Keywords

1 Introduction

1.1 Isotopes and Isotopic Labeling

Isotopes of a particular element have an identical number of protons in their respective nuclei but possess an unequal number of neutrons. Namely, they share the same atomic number but have different mass numbers, as exemplified for hydrogen (Scheme 1) [1]. The stability of an isotope is governed by the ratio of neutrons to protons within the nucleus, thus giving rise to two possible circumstances. Firstly, a heavy isotope of an element, such as 2H or 13C, has a stable nucleus and tends to be found in nature, albeit at lower abundances than their more common counterparts, 1H and 12C, respectively. In the alternative case, radioisotopes, such as 3H or 14C, have an unstable neutron/proton ratio and decay, via emission of radiation or particles, to form other elements, or different isotopes of the parent element.

Scheme 1
scheme 1

Simplified Bohr representations of the isotopes of hydrogen

The synthesis and supply of isotopically labeled molecules has a sustained importance in the study of metabolic processes, among myriad other processes [2]. It is therefore unsurprising that there is a large and growing body of research dedicated to the synthesis of isotopically labeled compounds. The labeling of molecules with 13C or 14C is most readily achieved through the use of commercially available, isotopically enriched starting materials. While such a technique ensures a regiospecific label will be present in the desired target molecule, it ultimately comes at the price of unwanted additional steps in the synthesis [3].

Research into deuterium (2H or D) and tritium (3H or T) labeling is more substantial than that for other isotopes and has been developed on a number of fronts over the past 60 years [2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22]. Further to this, key developments in synthetic strategies and analytical techniques over the past three decades are gradually making tritium labeling the preferred technique in many absorption, distribution, metabolism, excretion, and toxicology (ADMET) studies [10]. In one particularly active branch of such research, hydrogen isotope exchange (HIE) is commonly employed to deliver deuterium or radioactive tritium to pharmaceutical drug candidates in one synthetic step.

1.2 Applications of Hydrogen Isotope Exchange

The importance of hydrogen isotope exchange (HIE), for iridium catalysts and beyond, is manifest in the wealth of reviews published in the area [2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36]. As well as circumventing the requirement for isotopically enriched starting materials in synthesizing tritiated drug candidates [3, 10], HIE can also provide analogous deuterated compounds for use as internal standards for mass spectrometry [29, 37], for kinetic isotope studies, [21, 38, 39], and for the alteration of reaction pathways in total syntheses [40]. Additionally, HIE is applied within almost every sub-discipline in life science, in nuclear science, and beyond [2]. The ability for precise measurement of isotope ratios promotes a dynamic view on biosynthetic pathways, protein turnover, and systems-wide metabolic networks and, thus, has paved the way for a number of scientific breakthroughs in biomedical research. In assessing a drug candidate’s metabolic fate, the chemist must first have a flexible technique with which to study it. Consequently, isotopic labeling is the gold standard method by which early stage drug discovery processes are optimized. The numerous application areas for HIE are summarized in Scheme 2.

Scheme 2
scheme 2

Application areas served by hydrogen isotope exchange (HIE) technology

1.3 Synthetic Methods in HIE

With a broad range of existing HIE applications, there exists a wide range of synthetic methods to achieve hydrogen isotope incorporation in an ever-expanding array of substrates. While the full gamut of chemistries developed for HIE is beyond the primary focus of this chapter, it is worth covering these in brief. Firstly, the source of deuterium and tritium has varied extensively from method to method; however, some patterns exist. For deuteration, many methods have applied D2 gas, D2O (heavy water), DCl, benzene-d6, DMSO-d6, and numerous deuterated alcohols. Of these isotope sources, and of direct relevance to the focus of this chapter, D2 gas has been the preferred isotope source as it directly maps onto the preferred use of tritium gas for radiolabeling protocols [5, 7, 10, 17].

Hand in hand with the range of hydrogen isotope sources is a wide range of metal-mediated and other mediated processes for HIE (Scheme 3). Classically, these include acid- and base-mediated reactions, as well as modern variations using frustrated Lewis pairs (FLPs). Aryl labeling is most common, but many common organic transformations have been pivoted into labeling protocols. Nonetheless, metal-catalyzed HIE is dominant in HIE, covering heterogeneous and homogeneous catalysis. Such methods have been more fully reviewed elsewhere [11, 12].

