5252
Ind. Eng. Chem. Res. 2007, 46, 5252-5257
APPLIED CHEMISTRY Self-Neutralizing in Situ Acid Catalysis for Single-Pot Synthesis of Iodobenzene and Methyl Yellow in CO2-Expanded Methanol Ross R. Weikel, Jason P. Hallett, Charles L. Liotta,* and Charles A. Eckert* Schools of Chemical & Biomolecular Engineering and Chemistry and Biochemistry and Specialty Separations Center, Georgia Institute of Technology, Atlanta, Georgia 30332-0100
Despite widespread use, homogeneous acid catalysis has the drawback of requiring downstream neutralization, resulting in salt waste. Methylcarbonic acid is a self-neutralizing acid that forms in situ in methanol/CO2 systems at mild pressures (10-47 bar) and then decomposes by depressurization. In the work presented here, methylcarbonic acid catalyzes the diazotization of aniline, which is either coupled with N,N-dimethyl aniline to form methyl yellow or reacted with iodide to form iodobenzene. The syntheses of methyl yellow and iodobenzene represent a class of industrially important reactions. 1. Introduction The diazotization of aniline and aniline derivatives is a useful reaction for a wide range of products. The vast majority of all processes use a strong acid such as HCl or H2SO4 to promote the diazotization reaction.1-3 At the end of the reaction, base must be added to neutralize the mineral acid. This results in significant quantities of contaminated salts that must be removed from the process stream and treated as waste. The disposal of waste salts represents a growing problem with the industrial use of mineral acids, which represent the most common of industrial catalytic processes. Therefore, there is significant demand for an easily recycled, environmentally benign acid. Our approach to this problem has been the development of selfneutralizing acids that provide catalysis without producing large quantities of salt waste as waste products. The implementation of alternative acid technologies, such as solid acids and selfneutralizing acids, would potentially provide both economic and environmental benefits for a wide variety of large-scale industrial chemical processes. The strong acid used in the diazotization process reacts with a nitrite to form nitrous acid in situ. Nitrous acid releases the active species, nitrosyl cation, which reacts with an amine to form the diazonium cation through several steps. Recent research has aimed to simplify this process by using NO directly as a gas4 or as NOCl,5 but these methods have not gained industrial acceptance. After the diazonium ion is formed, there are at least two possible outcomes: substitution with a nucleophile or coupling of the diazonium with an activated arene. The former is represented in this article by the synthesis of iodobenzene and the latter by synthesis of methyl yellow (N,N-dimethyl-4aminoazobenzene). Activated arenes, including N,N-dialkylanilines and phenol derivatives, are used to produce a wide variety of azo products, which are commonly used in dyes for textiles6,7 and inks for jet printing.8 The textile industry has recently faced pressure to reduce the amount of electrolytes and colored wastewater * To whom correspondence should be addressed. Tel.: 404-8947070. Fax: 404-894-9085. E-mail:
[email protected].
Scheme 1. Alkylcarbonic Acid Formation and Proton Dissociation Equilibria
released into the environment, thus stimulating research into ways to alleviate these concerns. These ideas include use of ion-exchange resins;9 solid-solid reactions;5 polyethylene glycol solvent;10 clay catalysts,7 polymers,11 or resins;12 or carbonic acid from water and CO2 for catalysis.6,13,14 The reaction of diazonium salts to form chlorobenzene, bromobenzene, or benzonitrile is catalyzed by copper salts, whereas other substitution reactions take place simply with alkali metal salts of the nucleophilic anion. In the particular case of iodobenzene, only an iodide salt (generally sodium or potassium iodide) is needed to form the product from the diazonium salt. The work reported here proposes the use of methylcarbonic acid from methanol and CO2 (Scheme 1) to catalyze the synthesis of diazonium cation. The use of CO2 as a benign component of a reaction medium is well-known.15,16 The development of gas-expanded liquids (GXLs) from CO2 and organic solvents has also seen significant research for both reactions17,18 and separations.19,20 Alkylcarbonic acids were first observed in alcohol-CO2 GXL systems by West et al.21 and recently further characterized.22 Alkylcarbonic acids have been used in the catalysis of acetal formation,23 in the hydrolysis of β-pinene,24 and in Ugi reactions.25 Alkylcarbonic acids form when alcohols are in the presence of CO2 (even at low pressure) and naturally decompose when CO2 is removed. There is therefore no need for addition of a base to neutralize the acid catalyst, thus eliminating the contaminated salt waste normally produced when neutralization of mineral acid catalysts is required.
