Selective One-Pot Multicomponent Synthesis of N-Substituted 2,3,5-Functionalized 3-Cyanopyrroles via the Reaction between α-Hydroxyketones, Oxoacetonitriles, and Primary Amines

A one-step, three-component reaction between α-hydroxyketones, oxoacetonitriles, and primary amines gives N-substituted 2,3,5-functionalized 3-cyanopyrroles with complete selectivity in up to 90% isolated yields. The reaction worked on a wide substrate scope under mild reaction conditions (AcOH as a catalyst, EtOH, 70 °C, 3 h). The reaction proceeded with very high atom efficiency as water is the only molecule lost during the reaction. The practicality of the reaction was demonstrated on a large gram scale. The structures of the 3-cyanopyrroles were confirmed by single-crystal X-ray diffraction and NMR; this work provides a general and practical entry to pyrrole scaffolds suitably decorated for the synthesis of various bioactive pyrroles in a concise manner.


Introduction
Pyrroles are a class of nitrogen heterocycles with important applications in the pharmaceutical and material science industries [1][2][3]. Owing to their diverse bioactivities such as anti-inflammatory [4], anti-bacterial [5], anti-tumor [6], and anti-oxidative bioactivities [7], several pyrrole-based drugs have been successfully marketed to treat various conditions ( Figure 1). Additionally, many pyrrole-based drug candidates have shown great promise in treating various conditions ( Figure 1). Therefore, continuous efforts have been invested in designing simple, efficient, and sustainable routes to synthesize suitably functionalized pyrroles with specific "handles" for easy transformation to the final dug compounds. In this regard, special attention is given to synthesizing 3-cyanopyrroles since the cyano handle can easily be converted to -CXN functionality (e.g., X = H, H 2 , O, or simply nothing); this -CXN functionality appears in many drugs and drug candidates ( Figure 1) [5,[8][9][10][11][12].
Previous efforts to synthesize 3-cyanopyrroles include (i) the nickel(II) bis(acetylacetonate)catalyzed reaction between azirines and active methylene compounds reported in 1977 ( Figure 2, (i)) [13]; (ii) the reaction between 1-nitro-1-cyclopropyl ketones and primary amines followed by oxidizing the dihydropyrrole products using 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ) to give the desired 3-cyanopyrroles ( Figure 2, (ii) [14]; and (iii)) the reaction between α-hydroxyketones, malononitrile and primary amines ( Figure 2, (iii)) [15]; however, the first two reactions require extra steps to prepare the complex starting materials, use metals to catalyze the reactions, and are multistep syntheses. The last reaction works when the substituents at the α-hydroxyketones are similar (R 1 and R 1, both either methyl or phenyl, Figure 1, (iii)) and the starting material itself is not considered sustainable. In this reaction, when the substituents are different, several products are formed, which reduces the yields and complicates separation [16]; moreover, these reactions are conducted at high temperatures > 90 • C in toluene. Additionally, other inefficient multistep reactions Previous efforts to synthesize 3-cyanopyrroles include (i) the nickel(II) bis(acetylacetonate)-catalyzed reaction between azirines and active methylene compounds reported in 1977 ( Figure 2, (i)) [13]; (ii) the reaction between 1-nitro-1-cyclopropyl ketones and primary amines followed by oxidizing the dihydropyrrole products using 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ) to give the desired 3-cyanopyrroles ( Figure  2, (ii) [14]; and (iii)) the reaction between α-hydroxyketones, malononitrile and primary amines ( Figure 2, (iii)) [15]; however, the first two reactions require extra steps to prepare the complex starting materials, use metals to catalyze the reactions, and are multistep syntheses. The last reaction works when the substituents at the α-hydroxyketones are similar (R 1 and R 1, both either methyl or phenyl, Figure 1, (iii)) and the starting material itself is not considered sustainable. In this reaction, when the substituents are different, several products are formed, which reduces the yields and complicates separation [16]; moreover, these reactions are conducted at high temperatures > 90 °C in toluene. Additionally, other inefficient multistep reactions introduced the cyano handle on a preformed pyrrole and usually used heavy metals or toxic reagents [17]. These disadvantages encourage the search for a more robust, practical, and sustainable methodology. Earlier, we developed a cascade reaction between aldoses, oxoacetonitriles, and NH4OAc to synthesize N-unsubstituted 3-cyanopyrroles ( Figure 2, (iv)) [18]. Based on our previous experience [18,19], we envisioned that ketoses should also be ideal substrates for the synthesis of functionalized 3-cyanopyrroles ( Figure 1, (iii)). Ketoses are cheap, sustainable materials and can introduce chirality in the pyrrole framework; moreover, their polyhydroxyalkyl chain can easily be converted into several functional groups; it is also ben- Earlier, we developed a cascade reaction between aldoses, oxoacetonitriles, and NH 4 OAc to synthesize N-unsubstituted 3-cyanopyrroles ( Figure 2, (iv)) [18]. Based on our previous experience [18,19], we envisioned that ketoses should also be ideal substrates for the synthesis of functionalized 3-cyanopyrroles ( Figure 1, (iii)). Ketoses are cheap, sustainable materials and can introduce chirality in the pyrrole framework; moreover, their polyhydroxyalkyl chain can easily be converted into several functional groups; it is also beneficial to establish general reaction conditions that work for other α-hydroxyketones (ketoses are, in fact, masked α-hydroxyketones) such as phenacyl alcohols to add diversity to the substituents on the core 3-cyanopyrrole structure (Figure 1, (v)). Another important objective in this work is to achieve complete selectivity while using differently substituted α-hydroxyketones (R 1 = R 1 , Figure 1, (iii)) which has proven to be challenging in previous reports [16].
Herein, we report an AcOH-catalyzed selective one-pot three-component reaction between α-hydroxyketones (ketoses and phenacyl alcohols), oxoacetonitriles, and primary amines to synthesize densely functionalized 3-cyanopyrroles; this reaction gave the desired 3-cyanopyrroles in 53-90% yields and worked successfully on a gram scale. The structures of the 3-cyanopyrroles were confirmed using single-crystal X-ray analysis, and a plausible reaction mechanism was also provided.

