Rhodium‐Catalyzed Formylation of Unactivated Alkyl Chlorides to Aldehydes

Abstract

The first rhodium‐catalyzed formylation of non‐activated alkyl chlorides with syn gas (H₂/CO) allows to produce aldehydes in high yields (25 examples). A catalyst optimization study revealed Rh(acac)(CO)₂ in the presence of 1,3‐bisdiphenylphosphinopropane (DPPP) as the most active catalyst system for this transformation. Key for the success of the reaction is the addition of sodium iodide (NaI) to the reaction system, which leads to the formation of activated alkyl iodides as intermediates. Depending on the reaction conditions, either the linear or branched aldehydes can be preferentially obtained, which is explained by a different mechanism.

An efficient and convenient rhodium‐catalyzed formylation of alkyl chlorides to synthesize aldehydes has been developed. Depending on the conditions both linear and branched aldehydes can be obtained in a selective manner from readily available substrates.

Introduction

The reductive formylation of C−X bonds (X=Cl, Br, I) using syn gas offers a straightforward possibility to prepare versatile aldehydes from easily available alkyl and aryl halides. While the palladium‐catalyzed formylation of the latter substrates has been developed to some extent, ^[1] alkyl halides remained basically unknown as starting materials. Apart from stoichiometric multi‐step transformations (Scheme 1a),[ ² , ³ , ⁴ , ⁵ ] catalytic reductive formylations of alkyl iodides and bromides are only known using either platinum complexes or radical reactions (Scheme 1b).[ ⁶ , ⁷ ] More specifically, in 1986 Watanabe and co‐workers published a platinum complex catalyzing the formylation of alkyl iodides to give the corresponding aldehydes at high pressure of syn gas (50 bar). Here, the δ‐alkyl platinum species are the key intermediates in the reaction which will not undergo β‐elimination prior to CO insertion. ^[6] The same year, Ryu and co‐workers reported a free‐radical carbonylation of alkyl bromides to form aldehydes by using Bu₃SnH as hydrogen source, which also promotes the generation of the starting radical species at high CO pressure (65–80 bar). ^[7]

Scheme 1.

Selected known formylations of alkyl halides: a) Stoichiometric multi‐step transformations; b) catalytic reductive formylations; c) rhodium‐catalyzed formylation of alkyl chlorides.

To the best of our knowledge, no investigations of reductive carbonylations of alkyl chlorides using syn gas have been performed up to now. Due to this gap of knowledge and the importance of aldehydes in organic synthesis and the fine chemical industry,[ ⁸ , ⁹ ] we became interested in this transformation. Herein, we report the first metal‐catalyzed formylation of primary alkyl chlorides to produce aldehydes in good yields and high chemical selectivity (Scheme 1c). Depending on the solvent and base the regioselectivity of the formed aldehyde can be controlled, which is explained by a changing mechanism. Crucial for this transformation is also the use of sodium iodide (NaI) additive, which activates the alkyl chlorides by forming the respective alkyl iodides as intermediates.

Results and Discussion

To identify a suitable catalyst system for the formylation of alkyl chlorides, (2‐chloroethyl)benzene was selected as a model substrate. Initially, we investigated the catalytic activity of different metal complexes of palladium, ruthenium, iridium, and rhodium in the presence of 1,3‐bisdiphenylphosphinopropane (DPPP) (Supporting Information, Table S1). This ligand was selected due to its superior performance in our recent work on the alkoxycarbonylation of alkyl chlorides. ^[10] Notably, using Rh(acac)(CO)₂ best activity as well as selectivity was observed. Next, different phosphine ligands were tested in the presence of this metal precursor. Here, both monodentate and bidentate ligands such as PPh₃, Xantphos, etc. were selected, which have been applied for carbonylation of aryl chlorides.[ ¹¹ , ¹² ] As shown in Table 1, high conversion was obtained in the presence of all tested ligands. However, depending on the ligand, the regioselectivity of linear and branched aldehydes 2  a/3  a as well as the chemoselectivity of the reaction were strongly influenced. While Xantphos and DPPP showed good selectivity for the linear aldehyde 2  a, only the latter ligand provided high selectivity (conversion=99 %, chemoselectivity=92 %, regioselectivity=93/7, 76 % isolated yield) for 2  a (Table 1, entries 5 and 9). In case of high conversion, but low aldehyde yields, phenylethane 4, styrene 5, and (2‐iodoethyl)benzene 6 were identified as typical by‐products. Further optimization revealed the following reaction conditions to be optimal to obtain the linear aldehyde: Substrate (0.2 mmol), Rh(acac)(CO)₂ (0.01 mmol), DPPP (0.03 mmol), NaI (0.20 mmol), Na₂CO₃ (0.24 mmol), CO/H₂ (2.0 MPa), 150 °C, 1,4‐dioxane (0.5 mL), 20 h.

