Burgess reagent: what it dehydrates, what else it does and why a reaction fails
Share
Burgess reagent: what it dehydrates, what else it does and why a reaction fails
How the reagent works, which product each type of substrate gives, worked examples with yields, and how to check a bottle by 1H NMR
The Burgess reagent dehydrates secondary and tertiary alcohols to alkenes under mild conditions. It also converts primary amides and aldoximes into nitriles, formamides into isocyanides and β-hydroxy amides into oxazolines. Other reactions give sulfamidates, sulfamides, glycosylamines, guanidines and carbonyl compounds. The reagent is sensitive to moisture. Yields depend on the batch of reagent.[1]
How the examples are given. Each example comes from one publication. The yield is the yield reported there. One example does not predict the yield for another substrate. For a series of substrates, we give the range. Four examples include all amounts and conditions. We chose them because each is representative of its reaction.
Terms. Sulfamate ester: the alcohol after reaction with the reagent, R–O–SO2–N−–CO2Me, as the triethylammonium salt. Syn elimination: the hydrogen and the leaving group are removed from the same side of the C–C bond. Equiv: moles of reagent per mole of substrate. THF: tetrahydrofuran. DMSO: dimethyl sulfoxide.
Overview of the reactions
| Substrate | Product | Section | |
|---|---|---|---|
|
Secondary or tertiary alcohol
|
→ |
Alkene, by syn elimination
|
2 |
|
Primary alcohol
|
→ |
Methyl N-alkylcarbamate (urethane)
|
3 |
|
Primary amide
|
→ |
Nitrile
|
4 |
|
Aldoxime
|
→ |
Nitrile
|
4 |
|
Formamide
|
→ |
Isocyanide
|
4 |
|
Primary nitroalkane
|
→ |
Nitrile oxide
|
4 |
|
N-Trityl-N′-aryl urea
|
→ |
Carbodiimide
|
4 |
|
β-Hydroxy amide (serine, threonine)
|
→ |
Oxazoline, with inversion at the β-carbon
|
5 |
|
β-Hydroxy thioamide
|
→ |
Thiazoline
|
5 |
|
2-Acylamino ketone
|
→ |
Oxazole
|
5 |
|
1,2-Diacylhydrazine
|
→ |
1,3,4-Oxadiazole
|
5 |
|
N′-(Pyridin-2-yl)hydrazide
|
→ |
[1,2,4]Triazolo[4,3-a]pyridine
|
5 |
|
1,2-Diol
|
→ |
Cyclic sulfamidate, then β-amino alcohol
|
6 |
|
Epoxide
|
→ |
Cyclic sulfamidate (five- or seven-membered ring)
|
6 |
|
β-Amino alcohol
|
→ |
Cyclic sulfamide, with inversion at the carbinol carbon
|
6 |
|
Protected sugar with a free 1-OH
|
→ |
Glycosylamine, as the methyl carbamate
|
6 |
|
Oxycodone or oxymorphone N-oxide
|
→ |
Oxazolidine, then the N-demethylated amine
|
6 |
|
Thiourea with a carbamoyl group
|
→ |
Carbamate-protected guanidine
|
6 |
|
Indole, with a Lewis acid
|
→ |
Indole-3-sulfonyl carbamate
|
6 |
|
Alcohol, in DMSO
|
→ |
Aldehyde or ketone
|
6 |
|
H2O
Water
|
→ |
Et3NH+
Hydrolysis product
|
7, 8 |
The miniatures show the reacting group. R is any substituent. For the opioid N-oxides, the rest of the morphinan skeleton is not drawn.
1. What the reagent is
The Burgess reagent is methyl N-(triethylammoniumsulfonyl)carbamate, C8H18N2O4S. It is an inner salt. The positive charge is on the triethylammonium nitrogen. The negative charge is on the carbamate nitrogen (Figure 1, left). The reagent dissolves in most common organic solvents, including non-polar ones.[1][2] The review gives a melting point of 70–72 °C.[1] The crystal-structure paper gives 67.2–69.5 °C.[3]
In the crystal, the N–S distance between triethylamine and sulfur is 1.90 Å. The authors call this bond surprisingly long. The geometry at sulfur is between planar and tetrahedral.[3] The solid is therefore close to triethylamine bound to a sulfonylimine, O2S=N–CO2Me (Figure 1, right). The authors propose that the reagent first loses triethylamine. The alcohol then adds to the sulfonylimine.[3]
Figure 1. The Burgess reagent drawn as an inner salt (left) and as found in the crystal (right).[3] The triethylamine nitrogen is 1.90 Å from the sulfur of the sulfonylimine. The dashed gold line shows this long N···S bond.
The alcohol becomes an alkyl sulfamate ester, R–O–SO2–N−–CO2Me, as its triethylammonium salt.[1] The sulfamate is a leaving group. The product depends on the groups near it:
- If a β-hydrogen leaves with the sulfamate, an alkene forms (section 2).
