DOI: 10.1002/chem.

201601201 Full Paper

& Structure Elucidation

The Conformational Map of Volatile Anesthetics:

Enflurane Revisited
Cristûbal P¦rez,[a, f] Elena Caballero-Mancebo,[b] Alberto Lesarri,*[b] Emilio J. Cocinero,[c]
Ibon Alkorta,[d] Richard D. Suenram,[a] Jens-Uwe Grabow,[e] and Brooks H. Pate*[a]

Abstract: Previous ambiguities in the conformational and provided structural information using the substitution and
structural landscape of the volatile anesthetic enflurane have effective procedures. The structural preferences were ration-
been solved combining microwave spectroscopy in a jet ex- alized with additional ab initio, natural-bond-orbital and
pansion and ab initio calculations. The broadband (2– non-covalent-interaction analysis, which suggest that plausi-
18 GHz) rotational spectra identified three different rotamers, ble anomeric effects at the difluoromethyl group could be
sharing a common trans ether skeleton but differing in the overridden by other intramolecular effects. The difluoro-
œ gauche/trans position of the terminal chlorine atom. For methyl orientation thus reflects a minimization of inter-fluo-
each chlorine conformation two different gauche orienta- rine repulsions while maximizing F···H attractive interactions.
tions were predicted for the opposite difluoromethyl group, A comparison with previous electron diffraction and spectro-
but only one is experimentally observable due to collisional scopic data in the gas and condensed phases finally resulted
relaxation in the jet. The experimental dataset comprised in a comprehensive description of this ether, completing
nine different isotopologues (35Cl, 37Cl, 13C) and a large a rotational description of the most common multi-halogen-
number (> 6500) of rotational transitions. The inertial data ated anesthetics.

Introduction pounds leads to reduced metabolism and flammability, making

them appropriate as inhalational anesthetics. Volatile anesthet-
Many halogenated hydrocarbons are effective to induce gener- ics offer different reasons of interest. First of all, the molecular
al anesthesia. Following the initial development of halothane, mechanism of general anesthetics are not totally understood.[2]
fluoroxene and methoxyflurane in 1951–59, several methyl- The lack of experimental structures of the putative membrane
ethyl and methyl-isopropyl polyfluorinated ethers like sevoflur- receptors at atomic resolution has been recognized as the
ane, isoflurane, enflurane or desflurane were introduced in the “main stumbling block” for location and characterization of an-
clinical practice in the late 1960s.[1, 2] The presence of fluorine esthetic binding sites.[3] Because of crystallization problems,
atoms in these chemically stable and easy volatilizing com- many X-ray diffraction studies simply target model systems not
directly mediating anesthetic action, like soluble proteins.[4]
[a] Dr. C. P¦rez, Dr. R. D. Suenram, Prof. B. H. Pate Secondly, the structural analysis of the bare molecules or ag-
Department of Chemistry, University of Virginia gregates involving residues in the binding pockets are scarce.
McCormick Rd., Charlottesville, Virginia 22904 (USA) In particular, the high-resolution methods of vibrational[5–7] or
E-mail: brookspate@virginia.edu
rotationally resolved[8–10] spectroscopy have been applied to
[b] E. Caballero-Mancebo, Prof. Dr. A. Lesarri
Departamento de Qu†mica F†sica y Qu†mica Inorg‚nica
a few anesthetics and relatively small systems. Finally, the anal-
Universidad de Valladolid, 47011 Valladolid (Spain) ysis of molecular aggregation requires a good knowledge of
E-mail: lesarri@qf.uva.es the conformational landscapes of the free anesthetic molecule,
[c] Dr. E. J. Cocinero not always available and requiring both experimental and the-
Departamento de Qu†mica F†sica
oretical information. Clearly, the reliability of experimental data
Universidad del Pa†s Vasco, Ap. 644, 48080 Bilbao (Spain)
is the basis for extension of the theoretical molecular models
[d] Prof. I. Alkorta
Instituto de Qu†mica M¦dica, CSIC, Madrid (Spain) to larger clusters and interpretation of the origin of the struc-
[e] Prof. Dr. J.-U. Grabow tural preferences.
Institut fìr Physikalische Chemie and Elektrochemie Here we report on the conformational and structural proper-
Lehrgebiet A, Gottfried-Wilhelm-Leibniz-Universit•t ties of enflurane, probed in the cold and effectively isolated
Callinstrasse 3A, 30167 Hannover (Germany)
conditions of a supersonic jet expansion. This work is primarily
[f] Dr. C. P¦rez
intended to remove the structural ambiguities and inconsisten-
Present address: Max-Planck Institute for the Structure and
Dynamics of Matter, Luruper Chaussee 149, 22761 Hamburg (Germany) cies observed in previous gas electron diffraction[11] (GED) and
Supporting information and ORCID from authors for this article are avail- condensed-phase studies, which gave conflicting information
able on the WWW under http://dx.doi.org/10.1002/chem.201601201. of its molecular structural properties. Previous studies included

Chem. Eur. J. 2016, 22, 9804 – 9811 9804 Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
 Full Paper

