Two-Carbon Homologation of Aldehydes and Ketones to α,β-Unsaturated Aldehydes

Phosphonate reagents were developed for the two-carbon homologation of aldehydes or ketones to unbranched- or methyl-branched α,β-unsaturated aldehydes. The phosphonate reagents, diethyl methylformyl-2-phosphonate dimethylhydrazone and diethyl ethylformyl-2-phosphonate dimethylhydrazone, contained a protected aldehyde group instead of the usual ester group. A homologation cycle entailed condensation of the reagent with the starting aldehyde, followed by removal of the dimethylhydrazone protective group with a biphasic mixture of 1 M HCl and petroleum ether. This robust two-step process worked with a variety of aldehydes and ketones. Overall isolated yields of unsaturated aldehyde products ranged from 71% to 86% after the condensation and deprotection steps.


Introduction
A recurring theme in the synthesis of various polyketide natural products and other compounds is two-carbon homologation of aldehydes or ketones to unbranched or branched (e.g., α-methyl or β-methyl) α,β-unsaturated aldehydes. A typical approach based on the Horner-Wadsworth-Emmons condensation [1] uses a phosphonate ester, such as triethyl-2-phosphonoacetate or triethyl-2-

OPEN ACCESS
phosphonopropionate, and involves three synthetic steps: The phosphonate condensation step extends the chain by two carbons while introducing an (E) double bond, an α-alkyl branch (size depending on the choice of phosphonate reagent), and a terminal ester group. The ester is then reduced to an α,β-unsaturated alcohol, and the alcohol is partially oxidized to the desired α,β-unsaturated aldehyde. One way to shorten this scheme would be to use a homologating reagent that keeps the oxidation state of an aldehyde throughout.
A variety of options were considered. Aldehyde Wittig reagents, such as α-formylethylidene triphenylphosphorane are known [2,3], but in our hands the reagent was not reactive toward α,β-branched unsaturated aldehydes, such as (2E)-2-methyl-2-butenal [4]. Phosphonate anions are generally more reactive toward aldehydes or ketones than the analogous phosporanes [5]. Aldehyde phosphonates (where the ester group is replaced by an aldehyde) can be readily prepared [6], but using them directly in Wittig-Horner reactions is not practicable because of self-condensation. Reactions using aldehyde phosphonates protected as imines and acetals have been reported, but these reagents also have problems with side reactions, insufficient reactivity, and lability of products [7]. The silylaldimines are the functional equivalents of protected aldehyde phosphonates and have been used in Wittig-Horner-like reactions [8,9]; however, reaction conditions to form and use silylaldimines are not operationally convenient [9]. Solid base-promoted cross-aldol condensations have been successful, but only aromatic or heteroaromatic aldehydes were suitable as reactants [10]. Recently, acidic ionic liquid-catalyzed homologation of α,β-enals was reported, but the scope of the reaction did not include aliphatic aldehydes [11].We chose to further explore phosphonate chemistry with a better carbonyl protective group because the scope of the reaction would include a variety of aldehydes and ketones. Protection of carbonyl compounds as dimethylhydrazones is a common practice, and the incorporation and removal of this group have been studied extensively [12][13][14].
The aldehyde functional group was then protected as the dimethylhydrazone derivative, with N,N-dimethylhydrazine, as done previously [19]. Phosphonate 1 was prepared in 82% overall yield from the starting material. The product still contained about 8% unused starting material, carried through the synthesis. The presence of starting material did not pose a problem when the phosphonate reagent was used for aldehyde homologation or ketone to aldehyde homologation.
Phosphonate reagent 1, could also be prepared from commercially available diethyl methylphosphonate, by lithiation of diethyl methylphosphonate, formylation with dimethylformamide [19], followed by protection of the aldehyde function as the dimethylhydrazone derivative [13]. The overall yield for this synthetic pathway was only 50% (because of difficulty extracting water-soluble ethylformyl-2-phosphonate), so the other preparative route was preferred.
Phosphonate reagent 2 was prepared from diethyl ethylphosphonate by formylation, to prepare diethyl 1-propylformyl-2-phosphonate [19], and then the aldehyde functional group was protected as the dimethylhydrazone derivative. The overall yield was usually greater than 95%.