10.1021/ie061386a CCC: $37.00 © 2007 American Chemical Society Published on Web 07/11/2007
Ind. Eng. Chem. Res., Vol. 46, No. 16, 2007 5253 Scheme 2. Synthesis and Mechanism for Formation of Methyl Yellow from Aniline
2. Experimental Methods 2.1. Materials. All chemicals were obtained from Acros and used as purchased unless otherwise stated. These chemicals included aniline (99.5%), methanol (HPLC-grade, 99.93%), sodium nitrite (Sigma, 99.5%), N,N-dimethyl aniline (99%), potassium iodide (Sigma, 99.0%), toluene internal standard (HPLC), tetrahydrofuran (anhydrous 99.9%), and isoamyl nitrite (Sigma, 97%). The carbon dioxide (SFC/SFE grade) was obtained from Airgas and was filtered prior to use. 2.2. Reactions. All reactions were run in a 300 mL Parr model 4560 pressure vessel with windows and overhead stirring.
The temperature was regulated to within 1 °C of the set point using the accompanying Parr 4842 controller. The windows allowed for simple visual observation of dye formation from colorless reactants. All reactions were run for 24 h with the following loading unless specifically noted otherwise: 140 mL of methanol (3.9 mol), 0.6 mL of aniline (6.5 mmol), and 1.0 g of sodium nitrite (14.5 mmol). The coupling reaction systems contained 2.5 mL of N,N-dimethyl aniline (19.6 mmol), whereas the halogenation reaction systems contained 3.3 g of potassium iodide (19.9 mmol). For conditions containing different excess sodium nitrite loadings, the aniline concentration was held
5254
Ind. Eng. Chem. Res., Vol. 46, No. 16, 2007
Scheme 3. Carbamate Acid Formation from Aniline and CO2
Scheme 4. Deadiazotization of Diazobenzene
Figure 1. Effect of nitrite loading on yield at T ) 5 °C and CO2 loading of 0.54 mol, corresponding to P ) 10 bar.
Table 1. Results from the Synthesis of Methyl Yellow Using Methylcarbonic Acida NaNO2 (equiv)
T (°C)
CO2 (mol)
conversion (%)
yield (%)
1.04 2.2 2.2 2.2 2.2
5 5 5 25 50
0.54 0.54 1.25 1.25 1.25
93.0 ( 4.8 94.3 ( 2.0 97.5 ( 0.2 98.1 ( 0.7 99.5 ( 0.4
20.1 ( 7.0 63.3 ( 1.5 72.0 ( 4.3 65.3 ( 4.5 25.5 ( 3.1
Figure 2. Effect of CO2 loading on yield at T ) 5 °C and with 2.2 equiv of NaNO2.
a All reactions used methanol as the solvent. Control results noted in the text. Standard deviations are the results of three independent experiments.
Table 2. Results from the Synthesis of Iodobenzene Using Methylcarbonic Acida NaNO2 (equiv)
T (°C)
CO2 (mol)
conversion (%)
yield (%)
1.04 2.2 2.2 2.2 2.2
5 5 5 25 50
0.54 0.54 1.25 1.25 1.25
95.0 ( 2.0 95.0 ( 2.0 94.0 ( 2.0 99.7 ( 0.1 100 ( 0.0
27.0 ( 2.0 30.0 ( 2.0 30.5 ( 2.0 56.8 ( 2.0 72.2 ( 0.7
a All reactions used methanol as the solvent. Control results noted in the text. Standard deviations are the results of three independent experiments.
