Introduction

Optimization of cross-coupling reaction conditions commonly treats the “type of base” and the “equivalents of base” as clearly defined variables. For example, “sodium ethoxide (NaOEt), 3.0 equiv” means that the corresponding mass is weighed according to the molecular weight of NaOEt and is then used to compare the outcomes obtained under different conditions.

This comparison rests on two chemical assumptions: the added solid still consists primarily of NaOEt, and increasing the amount of NaOEt charged mainly increases the amount of NaOEt actually present in the reaction system.

The study of solid sodium alkoxides by Wethman et al. showed that NaOEt and sodium methoxide (NaOMe) exposed to air may no longer satisfy these assumptions. Reagents from different batches or stored for different lengths of time may contain different actual amounts of sodium alkoxide, together with other sodium salts formed through decomposition in air. When changes in reagent composition are superimposed on the equilibria involving the base in cross-coupling reactions, they can alter the outcomes of condition screening and loading optimization and may also make experiments performed with different batches or at different times difficult to reproduce.[1]

1. Why Can the Same Nominal Amount of NaOEt Produce Opposite Results?

The authors investigated a palladium-catalyzed C(sp²)–C(sp³) Suzuki–Miyaura cross-coupling in which an alkyl-9-borabicyclo[3.3.1]nonane (9-BBN) served as the nucleophilic coupling partner, an electron-deficient aryl chloride served as the other coupling partner, and NaOEt was used as the base.[1]

Four sets of data revealed the problem:

Effects of NaOEt Sample Condition and Nominal Loading on the Yield of the Suzuki–Miyaura Coupling

Note: The 51%, 96%, and 58% values are ¹H NMR yields; the isolated yield under the 96% condition was 95%. The two experiments using 3.0 equiv were single runs, whereas the remaining yields were averages of two independent experiments.[1]

First, consider the NaOEt sample with a confirmed assay of 99%. When the nominal loading was increased from 1.5 to 3.0 equiv, the yield decreased from 96% to less than 5%. In this specific coupling system, increasing the NaOEt loading had a pronounced negative effect.

The severely degraded commercial sample contained less than 1% residual NaOEt. This sample afforded yields of 51% and 58% at nominal loadings of 1.5 and 3.0 equiv, respectively. Because the 3.0 equiv condition was a single experiment and the NaOEt content of the sample was already very low, these two results cannot be used to determine whether increasing the amount of NaOEt improved the reaction.

At the same nominal loading of 3.0 equiv, 99% NaOEt gave a yield of less than 5%, whereas the severely degraded sample gave 58%. This difference indicates that reagent degradation altered the actual reaction conditions. The current experiments do not distinguish how much of the effect arose from the reduced NaOEt loading and how much arose from the degradant salts.[1]

Assay analysis showed that the commercial sample without appreciable degradation contained 99% NaOEt, whereas the severely degraded commercial sample contained less than 1% residual NaOEt. The severely degraded sample afforded yields of 51% and 58% at nominal loadings of 1.5 and 3.0 equiv, respectively. Because most of the solid in this sample was no longer NaOEt, these two results reflect the overall reaction performance in the simultaneous presence of residual NaOEt and the solid components formed after degradation. They cannot be interpreted as a loading–yield relationship for pure NaOEt.[1]

Accordingly, these experiments involve two distinct issues:

1. NaOEt itself exhibits a pronounced loading effect;

2. After NaOEt degradation, the nominal equivalents no longer represent the actual sodium alkoxide composition.

The second issue further complicates interpretation of the first.

2. Why Does Increasing NaOEt Not Necessarily Improve Suzuki–Miyaura Coupling?

2.1 The Base Can Help Determine Which Species Undergoes Transmetalation

Suzuki–Miyaura coupling generally proceeds through oxidative addition, transmetalation, and reductive elimination. In many systems, the base does more than simply neutralize byproducts; it participates in the chemical equilibria between palladium species and organoboron species and thereby influences transmetalation.[2–4]

In the arylboronic acid, weak-base, and aqueous-solvent systems studied by Carrow and Hartwig, experimental evidence supported transmetalation between an arylpalladium hydroxo species and a boronic acid:

Ar–Pd–X → Ar–Pd–OH

Ar–Pd–OH + Ar′B(OH)₂ → Ar–Pd–Ar′ → Ar–Ar′

This finding shows that a base can directly affect the species entering the transmetalation step by changing the form of the palladium intermediate.[3]