The new method requires the phosphonate reagents to be synthesized, whereas those for the original method are available commercially. Nevertheless, large batches of the new reagents can be prepared and stored, and these stocks could subsequently serve for syntheses of a variety of final products. Diethyl methylformylphosphonate can also be named (2-oxo-ethyl)-phosphonic acid diethyl ester and diethyl ethylformyl-2-phosphonate can also be named (1-methyl-2-oxo-ethyl)-phosphonic acid diethyl ester. Diethyl methylformylphosphonate dimethylhydrazone can also be named [2-(dimethylhydrazino)-ethyl]-phosphonic acid diethyl ester.and diethyl ethylformyl-2-phosphonate dimethylhydrazone can also be named [2-(dimethyl-hydrazino)-1-methyl-ethyl]-phosphonic acid diethyl ester. Alternative names for the aldehyde phosphonates and aldehyde phosphonate dimethylhydrazones 1 and 2 are given, but the simpler, and more intuitive, names were used in this paper.

Use of the Phosphonate 1 and 2 Reagents in Synthesis
Both 1 and 2 were readily converted to their corresponding anions with the base LDA and reacted smoothly with aldehydes or ketones, added in excess, to give unbranched-or methyl-branched, unsaturated aldehyde dimethylhydrazones with chains having two additional carbons (Scheme 2). Yields for the condensation step ranged from 77%-99%.
Deprotection of unsaturated aldehyde dimethylhydrazones was accomplished in 81%-96% yield, using a biphasic mixture of 1 M HCl and petroleum ether [14]. In the deprotection reaction, the dimethylhydrazone is initially converted to a water-soluble hydrochloride salt, and the free aldehyde then forms under the acidic conditions. Successful deprotection requires immediate migration of the liberated, but sometimes labile, aldehyde from the aqueous phase to the neutral petroleum ether phase.
In the absence of petroleum ether, deprotection of unsaturated aldehyde dimethylhydrazones 9 and 10 gave a complex mixture of reaction products.
Using these reagents led to a two-step aldehyde homologation cycle (condensation and deprotection) instead of the previous three-step cycle. Overall yields, of unsaturated aldehyde products, for the two-step sequence ranged from 71% to 86%. Homologation of unbranched aldehyde starting materials with phosphonate 1 resulted in disubstituted α,β-unsaturated aldehyde products whereas homologation of unbranched aldehyde starting materials with phosphonate 2 resulted in trisubstituted α,β-unsaturated aldehyde products, with a methyl group at the 2-position. Homologation of methyl ketones with phosphonate 1 resulted in trisubstituted α,β-unsaturated aldehyde products, with a methyl group at the 3-position. The two-step homologation procedure gave a favorable (9.5:1 to 170:1) ratio of 2E to 2Z products from aldehyde starting materials and a ratio (2.0:1 to 2.2:1) of 2E to 2Z products from ketone starting materials. Homologation of cycloheptanone with phosphonate 2 resulted in an improved synthesis of cycloheptylidene-acetaldehyde 26. The overall yield for the preparation of 26 increased from 42% [20] to 74%.

Assignment of Double-Bond Configuration
The major isomer configuration, at the α,β-double bond of disubstituted alkene compounds 4, 6, 9, 11, 14 and 16 (Scheme 2) was readily determined to be E by the coupling constant of J 2-3 =15.1 to J 2-3 = 16.1 Hz. The major isomer configuration of all trisubstituted double bonds, of compounds 5, 7, 10, 12, 15 and 17, was determined by nuclear Overhauser enhancement (nOe) NMR methods. Irradiation at either H-1 or H-3 caused a nuclear Overhauser enhancement of 2.1% to 4.3% at H-1 or H-3 for compounds 5, 7, 10, 12, 15 and 17, but only a nOe of 0% to 0.2% was observed for the methyl group protons for any of the compounds. Irradiation of the methyl group protons caused a nOe of just 0% to 0.5% for either H-1 or H-3. The data support the assignment of 2E configuration for compounds 5, 7, 10, 12, 15 and 17, because the H-1 and H-3 protons are closer to each other than to the protons of the methyl group, attached to C-2.
The proportion of 2Z-isomer product was identified by its shorter GC retention time and nearly identical mass spectra, compared to the 2E-isomer [21], for compounds 19, and 20. The major isomer configuration of the trisubstituted double bonds, of compounds 19, and 20 was determined to be E because of the later GC retention time, previously observed for (2E)-N,N-Dimethyl-N'-(dec-2enylidene)-hydrazine and (2E)-2-Decenal [19]. Also, the reported proton NMR chemical shifts more closely matched the E-isomer of 20 reported previously [22]. We could not use nuclear Overhauser enhancement (nOe) NMR methods for additional verification because the proton chemical shifts for the methyl branch and protons on adjacent carbons were too close together for the both isomers.