NMR spectroscopy,[12] vibrational circular dichroism (VCD),[13] 25 kJ mol¢1. All these conformations were further re-optimized
low-resolution IR and Raman analysis[14–16] and laser photoa- by using ab initio methods. The conformational composition
coustic spectra.[17] using second-order Møller–Plesset theory (basis set 6/311 + +
Enflurane is a structural isomer of isoflurane,[9] with the G(2df,p)) includes 23 different structures, summarized in
chlorine atom displaced to the terminal ethyl carbon Table S1 and Figure S1 (Supporting Information). Vibrational
(Scheme 1). This change results in a more complicated confor- frequency calculations characterized these stationary points as
mational landscape and different conformational preferences. local minima. The six most stable conformations of enflurane
Isoflurane and its fluorinated analogue desflurane share an are close in energy, differing by less than 1 kJ mol¢1 (Table S2,
anti-arrangement of the carbon skeleton for their most stable Supporting Information). For these conformations the quan-
conformation (t(C1-C2-O-C3) = 137.88 and 145.48, respectively). tum chemical predictions suggest a common trans carbon
Characteristically, hyperconjugative anomeric effects play skeleton (T, dihedral f2(C1-C2-O-C3) ca. 1808). The threefold in-
a major role in the conformational stability of the haloeth- ternal rotation of the terminal chloro-fluoromethyl group gen-
ers,[9, 14, 18] making those systems difficult to disentangle. On the erates three alternative gauche + /gauche¢/trans (G + /G¢/T)
other hand, the previous GED work on enflurane reported that conformations, with staggered orientations with respect to the
the carbon skeleton was trans, as in the ethyl methyl ether neighboring difluoromethylene group (f1(Cl-C1-C2-O) ca. œ 608,
prototype. The GED conformational composition was repro- 1808). The opposed difluoromethyl group adopts two orienta-
duced with a mixture of three conformers differing in the tions slightly gauche + /gauche¢ (g + /g¢ , f3(C2-O-C3-H) ca.
chlorine position, but it could not distinguish between two of œ 208) that are practically isoenergetic, finally resulting in the
them and admitted that the molecule is “really underdeter- six lowest-lying conformations in Figure 1 ((S)-enflurane stereo-
mined experimentally”.[11] Moreover, the trans carbon skeleton chemistry was adopted throughout). Initial MP2 calculations
conflicts with the claim of a gauche conformation initially sug- suggested that conformer G¢Tg¢ (dihedrals f1/f2/f3, later ab-
gested from NMR coupling constants.[12] Later low-resolution breviated to G¢) is the global minimum, followed by his coun-
VCD[13] and IR studies[14, 15] based their conclusions on ab initio terpart G¢Tg + and the two pairs G + Tg + /G + Tg¢ and TTg¢/
predictions and are not conclusive. Moreover, the very small TTg + . Additional CCSD(T) calculations later confirmed this
energy difference between the most stable conformations (< energy ordering (B3LYP reverses the two higher energy con-
0.4 kJ mol¢1) has made it difficult to establish the global mini- formers). Molecular conformations with non-trans carbon skele-
mum, and a more recent calculation suggested a reversed tons start at relative higher energies above 4.6 kJ mol¢1 (MP2).
order for the two most stable species with a small change in Consequently, several species were initially expected to be
the basis set.[19] The latest IR and Raman spectra offered other populated in the spectrum. Atoms-in-molecules[24] (AIM), non-
improved ab initio calculations.[16] Recently, a conformational covalent index[25] (NCI) and natural bond orbitals[26] (NBO)
exploration of enflurane in solution was also reported.[20] We methodologies were additionally applied to the most stable
address here all these issues exploiting the sub-Doppler resolu- conformations.
tion [kHz], isotopologue-sensitivity and broadband capabilities
of modern microwave spectroscopy,[21, 22] supplemented with
high-level ab initio calculations. The results now establish un-
2. Rotational spectrum
ambiguously the conformational map and molecular structure
of enflurane, paving the way for future analysis of enflurane The quantum mechanical predictions were checked against
clustering properties with high-resolution techniques. the rotational spectrum, recorded in a jet-cooled supersonic

Scheme 1. The volatile anesthetics desflurane, isoflurane and enflurane

(WHO ATC codes in parentheses).

Results
1. Conformational space
The conformational space of enflurane was explored systemati-
cally combining molecular mechanics search algorithms and
ab initio calculations. The conformational search (MMFFs) used
Figure 1. The six most stable conformers of enflurane differ in the orienta-
a specific combination of large-scale low-mode and Monte
tion of the chlorofluoro methyl (gauche¢, gauche + , trans, or G¢ , G + , T) and
Carlo multiple minimization procedures[23] and resulted in a set difluoromethyl groups (gauche œ or g + /g¢). The conformations observed ex-
of 43 different structures within an energy window of perimentally are labeled as A, B, C (MP2 relative energies).

