Hybrid Colloids Made with Polymers

Abstract

Polymers adsorb onto nanoparticles, NPs, by different mechanisms. Thus, they reduce coagulation, avoid undesired phase separation or clustering, and give rise to hybrid colloids. These find uses in many applications. In cases of noncovalent interactions, polymers adsorb onto nanoparticles, which protrude from their surface; the polymer in excess remains in the medium. In covalent mode, conversely, polymers form permanent links with functional groups facing outward from the NPs’ surface. Polymers in contact with the solvent minimize attractive interactions among the NPs. Many contributions stabilize such adducts: the NP–polymer, polymer–polymer, and polymer–solvent interaction modes are the most relevant. Changes in the degrees of freedom of surface-bound polymer portions control the stability of the adducts they form with NPs. Wrapped, free, and protruding polymer parts favor depletion and control the adducts’ properties if surface adsorption is undesired. The binding of surfactants onto NPs takes place too, but their stabilizing effect is much less effective than the one due to polymers. The underlying reason for this is that surfactants easily adsorb onto surfaces, but they desorb if the resulting adducts are not properly stabilized. Polymers interact with surfactants, both when the latter are in molecular or associated forms. The interactions occur between polymers and ionic surfactants or amphiphiles associated with vesicular entities. Hybrids obtained in these ways differ each from each other. The mechanisms governing hybrid formation are manifold and span from being purely electrostatic to other modes. The adducts that do form are quite diverse in their sizes, shapes, and features, and depend significantly on composition and mole ratios. Simple approaches clarify the interactions among different particle types that yield hybrids.

1. Introduction

The number of articles reporting on hybrid colloids drastically increased in the last few years [1,2,3,4,5]; think of the variety of composites made up of different colloidal units joined together. The resulting entities are quite stable, irrespective of whether this stability is thermodynamic or kinetic in nature. Composites obtained by mixing polymers, surfactants, and/or nanoparticles in different proportions are well known. Detailed studies and applications are manifold and find extensive use in advanced technologies, taking advantage of the different morphologies peculiar to these systems [6,7,8].

When faced with such a research field, terms such as sticky particles [9] or other definitions [10] are used. The denominations ascribed to hybrids are taken from daily life, the previous scientific literature, or from other sources. The etymology of the term hybrid is unknown, although its meaning is common knowledge. Thus, a historical background of that term helps to clarify its actual meaning.

2. A Historical Background

The concept originates from the early stages of human civilization, when our ancestors started breeding animals and growing comestible plants [11]. These processes developed in China, India, Egypt, and in Mesopotamia about 8000 years ago. Our ancestors were impressed with the effects of hybridization on animals, which looked like magic to them. As a consequence, they invented mythology, and imagined a series of hybrids: sphinxes, medusae, minotaurs, chimerae, centaurs, and so forth. N.B. Sphinxes have a human head and a lion body; minotaurs have a bull head and a human body; chimerae are lion-bodied with a goat’s head protruding from their backs and a tail ending in a snake’s head; medusae have a head surrounded by snakes, decorated with boar tusks, and have bronze hands, golden wings, and petrifying flashing eyes; centaurs have four legs, a horse rump, and a human trunk thereon. All mythological hybrids had the contemporary attitudes to run, swim, and fly, and conflicting characteristics such as wings and horns, or beaks and hoofs [12]. Our ancestors believed that hybrids were proof of the power, or wrath, of the gods. The concept was retained, revised, and widened over time. In the XVIth–XVIIIth centuries, for instance, hybrids were collected in “wunderkammers”; in addition, painters such as Hieronymus Bosch (1453–1516) drew some hybrid entities.

At the end of the XVIth century, Ulysses Aldrovandi (1522–1605), Professor of the Natural Philosophy chair at Bologna, described some of them [13]. The term was later used by Linnaeus in the “Systema Naturae, Sive Regna tria Naturae Systematice Proposita per Classes, Ordines, Genera, & Species” in 1735. Linnaeus applied the term to botany and zoology [14]. Its use in chemistry is controversial, but the resonance of benzene [15] and Pauling’s LCAO-MO theory for atomic and molecular orbitals (dating back to 1939) are defined on similar grounds.

