https://doi.org/10.

1038/s41586-024-07987-9

Accelerated Article Preview

W
Reductive alkyl-alkyl coupling from isolable
E
I
nickel-alkyl complexes
EV
PR
Received: 24 May 2024

Accepted: 22 August 2024

Accelerated Article Preview

Samir Al Zubaydi, Shivam Waske, Volkan Akyildiz, Hunter F. Starbuck, Mayukh Majumder,
Curtis E. Moore, Dipannita Kalyani & Christo S. Sevov

E
Published online xx xx xxxx This is a PDF file of a peer-reviewed paper that has been accepted for publication.
Although unedited, the content has been subjected to preliminary formatting. Nature

L
Cite this article as: Zubaydi, S. A. et al.
Reductive alkyl-alkyl coupling from isolable is providing this early version of the typeset paper as a service to our authors and

C
nickel-alkyl complexes. Nature https://doi. readers. The text and figures will undergo copyediting and a proof review before the
org/10.1038/s41586-024-07987-9 (2024)

I
paper is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers

T
apply.

AR
ED
AT
R
E
EL
C
C
A

Nature | www.nature.com
 1 Reductive alkyl-alkyl coupling from isolable nickel-alkyl complexes
2 Samir Al Zubaydi,1 Shivam Waske,1,§ Volkan Akyildiz,1,2,§ Hunter F. Starbuck,1 Mayukh Majumder,1 Curtis E. Moore,1
3 Dipannita Kalyani,3,* Christo S. Sevov1,*

1
4 Department of Chemistry and Biochemistry, The Ohio State University, 151 W Woodruff Avenue, Columbus, OH 43210,

W
5 United States. 2Department of Chemistry, Faculty of Science, Atatürk University, Erzurum 25240, Turkey. 3Discovery
6 Chemistry, Merck & Co., Inc., Rahway, NJ 07065, United States.

IE
§
7 Authors contributed equally
8 *Corresponding authors

EV
9
10 The selective cross-coupling of two alkyl electrophiles to construct complex molecules remains a challenge in organic
11 synthesis.1,2 Known reactions are optimized for specific electrophiles and are not amenable to interchangeably varying
12 electrophilic substrates that are sourced from common alkyl building blocks, such as amines, carboxylic acids, and

PR
13 halides.3–5 These limitations restrict the types of alkyl substrates that can be modified and, ultimately, the chemical space
14 that can be explored.6 Here we report a general solution to these limitations that enables a combinatorial approach to alkyl-
15 alkyl cross-coupling reactions. This methodology relies on the discovery of unusually persistent Ni(alkyl) complexes that
16 can be formed directly by oxidative addition of alkyl halides, redox-active esters, or pyridinium salts. The resulting alkyl
17 complexes can be isolated or directly telescoped to couple with a second alkyl electrophile, which represent cross-selective

LE
18 reactions that were previously unknown. The utility of this synthetic capability is showcased in the rapid diversification of
19 amino acids, natural products, pharmaceuticals, and drug-like building blocks by various combinations of dehalogenative,
20 decarboxylative, or deaminative coupling. In addition to a robust scope, this work provides insights into the organometallic

C
21 chemistry of synthetically relevant Ni(alkyl) complexes through crystallographic analysis, stereochemical probes, and
22 spectroscopic studies.
23
TI
24 Introduction
25 Reductive coupling reactions to construct C-C bonds directly from abundant carbon electrophiles have emerged as attractive
AR

26 methods for rapid molecular diversification because they circumvent the need to source unstable or incompatible organometallic
27 substrates.1,2,7 Such reactions are known as cross-electrophile coupling (XEC) reactions and are becoming increasingly reliable,
28 particularly for cross-coupling reactions of aryl/vinyl and alkyl electrophiles.8–13 Selective formation of those Csp2-Csp3 cross
29 products stems from the distinct mechanisms and rates of reaction of a metal catalyst with either the aryl or alkyl electrophile. In
D

30 contrast, cross-product formation from XEC reactions between similar electrophiles, such as those of two alkyl electrophiles to
31 form Csp3-Csp3 bonds (Fig. 1a), is challenging because catalysts often fail to distinguish one alkyl electrophile from the other.
Unlike redox neutral reactions of coupling partners that undergo complementary activation by oxidation and reduction, 6,14–16
E

32
33 reductive coupling of two electrophiles often results in unpredictable mixtures of cross- and homo-coupled products. As a result,
reactions are often performed with an excess of one electrophile to favor cross-product selectivity.5,17 The challenges to developing
AT

34
35 selective Csp3-Csp3 XEC reactions are further compounded by the greater instabilities and lower rates for C-C coupling of metal-
36 alkyl complexes than of analogous metal-aryl complexes.18
37 This synthetic disconnection has motivated recent efforts to develop cross-alkyl XEC reactions.19 Distinguishing alkyl
ER

38 fragments by size has emerged as one of the most successful approaches to XEC of alkyl substrates. MacMillan and coworkers
39 leveraged this strategy to sort large alkyl radical fragments from small alkyl radicals to effect selective Csp3-Csp3 coupling of 3°
40 alkyl redox-active esters (RAEs) and 1° alkyl bromides.20 A similar sorting strategy was utilized for reductive coupling reactions
41 between 3° and 1° alkyl RAEs by Kawamata, Baran, and Shenvi.21 Finally, a closed-shell, 2e– analog that also relies on steric
EL

42 control for alkyl XEC was demonstrated by See and Lin.22 There, stabilized 3° carbanions are generated upon controlled
43 electroreduction of activated 3° alkyl bromides and preferentially couple with 1° over 3° alkyl halides by an SN2 reaction. However,
44 selective XEC of alkyl electrophiles that have similar steric properties remains challenging. Moreover, known alkyl-alkyl XEC
C

45 reactions are optimized for specific combinations of alkyl electrophiles. As a result, many reactions of substrates from widely
46 available alkyl building blocks (e.g., amines, carboxylic acids, halides) are incompatible with current methods, and combinatorial
AC

47 strategies for dehalogenative, decarboxylative, or deaminative XEC are unknown.

