeNe Laser (632.8 nm) and an APD-based single photon detector coupled to an ALV/LSE-5004 auto-correlator. It has a minimum real time sampling time of 0.1 µs and a maximum of about 50 s. For all experiments temperature was maintained at 25 ˚C. The experimental autocorrelation function was measured at different angles from 30 to 150° with a 10° step. The time correlation function g2(q,τ) - 1 showed two modes for unfiltered H4SiW solutions and one mode for filtered ones. The autocorrelation functions were fitted by the least squares method applying the Eq. (S1 and S2) and by using the Solver in Microsoft Excel. For H3PW samples two modes were always observed, even after filtration. The origin of the second slow mode is likely due to the presence of nano-aggregate as observed by Bera et al. for HPW at high acidicity.53 The correlation functions were first analyzed using two exponentials and only the faster mode, related to the free POMs in solution, was used to determine the diffusion coefficient. NMR measurements were recorded at 25 °C on a Bruker 400 Avance III spectrometer, operating at 400.13 MHz for 1H and 161.98 MHz for 31P. It was equipped with a z-gradient 5 mm BBFO probe. Chemical shifts for 1H and 31P NMR are reported in parts per million respectively to tetramethylsilane (TMS) and 85% phosphoric acid. The 1H spectra were collected with a 3.98 s

ACS Paragon Plus Environment

13

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 43

acquisition time, 30 s relaxation delay and 7.5 µs 90° pulse width. The 31P spectra were collected with a 0.51 s acquisition time, 120 s relaxation delay and 9.5 µs 90° pulse width. DOSY experiments were performed with the bipolar gradient pulse pair longitudinal eddy-current delay sequence (ledbpgp2s). Methanol was used to calibrate the temperature of the probe. The gradient strength was calibrated with a D2O sample. The diffusion coefficients of PW3- and C8G1 were obtained from corresponding peaks located at around -15.3 ppm for 31P NMR spectra and at around 4.4, 1.2 and 0.8 ppm for 1H NMR spectra. RESULTS AND DISCUSSIONS POMs as charged nano-colloids Self-diffusion of POMs by DLS and NMR The analysis of the dynamic light scattering (DLS) results obtained from the H4SiW filtered solution yielded a monomodal correlation function with a fast mode indicating the presence of small particles (Fig. S1 in SI). The average exponential decay, Γ, of the auto-correlation functions has a linear q2 dependence which is the sign of a purely diffusive process and that can be attributed to the self-diffusion of POMs. The self-diffusion coefficients are plotted in Fig. 1 (a) (black squares) as a function of H4SiW concentration. The POM diffusion coefficient only slightly varies by increasing H4SiW concentration, except for the lowest (measurable) concentration of 10 mM and equals in average to 12 ⋅ 10-10 m2/s. This value is almost twice higher than the ones previously published, see Table 1 for comparison. This discrepancy can be attributed to the high electrostatic interactions between POMs in pure water. Indeed it is well known that electrostatic repulsions between particles leads to increase D values compared to non-interacting particles.61 This assumption can be tested by two experimental methods: the first one is to dilute the system to

ACS Paragon Plus Environment

14

Page 15 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

decrease POM-POM interactions, however further dilution below 10 mM (0.5% v/v) would lead to a too low scattered intensity, below the sensitivity limit of the DLS apparatus. The second approach is to screen electrostatic interactions by adding a supporting electrolyte, such as sodium chloride. In Fig. 1 (b) the diffusion coefficient D for 10 and 20 mM of H4SiW was plotted as a function of added salt concentration. The D values obtained for 20 mM are higher than for 10 mM, as expected from stronger POM-POM electrostatic repulsions. The addition of salt leads to a decrease in D value which can be attributed to the vanishing of the repulsive interactions. At around 100 mM NaCl full screening of the electrostatic repulsions is reached as D tends to a constant value (4.9 10-10 m2/s). As the volume fraction is only of 0.5 %v/v at 10 mM, hard sphere repulsions can be safely considered as negligible meaning that this D value is the one expected at infinite dilution, D0. This value was reported in Fig. 1 (a) (empty crossed square) and can be also retrieved from the diffusion coefficient measurements without added salt by extrapolation of D to zero H4SiW concentration as shown in Fig. 1 (a) by the dotted guide line. Thus, the D0 value of H4SiW in the presence of sufficient salt concentration or by extrapolation at low H4SiW concentration is now in a good agreement with published values (Table 1, see ref.32-33). The conversion of this value of D0 into hydrodynamic radii (Rh) from the Stokes-Einstein relation gives a Rh value of around 0.50 nm, which is close to geometrical radius (R ~ 0.45 nm) and hydrodynamic radii obtained in the publications mentioned in Table 1 or determined from small angle x-ray scattering experiments.2, 22 Note, that the viscosity of 10 mM of H4SiW in the presence of salt (0.90 ± 0.01 ⋅ 10-3 Pa/s) was used to calculate the hydrodynamic radius. It is then confirmed that the high values of D obtained here are due to significant POM-POM electrostatic repulsive interactions. For POM this effect is significant even at low concentrations since H4SiW is fully dissociated in aqueous solutions with SiW carrying four negative charges.

ACS Paragon Plus Environment

15

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 43

Figure 1. Diffusion coefficient D of SiW4- as a function of concentration in water (a), and as a function of sodium chloride concentration for [H4SiW] = 10 and 20 mM (b). The value of D for [H4SiW] = 10 mM with 100 mM of NaCl is assumed to be the value at infinite dilution and is reported in (a) at [H4SiW] = 0 (empty crossed square). For H3PW two modes in the auto-correlation functions were obtained (see Experimental Section). Though the fast one can be attributed to the POM self-diffusion process, the origin of the slow mode is unclear. It may result from the presence of small aggregates of PW3-, as observed in a highly acidic medium by Bera et al. in a recent contribution.53 These authors also investigated

ACS Paragon Plus Environment

16

Page 17 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Keggin’s POMs with higher electrical charges, SiW4- and AlW5-, for which no aggregation was detected. In a recent investigation it was shown that the super-chaotropic behavior of PW3- is more pronounced compared to SiW4-, which suggests here that the super-chaotropic behavior may be related to the aggregate formation for PW3-.22 Indeed the super-chaotropic property of Keggin’s POMs is related to their high polarizability, which may induce strong POM-POM attractions (dispersion forces)62-63 able to fight against electrostatic repulsions and ultimately lead to POM aggregation. It was shown theoretically for nitrate (NO3-), a typical chaotropic anion, that significant anion-anion attraction forces emerge from dispersion forces at high concentrations (in the molar range) but without aggregation formation.62, 64 Nevertheless, from the fast mode and for a solution containing 100 mM NaCl, the D0 and Rh values obtained for H3PW are close to the ones of H4SiW, respectively 4.3 m2/s (0.57 nm) compared to 4.9 m2/s (0.50 nm). This was expected from the same size and ionic weight of SiW4- and PW3- and this is in agreement with previous measurements (see Table S1 and experimental section/DLS for more details). 31

