Diverse Methods to Nanomanufacture Colloidal Dispersions of Polyaniline without Templates

: Different methods which could be used to produce colloidal dispersions of polyaniline (PANI) nano-objects without templates are described. While the methods are non-deterministic, different nano-objects (nanospheres, nanoﬁbers, nanobelts, nanorice, nanotubes, nanorods, nanodisks, etc.) can be produced. Those most used are: (i) solution polymerization with steric stabilizers (SPS) to produce nanospheres, (ii) interfacial polymerization (IP) to produce nanoﬁbers and (iii) solution polymerization in the presence of additives (SPA) to produce nanotubes. Oxidation of aniline in aqueous solution could produce nanotubes, nanoﬁbers and other shapes by controlling mass transport/concentration of reactants, pH, and the presence of oligomers/additives. The different models proposed to explain the formation of various nano-objects are discussed. Mechanochemical polymerization (MCP) could produce nanoﬁbers or nanospheres by controlling the aniline/oxidant ratio. PANI nanospheres of tunable sizes can also be produced by nanoprecipitation (NPT) of preformed PANI from its solutions using an antisolvent. The geometrical constraints to the small nano-objects made of high-molecular-weight rigid polymers are described. The conditions to produce nanostructures also affect the intrinsic properties of PANI (conductivity, crystallinity, and electroactivity). Selected technological applications of PANI nano-objects manufactured as colloidal dispersions without templates are discussed. Based on the reviewed work and models, future lines of work are proposed.


Introduction
Polyaniline (PANI) is the most widely studied and technologically used conducting polymer [1]. The history of conducting polymers began with the publication by Shirakawa et al. (1979), describing the huge increase in conductivity upon p-doping with the halogen of polyacetylene [2]. However, the products of aniline oxidation (so called "aniline blacks") were discovered by Runge in 1834 [3]. That PANI was a conductive polymer, electroactive and could be doped by protons was known in the 1960s [4]. Its role as a conducting polymer was rediscovered by MacDiarmid in the 1980s [5]. Unlike polyacetylene, aniline can be polymerized by oxidation in aqueous solution without catalysts and the resulting polymer is much more stable in air [6]. However, similar to other conducting polymers, it is insoluble in most common solvents and does not show Tg or Tm below decomposition. To use PANI in technological applications (anticorrosive coatings, electronic thin films, electrochromics, conductive ink, etc.), the polymer has to be processible. While thin films can be produced by in situ polymerization [7], the polymerization mixture is quite corrosive and only thin (<300 nm) films can be produced. With the advent of nanotechnology, a solution is at hand in the form of a colloidal solution of conducting nanoparticles. In addition to the technological applications as films, nanoparticles could be used for a variety of applications: to form multilayers, porous films for sensors, for photothermal therapy and even to produce N-doped carbon nanoparticles [8]. The published work on photothermal therapy and even to produce N-doped carbon nanoparticles [8]. The published work on polyaniline nanoparticles exceeds 4000 results. However, the vast majority involves the formation of mixed nanoparticles or nanocomposites together with metals [9], oxides [10], polymers [11], or nanocarbons [12]. Since the properties of those composite nanoparticles and nanocomposites depend on both components, this review only deals with the synthesis, characterization and applications of colloidal dispersions of PANI nanoparticles. One clear implication is that the large body of work on electrochemical polymerization of anilines will not be discussed [13]. Moreover, some widely used methods to manufacture PANI nanoparticles involves using porous hard membranes (e.g., alumina membranes) [14], or organized systems (e.g., micelles) as templates [15]. Those methods require a removal step for the template and, in principle, are deterministic in the sense that the shape/size of the nano-object is determined directly by the template.
On the other hand, the synthesis of nanotechnological species is dominated by selfformed non-deterministic systems, being metallic [16], silica [17] or polymeric [18] nanoparticles. Analogously, carbon nanotubes [19], graphene [20], and graphene-like materials [21] are made by non-deterministic processes. A similar set of methods, without templates have been shown to be effective [22], and even industrially useful [23] to produce PANI colloidal dispersions. Therefore, the published work on the nanomanufacturing of PANI as colloidal dispersions is analyzed in terms of the proposed mechanism of synthesis.
The main goal of the manufacturing of PANI nanomaterials is their use in technological applications. Since PANI is not soluble in water or common organic solvents, the manufacturing of colloidal dispersions enables producing a PANI conductive "ink" which can be used directly or as a way to form other structures, such as films. The electronic conductivity of PANI allows the manufacturing of electrochemical sensors, anticorrosion coatings, optoelectronic films, etc. The synergy of PANI conductivity with the large surface area of PANI nano-objects allows building double-layer supercapacitors where the intrinsic electroactivity of PANI adds a pseudocapacitance contribution. Moreover, the chemical nature of PANI gives a rich surface chemistry which could be used in adsorption of charged species. The technological applications of PANI nano-objects, produced as colloidal dispersions by template-free methods, will be summarily described. The advantages observed on the use of PANI in those applications will be discussed.

Template-Free Synthesis of PANI as Colloidal Dispersions
The oxidative polymerization of aniline begins in solution, but the growing chain becomes insoluble and precipitates by aggregation with other chains [24]. The formation of nanoparticle dispersion involves the formation of the solid form and the stabilization against aggregation. In metallic nanoparticles, the reduction of metal ions form the nuclei which grows and some molecules, which have stabilizing moieties such as charged groups or solvophilic polymer chains, adsorb on the formed nanoparticle, stabilizing it against aggregation. The usual shape obtained is granules that are more or less spherical. Since they are crystalline, rods with different aspect ratios (length/diameter) could be obtained by using additives which adsorb on some crystalline planes. The method can be used further to form very complex and useful shapes [25]. The size depends on the number of atoms and could be of a few nanometers for metal clusters [26]. Vinyl polymers have free rotation at the C-C bonds, allowing changes in conformations to form globules in the solid state which could be quite compact. The size of the globule does not depend directly on the size of the monomer unit [27].
On the other hand, if PANI chains are rigid, due to the extended conjugation of the quinonimine units N N , the aggregation of different chains should produce flat ribbons, the minimum of which size is the length of the chain. Therefore, from purely geometrical analysis, it is possible to calculate a minimum size of nanospheres of PANI or length of nanofibers. The molecular mechanics (MM2 [28]), package of Chem3D Ultra 8.0 ® was used to minimize the geometry of a fully oxidized PANI (pernigraniline base, PN) In a first view, these smaller nanoparticles could be formed only in the manufacturing procedure where PANI (EB) aggregates to form the nanoparticles. This is true for the nanoprecipitation (Section 2.1.3). However, while PANI chains grow as PN, it is converted into EB in the termination step. Therefore, the NP could grow in an expanded linear form (PN) and then contract to the ring-like form (EB) form by the termination reaction.

Manufacturing Colloidal Dispersions of PANI Nano-Objects by Polymerization without Templates
Aniline polymerization can be made by a variety of methods, including plasma oxidation and mechanochemistry. However, the most commonly used is so called "solution" On the other hand, polyaniline show a large polydispersity (δ = 2.18 [30]). Therefore, short chains could form small nanoparticles. Moreover, the polymerization chains terminate when the oxidant is consumed [29]. Therefore, in conditions of low local oxidant concentration, shorter chains will predominate.
Such constraints are valid for nanoparticles made of PANI (PN), with a rigid chain of planar geometry ( Figure 1A). The minimization of PANI (EB) (16 rings) shows a circle-like form due to the angle the nanoparticles manufactured by oxidative polymerization should have mean sizes in that order or larger.
On the other hand, polyaniline show a large polydispersity (δ = 2.18 [30]). Therefore, short chains could form small nanoparticles. Moreover, the polymerization chains terminate when the oxidant is consumed [29]. Therefore, in conditions of low local oxidant concentration, shorter chains will predominate.
Such constraints are valid for nanoparticles made of PANI (PN), with a rigid chain of planar geometry ( Figure 1A). The minimization of PANI (EB) (16 rings) shows a circlelike form due to the angle N H of the secondary amine (diphenylamine) units ( Figure 1B). The diameter of the circle is as low as 96 nm, allowing for small nanoparticles, or small-diameter nanotubes, to be formed. of the secondary amine (diphenylamine) units ( Figure 1B). The diameter of the circle is as low as 96 nm, allowing for small nanoparticles, or small-diameter nanotubes, to be formed.
In a first view, these smaller nanoparticles could be formed only in the manufacturing procedure where PANI (EB) aggregates to form the nanoparticles. This is true for the nanoprecipitation (Section 2.1.3). However, while PANI chains grow as PN, it is converted into EB in the termination step. Therefore, the NP could grow in an expanded linear form (PN) and then contract to the ring-like form (EB) form by the termination reaction.

Manufacturing Colloidal Dispersions of PANI Nano-Objects by Polymerization without Templates
Aniline polymerization can be made by a variety of methods, including plasma oxidation and mechanochemistry. However, the most commonly used is so called "solution" polymerization [31]. In fact, while the polymerization begins by mixing a solution of anilinium ion with another of an oxidant, the chain becomes insoluble and particles are formed. In the absence of special conditions, a solid precipitates. Another way involves interfacial polymerization. In this method, the reaction and chain growth occurs at the interface between two immiscible solvents, being a flat interface or an emulsion of one solvent dispersed in other. Finally, a method known to produce nanoparticles is the mechanochemical (solvent-free) polymerization. Since the reactants are solid, the produced nanoparticles are not made as a dispersion but can be dispersed in a solvent (e.g., water) during washing.

Solution Polymerization (SP)
Polymerization of aniline is made by mixing a solution of aniline salt with a solution of an oxidant [31]. After the reaction is finished, the polymer is filtered out of the solution.
The polymerization mechanism is shown in Scheme 1. As can be seen, the polymer grows in the pernigraniline (PN, fully oxidized form) by addition of aniline molecules at the extreme of the chain [32].
Nanomanufacturing 2023, 3, FOR PEER REVIEW 5 anilinium ion with another of an oxidant, the chain becomes insoluble and particles are formed. In the absence of special conditions, a solid precipitates. Another way involves interfacial polymerization. In this method, the reaction and chain growth occurs at the interface between two immiscible solvents, being a flat interface or an emulsion of one solvent dispersed in other. Finally, a method known to produce nanoparticles is the mechanochemical (solvent-free) polymerization. Since the reactants are solid, the produced nanoparticles are not made as a dispersion but can be dispersed in a solvent (e.g., water) during washing.