Substrate Scope
Initially, the scope of the reaction was investigated by reacting different ketoses, including D-(+)-fructose 1 and isomaltulose (also known as palatinose) 4 with primary amines 3 and 5-12 and oxoacetonitriles 2 and 13, (Table 2) under the optimized condition (Table 1, entry 1). The selected ketoses were representative of mono-and disaccharides, the amines were representative of primary, secondary, unsaturated, aliphatic, aromatic, and heterocyclic amines, while the oxoacetonitriles were representative of aliphatic and aromatic nitriles. Ketoses 1 and 4 reacted selectively to give the N-substituted 2,3,5-functionalized 3-cyanopyrroles 1a-k and 4a-c in 55-86% isolated yields. In particular, 3-cyanopyrroles 4a-c are interesting structures since they combine both a pyrrole moiety and a sugar moiety and are water-soluble [20]. The yield of the 3-cyanopyrroles 1a-k and 4a-c was affected by the type of sugar and the type of the primary amine. Hence, D-(+)-fructose/amines provide higher yields than isomaltulose/amines ( Table 2, entries 1-6 vs. [12][13][14], and aliphatic amines provided higher yields than aromatic amines due to their higher nucleophilicity ( Table 2, entries 1-6 vs. 7-9). All the 3-cyanopyrroles 1a-k and 4a-c were purified by column chromatography using a mixture of MeOH/DCM (5-10% MeOH/DCM for 1a-k, and 15-20% MeOH/DCM for 4a-c). The high polarity of the 3-cyanopyrroles is attributed to the extensive hydrogen bonding in the polyhydroxyalkyl chains. sition products indicated by a darkening of the reaction mixture. Next, several acidic and basic catalysts were examined. Except for ZnCl2, which gave 1a in 46% yield, other acids, including p-toluenesulfonic acid (PTSA) and camphor sulfonic acid (CSA), and bases including triethylamine (Et3N), sodium hydroxide (NaOH), potassium carbonate (K2CO3) and sodium methoxide (NaOMe) did not give the desired product, and the starting materials were recovered back (Table 1, entry 2-8). Changing EtOH solvent to acetonitrile (ACN) or dimethylformamide (DMF) did not improve the yield of 1a (Table 1, entry 1 vs. 9 vs. 10). Further, attempts to reduce the equivalents of AcOH from 1 to 0.5 reduced the yield of 1a dramatically (Table 1, entry 1 vs. 11). Attempts to reduce reaction temperature to 40 °C gave 1a in trace amount (Table 1, entry 12). Therefore, we concluded the optimum conditions to be 1 equivalent of AcOH in EtOH at 70 °C for 3 h (Table 1, entry 1). Table 1. Optimization of the reaction conditions for the three-component synthesis of 3-cyanopyrrole 1a. Table 2. Substrate scope of the AcOH-catalyzed three-component reaction between ketoses 1, and 4, primary amines 3 and 5-12, and oxoacetonitriles 2 and 13 for the selective synthesis of 3-cyanopyrroles 1a-k and 4a-c [a] .