P DOSY NMR measurements were carried out on this system, by keeping in mind that the

H3PW self-diffusion coefficient value obtained from DLS measurements can be affected by the presence of the second (slow) mode contribution to the time correlation function. The diffusion coefficient for H3PW in water was measured by Poulos et al. using 31P PGSE-NMR at 5%v/v (~ 94 mM).27 The self-diffusion of H3PW was investigated with the same procedure but here at a lower concentration (30 mM, 1.6 %v/v) and in the presence of 300 mM of NaCl to prevent the effect of electrostatic interactions on D. Here a value of 3.3 x 10-10 m2/s was determined, corresponding to an apparent hydrodynamic radius of 0.52 nm, taking into account the relatively high viscosity of the solution (1.30 x 10-3 Pa/s). This value is very close to the one obtained for H4SiW by DLS (0.50 nm) and in agreement with the size of POM and earlier results (Table 1).

ACS Paragon Plus Environment

17

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 43

POM-POM electrostatic interactions From Fig.1 (b) it was shown that D depends strongly on both the POM and the salt concentrations as expected from electrostatics. An attempt was made here to rationalize the variations in the D values according to electrostatic considerations. The aim is to collect in a master curve all the experimental data (D) obtained for different compositions: changing the type of POM (SiW4- or PW3-) and the POM/background salt concentrations. To fulfill this goal, electrostatics effects are evaluated by comparing the POM-POM average distance in solution to the Debye lengths that characterize long-range interactions in the system. Debye lengths λD were calculated from Eq. 1:      

 =  





!" 

(1)

where ε0 is the vacuum permittivity, εr - the relative permittivity, kB - the Boltzmann's constant, T – temperature, NA – the Avogadro number, e – the elementary charge and I – the ionic strength. The values of Debye lengths depend on the ionic strength which can be calculated in different ways. Indeed the POM’s counter-ions may or may not be taken into account in the calculation of I. The general consensus is that the ionic strength, that should be considered in the calculation of the Debye length, is the one at the mid-plane between colloids, i.e. the ionic strength at equidistance between the charged particles. Therefore for diluted charged colloids, for which the average colloid-colloid distance is large, only the background salt (and not the colloid’s counterions) contributes to the ionic strength of the medium. However we express in the following the ionic strength for the general case as:

ACS Paragon Plus Environment

18

Page 19 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

!

 3  # = #$% +  '([) * ]+, - + [./ * ]+0 - + [12 ]+45 6 7

(2)

where, Ibkg is the ionic strength of milliQ water (which is typically ~ 10-5 M for air-equilibrated water), zi is the charge number of the ion, i, and α is a factor ranging from 0 to 1 that represents the contribution of the POM’s counterions to I. For α = 1, POM’s counterions contribute fully to I and for α = 0, POM’s counterions do not contribute to I, as it is classically considered for charged colloids. In order to estimate the strength of POM-POM interactions we define a unitless ratio d/2λD, where d is the average distance between the surfaces of POMs for a given concentration, assuming that POMs are distributed on a cubic lattice. Therefore d = dPOM-POM – 2RPOM with dPOM-POM the average POM-POM distance and RPOM the POM radius. For d/2λD > 1, POMs are considered to be too far from each other to interact and for d/2λD < 1, the double layers overlap which implies that significant electrostatic repulsions take place. In Fig.2, D/D0 ratios are plotted as a function of d/2λD for the different compositions: [H4SiW] = 10 mM with [NaCl] = 0, 20, 50, 75, 100 and 250 mM, [H4SiW] = 20 mM with [NaCl] = 0, 20, 50, 100 and 150 mM and [H3PW] = 10 mM in the presence of 100 mM of NaCl. The self-diffusion coefficient at infinite dilution D0 for H4SiW and H3PW was assumed to be equal to the diffusion coefficient of H4SiW (10 mM) screened by 250 mM of NaCl and diffusion coefficient of H3PW (10 mM) screened by 100 mM of NaCl. λD was calculated with different values of α = 0 and 1, shown respectively in Fig.2(a) and 2(b). For α = 0, all points set on a master curve except for the two points with the lowest d/2λD values (close to 0). These two lower d/2λD values in Fig. 2(a) were obtained with [H4SiW] = 10 and 20 mM, their low values arise from the high Debye lengths, which diverge at low ionic strength. This discrepancy at very low d/2λD values is likely to arise

ACS Paragon Plus Environment

19

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 43

from a too strong influence of the repulsive interactions on the diffusion coefficient. The mutual diffusion coefficient should be then considered rather than the self-diffusion coefficient.65 A loglog representation of Fig.2, given in SI (see Fig. S9), makes a break (for α = 0) more apparent between the two regions for d/2λD < 1 (D > D0) and d/2 λD > 1 (D ≈ D0) i.e. between the concentrated (and diluted regimes) where POM-POM interactions play (and do not) play a role in the self-diffusion of POMs. For α = 1, Fig. 2(b) no self-agreement between all the experimental data was obtained, i.e. no master curve was obtained. Therefore POM’s counterions should not be taken into account in the calculation of I, as it is the case for diluted and charged colloids. As a conclusion, by considering POMs as (nano-) colloids, basic electrostatic considerations enable to rationalize the variation of the POMs diffusion coefficient for different POMs and concentrations of POM and supporting salt.

ACS Paragon Plus Environment

20

Page 21 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Figure 2. D/D0 as a function of d/2λD for: [H4SiW] = 10 mM with [NaCl] = 0, 20, 50, 75, 100 and 250 mM, [H4SiW] = 20 mM with [NaCl] = 0, 20, 50, 100 and 150 mM and [H3PW] = 10 mM in the presence of 100 mM of NaCl. D0 is the self-diffusion coefficients at infinite dilution. We made the assumption that D0 values for H4SiW and H3PW equal the ones obtained for [H4SiW] = 10 mM with [NaCl] = 250 mM and [H3PW] = 10 mM with [NaCl] = 100 mM, respectively. The curves are calculated for α = 0 (a) and α = 1 (b). For α = 0, a master curve is observed with all the experimental data obtained here except for d/2λD close to zero. The investigation of the self-diffusion of POM using DLS technique emphasized the electrostatic interactions between POMs. The next step was to investigate POM-POM interaction in static conditions using SAXS. In our previous works, SAXS spectra of POMs in solution have been already shown, either free in solution21-22 or adsorbed on micelles.22 The previous SAXS analysis was performed by considering POMs as homogeneous spheres. However, here we focused on the POM-POM interactions through the specific POM structure factor, S(q), and its evolution as a function of POM concentration and ionic strength.

