Homogeneous catalytic C(sp 3 )–H functionalization of gaseous alkanes

The conversion of light alkanes into bulk chemicals is becoming an important challenge as it eﬀectively avoids the use of prefunctionalized alkylating reagents. The implementation of such processes is, however, hampered by their gaseous nature and low solubility, as well as the low reactivity of the C–H bonds. Efforts have been made to enable both polar and radical processes to activate these inert compounds. In addition, these methodologies also benefit significantly from the development of a suitable reactor technology that intensifies gas–liquid mass transfer. In this review, we critically highlight these developments, both from a conceptual and a practical point of view. The recent expansion of these mechanistically-different methods have enabled the use of various gaseous alkanes for the development of different bond-forming reactions, including C–C, C–B, C–N, C–Si and C–S bonds.


Introduction
Gaseous alkanes are among the most abundant and cheapest carbon-based feedstocks, which are generally used as a source of energy for heating, propulsion or electricity generation. 1 In conjunction with these uses, methane, ethane, and heavier analogues could also be employed as building blocks for chemical synthesis. 2 This is an increasingly fascinating area of research as the conversion of greenhouse gases into chemical feedstocks and/or bulk chemicals would be of extreme value for the chemical industry as well as for the environment. Indeed, the direct use of alkanes would also bypass the need for functionalized reagents, e.g. haloalkanes and organometallic reagents. 3 However, there are several issues that need to be addressed in order to realize this goal ( Fig. 1), such as the general inertness of the C-H bonds and the poor solubility of the gases in common organic solvents. Moreover, the desired products are often easier to activate than the gaseous alkanes, leading to issues related to over-reaction. 4 While huge attention has been directed to the use of methane to produce methanol, 5 less focus has been devoted to its valorization for the production of different methylcontaining derivatives. Additionally, the vast majority of these processes are carried out under heterogeneous conditions, 5a,c which are well known to pose an additional problem in terms of selectivity. Notably, reports that exploit the use of a homogeneous catalyst, outside of the field of methanol formation, 5b are rare and remain largely underdeveloped.
The scope of this article is to cover the state-of-the-art in the functionalization of gaseous alkanes under homogeneous catalytic conditions, with the aim of showcasing the main achievements, as well as the different activation mechanisms and the future challenges within the field. We have divided the review based on the nature of the bond formed after the functionalization of the gaseous alkane, namely C-B, C-C, C-N, C-Si and C-S. Specifically for the C-C bond formation, processes able to induce chirality have also been highlighted. Regarding the formation of C-O bonds, possibly the most studied reaction for the functionalization of methane and other gaseous alkanes, this topic has been recently reviewed by Periana and co-workers 5b and will not be discussed within this article.

C-B bond formation
Protocols capable of converting feedstock chemicals such as light alkanes in added-value building blocks are highly sought after. In the field of catalytic C-H functionalization, particular interest has been directed towards the borylation of inert aliphatic substrates. 6 This is motivated by the high synthetic versatility of the resulting organoboron compound, which can be easily converted in a wide range of other functional groups such as amines, alkenes and alcohols or can be engaged in C-C bond forming reactions. 7 Earlier works demonstrated that Re, Rh, Ru and Ir are all competent catalysts for the homogeneous borylation of liquid alkanes. 6f,h,j Specifically, Rh-based catalysts are capable of forming metal bis(boryl) complexes after oxidative addition of a diboron precursor (B 2 pin 2 , Fig. 2). 6a,b These species have been reported to react with alkanes under neat conditions, leading to the formation of alkyl-metal boryl complex I. This intermediate undergoes reductive elimination to produce one molecule of the borylated product and metal hydride II. The latter reacts with the borylating reagent affording intermediate III that, upon extrusion of HBpin, 6h is able to activate C(sp 3 )-H bonds to restore active species I. Another plausible and concomitantly operating reaction pathway concerns the oxidative addition of the formed HBpin, leading in this case to the formation of the borylated alkyl species along with molecular hydrogen. 6a However, from a thermodynamic viewpoint, diboron species determine a greater enthalpic driving force due to the weaker B-B bond compared to the B-H bond in boranes. 8 Despite these literature precedents, translating these thermal catalytic protocols to the activation of C-H bonds within light alkanes faced some major challenges. The first problem to address is the choice of an appropriate solvent that should not outcompete the gaseous alkane during the C-H activation step.