In a kinetic study of another class of Suzuki–Miyaura systems, Amatore, Jutand, and Le Duc found that OH⁻ exerted both promoting and inhibitory effects. OH⁻ promoted the formation of the reactive palladium hydroxo species and accelerated reductive elimination. At higher concentrations, however, it converted the arylboronic acid into the less reactive anionic aryltrihydroxyborate. In that system, these effects collectively produced a nonmonotonic dependence of the observed reaction rate on the OH⁻ concentration: the rate could first increase and then decrease as the OH⁻ concentration increased.[2]

The competing relationships can be represented in simplified form as follows:

Promoting pathway:

Ar–Pd–X → Ar–Pd–OH → Transmetalation

Inhibitory pathway:

Ar′B(OH)₂ + OH⁻ ⇌ [Ar′B(OH)₃]⁻

This leads to an important mechanistic conclusion: an increase in base concentration does not inherently correspond to faster coupling or a higher yield; the reaction outcome depends on the equilibria among different palladium-containing and boron-containing species.

2.2 This Mechanism Cannot Be Applied Directly to the NaOEt/9-BBN System

Wethman et al. studied a system containing NaOEt, alkyl-9-BBN, an aryl chloride, and tetrahydrofuran, whereas the kinetic studies discussed above primarily examined arylboronic acids under OH⁻ conditions. The structures of the organoboron reagents, solvents, and bases were different, and the activation modes of organoboron reagents are strongly system-dependent.[4]

Therefore, the low yield obtained with 3.0 equiv NaOEt cannot currently be attributed directly to a particular palladium species or a particular borate intermediate.

What can be established is that the base in Suzuki–Miyaura coupling can participate in species equilibria associated with transmetalation, and increasing the amount of base is not a simple positive variable. Wethman’s experiments also showed that this particular reaction exhibited a pronounced NaOEt loading effect over the range of 1.5–3.0 equiv.[1]

3. What Changes in the Reaction Conditions After NaOEt Degradation?

NaOEt degradation produces two direct changes in the reaction conditions.

3.1 The Actual Content of Undegraded NaOEt Decreases

In the original study, the high-quality commercial NaOEt used in the Suzuki–Miyaura coupling had an assay of 99%, whereas the NaOEt content of the severely degraded sample was below 1%.[1]

Consequently, weighing two solids at the same “nominal equivalents” according to the molecular weight of NaOEt does not provide the same amount of NaOEt.

This point is especially important for a reaction that has already shown a pronounced NaOEt loading effect:

99% NaOEt, 3.0 equiv → <5%

This result indicates that a higher NaOEt loading itself is associated with a low yield. After the NaOEt content of the severely degraded sample had decreased substantially, the reaction had effectively moved away from the original high-NaOEt composition regime, even though the experimental record still stated “3.0 equiv.”

3.2 The Original NaOEt Is Replaced by New Sodium Salts

The authors further found that NaOEt can form sodium acetate (NaOAc) and sodium formate (HCO₂Na) during storage in air.[1]

The actual charging conditions for the severely degraded sample can therefore be represented as follows:

Nominal condition: NaOEt, 3.0 equiv

↓

Actual composition: a small amount of residual NaOEt + other sodium salts formed through decomposition in air

This means that reagent degradation simultaneously changes:

① the actual NaOEt content;

② the identities and proportions of the oxygen-containing anions introduced into the reaction system.

NaOAc, HCO₂Na, and NaOEt differ in basicity, dissolution behavior, and chemical reactivity. Wethman et al. did not establish whether these salts further affected 9-BBN, palladium catalytic intermediates, or other reaction processes.[1] The 58% result corresponds simultaneously to a decrease in NaOEt content and the introduction of degradant salts into the reaction system, but the respective contributions of these two compositional changes to the yield have not been separated.

4. Why Do Degraded Samples Sometimes Increase the Yield and Sometimes Decrease It?

Sodium alkoxide degradation both lowers the content of the original base and introduces new sodium salts. The three model reactions examined in the original study responded differently to these compositional changes.[1]

The different effects of sodium alkoxide compositional changes in the three cross-coupling systems are shown below: Data source: [1].