Scheme 3
scheme 3

Common synthetic transformations toward the installation of hydrogen isotopes in organic substrates

2 Ortho-Directed Iridium-Catalyzed HIE

Among all transition metals employed in homogeneous HIE methods, iridium is arguably the most widely studied [2, 3, 5, 6, 11,12,13, 15, 17, 18, 20, 22, 26, 35, 36, 41, 42], which is, in part, due to the vast and ever-expanding literature precedent in related hydrogenation reactions [31, 43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68]. Iridium was also present in some of the earliest metal-centered catalysts applied to HIE chemistries [69]. In support of this analysis of iridium’s popularity in HIE, Oro and co-workers estimated that iridium accounted for 33% of all reported HIE methods, greater than for any other metal [11]. While iridium catalysts have also been applied in heterogeneous catalysis [70, 71], the focus of this chapter is on the far more expansive homogeneous iridium catalysis developments.

2.1 Early Developments in Ortho-Directed HIE

There is a clear dominance of ortho-HIE in the homogeneous iridium catalyst literature. In 1992, Heys demonstrated the successful ortho-directed deuteration of several substituted aromatic compounds using the 18-electron Ir(III) bis-phosphine dihydride complex 1 under very mild conditions (2 → 3, Scheme 4) [72]. Crucially for the time, Heys’ investigations marked a significant advancement from Lockley’s ortho-labeling work (with rhodium and ruthenium catalysts) [73,74,75]: D2 gas replaced D2O as the deuterium source (an advantage when considering the use of tritium), reactions operated efficiently at room temperature, and, perhaps most importantly, catalyst loadings were significantly reduced from 50 mol% to 2 mol%. Interestingly, it was noted that labeling was significantly affected by steric or electronic aspects of the substituents present on arene substrates. For example, meta-substituted ethyl benzoates, such as 5, showed a consistent preference for labeling at C-2 over the less hindered C-6 position, presumably due to additional coordination assistance from meta-substituent lone pairs [76]. Steric hindrance from ortho-substitution reduced labeling efficiency (4 vs. 9); however, bulky α-substituted ketones such as 6 were not so adversely affected. Further to this, where substrates possessed more than one carbonyl directing group, the labeling site(s) changed according to which substituent could coordinate to the catalyst to the greatest extent (e.g., 4 vs. 10).

Scheme 4
scheme 4

Heys’ Ir-catalyzed ortho-directed deuteration of aromatic compounds

The mild labeling conditions pioneered by Heys and co-workers, coupled with intriguing substrate-dependent regioselectivity, captured the combined interest of the industrial and academic HIE communities, resulting in a large number of subsequent studies aimed at understanding the catalytic properties of 1 and related Ir-based HIE catalysts. Firstly, Heys followed up his initial study with a more in-depth assessment of the aryl substituent effects in the labeling efficiency of ethyl benzoates and N,N-dimethylbenzamide substrates (Scheme 5) [72]. In a rather unexpected outcome, para-substitution improved the rate of labeling in both substrate types, irrespective of substituent electronics (e.g., 11a vs. 11b and 11c). In an attempt to explain this effect, Heys monitored the rate of labeling in both rings of several monosubstituted benzophenones [72]. The substituted ring was labeled faster in every instance (12a12c). As both rings are connected to the same carbonyl functionality, it appeared that the rate-limiting step of the overall reaction could not be ascribed to the initial coordination of the substrate, nor hydride fluxionality [23]. Instead, Heys suspected that some aspect of the C–H bond cleavage was rate-limiting, proposing key intermediates 13 and 14 based on available literature. At this time, the formal oxidation state of iridium intermediates involved in the C–H bond cleavage (IrI /IrIII or IrIII/IrV) was not clear.

