-lactam Formation via a Reformatsky-like Reaction Under High Intensity Ultrasound
-lactam Formation via a Reformatsky-like Reaction Under High Intensity Ultrasound
Robert R. MacGregor
Texas Tech University
2001
Abstract
This research was proposed to synthesize various -lactams from imines and α-bromoesters under High Intensity Ultrasound and to test the extent of such reactions. Moderately high yields were achieved in all but the most hindered imines and the duration of the reactions was limited to 5 minutes; among the fastest -lactam formation reactions to date. Various imines were tested to identify any substituent effect in the reaction, which was discovered to be negligible except when affecting the nucleophicity of the nitrogen atoms. In addition, various hindered imines producing mixed results were tested to see the extent to which the reaction would be successful.
-lactam Formation via a Reformatsky-like Reaction Under High Intensity Ultrasound
Introduction
-lactams are one of the most successful groups of antibiotics including both penicillins and cephalosporins. Unfortunately, the research and commercial production of -lactams have been limited by both low yields and lengthy reaction duration. -lactams can be formed via a Reformatsky-like reaction, that is, a reaction in which an -bromoester reacts with zinc to form an organozinc intermediate before combining with an imine to form a closed -lactam and/or an open -aminoester (Scheme 1). Since the Reformatsky reaction was among the most successful reactions first studied under ultrasound, it was a natural step to attempt to produce -lactams under High Intensity Ultrasound (HIU) by this method.2 HIU produces cavitation bubbles that can reach temperatures and pressures as high as 5000 K and 500 atm respectively.5 This localized cavitation effect provides the activation energy necessary to form a closed -lactam instead of an open amino-ester. Therefore, by exposing the reaction to ultrasound a larger proportion of -lactam is formed in a shorter period than in traditional methods. In order to better understand the mechanism of reaction, differing substituent groups were added to one of the two phenyl groups of the parent imine, N-Benzylideneaniline (Figure 1). The substituents produced no effect when placed on the phenyl group attached to the double bonded carbon atom, but heavily effected the reaction when placed on the phenyl group directly attached to the nitrogen atom, suggesting a high sensitivity of the nitrogen atom to inductance effects. Other hindered imines were tested which either produced a majority of the -aminoester or failed to react entirely. To analyze the products of the reactions, proton nuclear magnetic resonance (1H nmr) spectra were recorded on a Bruker AF-300 instrument (300.133 Mhz). Chemical shifts are given in values of (ppm).
Scheme 1: A Reformatsky-like Reaction
Procedure
Formation of imines
Various imines had to be produced as they were not commercially available. The general procedure for the production of these imines is as follows. 100mmol of a ketone or aldehyde was weighed out and combined with 100 mmol of an amine in a round bottom flask. Forty grams of 5 Å activated molecular sieves were then added along with 40 mL of benzene.6 The mixture was stirred using a glass stir bar for 1 to 3 days or until IR spectroscopy showed desirable characteristics. The mixture was then filtered using a sintered glass funnel and washed with dichloromethane. A drying agent, MgSO4, was then added and filtered off. The resulting solution was placed on a rotary evaporator. The imines produced this way were N-p-chloroBenzylideneaniline (Figure 2), N-Benzylidene-p-chloroaniline (Figure 3), N-p-methoxyBenzylideneaniline (Figure 4), N-Benzylidene-p-methoxyaniline (Figure 5), N-p-trifluoromethylBenzylideneaniline (Figure 6), N-Benzylidene-p-trifluoromethyaniline (Figure 7), DimethylaminoBenzylideneaniline (Figure 8), N-Benzylidene-o-methoxyaniline (Figure 9), Benzophenone anil (Figure 10), and Acetophenone anil (Figure 11).