48 We report a general approach for reductive cross-coupling of alkyl electrophiles using an inexpensive organometallic platform
49 (Fig. 1b). NiII(alkyl) complexes can be formed by oxidative addition (OA) between a NiII precursor and alkyl halides, alkyl RAEs,4
50 or alkyl pyridinium salts under reducing conditions. The resulting alkyl complexes exhibit unusual persistence and can be isolated
51 or directly used for cross-coupling with a second alkyl electrophile. This synthetic platform enables previously unknown alkyl
52 XEC reactions and cross-selectivity from the alkyl electrophiles shown in Fig. 1b. Finally, access to persistent Ni(alkyl) complexes
53 enabled parallel coupling reactions, late-stage diversification of complex alkyl electrophiles, and fundamental insights of
 54 synthetically relevant organometallic complexes.
55
56 Results and Discussion
57 We sought to generate Ni(alkyl) complexes directly from the combination of an alkyl electrophile and a NiII precursor under
58 reducing conditions. These parameters are critical to the synthetic utility of the overall XEC reaction because reactions can be performed
59 as one-pot, two-step processes with shelf-stable reagents. The approach also circumvents the need for expensive Ni(0) precursors and
leverages our group’s experience in reductive synthesis of organometallic complexes.23 In particular, we recently identified

W
60
61 (MeBPI)NiII(OAc) complex 1 as an inexpensive reagent for OA of aryl halides under electroreducing conditions to form persistent
62 NiII(aryl) complexes. The resulting aryl complexes undergo reductive coupling with a wide range of alkyl electrophiles and the complex

IE
63 1 can be recovered. The true utility of this strategy is that it enables the exploration of previously inaccessible chemical space.
64 While persistent organonickel complexes of complex aryl fragments remain uncommon, the alkyl analogs are even rarer. Known

EV
65 Ni(alkyl) complexes are generally of the form with alkyl substituents that lack beta-hydrogen atoms or with bulky ligands, such as
66 bis(phosphines)24 or salicylaldiminates,25 which limit the synthetic utility of the system. As a result, Csp3-Csp3 coupling reactions from
67 well-defined organonickel complexes of even simple alkyl fragments – let alone alkyl fragments of drug-like complexity – are rare.26–
29
68 Developing such coupling reactions from persistent organonickel complexes that can be easily prepared by OA of drug-like substrates

PR
69 would establish important synthetic capabilities and simultaneously provide rare insights into the organometallic chemistry of
70 synthetically-relevant complexes.
71 Before directly testing the feasibility of reductively-driven OA to form Ni(alkyl) complexes, we first probed whether such Ni(alkyl)
72 complexes could be formed at all and under what conditions they persist in solution. We hypothesized that transmetalation between
73 (cyclobutyl)ZnBr and 1 offered a more straightforward – albeit synthetically limiting – approach to generating Ni(alkyl) complexes

LE
74 than by OA (Fig. 1c, top route). Addition of the alkyl zinc reagent to the yellow suspension of 1 in THF at -10 °C caused a rapid color
75 change to purple and formed a homogeneous solution. A dark purple solid was isolated after filtration of the solution over basic alumina
76 in the glovebox and evaporation of the solvent from the filtrate. 1H NMR spectroscopy of the solid revealed diamagnetic resonances

C
77 that are consistent with the alkyl complex 2-cyBu. The secondary alkyl complex was isolated in 98% yield and found to be stable for
78 weeks in both solution and solid states. Most importantly, the isolated complex 2-cyBu was found to undergo Csp3-Csp3 coupling with
TI
79 an alkyl RAE under reducing conditions to generate the cross-coupled product 3 in 86% yield (Fig. 1c, right).
80 Bolstered by the persistence and synthetic utility of the isolated alkyl complex, we explored methods to prepare 2-cyBu that are
81 more attractive than by transmetalation and that could be compatible in late-stage diversification reactions of drug-like molecules.
AR

82 Electrolysis of 1 and cyclobutyl bromide (Fig. 1c, middle route) under the conditions developed for the synthesis of aryl complexes
83 formed a brown solution that contained an equimolar mixture of the yellow starting material (1) and purple product (2-cyBu) as
84 determined by 1H NMR spectroscopy. Despite the incomplete formation of the Ni(alkyl) complex, the alkyl bromide was completely
85 consumed. We hypothesized that a key difference between the success of electrochemically-driven OA of aryl halides versus the
inefficiency of OA of alkyl halides stems from the mechanism of OA itself.29,30 Specifically, OA of aryl bromides occurs as a concerted
D

86
87 process at an electrochemically-generated Ni(I) from 1. In contrast, OA of alkyl halides by 1 occurs by a stepwise, radical pathway
E

88 initiated by halogen abstraction with Ni(I). The resulting Ni(II) complex can recombine with the alkyl radical for productive OA, but
89 the Ni(II) can also be reduced again at the polarized cathode to Ni(I) to then mediate another halogen abstraction (see the SI, Fig. S6E).
AT

90 This latter process represents an electrocatalytic reaction that unproductively consumes the alkyl electrophile, rather than a
91 stoichiometric OA reaction. The difference in OA mechanisms is evidenced by CVs of 1 and aryl halides (see the CVs in SI Fig. S6A)
92 that reveal limited current responses due to the consumption of 1 to form Ni(aryl), while CVs of 1 and alkyl halides reveal catalytic
93 currents resulting from continuous reduction and turnover of 1.
ER

94 These insights led us to evaluate OA reactions of alkyl halides at low current densities (mA/cm2 electrode) to limit the overreduction
95 of Ni(II) and with excess alkyl bromide (up to 2 equiv). Electrolysis under these conditions generated 2-cyBu in synthetically useful
96 yields of up to 84%, but the reactions were found to be highly substrate dependent. We instead turned to metal powders as weak
97 reductants (relative to 1) with high surface area that could transiently reduce complex 1 to initiate OA. In this way, an accessible but
EL

98 unfavorable electron transfer from a metal powder mimics a high surface area cathode operating at a low current. Zn powders fail to
99 chemically reduce 1 in solvents typically employed for electrosynthesis or XEC (dimethyl acetamide/formamide or acetonitrile) because
100 of the very negative reduction potential of 1 (E1/2 = -1.75 V vs Fc/Fc+), and no OA reaction was observed. We discovered that Zn
C

101 promotes OA of cyBuBr by 1 when the reaction is performed in solvent mixtures comprised predominantly of THF with LiCl as an
102 additive (Fig. 1c, bottom route). Reactions with Zn were completed in 4 h and formed 2-cyBu in near quantitative yields. These
AC