Solution Polymerization (SP)
Polymerization of aniline is made by mixing a solution of aniline salt with a solution of an oxidant [31]. After the reaction is finished, the polymer is filtered out of the solution.
The polymerization mechanism is shown in Scheme 1. As can be seen, the polymer grows in the pernigraniline (PN, fully oxidized form) by addition of aniline molecules at the extreme of the chain [32].  However, changing the polymerization parameters, defined nanostructures can be produced. Kaner and coworkers found the formation of PANI nanofiber dispersions by controlling the mixing of the reactants and stirring of the solution [33] (Scheme 2). Then, they extensively studied the mechanism of nanofiber formation [34][35][36][37]. The authors propose a general model of nucleation and growth of nanofibers, which occurs during any aniline polymerization. Since they assume that mass transfer is critical to the formation of nanostructures, the reaction model will be called Mass Transport Determining Polymerization (MTDP). The nuclei are formed homogenously by the oxidation of aniline. From the nuclei, chain growth produces nanofibers. Then, new chains are formed from the nanofibers, making large PANI aggregates. If the solution is stirred, the aggregation is favored However, changing the polymerization parameters, defined nanostructures can be produced. Kaner and coworkers found the formation of PANI nanofiber dispersions by controlling the mixing of the reactants and stirring of the solution [33] (Scheme 2). Then, they extensively studied the mechanism of nanofiber formation [34][35][36][37]. The authors propose a general model of nucleation and growth of nanofibers, which occurs during any aniline polymerization. Since they assume that mass transfer is critical to the formation of nanostructures, the reaction model will be called Mass Transport Determining Polymerization (MTDP). The nuclei are formed homogenously by the oxidation of aniline. From the nuclei, chain growth produces nanofibers. Then, new chains are formed from the nanofibers, making large PANI aggregates. If the solution is stirred, the aggregation is favored and a macroscopic precipitate is produced. Nanofiber dispersions can be readily obtained by rapidly mixing aniline, in an aqueous acidic solution, with an oxidant at higher temperatures (>25 • C) and in the absence of mechanical agitation. The kind of counterion influences the size of the nanofibers. While growth in HCl gives a small diameter (ca. 30 nm), using CSA produces a larger diameter (ca. 50 nm), and HClO 4 renders the largest sizes (>120 nm in diameter). Since through fast growth of chains and/or the presence of a lot of nuclei, a large number of independent fibers grow, the fibers stay dispersed in the solution and no precipitate is formed. The best conditions are: relatively high temperatures (>50 • C), no agitation of the solution and addition of nucleation, forming molecules [38]. and coworkers, acid (1 M) solution is used and the nanofibers are protonated not only in the imine nitrogen but also in the secondary amine groups. Therefore, the polymer has positive net charges and an electrical double layer made mainly of the mobile counterions. The repulsion of charged double layers could account for the stability of the nanofiber dispersions. In fact, increasing the pH to 5-7 induces aggregation, often irreversibly. On the other hand, the nanofiber dispersion could be purified/concentrated by centrifugation and the loose precipitate dispersed by ultrasound. Kaner and coworkers use the nucleation and growth model to explain the results assuming that all chains are linear (no ramification) and other parameters (oxidant/monomer ratio, kind and concentration of acid, etc.) have little effect. However, various groups obtain different nano-objects by polymerization of aniline in quite similar conditions, suggesting that other effects have some influence. The products are micrometric nanofibers (diameter < 100 nm). It should be noted that the nanofibers are not compact cylinders but have a ribbon-like shape and are not straight but contain several kinks. Li et al. synthesize PANI "nanosticks" by oxidation of ANIHCl with APS [41]. The microscopic images of the nanosticks resemble Kaner's nanofibers albeit shorter. The authors claim that ANIHCl form rod-like micelles, with neutral aniline inside, and direct the formation of the polymer as rods. They have no experimental proof of the existence of such micelles which have not been observed in the hundreds of studies of aniline polymerization. Moreover, they disregard the fact that polymerization with APS produces sulfuric acid and attribute the low conductivity of nanosticks produced in water to low doping. Stejskal and coworkers extensively studied the polymerization of aniline and the formation of nano-objects in the absence of templates [42][43][44][45][46][47][48][49]. They obtained nanotubes, nanospheres and other shapes. The main difference with the work of Kaner and coworkers is the use of different pH values (from ca. 0 to 10-11) and changing the dopant acid. They It is known [39] that the presence of some molecules (e.g., p-phenylenediamine) affects the size and crosslinking of the fibers which constitute the PANI precipitate. Therefore, addition to the solution promotes nucleation and nanofiber formation. The method is useful in the case of PANI, but it is essential for the polymerization of substituted anilines [40]. In addition to the mechanism of nanofiber formation remains the question of dispersion stability against aggregation in water without stabilizers. In the work of Kaner and coworkers, acid (1 M) solution is used and the nanofibers are protonated not only in the imine nitrogen but also in the secondary amine groups. Therefore, the polymer has positive net charges and an electrical double layer made mainly of the mobile counterions. The repulsion of charged double layers could account for the stability of the nanofiber dispersions. In fact, increasing the pH to 5-7 induces aggregation, often irreversibly. On the other hand, the nanofiber dispersion could be purified/concentrated by centrifugation and the loose precipitate dispersed by ultrasound. Kaner and coworkers use the nucleation and growth model to explain the results assuming that all chains are linear (no ramification) and other parameters (oxidant/monomer ratio, kind and concentration of acid, etc.) have little effect. However, various groups obtain different nano-objects by polymerization of aniline in quite similar conditions, suggesting that other effects have some influence. The products are micrometric nanofibers (diameter < 100 nm). It should be noted that the nanofibers are not compact cylinders but have a ribbon-like shape and are not straight but contain several kinks. Li et al. synthesize PANI "nanosticks" by oxidation of ANIHCl with APS [41]. The microscopic images of the nanosticks resemble Kaner's nanofibers albeit shorter. The authors claim that ANIHCl form rod-like micelles, with neutral aniline inside, and direct the formation of the polymer as rods. They have no experimental proof of the existence of such micelles which have not been observed in the hundreds of studies of aniline polymerization. Moreover, they disregard the fact that polymerization with APS produces sulfuric acid and attribute the low conductivity of nanosticks produced in water to low doping.
Stejskal and coworkers extensively studied the polymerization of aniline and the formation of nano-objects in the absence of templates [42][43][44][45][46][47][48][49]. They obtained nanotubes, nanospheres and other shapes. The main difference with the work of Kaner and coworkers is the use of different pH values (from ca. 0 to 10-11) and changing the dopant acid. They recognize, which is unfortunately not common in the field, that oxidation with APS produces sulfuric acid, made by sulfate anions produced by persulfate reduction and protons from the electrophilic aromatic substitution in the aniline ring. Therefore, unless the solution is buffered, with a large buffer capacity), the pH will become acidic when the conversion increases. If a weak acid (e.g., acetic) is used to acidify the initial solution, the weak acid will become protonated and it will not act as a dopant of the produced PANI. Instead, the sulfate (and bisulfate) ions will remain as counterion of PANI(ES). Moreover, the fact that aniline is oxidized at neutral (or basic) pH at the beginning of the reaction gives rise to oligomers which are not linear head-to-tail para linked chains but also contain ortho linkages and/or phenazinic rings (Scheme 3).
Nanomanufacturing 2023, 3, FOR PEER REVIEW 7 recognize, which is unfortunately not common in the field, that oxidation with APS produces sulfuric acid, made by sulfate anions produced by persulfate reduction and protons from the electrophilic aromatic substitution in the aniline ring. Therefore, unless the solution is buffered, with a large buffer capacity), the pH will become acidic when the conversion increases. If a weak acid (e.g., acetic) is used to acidify the initial solution, the weak acid will become protonated and it will not act as a dopant of the produced PANI. Instead, the sulfate (and bisulfate) ions will remain as counterion of PANI(ES). Moreover, the fact that aniline is oxidized at neutral (or basic) pH at the beginning of the reaction gives rise to oligomers which are not linear head-to-tail para linked chains but also contain ortho linkages and/or phenazinic rings (Scheme 3).

