IChO 2016 · Theory problem 7

Synthesis of dolasetron mesylate

This question looks at the synthesis of dolasetron mesylate, Z (shown right), a drug sold under the tradename Anzemet and used to treat post-operative nausea and vomiting.

Figure 1

The synthesis begins as shown below.

Figure 2

First cyclic compound A is made, which contains \(\ce{C}\), \(\ce{H}\), and \(\ce{O}\) only. Compound G is achiral and can be prepared directly from D using ozone under reductive conditions, or via stereoisomers E1 and E2 using \(\ce{OsO4}\), or via stereoisomers F1 and F2 using the peracid shown.

7.1 Determine the empirical formula of G from the percentage masses given.

7.2 Give the structures of A, B, C, D, E1, E2, F1, F2 and G.

Compound G is used in the next stage of the synthesis, under buffered conditions, to form H (as a mixture of two achiral diastereoisomers). Reduction of H with \(\ce{NaBH4}\) gives alcohol I (as a mixture of four achiral diastereoisomers). I reacts with acidified dihydropyran to form J (as a mixture of even more diastereoisomers). J is then treated first with t-butoxide base, then refluxed with acid before finally extracting under weakly basic conditions to form K as a mix of two diastereomers, K1 (major product) and K2 (minor product). These could be separated, and K1 was used in the final stages of the synthesis.

Figure 3

7.3.1 Give the structures of H, I, and J. There is no need to show the different diastereoisomers formed.

7.3.2 Give the structures of diastereoisomers K1, and K2.

In the final stage of the synthesis, L and M react to form intermediate N. N then reacts with K1 to form, after extraction, the neutral amine which gives the target compound upon protonation with \(\ce{CH3SO3H}\).

Figure 4

7.4 Give the structure of N.

Official English text of the problem.

Answers

  • 7.1: The empirical formula of G is \(\ce{C2H3O}\).

  • 7.2: The structures are listed below. The numerical labels within each E or F pair can be interchanged.

Compound Name and stereochemical description SMILES
A Diethyl cyclopent-3-ene-1,1-dicarboxylate CCOC(=O)C1(C(=O)OCC)CC=CC1
B Cyclopent-3-ene-1-carboxylic acid O=C(O)C1CC=CC1
C Cyclopent-3-ene-1-carbonyl chloride O=C(Cl)C1CC=CC1
D Ethyl cyclopent-3-ene-1-carboxylate CCOC(=O)C1CC=CC1
E1 Ethyl 3,4-dihydroxycyclopentane-1-carboxylate; both OH groups and the ester substituent are mutually cis CCOC(=O)[C@@H]1C[C@H](O)[C@H](O)C1
E2 Ethyl 3,4-dihydroxycyclopentane-1-carboxylate; the OH groups are cis to each other and both trans to the ester substituent CCOC(=O)[C@H]1C[C@H](O)[C@H](O)C1
F1 Ethyl 6-oxabicyclo[3.1.0]hexane-3-carboxylate; ester substituent syn to the epoxide oxygen bridge CCOC(=O)[C@H]1C[C@H]2O[C@H]2C1
F2 Ethyl 6-oxabicyclo[3.1.0]hexane-3-carboxylate; ester substituent anti to the epoxide oxygen bridge CCOC(=O)[C@@H]1C[C@H]2O[C@H]2C1
G Ethyl 2-(2-oxoethyl)-4-oxobutanoate CCOC(=O)C(CC=O)CC=O
  • 7.3.1: Stereochemistry is omitted as permitted in the question.
Compound Name SMILES
H Ethyl 9-(2-ethoxy-2-oxoethyl)-7-oxo-9-azabicyclo[3.3.1]nonane-3-carboxylate CCOC(=O)CN1C2CC(=O)CC1CC(C(=O)OCC)C2
I Ethyl 9-(2-ethoxy-2-oxoethyl)-7-hydroxy-9-azabicyclo[3.3.1]nonane-3-carboxylate CCOC(=O)CN1C2CC(O)CC1CC(C(=O)OCC)C2
J Ethyl 9-(2-ethoxy-2-oxoethyl)-7-[(oxan-2-yl)oxy]-9-azabicyclo[3.3.1]nonane-3-carboxylate—the THP ether of I CCOC(=O)CN1C2CC(OC3OCCCC3)CC1CC(C(=O)OCC)C2
  • 7.3.2: Both products are stereoisomers of 5-hydroxy-8-azatricyclo\([5.3.1.0^{3,8}]\)undecan-10-one. “Anti” and “syn” below refer to the OH group relative to the nitrogen bridge of the original bicyclic framework.
Compound Configuration SMILES
K1, major Anti-OH isomer; its alcohol configuration matches the ester-bearing centre in Z O=C1C[N@@]2[C@@H]3C[C@@H](O)C[C@H]2C[C@H]1C3
K2, minor Syn-OH isomer; the alcohol epimer of K1 O=C1C[N@@]2[C@@H]3C[C@H](O)C[C@H]2C[C@H]1C3
  • 7.4: N is the mixed anhydride of 1H-indole-3-carboxylic acid and trifluoroacetic acid: O=C(OC(=O)C(F)(F)F)c1c[nH]c2ccccc12.