Chem. Eur. J. 2016, 22, 9804 – 9811 www.chemeurj.org 9805 Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
 Full Paper

duced to experimental accuracy with a Watson semirigid-rotor

Hamiltonian[30] supplemented with nuclear quadrupole cou-
pling terms.[31]
Rotational constants, centrifugal distortion parameters and
coupling coefficients in Table 1 were obtained using Pickett’s
spherical tensor formulation of rotation and spin operators[32]
(a comparison between asymmetric and symmetric reduction
fits is presented in Tables S3–S5, Supporting Information). We
determined all the elements of the coupling tensor c (except
cac), which accounts for the interaction between the quadrupo-
lar moment of the 35Cl nucleus and the molecular electric field
gradient at the chorine position.[31]
The rotational study was immediately extended to the 37Cl
and all three different 13C isotopologues in natural abundance
Figure 2. The broadband rotational spectra of enflurane in the 2–8 GHz (24.2 and 1.1 %, respectively), which give independent spectra
region using a chirp-pulse microwave spectrometer (1.9 m averaging cycles). and were analyzed similarly (Table 2 and Table S6 in Supporting
Information). The experimental results for the isotopic species
expansion. The spectrum was measured in several periods are consistent with the predicted variation of the rotational
using both Balle–Flygare-type[27, 28] Fourier transform micro- and centrifugal distortion constants. As expected, the coupling
wave (FTMW) spectrometers and chirped-pulsed (CP-FTMW) constants of the 35Cl and 37Cl species scaled as the ratio of
broadband techniques,[29] described below. The 2–8 and 6– their quadrupolar moments (1.2686[33]). However, it was imme-
18 GHz full-band spectra are shown in Figure 2 and Figure S2 diately noticed that conformation A, despite apparently the
in the Supporting Information, respectively. Despite the drastic most intense in the spectrum, did not correspond to the pre-
reduction in rotational temperatures in the jet (< 2 K), the dicted global minimum, and the search was extended to other
combination of multiple conformations, two abundant chlorine species. The subtraction of conformer A from the original spec-
isotopologues, minor carbon isotopologues, hyperfine effects trum revealed many remaining transitions. The assignment
and differences in magnitude of the electric dipole moments process, illustrated in Figure S3, led to the identification of
rendered the spectrum relatively dense and not straightfor- a second conformer B, showing similar quadrupole hyperfine
ward to assign. splittings. The intensity of the spectrum permitted the detec-
The first spectral surveys searched the most intense transi- tion of both the 35Cl and 37Cl isotopologues, which were ana-
tions in the 8–18 GHz region, leading to the assignment of an lyzed with the same Hamiltonian. Noticeably, many transitions
asymmetric rotor with all selection rules active, denoted con- were very sensitive to the off-diagonal quadrupole coupling
former A. Nuclear quadrupole coupling hyperfine splittings terms, so the fitted rotational results in Table 1 include the full
arising from the chlorine nucleus were clearly visible in all tran- coupling tensor. Interestingly, this second conformation did
sitions, spanning in some cases more than 30 MHz. Typical hy- not correspond to the predicted global minimum either,
perfine patterns are illustrated in Figure 3 and Figure S3 (Sup- prompting the extension of the spectral search down to
porting Information). The experimental transitions were repro- 2 GHz. Using broadband spectroscopy and additional integra-
tion time, we finally assigned the rotational transitions from
a third conformer C. Despite the lower intensity we observed
both the 35Cl and 37Cl species, again split by nuclear quadru-
pole coupling effects. The resulting rotational parameters and
full coupling tensor are also collected in Table 1. The full set of
experimental measurements for the nine different species re-
corded for enflurane finally comprised more than 6500 individ-
ual transitions, which are collected in Tables S7–S15 in the Sup-
porting Information. No additional rotamers were detected in
the spectrum.
Three arguments supported the conformational assignment,
including the isotopic structural information (below), the rota-
tional and centrifugal distortion parameters and the nuclear
quadrupole coupling tensor values. All evidence conclusively
indicates that the detected conformers can be identified with
three different gauche + (G + ), trans (T) and gauche¢ (G¢) ori-
Figure 3. A section of the rotational spectrum of enflurane showing a com- entations of the chlorine atom, the weakest detected rotamer
parison between the experimental and predicted spectral intensities for
being the predicted global minimum (A = G + Tg¢ , B = TTg + ,
a conformational composition 1:1:1 (each transition is split typically in four
more intense hyperfine components by nuclear quadrupole coupling, color C = G¢Tg¢). On the other hand, a single conformation was ob-
version in Figure S4). served in each case for the difluoromethyl group (g¢ for the

Chem. Eur. J. 2016, 22, 9804 – 9811 www.chemeurj.org 9806 Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
 Full Paper

Table 1. Rotational parameters of enflurane (35Cl species): experiment versus theory.

Conf. C (G¢Tg¢) Conf. A (G + Tg¢) Conf. B (TTg + )