Scheme 3.
Mechanism of ortho addition of aniline to PANI (PN form) and formation of phenazinic rings. Only half of the pernigraniline tetramer unit is shown for simplicity.
The extended conjugation of linear PANI will be broken and the conductivity decreased. Moreover, packing of chains flat on each other (PN form) will not be possible. The oligomers are soluble in the solution and can self-assemble by π−π interactions and hydrogen bonding, creating insoluble solids which could template the assembly of more complex structures such as nanotubes. Therefore, two exothermic phases of nano-object formation are present: (i) formation of oligomers at high pH; (ii) PANI growth at low pH. The latter phase leads to large aggregates. Since the pH and the temperature can be continuously monitored, the reaction could be stopped at any point and the nano-objects isolated. The detailed study by Stejskal et al. has more evidence of the chemical structures, oligomers, etc., than other studies (Scheme 4) [42]. Other authors have supported the mechanism. Luo et al. polymerized different substituted anilines and found an even higher amount of phenazinic units [50]. At the same time, they was able to produce different nano-objects from those anilines. The extended conjugation of linear PANI will be broken and the conductivity decreased. Moreover, packing of chains flat on each other (PN form) will not be possible. The oligomers are soluble in the solution and can self-assemble by π−π interactions and hydrogen bonding, creating insoluble solids which could template the assembly of more complex structures such as nanotubes. Therefore, two exothermic phases of nano-object formation are present: (i) formation of oligomers at high pH; (ii) PANI growth at low pH. The latter phase leads to large aggregates. Since the pH and the temperature can be continuously monitored, the reaction could be stopped at any point and the nano-objects isolated. The detailed study by Stejskal et al. has more evidence of the chemical structures, oligomers, etc., than other studies (Scheme 4) [42]. Other authors have supported the mechanism. Luo et al. polymerized different substituted anilines and found an even higher amount of phenazinic units [50]. At the same time, they was able to produce different nano-objects from those anilines. Nanomanufacturing 2023, 3, FOR PEER REVIEW 8 Scheme 4. Reaction steps underlying the pH-Dependent Polymerization (pHDP) model.
The most interesting new nano-object obtained by Stejskal and coworkers are PANI open nanotubes. Since the key factor is the evolution of the pH of the solution, the model will be called pH-Dependent Polymerization (pHDP). Feng at al. observe the self-assembly of PANI oligomers and its effect on the formation of polymer nanostructures [51]. In that way, their results support the pHDP. Stejskal and coworkers used unbuffered solutions which become increasingly acidic during polymerization. Laslau et al. used a pH stat to maintain constant pH values (1, 2, 3, 4, 5 and 6) without adding different ions (buffer) which could affect the polymerization [52]. They partially confirmed and extended the pHDP model. While PANI nanotubes are formed at a constant acidity, the morphology changes with time even at constant pH. More different structures are observed, including spheres, flakes and ribbons at pH = 6, fibers and tubes at pH = 3-5, and large grains at pH < 2. Conducting nano-objects are produced at pH 1-3 while insulating structures are made at pH > 4. Interestingly, no effect of stirring was observed at any pH. The increase in pH produced decreased protonation, and increased ortho coupling, phenazine content, and sulfonation. [53]. MacDiarmid and coworkers give a name to the phenazinic (Scheme 3)-containing polymers: azanes [54]. They investigated the properties of the materials and found that they have low conductivity but could form different shaped nano-objects [55]. In fact, the structures have been known for a long time since they likely constitute the "ungreeneable aniline blacks" fabricated in the 19th century [56]. The "ungreeneable" adjective implies that they could not be reduced, unlike emeraldine or nigraniline which can be reduced to green polymers [57]. Aniline black is still a commercial dye produced by oxidation of aniline, whose structure contains several phenazine units [58]. The molecule cannot be reduced nor is it conductive. Li et al. make an excellent review on the state of knowledge on the formation of PANI structures in aqueous solution [59].
The MTDP model requires controlling the stirring to produce PANI nanofibers while the pHDP requires a high initial pH to produce the oligomers which induce the formation of nanotubes. In both cases, after formation of the nanostructures, some repulsive interaction should exist (e.g., double-layer formation by protonation) to stabilize the dispersions. Otherwise, large PANI aggregates are produced. However, McDiarmid and coworkers found that it is not the synthesis processes responsible for the aggregation but the The most interesting new nano-object obtained by Stejskal and coworkers are PANI open nanotubes. Since the key factor is the evolution of the pH of the solution, the model will be called pH-Dependent Polymerization (pHDP). Feng at al. observe the self-assembly of PANI oligomers and its effect on the formation of polymer nanostructures [51]. In that way, their results support the pHDP. Stejskal and coworkers used unbuffered solutions which become increasingly acidic during polymerization. Laslau et al. used a pH stat to maintain constant pH values (1, 2, 3, 4, 5 and 6) without adding different ions (buffer) which could affect the polymerization [52]. They partially confirmed and extended the pHDP model. While PANI nanotubes are formed at a constant acidity, the morphology changes with time even at constant pH. More different structures are observed, including spheres, flakes and ribbons at pH = 6, fibers and tubes at pH = 3-5, and large grains at pH < 2. Conducting nano-objects are produced at pH 1-3 while insulating structures are made at pH > 4. Interestingly, no effect of stirring was observed at any pH. The increase in pH produced decreased protonation, and increased ortho coupling, phenazine content, and sulfonation [53]. MacDiarmid and coworkers give a name to the phenazinic (Scheme 3)containing polymers: azanes [54]. They investigated the properties of the materials and found that they have low conductivity but could form different shaped nano-objects [55]. In fact, the structures have been known for a long time since they likely constitute the "ungreeneable aniline blacks" fabricated in the 19th century [56]. The "ungreeneable" adjective implies that they could not be reduced, unlike emeraldine or nigraniline which can be reduced to green polymers [57]. Aniline black is still a commercial dye produced by oxidation of aniline, whose structure contains several phenazine units [58]. The molecule cannot be reduced nor is it conductive. Li et al. make an excellent review on the state of knowledge on the formation of PANI structures in aqueous solution [59].
The MTDP model requires controlling the stirring to produce PANI nanofibers while the pHDP requires a high initial pH to produce the oligomers which induce the formation of nanotubes. In both cases, after formation of the nanostructures, some repulsive interaction should exist (e.g., double-layer formation by protonation) to stabilize the dispersions. Otherwise, large PANI aggregates are produced. However, McDiarmid and coworkers found that it is not the synthesis processes responsible for the aggregation but the separation/purification [60]. Using conventional filtration promotes aggregation while purifying by dialysis allows maintaining nanofiber dispersion. Another factor which has to be taken into account is that Manohar and coworkers detect the formation of micelles between aniline and persulfate ions, during the induction time of polymerization. They claimed that such micelles induce the formation of nanostructures [61]. Neelgund and Oki, polymerize aniline with APS (1:1 ratio) at 0-5 • C and a higher concentration (0.35 M) than other work, without adding acids [62]. They obtained nanospheres of 8 ± 3 nm. It is noteworthy that aniline solubility in pure water is 0.353 M at 8.6 • C [63]. In a solution which also contains 0.35 M salt (APS) and at 0-5 • C, the solubility is likely lower and spherical droplets of aniline should be present. Obviously, as the polymerization proceeds, the aniline concentration decreases and pH decreases, making the aniline protonated and soluble. In that way, the droplets dissolve. However, initially, a thin film of PANI could be formed around each droplet, creating the nanospheres. Ayad et al. polymerize aniline in water (0.07 M) by oxidation with APS, under stirring at 0-5 • C for 24 h [64]. They obtain spherical nanoparticles (ca. 20 nm) aggregated in long nanofibers/nanotubes. The synthetic procedure was corroborated by Neira-Carrillo et al. [65]. It seems that using a relatively lower concentration of monomer avoids formation of large aggregates. Li et al. manufactured PANI nanodisks by polymerization in water (at 5 • C), at a low concentration (ca. 0.04 M) of aniline and a high (1:0.5) ANI/APS ratio [66]. While the FTIR spectrum shows several bands that could be attributed to PANI, both the UV-vis spectrum and the cyclic voltammogram are quite different than those of PANI. Mahato et al. polymerized aniline at an even lower concentration (0.11 × 10 −3 M) and obtained highly crystalline nanospheres [67], which are likely the nucleation centers of nanofiber growth. They seem to use a higher temperature (70 • C) to polymerize in a reasonable amount of time (2 h). They claim that, since aniline polymerization is exothermic, it has to be heated to decrease the rate. Obviously, they confuse thermodynamics with kinetics and it could only be hoped they do not try with higher concentrations of aniline or other exothermic reactions. Wang et al. claimed that aniline concentration is the main factor that determines the morphology of PANI nanostructures [68]. At a lower aniline concentration, PANI nanofibers were formed, while larger PANI aggregates were obtained at a higher aniline concentration. The aggregates can be dislodged using ultrasonic irradiation.

Solution Polymerization Using Oxidants Other Than APS
Most of the synthesis of PANI, including nanostructures, has been performed using APS as an oxidant. However, other oxidants can be used which also promote the formation of PANI nanostructures. Amarnath et al. polymerizes aniline with selenious (H 2 SeO 3 ) acid as an oxidant [69] The shape of the nano-object depends on the oxidant/aniline mole ratio. When the mole ratio (MR) is very low (0.03-0.125), flakes of PANI are formed. Nanorods (150 nm diameter) are produced at low MR (0.5). At MR = 1, both nanorods and nanospheres were formed. Nanospheres are mainly favored for MR > 1. It should be noted that aniline requires 2.5 e − to oxidize to PANI. The half reaction for the reduction of selenious acid is: Therefore, 0.5 MR implies 2 mol of e − (less than the required by stoichiometric ratio) and MR = 1 implies a large excess of oxidant. Moreover, the fate of Se, which is not soluble in water, should be of interest. Lin et al polymerize aniline (0.5 mmol) using a mixture of FeCl 3 (8 mmol) and CuCl 2 (1 mmol) as oxidants [70]. The stoichiometric ratio for complete conversion of aniline into PANI requires 2.5 mol of electrons (i.e., 2.5 mol of FeCl 3 ) [71]. They claimed that nanostructures are only formed at an oxidant/aniline ratio between 6 and 10. The authors carry out the polymerization in water but assume that FeCl 3 acts as Lewis acid dopant (Figure 1g,h [65]), disregarding the production of HCl (protons from the polymerization reaction and Cl − from the reduction of FeCl 3 to FeCl 2 ). They obtained nanorice structures [72], which are quite useful in photothermal therapy. Another innovative way to produce PANI nanoparticles uses H 2 O 2 as an oxidant with an enzyme (glucose oxidase) as a catalyst [73]. The slow reaction produces nanoparticles of up to ca. 143 nm in diameter (after 162 h). Recently, Zeng et al. manufactured PANI nanobelts using ferric chloride and cumene hydroperoxide as oxidants in the presence of different inorganic acids, which influenced the microstructure [74]. The materials were tested as supercapacitor electrodes using galvanostatic charge/discharge. The best nanomaterial was produced in the presence of sulfuric acid and shows a S cap of 584.1 F g −1 (current density of 1 A g −1 ).

Solution Polymerization with Additives (SPA)
It has been observed that the presence of different ions and molecules induces the formation of PANI nanostructures. Different organic acids can act as additives to produce nanostructured PANI. Among them are AcOH [75], SA [76], DLTA [77], DBSA [78,79], DBSA+NSA [80], NSA [81], CSA [82], PSA [83], and DNSA [84]. In some cases, the organic acid could form micelles alone or as a salt of aniline. However, there is little evidence that micelle, if actually formed, shape and size determine the properties of the PANI nanostructures. Moreover, non-surfactive species: AMPSA [85], AMYGG [86], aminoacids [87] and sulfosalicylate [88] also induce nanostructuring. Moreover, inorganic acids with no amphiphilic nature could also induce nanostructure formation [89]. Other species such as 12tungstophosphate [90], sucrose stearate [91], sucrose octaacetate [92], AlCl3 [BuMe2Im]Cl; [BuMePyrrolidin]Cl [93], and β-CD [94,95] also induce formation of PANI nano-objects. Yang et al. polymerized aniline in the presence of 4-sulfobenzoic acid monopotassium salt (4SBAMPS) which delays the decrease in pH due to the polymerization [96]. At a low concentration of 4SBAMPS), PANI plates are obtained which roll into tubes at a higher concentration. The production of oligomers is also monitored and found that oligomers are produced at high pH (>2.6) while the polymer is produced below that pH value. These data confirm the pHDP mechanism. The polymers obtained at a higher pH show lower conductivity. Not only could the shape of the nano-objects be changed by the presence of the additive but also their chirality. Using enantiomerically pure CSA, the nanostructuring effect of the acid also induces chirality of the nanofibers/nanotubes [97][98][99] [101]. The morphology of PANI nanostructures could be changed from nano-needles or nanowires (50-100 nm in diameter) to hollow microspheres (ca 400 nm in outer diameter) by changing the concentration of the additive. Manohar and coworkers observed the formation of nanofibers by "seeding" with all kind of fibers [102]. However, nanotubes of different materials (e.g., single-wall carbon nanotubes) nanotubes are produced while using nanospheres of another CP (e.g., polypyrrole), PANI nanospheres are produced. Therefore, the seed method belongs to the use of templates. Recently, Xu et al. used IPA as an additive to induce the formation of nanofibrous PANI in a flow reactor [103]. The nanomaterial was tested in an electrochemical cell as a supercapacitor electrode. The nanofibers prepared with 5 wt% IPA show a large specific capacitance (S cap = 1082 F g −1 ).

