Hydrogen bonded arrays: The power of multiple hydrogen bonds

Article abstract of DOI:10.1021/ja2081907

Hydrogen bond interactions in small covalent model compounds (i.e., deprotonated polyhydroxy alcohols) were measured by negative ion photoelectron spectroscopy. The experimentally determined vertical and adiabatic electron detachment energies for (HOCH

Full text of DOI:10.1021/ja2081907

Article
pubs.acs.org/JACS
Hydrogen Bonded Arrays: The Power of Multiple Hydrogen Bonds
Alireza Shokri,† Jacob Schmidt,† Xue-Bin Wang,*,‡ and Steven R. Kass*,†
†Department of Chemistry, University of Minnesota, Minneapolis, Minnesota 55455, United States
‡Chemical & Materials Sciences Division, Pacific Northwest National Laboratory, P.O. Box 999, MS K8-88 Richland,
Washington 99352, United States, and Department of Physics, Washington State University, 2710 University Drive, Richland,
Washington 99354, United States
S
* Supporting Information
ABSTRACT: Hydrogen bond interactions in small covalent model com-
pounds (i.e., deprotonated polyhydroxy alcohols) were measured by nega-
tive ion photoelectron spectroscopy. The experimentally determined verti-
cal and adiabatic electron detachment energies for (HOCH2CH2)2CHO−
(2a), (HOCH2CH2)3CO− (3a), and (HOCH2CH2CH(OH)CH2)3CO− (4a)
reveal that hydrogen-bonded networks can provide enormous stabili-
zations and that a single charge center not only can be stabilized by up to three hydrogen bonds but also can increase the inter-
action energy between noncharged OH groups by 5.8 kcal mol−1 or more per hydrogen bond. This can lead to pKa values that are
very different from those in water and can provide some of the impetus for catalytic processes.
bond arrays systematically increase with the number of hydro-
xyl groups from 1 to 3 and for 6. That is, an oxygen anion
center is stabilized most effectively by up to 3 hydrogen bond
donors, but in the process the donor groups become better
hydrogen bond acceptors. The resulting hydrogen-bonded net-
work can provide a large energetic stabilization that may lead to
catalytic rate enhancements and greater acidities and basicities
than those measured in water.
INTRODUCTION
■
Many enzymes catalyze a wide variety of chemical processes by
using two or even three hydrogen bonds to a single oxygen
atom in what is commonly referred to as an oxyanion hole
(Figure 1).1 The strength of the individual hydrogen bonds can
EXPERIMENTAL SECTION
■
1
General. H and 13C NMR spectra were recorded on Varian VI-
300, VXR-300, and VI-500 spectrometers and are reported in parts per
million (δ). High-resolution mass spectral analyses were performed
with a Bruker BioTof II electrospray ionization−time-of-flight mass
spectrometer with an internal standard; a methanol solution of PEG
polymer was used for this purpose. Medium pressure liquid
chromatography (25−60 psi) was carried out on silica gel using a
Biotage Isolera 1. Thin-layer chromatography (TLC) was performed
on 0.25 mm Masherey-Nagel silica gel plates, and compounds were
visualized using aqueous potassium permanganate or p-anisaldehyde
(1% EtOH solution) stains.
2-Benzyl-1,3-diphenylpropan-2-ol. A dry three-necked round-
bottomed flask was equipped with a magnetic stirring bar and an
addition funnel, a reflux condenser with an argon inlet, and a rubber
septum. The flask was charged with 6.90 g (0.283 mol) of magnesium
turnings, and 200 mL of anhydrous diethyl ether. A solution of 17.56 g
(0.156 mol) of benzyl chloride (Alfa Aesar) in 20 mL of diethyl ether
was added dropwise over 45 min so as to maintain a gentle reflux.
Subsequently, 4.50 g (0.050 mol) of dimethylcarbonate was added
dropwise over a 20 min period and the resulting reaction mixture was
stirred for an additional 20 h at room temperature before being
quenched with 100 mL of 10% aqueous HCl. The resulting two layers
were separated, and the aqueous solution was extracted three times
with 100 mL of diethyl ether. The combined organic material was
Figure 1. Proposed transition state for the enzymatic hydrolysis of
acetylcholine with acetylcholinesterase via a three-pronged oxyanion
hole (i.e., 1−3); see ref 1d for additional details.