experiment[e] theory[f] experiment theory experiment theory

A [MHz][a] 1857.35494(28) 1865.67 1656.134974(92) 1643.63 2383.59171(19) 2389.65
B [MHz] 722.78593(18) 727.54 740.909748(52) 749.86 631.455371(79) 636.08
C [MHz] 605.95878(16) 609.31 683.075497(47) 694.49 574.955289(88) 578.37
DJ [kHz] 0.0518(17) 0.0391 0.39390(19) 0.0791 0.02606(33) 0.0213
DJK [kHz] ¢0.0410(88) ¢0.0238 2.0642(14) ¢0.0947 0.1053(16) 0.0714
DK [kHz] 0.507(14) 0.388 2.2842(22) 0.351 0.154(10) 0.184
dJ [kHz] 0.00659(92) 0.00953 -0.12865(11) 0.00847 0.00462(14) 0.00381
dK [kHz] 0.263(30) 0.191 5.7977(82) 0.473 0.632(19) 0.353
caa [MHz][b] 23.9679(43) 21.98 27.6749(28) 28.56 ¢46.9715(36) ¢43.78
cbb [MHz] ¢58.6093(58) ¢54.37 ¢61.5755(40) ¢59.32 18.4503(50) 16.58
ccc [MHz] 34.6415(101) 32.39 33.9005(68) 30.76 28.5211(86) 27.20
j cab j [MHz] 33.919(72) 31.98 35.29(12) 28.16 37.61(16) 36.03
j cac j [MHz] 8.82(29) 8.19 [0.] 6.67 31.36(20) 28.88
j cbc j /MHz 21.847(95) 19.92 13.648(44) 15.80 13.742(28) 13.05
j ma j [D] 0.12 0.63 0.31
j mb j /[D] 0.28 1.03 0.89
j mc j /[D] 0.21 1.73 0.09
mTOT/[D] 0.37 2.11 0.94
DEMP2 [kJ mol¢1][c] 0.0 0.7 0.9
DECCSD(T) [kJ mol¢1] 0.0 0.3 1.1
DEB3LYP [kJ mol¢1] 0.0 0.7 0.4
DGMP2 [kJ mol¢1] 0.0 0.7 0.8
DGB3LYP [kJ mol¢1] 0.0 2.1 1.5
N[d] 776 2119 965
s [kHz] 12.1 12.2 12.6

[a] Rotational constants (A, B, C) and Watson’s A-reduction centrifugal distortion constants (DJ, DJK, DK, dJ, dK). [b] Nuclear quadrupole coupling tensor cab
(a, b = a, b, c) and electric dipole moment ma (a = a, b, c). [c] Electronic (DE) and Gibbs free energies (DG, 298 K). [d] Number of fitted transitions (N) and mi-
crowave standard deviation (s) of the fit. [e] Standard errors in parentheses in units of the last digit. [f] MP2 structural and electrical parameters. The
CCSD(T) calculations used the MP2 geometry.

two most stable species and g + for the third one). The com-
Table 2. Rotational parameters of the 37Cl isotopologues of enflurane. parison between the ground-state rotational constants and the
Conf. C (G¢Tg¢) Conf. A (G + Tg¢) Conf. B (TTg + ) equilibrium ab initio values in Table 1 and Table S2 (Supporting
37
Cl 37
Cl 37
Cl Information) shows a definitive agreement for the proposed
A [MHz][a] 1830.71055(48) 1640.53612(16) 2380.92342(24) assignment. This conclusion is particularly reinforced by the
B [MHz] 714.22486(35) 730.734776(95) 619.03595(14) comparison of the nuclear quadrupole coupling tensor ele-
C [MHz] 597.21181(29) 671.970812(78) 564.49445(12) ments, which are extremely sensitivity to the molecular confor-
DJ [kHz] 0.0496(35) 0.40550(60) 0.02506(59) mation.[34] This fact allowed distinguishing even the minor dif-
DJK [kHz] ¢0.062(19) 1.6930(27) 0.1066(41)
DK [kHz] 0.517(29) 3.1987(64) 0.159(11) ferences between the difluoromethyl g¢/g + gauche orienta-
dJ [kHz] 0.0061(20) ¢0.12472(30) 0.00422(26) tions. The theoretical predictions in Table 1 additionally contain
dK [kHz] 0.285(68) 5.106(15) 0.656(41) conformational energies for three molecular models, including
caa [MHz] 18.2100(73) 21.0383(28) ¢37.1858(47) MP2, B3LYP and CCSD(T). The relative energies can be com-
cbb [MHz] ¢45.3488(99) ¢47.9635(40) 14.7407(63)
ccc [MHz] 27.1388(172) 26.9253(68) 22.4451(110) pared with the conformer populations in the jet, which were
j cab j [MHz] 27.67(14) 28.70(12) 29.33(15) estimated from relative intensity measurements.[35] This calcula-
j cac j [MHz] 7.41(55) [0.0000] 24.91(17) tion assumed a square dependence of the spectral intensities
j cbc j [MHz] 17.30(19) 9.992(75) 10.61(14) with the electric dipole moments for the chirped-pulse spectra
N 447 1341 705
s [kHz] 14.6 10.8 13.1 (weak pulse limit).[21] The dipole components were taken from
the MP2 ab initio predictions. Following different simulations
[a] Parameter definition as in Table 1.
for the three different conformers the experimental spectrum

Chem. Eur. J. 2016, 22, 9804 – 9811 www.chemeurj.org 9807 Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
 Full Paper

was finally reproduced satisfactorily with a conformational 4. Stereoelectronic effects