Solution Polymerization with an External Energy Input
Aniline polymerization is an exothermic reaction. Therefore, no external energy input is required to sustain the reaction. However, the oxidative polymerization of aniline with APS shows an induction period which could take several minutes. Since the induction time disappears when aniline oligomers (e.g., 4ADA) are added [104], it seems that the initial dimerization is a slow reaction. However, the initiation period could be eliminated by light irradiation [105], allowing direct photolithography of PANI. Therefore, instead of slow formation of few nucleation centers, large amount of centers are formed in the irradiated area. Pillalamarri et al. use gamma rays to induce the polymerization of aniline with APS [106]. They obtain nanofibers with diameters that depend on the concentration of aniline and APS. Li et al. use UV light to produce a similar array of nanofibers to use as sensors [107]. On the other hand, Liu et al. use UV light to produce PANI nanofibers but with the addition of CTAB as a "soft template" [108]. It is unclear if the surfactant is required to nanostructure PANI. Li et al. polymerized aniline with APS and induced the formation of nanostructures by irradiation with UV light [109]. They control the diameter of the nanostructures by changing the oxidant/monomer ratio and the overall concentrations. In that way, they are able to produce nanowires and nanofibers. The most widely used way to induce polymerization externally is the use of ultrasound. Jing et al. manufactured nanofibers by polymerization of aniline with APS and applying ultrasound with a cleaning bath [110]. The diameter of the nanofibers decreases with an increasing APS/ANI ratio. Wang et al. polymerized aniline with APS, adding HCl, citric acid, sulfamic acid and taurine and applying ultrasound [111]. While no nanostructures are observed in HCl, all other additives induce the formation of nanofibers of different sizes. Ganesan et al. use potentiostatic electrochemical oxidation on a Ti ultrasound horn [112]. Applying a short ultrasound pulse after polymerization dislodges the nanoparticles from the electrode, creating a colloidal dispersion. The electrochemical properties of the dispersions are tested by re-depositing the nanoparticles with a binder (Nafion ® ) by evaporation. While the FTIIR and UV-vis spectra are similar to conventional PANI, the cyclic voltammograms show one system of broad peaks without evidence of the two redox processes present in the electrochemistry of PANI [113]. Recently, Qiu et al. used microwaves to induce the formation of PANI nanostructures [114]. The material had 7-10-fold higher electrical conductivity than conventional PANI. Nanosheets were fabricated in weakly acidic media while NFs are formed in highly acidic solutions. Those results are in agreement with the pHDP model. Moreover, both electrical conductivity and electrochemical charge storage are improved when HCl and APS concentrations are increased.
In summary, nanospheres, nanofibers, nanowires, nanobelts, nanodisks, round and square nanotubes, nanobelt and nanodisks can be made by simple oxidative polymerization, changing the conditions.

Solution Polymerization with Polymeric Stabilizer (SPS)
The oldest method to produce polyaniline is polymerization in solution in the presence of a stabilizer [115], where PANI fibers were observed using PEO (30K) as a stabilizer. This methods is well known for the fabrication of micrometric and nanometric particles of conventional polymers [116]. When anionic polymers (e.g., acrylic acid) are used to polymerize aniline, the polymer acts as a semi-rigid (due to charge repulsion) template which determines the shape of the PANI, usually a fibril, and not as a stabilizer [117]. Armes and Aldissi use a copolymer, poly(2-vinylpyridine-co-(p-aminostyrene)), as a stabilizer [118,119]. Since the stabilizer contains linked aniline units, it seems that PANI chains graft onto the stabilizer. However, polymerization with APS produces insoluble gels. Therefore, they use KIO 3 as an oxidant which produces nanorice particles, likely due to the dual role (stabilizer and grafting center) of the copolymer. This work was then extended by Armes et al., with a detailed study of the colloidal dispersions produced [120]. At the time, Armes and coworkers thought that this grafting procedure was the only way to obtain colloidal PANI [121,122]. Therefore, Bay et al. use poly(1-vinylimidazole-co-4-aminostyrene) as a stabilizer for polyaniline colloidal nanoparticles [123].
However, Vincent et al. produce PANI nanospheres [124], using a tailor-made polymer with an ethylacrlate/styrene copolymer backbone, with pendant PEO side-chains (SGPEO). It seems that weakly adsorbing PEO allows the fiber to grow while the strongly adsorbing SGPEO inhibits growth and the polymer remains at the nucleation stage. Stejskal and coworkers used PVA (with different remaining acetylation degree) to produce spheroidal nanoparticles of PANI [125,126]. They determine the size of the particles (R hidrodynamic = 116 nm) and an estimation of the mean molecular weight of the particles (520 × 10 6 g/mol) from a combination of SLS and DLS. Stejskal et al. show that it was possible to stabilize PANI NPs using the current "standard" stabilizer: poly(N-vinylpyrrolidone) (PVP) [127].
Banerjee et al. use poly(vinyl methyl ether) (PVME) as a steric stabilizer of PANI nanoparticles [128]. Since the stabilizer is only slightly soluble in water at 25 • C, they tried different solvents and oxidants. The best results were obtained with 50% ethanol and APS. Wallace and coworkers produce nanoparticles by electrochemical oxidation of aniline in a hydrodynamic cell [129,130]. Camphorsulfonic acid is used as a dopant. Using the (+) or (−) enantiomer, chiral colloids can be produced. Since 1 H NMR signal of oligoanilines and polyaniline do not give sharp peaks due to slow relaxation, the depletion of aniline can be monitored from the spectra taken at different times [131]. The kinetics shows a longer induction time when APS is used than KIO 3 . However, after induction, the reaction is faster in APS. The addition of PEO as a steric stabilizer does not significantly change the kinetics of the polymerization PANI colloidal dispersions that are obtained by aniline oxidation with APS in an acidic aqueous solution with hydroxypropylcellulose as a stabilizer [132]. Spherical polyaniline particles with small polydisprsity are made at 0 • C. On the other hand, at 40 • C the nano-objects are cylinders with coral-like morphology. The difference arises from the faster reaction rate at 40 • C. At 0 • C, faster rates can be achieved by addition of p-phenylenediamine. Chin and Park manufacture colloidal dispersions of PANI stabilized by PVME [133]. The more hydrophobic stabilizer allows dispersing the colloid in silicone oil. In that way, an electrorheological fluid with a large ER property (normalized yield stress in an electric field) is produced.
Stejskal review the work on the formation of colloidal dispersions of conducting polymers until 2001 [134]. The role of the steric stabilizer is well accepted in conventional polymers. However, in PANI, Stejskal and Sapurina show evidence pointing out that the stabilizer molecule could act as initiation nuclei of PANI chain self-assembly than be adsorbed on the surface of the nanoparticles [135]. The main evidence is the fact that the stabilizer is required at the beginning of the polymerization when no nanoparticle has been formed to be adsorbed to. Stejskal and Sapurina prepare an IUPAC technical report, where a standard procedure of synthesis of PANI using PVP as a stabilizer was performed by six research groups around the world [136]. Then, the samples were collected and analyzed. The dispersion of properties was low, attesting that the procedure is reproducible.
PANI NP dispersions (with PVP as a stabilizer) are tested in acidic, neutral and alkaline aqueous media, at temperatures of 105 • C (1000 h), sonication, cyclic freezing-melting cycles, or irradiation with UV-visible light [137]. PANI nanoparticles, stabilized with PVA were made into monolayers by the Langmuir-Blodgett technique [138]. PANI NPs stabilized with partially phosphorylated poly(vinyl alcohol) (P-PVA) were prepared by the chemical oxidative dispersion polymerization [139]. The P-PVA acts as a stabilizer and a codopant. The P-PVA/aniline ratio affects the morphology, dispersion stability and electrical conductivity of the nanoparticles. Using a P-PVA/ANI ratio between 50 and 60 wt%, nanospheres were produced which can be dispersed in water and have a high electrical conductivity (up to 7 S/cm). PANI dispersions stabilized by PVP were prepared and thermally characterized by TG-DTA analysis [140]. The composites are stable up to approximately 160 • C. Degradation of PANI chains follows at temperatures exceeding 250 • C. Zimmermann et al. produce colloidal dispersions of PANI nanospheres using PSS as a stabilizer [141]. Interestingly, they did not observe a templating effect of the polyacid, producing nanofibrils, but a stabilizing effect of nanospheres.
Park et al. show that the shape of the PANI nanostructure depends on the concentration of the stabilizer (PVP) [142]. They were able to produce nanospheres, nanorods and nanofibers. The results suggest that PVP adsorbs on the growing chain, inhibiting the growth. At low stabilizer concentrations, PANI could grow into ratio nanostructures with larger aspect ratios: nanorods and nanofibers.

Biopolymers and Smart Polymers as Stabilizers
Anbalagan and Sawant manufacture PANI nanoparticles stabilized by a biopolymer (pectin) to produce a material for supercapacitors [143]. Anjali et al. use the same stabilizer and deposited NPs on paper to use as sensors of bacteria [144]. Gonçalves et al. produce PANI nanoparticle dispersions using gum Arabic and show its biocompatibility [145]. Cruz-Silva et al, manufacture "smart" colloidal dispersions of PANI NPs stabilized by chitosan (CHI), PVA or PNIPAM [146]. The polymerization was made in weakly acid media using enzymatic oxidation. The NPs stabilized by CHI are stable in acid/neutral media but aggregate in basic media due to the deprotonation of the glucosamine units. The NPs stabilized by PNIPAM flocculate when the LCST temperature of PNIPAM (32 • C) is reached. The colloid size was ca. 50 nm for chitosan or PVA and up to 350 nm for the particles synthesized using PNIPAM. Jasenka' et al. use the enzymatic polymerization method to produce PANI NPs stabilized by CHI or PVA [147]. Abel et al. manufacture colloidal dispersions of PANI NPs stabilized by oxidation with APS in acid media [148]. The dispersions were stabilized by PVP, PNIPAM and hydroxyproplycellulose (HPC). While the PVP-stabilized dispersions are stable at all temperatures, the ones stabilized by PNIPAM flocculate at 32 • C and those stabilized by HPC at 42 • C. The temperatures agree with the LCST transition of the stabilizers. DLS measurement shows the change in size of the nanoparticles upon LCST transition of PNIPAM. Moreover, since PANI absorbs strongly in the NIR range, the thermosensitive dispersions (stabilized by PNIPAM and HPC) also flocculate at 25 • C when irradiated with NIR light due to a local photothermal effect. All the results with these "smart" nanoparticles support the "classic" model of a PANI sphere surrounded by adsorbed hydrophilic polymers which are extended (coil configuration) into the solution. Upon LCST transition, the polymer (e.g., PNIPAM) forms a globular configuration, decreasing the steric stabilizing effect. On the other hand, in the self-assembly model [135], where the hydrophilic polymer is composited with PANI, it is not clear how the LCST transition would decrease the stabilizing effect. The "smart" nanoparticles could have biological applications. Therefore, the interactions with biological entities should be evaluated.

