Research Article pubs.acs.org/journal/ascecg
Formation of a Renewable Amine and an Alcohol via Transformations of 3‑Acetamido-5-acetylfuran Yi Liu, Cosima Staḧ ler, Jennifer N. Murphy, Brandon J. Furlong, and Francesca M. Kerton* Department of Chemistry, Memorial University of Newfoundland, 283 Prince Philip Drive, St. John’s, Newfoundland and Labrador A1B 3X7, Canada S Supporting Information *
ABSTRACT: The reactivity of the renewable amide 3acetamido-5-acetylfuran (3A5AF) was explored. Hydrolysis of the amido group to yield the amino-substituted furan, 2acetyl-4-aminofuran (1), was achieved via NaOH catalysis. Reduction of the acetyl group could be achieved stoichiometrically using NaBH4 or catalytically via transfer hydrogenation using an Ir catalyst. The product alcohol, 3acetamido-5-(1-hydroxylethyl)furan (2), underwent dehydration during analysis via GC-MS to yield an alkene (3). The potential reactivity of 3A5AF and 1 toward carbon dioxide was studied, but no reaction was observed due to the inherent acidity of 3A5AF and 1 despite the latter being an amine. The computationally determined pKa values for 3A5AF and 1 are reported. Interestingly, in this work, tautomerism of 3A5AF was observed in CD3OD as evidenced by H−D exchange within the acetyl group. KEYWORDS: Biomass, Furans, Reduction, Hydrolysis
■
INTRODUCTION In recent years, the use of renewable feedstocks to produce chemicals and biofuels has become a significant research area,1−5 due to the depletion of fossil fuels and climate change. So far, most studies about biomass transformations are limited to carbohydrates containing only C, H, and O atoms, and there is a lack of research on heteroatom-containing carbohydrates.6,7 Chitin is the second most abundant biopolymer after cellulose on earth. N-Acetyl-D-glucosamine (NAG) is the monomer of chitin and has been successfully converted to a nitrogencontaining product, 3-acetamido-5-acetylfuran (3A5AF) (Scheme 1). 3A5AF was first obtained from NAG using pyrolysis methods with quite low yields of 2 and 0.04%.8,9 In recent research by Omari et al., the yield of 3A5AF was increased to 60% through microwave irradiation with dimethylacetamide as the solvent and sodium chloride and boric acid as additives.10 In a study by Drover et al., ionic liquids were used as solvents in the conversion of NAG, and the highest 3A5AF yield achieved was also 60%.11 In 2014, Chen et al. performed the direct conversion of chitin to 3A5AF in Nmethyl-2-pyrrolidone, and the best yield obtained was 7.5%.12 In 2015, the same group switched to ionic liquids as solvents in the dehydration of chitin to 3A5AF.13 The use of 1-butyl-3methylimidazolium chloride together with boric acid and hydrochloric acid led to a maximum 3A5AF yield of 6.2%. To date, besides its preparation there is little information on the properties of 3A5AF,14 including its reactivity and conversion into other chemicals. As a promising platform chemical for a variety of useful compounds (Scheme 2), © 2017 American Chemical Society
reactions of 3A5AF should be investigated. Herein we report the formation of a renewable amine product, 2-acetyl-4aminofuran (1, Scheme 3), from the hydrolysis of 3A5AF. Also, the reduction of 3A5AF was performed using sodium borohydride (NaBH4) or via transfer hydrogenation, and an alcohol was obtained (2, Scheme 4). As amine-scrubbing technology is one of the main tools employed in CO2 capture, the potential reactivity of 3A5AF and 1 with CO2 was investigated. CO2 capture is an essential step in reducing emissions from the gas, oil, and chemical industries.
■
EXPERIMENTAL SECTION
Synthesis and Characterization of 2-Acetyl-4-aminofuran (1). First, 15.0 mg (0.0900 mmol) of 3A5AF was dissolved in 1 mL of methanol, and 3 mL of NaOH solution (concentrations as per Table 1) was added. The mixture was heated at the desired temperature in an oil bath under reflux for a fixed time. After the reaction was complete, 37% aqueous HCl was added dropwise to the reaction mixture until the solution was slightly acidic (i.e., a pH between 6 and 7 was obtained). Following this, 28% aqueous NH4OH was added to neutralize the solution. The solvent was evaporated on a Blowdown Evaporator at 55 °C. The obtained orange solid was redissolved in methanol and filtered. The orange filtrate was concentrated on a Rotavap at 300 mbar and 40 °C. The obtained orange solid was redissolved using EtOAc and filtered. The golden solution was concentrated on a Rotavap at 180 mbar and 40 °C until a golden solid Received: February 1, 2017 Revised: April 3, 2017 Published: April 13, 2017 4916
DOI: 10.1021/acssuschemeng.7b00323 ACS Sustainable Chem. Eng. 2017, 5, 4916−4922
Research Article
ACS Sustainable Chemistry & Engineering Scheme 1. Conversion of Chitin and NAG to Yield an Amido-furan (3A5AF)
Scheme 2. Proposed Reactions of 3A5AF
Table 1. Hydrolysis of 3A5AF Catalyzed by NaOH factors
A
B
C
entry
NaOH concentration (mol/L)
T (°C)
t (h)
conversion of 3A5AF (%)a
yield of 1 (%)a
1 2 3 4 5 6 7 8 9 10 11 12b
5 6 7 7 7 7 9 9 9 9 8 9
80 80 60 60 80 80 60 60 80 80 70 100
1 1 1 2 1 2 1 2 1 2 1.5 0.25
5.2 67.3 16.6 64.4 72.1 100 45.5 78.8 96.6 100 84.1 >99%
0 9.4 1.8 8.3 26.7 0 8.5 24.9 57.1 0 28.7 34.7
a
Determined by 1H NMR spectroscopy. bPerformed in a microwave reactor.
