Chem. Mater. 2007, 19, 4761-4768
4761
Reversible, Room-Temperature, Chiral Ionic Liquids. Amidinium Carbamates Derived from Amidines and Amino-Acid Esters with Carbon Dioxide† Taisuke Yamada, Paul Joseph Lukac, Tao Yu, and Richard G. Weiss* Department of Chemistry, Georgetown UniVersity, Washington, DC 20057-1227 ReceiVed May 18, 2007. ReVised Manuscript ReceiVed June 28, 2007
The properties of a new class of chiral, room-temperature, ionic liquids (RTILs) are described. They are made from easily synthesized, readily available materials and can be transformed reversibly to their nonionic liquid states. The nonionic liquids consist of neat equimolar mixtures of a N′-alkyl-N,Ndimethylacetamidine (L) and an alkyl ester of a naturally occurring amino acid (n). When exposed to 1 atm of CO2 gas, the L/n solutions become cationic-anionic pairs, amidinium carbamates. Of the 50 L/n combinations examined, all except those involving the methyl ester of tyrosine (which was immiscible with the amidines) form RTIL states under CO2 atmospheres, and several remain liquids to at least -18 °C. Heating the ionic liquids in air at ca. 50 °C or bubbling N2 gas through them at ambient temperatures for protracted periods displaces the CO2 and re-establishes the nonionic L/n states. As an example of the changes effected by cycling between the two liquid states, a spectroscopic probe, 1-(pdimethylaminophenyl)-2-nitroethylene, senses a polarity like that of toluene before a mixture of N′octyl-N,N-dimethylacetamidine/isoleucine methyl ester is exposed to CO2 and a polarity like that of N,Ndimethylformamide afterward; whereas a low-polarity solvent, decane, is solublized readily by the nonionic L/n mixture, it is immiscible with the RTIL. Thermal and spectroscopic properties of both the nonionic and ionic phases are reported and compared. Several possible applications for these RTILs can be envisioned because, unlike many other ionic liquids, these need not be prepared and handled under scrupulously dry conditions and they can be cycled repeatedly between high- and low-polarity states.
Introduction Ionic liquids (ILs)1-3 have become increasingly popular as solvents in academic and industrial applications during the past decade because of their environmentally friendly characteristics and ability to support a large variety of reactions by solutes.4 However, many ionic liquids are not inert under the reaction conditions of their solutes,5 and the effort to find nonreactive ILs is a topic of great interest. Room-temperature ionic liquids (RTILs) are especially interesting in this regard.3,6 Chiral ILs7 add opportunities as media for enantioselective reactions of solutes,8 stereoselec* To whom correspondence should be addressed. E-mail: weissr@ georgetown.edu. Phone: 202-687-6013. Fax: 202-687-6209. † This article is dedicated to the memory of Prof. Dmitry Rudkevich, a recently departed friend, colleague, and creative scientist. He will be missed.
(1) Welton, T. Chem. ReV. 1999, 99, 2071-2083. (2) (a) Ionic Liquids: Industrial Applications to Green Chemistry; Rogers, R. D., Seddon, K. R., Eds.; ACS Symposium Series 818; American Chemical Society: Washington, DC, 2002. (b) Ionic Liquids IIIA and IIIB: Fundamentals, Progress, Challenges, and Opportunitities; Rogers, R. D., Seddon, K. R., Eds.; ACS Symposium Series 901 and 902; American Chemical Society: Washington, DC, 2005. (3) (a) Zhao, H. Chem. Eng. Commun. 2006, 193, 1660-1677. (b) Hagiwara, R.; Lee, J. S. Electrochemistry 2007, 75, 23-34. (4) (a) Dupont, J.; de Souza, R. F.; Suarez, P. A. Z. Chem. ReV. 2002, 102, 3667-3692. (b) Welton, T. Coord. Chem. ReV. 2004, 248, 24592477. (c) Parvulescu, V. I.; Hardacre, C. Chem. ReV. 2007, 107, 26152665. (d) Binnemans, K. Chem. ReV. 2007, 107, 2592-2614. (5) Chowdhury, S.; Mohan, R. S.; Scott, J. L. Tetrahedron 2007, 63, 2363-2389. (6) (a) Chiappe, C.; Pieraccini, D. J. Phys. Org. Chem. 2005, 18, 275297. (b) Flannigan, D. J.; Hopkins, S. D.; Suslick, K. S. J. Organomet. Chem. 2005, 690, 3513-3517. (c) Handy, S. T. Curr. Org. Chem. 2005, 9, 959-988.
tive polymerizations,9 and enantiomeric separations10 and purifications.11 Among these are the ILs derived from the anions of natural amino acids and tetraalkylphosphonium cations12 (some of which are able to fix reversibly CO2 while remaining an IL12c) or the cations of amino acids and esters with various anions.13 Another desirable attribute would be the ability to cycle easily and reversibly between an RTIL and a corresponding nonionic form because such materials allow the easy separation of solvent mixtures, catalysts, etc. This goal has been achieved by alternating the bubbling of (7) (a) Baudequin, C.; Bre´geon, D.; Levillain, J. L.; Guillen, F.; Plaquevent, J.-C.; Gaumont, A.-C. Tetrahedron: Asymmetry 2005, 16, 3921-3945. (b) Ding, J.; Armstrong, D. W. Chirality 2005, 17, 281-292. (c) Branco, L. C.; Gois, P. M. P.; Lourenc¸ o, N. M. T.; Kurteva, V. B.; Afonso, C. A. M. Chem. Commun. 2006, 2371-2372. (d) Liu, W.S.; Tao, G.-H.; He, L.; Yuan, K. Chin. J. Org. Chem. 2006, 26, 10311038. (8) (a) Pegot, B.; Vo-Thanh, G.; Gori, D.; Loupy, A. Tetrahedron Lett. 2004, 45, 6425-6428. (b) Doherty, S.; Goodrich, P.; Hardacre, C.; Luo, H.-K.; Roony, D. W.; Seddon, K. R.; Styring, P. Green Chem. 2004, 6, 63-67. (c) Luo, S.; Mi, X.; Zhang, L.; Liu, S.; Hu, H.; Cheng, J.-P. Angew. Chem., Int. Ed. 2006, 45, 3093-3097. (9) Ma, H.-Y.; Wan, X.-H; Chen, X.-F.; Zhou, Q.-F. Chin. J. Polym. Sci. 2003, 21, 265-270. (10) Ding, J.; Welton, T.; Armstrong, D. W. Anal. Chem. 2004, 76, 78197822. (11) Tran, C. D.; Oliveira, D. Anal. Biochem. 2006, 356, 51-58. (12) (a) Fukumoto, K.; Yoshizawa, M.; Ohno, H. J. Am. Chem. Soc. 2005, 127, 2398-2399. (b) Kagimoto, J.; Fukumoto, K.; Ohno, H. Chem. Commun. 2006, 2254-2256. (c) Fukumoto, K.; Kouno, Y.; Ohno, H. Chem. Lett. 2006, 35, 1252-1253. (d) Zhang, J.; Zhang, S.; Dong, K.; Zhang, Y.; Shen, Y.; Lv, X. Chem.sEur. J. 2006, 12, 40214026. (13) Tao, G.-H.; He, L.; Sun, N.; Kou, Y. Chem. Commun. 2005, 35623564.
10.1021/cm0713531 CCC: $37.00 © 2007 American Chemical Society Published on Web 08/17/2007
4762 Chem. Mater., Vol. 19, No. 19, 2007 Scheme 1
CO2 and a displacing gas (e.g., argon or nitrogen), respectively, through mixtures of an amidine, 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU), and 1-hexanol14 or an amidine and an alkyl amine.15 Here, we describe a diverse class of RTILs which are reVersible, chiral, and easily formed. They are made by exposing a 1:1 mixture of an easily synthesized amidine (N′alkyl-N,N-dimethylacetamidine; L) and an alkyl ester of a naturally occurring amino acid (n) to carbon dioxide gas, and they can be returned to their original, nonionic state when nitrogen gas is passed through the RTILs for a protracted period or they are heated to above ca. 50 °C in air (Scheme 1); they remain RTILs under a CO2 atmosphere. All of the 50 amidine/amino acid ester combinations (L/n) examined here, except those with the methyl ester of tyrosine (which did not dissolve in the amidines), form reversible, chiral RTILs (L-n-C) directly upon exposure to CO2; the immiscible mixtures require an indirect methodology to convert them to RTILs. In addition, many of these RTILs persist as liquids to far below 0 °C. We have examined esters of 6 amino acids (Pro, Leu, Ile, Val, Phe, Tyr) here; many more (including the 20 “standard” ones commonly found in most proteins) are available for study, and each offers potentially different properties and applications. In addition, this research adds to the growing list of uses for a current “nemesis’’, carbon dioxide.16 Experimental Section Materials. Unless stated otherwise, all reagents were used as received. L-Proline (Pro; 99%), l-leucine (Leu; 99%), L-isoleucine (Ile; 99%), L-phenylalanine (Phe; 98.5%), L-tyrosine (Tyr; 99%), and thionyl chloride (99.5%) were from Acros; triethylamine (99.9%) was from Alfa Aesor; 1-octanol (99%), 1-octadecanol (95%), L-valine (Val; 98%), and p-toluenesulfonic acid monohydride (98.5%) were from Aldrich. Methanol (Aldrich, 99.8%) was dried by Vogel’s method.17 Toluene (Aldrich, 99.9%) was dried by refluxing over sodium metal for 5 h, followed by distillation. Carbon dioxide gas, generated by warming dry ice, was dried by passing it through a tube filled with indicating Drierite. (14) (a) Jessop, P. G.; Heldebrant, D. J.; Li, X.; Eckert, C. A.; Liotta, C. L. Nature 2005, 436, 1102. (b) Unfortunately, the uptake of CO2 in DBU/alcohol systems is not quantitative at 1 atm of pressure, at least in toluene solutions. Hori, Y; Nagano, Y; Nakao, J; Taniguchi, H. Chem. Express 1986, 1, 173-176. (c) Hori, Y.; Nagano, Y.; Nakao, J.; Fukuhara, T.; Taniguchi, T. Chem. Express 1986, 1, 224-227. (15) Yamada, T.; Lukac, P. J.; George, M.; Weiss, R. G. Chem. Mater. 2007, 19, 967-969. (16) (a) Ritter, S. A. Chem. Eng. News 2007, 85, 11-17. (b) Rayner, C. M. Org. Process Res. DeV. 2007, 11, 121-132. (17) Vogel, A. I. Textbook of Practical Organic Chemistry, 5th ed.; Longman Scientific & Technical: New York, 1989; pp 401-402.
Yamada et al. Instrumentation. IR spectra were obtained on a Perkin-Elmer Spectrum One FTIR spectrometer interfaced to a PC, using an attenuated total reflection accessory or NaCl plates. 1H (referenced to internal TMS) and 13C (referenced to chloroform-d at 77.2 ppm) NMR spectra were recorded on a Varian 300 MHz spectrometer interfaced to a Sparc UNIX computer using Mercury software. Gas chromatographic (GC) analyses were performed on a HewlettPackard 5890A gas chromatograph equipped with flame ionization detectors and a DB-5 (15 m × 0.25 mm) column (J & W Scientific, Inc.). UV-vis spectra were recorded on a Varian CARY 300 Bio UV-vis spectrophotometer in Hellma quartz cells with 0.1 or 0.2 mm path lengths. Thermal gravimetric analysis (TGA) measurements were performed on a TGA 2050 thermogravimetric analyzer (TA Instruments) interfaced to a computer. Heating rates were 5 °C/min. Crystallization and melting temperatures of RTILs were measured by differential scanning calorimetry (DSC) using a TA 2910 differential scanning calorimeter interfaced to a TA Thermal Analyst 3100 controller equipped with a hollowed aluminum cooling block onto which dry ice was placed for subambient measurements. Polarizing optical micrographs and solidification temperatures of ionic liquids, sandwiched between thin cover slides, were recorded on a Leitz 585 SM-LUX-POL microscope equipped with crossed polars, a Leitz 350 heating stage, a Photometrics CCD camera interfaced to a computer, and an Omega HH503 microprocessor thermometer connected to a J-K-T thermocouple. The samples were cooled by passing cold nitrogen gas (prepared by flowing it through a coil of copper tubing immersed in liquid nitrogen) through the Leitz 350 heating stage. GC-MS measurements were obtained on a Shimadzu GC-17A gas chromatograph connected to a Shimadzu QP-5000 mass spectrometer instrument using a 0.25 µm SGE BPX5 (15 m × 0.25 mm) column and a flame ionization detector. Optical rotations were recorded on a Rudolph Instruments DigiPal 781 automatic polarimeter at 589 nm in Hellma quartz cells with 0.1 or 0.2 mm path lengths (l). Observed rotations, R ((0.0005°), were averaged from five determinations each along three different axes of the cylindrical cells and are corrected for residual rotations by the empty cells. Specific rotations of neat amino acid esters or amino acid ester/amidine mixtures, [Rtotal], are from eq 1 and those of the optically active component in the mixtures, [Ropt], are from eq 2. Densities, F (g‚cm-3), were estimated from the slopes of graphs of sample weights versus volumes at ca. 25 °C (see the Supporting Information for details) and were used to calculate c (g‚dL-1). [R total] ) R/(lF)
(1)
[Ropt] ) 100R/(lc)
(2)
Syntheses. The amidines and 1-(p-dimethylaminophenyl)-2-nitroethylene (DAPNE) were available from previous investigations.15 Their physical characteristics and purities are collected in the Supporting Information. The general procedure for the synthesis of the amino acid methyl esters is presented in detail for LeuC1; minor variations were adopted for the other amino acid esters. L-Leucine Methyl Ester (LeuC1).18 Under a dry atmosphere, thionyl chloride (15.0 g, 126 mmol) was added dropwise, with stirring, over a period of 10 min, to 50 mL of absolute methanol which was cooled by an ice-salt bath (ca. -10 °C). The solution was stirred for another 10 min, 10.0 g (76 mmol) of L-leucine was (18) (a) Webb, R. G.; Haskell, M. W.; Stammer, C. H. J. Org. Chem. 1969, 34, 576-580. (b) Smith, C. S.; Brown, A. E. J. Am. Chem. Soc. 1941, 63, 2605-2606.
ReVersible, Room-Temperature, Chiral Ionic Liquids added, and the solution was slowly heated to reflux and left there for 5 h. Excess thionyl chloride was removed by distillation and then subsequent distillations in which 10 mL aliquots of dichloromethane were added. The residue was placed under house vacuum at 70-75 °C for 2 h. (A trap of crushed sodium hydroxide/Drierite was placed between the house vacuum and the drying apparatus to remove any acid vapors.) The yellowish solid was dissolved in 20 mL of water and extracted with 25 mL of diethyl ether. The aqueous phase was placed under 50 mL of diethyl ether, and 30% aqueous sodium hydroxide was added until pH 8 was reached. The layers were separated, and most of the organic liquid was removed on a rotary evaporator. The remaining material was distilled to yield 7.44 g (41%) of L-leucine methyl ester (bp 32-33 °C/0.35 Torr) as a clear liquid of 99% purity (GC). IR: 3381, 3313, 2956, 2871, 1736, 1679, 1619, 1468, 1437 cm-1. 1H NMR: δ 3.71 (s, 3H, C(dO)OCH3), 3.48 (t, 1H, JHH 6.3 Hz, C*H), 1.77-1.36 (m, 5H, NH2,CH2CH(CH)3), 0.93 (m, 6H, (CH3)2) (lit.19 3.72 (3H), 3.48 (1H), 1.78 (1H), 1.56 (1H), 1.43 (1H), 0.94 (3H), 0.92 (3H)). 13C NMR: δ 177.24, 52.92, 51.95, 44.23, 24.84, 23.01, 21.90. L-Isoleucine Methyl Ester (IleC1).18 Yield: 37% of a clear liquid of 99% purity (GC), bp 34-35 °C/0.5 Torr. IR: 3386, 3313, 2963, 2935, 2877, 1735, 1619, 1459, 1436 cm-1. 1H NMR: δ 3.72 (s, 3H, C(dO)OCH3), 3.35 (d, 1H, JHH 5.1 Hz, C*H), 1.74 (m, 1H, C*H-CH), 1.45 (m, 2H, NH2), 0.88-1.2 (m, 8H, CH(CH3)CH2CH3). 13C NMR: δ 176.13, 51.63, 39.20, 39.20, 24.67, 15.73, 11.63. L-Proline Methyl Ester (ProC1).20 Yield: 36% of a colorless clear liquid, bp 30 °C/0.15 Torr. Purity: 99% (GC). IR: 3351, 2955, 2876, 1733, 1436 cm-1. 1H NMR: δ 3.73 (s, 3H, C(dO)OCH3), 3.77-3.75 and 3.57-3.52 (m, 1H, C*H-), 2.87-2.95 and 3.043.12 (m, 2H, -CH2-NH-), 2.61 (br, 1H, NH), 1.34-2.36 (m, 4H, CH2CH2-C*H) (lit.19 3.77 (3H), 3.74 (3H), 3.03 (1H), 2.91 (1H), 2.13 (1H), 1.85 (1H), 1.76 (2H)). 13C NMR: δ 175.76; 59.50, 51.93, 46.89, 45.07, 30.10, 25.40. L-Valine Methyl Ester (ValC1).21 Yield: 25% of a clear liquid of 99% purity (GC), bp 30 min. Then, a flask with a weighed amount of neat amidine/amine ester mixture was attached to the burette. The liquid was stirred with a Teflon-coated spin bar while the volume of CO2 adsorbed was recorded as a function of time until no volume change was discernible. This procedure for each mixture was repeated 3 times using different aliquots of sample. The percentage of the theoretical amount of CO2 taken up (%CO2) was calculated as follows. The volume of CO2 taken up by the amine/amidine solution (V) was calculated from eq 3, where Vblank is the volume decrease in the burette when the flask contained only CO2 and Vobs is the total volume decrease measured by the burette; the CO2 pressure within the burette was kept at 1 atm by raising or lowering the mercury reservoir throughout each experiment. V ) (Vobs - Vblank)
(3)
The theoretical volume of CO2 (VP,T) taken up by M moles of amidine/amine at a known temperature (T, K) and pressure (P, Torr), assuming complete conversion to amidinium carbamate, was calculated using eq 4. The data are included in Table S1 in the Supporting Information. VP,T ) [22.4(760/P)(T/273.15)]M
(4)
%CO2 ) [V/VP,T]100
(5)
Then
Results and Discussion The various L/n combinations examined here are shown in Table 1. All except those with the methyl ester of tyrosine (which was immiscible with the amidines even when the mixtures were heated to 80 °C) yielded ionic liquids at room temperature after exposure to CO2 gas despite the fact that some of the original mixtures are not liquids, and all but those containing the amino acids with octadecyl esters remain liquids to -18 °C. By contrast, all but one of the amino acid esters investigated solidify when they are exposed as neat liquids to CO2. This solidification behavior has been observed during the transformation of many simple achiral amines to ammonium carbamates upon the addition of CO2, as well (eq 6).15,24 The maximum temperatures at which the (23) House, H. O. Modern Synthetic Reactions, 2nd ed.; W. A. Benjamin: Menlo Park, CA, 1972; p 2. (24) (a) George, M.; Weiss, R. G. J. Am. Chem. Soc. 2001, 123, 1239310394. (b) George, M.; Weiss, R. G. Langmuir 2002, 18, 7124-7135. (c) George, M.; Weiss, R. G. Langmuir 2003, 19, 1017-1025.
L-n-C combinations persist without losing CO2 is ca. 50 °C according to TGA (vide infra). Solutions of 1:1 or 2:1 C8/TyrC1 in ethyl acetate became turbid initially and cleared after ca. 5 min of CO2 bubbling. After the ethyl acetate was removed by continued bubbling of CO2, the residues became waxy. The 2:1 mixtures were examined because TyrC1 can potentially form both a carbamate and a carbonate, adding two molecules of CO2. CO2
2R-NH2 {\ } R-NH-CO2- +H3N-R N
(6)
2
The diversity of the RTILs that can be formed with the amidines is demonstrated by the fact that the phenylsubstituted amino acid methyl ester, PheC1, also yields liquids, although they appeared qualitatively to be more viscous than those with the alkyl-substituted amino acid methyl esters. Clearly, there are many more possibilities for other L-n-C combinations just from the list of naturally occurring amino acids. Although the list is too extensive to be studied exhaustively here, each is capable of imparting a different set of characteristics to its L-n-C and of affecting the stereochemical consequences of solute reactions, for instance. The results in Table 1 with IleC8, LeuC8, IleC18, and LeuC18 illustrate the ability of RTILs to be formed with very long alkyl ester chains, as well. However, none of the octyl or octadecyl esters exhibited a birefringent pattern when observed under a polarizing microscope before or after shearing, and therefore, none is liquid crystalline.25 Taken in toto, these results indicate that an exceedingly wide variety of RTILs can be made easily from much more diverse L/n mixtures than the structures in Scheme 1. Furthermore, although the reactions to form the L-n-C combinations and the L/n mixtures from them were performed under dry conditions, the systems appear to be able to tolerate a significant amount of water: the addition of up to 10 wt % of H2O to a C6/LeuC1 mixture did not impede the formation of its RTIL upon exposure to CO2 (yielding C6-LeuC1-C). The samples remained clear liquids after the addition of water and their subsequent exposure to CO2. The (25) (a) Binnemans, K. Chem. ReV. 2005, 105, 4148-4204. (b) Bowlas, C. J.; Bruce, D. W.; Seddon, K. R. Chem. Commun. 1996, 16251626. (c) Lee, C. K.; Chen, J. C. C.; Lee, K. M.; Liu, C. W.; Lin, I. J. B. Chem. Mater. 1999, 11, 1237-1242. (d) Abdallah, D. J.; Robertson, A.; Hsu, H.-F.; Weiss, R. G. J. Am. Chem. Soc. 2000, 122, 3053-3062. (e) Chen, H.; Kwait, D. C.; Go¨nen, Z. S.; Weslowski, B. T.; Abdallah, D. J.; Weiss, R. G. Chem. Mater. 2002, 14, 4063-4072. (f) Kanazawa, A.; Ikeda, T.; Abe, J. Angew. Chem., Int. Ed. 2000, 39, 612-615. (g) Buruiana, E. C.; Buruiana, T.; Melinte, V.; Negulescu, I. ReV. Roum. Chim. 2005, 50, 465-470.
ReVersible, Room-Temperature, Chiral Ionic Liquids
Figure 1. Percentage uptake of CO2 (%CO2) by C8/IleC1 (b), C8/LeuC1 (2), C8/IleC8 (O), and C8/ProC1 (4) as a function of time upon exposure to 1 atm pressure of CO2.
final uptake of CO2 increased by ∼4% (equivalents based on the C6/LeuC1 content) per wt % of water: %CO2/wt % water ) 107:0, 121:3, 138:6, 152:10. These results and the previous investigations of changes effected by bubbling CO2 through dilute solutions of amidines in water26 indicate that significant amounts of bicarbonate salts were present, as well in the RTIL samples containing water. The time-dependence of CO2 uptake by some of the amidine/primary acid ester mixtures was followed by using a gas burette containing 1 atm pressure of CO2. For example, as was observed for the uptake of CO2 by amidine/alkyl amine mixtures,15 C8/IleC1, C8/LeuC1, and C8/IleC8 added CO2 rapidly,27 reaching a constant value near 100% (Figure 1 and Table S1 in the Supporting Information), as expected if the reaction in Scheme 1 (leading to RTIL formation) goes to completion. In fact, the actual uptake by each of these L-n-C combinations exceeds the stoichiometric amount by ca. 10%. CO2 gas is known to be solubilized readily by several ionic liquids,28 including the amidine/alkylamine mixtures.15 In classic (nonprotic) ionic liquids, the physical absorption of CO2 has been attributed primarily to anionCO2 interactions;28a in our L-n-C systems, H-bonding is also a probable contributor. The uptake of CO2 by the C8/ProC1 system, containing a secondary amino acid methyl ester, stopped after ∼90% of the amount calculated for 1 equiv. The fractions that are chemically fixed and which are physically absorbed are not known at this time. If one assumes that 80% uptake indicates that the C8/ProC1 system produces a useful RTIL. (26) (a) Hori, Y.; Nagano, Y.; Miyake, S.; Teramoto, S.; Taniguchi, T. Chem. Express 1986, 1, 311-314. (b) Heldebrant, D. J.; Jessop, P. G.; Thomas, C. A.; Eckert, C. A.; Liotta, C. L. J. Org. Chem. 2005, 70, 5335-5338. (c) Liu, Y.; Jessop, P. G.; Cunningham, M.; Eckert, C. A.; Liotta, C. L. Science 2006, 313, 958-960. (27) The slope difference between C8/IleC1 and C8/ProC1 samples depends on the surface area of the L/n samples that is exposed to the CO2 gas as well as the rate at which the liquids are stirred. Thus, the rate of CO2 uptake by C8/IleC1, but not the ultimate %CO2 plateau value, was varied by altering the size of the vessel in which the experiment was conducted. (28) (a) Cadena, C; Anthony, J. L.; Shah, J. K.; Morrow, T. I.; Brenneck, J. F.; Maginn, E. J. J. Am. Chem. Soc. 2004, 126, 5300-5308. (b) Shiflett, M.; Yokozaki, A. Ind. Eng. Chem. Res. 2005, 44, 44534464.
Chem. Mater., Vol. 19, No. 19, 2007 4765
Figure 2. TGA curves for neat (a) C8/LeuC1 and (b) C8-LeuC1-C heated from room temperature. Curve c is curve a minus curve b.
We anticipated a lower uptake of CO2 by the L/ProC1 systems because primary amines are known to react more readily with CO2 than secondary (or tertiary) amines.29 The L-n-C combinations are not stable to acid treatment because protonation of the carbamate moieties leads to rapid loss of CO2 from the carbamic acids (at least at 1 atm pressures of CO2).30 TGA measurements of weight loss from heating C8/LeuC1 and C8-LeuC1-C are shown in Figure 2. They indicate that some of the amidine and amino acid ester are evaporated in the stream of nitrogen gas, which passes over the sample at elevated temperatures. The subtracted curve, Figure 2c, has a plateaulike region above ca. 50 °C, corresponding to the loss of 10.9% sample weight. This amount is in excellent agreement with the calculated weight for the loss of 1 mol equiv of CO2, 11.5%. The calculated CO2 weight loss from an isothermal run at 30 °C on the same compounds was 15.4% (Figure S8 of the Supporting Information); it occurred during ca. 20 min. These results demonstrate that the loss of CO2 gas from the L-n-C combinations becomes important in air or nitrogen atmospheres above room temperature. The RTILs are more stable to heating when placed under CO2 atmospheres. We have also explored the temperatures at which some of the L-n-C RTILs crystallize. Except for those from amidine/octadecyl esters, all of the L-n-C combinations remained liquids at -18 °C. Solidification of C6-LeuC18-C and C6-IleC18-C was observed by polarizing optical microscopy (Figure 3) as samples were cooled from room temperature and then heated subsequently from below their melting temperatures. For C6-LeuC18-C, solidification and melting occurred at 14.3-12.6 and 11.8-12.5 °C, respectively; for C6-IleC18-C, the corresponding temperatures were 11.6-9.2 and 9.5-11.0 °C. The freezing and melting temperatures for C6-LeuC18-C are in reasonably good agreement with the values determined by DSC (See Figure S1 of the Supporting Information) where the onsets of the (29) (a) Hoerr, C. W.; Harwood, H. J.; Ralston, A. W. J. Org. Chem. 1944, 9, 201-210. (b) Lallau, J.-P.; Masson, J.; Guerin, H. Bull. Soc. Chim. Fr. 1972, 3111-3112. (c) Nakamura, N.; Okada, M.; Okada, Y.; Sugita, K. Mol. Cryst. Liq. Cryst. 1985, 116, 181-186. (30) (a) Carretti, E.; Dei, L.; Baglioni, P.; Weiss, R. G. J. Am. Chem. Soc. 2003, 125, 5121-5129. (b) Carretti, E.; Macherelli, A.; Dei, L.; Weiss, R. G. Langmuir 2004, 20, 8114-8118.
4766 Chem. Mater., Vol. 19, No. 19, 2007
Yamada et al.
Figure 3. Optical micrographs of solidified C6-LeuC18-C (left) and (b) C6-IleC18-C (right) at 8 °C. Scale bars are 400 µm in width. Table 2. Specific Rotationsa of Alkyl Amines and Amino Acid Esters and Their Amidine Mixtures before and after Exposure to 1 atm CO2b sample
[R]total
[R]opt
sample
[R]total
[R]opt
LeuC1 C6/LeuC1 C6-LeuC1-C
17.1 6.1 -9.1
8.9 -13.1
LeuC8 C6/LeuC8 C6-LeuC8-C
6.2 6.5 -12.9
2.7 -5.3
IleC1 C6/IleC1 C6-IleC1-C
44.7 16.7 -1.0
19.8 -1.2
a Notation: [R] total, specific optical rotation for neat samples and mixtures; [R]opt, specific optical rotation for the optical active component in amidine/ amino acid ester mixtures. b The measured rotations, R, are collected as Table S4 in the Supporting Information; [R] values are (1°.