vary considerably, and the presence of a strong low-barrier or
single-well hydrogen bond has been used to explain a variety of
enzyme mechanism pathways (e.g., mandelate racemase, triose-
phosphate isomerase, citrate synthase, and glycolate oxidase).2
Additive effects involving two and three hydrogen bonds to one
charged center also have been shown in several enzyme model
systems.3,4 However, networks of hydrogen bonds are em-
ployed by enzymes, and the energies of these additional inter-
actions are important in catalysis but are not well under-
stood.5 In this report, negative-ion photoelectron spectroscopy
is used to directly probe the energetic consequences of hydro-
gen bond arrays in small covalently bound model compounds
and the experimental results are compared to computational
predictions. We find that in a series of monodeprotonated poly-
hydroxy alcohols (i.e., polyols) the strengths of the hydrogen
Received: September 2, 2011
Published: December 22, 2011
© 2011 American Chemical Society
2094
dx.doi.org/10.1021/ja2081907 | J. Am. Chem.Soc. 2012, 134, 2094−2099

Journal of the American Chemical Society
Article
dried with anhydrous MgSO4, filtered, and concentrated under
aspirator pressure. Recyrstallization of the resulting solid from absolute
ethanol afforded 11.92 g (79%) of the tribenzyl alcohol as a white
solid. 1H NMR (300 MHz, CDCl3) δ 1.51 (1H, s), 2.79 (6H, s), 7.28
(15H, m). 13C NMR (75 MHz, CDCl3) δ 45.7, 73.9, 126.4, 128.1,
131.0, 137.3. HRMS-ESI: calc for C22H22NaO+ (M + Na)+ 325.1563,
background subtraction to be carried out. Photoelectrons were
collected at nearly 100% efficiency with the magnetic bottle and
analyzed in a 5.2 m long electron flight tube. Time-of-flight
photoelectron spectra were collected and converted to kinetic energy
spectra, calibrated by the known spectra of I− and ClO2−. The electron
binding energy spectra were obtained by subtracting the kinetic energy
spectra from the detachment photon energies. The resolution (ΔE/E)
of the resulting spectra was approximately 2% or 20 meV at 1 eV, as
measured for I− at 355 nm. Photoelectron spectra for deprotonated
triol 2 and tetraol 3 with the higher energy light source are given in
Figure S1 in the Supporting Information.
Computational Methods. Extensive geometry optimizations were
previously carried out on all of the anions studied in this work using
the Becke three-parameter hybrid exchange and Lee−Yang−Parr
correlation density functional (B3LYP)7 along with the Dunning
augmented correlation-consistent double-ζ basis set (aug-cc-pVDZ).8
In this investigation, the lowest energy structure for each ion was
reoptimized using the newer Minnesota 2006 density functional
(M06-2X)9 and served as the starting point for optimizing the
structure of the corresponding radical with both functionals. The aug-
cc-pVDZ basis set was used as before, but additional M06-2X single-
point energies with the maug-cc-pVT(+d)Z basis set (i.e., M06-2X/
maug-cc-pVT(+d)Z//M06-2X/aug-cc-pVDZ) were also computed.10
Vibrational frequencies were calculated in both cases, and the unscaled
values provided zero-point energies and thermal corrections to the
enthalpies. Vertical detachment energies (VDEs) were calculated by
taking the electronic energy differences between the different anions
and their corresponding radicals with the same geometries as the ions.
Adiabatic detachment energies (ADEs) are reported as enthalpies at
0 K and were obtained using the optimized structures for both the
anions and the radicals. All of the other energies reported in this work
are given as enthalpies at 298 K. Gaussian 0911 was used to carry out
all of the calculations reported in this study, and the resulting geo-
metries and energies are provided in the Supporting Information.
found 325.1581.
1,3-Di(cyclohexa-1,4-dienyl)-2-(cyclohexa-1,4-dienylmethyl)-
propan-2-ol. Tribenzylmethanol (2.00 g, 6.61 mmol) was placed in a
500 mL flask that was equipped with a Claisen adapter and a dry ice−
acetone condenser. The reaction flask was charged with 200 mL of
liquid ammonia followed by 40 g (0.54 mol) of tert-butanol, and then
0.80 g (0.11 mol) of small pieces of lithium wire (Sigma Aldrich) were
added over 4 h with stirring. By allowing the reaction mixture to warm
to room temperature overnight, the ammonia was evaporated to afford
a white solid. It was dissolved in 100 mL of water and extracted three
times with 100 mL of diethyl ether. The combined organic layers were
dried over MgSO4 and concentrated under reduced pressure, and the
crude product was purified by flash column chromatography (20:1
hexanes/diethyl ether) to give 1.54 g (76%) of the desired alcohol as
a colorless oil. 1H NMR (300 MHz, CDCl3) δ 1.76 (1H, s), 2.14 (6H,
br s), 2.75 (12H, br s), 5.49 (3H, br s), 5.69 (6H, br s). 13C NMR
(75 MHz, CDCl3) δ 26.7, 31.1, 48.0, 74.2, 123.1, 123.4, 124.5, 132.2.