Scheme 3. 3A5AF Hydrolysis Catalyzed by NaOH
78 (4), 69 (6), 55 (12). HRMS calculated exact mass for 1 (C6H7NO2) = 125.0477, found = 125.0478, difference = 0.84 ppm. Selected IR data (cm−1) solid 3139.15, 3046.29 (N−H stretch), 1618.88 (CO stretch), 1570.92 (N−H bending). 3A5AF Reduction Using NaBH4. First, 31.5 mg (0.188 mmol) of 3A5AF was dissolved in 5 mL of dry ethanol in a Schlenk tube. Then, 21.2 mg (0.560 mmol) of NaBH4 was weighed in a Schlenk tube in a glovebox. After being transferred out of the glovebox, the NaBH4 solid was dissolved in dry ethanol under N2 flow. Under N2 flow and with stirring, the NaBH4 solution was added dropwise to the 3A5AF solution at 0 °C in an ice bath. Afterward, the ice bath was removed, and the starting materials were stirred at room temperature overnight under N2 flow. After the reaction, the mixture was filtered, and the filtrate was concentrated on a Rotavap at 130 mbar and 40 °C. A yellow oil was obtained. Two aliquots were taken and dissolved in 500 μL of EtOAc and 800 μL of CDCl3, respectively, for GC-MS and NMR analyses. For GC, 1 μL of the sample was injected through a 7683B Series Injector using a split mode of 50%. The GC separation was done using a DB5 column at a flow rate of 1 mL/min He 99.999%. The oven temperature was programmed as follows: 50 °C (hold 1 min), 65 °C/min to 215 °C, 2.5 °C/min to 225 °C, and 20 °C/min to 250 °C for 1.2 min. (The total run time was 10 min). Products were detected at m/z 50−250 range. Under these conditions, the retention times of 2 and 3 were 5.252 and 4.689 min, respectively.
was obtained. After being weighed, the solid was analyzed by GC-MS and NMR spectroscopy. In the reaction performed under microwave irradiation, the starting materials were placed in a 2−5 mL vial and heated in the microwave reactor at 100 °C for 15 min. In 1H NMR spectra of samples, the methyl groups of the acetyl moieties in the unreacted 3A5AF and 1 were integrated. The integration ratio of these peaks is equal to the molar ratio of unreacted 3A5AF to 1 in the solid obtained. Since there were no peaks of other impurities observed, it was assumed that the final solid was composed of only 1 and unreacted 3A5AF. Using this NMR-determined molar ratio of 1/3A5AF in combination with the weight of the solid, the masses of 1 and unreacted 3A5AF in the final product mixture were calculated, which were used to calculate the yield of 1 and the conversion of 3A5AF (Table 1). Characterization of 1. 1H NMR δH (298 K, 300 MHz; DMSO-d6; Me4Si) 2.09 (s, 3H), 6.33 (s, 1H), 8.39 (s, 1H), 9.11 (s, 1H), 11.43 (s, 1H). 13C NMR δC (298 K, 75 MHz; DMSO-d6) 9.65, 123.58, 126.51, 142.23, 149.94, 176.48. MS m/z (% ion) 125 (100), 124 (72), 96 (31),
Scheme 4. Synthesis of Furyl-alcohol 2 via (a) Reduction of 3A5AF with NaBH4 or (b) Transfer Hydrogenation of 3A5AF
4917
DOI: 10.1021/acssuschemeng.7b00323 ACS Sustainable Chem. Eng. 2017, 5, 4916−4922
Research Article
ACS Sustainable Chemistry & Engineering Characterization of 2. 1H NMR δH (298 K, 300 MHz; CDCl3; Me4Si) 1.44 (d, 3H), 2.05 (s, 3H), 2.75 (s, 1H), 4.71−4.78 (q, 1H), 6.12 (s, 1H), 7.82 (s, 1H), 7.99 (s, 1H). 13C NMR δC (298 K, 75 MHz; CDCl3) 21.24, 23.14, 63.54, 100.28, 124.72, 131.43, 156.35, 168.14. MS m/z (% ion) 169 (20), 151 (64), 127 (16), 109 (71), 97 (13), 80 (100), 68 (9), 53 (24). HRMS calculated exact mass for 2 (C8H11NO3) = 169.0739, found = 169.0746, difference = 0.7 ppm. Ir-Catalyzed Reduction of 5-HMF and 3A5AF. The reactions were performed using standard glovebox techniques (with circulation turned off during manipulation of the alcohol). 5-HMF or 3A5AF (10−200 mg) were weighed into a vial and dissolved in isopropyl alcohol. Solutions of the iridium complex IrH2Cl[(iPr2PC2H4)2NH] (4) (100 mg in 25 mL isopropyl alcohol) and potassium tert-butoxide (KOtBu) (300 mg in 25 mL isopropyl alcohol) were prepared. The active catalyst was prepared by mixing these two solutions together in a vial in a mole ratio of 1:15 Ir/KOtBu. The solution containing 5HMF or 3A5AF was added dropwise to the vial containing 4/KOtBu, at a substrate/Ir ratio of 100:1. The reaction mixture was stirred for 2.5 h at room temperature. After this, it was removed from the glovebox, filtered through a short plug of silica, rinsed with EtOAc, and the solvents were removed under vacuum. In the case of 2,5bis(hydroxymethyl)furan (BHMF), the resulting pale yellow solid was isolated in 95% yield, and analytical data were consistent with the literature.15 In the case of 2, a yellow oil was isolated (80% yield) with 1 H and 13C NMR spectra identical to those obtained via NaBH4 reduction described herein. Characterization of BHMF. 1H NMR δH (298 K, 300 MHz; C6D6; Me4Si) 4.46 (s, 4H, CH2OH), 6.02 (s, 2H, furyl). 