Download PDF
Synlett 2014; 25(10): 1409-1412
DOI: 10.1055/s-0033-1338634
letter
© Georg Thieme Verlag Stuttgart · New York

Continuous-Flow Hydroxylation of Aryl Iodides Promoted by Copper Tubing

Patrick Cyr
Centre in Green Chemistry and Catalysis, Faculty of Arts and Sciences, Department of Chemistry, Université de Montréal, P.O. Box 6128 Station Downtown, Montréal, Québec, H3C 3J7, Canada   Email: andre.charette@umontreal.ca
,
André B. Charette*
Centre in Green Chemistry and Catalysis, Faculty of Arts and Sciences, Department of Chemistry, Université de Montréal, P.O. Box 6128 Station Downtown, Montréal, Québec, H3C 3J7, Canada   Email: andre.charette@umontreal.ca
› Author Affiliations
Further Information

Publication History

Received: 21 February 2014

Accepted after revision: 28 March 2014

Publication Date:
08 May 2014 (online)

 
 PDF Download Permissions and Reprints All articles of this category


Abstract

A simple and ligand-free synthesis of phenols from the corresponding aryl iodides in a continuous-flow system is described. The reaction is complete in only 4 to 20 minutes when heated between 150 to 165 °C in a reactor consisting of a commercially available copper coil. An example of trapping of the phenoxide in situ is also shown.


#

Key words

phenols - copper - coupling - continuous flow - hydroxylation

Phenols are widely used as key intermediates in several chemical disciplines, including natural products and pharmaceutical targets.[1] Consequently, the development of a general and efficient method for their synthesis has attracted considerable attention within the synthetic community. More precisely, an appealing and non-oxidative approach to these building blocks consists of a nucleophilic substitution of activated aryl halides.[2] However, this process is limited by the specific functional groups required on the starting aryl halide ring. In this regard, Buchwald developed the first palladium-catalyzed direct hydroxylation of aryl halides using a hindered monodentate ligand (t-BuXPhos).[3] This method was also validated by other groups[4] and is considered to be one of the most robust and rapid methods available to access substituted phenols. Nonetheless, the toxicity and the cost associated with these palladium catalysts is problematic, thus a more economical system is desired. To circumvent the use of expensive late-transition-metal catalysts, Taillefer[5] and You[6] independently developed an efficient copper-catalyzed hydroxylation of aryl halides in the presence of a bidentate ligand (dibenzoylmethane and phenanthroline, respectively). Even though these achievements have garnered considerable recent attention,[7] there remains substantial drawbacks that impair the large-scale application of this approach. As such, the reaction employs high catalyst (5–20 mol%) and ligand (20–50 mol%) loadings, which lower the reaction TON. Moreover, extended reaction times (6–36 h) are typically required to achieve high conversion.[8]

To address these issues, and given our interest in using copper reactors,[9] we envisaged that continuous-flow chemistry would provide an ideal solution. Performing reactions under flow conditions enables improved temperature and mixing control. As such, additional control of reaction parameters is possible. In metal-catalyzed processes, this improved mass and heat transfer often leads to reduced ligand and catalyst loading. Flow methods allow for the continuous preparation of material. This ‘scale out’ as opposed to ‘scale up’ of reaction provides a safer environment and reduces the possibility of the reaction running out of control.[10]

Recently, Jamison developed an elegant method with which to generate phenols from their corresponding aryl Grignard by using flow technology.[11] While using only air as the oxidant, this methodology suffers from the tedious preparation of aryl Grignard reagents and presents significant safety hazards on a large scale. A survey of the literature revealed that continuous flow has, to the best of our knowledge, not been used in the copper-catalyzed hydroxylation of aryl halides. The reaction is known to be highly dependent on solvent, generally requiring a mixture of DMSO and water to proceed.[12] This feature limits the broad optimization of the reactant hydroxylation and makes it quite challenging to develop a system in which a nonpolar starting material, a hydroxide salt and a copper source are all sufficiently soluble to be mixed in a small reactor.

In previous accounts, it was disclosed that copper powder could be used to catalyze the hydroxylation reaction, although in each case a copper salt was found to be more reactive.[13] [7g] We, therefore, looked into the use of copper tubing to perform the transformation, where no additional catalyst would be needed. This hypothesis was supported by literature precedents in which a copper reactor has been successfully applied for alkyne–azide cycloaddition,[14] Ullmann condensation, Sonogashira coupling, and decarboxylation reactions.[9,15]

We started this study by comparing the solubility of different hydroxide salts in DMSO–water mixtures and found that tetrabutylammonium hydroxide (n-Bu4NOH) was an ideal candidate. This salt has already been used in copper-catalyzed hydroxylation reactions under standard batch conditions.[7a] [7d]

Table 1 Optimization of the Flow Hydroxylation of 4-Iodotoluene Using a Copper Reactora

Entry

Solvent

n-Bu4NOH (equiv)

Temp (°C)

Yield (%)b of 2a

 1

MeCN–H2O (1:1)

6

150

 2

 2

DMSO–H2O (1:1)

6

150

64

 3

DMSO–H2O (2:1)

3

150

38

 4

DMSO–H2O (1:1)

6

180

 1c

 5

DMSO–H2O (1:1)

6

150

55d

 6

DMSO–H2O (1:1)

6

165

75d

 7

DMSO–H2O (1:1)

6

200

56d

 8

DMSO–H2O (1:1)

6

165

76

 9

DMSO–H2O (1:1)

6

165

87e

10

DMSO–H2O (1:1)

5

165

71e,f

11

DMSO–H2O (1:1)

7

165

65e,f

12

DMSO–H2O (1:1)

6

150

85e,g

a Reaction conditions: 4-iodotoluene (0.275 mmol), n-Bu4NOH, 1:1 solvent mix (0.125 M vs. 1a), 15 minutes residence time.

b Yield obtained by 1H NMR analysis with triphenylmethane as internal standard.

c A 10 mL stainless steel coil was used instead of copper.

d Residence time: 10 min.

e Perfluorodecalin (0.5 mL) was injected before and after the reaction mixture as a fluorous spacer.

f Solvent concentration of 0.150 M vs. 1 was used.

g Residence time: 20 min.

The optimization of the reaction was performed on the commercially available Vapourtec R-Series flow system along with a 10 mL copper coil[16] employing 4-iodotoluene as the model substrate. A 4:1 ratio of DMSO–H2O was adopted as the transport solvent using standard segmented flow techniques (Table [1]).[17] We initially varied the solvent source and mixture ratio and found that the reaction proved to be sensitive to the solvent employed (entries 1–3), as observed in related precedents.[5] [6] [7] In the absence of any ligand, a 1:1 mixture of DMSO–H2O appeared to be the optimal solvent ratio, providing 64% yield with a residence time of only 15 minutes (entry 2). This interesting result led us to gain further insight on this ligand-free procedure. A control reaction was performed with a stainless steel coil to verify the hypothesis that the copper tubing was indeed promoting the reaction (entry 4). Screening different temperatures (entries 5–7) with a 10 minute residence time allowed the optimal temperature (165 °C) for this reaction to be determined. At this point, the yield for the desired product reached a 75% plateau (entry 6), with 16% remaining starting material as detected by analysis of the crude mixture by 1H NMR analysis. To our surprise, when the residence time was increased to 15 minutes (entry 8), the yield obtained was comparable to that observed after a 10 minute run. A possible explanation for this phenomenon is an increased level of degradation of n-Bu4NOH after 10 minutes heating. The presence of remaining starting material could be rationalized by the diffusion of the reaction mixture into the transport solvent.[18] To investigate this theory, an immiscible solvent spacer (perfluorodecalin) was used between the reaction mixture and the transport solvent.[19]

This modification resulted in an increase in yield (87%; Table [1] entry 9) and supported the conclusion that diffusion in a segmented flow reaction was important. Gratifyingly, it was also possible to reduce the temperature to 150 °C while increasing the residence time to 20 minutes to obtain a comparable yield (85%; entry 12). Increasing or decreasing the stoichiometry of n-Bu4NOH resulted in significant reduction in yield (entries 10 and 11), so the use of 6 equivalents was deemed optimal.

With the optimized conditions in hand, the scope of the reaction was then explored on a gram scale (Table [2]). In this manner, it was possible to benefit from the robustness of a scale-out process using flow technology. Furthermore, the use of a solvent spacer was no longer required because diffusion became negligible on this scale. Activated aryl iodides bearing electron-withdrawing groups (entries 2, 5, and 9–10) were readily converted into their corresponding phenols in only 5 minutes residence time within the reactor (conditions A: 5 min, 165 °C). Interestingly, electron-neutral and slightly electron-rich substrates (entries 1, 3–4, and 6–8) were synthesized in higher yields when the temperature was reduced (conditions B: 20 min, 150 °C). Unfortunately, applying these conditions to a more electron-rich substrate led to a poor yield (40%; entry 11), indicating that degradation of n-Bu4NOH competed with the desired reaction.

Table 2 Gram-Scale Flow Hydroxylation of Aryl Iodides Using a Copper Reactora

Entry

Aryl iodide

Conditions

Yield (%)b

1

1a

B

81

2

1b

A

88

3

1c

B

84

4

1d

B

78

5

1e

A

90

 6

1f

B

86

 7

1g

B

76

 8

1h

B

87

 9

1i

A

83c

10

1j

A

64 (75)d,e

11

1k

B

40

a Reaction conditions: aryl iodide (4.63 mmol), n-Bu4NOH (27.79 mmol, 6 equiv), 1:1 DMSO–H2O mix (0.125 M vs. 1).

b Isolated yield.

c Reaction mixture was kept in an ultrasonic bath during injection.

d A 4 minute residence time was used.

e Aryl iodide (1.00 g, 3.03 mmol) was used.

The direct mono-hydroxylation of a diiodobenzene compound has never been previously achieved, except for certain specific substrates.[20] In fact, this transformation is usually performed using a three-step procedure employing a halogen-exchange reaction followed by trapping with a boron source and hydrolysis through an oxidative quench.[21] In this instance, an experimentally easier and shorter alternative method would be desirable. We were pleased to see that our standard conditions could offer a solution to this problem because the desired mono-phenol could be prepared in 64% yield (conditions A; Table [2], entry 10). The yield could be improved to 75% by simply reducing the residence time to 4 minutes, constituting the first practical selective direct hydroxylation of a di-iodobenzene derivative.

To demonstrate further the utility of this methodology, the synthesized phenoxide ion was quenched in situ with benzyl bromide to make the corresponding aryl ether in an unoptimized 84% yield (Scheme [1]). This two-step procedure was performed with a second reactor connected in series with the hydrolysis output. This hydrolysis–benzylation sequence required only 14 minutes residence time in the reactors and illustrates well how powerful flow chemistry can be as compared with a similar flask reactor. Indeed, the batchwise equivalent usually requires a cooling step before quenching, extra amounts of base, a catalytic amount of tetrabutylammonium iodide, and long reaction times (4–21 h for the nucleophilic substitution step), which makes it comparatively less appealing on a large scale.[6] [7a] [e]

Zoom Image
Scheme 1 Example of trapping of the phenoxide ion in situ within the flow system

In conclusion, a straightforward continuous-flow ligand-free hydroxylation of aryl iodides promoted by copper tubing was developed.[22] [23] The presented methodology offers a rapid synthesis of phenols (4–20 min in the reactor) including the first efficient direct mono-hydroxylation of a di-iodobenzene derivative. Moreover, the use of a flow system allowed the safe and facile quench of the phenoxide ion in situ by an electrophile in only a few minutes and without extra cooling time.


#

Acknowledgment

This work has been supported by the Université de Montréal, the Natural Science and Engineering Council of Canada (NSERC), the Canada Research Chair Program, the Canada Foundation for Innovation and the Centre in Green Chemistry and Catalysis (CGCC). P.C. is grateful to NSERC and the Université de Montréal for postgraduate scholarships.

Supporting Information

    for this article is available online at http://www.thieme-connect.com.accesdistant.sorbonne-universite.fr/products/ejournals/journal/ 10.1055/s-00000083.
  • Supporting Information

  • References

  • 3 Anderson KW, Ikawa T, Tundel RE, Buchwald SL. J. Am. Chem. Soc. 2006; 128: 10694
    CrossrefPubMedSearch in Google Scholar
  • 5 Tlili A, Xia N, Monnier F, Taillefer M. Angew. Chem. Int. Ed. 2009; 48: 8725
    Search in Google Scholar
  • 6 Zhao D, Wu N, Zhang S, Xi P, Su X, Lan J, You J. Angew. Chem. Int. Ed. 2009; 48: 8729
    CrossrefSearch in Google Scholar
  • 8 For an example using microwave heating, see: Mehmood A, Leadbeater NE. Catal. Commun. 2010; 12: 64
    CrossrefSearch in Google Scholar
  • 9 Mousseau JJ, Bull JA, Charette AB. Angew. Chem. Int. Ed. 2010; 49: 1115
    CrossrefSearch in Google Scholar
  • 11 He Z, Jamison TF. Angew. Chem. Int. Ed. 2014; 53: 3353
    CrossrefPubMedSearch in Google Scholar
  • 13 Kormos CM, Leadbeater NE. Tetrahedron 2006; 62: 4728
    CrossrefSearch in Google Scholar
  • 14 Bogdan AR, Sach N. Adv. Synth. Catal. 2009; 351: 849
    CrossrefSearch in Google Scholar
  • 15 Zhang Y, Jamison TF, Patel S, Mainolfi N. Org. Lett. 2011; 13: 280
    CrossrefSearch in Google Scholar
  • 16 Copper coil flow reactors are available from Vapourtec Ltd., see: www.vapourtec.co.uk.
  • 17 See the Supporting Information for more details.
  • 18 Song H, Tice JD, Ismagilov RF. Angew. Chem. Int. Ed. 2003; 42: 767
    Search in Google Scholar
  • 22 Phenol Synthesis; General Procedure: A solution of the aryl iodide (4.83 mmol) in a mixture of aq n-Bu4NOH (1.5 M, 19.3 mL) and DMSO (19.3 mL, for an overall 0.125 M solution vs. aryl iodide) was prepared. The reaction solution (37 mL) was injected by using direct injection mode and the reagent stream was pumped into the 10-mL copper reactor (1.0 mm i.d.) at the desired temperature for the needed residence time, then 48 mL of the crude reaction solution was collected, acidified to pH 1 with 2 M HCl. H2O (150 mL) was added and the mixture was extracted with Et2O (3 × 150 mL). The combined organic layers were dried over anhydrous MgSO4, filtered, and concentrated under vacuum. The crude mixture was purified by flash chromatography (CH2Cl2–hexanes gradient) to afford the desired product (see the Supporting Information for more details).
  • 23 Benzyl-3-chlorophenyl Ether (3); Typical Procedure: A first solution (A) of 1-chloro-3-iodobenzene (1b; 0.50 g, 2.10 mmol) in a mixture of aq n-Bu4NOH (1.5 M, 8.4 mL) and DMSO (8.4 mL; for an overall 0.125 M solution vs. 1-chloro-3-iodobenzene) was prepared. A second solution (B) consisting of benzyl bromide (0.62 mL, 6.30 mmol) in DMSO (8 mL) was prepared. Solution A was injected using the direct injection mode and the reagent stream was pumped into the first reactor (10 mL copper coil of 1.0 mm i.d.) at 165 °C. Then, solution B was injected by using the direct injection mode and mixed in a T-mixer with solution A after the latter had passed the first reactor. The mixture then passed the second reactor (5 mL PFA coil of 1.0 mm i.d.) at 150 °C. 23 mL of the crude reaction solution was then collected, diluted with H2O (100 mL) and extracted with Et2O (3 × 100 mL). The combined organic layers were dried over anhydrous MgSO4, filtered, and concentrated under vacuum. The crude mixture was purified by flash chromatography (CH2Cl2–hexanes, 0–30%) to afford 3 (385 mg, 84% yield) as a yellow solid (see the Supporting Information for more details).

  • References

  • 3 Anderson KW, Ikawa T, Tundel RE, Buchwald SL. J. Am. Chem. Soc. 2006; 128: 10694
    CrossrefPubMedSearch in Google Scholar
  • 5 Tlili A, Xia N, Monnier F, Taillefer M. Angew. Chem. Int. Ed. 2009; 48: 8725
    Search in Google Scholar
  • 6 Zhao D, Wu N, Zhang S, Xi P, Su X, Lan J, You J. Angew. Chem. Int. Ed. 2009; 48: 8729
    CrossrefSearch in Google Scholar
  • 8 For an example using microwave heating, see: Mehmood A, Leadbeater NE. Catal. Commun. 2010; 12: 64
    CrossrefSearch in Google Scholar
  • 9 Mousseau JJ, Bull JA, Charette AB. Angew. Chem. Int. Ed. 2010; 49: 1115
    CrossrefSearch in Google Scholar
  • 11 He Z, Jamison TF. Angew. Chem. Int. Ed. 2014; 53: 3353
    CrossrefPubMedSearch in Google Scholar
  • 13 Kormos CM, Leadbeater NE. Tetrahedron 2006; 62: 4728
    CrossrefSearch in Google Scholar
  • 14 Bogdan AR, Sach N. Adv. Synth. Catal. 2009; 351: 849
    CrossrefSearch in Google Scholar
  • 15 Zhang Y, Jamison TF, Patel S, Mainolfi N. Org. Lett. 2011; 13: 280
    CrossrefSearch in Google Scholar
  • 16 Copper coil flow reactors are available from Vapourtec Ltd., see: www.vapourtec.co.uk.
  • 17 See the Supporting Information for more details.
  • 18 Song H, Tice JD, Ismagilov RF. Angew. Chem. Int. Ed. 2003; 42: 767
    Search in Google Scholar
  • 22 Phenol Synthesis; General Procedure: A solution of the aryl iodide (4.83 mmol) in a mixture of aq n-Bu4NOH (1.5 M, 19.3 mL) and DMSO (19.3 mL, for an overall 0.125 M solution vs. aryl iodide) was prepared. The reaction solution (37 mL) was injected by using direct injection mode and the reagent stream was pumped into the 10-mL copper reactor (1.0 mm i.d.) at the desired temperature for the needed residence time, then 48 mL of the crude reaction solution was collected, acidified to pH 1 with 2 M HCl. H2O (150 mL) was added and the mixture was extracted with Et2O (3 × 150 mL). The combined organic layers were dried over anhydrous MgSO4, filtered, and concentrated under vacuum. The crude mixture was purified by flash chromatography (CH2Cl2–hexanes gradient) to afford the desired product (see the Supporting Information for more details).
  • 23 Benzyl-3-chlorophenyl Ether (3); Typical Procedure: A first solution (A) of 1-chloro-3-iodobenzene (1b; 0.50 g, 2.10 mmol) in a mixture of aq n-Bu4NOH (1.5 M, 8.4 mL) and DMSO (8.4 mL; for an overall 0.125 M solution vs. 1-chloro-3-iodobenzene) was prepared. A second solution (B) consisting of benzyl bromide (0.62 mL, 6.30 mmol) in DMSO (8 mL) was prepared. Solution A was injected using the direct injection mode and the reagent stream was pumped into the first reactor (10 mL copper coil of 1.0 mm i.d.) at 165 °C. Then, solution B was injected by using the direct injection mode and mixed in a T-mixer with solution A after the latter had passed the first reactor. The mixture then passed the second reactor (5 mL PFA coil of 1.0 mm i.d.) at 150 °C. 23 mL of the crude reaction solution was then collected, diluted with H2O (100 mL) and extracted with Et2O (3 × 100 mL). The combined organic layers were dried over anhydrous MgSO4, filtered, and concentrated under vacuum. The crude mixture was purified by flash chromatography (CH2Cl2–hexanes, 0–30%) to afford 3 (385 mg, 84% yield) as a yellow solid (see the Supporting Information for more details).

   Permissions and Reprints All articles of this category
Zoom Image
Scheme 1 Example of trapping of the phenoxide ion in situ within the flow system