HRMS-ESI: calc for C22H28NaO+ (M + Na)+ 331.2032, found
331.2029.
5-(2,4-Dihydroxybutyl)nonane-1,3,5,7,9-pentaol (4). A solution of
1.23 g (4.00 mmol) of 1,3-di(cyclohexa-1,4-dienyl)-2-(cyclohexa-1,4-
dienylmethyl)propan-2-ol in 100 mL of a 1:1 mixture of methanol/
dichloromethane was cooled to −78 °C, and then ozone was bubbled
through it for 1 h until a blue color persisted. Excess ozone was
removed from the reaction mixture by passing oxygen through it, and
then 7.40 g (195 mmol) of sodium borohydride was slowly added with
stirring over the course of 1 h, all the while maintaining the temper-
ature at −78 °C. The resulting solution was allowed to warm to room
temperature and was maintained at 20 °C for 14 h. Acidification of the
reaction mixture was carried out by slowly adding concentrated HCl
until the pH was lowered to ∼4 and a white precipitate formed. The
resulting methanolic solution was filtered and concentrated at aspirator
pressure with a rotary evaporator to afford a residue which was filtered,
redissolved in 100 mL of methanol and concentrated again. This
step was repeated several times to facilitate the removal of boron-
containing compounds as volatile byproducts. The remaining material
was purified on a silica gel column by flash chromatography (4:1 ethyl
acetate/methanol) to afford 0.29 g (25%) of 5-(2,4-dihydroxybutyl)-
nonane-1,3,5,7,9-pentaol as an ∼3: 1 mixture of the anti (RRS/SSR)
RESULTS AND DISCUSSION
■
For an enzyme to catalyze a reaction, it must increase the rate
relative to the uncatalyzed background process by reducing the
energy difference between the reactant and the transition state.
This can be accomplished by destabilizing the energy of the
substrate in the enzyme bound complex or by lowering the
transition state energy. In a similar manner, the electron bind-
ing energy of an anion can be increased by one or more hydro-
gen bonds if the stabilizing interaction(s) is greater in the ion
than in the corresponding radical. To directly probe the effects
of hydrogen bonds on a single-charged center, HOCH2CH2-
CH2OH (1), (HOCH2CH2)2CHOH (2), and (HOCH2CH2)3-
COH (3) were electrosprayed into a home-built variable tem-
perature photoelectron spectrometer which has been described
previously.6 The (M − 1)− ions of 2 and 3 (2a and 3a) were
produced, but no signal was observed for the diol because it is
not much more acidic than methanol.12 In the triol and tetraol,
there are two hydroxyl groups that could be ionized, but depro-
tonation of the internal one (i.e., the secondary and tertiary
alcohol sites in 2 and 3, respectively) leads to a more favorable
hydrogen bonding network (Figure 2). For example, the pri-
mary alkoxide derived from the tetraol can only form two
hydrogen bonds to the charged site whereas the tertiary alk-
oxide can form three. This leads to a computed 6.0 kcal mol−1
energy difference (B3LYP/aug-cc-pVDZ) between the two iso-
mers, and since they should be able to readily interconvert, it is
reasonable to conclude that the ion composition is largely, if
not entirely, made up of the more stable anion.
1
and syn (RRR/SSS) isomers. H NMR (300 MHz, CD3OD) δ 1.80
(12H, m), 3.68 (6H, t, J = 6.0 Hz), 4.10 (3H, m). 13C NMR (75 MHz,
CD3OD) δ 42.5 (anti), 42.6 (anti and syn), 42.7 (anti), 46.3 (anti),
46.6 (syn), 47.2 (anti), 47.9 (anti), 60.2 (syn), 60.3 (anti), 60.4 (anti),
60.5 (anti), 67.3 (syn), 67.57 (anti), 67.60 (anti), 77.0 (anti), 77.4
(syn); the syn isomer was isolated from its diastereomer by flash
chromatography using a 4:1 dichloromethane/methanol solvent mix-
ture after the PES experiments were carried out. HRMS-ESI: calc for
+
C13H28NaO7 (M + Na)+ 319.1727, found 319.1732.
Photoelectron Spectroscopy. Photoelectron spectra were
obtained at 20 K with a home-built variable temperature photoelectron
spectrometer coupled with an electrospray ion source and a cryogenic
ion-trap which has been previously described.6 In this work, the alco-
hols were dissolved in a methanol/water solution (7:3 v/v) and
electrospray ionization of ∼10−3 M solutions afforded the (M − 1)−
ions, except for 1,3-propanediol. The ions were trapped and cooled for
a period of 20−80 ms in the trap before being pulsed out into the
extraction zone of a time-of-flight mass spectrometer with a repetition
rate of 10 Hz. The desired (M − 1)− ions were mass-selected and
decelerated before being intercepted by a probe laser beam in the
photodetachment zone of a magnetic bottle photoelectron analyzer.
An excimer laser (193 nm, 6.424 eV) and a Nd:YAG laser (266 nm,
4.661 eV) were used in this study, and all three ions were examined at
both wavelengths. The laser was operated at a 20 Hz repetition rate
with the ion beam off on alternating laser shots to enable shot-by-shot
Photoelectron spectra were recorded for 2a and 3a at
193 nm (6.424 eV) and 266 nm (4.661 eV) at 20 K, and the
latter data are given in Figure 3; the 193 nm spectra have an
2095
dx.doi.org/10.1021/ja2081907 | J. Am. Chem.Soc. 2012, 134, 2094−2099

Journal of the American Chemical Society
Article
set (i.e., B3LYP/aug-cc-pVDZ).7,8 The newer and appa-
rently superior Minnesota 06 functional of Truhlar et al. was
also used,9 and since the energetics obtained using this method
are more sensitive to the basis set, single-point energies were
computed using the recently reported minimally aug-
mented triple-ζ+d basis set (i.e., maug-cc-pVT(+d)Z)10 on
the aug-cc-pVDZ geometries (i.e., M06-2X/maug-cc-pVT(+d)-
Z//M06-2X/aug-cc-pVDZ). In general, both computational
methods do well in reproducing the experimental data, but
the M06-2X approach with the maug-cc-pVT(+d)Z basis set
gives the best results. These values, however, are consistently
too small by 0.16 eV (i.e., the errors span from 0.12 to 0.19 eV). As
a result, this leads to an estimate of 2.63 eV for the ADE of
HOCH2CH2CH2O−.15
The differences between the ADEs of the conjugate bases of
1-propanol and the polyols provide a direct measure of the
consequences of multiple hydrogen bonds to a single charged
center. This is not to say that these energy differences are the
hydrogen bond strengths, since there is no way to uniquely
partition the various contributions that lead to the overall stabi-
lity of the ion. In any case, total stabilizations as given by ADE
((HOCH2CH2)nCH3−nO− − CH3CH2CH2O−) for n = 1−3
lead to 19.4, 34.8, and 47.5 kcal mol−1, respectively, and stabili-
zation energies of 19.4, 17.4, and 15.8 kcal mol−1 per hydrogen
bond; the latter numbers were obtained from the total stabili-
zations by dividing them by n, the number of hydrogen bonds
in the ion. These findings are in accord with earlier results by
Herschlag et al. on ortho-substituted benzoic acids obtained by
measuring their pKa’s in dimethysulfoxide (DMSO),3 our
earlier work on the acidities of 1−3 in DMSO and the gas
phase,4 and a variety of reports describing the stabilization of
polyfunctional ions via hydrogen bonds.16 Our data herein on
simple covalently bound model systems for oxyanion holes with
up to three hydrogen bonds reveal that up to 47.5 kcal mol−1 is
available for catalysis, to the extent that Nature can mimic the
gas phase results. The energy gain is 44% of this value, however,
if DMSO is used as the reference medium.17
Figure 2. Most favorable hydrogen bonding patterns found for
HOCH2CH2O− (1a), (HOCH2CH2)2CHO− (2a), (HOCH2CH2)3CO−
(3a), and (HOCH2CH2CH(OH)CH2)3CO− (4a). Bond lengths for
the anion and radical (in italics) are in angstroms and are from the
M06-2X/aug-cc-pVDZ geometries; the values for 3r, 4a, and 4r are
averages of the three different bond distances.
Figure 3. Low temperature (20 K) photoelectron spectra of
(HOCH2CH2)2CHO− (2a, top) and (HOCH2CH2)3CO − (3a, bottom)
at 266 nm (4.661 eV).
To this point, only hydrogen bonds to a charged center are
present, but additional interactions between noncharged groups
may also contribute to an enzyme’s catalytic ability. This can be
energetically significant, particularly since hydrogen bond
donors to a charged group also will be better hydrogen bond
acceptors than they would be in the absence of the charge. A
more effective hydrogen bond array, consequently, could result.
To assess the energetics that can be brought to bear in such a
network, the conjugate base of an alcohol with seven OH
groups (i.e., (HOCH2CH2CH(OH)CH2)3COH, 4) was
examined. This ion (4a) can be stabilized by a total of six
hydrogen bonds, and the most favorable structure that was
additional band at higher energy corresponding to the forma-
tion of the radical in an excited state and are provided in the
Supporting Information. The top of the bands in the 266 nm
spectra provide the vertical detachment energies (VDEs) of the
anions, and a linear extrapolation of the fast rising onset region
leads to what should be a good estimate of the adiabatic detach-
ment energies (ADEs).13 The results are presented in Table 1
along with the literature value for 1-propoxide (CH3CH2-
CH2O−)14 and B3LYP predictions with the aug-cc-pVDZ basis
Table 1. Experimental and Computed Adiabatic (ADE) and Vertical (VDE) Electron Binding Energies for a Series of Alkoxides
in eV
a
exptl
calcd
−
cmpd (RO )
ADE
VDE
ADE
VDE
1.81
−
CH3CH2CH2O
1.789 0.033
1.66 (1.57) [1.63]
2.41 (2.40) [2.47]
3.07 (3.09) [3.18]
3.53 (3.58) [3.66]
4.28 (4.36) [4.45]
−
HOCH2CH2CH2O (1a)
2.69 (2.87)
3.53 (3.70)
4.00 (4.18)
4.76 (5.05)
−
(HOCH2CH2)2CHO (2a)
3.30
3.85
4.60
3.63
4.18
5.06
−
(HOCH2CH2)3CO (3a)
(HOCH2CH2CH(OH)CH2)3CO− (4a)
a
Computed values are at 0 K and correspond to B3LYP/aug-cc-pVDZ, M06-2X/aug-cc-pVDZ (in parentheses), and M06-2X/maug-cc-
pVT(+d)Z//M06-2X/aug-cc-pVDZ (in brackets) energies; the difference between computed values at 0 and 20 K is negligible (i.e., <0.01 eV).
2096
dx.doi.org/10.1021/ja2081907 | J. Am. Chem.Soc. 2012, 134, 2094−2099

Journal of the American Chemical Society
Article
computationally located has three hydrogen bonds to the
charged center and three additional ones between noncharged
OH groups (Figure 2). In this species the internal secondary
hydroxyl groups serve as both hydrogen bond donors and
acceptors. Conformers with four OH groups interacting with
the oxyanion center were located, but the most favorable one is
7.0 (M06-2X/maug-cc-pVT(+d)Z//M06-2X/aug-cc-pVDZ)
kcal mol−1 less stable than the structure shown in Figure 2
(and Figure 4). This is consistent with a hydration shell
number of 3 for hydroxide, as determined by Meot-Ner and
Speller.18
noncharged OH groups. These intramolecular interactions and
those in the conjugate bases of 1−3 can be compared to the
intermolecular hydrogen-bonded networks in OH−(H2O)n,
where n = 1−6 (Figure 5 and Table 2).18,24,25 As can be seen
Heptaol 4 is an unknown compound, but it can be prepared
as a mixture of two diastereomers as outlined in Scheme 1. The
Scheme 1. Synthetic Route for the Preparation of Heptaol 4
Figure 5. Dissociation energies versus hydrogen bond numbers.
Diamonds are for polyol (M − 1)− ion ΔADEs relative to CH3CH2-
CH2O− (blue), circles represent the total dehydration energies of
OH−(H2O)n (black), and squares are for the CH3O−(CH3OH)n
cluster energies (red).
Table 2. Energetic Consequences of Intramolecular vs
a
Intermolecular Hydrogen Bonds
resulting product affords an abundant (M − 1)− ion signal via
electrospray ionization, but irradiation at 266 nm does not lead
to electron detachment. A low temperature spectrum at 20 K
was recorded (Figure 4), however, when more energetic
−
ΔH°(RO (ROH)n →
−
RO + nROH)/n,
intramol/
ΔADE/n
R = H and CH3
intermol
b
c
n
(intramolecular)
(intermolecular)
(%)
−
1
2
3
4
5
6
(1a − CH3(CH2)2O )/n = 19.4
26.5 [28.8]
22.1 [25.1]
20.1 [21.7]
12.0 [11.4]
11.8
73
79
79
(2a − 1a)/n = 17.4
(3a − 1a)/n = 15.8
(4a − 3a)/(n − 3) = 5.8
11.6
50
a
b
All of the energies are in kcal mol−1. These data come from refs 18
and 24, and the first value is for R = H and the latter one (in brackets)
is for R = CH3. When n ≥ 4, then the given energies are for
ΔH°(RO−(ROH)n → RO−(ROH)3 + mROH)/m, where m = n − 3.
c
The water cluster intermolecular values were used for this comparison.
for n = 1−3 in Figure 5, the intramolecular hydrogen bond
system mirrors the intermolecular interaction energies closely
but is ∼20−25% smaller. This is presumably because of
geometric constraints in the polyols that restrict the O−H···O
hydrogen bond angles from the ideal value of 180° to 153.0−
158.5° (B3LYP) or 151.0−160.4° (M06-2X) when n = 1−3.