In the first system, severely degraded NaOEt gave a higher yield than high-purity NaOEt at the nominal loading of 3.0 equiv. In the other two nickel-catalyzed systems, further degradation of NaOMe was accompanied by a pronounced decline in reaction performance.

This cross-system comparison shows that the “extent of degradation” alone cannot predict the direction of the change in reaction yield.

Different reactions require different base compositions and base loadings. One system may be sensitive to an excessively high sodium alkoxide concentration, whereas another may require a sufficient concentration of NaOMe or may be more sensitive to the salts formed after degradation.

The authors also examined alternative conditions that avoided compositional changes in solid primary sodium alkoxides. In the Pd-catalyzed alkyl-9-BBN/aryl chloride coupling, 1.5 equiv K₃PO₄ with 5 equiv H₂O afforded a 99% ¹H NMR yield and a 96% isolated yield. In the Ni-catalyzed alkene 1,2-diarylation, 2.0 equiv NaOH afforded a 91% ¹H NMR yield and a 91% isolated yield.[1] These results show that eliminating the compositional variable associated with unstable sodium alkoxides can improve reaction reproducibility, although the alternative base must be determined separately for each reaction system.

5. Why Do CO₂ and the Solid–Gas Interface Cause the Reagent Composition to Change Continuously?

The authors used NaOMe as a model to investigate the degradation of sodium alkoxides in air and confirmed that carbon dioxide (CO₂) in air is a key reactant.[1]

After fresh NaOMe had been exposed to air for five days, a new major species, sodium methyl carbonate (SMC), was detected, together with a small amount of sodium formate. Sodium methyl carbonate (MeOCO₂Na) can form through the direct reaction of NaOMe with CO₂:[1]

NaOMe + CO₂ → MeOCO₂Na

Sodium methyl carbonate can subsequently undergo further transformations, with the final products depending on the CO₂ concentration, atmospheric composition, and sample-handling conditions. Sodium formate and sodium carbonate were subsequently detected in degraded NaOMe that had been stored for extended periods.[1]

Raman spectroscopy further showed that this process begins at the solid–air interface. Spectral changes could be observed within several hours after NaOMe was exposed to air. Degradation was more pronounced near the air interface, whereas the interior of the solid remained less degraded before air penetrated deeper into the sample. No appreciable degradation was detected over several days under dry nitrogen.[1]

Degraded NaOMe and NaOEt may retain the appearance of white or off-white, free-flowing solids. Reagent composition therefore cannot be determined from appearance alone. On this basis, the authors recommended storing solid NaOMe and NaOEt under an inert atmosphere. Before quantitative comparisons of reaction conditions are made, the sample composition can first be confirmed using appropriate analytical methods.[1]

The composition of a bottle of sodium alkoxide depends not only on the production batch but also on its subsequent history of exposure to air and storage. The same label, weighing method, and nominal equivalents do not guarantee that a reagent used at different times will have the same composition.

6. How Can the Decrease in “Effective NaOEt” Be Distinguished from the Effects of Degradants?

The results of Wethman et al. demonstrated that changes in sodium alkoxide composition can markedly alter cross-coupling outcomes, but they do not answer a more specific mechanistic question:

How much of the change in reactivity produced by a degraded sample arises from the decrease in NaOEt itself, and how much arises from newly formed salts?

These two factors can be investigated separately through composition-reconstruction experiments.

6.1 Establishing the Relationship Between Actual NaOEt Loading and Reactivity

First, NaOEt with a confirmed assay should be used, and only the actual molar amount of NaOEt should be varied systematically while all other reaction conditions are kept unchanged.

For example, a range extending below 1.5 equiv, through 1.5 equiv, and up to 3.0 equiv could be examined to determine the complete changes in conversion and yield.

This experiment could determine:

① whether the reaction has a suitable NaOEt concentration range;

② where the sharp deactivation observed upon increasing the loading from 1.5 to 3.0 equiv begins;

③ whether the result obtained with the severely degraded sample can be explained primarily by the decrease in the actual NaOEt content.

If lowering the amount of pure NaOEt reproduces a reaction outcome similar to that obtained with the degraded sample, this would provide strong support for the contribution of “decreased effective NaOEt.”

6.2 Reconstructing the Degradant Salt Composition at Constant NaOEt Loading

The actual molar amount of NaOEt should then be held constant while the following species confirmed in the NaOEt degradation study are added separately:

· NaOAc;

· HCO₂Na;

· a mixture of the two with a defined composition.