-lactam Formation under HIU
10 mmol of an imine was combined in a custom fitted flask with 18 mmol of Zinc dust and 2 mmol of Iodine, acting as catalysts. Nitrogen was slowly pumped into the flask while 15 mmol of the bromoester was added and 25 mL of dioxane was poured into the flask.2 The flask was screwed onto the ultrasound probe and placed in a thermally regulated 20.00 C water bath (the ratio of -lactam to -aminoester has been found to be temperature dependent).1 The solution was exposed to approximately 90 watts (electrical input) of 20kHz ultrasound for 1 minute continuously and then to 6 second pulses for 4 additional minutes. The contents of the flask were then dumped into an ice water bath and poured into a separatory funnel and dichloromethane was added. The funnel was shaken vigorously for 20 seconds and the dichloromethane solution was poured off into an Erlenmeyer flask. Two scoops of a drying agent, MgSO4, were added and the solution was filtered into a round bottom flask. The flask was placed on a rotary evaporator and the dichloromethane was evaporated. Approximately 100mL of ethyl acetate was added and the flask was again placed on a rotary evaporator to remove the dioxane through mass action. More dichloromethane was then added to evaporate the ethyl acetate. The product was then dissolved in a 1:1 ethyl acetate to hexanes solution and poured through a filtration column consisting of glass wool, sand, and approximately 4 inches of alumina. Thin Layer Chromatography was used to determine when all of the solution had been filtered. The ethyl acetate and hexanes solution was evaporated off using a rotary evaporator and the remaining product was first purified in vacuo and then recrystallized in hexanes to remove any impurities. If the product remained a liquid it was first distilled using a Kugelrohr distillation apparatus and then recrystallized.
Results and Conclusions
Before beginning the substituent tests, a suitable -bromoester needed to be found. This was accomplished by reacting the parent imine, N-Benzylideneaniline (Figure 1), with ethyl bromoacetate (Figure 27), ethyl-2-bromopropionate (Figure 27), and ethyl-2-bromoisobutyrate (Figure 27). Although one would expect ethyl-2-bromoisobutyrate to be hindered by steric effects, it produced the highest yield of -lactam of the three; 94%. Ethyl bromoacetate yielded a 1.0:3.7 ratio of amino-ester to -lactam and ethyl-2-bromopropionate yielded a 4.7:1.0 ratio, favoring the amino-ester. Therefore, ethyl-2-bromoisobutyrate was used as the -bromoester for all of the remaining reactions.
The core goal of the project was to determine the exact effect substituent groups would have on the ratio of -lactams to -aminoesters. The results shown in Table 1 draw a clear picture of that effect. Three different groups were added to each of the two phenyl rings for a series of 6 reactions. These groups; CF3, Cl, and OCH3 have the characteristics of being electron withdrawing, neutral, and electron donating respectively. They were used to alter the nucleophicity of the nitrogen atom in the imine in order to hinder or promote a closed ring -lactam. When the substituents were placed on the phenyl group attached to the carbon atom (R4) they had little effect as the nitrogen atom was shielded. A dimethylamine group was also tested which produced a 72% yield of its corresponding -lactam. All of the substituents on the carbon phenyl group yielded between 72% and 86% of the -lactam, demonstrating the small effect of the substituents. However, when the substituents were placed on the phenyl group attached to the nitrogen (R3), they greatly altered the reaction. The Cl group resulted in a mixture of -aminoester and -lactam (1.0:3.0), while the methoxy group produced a majority of the -lactam (79% isolated yield). Both are in accordance with the Reformatsky reaction mechanism, i.e., the methoxy group promotes -lactam formation because of its electron donating characteristic, while Cl favors neither to a large extent. The CF3 group not only produces a majority of the -aminoester as expected (82% isolated yield), it halts the formation of the -lactam nearly completely.
In addition to substituent effects, hindrance effects were tested with three different substituents: an ortho-methoxy group (Figure 8), an additional phenyl group (Figure 9), and an additional methyl group (Figure 11). The ortho-methoxy group produced a majority of amino-ester, which has previously been explained through an inductive effect of the electronegative oxygen atom, not through steric effects.1 This was shown by Bose, Gupta, and Manhas by reacting an ortho-ethyl group in place of the ortho-methoxy group which increased the yield of -lactam.1 Adding an additional phenyl group prevented any reaction from occurring as only starting materials were recovered using both ethyl bromoacetate and ethyl-2-bromoisobutyrate. The additional methyl group added to N-Benzylideneaniline produced the same results.