Interfacial Polymerization (IP)
Michaelson and McEvoy polymerize aniline at the interface between an aniline solution in CCl 4 (aided by a surfactant) and an aqueous solution of an oxidant (APS) with acid [149]. Since their interest was to form PANI films, they use conditions where a nanofibrillar mat was produced at the interface. Since CCl 4 is denser than water, the mat floats at the interface and can be attached to an electrode to perform electrochemistry. Moreover, the authors use an additional organic layer (toluene) on top of the water layer, where aniline and oxidant dissolve. Therefore, the reactants are confined in the thin water layer promoting aggregation of the nanofibers into a mat.
Kaner and coworkers use the same technique but with a more polar organic solvent (CHCl 3 ) [150], which dissolves aniline without surfactants. Only two phases are used and the nanofibers grow into the aqueous layer manufacturing a dispersion. The fibers were characterized by TEM and SEM (backscattered electrons) as well as UV-visible spectroscopy at different pH. The nanofibers have a nearly uniform diameter (30-50 nm) with lengths varying from 500 nm to several micrometers. Upon drying of the dispersions, a nanofiber mat is formed (0.2-5 µm) which was tested as a sensor of acid (HCl) and basic (NH 3 ) gases. The sensors show clear changes in resistivity due to the protonic acid doping/dedoping. The response time is much faster than PANI films since the mass transport pathlength inside the solid is smaller (30-50 nm vs. 200 nm), and the geometry of the fibers allows for gas intake from the periphery of the fiber, also changing the fiber-to-fiber contact. PANI nanofibers grow at the liquid-liquid interface and are collected in the aqueous phase (Scheme 5). The organic byproducts of the polymerization dissolve in the organic phase.
Taking into account the MTDP and pHDP models, PANI nanofibers should grow at the interface because: (i) the slow mass transport of aniline from the organic solvent to the interface means the local concentration is low, similar to dilution; (ii) using inorganic acids in the aqueous phase (e.g., HCl) which are less soluble in organic solvents means that the local acidity is low, equivalent to a high enough pH, where slow accumulation of nucleates and phenazinic moieties appear. Pramanik et al. studied the interfacial polymerization of aniline in different solvents [151]. They use the difference of solvation parameters [152], to ascertain the effect of the interface on the nanofibers properties. There are little morphologi-cal properties in the nanofibers. On the other hand, the electronic properties related to their conductivity and UV-vis spectra, the doping measured by quinoid/benzenoid bands in FTIR and the crystallinity (measured by XRD) are strongly affected. It is unclear how the interface affects such properties. One possible explanation is that most of the reaction occurs in the organic solvent, since APS is soluble in different organic solvents, and the aqueous solution only act as reservoir of APS and sink of nanofibers. Su et al. study the effect of acid (HClO 4 ) concentration and the ratio ANI/APS on the formation of nanofibers by IP [153]. At a low [APS]/[ANI] ratio (≤1/4) and a high HClO 4 concentration (≥1 M), uniform PANI nanofibers are produced with a low content of phenazine-like units and ANI oligomers. Therefore, the nanofibers show higher conductivity than typical nanofibers made by IP. Chen et al. manufacture PANI nano-objects by the same interfacial polymerization but adding chitosan to the aqueous solution [154]. The biopolymer will adsorbs at the interface and modulates the shape of the nanostructures, at a low (1.5%) concentration of chitisan, nanofibers are produced while only nanospheres are produced at a high (2.5%) chitosan concentration. Nuraje et al. polymerize aniline using FeCl 3 (unknowm concentration), a low aniline concentration (ca. 0.01 M) and no acid [155]. They obtained highly crystalline PANI nanorice, likely due to the low concentration and slow reaction. Combining the use of additives with interfacial polymerization, Oueiny et al. obtain nanofibers with benzylphosphonic acid (BPA) and nanotubes with decylphosphonic acid (DPA), both present in the aqueous phase [156]. While at least DPA form micelles, the authors exclude templating by micelles as explanation of the nano-object shape. Zhang et al. polymerize aniline at the interface water/toluene with APS [157]. Two acids were used: CSA and AMPS. Moreover, twin-tailed surfactants were adsorbed at the interface to control the size of the nanofibers. The opposite effect of the surfactants on the size is observed in CSA and AMPS. It is likely that CSA is more amphiphilic, creating a more ordered interphase. Taking into account the MTDP and pHDP models, PANI nanofibers should grow at the interface because: (i) the slow mass transport of aniline from the organic solvent to the interface means the local concentration is low, similar to dilution; (ii) using inorganic acids in the aqueous phase (e.g., HCl) which are less soluble in organic solvents means that the local acidity is low, equivalent to a high enough pH, where slow accumulation of nucleates and phenazinic moieties appear. Pramanik et al. studied the interfacial polymerization of aniline in different solvents [151]. They use the difference of solvation parameters [152], to ascertain the effect of the interface on the nanofibers properties. There are little morphological properties in the nanofibers. On the other hand, the electronic properties related to their conductivity and UV-vis spectra, the doping measured by quinoid/benzenoid bands in FTIR and the crystallinity (measured by XRD) are strongly affected. It is  [158]. They found out that the centrifugal force presses the nanofibers onto the interface, making a fiber mat. Ding et al. use interfacial polymerization with DBSA as acid to produce PANI nanorods [159]. The oxidant/aniline ratio, DBSA concentration and temperature affect the diameters which can be tuned from 40 to 10,000 nm. A lower DBSA concentration and low temperatures produce nanorods of small diameter. Most interfacial polymerization is carried out at flat, quiet interfaces. However, Xing et al. show that nanofibers could be extended to drops of xylene dispersed in water by strong stirring [160]. The novel method has a larger interface area and could produce larger amount of nanofibers with the same solvent volume. Kamarudin et al. use a homogenizer to stir the mixture, reaching 30,000 rpm [161]. They observe that fast stirring produces shorter and thinner nanofibers. Sun et al. manufacture nanotubes of PANI by IP in the presence of dl-tartaric acid [162]. Above a threshold, the walls of the tubes become thicker and square-shaped nanotubes appear. Dallas et al. use IP to produce PANI nanostructures with SDS (anionic) or DTAB (cationic) additives in the aqueous solution [163]. While IP in the absence of additives gives nanofibers, the presence of these additives induces the formation of small nanospheres. Do Nascimento et al. study the structure and vibrational bands pf PANI nanofibers prepared by IP, varying the concentrations of HCl, in the water phase [164]. XRD and SAXS showed that the PANI crystallinity is reduced at high concentrations of HCl. The changes correlate with the occurrence of granular morphology, as observed by SEM. The resonance Raman and FTIR spectra, as well as the EPR data show an increase in the torsion angles of C ring −N−C ring segments. It is noteworthy that PANI incorporates chlorine atoms via nucleophilic addition at a high HCl concentration [165]. The steric effect of the chlorine atoms should increase the torsion angle. Zeng et al. manufacture PANI nanostructures with mixed oxidants (APS/FeCl 3 and APS/Cr 2 O 7 2− ) [166]. The use of dichromate, together with APS, produces petal-like nanofibers with a large surface area. Accordingly, the material shows a large specific capacitance (692.4 F/g) in an electrochemical supercapacitor. Cui et al. use a ANI/APS molar ratio of 1/0.25 [167]. Due to the low ANI concentration, petal-like ("cauliflower") structures are produced. Using PDMS, the nano-object is made hydrophobic. The combination of chemical hydrophobicity and topographic effects (Cassie-Baxter effect [168]) creates superhydrophobic surfaces.
Manuel et al. prepare PANI nanofibers by IP using CCl 4 or CHCl 3 as solvents [169]. They tested the fibers in a lithium ion electrochemical cell with LiPF 6 as an electrolyte. The PANI shows a good charge storage capacity. However, the authors claim that PANI exchange Li + during oxidation/reduction while only PF 6 − is exchanged as a counterion of the positively charged oxidized state of PANI [170]. Goel et al. study the FTIR spectra and XRD data of PANI nanofibers at different times during IP and compare the results with conventional PANI [171]. The results suggest that the nanofibers have higher crystallinity than PANI made in solution. Mu, manufacture PANI nanofibers by IP and tested its electrochemical behavior at different pH values [172]. While conventional PANI lost electroactivity at neutral pH, the nanofibers maintain some electroactivity at neutral pH. Jin et al. manufacture PANI nanofibers by IP with the addition of small alcohol molecules which adsorb at the interface between water and the organic solvent [173]. The changes in the interface affect the morphology, microstructure, and electrochemical properties of the PANI NFs. The highest S cap measured was 1042 F g −1 (scan rate of 5 mV s −1 ) and 489 F g −1 (current density of 1 A g −1 ) for the nanomaterial made with ethanol addition.

Other Immiscible Phases
Instead of toxic organic solvents (chloroform, toluene, CH 2 Cl 2 , benzene, CCl 4 ), other phases, immiscible with water, can be used. Akbarinezhad et al. use supercritical CO 2 , which is immiscible with water [174]. PANI nanofibers with ca. 60-80 nm diameter and several micrometer lengths are produced. Gao et al. use an interface between an ionic liquid (1-butyl-3-methylimidazolium hexafluorophosphate ([C 4 mim] PF 6 ) and water to produce PANI nano-objects [175]. Instead of nanofibers, small (30-80 nm diameters) nanospheres are produced. Nanoparticles of MgCO 3 and CaCO 3 were dispersed in the aqueous phase of aniline IP [176]. When MgCO 3 nanoparticles were used, polyaniline nanofibers are obtained while nanotubes are produced with CaCO 3 particles. Another immiscible phase is the gas phase. Bhadra and Lee vaporize aniline at 200 • C and put the vapor in contact with an acid solution of APS [177]. Nanofibers are produced, which are more soluble than conventionally prepared NF. Yang et al. manufacture nanofibers at the interface between frozen APS/acid/water and aniline/water solution [178]. Different shapes were obtained changing the concentration or the order of reactant addition. Gao et al. reviewed the manufacture of nano-objects by IP [179]. Dallas et al. wrote an historical perspective of the interfacial polymerization field [180].

Mechanochemical Polymerization
A novel method of aniline polymerization involves the solid-state (solvent-free) polymerization of aniline salts with oxidants (e.g., APS) under mechanochemical conditions [181]. While the solid-state reaction produces a solid, nano-objects can be found during the washing of the solid. Zhou [183]. The reaction occurs in slurry produced because the iron chloride hydrate melts at 37 • C, which is easily reached by the exothermic polymerization. The MCP is carried with a large ANI.HCl/oxidant ratio which seems required to produce nanofibers. Bhadra et al. manufacture PANI nanospheres (diameter 30-60 nm) by MCP of ANI.HCl with APS under mortar grinding [184]. The polymer inside the NPs shows higher crystallinity that conventional PANI but lower conductivity. Shao et al. manufacture PANI nanospheres (ca. 60 nm) by MCP of ANI.HCl with APS [185]. They studied the effect of water addition to the solid mixture, finding an effect on the doping and electronic properties but not on the nanoparticle size/shape. Bhandari et al. produce PANI nanofibers by MCP of aniline (as sulfate salt) and APS, with the addition of citric acid [186,187]. While it is likely that citric acid affects the nano-structure shape and size, the author's claim that citric acid doped the PANI is unattainable since the polymerization reaction produces sulfuric acid which will protonate the weak citric acid. Acevedo and Barbero show that MCP of ANI.HCl/APS produces a solid consisting of nanoparticles (ca. 70 nm diameter) [188]. From the size and distribution, a S BET of 69 m 2 g −1 can be calculated which agrees with the one experimentally measured by Kaner and coworkers [181].