ACS Paragon Plus Environment

21

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 43

In Fig. 3 (a) the SAXS spectra of H4SiW are plotted in absolute intensity, at different POM concentrations in water. In such isotropic media, the scattered intensity by POMs as a function of q can be simply expressed as: #() = 89:; ) ?(, A9:; )B()

(3)

where ФPOM is the POM volume fraction, VPOM the volume of the POM ion with is the electronic contrast, which

equals to the difference in scattering length densities between POM and water, >POM and >water respectively. P(q, RPOM) is the sphere form factor and S(q) is the structure factor that accounts for POM-POM interactions in solution. The plot of I(q)/ФPOM vs. q (see Fig. S4) shows first a superposition of the spectra at large qvalues (q > 2 nm-1). In this q-range the form factor of POM, P(q), is probed and is independent of POM concentration. Second, the decrease in scattered intensity at low q-values (q < 2nm-1), which gets stronger as the POM concentration increases, is related to the structure factor and can be attributed to growing electrostatic repulsions between POMs. Therefore, the shape of the spectrum is only related to P(q) when S(q) equals to one, i.e. when POM-POM repulsive interactions are negligible. This condition is met either when the POM is diluted or when salt is added at a concentration sufficiently high to screen fully electrostatics repulsions. The SAXS instrument is not sensitive enough to obtain a good scattering signal at high POMs dilution. Therefore, SAXS spectra of 10 mM H4SiW solutions were collected at different salt concentrations, see Fig. 3 (b). The increase in the scattered intensity at low q-values is observed by adding salt until saturation ([NaCl] > 100 mM), indicating the full screening of the repulsive interactions between POMs (Fig. 3 b). The scattering curve at 250 mM NaCl can be well fitted using a P(q) of homogeneous

ACS Paragon Plus Environment

22

Page 23 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

spheres characterized by a radius of A9:; = 0.46 nm, as already determined in previous studies. Then, S(q) were obtained for different POMs and salts concentrations by dividing all the scattering spectra by 89:; ) ?(, A9:; = 0.46 IJ) (see Fig. 3 c and d). As expected from repulsive particles, the structure factors equal to 1 at large q-values and decreases at lower qvalues. These profiles were analyzed considering a pair potential of spherical charged particles such as the one developed by Verwey-Overbeek66 and by using the well-known Hayter mean spherical approximation (see Fig. 4).67 This model requires 4 parameters: the effective charge of the colloid (zeff), an effective radius of the sphere (Reff), the volume fraction of particle Ф and the ionic strength (or salt concentration for monovalent salt) I. zeff and Reff were fitted, whereas Ф and I were considered as input parameters and set to their experimental values. An attempt was made to fit the experimental S(q) by the rescaled MSA model but it gave much less agreement, being always out of the model convergence.

ACS Paragon Plus Environment

23

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 43

ACS Paragon Plus Environment

24

Page 25 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Figure 3. SAXS spectra of H4SiW at different concentrations: 10, 15, 20, 25, 30 and 35 mM (a) and at a fixed concentration of 10 mM varying the NaCl concentration: 25, 50, 75, 100 and 250 mM (b). The corresponding S(q) profiles are in (c) and (d). The fit of the S(q) for [H4SiW] = 10 mM without added salt shows a good agreement with the experimental data, see Fig. 4 (a), with only two fitting parameters, zeff = 3.6 and Reff = 0.77 nm. These values are reasonable compared to the nominal ones, zPOM = 4 and RPOM = 0.46 nm. For higher POM concentrations (see Fig. 4 b), the adjustment of the model to the experimental S(q) is more critical and it is then difficult to achieve the convergence of the fit even if similar zeff and Reff values as for the system at 10 mM are obtained from the fitting process (see Table 3). When the

ACS Paragon Plus Environment

25

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 43

ionic strength is varied (see Fig. 4 c), the model fits much better to the experimental data. However, a slight increase in the effective charge as well as a more pronounced variation of the size is found by increasing salt concentration. This deviation from reasonable Reff values shows here the limit of this model for nano-objects characterized by a very high surface charge density in highly diluted conditions. It is also possible that some part of the POM forms nano-aggregates, as noticed by Bera et al. in a previous contribution (only) for H3PW in highly acidic medium.53 The presence of small amount of POM nano-assemblies could also explain the deviation from reasonable Reff values, compared to the radius of a POM anion. Table 3. Parameters resulting from the fits of S(q), obtained from SAXS, by using a pair potential of spherical charged particles within the Hayter mean spherical approximation.

[H4SiW] [NaCl] mM

mM

Fitting

Input

parameters

parameters

zeff

Reff

I

Ф

nm

mM

%v/v

10

0

3.6

0.77

0

0.52

20

0

4

0.77

0

1.04

30

0

4.4

0.78

0

1.56

10

25

3.8

0.86

25

0.52

10

100

4.4

1.05

100

0.52

ACS Paragon Plus Environment

26

Page 27 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Figure 4. Structure factors as a function of q for a) [H4SiW] = 10 mM in milliQ water, data were fitted using MSA Hayter model with two fitting parameters, the effective charge (zeff = 3.6) and the effective radius (Reff = 0.77 nm) setting the POM volume fraction and salt concentration to their experimental values. b) [H4SiW] = 20 and 30 mM in milliQ water, the fitting process gave

ACS Paragon Plus Environment

27

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 43

zeff = 4 and 4.4 and Reff = 0.77 and 0.78 nm respectively. c) [H4SiW] = 10 mM with [NaCl] = 25 and 100 mM, the fitting process gave zeff = 3.8 and 4.4 and Reff = 0.86 and 1.05 nm respectively. These results are in agreement with the ones obtained by DLS experiments and point out that the electrostatic repulsions between the POMs are almost completely screened for salt concentrations above 100 mM. Moreover, it is confirmed that (i) POMs are fully dissociated from their counterions, here H+, as already concluded from previous conductivity measurements,33 and that (ii) POMs behave as highly charged nano-colloids, as deduced in the previous section. The SAXS analysis was also applied for H3PW (Fig. S5 a). The fit of the scattering spectrum of the full screened system, 10 mM of POM with 100 mM NaCl, using a sphere form factor, gives a radius of 0.47 nm. This result is very similar to the one obtained for H4SiW, as expected from the isostructure of PW3- and SiW4- and in agreement with the literature.22, 27, 68 Meanwhile, the structure factor profiles differ from the ones obtained with H4SiW, see Fig. S5 (b), and show two steps versus q whatever the ionic strength in the system. Addition of salt leads to a screening effect of the electrostatic repulsions as for H4SiW. However for H3PW, a salt concentration of 100 mM is sufficient to obtain a full screening, meaning that the effective charge of H3PW is weaker than for H4SiW, as expected from the lower formal charge of PW3- compared to SiW4-. The peculiarity in the profile of the H3PW structure factors may be related to POMs cluster formation, that could be at the origin of the slow mode observed in DLS. POMs as (super-chaotropic) anions: Adsorption on neutral micellar surfaces In a recent study by Naskar et al. it was shown that POMs, H3PW and H4SiW, spontaneously adsorb at the surface of non-ionic micelles.22 The POM adsorption process on the micelle affect the monomer-micelle equilibrium as observed in the evolution of the critical micellar