constant at 6.5 mmol, and the sodium nitrite loading was varied. All reactants were loaded into the autoclave and then heated or cooled to temperature at a stirring rate of 725 rpm for 1 h. Then, CO2 was added (either 0.54 mol for low loading or 1.25 mol for high loading) by an ISCO 500D syringe pump with series D controller, and the reaction started. The reaction was run at constant temperature, pressure, and stirring speed for 24 h, and then the reaction system was depressurized, which generally required approximately 1 h. Next, 0.69 mL of toluene internal standard (6.5 mmol) was added, and samples were collected. These samples were analyzed using a Hewlett-Packard 6890 gas chromatograph with flame ionization detector (GC-FID) for quantification of yield. Samples were also run on a HewlettPackard GC with mass-selective detector 5972 (GC-MS) for peak indentification. All yields reported are averages of three runs. 3. Results and Discussion 3.1. Synthesis of Methyl Yellow. Methyl yellow was synthesized in a single-pot reaction from aniline, sodium nitrite, N,N-dimethyl aniline, and CO2 in methanol as shown in Scheme 2. Aniline was the limiting reagent, with both the nitrite and N,N-dimethyl aniline in excess. All reactants except CO2 were
Figure 3. Effect of temperature on yield with 2.2 equiv of NaNO2. The pressures varied from P ) 10 bar (5 °C, 0.54 mol of CO2) to P ) 47 bar (50 °C, 1.25 mol of CO2).
loaded and stirred for 1 h to allow the temperature to equilibrate. Then, CO2 was added, and the reaction began. The CO2 and methanol form methylcarbonic acid, which reacts with the nitrite to form nitrous acid. The nitrous acid then reacts with the aniline to form a diazonium salt that couples with N,N-dimethyl aniline to form methyl yellow. The most successful coupling reaction conditions were a temperature of 5 °C with excess nitrite salt and a high CO2 loading. The reaction was run for 24 h and produced an average yield of 72% methyl yellow and 97% conversion of aniline. Two different nitrite loadings, two different CO2 loadings, and three temperatures were run for the coupling reaction, all in triplicate. Also, several experiments were run to ensure that the catalysis was due to methylcarbonic acid. The first experiment was to run the normal loading without any added CO2. This resulted in no visual color change and no measurable yield. Next, the reaction was run with tetrahydrofuran (THF) as the solvent rather than methanol because THF should not form an in situ acid with CO2. This reaction resulted in a yield of 0.3%, which was likely due to the formation of carbamates from aniline and CO226 (Scheme 3). To verify this result, the reaction was run without any solvent (in CO2-expanded aniline) and resulted in a methyl yellow yield of 14%. Thus, although the aniline/CO2
Ind. Eng. Chem. Res., Vol. 46, No. 16, 2007 5255 Scheme 5. Synthesis and Mechanism for Formation of Iodobenzene from Aniline
complex can catalyze the reaction to trace yield, in the methanol solvent system, the methylcarbonic acid clearly provides the major catalysis. The effects of the nitrite source and loading were also investigated. Initially, both isoamyl nitrite and sodium nitrite were used in slight excess. The isoamyl nitrite gave yields roughly one-half that obtained with sodium nitrite. The reason for this observation is not understood at this juncture. All further reactions were run with sodium nitrite. Next, the effect of varying the amount of excess sodium nitrite was investigated by comparing the yields and rates for a slight excess (1.04 times) and for a 2.2-times excess. The results in Figure 1 show a significant increase with the large excess of nitrite. The amount of CO2 loaded was also varied. A small loading of 0.54 mol was compared to a higher loading of 1.25 mol.