Scheme 5
scheme 5

Mechanistic investigations into Heys’ Ir-catalyzed HIE protocol

Inspired by Heys, Hesk and co-workers probed the efficacy of the commercially available Crabtree hydrogenation catalyst, 15 [43], in labeling acetanilide derivatives, the first such substrates to be effectively labeled via a 6- rather than a 5-membered metallocyclic intermediate (mmi) [77]. Consistent with Heys’ work, Hesk reported that deuteration was directed ortho to the coordinating functionality. Moreover, no clear relationship emerged regarding the electronics of para-substituents and labeling efficacy. Ketones 17 and 18 were also compatible with this mild labeling method; however, weakly coordinating benzenesulfonamide 19 and benzoic acid 20 were not (Scheme 6).

Scheme 6
scheme 6

Hesk’s application of Crabtree’s catalyst in ortho-HIE

Since Heys’ and Hesk’s respective discoveries of iridium catalysts for ortho-directed HIE, complexes 1 and 15 (and derivatives thereof, vide infra) have remained topics of high interest in HIE research [26, 78,79,80,81]. In a further key development by Heys, 21, a precatalytic Ir(I) variant of Ir(III) catalyst 1 was compared to related bidentate pre-catalyst, 22 (Scheme 7) [82]. By the mid-1990s, it had already been hypothesized by several researchers that both 5- and 6-mmis could be formed during the C–H bond cleavage step in the ortho-deuteration process (23 vs. 24), depending on the substrate being studied; this was to be the platform on which to compare iridium catalysts 21 and 22.

Scheme 7
scheme 7

Mono- vs. bidentate Ir–phosphine catalysts to study ortho-deuteration via 5- and 6-mmis

Labeling a range of substrates enabled a comparison of the mono- and bidentate phosphine complexes to be made, highlighting a preference for monodentate 21 to react through a 5-mmi only, whereas bidentate 22 could react through both a 5- and a 6-mmi. This result was exemplified in the labeling of ethyl 1-naphthoate, 25 (Scheme 8, top). Of the two available labeling sites, the monodentate complex, 21, labeled solely at C-2. Conversely, bidentate complex 22 demonstrated the capability to direct labeling at both C-2 and C-8. When Crabtree’s catalyst, 15, was exposed to similar reaction conditions, the regioselectivity in labeling was similar to monodentate complex 21, albeit with reduced labeling efficiency. Labeling through a 6-mmi only was also investigated. Perhaps the most remarkable findings from this study were those concerning the labeling of N-phenyl phenylacetamide, 26 (Scheme 8, bottom). Interestingly, the less active monodentate complex, 21, showed selectivity for the aromatic ring adjacent to the nitrogen, 26a, an effect emulated more efficiently by Crabtree’s catalyst in 26c. However, the bidentate catalyst 22 was able to label both rings of 26 almost indiscriminately (see 26b). This served to show that there was potential to distinguish not only between a 5- and 6-mmi, but also between different types of 6-mmi, depending on the ancillary ligands employed.

Scheme 8
scheme 8

Heys’ vs. Hesk’s ortho-HIE methods for 5- and 6-mmi substrates [77, 82]

On accumulation of these data, Heys proposed a catalytic cycle by which these iridium complexes may be affecting the observed regioselective hydrogen isotope exchange (Scheme 9) [82]. Upon treatment of the Ir(I) pre-catalyst, 27, with deuterium gas, hydrogenolysis of cyclooctadiene (COD) as d4-cyclooctane generates the active Ir(III) catalyst, 28, where ligands (L) are assumed to be arranged trans to one another when monodentate. Coordination of substrate displaces a solvent molecule (S) and is thus accepted into the coordination sphere of the iridium catalyst to give 29. A second solvent molecule can then be displaced, allowing iridium to cleave the nearby ortho C–H bond of the aryl ring to yield 30. Transformation of species 31 to 32 is driven by a process known as hydride fluxionality and is central to the isotope exchange process [23]. The overall effect brings a deuteride and the activated aryl carbon into a cis arrangement. Subsequent C–D elimination furnishes 32, with a deuterium atom now installed ortho to the directing group. Finally, the release of deuterated substrate, 33, regenerates the resting catalytic intermediate, 28. This mechanism invokes an all-Ir(III) catalytic cycle with C–H activation as the rate-limiting step, supporting evidence for which would take another decade to accumulate. Said evidence involved isolation and crystallographic characterization of 34 (an acetonitrile-solvated analogue of 30) and spectroscopic studies on the evolving nature of iridium hydride equilibria as a function of ancillary ligand electronics (Scheme 9, inset) [20].