The major isomer configuration of the trisubstituted double bonds, of compounds 22 and 23, derived from methyl ketone starting materials, 6, and 7 was determined by nuclear Overhauser enhancement (nOe) NMR methods. A moderate nOe enhancement was observed (2D NOESY experiments) between the methyl group protons and H-2, in the case of the minor isomer for both compounds. Only weak nOe enhancement was observed between the methyl group protons and H-2, in the case of the major isomer of compound 6 and no nOe enhancement was observed between the methyl group protons and H-2, in the case of the major isomer of compound 23. The data support the assignment of 2Z configuration for the minor isomer of compounds 22 and 23 because the H-2 protons are closer to the protons of the methyl group, attached to C-3. Therefore, the major isomer has the E-configuration. The GC retention time for the major isomer was also later than that observed for the minor isomer, consistent with the assignment of E-configuration for the major isomer. Compound 23 has been prepared previously [23][24][25] and the chemical shift for the methyl branch carbon, attached to C-3, was δ 16.4 for the major E-isomer and δ 26.4 for the minor Z-isomer, consistent with reported 13 C-NMR data [25].

Experimental
General Diethyl-2,2-(diethoxy)ethylphosphonate, diethyl methylphosphonate and diethylethylphosphonate were purchased from Alfa Aesar (Ward Hill, MA, USA). Diethyl ethylphosphonate can also be prepared in 98% yield by refluxing a mixture of triethyl phosphite and ethyl iodide for 3 h [26]. THF was purified by distillation from sodium benzophenone ketyl. Lithium diisopropylamide (LDA) was purchased from the Aldrich Chemical Co (Milwaukee, WI, USA) as a 2 M solution in a mixture of heptane/THF/ethylbenzene. Other organic reagents were obtained from Aldrich and were used without further purification. Non-aqueous reactions were performed under an atmosphere of dry argon in oven-dried glassware. Removal of solvent was accomplished by rotary evaporation at water aspirator pressure.
Progress of synthetic reactions was monitored by gas chromatography (GC). All reaction products were identified by GC/MS and the structures were additionally verified by NMR. Yields were corrected for purity. Both 2E and 2Z isomers were considered to be the product for all product compounds; however, the 2E:2Z isomer ratios are reported.
The Hewlett-Packard (HP) 5890 Series II gas chromatograph was equipped with flame ionization detector and split/splitless inlet and was interfaced to an HP ChemStation data system. The column was a DB-5 capillary (30 m × 0.25 mm, 0.25-μm film thickness, J&W Scientific, Folsom, CA, USA). Carrier gas was He. The oven temperature was programmed from 50 to 280 °C at 10 °C/min, and the detector temperature was 280 °C. The inlet temperature was 220 °C, and 1.0 µL sample injections were made in splitless mode. The inlet temperature was 120 °C, and 1.0 µL sample injections were made in splitless mode for analysis for compounds 9 and 10 because considerable degradation of unsaturated aldehyde dimethylhydrazones was observed during GC runs with higher inlet temperatures.
Electron impact mass spectra (70 eV) were obtained with an HP 5973 MSD instrument, interfaced to an HP 6890 GC, equipped with a splitless inlet. Several columns were used, but gave comparable results to that used for GC. The oven temperature was programmed from 50 to 250 °C at 10 °C/min; the inlet temperature was 220 °C and the transfer line temperature was 250 °C. The inlet temperature was 120 °C for compounds 9 and 10.
