The main protecting groups for alcohols are ethers, silyl ethers and acetals. Here’s an overview of their structures, protection and deprotection conditions.
For more details on some of these protecting groups, check out these articles. For each group, we explain the context, mechanisms for protections and deprotections, 3D models and examples (ranging from easy to hard). O-PMB | O-Trityl | | O-TBS | O-MOM | O-THP | O-MEM | O-SEM
When are Alcohol Groups Protected During Organic reactions?
This might be an obvious question we explain here already: Why do we need protecting groups to begin with?
It’s all about getting what you want, while avoiding what you don’t want.
Protecting groups are temporary shields for reactive functional groups to prevent side reactions and maximize chemoselectivity of reactions.
When we modify more complex molecules, we might want a specific functional group to react selectively while preserving other hydroxyl groups presentin the molecule. The synthesis of carbohydrates (and other natural products, like oxygenated terpenoids) thus often requires protecting groups for alcohols.
The issue comes from one of the first reactivities we learn about in organic chemistry: the chemistry of water (H2O) and alcohols, like methanol (CH3OH). The “-OH” hydroxyl group is a nucleophile and can interfere with reactions of other functional groups.
It is somewhat possible to selectively functionalize some groups without protecting groups, for instance by leveraging relative reactivity due to steric hindrance. Primary hydroxyl groups react faster than secondary hydroxyls. However, for the assembly of complex molecules with dozens of reactive functional groups, protecting groups are basically impossible to avoid.
Let me know if you would like additional summaries of protecting group classes (e.g., for amines etc.). If you’re interested in practice problems in organic chemistry, check out the free problem sets.
The Alloc protecting group is used to protect amines. It is removed with catalytic Pd(PPh3)4 and thus orthogonal to many PGs.
This article explains Alloc protection and its unique deprotection mechanism.
👀 In the 3D structure, you see that the allyl group is not planar with the carbonyl!
What is the Alloc Protecting Group?
The Alloc protecting group is related to the Allyl protecting group and orthogonal to almost every other protecting group out there.
It might seem complicated, but it’s just a fancy version of a carbamate protecting group based on two structures: 1) A carbamate which if present in its free form (-OH), can decarboxylate to release the original amino group. Other carbamate protecting groups we’ve seen like Boc, Cbz or Fmoc work this way. They just have a different “trigger” group. 2) An O-allyl group that is activated towards nucleophilic attack of palladium(0) to form of allyl-palladium complexes. Instead of e.g., a tBu group present in Boc which gets “triggered” by acid, this one is triggered by Pd(0)!
Due to the mild deprotection conditions, the group has seen significant application in the synthesis of complex peptides and carbohydrates as a solid alternative to e.g., Boc.
Alloc Protection Mechanism
Alloc protection is trivial and follows the same logic like other carbamates: Nucleophilic attack of the amine to some sort of activated Alloc reagent, typically AllocCl (allyl chloroformate – so like CbzCl for Cbz) or Alloc2O (diallyl dicarbonate – like Boc2O!).
Exemplary conditions are i) AllocCl, pyridine in THF; ii) AllocCl, DMAP, NEt3 in CH3CN; iii) Alloc2O in dioxane, H2O or in CH2Cl2; iv) Alloc-OSu, NEt3, CH2Cl2
Alloc deprotection mechanism
As already alluded to, the magic of this group is in the activated allyl group (given it’s connected to the oxygen of the carbamate). The deprotection is a catalytic cycle, initiated by coordination of Pd(0) and oxidative addition to form an allyl-palladium(II) complex.
The carbamate ligand can dissociate from the complex and decarboxylate to give our desired deprotected amine (again, the same logic as we saw with Boc, Cbz and Fmoc).
But how do we regenerate our catalyst Pd(0) and get rid of the allyl group? There are two options: 1) Nucleophiles can attack the complex and transfer the allyl group, reducing Pd(II) to give Pd(0). Morpholine is one of the key allyl transfer reagent in the text books, but other amines (Me2NH•BH3), C-H acids (e.g., dimedone, barbituric acid)… can be used. 2) Hydride donors can reduce the allyl group via reductive elimination, giving butene. Silanes (such as PhSiH3 / phenylsilane) are common but other hydride donors such as formic acid, SnBu3H (tributylstannane) or sodium borohydride (NaBH4) exist.
If no additional allyl transfer reagent / scavenger would be added, the cycle would not be closed and we would see undesired allylation of our deprotected amine (because it’s a nucleophile).
If you are crazy but want some of that OG E. J. Corey chemistry swag [1], you can also use Ni(CO)4. This compound is extremely toxic and highly volatile… but why would you if you can use palladium?! Fun fact: The name of Corey’s co-worker in the paper [1] is “Suggs” – so yeah, working with nickel carbonyl suggs!
Pd(0) is unstable. To not prepare it fresh, it can be formed from Pd(II) precursorcatalysts like Pd(PPh3)2Cl2 and reductants such as silanes. Not all Pd(0) reactions start with Pd(0)!
Examples of Alloc PROTECTION in Organic Synthesis
This first example [2] shows the orthogonality of Alloc – here, with Fmoc and a methyl ester. In this case, the borane-dimethylamine complex is used as a nucleophilic allyl transfer reagent.
Our second example [3] is from the total synthesis of antillatoxin. This marine natural product was isolated from some exotic cyanobacteria and, in addition to being toxic to shrimp (lol), might have interesting bioactivity (antiproliferation of cells via inhibition of tubulin polymerization).
You’ll see that two allyl groups were deprotected in one step – one from an amino group, and one from an ester. This set up the final intramolecular cyclization step to form the lactam (cyclic amide) in the product.
That’s All(oc) for this article! (ok, bad pun…) Feel free to check out my other articles on protecting groups, my page or my videos!
Alloc Protection experimental procedure [4]
“A mixture of amine (0.0842 mmol), NaHCO3(44 mg, 0.53 mmol, 6 equiv), THF (3 mL), and H2O (3 mL) at room temperature was treated with allyl chloroformate (28 μL, 0.26 mmol, 3 equiv). The reaction mixture was stirred at room temperature for 12 h, extracted with EtOAc (200 mL, 100 mL), and the combined organic layers were washed with saturated aqueous NaCl (200 mL), dried over Na2SO4, and concentrated in vacuo. Column chromatography provided 38 (48.8 mg, 87% over 2 steps) as a white foam.”
Alloc deprotection experimental procedure [4]
“A stirred solution of 40 (8.61 g, 8.2 mmol, 1.0 equiv) in CH2Cl2 (82 mL) at 0 ̊C under Ar was treated with PhSiH3 (7.1 ml, 57 mmol, 7.0 equiv) followed by Pd(PPh3)4 (0.95 g, 0.82 mmol, 10 mol %). The reaction mixture was stirred at 0 °C for 1 h and concentrated under reduced pressure. Column chromatography provided the semi-pure amine (7.79 g) as a yellow solid.”
Alloc Protecting Group References
General: P. G. M. Wuts, T. W. Greene: Greene’s Protective in Organic Synthesis (Wiley)
[1] Cleavage of allyloxycarbonyl protecting group from oxygen and nitrogen under mild conditions by nickel carbonyl | E. J. Corey, J. William Suggs | J. Org. Chem. 1973, 38, 3223
[2] P. J. Kocienski: Protecting Groups (Thieme)
[3] Total Synthesis and Revision of Absolute Stereochemistry of Antillatoxin, an Ichthyotoxic Cyclic Lipopeptide from Marine Cyanobacterium Lyngbya majuscula | Fumiaki Yokokawa, Hideyasu Fujiwara, Takayuki Shioiri | Tetrahedron 2000, 56, 1759
[4] Next-Generation Total Synthesis of Vancomycin | Maxwell J. Moore, Shiwei Qu, Ceheng Tan, Yu Cai, Yuzo Mogi, D. Jamin Keith, Dale L. Boger | J. Am. Chem. Soc. 2020, 142, 16039
The SEM protecting group is used to protect alcohols and amines. SEM is removed with fluoride (F–) or acid (H+).
2-(Trimethylsilyl)ethoxymethyl (SEM) is an acetal-type protecting group for alcohols, but can be used for other nucleophiles like amines as well. Here we cover SEM protection and deprotection mechanisms, as well as examples.
No surprise, SEM is related to MEM and the simpler MOM protecting group. But note the TMS!
👀 Here is an interactive 3D structure of SEM.
What is the SEM Protecting Group?
The SEM protecting group is essentially a combination of the MEM group and a TMS group. The presence of silyl group means that fluoride can play a part in deprotection as well.
When attached to an alcohol, it forms a much less reactive acetal. However, just like other similar protecting groups, it can also be added to other nucleophiles like amines. SEM is stable under various conditions, including bases, reductants, organometallic reagents, oxidants and mild acids.
The SEM group was invented in 1980 by Lipshutz and Pegram [1], so just a few years after introduction of MEM by E. J. Corey.
SEM Protection Mechanism
The protection is MOM and MEM all over again. 1. Option: Treatment with SEM chloride and DIPEA (N, N-diisopropylethylamine) or another weak base. Deprotonation occurs after nucleophilic attack. 2. Option: Treatment with a strong base like NaH (KH, n-BuLi, …)and SEMchloride. Experimentally / in the lab, only base is added to the alcohol first – and only after some time (e.g., 1h to ensure the alkoxide is formed), SEMCl is added.
Again, note the activation and higher reactivity of such alkylating agents due to the adjacent oxygen.
SEM deprotection mechanism 1: Fluoride (F–)
The presence of silicon in SEM allows for deprotection with fluoride anions (high thermodynamic affinity for the very strong Si-F bond). These mechanisms proceed via formation of the pentavalent siliconate intermediate.