Manufacturing Colloidal Dispersions of PANI Nano-Objects by Nanoprecipitation of a Preformed Polymer
Since PANI is soluble in some solvents, an additional method to produce nanoparticles involves adding a solution of PANI in a good solvent (e.g., NMP) into a poor solvent of PANI (e.g., water), which is miscible with the good solvent and contains a stabilizer (e.g., PVP). The PANI chains precipitate and form particles where the stabilizer adsorbs and avoid further aggregation. An advantage of this method is that the polymerization and nanoparticle formation step are separated, allowing detailed characterization of the base polymer. This is relevant since the conditions to produce nanostructures by polymerization (acidity, oxidant/monomer ratio, the presence of template anions or neutral molecules) not only affect the shape of the polymer but also the chemical structure. This method is widely used to produce nanoparticles of conventional polymers (e.g., PLA) and involves nanoprecipitation [189]. It is surprising the limited application of the method to purposely manufacturing PANI nano-objects.
Rubner and coworkers prepare a "solution" of PANI (EB) by diluting a true solution of PANI (EB) in DMF (good solvent) in water (poor solvent) [190,191]. The authors report that the "solutions" were unstable and have to be prepared fresh daily. Since the goal was to selfassemble PANI species by hydrogen bonding or charge, low stability will help the assembly process. It is likely that unstable colloidal dispersions of PANI particles were prepared and assembled. The dispersions were unstable because PANI (EB) is neutral. Therefore, no charge was present and no steric stabilizer was added. In a similar fashion, Li et al. mix a solution of PANI in formic acid (good solvent) with acetonitrile (poor solvent) [191]. They produce a dispersion of PANI NPs which are then deposited electrophoretically on Pt an ITO. In this case, the dispersion is more stable since the nanoparticles are charged by doping/protonation of formic acid which also provides the electrolyte. The double layer around the charged PANI nanospheres should be large since acetonitrile has a relatively low dielectric constant. Repulsion of the equally charged double layers of different nanoparticles avoids aggregation.
On the other hand, Abel et al. studied in detail the manufacturing process of PANI nanospheres by mixing a solution of PANI (EB) in NMP (good solvent) with water (poor solvent) which contains PVP (a steric stabilizer) [192]. The size of the NPs is determined by the concentration of PANI in NMP and the amount of solution in NMP added to the water. Nanospheres from 40 to 250 nm can be produced in that way. Since the emeraldine base form is used, it is possible for PANI chains to somewhat coil and produce small nanoparticles of mean-molecular-weight PANI. Nanoparticles can be made by mixtures of PVP with other stabilizers (e.g., PNIPAM) but not without PVP and the size depends also on the ratio PVP/PNIPAM. The method was used to produce fluorescent NPs by nanoprecipitation of PANI functionalized with fluorescent groups (dansyl).

Applications of Nano-Objects Synthesized by Template-Free Polymerization
The main goal for template-free manufacturing of PANI nano-objects is to produce technological applications. In various applications (e.g., double-layer capacitors), the specific surface area of the nano-object ensemble (e.g., nanofiber mat) is a relevant parameter. In Figure 2 is shown the calculated S BET for PANI nanospheres or nanocylinders (nanorods, nanofibers) of different diameters. The Area/volume ratio (which is related to S BET through the density) only depends on the radius of the object. As can be seen, nanospheres have a larger area than nanocylinders for the same diameter but the difference decreases with diameter. Moreover, quite small nanoparticles have to be manufactured to obtain large S BET values. In the case of nanospheres, a particle of 10 nm in diameter has a S BET of 400 m 2 g −1 , while a nanocylinder of 10 nm in diameter has only a S BET of 267 m 2 g −1 . However, the calculation assumes a compact nano-object while solid PANI is made of polymer chains with a non-compact interface. Therefore, experimental values of S BET could be larger. Another parameter of interest is the time required for a mobile species (counterion, analyte, adsorbate) to reach the center of the nano-object. This parameter is related with the response time of batteries and supercapacitors (where redox pseudocapacitance has a major contribution), response time of sensors and loading time of adsorption nanomaterials.
In Figure 3 is shown the diffusional response time of the nanospheres/nanocylinders for different diffusion coefficients (typical of solids). In all cases, the response time is below one second, even for relatively large diameters (200 nm) and slow diffusion (Do = 10 −10 cm 2 s −1 ). Therefore, the use of nano-objects ensembles could improve significantly the be- Another parameter of interest is the time required for a mobile species (counterion, analyte, adsorbate) to reach the center of the nano-object. This parameter is related with the response time of batteries and supercapacitors (where redox pseudocapacitance has a major contribution), response time of sensors and loading time of adsorption nanomaterials.
In Figure 3 is shown the diffusional response time of the nanospheres/nanocylinders for different diffusion coefficients (typical of solids). In all cases, the response time is below one second, even for relatively large diameters (200 nm) and slow diffusion (Do = 10 −10 cm 2 s −1 ). Therefore, the use of nano-objects ensembles could improve significantly the behavior of the device, even for larger diameters.

Sensors
Polyaniline is a material widely used for sensors [193][194][195]. It can be used as conductor in electrochemical or resistive sensors. Moreover, the electroactivity and UV-visible spectra depend on pH. The surface can be modified by simple organic chemistry reactions to bind enzymes, antibodies, aptamers, etc. Using PANI nanomaterials, it is possible to increase the surface area ( Figure 2) and decrease the response time ( Figure 3). Indeed, the first application of PANI nanofibers was as an ammonia resistive sensor [150], where the time response was faster than for PANI films. PANI nanofibers prepared by the MTDP method were used as reductants of Ag + to produce Ag 0 nanoparticles [196]. The large surface area of the nanofiber mat allows producing small and dispersed Ag nanoparticles which show SERS activity. Popov et al. manufacture PANI nanofibers by IP [197]. They combine the NF with rGO, Nafion ® and glucose oxidase to manufacture electrochemical glucose sensors. Nate et al. manufacture PANI nanofibers loaded with cobalt oxide (Co3O4) nanoparticles by IP [198]. To that purpose, Co3O4 NPs are dispersed in the aqueous phase of IP. Thin films of the nanomaterial are then deposited onto GC by drop casting. The films are used to detect electrochemically antimalarial drugs, where the PANI NF mat act as conductive matrix and the cobalt oxide as electrocatalyst. Zou et al. produce PANI nanotubes by SP and combine with graphene and aramid nanofibers to produce aerogels [199]. The elastic and conductive aerogels show excellent properties as pressure sensors. Thakur et al. manufacture biocompatible PANI nanospheres by SPS using a biopolymer (pectin) [200]. The colloidal dispersion is used to sense the growth of Escherichia Coli through the change in visible spectra of the emeraldine base (blue) to emeradine salt (green). The organic acids produced by the bacteria protonate the PANI backbone. Zhao et al. manufacture PANI nanofibers by IP [201]. An enzyme (glucose oxidase) was covalently bonded to the NFs by amidation and used as an amperometric sensor of glucose. The field of nanostructured conducting polymers applied in sensors was previously reviewed [202]  Subsequently, selected applications in each different field will be described.

Sensors
Polyaniline is a material widely used for sensors [193][194][195]. It can be used as conductor in electrochemical or resistive sensors. Moreover, the electroactivity and UV-visible spectra depend on pH. The surface can be modified by simple organic chemistry reactions to bind enzymes, antibodies, aptamers, etc. Using PANI nanomaterials, it is possible to increase the surface area ( Figure 2) and decrease the response time ( Figure 3). Indeed, the first application of PANI nanofibers was as an ammonia resistive sensor [150], where the time response was faster than for PANI films. PANI nanofibers prepared by the MTDP method were used as reductants of Ag + to produce Ag 0 nanoparticles [196]. The large surface area of the nanofiber mat allows producing small and dispersed Ag nanoparticles which show SERS activity. Popov et al. manufacture PANI nanofibers by IP [197]. They combine the NF with rGO, Nafion ® and glucose oxidase to manufacture electrochemical glucose sensors. Nate et al. manufacture PANI nanofibers loaded with cobalt oxide (Co 3 O 4 ) nanoparticles by IP [198]. To that purpose, Co 3 O 4 NPs are dispersed in the aqueous phase of IP. Thin films of the nanomaterial are then deposited onto GC by drop casting. The films are used to detect electrochemically antimalarial drugs, where the PANI NF mat act as conductive matrix and the cobalt oxide as electrocatalyst. Zou et al. produce PANI nanotubes by SP and combine with graphene and aramid nanofibers to produce aerogels [199]. The elastic and conductive aerogels show excellent properties as pressure sensors. Thakur et al. manufacture biocompatible PANI nanospheres by SPS using a biopolymer (pectin) [200]. The colloidal dispersion is used to sense the growth of Escherichia coli through the change in visible spectra of the emeraldine base (blue) to emeradine salt (green). The organic acids produced by the bacteria protonate the PANI backbone. Zhao et al. manufacture PANI nanofibers by IP [201]. An enzyme (glucose oxidase) was covalently bonded to the NFs by amidation and used as an amperometric sensor of glucose. The field of nanostructured conducting polymers applied in sensors was previously reviewed [202].

Electrochemical Energy Storage
Energy storage applications (batteries/supercapacitors) of electroactive PANI require preparation of solid electrodes with a large surface area/small solid size and macro/ mesoporosity [203]. The large surface area (Figure 2) allows development of a double layer for charge storage in supercapacitors while the small solid size of the nano-objects decrease the pathlength of slow mass transport of ions inside the solid (Figure 3), which is the rate controlling processes in batteries. Moreover, macro/mesoporosity is required to provide/sink ions in the solution (fast mass transport). Therefore, hierarchical structures are especially suited for energy storage [204]. Nanofibers mats or nanocomposites could fulfill those requirements [205]. Kim et al. manufacture PANI nanofibers by MTDP and disperse them in a polyelectrolyte (Nafion ® ) [206]. The manufactured supercapacitor show good specific capacitance and high flexibility. Gao et al. produce nitrogen-enriched carbon nanofibers by carbonization of PANI NFs [207]. The polymeric nanomaterial was made by SP using the MTDP method. The resulting N-enriched (7.59 at %) carbon nanofibers have diameters in the range 100-150 nm with a S BET of 915 m 2 g −1 . The S cap is of 172 F g −1 (at 2 mV s −1 ). It shows a good cycling stability (89% S cap retention after 5000 cycles). Udayan et al. manufacture PANI nanofibers by IP, and combine them with a MOF (ZIF-8) to produce a flexible electrode material for supercapacitors [208]. Due to its hierarchical structure, ZIF-8/PANI has a S BET of 610.8 m 2 g −1 and a S cap of 395.4 F g −1 (at a current density 0.2 A g −1 ). The field of application of PANI, including nanomaterials, in supercapacitors has been reviewed [209].