cooling exotherm (-24.0 J/g) and heating endotherm (19.2 J/g) were at 7.6 and -3.1 °C, respectively; for C6-IleC18C, the corresponding temperatures (and heats) were 9.0 °C (-22.0 J/g) and -5.5 °C (24.9 J/g). In subsequent cooling and heating cycles, the heats of transitions became progressively lower. We ascribe this behavior to loss of some CO2 while the sample was heated to 30 °C. Furthermore, C6LeuC1-C, to which had been added 3 wt % of water, also remained a liquid at -18 °C. The formation of amidinium carbamate is also evidenced by IR spectral measurements. The typical NdC amidine stretch at 1629 cm-1 31 in the FT-IR spectrum of a neat mixture of C8/IleC1 was replaced by bands at 1646 and 1575 cm-1, assigned to protonated amidine31 and carbamate24 stretches, respectively, of C8-IleC1-C (Figure 4). In addition, there is a broad band centered at 3391 cm-1 that we attribute to the amidinium and carbamate N-H stretching modes; the broad ammonium N-H stretching band of protonated C8 (by 1 equiv of trifluoroacetic acid in chloroform) is centered at 3423 cm-1. Because the amino acid esters are optical active, we have investigated the changes that occur in their specific optical rotations at 589 nm before and after the addition of CO2 (Table 2). As expected, the [R]total values of achiral amidine/
amine mixtures remain within the experimental error of 0° before and after the addition of CO2. However, [R]total values of the mixtures with an amino acid ester, especially those with leucine, changed significantly in magnitude and changed sign upon the addition of CO2. Somewhat surprisingly, the [R]opt values of C6/LeuC1 and C6/IleC1 are lower than the [R]total values of their neat optically active components, LeuC1 and IleC1, respectively. This result suggests that even before the addition of CO2 to the L/n mixtures, amidine and amino acid ester molecules interact strongly. The specific nature of that interaction is not obvious from any of the other measurements we have made. However, we suspect that there is a strong interaction between the relatively acidic amino protons and the basic amidine functionality. An H-bonding interaction of his sort could alter the conformation at the chiral center of the amino acid esters and lead to the changes observed.32 These interactions are in addition to the microdomains that may exist if the charged centers segregate themselves from the less-polar groups.33 A significantly different (and stronger) interaction between the two molecular components upon the addition of CO2 (and amidinium carbamate formation) is signaled by the change from positive to negative [R]total values for the C6/LeuC1 and C6-LeuC1-C systems and the large decrease of [R]total to nearly 0° for the C6/IleC1 system. In addition, the [R]total values of C6/LeuC1 and C6/LeuC8 before and after CO2 exposure indicate that the conformation at the chiral center
(31) The wavenumber assignment of the protonated amidine was confirmed by the IR spectrum of a 1:1 C8/trifluoroacetic acid mixture which showed a CdN stretch at 1648 cm-1. Protonation of the amidines has been reported: Corset, C.; Froment, F. J. Phys. Chem. 1990, 94, 6908-6911.
(32) (a) Circular Dichroism; Berova, N., Nakanishi, K., Woody, R. W., Eds.; Wiley-VCH: New York, 2000. (b) Lambert, J. B.; Shurvell, H. F.; Verbit, L.; Cooks, R. G.; Stout, G. H. Organic Structural Analysis; Macmillan: New York, 1976; Part 3, Chapter 3. (33) Wang, Y.; Voth, G. A. J. Phys. Chem. B 2006, 110, 18601-18608.
Figure 4. IR spectra of neat C8/IleC1 (a) before and (b) after exposure to CO2.
ReVersible, Room-Temperature, Chiral Ionic Liquids
of the amino acid ester is affected grossly by the absence or presence of CO2 but not by the length of the alkyl ester chain. Attempts to investigate the chirality of the RTILs and their L/n precursors by circular dichroism (CD) were unsuccessful because the achiral amidines absorb much more strongly in the UV region than do the chiral amino acid esters. Apparently, the aforementioned interactions between the amidines and the amino acid esters, as well as those between the amidinium and carbamate components of the RTILs, do not induce significant chirality in the (planar) amidine parts (see Figures S3 and S4 in the Supporting Information). Thus, both C6/IleC1 and C6/LeuC1 showed no clear CD bands at wavelengths corresponding to their absorption bands at wavelengths where the optical densities were 290 nm), led to no discernible induced CD bands, as well, when they were added as solutes to C6/IleC1 and C6/LeuC1 and their RTILs (see Figures S5 and S6 in the Supporting Information). A major attribute of these systems is their large polarity changes when exposed to CO2 and N2 gases.15 As an example, the polarity of one of our reversible amidine/amino acid ester systems was estimated using the solvatochromic dye,34 DAPNE, as a probe. Its 425 nm absorbance maximum in C8/IleC1 indicates an environment like that afforded by toluene (λmax ) 425 nm).35 After CO2 exposure, the λmax value is shifted to 443 nm, indicating a medium slightly less polar than N,N-dimethylformamide (DMF; λmax ) 446 nm).35 Consistent with the presence of the ester functionality in the C8/IleC1 system, it is slightly more polar than the C8/1-hexylamine system in which DAPNE exhibited a λmax ) 423 nm before and λmax ) 438 nm after the addition of CO2 (See Table S5 in the Supporting Information).15 As such, it is possible to dissolve reversibly low-polarity materials into the L/n phase and separate them from the L-n-C phase or, by bubbling through N2 gas, redissolve them into the L/n phase. An example of this strategy is shown in Figure 5, and a similar separation-dissolution strategy has been reported with DBU and 1-hexanol.14a A solution of C8/ LeuC1 and n-decane containing Eosin Y became opaque after bubbling CO2 through it for