- A nucleophile can displace the sulfamate. With primary alcohols, the nucleophile is a nitrogen atom (section 3). In other substrates, it is an amide oxygen, a sulfamide nitrogen or a hydroxyl group in the same molecule (sections 5 and 6).
Water reacts like an alcohol. The product is triethylammonium N-(methoxycarbonyl)sulfaminate. Gentle warming under vacuum does not turn this salt back into the reagent.[3] Each molecule of water consumes one molecule of reagent irreversibly.
The reagent converts an OH group into a leaving group under mild conditions. Water hydrolyses the reagent irreversibly.
Product: Burgess reagent
2. How an alcohol becomes an alkene
A secondary or tertiary alcohol loses water in two stages (Figure 2):[1][4]
- Stage 1: ester formation. The sulfamate ester forms at or below 30 °C in hydrocarbon solvents.
- Stage 2: elimination. On heating, the ester gives the alkene. The review describes this stage as a pyrolysis below 100 °C. Microwave heating can also be used.
The two stages can be separated by temperature. In a 2025 study, a secondary alcohol formed its sulfamate in refluxing THF. The sulfamate did not eliminate at this temperature. The crude 1H NMR spectrum showed the sulfamate CH at δ 4.71 and no alkene signals. The sulfamate could not be isolated. In refluxing toluene, the same alcohol gave the alkene in 25% (single example).[5]
Figure 2. The two stages of the dehydration.[1] By mass balance, the by-product is the same triethylammonium salt that water gives.
The rate-determining step is the ionisation of the C–O bond. This gives an ion pair. The anion then quickly removes a β-hydrogen from the same side of the molecule. The elimination is therefore syn. The review calls this a variant of the Ei mechanism (intramolecular elimination).[1]
A deuterium experiment tests this (Figure 3). The erythro and threo isomers of 2-deuterio-1,2-diphenylethyl sulfamate both give (E)-stilbene. The erythro isomer gives stilbene with deuterium. The threo isomer gives stilbene without deuterium. This is the result expected for a syn elimination.[1][4]
Figure 3. The deuterium test for syn elimination.[1] In each isomer, the conformation that leads to (E)-stilbene places a different atom on the same side as the leaving group. In the erythro isomer, this atom is H. In the threo isomer, it is D. An anti elimination would give the opposite result.
In most cases, the more substituted alkene forms (Zaitsev’s rule).[1] Some substrates behave differently:
- One steroidal alcohol gave anti elimination.[1][6]
- Tertiary alcohols react under milder conditions. They can rearrange before they eliminate.[1]
- With an allylic sulfamate, the product depended on the conditions. Its sodium salt gave the rearranged urethane (94%) when heated as a solid at 80 °C. In triglyme solution, the same salt gave the diene (73%).[1][4]
Table 1 shows the yields for one series of steroid alcohols. The review gives the products and yields, but not the conditions.[1][7]
Table 1. Single examples: steroid alcohols
| Alcohol | Alkene | Yield |
|---|---|---|
| 3α-Hydroxy-5α-androstan-17-one (androsterone) | 5α-Androst-2-en-17-one | 75% |
| 3β-Hydroxy-5α-androstan-17-one (epiandrosterone) | 5α-Androst-2-en-17-one | 52% |
| 5α-Cholestan-3β-ol | 5α-Cholest-2-ene | 63% |
| Cholesterol | Cholesta-3,5-diene | 27% |
| An 11β-hydroxy steroid (pregnane series) | The Δ9(11)-alkene | 96% |
Data as tabulated in the review.[1] The C-3 epimers androsterone and epiandrosterone give the same alkene in different yields.
Secondary and tertiary alcohols give alkenes by syn elimination. Tertiary and allylic alcohols can also give rearranged products.
3. Which alcohol gives which product
The type of alcohol determines the main product (Table 2).
Table 2. Type of alcohol and main product
| Alcohol | Main product | Note |
|---|---|---|
| Primary | Methyl N-alkylcarbamate (urethane) | Substitution, not elimination (see below) |
| Secondary | Alkene | Syn elimination, mostly to the more substituted alkene |
| Tertiary | Alkene | Reacts under milder conditions. Rearrangement is possible. |
| Allylic | Diene or rearranged urethane | The product depends on the conditions (section 2). |
| β-Hydroxy amide or thioamide | Oxazoline or thiazoline | Ring closure (section 5) |
| 1,2-Diol | Cyclic sulfamidate | Section 6 |
A primary alcohol does not give a terminal alkene. Instead, a nitrogen atom displaces the sulfamate in an SN2 reaction. The product is a methyl N-alkylcarbamate (a urethane).[1][8] The Organic Syntheses procedure is limited to primary alcohols. Secondary and tertiary alcohols are dehydrated under the same conditions.[8]
Hydrolysis of the urethane gives the amine. The two steps convert a primary alcohol into a primary amine.[8] If the primary carbon is too hindered for substitution, the review reports rearrangement and elimination at higher temperature.[1]
Figure 4 gives the full Organic Syntheses procedure. Organic Syntheses procedures are checked in a second laboratory. The reaction uses no solvent and a small excess of reagent.[8]
Figure 4. Full example: methyl N-hexylcarbamate from 1-hexanol.[8] 1-Hexanol (6.0 g, 58 mmol) and the reagent (14.8 g, 62.2 mmol, 1.07 equiv) are heated without solvent in an oil bath at 95 °C for 1 h. The flask is fitted with a calcium chloride drying tube. The start of the reaction is mildly exothermic. Sometimes there is an induction period of about 5 minutes. The work-up uses water and dichloromethane, a wash with dilute hydrochloric acid, trituration with ether and short-path distillation. The yield is 4.8–4.9 g (51–52%), b.p. 59–60 °C at 0.08 mmHg.