Table 1.

Rh‐catalyzed formylation of (2‐chloroethyl)benzene using different ligands, temperature and catalyst loading.^[a]

Entry	P‐ligand	Additive (1.0 equiv.)	Conv. [%][b]	Sel. 2 a/3 a [%][b]	Sel. 4/5/6 [%][b]
1	–	–	0	0	0/0/0
2	–	NaI	45	0	4/2/94
3	DPPP	–	0	0	0/0/0
4	PPh3	NaI	99	8 (72/28)	59/5/28
5	Xantphos	NaI	99	11 (82/18)	83/0/6
6	DPPE	NaI	99	13 (56/44)	2/75/10
7	DPPB	NaI	99	53 (64/36)	27/0/20
8	DPPF	NaI	99	12 (82/18)	88/0/0
9[c]	DPPP	NaI	99	92 (93/7)	8/0/0
10[d]	DPPP	NaI	36	0	100/0/0
11[e]	DPPP	NaI	96	10 (93/7)	12/18/60
12[f]	DPPP	NaI	99	64 (78/22)	36/0/0
13[g]	DPPP	NaI	99	75 (88/12)	25/0/0
14[h]	DPPP	NaI	99	31 (97/3)	4/3/62
15[i]	DPPP	NaI	21	0	0/0/100

[a] (2‐Chloroethyl)benzene (0.2 mmol), Rh(acac)(CO)₂ (5 mol %), S/Rh mol ratio=20, P‐ligand (15 mol %), [Rh]/P‐ligand molar ratio=1/6, Na₂CO₃ as base (0.24 mmol,1.2 equiv.), NaI as additive (0.20 mmol,1.0 equiv.), CO/H₂=20 bar, 150 °C, 1,4‐dioxane as solvent 0.5 mL, 20 h. [b] Determined by GC and GC‐MS, n‐dodecane as internal standard. [c] Isolated yield of 2  a is 76 %. [d] [Rh] 2.5 mol %, S/Rh mol ratio=40. [e] CO/H₂=2 bar. [f] CO/H₂=10 bar. [g] CO//H₂=30 bar. [h] 130 °C. [i] 100 °C.

Interestingly, we found that no carbonylation of the substrate took place in the absence of NaI (Table 1, entries 1 and 3). When NaI was added alone, mainly (2‐iodoethyl)benzene is formed by nucleophilic substitution reaction with high selectivity (94 %, Table 1, entry 2). Utilizing a lower catalyst loading (2.5 mol %) decreased the rate of the carbonylative transformation and elimination of HX (X=Cl, I) occurred predominantly (Table 1, entry 10). Noteworthy, the model reaction showed a strong influence with respect to the CO/H₂ pressure and temperature (Table 1, entries 11–15). Thus, below 10 bar CO/H₂ pressure and 150 °C a significantly reduced rate of the carbonylative procedure is observed.