Scheme 9
scheme 9

Heys’ mechanistic analysis for homogeneous Ir-catalyzed ortho-HIE

In an extension of the theory of ortho-directed HIE, Heys postulated that the preference for the monodentate phosphine complex, 21, to react only via a 5-mmi, 38, as opposed to a 6-mmi, 37, was based on steric effects (Scheme 10) [82]. By contrast, the bidentate complex, 22, is forced to arrange the phosphines cis to one another. For substrates such as 25, this opens up a second face on the iridium complex, offering greater spatial freedom for the formation of the less planar 6-mmi, 40, as well as the 5-mmi, 39. By the same thought, the monodentate Crabtree’s catalyst, 15, can facilitate labeling through a 6-mmi as the pyridine and tricyclohexylphosphine ligands present less steric bulk than the two triphenylphosphine ligands of complex 21 and may thus exist in cis or trans form. Herbert later capitalized on this rationale to further improve bidentate catalyst 22 in the labeling of 6-mmi substrates, changing the diphenylphosphinoethane (dppe) ligand for the sterically less encumbered arsine analogue [81].

Scheme 10
scheme 10

Rationale for 5- vs. 6-mmi labeling selectivity with mono-/bidentate phosphine catalysts

2.2 Contemporary Methods in Ortho-Directed HIE

Further synthetic developments by Herbert [28, 78, 83] and later Salter [26] showed that bis-phosphine catalysts like 22 may be generated in situ from the appropriate free phosphine and commercial iridium dimer, [Ir(COD)Cl]2, with comparable activity to the isolated complexes. The same authors are also separately responsible for detailed studies into alteration of the phosphine structure [26, 78, 81]. However, both parties have remarked that strong correlations between ligand properties (such as sterics or electronics) and catalyst activity are difficult to detect. The number of such ligands applied to iridium-catalyzed HIE is now extensive and includes more elaborate catalyst system like 41 (Scheme 11).

Scheme 11
scheme 11

An overview of bis-phosphine-ligated iridium catalysts applied in HIE

Parallel with studies into bis-phosphine systems, Crabtree’s catalyst 15 has also been the subject of intense study in deuteration and tritiation, since Hesk’s discovery [76, 84,85,86,87,88,89]. In one of the largest of any such study, Herbert explored an expansive substrate scope, including ketones, amides, anilides, and various heterocycles [83]. Despite the impressive array of examples reported, this study employed at least stoichiometric quantities of 15 and a dual D2/D2O isotope source, making comparisons to related ortho-labeling methods difficult.

In a notable crossover between bis-phosphine catalysts and Crabtree’s catalyst, Hickey and co-workers developed a polymer-supported variant of Heys’ bis-phosphine catalyst, 42, which showed comparable ortho-HIE activity to 15 and 22, but with the practical benefit of simple catalyst filtration at the end of the reaction (43 vs. 44 vs. 45; Scheme 12) [71]. Solid-supported iridium catalysts for HIE have now been adapted to flow systems [80].

Scheme 12
scheme 12

Polymer-supported iridium catalyst in ortho-HIE

Exploring an altogether different ligand architecture, Lockley reported the application of hexafluoroacetylacetonate (hfacac)-ligated Ir(I) complex, 46, in ortho-HIE (Scheme 13) [8, 10, 22, 90,91,92]. This catalyst has been successfully applied in the labeling of benzylic amines, benzoic acids, and primary sulfonamides, where few other Ir-based HIE catalysts have succeeded. The catalyst is one of the few iridium HIE catalysts operational in highly polar solvents such as DMF (desirable for poorly soluble drug molecules) and displays different labeling regioselectivities depending on the choice of isotope source (D2 or D2O; see 47 → 48 vs. 47 → 49).