ACS Paragon Plus Environment

28

Page 29 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

concentration (cmc) of the surfactant by addition of POM, see the results for C8G1 with H3PW and H4SiW in Fig. S6. H4SiW slightly increases the cmc, whereas H3PW has the opposite effect. This result suggests that H4SiW shifts the monomer-micelle equilibrium towards the formation of monomer, whereas H3PW favors micellization due to a stronger affinity of PW3- for the surface of the micelle. Indeed, the investigation of the micelles by SAXS and the variation on the cloud points of a polyethoxy surfactant with POMs have concluded that PW3- adsorbs more strongly on the micellar surface than SiW4-.22 In the present work, we investigate the evolution of the interactions between H3PW and H4SiW with C8G1 micelles, using 31P NMR, 31P and 1H DOSY NMR for H3PW and 1H DOSY NMR for H4SiW. The advantage of combining 31P and 1H DOSY NMR is to obtain information on the dynamics of both H3PW and C8G1. For H4SiW only 1H DOSY could be conducted on C8G1. DOSY NMR informs here on the dynamics of the systems, i.e. self-diffusion of the POM and the surfactant micelle, and enables to investigate the effect of adsorption on the self-diffusion of the POM. The chemical shift of the 31P NMR signal of PW3- for [H3PW] = 30 mM was recorded with increasing C8G1 concentration from 0 up to 500 mM, see Fig. 5 (a). In this concentration range C8G1 mostly form micelles, as the cmc ranges from 10 to 26 mM (Fig. S6), and PW3- adsorbs on the micelles, as clearly monitored by SAXS (see Fig. S7). The SAXS intensity profiles indeed show large oscillations indicating the presence of the core-shell. The NMR signal becomes constant for [C8G1] > 200 mM, meaning that above this concentration the chemical environment around the POM does not change with further addition of the surfactant. This indicates that all POMs are adsorbed on micelles when [C8G1] > 200 mM.

ACS Paragon Plus Environment

29

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 43

Therefore at this specific concentration, [C8G1] = 200 mM, the average C8G1/PW3- ratio (6.7) corresponds to the lowest accessible ratio in micelles, i.e. the maximum coverage of PW3- on the micellar surface. This value is in good agreement with the C8G1/PW3- ratios determined by SAXS (4.3) and by ion flottation experiment (5.4) in a previous study.22 As a consequence when the C8G1 concentration increases above 200 mM, the surfactant/PW3- ratio increases which means that the micelle surface charge, brought by the adsorbed PW3-, decreases.

ACS Paragon Plus Environment

30

Page 31 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Figure 5. (a) Chemical shift of the

31

P peak from PW3- as a function of C8G1 concentration. (b)

Fraction of H3PW adsorbed on the micellar surface versus added C8G1 concentration (i) experimental data estimated using Eq. (4) (black squares), (ii) fit for the K using Langmuir isotherm (red line) and (iii) fit obtained from the Langmuir isotherm substituting K stemmed from the 1:1 binding model (blue line). 0LM The mole fraction of PW3- adsorbed at the micellar surface, K9:; , can be simply expressed as a

function of the chemical shift as: (N3N )

0LM K9:; = (N

OPQ. 3N )

(4)

by assuming that (i) the chemical shift without C8G1 (δ0 = 15.298 ppm) corresponds to the free PW3- and that (ii) the maximum chemical shift obtained for [C8G1] > 200 mM (δads. = 15.427 ppm) corresponds to PW3- adsorbed at the micellar surface. 0LM By assuming that K9:; corresponds to the surface coverage on the micelles by POM, with 0LM 0LM K9:; = 1 in the case of maximum coverage, then a plot of K9:; vs. [C8G1], see Fig.5 (b), can be

assimilated to an adsorption isotherm. A simple Langmuir model for monolayer adsorption was 0LM applied to fit the K9:; vs. [C8G1] curve, see Fig.5 (b). However the significant deviation from the

Langmuir model, obtained at low surfactant concentrations, is likely to arise from PW3--PW3electrostatics repulsions at the micellar surface. Therefore the best fit gave only a rough estimate of the adsorption constant (K ≈ 40 M-1) of PW3- on the micellar surface. Moreover, at low surfactant concentrations the surfactant monomers, involved in the monomer-micelle equilibrium, may compete and decrease artificially the adsorption constant. This latter effect should vanish at higher surfactant concentrations well above the cmc (10 mM) i.e. [C8G1] ≫ cmc.

ACS Paragon Plus Environment

31

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 32 of 43

A 1:1 steochiometric binding model69 (see SI for more details) was also applied and gave a good fit of the data with K = 9 M-1, see Fig. S8. However this model assumes the formation of 1:1 PW3:C8G1 complex which is not relevant here as no well-defined stoechiometry is expected. Nevertheless, this model gives the order of magnitude of the association constant. The simulation of a Langmuir model with K = 9 M-1 was plotted in Fig.5 (b) (blue line). It produces fPOM values much lower than the ones obtained experimentally for the high surfactant concentrations but a good agreement is obtained in the low surfactant concentrations. Consequently, the adsorption constant of PW3- on C8G1 micellar surface lies in the range of 9 < K < 40 M-1 confirming that POMs strongly adsorb on polar neutral surfaces in aqueous phases, as it was qualitatively demonstrated by SAXS in a previous study.22 The diffusion coefficient values of PW3- (T9U V6 ) and C8G1 (T4W XY ) as a function of surfactant concentration, for 30 mM of H3PW, obtained respectively by using 31P and 1H DOSY NMR, are presented in Fig. 6. T9U V6 strongly decreases from 3.48 10-10 to 7.5 10-11 m2/s for 0 < [C8G1] < 200 mM and then it only slightly decreases for [C8G1] > 200 up to a value of 3.67 10-11 m2/s for [C8G1] = 300 mM. This result is in full agreement with the evolution in the 31P chemical shift (Fig.5 a) which informed on the maximum coverage of micelles by PW3- for [C8G1] > 200 mM i.e. for C8G1/PW3- > 6.7. The evolution of T4W XY by increasing [C8G1] has a similar shape as for T9U V6 , see Fig. 6. However, T4W XY is much smaller than T9U V6 for [C8G1] → 0, as expected from the large difference in size between C8G1 micelles (R > 2 nm)70-71 and PW3- (R = 0.46 nm). By increasing [C8G1] from 0 to 200 mM, T9U V6 becomes comparable to T4W XY indicating that PW3sticks strongly on the micelles. The diffusion coefficients of C8G1 in water, i.e. without PW3-, were also determined for comparison (Fig.6 inverted triangles) and show a decrease by increasing [C8G1]. Nilsson et al. have reported that the decrease in the diffusion coefficient was related to the