Substrate Scope
Initially, the scope of the reaction was investigated by reacting different ketoses, including D-(+)-fructose 1 and isomaltulose (also known as palatinose) 4 with primary amines 3 and 5-12 and oxoacetonitriles 2 and 13, (Table 2) under the optimized condition ( Table 1, entry 1). The selected ketoses were representative of mono-and disaccharides, the amines were representative of primary, secondary, unsaturated, aliphatic, aromatic, and heterocyclic amines, while the oxoacetonitriles were representative of aliphatic and aromatic nitriles. Ketoses 1 and 4 reacted selectively to give the N-substituted 2,3,5-functionalized 3-cyanopyrroles 1a-k and 4a-c in 55-86% isolated yields. In particular, 3-cyanopyrroles 4a-c are interesting structures since they combine both a pyrrole moiety and a sugar moiety and are water-soluble [20]. The yield of the 3-cyanopyrroles 1a-k and 4a-c was affected by the type of sugar and the type of the primary amine. Hence, D-(+)-fructose/amines provide higher yields than isomaltulose/amines ( Table 2, entries 1-6 vs. [12][13][14], and aliphatic amines provided higher yields than aromatic amines due to their higher nucleophilicity ( Table 2, entries 1-6 vs. 7-9). All the 3-cyanopyrroles 1a-k and 4a-c were purified by column chromatography using a mixture of MeOH/DCM (5-10% MeOH/DCM for 1a-k, and 15-20% MeOH/DCM for 4a-c). The high polarity of the 3-cyanopyrroles is attributed to the extensive hydrogen bonding in the polyhydroxyalkyl chains. Table 2. Substrate scope of the AcOH-catalyzed three-component reaction between ketoses 1, and 4, primary amines 3 and 5-12, and oxoacetonitriles 2 and 13 for the selective synthesis of 3-cyanopyrroles 1a-k and 4a-c. [   Next, the reactivity of phenacyl alcohols 14-17 as the α-hydroxyketone was tested under the optimum conditions (Table 1, entry 1) to further extend the scope of this reaction and examine the selectivity. Additionally, phenacyl alcohols would also give needed diversity to the substituents on the 3-cyanopyrrole core. The reaction between phenacyl Next, the reactivity of phenacyl alcohols 14-17 as the α-hydroxyketone was tested under the optimum conditions (Table 1, entry 1) to further extend the scope of this reaction and examine the selectivity. Additionally, phenacyl alcohols would also give needed diversity to the substituents on the 3-cyanopyrrole core. The reaction between phenacyl Next, the reactivity of phenacyl alcohols 14-17 as the α-hydroxyketone was tested under the optimum conditions (Table 1, entry 1) to further extend the scope of this reaction and examine the selectivity. Additionally, phenacyl alcohols would also give needed diversity to the substituents on the 3-cyanopyrrole core. The reaction between phenacyl Next, the reactivity of phenacyl alcohols 14-17 as the α-hydroxyketone was tested under the optimum conditions (Table 1, entry 1) to further extend the scope of this reaction and examine the selectivity. Additionally, phenacyl alcohols would also give needed diversity to the substituents on the 3-cyanopyrrole core. The reaction between phenacyl Next, the reactivity of phenacyl alcohols 14-17 as the α-hydroxyketone was tested under the optimum conditions (Table 1, entry 1) to further extend the scope of this reaction and examine the selectivity. Additionally, phenacyl alcohols would also give needed diversity to the substituents on the 3-cyanopyrrole core. The reaction between phenacyl Next, the reactivity of phenacyl alcohols 14-17 as the α-hydroxyketone was tested under the optimum conditions (Table 1, entry 1) to further extend the scope of this reaction and examine the selectivity. Additionally, phenacyl alcohols would also give needed diversity to the substituents on the 3-cyanopyrrole core. The reaction between phenacyl Next, the reactivity of phenacyl alcohols 14-17 as the α-hydroxyketone was tested under the optimum conditions (Table 1, entry 1) to further extend the scope of this reaction and examine the selectivity. Additionally, phenacyl alcohols would also give needed diversity to the substituents on the 3-cyanopyrrole core. The reaction between phenacyl Next, the reactivity of phenacyl alcohols 14-17 as the α-hydroxyketone was tested under the optimum conditions (Table 1, entry 1) to further extend the scope of this reaction and examine the selectivity. Additionally, phenacyl alcohols would also give needed diversity to the substituents on the 3-cyanopyrrole core. The reaction between phenacyl Next, the reactivity of phenacyl alcohols 14-17 as the α-hydroxyketone was tested under the optimum conditions (Table 1, entry 1) to further extend the scope of this reaction and examine the selectivity. Additionally, phenacyl alcohols would also give needed diversity to the substituents on the 3-cyanopyrrole core. The reaction between phenacyl Next, the reactivity of phenacyl alcohols 14-17 as the α-hydroxyketone was tested under the optimum conditions (Table 1, entry 1) to further extend the scope of this reaction and examine the selectivity. Additionally, phenacyl alcohols would also give needed diversity to the substituents on the 3-cyanopyrrole core. The