Primary alcohols give carbamates. Only secondary and tertiary alcohols give alkenes.
4. Dehydrations at nitrogen
The reagent also dehydrates five classes of nitrogen compound (Figure 5).
Figure 5. Five dehydrations at nitrogen.[1][9]
- Primary amides to nitriles. The reaction is chemoselective. Other functional groups need no protection.[2][10] In one series, six amides gave 82–92%. A secondary alcohol in the substrate was not affected (82%). An epoxide was not opened (85%). No racemisation (92%) or epimerisation (88% and 91%) occurred.[1][10] The review suggests that amides react with the reagent faster than secondary alcohols.[1]
- Aldoximes to nitriles. Aliphatic and aromatic aldoximes give nitriles in excellent yields. The conditions are 1.5 equiv of the reagent in THF with heating. A version of the reagent bound to polyethylene glycol (PEG) also gives nitriles.[11] Acid-sensitive groups are not affected.[2]
- Formamides to isocyanides. The yields are high. The method works well for substrates with trimethylsilyl ethers. These groups are sensitive to halide.[12] The reagent contains no halogen.
- Primary nitroalkanes to nitrile oxides. In a comparison of reagents, the Burgess reagent and DAST (diethylaminosulfur trifluoride) were the most effective.[1][13]
- Ureas to carbodiimides. N-Trityl-N′-aryl ureas give carbodiimides. These were converted into N,N′-diarylguanidines.[9]
The review gives no mechanism for these reactions. Figure 5 shows only substrates and products.[1]
A full example on a 100-gram scale
Nirmatrelvir is a SARS-CoV-2 main-protease inhibitor. In its synthesis published in 2021, the last step is a Burgess dehydration of a primary amide. The supporting information describes this step in detail (Figure 6).[14]
The step was run in two parallel batches. In each batch, the reagent (69.3 g, 276 mmol, 2.5 equiv) was added to the primary amide (61 g, 111 mmol) in dichloromethane (550 mL). The mixture was stirred at 25 °C for 1 h. The work-up included aqueous washes, filtration through silica and crystallisation from ethyl acetate and methyl tert-butyl ether (MTBE). The two batches together gave 104 g of nirmatrelvir as the MTBE solvate. The yield was 75% after correction for potency.[14]
Figure 6. Full example: the last step of a published nirmatrelvir synthesis.[14] The gold circle marks the carbon that changes from amide to nitrile.
The substrate has six stereocentres. It has four other amide groups, including a lactam and a trifluoroacetamide. Only the primary amide reacted.[14] Our article on coupling reagents beyond the amide bond lists other reagents for this step.
The amide dehydration is chemoselective. In these examples, a secondary alcohol, an epoxide, other amide groups and the stereocentres were not affected.
5. Ring closures
If the substrate has a nucleophile in a suitable position, the reaction forms a ring (Figures 7 and 8).
Oxazolines from serine and threonine amides
β-Hydroxy amides cyclise to oxazolines (4,5-dihydrooxazoles) under mild, neutral conditions. Examples are N-acyl serine and threonine esters.[1][15] The cyclisation inverts the configuration at the carbon that had the OH group (the β-carbon).[1] This inversion was used to convert threonine in peptides into allo-threonine.[16]
The review gives no mechanism. The inversion fits an SN2 displacement of the sulfamate by the amide oxygen. According to the review, no β-lactam, dehydroamino acid or aziridine side products form.[1]
Figure 7 gives a full example from a patent. N-Benzoyl-L-threonine methyl ester (1.07 g, 4.5 mmol) and the reagent (1.18 g, 5.0 mmol, 1.1 equiv) were heated in THF (10 mL) at 70–80 °C for 2 h. Water was added, and the product was extracted with dichloromethane. After drying, the oxazoline was an oil (0.890 g). It was used without purification.[17] This mass is 90% of theory. The patent gives 95%.
The product is the (4S,5S)-oxazoline. Its 5-methyl and 4-ester groups are cis.[17] L-Threonine is (2S,3R). The carbon that had the OH group is C-3 in threonine and C-5 in the oxazoline. Its Cahn–Ingold–Prelog (CIP) priorities have the same order in both compounds. The change from R to S therefore shows inversion.
