Transforming the Buckyball: Regioselective Synthesis of Water-Soluble [60]Fullerene Derivatives for Biomedical Applications

Abstract

Water-soluble fullerene derivatives exhibit a wide range of fascinating biological properties, including antioxidant, antiviral, antitumor, antibacterial, and myogenic effects. During the initial stage of research, most of the reported data on the biological activity of fullerenes were obtained using complex, inseparable mixtures of regiomers with a big focus on fullerenols as the most accessible form of water-soluble fullerene-based compounds. However, during the past decade, significant progress has been made in the synthesis of various isomerically pure water-soluble fullerene derivatives, which opens up possibilities for more directed investigations of their biological activity. In this review, we will highlight current methods for the straightforward synthesis of different types of water-soluble fullerene derivatives with well-defined molecular structures. Special attention will be paid to the possibilities of the precise control of the number, types, and positions of functional groups on the fullerene cage. We will also discuss the opportunities for and challenges within the biomedical applications of water-soluble fullerene derivatives.

1. Introduction

Fullerenes are well known as the first isolated stable molecular form of carbon, which not only launched a new field of organic chemistry but also, through the development of carbon nanotubes and other carbon forms, a new field of materials science [1].

Soon after their discovery during mass spectrometry analysis of the products of graphite laser evaporation, fullerenes C₆₀ and C₇₀ became available at the gram scale using the approach proposed by Huffman and Krätschmer in 1990 [2].

After the problem of fullerene mass production and purification was solved, the era of covalent modification of the fullerene cage began, and other important fundamental challenges have emerged. Namely, it was discovered that the uncontrolled addition of functional groups to the fullerene cage often led to the formation of complex mixtures of regioisomers and derivatives with different numbers of addends. For example, monofunctionalized C₆₀ has nine symmetrically non-equal 6-6 bonds that can react in a second addition (Figure 1) [3]. E,e,e-isomers are the most prominent members in the family of tris-substituted fullerenes, since they can be rather selectively synthesized under specific reaction conditions [4,5]. The functionalization of C₇₀ is even more challenging. Unlike C₆₀—whose perfect spherical geometry yields just one possible mono-adduct isomer—C₇₀ with D_5h symmetry has four inequivalent types of [6,6]-bonds defined as α, β, γ, and δ, leading to the C₇₀-based bis-adducts having 38 possible isomers on the condition of identical groups and additions occurring exclusively on [6,6]-bonds [6,7]. This inherent complexity means that the regioselective bis-functionalization of C₇₀ is a considerably more difficult task than it is for C₆₀.

Figure 1. Positional isomerism of the fullerene C₆₀ bis-adducts and types of double bonds in C₇₀.

This limitation led most researchers to focus on synthesizing monofunctionalized fullerene derivatives—products of addition reactions involving only one double bond of C₆₀. Since all double bonds on the fullerene cage are equivalent, these derivatives can be readily isolated as pure compounds. Classical functionalization methods, such as the Bingel–Hirsch [4] and Prato [8] reactions, were developed and widely adopted for this purpose. Some of the pioneers of fullerene chemistry were willing to embrace the complexity of multiple functionalization of the fullerene cage, which led to the discovery of novel reactions and addition patterns in fullerene chemistry.

Prof. R. Taylor and his team from the University of Sussex have left a remarkable legacy on the arylation [9,10], alkylation [11], alkoxylation [12], halogenation [13,14] of fullerenes and further transformation of halogenated fullerenes. Prof. L. Gang from Peking University has contributed to the peroxide chemistry of fullerenes [15] and has developed unique methods for the synthesis of open-cage [16,17], endohedral [18] and heterofullerenes [19]. Prof. E. Nakamura and colleagues discovered the unique pentaaddition reaction of organocopper reagents to C₆₀ [20,21]. Prof. S. Troyanov has developed an efficient high-temperature ampule synthesis method for a broad variety of highly functionalized derivatives of fullerenes and endometallofullerenes and succeeded in the X-ray structural characterization of >100 unique compounds bearing different halogen atoms and/or perfluroalkyl groups attached to the carbon cage [22,23,24,25].

The extreme hydrophobicity of pristine fullerenes prevents their effective dispersion in physiological fluids, hindering in vivo applications. The addition of hydrophilic polar groups, which are capable of hydrogen bonding with water molecules, is the key to achieving water solubility. This functionalization not only confers water solubility but also enhances bloodstream retention time by reducing immune system recognition of the fullerene derivatives as foreign particles [26].

The number, chemical nature, and relative position of organic addends attached to the fullerene cage determine solubility in aqueous media, tendency to aggregation, biological activity profile, and the electron affinity of fullerene derivatives.

To obtain truly water-soluble covalently modified fullerene derivatives, several polar hydrophilic groups have to be attached to the cage directly or through a linker. Depending on the type of the solubilizing groups, fullerene derivatives can be divided into cationic (mostly ammonium salts) or anionic (with attached carboxylic, phosphonate, sulfonate or hydroxyl groups). Typically, good solubility in water can be reached when the number of ionic groups exceeds four. For example, potassium salts of fullerene-based polycarboxylic acids and aminofullerenes with five cationic groups were reported to have high solubility in water (>50 mg/mL) [27]. Non-ionic individual fullerene derivatives, such as, for example, fullerene-based polyols with 24 hydroxyl groups reach solubilities of up to 75 mg/mL [28].

Thus, two different approaches to the synthesis of individual water-soluble fullerene derivatives can be proposed: (1) to utilize a branched reagent that has a required amount of hydrophilic groups in its structure and can be attached via functionalization of only one fullerene double bond; (2) to use a substituent bearing one or two polar functionalities and a protocol that provides a regioselective addition of several addends (Figure 2).

Figure 2. Two different approaches to the synthesis of the water-soluble fullerene derivatives.

The first approach was utilized to obtain one of the first water-soluble individual dendro [60]fullerene derivatives containing 18 carboxyl groups using the reaction of C₆₀ with branched malonate [29,30]. This compound was reported as one of the most effective inhibitors of HIV-1 protease in vitro.

However, the limitations of this method, based mostly on the low accessibility of dendrite hydrophilic precursors, were soon realized, and focus was shifted to the second approach.

Thus, the search for regioselective methods for the multiple functionalization of the fullerene cage is still underway and represents an important synthetic task.

In this review, we will introduce general approaches to the synthesis of individual fullerene derivatives (Section 2), discuss the most popular synthetic routes to water-soluble fullerene-based compounds, such as Bingel–Hirsch (Section 3) and Prato (Section 4) reactions, as well as the addition of organocopper reagents (Section 5). In addition, we will discuss less conventional approaches based on the substitution of halogen atoms in halogenated fullerenes (Section 6). We will also consider protocols that allow one to obtain water-soluble fullerenes with different types of functional addends, tailoring à la carte the properties of the resulting molecules (Section 7).

Finally, we will showcase the most promising results from investigations of the biological activity of individual fullerene derivatives (Section 8).

2. General Approaches to the Synthesis of Isomerically Pure Fullerene Derivatives

Despite the fact that the synthesis of numerous individual fullerene derivatives has been reported, most of them are devoted to the preparation of mono-adducts, while the simple and selective synthesis of polyaddition products still remains a non-trivial task.

Classical approaches to the functionalization of the fullerene cage allow the synthesis of compounds with 1–2 addends with acceptable selectivity. Attempts to use these methods for the synthesis of polyfunctionalized fullerene derivatives often lead to the formation of inseparable mixtures of isomers [31].

The problem of the regioselective attachment of multiple addends to the fullerene cage is being actively addressed using various approaches, which can be formally divided into covalent and supramolecular (Figure 3).

Figure 3. General approaches to regioselective synthesis of fullerene polyadducts.

The covalent approach includes methods based on both addition and substitution reactions. Several examples of regioselective addition reactions are presented in Figure 3. For example, extensive cyclopropanation of the fullerene cage, which involves five or six double bonds, leads to the selective formation of a single isomer in each case. Covalent methods that use different “guiding” spacers to provide the regioselective formation of bis- and tris-adducts of fullerenes have been actively elaborated [31].

Methods based on removable templates, such as anthracene attached to the fullerene cage, have been proposed. Template molecules are often removed using a retro-Diels–Alder reaction. This approach can be illustrated by the conversion of the bis-anthracene adduct of fullerene into e,e,e,e-C₆₀[C(CO₂Et)₂]₄ [32,33].

Another regioselective method of covalent functionalization is based on the fusion of C₆₀ [34] and C₇₀ [35] fullerenes with anthracene at high temperature, during which so-called topochemical orthogonal transposition occurs. The lattice of a fullerene derivative can serve as a template that fixes reagent molecules in the necessary position. Thus, heating a crystalline adduct of [60]fullerene with anthracene afforded trans-1-bis-adduct and fullerene in virtually quantitative yields. Such an unexpected mechanism was associated with the packing of molecules in the crystal according to the ‘head-to-tail’ type.