Scheme 13
scheme 13

Ir(I)-hfacac ortho-HIE catalyst and isotope source-dependent regioselectivity switch

In the early 2000s, increasing interest in Crabtree’s catalyst, 15, in HIE was paralleled with investigations by other researchers to improve efficiency and chemoselectivity in iridium-catalyzed olefin hydrogenation reactions [61]. Despite its widely reported success, 15 is known to suffer from thermal deactivation via the formation of inactive, hydride-bridged, iridium clusters (50, Scheme 14) [54]. Similar effects have been documented for other iridium-based complexes [66, 93].

Scheme 14
scheme 14

Trimeric iridium cluster formed from thermal deactivation of Crabtree’s catalyst

Separate investigations by Nolan [62] and Buriak [94] toward improved thermal stability and predictable chemoselectivity of Crabtree-like hydrogenation catalysts resulted in a plethora of highly promising electron-rich, N-heterocyclic carbene (NHC)-ligated iridium catalysts (Scheme 15). Such species were first applied and published in ortho-HIE processes by Powell and co-workers [95]. In Powell’s study, complexes 51a and 52a52c were employed under stoichiometric (industrial “tritiation-like”) conditions, with the most active variant, 52c, shown to be superior to Crabtree’s catalyst across the entire substrate range.

Scheme 15
scheme 15

NHC-ligated iridium catalysts for hydrogenation later explored by HIE chemists

In a more interesting variant of this work, Kerr and co-workers studied the catalytic activity of complexes 51b–51f, showing most active complex, 51e, to be highly active over an appreciable substrate scope (5 mol% [Ir], 16 h, rt) and displaying a higher turnover frequency (TOF) than Heys’ bis-phosphine catalyst, 22. Interestingly, the smaller complexes in the series studied by Kerr (51b and 51c) were completely inactive as HIE catalysts [86]. Similar investigations by the same group led to the discovery that small NHC/phosphine complexes such as 52c were inactive as HIE catalysts, but larger variants 52d and 52e were active across a limited substrate scope [96].

The exploration of NHC-ligated iridium HIE catalysts had revealed promising (proof-of-concept) developments beyond the popular and established works of Hesk and Heys. Kerr and co-workers later developed a synthesis of previously unattainable complexes 53a53c, bearing large phosphine and large NHC ligands (Scheme 16) [97]. These complexes have proven seminal within the ortho-HIE domain and have among the highest activity [98], substrate/solvent scope [99, 100], and tritiation reaction cleanliness of any such catalyst reported to date. Additionally, ortho-HIE process with these complexes has been studied experimentally and computationally, strengthening the case for a Ir(III)-based reaction mechanism akin to that proposed by Heys [98]. More specifically, kinetic isotope effect (KIE) measurements [101] revealed that C–H bond cleavage was the rate-limiting step of the reaction (54 → 55), and detailed NMR studies revealed (via 2JP–C coupling information) the trans-geometry of the ancillary ligands [98]. The same study was also able to reveal the origins of the selective reactivity of such catalysts for 5- over 6-mmi substrates, citing dual kinetic and thermodynamic favorability for the 5-mmi. The calculated transition states 56 and 57 revealed, for the first time, the sigma-bond-assisted metathesis (sigma-CAM) process at the heart of the all-Ir(III) C–H activation step [98, 102].

Scheme 16
scheme 16

Highly active NHC/phosphine ortho-HIE iridium catalysts

While developing a rare method for labeling primary sulfonamides, Kerr and co-workers considered directing group chemoselectivity in detail [102]. It was observed that the sulfonamide vs. pyrazole selectivity in celecoxib 58 varied dramatically with catalyst choice (Scheme 17). Whereas encumbered and most-often used NHC/phosphine catalysts facilitated labeling adjacent to the pyrazole moiety, giving 58b, neutral NHC/Cl catalysts, such as 59, facilitated selective sulfonamide labeling, delivering 58a, for the first time. Accompanying DFT studies revealed that the substrate binding event was likely to be product-determining (60a vs. 60b), even though C–H activation remained rate-limiting (Scheme 17). A similar rationale was presented for multifunctional molecules containing esters as the targeted directing group [103]. Following this, Derdau and co-workers significantly expanded on the HIE studies of competing directing groups, showing once again that calculated binding energies could serve as a semiquantitative and predictive tool for rationalizing directing group chemoselectivity in HIE [42].