The siliconate is unstable and can trigger a beta-elimination decomposition. This releases three neutral molecules: TMSF, ethylene and formaldehyde. We can just draw everything in one single step and take a proton from the solution (under acidic conditions like HF, the oxygen might be protonated already prior to decomposition).
The benefit of fluoride is that it the conditions are often orthogonal / compatible with many functional and protecting groups. However, compared to ordinary silyl ethers (e.g., TMS), SEM deprotection tends to require higher temperatures, longer reaction times, or in some cases the use of additives (HMPA).
Exemplary fluoride deprotection conditions for SEM are i) TBAF, DMF; ii) HF, MeCN; iii) LiBF4, (MeCN-H2O).
SEM deprotection mechanism 2: Bronsted Acid (H+)
Though less preferred than fluoride-mediated deprotection, acidic hydrolysis also works for SEM. This is another parallel to MOM, MEM and THP. There are more mechanistic options , but the direct path is probably the most likely one.
i) Direct path: The oxygen of our protected alcohol is protonated and achieves the deprotection in the simplest manner.
ii) Indirect path: The other ether oxygen is protonated, and goes via the hemiacetal intermediate (formaldehyde is lost in the solution or upon work-up). This is what we’ve seen for MEM or MOM.
iii) Beta elimination: What we’ve seen during the fluoride deprotection, but now just in the acidic variant.
What about relative stability? According to Kocienski [2], SEM is more labile than MOM and MEM under acidic conditions. (Obviously, fluoride does not remove MOM). On the other hand, Greene [3] notes that SEM are very robust groups and often difficult to remove.
The truth is probably more the latter. Here is an example from the total synthesis of taxol by Kuwajima [4].
“We also investigated the Birch reduction of derivatives of 25a with various protecting groups on the C2-OH: reaction of the di-tert-butylsilylene derivative 25b induced C2−O bond cleavage predominantly (eq 3). Use of the C2-O-SEM substrate 25c gave a much more satisfactory result (eq 4), but we encountered much difficulty in removing the SEM group at a later stage; the SEM group was thus deemed to be unsuitable for the present purpose.”
Lesson: Higher yield is not always better! Sometimes, we have to accept lower yields in the early steps of a synthesis to finish it!
Examples of SEM PROTECTION in Organic Synthesis
Our first example shows the orthogonality of SEM and MOM [3]. Harsher conditions are needed to remove SEM, but our MOM groups survive at 100 °C without issue.
The second example is really interesting.
Question: Although this reaction uses similar conditions as our first example, we are far from our nice 98% yield of a single product. Can you guess the structures of the three products?
Solution: OK, first of all we do see a MOM group but the first example taught us that these do not fly off when exposed to fluoride sources (unless we cook them in e.g., HF).The first insight is that the triethylsilyl (TES) group might also be a victim of the fluoride deprotection. So we might expect a mixture of SEM- and TES-deprotected products.
Indeed, product 1 is SEM- and TES-deprotected one whereas product 2 is only TES-deprotected.
But what about product 3?
Beware of interrupted deprotections!
Instead of beta-elimination and release of ethylene (and formaldehyde), it looks like we have a proto-desilylation. The ethyl group remains on the ether, and we basically have a one-carbon extended MOM group now. Why could this SEM group just not be bothered to leave? Uhm, we do not know. That’s chemistry for you.
We have seen a few of these tricky questions with other protecting groups, where, e.g., the presence of an intramolecular nucleophile can lead to side products.
As we mention often, many alcohol protecting groups can also be used to protect carboxylic acids or amines. As we see in our third example [5], SEM can also used to protect the nucleophilic nitrogen in heterocycles like imidazole.
This acidic deprotection was particularly sluggish. At 25 °C, >200 equivalents of TFA were added over two batches and ultimately just gave 20% yield. As we see, SEM is indeed hard to remove!
We’re done! If you learned something, make sure to check out my other articles on protecting groups, my page or my videos!
SEM Protection experimental procedure [6]
“To a dry 100 mL round-bottom flask under argon was added anhydrous DMF (40 mL) and NaH (0.293 g, 60%, 7.31 mmol); then the solution was cooled to 0 °C. In a separate flask the alcohol (1.021 g, 4.87 mmol) was dissolved in DMF (10 mL), and then this solution was added dropwise by cannula to the NaH/DMF mixture. The reaction mixture was stirred at 0 °C for 2 h, and then 2-(trimethylsilyl)ethoxymethyl chloride (1.067 g, 6.33 mmol) was added. After 10 h saturated NH4Cl solution (10 mL) was added, and this mixture was extracted with ethyl acetate. The combined organic layers were washed with water (3 × 25 mL) and brine (1 × 25 mL), dried over Na2SO4, and then concentrated to give a crude solid which was purified using flash chromatography (hexanes–ethyl acetate, 1:1) to give 12 as a light brown solid (1.25 g, 78%).”
SEM deprotection experimental procedure [6]
“To a 100 mL round-bottom flask, the alcohol was added (0.396 g, 0.84 mmol) and dissolved in DMF (50 mL). To this was added tetramethylethylenediamine (0.293 g, 2.53 mmol) and TBAF (1.0 M in THF, 2.53 mL, 2.53 mmol), attached a reflux condenser and set the reaction for heating at 45 °C for 20 h. After confirming the completion of the reaction by LCMS (m/z 340), the reaction was allowed to cool to room temperature and to it was added saturated solution of NH4Cl. The contents were transferred to a separatory funnel containing 100 mL of water. Extraction was done using ethyl acetate, and the combined organic layer was washed with water, followed by brine, dried over sodium sulfate, and evaporated the solvents to give a crude solid which after flash chromatography purification using hexane–ethyl acetate (1:1) gave 8 as a buff solid (0.202 g, 71%).”
SEM Protecting Group References
[1] β-(Trimethylsilyl)ethoxymethyl chloride. A new reagent for the protection of the hydroxyl group | Bruce H. Lipshutz, Joseph J. Pegram | Tetrahedron Letters 1980, 21, 3343
[2] P. J. Kocienski: Protecting Groups (Thieme)
[3] P. G. M. Wuts, T. W. Greene: Greene’s Protective in Organic Synthesis (Wiley)
[5] Discovery of Bis-imidazolecarboxamide Derivatives as Novel, Potent, and Selective TNIK Inhibitors for the Treatment of Idiopathic Pulmonary Fibrosis | Vladimir Aladinskiy, Chris Kruse, Luoheng Qin, Eugene Babin, Yaya Fan, Georgiy Andreev, Heng Zhao, Yanyun Fu, Man Zhang, Yan Ivanenkov, Alex Aliper, Alex Zhavoronkov, Feng Ren | J. Med. Chem. 2024, 67, 19121
[6] Tale of Two Protecting Groups—Boc vs SEM—for Directed Lithiation and C–C Bond Formation on a Pyrrolopyridazinone Core | Reji N. Nair, Thomas D. Bannister | Org. Process Res. Dev. 2016, 20, 1370
The MEM protecting group protects alcohols as stable acetals. MEM is removed with protic acids (H+) and more importantly, Lewis acids.
2-Methoxyethoxymethyl (MEM) is an acetal-type protecting group for alcohols. This article explains MEM protection and deprotection mechanisms, as well as examples.
MEM is related to the simpler MOM protecting group. It’s rather oldschool, but the additional ether group brings an interesting twist!
👀 Quite a few oxygens – I hope this interactive 3D model helps!
What is the MEM Protecting Group?
The MEM group belongs to the class of acetal (‘double-ether’) protecting groups which are much less reactive than their free alcohol counterparts. Fun fact: The MEM group was introduced in 1976 by the legendary chemist E. J. Corey and co-workers [1], just like the TBS/TBDMS and allyl protecting groups.
MEM is stable under various conditions, including strong bases, reductants, organometallic reagents, oxidants and mild acids. What makes it unique is the selective cleavage with Lewis acids (see below).
MEM Protection Mechanism
The protection conditions are identical to MOM: 1) Treatment with MEM chloride and DIPEA (N, N-diisopropylethylamine) or another weak base. Deprotonation occurs after nucleophilic attack. 2) Treatment with a strong base like NaH and MEM chloride. Experimentally / in the lab, only base is added to the alcohol first – and only after some time (e.g., 1h to ensure the alkoxide is formed), MEMCl is added.
MEMCl is another case of an activated alkylating agent. The adjacent oxygen can facilitate departure of the chloride, creating a highly electrophilic oxonium ion. (No surprise it’s also bad for you – it alkylates your DNA!)
MEM deprotection mechanism 1: Bronsted Acid (H+)
Just like MOM or THP, we can get rid of MEM with acid, tough typically stronger / more forcing conditions are required. Thus, it is possible to selectively remove THP, MOM or PMB groups in the presence of MEM. The simplest mechanism starts with protonation of our protected oxygen in the acetal system. This cation can release of our free alcohol and form some type of byproduct (depending on solvent).
The alternative / indirect mechanismwould be protonation of the other oxygen atom, going through a hemiacetal intermediate. With water in the solvent or upon work-up, this unstable species releases formaldehyde and gives our free alcohol.
So what makes MEM unique versus MOM? Due to the presence of an additional ether oxygen, the acetal is much more labile when exposed to Lewis acids. Due to this additional lone pair, we have bidentate coordination to metals such as i) ZnBr2, ii)TiCl4, – or like we’ve seen for MOM, also more Lewis acidic species like iii)TMSI.
The productive mechanism is now kind of the opposite as before – it is the oxygen of our protected alcohol that cleaves the protecting group. This oxonium intermediate is ultimately hydrolysed (either water present in the reaction or upon work-up), releasing formaldehyde and giving our deprotected product.
Examples of MEM in Organic Synthesis
Our first example [2] is really chill. Simple introduction of MEM, just like we’ve talked about.