For an example of the advantages of ultrasound used in this reaction, Adrian, Barkin, and Hassib reported only a 45:55 ratio of -lactam to -aminoester when reacting N-Benzylidene-p-methoxyaniline with ethyl bromoacetate over a period of 3.5 hours using conventional means. Our methods produced an 86% isolated yield of -lactam in a similar reaction in just five minutes.1 Also, Gluchowski, Cooper, Bergbreiter, and Newcomb failed to produce a -lactam from Benzylidene-ethyl-amine and Ethyl-2-bromoisobutyrate via Lithium Ester Enolate-Imine Condensation, while our methods yielded a 1:8 ratio of -aminoester to -lactam using Benzylidene-methyl-amine.3 HIU was chosen in favor of Low Intensity Ultrasound (LIU) since LIU has been reported by Bose, Gupta, and Manhas to be unsuccessful in reacting N-Benzylideneaniline with ethyl-2-bromopropionate, while our methods achieved a complete reaction, producing a mixture of -lactam and -aminoester (4.7:1.0).2 In addition, the procedure described by Bose, Gupta, and Manhas proved unsuccessful when an attempting to verify their results using N-Benzylideneaniline and ethyl bromoacetate.
Table 1: Substituents and their effects
BromoesterImine
R1R2R3R4TimeBath TempYieldRatio*
(min)(C) (-amino : -lactam)
H / H / Ph / Ph / 60 / 20 / N/A / 1.0 : 2.7H / H / Ph / Ph / 5 / 20 / N/A / 1.0 : 1.3
H / H / Ph / Ph / 5 / 0 / N/A / Major : Trace
H / H / Ph / Ph / 5/10 / 20/0 / N/A / Major : Trace
H / H / Ph / Ph / 60 / 50 / N/A / 1.0 : 3.7
H / CH3 / Ph / Ph / 5 / 20 / N/A / 4.7 : 1.0
CH3 / CH3 / Ph / Ph / 5 / 20 / 94% / Trace : Major
CH3 / CH3 / p-ClPh / Ph / 5 / 20 / 72% / Trace : Major
CH3 / CH3 / p-MeOPh / Ph / 5 / 20 / 86% / Trace : Major
CH3 / CH3 / p-CF3Ph / Ph / 5 / 20 / 79% / Trace : Major
CH3 / CH3 / p-N(CH3)2 / Ph / 5 / 20 / 72% / Trace : Major
CH3 / CH3 / Ph / p-ClPh / 5 / 20 / 69% / 1.0 : 3.0
CH3 / CH3 / Ph / p-MeOPh / 5 / 20 / 79% / Trace : Major
CH3 / CH3 / Ph / p-CF3Ph / 5 / 20 / N/A / Major : Trace
CH3 / CH3 / Ph / o-MeOPh / 5 / 20 / N/A / Major : Trace
CH3 / CH3 / Ph / Ph, Ph / 5 / 20 / N/A / No reaction
H / H / Ph / Ph, Ph / 5 / 20 / N/A / No reaction
CH3 / CH3 / Ph / CH3, Ph / 5 / 20 / N/A / No reaction
CH3 / CH3 / Ph / CH3 / 5 / 20 / N/A / 1 : 8
*Ratios determined by 1H nmr of crude products
Works Cited
- Adrian, James C.; Barkin, Julia L.; and Hassib, Lamyaa. “-amino Esters Via the Reformatsky Reaction: Restraining Effects of the ortho-Methoxyphenyl Substituent,” Tetrahedron Letters, Vol. 40, pp. 2457-2460, 1999.
- Bose, Ajay K.; Gupta, Kavita; and Manhas, M.S. “-Lactam Formation by Ultrasound-promoted Reformatsky Type Reaction,” J. Chem. Soc., Chem. Commun., pp. 86-87, 1984.
- Gluchowski, Charles; Cooper, Lynn; Bergbreiter, David E.; Newcomb, Martin. “Preparation of -Lactams by the Condensation of Lithium Ester Enolates with Aryl Aldimines,” J. Org. Chem.1980, 45, 3413-3416.