13C NMR δC (298 K, 75 MHz; C6D6) 56.9 (2 × CH2OH), 107.9 (2 × 3,4-furyl), 155.0 (2 × 2,5-furyl) Attempted CO2 Capture Experiments. First 50.0 mg of dry 3A5AF (or 1) solid was dissolved in 1 mL of dry methanol-d4 (CD3OD) and analyzed by 1H and 13C NMR spectroscopy. Afterward, the sample was transferred to a Schlenk flask. The flask was freeze− pump−thawed three times to degas. In the first experiment, the solution was stirred at room temperature under CO2 flow for 1 h. The second test was performed at 40 °C in an oil bath for 1 h. Subsequently, 30 μL of deionized water was added into the solution. The mixture was stirred at room temperature and at 40 °C under CO2 flow, each for 1 h. After each experiment, the solution was analyzed by NMR spectroscopy, and the results were compared with the original NMR spectra of 3A5AF or 1. In order to further study the deuterium exchange within 3A5AF, the CD3OD was removed from the NMR sample using a Rotavap, and the solid obtained was dissolved in 5 mL of methanol. The solution was stirred overnight to allow the deuterium in 3A5AF to have sufficient exchange with protons in methanol. Then, the solvent was removed using a Rotavap, and the solid obtained was dissolved in 500 μL of EtOAc for GC-MS and 750 μL of CDCl3 for NMR analyses.
experiments an increase in NaOH concentration resulted in an increase in the yield of 1 (Table 1, entries 1, 2, 5 and 9), leading us to think that this might be a significant factor in the reaction process. Each factor (A−C) was studied at a high (+1) and low (−1) value: concentration of NaOH was investigated at 7 and 9 mol/L; temperature was investigated at 60 and 80 °C; and time was investigated at 1 and 2 h. The yield of 1 was analyzed as the response. The total number of experiments for the factorial design was 9 including one center point (Table 1, entry 11). In a 2-level factorial design, the resulting model can only be linear. Testing a center point where all factors are at the 0 level allows us to see if the model has curvature, i.e., has a higher order. In a typical reaction, 15 mg (0.090 mmol) of 3A5AF was dissolved in 1 mL of methanol. After 3 mL of the aqueous NaOH solution was added, the reaction mixture was heated in an oil bath under reflux at the desired temperature for a fixed time. After workup, a golden solid was obtained, which was analyzed by GC-MS, and 1H and 13C NMR spectroscopy. The highest yield of 1 obtained was 57.1% (6.02 mg), when the reaction was catalyzed by the 9 mol/L NaOH solution and was heated at 80 °C for 1 h (Table 1, entry 9). The results were analyzed using Design Expert. Analysis of variance (ANOVA) of the model with 5% significance showed that curvature was significant as well as the two-factor interaction between temperature and time (factors B and C). When both the temperature and time are low, the yield of 1 is low. When the temperature is high, the yield is high only when the reaction time is low. This is illustrated in Figure 1, and the
■
RESULTS AND DISCUSSION 3A5AF Hydrolysis: 23 Factorial Design. As a relatively simple reaction, the hydrolysis of 3A5AF was studied first. An initial screening was performed using a number of mineral acids and bases. From the initial test reactions, it became clear that sodium hydroxide (NaOH) gave superior reactivity in this reaction and so was studied in more detail. The hydrolysis of 3A5AF was performed in aqueous media catalyzed by NaOH (Scheme 3, Table 1). A Design of Experiments approach was used to ascertain the most important variables in this reaction and any potential interactions between them, which might be overlooked using a traditional one variable (factor) at a time approach. A 23 factorial design with one center point was used to study the effects of concentration of NaOH (factor A), temperature (factor B), and time (factor C). The incorporation of NaOH concentration as a factor for study was because in initial
Figure 1. Response surface plot of yield of 1 based on the factors temperature (B) and time (C) in the NaOH-catalyzed hydrolysis of 3A5AF.