103 conditions could be extended to OA reactions of alkyl Katritzky pyridinium salts and alkyl RAEs to generate the corresponding
104 complexes in respective yields of 48% and 95%.
105 The rarity of well-defined organonickel complexes with synthetic relevance to C(sp3)-C(sp3) coupling inspired us to evaluate
106 the structural properties of Ni complexes bearing a range of alkyl substituents. In addition to 2-cyBu, we synthesized and
107 crystallographically characterized Ni complexes with methyl (2-Me), (3-phenyl)propyl (2-Pr), cyclopropyl (2-cyPr), and
108 cyclohexyl (2-Cy) substituents. We included crystallographic data from the phenyl analog (2-Ph) as a direct comparison of
109 Ni(alkyl) and Ni(aryl)8 complexes. Finally, we performed reductive coupling reactions of all six organonickel complexes under
110 unoptimized conditions with an alkyl RAE. The comparative structure-reactivity data are tabulated in Figure 1d. Crystallographic
111 analysis of the complexes reveals that the alkyl or aryl substituents in all complexes are bent out of the square plane (N-Ni-C angle
112 < 180°). As a result, the organonickel complexes adopt a distorted square planar geometry. Distortion from a square planar to
113 tetrahedral geometry has been implicated in low yields of C(sp2)-C(sp3) XEC reactions30 and can be quantified by the 𝜏4 parameter:
114 0 for square planar to 1 for tetrahedral geometries (see the SI, Fig. S1). Geometric distortion beyond a certain threshold, rather than
115 steric encumbrance around the Ni center (calculated by %Vbur), may contribute to the low yields of C(sp3)-C(sp3) coupled products
116 from stoichiometric reactions (Fig. 1D, bottom). Reactions of complexes with 𝜏4 ≤ 0.3 resulted in highest yields of cross-coupled

W
117 products. In contrast, reactions with the cyclohexyl analog 2-Cy (𝜏4 = 0.36) generated coupled products in just 12% yield before
118 further reaction development (vide infra).

IE
119 We evaluated the reactivity of Ni complexes of unsubstituted (2-Me), primary (2-Pr), and secondary (2-cyPr) alkyl fragments
120 with common radical precursors of 1°, 2°, and 3° alkyl electrophiles (halides, RAEs, and pyridiniums) to establish the reaction
121 conditions for each substrate combination (Fig. 2a). Coupling reactions of each isolated complex and 1.5 equivalents of electrophile

EV
122 were performed under either electrochemical or chemical reduction. Yields from these stoichiometric reactions (left box, stoich
123 yield) are compared to those from the analogous reactions that subject both alkyl electrophiles to reductive coupling in a single
124 step with 10 mol% of 1 as catalyst under matching conditions (right box, catal yield). Finally, reaction yields are color-coded
125 according to the bar chart in Fig. 2a. This ternary color scheme highlights which substrate combinations are coupled in yields that

PR
126 exceed (blue), match (white), or fall below (red) the statistically-expected yields of cross-products from random activation and
127 coupling of two alkyl electrophiles.
128 Many of the coupling reactions can be conducted under the same conditions as those for oxidative addition using the
129 combination of THF/LiCl and powdered Zn, which creates the potential for a two-step, one-pot procedure (vide infra). This simple

LE
130 protocol for coupling was most successful for reactions of unhindered alkyl fragments. Specifically, 1° or 2° alkyl halides and
131 RAEs coupled with 2-Me and 2-Pr in high yields (4a-b and 5a-b). However, reactions between hindered alkyl partners to form
132 coupled products between two 2° alkyl fragments (5c) or with 3° alkyl groups (6a-b) were generally low yielding when performed
133 with Zn powder (<20% yield, see the SI Fig. S3). Chemical reduction also failed to promote coupling reactions of 1° alkyl

C
134 pyridinium salts. We tested electrochemistry as an alternative strategy to initiate coupling that could controllably deliver reducing
135 equivalents. Electrochemical reactions were performed at a constant current in THF/DMA with NaI as electrolyte. Product yields
TI
136 from electrochemical reactions of alkyl pyridinium salts (4a-c pyridinium couplings) or hindered alkyl electrophiles (5c, 6a-b)
137 improved dramatically over those from reactions performed with Zn powder alone (<20% to >90%; see the SI Fig. S3 for
138 comparison). In general, yields from electroreductive reactions were found to be equal to – or better than – those from coupling
AR

139 reactions performed with Zn powder. Despite this, we elected to report yields from Zn-promoted reactions for substrate
140 combinations that are high-yielding regardless of reductive method because the accessibility of chemical reduction often outweighs
141 the marginal improvement in yield that electrolysis might provide. Nonetheless, electrochemistry is critical to expanding the
142 generality of the coupling methodology to diverse combinations of alkyl electrophiles.
D

143 These studies established conditions for high-yielding reactions of 3° alkyl RAEs and Me (6a), 1° (6b), or 2° (6c) alkyl
144 fragments from the corresponding alkyl complexes. In contrast to reactions of 3° alkyl RAEs, reactions of 3° alkyl bromides
E

145 uniquely generated a mixture of constitutional isomers from cross-coupling of isomerized alkyl intermediates (<25% yield of the
146 3° coupled fragment). Reactions of 1° or 2° alkyl bromides undergo Csp3-Csp3 coupling without isomerization and form products
AT

147 4 and 5. Finally, reactions to form C2°–C3° or C2°–C2° alkyl bonds such as those in products 6c of 5c are particularly challenging
148 because the electrophilic substrates have similar steric properties. In all cases, two-step reactions through a Ni(alkyl) intermediate
149 generated products in higher yields than the catalytic coupling of both electrophiles in a single step. In the best cases, the catalytic
150 reactions generated products in 40-60% yield (white coloring) from substrates of similar steric properties and of the same
ER

151 electrophilic class (1° + 1° alkyl bromide 4b). These cross-product yields reflect a statistical distribution of randomized cross-
152 versus homo-coupling. However, catalytic reactions of different classes of electrophile, such as alkyl-RAEs or -pyridiniums with
153 alkyl halides, generate only trace quantities of cross-products (red colors). These reactions suffer from mismatched rates of
154 substrate activation and typically result in homocoupling of the more easily-activated electrophile. Reactions with persistent
EL