The low CO2 loading at low temperatures gave a pressure close to 10 bar, whereas a high CO2 loading and high temperature gave pressures as high as 47 bar. Figure 2 shows that the higher CO2 loading gave improved yields. This was expected because an increased CO2 concentration increases the alkylcarbonic acid concentration. However, there was the possibility that the increased CO2 concentration could sufficiently lower the dielectric constant of the solvent to hinder proton dissociation.22 A range of temperatures was also employed for the coupling reaction. Typical diazotizations are performed at low temperature because of the instability of the diazonium salt. After the coupling agent or substitution reactant is added, the temperature is allowed to rise to increase the reaction rate. Therefore, in a one-pot synthesis, it is assumed that a low temperature will do best. However, this was not the case in the coupling to form
5256
Ind. Eng. Chem. Res., Vol. 46, No. 16, 2007
N,N-diethyl-4-[(4-nitrophenyl)azo]aniline (DENAB) as reported by Hooker et al., where the highest yields were found at 80 °C.6 Figure 3 shows the results that we obtained using methylcarbonic acid at temperatures of 5, 25, and 50 °C, with the highest yields at 5 °C and decreasing yield with increased temperature. This shows that the stability of the diazonium cation is limiting compared to the reaction kinetics in this case. In the case of the iodobenzene synthesis, the Arrhenius effect on the rate constant is dominant; therefore, the yield does increase uniformly with temperature. Additionally, the reaction was run under conditions corresponding to those of Hooker et al., who synthesized DENAB in a water/CO2 system with carbonic acid in yields up to 91%.6 When the reaction setup of Hooker et al. (CO2, 25 °C, 59 bar, 0.02 M in each reactant) was replicated for methyl yellow, the resulting yield was only 11%. This is despite the lower water solubility of aniline compared to 4-nitroaniline,27 and we therefore conjecture that the lack of the activating 4-nitro group on the aromatic ring decreases the diazonium ion yield when aniline is the substrate. The Hooker et al. reaction setup could not be duplicated with DENAB using our analytical procedure because the high melting point of DENAB prevented it from being run through a gas chromatograph and the substrate and product peaks overlapped in the UV-vis spectrum. All reactions had similar byproducts and product ratios. The major byproduct from all reactions by GC-MS was benzene formed through dediazotization of the diazonium cation (see Scheme 4), which was expected.28 Other common byproducts were azobenzene, diphenyl amine, aminobiphenyl, and biphenyl, although all were generally present in small amounts. The conversion of aniline was reproducible for each set of reaction conditions, with the value varying only from 93% to 99% for all coupling reaction conditions. A summary of all methyl yellow results is provided in Table 1. 3.2. Synthesis of Iodobenzene. The synthesis of iodobenzene proceeded just as the coupling reaction in a single pot with potassium iodide in place of N,N-dimethyl aniline (Scheme 5). All reactants but CO2 were loaded into the pressure vessel and stirred for 1 h to reach thermal equilibrium. When the CO2 was added, the reaction began, and the same process occurred to form the diazonium ion. In this case, the diazonium ion reacted with the nucleophilic iodide to form iodobenzene. The most effective conditions for the synthesis of iodobenzene were similar to those for the coupling reaction except for the temperature. The best yield was an average of 72% for three runs at a high nitrite loading (2.2 times excess) and CO2 loading (1.25 mol) and at high temperature (50 °C). The effects of nitrite, temperature, and CO2 loading were tested in triplicate as well. Additionally, several runs were carried out to show that alkylcarbonic acids were catalyzing the reaction. The normal loading at 5 °C without CO2 resulted in a yield of 5%, which, though higher than expected, is still quite small. The reaction was run in tetrahydrofuran with CO2 pressure, and no product was observed. This suggests that perhaps the methanol is acidic enough to catalyze the reaction slightly but that alkylcarbonic acids significantly increase the rate. The effect of nitrite loading was not as pronounced for iodobenzene as it was for methyl yellow. The increase in yield shown in Figure 1 was very slight (3%). Because the solubility of potassium iodide in methanol is much lower than that of N,N-dimethyl aniline, an increase in loading of the salt KI would not have a significant impact on the actual concentration in solution, as opposed to the case of the organic reagent N,Ndimethyl aniline. The results for increased CO2 loading agree
with this hypothesis because more CO2 (and thus more acid) did not have an effect on the yield (Figure 2). However, for iodobenzene, an increase in temperature had a significant impact on yield, as shown in Figure 3. The yield increased with temperature, which could be an effect of both an increased reaction rate and an improved potassium iodide solubility. The only observed byproduct for the substitution reactions was benzene. The conversion of aniline was reproducible but again only varied from 94% to 100% for all reactions. A summary of all iodobenzene results is provided in Table 2. Some limited conversion occurred in the “blank” experiments. An 11% yield was obtained when no CO2 was added (5 °C, 24 h), and 0% yield was observed when THF was the solvent instead of methanol (5 °C, 24 h, 13 bar CO2). The limited yield in the absence of CO2 indicates that methanol is protic enough to effect a small amount of diazonium formation. The lack of conversion in the THF/CO2 system indicates that an acid source must be present for coupling to occur. Further, no coupling was observed in the absence of sodium nitrite, and trace azobenzene yield was observed at long reaction times (>24 h), strongly supporting the stated necessity for diazonium ion formation to precede coupling. Conclusion Methylcarbonic acid is demonstrated as a green substitute for strong acids for the diazotization of aniline and subsequent coupling to form methyl yellow or substitution to form benzyl iodide. The use of methylcarbonic acid eliminates the salt waste associated with use of strong acids generally employed in the synthesis of these chemicals. Several variables were examined to improve yields, and optimal conditions for both reactions were found at reasonable temperatures and pressures (below the critical pressure of CO2). The demonstrated utility of alkylcarbonic acids could lead to other applications as a benign substitute for mineral acids. Acknowledgment The authors acknowledge the Department of Energy (DEFG02-04ERI5521) and J. Erskine Love Institute Chair for their financial support. The authors also thank Craig Simpson and Kyle Ross for their laboratory assistance. Supporting Information Available: Typical GC-MS trace with the product and reactant peaks identified. This material is available free of charge via the Internet at http://pubs.acs.org. Literature Cited (1) Hartmann, H.; Zug, I. On the Coupling of Aryldiazonium Salts with N,N-Disubstituted 2-Aminothiophenes and Some of their Carboxylic and Heterocyclic Analogues. J. Chem. Soc., Perkin Trans. 1 2000, 4316. (2) Heemstra, J. M.; Moore, J. S. A Novel Indicator Series for Measuring pKa Values in Acetonitrile. Tetrahedron 2004, 60, 7287. (3) Leverenz, K. Process for Diazotizing Aromatic Amines. U.S. Patent 4,250,089, 1981. (4) Itoh, T.; Nagata, K.; Matsuya, Y.; Miyazaki, M.; Ohsawa, A. Reaction of Nitric Oxide with Amines. J. Org. Chem. 1997, 62, 3582. (5) Kaupp, G.; Herrmann, A.; Schmeyers, J. Waste-Free Chemistry of Diazonium Salts and Benign Separation of Coupling Products in Solid Salt Reactions. Chem. Eur. J. 2002, 8, 1395. (6) Hooker, J.; Hinks, D.; Montero, G.; Conlee, C. Synthesis of N,NDiethyl-N-{4-[(E)-(4-nitrophenyl)diazenyl]phenyl}Amine via in Situ Diazotisation of Coupling in Supercritical Carbon Dioxide. Color Technol. 2002, 118, 273. (7) Bahulayan, D.; John, L.; Lalithambika, M. Modified Clays as Efficient Acid-Base Catalyst Systems for Diazotization and Diazocoupling Reactions. Synth. Commun. 2003, 33, 863.
Ind. Eng. Chem. Res., Vol. 46, No. 16, 2007 5257 (8) Heinz, P. Dyes for Ink Jet Printing. U.S. Patent 5,989,326, 1999. (9) Wolff, J.; Wolf, K.; Klipper, R. M.; Lange, P. M. Process For Preparing Aromatic Diazonium Salts by Diazotization of Amines. U.S. 4,754,020, 1988. (10) Suzuki, N.; Azuma, T.; Kaneko, Y.; Izawa, Y.; Tomioka, H.; Nomoto, T. Diazotization and Sandmeyer Reactions of Arylamines in Poly(ethylene glycol)-Methylene Dichloride: Usefulness of PEG in Synthetic Reactions. J. Chem. Soc., Perkin Trans. 1 1987, 645. (11) Calderelli, M.; Baxendale, I. R.; Ley, S. V. Clean and Efficient Synthesis of Azo Dyes Using Polymer-Supported Reagents. Green Chem. 2000, 2, 43. (12) Merrington, J.; James, M.; Bradley, M. Supported diazonium saltss Convenient reagents for the combinatorial synthesis of azo dye. Chem. Commun. 2002, 140. (13) Raue, R.; Brack, A.; Lange, K.-H. Process for the Preparation of Dyestuffs. U.S. Patent 5,578,711, 1996. (14) Raue, R.; Brack, A.; Lange, K.-H. Process for the Preparation of Dyestuffs. U.S. Patent 5,541,299, 1996. (15) Eckert, C. A.; Knutson, B. L.; Debenedietti, P. G. Supercritical Fluids as Solvents for Chemical and Materials Processing. Nature 1996, 383, 313. (16) Eckert, C. A.; Bush, D.; Brown, J. S.; Liotta, C. L. Tuning Solvents for Sustainable Technology. Ind. Eng. Chem. Res. 2000, 39, 4615. (17) Musie, G.; Wei, M.; Subramaniam, B.; Busch, D. H. Catalytic Oxidations in Carbon Dioxide-Based Reaction Media, Including Novel CO2Expanded Phases. Coord. Chem. ReV. 2001, 219-221, 789. (18) Subramaniam, B.; Lyon, C. J.; Arunajatesan, V. Environmentally benign multiphase catalysis with dense phase carbon dioxide. Appl. Catal. B 2002, 37, 279. (19) Eckert, C. A.; Liotta, C. L.; Bush, D.; Brown, J. S.; Hallett, J. P. Sustainable Reactions in Tunable Solvents. J. Phys. Chem. B 2004, 108, 18108. (20) Xie, X.; Brown, J. S.; Joseph, P. J.; Liotta, C. L.; Eckert, C. A. Phase-Transfer Catalyst Separation by CO2 Enhanced Aqueous Extraction. Chem. Commun. 2002, 1156.
(21) West, K. N.; Culp, C.; McCarney, J.; Griffith, K.; Bush, D.; Liotta, C. L.; Eckert, C. A. In Situ Formation of Alkylcarbonic Acid with CO2. J. Phys. Chem. A 2001, 105, 3947. (22) Weikel, R. R.; Hallett, J. P.; Liotta, C. L.; Eckert, C. A. Self-Neutralizing in Situ Acid Catalysts from CO2. Top. Catal. 2005, 37, 75. (23) Xie, X.; Liotta, C. L.; Eckert, C. A. CO2-Catalyzed Acetal Formation in CO2-Expanded Methanol and Ethylene Glycol. Ind. Eng. Chem. Res. 2004, 43, 2605. (24) Chamblee, T. S.; Weikel, R. R.; Nolen, S. A.; Liotta, C. L.; Eckert, C. A. Reversible in Situ Acid Formation for β-Pinene Hydrolysis Using CO2 Expanded Liquids and Hot Water. Green Chem. 2004, 6, 382. (25) Hulme, C.; Ma, L.; Romano, J. J.; Morton, G.; Tang, S.-Y.; Cherrier, M.-P.; Choi, S.; Salvino, J.; Labaudiniere, R. Novel applications of carbon dioxide/MeOH for the synthesis of hydantions and cyclic ureas via the Ugi reaction. Tetrahedron Lett. 2000, 41, 1889. (26) Salvatore, R. N.; Flanders, V. L.; Ha, D.; Jung, K. W. Cs2CO3Promoted Efficient Carbonate and Carbamate Synthesis on Solid Phase. Org. Lett. 2000, 2, 2797. (27) Abraham, M. H.; Lee, J. The Correlation and Prediction of the Solubility of Compounds in Water Using an Amended Solvation Energy Relationship. J. Pharm. Sci. 1999, 88, 868. (28) DeTar, D. F.; Kosuge, T. Mechanisms of Diazonium Salt Reactions. VI. The Reactions of Diazonium Salts with Alcohols under Acidic Conditions: Evidence for Hydride Transfer. J. Am. Chem. Soc. 1958, 80, 6072.
ReceiVed for reView October 27, 2006 ReVised manuscript receiVed March 29, 2007 Accepted June 4, 2007 IE061386A