response surface for the model shows that a yield of around 75% is predicted at a temperature of 80 °C and 1 h reaction time. Surprisingly, the concentration of NaOH (factor A) was not determined as a significant factor. This was different from our original expectation based on initial trial reactions. Therefore, a factorial design is important in designing and optimizing chemical processes since it allows researchers to effectively considers factor combinations and analyze all factors 4918
DOI: 10.1021/acssuschemeng.7b00323 ACS Sustainable Chem. Eng. 2017, 5, 4916−4922
Research Article
ACS Sustainable Chemistry & Engineering
The IR spectrum was obtained for a solid sample of 1 and compared with that of pure 3A5AF (Figure S11). 1 exhibits amine stretching bands (vN−H, stretch) at 3139.2 and 3046.3 cm−1 and an amine bending band (vN−H, bending) at 1570.9 cm−1. The peak at 1618.9 cm−1 in 1 was assigned to the carbonyl stretching frequency (vCO). In 3A5AF, the band at 1657.9 cm−1 is attributed to the combination of the vCO of the secondary amide and ketone functional groups. The shift of vN−H and vCO in 1 compared with 3A5AF toward lower frequencies may be the result of intermolecular hydrogen bonding. 3A5AF Reduction. The reduction of 3A5AF was performed with the aim of obtaining a N-containing compound with hydroxyl groups instead of carbonyl groups (Scheme 4). A mixture of 3A5AF and NaBH4 was stirred overnight under N2 flow at room temperature. After workup, a yellow oil was obtained and analyzed by GC-MS, HRMS, and 1H and 13C NMR spectroscopy. The use of NaBH4 led to the reduction of the carbonyl in the acetyl group to a hydroxyl. The compound obtained, 3-acetamido-5-(1-hydroxylethyl)furan (2), was previously observed in a study by Kuhn et al. in 1958.16 In the GCMS and NMR analyses no 3A5AF was detected, indicating its 100% conversion. However, the amide functional group remained intact. Another product, 3 (m/z 151.0 mol−1), was observed besides 2 (m/z 169.1 g mol−1) in GC-MS analysis, which resulted from the dehydration of 2. This dehydration happened in a similar way as was observed previously in the reaction of 3A5AF with a Grignard reagent.14 3 could not be seen in the NMR spectra possibly because its amount was too small or it was formed in situ within the inlet of the GC-MS instrument. In the 1H NMR spectrum of 2 (Figure S12), a broad peak at 2.75 ppm was assigned to δ(OH). In the 13C NMR spectrum (Figure S13), the reduced carbon atom had a chemical shift of 63.54 ppm. The reducing ability of NaBH4 is not sufficient to reduce the amide functional group. Therefore, lithium aluminum hydride (LiAlH4), a stronger reducing agent, was studied. The products obtained were analyzed by 1H and 13C NMR spectroscopy. The 13 C NMR spectrum showed that all significant signals were in the range of 10−70 ppm, indicating that the furan ring had opened to yield linear acyclic products. Unfortunately, a single product from the reaction mixture could not be isolated and characterized. However, it can be concluded that in the reaction using LiAlH4 instead of the generation of functional (aminoalcohol) furan products as expected the furan ring in 3A5AF was also reduced and ring-opened. Ir-catalyzed transfer hydrogenation of 3A5AF was also studied as a more efficient route to 2. As 3A5AF is not commercially available, the reaction was initially performed using 5-hydroxymethylfurfural (5-HMF) and then the same conditions employed for 3A5AF. The iridium complex IrH2Cl[(iPr2PC2H4)2NH] (4) has been reported as an exceptionally active catalyst in the presence of base for the transfer hydrogenation of acetophenone to phenylethanol.17 The reaction conditions were mild (room temperature, 2 h), and a conversion of acetophenone over 99% and a phenylethanol yield of 98% were obtained. The catalytic reduction of 5-HMF to bis(2,5-hydroxymethyl)furan (BHMF) has previously been performed using metal (Ru, Pt, and Zr) catalysts and H2 as the reducing agent.18−20 Transfer hydrogenation of 5-HMF has also been achieved with ethanol or formic acid as the hydrogen donor catalyzed by metal (Zr and Ni−Co) catalysts.21,22 Herein, reactions were performed using the
systematically, instead of the traditional method of studying only one factor at a time. The reaction in entry 12 was heated through microwave irradiation, and the result was compared with that of entry 9 (Table 1). In the 1H NMR spectrum of the reaction mixture from entry 12, only a trace amount of 3A5AF was present indicating that the conversion of 3A5AF reached almost 100%. However, the yield of 1 (34.7%) was lower than that from entry 9 (57.1%). Among the conventionally heated reactions, when the temperature was high (80 °C) and the reaction time was long (2 h), 1 could not be isolated despite 100% conversion of 3A5AF (entry 6 and 10). These results indicate that side reactions of 3A5AF or 1 happened at high temperatures and that the selectivity of 3A5AF hydrolysis toward 1 was affected by the reaction conditions. For example, the amine group in 1 could react with the carbonyl group in another molecule of 1 or 3A5AF to produce an