Corrosion Protection
It is well known that PANI coatings protect metals from corrosion [210,211]. However, since PANI is not soluble in common solvents, the formation of coatings on large surfaces is difficult. Using colloidal dispersions is a well-known method to produce protective metal coatings. Therefore, dispersions of PANI nano-objects could be used to produce protective coatings. Fuseini et al. deposited PANI colloids by electrophoresis onto copper [212]. The colloidal dispersions were made by nanoprecipitation from a formic acid (good solvent) solution of PANI using ACN as poor solvent and ultrasonication to aid the dispersion. The dispersion is stabilized by protonation as shown by zeta potential measurements (ζ = + 32.9 mV, D h = 542 nm). The charged colloids are driven by the electric field (15 V) into the negative Cu electrode. The produced coating resembles PANI chemical or electrochemical films, suggesting that the colloidal particles aggregate during deposition. Indeed, the PANI coating acts as a barrier for electrolyte reaching the Cu, inhibiting the metal corrosion. Abel et al. synthesize PANI nanofibers by IP [213]. Then, the NFs were surface functionalized by addition [214], of different nucleophiles, producing nanofibers with different surface properties. While PANI NF dispersions are stable in acid media due to charge repulsion of the protonated fibers, they aggregate into porous mats at neutral pH. Therefore, conducting mats of functionalized nanofibers can be deposited onto interdigitated metal electrodes. The modified electrodes show different conductivity changes when exposed to volatile organic contaminants. Combining the signal of several electrodes, electronic noses can be manufactured. While conventional PANI lost electroactivity at neutral pH, the nanofibers maintain some electroactivity at neutral pH.

Biomedical Applications
While electronic properties are relevant for battery or sensor research, biocompatibility is required to use nanofibers in biomedicine (e.g., photothermal therapy). Yslas et al. studied the toxicity and teratogenicity, using toad embryos, of PANI NF produced by IP (using HCCl 3 as organic phase) [215]. The NF shows low toxicity and dose-dependent teratogenicity. Ibarra [218]. The stable dispersions are absorbed into macroporous hydrogels of PNIPAM made by cryogelation. The PANI NPs, retained inside the gels, absorb electromagnetic radiation (e.g., NIR light) and heat up, triggering the volume collapse of the thermosensitive hydrogel. Jasenka' et al. manufacture PANI NPs stabilized by CHI or PVA [147]. The NPs showed in vitro antioxidation activity. Therefore, the NPs were applied to reduce oxidative stress and inhibit the production of ROS by neutrophils. The NPs also reduce the production of inflammatory cytokines by macrophages.
Ghosh et al. manufacture PANI (and PEDOT) nanofibers by SP using an additive (CTAB) [219]. Thin films were produced by drop casting. The films exhibited strong antimicrobial activity, under UVA irradiation, against Gram-positive (Staphylococcus aureus) and Gram-negative (Escherichia coli) bacteria. Pang et al. use PANI NPs, bearing vinyl groups, to crosslink hydrogels [220]. The nanocomposites show good mechanical and antibacterial properties which allow their use in wound dressing. The field of stimuli-responsive nanogels, which include those loaded with PANI, has been already reviewed [221].

Optoelectronics
Since nanoparticles are dispersible in organic solvents, nanocomposites of soluble dielectric polymers and PANI nano-objects can be made by casting films from mixed solutions. Araujo et al. synthesize nanofibers by IP (CH 2 Cl 2 (with SDS)/water [222]. The nanofibers were dispersed in butanonone and mixed with a solution of PMMA in the same solvent under sonication. By evaporation of the solvent, homogenous nanocomposite films can be casted. Ghaleghafi et al. used three different methods to produce nanocomposites of 2D MoS 2 and 1D PANI (nanofibers) [223]. One of the methods involve synthesizing PANI nanofibers and mixing with nano MoS 2 (synthesized by hydrothermal (HT) reaction of molybdate ion with thiourea). Another involves the HT synthesis of nano MoS 2 in the presence of PANI nanofibers. The optoelectronic properties are poorer than the nanocomposite made by in situ polymerization of aniline in the presence of nano MoS 2 , suggesting that mixing the nanomaterials produce phase separation. On the other hand, PANI nanofibers likely degrade under the hydrothermal conditions (210 • C for 24 h) used to produce nano MoS 2 . Yarmohamadi-Vasel et al. manufacture PANI nanofibers by MCP using iron chloride as an oxidant [224]. The NFs are combined with TiO 2 NPs, in different architectures, producing photovoltaic cells. The deposit of PANI NFs acts efficiently as the hole-injection layer.

Microwave Absorption
As an electronic conductor, PANI interacts strongly with oscillating electromagnetic fields. In that way, it can absorb and reflect microwaves [225]. Using PANI dispersion allows producing composites with structural polymers. Zhang et al. manufacture PANI nanoparticles by SP with polymer (PVP) stabilizer [226]. PANI nanoparticles exhibit very strong reflection losses (up to −40.5 dB at 5.8 GHz) with a wide bandwidth (up to 3.2-18 GHz over −10 dB. The values are comparable to magnetic particles. Yang et al. manufacture nanorod-coated PANI microtubes [227]. The hierarchical structures are made by SP using SDS as an additive and adjusting the HCl concentration. The material show a large absorption of −43.6 dB (1.55 mm thick) and a wide absorption band of 5.84 GHz. Su el al, manufactured PANI nanotubes by SP [228]. To form the nanotubes, acetic acid was used as an additive. Since ferric chloride is added to the polymerization solution, the polymer nanotubes retain iron. Upon carbonization in the absence of oxygen, nitrogen enriched carbon nanostructures loaded with iron are produced. The presence of Fe improves the microwave absorption properties of carbon. The reflection loss of the Fe/carbon fibers is of −44.09 at 6.62 GHz, with a bandwidth of 6 GHz.

Adsorption of Contaminants
The chemical nature of PANI, with aromatic rings, hydrogen bonding centers and positive charges (cation radicals) allows interaction with charged, hydrogen bonding donor/acceptors and anions. PANI nanomaterials show large surface areas ( Figure 2) to interact with a large molecules (e.g., proteins). On the other hand, small molecules could penetrate inside the solid nanomaterial and the short pathlength allows for short response times (Figure 3). Bayat et al. manufacture PANI nanofibers by emulsion polymerization using CTAB [229]. Then, SiO 2 is formed onto the nanofibers and linked to PANI using formaldehyde. The nanocomposite is dispersed in a PVA/CHI solution to produce fibers by electrospinning. The presence of PANI promotes the adsorption of anionic dyes on the cation radical charges of the CP backbone. Lyu et al. use the same property of PANI to adsorb acid red G dye on PANI nanofibers [230]. PANI nanofibers were produced using FeCl 3 as an oxidant and varying the reaction temperature and aniline/to oxidant ratio (R AOx from 1:2 to 1:12). The authors use a relatively large acid (HNO 3 ) concentration (ca. 0.4 M) and assume that the counterion of the as produced PANI is NO 3 − . However, the aniline concentration is ca. 0.18 M. Therefore, chloride (from FeCl 3 ) concentration is in the range 1 (ratio 1:2) to 6.5 (ratio 1:12) M and Cl − should be dominant as counterion. The retention of the anionic dye reaches up to 491 mg g −1 for a material produced at 20 • C with a R Aox of 1:4. Such material has a S BET of 25 m 2 g −1 . The adsorbed dye could be removed by deprotonation of PANI and the removal is reversible for 10 cycles.

Photothermal Applications
PANI absorbs light in the UV-visible-NIR range [231], heating up the polymer [232]. Additionally, an alternating electromagnetic field (RF or MW) interacts with the charge carriers, creating Eddy currents inside the conductive nanoparticles which locally heat up them. It has been shown that PANI have to be in the conductive state (protonated) to interact with electromagnetic field (such as MW or RF), but NIR light is also absorbed by the deprotonated state [233]. Abel et al. grow thermosensitive hydrogels around PANI nanofibers (made by IP) and nanospheres (made by SPS with PVP) [234]. The LCST transition of the hydrogel matrix was induced by application of RF (30 kHz, 1 kW) or MW (2.4 GHz, 700 W). Bongiovanni Abel et al. synthesize PANI nanospheres by SPS, using a thermosensitive polymer (PNIPAM) as a stabilizer [235]. The nanoparticles are then used to stabilize Pickering emulsions of water immiscible organic solvents. Upon heating, the nanoparticles become destabilized and the emulsion breaks. Moreover, PANI NPs could be locally heat up by applying RF, inducing the emulsion break. Dong et al. synthesize PANI nanofibers by IP [236]. The nanomaterial is dispersed in epoxy curing mixture and the healing of coating damage is activated photothermically by NIR (808 nm) light. PANI nanofibers absorb the light, and heat up, increasing the rate of epoxy curing.

Electrorheological Fluids
Since conductive PANI nanoparticles bear charges, they will be affected by strong electric fields. Specifically, the flux of a charged colloidal dispersion is hindered by the electrostatic force exerted by the external field on the colloid. Such electrorheological effect could be used in non-mechanical brakes and variable transmission devices [237]. Quadrat and Stejskal reviewed the application of PANI nano-objects, manufactured without templates, in electrorheology [238].
Wang reviewed the methods of fabrication of PANI nanofibers, including templatefree processes [239]. He also described and discussed technological applications of PANI nanofibers.