1 H-NMR, 13 C-NMR and 2D NMR spectra were collected on a Bruker (Bellerica, MA, USA) Avance 500 spectrometer using a 5 mm broadband probe. Samples were dissolved in either CDCl 3 or C 6 D 6 (where indicated) and all spectra ( 1 H and COSY at 500 MHz, 13  Diethyl methylformylphosphonate dimethylhydrazone (1). Diethyl-2,2-(diethoxy)ethylphosphonate (5.31 g, 20.9 mmol) was added to acetone (200 mL), followed by water (3 mL) and p-toluenesulfonic acid (1.3 g, 6.8 mmol). The mixture was stirred at r.t., under an argon atmosphere for 4 days. A solution of NaHCO 3 (570 mg, 6.8 mmol in 6 mL water) was added with stirring to quench the reaction and neutralize acid present. The pH was checked and found to be 7. After removal of solvent, aqueous oil remained. The oil was saturated by the addition of NaCl (25 g). The aqueous oil was repeatedly extracted with CH 2 Cl 2 (100 mL then 2 × 50 mL), and the combined CH 2 Cl 2 extracts were dried over anhydrous Na 2 SO 4 and filtered. Solvent was removed to afford 5.59 g of crude diethyl methylformyl-2-phosphonate. Without further purification, the crude product was converted to the dimethylhydrazone (DMH) derivative by adding it to a mixture of CH 2 Cl 2 (75 mL) containing anhydrous MgSO 4 (5. . The MS and NMR spectral data were consistent with the reference spectral data from previous work [19]. Alternative preparation of diethyl methylformylphosphonate dimethylhydrazone (1). A solution of n-butyllithium (2.5 M, 25 mL, 0.213 mol, slight excess) was added to dry THF (80 mL) and the mixture was cooled in an argon atmosphere to dry ice-ethanol bath temperature (approx −78 °C). Diethyl methylphosphonate (9.13 g, 8.8 mL, 0.06 mol) was added dropwise and the mixture was stirred for an additional hr before dry dimethylformamide (DMF, distilled from CaH 2 , 4.8 mL, 0.066 mol) was added. After warming to 20 °C, ice cold 3 M aqueous HCl (300 mL) was added and the mixture was stirred 5 min. A fine white precipitate formed initially, and then the solution cleared and separated into two phases (liberation of the free phosphonate aldehyde). The phases were separated, and the organic phase was washed with 10 mL of alkaline brine solution (0.5 % NaHCO 3 in saturated aqueous NaCl), dried over anhydrous MgSO 4 , and filtered. The acidic aqueous phase was repeatedly extracted with CH 2 Cl 2 (5 × 50 mL), and the combined CH 2 Cl 2 extracts were washed with alkaline brine solution (2 × 10 mL). The alkaline brine washes were combined and extracted again with CH 2 Cl 2 (4 × 15 mL). All CH 2 Cl 2 extracts were combined, dried over anhydrous MgSO 4 , and filtered. Solvent was removed from the dried organic phase and all dried CH 2 Cl 2 extracts to afford 9.8 g of crude diethyl    (2E)-2-Decenal (6). Deprotection was initiated by stirring 4 (1.10 g, containing 0.98 g 4, 5.0 mmol) with 1 M HCl (50 mL) at room temperature for 5 min (4 dissolves as the hydrochloride salt forms), and then petroleum ether (b.p. 35 °C to 60 °C, 50 mL) was added and stirring continued. After 4 h, the phases were separated and the aqueous phase was returned to the reaction vessel. Petroleum ether (30 mL) was added and the mixture was stirred at r.t. for an additional 1 h before the phases were separated. The combined petroleum ether phases were washed with alkaline brine solution (2 × 10 mL), and the brine phase was back-extracted once with petroleum ether (20 mL). The combined organic phases were dried, filtered, and removal of solvent afforded 0.75 g of the free aldehyde 6. Purity by GC was 92%, and yield corrected for purity was 90%. The 2E:2Z isomer ratio was 25 to 1. The MS and NMR spectral data were consistent with the reference spectral data from previous work [9].  (7). Deprotection procedure was similar to that used for 6. A solution containing 5 (1.15 g, containing 1.04 g 3c, 5.0 mmol) and 1 M HCl (60 mL), to which petroleum ether (b.p. 35 °C to 60 °C, 60 mL) was added after 5 min, was stirred at room temperature for 6 h. The phases were separated and the aqueous phase was returned to the reaction vessel. Petroleum ether (40 mL) was added and the mixture was stirred at r.t. for an additional 1 h before the phases were separated. Workup of the reaction as for 6 resulted in 0.85 g of the free aldehyde 7 (purity by GC showed 90%, yield corrected for purity 91%). The 2E:2Z isomer ratio was 24 to 1. The MS and NMR spectral data were consistent with the reference spectral data from previous work [28]. (2E,4E)-2,4-undecadienal (11). Deprotection procedure was similar to that used for 6. A solution containing 9 (1.16 g, containing 1.07 g 9, 5.1 mmol) and 1 M HCl (50 mL), to which petroleum ether (b.p. 35 °C to 60 °C, 50 mL) was added after 5 min, was stirred at room temperature for 24 h. The phases were separated and the aqueous phase was returned to the reaction vessel. Petroleum ether (40 mL) was added and the mixture was stirred at r.t. for an additional 1 h before the phases were separated. Workup of the reaction as for 6 resulted in 0.81 g of the free aldehyde 11 (purity by GC showed 93%, yield corrected for purity 89%). The 2E:2Z isomer ratio was 9. (2E, 4E)-2-Methyl-2,4-undecadienal (12). Deprotection procedure was similar to that used for 6. A solution containing 10 (1.29 g, containing 1.10 g 10, 5.0 mmol) and 1 M HCl (50 mL), to which petroleum ether (b.p. 35 °C to 60 °C, 50 mL) was added after 5 min, was stirred at room temperature for 48 h. The phases were separated and the aqueous phase was returned to the reaction vessel. Petroleum ether (40 mL) was added and the mixture was stirred at rt for an additional 1 h before the phases were separated. Workup of the reaction as for 6 resulted in 0.91 g of the free aldehyde 12 (purity by GC showed 88%, yield corrected for purity 89%). The 2E:2Z isomer ratio was 30 to 1. (88% purity by GC, yield corrected for purity 88%). By GC and GC-MS, the 2E/2Z isomer ratio was 50 to 1. The MS and NMR spectral data were consistent with the reference spectral data from previous work [13].