reaction between phenacyl Next, the reactivity of phenacyl alcohols 14-17 as the α-hydroxyketone was tested under the optimum conditions (Table 1, entry 1) to further extend the scope of this reaction and examine the selectivity. Additionally, phenacyl alcohols would also give needed diversity to the substituents on the 3-cyanopyrrole core. The reaction between phenacyl Next, the reactivity of phenacyl alcohols 14-17 as the α-hydroxyketone was tested under the optimum conditions (Table 1, entry 1) to further extend the scope of this reaction and examine the selectivity. Additionally, phenacyl alcohols would also give needed diversity to the substituents on the 3-cyanopyrrole core. The reaction between phenacyl Next, the reactivity of phenacyl alcohols 14-17 as the α-hydroxyketone was tested under the optimum conditions (Table 1, entry 1) to further extend the scope of this reaction and examine the selectivity. Additionally, phenacyl alcohols would also give needed diversity to the substituents on the 3-cyanopyrrole core. The reaction between phenacyl Next, the reactivity of phenacyl alcohols 14-17 as the α-hydroxyketone was tested under the optimum conditions (Table 1, entry 1) to further extend the scope of this reaction and examine the selectivity. Additionally, phenacyl alcohols would also give needed diversity to the substituents on the 3-cyanopyrrole core. The reaction between phenacyl Next, the reactivity of phenacyl alcohols 14-17 as the α-hydroxyketone was tested under the optimum conditions (Table 1, entry 1) to further extend the scope of this reaction and examine the selectivity. Additionally, phenacyl alcohols would also give needed diversity to the substituents on the 3-cyanopyrrole core. The reaction between phenacyl Next, the reactivity of phenacyl alcohols 14-17 as the α-hydroxyketone was tested under the optimum conditions (Table 1, entry 1) to further extend the scope of this reaction and examine the selectivity. Additionally, phenacyl alcohols would also give needed diversity to the substituents on the 3-cyanopyrrole core. The reaction between phenacyl Next, the reactivity of phenacyl alcohols 14-17 as the α-hydroxyketone was tested under the optimum conditions (Table 1, entry 1) to further extend the scope of this reaction and examine the selectivity. Additionally, phenacyl alcohols would also give needed diversity to the substituents on the 3-cyanopyrrole core. The reaction between phenacyl Next, the reactivity of phenacyl alcohols 14-17 as the α-hydroxyketone was tested under the optimum conditions (Table 1, entry 1) to further extend the scope of this reaction and examine the selectivity. Additionally, phenacyl alcohols would also give needed diversity to the substituents on the 3-cyanopyrrole core. The reaction between phenacyl Next, the reactivity of phenacyl alcohols 14-17 as the α-hydroxyketone was tested under the optimum conditions (Table 1, entry 1) to further extend the scope of this reaction and examine the selectivity. Additionally, phenacyl alcohols would also give needed diversity to the substituents on the 3-cyanopyrrole core. The reaction between phenacyl Next, the reactivity of phenacyl alcohols 14-17 as the α-hydroxyketone was tested under the optimum conditions (Table 1, entry 1) to further extend the scope of this reaction and examine the selectivity. Additionally, phenacyl alcohols would also give needed diversity to the substituents on the 3-cyanopyrrole core. The reaction between phenacyl Next, the reactivity of phenacyl alcohols 14-17 as the α-hydroxyketone was tested under the optimum conditions (Table 1, entry 1) to further extend the scope of this reaction and examine the selectivity. Additionally, phenacyl alcohols would also give needed diversity to the substituents on the 3-cyanopyrrole core. The reaction between phenacyl alcohols 14-17 [21] oxoacetonitriles 2, 13, and 18-20, and primary amines 3, 5, 6, 10, 11, and 21 selectively gave the desired N-substituted 2,3,5-functionalized 3- cyanopyrroles  14a-e, 15a, 16a-b, and 17a-b in excellent 70-90% isolated yields (Table 3). Again, aromatic amines generally gave lower yields than aliphatic amines due to their lower nucleophilicity. 3 -Cyanopyrroles 14a-e, 15a, 16a-b, and 17a-b were purified using silica gel column chromatography using a mixture of 5-35% EtOAc/Hexane. These 3-cyanopyrroles showed the expected lower polarity on silica gel TLC and column chromatography compared to 3-cyanopyrroles 1a-k and 4a-c due to lower hydrogen bonding.  14-17, oxoacetonitriles 2, 13, and 18-20, and primary amines 3, 5, 6, 10, 11, and 21 for the selective synthesis of 3-cyanopyrroles 14a-e, 15a, 16a-b, and 17a-b [a] .  14-17, oxoacetonitriles 2, 13, and 18-20, and primary amines 3, 5, 6, 10, 11, and 21 for the selective synthesis of 3-cyanopyrroles 14a-e, 15a, 16a-b, and 17a-b. [a].   14a-e, 15a, 16a-b, and 17a-b. [a].   14a-e, 15a, 16a-b, and 17a-b. [a].   14a-e, 15a, 16a-b, and 17a-b. [a].   14a-e, 15a, 16a-b, and 17a-b. [a].   14a-e, 15a, 16a-b, and 17a-b. [a].    14a-e, 15a, 16a-b, and 17a-b. [a].   14a-e, 15a, 16a-b, and 17a-b. [a].        Table 3. Cont.   14a-e, 15a, 16a-b, and 17a-