Figure 7. Full example: the cis-oxazoline from N-benzoyl-L-threonine methyl ester.[17] The gold circles mark C-3 of threonine. This carbon had the OH group. It becomes C-5 of the oxazoline, with inverted configuration. 1H NMR (CDCl3, 400 MHz): δ 5.07 (dq, H-5), 4.98 (d, H-4), 3.78 (s, OMe), 1.38 (d, 5-Me).
A second patent gives three more examples. In the first, N-[2-(benzyloxy)benzoyl]-L-threonine benzyl ester was treated with 1.5 equiv of the reagent in THF. The mixture was kept at room temperature overnight and then heated at reflux for 1 h. After chromatography, the oxazoline was obtained in 76%. The corresponding methyl ester gave 80.5% under the same conditions. The L-serine benzyl ester analogue gave 66% with 1.1 equiv of the reagent.[18]
Thiazolines
β-Hydroxy thioamides cyclise to thiazolines.[1][19] One study compared four reagents on one peptide β-hydroxy thioamide (Table 3). Only the Burgess reagent gave a diastereomer ratio above 97 : 3. The reaction took 10 minutes.[1][19]
Table 3. One substrate, four reagents: thiazoline formation
| Reagent and conditions | Yield | Diastereomer ratio |
|---|---|---|
| TsCl, Et3N, CH2Cl2, 42 °C, 1 h | 40% | 1 : 1 |
| SOCl2, 0 °C, 2 h, then pyridine, THF | 49% | 1 : 1 |
| Ph3P, DIAD, CH2Cl2, −78 to 22 °C | 80% | 78 : 22 |
| Burgess reagent, THF, 65 °C, 10 min | 96% | > 97 : 3 |
Data as tabulated in the review.[1] TsCl: p-toluenesulfonyl chloride. DIAD: diisopropyl azodicarboxylate.
Oxazoles and 1,3,4-oxadiazoles
2-Acylamino ketones give oxazoles. According to the review, these cyclisations need single-mode microwave heating.[1]
1,2-Diacylhydrazines give 1,3,4-oxadiazoles. The parent reagent gave moderate to poor yields. A polymer-supported reagent with single-mode microwave heating gave 70–96% in one series. The data below are from the review:[1][20]
- 2,5-diphenyl-1,3,4-oxadiazole: 96%
- 2-methyl-5-(2-nitrophenyl)-1,3,4-oxadiazole: 95%
- 2-(3-methoxyphenyl)-5-methyl-1,3,4-oxadiazole: 95%
- 2-(2-furyl)-5-phenyl-1,3,4-oxadiazole: 86%
- 2-(2-chlorophenyl)-5-methyl-1,3,4-oxadiazole: 70%
Fused pyridines and a bicyclic ether
Cyclodehydration also gives pyrrolo-, imidazo- and triazolopyridines with a nitrogen at the ring junction. The conditions are mild and neutral. Acid-sensitive groups are not affected.[21] The same paper reports a side reaction. A second equivalent of the reagent adds to the electron-rich pyrrolo- or imidazopyridine product. The result is a sulfonylcarbamate adduct.[21]
A patent describes two triazolopyridine syntheses from N′-(pyridin-2-yl)hydrazides. Both were run in acetonitrile–dioxane in a sealed vial. The first used 2.1 equiv of the reagent at 85 °C for 18 h. It gave 85% after chromatography. The second used 2.5 equiv at 110 °C for 4 h. It gave 45%. In the second synthesis, the product precipitated and was collected by filtration.[22]
A 2025 paper uses the reagent for a two-step ring closure. Each substrate has a free secondary alcohol. Its cyclohexane ring has two hydroxyl groups protected as tert-butyldimethylsilyl (TBS) ethers. The reagent converted the secondary alcohol into its sulfamate in refluxing THF. HCl in methanol at room temperature then removed the TBS groups. The ring hydroxyl in the δ-position displaced the sulfamate. The product was an oxabicyclo[3.2.1]octan-3-ol.[5]
The first example in the paper gave the bicyclic alcohol in 74% over the two steps. In the other examples, the yields were 12–79%. Some reactions gave two epimers. A route through the tosylate and sodium hydride gave mainly alkenes.[5]
Figure 8. Four other ring closures, shown as substrate classes and products.[1][20][22]
The oxazoline cyclisation inverts the configuration at the carbon that had the OH group, so L-threonine gives the cis-oxazoline.