Scheme 17
scheme 17

DFT-calculated rationale for directing group selectivity using catalyst 59 and sulfonamide drug 58

Building on Kerr’s work, Ir(III)-catalyzed ortho-HIE has continued to flourish [3, 11, 13, 15, 35, 41, 42]. From the same group, and others applying the developments therefrom, the application of bulky NHC–phosphine systems in HIE has steadily advanced in terms of the applicable substrate and solvent scope [42, 102, 104,105,106,107,108]. With regard to solvent scope, Kerr and Tamm have reported complementary strategies toward modifying the solubility profile of existing iridium HIE catalysts. On the one hand, Kerr explored the use of the bulky tetrakis[3,5-bis(trifluoromethyl)phenyl]borate (BArF) counterion in place of the standard hexafluorophosphate (PF6) [104, 109], and on the other, Tamm integrated a related borate anion into the backbone of an anionic carbene ligand (Scheme 18) [110]. The wide range of solvents made applicable in extending the Kerr catalyst series through 61a61d evidenced new opportunities to tune HIE regioselectivity through simple solvent switching [104]. From Tamm’s most recent developments, catalysts 62a, 62b, and 62e have been identified as competent HIE catalysts in hexane and cyclohexane for the first time [110].

Scheme 18
scheme 18

Counter-anion effects explored in iridium-catalyzed ortho-directed HIE

A growing community of researchers have, in more recent times, contributed a wider range of elaborated ligand spheres around tractable iridium(I) pre-catalysts. In turn, more iridium HIE catalysts have enabled applications using more challenging directing groups. A recent contribution from Pfaltz and Muri showed the application of P,N-derived bidentate ligands [111]. Most notably, these latest iridium-based HIE catalysts have been developed to be able to label ortho to secondary benzenesulfonamides for the first time, albeit using high temperatures and synthetically intricate ligands [111]. Along similar lines, Tamm and Derdau have reported complementary P,N- and C,N-ligated iridium catalysts able to further expand the range of accessible directing groups applicable in ortho-directed HIE processes (6365, Scheme 19) [110, 112, 113].

Scheme 19
scheme 19

Modern chelated iridium catalysts expanding the range of accessible ortho-directing groups in HIE

3 Beyond Ortho-Directed HIE

Far from the humble beginnings of homogeneous iridium-catalyzed HIE [69], labeling of organic molecules has continued to advance along complementary lines to ortho-directed HIE. While some instances have been discovered as unintended by-products of desired ortho-labeling, [111] or to assess non-innocent ancillary ligand behaviors, [114,115,116,117,118,119,120,121,122,123,124,125] contributions have been made to labeling global aromatic, sp3, vinyl, formyl, and heteroatom positions in a strategic manner (cf. Scheme 3). In the application domain, such developments have given industrialists a more diverse palette of methods with which to incorporate hydrogen isotopes into an increasingly elaborate array of drug candidates.

3.1 Directed sp3 C–H HIE Methods

Somewhat inspired by the deep understanding of iridium catalysts and compatible directing groups for ortho-directed HIE protocols, significant contributions have emerged toward labeling sp3 centers rather than aromatic sp2 centers [15, 111, 126, 127].

Using Kerr’s commercially available catalyst 53a and 61a, Derdau and Kerr have developed expansions of the original ortho-labeling methodologies, showing that the same catalyst systems can effectively label sp3 C–H positions in complex amides and a range of drug molecules (Scheme 20).

Scheme 20
scheme 20

Directed sp3 HIE using the commercially-available iridium catalysts

In a new paradigm for the field, MacMillan and co-workers developed a photoredox- and hydrogen atom transfer (HAT)-catalyzed method, employing an iridium(III) photocatalyst Ir(F-Meppy)2 (dtbbpy)PF6 [F-Meppy, 2-(4-fluorophenyl)-5-(methyl)pyridine; dtbbpy, 4,4′-di-tert-butyl-2,2′-bipyridine], 66 [128]. In combination with labeled water (D2O or T2O) as the isotope source, and a suitable hydrogen atom donor, this method selectively delivered isotope incorporation to the sp3 a-amino sites in 18 drug molecules (Scheme 21).

Scheme 21
scheme 21

Selected examples from MacMillan’s photoredox and hydrogen atom transfer (HAT)-mediated alpha-selective sp3 HIE process for drug-like amines

The reaction is proposed to operate via coupled photoredox and hydrogen atom transfer (HAT) cycles (Scheme 22). The photoredox catalyst 66 is excited by the blue light-emitting diode (LED) to generate a long-lived excited state triplet 67, a strong single electron oxidant. The catalyst then generates an alpha-amino radical 70 from 69, and the reduced Ir(II) catalyst 68, which is now a strong reductant. Isotopic scrambling between the labeled water source and added thiol delivers the on-cycle labeled thiol 72 from 71, judiciously chosen due to the favorably weak S–H bond. Labeled thiol 72 (polarity matched with the nucleophilic amino radical 69) undergoes a HAT process to generate the alpha-labeled amine product 74 and thiol radical 73. Thereafter, the photoredox and HAT catalytic cycles converge to generate the thiolate anion 75 and regenerate the photoredox catalyst 66. Through adjustments in the choice of photocatalyst and thiol source, this method was applicable to both deuteration and tritiation processes.

Scheme 22
scheme 22

Hypothesized mechanism for photoredox- and HAT-mediated HIE

3.2 Non-ortho-HIE on Aromatic Substrates

A range of cyclopentadienyl (Cp, and derivatives thereof)-ligated iridium complexes have been shown to be active in HIE (7684, Scheme 23). Principally, several nondirected and global aromatic C–H deuteration strategies have been reported and improved over several iterations of catalyst design [114, 129,130,131,132,133,134,135,136]. In 2001, Bergman and co-workers showed that complexes of the type [(Cp*)Ir(PR3)(H)(DCM)], such as 76, and, later, [(Cp*)Ir(PMe3)(H)3]OTf, were active in HIE across a range of aromatic and aliphatic substrates [114, 129,130,131]. In further iterations, Peris [132] and Ison [134, 135] reported a range of NHC-ligated complexes based on the Cp-I core. In more practically facing contributions, Thieuleux and collaborators divulged solid-supported variants of [(Cp*)Ir(NHC)] cores, 82–84 [133, 136]. Across this series of publications, mechanisms of HIE were hypothesized to vary with deuterium source, solvent, and ancillary ligand combination (see Scheme 24 or exemplar transformations).

Scheme 23
scheme 23

Overview of Cp*Ir complexes applied to HIE processes

Scheme 24
scheme 24

Exemplar HIE processes enabled by Cp*Ir complexes

3.3 Vinyl HIE Processes

Expanding sp2 labeling protocols beyond simple aromatic systems, a number of recent reports have shown the possibility of selectively labeling vinyl groups. Because many modern iridium HIE catalysts of the type [(COD)Ir(L1)(L2)]X evolved from the hydrogenation literature [54, 62, 67], the labeling community has been aware of (and exploited) the reductive power of these catalyst systems to install isotopes across unsaturated moieties [10]. However, the dual HIE and hydrogenation reactivity of these iridium systems presents a challenge if the same catalyst is targeted for an HIE application, and not a hydrogenation. While designing HIE methods for labeling α,β-unsaturated substrates, Kerr and co-workers hypothesized that the competing reactivity could be rationalized by a equilibrating C–C bond rotation 85 to 86 upon substrate coordination (Scheme 25). For larger ligand spheres such as in catalyst 53a, intermediate 86 would be favored, driving HIE (86 → 87 → 88 → 89). For smaller ligand systems, as has been observed in attempts to use Crabtree’s catalyst for similar transformations [137], intermediate 85 is favored, driving hydrogenation over HIE (85 → 90 → 91 → 92).