Structural Modification of the Synthesized Pyrroles
The polyhydroxyalkyl chains of 3-cyanopyrroles 1a-k are useful handles for easy modifications into other functional groups. For example, the polyhydroxyalkyl chain of pyrrole 1c was oxidized effectively using sodium periodate to an aldehyde moiety, thus Scheme 1. Large-scale reactions for the synthesis of 3-cyanopyrroles 1c, 4a, and 16b.

Structural Modification of the Synthesized Pyrroles
The polyhydroxyalkyl chains of 3-cyanopyrroles 1a-k are useful handles for easy modifications into other functional groups. For example, the polyhydroxyalkyl chain of pyrrole 1c was oxidized effectively using sodium periodate to an aldehyde moiety, thus generating 5-formyl 3-cyanopyrroles 22 in 98% yield (Scheme 2) [22,23]. Further, the reduction of the formyl group of 22 using sodium borohydride gave 5-hydroxymethyl 3-cyanopyrroles 23 in 95% yield (Scheme 2) [24]. Both the aldehyde and hydroxyl groups can be used for homologation reactions or transformations to other functional groups. The 5-formyl and 5-hydroxymethyl groups are important functional substitutions that enable further modifications of 3-cyanopyrroles to become bioactive molecules or natural products [25,26]. Scheme 1. Large-scale reactions for the synthesis of 3-cyanopyrroles 1c, 4a, and 16b.