composition of 1:1:1, as exemplified in Figure 3 and Figure S4
(Supporting Information). The jet rotational temperature was The AIM, NCI and NBO methodologies offer information on the
assumed to be 2 K, following previous observations.[36] The origin of the structural preferences of enflurane. The AIM mo-
spectral intensity differences between the three conformations lecular graphs in Figure S5 (Supporting Information) show the
are thus attributable to the different electric dipole moments. lines of maximum electron density linking bonded nuclei and
the bond critical points for the three most stable conforma-
tions. This analysis does not reveal any unexpected interactions
in these conformations. The NBO calculation of electronic
3. Structure
charge transfer from occupied to empty orbitals is shown in
The substitution (Kraitchman) and effective methods offer al- Table S20, and mostly involves the two oxygen lone-pairs LP(1)
ternative operational definitions of the equilibrium and and LP(2) in Figure 4 and Figure S6 (Supporting Information)
ground-state structures,[37] which can be directly compared to as electron donors. For the three detected conformations of
the theoretical results for the isolated molecule. We applied enflurane LP(1) is predicted to poorly interact with the sur-
both methods to the most intense enflurane G + conformation, rounding antibonding s*(C-F) orbitals. In contrast, LP(2) shows
where an experimental set of fifteen ground-state moments of large stabilization energies with both difluoromethylene s*(C-
inertia was available. The substitution coordinates do not pro- F2) and (C-F3) orbitals and one of the (C-F4) or (C-F5), depend-
vide a full structure here, as the five fluorine atoms are monoi- ing on the conformation. NCI results are commented below.
sotopic and the low-abundance 18O species was not detected
in the spectrum. Nevertheless, the atomic coordinates of the
carbon and chlorine atoms in Table S16 (Supporting Informa-
tion) constitute a definitive proof of the identification of con-
former G + . For comparison, the three sets of theoretical coor-
dinates for conformers G + , G¢ and T are shown in Tables S17–
S19 (Supporting Information). In the effective structure calcula-
tion we floated a total of five valence angles and dihedrals, the
bond lengths being constrained to the ab initio values. The re-
sulting molecular structure of enflurane G + is compared in
Table 3 with the ab initio MP2 structural data. This structure
reproduces the experimental rotational constants below
0.2 MHz.

Table 3. Experimental and ab initio (MP2) structures of enflurane.

Conf. G + Tg¢ Conf. G¢Tg¢ Conf. TTg +

effective (r0) ab initio (re) ab initio (re) ab initio (re)
r(C1-C2) [æ] 1.526 1.527 1.527
r(C2-O) [æ] 1.357 1.363 1.366 Figure 4. Non-bonded molecular orbitals (LP(1) and LP(2)) at the oxygen
r(C3-O) [æ] 1.389 1.389 1.389 atom of enflurane involved in hyperconjugative interactions with the vicinal
r(C1-Cl) [æ] 1.753 1.754 1.755 fluorine atoms (color version in Figure S6).
r(C1-F1) [æ] 1.354 1.355 1.353
r(C2-F2) [æ] 1.348 1.350 1.344
r(C2-F3) [æ] 1.354 1.344 1.347
r(C3-F4) [æ] 1.340 1.340 1.333 Discussion
r(C3-F5) [æ] 1.333 1.334 1.341
The conformational and structural map of enflurane has been
a(C1-C2-O) 111.1(1) 110.2 108.4 107.0
a(C2-O-C3) 116.8(4) 115.9 115.7 115.7 fully resolved. Despite the small size of the molecule, the large
a(Cl-C1-C2) 113.5(3) 110.8 110.8 110.3 number of low-energy conformations posed in the past experi-
a(F1-C1-C2) 108.4 107.7 108.3 mental and theoretical problems. The initial gas-phase investi-
a(C1-C2-F2) 109.5 108.5 111.2
gation of enflurane using electron diffraction[11] did not fully
a(C1-C2-F3) 108.1 111.2 109.9
a(O-C3-F4) 109.1 109.0 106.8 discriminate between the most stable conformations and, as
a(O-C3-F5) 106.7 106.8 109.1 a consequence, the conformational populations were con-
f1 = t(Cl-C1-C2-O) 63(1) 60.7 ¢62.2 ¢178.9 fused. Moreover, the GED work used an incomplete set of ab
f2 = t(C1-C2-O-C3) ¢172.4(5) ¢176.8 ¢174.3 176.2
t(F1-C1-C2-O) ¢60.9 177.0 60.3
initio predictions, missing some of the most stable conforma-
t(F2-C2-O-C3) ¢54.7 ¢55.2 ¢62.3 tions and offering inconclusive conformational energies. These
t(F3-C2-O-C3) 63.5 63.5 56.3 inconsistencies were transmitted along the years to other low-
t(F4-C3-O-C2) 99.8 101.1 143.5 resolution spectroscopy works,[12–17] biasing the interpretation
t(F5-C3-O-C2) ¢143.6 ¢142.5 ¢100.0
f3 = t(H-C3-O-C2) ¢22.3 ¢21.0 22.0
of the experimental results. We now definitively establish
based on experimental evidence that the conformational map

Chem. Eur. J. 2016, 22, 9804 – 9811 www.chemeurj.org 9808 Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
 Full Paper