Timothy Noël
Timothy Noël currently holds a position as Full Professor at the University of Amsterdam, where he is the chair of Flow Chemistry. His research interest ranges from organic chemistry to chemical engineering and encompass more specifically flow chemistry, organic synthesis and synthetic catalytic methodology development. His work received several awards, including the DECHEMA prize (2017), the Hoogewerff Jongerenprijs (2019), the IUPAC-ThalesNano prize in flow chemistry and microfluidics (2020) and the KNCV Gold Medal (2021). He also serves as editor-in-chief of Journal of Flow Chemistry.
Specifically, the higher BDE of methane (bond dissociation energy, BDE = 105 kcal mol À1 for methane) compared to other liquid alkanes poses a unique challenge. 5b In addition, the over-functionalization of the product is another likely scenario given the presence of more labile C-H bonds compared to the ones present in the gaseous alkane. [9][10][11] The borylation of gaseous hydrocarbons was pioneered by the groups of Sanford 9 and Mindiola, 10 who simultaneously reported two transition metal-catalyzed methodologies capable of borylating methane and ethane under thermal conditions.
In their report, Sandford and coworkers 9 identified optimal conditions for the borylation of methane using a high-pressure batch reactor employing Rh-based catalyst [Cp*Rh(Z 4 -C 6 Me 6 )], 35 bar of methane and B 2 pin 2 1 as the limiting reagent in cyclohexane at 150 1C (Fig. 3). The choice of the catalyst was made based on earlier density functional theory (DFT) studies reported by Hall et al., proving the theoretical feasibility of the process. 12 Crucial for the success of the desired transformation was the choice of the solvent: cyclohexane granted good solubility of the catalyst without compromising the selectivity. Indeed, due to the more strained and hindered nature of the C-H bonds in cyclohexane, only traces of the borylated solvent byproduct 3 were observed. 13 Under the optimized reaction conditions, high selectivity and yield were obtained for the mono-borylated methane 2. Additionally, by testing other known catalysts previously employed in the borylation of heavy alkanes, the authors demonstrated that the fine-tuning of the catalyst structure had significant impact on the selectivity of the process. 6a The relationship between steric hindrance and selectivity was further corroborated by a competition experiment using known molar quantities of methane and ethane: all the tested catalysts preferentially functionalized the less sterically-encumbered methane.
Simultaneously, the group of Mindiola 10 reported an elegant iridium-catalyzed borylation of methane (Fig. 4). Given the wellknown efficacy of iridium-based catalytic systems for the borylation of inert C-H bonds, 6f,14 the authors conducted an extensive screening of commercially available iridium catalysts and nitrogen-based ligands. Preliminary results revealed that the highest conversion and selectivity for the monoborylated product 2 were obtained in the presence of an Ir(I) precatalyst and L 1 (3,4,7,8-tetramethyl-1,10-phenanthroline) as ligand using tetrahydrofuran as a solvent under 35 bar of methane at 120 1C. Under these conditions, the desired product was obtained in 4.1% yield (determined via gas-chromatography (GC) analysis). The mechanism, as predicted by DFT calculations, begins with the substitution of the ligand to afford Ir(III) complex V, which is the putative resting state of the catalyst. After methane coordination at one of the vacant sites of the complex, the rate-determining oxidative addition occurs via transition state A yielding intermediate VII. Ligand reorganization, followed by reductive elimination, affords the desired product 2 and complex IX, which further regenerates the catalyst resting state by reacting with B 2 pin 2 . Since the C-H bond cleavage implies the oxidation of Ir(III) to Ir(IV) within intermediate VI (transition state A), the authors postulated that, in light of their reduced Lewis basicity, phosphine ligands might enhance the polarizability of the metal center, thus facilitating oxidative addition. Indeed, employing L 2 as ligand under the optimized conditions, the desired product was  obtained in 52% yield with a 3 : 1 chemoselectivity over the monofunctionalized compound. Isotopic labeling studies using 13 CH 4 confirmed that the light alkane is the methylating reagent under the described reaction conditions.

C-C bond formation
Carbon-carbon bond-forming reactions are essential tools for the synthetic community as these bonds constitute the framework of every organic molecule. Within this field, the installment of light alkyl fragments, especially the methyl one, can strongly alter the binding affinities of the parent molecule, boosting its potency as a drug, and has therefore deep implications in medicinal chemistry. 15 However, these reactions are generally performed using reactive prefunctionalized alkyl derivatives, either halides or organometallic ones. As such, the direct use of gaseous alkanes would avoid unnecessary functionalization steps and would generally provide a more sustainable process.