ACS Paragon Plus Environment

32

Page 33 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

increase in the C8G1 micelles size by increasing the surfactant concentration.71 However, it was also shown that the adsorption of PW3- on C8G1 micelles leads to decrease the micellar size and to form spherical micelles.22 The fraction of POM adsorbed on the surface of micelles can be estimated from the PW3diffusion coefficient, with the same assumptions made for Eq. (4): (3 )

0LM K9:; = (

OPQ. 3 )

(5)

where D is the diffusion coefficient of PW3- at a given C8G1 concentration, D0 is the diffusion coefficient of PW3- without surfactant and Dads. is the diffusion coefficient of PW3- when all the POM is adsorbed on the micelles. The simulations of the D vs. [C8G1] curves by using the Langmuir isotherm model with the K values found above, 9 and 40 M-1, was added in Fig.6. The agreement between the two simulation curves and the experimental data is comparable to the one obtained above for the fit of the 31P chemical shifts. This result indicates that the range of K values obtained above is valid. Therefore, the evolution in the 31P chemical shifts and in the C8G1/PW3diffusion coefficients with [C8G1] are in good agreement.

ACS Paragon Plus Environment

33

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 34 of 43

Figure 6. Diffusion coefficients of PW3- (dark squares), C8G1 in the presence of [H3PW] = 30 mM (red cicrles), and C8G1 in pure water (magenta inverted triangles) as a function of surfactant concentration. Simulations of D using Eq. 5. with two different values of adsorption constants: K = 9 M-1 (solid line) and K = 40 M-1 (dashed line). All data were obtained at 25 ˚C. Comparison H4SiW and H3PW. The diffusion coefficients of C8G1 (T4W XY ) as a function of [C8G1] in the presence of 30 mM H4SiW and 30 mM H3PW are presented in Fig. 7. Both H4SiW and H3PW lead to decrease the C8G1 diffusion coefficent but the decrease is more pronounced with H3PW. As already stated, H3PW has a stronger affinity for the C8G1 micelle surface than H4SiW.22 This difference in their surface affinity is reflected in the evolution of the C8G1 cmc by addition of POMs, see Fig. S7: (i) H4SiW shifts the monomer – micelle equilibrium towards the formation of monomers, i.e. it increases the cmc, whereas (ii) H3PW, due to its higher affinity to the micellar surface, decreases the cmc, then promoting micelle formation. Moreover the T4W XY values, as measured by 1H DOSY NMR, are averaged values between the contributions of C8G1 as monomers (T4Z[\[Z] ) and as micelles (T4Z^_55 ), with T4Z[\[Z] > T4Z^_55 due to the larger W XY W XY W XY W XY size of the micelle compared to the monomer. Therefore, the contribution of monomers to T4W XY is more significant in the low surfactant concentration range, [C8G1] < 200 mM, for H3PW compared to H4SiW. This effect explains the origin of the stronger decrease in the diffusion coefficient oberved with H3PW compared to H4SiW. For higher surfactant concentrations, [C8G1] > 200 mM, the contribution of monomers to T4W XY becomes negligible for both H3PW and H4SiW. However the T4W XY values remain lower with H3PW than with H4SiW. This difference is likely to arise from the higher charge of the micelles covered by SiW4- compared to PW3-. Indeed, stronger repulsive interactions between micelles, arising from the POMs adsorbed at their surface, are expected to increase the micelle diffusion coefficient. Unfortunately, zeta potential measurement,

ACS Paragon Plus Environment

34

Page 35 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

giving information on the micellar surface charge, could not be conducted because of electrochemical reactions of POMs at the electrodes, resulting in the formation of blue dark coloration due to the presence of reduced POMs species.

Figure 7. Diffusion coefficients as a function of surfactant concentration for 30 mM H3PW without surfactant (black squares); C8G1 in the presence of 30 mM H3PW (red squares); C8G1 in the presence of 30 mM H4SiW (blue squares) and C8G1 alone (magenta inverted triangles). All data were obtained at 25 ˚C. CONCLUSIONS The POM-POM interactions as well as the adsorption of POM on neutral micellar surface were studied by combining dynamic and static approaches. This comprehensive analysis demonstrated the importance of electrostatic repulsions between POMs and its impact on the dynamic behaviour of POM anions in aqueous solutions. It was shown that the self-diffusion of SiW4- and PW3-, at different POM/salt concentrations, could be rationalized on a master, based on electrostatic considerations and without taking into account POM’s counterions in the Debye length

ACS Paragon Plus Environment

35

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 36 of 43

calculation, as for classical charged colloids. The electrostatic inter-POM interactions obtained by SAXS confirmed that a classical colloidal approach can be used to describe POMs in solution. The strength of the spontaneous adsorption of POMs on polar surfaces, i.e. their behavior as (chaotropic-)anions, was also evaluated by fitting the combined data obtained by 1H/31P NMR and 1

H/31P DOSY NMR by adsorption isotherm models. This confirmed that the stickiness of POMs

towards polar (hydrated) interfaces is very high and that POMs can be classified as superchaotrope. As a conclusion Keggin’s POMs has a dual behavior as anions and as colloids: (i) POMs behave as anions, because they adsorb at surfaces as chaotropic anions, but (ii) POMs can also be considered as colloids, from their size (close to 1nm) which lies at the border of the IUPAC definition of colloidal systems and they can therefore be investigated using classical experimental approaches for charged colloids e.g. by investigating POM-POM structure factor in solution determined by SAXS. ASSOCIATED CONTENT (Word Style “TE_Supporting_Information”). Supporting Information. The Supporting information is available free of charge on the ACS Publication website. The following files are available free of charge. Dynamic light scattering (PDF) Debye length (PDF) SAXS measurements (PDF) Adsorption constant. 1:1 Binding model (PDF) AUTHOR INFORMATION Corresponding Author

ACS Paragon Plus Environment

36

Page 37 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

* E-mail: [email protected]. Present Addresses †If an author’s address is different than the one given in the affiliation line, this information may be included here. Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. ‡These authors contributed equally. (match statement to author names with a symbol) Funding Sources Any funds used to support the research of the manuscript should be placed here (per journal style). Notes Any additional relevant notes should be placed here. ACKNOWLEDGMENT The authors would like to thank Pr. J-F. Dufrêche, Dr. S. Gourdin and Th. Buchecker for fruitful discussions. The national funding ANR-12-BS08-0017 (CELADYCT) supported this work. ABBREVIATIONS POM, polyoxometalate; SiW4-, silicotungstate; PW3-, phosphotungstate; H4SiW, silicotungstic acid; H3PW, phosphotungstic acid; C8G1, octyl-beta-glucoside; PEG, polyethylene glycol; DLS, dynamic light scattering; SAXS, small angle X-ray scattering; PGSE NMR, pulsed gradient spin echo nuclear magnetic resonance; DOSY NMR, diffusion ordered spectroscopy nuclear magnetic resonance; MSA, mean spherical approximation; MD, molecular dynamics.