Product: Burgess reagent
6. Reactions other than dehydration
Sulfamidates from 1,2-diols and epoxides
Enantiopure 1,2-diols give cyclic sulfamidates (Figure 9, top). These are five-membered rings with an O–SO2–N unit. The nitrogen has a methoxycarbonyl group. The nitrogen replaces one hydroxyl group with inversion of configuration. The other carbon keeps its oxygen. Aqueous HCl (4 M) in dioxane then gives the β-amino alcohols as methyl carbamates.[23]
Epoxides and epoxy alcohols also give sulfamidates.[24][25] Epoxides give a five-membered ring, a seven-membered ring or both (Figure 9, bottom). The seven-membered ring contains the carbamate group of the reagent. Aryl epoxides and styrene diol give mainly the seven-membered ring.[25][26] Its proportion depends on the electron density of the aromatic ring.[26]
A chiral version of the reagent has a menthyl carbamate in place of the methyl carbamate. With epoxides, it gave diastereomeric sulfamidates. These were converted into both enantiomers of protected trans-amino alcohols (84–98% enantiomeric excess).[26]
Figure 9. Sulfamidates. Top: an enantiopure 1,2-diol gives a cyclic sulfamidate. The structures follow the graphical abstract of the paper.[23] The gold circles mark the carbon that receives the nitrogen. Its configuration is inverted. Bottom: epoxides give a five-membered ring, a seven-membered ring or both. The structures follow Scheme 1 of the paper on the chiral reagent.[26]
Cyclic sulfamides from β-amino alcohols
β-Amino alcohols with a secondary amine give cyclic sulfamides in one step (Figure 10, top). The amine becomes one nitrogen of the sulfamide. The reagent supplies the other nitrogen, with its methoxycarbonyl group. This nitrogen displaces the activated hydroxyl group in an SN2 reaction. The configuration at the carbon that had the OH group is inverted.[24][27]
In the patent, eight β-amino alcohols gave 69–92%. The conditions were 2.5 equiv of the reagent in refluxing THF for 8 h. (R)-Propranolol gave the (S)-sulfamide in 89% after 21 h (Figure 10, bottom). 2-(2-Aminophenyl)ethanol gave a seven-membered ring in 90% after 2 h. 2-Aminobenzyl alcohol gave 45% at 0 °C and then at room temperature. Primary aliphatic amino alcohols gave “rather disappointing” yields. Decomposition and rearranged side products formed. According to the patent, their nitrogen is less nucleophilic, so the reaction needs more time.[27]
Figure 10. Cyclic sulfamides from β-amino alcohols, drawn from the patent figures.[27] Top: the general reaction. Bottom: (R)-propranolol. The gold circles mark the carbon that had the OH group. Its configuration is inverted.
Glycosylamines
Protected sugars with a free anomeric hydroxyl group give glycosylamines (Figure 11). The amine is protected as a methyl carbamate.[28] If all other hydroxyl groups are protected, the product is the β-glycosylamine. If the 2-hydroxyl group is free, a 1,2-fused cyclic sulfamidate forms. Fluoride, nitrogen and sulfur nucleophiles open this ring at C-2 with inversion. The products are α-glycosylamines. Both reactions use THF–CH2Cl2 with heating for 6 h.[28]
Figure 11. Glycosylamines, drawn from the visual abstract of the paper.[28] R is a protecting group. Nu is a fluoride, nitrogen or sulfur nucleophile. The wavy bond marks a mixture of anomers. The carbamate on nitrogen comes from the reagent.
N-Demethylation of opioids through oxazolidines
The N-oxides of oxycodone and of O-acylated oxymorphone give oxazolidines in high yields. In the oxazolidine, the former N-methyl carbon is bonded to the nitrogen and to the 14-oxygen. Hydrolysis removes this carbon. The product is the N-demethylated amine. From oxymorphone, this is noroxymorphone. Noroxymorphone is the starting material for naltrexone, naloxone and nalbuphone.[29]
Figure 12 gives a full example from a patent. Oxymorphone (600 mg, 1.99 mmol) was acetylated at the phenol. It was then oxidised to the N-oxide with meta-chloroperoxybenzoic acid (mCPBA). The reagent (593 mg, 2.49 mmol, 1.25 equiv) was added at −20 °C. All three steps were run in one pot. The crude yield of the oxazolidine was 93% (purity 92–95%). After crystallisation, the yield was 76%. A repeat on a 1 g scale gave 78%. Hydrolysis in an acetic acid–ammonia buffer at pH 9 and 50 °C gave noroxymorphone (82%). N-Alkylation then gave naltrexone (87%), naloxone (84%) and nalbuphone (66%).[30]
Figure 12. Full example: N-demethylation of oxymorphone through the oxazolidine, from the patent examples.[30] The one-pot sequence goes through the 3-O-acetyl N-oxide. The gold circles mark the carbon of the N-methyl group. In the oxazolidine, this carbon is bonded to the nitrogen and to the 14-oxygen. Hydrolysis removes it.
Guanidines, indoles and oxidation in DMSO
- Guanidines from thioureas. Thioureas from primary amines and carbamoyl isothiocyanates give guanidines (Figure 13, top). The reagent supplies the third nitrogen with its methoxycarbonyl group. Burgess-type reagents with other carbamate groups give guanidines with two different protecting groups. Removal of one group gives N-monoprotected guanidines. The paper reports 18 examples in up to 86% yield.[31]
- Sulfonylation of indoles. With a Lewis acid, the reagent sulfonylates indoles at C-3 (Figure 13, middle). The products are indole-3-sulfonyl carbamates and sulfonamides. The method was developed for an intermediate of a kinase inhibitor.[32] The side reaction in section 5 is the same type of reaction.