Biology barriers, be they genetic, physiological, or embryonal in nature, hinder crossbreeding. The onset of hybrids, thus, is possible only among very closely related animal species. In colloid sciences, conversely, hybrid formation may occur among entities differing in their chemical natures, packing modes, compositions, and so on. It could have been troublesome to introduce this concept at the early beginning of colloid sciences, since scientists ruled out the possibility of joining surfactants, inorganic species, and polymers into a whole class [16]. The modern definition came later, thanks to Derjaguin, Landau, Verwey, and Overbeek, and ended in the so-called DLVO theory [17,18,19,20,21]; a clear-cut definition of soft matter colloids was proposed even later [22]. It is, thus, evident that tendencies in science move from complexity to simplicity, and back to complexity once our knowledge has clarified some obscure aspects.

3. What Are Hybrid Colloids?

What we actually know about hybrid colloids includes the following:

(1)

Single domains show diverse features, each with respect to others;

(2)

Their components self-organize, self-sustain, self-help, form composites, and follow the constraints dictated by thermodynamic or kinetic stability;

(3)

The domains in a hybrid are held together by covalent or noncovalent forces. The composites retain some properties of the parent entities, but also gain completely new ones.

The map in Figure 1 gives an indication of the combinations among different species leading to the formation of hybrids. Binary and three-component systems are possible.

Hybrids are stabilized from the combination of van der Waals (vdW), electrostatic, osmotic, and steric forces, and so forth. These facts allow to hybrids to assume many physical forms. Hybrids find uses in electronic materials, surface coatings, catalysis, gene therapy, food sciences, bioanalytical methods, molecular medicine, etc. [23,24,25,26].

Hybridization is strongly recommended since naked colloids almost always adsorb ions, solvents, interact with each other and with hosts [27], self-associate [28], and phase-separate [29]. These processes are detrimental for most practical purposes and must be avoided.

To mind but a few, let us note functionalized nanoparticles, polymer–surfactant systems, coated vesicles, colloids hold together by host–guest interactions, and biological entities, as well. Hybrids are built up using “bottom-up” strategies. They are made by two or more components, as indicated in the sections below; the forces governing their onset are briefly outlined.

4. Hybrids Containing Surfactants and Polymers

The objects are obtained by adsorption of one species onto formerly existing colloid units. This is what happens in polymer–surfactant systems, where the surfactant adsorbs on the polymer. Surfactant ions bind onto selected loci of (bio)polymers until new entities are attained. The “string of beads” structures obtained accordingly are termed polymer–surfactant systems, reported in Figure 2. Conversely, vesicles surface-adsorb polymers. The latter pathways are controlled by interactions among polar head groups facing outside vesicles and the polymer: generally, they are electrostatics-controlled [30]. The two subclasses, thus, are substantially different, and are described as separate items.

5. Polymer–Surfactant Systems

A comprehensive description of polymer–surfactant systems dates back to Goddard and Ananthapadmanabhan [31]. Thereafter, many researchers focused on these items. According to a huge number of experiments, uptake of surfactant ions onto polymer sites continues up to saturation of the macromolecular chain. Small aggregates, emi-micelles [32], nucleate and grow on the polymer chain until site binding is completed. The process largely modifies the polymer nature, which unrolls and changes its conformation. The resulting complexes do exist in the whole interaction region [33]. The horizontal line in Figure 2 indicates where the above composites dominate. The phase map is a pseudo-binary diagram, and is commonly used in dilute regimes [31].

Detailed studies place in evidence the presence of two different concentration regimes, with occurrence of physical discontinuities at the cac (critical association concentration) and at the cmc* (critical micellar concentration in presence of saturated polymer). Typical examples come from surface tension plots (Figure 3). In terms of efficiency, anionic surfactants are much more effective in binding compared to cationic; nonionic are not used.