ACS Paragon Plus Environment

37

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 38 of 43

REFERENCES 1. Hill, C. L., Introduction: Polyoxometalates - Multicomponent molecular vehicles to probe fundamental issues and practical problems. Chemical Reviews 1998, 98 (1), 1-2. 2. Antonio, M. R.; Chiang, M.-H.; Seifert, S.; Tiede, D. M.; Thiyagarajan, P., In situ measurement of the Preyssler polyoxometalate morphology upon electrochemical reduction: A redox system with Born electrostatic ion solvation behavior. J. Electroanal. Chem. 2009, 626 (12), 103-110. 3. Long, D.-L.; Tsunashima, R.; Cronin, L., Polyoxometalates: Building Blocks for Functional Nanoscale Systems. Angew Chem Int Edit 2010, 49 (10), 1736-1758. 4. Kozhevnikov, I. V., Catalysis by heteropoly acids and multicomponent polyoxometalates in liquid-phase reactions. Chemical Reviews 1998, 98 (1), 171-198. 5. Misono, M., Heterogeneous Catalysis by Heteropoly Compounds of Molybdenum and Tungsten. Catal Rev 1987, 29 (2-3), 269-321. 6. Oms, O.; Yang, S.; Salomon, W.; Marrot, J.; Dolbecq, A.; Riviere, E.; Bonnefont, A.; Ruhlmann, L.; Mialane, P., Heteroanionic Materials Based on Copper Clusters, Bisphosphonates, and Polyoxometalates: Magnetic Properties and Comparative Electrocatalytic NOx Reduction Studies. Inorganic Chemistry 2016, 55 (4), 1551-1561. 7. Bernardini, G.; Wedd, A. G.; Zhao, C.; Bond, A. M., Photochemical oxidation of water and reduction of polyoxometalate anions at interfaces of water with ionic liquids or diethylether. Proceedings of the National Academy of Sciences 2012, 109 (29), 11552-11557. 8. Clemente-Juan, J. M.; Coronado, E.; Gaita-Arino, A., Magnetic polyoxometalates: from molecular magnetism to molecular spintronics and quantum computing. Chemical Society Reviews 2012, 41 (22), 7464-7478. 9. Musumeci, C.; Rosnes, M. H.; Giannazzo, F.; Symes, M. D.; Cronin, L.; Pignataro, B., Smart High-κ Nanodielectrics Using Solid Supported Polyoxometalate-Rich Nanostructures. Acs Nano 2011, 5 (12), 9992-9999. 10. Xu, L.; Zhang, H.; Wang, E.; Kurth, D. G.; Li, Z., Photoluminescent multilayer films based on polyoxometalates. Journal of Materials Chemistry 2002, 12 (3), 654-657. 11. Zhang, H.; Lin, X.; Yan, Y.; Wu, L., Luminescent logic function of a surfactantencapsulated polyoxometalate complex. Chem. Commun. 2006, (44), 4575-4577. 12. Poulos, A. S.; Constantin, D.; Davidson, P.; Imperor, M.; Pansu, B.; Panine, P.; Nicole, L.; Sanchez, C., Photochromic hybrid organic-inorganic liquid-crystalline materials built from nonionic surfactants and polyoxometalates: Elaboration and structural study. Langmuir 2008, 24 (12), 6285-6291. 13. Pope, M. T.; Müller, A., Polyoxometalate Chemistry: An Old Field with New Dimensions in Several Disciplines. Angewandte Chemie International Edition in English 1991, 30 (1), 34-48. 14. Neumann, R., Applications of Polyoxometalates in Homogeneous Catalysis. In Polyoxometalate Molecular Science, Borrás-Almenar, J. J.; Coronado, E.; Müller, A.; Pope, M., Eds. Springer Netherlands: Dordrecht, 2003; pp 327-349. 15. Geletii, Y. V.; Hill, C. L.; Bailey, A. J.; Hardcastle, K. I.; Atalla, R. H.; Weinstock, I. A., Electron Exchange between α-Keggin Tungstoaluminates and a Well-Defined Cluster-Anion Probe for Studies in Electron Transfer. Inorganic Chemistry 2005, 44 (24), 8955-8966. 16. Hasenknopf, B., Polyoxometalates: introduction to a class of inorganic compounds and their biomedical applications. Front. Biosci. 2005, 10, 275–287.