Too easy for you? Well, check this one out:
Upon treatment of the compound above, two products were observed in significant yield. What are they?
One of the products surely is the normal MEM-deprotected alcohol. After all, this is an article about MEM and we’ve just seen than ZnBr2 will act as a Lewis acid and remove the group. We cannot say with certainty, but this is probably the major product 1.
But what about product 2? If you have no ideas, go back to the mechanism above and look at what happens after initial break-down of MEM!
Always watch out for neighbouring group participation!
The answer is very cool: Our starting material has a sneaky hydroxyl group that can attack our oxonium intermediate. This occurs because the intramolecular reaction is very fast – even despite the tertiary alcohol being very hindered. This means that we do not go through the intermolecular bromide-hydrolysis pathway, but rather form a cyclic acetal!
That’s it! If you learned something, make sure to check out my other articles on protecting groups, my page or my videos!
MEM Protection experimental procedure [2]
“To a solution of alcohol (4g, 40 mmol) in CH2Cl2 (100 mL)under N2 was added MEM chloride (7.5g, 60mmol) and DIPEA (7.8g, 60mmol) at 25 °C and the reaction mixture was stirred at rt for 5 h. Water (30 mL) was added to the reaction mixture and CH2Cl2 was used to extract the mixture. The combined organic layers were washed with brine, dried over Na2SO4 and concentrated to give 7.4g of crude product. Flash chromatography of the residue on silica gel gave 6.1g (80%) of MEM protected alcohol as a colorless liquid.”
MEM Protecting Group References
General: P. G. M. Wuts, T. W. Greene: Greene’s Protective in Organic Synthesis (Wiley)
[1] A new general method for protection of the hydroxyl function | E.J. Corey, Jean-Louis Gras, Peter Ulrich | Tetrahedron Letters 1976, 17, 809
[2] Total Synthesis of (±)-Deoxypenostatin A. Approaches to the Syntheses of Penostatins A and B | Barry B. Snider, Tao Liu | J. Org. Chem. 2000, 65, 8490
The MOM protecting group protects alcohols as acetals. MOM is introduced with MOMCl and removed with acid (H+).
Methoxymethyl is the simplest acetal protecting group for alcohols. This article covers mechanisms for MOM protection and deprotection, and examples.
MOM is quite common and similar to protecting groups such as THP, so it’s a “must-know” for students!
👀 Here’s an interactive 3D model of the MOM protecting group.
What is the MOM Protecting Group?
Protecting groups serve to minimize unwanted side reactions during organic synthesis. When protecting alcohols, methoxymethyl ethers actually form a type of acetal (‘double-ether’). These are much less reactive than the free alcohol.
Remember – just like with other protecting groups – that different nucleophiles can be protected with the same protecting group. This is why MOM is also used to protect phenols, carboxylic acids (still a nucleophilic oxygen) as well as amines (nucleophilic nitrogen). Introduction and removal (typically acid) follow the same logic, so most study materials focus on just the protection of alcohols.
MOM Protection Mechanism
Two protection conditions are most common for alcohols: 1) Treatment with MOM chloride and DIPEA (N, N-diisopropylethylamine) or another weak base. Here, deprotonation occurs after nucleophilic attack. 2) Treatment with the strong base NaH (or KH) and MOM chloride. Here, deprotonation occurs first and nucleophilic attack is second. Experimentally / in the lab, only base is added to the alcohol first – and only after some time (e.g., 1h to ensure the alkoxide is formed), MOMCl is added.
Note that the lone pairs on the oxygen on MOMCl actually activate the departure of the chloride. This creates a highly reactive, electrophilic oxonium ion which is captured by the alcohol. This makes MOMCl a very powerful alkylating agent and carcinogenic (it alkylates your DNA base pairs which is not what you want). So, special care in the lab is required.
A safer alternative protocol is the following: 3) Use of dimethoxy methane and an acid. This reaction is different as it is an acetal exchange reaction and uses an excess of reagent to drive the equilibrium.
mOm deprotection mechanism
The standard MOM deprotection is acidic hydrolysis. Protonation activates the acetal system towards release of our free alcohol. Less importantly, the remaining stabilized cation forms a byproduct after trapping by the solvent. This is pretty similar to the THP deprotection.
Note that you can also draw the alternative / indirect mechanism with protonation at the other oxygen, leading to a hemiacetal intermediate and ultimately our free alcohol upon elimination of formaldehyde.
Exemplary deprotection conditions are: i)HCl in aqueous EtOH; ii) TFA (trifluoroacetic acid) in dichloromethane; iii) PPTS (pyridinium p-toluenesulfonate) in tBuOH.
Alternatively, reactive electrophiles / Lewis acids like TMS+ can also be used to remove MOM groups. A key method is use of TMSBr (the mechanism follows a similar logic of nucleophilic attack).
Classes often teach you a single method (to preserve your sanity) of protection or deprotection, but having different options at our disposal helps in the lab when we deal with complex molecules that have other protecting groups!
Examples of MOM in Organic Synthesis
Now that we now the basics, let’s check out two use cases of MOM and connect it to other protecting groups we have learned about.
The first example [1] shows that lability does not equal lability. We’ve learned that the PMB protecting group can be cleaved under acidic conditions. However, HCl (generated in situ from AcCl and MeOH) here selectively removed the MOM group while retaining the PMB group. Chemistry is pretty experimental!
The second example [2] is a fancy regioselective introduction of the MOM group. This procedure achieves MOM protection at the more sterically hindered alcohol of a vicine diol which should be less reactive with reagents like MOM-Cl (secondary vs. primary alcohol). The trick here is to create the orthoester first as a detour. It can then be reduced with DIBAL-H which exhibits preference in coordination and hydride reduction, leaving the MOM group hanging at the more hindered alcohol!
Appreciate you reading the full article! Feel free to check out other protecting groups, my page or my videos!
MOM Protection experimental procedure [3]
“An oven-dried 3-neck 500 mL round-bottom flask equipped with a stir bar was charged with alcohol 16 (15.5 g, 78.18 mmol, 1.0 eq.), DIPEA (40.41 g, 312.72 mmol, 4.0 eq.) and DCM (160 mL) under Ar. The resulting suspension was cooled down to 0 °C and freshly distilled MOMCl (18.88 g, 234.50 mmol, 3.0 eq.) was added dropwise over a period of 10 min. NaI (5.80 g, 39.09 mmol, 0.5 eq) was added to the reaction solution, and the resulting mixture was allowed to warm to 25 °C, and stirred for 16 h. After completion of the reaction, the reaction mixture was quenched with saturated ammonium chloride solution (300 mL) and diluted with DCM (100 mL).The two layers were separated and aqueous layer was extracted with DCM (2 × 100 mL). Combined organic phases were washed with saturated sodium chloride solution (1 × 100 mL)，dried over anhydrous magnesium sulfate and concentrated under reduced pressure. The crude product was purified by silica gel flash chromatography to give 17 (17.5 g, 92% yield) as colorless liquid.”
MOM deprotection experimental procedure [3]
“31 (68 mg, 0.142 mmol, 1.0 eq.) was dissolved in DCM/TFA = 15/1 (3 mL) at 25 °C. The resulting suspension was stirred at 25 °C for 12 h, when TLC analysis of the crude mixture showed full conversion. The reaction mixture was diluted with DCM (2 mL) and treated with sat. aq. NaHCO3 (4 mL). The layers were separated and the aqueous phase was extracted with DCM (2 x 3 mL). Combined organic phases were washed with sat. aq. NaCl (1 × 5 mL)，dried over anhydrous MgSO4 and concentrated under reduced pressure to give the crude product. The crude residue was purified by preparative TLC to provide isorosthin L (35 mg, 71% yield) as a white solid.”
mom Protecting Group References
[1] P. G. M. Wuts, T. W. Greene: Greene’s Protective in Organic Synthesis (Wiley)
[2] A convenient procedure for the regioselective monoprotection of 1,n-diols | M. Takasu, Y. Naruse, H. Yamamoto | Tetrahedron Lett. 1988, 29, 194
Summary of electrophilic aromatic substitution reactions
1. Mechanism: A nucleophilic aromatic ring attacks the electrophileE+, giving a cation, and loses H+ to regain its aromaticity. This substitutes C-H for C-E.
2. Key reactions: Bromination, chlorination, Friedel-Crafts alkylation (R) & acylation (C(O)R). nitration (NO2), sulfonation (SO3H). Most reactions utilize acid or Lewis acids to create reactive electrophiles.
3. Substituents – Activation: Electron-donating groups at a ring increase nucleophilicity and reactivity. Electron-withdrawing groups work the opposite.
4. Substituent – Regioselectivity: Activators lead to ortho and para products. Deactivating groups direct meta (excl. halogens).
5. Examples: TNT synthesis nicely exemplifies ring activation. Pharmaceuticals use electrophilic aromatic substitution – sometimes with heterocycles!
6. Orbitals: Addition is fastest where overlap of donor (HOMO) and acceptor (LUMO) is best. However, orbital models do not explain the full story!
Common mistakes and reaction examples at the end!
1. mechanism of electrophilic aromatic substitution
So what is electrophilic aromatic substitution? Substitution as it replaces a C-H bond at an aromatic ring with a new C-E bond. Electrophilic because our new E group was introduced as an electrophile.
Due to its aromaticity, benzene is relatively unreactive. Unlike alkenes, strong electrophiles give substitution products and not addition products.
Because our new group is an electrophile, the double bond of the aromatic ring has to be the nucleophile (just like in electrophilic additions)! Here’s how to identify nucleophiles vs electrophiles. Note that the electrophile can be a neutral molecule that’s positively polarized, it does not have to be positively charged.