imine compound. However, no identifiable signals from soluble side-products were observed in GC-MS and NMR analyses in the current study, so no definite conclusions could be made herein. We also note that no insoluble byproducts, i.e., polymeric materials, were seen during reaction workup. Further studies are needed in order to maximize the yields of 1 and facilitate catalyst reuse. Characterization of 1. The yield of 1 was the highest (57.1%) from the experiment in entry 9 (Table 1), and the portion of 1 in the solid obtained reached 92.6 wt % (determined by 1H NMR spectroscopy). Therefore, this sample was analyzed for the characterization of 1. In the 1H NMR spectrum of the sample (Figure S5), impurities were present at negligible levels, and there was only approximately 5.7 mol % unreacted 3A5AF. Therefore, this sample reached 94.3 mol % purity of 1. It is surprising to find that the amine group of 1 had two separate signals at 8.39 and 11.43 ppm. In order to verify this, 10 μL of deuterium oxide (D2O) was added to the NMR sample. After being shaken and equilibrated for 1 h, the sample was analyzed via 1H NMR (Figure S8). Both peaks at 8.39 and 11.43 ppm disappeared because of deuterium exchange, indicating that they are both exchangeable protons and potentially both from the amine group. It is assumed that the presence of two proton environments in this region is due to them having different amounts of intermolecular hydrogen bonding with another molecule of 1, which results in their different chemical shifts in the 1H NMR spectra. We have previously seen significant hydrogen-bonding in samples of 3A5AF;14 therefore, it is not surprising that 1 would also be affected by such phenomena. A further study should be carried out in the future to get more insight. The tertiary and quaternary carbon atoms (C3−C6) on the furan ring of 1 gave rise to very weak resonances in the 13C NMR spectrum (Figure S6). Computational calculations were performed for NMR prediction of 1 (see Supporting Information for more details). The predicted 13C NMR spectrum (Figure S7) is in acceptable agreement with the experimental one, and it confirms the assignment of C3−C6 peaks in the experimental spectrum. The GC-MS analysis of the sample (Figure S9) showed that besides a small amount of unreacted 3A5AF at 5.314 min (m/z 167.0) the main component was the amine product 1 at 4.348 min with m/z 125.0. In the HRMS spectrum, the peak of (1+H)+ (m/z 126.0550) had an overwhelming abundance. The molecular ion of 1 appeared at m/z 125.0478, which has only 0.84 ppm difference from the theoretical value (125.0477), thus supporting the identification of 1. 4919
DOI: 10.1021/acssuschemeng.7b00323 ACS Sustainable Chem. Eng. 2017, 5, 4916−4922
Research Article
ACS Sustainable Chemistry & Engineering combination of 4 and KOtBu in dry isopropyl alcohol (iPrOH) at room temperature, and aliquots were taken for monitoring via 1H NMR spectroscopy. After 2.5 h, the peak corresponding to the aldehyde moiety in 5-HMF had disappeared, so it was assumed that complete conversion was achieved. Filtration of the reaction solution through a plug of silica and removal of the solvent under vacuum allowed the isolation of BHMF in 95% yield. Similarly, using a 3A5AF/Ir ratio of 100:1, 2 could be obtained in 80% yield. Given the exceptional reactivity of 4 toward ketone reduction, it may be possible to use a lower catalyst loading. However, in our study, we wanted to perform the reaction at room temperature in a minimal amount of time, so relatively high catalyst loadings of 1 mol % were used. Compared with the yield of BHMF from 5-HMF (95%), the yield of 2 from 3A5AF is lower (80%). This may be due to a number of factors. For example, the catalyst might be inhibited by the amide functional group of 3A5AF. In addition, reduction of a ketone is more challenging than an aldehyde because of steric and electronic effects. Carbon Dioxide (CO2) Capture Investigation. Emissions of CO2 and its subsequent influence on the climate have become serious problems that can no longer be neglected. CO2 capture is an important solution to relieve the situations. As a solvent-based approach to CO2 capture, amine-scrubbing technology has been used for many years in the oil and chemical industries for removal of CO2 and hydrogen sulfide (H2S) from gas streams.23−25 Monoethanolamine (MEA) is a widely used amine for CO2 and H2S capture. Aqueous 30% (w/ v) MEA solution (pKa = 9.50)26 was optimized by Moser et al. as a benchmark solvent with a CO2 removal rate up to 90%.27 However, there are still concerns about MEA application in CO2 capture. Amine solvents are volatile, so during the process, part of the amine will evaporate to the atmosphere and produce some environmentally hazardous compounds.28 MEA is vulnerable to thermal and oxidative degradation in the presence of O2, CO2, SOx, and NOx.29 This will have negative influence on its CO2 capture capacity, and the compounds generated cause equipment corrosion, solvent contamination, viscosity increases, and both environmental and human health hazards.30,31 In addition, a great amount