The regioselective synthesis of fullerene derivatives via substitution reactions represents a very convenient and versatile approach but also has several limitations, which will be discussed later in this article. Since the regioselective chlorination of fullerenes with the formation of the products C₆₀Cl₆, C₇₀Cl₈ and C₇₀Cl₁₀ has been reported, these molecules became popular precursors for the synthesis of polyfunctional fullerene derivatives through the replacement of labile chlorine atoms with organic addends bearing hydrophilic groups [36]. Alkoxy substituents were also reported as good leaving groups that can be substituted by thiol residues on the fullerene cage [37].

The supramolecular approach is based on the preliminary preparation of host–guest fullerene complexes with cycloparaphenylenes and other macrocyclic molecules, followed by the regioselective functionalization of the resulting complexes [33]. The limitations of these approaches include the specific chemical nature of substrates, multi-stage processes, and the use of complex structures to obtain supramolecular complexes.

The supramolecular approach was actively developed by several research groups, including the teams of Prof. X. Ribas and Prof. M. Costas (Universitat de Girona) [38], Prof. Y. Xu (Henan University) and Prof. M von Delius (Ulm University) [39], and Prof. G.H. Clever (TU Dortmund University) [40].

For example, the complex between C₆₀ or its derivatives and [10]CPP represents the perfect fit regarding the size of guest and host (Figure 3) [41]. The addition of one equivalent of [10]CPP to fullerene monomalonate leads to the corresponding inclusion complex, which yields two major isomers (trans-2 and trans-3) upon addition of the next malonate equivalent [42].

Another example of a supramolecular approach is the complexation of a supramolecular tetragonal prismatic nanocapsule bearing two Zn-porphyrin units and four macrocyclic Pd-based molecular clips with C₆₀, C₇₀, and C₈₄ [43]. This nanocapsule was later used as a supramolecular mask, leading to the formation of tetrakis-adduct (D_2h-symmetric e,e,e,e-regioisomer).

Another fascinating example of the supramolecular approach has been recently reported, which allowed for the selective preparation of fullerene hetero-tris-adducts. After the first addend was introduced onto C₆₀ through a cascade process involving Ru-catalyzed cycloisomerization of 1,6-enynes followed by a Diels–Alder reaction, the supramolecular mask was introduced and two pentacene addends were regioselectively attached [44].

The combination of supramolecular and covalent chemistry of fullerene yielded a series of beautiful supramolecular fullerene-based architectures, including [10]cycloparaphenylene-fullerene [2]rotaxanes [42], matryoshka-like complexes, etc. [45]. One of the pioneers in this field, Prof. F. Diederich, systematically explored the chemistry of fullerenes with a focus on functional and switchable supramolecular nanosystems [46]. The history of the development of this research area and its perspectives are summarized in several review papers [39]. In this review article, we only cover the basic idea of the supramolecular approach, since methods utilized for the further functionalization of supramolecular fullerene complexes are similar to those discussed below.

3. Cyclopropanation Reactions for the Synthesis of Individual Water-Soluble Fullerene Derivatives

The protocol for the cyclopropanation of the fullerene developed by C. Bingel and A. Hirsch is one of the most popular approaches for functionalized derivatives (Figure 4) [4]. This reaction is widely utilized for the synthesis of water-soluble fullerene derivatives, including mono- and poly-adducts. From the perspective of medicinal chemistry, monoadducts and hexakis-adducts appear to be particularly relevant, due to their occurrence as single regioisomers.

Figure 4. Bingel–Hirsch cyclopropanation reaction.

Polycationic, polyanionic, and neutral mono-adducts have been obtained over the years (Figure 5). Namely, water-soluble decacationically armed C₆₀ decaiodide mono-adduct 1 was obtained via the Bingel–Hirsch approach. The authors also reported that it can be applied as a photosensitizer and potentially a nano-PDT (photodynamic therapy) agent against pathogenic bacteria and cancer cells [47]. Highly water-soluble C₆₀-anchored multiarmed poly(acrylic acids) 2–4 were obtained through the complex multistep synthetic protocol involving fullerene cyclopropanation with a polymeric malonate precursor [48]. Another approach to the water-soluble fullerene was based on the synthesis of fullerene–cyclodextrin conjugates (compounds 5–7) [49,50].

Figure 5. Water-soluble fullerene mono-adducts obtained using Bingel–Hirsch cyclopropanation.

Further cyclopropanation of mono-adducts with the formation of 2 to 4 cyclopropane rings on the fullerene cage yields mixtures of isomeric products that require chromatographic separation (Figure 1). For bis-adducts, the formation of nine isomers is possible if addends are different, and eight isomers can be obtained if two identical addends are attached. Cyclopropanation of trans-3-bis-adduct (trans-3-C₆₀A₂) results in the preferential formation of a D₃-symmetrical tris-adduct trans-3-C₆₀A₃, in which all three addends are in trans-3-positions.

The first actual synthesis of water-soluble C₆₀ can also be attributed to Lamparth and Hirsch [51], who emphasized the critical importance of isolating single regioisomers for biomedical evaluation. They synthesized polycarboxylic tris-adducts C₃-C₆₀A₃ (compound 8) and D₃-C₆₀A₃ via NaH-promoted hydrolysis of the respective malonic esters. Respective malonic ethyl esters were obtained using cyclopropanation of individual bis-adducts followed by HPLC-assisted purification [4]. These tris-adducts have been extensively studied for a variety of biomedical applications (Figure 6) [52,53].

Figure 6. Examples of individual fullerene multiadducts obtained via cyclopropanation reaction.

Water-soluble compounds 9a–c functionalized with two diphosphonate groups were obtained via hydrolysis of the respective chromatographically separated isomeric ethyl esters [54].

Fortunately, for polyadducts with five and six cyclopropane rings, only one isomer is possible. This tendency was actively used for the synthesis of individual water-soluble penta- and hexa-adducts of fullerenes, such as compounds 10–16. Polyol methanofullerenes 10–12 were obtained via exhaustive cyclopropanation of C₆₀ with acetonide-protected precursors, followed by deprotection under acidic conditions [28].

The water-soluble glycofullerenes 13 and 14 containing D-glucosamine fragments were synthesized using the two-step modified Bingel–Hirsch methodology [55].

Polycationic fullerene hexakis adducts 15 and 16 were also obtained using the approach based on exhaustive or two-step cyclopropanation of the fullerene cage [56].

To obtain the first fullerene-based X-ray contrast agent 17, a single addend containing six iodine atoms and eight protected hydroxyl groups was appended to C₆₀ followed by the addition of four more addends, each containing four protected hydroxyl groups. Final deprotection afforded the perfectly water-soluble (>460 mg mL⁻¹), non-ionic, highly iodinated (24% of iodine) individual fullerene derivative [57].

The development of click chemistry gave a new direction to the synthesis of water-soluble fullerenes. The copper-catalyzed azide–alkyne [2+3] cycloaddition (CuAAC) approach was utilized for the synthesis of fullerene conjugates with carbohydrates, such as compounds 18 [58], 19 [59] and 20 (Figure 7) [60], bearing mannose residues.

Figure 7. Examples of individual fullerene multiadducts obtained via CuAAC conjugation.

Since the cyclopropanation reaction allows for the regioselective synthesis of C₆₀A₅A′ hexakis adducts with two different types of addends, this synthetic method was used for the CuAAC conjugation of functionalized water-soluble fullerene C₆₀A₅ moiety with biological substrates such as peptides (compound 21) and other fullerene-based residues (compound 22), thus producing supramolecular structures and so-called giant superballs [61,62].

J. Zhang and colleagues reported the synthesis of metallobuckytrios, three buckyball systems, obtained via a CuAAC click reaction (compounds 22a–f) (Figure 7) [63].

4. [2+3]Cycloaddition Reactions for the Synthesis of Individual Water-Soluble Fullerene Derivatives

Fullerene C₆₀ readily undergoes [2+3] cycloaddition with different 1,3-dipolar reagents, namely, azomethine ylides, azides, nitrile ylides, diazo compounds, nitrile imines, thiocarbonyl ylides, nitrile oxides, and carbonyl ylides [64]. Among those reactions, the method developed by Prof. M. Prato and based on the addition of azomethine ylides to the fullerene cage is one of the most convenient and high-yielding (Figure 8) [8]. It affords fulleropyrrolidines and is often utilized for the synthesis of water-soluble fullerenes. Azomethine ylides are typically obtained from α-amino acids or their derivatives and carbonyl compounds [65]. Similarly to other mono-adducts, [2+3]cycloaddition products with a single pyrrolidine unit annealed to the [60]fullerene cage produce only one regioisomer. In a comparative analysis of multiaddition regioselectivity, Lu et al. demonstrated that bis-addition via the Prato reaction offers less control over product formation compared to the Bingel–Hirsch reaction [66]. Additional difficulty in individual product purification arises from the cis/trans isomerism present in pyrollidinofullerenes due to the chiral carbon atoms in the ring. This becomes particularly important for biomedical applications, as even slight differences in molecular structure (stereoisomerism) can dramatically alter how these fullerene compounds behave in living systems, affecting pharmacokinetics and pharmacodynamics [67].