(2E)-2-Methyl-2-Decenal
(2E)-Cinnamaldehyde (16). Deprotection procedure was similar to that used for 6. A solution containing 14 (0.98 g, containing 0.86 g 14, 4.9 mmol) and 1 M HCl (50 mL), to which petroleum ether (b.p. 35 °C to 60 °C, 50 mL) was added after 5 min, was stirred at room temperature for 18 h. The phases were separated and the aqueous phase was returned to the reaction vessel. Petroleum ether (50 mL) was added and the mixture was stirred at r.t. for an additional 1 h before the phases were separated. Workup of the reaction as for 6 resulted in 0.61 g of the free aldehyde 16 (purity by GC showed 86%, yield corrected for purity 81%). The 2E:2Z isomer ratio was 170 to 1. The MS and NMR spectral data were consistent with the reference spectral data from previous work [9]. (89% purity by GC, yield corrected for purity 88%). By GC and GC-MS, the 2E/2Z isomer ratio was 25 to 1. The MS and NMR spectral data were consistent with the reference spectral data from previous work [13].
(2E)-Methyl cinnamaldehyde (17). Deprotection procedure was similar to that used for 6. A solution containing 15 (1.12 g, containing 1.00 g 15, 5.3 mmol) and 1 M HCl (50 mL), to which petroleum ether (b.p. 35 °C to 60 °C, 50 mL) was added after 5 min, was stirred at r.t. for 48 h. The phases were separated and the aqueous phase was returned to the reaction vessel. Petroleum ether (40 mL) was added and the mixture was stirred at r.t. for an additional 1 hr before the phases were separated. Workup of the reaction as for 6 resulted in 0.78 g of the free aldehyde 17 (purity by GC showed 96%, yield corrected for purity 96%). The 2E:2Z isomer ratio was 56 to 1. The MS and NMR spectral data were consistent with the reference spectral data from previous work [9].
N,N-Dimethyl-N'-(3-methyl-non-2-enylidene)-hydrazine (19  . This compound has been prepared previously [23]. (23). The deprotection procedure was similar to that used for 19. A solution containing 22 (1.49 g, containing 1.28 g 22, 6.8 mmol) and 1 M HCl (100 mL), to which hexane (100 mL) was added after 5 min, was stirred at room temperature for 3 h. The phases were separated and the aqueous phase was returned to the reaction vessel. Hexane (100 mL) was added and the mixture was stirred at r.t. for an additional 1 h before the phases were separated. Workup of the reaction as for 19 resulted in 0.95 g of the free aldehyde 23 (purity by GC showed 93%, yield corrected for purity 90%). The 2E:2Z isomer ratio was 2. . The MS and NMR spectral data were consistent with the reference spectral data from previous work [24,25].

Conclusions
The new phosphonate reagents 1 and 2 were prepared without difficulty from sufficiently inexpensive starting materials and were stable enough to be distilled. The starting materials are thus inexpensive and the homologation reactions are sufficiently robust, so use of this two-step process can be commercially viable. Subsequent removal of the dimethylhydrazone protective group could be easily accomplished, even for delicate product aldehydes. Furthermore, the new method eliminates the use and disposal of some metal-containing reagents (powerful reducing agents such as lithium aluminum hydride or diisobutylaluminum hydride or oxidants such as pyridinium dichromate or manganese dioxide). This more concise homologation chemistry could be applied to the practical synthesis of a variety of natural products, pharmaceuticals, and other compounds. This new protocol should replace the usual three-step Horner-Wadsworth-Emmons [1] aldehyde or ketone homologation scheme.