Note: The order is ‘form C-E, then break C-H’. While the other way around can work, it requires different reaction conditions (very strong base to deprotonate C-H) and is not as broadly applicable.
Let’s see what we can learn from this energy diagram. 1. One of the double bonds attacks the electrophile (here: NO2+). Because this breaks the aromaticity of the starting material, the step requires quite a bit of activation energy (the molecule has to have energy to climb the hill of TS1).
2. The addition product is a carbocation(the double bond was ‘invested’ by one carbon to form the bond with E+, the other carbon has one electron less in its proximity and is now a non-aromatic carbocation). The positive charge sits next to the other double bonds , so it is somewhat stabilized and delocalized. This is a real intermediate whereas the transition states are just “snapshots at maximum energy”. This is also called Wheland intermediate or sigma complex.
3. The deprotonation occurs quite easily (the hill to TS2 is lower than the first one) and gives the stable substituted aromatic product. By the way, you don’t need to add base (e.g., KOH) for this step; the proton will get lost to other molecules present in the reaction or solvent. If the final product is lower in energy vs. starting material (more specifically, “enthalpy”), the reaction is exothermic.
But benzene is very simple and boring. In sections 2-4, we cover: 2. What types of electrophiles we can add and how we generate them 3. How other ring substituents can accelerate or slow down this reaction (activation) 4. How other ring substituents influence where the reaction occurs (regioselecitvity)
By the way, this is an old reaction! Pioneering work came from the Brit Armstrong in the 1880s already, and the American George Wheland in the 20th century.
Source: J. Chem. Soc., Trans., 1887, 51, 258
From our videos, we know that the early days of chemistry were was a bit crazy. Even simple benzene was disputed (left: Armstrong’s “centroid formula”). We can’t blame these OG chemists as our modern knowledge evolved out of the many dubious theories.
1. Bromination or chlorination: Uses molecular bromine or chlorine to form a C-Br or C-Cl bond, needs Lewis acid 2. Friedel Crafts alkylation & acylation: Forms a C-C bond, needs Lewis acid 3. Nitration: Incorporation of a nitro group (-NO2), needs acid 4. Sulfonation: Incorporation of a sulfonic acid (-SO3H), needs acid
These reactions have the same fundamental mechanism… but they differ in how they generate their reactive electrophile. They all use some sort of reagent (green/purple boxes) to get to the true electrophilic species that is attacked by the aromatic ring.
Bromination, chlorination and Friedel-Crafts reactions use a Lewis acid (=electrophilic metal compounds that things with electron pairs can coordinate to). This makes the Br, Cl, R or C(O)R centers more positively polarized, so they are more reactive with the double bond.
Nitration and sulfonation use a ‘normal’ acid (e.g., sulfuric acid) to protonate the reactants and activate them into more electrophilic species. Nitration is a particularly useful reaction as the group can converted to amino or hydroxyl groups.
After initial electrophile generation, all reactions follow the same steps and experience the same substituent effects (Sections 3. and 4. below). For full clarity, I might explain the nuances of each reaction in more detail in future posts.
There are two important points I want to point out on the halogenation reactions above: bromination and chlorination.
1) The typical conditions are bromine or chlorine + Lewis acid, but students will see in more advanced classes that there are other sources of electrophilic bromine or chlorine, like NBS or NCS (N-bromo/chloro-succinimide). I’ll probably explain these reagents more in separate posts.
In organic chemistry, there are always multiple ways of achieving a certain reaction or generating a certain reactive species, e.g., an electrophile.
2) Chemistry class syllabus usually ignores the other two cousins in the halogen family: iodine and fluorine. But that doesn’t mean that they cannot participate in these reactions! The point is more that these reactions are less common and straightforward, so they are usually not discussed. If you can somehow generate activated/ polarized electrophiles (I+ or F+), they will also react with electron-rich aromatic rings just like Cl+ or Br+ shown above. While there are several options for iodination (using I2 and additional reactants like acids), fluorination is more challenging and requires special reagents (not F2). Aryl fluorides are thus usually made with nucleophilic aromatic substitution reactions like the Balz-Schiemann reaction.
3. Activating and deactivating groups
Recall: Our aromatic ring is a nucleophile, attracted to positive charge or polarization (which are electrophiles) because it has a decent amount of electron density.
The higher the electron-density in a ring, the more nucleophilic and more reactive it will be. This accelerates aromatic electrophilic substitution (higher reaction rate).
The electron-density of aromatic systems is influenced by the ring substituents. For example, nitrobenzene does not undergo normal Friedel-Crafts acylations.
Activating groups are electron-donating groups that increase nucleophilicity and reactivity of an aromatic ring, usually through mesomeric πeffects (resonance). Deactivating groups are electron-withdrawing groups that decrease nucleophilicity and reactivity either via inductiveσ or mesomeric πeffects.
Here are the key groups, (directional) strength and primary drivers. Activators tend to be π donors (they push/ delocalize their electron pair into the ring) while deactivators tend to be π acceptors (they accept delocalized electrons, acting as an electron sink). If this doesn’t make sense, please read up on resonance.
Note the key counterintuitive example: halogens deactivate rings more strongly (due to their negative inductive σ effect, think high electronegativity) than they activate them (due to their mesomeric π effect, they have a free electron pair).
Another way to think about these groups is to revisit our energy diagram. The TS1 is partially positively charged and the intermediate is fully positively charged. Obviously, this positive charge will be very comfortable if there is a lot of electron density in the ring (charges like to be spread across and stabilized across atoms). Inversely, deactivators would make the intermediate unhappy. It’s like robbing someone who’s already very poor – not the best look.
The stabilizing effects for activating groups mean that TS1 requires a lower activation energy (because the ring is more nucleophilic) and our carbocation intermediate is also lower in energy (the charge is better stabilized by the electron-rich ring).
Note: It’s critical that these groups are directly attached. If there is an “alkyl spacer” between the ring and a strong (de)activator, the effect will vanish. The donor or acceptor groups need to be conjugated with the ring! The substituents here would be more like σ donors.
4. Regioselectivity of aromatic electrophilic substitutions (Directing Effects)
All activating groups and halogens direct/favor substitution at ortho& para positions. All deactivating groups excluding halogens direct/favor substitution at meta.
This regioselectivity can be rationalized by looking at the delocalization/ stabilization of the carbocation intermediate. Ortho and para additions favor carbocations where the positive charge sits next to the R group (either directly in case of ortho or after delocalization in the case of para). If R can stabilize this positive charge as a σ or π electron donor, it will lower the energy of the carbocation and prefer these pathways over the meta addition.
Well if R is an electron-withdrawing group, the positive charge will absolutely hate being next to it (electron-sink next to something that already doesn’t have enough electrons). Thus, deactivating groups avoid ortho and para substitution and show meta selectivity.
Halogens are again sneaky. Regarding ring activation, their negative inductive effect (electron-withdrawal) overrides their positive mesomeric effect (electron pair donation). Regarding regioselectivity, the mesomeric effect is more important (deal with it). Outside of this special case, all groups follow the same rules.
5. Examples: Tnt & pHARMACEUTICALS
Our first example is 2,4,6-trinitrotoluene or TNT. One of many ways of synthesizing TNT is exactly what you think: take toluene and nitrate the heck out of it. The issue? Substituent effects! You see, while toluene is decently nucleophilic (it’s slightly activated through the methyl group), it becomes less and less reactive the more nitro groups are attached.
The first addition gives a mix of ortho and para substituted products. final nitration only occurs at high temperatures and use of fuming sulfuric acid, a particularly reactive form of sulfuric acid containing additional sulfur trioxide. There are easier and more clever ways to make TNT, but that shouldn’t be our focus.
You also see that just because TNT is deactivated towards electrophilic aromatic substitution, it still likes to explode violently. Reactivity is always relative. This is because detonation does not depend on the nucleophilicity of the aromatic ring 🙂
All reactions can happen either between two molecules (intermolecularly) or also within a single molecule (intramolecularly), if it has all the needed functional groups.
Most pharmaceuticals contain aromatic rings so we can find these reactions in their syntheses. One example is the drug lifitegrast. Its synthesis uses an intramolecular Friedel-Crafts alkylation to create the bicyclic core.
Final lesson: Heterocycles can also be nucleophiles! (aromatic ring with heteroatom)
The synthesis of the antibiotic metronidazole is very simple. The first step is nitration of the ring with carbon as the nucleophile. In the second step, the nitrogen acts as the nucleophile and undergoes a SN2 reaction with the alkyl chloride.
The five-membered rings pyrrole (heteroatom: N) and furan (heteroatom: O) are very electron-rich so we also often see them engage in electrophilic substitution reactions.
This is because the electron pair can stabilize the positive charge during the reaction mechanism. It’s like having an activating substituent effect within the ring!
6. Orbitals in aromatic substitutions
You probably shouldn’t read this section if you’re just learning about aromatic substitutions. Drawing resonance structures is useful for us mortal humans. However, they are just tools and don’t correspond to reality. After all, we’re taught that “real structure” of a molecule is a mix of all “good” resonance structures. Advanced readers might ask themselves if orbitals can be useful predictors for aromatic reactions?
We won’t go through orbital theory here, but in short: When atoms form bonds in a molecule, their atomic orbitals combine to molecular orbitals. These can be filled with electrons or empty (or half-empty, if it’s a radical). The highest-energy filled orbital (HOMO, occupied) influences nucleophilic behavior and the lowest-energy empty orbital (LUMO, unoccupied) influences electrophilic behavior.
Looking at orbitals can be partially insightful. Here’s a 3D visualization of the early days of an electrophilic aromatic nitration. The HOMO of chlorobenzene (nucleophile) is getting close to the LUMO of the nitronium cation (electrophile). How will they add?