In thermodynamic terms, the binding of surfactant ions on polymers is a multiple equilibrium, with adsorption of amphiphiles onto a chain having several adjacent sites (over 100, usually); very presumably, the process is strongly cooperative. Significant consequences are concomitant to that process; for instance, the kinetics of surfactant exchange to/from the polymer binding sites is nearly constant [34]. This fact is in agreement with the equilibrium hypothesis depicted above. It is immaterial if surfactants associate in small entities; it is the affinity towards the polymer sites that is responsible for the mentioned behavior. Surfactant binding onto proteins occurs nearly in the same way, as observed in SDS uptake on the hydrophobic tasks of albumin [35,36].

Polymer–surfactant systems are efficient viscosity modulators and detergents. Blends of the two components are combined in several pharmaceuticals (see the leaflets in most pharmaceutical boxes), or in nappies. The related formulations, usually prepared in nonaqueous gel form, adsorb significant amounts of urine; the volume of the final gels may increase more than ten times compared to the original conditions.

6. (Bio)Polymer Binding onto Preformed Surfactant Aggregates

The process is a surface nucleation and depends on (bio)polymer content and on the surface charge density of aggregates, which are natural, or synthetic, vesicles. The latter are by far preferred by experimentalists for many practical reasons. First, the stoichiometry is well defined. Second, the area on which proteins bind is considerably large [37]. Third, the surface charge density of synthetic vesicles can be modulated at will [38]; this holds true in synthetic cat-anionic vesicles, which can be negatively/positively charged [39,40]. These entities are built up by mixing oppositely charged surfactants in proper amounts, provided the mixture is not stoichiometric; in the latter eventuality, charge neutralization implies precipitation. The molecular surfactant concentration in cat-anionic dispersions is much lower than the one pertinent to the corresponding micellar solutions. N.B. The term CAT-ANs is an acronym for mixtures made of cationic plus anionic species. Possible surfactants binding to proteins, which could be a secondary reaction, is not significant. Such considerations come from statements of H. Wennerstroem [41] and C. Pucci [42]. Binding induces noticeable changes in the size of vesicles and large modifications in the net surface charge density of the adducts. Let us mention, as an example, BSA (bovine serum albumin) binding onto positively charged vesicles. The process is strongly pH-sensitive, because of charge modulation of the protein. For pH lower than the protein isoelectric point, BSA is positively charged, and no interactions take place. BSA has an isoelectric point close to 4.5 and is negatively charged above that threshold. If negatively charged CAT-AN vesicles interact with positively charged proteins, the reverse behavior holds true. In all cases, the width of the interaction region and the adducts sizes that do form largely increase compared to bare vesicles; see, for instance, Figure 4. Such vesicular dispersions find extensive use in biopolymer transfection technologies, as demonstrated by in vitro and in vivo studies.

As shown above, systems made up of polymers and surfactants largely differ each from the other; the preparation sequence controls the formation and properties of adducts that do form. Not only the stoichiometry but, also, the intermediate pathways are relevant in the fate of the objects that are formed. Not much is known about the possibilities offered by forming vesicles onto polymer surfactant systems, which could be a good demonstration of the strict similarities existing among the two subclasses mentioned above.

7. Hybrids Made of Nanoparticles

Covalent and noncovalent functionalization modes give rise to such hybrids. The noncovalent ones occur through physical adsorption of polymers, or surfactants, on NPs. The above two adsorbents share several points in common, and substantial differences. Surfactants easily adsorb, but tend to be released the same; therefore, polymers are by far preferred. Polymer adsorption occurs by NP dipping, eventually with co-solvents. Such preparation procedures date back to old times. Consider that kajal, decorating the lashes and brows of Queen Nefertiti [43], is a very fine powder obtained by mixing galena, malachite, and antimony, and dispersing it in acacia resin, a polysaccharide. Kajal is, thus, a mixture of three different hybrids, and is still in use for aesthetic and eye-disinfectant purposes.

Surface coverage prevents NP coalescence (Figure 5). Polymer-wrapped NPs, PW–NPs, are relatively stable and remain as such for a long time. They are obtained in different ways: information on the preparation is outside our purposes, and is given in selected articles and reviews [44,45,46,47,48].