ACS Paragon Plus Environment

38

Page 39 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

17. Liu, T., Hydrophilic Macroionic Solutions: What Happens When Soluble Ions Reach the Size of Nanometer Scale? Langmuir 2010, 26 (12), 9202-9213. 18. Santoni, M. P.; Pal, A. K.; Hanan, G. S.; Tang, M. C.; Venne, K.; Furtos, A.; MenardTremblay, P.; Malveau, C.; Hasenknopf, B., Coordination-driven self-assembly of polyoxometalates into discrete supramolecular triangles. Chemical Communications 2012, 48 (2), 200-202. 19. Giner-Casares, J. J.; Brezesinski, G.; Moehwald, H.; Landsmann, S.; Polarz, S., Polyoxometalate Surfactants as Unique Molecules for Interfacial Self-Assembly. J. Phys. Chem. Lett. 2012, 3 (3), 322-326. 20. Landsmann, S.; Wessig, M.; Schmid, M.; Cölfen, H.; Polarz, S., Smart Inorganic Surfactants: More than Surface Tension. Angewandte Chemie International Edition 2012, 51 (24), 5995-5999. 21. Buchecker, T.; Le Goff, X.; Naskar, B.; Pfitzner, A.; Diat, O.; Bauduin, P., Polyoxometalate/Polyethylene Glycol Interactions in Water: From Nanoassemblies in Water to Crystal Formation by Electrostatic Screening. Chem. - Eur. J. 2017, 8434-8442. 22. Naskar, B.; Diat, O.; Nardello-Rataj, V.; Bauduin, P., Nanometer-Size Polyoxometalate Anions Adsorb Strongly on Neutral Soft Surfaces. Journal of Physical Chemistry C 2015, 119 (36), 20985-20992. 23. de Viguerie, L.; Mouret, A.; Brau, H. P.; Nardello-Rataj, V.; Proust, A.; Bauduin, P., Surface pressure induced 2D-crystallization of POM-based surfactants: preparation of nanostructured thin films. Crystengcomm 2012, 14 (24), 8446-8453. 24. Song, Y.-F.; McMillan, N.; Long, D.-L.; Thiel, J.; Ding, Y.; Chen, H.; Gadegaard, N.; Cronin, L., Design of Hydrophobic Polyoxometalate Hybrid Assemblies Beyond Surfactant Encapsulation. Chemistry - A European Journal 2008, 14 (8), 2349-2354. 25. Baker, L. C. W.; Pope, M. T., The Identical Diffusion Coefficients of Isostructural Heteropoly Anions - the Complete Independence of D from Ionic Weight. Journal of the American Chemical Society 1960, 82 (16), 4176-4179. 26. Pope, M. T.; Varga, G. M., Heteropoly Blues .I. Reduction Stoichiometries and Reduction Potentials of Some 12-Tungstates. Inorganic Chemistry 1966, 5 (7), 1249-&. 27. Poulos, A. S.; Constantin, D.; Davidson, P.; Imperor, M.; Judeinstein, P.; Pansu, B., A PGSE-NMR Study of Molecular Self-Diffusion in Lamellar Phases Doped with Polyoxometalates. Journal of Physical Chemistry B 2010, 114 (1), 220-227. 28. Lopez, X.; Nieto-Draghi, C.; Bo, C.; Avalos, J. B.; Poblet, J. M., Polyoxometalates in solution: Molecular dynamics simulations on the alpha-PW12O403- keggin anion in aqueous media. J Phys Chem A 2005, 109 (6), 1216-1222. 29. Leroy, F.; Miro, P.; Poblet, J. M.; Bo, C.; Avalos, J. B., Keggin polyoxoanions in aqueous solution: Ion pairing and its effect on dynamic properties by molecular dynamics simulations. Journal of Physical Chemistry B 2008, 112 (29), 8591-8599. 30. Grigoriev, V. A.; Hill, C. L.; Weinstock, I. A., Role of cation size in the energy of electron transfer to 1 : 1 polyoxometalate ion pairs {(M+)(Xn+VW11O40)}((8-n)-) (M = Li, Na, K). Journal of the American Chemical Society 2000, 122 (14), 3544-3545. 31. Ostroushko, A. A.; Tonkushina, M. O.; Korotaev, V. Y.; Prokof’eva, A. V.; Kutyashev, I. B.; Vazhenin, V. A.; Danilova, I. G.; Men’shikov, S. Y., Stability of the Mo72Fe30 polyoxometalate buckyball in solution. Russian Journal of Inorganic Chemistry 2012, 57 (9), 1210-1213.

ACS Paragon Plus Environment

39

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 40 of 43

32. Horky, A.; Kherani, N. P.; Xu, G., Conductivity and transport properties of aqueous phosphotungstic and silicotungstic acid electrolytes for room-temperature fuel cells. J Electrochem Soc 2003, 150 (9), A1219-A1224. 33. Olynyk, T.; Jardat, M.; Krulic, D.; Turq, P., Transport coefficients of an inorganic Brownian particle in solution: The tungstosilicate anion. Journal of Physical Chemistry B 2001, 105 (31), 7394-7398. 34. Olynyk, T. Coefficients de transport de particules browniennes en solution : approche experimentale et modelisation. Universit´e Pierre et Marie Curie - Paris VI, 2001. 35. Sures, D. J.; Serapian, S. A.; Kozma, K.; Molina, P. I.; Bo, C.; Nyman, M., Electronic and relativistic contributions to ion-pairing in polyoxometalate model systems. Physical Chemistry Chemical Physics 2017, 19 (13), 8715-8725. 36. Antonio, M. R.; Nyman, M.; Anderson, T. M., Direct Observation of Contact Ion-Pair Formation in Aqueous Solution. Angewandte Chemie International Edition 2009, 48 (33), 61366140. 37. Li, D.; Yin, P.; Liu, T., When Giants Meet Dwarves in the Same Pond — Unique Solution Physical Chemistry Opportunities Offered by Polyoxometalate Macroions. In Polyoxometalate Chemistry: Some Recent Trends, Scientific, W., Ed. 2013; pp 49-100. 38. Hyde, S. T.; Andersson, S.; Larsson, K.; Blum, Z.; Landh, T.; Lidin, S.; Ninham, B. W., The Language of Shape. Elsevier: Amsterdam, 1997. 39. Landsmann, S.; Wessig, M.; Schmid, M.; Coelfen, H.; Polarz, S., Smart Inorganic Surfactants: More than Surface Tension. Angewandte Chemie-International Edition 2012, 51 (24), 5995-5999. 40. Tan, C., Self-assembly and morphology change of four organic-polyoxometalate hybrids with different solid structures from 2D lamellar to 3D hexagonal forms. Journal of Molecular Structure 2017, 1130, 276-283. 41. Izzet, G.; Abécassis, B.; Brouri, D.; Piot, M.; Matt, B.; Serapian, S. A.; Bo, C.; Proust, A., Hierarchical Self-Assembly of Polyoxometalate-Based Hybrids Driven by Metal Coordination and Electrostatic Interactions: From Discrete Supramolecular Species to Dense Monodisperse Nanoparticles. Journal of the American Chemical Society 2016, 138 (15), 5093-5099. 42. Assaf, K. I.; Ural, M. S.; Pan, F.; Georgiev, T.; Simova, S.; Rissanen, K.; Gabel, D.; Nau, W. M., Water Structure Recovery in Chaotropic Anion Recognition: High-Affinity Binding of Dodecaborate Clusters to γ-Cyclodextrin. Angewandte Chemie International Edition 2015, 54 (23), 6852-6856. 43. Bauduin, P.; Prevost, S.; Farras, P.; Teixidor, F.; Diat, O.; Zemb, T., A Theta-Shaped Amphiphilic Cobaltabisdicarbollide Anion: Transition From Monolayer Vesicles to Micelles. Angewandte Chemie-International Edition 2011, 50 (23), 5298-5300. 44. Bauduin, P.; Zemb, T., Perpendicular and lateral equations of state in layered systems of amphiphiles. Current Opinion in Colloid & Interface Science 2014, 19 (1), 9-16. 45. Gassin, P.-M.; Girard, L.; Martin-Gassin, G.; Brusselle, D.; Jonchere, A.; Diat, O.; Vinas, C.; Teixidor, F.; Bauduin, P., Surface Activity and Molecular Organization of Metallacarboranes at the Air-Water Interface Revealed by Nonlinear Optics. Langmuir 2015, 31 (8), 2297-2303. 46. Everett, D. H., Manual of Symbols and Terminology for Physicochemical Quantities and Units, Appendix II: Definitions, Terminology and Symbols in Colloid and Surface Chemistry. In Pure and Applied Chemistry, 1972; Vol. 31, p 577. 47. Israelachvili, J. N., Intermolecular and Surface Forces Revised Third Edition. Academic Press 2011.