of enflurane primarily originates from the internal rotation of (gauche F) to the CO bond, to allow for optimal hyperconjuga-
the chlorofluoromethyl group, in combination with two pre- tive interactions between the oxygen lone pairs in Figure 4
ferred gauche orientations at the difluoromethyl side. The and the s*(C-F) orbitals.[40] The fluorinated methyl ethers are
gauche carbon structure suggested by the NMR study[12] is de- consistent with this effect, except difluoromethyl ether.[41] In
finitively dismissed, as the six most stable conformers of free enflurane, steric[11] and electrostatic[14] repulsions have been
enflurane share a trans skeleton. We observed three different proposed to override the anomeric effect, which could also ex-
chlorine conformations, suggesting that relatively high poten- plain that trifluoromethyl difluoromethyl ether (CF3OCF2H) is
tial barriers are hindering this internal rotation. Our ab initio stabilized in the H gauche form. We checked this hypothesis
(MP2) modeling in Figure S7 (Supporting Information) consis- combining the information from the NBO and NCI methods.
tently gave torsional barriers between 19.9 and 22.9 kJ mol¢1 Our NBO hyperconjugation results for the most stable con-
for the G¢ !T and G¢ !G + interconversions. The global mini- formers of enflurane in Table S20 are similar to those previous-
mum has a gauche G¢ chlorine orientation, while the gauche ly reported[14, 20] and confirm sizeable hyperconjugation ener-
G + and trans T species are slightly less stable (CCSD(T): 0.3 and gies in the three conformers. However, the NCI plots in
1.1 kJ mol¢1, respectively). The GED work incorrectly reported Figure 5 and Figure S8 (Supporting Information), which dis-
the T species as most stable, based on their theoretical calcula- criminate between hydrogen bonds, weak delocalized interac-
tions. Noticeably, two quasi isoenergetic possibilities exist for tions or destabilizing repulsions on the basis of the reduced
each of the three chlorine rotamers, differing in the gauche electron density gradient, offer additional information. The NCI
g + /g¢ orientations of the difluoromethyl group. In the jet analysis of CF3OCF2H and the three most stable conformers of
spectrum only one of these g + /g¢ possibilities is experimental- enflurane shows the presence of a weak attractive interaction
ly detectable, as the small potential barrier within each pair (green surface in Figure S8) between the hydrogen atom of
(calculated as 0.13 kJ mol¢1 in Figure S7) surely produces a con- the difluoromethyl group with the fluorine atoms in the adja-
formational relaxation of the higher-energy conformation to cent carbon. In consequence, the conformation of the CF2H
the most stable rotamers,[38] resulting in three detectable spe- group results from a compromise between the stabilization of
cies. The jet spectrum thus offers unquestionable evidence of the anomeric effect with the oxygen lone pairs and the inter-
the most stable species of enflurane. In the second G + confor- actions of the atoms of this group with the rest of the mole-
mation the difluoromethyl group adopts the same orientation cule. In the same way, the CF3OCF2H molecule shows a trans/
than in the global minimum, while it is different in the third T gauche conformation of the F atoms as the most stable struc-
conformer. The tiny energy difference between the two difluor- ture which simultaneously minimizes the F···F repulsion and
omethyl alternatives may justify this observation. optimizes a F···H attractive interaction.
The fact that the global minimum was the weaker species in The synergy between broadband and high-resolution rota-
the microwave spectrum put difficulties to its identification, tional data with the ab initio theoretical calculations has thus
but can be rationalized by the differences in the magnitudes proved very effective to disentangle the molecular properties
of the electric dipole moments (nearly one order of magnitude of enflurane, providing conformational populations and struc-
larger for G + compared to G¢). The situation is commonly ob- tural data which will be valuable for the analysis of the inter-
served in rotational spectra, an example being the difficulty to molecular clusters of this molecule. This analysis completes
initially detect the (low-dipole) global minimum of glycine.[39] a description of the conformational landscapes of the most rel-
The experimental spectrum was reproduced with a semirigid evant volatile anesthetics, permitting also a comparison be-
rotor model, but the sensitivity to the coupling constants tween enflurane and the molecules of sevoflurane, isoflurane,
might suggest some quadrupole mediated perturbations. desflurane and fluoroxene.
The experimental conformational composition obtained
from relative intensity measurements is now estimated as G¢/
G + /T = 1:1:1. This population ratio strikingly compares with the Experimental Section
inversed ratio of G¢/G + /T = 10:32:58 deduced in the electron
diffraction work.[11] Our new estimations are compatible with The rotational spectrum was measured initially with resonator-
based Balle–Flygare Fourier-transform microwave spectrome-
the ab initio predictions at room temperature (G¢/G + /T =
ters[27, 28] (8–26 GHz) and later with a chirped-pulse broadband
10:8:7), not only because of possible uncertainties in the rela- spectrometer[29] (2–18 GHz), fully described before. In these instru-
tive spectral intensities and/or ab initio data, but also on con- ments either a monochromatic or a chirped pulse are used to
sideration of possible conformational population transfer in excite a coherent polarization in a gaseous polar sample. The re-
the jet expansion. A supersonic jet cools down the vibrational sulting transient spontaneous emission is digitized in the time
populations under kinetic collisional control, so the pre-expan- domain and Fourier transformed to obtain the frequency domain
sion equilibrium populations within each vibrational potential spectrum. In the Balle–Flygare spectrometer the use of a microwave
well are only maintained in case of no conformational relaxa- resonator limits the operating bandwidth to about 1 MHz, requir-
ing mechanical retuning for frequency scans. The chirped-pulse
tion.[38]
spectrometer simultaneously excites the full spectral bandwidth. In
The preference for a gauche orientation of the hydrogen all cases the sample was prepared as a molecular jet by co-expan-
atom at the difluoromethyl side is observed in the three de- sion with an inert carrier gas (backing pressures < 3 bar). The race-
tected conformations. However, it might be expected that the mic samples of (RS)-enflurane were obtained commercially and
enatiomeric effects would stabilize a trans H orientation used without further purification. Enflurane is a volatile liquid (b.p.

Chem. Eur. J. 2016, 22, 9804 – 9811 www.chemeurj.org 9809 Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
 Full Paper

[1] R. C. Terrell, L. Speers, A. J. Szur, J. Treadwell, T. R. Ucciardi, J. Med. Chem.