Furthermore, the identity of the base and solvent proved to be critical for the success and the selectivity of the carbonylative reaction. As shown in Table 2, no desired products were formed under the standard conditions in THF, toluene, DMSO and DMF (Table 2, entries 1–4). Interestingly, when using DMAc as solvent the regioselectivity towards the aldehydes completely changed and instead of the linear isomer 2  a the branched aldehyde 3  a is formed preferentially (2  a/3  a=8/92, Table 2, entry 5). This surprising behavior is explained by a change of the mechanism (see below mechanistic discussion). Furthermore, we tested the effect of base for this carbonylative procedure (Table 2, entries 8–13). In the presence of various amine bases, the carbonylation was inhibited and styrene or ethylbenzene are formed as major products (Table 2, entries 8–11). The addition of t‐BuOK or t‐BuONa also gave favorably the branched regioisomer (Table 2, entries 12 and 13). Combining t‐BuOK as base and DMAc as solvent allowed for superior catalytic activity and excellent branched regioselectivity (86 %, 2  a/3  a=4/96, Table 2, entry 14).

Table 2.

Rh‐catalyzed formylation of (2‐chloroethyl)benzene: Testing different solvents and bases.^[a]

Entry	Base (1.2 equiv.)	Solvent (0.5 mL)	Conv. [%][b]	Sel. 2 a/3 a [%][b]	Sel. 4/5/6 [%][b]
1	Na2CO3	THF	80	0	14/0/86
2	Na2CO3	toluene	98	0	0/11/89
3	Na2CO3	DMSO	20	0	0/20/80
4	Na2CO3	DMF	78	0	26/74/0
5	Na2CO3	DMAc	99	65 (8/92)	35/0/0
6	Na2CO3	CH3CN	99	51(53/47)	28/21/0
7	Na2CO3	cyclohexane	74	20 (65/35)	26/4/50
8	Et3N	1,4‐dioxane	99	0	21/51/28
9	Bu3N	1,4‐dioxane	99	0	78/22/0
10	TBEA	1,4‐dioxane	99	0	80/20/0
11	TMEDA	1,4‐dioxane	99	0	29/46/25
12	t‐BuOK	1,4‐dioxane	99	68 (15/85)	28/4/0
13	t‐BuONa	1,4‐dioxane	99	50 (10/90)	50/0/0
14	t‐BuOK	DMAc	99	86 (4/96)	14/0/0

[a] (2‐Chloroethyl)benzene (0.2 mmol), Rh(acac)(CO)₂ (5 mol%), S/Rh mol ratio=20, DPPP as P‐ligand (15 mol%), [Pd]/P‐ligand molar ratio=1/6, base (0.24 mmol,1.2 equiv.), NaI as additive (0.20 mmol, 1.0 equiv.), CO/H₂=20 bar, 150 °C, solvent 0.5 mL, 20 h. [b] Determined by GC and GC‐MS, n‐dodecane as internal standard.

Mechanism and control experiments

To gain insight into the mechanism of the reaction and to understand the surprising influence of base and solvent on the regioselectivity, we performed several control experiments. At first, a kinetic profile of the formylation of (2‐chloroethyl)benzene with CO/H₂ was carried out under the standard conditions (Rh(acac)(CO)₂/DPPP catalyst in 1,4‐dioxane). As illustrated in Figure 1, in the first 2 h, 80 % of (2‐chloroethyl)benzene is converted to (2‐iodoethyl)benzene and only little product 2  a (7 %) is detected, which indicates (2‐iodoethyl)benzene as an important reaction intermediate. As the reaction continued, the conversion of (2‐chloroethyl)benzene increased gradually along with the selective formation of product 2  a. After 8 h, the substrate is completely converted and chemical selectivity of the aldehyde 2  a increased to 52 % while the branched aldehyde was obtained in 5 % selectivity. At the end of the reaction (20 h), the aldehyde 2  a increased to 86 %, while 3  a was 6 %.

Figure 1.

Kinetic profile of the rhodium‐catalyzed formylation of (2‐chloroethyl)benzene in 1,4‐dioxane (products determined by GC; n‐dodecane as internal standard).

Throughout the reaction small amounts of styrene are formed by elimination of hydrogen chloride. Obviously, styrene can also be formed after oxidative addition of the Rh complex to C−X and subsequent β‐H elimination of the alky RhL _n complex. The formed styrene might be carbonylated in a follow up reaction. Indeed, styrene disappeared during the reaction, while 3  a kept around 6 % unchanged.