- Love, Brian E. and Ren, Jianhua. “Synthesis of Sterically Hindered Imines,” J. Org. Chem. 1993, 58, 5556-5557.
- Luche, Jean-Lous. Synthetic Organic Sonochemistry. Plenum Press New York and London, 1998, pp. 1-30.
- Taguchi, Kazuo and Westheimer, F.H. “Catalysis by Molecular Sieves in the Preparation of Ketimines and Enamines.” J. Org. Chem., Vol. 36, No. 11, 1971.
Acknowledgments
I would like to take this opportunity to thank the Welch foundation and particularly Dr. Norman Hackerman for providing a truly rewarding experience to me and to my fellow Welch Scholars. I would also like to extend my gratitude to Dr. Richard Bartsch for allowing me to work in his laboratory and to Mr. Nathan Ross for showing me the ropes. This program wouldn’t have been a success without the hard work and dedication of both Dr. Steven Tomlinson, his wife Jocelyn, and Mrs. Cheryl Blasingame. I would also like to thank my newfound friends (in order of appearance) Gunnar “Guns” Ristroph, Jon Song, Haichen “the Wang” Johnnie, Nimran “Mr. PDA” Ali, Chirag “Smelly Feet” Bhatia, Steven “Unshakeable” Palmer, Wendy “Giggles” Lai, Joseph “Broken toe” Sanford, Sheetal “Mrs. PDA” Wadera, and Lindsey Hoover for making it all worthwhile. But above all I would like to thank my teachers (especially Mrs. Ann Welch, Mr. Rudy Pirovitz, and Mrs. Kelly Saenz) friends, and family who put me in a position to be a part of this program and who I will always be indebted to.
Appendix A: Imines used as reactants.
Figure 1: N-Benzylideneaniline. Used to test substituent effects and to choose the bromoester for the reactions. Available commercially.
Figure 2: N-p-chloroBenzylideneaniline. An imine with an added chlorine group. Produced on-site from the parent imine. 95.8% yield achieved through described procedure. 1H nmr (CDCl3): = 7.08-7.22 (t, 3H), 7.32-7.46 (m, 4H), 7.74-7.81 (d, 2H), 8.33 (s, 1H).
Figure 3: N-Benzylidene-p-chloroaniline. An imine with an added chlorine group. Produced on-site. 97% yield achieved through described procedure. 1H nmr (CDCl3): = 7.07-7.11 (m, 2H), 7.28-7.31 (m, 2H), 7.41-7.44 (m, 3H), 7.83-7.86 (d, 2H), 8.34 (s, 1H).
Figure 4: N-p-methoxyBenzylideneaniline. An imine with an added methoxy group. Produced on-site. 96% yield achieved through described procedure. 1H nmr (CDCl3): = 3.78 (s, 3H), 6.96-7.00 (d, 2H), 7.23-7.28 (m, 3H), 7.39-7.45 (m, 2H), 7.87-7.90 (d, 2H), 8.39 (s, 1H).
Figure 5: N-Benzylidene-p-methoxyaniline. An imine with an added methoxy group. Produced on-site. 100% yield achieved through described procedure. 1H nmr (CDCl3): = 3.75 (s, 3H), 6.87-6.91 (d, 2H), 7.18-7.22 (d, 2H), 7.38-7.42 (m, 3H), 7.84-7.87 (m, 2H), 8.42 (s, 1H).
Figure 6: N-p-trifluoromethylBenzylideneaniline. An imine with a CF3 group added. Produced on-site. 98% yield achieved through described procedure. 1H nmr (CDCl3): = 7.17-7.25 (m, 3H), 7.34-7.39 (m, 2H), 7.64-7.67 (d, 2H), 7.87-7.95 (d, 2H), 8.41 (s, 1H).
Figure 7: N-Benzylidene-p-trifluoromethyaniline. An imine with a CF3 group added. Produced on-site. 99% yield achieved through described procedure. 1H nmr (CDCl3): = 7.15-7.18 (d, 2H), 7.38-7.47 (m, 3H), 7.55-7.58 (d, 2H), 7.83-7.88 (d, 2H), 8.32 (s, 1H).
Figure 8: N-p-DimethylaminoBenzylideneaniline. An imine with a trimethyl group added. Produced on-site. 99.3% yield achieved through described procedure. 1H nmr (CDCl3): = 2.94 (s, 6H), 6.65-6.68 (d, 2H), 7.11-7.18 (m, 3H), 7.31-7.36 (m, 2H), 7.72-7.75 (d, 2H), 8.28 (s, 1H).
Figure 9: N-Benzylidene-o-methoxyaniline. An imine with an added ortho-methoxy group to test hindrance. Produced on-site. 91% yield achieved through described procedure. 1H nmr (CDCl3): = 3.79 (s, 3H), 6.90-6.99 (m, 3H), 7.15 (s, 3H), 7.88-7.92 (m, 2H), 8.43 (s, 1H).
Figure 10: Benzophenone anil. An imine with an added phenyl group to test hindrance. Produced on-site. 99.5% yield achieved through described procedure. 1H nmr (CDCl3): = 6.65-6.73 (m, 2H), 7.09-7.25 (m, 7H), 7.39-7.41 (m, 3H), 7.73-7.76 (m, 1.972).
Figure 11: Acetophenone anil. An imine with an added methyl group to test hindrance and to compare with benzophenone anil (Figure 10). Produced on-site. 98% yield achieved through described procedure. 1H nmr (CDCl3): = 2.20 (s, 3H), 6.67-6.80 (d, 1H), 7.31-7.44 (m, 6H), 7.95-7.99 (m, 2H).
Figure 12: Benzylidene-methyl-amine. Available commercially. Used to test the extent of the reaction.
Appendix B: The -lactams produced and their corresponding amino-esters (if applicable).
Figure 13: The -lactam (1,4-Diphenyl-azetidin-2-one) and amino-ester (3-Phenyl-3-phenylamino-propionic acid ethyl ester) formed from the parent imine Benzylideneaniline (Figure 1) and Ethyl bromoacetate (Figure 27).
Figure 14: The -lactam (3-Methyl-1,4-diphenyl-azetidin-2-one) and amino-ester (2-Methyl-3-phenyl-3-phenylamino-propionic acid ethyl ester) formed from the parent imine Benzylideneaniline (Figure 1) and Ethyl-2-bromoporpionate (Figure 27).
Figure 15: The -lactam (3,3-Dimethyl-1,4-diphenyl-azetidin-2-one) formed from the parent imine Benzylideneaniline (Figure 1) and Ethyl-2-bromoisobutyrate (Figure 27).
Figure 16: The -lactam (4-(4-Chloro-phenyl)-3,3-dimethyl-1-phenyl-azetidin-2-one) formed from N-p-chloroBenzylideneaniline (Figure 2) and Ethyl-2-bromoisobutyrate (Figure 27). 1H nmr (CDCl3): = 0.84 (s, 3H), 1.51 (s, 3H), 4.78 (s, 1H), 7.01-7.06 (m, 1H), 7.12-7.15 (m, 2H), 7.21-7.32 (m, 6H).
Figure 17: The -lactam (1-(4-Chloro-phenyl)-3,3-dimethyl-4-phenyl-azetidin-2-one) and amino-ester (3-(4-Chloro-phenylamino)-2,2-dimethyl-3-phenyl-propionic acid ethyl ester) formed from N-Benzylidene-p-chloroaniline (Figure 3) and Ethyl-2-bromoisobutyrate (Figure 27). 1H nmr (CDCl3): = 0.84 (s, 3H), 1.14-1.19 (t, 1H), 1.26 (s, 1H), 1.52 (s, 3H), 4.78 (s, 1H), 5.26 (s, 1H), 6.40-6.42 (d, 1H), 6.57-6.94 (m, 1H), 7.15-7.36 (m, 10H).
Figure 18: The -lactam (4-(4-Methoxy-phenyl)-3,3-dimethyl-1-phenyl-azetidin-2-one) formed from N-p-methoxyBenzylideneaniline (Figure 4) and Ethyl-2-bromoisobutyrate (Figure 27). 1H nmr (CDCl3): = .86 (s, 3H), 1.50 (s, 3H), 3.78 (s, 3H), 4.75 (s, 1H), 6.86-6.89 (m, 2H), 7.00-7.05 (m, 3H), 7.21-7.33 (m, 4H).
Figure 19: The -lactam (1-(4-Methoxy-phenyl)-3,3-dimethyl-4-phenyl-azetidin-2-one) formed from N-Benzylidene-p-methoxyaniline (Figure 5) and Ethyl-2-bromoisobutyrate (Figure 27). 1H nmr (CDCl3): = 0.84 (s, 3H), 1.51 (s, 3H), 3.73 (s, 3H), 4.77 (s, 1H), 6.78-6.82 (d, 2H), 7.17-7.37 (m, 8H).
Figure 20: The -lactam (3,3-Dimethyl-1-phenyl-4-(4-trifluoromethyl-phenyl)-azetidin-2-one) formed from N-p-trifluoromethylBenzylideneaniline (Figure 6) and Ethyl-2-bromoisobutyrate (Figure 27). 1H nmr (CDCl3): = 0.85 (s, 3H), 1.55 (s, 3H), 4.89 (s, 1H), 7.05-7.10 (m, 1H), 7.24-7.35 (m, 6H), 7.61-7.64 (d, 2H).
Figure 21: The amino-ester (2,2-Dimethyl-3-phenyl-3-(4-trifluoromethyl-phenylamino)-propionic acid ethyl ester) formed from N-Benzylidene-p-trifluoromethyaniline (figure 7) and ethyl-2-bromoisobutyrate (Figure 27). 1H nmr (CDCl3): = 1.14-1.19 (t/s overlap, 6H), 1.30 (s, 3H), 4.08-4.16 (m, 2H), 4.45-4.47 (d, 1H), 5.31-5.34 (d, 1H), 6.48-6.51 (d, 2H), 7.24-7.28 (m, 7H).
Figure 22: The amino-ester (3-(2-Methoxy-phenylamino)-2,2-dimethyl-3-phenyl-propionic acid ethyl ester) formed from N-Benzylidene-o-methoxyaniline (figure 9) and ethyl-2-bromoisobutyrate (Figure 27). 1H nmr (CDCl3): = 0.86-0.88 (s, 1H), 1.15-1.20 (t/s overlap, 5H), 1.25 (s, 3H), 3.84 (s, 3H), 4.07-4.16 (m, 2H), 4.53-4.56 (d, 2H), 5.35-5.38 (d, 2H), 6.29-6.33 (d, 1H), 6.54-6.72 (m, 3H), 7.18-7.30 (m, 6H).
Figure 23: The -lactam (4-(3-Dimethylamino-phenyl)-3,3-dimethyl-1-phenyl-azetidin-2-one) formed from N-p-DimethylaminoBenzylideneaniline (figure 10) and ethyl-2-bromoisobutyrate (Figure 27). 1H nmr (CDCl3): = 0.86 (s, 3H), 1.47 (s, 3H), 2.92 (s, 6H), 4.70 (s, 1H), 6.65-6.68 (d, 2H), 6.97-7.06 (m, 3H), 7.18-7.23 (t, 2H), 7.32-7.35 (d, 2H).
Figure 24: The -lactam (1-Methyl-4-phenyl-azetidin-2-one) formed from Benzylidene-methyl-amine (Figure 11) and ethyl-2-bromoisobutyrate (Figure 27). 1H nmr (CDCl3): = .76 (s, 3H), 1.10-1.16 (d, 1H), 1.24-2.31 (m, 3H), 1.43 (s, 3H), 2.86 (s, 3H), 4.32 (s, 1H), 7.16-7.19 (m, 2H), 7.32-7.42 (m, 4H).
Appendix C: Bromoesters used.
Figure 25: Ethyl bromoacetate.
Figure 26: Ethyl-2-bromopropionate.
Figure 27: Ethyl-2-bromoisobutyrate.
35-1
WSSP 2001