Electrophilic attack will occur where p orbitals contribute most to the π HOMO. The higher the overlap of donor (HOMO) and acceptor (LUMO), the faster the reaction.
You can see that on the ring, para and (to a slightly lesser degree) ortho positions have high contributions, while there is “almost no orbital” on meta. Any electrophilic attack at meta would thus be disfavored (low orbital overlap). Our nitronium electrophile has a slightly higher LUMO localization at the central nitrogen, meaning the nitrogen is more electrophilic than the oxygens!
So, it looks like molecular orbitals can nicely predict the regioselectivity that we observe for chlorobenzene. So why don’t people talk about orbitals in the same fashion as they do e.g., for the Diels-Alder reaction?
Aromatic substitutions are one of the first reactions students, so going into orbitals would be an overkill that makes everyone hate chemistry even more.
More importantly, they don’t always give the full picture. Consider the HOMO of nitrobenzene below. This aromatic ring is deactivated and should attack meta.
We do see that the para position is not contributing to the HOMO at all, but the orbital structure does not tell us why there is no substitution at ortho! In this case, our resonance structure analysis is actually more helpful because it tells us that the ortho-position of nitro groups is particularly electron-poor.
There are other ways to analyse this, e.g., by computing charge density. Below you can find an atomic charge analysis for benzene, aniline and nitrobenzene. Ignore the method or units – simply put, a more negative number equals more electron-density at a given atom. More electron-density means more nucleophilic and more reactive towards electrophiles.
Question to you: What are important takeaways from this charge analysis?
Source: Phys. Chem. Chem. Phys., 2016, 18, 11624
Here are some thoughts:
Benzene: Obviously, hydrogens are more electropositive than carbons (they have a positive charge number). The molecule is symmetrical.
Aniline: Electron-density is higher at ortho and pararelative to meta, explaining the directing effect. Relative to benzene, ortho and para have higher electron density, i.e., the aniline ring is more activatedand reactive . Note that there is a lot of electron density on the amine as well! (Aniline can also react as a N-nucleophile! See common mistakes)
Nitrobenzene: Electron-density is highest at meta relative to ortho and para, explaining the directing effect. Now, the carbons all have lower electron density, in line with deactivation / lowering of reactivity due to the nitro group. Sidenote: The nitro group’s electron-withdrawing effect further polarizes the C-H hydrogens as they are more electron-poor in nitrobenzene.
So, orbital theory is sometimes more, and sometimes less useful. Generally, combining multiple models and perspectives gives us a better appreciation. For now, you can happily ignore orbitals for these reactions (but not for others!).
Common Mistakes in Electrophilic Aromatic Substitutions
Students can feel overwhelmed trying to remember effects or specific aromatic substitutions. This can lead to common mistakes or errors, including:
Mixing up ortho, para and meta: This is the first hurdle for new students. If you’re struggling with this, try remembering it this way: Meta is in the middle between the other two. The letter o comes before p in the alphabet, so ortho is closer to R.
Thinking too superficially: Although electrophilic substitution is the most common reaction of aromatic compounds you’ll see, don’t jump to conclusions too quickly. In this example, an acyl chloride (electrophile) would lead to O-acylation instead of electrophilic aromatic substitution. The benzene ring is deactivated and thus not nucleophilic enough.
Along the same lines: In section 6. Orbitals, we saw that charge analysis suggests high electron density at the nitrogen of aniline. Again, under many conditions, anilines can react as N nucleophiles instead of C nucleophiles (e.g., alkylation at nitrogen instead of at carbon).
Electrophilic or nucleophilic?: Get your terms right. Nucleophilic aromatic substitution reaction does not substitute C-H but rather C-X, where X is a good leaving group. This is the opposite of what saw for the electrophilic substitution: our aromatic ring is the electrophile here. This reaction is explained here.
Misunderstanding substituent effects: Abbreviations can be confusing, and we might wrongly believe a specific atom or an abbreviated group to be activating or deactivating. Before you go with it, double-check whya substituent would be an electron-donating group or an electron-withdrawing group. E.g., can you really find an electron-pair that is driving a positive mesomeric effect?
Remember to consider if a newly introduced functional group is more or less activating. Reactions often use an excess of reagents (more moles of reagent than starting material). Adding one activating group? The ring might add more (e.g., polyalkylation in Friedel-Crafts alkylations). Want to add more than one deactivating group? This might be challenging (remember our TNT example)!
Electrophilic Aromatic Substitution Examples
Here are additional questions to test your knowledge of the reaction mechanism and directing effects of substituents.
The first exercise is pretty straight-forward, with the fluoro-substituted ring being less activated (thus, bromination occurs on the other one, para to the aryl-aryl bond). The same idea applies to the question on the right, where we expect the inductive acceptor effect of the trifluoromethyl group to deactivate the ring towards sulfonation. The Friedel-Crafts acylation is pretty obvious (para to the electron-donating ether group). Last, we have a Friedel-Crafts like C-C bond formation. Instead of a alkyl halide and Lewis acid, we generate our carbon electrophile by protonation of the double bond with sulfuric acid.
Hope this summary helped you get a good overview of this class of aromatic reactions. Feel free to check out my page or my videos for more learning and fun content!
Bredt’s rule states that double bonds do not occur at bridgeheads of bridged rings.
BUT: “Anti-Bredt olefins” defy this rule. They are not often talked about, but chemists have known them for decades. In this post, we explain when Bredt’s rule becomes invalid – a good read for any student willing to go beyond their curriculum. This is particularly relevant as recent news headlines touted the “synthesis of impossible molecules” by chemists. More visuals can be found in my video on this topic.
Interactive 3D model of 1-norbornene, an anti-Bredt olefin
BredT’s rule: History and its original form
To understand Bredt’s rule, we need to go back in time. Julius Bredt was one of the oldschool German chemists of the 19th century, and much of his early work centered on the molecule camphor. This pleasantly smelling terpenoid natural product is found in camphor and other trees, and saw a move towards large scale use for the plastics industry in the 1870s. Various chemists proposed pretty freaky structures, including the legendary August Kekulé.
However, it was Julius Bredt who identified camphor’s real structure with a bridged bicyclic ring system. He would spend years validating his findings and correcting the errors of his predecessors.
Bredt’s famous rule originated from his in-depth camphor studies, as he was not able to synthesize certain derivatives as planned. One of the first examples was this anhydride with a double bond at the bridged position. Elimination of a brominated precursor, or alternatively a condensation reaction of an unsaturated precursor did not prove fruitful. After many synthetic attempts and encountering more quote unquote impossible olefins, Bredt formalized his guideline: bridgehead atoms of bridged rings cannot engage in double bonds.
Ring Strain in Bredt’s Rule
In his seminal 1924 paper, Julius Bredt had already correctly stated that ring strain is behind the apparent impossibility of these olefin compounds. But what is the source of that ring strain, and when is this rule really valid? And how strained are these forbidden rings really?
So we all know that a normal olefins are completely flat with a co-planar arrangement of substituents. However, involving a bridgehead creates an (E)-alkene where a substituent on one carbon is connected to a substituent on the other carbon.
If there are only few carbon atoms in this bridge, they would need to stretch out to ridiculous proportions to accommodate for this planar geometry. So, the molecule tries to find a middle ground regarding ring strain and distortion. This means our double bond is not coplanar but actually twisted. This twisting leads to a suboptimal overlap of the p-orbitals in the pi-bond, which is also destabilizing.
Speaking of p-orbitals, remember how normal alkenes are sp2 hybridized? In the anti-Bredt olefin, we see significant rehybridization to sp3 character as the carbons are pyramidal. This reorientation is key to boost the bond order to a surprisingly high value of 1.86. We have good evidence to believe this is a legit double bond and not something like a di-radical – more on this later.
We have seen why this system is destabilized or strained, but how strained is it? You might know that some thermodynamics shenanigans called homodesmotic equations can help out. Essentially, we are comparing the total strain energies of the saturated and unsaturated molecules to evaluate the pure strain contribution of the anti-Bredt double bond.
Such computations were established in the 1980s, as physical organic chemists tried to understand the limits of these olefins based on their size and strain energy. Here it becomes useful to quantify the size of a bridged ring by simply adding up the number of atoms in the bridges (referred to as S).
The pioneers found that the ring strain estimation can pretty accurately predict if an olefin is stable at room temperature, if it’s olefin observable only at lower temperatures, or if it’s so unstable that we can’t form it or observe it at all.
Key takeaway: Small bridged rings are very strained while larger bridged rings can increasingly accommodate the desired olefin geometry without all the energy penalties. This is why Bredt’s rule loses its validity in larger rings!
Quick knowledge test
Let’s briefly check whether you understood the concept or not. A typical exam question goes along the lines of: Does the alkene shown below violate Bredt’s rule? Look at the four olefins below and identify if they violate Bredt’s rule. Why (not)?
First examples of Anti-Bredt Olefins
The quest to find such larger rings (which don’t obey Bredt’s rule) started in the mid-20th century, with notable efforts from chemists Prelog and Ruzicka. Over time, chemists figured out that such rings even occur in natural products. Prominent examples are the anti-cancer compound taxol or the CP molecules which have a ring size S = 8.
Over the last decades, chemists brought forward clear evidence for these anti-Bredt olefins, even ones with S values below 7. Most of these used trapping experiments to capture short-lived anti-Bredt olefins, like 1-norbornene. In this case, the chemists ran a lithium-halogen exchange reaction of a di-halo precursor in the presence of furan. As they isolated the corresponding Diels-Alder adducts, it seems reasonable to assume that an anti-Bredt olefin with S = 5 was formed and very rapidly intercepted by furan.