Hybrids are formed by silica and titania nanoparticles, carbon-based entities, fullerenes, graphenes, carbon nanotubes, glass spheres, polymeric nanoparticles, such as PMMA (polymethyl–methacrylate), inorganic oxides, metals (including gold and silver), and so on. Among the many possibilities inherent to the above particles, let us refer to silica, titania, and carbon nanotubes (CNTS). The substantial differences among them are related to their different physical properties. Silica and titania are formed by reaction of their alcoxides with ammonia in ethanol. Bare CNTS, conversely, have very high aspect ratios (mostly if they are single-walled), are strongly hydrophobic, and form bundles in polar media. To reduce their aspect ratios, they are functionalized [49,50]. On this purpose, CNTS are reacted with H₂SO₄-HNO₃ mixtures, or with H₂O₂. As a consequence of drastic oxidation, the axial ratios reduce significantly and polar groups, usually carboxylates, are introduced on the outer CNTS surface. Further functionalization modes are possible by adsorption of oppositely charged species or by covalent reaction with proper species.

Physical wrapping is concomitant to polymer partition with the bulk; depletion may occur [51,52,53,54]. The latter phenomenon is due to unbalanced osmotic effects and dispersant flow from regions between the adducts; it depends on the polymer volume fraction, φ, and on that of NPs. Depletion was observed in solutions of homo-polymers, co-polymers, poly-electrolytes, polymer–surfactant systems, surfactants, and their mixtures.

Wrapping is controlled by electrostatic, steric, hydrophobic, vdW forces, and many combinations thereof [55]. Wrapping gives rough surfaces, with solvent molecules located in the polymer layers. Some polymer portions surface-adsorb; others protrude to the bulk. The protrusion probability is high; cases are known where extended coronas self-assemble around NPs [56,57]; the wrapped layer can be some nanometers thick and uniform. Wrapping is dense or loose, depending on the polymer affinity towards NPs, on the interaction strength, and on other factors.

Concentrated polymer regimes occur on the NPs’ surfaces. Some portions pack in compact domains; loose packing is possible. The transitions between two such regimes depends on the polymer content, and on its affinity towards NPs. The conformation of protruding groups is controlled by polymer–solvent and polymer–polymer interactions. Protruding portions repel each other, giving rise to different structures, as in PEO-based [58], or PPO-based systems. N.B. The term PEO stands for poly-ethylene-oxide, and PPO for poly-propylene-oxide. Wrapping minimizes biopolymers’ coagulation onto NPs [59]. Below are reported the quintessential aspects of wrapping. PEO wrapping onto silica NPs is discussed in more detail below, and an analytical solution to the problem is described. Interested readers may find the statistical thermodynamics model in References [60,61].

8. Polymer Wrapping

The [wrapping/protruding] ratio is >0; the partition implies that the two states differ in energy. Surface-bound segments are termed α (≤1), the protruding ones (trains) ε. X_b is the amount of bound polymer, which moves as a whole kinetic entity with NPs. Multilayer adsorption is possible. Wrapping can be analyzed by Langmuir-like isotherms [59]. It applies to SiO₂, TiO₂, carbon nanotubes, CNTs, graphenes, latexes, droplets, vesicles, etc. Wrapping holds, therefore, in both hard- and soft-matter matrices.

It is assumed that:

(a)

The polymer length is ≤NPs’ diameter, and wrapping length is lower than the extended polymer one. Different parts of the same polymer chain may wrap;

(b)

Polymers and NPs are size mono-disperse;

(c)

Polymer partition between bulk and surface states is possible, and can be substantial. The amount of surface-bound polymer is lower than X_tot, its overall mole fraction in the medium. This implies some partition with the bulk.

Polymer chains consist in loops, α, whose physicochemical properties differ from those of the protruding ones, ε; T is the number of segments in a single polymer chain. The interaction energy of NP–polymer systems differs from those of polymer–polymer and polymer–solvent ones. The solvent is ubiquitous in the coronas around NPs. From the amount of free and bound polymer, one obtains the equilibrium between α and ε states.