Solution

For redox steps whose complete work-up is unspecified, the equations below use balanced organic net reactions. Here \(\ce{[O]}\) denotes one oxygen-atom equivalent and \(\ce{[H]}\) one hydrogen-atom equivalent.

7.1 — Empirical formula of G

For a \(100\ \mathrm{g}\) sample, using conventional atomic molar masses:

Element Mass Atomic molar mass Amount
Carbon \(55.81\ \mathrm{g}\) \(12.01\ \mathrm{g\,mol^{-1}}\) \(4.647\ \mathrm{mol}\)
Hydrogen \(7.03\ \mathrm{g}\) \(1.008\ \mathrm{g\,mol^{-1}}\) \(6.97\ \mathrm{mol}\)
Oxygen \(37.17\ \mathrm{g}\) \(16.00\ \mathrm{g\,mol^{-1}}\) \(2.323\ \mathrm{mol}\)

Dividing by the smallest amount gives

\[ n_{\ce C}:n_{\ce H}:n_{\ce O} = \frac{4.647}{2.323}: \frac{6.97}{2.323}:1 \approx 2.00:3.00:1.00. \]

Therefore,

\[ \boxed{\text{Empirical formula of G}=\ce{C2H3O}}. \]

The molecular structure established below has formula \(\ce{C8H12O4}\), four times this empirical formula.

7.2 — Identification of A–G

Formation of A. The starting materials are diethyl malonate, CCOC(=O)CC(=O)OCC, and \((Z)\)-1,4-dichlorobut-2-ene, ClC/C=C\CCl.

Lithium hydride removes a proton from the methylene group between the two ester carbonyls. The resulting carbon nucleophile displaces chloride from one terminal chloromethyl group. Deprotonation of its remaining malonate proton permits a second, intramolecular substitution at the other chloromethyl group.

The malonate carbon and the four-carbon chain therefore form a five-membered ring, giving A. Both ester groups remain attached to the same ring carbon.

The stoichiometric balance is

\[ \ce{C7H12O4 + C4H6Cl2 + 2LiH -> C11H16O4 + 2LiCl + 2H2}. \]

The supplied \(2.5\) equivalents of \(\ce{LiH}\) exceed the two equivalents required by this net transformation.

A to B. Both esters undergo alkaline hydrolysis. Acidification gives cyclopent-3-ene-1,1-dicarboxylic acid, O=C(O)C1(C(=O)O)CC=CC1. This is a substituted malonic acid and loses one carboxyl group on heating, giving B.

Combining hydrolysis, acidification and decarboxylation:

\[ \ce{C11H16O4 + 2NaOH + 2HCl -> C6H8O2 + 2C2H5OH + CO2 + 2NaCl}. \]

B to C. Thionyl chloride converts the carboxylic acid into its acyl chloride:

\[ \ce{C6H8O2 + SOCl2 -> C6H7ClO + SO2 + HCl}. \]

Thus C is cyclopent-3-ene-1-carbonyl chloride.

C to D. Ethanol substitutes for chloride at the acyl carbon, producing the ethyl ester D:

\[ \ce{C6H7ClO + C2H5OH -> C8H12O2 + HCl}. \]

Pyridine, n1ccccc1, captures the liberated acid as pyridinium chloride, [Cl-].[nH+]1ccccc1:

\[ \ce{C5H5N + HCl -> [C5H6N]+ + Cl-}. \]

D to E1 and E2. Osmium tetroxide adds across the alkene through a cyclic osmate intermediate. Hydrolysis gives syn dihydroxylation: the two OH groups must be on the same face of the ring.