Several types of C-H functionalization strategies have been exploited to engage volatile alkanes directly in carbon-carbon bond-forming reactions. The first one, reported by Fujiwara and co-workers, 16 is based on a transition-metal-catalyzed C-H activation mechanism. 17 In these seminal contributions, the authors proved how light alkanes could be converted into carboxylic acids 6 through the action of palladium and/or copper catalysis, using CO as the carboxylating reagent and K 2 S 2 O 8 as the oxidant in an autoclave (Fig. 5). Specifically for the palladium catalysis, the authors suggest an electrophilic C-H activation of the gaseous alkane by the catalyst to afford intermediate XI. The latter can undergo CO insertion to yield carbonyl-palladium species XII that ultimately is converted into the target carboxylic acid 6 while the palladium catalyst X is regenerated by the oxidant (1-41% yield). Notably, the same group also expanded this mode of action to the aminomethylation of gaseous alkanes by employing in situ generated iminium ions in place of carbon monoxide as electrophilic partners. 18 Later on, Lin and Sen 19 described a rhodium-based catalytic system able to promote the same carboxylation reaction of methane under much milder conditions, employing water instead of trifluoroacetic acid as solvent and oxygen as the oxidant.
In 2020, Chang and co-workers reported a copper-catalyzed process to promote the cross-dehydrogenative-coupling (CDC) 20 between gaseous alkanes and polyfluorinated arenes 7 (Fig. 6). 21 After an extensive optimization screening, the authors identified b-diketimine L 3 as the optimal ligand for the copper catalyst, t-butyl peroxide as the oxidant, t-BuONa as the base in benzene at 80 1C. Apart from the efficient functionalization of liquid and solid alkanes, the authors demonstrated how, in a high-pressure reactor, also n-butane and propane could be productively converted into the corresponding arylated products 8 (39-51% yield).
The authors performed a thorough mechanistic investigation, encompassing both experimental and computational work. Overall, the process relies on a dual activation mode, where the copper catalyst is activating the C-H bonds of both polyfluorinated arene 7 and the alkane. Specifically, active copper complex XII reacts with t-butyl peroxide to afford tBuO and Cu(II) intermediate XIII. The latter is responsible for both the activation of the alkane through a radical pathway and of the fluoroarene via a polar mechanism. The ensuing Cu(III) species XVIII undergoes reductive elimination to afford product 8 and close the catalytic cycle.
Apart from direct C-H activation methods, the groups of Asensio, Etienne and Pérez 22 reported in 2011 a method that takes advantage of the formation of a metal carbenoid species 23 between a silver-based catalyst L 4 Ag and ethyl diazoacetate 9 (Fig. 7). This intermediate is capable of inserting into the C-H  bond of methane and ethane, producing ethyl acetate and propionate respectively. Two parameters were crucial for the good outcome of the whole process. First of all, because of the pronounced reactivity of metal carbenoids and of the strength of the C-H bonds within methane and ethane, as well as the low solubility of the gaseous reagents, the choice of the reaction solvent becomes extremely important. Indeed, the use of a solvent possessing weaker C-H bonds with respect to methane would lead to the functionalization of the former. Consequently, the authors used supercritical CO 2 (ScCO 2 ) in a highpressure reactor thermostated at 40 1C; ScCO 2 provides the required inertness and displays complete miscibility with methane. The second crucial parameter is the choice of the ligand to tune the reactivity of the metal catalyst. In order to activate a strong chemical bond such as the C-H bonds in methane and ethane, a highly electrophilic metal carbenoid is required. The authors, based on their experience in the development of coinage metal complexes to catalyze C-H insertion reactions, 24 selected catalyst L 4 Ag due to the enhanced electrophilic properties granted by the ligand. Under the optimized conditions, methane and ethane were converted into ethyl acetate and propionate respectively with good efficiency (21-30% yield) taking into account the difficulty of the transformation. As for the mechanism, in analogy with other metal carbenoid-promoted C-H functionalizations, the authors proposed a metal-catalyzed N 2 extrusion from compound 9 to form silver carbenoid XIX that subsequently undergoes carbene transfer onto methane or ethane to give the corresponding products 10.