- Oxidation of alcohols in DMSO. In DMSO, the reagent oxidises primary and secondary alcohols to aldehydes and ketones. The reaction takes 5 minutes at room temperature (Figure 13, bottom). It can be combined with a Wittig olefination. The authors propose a mechanism like that of the Pfitzner–Moffatt and Swern oxidations.[33]
Figure 13. Top: guanidines from thioureas, drawn from the visual abstract of the paper.[31] Middle: C-3 sulfonylation of indole, drawn from the title and abstract of the paper only.[32] Bottom: oxidation in DMSO, with the conditions from the graphical abstract.[33]
The reagent gives sulfamidates, sulfamides, glycosylamines and guanidines. It N-demethylates opioid N-oxides, sulfonylates indoles and, in DMSO, oxidises alcohols.
7. Why a Burgess reaction fails
A poor result has five documented causes. They are water in the reaction, a hydrolysed bottle, the type of substrate, the temperature and the amount of reagent (Table 4).
Table 4. Observations, likely causes and what to do
| What you see | Likely cause | What to do |
|---|---|---|
| Lower yield with an older bottle, or yields that vary between batches | Part of the reagent is hydrolysed. The review notes that yields depend on the batch.[1] | Check the bottle by 1H NMR (section 8). |
| Part of the solid does not dissolve in benzene | The insoluble part is probably the hydrolysis product. It dissolves quickly in water. The intact reagent dissolves easily in benzene.[3] | Use a fresh bottle, or check this one by 1H NMR. |
| Incomplete conversion with a wet solvent or substrate | Each molecule of water consumes one molecule of reagent (section 1). | Dry the solvent, the substrate and the glassware. |
| A primary alcohol gives a carbamate instead of the alkene | This is the normal reaction of primary alcohols (section 3).[1][8] | Use another method for terminal alkenes. |
| A tertiary or allylic alcohol gives a rearranged product | Tertiary sulfamates react under milder conditions. The cation of the ion pair can rearrange before elimination (section 2). Allylic sulfamates gave a diene or a rearranged urethane, depending on the conditions.[1] | For an allylic alcohol, choose the conditions for the product you need (section 2). |
| The alkene is not the expected regioisomer | The elimination is syn. Only a β-hydrogen on the same side as the C–O bond can be removed.[1] | Check which β-hydrogens can be syn to the leaving group, especially in rings. |
| An electron-rich product reacts further | Excess reagent. A second equivalent added to pyrrolo- and imidazopyridine products and gave sulfonylcarbamate adducts.[21] With a Lewis acid, the reagent also sulfonylates indoles at C-3.[32] | Avoid a large excess. Monitor the reaction and stop it when the substrate is consumed. |
| An epoxide elsewhere in the substrate reacts | Epoxides give sulfamidates with the reagent (section 6).[25] In one amide dehydration, an epoxide was not affected.[1][10] | Run a test reaction. Keep the excess of reagent and the temperature low. |
| Elimination instead of substitution, or the reverse | Temperature. A secondary sulfamate was stable in refluxing THF. It eliminated in refluxing toluene (section 2).[5] | Choose the solvent and temperature for the step you need. |
| The oxazoline has the unexpected relative configuration | The cyclisation inverts the configuration at the carbon that had the OH group. L-Threonine gives the cis-oxazoline.[1][17] | For the trans-oxazoline, start from the other configuration at that carbon, for example allo-threonine. |
| Low yield of a 1,3,4-oxadiazole | The parent reagent gave moderate to poor yields.[1] | Use a polymer-supported reagent with single-mode microwave heating (section 5).[20] |
Entries in the last column without a literature reference are general laboratory practice or our inference from the cause in the same row.
Check the bottle before an important reaction, because a hydrolysed reagent cannot be recovered. The other problems in Table 4 come from the substrate, the conditions or excess reagent.
8. Checking a bottle by 1H NMR
Hydrolysis changes the 1H NMR spectrum of the reagent. A spectrum in CDCl3 shows whether a bottle is intact. One study measured the reagent and its hydrolysis product under the same conditions (Table 5, Figure 14).[3]
Table 5. NMR in CDCl3 at 600 MHz
| Signal | Reagent | Hydrolysis product |
|---|---|---|
| CH3 of the ethyl groups, t, 9H | 1.41 | 1.36 |
| NCH2, q, 6H | 3.47 | 3.22 |
| OCH3, s, 3H | 3.70 | 3.68 |
| SO2NHCO, br s, 1H | — | 7.31 |
| Et3NH+, br s, 1H | — | 9.14 |
| 13C (ppm) | 9.38, 50.43, 53.26, 158.2 | 8.47, 46.41, 52.26, 153.7 |
Chemical shifts in ppm. J = 7.2 Hz in the reagent and 6.6 Hz in the hydrolysis product, as reported.[3]
Figure 14. 1H NMR signals of the reagent (top) and of its hydrolysis product (bottom) in CDCl3, drawn from the published chemical shifts. It is not a measured spectrum.[3] The letters a–e show which protons give each signal. The heights are proportional to the number of protons.