Addition on either face relative to the ester substituent gives:

  • E1: the ester substituent and both OH groups on the same face;
  • E2: both OH groups on the face opposite the ester substituent.

To sketch them, draw a cyclopentane ring bearing the ester at C1 and OH groups at C3 and C4. Give the two OH bonds matching wedges; use a matching ester wedge for E1 and an opposite ester wedge for E2. Both products are achiral.

The organic net balance is

\[ \ce{C8H12O2 + H2O + [O] -> C8H14O4}. \]

D to F1 and F2. The peracid is 3-chloroperoxybenzoic acid, or mCPBA, O=C(OO)c1cccc(Cl)c1. In the stated dichloromethane medium it converts the alkene into an epoxide.

Oxygen transfer from the two faces produces the two epoxide configurations in the answer table. The epoxide oxygen bridge and ester substituent are syn in F1 and anti in F2.

The by-product is 3-chlorobenzoic acid, O=C(O)c1cccc(Cl)c1:

\[ \ce{C8H12O2 + C7H5ClO3 -> C8H12O3 + C7H5ClO2}. \]

Formation of G. Each alkene carbon in D bears a hydrogen. Reductive ozonolysis therefore converts both into aldehydes. Cleavage opens the five-membered ring and gives

\[ \boxed{\ce{OHC-CH2-CH(CO2C2H5)-CH2-CHO}}. \]

This is G, ethyl 2-(2-oxoethyl)-4-oxobutanoate, CCOC(=O)C(CC=O)CC=O.

The central carbon has two identical \(\ce{-CH2CHO}\) substituents, so it is not stereogenic; G is achiral. Its formula is \(\ce{C8H12O4}\), consistent with part 7.1.

The organic net ozonolysis balance is

\[ \ce{C8H12O2 + 2[O] -> C8H12O4}. \]

Periodic acid cleaves the bond between the two hydroxyl-bearing carbons of either E isomer:

\[ \ce{C8H14O4 + HIO4 -> C8H12O4 + HIO3 + H2O}. \]

The epoxides F1/F2 also undergo cleavage in aqueous periodic acid, involving epoxide opening. Their overall balance is

\[ \ce{C8H12O3 + HIO4 -> C8H12O4 + HIO3}. \]

Thus all three routes give the same dialdehyde G.

7.3.1 — Structures and formation of H, I and J

Formation of H. The other reactants are:

  • 3-Oxopentanedioic acid, commonly acetonedicarboxylic acid: O=C(O)CC(=O)CC(=O)O;
  • Ethyl glycinate hydrochloride: CCOC(=O)C[NH3+].[Cl-].

This is a Robinson–Schöpf-type bicyclic aminoketone synthesis, involving two Mannich condensations.

The connectivity can be tracked as follows:

  1. The nitrogen of ethyl glycinate becomes bonded to both terminal aldehyde carbons of G.
  2. The two methylene carbons flanking the ketone of acetonedicarboxylic acid form carbon–carbon bonds to those same aldehyde-derived carbons.
  3. Both carboxylic acid groups of acetonedicarboxylic acid are lost as \(\ce{CO2}\).
  4. The nitrogen retains its \(\ce{-CH2CO2C2H5}\) substituent, and the ester originally present in G remains intact.

The resulting bridgeheads are connected by three paths containing three, three and one internal atoms: the 9-azabicyclo[3.3.1]nonane skeleton of H.

The overall balance, writing the acid equivalent separately from its capture by the buffer, is

\[ \ce{C8H12O4 + C5H6O5 + C4H10ClNO2 -> C15H23NO5 + 2CO2 + 2H2O + HCl}. \]

This reproduces the supplied formula \(\ce{C15H23NO5}\).

The two achiral diastereoisomers differ in the orientation of the ring-bound ester relative to the nitrogen bridge.

H to I. Sodium borohydride reduces the ring ketone to a secondary alcohol while retaining both esters:

\[ \ce{C15H23NO5 + 2[H] -> C15H25NO5}. \]

The two possible alcohol configurations, combined with the two configurations already present in H, account for the four achiral diastereoisomers of I.

The acid and carbonate work-up gives the free amino alcohol shown in the answer table.

I to J. Acid-catalysed reaction with 3,4-dihydro-2H-pyran, C1=COCCC1, protects the alcohol as a tetrahydropyranyl (THP) ether.