Figure 8. Synthesis of fulleropyrrolidines via addition of azomethine ylides.

An early attempt to synthesize water-soluble fullerenes via the Prato reaction was not completely successful [68]. The introduction of two triethylene glycol chains on the pyrrolidine ring of the mono-adduct (compound 23) gave a poor solubility of 3 × 10⁻⁵ M in a 9:1 water-DMSO mixture [69].

However, the introduction of several ionic groups such as carboxylic and positively charged amines (compounds 24–28) provided higher solubility of the monofunctionalized fulleropyrrolidines in water (Figure 9).

Figure 9. Several examples of water-soluble monofunctionalized fulleropyrrolidines.

A recent review by Zhang et al. summarizes the data on the biological application of fullerene mono-adducts [70].

Isomerically pure fulleropyrrolidines with multiple addends represent a versatile class of materials with tailored properties for a wide range of applications. Marchesan et al. synthesized isomeric bis-fulleropyrrolidinium salts (29–36) and pegylated polycationic fullerene derivatives (Figure 10) [71,72].

Figure 10. Examples of individual fullerene multiadducts obtained via [2+3]cycloaddition reaction.

5. Selective Multiaddition of Organocopper Reagents to Fullerene

In 1996, Prof. E. Nakamura and colleagues reported that a phenylcopper reagent derived from a Grignard reagent and a copper(I) halide adds five times to [60]fullerene and with a complete regioselectivity in quantitative yield to produce penta(aryl)[60]fullerenes C₆₀Ar₅H (Figure 11) [20,73]. Synthesized pentaaryl [60]fullerene anions C₆₀Ar₅⁻·K⁺ were reported to be soluble in water and form spherical bilayer vesicles [74].

Figure 11. Synthesis of penta(aryl)[60]fullerenes via the addition of phenylcopper reagents.

Pyridine-modified organocopper reagents are used for single-step decaaddition reactions to [60]fullerene to obtain deca(organo)[60]fullerenes C₆₀R₁₀H₂ (Figure 12) [73]. Importantly, this approach also allows for the sequential addition of two different types of aromatic addends. The developed protocol was efficiently utilized for the synthesis of deca(aryl)[60]fullerenes bearing five ferrocenyl and five ester groups [74]. Thus, it can also be used for the synthesis of water-soluble fullerene derivatives with unique addition patterns.

Figure 12. Synthesis of deca(aryl)[60]fullerenes.

Importantly, the regioselective multiaddition of organocopper reagents to the fullerene cage allowed for the introduction of useful functional groups such as hydroxyl, thiol, ester, and carboxylic groups. Examples of neutral, anionic, and cationic fullerene derivatives, obtained via the addition of organocopper compounds, are presented in Figure 13 [75,76,77].

Figure 13. Fullerene derivatives synthesized via selective multiaddition of organocopper reagents.

6. Chlorinated and Fluorinated Fullerenes as Precursors for the Synthesis of Water-Soluble Fullerene Derivatives

6.1. Chlorofullerenes as Precursors for Water-Soluble Fullerene Derivatives

Chlorofullerenes C₆₀Cl_n were obtained in 1991, thus becoming one of the first isolated fullerene derivatives [78,79]. In 1993, the group of Prof. R. Taylor synthesized and characterized isomerically pure chlorofullerene C₆₀Cl₆ using the reaction of fullerene with iodine monochloride in benzene [13]. Since then, the method for the synthesis of C₆₀Cl₆ has been optimized, chlorination reaction intermediates have been isolated and characterized, and the mechanism of the chlorination reaction has been proposed [80].

The optimized procedure (Figure 14) allows the quick synthesis of pure C₆₀Cl₆ (>97%, HPLC) with quantitative yields starting from the solution of C₆₀ in 1,2-dichlorobenzene [81]. The resulting compound is soluble in aromatic solvents such as toluene and chlorobenzenes.

Figure 14. Facile synthesis of chlorofullerene C₆₀Cl₆.

A similar procedure utilizing C₇₀ as a precursor allows us to obtain C₇₀Cl₈, which has the distinct feature of being insoluble in most organic solvents. However, it is possible to overcome this limitation using either a soluble analog C₇₀Cl₁₀ or performing the heterogeneous reaction under elevated temperatures with the gradual substitution of Cl atoms and solubilization of the formed products [14].

Various halogenofullerenes have been considered as precursors for the synthesis of functionalized fullerenes via substitution reactions. However, isomerically pure fluorinated fullerenes are not as easily accessible, bromofullerenes have limited stability and are poorly soluble, and iodofullerenes have not been isolated. Thus, chlorofullerene C₆₀Cl₆ is the optimal precursor for the synthesis of functional fullerene derivatives because it simultaneously possesses the following required properties: (1) can be quickly and easily obtained in isomerically pure forms with high yields in gram quantities; (2) has high solubility in organic solvents; (3) chlorine atoms in the structure can be easily substituted by various organic addends; (4) the original C_s-symmetrical (cyclopentadienyl-type) halogen addition pattern is preserved during the substitution reactions, allowing the regioselective formation of functional derivatives; (5) six introduced addends is enough to easily change the properties of the molecule (including solubility in water) in a desired manner via substitution reactions.

Additionally, C₆₀Cl₆ has a specific reactivity due to the presence of two chlorine atoms in the 1,2-position, allowing the selective replacement of five chlorine atoms with bulky addends, while the sixth halogen atom is typically not substituted due to steric hindrance. The substitution of the remaining chlorine atom can be carried out during further synthetic steps, which opens up possibilities for the synthesis of compounds with several types of addends. This approach will be discussed further in detail.

The chemical properties of C₆₀Cl₆ have been investigated rather intensively during previous years. Most of the approaches that allow us to obtain water-soluble fullerene derivatives from C₆₀Cl₆ are based on its reactions with nucleophiles containing hydrophilic groups in their structures.

6.2. Reactions with Amines and Amino Acids

The reaction of C₆₀Cl₆ with amines can be utilized as a method for the synthesis of water-soluble fullerene derivatives if ammonium groups are formed during the synthesis or during further quaternization or protonation reactions. For example, the interaction of C₆₀Cl₆ with N-substituted piperazines can be used as an effective method for the synthesis of polycationic water-soluble fullerene derivatives (Figure 15). The reaction is performed in toluene and leads to the immediate precipitation of polycationic products, which have shown exceptionally high solubility in water (>300 mg mL⁻¹) [82].

Figure 15. Preparation of polycationic water-soluble fullerene derivative with attached N-methylpiperazine hydrochloride residues.

Amino acid-functionalized water-soluble fullerene derivatives represent a particularly promising class of compounds, combining the inherent biological compatibility of amino acids with the diverse bioactive potential of fullerene systems.

While initial synthetic efforts primarily produced complex mixtures of fullerene–amino acid adducts [83,84,85], recent advances have enabled the preparation of isomerically pure derivatives. They can be obtained via the reaction of C₆₀Cl₆ with tert-butyl esters of amino acids followed by the cleavage of tert-butyl groups with a strong acid such as trifluoroacetic (Figure 16) [86].

Figure 16. Reaction of C₆₀Cl₆ with hydrochlorides and hydroacetates of tert-butyl esters of amino acids.

Chlorofullerene C₆₀Cl₆ also reacts with methyl esters of amino acids, forming compounds with similar structures. These functionalized fullerenes have been successfully employed to construct drug-loaded vesicles, demonstrating effective encapsulation of clinically relevant anticancer agents, including 5-fluorouracil, cyclophosphamide, and cisplatin [87]. However, methyl esters are not suitable for the synthesis of highly soluble polycarboxylic derivatives due to the absence of a synthetic protocol allowing the deprotection of ester groups without degrading the C-N bonds between the fullerene cage and amino acid-based addend.