Conclusions
The work reviewed suggests that it is possible to produce a variety of different shaped/sized PANI nanoparticles by polymerization or nanoprecipitation of the preformed polymer, without using templates. The widely used oxidative polymerization of aniline in solution (SP) produces nano-objects (e.g., nanofibers) by controlling the polymerization conditions. Two models are proposed to explain the formation of nano-objects, instead of macroscopic aggregates. One explains the formation of nanofiber by the restrictions in mass transport in a non-stirred solution (MTDP). In those conditions, fast nucleation and slow growth give rise to 1D nanostructures. The other model (pHDP) is based on the effect of pH during polymerization. At high pH, non-only linear (head-to-tail) poyaniline oligomers are formed but also chains with ortho substitutions and phenazinic rings are produced. The non-linear chains show lower conductivity and crystallinity but self-assemble, giving rise to nanofibers and nanotubes. On the other hand, the presence of aniline oligomers (e.g., 4-aminodiphenylamine) also makes the initiation/nucleation step, promoting the formation of 1D nanostructures. A variety of ions (e.g., SDS) or neutral species (e.g., IPA) could act as additives promoting the formation of nano-structures and controlling the shape/size of the nano-objects.
In addition to the formation of nano-objects, the manufacturing of high-solid-content colloidal dispersions of PANI nanoparticles requires stabilization. Hydrophilic polymers (e.g., PVP) which adsorb on the surface of the particles could act as steric stabilizers in aqueous solutions. Stable dispersions of spherical nanoparticles, produced by conventional solution polymerization, could be manufactured using such polymers, including biopolymers (e.g., pectin) or biobased polymers (e.g., hydroxypropylcellullose). Smart polymers (e.g., PNIPAM) have also used as stabilizers of PANI nanoparticles. The polymer chains react to external stimuli (e.g., temperature) by changing its spatial conformation from coil to globule (LCST transition). Since the stabilizing effect requires the polymer chain to be unfolded in solution, the colloidal dispersions can be flocculated by increasing the temperature above the LCST. The nanoparticles can be redisolved by cooling with string stirring. Moreover, by irradiation with electromagnetic radiation (e.g., NIR light), which is absorbed by PANI, the flocculation can also be induced since the PANI core heats up and triggers the LCST transition of the PNIPAM stabilizer shell.
The polymeric stabilization could also be used to produce PANI nanospheres by nanoprecipitation. Mixing a solution of PANI (EB) in a good solvent (e.g., dimethylformamide) with a poor solvent (e.g., water) induces the precipitation of PANI. Since the solvents are miscible, the good solvent is diluted in the poor solvent and the interactions with the polymer are weakened, promoting chain aggregation. The formed nanoparticles aggregate into macroscopic precipitates. However, if the poor solvent contains a steric stabilizer (e.g., PVA) or immobile charges are generated (e.g., by protonation), stable dispersions will be produced. Since the process is controlled by solubility equilibrium and mass transport, the size of the nanoparticles depends on the number of nuclei and the amount of available stabilizer. Therefore, small (down to ca. 40 nm) nanoparticles can be produced. The method has been scarcely used, when could be used to manufacture nanoparticles from PANI, copolymers and functionalized polyanilines [240]. Geometrical considerations preclude the formation of small nanoparticles of a high-molecular-weight rigid polymer, since the polymer could not fold into a small globule. Since PANI grows in the pernigraniline state, which is highly conjugated and rigid, the mean chains (Mn = 96800 g/mol) could not produce nanoparticles smaller than 280 nm. Therefore, very small nanoparticles (e.g., single-crystal nanorices) are likely made of low-molecular-weight chains. On the other hand, PANI chains in the emeraldine form are also rigid but the angle in the Ar-NH-Ar diphenylamine unit makes the polymer take a ring-like shape with a smaller diameter (down to 90 nm). These considerations explain why small-diameter (<50 nm) nanotubes can be built. The formation of nanospheres by precipitation of PANI (EB) from solution could be explained due to the relaxation of that constraint.
In summary, nanospheres can be prepared by solution polymerization with stabilizers and nanoprecipitation of preformed PAN I from solution. Nanofibers and single crystals (nanorice) are usually produced by interfacial polymerization. Nanotubes are produced by solution polymerization with varying pH or using additives. Other shapes can be produced (petal-like or square nanotubes, nanorice single crystals, nanobelts, nanodisks, etc.) by changing some parameters of the polymerization.

Future Directions
Several synthetic methods at solid interfaces produce polyaniline nanoparticles without templates: electrochemical [241], chemical [242], or photochemical polymerization produce solid aggregates which do not disperse in the solvent. While such behavior is useful to produce nanostructured solid materials, it precludes the application in dispersed form. A post-polymerization dispersion step, using polymeric stabilizers (e.g., PVP) and fast stirring or ultrasound will enable to produce stable dispersions and extend solid-state methods (including mechanochemistry) to the formation of dispersed PANI nano-objects.
While nanoprecipitation is a well-known method to produce nanoparticle dispersions of common polymers, it has been seldom used to produce PANI colloidal dispersions. Biopolymers (e.g., CHI) and even biomacromoecules (proteins, polysacharides, and nucleic acids) could be used as stabilizers. In that way, biocompatibility will be improved and new properties (e.g., immunogenicity) could be incorporated. It has been shown that conducting polymers functionalized with solvophilic groups could act as polymeric stabilizers of nanoobjects (e.g., carbon nanotubes [243,244]). In the case of PANI nano-objects, the use of such electronic conductive stabilizers should improve the bulk conductivity of films since the electron hopping between nano-objects should be less hindered than with a dielectric polymer stabilizer.
Interfacial polymerization is the preferred technique to produce PANI nanofibers. However, usually solvents which are harmful and/or environmentally deleterious (e.g., chloroform) are used. While other immiscible phases, such as supercritical CO 2 [174] or ionic liquids [175] have been applied, they are usually expensive and/or of complex operation. On the other hand, common organic solvents with a low environmental impact [245], which are immiscible with water, could be applied. Moreover, even solvents soluble in water (e.g., acetonitrile) or even pure liquid aniline form a separated aqueous phase when put in contact with concentrated aqueous solutions of hydrophilic species (e.g., CaCl 2 ). Using those kinds of solvents, a more ecofriendly way to produce PANI nano-objects by interfacial polymerization could be devised.
The mechanism of template-free polymerization on solution is still unclear and further investigation on the subject could provide useful information. However, from the technological point of view, the setting of clear procedures to produce PANI nano-objects in a reproducible way seems more relevant. Such an endeavor requires re-studying published work where sometimes the experimental details are not complete and/or the results could not be reproduced by unknown reasons. However, the systematic record of testing detailed synthetic procedures to produce all possible shapes (nanospheres, nanorice, nanobelts, nanofibers, nanotubes, nanodisks, etc.) would be very useful towards application of PANI nano-objects.
Once stable PANI colloidal dispersions are obtained, they can be used to build larger structures or nanocomposites. In general, nanocomposites have usually been produced by in situ polymerization of aniline inside the matrix [246], or in the presence of the dispersion of another nanomaterial [247]. While the method has been widely used [248][249][250], it precludes a complete characterization of the in situ produced nano PANI. Indeed, the materials produced by in situ polymerization of aniline inside hydrogels have usually been considered a kind of blend (semi-interpenetrated networks) [251]. It was experimentally shown that they are nanocomposites with PANI nanoparticles filling the pores of the polyacrylamide hydrogel [252]. However, the semi-interpenetration model is still believed to be true since the true nature of the PANI inside the matrix is not simple to characterize [253]. Therefore, even when manufacturing solid structures (thick and thin films, porous materials, etc.) is the main goal, colloidal dispersions could be useful. The building process requires assembly of the particles into a stable solid. In that sense, layer-by-layer self-assembly [254], the method pioneered by Rubner and coworkers using unstable PANI "solutions" in water [190], allows building films with controlled thickness. While they use hydrogen bonding as self-assembly interactions, protonated PANI (positive charged) could be assembled with negative polyelectrolytes. Moreover, PANI nanoparticles could be stabilized with polymers bearing charges (e.g., polyelectrolytes) or hydrogen bonding donors (e.g., CHI) or acceptors (e.g., PVP). By preparing nanoparticles, by in situ polymerization or nanoprecipitation, bearing both kinds of stabilizers, self-assembly of PANI multilayers can be achieved [255]. On the other hand, while each adsorption step adds mass to the film, it is not clear that nanoparticles behave as macromolecules [256]. In larger colloids, which can be observed in situ by optical microscopy, bidimensional (2D) covering resembling molecular adsorption was found [257]. In most applications of nanostructured films (e.g., supercapacitor electrodes), the 2D self-assembly should have similar properties than the multilayer.
Other ways to produce structures from dispersions involve electrophoresis, drop casting, and spin coating. Electrophoresis has only been used to produce films of undefined nanoparticles made by nanoprecipitation (without stabilizer [191]. On the other hand, it has been applied to deposit 1D materials, such as carbon nanotubes [258], and metal nanorods [259], as well as 2D materials, such as graphene [260]. The higher dimensionality of the nano-objects induces multiple dispersion interactions which fix the nanomaterial to the surface. PANI nano-objects strongly absorb light in the UV-visible-NIR wavelength range (200-1600 nm). On the other hand, heat conduction out of the nano-object is hindered by the low thermal conductivity of the surrounding media. Therefore, the nano-objects could reach higher temperatures than thick PANI films. Such an effect has been used by Kaner and coworkers to flash weld PANI nanofibers [261]. Moreover, PANI thin (200-300 nm) have been structured by laser ablation (DLIP [262]), or transferred by LIFT [263], using the photothermal effect. Applying such methods to different PANI nano-objects could be used to manufacture hierarchical structures and complex devices containing PANI nano-objects. When polymers are used as stabilizers, the film formed by aggregation of nano-objects contains the stabilizer as an interlayer.
Such a layer affects the properties (e.g., conductivity) and its nature depends on the drying process. The effects of the drying process on the film structure have been studied [264], and the results could be used as guidelines to improve the reproducibility and properties of films produced from colloidal dispersions of PANI nano-objects. On the other hand, macroscopic structures such as thick films/mats are likely the most useful form of nanomaterial for technological applications, such as supercapacitor electrodes and adsorption membranes. Early on, it was demonstrated that interfacial polymerization produces a porous membrane made of PANI nanofibers [149]. The method should be developed further and include the incorporation of other nanomaterials (e.g., carbon nanotubes). Another way, extensively used in the formation of carbon nanotubes and cellulose nanofibers mats [265], involves the deposition from vacuum filtration through a fine pore membrane. The method can be easily applied to PANI nanofibers, nanotubes or even nanorods. The thickness of the mat can be controlled by the amount of dispersion filtered and different 1D nanomaterials can be combined.
The low-cost, easy and reproducible synthesis of PANI nano-objects suggests their use as templates of solids to form nanoporous materials. Other well-known nanomaterials, such as silica nanospheres, have been used to template pores into solid materials [266], by in situ formation of the solid around the particles. However, the removal of the template requires corrosive solutions (e.g., HF). On the other hand, PANI nano-objects can be removed by dissolution (e.g., in NMP) for soft materials (e.g., crosslinked polyacrylamides) or by calcination for heat-resistant solids (e.g., metal oxides). Using PANI nanofibers, it is possible to template open pores in the solid. Such porosity is required for most chemical applications (e.g., adsorption of contaminants from solution) where the species have to reach the inside of the solid. Since the solid itself could have a porosity level, hierarchical structures can be formed.

Conflicts of Interest:
The author declares no conflict of interest.