The alcohol oxygen bonds to the carbon adjacent to the ring oxygen of the dihydropyran. This creates an acetal carbon and additional stereoisomers.

No small molecule is eliminated:

\[ \ce{C15H25NO5 + C5H8O -> C20H33NO6}. \]

The resulting formula agrees with the stated formula of J.

7.3.2 — Formation and stereochemistry of K1 and K2

Potassium tert-butoxide promotes an intramolecular Claisen, or Dieckmann, condensation between the two ester groups.

Deprotonation of the nitrogen-bound \(\ce{-CH2CO2C2H5}\) group gives an enolate. Its carbon attacks the carbonyl of the ester attached to the bicyclic ring, and ethoxide is expelled. The new carbon–carbon bond constructs the third ring.

A stereochemically unspecified representation of the resulting β-keto ester is:

Ethyl 5-[(oxan-2-yl)oxy]-10-oxo-8-azatricyclo\([5.3.1.0^{3,8}]\)undecane-9-carboxylate,
O=C1C(C(=O)OCC)N2C3CC(OC4OCCCC4)CC2CC1C3.

Calling this transient intermediate Q, its neutral organic balance after protonation is

\[ \ce{C20H33NO6 -> C18H27NO5 + C2H5OH}. \]

Heating with aqueous acid then accomplishes three transformations:

  • Hydrolysis of the remaining ethyl ester;
  • Decarboxylation of the resulting β-keto acid;
  • Removal of the THP protecting group, restoring the alcohol.

THP hydrolysis formally releases 5-hydroxypentanal, O=CCCCCO, which can equilibrate with its cyclic hemiacetal. The combined balance is

\[ \ce{C18H27NO5 + 2H2O -> C10H15NO2 + C2H5OH + CO2 + C5H10O2}. \]

Hence the overall organic balance from J is

\[ \ce{C20H33NO6 + 2H2O -> C10H15NO2 + 2C2H5OH + CO2 + C5H10O2}. \]

Basic extraction furnishes the neutral tertiary amine. The product formula,

\[ \boxed{\ce{C10H15NO2}}, \]

matches the problem.

The ketone in K comes from the ester condensation. The alcohol formed in the earlier borohydride reduction survives protection and deprotection, leaving two alcohol configurations:

  • K1: OH anti to the original nitrogen bridge;
  • K2: OH syn to that bridge.

For the requested drawings, use the same tricyclic cage for both and invert only the alcohol-bearing centre. The target drawing identifies K1: replacing its OH hydrogen with the indole-3-carbonyl group gives the ester orientation shown in Z. Their precise configurations are encoded in the answer-table SMILES.

7.4 — Structure of N and completion of the synthesis

L is 1H-indole-3-carboxylic acid, O=C(O)c1c[nH]c2ccccc12.

M is trifluoroacetic anhydride, O=C(OC(=O)C(F)(F)F)C(F)(F)F.

The carboxylic acid reacts with the anhydride to form the activated mixed anhydride N:

\[ \boxed{\ce{(1H-indol-3-yl)-C(=O)-O-C(=O)-CF3}}. \]

Its SMILES is O=C(OC(=O)C(F)(F)F)c1c[nH]c2ccccc12.

The by-product is trifluoroacetic acid, O=C(O)C(F)(F)F. The balanced equation is

\[ \ce{C9H7NO2 + C4F6O3 -> C11H6F3NO3 + C2HF3O2}. \]

The alcohol oxygen of K1 attacks the indole-derived acyl carbonyl of N, and trifluoroacetate leaves. After extraction, this gives neutral dolasetron:

O=C1C[N@@]2[C@@H]3C[C@@H](OC(=O)c4c[nH]c5ccccc45)C[C@H]2C[C@H]1C3.

The net esterification balance is

\[ \ce{C10H15NO2 + C11H6F3NO3 -> C19H20N2O3 + C2HF3O2}. \]

Finally, methanesulfonic acid, CS(=O)(=O)O, protonates the tertiary amine:

\[ \ce{C19H20N2O3 + CH3SO3H -> [C19H21N2O3]+ + CH3SO3-}. \]

The product is dolasetron mesylate, Z:

CS(=O)(=O)[O-].O=C1C[N@H+]2[C@@H]3C[C@@H](OC(=O)c4c[nH]c5ccccc45)C[C@H]2C[C@H]1C3.

The answers were checked against the official answers published by the organisers.