Even more impressive is the case of this anti-Bredt olefin with S = 7. Considering the delicate nature of the product, it seems really ironic that the chemists made it by brute-force pyrolyzing this quaternary ammonium precursor through a Hofmann elimination. They also did some nice trapping, but even managed to get NMR data at -80 °C. This is really remarkable and only case to date of a theoretically unstable anti-Bredt olefin being observed directly. Evidently, using the total ring size S as proxy for stability does not work every time.
“Solution to the Anti-Bredt Olefin Synthesis Problem”?
So, we see that the question is not if anti-Bredt olefins can be made, but rather if there are more practical and useful approaches than what we know already. We might not be content with having to incinerate our molecules to get crappy yields of product in a soup of unwanted side products (e.g., due to their low stability, anti-Bredt olefins can intramolecularly rearrange).
Recent research by Garg used fluoride-mediated elimination to create the anti-Bredt olefins, but found that the relative stereochemistry was critical. You see, the precursor with an equatorial silyl group proved unreactive even under forcing conditions while its diastereomer was more useful. By computing the structures, the chemists realized that the reactive diastereomer features a relatively smaller angle between the silyl electrofuge and the leaving group. Because this is a syn-elimination that requires overlap of the carbon silicon sigma and the carbon-oxygen sigma star orbitals, a narrow angle is better. The equatorial diastereomer has a very large angle and low overlap, suppressing any reactivity.
The chemists optimized the conditions by using anthracene as a trapping agent. They found success with a low temperature option using the classic fluoride source TBAF, or alternatively a high-temperature option using a slow-release source of TBAF which helps to control the reaction. The high yield is remarkable as you would expect 16 hours at 120 °C to cause quite some damage. Well, it does not, and the experimental procedure is pretty simple.
So what’s the scope of this method? Other trapping agents were less efficient but still a major improvement to what we knew before – we have furan and its aromatic friends, as well as 1,3-dipoles. They also saw good breadth in terms of anti-Bredt olefins, with some larger and functionalized rings being tolerated. If you are paying attention, you will have noticed that the figure reports diastereomeric ratios of products. But what stereoselectivity does this even refer to?
Our anti-Bredt olefin is actually one specific diastereomer with the hydrogen retaining a pseudo-axial position after the initial elimination.
Remember how we said that instead of a coplanar geometry, we said the system is twisted to a significant degree? Well, the epi diastereomer with a pseudo-equatorial hydrogen would be even more distorted with larger twisting angles and pyramidalization. This elevates its energy and explains why we don’t see any cycloaddition adducts with an equatorial hydrogen at this carbon.
What happens if we use an olefin precursor with just one chiral center?
Here, the chemists synthesized the symmetrical [2.2.2] ring with high enantiometric excess through separation of diastereomeric derivatives. We’ve already seen the diastereoselectivity and stereospecificity of the reactions so it shouldn’t be too surprising: The chirality is fully transferred to the cycloadduct. This once again suggests a concerted elimination step and a chiral alkene with high barrier to racemization and a low, if any, diradical character of the anti-Bredt olefin.
Conclusion
So, we’ve known about small anti-Bredt olefins for a long time. This means Bredt’s rule was already plenty broken in the past. The new research confirms the current interpretation of Bredt’s rule: small olefins are unstable but not necessarily impossible to form. However, the new research adds very practical experimental methods and deeper computation understanding. Thus, anti-Bredt olefin intermediates might get into reach of synthetic chemists. However, the structures of the trapping products are not common so the utility is currently debatable. Maybe, this will encourage more rule-breaking in other areas. Only time will tell!
References on Bredt’s Rule
Generation of strained alkene by the elimination of .beta.-halosilane. On the nature of the double bond of a bicyclo[2.2.2] bridgehead alkene | JACS 1977, 99, 936
Did you know that carbon was allegedly found to engage in six bonds? The strange hexamethylbenzene dication has attracted significant attention due to structure featuring a hyper-coordinated carbon. But can carbon be hypervalent?
Interactive 3D model of the hexamethylbenzene dication
To add more confusion, this system is apparently like a transition-metal complex – though no metal can be found. If you want to make this at home, you just need some equally cursed Dewar benzene, one of the strongest super-acids we know, some hydrofluoric acid for good measure – and the utmost care to not turn your reaction into black tar. Here, we explain if this cursed molecule is an exception to the octet rule.
Octet rule definition
To refresh our memory, elements in the second period typically form compounds that satisfy the Octet rule which states that atoms strive to surround themselves with 8 valence electrons or 4 bond equivalents. However, once we go to lower periods, larger elements start to misbehave. For example, the halogen fluorine can’t expand its octet, but hypervalent iodine compounds like the Dess-Martin periodinane oxidant are common. Similarly, nitrogen is limited by the octet rule but it’s heavier and strange cousin antimony can form more than the four permissible bonds.
But what about carbon, the favorite element of organic chemists? Does carbon follow the octet rule?
We have previously discussed several exotic carbon species, like carbon 2+. This compound is an exception to the Octet rule but it’s hypovalent, not hypervalent. It’s made by oxidizing an already electron-deficient carbene, leaving it with only four instead of the desired eight valence electrons. To pick another example, we’ve also talked about carbones when looking at a dual Diazo-Wittig reagent. This stabilized bis-ylide looks very strange but is not out of order if you count its valence electrons.
CH5+ (Methanium) and 2-Norbornyl cations
To understand other compounds where carbon appears to have more than four bonds, we need to watch out for non-classical bonding. The simplest case, if you can call it simple, is protonated methane or CH5+. On first sight, you might think this is a pentavalent carbon – but beware, as carbon only has four available orbitals. The solution is that instead of having 5 2-center 2-electron bonds, two hydrogens are bound as H2 and partially share their sigma bond with carbon in a 3-center 2-electron bond. This means CH5+ looks more like H2 coordinated to CH3+.
Because of their elusive nature, such non-classical cations were typically not isolated and just observed with methods like spectroscopy. One major breakthrough was in 2013, when a team of chemists finally managed to isolate the 2-norbonyl cation. It was long debated whether this was a rapidly equilibrating classical cation or rather non-classical, penta-coordinated carbon. The chemists cleverly used bromide abstraction from this precursor to generate the cation.
The byproduct bromoaluminate anion nicely stabilizes this species, allowing the growth of, as they put it, slightly brownish but nearly colourless crystals. These were extremely labile towards normal atmosphere and room temperature, so the x-ray analysis had to be highly controlled at low temperatures of just 120 and even 40K.
Still, they somehow made it work and confirmed the non-classical, symmetrical structure. The C-C double bond interacts with the primary carbocation through a 3-center 2-electron bond. As you would expect, this interaction is longer than a normal C-C single bond.
The Hexamethylbenzene dication
So compared to the disputed norbornyl cation, the hexamethylbenzene dication received much less attention. The cursed structure was however directly proposed upon its first synthesis 50 years ago, which I find quite impressive, based on just NMR and some ancient computational methods. Still, it wasn’t clear which structure was present – and we still need to answer if the carbon is truly hypervalent here.
Fast forward to 2017, chemists finally rose to the challenge. Similar to the norbornyl cation, the key task was to stabilize the cation in a crystalline matrix that allowed characterization via x-ray analysis.
Synthesis of a hexa-coordinated carbon
You might think that the simplest way of generating the cation is to take hexamethylbenzene and oxidize the living hell out of it. This would be too easy for such an epic compound, so it doesn’t work. Instead, the strained, high-energy Dewar benzene isomer is more susceptible to the intended manipulations.
It looks strange itself but can be made by bicyclo-trimerization of alkynes through [2+2] additions with aluminium trichloride as a Lewis acid. Oxidizing one of the double bonds is pretty simple, with one equivalent of perbenzoic acid delivering the epoxide. If we count the atoms, we simply need to remove O2- to get to the correct sum formula.
Magic Acid for O2- abstraction
This calls for magic acid. This is a mixture of fluorosulfuric acid, a super acid with a pKa equivalent of -10, and the hypervalent antimony pentafluoride. This Lewis acid forms very stable adducts with anything that has electron pairs. This leads fluorosulfuric acid to react with itself, generating an acid so devilish it can protonate the C-H bond in methane to give CH5+, the compound we’ve talked about earlier.
The reaction is performed at extremely low temperatures given the dication instability. It requires masterful rinsing of frozen dewar benzene epoxide with the super acid mixture. If you swirl too fast – whatever that means – your precious reactants will turn into black tar. You know what we are missing? Of course, after forming the dication, we apparently need to add an excess of anhydrous HF to crystallize it out after some days at -80 °C. I like how the authors wrote ‘in the aftermath’, indicating this reaction is a wicked battle that will require attempts to get it just right.
But how does this work mechanistically? It’s just guessing, but likely, the protonation of the epoxide triggers a rearrangement from the Dewar benzene framework into the bridged five-membered ring. The positive charge ends up at the top of the pyramid, stabilized by a 3 center 2 electron bond, reminiscent of the 2-norbornyl example we covered.
This intramolecular rearrangement is very fast, but remember we are still in a soup of excess super-acid. Thus, another protonation finally gets rid of the oxygen, giving another carbocation. In similar fashion, we again oxidize the top carbon and create a cyclopentadiene at the bottom of the pyramid. This dication crystallizes with an entourage of two SbF6 anions and one molecule of fluorosulfuric acid.
Crystal Structure of the Hexamethylbenzene dication
The apical carbon binds the methyl group on the top and seems to interact with all five carbons of the pyramid base at a distance of roughly 1.7A. This is quite a bit longer than normal C-C single bonds, so you should not be surprised that these are not fully-fledged, classical two-electron bonds.
Based on orbital computations, the multi-center p-electron interactions can be seen as having bond orders of roughly one half each. In total, the apical carbon experiences just below four bond equivalents. This means that just because it’s hexa-coordinated does not mean its hexavalent.