Although reactions of chlorofullerene with N-nucleophiles allow us to obtain a wide variety of water-soluble derivatives, the formation of C-N bonds between the carbon cage and solubilizing addends is undesirable due to the comparatively low stability of these compounds in aqueous media. Aqueous solutions of fullerene derivatives with C-N bonds between the cage and addend typically degrade through partial hydrolysis and the depletion of attached residues, which leads to a loss of solubility. Thus, the development of alternative approaches to the synthesis of amino acid- and protein-modified fullerene derivatives is of great importance.

6.3. Reactions with Alcohols and Hydroxy Acids

The reaction of chlorofullerene with alcohols was one of the first reactions discovered and reported for chlorofullerenes. Prof. R. Taylor and colleagues demonstrated that prolonged heating of C₆₀Cl₆ solution in benzene and aliphatic alcohol leads to the partial elimination of chlorine atoms and the formation of 1,4-dialkoxy fullerene derivatives [12]. Notably, the reaction of C₆₀Cl₆ with alkali metal alcoholates leads to the formation of alkoxy fullerene derivatives C₆₀(OR)₅Cl with quite low yields (Figure 17).

Figure 17. Reactions of C₆₀Cl₆ with alcohols and alkali metal alcoholates.

The reaction of C₆₀Cl₆ with methanol under aerobic conditions in the presence of dimethyl sulfoxide (DMSO) yields the methoxylated derivative C₆₀(OCH₃)₄ (Figure 18) [88]. This transformation involves the elimination of two chlorine atoms, generating a reactive double bond that demonstrates remarkable selectivity in subsequent reactions with azides and tosylhydrazones.

Figure 18. Synthesis and functionalization of the fullerene derivative C₆₀(OCH₃)₄.

Fullerene derivatives bearing five alkoxy residues attached to the fullerene cage can be obtained in fairly high yields using the reaction of chlorofullerene C₆₀Cl₆ with alcohols in the presence of tetrabutylammonium bromide or triethylamine. Depending on the reaction conditions, the developed method allows one to obtain alkoxy derivatives with the general formula C₆₀(OR)₅X (X=Cl, Br, H), C₆₀(OR)₄O or C₆₀(OR)₂ (Figure 19) [89].

Figure 19. Synthetic routes to diverse alkoxy-functionalized fullerene derivatives.

Although reactions of C₆₀Cl₆ with alcohols do not directly yield water-soluble fullerene derivatives, the observed reactivity patterns become particularly valuable when employing hydroxy acid esters instead of aliphatic alcohols. This approach enabled the synthesis of the first water-soluble polycarboxylic fullerene derivative featuring C-O bonds between the carbon cage and solubilizing groups (Figure 20) [89].

Figure 20. Synthesis of the fullerene derivative with the attached residues of hydroxy acids.

Fullerenols 1,4-C₆₀(OH)₂ and C₆₀(OH)₅Br were also synthesized via tetrabutylammonium bromide-promoted reactions of C₆₀Cl₆ with water [90]. Performing the synthesis in wet chlorobenzene produced 1,4-C₆₀(OH)₂ as the only isolable product along with substantial amounts of non-functionalized C₆₀. However, the addition of DMSO as a co-solvent altered the reaction pathway, leading to the formation of C₆₀(OH)₅Br as the main product (Figure 21).

Figure 21. Synthesis of fullerenols 1,4-C₆₀(OH)₂ and C₆₀(OH)₅Br.

In the presence of tetramethylethylenediamine (TMEDA), C₆₀Cl₆ also reacts with alcohols, with the formation of alkoxy derivatives C₆₀(OR)₅Cl and C₆₀(OR)₄(OH)Cl (Figure 22) [91].

Figure 22. Preparation of alkoxy derivatives of fullerene in the presence of TMEDA.

The obtained fullerene derivatives 58a,d with four alcohol residues and a hydroxyl group can be used as precursors for the facile synthesis of open-cage fullerenes (Figure 23) [91].

Figure 23. Synthesis of open-cage fullerene derivatives with alkoxy residues.

Thus, the substitution of chlorine atoms in C₆₀Cl₆ with functionalized alcohols can also be considered an efficient pathway to water-soluble fullerenes with interesting molecular structures, including open-cage compounds. However, partial hydrolysis (the depletion of -OR groups with the formation of fullerenols) should be considered a limitation of this approach.

6.4. Reactions with Thiols and Mercapto Acids

Chlorofullerene C₆₀Cl₆ reacts readily with thiols and mercapto acids to yield derivatives containing five organic addends and a hydrogen atom attached to the fullerene cage. This method offers a fast, simple, and highly selective route to water-soluble fullerene derivatives. The transformation proceeds under exceptionally mild conditions (ambient temperature, aerobic atmosphere) using diisopropylethylamine as a base, yielding C₆₀(SR)₅H derivatives in excellent yields (47–95%). Notably, these products typically require no chromatographic purification, significantly simplifying the isolation process (Figure 24) [92].

Figure 24. Reaction of C₆₀Cl₆ with functional thiols and mercapto acids.

Under slightly modified conditions (15-crown-5 ether, no base), C₆₀Cl₆ reacts with sodium sulfanylsulfonates to yield fullerene derivatives bearing sulfonate groups. Similar to their polycarboxylic analogs, compounds containing five SO₃Na groups precipitate directly from the reaction mixture and can be isolated in pure form without chromatographic purification (Figure 25) [92].

Figure 25. Synthesis of water-soluble fullerene derivatives using the reaction of C₆₀Cl₆ with sulfanylsulfonates.

The main limitation of this method is the poor long-term stability of the resulting compounds, both as solid powders and in aqueous solution. Freshly prepared C₆₀(SR)₅H derivatives gradually decompose to yield pristine C₆₀, thiols, and disulfides (Figure 26). Although this instability makes them less suitable for biomedical applications, thermocleavable C₆₀(SR)₅H derivatives have found their application as precursors for the environmentally friendly fabrication of fullerene-based organic field-effect transistors [93].

Figure 26. Thermal decomposition of fullerene derivatives with attached thiol residues.

6.5. Reactions with Phosphites

Similar to the classic Arbuzov (Michaelis–Arbuzov) reaction, where alkyl halides (RX) react with trialkyl phosphites (P(OR′)₃ to form phosphonates (RP(O)(OR′)₂) [94], chlorofullerene C₆₀Cl₆ undergoes an analogous transformation during the reaction with trialkyl phosphites. This reaction yields fullerene phosphonate derivatives C₆₀[P(O)(OR)₂]₅H in good to excellent yields (23–94%) (Figure 27) [95].

Figure 27. Reaction of C₆₀Cl₆ with trialkyl phosphites.

Notably, fullerene derivatives bearing some alkyl phosphonate groups exhibit good water solubility. Although the phosphonate moieties (-PO(OR)₂) can be hydrolyzed to phosphonic acid groups (-PO(OH)₂) using bromotrimethylsilane to yield anionic water-soluble compounds and enhance solubility (Figure 28), this deprotection step proceeds under harsh conditions.

Figure 28. Strategies for the synthesis of phosphonate water-soluble fullerene derivatives.

To avoid harsh deprotection conditions, a more direct approach based on the C₆₀Cl₆ reaction with tris(trimethylsilyl) phosphite (P(OTMS)₃) was proposed. TMS groups in the structure of the fullerene derivative are easily depleted via water/acid treatment [96].

6.6. Reactions with C-H Acids and Enol Silyl Ester

Chlorofullerene C₆₀Cl₆ also regioselectively reacts with CH-acids and enol silyl ester (Figure 29) [97]. Both reactions lead to the formation of C_s-symmetrical compounds with four aliphatic addends that can be further selectively converted to fullerene derivatives with a cyclopentadienyl-type addition pattern, similarly to C₆₀(OCH₃)₄ derivatives, shown in Figure 18.

Figure 29. Reaction of C₆₀Cl₆ with CH-acids and enol silyl ester.

6.7. Reactions with Aromatic Compounds

As demonstrated above, halogen atoms on the fullerene cage can be easily substituted with a broad range of organic residues attached to the fullerene cage via C–N, C–P, C–S and C–O bonds. Although most of these reactions proceed in one step and with nearly quantitative yields, many of the resulting fullerene derivatives appeared to be not fully stable if stored as aqueous solutions.

In contrast, water-soluble fullerene derivatives with organic addends attached to the carbon cage via C–C bonds are inherently stable, can be stored in aqueous media, and, therefore, appear to be one of the most promising types of water-soluble fullerene derivatives for biomedical applications.

Several approaches to form C-C bonds with a fullerene cage via the substitution of chlorine atoms have been reported, including C₆₀Cl₆ reactions with organolithium compounds [98,99,100], allyltrimethylsilane [101], arylacetylenes [102], and organic cyanides [103].

One of the most versatile approaches to creating water-soluble compounds is based on Friedel–Crafts arylation with benzene- or thiophene-based compounds. C₆₀Cl₆ reacts with aromatic compounds via the electrophilic substitution reaction in the aromatic ring. The substitution of chlorine atoms by aryl or heteroaryl residues (Ar) occurs directly in the aromatic hydrocarbon medium or in a suitable inert solvent (Figure 30) [9,104]. When substituted benzenes are used as substrates, the reaction predominantly yields para-substituted products with the general structure C₆₀Ar₅X (where X = Cl or H).

Figure 30. Friedel–Crafts arylation reaction of chlorofullerene C₆₀Cl₆.

Notably, when phenol or hydroquinone derivatives are used in arylation reactions, several products with non-typical structures can be formed (Figure 31) [105].

Figure 31. Friedel–Crafts arylation reaction of chlorofullerene C₆₀Cl₆ with phenol ethers.

In particular, the reaction of C₆₀Cl₆ with alkyl phenols and alkyl phenyl ethers allows us to obtain non-classical products with a dihydrobenzofuran fragment fused to the cage (Figure 32) [106].

Figure 32. Interaction of C₆₀Cl₆ with phenols in the presence of a base.

The arylation of chlorofullerene C₆₀Cl₆ with dimethyl-2,2′-(1,4-phenylenebis(oxy))-diacetate, catalyzed by SnCl₄, leads to the formation of a mixture of five different products, which were isolated using preparative HPLC. In addition to the two classical products of the reaction C₆₀Ar₅X (X=H, Cl), two derivatives with dihydrobenzofuran fragments and one compound with a cyclobutane fragment were also isolated (Figure 33) [107].

Figure 33. The reaction of C₆₀Cl₆ with hydroquinone ether leading to the formation of a mixture of products with atypical structure.

Water-soluble polycarboxylic derivatives can be synthesized via an arylation reaction using ArH substrates containing ester groups. The subsequent hydrolysis of these esters to carboxylic acids, followed by treatment with K₂CO₃ or Na₂CO₃, yields the target water-soluble salts.

Recently, a more efficient approach based on the direct arylation of C₆₀Cl₆ with unprotected carboxylic acids was developed (Figure 34) [108]. This method significantly reduces the number of synthetic steps and the overall reaction time. Moreover, since it does not include an acidic hydrolysis step, this strategy is suitable for the synthesis of fullerene–amino acid conjugates.

Figure 34. Strategies for the preparation of water-soluble fullerene derivatives with aromatic addends.

However, the scope of this method is severely limited to those substrates that could tolerate the rather harsh conditions of the Friedel–Crafts arylation reaction (high temperature, the presence of a Lewis acid).

That is why an alternative approach to the synthesis of fullerene derivatives loaded with amino acid residues through the acylation of amino acid tert-butyl esters with in situ-formed fullerene-based chloroanhydride was proposed (Figure 35) [109]. The optimized conditions of this reaction do not impose a limitation on the structure of the amino acid and allow the facile synthesis of numerous adducts starting from a single fullerene-based precursor.

Figure 35. Acylation of tert-butyl esters of amino acids with fullerene-based chloroanhydrides formed in situ.

Since the sixth chlorine atom attached to the fullerene cage usually cannot be substituted with a large aromatic residue due to steric hindrances, the possibility of its substitution with smaller functional groups and various rearrangements was investigated.

One of the first discovered substitution reactions was the interaction of pentaarylated fullerene derivative C₆₀Ar₅Cl with water and triphenylphosphine with the formation of C₆₀Ar₅H adducts [6]. The substitution of the Cl atom with alkyl residue is possible if the fullerene derivative C₆₀Ar₅Cl is treated with trialkylphosphite under “abnormal” Arbuzov reaction conditions (Figure 36) [110].

Figure 36. Chemical transformations of fullerene derivatives C₆₀Ar₅Cl.

When C₆₀Ph₅Cl interacts with the Lewis acid, heterolytic dissociation of the C-Cl bond leads to the formation of the antiaromatic C_5v-symmetric carbocation [C₆₀Ph₅]⁺. One of the phenyl groups undergoes a 1,2-shift, forming a more stable carbocation C_s-[C₆₀Ph₅]⁺. The position of the positive charge in the C_s-[C₆₀Ph₅]⁺ cation was confirmed using the isolation and characterization of the fullerene derivative C_s-[C₆₀Ph₅]CN obtained via the treatment of C_s-[C₆₀Ph₅]⁺ with TMSCN reagent; however, this result was only mentioned in a review paper (Figure 36) [111].

The reaction of C₆₀Ph₅Cl with BrCN in the presence of FeCl₃ results in the substitution of the chlorine atom followed by the substitution of the hydrogen atom of the adjacent phenyl ring (Figure 36). The treatment of the product with alkali results in the formation of the 1-isoquinolone derivative [112].

The treatment of fullerene derivatives C₆₀Ar₅Cl with hydroxylammonium sulfate and TMEDA (Figure 37) yields C₆₀Ar₅(NHOH) derivatives (82–83a) [113]. The subsequent reaction of 82–83a with phosphorus pentachloride produces aziridine-containing derivatives 82–83b, which feature both an aziridine ring and a chlorine substituent.

Figure 37. Synthesis and transformation of fullerene derivatives C₆₀Ar₅(NHOH).

Unexpectedly, attempts to replace the chlorine atom in 82–83b with an –NHOH group (using hydroxylammonium sulfate/TMEDA) did not yield the anticipated products. Instead, this transformation led to the formation of derivatives 82–83c containing a pyridinone fragment.

Further reactivity studies revealed that the treatment of 82–83b with Lewis acids (FeCl₃ or SnCl₂) results in partial conversion to 82–83c, along with the formation of unique derivatives 82–83d–f.

Gan et al. demonstrated that chlorofullerene C₆₀Cl₆ reacts with pyrroles and pyrazines, forming products with five residues attached via either C-C or C-N bonds (Figure 38). Pyrroles and pyrazines bearing carboxylic or other polar groups can potentially be utilized for the synthesis of water-soluble derivatives [114].

Figure 38. Reactions of C₆₀Cl₆ with pyrroles and pyrazoles.

Thus, chlorofullerenes, particularly C₆₀Cl₆, serve as highly reactive intermediates for the efficient synthesis of water-soluble derivatives. Their multiple chlorine substituents enable diverse functionalization pathways, facilitating attachment of various functional groups with versatile addition patterns.

6.8. Fluorinated Fullerenes as Precursors for Water-Soluble Fullerene Derivatives

It should be noted that efficient gram-scale synthetic methods are currently available only for the halogenated fullerenes C₆₀F₁₈, C₆₀Cl₆, C₇₀Cl₈ and C₇₀Cl₁₀. Similarly to C₆₀Cl₆, C₆₀F₁₈ also exhibits high selectivity in reactions with nucleophiles.

The reactivity of fluorofullerene C₆₀F₁₈ was actively investigated by Prof. R. Taylor and colleagues. C₆₀F₁₈ can be prepared using the reaction between C₆₀ and potassium hexafluoroplatinate under elevated temperatures [115]. They revealed that the fluorine atoms in C₆₀F₁₈ can be substituted with aryl groups during the reaction with ArH in the presence of FeCl₃, yielding C₆₀F₁₅Ar₃ (Figure 39) [116]. This reaction holds significant potential for the synthesis of water-soluble fullerene derivatives if the Ar substituent incorporates two or more polar hydrophilic groups, such as a carboxylic group.

Figure 39. Arylation of fluorofullerene C₆₀F₁₈.

The reaction of C₆₀F₁₈ with carbanionic nucleophiles follows a distinctive S_N2″ mechanism, leading to the formation of trannulenes with the general formula C₆₀F₁₅R₃. This S_N2″ pathway is exclusive to the chemistry of fluorinated fullerenes and involves nucleophilic attack at a δ-carbon atom, triggering a shift in two conjugated double bonds and subsequent elimination of fluorine (Figure 40) [117]. Several unique water-soluble trannulenes were synthesized using this approach.

Figure 40. Synthesis of water-soluble trannulene from C₆₀F₁₈.

The 18-membered aromatic macrocycle provides trannulenes with high stability and unique optoelectronic properties. However, they were later found to undergo thermal isomerization to triumphenes with high yields (70–95%) [118]. Thus, thermal isomerization of trannulenes offers a pathway to the synthesis of new triumphenes with aliphatic addends, which cannot be obtained directly from fluorofullerene C₆₀F₁₈ (Figure 41). Further hydrolysis of ester groups can lead to the formation of water-soluble triumphenes with aliphatic addends.

Figure 41. Synthesis of triumphenes from trannulenes.

7. Fullerene Derivatives with Several Types of Addends

Since even the introduction of several identical addends to the fullerene cage often leads to the formation of isomeric mixtures, the regioselective synthesis of fullerene derivatives bearing several types of addends represents an even more complex synthetic challenge.

While several examples of statistical synthesis for fullerene derivatives bearing two distinct functional groups have been reported [119], these methods suffer from significant limitations. The lack of regioselectivity in these reactions typically results in complex product mixtures, making the isolation of individual compounds challenging even with preparative HPLC.

Among the approaches that allow us to introduce two types of addends on the fullerene cage in a regioselective manner, sequential Bingel–Hirsch reactions (Figure 42) can be given as an example [4]. First, the mono-adduct is obtained in good yield by carefully controlling the stoichiometry of the malonate. In the second step, adding an excess of another 1,3-dicarbonyl compound completes the octahedral addition pattern.

Figure 42. Sequential Bingel–Hirsch reactions for the regioselective synthesis of polyfunctionalized fullerenes.

The malonate addition sequence can be reversed to 5 + 1 if the strategy based on the retro dipolar cycloaddition of nitrile oxides to the fullerene cage is applied (Figure 43) [120].

Figure 43. Synthesis of fullerene derivatives with two types of addends using retro dipolar cycloaddition of nitrile oxides.

Fullerene derivatives featuring two different addends can also be prepared via palladium-catalyzed three-component C₆₀ alkoxyarylation (Figure 44) [121]. However, this approach yields only 1,4-adducts with a low degree of functionalization, precluding its application for the synthesis of water-soluble fullerene derivatives.

Figure 44. Palladium-catalyzed three-component alkoxyarylation of C₆₀.

Heteroannulation of C₆₀ with α-bromo acetamides or anilides followed by electrochemical functionalization also yields fullerene derivatives with both benzyl-type addends and a fused lactam moiety (Figure 45) [122].

Figure 45. Cu(I)-catalyzed synthesis of fullerene-fused lactams and further electrochemical functionalization.

The synthesis of fullerene derivatives bearing three types of functional addends on the cage is so far represented by the reaction of C₆₀Ph₄(CH₂SiMe₃)H with [CpFe(CO)₂]₂ (Figure 46) [123] and the reaction of C₆₀ with acetone under basic conditions coupled with intermediate quenching with benzylbromide derivatives [124].

Figure 46. Regioselective synthesis of fullerene derivative with three types of addends.

Most of these reactions have a limited synthetic scope, and, with a couple of exceptions, they can hardly be used for the introduction of the solubilizing polar groups required for the synthesis of water-soluble compounds with tailored biological properties.

Water-soluble fullerene derivatives with two or more types of addends were synthesized either via sequential cyclopropanation, as discussed in Section 2, or via the approach based on C₆₀Ar₅Cl reactions with substituted thiophenes, which enables the substitution of Cl with thiophene residues and the formation of the novel family of C₁-symmetrical C₆₀ fullerene derivatives with six functional addends C₆₀Ar₅Th. Fullerene derivatives C₁-C₆₀Ar₅Th’ also react with another equivalent of substituted thiophene Th″H, leading to the formation of the unique family of water-soluble compounds C₁-C₆₀Ar₅Th′Th″H with three types of functional addends (Figure 47) [125].

Figure 47. Synthesis of novel fullerene derivatives C₁-C₆₀Ar₅Th′₂H and C₁-C₆₀Ar₅Th′Th″H.

8. Biological Activity of Individual Water-Soluble Fullerene Derivatives

Despite two decades of active development of synthetic methods for individual polyfunctionalized fullerene derivatives, most studies on the biological activity of water-soluble fullerene-based compounds have been conducted using complex, inseparable mixtures of fullerenols and related substances. However, when the goal is mechanistic understanding, therapeutic development, or regulatory approval, isomerically pure fullerene compounds should be used. Mixed-structure materials are acceptable only for early-stage exploratory work, non-clinical uses, or bulk applications—situations where high reproducibility and detailed mechanistic knowledge are not critical.

Regardless their rather limited synthetic accessibility, individual water-soluble fullerenes were investigated both in vitro and in vivo. Key research directions include antiviral and antimicrobial activity, antitumor, antioxidant and antiischemic properties, neuroprotection, radiation protection, drug delivery systems, bioimaging, and stem cell modulation (Figure 48).

Figure 48. Biological activity of water-soluble fullerene derivatives: general research directions.

8.1. Fullerene–Cell Interactions and Localization

Several studies implicated energy-dependent endocytotic pathways in the uptake of functionalized fullerenes. For example, Curley et al. have reported this for the individual water-soluble compound 102, which after uptake was preferentially localized in the cell nucleus (Figure 49) [126]. The uptake of an individual water-soluble fullerene-substituted phenylalanine poly-lysine derivative with an FITC label 103 involves a complex, ATP-dependent endocytic mechanism mediated by caveolae/lipid rafts, F-actin, and microtubules, likely proceeding through a scavenger receptor-specific pathway [127].

Figure 49. Individual fullerene derivative with endocytotic uptake pathway.

Malonic acid-derivatized mono-adduct C₆₀(COOH)₂ and hexa-adduct C₆₀(COOH)₆ localize predominantly to the mitochondria [128,129], whereas the isomerically pure bis-adduct C₆₀(C(COOH)₂)₂ localizes to lysosomes [130].

The protein corona fundamentally redefines the biological identity, cellular interactions, and in vivo fate of nanoparticles. The complex interactions between fullerenes and proteins fall into three primary subgroups: π–π stacking involving fullerene sp² carbons and aromatic amino acid residues, van der Waals contacts between the C₆₀ cage and the protein surface, and hydrophobic interactions (nonpolar solvation) [131]. Fullerenes can interact with serum proteins so that a protein corona is formed around the nanoparticles, but the detailed composition of corona is still unknown [132].

Research has shown that protein binding with the C₆₀ nanocomplex did not change the charge properties of the protein. The proteins induced the aggregation of the C₆₀ nanocomplex [133]. Bai et al. have recently developed an artificial intelligence model for the accurate prediction of protein targets of individual isomerically pure amphiphilic fullerene derivatives, which will be extremely promising for the rational design of fullerene-based therapeutics, drug discovery, and precision nanomedicine [134].

8.2. Acute Toxicity

The pharmacokinetic behavior and toxicological effects of fullerenes are highly influenced by their route of administration and are closely related to their functionalization [135]. Extensive studies confirm that pristine C₆₀ exhibits no acute or sub-acute toxicity across diverse biological systems, including bacteria, fungi, human leukocytes, and model organisms such as Drosophila, mice, rats, and guinea pigs [136].

Short-term in vitro studies have revealed that high concentrations of unmodified C₆₀ can induce rapid cell membrane damage and apoptosis in certain cell lines [137], but this effect is mitigated when fullerenes are properly solubilized. Thus, early claims of the toxicity of fullerene derivatives can be associated with the difficulty of their removal from organisms and the use of co-solvents, such as tetrahydrofuran, for solubilization.

Acute toxicity in mice was investigated for several individual water-soluble fullerenes, including arylated compounds, compounds with attached residues of aminoacids, thioacids, etc. (Figure 50) [138]. Most of them demonstrated low toxicity with LD₅₀ values exceeding 300 mg/kg.

Figure 50. Acute toxicity of individual water-soluble fullerene derivatives.

8.3. Antiviral Properties

The active investigation of their anti-HIV activity began in 1993 when Shinazi and colleagues [139] reported the anti-HIV activity of a water-soluble fullerene derivative in infected cells, while Friedman et al. [140] proposed allosteric HIV-1 protease inhibition as a possible mechanism of antiviral activity. According to molecular docking results, C₆₀ bis(phenethylaminosuccinate) was able to interact with the active site of HIV-1 protease, and these theoretical findings were supported by cell-free HIV-1 protease inhibition assays [140]. Further works by Toniolo et al. [141], Friedman et al. [142], and Strom et al. [143] revealed a wide range of fullerene derivatives able to inhibit the catalytic activity of HIV-1 protease in micro- and nanomolar concentrations in cell-free experiments. Marchesan et al. showed that five isomeric bis-fulleropyrrolidinium salts 29–33 are inhibitors of HIV-1 and HIV-2, finding that the positions of the side chains influence antiviral activity, and trans- isomers of the fullerene bis-adducts are more active than the cis- one [71]. The same group also investigated the anti-HIV properties of pegylated polycationic fullerene derivatives 34–36 and showed that positive charges near to the C₆₀ backbone could increase potency, while longer solubilizing chains seem to induce cytotoxicity [72]. Polycarboxylic fullerene derivatives derived from chlorofullerenes C₆₀Cl₆ and C₇₀Cl₈ were investigated as broad-spectrum antivirals. The most efficient compounds 104a–c inhibited HIV at concentrations of 150–350 nM.