of energy is needed for the regeneration of the amine.32 Therefore, research about improved methods and alternative materials for CO2 capture is being constantly performed.33−35 As 3A5AF contains an amide group and 1 has an amine group in their structures, they were expected to capture CO2 via the reactions between the functional groups and CO2. In a typical experiment, 50 mg of dry solid was dissolved in 1 mL of dry methanol-d4 (CD3OD) (for 3A5AF) or dry dimethyl sulfoxide-d6 (for 1), and the solution was degassed under vacuum using a Schlenk line. The test was performed for both the dry solution and the solution with 30 μL of deionized water added. In each situation, the solution was stirred under CO2 flow at room temperature and at 40 °C, each for 1 h. After each test, the solution was analyzed by 1H and 13C NMR spectroscopy. Unfortunately, neither 3A5AF nor 1 showed the ability to capture CO2. The aqueous pKa values of 3A5AF and 1 were calculated computationally, using an approach that we have used previously,14 in order to provide a reasonable explanation for their negative performance. When acetanilide (pKa = 0.5) was used as the reference, the calculated pKa value of 3A5AF was −6.5. The pKa value of 1 was 2.62 with aniline (pKa = 4.87) as the reference. Aqueous carbon dioxide (CO2 dissolved in water) has a pKa of 6.36. Therefore, 3A5AF and 1
are stronger acids than CO2 (aqueous) in water, and this makes it a challenge for them to react with CO2. The strong acidity of 3A5AF and 1 is attributed to resonance effects and the delocalization of the lone pair on the N atom (Scheme 5). Scheme 5. Resonance Structures Leading to Delocalization of the Lone Pair on the N Atom of 3A5AF
One unexpected finding during the current studies of 3A5AF was the ability of the protons of the methyl group in the acetyl moiety to undergo deuterium exchange. In the 1H NMR spectrum of 3A5AF in CD3OD after it was mixed with CO2, the signal of δ(CH3) of the acetyl methyl group disappeared (Figure 2, see the full spectrum in Supporting Information). At first it was thought that 3A5AF had undergone hydrolysis catalyzed by the in situ formed carbonic acid. The solvent (CD3OD) in this NMR sample was removed using a Rotavap, and the solid obtained was dissolved in methanol, CH3OH, with the aim of performing H−D exchange on the mixture. When this solution was injected into the GC-MS system, it showed only one component, which was 3A5AF (m/z 167.0). The methanol was removed using a Rotavap and redissolved in chloroform-d for NMR analysis. In the 1H NMR spectrum, the expected signals for 3A5AF were all present. These results implied that the acetyl protons had undergone exchange with the deuterium in CD3OD, which is rare since usually the protons in a methyl group do not exchange. This can be explained by the fact that 3A5AF undergoes keto−enol tautomerism in solution (Scheme 6). All protons in the methyl group will be replaced by deuterium during this process, thus resulting in the disappearance of the signal in the 1H NMR spectrum. In addition, the carbonyl group is strongly electronwithdrawing, which activates the protons in its neighboring methyl group. As a consequence, the protons become acidic and more readily undergo deuterium exchange. In the corresponding 13C NMR spectrum of 3A5AF in CD3OD after deuterium exchange, the peak of the carbon atom of this methyl group is split into a multiplet caused by C−D coupling (Figure 3).
■
CONCLUSIONS An amine product, 2-acetyl-4-aminofuran (1), was successfully synthesized via the hydrolysis of 3A5AF catalyzed by NaOH. The highest yield of 1 obtained was 57.12% with 94.3 mol % purity when the reaction was performed at 80 °C for 1 h. The golden solid was characterized using GC-MS, HRMS, NMR, and IR spectroscopies. The reduction of 3A5AF was performed using NaBH 4 . A secondary alcohol, 3-acetamido-5-(1hydroxylethyl)furan (2), was obtained and an alkene (3) was observed from the subsequent dehydration of 2 during GC-MS analysis. The ability of 3A5AF and 1 in CO2 capture was investigated, but they are not effective absorbents due to their strong acidity. We hope these initial reactions can be a good starting point for the development of 3A5AF as a platform chemical to yield products containing naturally fixed nitrogen. Further studies on the applications of both 3A5AF and the obtained products are needed to make full use of these biomass-derived compounds. 4920
DOI: 10.1021/acssuschemeng.7b00323 ACS Sustainable Chem. Eng. 2017, 5, 4916−4922
Research Article
ACS Sustainable Chemistry & Engineering
Figure 2. Selected regions of the 1H NMR spectra of 3A5AF in CD3OD (a) initially (t = 0 min) before and (b) after deuterium exchange.
■
Scheme 6. Deuterium Exchange of the Methyl Group in 3A5AF with CD3ODa
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acssuschemeng.7b00323. Synthesis and identification of 3A5AF, spectroscopic and analytical data for 3A5AF, 1, and 2, 23 factorial design data, and 1H and 13C NMR spectra of deuterium exchange between 3A5AF and CD3OD (PDF)
■
AUTHOR INFORMATION
Corresponding Author a
*E-mail:
[email protected].