A third strategy used to drive the C-H functionalization of gaseous alkanes relies on the formation of radical intermediates. The advent of photoredox catalysis 25 has enabled the generation of open-shell species under extremely mild conditions, namely light illumination in the presence of substoichiometric quantities of a photocatalyst at room temperature. 26 Upon photoexcitation, the catalyst has enhanced redox properties and can either promote single electron transfer (SET) events or directly cleave C(sp 3 )-H bonds through a hydrogen atom transfer (HAT) pathway. 27 Notably, both these mechanisms of activation have been used with the aim of functionalizing gaseous alkanes.
In 2018, the group of Wu reported a photocatalytic process that could promote the functionalization of ethane by using hydrochloric acid as an efficient HAT catalyst (Fig. 8). 28 The authors, based on previous findings on the photocatalytic value of chlorine radicals as HAT reagents, 29 envisioned that, upon single electron oxidation, HCl itself could be used to generate Cl under catalytic conditions. Nonetheless, the development of such a process poses several challenges. First of all, because of the gaseous nature of both ethane and hydrochloric acid, the use of classical batch reactors was not a viable option. Moreover, because of the necessity to establish uniform irradiation to efficiently promote the photochemical steps, the use of high pressure reactors was also discarded. Within the field of photocatalytic transformations, microreactor technologies have provided a platform based on short length scales which ensure homogeneous irradiation of the reaction mixture while simultaneously guaranteeing efficient gas-liquid mass transfer to saturate the gaseous reagents in the reaction mixture. 30 Taking this into consideration, the group of Wu decided to develop the aforementioned process under micro-scale conditions using narrow capillaries. Specifically, the authors employed a so-called ''stop-flow'' microtubing (SFMT) reactor, 31 consisting of a capillary equipped with two shut-off valves at each end of the reactor (PFA (perfluoroalkoxy alkanes), I.D. = 762 mm, V = 3.0 mL). Under the optimized conditions, several electronpoor olefins 11 could be functionalized with ethane, affording  adducts 12 in good yields (45-95% yield). Mechanistically, the reaction kicks off with the photoexcitation of the organic photocatalyst Mes-AcrClO 4 to afford species *Mes-AcrClO 4 that features enhanced oxidizing properties (E RED = 2.05 V vs. SCE) 32 and is therefore able to take one electron from the chloride anion (E OX = 2.03 V vs. SCE), 33 affording Cl . The latter can perform HAT on ethane to produce hydrochloric acid alongside the nucleophilic ethyl radical XXII that is intercepted by electron-poor olefin 11 to afford radical XXIII. This intermediate is prone to single electron reduction by the reduced form of the photocatalyst (XXI) to yield, after protonation by hydrochloric acid, product 12 while regenerating both the catalysts. As a general limitation, only strongly electron-poor olefins could be employed due to the fleeting nature of the ethyl radical, which requires a fast trapping by the acceptor, and due to the limited oxidizing properties of the radical adducts to turn-over the photocatalytic cycle. Moreover, under these conditions, the functionalization of methane proved to be unsuccessful, presumably due to the too high BDE of its C-H bonds.
The group of Noël developed a direct photocatalytic HAT method, catalyzed by tetrabutylammonium decatungstate (TBADT, Fig. 9). 34 Addressing the same problems faced by the group of Wu, namely the need for a uniform irradiation and an efficient gas-liquid mass transfer, the authors developed a continuous-flow procedure, using the commercially available Thanks to the good photocatalytic properties of TBADT, 35 as well as the use of back-pressure regulators to enable higher pressures within the microflow reactor that completely solubilizes the gaseous alkane into the liquid phase, these processes enabled the functionalization of several substrates. The scope of the reaction is quite broad for what concerns the gaseous alkanes: i-butane, propane, ethane and even methane could all be employed to afford the corresponding products 13 in good to excellent yields (38-93% yield). Mechanistically, the process is initiated by the absorption of a near-UV light irradiation by TBADT to reach an electronically excited state. Upon photoexcitation, the ensuing TBADT* is able to split C(sp 3 )-H bonds within the alkane substrate, affording carboncentered nucleophilic radical XXV which is trapped by the electronpoor olefin 11 to generate the second radical intermediate XXV. The latter, having now an opposite polarity, is able to take back the hydrogen atom from the reduced form of the photocatalyst, TBADTH, to afford product 13 and turn over the catalyst.