The energy gap in Figure 5 is larger for n = 6, and it is tempting
to attribute this to water−water interactions in OH−(H2O)6,
especially since computations indicate that the water molecules
hydrogen bond with each other when n ≥ 3.26 Methoxide−
methanol cluster energies (the squares in Figure 5),24 however,
follow the same trend as the hydroxide−water complexes even
though the hydrogen-bonded network is more limited in the
former case, since alcohols have one less O−H bond than in
water. This suggests that the fall off for 4a is not due to water−
water interactions in OH−(H2O)6 but is instead the result of its
more limited conformational flexibility. In accord with this
hypothesis, the computed O−H···O hydrogen bond angles
between the terminal hydroxyl groups and the oxygen atoms of
the internal hydroxyl groups are 144−145° (B3LYP) and 142−
143° (M06-2X) or ∼10−15° smaller than the O−H···O− angles
in 1a−4a. This leads to weaker hydrogen bonds in the outer
Figure 4. Low temperature (20 K) photoelectron spectrum of
(HOCH2CH2CH(OH)CH2)3CO− (4a) at 193 nm (6.424 eV).
193 nm photons were used. As with the smaller polyol anions
at this wavelength, two features are apparent in the spectrum,
and the lower energy band provides the adiabatic and vertical
electron detachment energies. These values are 4.60 (ADE)
and 5.06 eV (VDE), which are remarkably large for a saturated
alkoxide anion that lacks electron withdrawing groups and is
not stabilized by resonance. To put these record breaking
electron detachment energies in context, it is worth noting that
the conjugate bases of strong acids such as CH3CO2H, HCl,
HNO3, and HClO3 all have smaller ADEs (i.e., 3.470 0.010,
3.613577
0.000044, 3.937
0.014, and 4.25
0.10 eV,
respectively)19−22 than the deprotonated heptaol, whereas
dihydrogen phosphate (H2PO4−) has essentially the same value
(i.e., 4.570 0.010 eV).23
The hydrogen bond array in the conjugate base of 4 leads to
a 64.8 kcal mol−1 stabilization relative to CH3CH2CH2O−, and
17.3 kcal mol−1 compared to the (M − 1)− ion of tetraol 3. The
latter difference can be attributed to the additional contribu-
tion from the three enhanced hydrogen bonds between the
2097
dx.doi.org/10.1021/ja2081907 | J. Am. Chem.Soc. 2012, 134, 2094−2099

Journal of the American Chemical Society
Article
shell of 4a that presumably can be strengthened by increasing
the flexibility in 4 (e.g., by adding a CH2 spacer between the
primary and secondary OH groups).
Environmental Research and located at Pacific Northwest
National Laboratory, which is operated by Battelle for the
DOE. We also thank Prof. Brian Miller for helpful suggestions.
Enzyme−substrate complexes enjoy a kinetic advantage over
bimolecular processes because intramolecular reactions are gene-
rally favored over intermolecular ones.27 Our simple covalently
bound models for the former species reveal that hydrogen-
bonded networks can be stabilized by the presence of a charged
site, and this thermodynamic effect can be used to perturb acidi-
ties and basicities and catalyze enzyme reactions.28 As a result,
aqueous pKa values may be a poor indication of the acidity or
basicity of a given group in a biological context, and proton
transfer processes that are currently viewed as being energetically
unfavorable and inaccessible actually may take place. Careful
control of hydrogen-bonded networks consequently is an attrac-
tive design strategy for molecular recognition and artificial
enzyme construction.29
REFERENCES
■
(1) (a) Childs, W.; Boxer, S. G. Biochemistry 2010, 49, 2725−2731.
(b) Simon, L.; Goodman, J. M. J. Org. Chem. 2010, 75, 1831−1840.
(c) Sigala, P. A.; Kraut, D. A.; Caaveiro, J. M. M.; Pybus, B.; Rubin, E.
A.; Ringe, D.; Petsko, G. A.; Herschlag, D. J. Am. Chem. Soc. 2008, 130,
13696−13708. (d) Zhang, Y.; Kua, J.; McCammon, J. A. J. Am. Chem.
Soc. 2002, 124, 10572−10577. (e) Whiting, A. K.; Peticolas, W. L.
Biochemistry 1994, 33, 552−561.
(2) (a) Gerlt, J. A.; Gassman, P. G. Biochemistry 1993, 32, 11943−
11952. (b) Gerlt, J. A.; Gassman, P. G. J. Am. Chem. Soc. 1993, 115,
11552−11568. (c) Cleland, W. W.; Kreevoy, M. M. Science 1994, 264,
1887−1890. (d) Frey, P. A.; Whitt, S. A.; Tobin, J. B. Science 1994,
264, 1927−1930. (e) Cleland, W. W.; Frey, P. A.; Gerlt, J. A. J. Biol.
Chem. 1998, 273, 25529−25532. (f) Schwans, J. P.; Kraut, D. A.;
Herschlag, D. Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 14271−14275.
(3) (a) Shan, S. O.; Herschlag, D. Proc. Natl. Acad. Sci. U. S. A. 1996,
93, 14474−14479. (b) Shan, S. O.; Herschlag, D. J. Am. Chem. Soc.
1996, 118, 5515−5518. (c) Shan, S. O.; Loh, S.; Herschlag, D. Science
1996, 272, 97−101. (d) Shan, S. O.; Herschlag, D. Methods Enzymol.
1999, 308, 246−276.
CONCLUSIONS
■
The photoelectron spectra of a series of deprotonated poly-
hydroxyalcohols were obtained at 20 K, and the adiabatic
electron detachment energies for HOCH2CH2CH2O− (1a),
(HOCH2CH2)2CHO− (2a), (HOCH2CH2)3CO− (3a), and
(HOCH2CH2CH(OH)CH2)3CO− (4a) are 2.63 (best esti-
mate, see text for details), 3.30, 3.85, and 4.60 eV, respectively.
These values are remarkably large and are bigger than the
experimental ADE for CH3CH2CH2O− (1.789 0.033 eV) by
19.4, 34.8, 47.5, and 64.8 kcal mol−1. These energy differences
are a consequence of the hydrogen-bonded networks in the
anions, and in the case of 4a this leads to an ADE that is greater
than those of the conjugate bases of CH3CO2H, HCl, and
HNO3. Its ADE is also 17.3 kcal mol−1 or 5.8 kcal mol−1 per
hydrogen bond greater than that for 3a, and this difference can
be attributed to the enhanced strength of the three hydrogen
bonds between the noncharged OH groups in 4a. The presence
of a charged center leads to a considerable increase in the
strength of a hydrogen-bonded network, and this undoubtedly
plays a key role in regulating the structure and function of a
wide range of biomolecules.
(4) Tian, Z.; Fattahi, A.; Lis, L.; Kass, S. R. J. Am. Chem. Soc. 2009,
131, 16984−16988.
(5) (a) Jencks, W. P. Adv. Enzymol. Relat. Areas Mol. Biol. 1975, 43,
219−410. (b) Jencks, W. P. Methods Enzymol. 1989, 171, 145−164.
(6) Wang, X. B.; Wang, L. S. Rev. Sci. Instrum. 2008, 79, 073108.
(7) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648−5652. (b) Lee, C.;
Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785−789.
(8) Dunning, T. H. Jr. J. Chem. Phys. 1989, 90, 1007−1023.
(9) (a) Zhao, Y.; Truhlar, D. G. J. Phys. Chem. A 2008, 112, 1095−
1099. (b) Zhao, Y.; Truhlar, D. G. Theor. Chem. Acc. 2008, 120, 215−
241. (c) Zhao, Y.; Truhlar, D. G. Acc. Chem. Res. 2008, 41, 157−167.
(10) Papajak, E.; Truhlar, D. G. J. Chem. Theory Comput. 2010, 6,
597−601.
(11) Frisch, M. J., et al.. Gaussian 09; Gaussian, Inc.: Pittsburgh, PA,
2009.
(12) Woolley, E. M.; Tomkins, J.; Hepler, L. G. J. Solution Chem.
1972, 1, 341−351.
(13) Vibrational resolution is required to obtain a precise value for
the electron affinity, but given that the anion and radical geometries
are predicted to be in similar bonding configurations, we anticipate
that the adiabatic electron affinities reported herein are accurate to
within 0.10 eV. Our M06-2X computations on 1a−4a indicate,
however, that the OH···O− distances increase upon formation of their
radicals, as one would expect.
ASSOCIATED CONTENT
* Supporting Information
Computed geometries and energies, along with the 20 K
photoelectron spectra of 2a and 3a at 193 nm (6.424 eV) and
the complete citation to ref 11. This material is available free of
charge via the Internet at http://pubs.acs.org.
■
S
(14) Ellison, G. B.; Engleking, P. C.; Lineberger, W. C. J. Phys. Chem.
1982, 86, 4873−4878.
(15) Both the B3LYP and M06-2X ADEs with the aug-cc-pVDZ basis
set are also underestimated, but in these cases, the average error is
0.23 eV. This leads to corrected predictions of 2.64 (B3LYP) and 2.63 eV
(M06-2X) for HOCH2CH2CH2O−, which are the same as the M06-2X/
maug-cc-pVT(+d)Z value given in the text (to within 0.01 eV).
(16) For examples, see: (a) Tian, Z.; Fattahi, A.; Lis, L.; Kass, S. R.
Croat. Chem. Acta 2009, 82, 41−45. (b) Meot-Ner (Mautner), M.
Chem. Rev. 2005, 105, 213−284.
AUTHOR INFORMATION
Corresponding Author
kass@umn.edu; xuebin.wang@pnnl.gov
■
ACKNOWLEDGMENTS
■
This work is dedicated to the memory of Donna Kass.
Generous support from the National Science Foundation, the
Petroleum Research Fund, and the Minnesota Supercomputer
Institute for Advanced Computational Research are gratefully
acknowledged. The photoelectron spectra work was supported
by the Division of Chemical Sciences, Geosciences and
Biosciences, Office of Basic Energy Sciences, U.S. Department
of Energy (DOE), and performed at the Environmental
Molecular Sciences Laboratory (EMSL), a national scientific
user facility sponsored by DOE’s Office of Biological and
(17) This percentage is the average of pKa(2 − 3)DMSO/ΔG°acid
(2 − 3)gas phase and pKa(3 − 4)DMSO/ΔG°acid(3 − 4)gas phase, where the
experimental values come from ref 4.
(18) Meot-Ner (Mautner), M.; Speller, C. V. J. Phys. Chem. 1986, 90,
6616−6624.
(19) Lu, Z.; Continetti, R. E. J. Phys. Chem. A 2004, 108, 9962−9969.
(20) Berzinsh, U.; Gustafsson, M.; Hanstorp, D.; Klinkmuller, A.;
Ljungblad, U.; Martensson-Pendrill, A. M. Phys. Rev. A 1995, 51, 231−
238.
2098
dx.doi.org/10.1021/ja2081907 | J. Am. Chem.Soc. 2012, 134, 2094−2099

Journal of the American Chemical Society
Article
(21) Weaver, A.; Arnold, D. W.; Bradforth, S. E.; Neumark, D. M.
J. Chem. Phys. 1991, 94, 1740−1751.
(22) Wang, X. B.; Wang, L. S. J. Chem. Phys. 2000, 113, 10928−
10933.
(23) Wang, X. B.; Vorpagel, E. R.; Yang, X.; Wang, L. S. J. Phys.
Chem. A 2001, 105, 10468−10474.
(24) Meot-Ner (Mautner), M. J. Am. Chem. Soc. 1986, 108, 6189−
6197.
(25) For related studies that report multiple intramolecular hydrogen
bonding interactions, see: (a) Wu, R.; McMahon, T. B. J. Phys. Chem.
B 2009, 113, 8767−8775. (b) Wu, R.; McMahon, T. B. J. Am. Chem.
Soc. 2007, 129, 11312−11313. (c) Norrman, K.; McMahon, T. B.
J. Phys. Chem. A 1999, 103, 7008−7016. (d) Meot-Ner (Mautner), M.;
Sieck, L. W.; Scheiner, S.; Duan, X. J. Am. Chem. Soc. 1994, 116,
7848−7856. (e) Campbell, S.; Rodgers, M. T.; Marzluff, E. M.;
Beauchamp, J. L. J. Am. Chem. Soc. 1995, 117, 12840−12854.
(f) Sharma, R. B.; Blades, A. T.; Kebarle, P. J. Am. Chem. Soc. 1984,
106, 510−516. (g) Meot-Ner (Mautner), M. J. Am. Chem. Soc. 1983,
105, 4906−4911.
(26) Xantheas, S. S. J. Am. Chem. Soc. 1995, 117, 10373−10380.
(27) Bruice, T. C.; Pandit, U. K. Proc. Natl. Acad. Sci. U.S.A. 1960, 46,
402−404.
(28) Ionic hydrogen bond networks were modeled for the trypsin
active site and were found to contribute 65 kcal mol−1 in the resting
state conformation and 32 kcal mol−1 in the transition structure. For
more details, see: Meot-Ner (Mautner), M. J. Am. Chem. Soc. 1988,
110, 3075−3080.
(29) For reviews, see: (a) Joyce, L. A.; Shabbir, S. H.; Anslyn, E. V.
Chem. Soc. Rev. 2010, 39, 3621−3632. (b) Takemoto, Y. Yuki Gosei
Kagaku Kyokaishi 2006, 64, 1139−1147. (c) Schreiner, P. R. Chem.
Soc. Rev. 2003, 32, 289−296. (d) Betschmann, P.; Lerner, C.; Sahli, S.;
Obst, U.; Diederich, F. Chimia 2000, 54, 633−639.
2099
dx.doi.org/10.1021/ja2081907 | J. Am. Chem.Soc. 2012, 134, 2094−2099