1971, 14, 517.
[2] A. S. Evers, C. M. Crowder, J. R. Balser, in Goodman & Gilman’s The Phar-
macological Basis of Therapeutics, 11th ed. (Eds.: L. L. Brunton, J. S. Lazo,
K. L. Parker), McGraw – Hill, New York, 2006, Chap. 13.
[3] P.-L. Chau, Br. J. Pharmacol. 2010, 161, 288.
[4] R. G. Eckenhoff, J. S. Johansson, Pharm. Rev. 1997, 49, 343.
[5] a) I. Leûn, J. Mill‚n, E. J. Cocinero, A. Lesarri, F. CastaÇo, J. A. Fern‚ndez,
Angew. Chem. Int. Ed. 2013, 52, 7772; Angew. Chem. 2013, 125, 7926;
b) I. Leûn, J. Mill‚n, E. J. Cocinero, A. Lesarri, J. A. Fern‚ndez, Angew.
Chem. Int. Ed. 2014, 53, 12480; Angew. Chem. 2014, 126, 12688; c) I.
Leûn, J. Mill‚n, E. J. Cocinero, A. Lesarri, J. A. Fern‚ndez, Phys. Chem.
Chem. Phys. 2014, 16, 23301.
[6] E. Aguado, I. Leûn, J. Mill‚n, E. J. Cocinero, S. Jaeqx, A. M. Rijs, A. Lesarri,
J. A. Fern‚ndez, J. Phys. Chem. B 2013, 117, 13472.
[7] E. Aguado, I. Leûn, E. J. Cocinero, A. Lesarri, J. A. Fern‚ndez, F. CastaÇo,
Phys. Chem. Chem. Phys. 2009, 11, 11608.
[8] A. Lesarri, A. Vega-Toribio, R. D. Suenram, D. J. Brugh, J.-U. Grabow,
Phys. Chem. Chem. Phys. 2010, 12, 9624.
[9] A. Lesarri, A. Vega-Toribio, R. D. Suenram, D. J. Brugh, D. Nori-Shargh,
J. E. Boggs, J.-U. Grabow, Phys. Chem. Chem. Phys. 2011, 13, 6610.
[10] I. Uriarte, P. Êcija, L. Spada, E. Zabalza, A. Lesarri, F. J. Basterretxea, J. A.
Fern‚ndez, W. Caminati, E. J. Cocinero, Phys. Chem. Chem. Phys. 2016,
18, 3966.
[11] A. Pfeiffer, H.-G. Mack, H. Oberhammer, J. Am. Chem. Soc. 1998, 120,
6384.
[12] P. E. Balonga, V. J. Kowalewski, R. H. Contreras, Spectrochim. Acta Part A
1988, 44, 819.
[13] C. Zhao, P. L. Polavarapu, H. Grosenik, V. Schurig, J. Mol. Struct. 2000,
550, 105.
[14] D. Michalska, D. C. Bienko, B. Czarnik-Matusewicz, M. Wierzejewska, C.
Sandorfy, T. Zeegers-Huyskens, J. Phys. Chem. B 2007, 111, 12228.
[15] S. M. Melikova, K. S. Rutkowski, B. Zarnik-Matusewicz, M. Rospenk,
Chem. Phys. Lett. 2014, 604, 68.
[16] S. M. Melikova, K. S. Rutkowski, E. Telkova, B. Zarnik-Matusewicz, M. Ro-
spenk, W. Herrebout, Chem. Phys. 2015, 26, 453.
[17] S. Hsieh, J. L. Vallejo, R. Vushe, Y. T. Tun, R. Sinha, A. Ahmad, J. Mol. Spec-
trosc. 2014, 303, 20.
Figure 5. Non-covalent-index (NCI) plots (color version in Figure S8): Gradi- [18] D. Sutradhar, T. Zeegers-Huyskens, A. K. Chandra, J. Phys. Chem. A 2013,
ent isosurfaces (s = 0.5 a.u.) for CF3-O-CF2H (lower panel) and the three en- 117, 8545.
flurane conformations (upper panels). The interaction regions are colored ac- [19] W. Zierkiewicz, D. Michalska, T. Zeegers-Huyskens, J. Mol. Model. 2013,
cording to the eigenvalues of the electron density Hessian (sign l21 from 19, 1399.
¢0.05 to 0.05 a.u.). The grey surfaces represent non-bonded interactions [20] L. A. F. Andrade, J. M. Silla, S. L. Stephens, K. Marat, E. F. F. da Cunha, T. C.
characterized as delocalized weak attractive interactions. Ramalho, J. van Wijngaarden, M. P. Freitas, J. Phys. Chem. A 2015, 119,
10735.
[21] S. T. Shipman, B. H. Pate, New Techniques in Microwave Spectroscopy, in
56.5 8C) and was diluted in neon gas mixtures below 1 %. Molecular Handbook of High Resolution Spectroscopy (Ed.: M. Quack, F. Merkt),
mechanics and conformational search procedures used MacroMo- Wiley, New York, 2011, pp. 801 – 828.
del.[42] Ab initio and DFT calculations were carried out with the [22] J.-U. Grabow, Fourier Transform Microwave Spectroscopy : Measurement
Gaussian 09.[43] The AIM and NCI analysis was carried out with the and Instrumentation, in Handbook of High Resolution Spectroscopy (Ed.:
AIMAll[44] and NCIPlot[45] programs, respectively. M. Quack, F. Merkt), Wiley, New York, 2011, pp. 723 – 800.
[23] G. M. Keser˜, I. Kolossv‚ry, J. Am. Chem. Soc. 2001, 123, 12708.
[24] R. F. W. Bader, Atoms in Molecules A Quantum Theory (Int. Ser. Mono-
Acknowledgements graphs Chemistry), Clarendom, 1994.
[25] E. R. Johnson, S. Keinan, P. Mori-S‚nchez, J. Contreras-Garc†a, A. J.
Cohen, W. Yang, J. Am. Chem. Soc. 2010, 132, 6498.
We thank the US National Science Foundation (NSF CHE- [26] A. E. Reed, L. A. Curtiss, F. Weinhold, Chem. Rev. 1988, 88, 899.
00960074), the Spanish MINECO (CTQ2014-54464-R, CTQ2015- [27] R. D. Suenram, J.-U. Grabow, A. Zuban, I. Leonov, Rev. Sci. Instrum. 1999,
68148-C2–2-P and CTQ2015-63997-C2-2-P), Comunidad Autûn- 70, 2127.
[28] a) J.-U. Grabow, W. Stahl, Z. Naturforsch. A 1990, 45, 1043; b) J.-U.
oma de Madrid (S2013/MIT-2841, Fotocarbon), the Deutsche
Grabow, W. Sathl, H. Dreizler, Rev. Sci. Instrum. 1996, 67, 4072.
Forschungsgemeinschaft (DFG) and the Land Niedersachsen [29] C. P¦rez, S. Lobsiger, N. A. Seifert, D. P. Zaleski, B. Temelso, G. C. Shields,
for research funds. R.D.S. thanks Jerry Fraser at NIST for experi- Z. Kisiel, B. H. Pate, Chem. Phys. Lett. 2013, 571, 1.
mental support. E.J.C. thanks the MINECO for a “Ramûn y [30] J. K. G. Watson, Aspects of Quartic and Sextic Centrifugal Effects on Rota-
Cajal” contract. tional Energy Levels, in Vibrational Spectra and Structure Vol. 6, Elsevier,
Amsterdam 1977, Chap. 1.
[31] W. Gordy, R. L. Cook, Microwave Molecular Spectra, Wiley, New York
Keywords: ab initio calculations · halogenated ethers · 1984.
[32] H. M. Pickett, J. Mol. Spectrosc. 1991, 148, 371.
rotational spectroscopy · supersonic jets · volatile anesthetics
[33] P. Pyykkç, Mol. Phys. 2008, 106, 1965.