Enantioselective C-C bond formation
The previously described catalytic processes have shown the potential of both transition metal and photocatalysis to enable the conversion of feedstock chemicals such as gaseous alkanes into added-value compounds. Despite these undeniable contributions, the development of asymmetric variants would be highly relevant as it could serve as a platform for the asymmetric installment of light alkyl fragments. Enantioselective catalysis is one of the most efficient chemical approaches to target chiral compounds with high optical purity and in an economical and environmental benign way. As such, the possibility to combine the aforementioned processes for the activation of gaseous alkanes and enantioselective catalysis would greatly enrich the chemical toolbox of the synthetic community.
The group of Davies 36 expanded the reactivity of catalytically generated metal carbenoid species 23 towards the asymmetric C(sp 3 )-H functionalization of alkane derivatives. 37 Within this seminal contribution, the authors also demonstrated the viability of i-butane as starting alkyl substrate (Fig. 10). As the reaction was performed in batch, for this specific compound, the process was performed at À12 1C to condense the gaseous reagent and use it directly as solvent. Moreover, in order to exert stereocontrol over the process, the authors used a chiral dirhodium-based catalyst, Rh 2 (S-DOSP), responsible for the formation of chiral carbenoid intermediate XXVII that can insert into the C-H bond of i-butane and ultimately yield the enantioenriched product 15 (20% yield, 75% ee).
Similarly to the organometallic approach, also the photocatalytic HAT approach has been combined with enantioselective catalysis. The development of asymmetric photocatalytic methods has been one of the major area of research in the past two decades, 38 however the enantioselective photochemical functionalization of gaseous reagents has remained underdeveloped so far. In 2019, Gong and co-workers 39 reported a procedure that merged photo-HAT activation and chiral Lewis acid catalysis to  enable the stereocontrolled functionalization of imine derivatives 16 with, amongst others, propane and i-butane (Fig. 11). Specifically, this cooperative system is composed by a chiral bisoxazoline ligand chelated to a copper metal center as the Lewis catalyst and pentacenetetrone (PT) as the photocatalyst under blue light irradiation for 48 hours. The mechanism of the reaction relies on the visible-light excitation of the photocatalyst to afford a biradical species that is able to perform HAT on the alkyl derivatives, affording radical intermediates XXIX and XXX. In parallel, the copper-based Lewis acid is chelated by imine substrate 16 affording complex XXXI which is reduced by XXX to generate the second open-shell intermediate XXXII and turnover the photocatalytic cycle. At this point, the authors propose a stereocontrolled intermolecular radical-radical coupling, driven by the chiral ligand of the Lewis acid catalyst, to afford the desired product 17 (35-79% yield, 39-69% ee). Despite the fact that both these methodologies are affected by either low chemical yields and/or limited enantiomeric excess, they serve as a valuable proof of concept for the development of asymmetric procedures for the functionalization of gaseous alkanes.

C-N bond formation
Given the widespread presence of carbon-nitrogen bonds in biologically active molecules and pharmaceuticals, new methods capable to rapidly establish these bonds are of high interest for the scientific community. 40 Despite remarkable achievements in this field, 41 the development of new catalytic methodologies addressing the direct conversion of earthabundant inert alkanes into valuable fine chemicals featuring a C-N bond is highly attractive from an industrial perspective. Moreover, installing nitrogen-containing functional groups in gaseous hydrocarbons by C-N bond forming processes has proved to be highly challenging.
The first example of such a transformation was reported by Crabtree and co-workers describing the formation of imines from methane and ammonia employing Hg photosensitization (Fig. 12). 42 The protocol takes advantage of the formation of an exciplex [Hg-NH 3 ]*, where ammonia is strongly bound to the mercury atom. 43 Indeed, the so-formed three electrons bond ensures proximity and facilitates an energy transfer process from mercury to ammonia, thus promoting N-H bond cleavage. 44 The authors envisioned that an electrophilic N-centered radical (N-H BDE = 107 kcal mol À1 ) might promote the C-H abstraction of inert gaseous alkanes. 45 In order to prevent further over-functionalization of the product, the authors devised a reactor where a peristaltic pump continuously introduces the gas mixture in a photolysis chamber and a cold trap (À20 1C), which ensures condensation of the products. When an equimolar mixture of alkane/ammonia,   containing a trace of mercury vapor was irradiated with UV light (254 nm), the corresponding imine 18 was obtained. Mechanistically, after C-H bond cleavage by the HAT reagent, the alkyl radical undergoes radical-radical coupling with the N-centered radical. The ensuing amine can undergo two consecutive HAT events to yield the final imine.