Scheme 25
scheme 25

Hypothesized competing HIE and hydrogenation pathways [99]

Beyond re-optimizing HIE the use of catalysts in which competing hydrogenation is an issue, several methods for the chemoselective labeling of alkenes have also appeared in the iridium literature. In 2008, Hartwig reported a method where pincer complex 93 was shown to label vinyl C–H positions with selectivity largely dependent on the specific steric environment of the substrate, albeit under air and moisture sensitive conditions (Scheme 26, left) [138]. Notably, this method was applied to a series of both simple and complex organic molecules and included global labeling of aromatic and heteroaromatic substrates. A more practical variant of this method was divulged by Nishimura and co-workers [139]. Using an in situ-derived Ir(III) monohydride, 94, and D2O as the isotope source, an attractive range of mono-substituted alkenes could selectively deuterated at the vinyl or methylidene positions (Scheme 26, right).

Scheme 26
scheme 26

Iridium-catalyzed vinyl HIE

In relation to vinyl HIE, formyl-selective methods of labeling benzaldehyde derivatives has been of notable interest, due, in part, to the synthetic handle of derivatization presented through the carbonyl functional group [140,141,142,143,144,145]. In 2010, Chapelle and co-workers showed that Crabtree’s catalyst was able to deliver formyl-labeled benzaldehyde derivatives, albeit with variable selectivity against competing aryl ring labeling [142]. Kerr and co-workers used this work as inspiration to compare Crabtree’s catalyst in formyl labeling vs. other competent ortho-HIE catalysts. Comparing catalysts 15 vs. 53b vs. 95, it was shown that the NHC/Cl system delivered superior formyl selectivity than either of the cationic iridium centers bearing larger ligand spheres (Scheme 27). The group accounted for these observations using a detailed mechanistic model centered around cis-trans isomerization of the activated Ir(III) catalyst. While intermediate 97 bearing trans ancillary ligands favors the approach trajectory of the aldehyde substrate that leads to aryl HIE, isomer 96 of the same catalyst enables the aldehyde to approach along a trajectory leading to formyl HIE [146].

Scheme 27
scheme 27

Toward formyl-selective iridium-catalyzed HIE processes

3.4 Beyond C–H Labeling

Some of the most recent developments in isotopic labeling employing iridium catalysis have been applied to X–H moieties. While comparatively rare when compared to C–H HIE methods, heteroatom labeling can be insightful en route to establishing new carbon–heteroatom bonding–forming processes. Specifically, Nolan and Grubbs have independently reported on silane labeling [147, 148]. Grubbs studied catalyst 98, while Nolan investigated 99 and 100 in Si–H and B–H labeling, respectively (Scheme 28) [149].

Scheme 28
scheme 28

Iridium-catalyzed HIE for Si–H and B–H bonds

4 Concluding Remarks

Notwithstanding earlier pioneering developments in the field [69, 73,74,75, 150,151,152], iridium-catalyzed HIE has undergone explosive growth since Heys’ use of bis-phosphine systems in the early 1990s [153]. The main thrust of developments in the field have been in ortho-directed HIE domain. Such is the maturity and underlying mechanistic understanding of the ortho-labeling subfield, that it is now influencing catalyst design strategies in the broader C–H functionalization field. Considered alteration of the iridium ligand sphere – for both Ir(I) and Ir(III) systems – has now expanded the field of HIE well beyond its ortho-labeling comfort zone. Iridium-catalyzed methods to install heavy and radioactive hydrogen isotopes now span global aromatic labeling, sp3 labeling, vinyl labeling, heteroatom labeling, and combinations thereof.

Iridium-catalyzed HIE is evolving at a time when computationally supported catalyst design is reaching unprecedented levels of sophistication [154,155,156,157,158]. It is expected, therefore, that forthcoming developments in iridium-catalyzed HIE will be enabled by deeper exploration of predictive methods of understanding substrate–catalyst compatibility.