155 Ni(alkyl) complexes bypass this substrate-controlled limitation by decoupling substrate activation into two steps.
156 A complementary strategy for cross-coupling is demonstrated in Fig. 2b. Rather than reactions between a series of alkyl
157 electrophiles and one Ni(alkyl) complex, reactions can be performed between a series of Ni(alkyl) complexes and one electrophile
C

158 of value that is used as the limiting reagent (represented by the alkyl RAE of aspartic acid). Inexpensive complexes of alkyl
159 fragments that are often varied to evaluate structure-activity relationships (Me, Et, iPr, etc) can be generated in situ by OA of the
AC

160 corresponding alkyl electrophile and used without further purification in slight excess relative to the valuable component. We
161 simultaneously evaluated the applicability of parallel electrolysis reactions for this diversification study. Coupling reactions were
162 miniaturized to 0.35 mL from 3 mL and electrolyzed in an HTe-Chem reactor.31 In all cases, reactions generated products in good
163 yields, as measured by gas chromatography, and provided rapid access to a library of unnatural amino acids.
164 Finally, tolerance studies revealed that reactions do not require rigorously dry conditions (see the SI, Fig. S7). Reaction
165 mixtures can be prepared under ambient conditions with undried solvents and simply degassed prior to reductive activation. Such
166 reactions generated products in only slightly lower yields (7d, 66% yield) than those conducted with rigorously inert and dry
167 conditions in a glovebox (83% yield). Additionally, reactions doped with additives containing various functional groups (nitriles,
168 amines, alcohols, etc. see the SI, Fig. S8) generated coupled products in similar yields to those without additives. Collectively,
169 these results highlight the practicality of the method and the mild conditions for coupling.
170 We applied the developed conditions to stoichiometric reactions of substrate classes that complement those of the
171 combinatorial studies in Fig. 2. Reactions in Fig. 3 were performed on 0.15 mmol scale to provide useful quantities of cross-
172 coupled products (10 - 60 mg). We first evaluated reactions of 3° alkyl RAEs to form quaternary carbon products from couplings
173 with 1° (8) and 2° (9, 10) alkyl electrophiles. Reactions of bulky substrates revealed that Csp3-Csp3 coupling is most reliable if the

W
174 less substituted substrate is activated first, and the more substituted substrate is introduced as the second electrophile (11, 70%).
175 Reversing this order, as illustrated in the convergent synthesis of 11, results in low yields (19%) prior to additional optimization
176 (43%, e-chem).

IE
177 The ability to cross-couple alkyl substrates of multiple electrophilic classes allowed us to derivatize amino acids (AAs) through
178 deaminative (lysine), decarboxylative (aspartic and glutamic acid, proline), and dehalogenative (serine) reactions. Coupling

EV
179 reactions of halogenated serine or activated aspartic acid converge to form products (12a-d) in good, isolated yields. Modifications
180 at gamma and epsilon positions of AAs were successful through decarboxylative or deaminative functionalization of glutamic acid
181 (13a-b) and lysine (14a-b) derivatives, respectively. These reactions all employ the AA fragment as the 2nd electrophile in
182 combination with a preformed organonickel complex. Reactions can also be performed in reverse order where organonickel

PR
183 complexes of proline and aspartic acid fragments are generated in situ by OA in the 1st step and coupled with alkyl electrophiles
184 in the 2nd step to form respective products 15a-b and 16. These studies highlight the operational flexibility of the coupling
185 methodology.
186 Finally, we performed reactions of alkyl electrophiles derived from pharmaceuticals or natural products. Brominated galactose
187 and ribose sugars undergo successful coupling with Ni(alkyl) intermediates to install both 1° (17b-c) and 2° (17a, 22a) alkyl

LE
188 fragments. Alkylation of the anomeric position of galactose occurs in high yield from the corresponding bromide to generate
189 methylated galactose 22d as a single stereoisomer. Alternatively, coupling reactions of an organonickel complex of galactose with
190 alkyl bromides rapidly generated sugar derivatives in good, isolated yields (22c-b). This approach was also successful for the

C
191 coupling of various alkyl pyridiniums (18a, 18c) and halides (18b) with the OA complex of halogenated Nopol. Derivatives of
192 thymidine can be prepared without protection of alcohols or acidic amines (20), and a medicinally relevant azetidine fragment32
193 can be easily installed into the backbone of an Oxaprozin derivative (21). Finally, a wide range of derivatives of citronellol (19) or
TI
194 vitamin E (23a-e) are easily accessible from couplings of the corresponding organonickel intermediates.
195 The systematic evaluation of reaction scope revealed a breadth of successful coupling reactions but also revealed synthetic
AR

196 limitations of the current organometallic platform. Such limitations, as well as the unexpected effect of electroreduction versus Zn-
197 promoted coupling, motivated an investigation of the mechanism of C-C coupling to aid in the development of more reactive
198 complexes and improved conditions. Fig. 4a illustrates two pathways for C-C bond formation form Ni(alkyl) complexes that are
199 frequently invoked. The first mechanism is a bimolecular homolytic substitution (SH2) of the alkyl fragment bound to Ni by a
200 transient alkyl radical. This outer-sphere mechanism for C-C bond formation has been proposed in reactions of CoIIIalkyl
D

201 complexes33 and in recent stoichiometric34 and catalytic reactions of Ni.20,35 The second mechanism involves capture of the alkyl
202 radical at NiII(alkyl) to generate a NiIII(dialkyl) species that undergoes Csp3-Csp3 bond-forming reductive elimination (RE). This
E

203 inner-sphere pathway mirrors the mechanism of Csp2-Csp3 coupling reactions and has also been proposed in recent cross-alkyl
204 coupling reactions.3,36,37
AT

205 Distinguishing between these two mechanisms identifies the site of radical capture. To probe the mechanism of the C–C bond-
206 forming step, we investigated the stereochemistry of cross-coupled products that were generated from reactions of a
207 stereochemically-defined Ni(alkyl) complex. We synthesized complex 24 as a 2:1 mixture of syn:anti diastereomers by OA from
ER

208 the corresponding alkyl bromide. The syn stereochemistry of the major diastereomer was confirmed by 1H NMR NOESY and by
209 X-ray crystallography. Coupling reactions by an outer-sphere SH2 mechanism of the 2:1 syn:anti mixture of 24 would invert the
210 stereochemistry at carbon and generate a 1:2 ratio of syn:anti products. In contrast, coupling by the inner-sphere pathway from a
211 NiIII(dialkyl) intermediate would retain the stereochemistry at carbon and translate the diastereomeric ratio (d.r.) of the starting
EL