Polymer affinity towards NPs is related to the forces acting among protruding chains. The statistical model for homopolymers wrapping considers the chain as a ribbon of constant width. Several equivalent polymer units adsorb on the particle, and form domains containing l_i subunits (with l_i > 1). The adsorption energy obtained by the model, E_ad, is proportional to the number of wrapped segments. The weight of each state is calculated from the distribution function. For the most probable number of l_i values, a maximum in the probability occurs (Figure 6). Wrapping can be cooperative, as in ss-DNA wrapping onto CNTs [62], or when poly-electrolytes adsorb on oppositely charged NPs [63,64]. These processes occur because attractive terms among adsorbed chains are much higher than repulsive ones_, and depend, thus, on the polymer affinity towards NPs.

Different block co-polymers are used in wrapping. Pluronics F127, with its peculiar PEO–PPO–PEO sequence [65], is a good example. F127 has different blocks in separate parts of the chain; its sequence-sensitive HLB (hydrophilic–lipophilic balance) depends on the molecular details. In this system, wrapping is related to HLB, T, and to the PPO/PEO number ratios. Poly-(L-lysine)–PEO block copolymers show a similar behavior [66]; there, wrapping depends also on the length and charge of the polypeptide moiety.

The PEO–SiO₂ system is an excellent model to quantify wrapping efficiency. This is because both components are finely mono-dispersed. In the following, we consider SiO₂ sizes of 200 nm, and 50–60 EO units in the range; the polymer length is nearly equivalent to the NPs diameter. The conformational energy is hardly quantified. The hydration one equals the transfer energy from water to the NP surface, and is about 5.5 kJ per EO unit [67]. Wrapping can be continuous (the EO units adsorb with continuity) or random. The polymer releases water upon adsorption; the process is similar to micelle formation. The reason is that wrapping is a phase transition from a hydrated to a less hydrated regime. The number of surface bound units (usually 8–10 EO groups) comes from the maximum of the distribution function in Figure 6. For the PEO–SiO₂ system, the number of wrapping units gives an interaction energy, W, in the range of 45 ± 5.0 kJ mol⁻¹.

Polymers increase the medium osmotic pressure, modulated by interactions with NPs. At constant NP content, the osmotic pressure, π, is due to free polymer, free NPs, and adducts. From the osmotic coefficients [68], one obtains the Gibbs energy, which depends on composition and surface adsorption. At saturation, there are chemical potential changes in slope (Figure 7). The curve shape gives information on the quality of polymer uptake.

Appealing are hydrophobically modified polysaccharides, HMPs, adsorbing onto SiO₂ NPs [69]. Depending on the co-solvent, the polymer chains adsorb and form wrinkled particles. The final appearance of these adducts gives images looking like blackberries (Figure 8). Crucial in the above process is the polymer adsorption kinetics.

Ancillary aspects of PW–NPs systems imply depletion and hosts adsorption in the wrapped segments. The former process occurs when the volume fraction of free polymer, N°X_f/ρ_polym = φ_f, is higher than a critical value, dependent on polymer content. Hosts adsorption in the coronas occurs; the partition coefficients can be somehow obtained.

To obtain information on such systems, studies take advantage of experimental facilities focusing on selected aspects peculiar to these systems. Thus, the state of charge and interfacial effects can be quantified by Z-potential or dielectric relaxation properties, when sizes are determined by DLS, SAXS, or SANS, in addition to microscopies.

9. Hybrid Composites

The latter are hybrids built up by reacting, preformed hybrids. In some sense, they can be considered super-hybrids. The reactive pathways imply weak interaction modes, as in the case of ss-DNA/CNTs interacting with surfactant-covered TiO₂ [8]. There are cases in which robust chemical links between the single components take place [70]. To account for that behavior, polystyrene particles differing in size were used. The two families are 1.0 and 0.150 μ large, respectively. Large ones are streptavidin-coated, and the small ones are biotin-coated. The interactions between the two families lead to the formation of chemical links between them; the resulting entities become progressively larger and larger. The components were chosen in such a way to observe particles growth by optical microscopy and DLS. Particles growth depends on the number ratios among the two NP families; the adducts grow from nano- to micro- and, finally, to meso-scales.