By the way, the carbon NMR spectrum is pretty interesting – we see the ring carbons, the apical carbon, the five methyl groups at the base and at a negative chemical shift, the top methyl group which is extremely shielded. The chemists initially proposed that the system corresponds to a cyclopentadiene cation interacting with an ethyl-ylidene cation through a total of six electrons in the pyramid.
However, subsequent research showed that there might be more than meets the eye here. Based on some fancy computations of effective oxidation states, they believe the cyclopentadienyl ligand is not cationic. Instead, it’s an aromatic cyclopentadiene donor anion and the top group is actually a formal methyl cation . This means the central carbon behaves like a transition metal with two modes of coordination. On one hand, it binds a donor anion as a Lewis acid, and on the other it donates an electron pair to the acceptor cation.
Overall, this is an awesome system that remarkably, is stable enough for crystallization and x-ray analysis. It definitely adds to the list of strange carbon species.
The Pentagonal-Pyramidal Hexamethylbenzene Dication: Many Shades of Coordination Chemistry at Carbon | Chem. Eur. J. 2018, 24, 12340
On the Aromaticity and 13C-NMR Pattern of Pentagonal-Pyramidal Hexamethylbenzene Dication [C6(CH3)6]2+: A {C5(CH3)5}−–{CCH3}3+ Aggregate | Chemistry 2021, 3, 1363
Crystal Structure Determination of the Nonclassical 2-Norbornyl Cation | Science 2013, 341, 62
The Diels Alder reaction is a [4+2] cycloaddition reaction between a conjugated diene and a dienophile. Here we explain its mechanism, orbitaltheory, stereoselectivity and endo rule using 3D models and examples.
3D model: Diene (top – ignore the tilt, it’s drunk) and dienophile (bottom) approaching for a Diels Alder reaction. Can you see the HOMO and LUMO orbital overlap?
Summary
1. Mechanism: The Diels Alder reaction is a concerted [4+2] cycloaddition of a diene in its s-cis conformation with a dienophile.
2. Selectivity: Diels Alder reactions can be regioselective and stereoselective. As there are no intermediates, they are also stereospecific.
3. Substituent effects: Normal electron demand DA reactions feature a nucleophilic, electron-rich diene reacting with an electrophilic, electron-poor dienophile. Resonance structures explain regioselectivity, while secondary orbital overlap explains the endo rule.
4. Orbitals: In the normal electron demand DA reaction, the diene’s HOMO reacts with the dienophile’s LUMO. Higher reactivity (substituents, Lewis acids) arises from smaller energy gap between those interacting orbitals.
5. Hetero Diels Alder reactions: Involve reactants with non-carbon heteroatoms
6. Inverse electron demand Diels Alder: Feature electron-poor dienes and electron-rich dienophiles.
In case you are working on retrosynthesis problems, the DA is often the easiest way to synthesize six-membered rings!
1. reaction mechanism
As the Diels Alder is a pericyclic reaction, it proceeds in one step / concerted. The diene, containing two conjugated alkenes, reacts with a dienophile which is usually an alkene or alkyne (the name means something that likes dienes; sorry, chemistry is strange). Both reactants are flat (double bonds and diene is conjugated) and approach each other in a face-to-face manner.
The driving force of the Diels-Alder reaction is the formation of two σ-bonds (σ-bonds are stronger than π-bonds) in the cyclohexene product. The four atoms linked to these sigma bonds all change from sp2to a sp3 hybridization in the product.
Note that the cycloaddition initially gives the cyclohexene in its boat conformation. Try to really visualize how the transition state. If the dienophile approaches from the bottom (as drawn here; the alternative approach from the top is equally likely for achiral molecules), we form the boat conformation where the hydrogen that was previously pointing away from the diene, the Hexo, “looks up” (wedge in the product). As we see below, the reaction is stereospecific so both Hexo and Hendo in the diene retain their relative orientation (i.e., they look the same way in the product).
The diene has to be in the s-cis conformation to react (bonds form simultaneously).
Unlike π-bonds in alkenes, the single bond can rotate and thus change its conformation. However, given the steric interactions that the CH2 groups experience when they are on the same same of the bond, the s-cis conformation is higher energy than s-trans. The energy difference is ~2.9 kcal/mol [ref 1], so only ~1-2% of butadiene molecules are actually in the s-cis form at a given time.
Note that some dienes like cyclopentadiene are extremely reactive because they are locked in the s-cis conformation.
By the way, we do not say [4+2] because 4 electrons from the diene react with 2 electrons from the dienophile. This is indeed the case – the transition state can actually be considered to be aromatic – but [4+2] refers to the number of atoms involved. For instance, there are so-called [3+2] dipolar cycloadditions which also operate with 6 electrons in total, but only 5 atoms are involved.
Chemistry would be so easy without substituents! By adding substituents on the diene and dienophile, we start to observe reaction selectivity and increased or decreased reactivity. Let’s first understand this in theory, and then check out the specific substituent effects.
2. Selectivity in Diels Alder reactions
Diels Alder reactions are usually regioselective, stereospecificand stereoselective.
Depending on which atoms of the diene and dienophile make contact, we can get two regioisomers (A is either next to E, or A is next to F). The actual regioselectivity is driven by the nature of the substituents (see below).
Because 1) there are no intermediates in the Diels Alder, and 2) both σ-bonds are created from the same side/face, the relative orientation of substituents is fully retained in the product. This is why the reaction is stereospecific.
Beyond the different orientation leading to regioselectivity, the dienophile can also be oriented outside or inside of the diene. This again gives two different products, exo and endo. Note that the example here shows just one regioisomer (it does not show products where A is next to F).
3. Substituent effects in diels alder reactions
Substituents critically influence selectivity and reactivity of Diels Alder reactions.
Normal electron demand DA reactions are cases where the diene is electron-rich, while the dienophile is electron-poor. Typically, the diene has electron-donating groups (EDG; e.g., -OMe) while the dienophile has electron-withdrawing groups ( EWG; e.g., -CO2Me). The majority of reactions follow this polarity, so let’s focus on such examples first. The opposite happens in inverse electron demand DA reactions (explained further below).
The regioselectivity of Diels-Alder reactions can be predicted by considering potential resonance structures of the two reactants. ➡️ The most nucleophilic position of the diene will prefer to add to the most electrophilic position of the dienophile.
Diels Alder reactions can give endo (kinetically preferred) or exo products (thermodynamically more stable). ➡️ If the dienophile carries substituents with π-electrons, the endo product is usually preferred. (explanation requires orbital theory, see below) ➡️ If you are struggling with drawing endo versus exo, draw the transition state in 3D (e.g., as here, diene adding from the top) and draw the resulting boat conformation to compare which groups point which way.
4. Frontier Molecular Orbital (FMO) Analysis
The Diels Alder reaction, just like many other reactions, can be explained by purely looking at the so-called frontier molecular orbitals (FMO). These are the HOMO (highest occupied molecular orbital; filled) and LUMO (lowest unoccupied molecular orbital; empty) of the π-system.
The Diels Alder reaction is exergonic (ΔG<0) – the Hammond postulate tells us that such a reaction has a relatively “early” transition state that looks similar to the starting materials. Thus, the orbitals of the starting materials are a good “proxy” for what will happen in the transition state.
In case you cannot construct the MO diagram for butadiene or ethylene, I refer you to other web sources that explain this.
In the normal electron demand scenario, the HOMO of the diene reacts with the LUMO of the dienophile. Notice how the orbital lobes (positive and negative, grey and black) that overlap have the same phase.
The closer the orbitals in energy (small HOMO-LUMO energy gap), the easier the reaction! This is because when orbitals of similar energies overlap, they experience a higher stabilization of the newly formed orbitals. We want as high of a stabilization as possible, as this decreases the energy of the electrons that we use to populate the new molecular orbital.
Electron-donating groups raise the energy of all molecular orbitals; electron-withdrawing groups decrease the energy of all molecular orbitals.
By having an electron-rich diene and an electron-deficient dienophile, we lower the HOMO-LUMO energy gap (new green line). This means the stabilization experienced by orbital interaction for the purple electrons will increase (they will go down in energy, which is good). This effect is additive – the more substituents we add, the easier and faster the reaction.
The stereoselectivity and endo rule can also be explained through orbital considerations. If the dienophile is oriented endo, the π-orbitals of the carbonyl group (or any other group that has π-system) can interact with the “central” diene orbitals. This lowers the transition state (TS) of the endo approach compared to the exo approach – and thus, the two substituents in our example above point in the same direction.
This might be a bit clearer from the 3D model here. Ignore the non-planarity of the butadiene (my orbital computation was trolling me).
Substituents do not only influence the energy of molecular orbitals, they also change the orbital coefficients at the different atoms. A larger coefficient means that an orbital is relatively more localized on a given atom – e.g., for the HOMO, the atom with the largest coefficient is most nucleophilic. This explains the regioselectivity of Diels Alder reactions more truthfully than the resonance structures which we have mentioned above. However, because this article is not intended to be fully exhaustive, we will not explain coefficients in detail here.
For dienophiles with basic sites on electron-withdrawing groups, Lewis acids can function as catalysts and increase the regio- and stereoselectivity. Coordination of the Lewis acid makes the dienophile even more electron-deficient, and increases the relative polarization (and orbital coefficient).
5. Hetero Diels Alder reaction
DA reactions are not limited to only carbon atoms. Both the diene or the dienophile can feature non-carbon heteroatoms. An example for such hetero Diels Alder reactions is the cycloaddition above with an aldehyde as the dienophile. This works well with the normal electron demand scenario as the carbonyl π-system has a low-energy LUMO.
6. Inverse electron demand Diels Alder
Inverse electron demand Diels Alder reactions are cases where the diene is electron-poor, while the dienophile is electron-rich. Now, the HOMO and LUMO are mapped the other way around: the electron-rich dienophile is nucleophilic and has a high-energy HOMO – while the electron-poor diene is electrophilic and has a low-energy LUMO.