212 material to the product. We tested this mechanistic hypothesis according to the reaction in Fig. 4b (right). The cross-coupled
213 product 25 was isolated in 67% yield as a 2:1 mixture of syn:anti. The structures of the major and minor diastereomers were
214 determined by 1H NMR NOESY. The retained stereochemistry from Ni(alkyl) complex to product supports an inner-sphere
215 pathway of radical capture at Ni, followed by Csp3-Csp3 bond-forming RE.
C

216 To support these studies, we confirmed that the Ni-alkyl bond does not reversibly homolyze in solution, which could scramble
AC

217 the stereochemistry of the complex and invalidate the stereochemical probe. First, complex 24 is persistent in solution, and the d.r.
218 of 24 does not change over time. Second, coupling reactions of a 2-hex-1-ene complex, which would undergo rapid 1,5-cyclization
219 upon Ni-alkyl bond homolysis, form only the linear products with no detection of products from radical-based cyclization (see the
220 SI, Fig. S4). These studies underscore the fidelity of the Ni-alkyl bond and validate the design of the stereochemical probe.
221 Our studies provide unambiguous support for Csp3-Csp3 bond-forming RE from a synthetically-relevant NiIII(dialkyl)
222 complex. Studies on Ni(alkyl) complexes by Diao and coworkers have implicated high barriers to this elementary organometallic
223 reaction.37 There, NiII(CH2TMS) complexes failed to undergo Csp3-Csp3 coupling with benzyl radicals, but the analogous
224 NiII(mesityl) complex readily reacted to form a Csp2-Csp3 bond. We computationally evaluated the barriers to RE our system for
225 both Ni(alkyl) and Ni(aryl) complexes. Calculations using density functional theory (DFT) predict a free energy barrier to RE of
226 19.8 kcal/mol from NiIII(dialkyl) 26 (Fig. 4b). The analogous reaction from 2-Ph is predicted to proceed with a barrier of only 16.3
227 kcal/mol. The computed barrier to Csp3-Csp3 coupling is consistent with the observation of a slow conversion of the purple solution
228 of Ni(alkyl) over the course of hours when coupling is performed at 0 °C (t1/2 ≈ 20 min for ∆G‡ = 19.8 kcal/mol at 0 °C).
229 These computational results and experimental observations suggested that NiIII(dialkyl) intermediates persist on the order of
230 minutes at ambient temperatures and might be detectable at low temperatures. We performed a coupling reaction between two

W
231 secondary alkyl fragments at low temperatures to slow the rate of RE (Fig. 4c). Coupling was initiated with Zn powder at -14 °C
232 for just 10 min, at which point an aliquot of the solution was collected and frozen for analysis by electron paramagnetic resonance
233 (EPR) spectroscopy at 77 K (Fig. 4C, right). Resonances consistent with a rhombic NiIII complex were observed,38 which we

IE
234 attribute to the NiIII(dialkyl) species 26.
235 The putative NiIII(dialkyl) detected by EPR spectroscopy is synthetically relevant based on studies in Fig. 2a that confirmed

EV
236 coupling of the same alkyl fragments to form 5c in near quantitative yield. However, these high yields were only achieved when
237 coupling was performed electrochemically. Coupling reactions promoted by Zn as the reductant generated 5c in just 18% yield.
238 Insights into the reactivity of NiIII(dialkyl) intermediates provide a possible explanation as to why yields are generally higher when
239 the coupling step is promoted by electrochemistry than Zn and why the differences in yields are greatest for coupling reactions of

PR
240 bulky alkyl fragments (see e-chem vs Zn reactions in Fig. 2a). Specifically, NiIII(dialkyl) intermediates that are slow to undergo
241 C(sp3)-C(sp3) RE may instead undergo competitive reduction39 by the slurry of high surface-area Zn powder. In contrast,
242 electroreduction occurs with restricted currents exclusively at an electrode. It is less likely that the low concentration of
243 NiIII(dialkyl) will encounter the cathode compared to the slurried Zn prior to RE. Reactions of bulky substrates that undergo
244 challenging RE reactions are most susceptible to reductive quenching and, therefore, benefit most from electroreductive control.

LE
245
246 Conclusions
247 Collectively, these findings underscore the importance of designing alkyl complexes with high persistence to undergo a

C
248 relatively slow RE reaction in preference to the myriad pathways for decomposition. Additionally, insights into the mechanisms
249 of OA and RE facilitated the reliable synthesis of persistent Ni(alkyl) complexes and improved the synthetic utility of the
TI
250 organonickel complexes for cross-coupling with myriad alkyl electrophiles. Critically, many of the examples in this work represent
251 previously unknown coupling reactions that enable the exploration of new chemical space. We anticipate these complexes will
252 serve as platforms for challenging coupling reactions as well as for fundamental studies into the organometallic chemistry of Ni-
AR

253 alkyl complexes.

254 References
255 1. Tasker, S. Z., Standley, E. A. & Jamison, T. F. Recent advances in homogeneous nickel catalysis. Nature 509, 299–309 (2014).
256 2. Wang, X., Dai, Y. & Gong, H. Nickel-Catalyzed Reductive Couplings. Top. Curr. Chem. 374, 43 (2016).
D

257 3. Yue, H. et al. Nickel-catalyzed C-N bond activation: Activated primary amines as alkylating reagents in reductive cross-coupling. Chem. Sci. 10,
258 4430–4435 (2019).
259 4. Huihui, K. M. M. et al. Decarboxylative Cross-Electrophile Coupling of N-Hydroxyphthalimide Esters with Aryl Iodides. J. Am. Chem. Soc. 138,
E

260 5016–5019 (2016).

261 5. Li, P. et al. Nickel-electrocatalysed C(sp3)–C(sp3) cross-coupling of unactivated alkyl halides. Nat. Catal. 15–17 (2024) doi:10.1038/s41929-024-
AT

262 01118-3.
263 6. Lyon, W. L. & MacMillan, D. W. C. Expedient Access to Underexplored Chemical Space: Deoxygenative C(sp3)-C(sp3) Cross-Coupling. J. Am.
264 Chem. Soc. 145, 7736–7742 (2023).
265 7. Weix, D. J. Methods and Mechanisms for Cross-Electrophile Coupling of Csp2 Halides with Alkyl Electrophiles. Acc. Chem. Res. 48, 1767–1775
266 (2015).
ER

267 8. Truesdell, B. L., Hamby, T. B. & Sevov, C. S. General C(sp2)-C(sp3) Cross-Electrophile Coupling Reactions Enabled by Overcharge Protection of
268 Homogeneous Electrocatalysts. J. Am. Chem. Soc. 142, 5884–5893 (2020).
269 9. Hamby, T. B., LaLama, M. J. & Sevov, C. S. Controlling Ni redox states by dynamic ligand exchange for electroreductive Csp3–Csp2 coupling.
270 Science 376, 410–416 (2022).
271 10. Zhang, P., Le, C. C. & MacMillan, D. W. C. Silyl Radical Activation of Alkyl Halides in Metallaphotoredox Catalysis: A Unique Pathway for Cross-
EL

272 Electrophile Coupling. J. Am. Chem. Soc. 138, 8084–8087 (2016).

273 11. Harwood, S. J. et al. Modular terpene synthesis enabled by mild electrochemical couplings. Science 375, 745–752 (2022).
274 12. Le, C. C. et al. A General Small-Scale Reactor to Enable Standardization and Acceleration of Photocatalytic Reactions. ACS Cent. Sci. 3, 647–653
275 (2017).
276 13. Hansen, E. C. et al. New ligands for nickel catalysis from diverse pharmaceutical heterocycle libraries. Nat. Chem. 8, 1126–1130 (2016).
C

277 14. Choi, J. & Fu, G. C. Transition metal–catalyzed alkyl-alkyl bond formation: Another dimension in cross-coupling chemistry. Science 356, (2017).
278 15. Johnston, C. P., Smith, R. T., Allmendinger, S. & MacMillan, D. W. C. Metallaphotoredox-catalysed sp3-sp3 cross-coupling of carboxylic acids with
AC

279 alkyl halides. Nature 536, 322–325 (2016).

280 16. Sakai, H. A. & Macmillan, D. W. C. Nontraditional Fragment Couplings of Alcohols and Carboxylic Acids: C(sp3)-C(sp3) Cross-Coupling via Radical
281 Sorting. J. Am. Chem. Soc. 144, 6185–6192 (2022).
282 17. Smith, R. T. et al. Metallaphotoredox-Catalyzed Cross-Electrophile C sp 3 –C sp 3 Coupling of Aliphatic Bromides. J. Am. Chem. Soc. 140, 17433–
283 17438 (2018).
284 18. Dawson, G. A., Spielvogel, E. H. & Diao, T. Nickel-Catalyzed Radical Mechanisms: Informing Cross-Coupling for Synthesizing Non-Canonical
285 Biomolecules. Acc. Chem. Res. 56, 3640–3653 (2023).
286 19. Kranthikumar, R. Recent Advances in C(sp3)-C(sp3) Cross-Coupling Chemistry: A Dominant Performance of Nickel Catalysts. Organometallics 41,
287 667–679 (2022).
288 20. Liu, W., Lavagnino, M. N., Gould, C. A., Alcázar, J. & MacMillan, D. W. C. A biomimetic SH2 cross-coupling mechanism for quaternary sp3-carbon
289 formation. Science 374, 1258–1263 (2021).
290 21. Gan, X. et al. Carbon quaternization of redox active esters and olefins by decarboxylative coupling. Science 384, 113–118 (2024).
291 22. Zhang, W. et al. Electrochemically driven cross-electrophile coupling of alkyl halides. Nature 604, 292–297 (2022).
292 23. Dinh, L. P. et al. Persistent organonickel complexes as general platforms for Csp2–Csp3 coupling reactions. Nat. Chem. (2024) doi:10.1038/s41557-
293 024-01528-7.
294 24. Kitiachvili, K. D., Mindiola, D. J. & Hillhouse, G. L. Preparation of Stable Alkyl Complexes of Ni(I) and Their One-Electron Oxidation to Ni(II)
295 Complex Cations. J. Am. Chem. Soc. 126, 10554–10555 (2004).
296 25. Wagner, C. L. & Diao, T. Nickel-Carbon σ-Bonded Complexes. in Comprehensive Organometallic Chemistry IV 271–356 (Elsevier, 2022).

W
297 doi:10.1016/B978-0-12-820206-7.00141-4.
298 26. Csok, Z., Vechorkin, O., Harkins, S. B., Scopelliti, R. & Hu, X. Nickel Complexes of a Pincer NN2 Ligand: Multiple Carbon−Chloride Activation of
299 CH2Cl2 and CHCl3 Leads to Selective Carbon−Carbon Bond Formation. J. Am. Chem. Soc. 130, 8156–8157 (2008).

IE
300 27. Anderson, T. J., Jones, G. D. & Vicic, D. A. Evidence for a NiI Active Species in the Catalytic Cross-Coupling of Alkyl Electrophiles. J. Am. Chem.
301 Soc. 126, 8100–8101 (2004).
302 28. Jones, G. D. et al. Ligand Redox Effects in the Synthesis, Electronic Structure, and Reactivity of an Alkyl−Alkyl Cross-Coupling Catalyst. J. Am.
303 Chem. Soc. 128, 13175–13183 (2006).

EV
304 29. Griego, L., Chae, J. B. & Mirica, L. M. A bulky 1,4,7-triazacyclononane and acetonitrile, a Goldilocks system for probing the role of NiIII and NiI
305 centers in cross-coupling catalysis. Chem 10, 867–881 (2024).
306 30. Akana, M. E. et al. Computational Methods Enable the Prediction of Improved Catalysts for Nickel-Catalyzed Cross-Electrophile Coupling. J. Am.
307 Chem. Soc. 146, 3043–3051 (2024).
308

PR
31. Rein, J. et al. Unlocking the Potential of High-Throughput Experimentation for Electrochemistry with a Standardized Microscale Reactor. ACS Cent.
309 Sci. 7, 1347–1355 (2021).
310 32. Gatazka, M. R., McFee, E. C., Ng, C. H., Wearing, E. R. & Schindler, C. S. New strategies for the synthesis of 1- and 2-azetines and their applications
311 as value-added building blocks. Org. Biomol. Chem. 20, 9052–9068 (2022).
312 33. Wang, Y. & Begley, T. P. Mechanistic Studies on CysS – A Vitamin B12-Dependent Radical SAM Methyltransferase Involved in the Biosynthesis of
313 the tert-Butyl Group of Cystobactamid. J. Am. Chem. Soc. 142, 9944–9954 (2020).
314 34. Bour, J. R., Ferguson, D. M., McClain, E. J., Kampf, J. W. & Sanford, M. S. Connecting Organometallic Ni(III) and Ni(IV): Reactions of Carbon-

LE
315 Centered Radicals with High-Valent Organonickel Complexes. J. Am. Chem. Soc. 141, 8914–8920 (2019).
316 35. Chen, R. et al. Alcohol-alcohol cross-coupling enabled by SH2 radical sorting. Science 383, 1350–1357 (2024).
317 36. Qin, T. et al. A general alkyl-alkyl cross-coupling enabled by redox-active esters and alkylzinc reagents. Science 352, 801–805 (2016).
318 37. Lin, Q., Spielvogel, E. H. & Diao, T. Carbon-centered radical capture at nickel(II) complexes: Spectroscopic evidence, rates, and selectivity. Chem 9,

C
319 1295–1308 (2023).
320 38. Watson, M. B., Rath, N. P. & Mirica, L. M. Oxidative C-C bond formation reactivity of organometallic Ni(II), Ni(III), and Ni(IV) complexes. J. Am.
321 Chem. Soc. 139, 35–38 (2017).
TI
322 39. Lin, Q., Dawson, G. & Diao, T. Experimental Electrochemical Potentials of Nickel Complexes. Synlett 32, 1606–1620 (2021).
323
324 Data Availability: All experimental data, analytical procedures, cell designs, copies of spectra, and CIF data are available in the supplementary information.
AR

325
326 Acknowledgments This work was supported by the National Institutes of Health (NIH R35 GM138373) and a Camille and Henry Dreyfus Teacher Scholar
327 Award to C.S.S. V.A. thanks the TÜBİTAK - BİDEB (2214-A International Research Fellowship Program for PhD Students) for a scholarship. We thank Dr.
328 Luke Lewis for assistance with EPR spectroscopic measurements.
329
330 Author Contributions: S.A., D.K., and C.S.S. conceived the work and designed the experiments. S.W. and V.A. contributed equally. S.A., S.W., V.A., H.F.S.,
D

331 and M.M. performed all experiments and collected all data. S.A., S.W., V.A., H.F.S., and M.M. synthesized all substrates. H.F.S. performed parallel electrolysis
332 reactions. C.E.M performed collection and refinement of crystallographic data. M.M. performed computational studies. All authors analyzed the data. S.A. and
333 C.S.S. wrote the manuscript and all authors provided revisions.
E

334
335 Competing Interests: The authors do not claim any competing interests.
AT

336
337 Author Information Correspondence should be addressed to [email protected] and [email protected]
338
339
ER

340 Fig. 1. Synthetic Approach, Synthesis, and Properties of Ni-Alkyl Complexes. (a) Selectivity in alkyl cross-coupling reactions. (b) Stoichiometric strategy for
341 cross-coupling of alkyl electrophiles. (c) Comparison of synthetic routes for Ni(alkyl) and subsequent reactivity. (d) Structural comparison of Ni(alkyl) complexes
342 from X-ray crystallographic analysis. a 1.5 equiv cyBu-[Py+], 10 equiv Zn, 100 mM LiCl, 1:1 DMA:THF (50 mM), 40 ˚C, 30 min.
343
EL

344 Fig. 2. Synthesis and Reactivity of Organonickel Complexes. (a) Analytical yields determined by gas chromatography from combinatorial cross-coupling
345 reactions of alkyl RAEs, bromides, and pyridinium salts. Yields highlighted in gray boxes are from electrochemical reactions, all others are performed with Zn.
346 (b) Parallel electrolysis reactions performed in HTe-Chem. Reported yields are analytical yields determined by gas chromatography.
347
C

348 Fig. 3. Application to Complex Molecules. Isolated yields are reported versus Ni loading (0.15 mmol). Purple circles highlight the carbon atoms of fragments
349 that were prepared as organonickel complexes in the 1st step. Light red circles highlight the carbon atoms of electrophilic fragments introduced in the 2 nd step. a Via
350
AC

e-chem.
351
352 Fig. 4. Mechanistic Studies. (a) Proposed pathways for C–C bond formation. (b) Synthesis and reaction of a stereochemically-defined complex as a probe into
353 the mechanism of C–C bond formation. Crystallographic data of the major diastereomer of 24 (syn) is represented as an ORTEP structure with ellipsoids at 50%
354 probability. (c) EPR spectroscopic analysis of a low temperature coupling reaction (9.45 GHz, THF, -196 ˚C, g1 = 2.237, g2 = 2.134, g3 = 2.082), and a proposed
355 pathway of undesirable overreduction.
356
 EV
PR
LE
a b general strategy for C(sp3)-C(sp3) cross-selectivity
X + X

[LnM] R-X R-X

+ 2e– NiII + R-CO2R* e– NiII + e– R-CO2R*
R-[Py+] R-[Py+]
persistent
Ni-Ralk

C
transmetalation
cyBu-ZnBr, >95%
Me Br cyBu-Br Nphth
THF, -10 ˚C, 5 min
Nphth
or O
N e-chem

N
N

Ni OAc +

Me
or
Ph

N+
O N

O O cyBu-RAE

Ph
TI cyBu-Br, 84%
1:1 DMF:THF, LiBr (250 mM)
Ni(-)Zn(+), 1 mA, 2 F/mol, r.t.
chemical reduction
cyBu-Br cyBu-RAE cyBu-[Py+]
92% 95% 48%a
+– Ni

2-cyBu
O NHP
(1.5 equiv)
LiCl, Zn
THF, r.t., 16 h
3, 86%
AR
cyBu-[Py+] 9:1 THF:DMF, LiCl (100 mM)
1, Ni OAc Ph Zn (5 equiv), 0 ˚C, 4 h

d Ni H Ni H Ni Ni Ni Ni
H H H
H H
H Ph
2-Me 2-Pr 2-cyPr 2-cyBu 2-Cy 2-Ph

N Me N2 Ni N3
N N N1
N
Ni
N H C1
ED

Me
RR
X-ray structure:
isolated yield: 89% 95% 83% 96% 83% 71%
∠N2-Ni-C1: 157.0˚ 160.2˚ 161.2˚ 158.1˚ 156.1˚ 156.9˚
τ4 : 0.30 0.28 0.26 0.29 0.36 0.27
Ni–C bond length: 1.952 Å 1.956 Å 1.934 Å 1.948 Å 2.012 Å 1.913 Å
Vbur [Ni]: 71.4% 74.2% 75.3% 76.0% 77.3% 75.6%