A water-soluble fullerene conjugate with cyclodextrin 5 was shown to be an efficient hepatitis C virus entry inhibitor [49].

The most promising inhibition of herpes simplex virus (HSV) was demonstrated for the compounds incorporating thiophene residues. For example, compound 105 demonstrated a selectivity index (SI) of 40,380 (Figure 51) [144].

Figure 51. Polycarboxylic fullerene derivatives demonstrating promising antiviral properties.

Water-soluble conjugates with cyclodextrin 6 and 7 efficiently inhibited the Influenza A (H1N1) virus in vitro in micromolar concentrations [49]. Among a wide range of pentacarboxylic fullerene derivatives, maximum virus-inhibiting activity and minimum toxicity were found for the compound containing residues of salts of 2-amino-3-cyclopropylpropanoic acid (CC₅₀ > 300 µg/mL, IC₅₀ = 4.73 µg/mL, SI = 64) [145].

Several recent reviews have identified fullerenes as a promising platform for developing next-generation antivirals against SARS-CoV-2, particularly through the inhibition of viral enzymes [146,147,148]. Computational studies have largely focused on either pristine C₆₀ and C₇₀ fullerenes [149,150] or their functionalized derivatives [151,152,153,154], probing interactions with a range of SARS-CoV-2 targets. These include the proteases Mpro (3CLpro) and PLpro, RNA-dependent RNA polymerase, the spike glycoprotein and its human receptor hACE2, the nucleocapsid protein’s RNA-binding domain, and even combinations of fullerene derivatives with anti-COVID drugs like favipiravir and chloroquine. Petitjean used atomic force microscopy to screen multivalent fullerene-based glycoclusters capable of disrupting the interaction between the SARS-CoV-2 spike protein and the glycan 9-O-acetylated sialic acid (9-AcSA), a molecule believed to be critical for viral attachment to host cells [155]. Despite strong interest in the antiviral potential of fullerene derivatives against coronaviruses—SARS-CoV-2 in particular—only few experimental studies have so far assessed their activity using either viral enzymes or infected cells. In one, Katagishi et al. tested several fullerene derivatives for their ability to inhibit the SARS-CoV-2 main protease [156]. A polycarboxylic malonic acid-type derivative 106 and pyrrolidino [60]fullerene derivative 107 emerged as the most effective, achieving 50% protease inhibition at just 0.53 and 0.98 μM. Another study introduced polycarboxylic fullerene derivatives 108a–c (Figure 50) with cilastatin-based addends representing promising antiviral agents for COVID-19 treatment (IC₅₀ = 390–620 nM) [157].

In vitro screening of the individual fullerene derivative C₆₀(P(O)(OK)₂)₅H showed that it demonstrates broad-spectrum antiviral activity against a panel of ten distinct viruses, including Varicella zoster virus, cytomegalovirus, HSV, cowpox virus, feline coronavirus, feline herpes virus, etc. [96].

Glycoconjugates 18–20 with mannose residues attached to a fullerene cage synthesized by Prof. Nierengarten and colleagues effectively inhibited the DC-SIGN-dependent (dendritic cell-specific intercellular adhesion molecule-3-grabbing nonintegrin) entry of Ebola virus envelope glycoprotein EboGP pseudotyped viral particles into cells. Most of the conjugates effectively bound to the DC-SIGN receptor, blocking its interaction with viral glycoproteins at a concentration as low as 0.667 nM [60].

Cell entry of Zika (ZIKV) and dengue viruses (DENVs) also involves several receptors, including DC-SIGN. Glycoconjugates containing 36–360 mannose residues in their structure were obtained via the strain-promoted cycloaddition of azides to alkynes (SPAAC) strategy. These compounds also efficiently blocked the DC-SIGN receptor through multivalent interactions. They demonstrated remarkable antiviral activity against Zika and dengue viruses with 50% inhibitory doses (IC₅₀) in the nano- and picomolar ranges. Tridecafullerene 109 (Figure 52) appended with 360 1,2-mannobioside showed the best IC₅₀ values for both the ZIKV (67 pM) and the DENV (35 pM) [61]. Considering that no vaccine or specific antiviral drug has yet been approved for these viral tropical fevers, the obtained results look promising and can be used for the development of effective antiviral drugs.

Figure 52. Tridecafullerene appended with 360 1,2-mannobioside units using the SPAAC synthetic pathway. Figure adapted from ref. [61].

8.4. Antitumor Properties

Some water-soluble fullerene derivatives exhibit promising antitumor activity through selective mechanisms that spare normal cells. For example, amino acid- and thioacid-functionalized derivatives 110–119 and 62o effectively inhibit lung cancer cell proliferation by inducing either autophagy or apoptosis (Figure 53) [86,158,159,160]. Structural analysis through QSAR modeling reveals that specific molecular features, particularly thiophene moieties, enhance their inhibitory effects against various non-small cell lung carcinoma cell lines [158]. Importantly, these compounds show low acute toxicity in mice, distinguishing them from conventional chemotherapeutics. Some derivatives demonstrate the ability to cross the blood–brain barrier, offering potential dual functionality—either promoting neural regeneration or suppressing glioblastoma growth, depending on their specific functional groups [159]. Their anticancer activity appears to be closely linked to molecular structure, with aromatic bonds, sulfur-containing aromatic rings, and oxygen atoms promoting the inhibitory effects on H460 and H1299 carcinoma cell lines [160]. These findings highlight the potential of fullerene derivatives as a versatile platform for developing targeted, low-toxicity anticancer therapies with possible applications in both peripheral and central nervous system diseases. The ability to fine-tune biological effects through the strategic modification of functional groups opens new possibilities for precision cancer treatment.

Figure 53. Water-soluble fullerene derivatives demonstrating promising anitumor properties.

Fullerenes serve as efficient photosensitizers in photodynamic therapy due to their unique ability to absorb light and generate cytotoxic reactive oxygen species (ROS), which selectively destroy cancer cells or pathogenic microorganisms. A number of [light-harvesting antenna]–fullerene structures have already been studied for their application in PDT [161,162], including covalently linked BODIPY−C₆₀ [163], methylpheophorbide−C₆₀ [164], and chlorin e6−C₆₀ dyads [165]. Fullerenes also show great promise as antitumor drug carriers due to their unique cage-like structure, which allows for the efficient encapsulation and delivery of chemotherapeutic agents [166].

8.5. Neuroprotective Properties

Fullerene derivatives, including several individual compounds, act as cell protectors under induced neurotoxicity conditions [167,168,169].

Additionally, fullerene derivatives were shown to mitigate ischemia–reperfusion injury by scavenging ROS and reactive nitrogen species (RNS) that are overproduced during oxygen deprivation [170,171].

PEGylated, hydroxylated fullerenes and individual polycarboxylic compounds 120–122 (Figure 54) inhibit β-amyloid aggregation and destabilize preformed fibrils in vitro and in transgenic Alzheimer’s mouse models [172,173]. Similar effects were observed with α-synuclein [174].

Figure 54. Fullerene derivatives as DNA vectors.

Fullerene derivatives 119–121 were observed to induce neural stem cell (NSC) proliferation in vitro and rescue the function of injured CNS in zebrafish. Individual water-soluble fullerenes with different functional groups act on mitochondria that can either induce or reduce the oxidation to achieve neuroprotective or antitumor effects in different types of cells [159].

8.6. Fullerene Derivatives as Radioprotectors

The remarkable antioxidant properties of fullerenes and their derivatives positioned them as promising radioprotective candidates in the early 2000s, when sufficient quantities became available for biological investigation [175]. Initial studies examined both pristine fullerenes and polyhydroxylated mixtures C₆₀(OH)_x and C₇₀(OH)_x, along with several individual compounds 111, 123–124 (Figure 55). The review paper by Krokosz et al. summarizes most of the data on the radioprotective activity of fullerene derivatives until 2016 [176].

Figure 55. Molecular structures of individual water-soluble fullerene derivatives investigated as potential radioprotectors.

More recent research discusses the influence of non-modified hydrated fullerene [177], and individual compound 124 [178], on cell cultures and animals treated with high-energy electrons, X-rays, and γ-rays. Kostuyk et al. demonstrated that polycarboxylic compound 111 has a stabilizing effect on cell nuclei under the action of 1 Gy radiation [179].

The comprehensive evaluation of these nanomaterials, conducted using in vitro and in vivo animal models, has demonstrated significant radioprotective activity, manifesting as enhanced survival rates in irradiated biological systems, attenuated oxidative stress markers, reduced amounts of radiation-induced DNA damage, and lower intracellular ROS levels.