Chem. Eur. J. 2016, 22, 9804 – 9811 www.chemeurj.org 9810 Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
 Full Paper

[34] P. Êcija, E. J. Cocinero, A. Lesarri, J. Mill‚n, F. Basterretxea, J. A. Fern‚n- A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara,
dez, F. CastaÇo, J. Chem. Phys. 2011, 134, 164311. K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.
[35] G. T. Fraser, R. D. Suenran, C. L. Lugez, J. Phys. Chem. A 2000, 104, 1141. Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro,
[36] N. A. Seifert, C. P¦rez, J. L. Neill, B. H. Pate, M. Vallejo-Lûpez, A. Lesarri, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Ko-
E. J. Cocinero, F. CastaÇo, Phys. Chem. Chem. Phys. 2015, 17, 18282. bayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyen-
[37] Equilibrium Molecular Structures: From Spectroscopy to Quantum Chemis- gar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B.
try (Eds.: J. Demaison, J. E. Boggs, A. G. Csaszar), CRC, Boca Raton, 2010. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
[38] a) P. D. Godfrey, R. D. Brown, J. Am. Chem. Soc. 1998, 120, 10724; b) R. S. O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,
Ruoff, T. D. Klots, T. Emilsson, H. S. Gutowsky, J. Chem. Phys. 1990, 93, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg,
3142. S. Dapprich, A. D. Daniels, ©. Farkas, J. B. Foresman, J. V. Ortiz, J. Cio-
[39] R. D. Suenram, F. J. Lovas, J. Am. Chem. Soc. 1980, 102, 7180. slowski, D. J. Fox, Gaussian, Inc., Wallingford CT, 2009.
[40] I. V. Alabugin, T. A. Zeidan, J. Am. Chem. Soc. 2002, 124, 3175. [44] AIMAll (Version 16.01.09), Todd A. Keith, TK Gristmill Software, Overland
[41] P. Cam Nam, M.-T. Nguyen, T. Zeegers-Huysken, J. Mol. Struct. 2007, 821, Park KS, USA, 2016 (aim.tkgristmill.com).
71. [45] J. Contreras-Garcia, E. R. Johnson, S. Keinan, R. Chaudret, J.-P. Piquemal,
[42] F. Mohamadi, N. G. J. Richard, W. C. Guida, R. Liskamp, M. Lipton, C. Cau- D. N. Beratan, W. Yang, J. Chem. Theory Comput. 2011, 7, 625 – 632.
field, G. Chang, T. Hendrickson, W. C. Still, J. Comput. Chem. 1990, 11,
440.
[43] Gaussian 09, Revision E.01, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Men- Received: March 14, 2016
nucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, Published online on June 3, 2016

Chem. Eur. J. 2016, 22, 9804 – 9811 www.chemeurj.org 9811 Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim