Here we present a protocol for performing reactions in simple reaction vessels under low-to-moderate pressures of CO2. The reactions can be performed in a variety of vessels simply by administering the carbon dioxide in the form of dry ice, without the need for costly or elaborate equipment or set-ups.
Herein is presented a general strategy to perform reactions under mild to moderate CO2 pressures with dry ice. This technique obviates the need for specialized equipment to achieve modest pressures, and can even be used to achieve higher pressures in more specialized equipment and sturdier reaction vessels. At the end of the reaction, the vials can be easily depressurized by opening at room temperature. In the present example CO2 serves as both a putative directing group as well as a way to passivate amine substrates, thereby preventing oxidation during the organometallic reaction. In addition to being easily added, the directing group is also removed under vacuum, obviating the need for extensive purification to remove the directing group. This strategy allows the facile γ-C(sp3)-H arylation of aliphatic amines and has the potential to be applied to a variety of other amine-based reactions.
The use of gaseous compounds in chemical reactions typically requires specialized equipment and procedures1,2. At bench scale, some gases can be added directly from a tank using a high pressure regulator3. An alternative method is to condense the gas under cryogenic conditions4,5. Although useful, these strategies require the use of specialized pressure reactors with valves, which can be cost prohibitive for running numerous reactions in parallel. This can therefore greatly slow the rate at which reaction screening can proceed. As a result, chemists have found it desirable to introduce these compounds using alternative methods. Ammonia can be added to reactions using different ammonium carboxylate salts, taking advantage of the weak equilibrium between these salts and free ammonia6. Transfer hydrogenation is an important strategy for reduction reactions of olefins, carbonyl, and nitro groups that circumvents the use of flammable hydrogen gas with compounds such as ammonium formate or hydrazine as carriers of H27. Another gas of interest in this area is carbon monoxide8 – CO can be generated in situ by liberation from metal carbonyl complexes9,10, or alternatively it can be generated by decarbonylation from sources such as formates and formamides11,12,13 or chloroform14,15.
One gas which has not enjoyed significant development in this respect is carbon dioxide16. One reason for this is that many transformations that involve CO2 also require high temperatures and pressures, and thus are automatically relegated to specialized reactors17,18. Recent efforts to develop more reactive catalysts, however, have facilitated running many of these reactions under atmospheric pressures of CO219,20,21,22. We recently discovered a reaction in which carbon dioxide could be used to mediate the γ-C(sp3)–H arylation of aliphatic amines23. This strategy was expected to combine the benefits of a static directing group approach including amide24,25,26,27,28, sulfonamide29,30,31,32, thiocarbonyl33,34, or hydrazone35-based directing groups (chemical robusticity), with the ease of a transient directing group (decreased step economy)36,37,38,39.
Although the reaction could occur under atmospheric pressure of CO2, the need for a Schlenk set-up to screen reactions proved prohibitively slow. Furthermore, increasing the pressure slightly led to improved reaction yields, but could not be easily achieved using a Schlenk line. We therefore sought an alternative strategy, and subsequently identified that dry ice could be easily used as a solid source of CO2 that could be added to a variety of reaction vessels to introduce the necessary amount of carbon dioxide to achieve moderate pressures (Figure 1). Though underutilized in synthesis, a similar strategy is fairly common as a method to generate liquid CO2 for chromatography and extraction applications40,41,42,43,44. Utilizing this strategy allowed our group to rapidly screen large numbers of reactions in parallel, while the ability to access moderate CO2 pressures of between 2-20 atmospheres were critical to improving the yields of the reactions. Under these conditions, both primary (1°) and secondary (2°) amines can be arylated with electron rich and electron poor aryl halides.
CAUTION: 1) The following protocols have been deemed safe through repeated trials. However, caution should be exercised when sealing vials, throughout the reaction, and especially when opening the reactions, as inhomogeneity in the reaction vials may lead to equipment failure. Vials should be inspected for physical defects prior to use. Vials should be placed behind some form of blast shield or hood sash immediately after sealing to prevent incidents should the vials fail. 2) Although there is little chance for asphyxiation due to the small quantities of CO2 used, reactions should be set-up as well as opened in a well-ventilated area or in a fume hood. 3) Dry ice is a cryogen and can cause serious tissue damage. Care should therefore be exercised while manipulating it to avoid frostbite, such as limiting direct contact or using cryogenic gloves. 4) Dry ice will condense water vapor, meaning that prior to use, the dry ice should be mechanically exfoliated to ensure the mass is of CO2(s) only. This can be achieved by simply rubbing the dry ice between one’s fingers, or more safely, rubbing it between one’s fingers with a protective layer such as a glove or towel.
1. Reaction in a 7.5 mL Vial (Air Not Excluded)
2. Reaction in a 7.5 mL Vial (Purging Conditions – Air Excluded)
3. Reaction in a 40 mL Vial (Air Not Excluded)
4. Reaction in a 35 mL Pressure Tube (Air not excluded)
Following these protocols, it is possible to charge a reaction vial with an appropriate amount of carbon dioxide to achieve chemical reactions that require CO2 atmospheres. The pressure achieved in Step 1 is calculated to be approximately 3 atmospheres (see discussion for determination of this value), although due to partial solvation, the observed pressure is in the vicinity of 2 atmospheres at room temperature, and should be approximately 2.6 atmospheres under the reaction conditions. Therefore, under the conditions in Step 1, 2-Methyl-4-phenyl-butanamine can be obtained in 69% yield (Figure 2). By first purging the flask of air through displacement by sublimating CO2 (Step 2), the yield can be increased slightly to 72%. To distinguish between these results at ~2.6 atmospheres of pressure, performing the reaction under 1 atmosphere of CO2 using a standard Schlenk set-up furnishes the desired product in only 49% isolated yield. If no CO2 is used, or the vial is not properly sealed and thus a stable CO2 atmosphere is not maintained, then <5% yield of the desired product is detected by 1H NMR (using 1,1,2,2-tetrachloroethane as reference standard). Meanwhile, scaling the reaction up by a factor of 5 while simultaneously using a larger reaction vial (Step 3) can still give product, albeit in a slightly decreased yield of 42%. The reactions can also be performed in pressure reaction tubes (Figure 1), in this case allowing the synthesis of 2-Methyl-N-(3-methylbenzyl)-4-phenylbutan-2-amine in 40% yield (Figure 3).
Figure 1. Reaction Vessels Used in this Study. From left to right: 2 dram Vial, 10 Dram Vial, 35 mL Pressure Tube). Please click here to view a larger version of this figure.
Figure 2. 1H NMR of 2-Methyl-4-phenyl-butanamine. 400 MHz, CDCl3, 298 K. Please click here to view a larger version of this figure.
Figure 3. 1H NMR of 2-Methyl-N-(3-methylbenzyl)-4-phenylbutan-2-amine. 400 MHz, CDCl3, 298 K. Please click here to view a larger version of this figure.
Carbon Dioxide Loading in Empty Vials | ||||||||||||
100 mg | 125 mg | 150 mg | 175 mg | 200 mg | 225 mg | 250 mg | 275 mg | 300 mg | Legend | |||
25ºC | √ | √ | √ | √ | √ | √ | √ | √ | √ | √ = Stable Under Conditions | ||
60ºC | √ | √ | √ | √ | √ | √ | √ | √ | √ | x = Unstable Under Conditions | ||
70ºC | √ | √ | √ | √ | √ | √ | √ | √ | √ | |||
80ºC | √ | √ | √ | √ | √ | √ | √ | √ | √ | |||
90ºC | √ | √ | √ | √ | √ | √ | √ | √ | √ | |||
Temperature | 100ºC | √ | √ | √ | √ | √ | √ | √ | √ | √ | ||
110ºC | √ | √ | √ | √ | √ | √ | √ | √ | √ | |||
120ºC | √ | √ | √ | √ | √ | √ | √ | √ | √ | |||
130ºC | √ | √ | √ | √ | √ | √ | √ | √ | √ | |||
140ºC | √ | √ | √ | √ | √ | √ | √ | √ | √ | |||
150ºC | √ | √ | √ | √ | √ | √ | √ | √ | √ | |||
160ºC | √ | √ | √ | √ | √ | √ | √ | √ | √ |
Table 1. Relative Stability of 7.5 mL Vials Based on CO2 Loading and Temperature. Vials were loaded with the requisite amount of dry ice, followed by immediately sealing with a PTFE-lined cap. The vials were immediately placed into Pie-blocks behind a blast shield in a fumehood, followed by heating to 60 °C, followed by raising 10 °C every hour to a peak of 160 °C. The vials were then cooled, and opened carefully to confirm no loss of CO2 pressure had occurred.
Carbon Dioxide Loading in Empty Vials | |||||||||||||
300 mg | 325 mg | 350 mg | 375 mg | 400 mg | 425 mg | 450 mg | 475 mg | 500 mg | 525 mg | Legend | |||
25ºC | √ | √ | √ | √ | √ | √ | √ | √ | √ | x | √ = Stable Under Conditions | ||
60ºC | √ | √ | √ | √ | √ | √ | √ | √ | √ | x | x = Unstable Under Conditions | ||
70ºC | √ | √ | √ | √ | √ | √ | √ | √ | √ | x | |||
80ºC | √ | √ | √ | √ | √ | √ | √ | √ | √ | x | |||
90ºC | √ | √ | √ | √ | √ | √ | √ | √ | √ | x | |||
Temperature | 100ºC | √ | √ | √ | √ | √ | √ | √ | √ | √ | x | ||
110ºC | √ | √ | √ | √ | √ | √ | √ | √ | √ | x | |||
120ºC | √ | √ | √ | √ | √ | √ | √ | √ | √ | x | |||
130ºC | √ | √ | √ | √ | √ | √ | √ | √ | √ | x | |||
140ºC | √ | √ | √ | √ | √ | √ | √ | √ | √ | x | |||
150ºC | √ | √ | √ | √ | √ | √ | √ | √ | √ | x | |||
160ºC | √ | √ | √ | √ | √ | √ | √ | √ | √ | x |
Table 2. Relative Stability of 40 mL Vials Based on CO2 Loading and Temperature. Vials were loaded with the requisite amount of dry ice, followed by immediately sealing with a PTFE-lined cap. The vials were immediately placed into Pie-blocks behind a blast shield in a fumehood, followed by heating to 60 °C, followed by raising 10 °C every hour to a peak of 160 °C. The vials were then cooled, and opened carefully to confirm no loss of CO2 pressure had occurred.
Using the van der Waals Equation of State, the approximate pressure of these systems can be calculated45
Eq. 1:
Under the conditions in Protocol 1, we can assume 26.3 mg of CO2 gives n =5.98 x 10-4 mols
As a rough estimate, this suggests that in Protocol 1 the reactions were performed under approximately 2.8 atmospheres of CO2. Assuming negligible displacement of the native atmosphere in the vessel (as noted above, a patina of frozen acetic acid will slow the initial sublimation of the dry ice, facilitating better accuracy in the measurement of added dry ice), however, the total pressure would then be expected to be modeled better by Dalton's law:
Eq. 2:
This model does not take into account that some of the gases will be dissolved in the solvent. In that case, it was necessary to attach a pressure gauge to adequately assess the pressure. By attaching a septum to the vial and inserting a pressure gauge, it was possible to measure the pressure at room temperature. The observed pressures over multiple reactions were only 15 ± 3 psi above atmospheric pressure (≈ 1 ± 0.2 atmospheres), or around 2 atmospheres total. Although the Henry constant of CO2 in neat acetic acid was not readily available for comparison, it is known that addition of acetic acid to water improves the solubility of carbon dioxide46. The estimated pressure at room temperature could be calculated using the previous approach:
The expected pressure would therefore be a slightly lower 3.3 atmospheres at room temperature in the absence of gas dissolving in the solvent. The difference between the observed and calculated pressures imply that CO2 has relatively high solubility in the organic solvent. Assuming negligible difference in the amount of dissolved CO2 over the temperature range, an increase in temperature from 298 K to the reaction temperature of 383 K would increase the pressure within the 2 dram vial to ~2.6 atmospheres.
To adequately assess the practical operating conditions, 2 dram vials were set-up with varying amounts of CO2, followed by screening these at different temperatures. To ensure operator safety, the vials were only heated after being placed behind a blast shield to contain any vial failures. If the vials blew up, the conditions were considered too harsh for the vials. Through these experiments it was determined that CO2 loading of up to 200 mg was tolerated at 110 °C for the 2 dram (7.5 mL) vials through consistent trials. This corresponds to approximately 20.7 atmospheres of pressure based on the previous approach, not withstanding the amount of gas dissolved, which may decrease the total pressure by a few atmospheres. Beyond 200 mg loading, however, the reaction vials would generally explode before reaching the target temperature of 110 °C. Caution should be exercised when modifying the conditions, however: In one scenario a related reaction was attempted at 160 °C with only 150 mg of CO2, but the vial failed before it had reached the target temperature. The greatest danger for modifying the reaction conditions would be from increasing the loading of CO2, as this can cause the vials to fail before safe engineering controls, such as blast shields, can be implemented.
A potential limitation to this strategy is the lack of data about the stability of the vials under different conditions. Therefore, it was necessary to screen the vials for their ability to withstand different pressures under a range of different temperatures. This was initiated with the 7.5 mL vials (Table 1). Each vial was charged with a pre-determined quantity of dry ice, followed by immediate sealing with a PTFE-lined screw cap. These vials were observed to be tolerant of these conditions, and none failed at room temperature. The temperature was then raised for all of the vials, and no explosions occurred during the experiment. Upon cooling, each vial was opened to confirm they had maintained pressurization with CO2. This suggests that the vials can tolerate upwards of 26.5 atmospheres of pressure, which is in contrast to the reaction conditions in which ~20.7 atmospheres of pressure was the consistent limit. It is therefore encouraged that solvent identity and volume be carefully considered in deviation from the disclosed method.
A similar screen for maximum tolerance was performed using 40 mL reaction vials (Table 2). In this case, an upper limit for the CO2 loading of empty vials was determined to be 500 mg. Above this the vials quickly failed at room temperature. Surprisingly, the calculated pressure of the samples that began to fail at room temperature was approximately 7 atmospheres and above. This is in contrast to the vial containing 500 mg of CO2, which was stable at 160 °C, which would correspond to a calculated pressure of just under 10.5 atmospheres. These results were reproducible across different vials, but there is no clear explanation for this phenomenon at this time. Under the conditions described under Protocol 3, only approximately 300 mg loadings of CO2 were tolerated. However, this is actually in line with the previous experiments, as under the conditions the pressure, uncorrected for potential absorption of carbon dioxide by the solvent, would be approximately 10 atmospheres. The decreased stability of the larger vials to pressure is expected, and suggests that these procedures are better performed in vessels that have smaller diameters and thicker walls47.
In summary, this protocol for using dry ice as a solid CO2 source in readily available glassware is expected to open new directions in the field of synthetic chemistry. By generating low to moderate pressures inside of sealed vials or pressure tubes, carbon dioxide fixation processes such as carboxylation48,49,50, as well as CO2 reduction51,52,53, can be achieved without the use of expensive specialized equipment. This newly adopted strategy will facilitate advances in the area of valorization of CO2 by incorporation into useful chemical feedstocks such as cyclic carbonates, poly carbonates, and carbamates54. Furthermore, the strategy of introducing CO2 as a solid may also be beneficial where mixtures of gases are desired, such as CO2 and CO, or CO2 and H2, as this facilitates the addition of both reagents in a non-gaseous form. Although use of dry ice to introduce liquid CO2 has been used for extractions and chromatography40,41,42,43,44, this protocol for introducing CO2 as a solid may also be useful for in situ generation of CO2(L) for use as a reaction solvent55,56,57, Future work exploring other uses for this approach, especially the combination of CO2 with other gas-precursors, are currently underway in our group.