More recently, the group of Zuo reported a photocatalytic C-N bond formation protocol capable of converting methane, ethane and higher alkanes into the corresponding protected N-alkylated hydrazines 20 by using a homogeneous ceriumbased photocatalyst CeCl 3 and an alcohol co-catalyst subjected to 400 nm light irradiation (Fig. 13). 46 Under the optimized conditions, using a standard pressure batch reactor, methane, ethane, propane and n-butane were all converted to the desired products in good to excellent yields (15-97% yield) using a variety of diazodicarboxylates 19 as radical traps. Additionally, the authors proved the scalability of the methodology by using a continuous-flow reactor. 47 For ethane, an excellent yield of 90% was obtained after 6 minutes of residence time and a high productivity of 2.0 mmol h À1 was accomplished at 15 bar pressure and ambient temperature. The same transformation using methane as the alkylating agent led to a maximum of 15% yield due to the low pressure limit of the reactor employed, which hampered adequate solubilization of the gas. Furthermore, the authors demonstrated the possibility to establish new C-C bonds by intercepting the nucleophilic radicals with both electron poor olefins and electron deficient heteroaromatic scaffolds. 48 Concerning the reaction mechanism, the authors initially postulated that a photoinduced ligand-to-metal charge transfer (LMCT) of a transient Ce(IV)-alkoxide species ([CeCl 5 (OCH 2 CCl 3 )] 2À ) could lead to the formation of an oxygen-centered radical, responsible for the HAT step. However, recent studies by Schelter, Walsh and co-authors convincingly showed that oxygen-centered radicals are not involved in the catalytic cycle. 49 Indeed, titration of cerium pre-catalyst ([NEt 4 ] 2 [CeCl 6 ]) with increasing amounts of CCl 3 CH 2 OH did not determine any change in its UV-visible absorption spectrum. Preparation of the Ce(IV)-alkoxide complex ([CeCl 5 (OCH 2 CCl 3 )] 2À ) was pursued, and UV-visible spectrum of this species displayed lower absorptivity and a blue shift, thus demonstrating the absence of this intermediate. Additionally, preliminary studies demonstrated that visible light irradiation of ([NEt 4 ] 2 [CeCl 6 ]), formed in situ from CeCl 3 and [NnBu 4 ]Cl, promotes the formation of Cl upon LMCT. 50 Kinetic studies showed that the initial rate of photolysis of ([NEt 4 ] 2 [CeCl 6 ]) has no dependence on the concentration of HOCH 2 CCl 3 . In addition, control experiments confirmed the presence of chlorine-radical trapping products. Based on these observations, an alternative mechanistic scenario was suggested in which a photoinduced LMCT of [Ce IV Cl 6 ] 2À generates a chlorine radical atom which is capable of performing HAT on the alkane substrate, thus generating radical XXXIV that is intercepted by the electrophilic nitrogen source 19. The formed N-centered radical intermediate XXXV is then reduced by [Ce III Cl 5 ] 2À through a single electron transfer and protonated by HCl affording the targeted N-alkylated protected hydrazine 20.
Recently, the groups of Hartwig and Pérez, reported a copper-catalyzed C-H amidation protocol using gaseous hydrocarbons as alkylating agents (Fig. 14). 51 Under the optimized conditions, using t-butyl peroxide as both the oxidant and HAT reagent and CuI-phenantroline complex as the catalyst, ethane, propane, n-butane and i-butane were smoothly converted to the corresponding N-alkyl amides 22 (6-94% yield). Regarding the   selectivity of the process, substrates containing two different C-H bonds were functionalized at both sites yielding mixtures of regioisomers. For all the tested gaseous alkanes, N-methylbenzamide 23 was detected as a side-product and accounted for the remaining mass balance. The recurring presence of 23 implies the involvement of a t-butoxy radical in the catalytic cycle, which is prone to undergo b-Me scission, 52 thus generating methyl radical XL which ultimately leads to the formation of the methylated side-product. Indeed, removal of the alkane substrate led to the selective formation of 23. In order to distinguish between the two pathways when using methane as a substrate, the amidation of the latter was carried out employing the perdeuterated analogue of t-butyl peroxide. Under these conditions, 3% yield of the product derived from methane incorporation was observed along with 50% yield of the side-product bearing a deuterated methyl unit. Mechanistically, in accordance with a previous report, 52a the Cu(I) complex XXXVII promotes the reductive fragmentation of the peroxide, thus leading to Cu(II) species XXXVIII and an oxygen centered radical. While the former reacts with one equivalent of amide 21, the latter can either abstract a hydrogen atom from the gaseous substrate to generate XXXIX or undergo fragmentation affording methyl radical XL. Radical combination with Cu(II) intermediate XXXVIII, followed by reductive elimination, affords the N-alkylated product 22 and regenerates the Cu(I) catalyst.
In the same year, Chang and co-workers reported a highly selective C-N bond formation methodology capable of converting a wide range of inactivated alkanes into the corresponding N-alkyl carbamic acid ester derivatives 25 (Fig. 15). 53 Among the tested compounds, gaseous substrates, such as n-butane, propane and ethane, afforded the desired products 25 in moderate to good yields (34-73% yield). The protocol takes advantage of the formation of an electrophilic cobalt-nitrenoid species XLI that is responsible for the putative hydrogen atom abstraction. The design of a novel Cp*Co(III)(L 7 ) catalyst, featuring enhanced electrophilicity and adequate steric hindrance, proved to be of key-importance to obtain the desired reactivity and siteselectivity. Concerning the first aspect, the authors envisioned that installing an electron-withdrawing group within the bidentate ligand scaffold, might boost the overall electrophilicity of the cobalt-nitrenoid intermediate XLI, enabling the intermolecular activation of unbiased alkanes. This constitutes a transformation which is still elusive via similar catalytic manifolds. 54 Another advantage of the disclosed method is the high siteselectivity: indeed, for n-butane and propane, high preference for the functionalization of the methylene unit over the methyl one was observed. Interestingly, neither undesired overfunctionalization nor solvent activation occurred. Based on a set of mechanistic studies, including electron paramagnetic resonance (EPR), DFT computations, radical clock experiments and kinetic isotope effect (KIE) studies, the authors proposed a cobalt(III)-nitrene radical activation mode in agreement with previous studies. 55 Coordination of Cp*Co(III)(L 7 ) to substrate 24 followed by a rate-determining nitrogen extrusion affords key metal-nitrenoid intermediate XLI. This species is then capable of abstracting a hydrogen atom from the aliphatic substrate, which in turn rapidly undergoes radical rebound affording the desired N-alkylated carbamic acid ester 25.

C-Si bond formation
The synthesis of organosilicon compounds is an increasingly important field because of its ramifications in both medicinal 56 and material chemistry. 57 Moreover, the ensuing compound is a useful synthon due to its feature of acting as a carbon-based nucleophile in cross-coupling reactions. 58 Taking this into account, several methods to perform the direct silylation of C(sp 3 )-H bonds have been reported. 59 Within this field, efforts for the installment of the silicon functionality on gaseous alkanes have been made. Specifically, the group of Tilley, 60 based on previous findings on the activation of methane by Cp* 2 ScR 2 complexes, 61 used this type of catalysts to install the silicon-based group on both methane and cyclopropane in a Parr high-pressure vessel (Fig. 16). After a thorough mechanistic study under both stoichiometric and catalytic conditions, the reaction has been proposed to initiate with an intracomplex C-H activation of one of the methyl substituents on the cyclopentadienyl ring and subsequent release of molecular hydrogen to afford complex XLIV. This intermediate is responsible for the activation of the alkane substrate, thus generating XLV which reacts with the silylating reagent 26, to yield product 27a and regenerate Cp* 2 ScH. Notwithstanding that the overall reactions are slow, this is an intriguing and significant achievement as it is one of the few processes catalyzed by an early metal that proceed under a non-radical mechanism. 62

C-S Bond formation
Owing to their involvement in many processes such as protein folding and electron transfer events, sulfur containing molecules occupy a privileged role in biochemistry. 63 Additionally, this class of compounds is routinely selected as novel candidates in the field of agricultural and pharmaceutical chemistry. 64 In the context of feedstock valorization, particular emphasis has been given on identifying chemoselective synthetic routes capable of converting methane into methane sulfonic acid (MSA). Indeed, MSA has recently emerged as an ideal electrolyte for many electrochemical industrial processes such as the electrodeposition of metals on electronic devices. 65 Early examples of methane conversion into MSA took advantage of either K 2 SO 4 or metal peroxides as radical initiators and SO 3 as the sulfonating agent, with sulfuric acid as solvent medium. 66 In the same period, inspired by an elegant vanadium-catalyzed sulfoxidation of inactivated alkanes, 67 Bell and coworkers published the first synthesis of MSA using methane as the alkylating agent and SO 2 as the sulfur source. 68 The reaction was carried out in triflic acid, with potassium persulfate and calcium salts used as, respectively, the oxidant and the promoter. Shortly after, the same group reported the first homogeneous catalytic sulfonation of methane using SO 2 . 69 The highest conversion of sulfur dioxide into MSA was achieved using a dual catalytic system of Pd(II) and Cu(II) salts, with O 2 acting as an effective oxidant in a Parr autoclave reactor under 83 bar of methane (Fig. 17, 20% yield). As for the solvent, only strong acids proved suitable for the desired transformation. Despite the mechanism was not fully elucidated, similarly to Fujiwara and co-workers, 16 the authors postulate that the Pd species might promote an electrophilic activation of methane, followed by sulfur dioxide insertion and oxidation. Cu(II) and oxygen are proposed to be involved in the reoxidation of Pd(0) to Pd(II). Indeed, omission of Cu(II) resulted in formation of Pd-black with consequent inhibition of the reactivity.