8.7. Fullerene Derivatives as DNA and RNA Vectors

DNA transfection is the process of intentionally introducing exogenous DNA into the interior of a living cell. Compound 8 (trans-3-C₆₀A₃) was utilized as a DNA vector and was able to transfect cells grown in culture with DNA carrying the green fluorescent protein (Figure 56) [53].

Figure 56. Fullerene derivatives as DNA/RNA vectors.

siRNA transfection is a molecular biology technique used to deliberately introduce small interfering RNA (siRNA) molecules into eukaryotic cells. siRNA transfection techniques have been actively developed for the tetra-piperazino-[60]fullerene epoxide 125 derivative [180]. Importantly, 125 is non-toxic and achieves lung-targeted in vivo siRNA delivery by forming micrometer-sized complexes with siRNA and serum proteins. These complexes stabilize siRNA, protecting it from degradation under physiological conditions [181]. Recently Wang et al. designed a fullerene-based nanocomplex based on 125 to achieve the pulmonary delivery of siRNA via inhalation. Through the encapsulation of programmed death-ligand 1 (PD-L1) siRNA, this nanocomplex significantly inhibited the progression of metastatic lung cancer in a mouse model, without causing obvious adverse events or toxicity [182]. Polycationic fullerene hexakis adducts 15 and 16 were tested as small interfering RNA (siRNA) vehicles in vitro using the prostate cancer cell line DU145 expressing green fluorescent protein. Compound 15 was an efficient siRNA transfection agent, while glycofullerene 16 was inactive in the transfection experiments due to its high zeta potential and the formation of an extremely stable complex with siRNA [56].

8.8. Bactericidal Effects of Fullerene Derivatives

In addition to their antiviral properties, certain fullerene monoderivatives have demonstrated significant antibacterial activity (Figure 57). For example, the fullerene–cyclen derivative 27 [183], exhibited potent antibacterial effects against E. coli and S. aureus. Furthermore, a fullerene–glycopeptide conjugate derived from teicoplanin 126, showed enhanced antibacterial activity against teicoplanin-resistant E. faecalis [184]. In vitro studies demonstrated that polycationic fullerene derivatives, including compound 45, exhibit potent bactericidal effects against both Gram-negative (E. coli) and Gram-positive (B. subtilis) bacterial strains. This activity arises from the compound’s ability to assemble on negatively charged cell walls through electrostatic interactions, ultimately disrupting cellular integrity [82].

Figure 57. Individual fullerene derivatives possessing antibacterial properties.

8.9. Fullerene Derivatives in Biomedical Imaging

The accumulation of fullerenes in cells is particularly difficult to measure using standard fluorescence microscopy because their fluorescence is barely detectable in aqueous media [185]. By making an otherwise-invisible nanoparticle visible, fluorescent fullerenes allow researchers to track both the location and activity of these particles simultaneously.

In contrast to pristine fullerenes, multi-functionalized fullerene derivatives exhibit remarkable fluorescent properties due to the partially reduced π-conjugation of the fullerene cage. One approach to achieving this involves the direct chemical modification of the fullerene cage through multi-functionalization. For instance, Rubin, Wudl, and colleagues reported a C₆₀ hexaadduct that exhibited significantly improved fluorescence compared to unmodified C₆₀ [186]. Similarly, Nakamura and co-workers prepared various C₆₀ multiadducts that showed greatly enhanced luminescence relative to pristine C₆₀, achieving a maximum quantum yield of 0.24 [187]. Gan et al. have reported that the emission wavelength of C₇₀ derivatives obtained via the functionalization of C₇₀(OMe)₈ can be fine-tuned over a wide range from blue to red [188]. Deng et al. have demonstrated that isomerically pure fullerene derivatives C₇₀(OCH₃)₁₀[C(COOEt)₂] and C₇₀(OCH₃)₁₀[C(COOEt)₂]₂ exhibited enhanced blue fluorescence emission compared to C₇₀(OCH₃)₁₀ [189]. The second approach to obtaining fluorescent fullerene derivatives is their conjugation with separate fluorescent molecules. For instance, a fullerene-based photosensitizer has been successfully synthesized by introducing a fullerene unit onto rhodamine B hydrazide [190]. Under acidic conditions, the resulting covalent conjugate, which initially adopts a non-fluorescent spirolactam structure, can be activated to its ring-opened form. This activation accomplishes enhanced visible-light absorbance, fluorescence turn-on, and triplet excited state generation. The fluorescence turn-on and efficient singlet oxygen generation enabled by pH-activated fullerene-based particles testify to their great potential applications for cancer diagnosis and treatment.

T. Cai et al. have developed tumor-targeted C₆₀ conjugates, which can be detectable using positron emission tomography (PET) imaging. Fullerene malonic acid tris-adduct was covalently conjugated with eight-arm amine-terminal branched polyethylene glycol. Water-soluble functionalized C₆₀ conjugates were radio-labeled with ⁶⁴Cu and modified with cyclo (Arg-GlyAsp) peptides (cRGD) to target integrin α_vβ₃ in glioblastoma. Although the obtained material was not an individual compound, this approach demonstrates the usability of individual tris-adducts for PET imaging [191]. Virta et al. have created ¹⁸F-radiolabelled molecular spherical nucleic acids, assembled on a [60]fullerene core via the Bingel–Hirsch reaction and click chemistry. The biodistribution of the compound was monitored in vivo using positron emission tomography/computed tomography (PET/CT) imaging [192].

Gadofullerenes, a class of gadolinium-based metallofullerenes, represent a promising new type of MRI contrast agents with high relaxivities [193]. However, only a few individual water-soluble gadofullerenes were synthesized and investigated. For example, water-soluble Gd metallobuckytrios 22a–f exhibited superior T₁ relaxivity compared to the typical Gd complexes, potentially superseding current clinical MRI contrast agents in both safety and efficiency. The Lu MBTs generated reactive oxygen species upon light irradiation, showing promise as photosensitizers [63].

Other examples are mostly represented by polyhydroxylated, carboxylated, and PEGylated–hydroxylated complex mixtures of gadolinium-based metallofullerenes [194].

8.10. Myogenic Activity

Phosphonate C₆₀ fullerene derivatives, especially compound 64, enhanced myogenic differentiation in human mesenchymal stem cells (MSCs) by upregulating key myogenic transcription factors (MYOD1, MYOG, MYF5, and MRF4) while suppressing osteogenic and adipogenic gene expression. The observed myogenic induction may be linked to autophagy stimulation, potentially triggered by intracellular alkalosis or transient oxidative stress caused by the fullerene derivatives. These findings suggest that water-soluble C₆₀ derivatives could be promising candidates for regenerative therapies targeting muscle disorders and injuries directing the myogenic differentiation of stem cells [96,195].

8.11. Other Types of Biological Activity

Compound 21 exhibits a reactive SpyTag fragment on the fullerene cage utilized for its facile conjugation with SpyCatcher-bearing proteins through the formation of an intermolecular isopeptide bond. The water solubility of this compound was ensured by the addition of multiple residues bearing hydroxyl groups. The efficient, cell-compatible SpyTag–SpyCatcher reaction enabled the native conjugation between fullerenes and protein molecules [62].

The water-soluble glycofullerenes 13 and 14 containing D-glucosamine fragments were proposed as a novel class of non-receptor tyrosine kinase inhibitors, particularly found to inhibit Fyn A tyrosine kinase and Bruton’s tyrosine kinase with IC₅₀ values in the low micromolar range [55]. Several individual polycarboxylic C₆₀ and C₇₀ fullerene derivatives were proposed as protein tyrosine phosphatase inhibitors [196].

9. Conclusions and Outlook

Fullerene C₆₀ and its functional derivatives exhibit unique properties such as high electron affinity, radical scavenging ability, photodynamic activity, ability to inhibit a wide range of pathogens, etc. However, their poor solubility in water limits their applications in the biomedical field. The development of efficient, tunable synthetic methodologies for water-soluble fullerenes is essential for bridging the gap between their remarkable theoretically predicted properties and real-world applications.

In this review, we have highlighted some general strategies and approaches to the synthesis of individual water-soluble fullerene-based compounds. Namely, classic approaches, such as Bingel–Hirsch and Prato reactions, were discussed; the regioselective multiaddition of organocopper reagents was also introduced as an efficient route to polyfunctional fullerenes. The role of chlorofullerenes as precursors for the synthesis of water-soluble fullerene derivatives was discussed. Importantly, recently developed methodologies for the regioselective synthesis of fullerene derivatives with several addend types were analyzed.

Moving forward, challenges for the biomedical applications of fullerenes remain in their regioselective synthesis and its scalability. The continued development of structurally defined fullerene derivatives holds great promise for drug discovery, nanomedicine, and beyond. Future research should focus on optimizing these materials for practical applications while deepening our understanding of their biological properties and interactions with biological matter.