Comparing the reaction rate, conversion, and yield before and after these additions could further determine whether the degradation products exert independent effects.

The experimental logic can be represented as follows:

Variation in the amount of pure NaOEt → Determine the sodium alkoxide concentration effect

Constant NaOEt + a single degradant salt → Determine the effect of the specific salt

Constant NaOEt + a reconstructed salt mixture → Determine the combined effect of the actual degradant composition

If adding a particular degradant salt at a constant molar amount of NaOEt still produces a pronounced change in the reaction, the salt does not act solely by “lowering the purity of NaOEt.” Subsequent studies of changes in the organoboron species and palladium or nickel species under conditions showing a pronounced effect would then be required to determine which reaction step is affected.

7. Classification Tables of Chemicals Related to Solid-State Sodium Alkoxide Decomposition, Composition Analysis, and Cross-Coupling Reaction Optimization

Table 1. Core Alkoxide Salts and Structural References

Table 2. Sodium Alkoxide Degradation Products, Analytical References, and Alternative Bases

Table 3. Substrates, Catalysts, and Ligands for the Cross-Coupling Model Reactions

Note: The products listed above are representative Aladdin products relevant to scientific research. Their specific uses should be determined according to the product specifications, batch COA, and the intended reaction or evaluation system. Additional information on product specifications, grades, and COAs can be found on the Aladdin website by searching for the “product name/CAS/catalog number.”

References

[1] Wethman, R.; Derosa, J.; Tran, V. T.; Kang, T.; Apolinar, O.; Abraham, A.; Kleinmans, R.; Wisniewski, S. R.; Coombs, J. R.; Engle, K. M. An Under-Appreciated Source of Reproducibility Issues in Cross-Coupling: Solid-State Decomposition of Primary Sodium Alkoxides in Air. ACS Catal. 2021, 11, 502–508. DOI: 10.1021/acscatal.0c04124.

[2] Amatore, C.; Jutand, A.; Le Duc, G. Kinetic Data for the Transmetalation/Reductive Elimination in Palladium-Catalyzed Suzuki–Miyaura Reactions: Unexpected Triple Role of Hydroxide Ions Used as Base. Chem. Eur. J. 2011, 17, 2492–2503. DOI: 10.1002/chem.201001911.

[3] Carrow, B. P.; Hartwig, J. F. Distinguishing Between Pathways for Transmetalation in Suzuki–Miyaura Reactions. J. Am. Chem. Soc. 2011, 133, 2116–2119. DOI: 10.1021/ja1108326.

[4] Lennox, A. J. J.; Lloyd-Jones, G. C. Selection of Boron Reagents for Suzuki–Miyaura Coupling. Chem. Soc. Rev. 2014, 43, 412–443. DOI: 10.1039/C3CS60197H.

For more related articles, see below:

Low-Pd Coupling of Aryl/Heteroaryl Chlorides (Ar–Cl): How to Make It Robust—Buchwald Ligand/Precatalyst Initiation Consistency and Scale-Up Reproducibility (with Controls and Selection Tables)

Sulfonyl Hydrazide-Mediated Introduction of Small C(sp³) Fragments: Nickel-Catalyzed Cross-Coupling, Medicinal Chemistry Applications, and Reagent Selection

Make “Aryl Chlorides + Low Pd Loading + Scale-Up Reproducibility” Reliable: The Initiation and Durability Logic of Pd–NHC (Palladium–N-Heterocyclic Carbene) Cross-Coupling (with Selection Navigation and Product Tables)

Experimental Decision-Making for Cross-Coupling Reactions: Target Bond Type, Substrate Combination, and Catalytic System Selection

Haloheterocycles and Cross-Coupling: A Research-Oriented Selection Framework from Substrate Identification to Bond-Forming Routes (Including Product Navigation and Tables 1–5)

Choosing Boron Sources to Make Reactions Robust: How Boronic Acids, Boronate Esters, BF₃K Salts, and MIDA Improve Suzuki–Miyaura Start-Up and Scale-Up Reproducibility (with Product Tables 1–5)

How to Make the Suzuki–Miyaura Reaction Robust: Pinpoint the Bottleneck and Lock in a Reproducible Operating Window (with Selection Navigation and Product Tables 1–5)