- The NCH2 quartet is the most useful signal. It is at 3.47 ppm in the reagent and at 3.22 ppm in the hydrolysis product. Both quartets integrate for 6H. The ratio of their integrals gives the molar ratio of reagent to hydrolysis product. For example, integrals of 6.0 at 3.47 ppm and 1.5 at 3.22 ppm mean 80 mol% reagent and 20 mol% hydrolysis product. Our article on NMR yield and isomer ratios explains how the delay between pulses affects such integrals.
- The NH signals confirm hydrolysis. Two broad signals at 7.31 and 9.14 ppm appear only in the hydrolysis product. The signal at 7.31 ppm is close to the CHCl3 signal at 7.26 ppm. The quartet is the more reliable test.
- The methyl singlet is less useful. It moves only from 3.70 ppm in the reagent to 3.68 ppm in the hydrolysis product.
- Use dry CDCl3 and record the spectrum immediately. The authors suggest that the reagent may hydrolyse in chloroform solution.[3]
If you have no NMR spectrometer, NorrChemica measures 1H spectra at 400 MHz as an analytical service (Package I).
A quartet at 3.22 ppm shows that the bottle is partly hydrolysed.
9. Storage and handling
- The reagent is sensitive to moisture and to oxidation. The review recommends storage at low temperature.[1] The shelf life is limited.[34]
- Let the closed bottle reach room temperature before you open it. Otherwise, water condenses on the cold solid. Weigh the reagent quickly and close the bottle again. Use dry nitrogen or argon if possible.
- Warming under vacuum does not recover a hydrolysed reagent (section 1). Stronger heating chars it.[3]
- Polymer-supported versions, for example on PEG, have a longer shelf life.[1][34]
- Hazards: GHS07, Warning. H315: causes skin irritation. H319: causes serious eye irritation. H335: may cause respiratory irritation. The safety data sheet gives the full information.
Store the bottle tightly closed at −20 °C.
Product: Burgess reagent
Burgess reagent from NorrChemica
Packs from 1 g to 250 g, each with CoA and SDS. Shipped from Finland, with DDP delivery worldwide.
- Burgess reagent, ≥95%CAS 29684-56-8
- 1H NMR measurement, Package I400 MHz
References
- Khapli S, Dey S, Mal D. Burgess reagent in organic synthesis. J. Indian Inst. Sci. 2001, 81, 461–476 (review). PDF
- Santra S. Burgess reagent: from oblivion to renaissance in organic synthesis. Synlett 2009, 328–329 (spotlight). DOI
- Arduengo AJ, Uchiyama Y, Dixon DA, Vasiliu M. Crystal structure of Burgess inner salts and their hydrolyzed ammonium sulfaminates. Aust. J. Chem. 2019, 72, 867–873. DOI
- Burgess EM, Penton HR, Taylor EA. Thermal reactions of alkyl N-carbomethoxysulfamate esters. J. Org. Chem. 1973, 38, 26–31 (data as reported in ref. 1). DOI
- Elmansy MF, dos Remedios JRD, Silverman RB. Synthesis of oxabicyclo[3.2.1]octan-3-ol scaffold via Burgess reagent mediated cyclodehydration of δ-diols under acidic conditions. Org. Lett. 2025, 27, 640–644 (author manuscript in PMC). DOI
- J. Org. Chem. 1974, 39, 2124 (steroidal alcohol with anti elimination; as cited in ref. 1).
- J. Org. Chem. 1970, 35, 2594 (steroid alcohols; yields as tabulated in ref. 1).
- Burgess EM, Penton HR, Taylor EA, Williams WM. Conversion of primary alcohols to urethanes via the inner salt of methyl (carboxysulfamoyl)triethylammonium hydroxide: methyl n-hexylcarbamate. Org. Synth. 1977, 56, 40; Coll. Vol. 6, 788. DOI
- Barvian MR, Showalter HDH, Doherty AM. Preparation of N,N′-bis(aryl)guanidines from electron deficient amines via masked carbodiimides. Tetrahedron Lett. 1997, 38, 6799–6802.
- Claremon DA, Phillips BT. An efficient chemoselective synthesis of nitriles from primary amides. Tetrahedron Lett. 1988, 29, 2155–2158 (yields as tabulated in ref. 1). DOI
- Miller CP, Kaufman DH. Mild and efficient dehydration of oximes to nitriles mediated by the Burgess reagent. Synlett 2000, 1169–1171. DOI
- Creedon SM, Crowley HK, McCarthy DG. Dehydration of formamides using the Burgess reagent: a new route to isocyanides. J. Chem. Soc., Perkin Trans. 1 1998, 1015–1018. DOI
- Maugein N, Wagner A, Mioskowski C. Tetrahedron Lett. 1997, 38, 1547 (nitrile oxides from primary nitroalkanes; as reported in refs. 1 and 2).