The deuterium exchange of the amide group is not shown for clarity.
ORCID
Francesca M. Kerton: 0000-0002-8165-473X
Figure 3. Selected regions of the 13C NMR spectra of 3A5AF in CD3OD (a) initially (t = 0 min) before and (b) after deuterium exchange. 4921
DOI: 10.1021/acssuschemeng.7b00323 ACS Sustainable Chem. Eng. 2017, 5, 4916−4922
Research Article
ACS Sustainable Chemistry & Engineering Notes
Product Role in the Shvo’s Catalyzed Selective Hydrogenation of the Platform Bio-Based Chemical 5-Hydroxymethylfurfural. Dalton Trans. 2014, 43 (26), 10224−10234. (19) Chatterjee, M.; Ishizaka, T.; Kawanami, H. Selective Hydrogenation of 5-Hydroxymethylfurfural to 2,5-Bis-(hydroxymethyl)furan Using Pt/MCM-41 in an Aqueous Medium: A Simple Approach. Green Chem. 2014, 16 (11), 4734−4739. (20) Han, J.; Kim, Y.-H.; Jang, H.-S.; Hwang, S.-Y.; Jegal, J.; Kim, J. W.; Lee, Y.-S. Heterogeneous Zirconia-Supported Ruthenium Catalyst for Highly Selective Hydrogenation of 5-Hydroxymethyl-2-furaldehyde to 2,5-Bis(hydroxymethyl)furans in Various n-Alcohol Solvents. RSC Adv. 2016, 6 (96), 93394−93397. (21) Hao, W.; Li, W.; Tang, X.; Zeng, X.; Sun, Y.; Liu, S.; Lin, L. Catalytic Transfer Hydrogenation of Biomass-Derived 5-Hydroxymethyl Furfural to the Building Block 2, 5-Bishydroxymethyl Furan. Green Chem. 2016, 18 (4), 1080−1088. (22) Yang, P.; Xia, Q.; Liu, X.; Wang, Y. Catalytic Transfer Hydrogenation/Hydrogenolysis of 5-Hydroxymethylfurfural to 2, 5Dimethylfuran over Ni-Co/C Catalyst. Fuel 2017, 187, 159−166. (23) CCP CO2 Capture Project. http://www.co2captureproject.org (accessed January 24, 2017). (24) MacDowell, N.; Florin, N.; Buchard, A.; Hallett, J.; Galindo, A.; Jackson, G.; Adjiman, C. S.; Williams, C. K.; Shah, N.; Fennell, P. An Overview of CO2 Capture Technologies. Energy Environ. Sci. 2010, 3 (11), 1645−1669. (25) Dutcher, B.; Fan, M.; Russell, A. G. Amine-Based CO2 Capture Technology Development from the Beginning of 2013-A Review. ACS Appl. Mater. Interfaces 2015, 7 (4), 2137−2148. (26) Hall, H., Jr Correlation of the Base Strengths of Amines. J. Am. Chem. Soc. 1957, 79 (20), 5441−5444. (27) Moser, P.; Schmidt, S.; Sieder, G.; Garcia, H.; Stoffregen, T.; Stamatov, V. The Post-Combustion Capture Pilot Plant Niederaussem−Results of the First Half of the Testing Programme. Energy Procedia 2011, 4, 1310−1316. (28) Nguyen, T.; Hilliard, M.; Rochelle, G. Volatility of Aqueous Amines in CO2 Capture. Energy Procedia 2011, 4, 1624−1630. (29) Mazari, S. A.; Ali, B. S.; Jan, B. M.; Saeed, I. M.; Nizamuddin, S. An Overview of Solvent Management and Emissions of Amine-Based CO2 Capture Technology. Int. J. Greenhouse Gas Control 2015, 34, 129−140. (30) Zoannou, K.-S.; Sapsford, D. J.; Griffiths, A. J. Thermal Degradation of Monoethanolamine and its Effect on CO2 Capture Capacity. Int. J. Greenhouse Gas Control 2013, 17, 423−430. (31) Vega, F.; Sanna, A.; Navarrete, B.; Maroto-Valer, M. M.; Cortés, V. J. Degradation of Amine-Based Solvents in CO2 Capture Process by Chemical Absorption. Greenhouse Gases: Sci. Technol. 2014, 4 (6), 707−733. (32) Wu, X.; Yu, Y.; Qin, Z.; Zhang, Z. The Advances of PostCombustion CO2 Capture with Chemical Solvents: Review and Guidelines. Energy Procedia 2014, 63, 1339−1346. (33) Chan, W. H.; Mazlee, M. N.; Ahmad, Z. A.; Ishak, M. A. M.; Shamsul, J. B. The Development of Low Cost Adsorbents from Clay and Waste Materials: A Review. J. Mater. Cycles Waste Manage. 2017, 19 (1), 1−14. (34) Mumford, K. A.; Wu, Y.; Smith, K. H.; Stevens, G. W. Review of Solvent Based Carbon-Dioxide Capture Technologies. Front. Chem. Sci. Eng. 2015, 9 (2), 125−141. (35) Bara, J. E.; Camper, D. E.; Gin, D. L.; Noble, R. D. RoomTemperature Ionic Liquids and Composite Materials: Platform Technologies for CO2 Capture. Acc. Chem. Res. 2010, 43 (1), 152− 159.