Next, some control experiments were performed to verify the role of base, solvent and catalyst system. As shown in the kinetic profile (2‐chloroethyl)benzene 1  a is initially converted to (2‐iodoethyl)benzene 6 in the presence of Na₂CO₃ in 1,4‐dioxane (Scheme 2A, conditions 1). Under these reaction conditions the linear aldehyde is formed exclusively. However, using t‐BuOK as base and DMAc as solvent the branched aldehyde is obtained regioselectively (Scheme 2A, conditions 2). This dramatic selectivity change can be explained by the exclusive formation (>99 % selectivity) of styrene 5 under conditions 2, which further underwent a classic hydroformylation process. Notably, the presence of the ligand DPPP improved the selectivity of the chloride/iodide exchange reaction (Scheme 2B, conditions 1), while the t‐BuOK/DMAc system still gave exclusively 5 (Scheme 2B, conditions 2). Not surprisingly, phenylethane 4 showed no reactivity under both conditions (Scheme 2C, conditions 1 and 2). Using (2‐iodoethyl)benzene 6 as substrate under conditions 1, phenylethane is formed in high selectivity; however, in the presence of t‐BuOK/DMAc again styrene is detected in >99 % selectivity (Scheme 2D). As a substrate (2‐iodoethyl)benzene 6 is relatively stable under conditions 1 compared to conditions 2. Indeed, without catalyst system only 6 % conversion was detected (Scheme 2E, conditions 1).

Scheme 2.

Mechanistic studies: Selected control experiments.

Using (2‐iodoethyl)benzene 6 the same chemo‐ and regioselectivity as with 1‐chloro‐2‐phenylethane is observed: with Na₂CO₃ in 1,4‐dioxane the linear product 2  a is favored, while with t‐BuOK in DMAc the branched isomer 3  a is dominant (Scheme 2F, conditions 1 and 2). As expected, the hydroformylation of styrene gave the branched regioisomer even when using the Na₂CO₃/1,4‐dioxane system (Scheme 2G, conditions 1–4). This is explained by the increased stability of the benzylic rhodium intermediate, which subsequently undergoes CO insertion.

Based on the above results and previous reports, ^[13] the following plausible mechanism is proposed (Scheme 3). The reaction starts by nucleophilic substitution of the alkyl chloride with NaI to form the corresponding alkyl iodide. This intermediate will either lead to the linear aldehyde by a direct Rh‐catalyzed formylation process or form an alkene via HI‐elimination. According to cycle I, oxidative addition of the alkyl iodide with a Rh(I) complex gives the linear alkyl Rh^IIIL_n intermediate B, similar to the Montsanto carbonylation process. ^[14] Next, elimination of HI, CO coordination and insertion forms the acyl Rh^IL_n complex C. Final oxidative addition with H₂ and subsequent reductive elimination forms the targeted linear aldehyde along with the [Rh^IHL_n] A species left for the next catalytic cycle. As a side reaction, the hydrogenation of the olefin leads to the observed alkanes. Notably, reaction of A with styrene favors the branched aldehyde because of the increased stability of the benzylic metal complex F (cycle II). ^[8] Apparently in the presence of Na₂CO₃ in 1,4‐dioxane cycle II is avoided because of a fast CO insertion to give the linear acyl rhodium complex (cycle I), while using t‐BuOK/DMAc elimination leads to the easier formation of styrene, which preferentially forms the branched product (cycle II).

Scheme 3.

Proposed mechanism for the formylation of unactivated alkyl chlorides.