Structural Modification of the Synthesized Pyrroles
The polyhydroxyalkyl chains of 3-cyanopyrroles 1a-k are useful handles for easy modifications into other functional groups. For example, the polyhydroxyalkyl chain of pyrrole 1c was oxidized effectively using sodium periodate to an aldehyde moiety, thus generating 5-formyl 3-cyanopyrroles 22 in 98% yield (Scheme 2) [22,23]. Further, the reduction of the formyl group of 22 using sodium borohydride gave 5-hydroxymethyl 3cyanopyrroles 23 in 95% yield (Scheme 2) [24]. Both the aldehyde and hydroxyl groups can be used for homologation reactions or transformations to other functional groups. The 5-formyl and 5-hydroxymethyl groups are important functional substitutions that enable further modifications of 3-cyanopyrroles to become bioactive molecules or natural products [25,26].

Mechanistic Considerations
Plausible reaction pathways are shown in Figure 4. In path A, the AcOH-catalyzed reaction between the α-hydroxyketones and amines gives the imine intermediate

Mechanistic Considerations
Plausible reaction pathways are shown in Figure 4. In path A, the AcOH-catalyzed reaction between the α-hydroxyketones and amines gives the imine intermediate I which tautomerizes to intermediate II. The reaction of II with oxoacetonitriles gives enaminone intermediate III, which upon cyclization and dehydration, gives intermediate IV. Finally, the aromatization of IV gives the product (Figure 4, path A) [15,27]. In path B, oxoacetonitriles react first with primary amines to give the enamine intermediate V, which condenses with the α-hydroxyketones to give VI. Intermediate VI tautomerizes to VII, which upon intermolecular cyclization and loss of water gives VIII. Aromatization of VIII then gives the product (Figure 4,   In all three pathways, the reaction proceeded with very high atom efficiency. Water was the only molecule lost during this three-component reaction. In all three pathways, the reaction proceeded with very high atom efficiency. Water was the only molecule lost during this three-component reaction.

Materials and Methods
All chemicals and AR grade solvents were obtained from Sigma-Aldrich (Saint Louis, MO, USA), Merck (Lebanon, NJ, USA) or Alfa Aesar (Tewksbury, MA, USA) and were used as received without further purification. IR spectra were recorded using Bruker MPA FT-IR machine (Karlsruhe, Germany). 1 H NMR spectra were recorded at 300 MHz on a Bruker Avance DPX 300 or at 400 MHz Bruker Avance III 400 (BBFO 400) (Karlsruhe, Germany). 13 C NMR spectra were recorded at 75.47 MHz on a Bruker Avance DPX 300 or at 101 MHz Bruker Avance III 400 (BBFO 400). Structural assignments were made with additional information from gCOSY, gHSQC, and gHMBC experiments. HRMS were measured using a hybrid Quadrupole Time-of-Flight (Q-TOF) (Waters Xevo G2-X2 MS) on Qstar XL MS/MS system (Milford, CT, USA). LCMS spectra were recorded using Agilent 6530 LCMS (Santa Clara, CA, USA). Single-crystal X-ray crystallographic analysis was done using Bruker D8 Quest (Karlsruhe, Germany). Analytical TLC was performed using Merck 60 F 254 precoated silica gel plates (0.2 mm thickness) (Oakville, ON, Canada). The plates were visualized using UV radiation (254 nm) or stained in ceric ammonium sulfate solution with heating to detect the reaction spots. Flash chromatography was performed using Merck silica gel 60 (230-400 mesh) (Oakville, ON, Canada). The general procedure of phenacyl alcohols preparation, synthesis of 3-cyanopyrroles 1a-k, 4a-c, 14a-e, 15a, 16a-b, 17a-b and compounds 22 and 23 are found below.

General Procedure-Preparation of Substituted Phenacyl Alcohols 14-17
A mixture of phenacyl bromide (20 mmol) and sodium formate (16 mmol) was stirred in ethanol/water (30 mL, EtOH:H 2 O = 9:1) at 90 • C for 12 hours. Upon completion of the reaction (TLC), the mixture was cooled to room temperature, and ethanol was evaporated under reduced pressure. The resulting concentrated mixture was diluted with water (8 mL) and extracted using ethyl acetate (10 mL × 3 times). The organic layers were dried over Mg 2 SO 4 , filtered, and evaporated to dryness under reduced pressure. The residual compound was purified using column chromatography using an EtOAc/Hexane mixture (1:4). This general procedure was used to prepare phenacyl alcohols 14-17 (Scheme 3).