A second procedure giving aggregates of hybrids relies on interactions governed by the surface charge density of the particles. As is well known, the electrical double layers around NPs may attract or repel. The most common case deals with entities of the same charge; it is also possible to obtain effective interactions among oddly charged entities. The latter are hardly rationalized from the Poisson–Boltzmann equation [71,72] (PB equation), which, therefore, must be modified. Accordingly, the electrical potential of the more charged colloids is transferred to the other particles. The equation is recalculated, taking into account the fact that the electric energy has been transferred [73]. This behavior was observed in some complex colloid mixtures and in biological systems as well [74,75].

It is also worth noting that hybrid colloids may form liquid crystalline phases, when the single particles self-organize to obtain nematic order in bulk [76], or in droplets [77]. In particular, ss-DNA-CNTs adducts dispersed in water/brine form nematic structures, characterized by a significant viscoelastic behavior. The transition from a random dispersion to an ordered, uniaxial, structure is assisted by the addition of polymer. Information on uniaxial order comes from ²H NMR spectral profiles. The formation of nematic droplets, conversely, is assisted by the addition of a positively charged protein, or a cationic surfactant, to ss-DNA-CNTs dispersions. A permanent, robust, peel is formed on the nematic domains, which remain confined in the aforementioned pellicle for very long times. Furthermore, such formulations resist osmotic pressure shocks.

10. Conclusions

Hybrids obtained by polymer adsorption onto NPs have a three-dimensional arrangement of the components. Very plausible scenarios for the practical application of hybrids imply depletion and binding of hosts; it is possible that PW–NPs interact with cells and biological tissues. Most such items have been described in selected articles, dealing, respectively, with the polymer organization in the coronas around NPs [78], and on their topology as well [79]. Other contributions deal with the organization modes of polymers, mostly PEO, onto the surfaces of NPs [80,81]. Even the biomolecular fate of properly functionalized NPs entering in biological tissues, including malignant ones, has been considered in some detail [82,83]. Future applications imply using more refined versions of the aforementioned adducts, bearing functionalities that allow their adhesion onto topical points of cells and tissues. These are the bases of modern molecular medicine approaches. A cogent analysis of such systems indicates that hybrids can be obtained by very different procedures. In terms of prime principles, the difference among physical and covalent adsorption is relevant.

The formation of hybrids can be explained in terms of statistical mechanics. Conceivably, the distribution functions responsible for physical adsorption are quite different compared to those of the covalent ones.

The reasons for using surface-covered hybrid NPs find origin in the fact that these entities always adsorb different molecules on their surface. Obviously, the effect is more significant the lower the size of NPs. This fact is detrimental in biomedicine. Consider that the surface adsorption of biopolymers onto injected raw NPs may lead to the formation of thrombuses and other diseases. To reduce such drawbacks, it is suggested to cover bare NPs. In particular, it was observed that PEO adsorbed onto NPs has an important role in reducing phagocytic uptake and plasma protein adsorption [44,75]. The reason is that the surfaces of PEO-functionalized NPs repel the adsorption of other biomolecules.

The approaches to wrapping presented here take into account most such physicochemical aspects. Presumably, the polymer affinity towards bare NPs is controlled by the interaction energies. These control the number of wrapping segments, packing density, and end at saturation. Studies reported to date suffer from the fact that the forces acting in the wrapped layer are badly defined. We are convinced that statistical thermodynamic approaches to wrapping will give more details on the process efficiency. Surely, the results will also benefit from more detailed experimental approaches.

Some important points deal with the formation of hybrids by chemical or direct interactions among preformed entities, or by properly modulating the electrostatic terms. It is conceivable that such interactions are hardly controlled, giving rise to entities of badly defined composition and/or shapes. Actual knowledge in the field does not allow a fine tuning of interactions among the components. It is possible, however, that “ad hoc” procedures shall be possible in a close future.

It is possible, too, that the most promising advantages shall arise in transfection technologies and molecular medicine. It must be kept in mind, however, that using bare nanoparticles in such technologies may be hazardous if the formulations are not stabilized and somehow made biocompatible.