Typically, inverse electron demand reactions as part of multi-step cascade reactions.
Draw the correct product for the given Diels Alder reaction!
So the heteroaromatic ring clearly has a diene, but where is our dienophile [ref 2]? The first step of the reaction is a condensation-enamine formation of the amine with our ketone. The enamine is an electron-rich olefin, so a perfect dienophile in the inverse electron demand scenario. To figure out which regioisomer will be preferred, we need to consider which resonance structures are stabilized by the EDG and EWG, respectively.
After the first Diels Alder reaction, we have a twist: The N=N group can be kicked out in a reverse or retro Diels Alder reaction. It works the exact other way around, with the very stable N2 driving the reaction forward (irreversible step). The last step is aromatization to the much more stable product after elimination of pyrrolidine. I will work on a separate article on the retro Diels Alder reaction in future.
If you learned something, feel free to check out my page or my videos!
Diels Alder Reaction: Exemplary procedure [ref 3]
To a Schlenk tube charged with 3,5-dibromo-2-pyrone [diene] (20, 561 mg, 2.21 mmol) in 7 mL of toluene was added alkenyl boronate [dienophile] (600 mg, 1.70 mmol) at rt. The resulting mixture was heated in an oil bath at 110 °C for 3 days. The reaction mixture was cooled to rt, concentrated in vacuo, and chromatographed (20:1 hexane/EtOAc → 5:1 hexane/EtOAc) to give cycloadduct as a white solid (720 mg, 70%).
Diels Alder reaction references
[1] Thermodynamics of conformational change in 1,3-butadiene studied by high-temperature ultraviolet absorption spectroscopy | Philip W. Mui, Ernest Grunwald | J. Am. Chem. Soc. 1982, 104, 24, 6562–6566
[2] Inverse electron demand Diels-Alder reactions of heterocyclic azadienes: formal total synthesis of streptonigrin | Dale L. Boger, James S. Panek | J. Am. Chem. Soc. 1985, 107, 20, 5745–5754
[3] Total Synthesis of (±)-Clivonine via Diels–Alder Reactions of 3,5-Dibromo-2-pyrone | Cheng-dong Wang, Qinyang Chen, Seunghoon Shin, Cheon-Gyu Cho | J. Org. Chem. 2020, 85, 15, 10035–10049
The Wittig reaction is an olefination reaction in organic chemistry. Let’s explain its mechanism and stereoselectivity using some examples and 3D models!
👀 Interactive 3D model of methylenetriphenylphosphorane (simplest Wittig reagent)
Wittig reaction mechanism
Step 1 – Wittig reagent generation: Every Wittig reaction is based on a carbonyl compound and a Wittig reagent. This is a phosphonium ylide species that can be drawn in two resonance structures: the neutral phosphorane structure or the ylide structure. Chemists generate this species by deprotonating a precursor, a phosphonium (positive charged phosphorous) salt, with strong base. The choice and strength of the base depends on the stabilization of the Wittig reagent (see below). A Wittig reagent that can stabilize the negative charge through other groups can be formed by using a milder base.
How does this work in the laboratory? To avoid cross-reactions (you might know that carbonyls can also be deprotonated by strong bases), chemists add the carbonyl to the reaction only after the base was used to deprotonate the phosphonium. At the very bottom, you can find two exemplary procedures. Some Wittig reagents are so stable that they can be isolated.
Step 2 – Olefination: The ylide/phosphorane is very nucleophilic at the carbon, so it can intermolecularly attack the electrophilic carbonyl carbon. At the same time, the oxygen is nucleophilic and attacks the electrophilic phosphorous. This step likely occurs in a one-step [2+2] addition[see reference 1]. The four-membered ring product is called an oxaphosphetane.
The ring can fragment the ‘opposite way that it was made’ – releasing triphenylphosphine oxide (the driving force) and our alkene product. This is why the Wittig reaction can be classified as an olefination reaction.
The orientation of substituents at the oxaphosphetane determines the stereoselectivity of the reaction. There are two chiral centres on the ring – so cis and trans diastereomers are formed. The stereochemistry of the oxaphosphetane translates into the product following the retro-[2+2]. Cis leads to (Z), trans leads to (E). Here, we note that because the Wittig reagent is a so-called non-stabilized ylide, the (Z) olefin is preferred (see below).
What type of betaine is this? What diastereomer do you expect in the product?
What is the Driving force in the Wittig Reaction?
This is a common question for students. The driving force of the Wittig reaction is the oxidation of triphenylphosphine to form triphenylphosphine oxide or Ph₃P=O. This new phosphorus-oxygen double bond is very strong, making its formation highly favorable from a energetic (thermodynamic) standpoint. You could say there is also a kinetic driving force as the ylide reagent is highly nucleophilic and thus, reactive with the electrophilic carbonyl starting material.
Stabilized ylides and non-stabilized ylides
Depending on the substituents attached to the α-carbon (next to the phosphorous), ylides are categorized into stabilized or non-stabilized ylides. The stability refers to the ability of the substituent to stabilize the negative charge.
Stabilized ylides have electron-withdrawing groups (EWGs) attached to the α-carbon. These can include carbonyls like esters, or nitriles. The mesomeric effect of electron-withdrawing groups stabilizes the negative charge through resonance. This makes the Wittig reagent less reactive but more selective, usually favouring the more thermodynamically stable (E)-alkene product. These Wittig reactions can operate at higher temperatures.
Non-stabilized ylides lack such electron-withdrawing groups and feature alkyl groups. Due to the lack of stability (there is no resonance with alkyl groups), non-stabilized ylides are more reactive. They usually form the kinetically favoured (Z)-alkene product. To maintain stability and the kinetic selectivity, reactions with these ylides are performed at low temperatures.
Semi-stabilized ylides aryl or alkenyl substituents. They fall between the stabilized and non-stabilized ylides and their stereoselectivity is typically poor, leading to similar (E) and (Z) alkene mixtures.
Question for you: Which type of ylide is easier to form by deprotonation? What are their rough pKa values in DMSO?
➡️ Stabilized ylides are more easily formed by deprotonation as they are by definition compounds that stabilize the negative charge on carbon. This means the respective conjugative acid has a lower pKa value (i.e., is more acidic).
Is the Wittig Reaction Concerted or Step-wise?
Over much of its history, the Wittig reaction has been described as a stepwise ionic process. Instead of concerted (one-step) [2+2] cycloaddition, it was assumed that the addition to the carbonyl proceeds step-wise, forming a betaine intermediate. However, modern research suggests that the cycloaddition is more likely [ref 1].
But hey, if your course teaches you the step-wise one, just write that one. Rather get full marks than trying to be right and a smart ass 🙂
Wittig Reaction: Advanced Example [Ref 2]
Here’s a final question (2nd year undergrad level):
What is the mechanism of this two-step sequence? Is the ylide stabilized?
The answer is given below to avoid spoilers.
If you liked this post, feel free to check out other articles on my page or my educational videos!
Under an N2 atmosphere, methyltriphenylphosphonium bromide (40 mg, 0.113 mmol) was suspended in dry THF (1 mL) in a Schlenk tube at 0 °C. BuLi (2.5 M in hexane, 45 μL, 0.113 mmol) was added dropwise. After stirring for 30 min at this temperature, the mixture was cooled to –78 °C, compound 20 (22 mg, 0.057 mmol) was added dropwise as a solution in THF (1 mL). The reaction was continued at the same temperature for 1 h and then at 0 °C for 30 min. Water (1 mL) was added to quench the reaction and the mixture was extracted with CH2Cl2 (3 × 2 mL). The organic extracts were then combined, dried with anhydrous Na2SO4, filtered, concentrated, and purified with by chromatography on silica gel (pentane/EtOAc, 2:1) to give compound 21 (22 mg, >99 % yield).
To a solution of aldehyde (+)-17 (1.40 g, 2.81 mmol, 1 equiv) in dry THF (35 mL) was added crystalline (α-carbomethoxyethylidene)triphenylphosphorane (1.96 g, 5.62 mmol, 2 equiv). The reaction mixture was stirred and heated from room temperature to 50 °C for 2 h under argon. The solvent was then removed under reduced pressure. The residue was purified by flash chromatography (PE/EtOAc 4/1) to give the ester (+)-18 as a white solid (1.31 g, 82%).
Wittig Reaction answer [Ref 2]:
The advanced example above is a Wittig reaction, followed by a Claisen rearrangement. The latter is a [3,3]-sigmatropic rearrangement, a type of pericyclic reaction. It converts allyl vinyl ethers to γ,δ-unsaturated carbonyls. This is an interesting transformation as it looks like it preserves the aldehyde, but moves it one carbon further away from the aromatic ring.
The ylide substituent (-OR) is a mixture of electron-withdrawing (sigma-effect) and electron-donating (mesomeric effect). Thus, one can argue the ylide is semi-stabilized. This is reflected in the rather close ratio of (E) to (Z) alkene after the Wittig. Fortunately, the mixture is not an issue as the Claisen rearrangement gives the same product regardless of diastereomer in this system.
[3] Concise Total Synthesis of Dihydrocorynanthenol, Protoemetinol, Protoemetine, 3-epi-Protoemetinol and Emetine | Shuangzheng Lin, Luca Deiana, Abrehet Tseggai, Armando Córdova | European Journal of Chemistry 2012, 2012, 2, 398-408
[4] Total Synthesis of (−)-Chanoclavine I and an Oxygen-Substituted Ergoline Derivative | Jia-Tian Lu, Zi-Fa Shi, Xiao-Ping Cao | J. Org. Chem. 2017, 82, 15, 7774–7782