O NHP
AT

Nphth Nphth Me Nphth Nphth Nphth Nphth Nphth

Ph
XEC rxn yield: 91% >95% 95% 86% 12% >95%
R
LE
a
e-chem: 250 mM LiCl, THF:DMF 1:1 R
Ni(-)/Zn(+) or Co(+), 5 mA, 2 F/mol, r.t. 10 mol% 1
Ni R + X X + Br/I R

C
stoich catal
chemical: Zn (5 equiv) same conditions
1.5 equiv 100 mM LiCl, THF, r.t.,12 h as stoich. rxn 1.5 equiv
0% 50% 100%
X = Br, RAE, or Py+ yield X = Br, RAE, or Py+
redox active esters bromides
Ni Me Ni Ph Ni Ni Me Ni Ph Ni

TI
2-Me 2-Pr 2-cyBu 2-Me 2-Pr 2-cyBu

O NHP
Nphth Nphth Me Nphth Nphth 4a: 1°-Me 4b: 1°-1° 4c: 1°-2°
Ph 1°-Br 79% 33% 72% 54% 98% 2%
1°-RAE 4a 91% 29% 4b 79% 12% 4c 98% 0%
stoich catal 5a: 2°-Me 5b: 2°-1° 5c: 2°-2°
O Ph

AR
Me 2°-Br 98% 58% 84% 15% 98% 0%
NHP
BzN BzN BzN BzN
pyridiniums
2°-RAE 5a 98% 52% 5b 81% 0% 5c 52% 0% 4a: 1°-Me 4b: 1°-1° 4c: 1°-2°
O Ph 1°-[Py+] 58% 16% 48% 6% 73% 0%
Me
NHP
5a: 2°-Me 5b: 2°-1° 5c: 2°-2°
NHBoc NHBoc NHBoc NHBoc 2°-[Py+] 87% 8% 95% 3% 95% 0%
3°-RAE 6a 95% 31% 6b 95% 0% 6c 70% 0%

b parallel electrolysis reactions Boc

Boc N
N
D
O NHP NBoc
R
NHCbz NHCbz NHCbz NHCbz NHCbz NHCbz NHCbz NHCbz
+ Ni
CO2Me HTe-Chem CO2Me CO2Me CO2Me CO2Me CO2Me CO2Me CO2Me
1 equiv 1.5 equiv 7a 73 7b 92 7c 44 7d 65 7e 61 7f 83 7g 53
 IE
EV
PR
LE
Ni + R–X
e-chem: 250 mM NaI, THF:DMA 1:1
2e–
Ni(-)/Zn(+) or Co(+), 5 mA, 2 F/mol, r.t.
Ni R or Ni R + X or R
isolated telescoped 1.5 equiv Zn powder: Zn (5 equiv)
100 mM LiCl, THF, r.t.,12 h
X = Br, RAE, or Py+

C
O 1° RX 19% 2° RX
Me 10b, 42% 70% a
NBoc MeO2C Ph (43%)
Ph BocHN Ph Ni Ni
8, 58% 9, 66% 10a, 52% 10c, 52% 2° RX 11 1° RX
amino acid derivatives
from L-serine

Br
O
OR
NHBoc

Ralk OR
Ni Ralk

O
from L-aspartic acid

NHP
O
O
OR
NHBoc

R = tBu or Bn
NHP
O O

TI
from L-glutamic acid

OR
NHCO2R
BocN
13a, 67%
R = Bn

13b, 52%
R = tBu
from L-proline

Ni Boc
N
BocHN

15a, 40%
Ph
Boc
N
BzN

15b, 40%
Boc
N
AR
N
NHBoc from L-lysine BocN NHCbz Ph
Me Ph O 14a, 56% Ni O NHCbz
12a, 78% 12c, 37% CO2Me
[Py+]
OtBu Me CO2Me
12b, 61% Nphth 12d, 52% NHBoc 14b, 81% 16, 36%

natural products & pharmaceuticals

from D-ribose from (-)-nopol NBoc

BzN [Py+]
Br Nphth
O 17a, 74% Ni 18a, 51% 19, 77%
MeO
Ph O
BocN I Ph O
O O 17b, 67% 18b, 92% HN O Ph N
O N
ED

Nphth Nphth [Py+] OH Ph

17c, 64% 18c, 86% 20, 61% 21, 71%

from D-galactose from vitamin E succinate

O
O -or- O OAc O
O O NHP
O Ni O AcO OAc
O
Ni O
O Br O OAc
O O NHCbz Me O
NHP Nphth
Br Ph CO2Bn Me
NBoc Ph
from α-Br
AT

O
22a, 81% 22b, 46% 22c, 68% 22d, >95% 23a, 72% 23b, 63% 23c, 57% 23d, 58% 23e, 80%
RE
EL
 E
PR
LE
a outer-sphere reductive elimination (SH2) inner-sphere reductive elimination

C
R
inversion of vs retention of
NiII σ* R R configuration NiII R NiIII R configuration

TI
Br CO2Me H H Nphth H H
Br
Ni OAc Nphth
Ni CO2Me CO2Me
Zn (5 equiv), Zn (5 equiv),
1 THF, LiCl, r.t., 16 h 24, 82% THF, LiCl, r.t., 16 h 25, 67%
2 : 1 syn:anti 2 : 1 syn:anti
retention
H H
AR
N3
C1 R• NiIII CO2Me reductive elimination
Ni
∆G‡DFT =19.8 kcal/mol
Nphth
N1
N2 24 (syn), X-ray

c 259 281 307 337 375

Br NBz e-chem: >95% BField (mT)
5 equiv Zn: 25% –– exp
Ni + NiIII NBz NBz
LiCl (35 mM), THF slow RE
2-cyPr Zn, -14 ˚C, 10 min, 26, EPR active 5c –– sim
then 77 K
ED

26 EPR spectrum
+ e– g-value
[LNiII(alk)2]–
competitive reduction w/ excess Zn unstable
AT
R
LE