To demonstrate the generality of this novel protocol for the synthesis of linear aldehydes, different alkyl chlorides were selected to test the suitability of the catalyst system (Scheme 4). Using linear alkyl chlorides with different chain length, formylation occurred to form linear aldehydes in good to high isolated yields (58–85 %, regioselectivity up to 91/9, 2  b–2  l). Formylation of sterically more hindered isopropyl and t‐butyl chloride also proceeded in good yield to give the linear products (60–62 %, 2  m–2  n). Phenyl‐substituted alkyl chlorides, like benzyl chloride and 1‐chloro‐3‐phenylpropane were converted to the target aldehydes in good yields (75–80 %, 2  o–2  p), too. The hydroformylation of allyl esters allows for an efficient access of synthetically interesting bifunctional compounds. The resulting ester‐substituted aldehydes can be easily converted into 1,4‐diols or 1,4‐aminoalcohols. Thus, we synthesized a series of alkyl chlorides containing ester groups. ^[15] In all cases the desired products were obtained in good isolated yields (2  q–2  w, 54–72 %). Finally, our procedure can be also applied for the modification of bioactive molecules. More specifically, gemfibrozil and a naproxen derivative delivered formylation products 2  x and 2  y in 46 % and 54 % yield, respectively.

Scheme 4.

General formylation of unactivated alkyl chlorides using Rh(acac)(CO)₂/DPPP catalytic system. [a] Alkyl chlorides (0.2 mmol), Rh(acac)(CO)₂ (0.01 mmol), DPPP (0.03 mmol), NaI (0.2 mmol), Na₂CO₃ (0.24 mmol), CO/H₂=20 bar, 150 °C, 1,4‐dioxane as solvent 0.5 mL, 20 h. [b] Determined by GC and GC‐MS (n‐dodecane was used as internal standard), isolated yield. [c] Chloroethane as substrate (2.0 mol/L in THF).

Conclusion

Here, we report a novel protocol for the formylation of unactivated alkyl chlorides to the corresponding linear/branched aldehydes for the first time. Utilizing a Rh(acac)(CO)₂/1,3‐bisdiphenyl‐phosphinopropane (DPPP) catalyst system in dioxane in the presence of sodium carbonate provided different linear aldehydes with high chemo‐ and regioselectivity and in good isolated yields. Modification of the solvent and base to DMAc, potassium tert‐butoxide allows to switch the regioselectivity from the linear to the branched aldehyde.

Experimental Section

Reagents and analysis: Chemical reagents were purchased from FluoroChem, Abcr Chemicals, TCI Chemicals and were used as received. ¹H and ¹³C NMR spectra were recorded on a Bruker 300 spectrometer. Chemical shifts δ (ppm) are given relative to solvent: references for CDCl₃ are 7.26 ppm (¹H NMR) and 77.16 ppm (¹³C NMR), Multiplets are assigned as s (singlet), d (doublet), t (triplet), q (quartet), dd (doublet of doublet), m (multiplet). Conversion and selectivity were determined by Agilent 6890 N Gas chromatography (GC) and Agilent 5973 Network GC‐MS, using splitless injection and FID detector (column: HP‐5 (30 m×320 μm×0.25 μm); inlet temperature: 250 °C; starting oven temperature: 50 °C; rate: 8 °C/min; final temperature: 260 °C, carrier gas: N₂ 25 mL/min).

General procedures for formylation of alkyl chlorides: In a typical experiment, (2‐chloroethyl)benzene (0.2 mmol), Rh(acac)(CO)₂ (0.01 mmol), the phosphine ligand (0.06 mmol monophosphine ligand or 0.03 mmol bisphosphine ligand), NaI as additive (0.20 mmol) and Na₂CO₃ as base (0.24 mmol) were added sequentially into 1,4‐dioxane (0.5 mL) in a 50 mL stainless steel autoclave. The obtained mixture was purged with syn gas for three times and pressured to 0.2 MPa. The reaction mixture was stirred at the appointed temperature for 20 h. Upon completion, the autoclave was cooled down to room temperature and depressurized carefully. The reaction solution was analyzed by GC and GC‐MS to determine the conversions and the chemical selectivity (n‐dodecane as internal standard). Then, the product was further purified by column chromatography. ¹H NMR and ¹³C NMR spectroscopy were used to identify the structure of the products.

Conflict of interest

The authors declare no conflict of interest.

Supporting information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supporting Information

Wang P., Wang Y., Neumann H., Beller M., Chem. Eur. J. 2023, 29, e202203342.

Data Availability Statement

The data that support the findings of this study are available in the supplementary material of this article.