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS We thank Memorial University, Natural Sciences and Engineering Research Council of Canada (NSERC), Research & Development Corporation Newfoundland and Labrador. We thank K. Hawboldt for advice on CO2 capture and funding via a Mitacs internship cluster with Suncor Energy. Calculations were performed using the ACEnet consortium of Compute Canada. C.S. thanks DAAD for a RISE worldwide internship.
■
REFERENCES
(1) Fiorentino, G.; Ripa, M.; Ulgiati, S. Chemicals from Biomass: Technological versus Environmental Feasibility. A Review. Biofuels, Bioprod. Biorefin. 2017, 11 (1), 195−214. (2) Zhang, X.; Lei, H.; Chen, S.; Wu, J. Catalytic Co-pyrolysis of Lignocellulosic Biomass with Polymers: A Critical Review. Green Chem. 2016, 18, 4145−4169. (3) Upton, B. M.; Kasko, A. M. Strategies for the Conversion of Lignin to High-Value Polymeric Materials: Review and Perspective. Chem. Rev. 2016, 116 (4), 2275−2306. (4) Chatterjee, S.; Saito, T. Lignin-Derived Advanced Carbon Materials. ChemSusChem 2015, 8 (23), 3941−3958. (5) Chen, X.; Zhang, B.; Wang, Y.; Yan, N. Valorization of Renewable Carbon Resources for Chemicals. Chimia 2015, 69 (3), 120−124. (6) Yan, N.; Chen, X. Sustainability: Don’t Waste Seafood Waste. Nature 2015, 524, 155−157. (7) Chen, X.; Yang, H.; Yan, N. Shell Biorefinery: Dream or Reality? Chem. - Eur. J. 2016, 22, 13402−13421. (8) Franich, R. A.; Goodin, S. J.; Wilkins, A. L. Acetamidofurans, Acetamidopyrones, and Acetamidoacetaldehyde from Pyrolysis of Chitin and N-Acetylglucosamine. J. Anal. Appl. Pyrolysis 1984, 7 (1− 2), 91−100. (9) Chen, J.; Wang, M.; Ho, C.-T. Volatile Compounds Generated from Thermal Degradation of N-Acetylglucosamine. J. Agric. Food Chem. 1998, 46 (8), 3207−3209. (10) Omari, K. W.; Dodot, L.; Kerton, F. M. A Simple One-Pot Dehydration Process to Convert N-Acetyl-D-glucosamine into a Nitrogen-Containing Compound, 3-Acetamido-5-acetylfuran. ChemSusChem 2012, 5 (9), 1767−72. (11) Drover, M. W.; Omari, K. W.; Murphy, J. N.; Kerton, F. M. Formation of a Renewable Amide, 3-Acetamido-5-acetylfuran, via Direct Conversion of N-Acetyl-D-glucosamine. RSC Adv. 2012, 2 (11), 4642−4644. (12) Chen, X.; Chew, S. L.; Kerton, F. M.; Yan, N. Direct Conversion of Chitin into a N-Containing Furan Derivative. Green Chem. 2014, 16, 2204−2212. (13) Chen, X.; Liu, Y.; Kerton, F. M.; Yan, N. Conversion of Chitin and N-Acetyl-D-glucosamine into a N-Containing Furan Derivative in Ionic Liquids. RSC Adv. 2015, 5 (26), 20073−20080. (14) Liu, Y.; Rowley, C. N.; Kerton, F. M. Combined Experimental and Computational Studies on the Physical and Chemical Properties of the Renewable Amide, 3-Acetamido-5-acetylfuran. ChemPhysChem 2014, 15 (18), 4087−4094. (15) Boyer, A.; Lautens, M. Rhodium-Catalyzed Domino Enantioselective Synthesis of Bicyclo[2.2.2]lactones. Angew. Chem., Int. Ed. 2011, 50 (32), 7346−7349. (16) Kuhn, R.; Bister, W.; Dafeldecker, W. Amino Sugar Syntheses. XIV. D- and L-Quinovosamines. Justus Liebigs Ann. Chem. 1958, 617, 115−28. (17) Clarke, Z. E.; Maragh, P. T.; Dasgupta, T. P.; Gusev, D. G.; Lough, A. J.; Abdur-Rashid, K. A Family of Active Iridium Catalysts for Transfer Hydrogenation of Ketones. Organometallics 2006, 25 (17), 4113−4117. (18) Pasini, T.; Solinas, G.; Zanotti, V.; Albonetti, S.; Cavani, F.; Vaccari, A.; Mazzanti, A.; Ranieri, S.; Mazzoni, R. Substrate and 4922
DOI: 10.1021/acssuschemeng.7b00323 ACS Sustainable Chem. Eng. 2017, 5, 4916−4922