- Owen DR, et al. An oral SARS-CoV-2 Mpro inhibitor clinical candidate for the treatment of COVID-19. Science 2021, 374, 1586–1593; Supplementary Materials, pp. 13 and 15. DOI
- Wipf P, Miller CP. A short, stereospecific synthesis of dihydrooxazoles from serine and threonine derivatives. Tetrahedron Lett. 1992, 33, 907–910. DOI
- Wipf P, Miller CP. Stereospecific synthesis of peptide analogs with allo-threonine and D-allo-threonine residues. J. Org. Chem. 1993, 58, 1575–1578. DOI
- US 6,632,954 B2 (divisional of US 6,498,256 B2), Phosphinite-oxazolines and metal complexes, Example A12b. Google Patents
- US 6,403,623 B1, Heterocyclic antituberculosis agents, Examples 1, 11 and 12. Google Patents
- Wipf P, Fritch PC. Tetrahedron Lett. 1994, 35, 5397–5400 (peptide thiazolines from β-hydroxythioamides; data as tabulated in ref. 1). DOI
- Brain CT, Paul JM, Loong Y, Oakley PJ. Tetrahedron Lett. 1999, 40, 3275–3278 (1,3,4-oxadiazoles with polymer-supported Burgess reagent under microwave heating; yields as tabulated in ref. 1).
- A synthesis of N-bridged 5,6-bicyclic pyridines via a mild cyclodehydration using the Burgess reagent and discovery of a novel carbamylsulfonylation reaction. Org. Lett. 2008, 10, 2897–2900. DOI
- US 9,738,642 B2, Triazolopyridine ether derivatives and their use in neurological and psychiatric disorders, Examples 3 (Step 3B) and 86 (Step 86B). Google Patents
- Nicolaou KC, Huang X, Snyder SA, Bheema Rao P, Bella M, Reddy MV. A novel regio- and stereoselective synthesis of sulfamidates from 1,2-diols using Burgess and related reagents: a facile entry into β-amino alcohols. Angew. Chem. Int. Ed. 2002, 41, 834–838. DOI
- Nicolaou KC, Snyder SA, Longbottom DA, Nalbandian AZ, Huang X. New uses for the Burgess reagent in chemical synthesis: methods for the facile and stereoselective formation of sulfamidates, glycosylamines, and sulfamides. Chem. Eur. J. 2004, 10, 5581–5606. DOI
- Rinner U, Adams DR, dos Santos ML, Abboud KA, Hudlicky T. New application of Burgess reagent in its reaction with epoxides. Synlett 2003, 1247–1252. DOI
- Leisch H, Saxon R, Sullivan B, Hudlicky T. Chiral version of the Burgess reagent and its reactions with epoxides. Synlett 2006, 445–449. DOI
- US 7,015,321 B2, Synthesis of non-symmetrical sulfamides using Burgess-type reagents. Google Patents
- Nicolaou KC, et al. A new method for the stereoselective synthesis of α- and β-glycosylamines using the Burgess reagent. J. Am. Chem. Soc. 2004, 126, 6234–6235. DOI
- Werner L, Wernerova M, Machara A, Endoma-Arias MA, Duchek J, Adams DR, Cox DP, Hudlicky T. Unexpected N-demethylation of oxymorphone and oxycodone N-oxides mediated by the Burgess reagent: direct synthesis of naltrexone, naloxone, and other antagonists from oxymorphone. Adv. Synth. Catal. 2012, 354, 2706–2712. DOI
- US 8,957,072 B2, Processes and intermediates in the preparation of morphine analogs via N-demethylation of N-oxides using cyclodehydration reagents, Examples 2, 3 and 5–8. Google Patents
- Maki T, Tsuritani T, Yasukata T. A mild method for the synthesis of carbamate-protected guanidines using the Burgess reagent. Org. Lett. 2014, 16, 1868–1871. DOI
- Vijayasaradhi S, Srinivas R, Ramesh U, Chandra Prakash K, Sathish K. A simple and effective Lewis acid assisted synthesis of indole-3-sulfonyl carbamates and sulfonamides using Burgess reagent. Results Chem. 2022, 4, 100467. DOI
- Sultane PR, Bielawski CW. Burgess reagent facilitated alcohol oxidations in DMSO. J. Org. Chem. 2017, 82, 1046–1052. DOI
- Burckhardt S. Methyl N-(triethylammonium-sulfonyl)carbamate: “Burgess reagent” (spotlight). Synlett 2000, 559. DOI
About this article. We read the full texts of these sources: the review (ref. 1), the crystal-structure paper (ref. 3), the Organic Syntheses procedure (ref. 8), the supporting information of ref. 14, the δ-diol paper (ref. 5, author manuscript) and the cited examples and figures in the patents (refs. 17, 18, 22, 27 and 30). For the other references, we read the abstracts, graphical abstracts, publisher pages or excerpts. All sources were read on 17 November 2026. Yields are as reported. Where we calculated a yield from the reported masses, the text says so. This article is not a protocol. The full procedures are in the references.
© 2026 NorrChemica. All rights reserved.
Content may not be reproduced without written permission.