ACS Paragon Plus Environment

40

Page 41 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

48. Bishop, K. J. M.; Wilmer, C. E.; Soh, S.; Grzybowski, B. A., Nanoscale Forces and Their Uses in Self-Assembly. Small 2009, 5 (14), 1600-1630. 49. Liu, Z.; Liu, T.; Tsige, M., Elucidating the Origin of the Attractive Force among Hydrophilic Macroions. Scientific Reports 2016, 6 (1). 50. Liu, G.; Liu, T., Strong Attraction among the Fully Hydrophilic {Mo72Fe30} Macroanions. Journal of the American Chemical Society 2005, 127 (19), 6942-6943. 51. Kistler, M. L.; Bhatt, A.; Liu, G.; Casa, D.; Liu, T., A Complete Macroion−“Blackberry” Assembly−Macroion Transition with Continuously Adjustable Assembly Sizes in {Mo132} Water/Acetone Systems. Journal of the American Chemical Society 2007, 129 (20), 6453-6460. 52. Koelsch, P.; Motschmann, H., A Method for Direct Determination of the Prevailing Counterion Distribution at a Charged Surface. The Journal of Physical Chemistry B 2004, 108, 18659-18664. 53. Bera, M. K.; Qiao, B.; Seifert, S.; Burton-Pye, B. P.; Olvera de la Cruz, M.; Antonio, M. R., Aggregation of Heteropolyanions in Aqueous Solutions Exhibiting Short-Range Attractions and Long-Range Repulsions. The Journal of Physical Chemistry C 2016, 120 (2), 1317-1327. 54. Liu, T.; Diemann, E.; Li, H.; Dress, A. W. M.; Muller, A., Self-assembly in aqueous solution of wheel-shaped Mo154 oxide clusters into vesicles. Nature 2003, 426 (6962), 59-62. 55. Verhoeff, A. A.; Kistler, M. L.; Bhatt, A.; Pigga, J.; Groenewold, J.; Klokkenburg, M.; Veen, S.; Roy, S.; Liu, T.; Kegel, W. K., Charge Regulation as a Stabilization Mechanism for Shell-Like Assemblies of Polyoxometalates. Physical Review Letters 2007, 99 (6). 56. Yin, P.; Li, D.; Liu, T., Solution behaviors and self-assembly of polyoxometalates as models of macroions and amphiphilic polyoxometalate-organic hybrids as novel surfactants. Chemical Society Reviews 2012, 41 (22), 7368-7383. 57. Pusset, R.; Gourdin-Bertin, S.; Dubois, E.; Chevalet, J.; Mériguet, G.; Bernard, O.; Dahirel, V.; Jardat, M.; Jacob., D., Nonideal effects in electroacoustics of solutions of charged particles: combined experimental and theoretical analysis from simple electrolytes to small nanoparticles. PCCP : Physical chemistry chemical physics 2015, 17 (17), 11779-11789. 58. Pope, M. T., Introduction to Polyoxometalate Chemistry. In Polyoxometalate Molecular Science, Borrás-Almenar, J. J.; Coronado, E.; Müller, A.; Pope, M., Eds. Springer Netherlands: Dordrecht, 2003; pp 3-31. 59. Zhu, Z.; Tain, R.; Rhodes, C., A study of the decomposition behaviour of 12tungstophosphate heteropolyacid in solution. Canadian Journal of Chemistry 2003, 81 (10), 10441050. 60. Bressler, I.; Kohlbrecher, J.; Thunemann, A. F., SASfit: a tool for small-angle scattering data analysis using a library of analytical expressions. Journal of Applied Crystallography 2015, 48 (5), 1587-1598. 61. Russel, W. B.; Saville, D. A.; Schowalter, W. R., Colloidal Dispersions. Cambridge University Press: 1989. 62. Kunz, W., Specific Ion Effects. World Scientific: Singapore, 2009; p 348. 63. Ninham, B. W.; Yaminsky, V., Ion Binding and Ion Specificity:  The Hofmeister Effect and Onsager and Lifshitz Theories. Langmuir 1997, 13 (7), 2097-2108. 64. Kunz, W., personal communication. 65. Dahirel, V.; Ancian, B.; Jardat, M.; Mériguet, G.; Turq, P.; Lequin, O., What can be learnt from the comparison of multiscale brownian dynamics simulations, nuclear magnetic resonance and light scattering experiments on charged micelles? Soft Matter 2010, 6 (3), 517-525.

ACS Paragon Plus Environment

41

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 42 of 43

66. Verwey, E. J. W.; Overbeek, J. T. G.; van Nes, K., Theory of the Stability of Lyophobic Colloids: The Interaction of Sol Particles Having an Electric Double Layer. Elsevier Publishing Company: 1948. 67. Hansen, J. P.; Hayter, J. B., A rescaled MSA structure factor for dilute charged colloidal dispersions. Mol. Phys. 1982, 46 (3), 651-656. 68. Brown, G. M.; Noe-Spirlet, M.-R.; Busing, W. R.; Levy, H. A., Dodecatungstophosphoric acid hexahydrate, (H5O2+)3(PW12O403-). The true structure of Keggin's `pentahydrate' from single-crystal X-ray and neutron diffraction data. Acta Crystallographica Section B 1977, 33 (4), 1038-1046. 69. Thordarson, P., Determining association constants from titration experiments in supramolecular chemistry. Chemical Society Reviews 2011, 40 (3), 1305-1323. 70. Bauer, C.; Bauduin, P.; Girard, L.; Diat, O.; Zemb, T., Hydration of sugar based surfactants under osmotic stress: A SAXS study. Colloids Surf. A Physicochem. Eng. Asp. 2012, 413, 92-100. 71. Nilsson, F.; Söderman, O.; Johansson, I., Physical−Chemical Properties of the n-Octyl βd-Glucoside/Water System. A Phase Diagram, Self-Diffusion NMR, and SAXS Study. Langmuir 1996, 12 (4), 902-908.

BRIEFS (Word Style “BH_Briefs”). If you are submitting your paper to a journal that requires a brief, provide a one-sentence synopsis for inclusion in the Table of Contents. SYNOPSIS (Word Style “SN_Synopsis_TOC”). If you are submitting your paper to a journal that requires a synopsis, see the journal’s Instructions for Authors for details.

ACS Paragon Plus Environment

42

Page 43 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

TOC

ACS Paragon Plus Environment

43