Recently, the chemical division of Grillo-Werke AG reported a highly selective thermal process (99% selectivity and yield) for the conversion of methane and sulfur trioxide into MSA (Fig. 18). 70 The protocol employs a highly electrophilic sulfonyl peroxide initiator, formed in situ from MSA, H 2 O 2 and Oleum (SO 3 36% weight in H 2 SO 4 ). The initial optimization, carried out in a high pressure Parr autoclave, revealed that the highest conversion was obtained at 50 1C using 100 bar of methane. Concerning the reaction mechanism, the authors carried out a series of experiments aimed at demonstrating the non-radical nature of the presented transformation. Given the known tendency of the sulfonylperoxides to undergo homolytic cleavage of the O-O bond, 71 the authors conducted the same transformation at ambient temperature under exposure to UV-light (190 nm). After eight hours of irradiation no change in the methane pressure was observed, thus excluding the involvement of a radical pathway. Furthermore, inclusion of molecular oxygen in the reaction mixture did not inhibit the desired reactivity. 72 Despite these evidences, the reaction mechanism of the catalytic transformation remains unclear and computational studies demonstrated the unfeasibility of a polar cationic pathway. 73 In light of the industrial relevance, the authors realized a pilot plant to scale-up the presented transformation with a projected capacity of 20 metric tons year À1 . As shown in Fig. 18, the facility features three stainless-steel reactors connected in series. In the first tank, methane and SO 3 are introduced and immediately reacted with the pre-formed electrophilic initiator. When the reaction mixture reaches the last reactor, the excess of SO 3 is quenched while methane is recycled. The crude    mixture obtained after quenching is transferred to a distillation apparatus (D, Fig. 18) which delivers pure MSA. The exhausted solution containing H 2 SO 4 and MSA is further recirculated to the first tank in order to form back the initial solution.

Conclusions
Over the past years, several groups have contributed to the homogeneous functionalization of gaseous alkanes. In this review, we have outlined the different catalytic tools used to perform this transformation, including transition metalcatalyzed C-H activation, carbenoid C-H functionalization and photocatalytic HAT, encompassing both polar and radical reactivity modes. Up to now, these methods have enabled the forging of C-B, C-C, C-N, C-O, C-Si and C-S bonds. Specifically for the C-C bond formation, enantioselective variants have been developed. Moreover, we have also detailed the importance of suitable setups, such as high-pressure Parr vessels or continuous-flow reactors. Future efforts should be directed to construct new chemical bonds, such as C-P or C-Se, with these gaseous coupling partners. In addition, the functional group tolerance of these transformations needs to be further increased to enable the use of volatile alkanes as alkylating reagents within the synthetic routes towards more complex molecules. Other catalytic tools and activation modes should be identified and exploited to promote the functionalization of these recalcitrant substrates. At the same time, as technology clearly plays a crucial role to support these transformations, future endeavours should be also directed towards the development of suitable reactors, which can handle these chemicals efficiently and safely, and can ensure the scale up of the developed transformations. Finally, to bridge the gap with asymmetric catalysis, more efforts are required to ensure high enantioselectivities across various bond-forming reactions.

Author contributions
The concept of this article was conceived and agreed upon after discussions with all authors. D. M. and A. P. carried out the literature study and wrote the article (equal contributions). The entire manuscript was corrected and edited by T. N.

Conflicts of interest
There are no conflicts to declare.