The authors have nothing to disclose.
The authors wish to acknowledge start-up funding from The University of Toledo, as well as funds from the American Chemical Society's Herman Frasch Foundation in partial support of this work. Mr. Thomas Kina is acknowledged for his assistance with developing a suitable pressure gauge for measuring the reaction pressures. Mr. Steve Modar is thanked for useful discussions.
7.5 mL Sample Vial with Screw Cap (Thermoset) | Qorpak | GLC-00984 | Can be reused. |
40 mL Sample Vial with Screw Cap (Thermoset) | Qorpak | GLC-01039 | Can be reused. |
Pressure Tube, #15 Thread, 7" Long, 25.4 mm O.D. | Ace Glass | 8648-06 | Can be reused. |
Pie-Block for 2 Dram Vials | ChemGlass | CG-1991-P14 | Can be reused. |
Pie-Block for 10 Dram Vials | ChemGlass | CG-1991-P12 | Can be reused. |
3.2 mm PTFE Disposable Stir Bars | Fisher | 14-513-93 | Can be reused. |
C-MAG HS 7 Control Hotplate | IKA | 20002695 | |
Analytical Weighing Balance | Sartorius | QUINTIX2241S | |
Double-Ended Micro-Tapered Spatula | Fisher Scientific | 21-401-10 | |
Hei-VAP Advantage – Hand Lift Model with G5 Dry Ice Condenser Rotary Evaporator | Heidolph | 561-01500-00 | |
Bump Trap 14/20 Joint | ChemGlass | CG-1322-01 | |
tert-Amyl amine | Alfa Aesar | B24639-14 | Used as received. |
2-Methyl-N-(3-methylbenzyl)butan-2-amine | N/A | N/A | Prepared from reductive amination of tert-amyl amine and 3-tolualdehyde in the presence of sodium borohydride in methanol. |
Palladium Acetate | Chem-Impex International, Inc. | 4898 | Used as received. |
Silver Trifluoroacetate | Oakwood Chemicals | 007271 | Used as received. |
Phenyl Iodide | Oakwood Chemicals | 003461 | Used as received. |
Acetic Acid | Fisher Chemical | A38 | Used as received. |
1,1,1,3,3,3-Hexafluoroisopropanol | Oakwood Chemicals | 003409 | Used as received. |
Deionized Water | Obtained from in-house deionized water system. | ||
Dry Ice | Carbonic Enterprises Dry Ice Inc. | Non-food grade dry ice. | |
Concentrated Hydrochloric Acid | Fisher Chemical | A144SI | Diluted to a 1.2 M solution prior to use. |
Diethyl Ether, Certified | Fisher Chemical | E138 | Used as received. |
Hexanes, Certified ACS | Fisher Chemical | H292 | Used as received. |
Saturated Ammonium Hydroxide | Fisher Chemical | A669 | Used as received. |
Dichloromethane | Fisher Chemical | D37 | Used as received. |
Sodium Sulfate, Anhydrous | Oakwood Chemicals | 044702 | Used as received. |
250 mL Separatory Funnel | Prepared in-house by staff glassblower. | ||
100 mL Round Bottom Flask | Prepared in-house by staff glassblower. | ||
Scientific Disposable Funnel | Caplugs | 2085136030 | |
Borosilicate Glass Scintillation Vials, 20 mL | Fisher Scientific | 03-337-15 | |
5 mm O.D. Thin Walled Precision NMR Tubes | Wilmad | 666000575 | |
Chloroform-d | Cambridge Isotope Laboratories, Inc. | DLM-7 | Used as received. |