08.06.2013 Views

行政院國家科學委員會補助專題研究計畫 期中進度報告

行政院國家科學委員會補助專題研究計畫 期中進度報告

行政院國家科學委員會補助專題研究計畫 期中進度報告

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

<strong>行政院國家科學委員會補助專題研究計畫</strong> <strong>期中進度報告</strong><br />

複雜氫鍵系統之物理化學性質及應用於合成之理論研究<br />

計畫類別: 個別型計畫 □ 整合型計畫<br />

計畫編號:NSC 97-2113-M-007-006-MY3<br />

執行期間: 98 年 8 月 1 日至 99 年 7 月 31 日<br />

計畫主持人:游靜惠<br />

共同主持人:<br />

計畫參與人員: 藍安嚳、黃國韜、劉家源、鄒沛剛、楊昇哲、吳事勳<br />

成果報告類型(依經費核定清單規定繳交):精簡報告 □完整報告<br />

本成果報告包括以下應繳交之附件:<br />

□赴國外出差或研習心得報告一份<br />

□赴大陸地區出差或研習心得報告一份<br />

出席國際學術會議心得報告及發表之論文各一份<br />

□國際合作研究計畫國外研究報告書一份<br />

備註:本計畫為期中報告,不提供公開查詢<br />

執行單位:國立清華大學化學系<br />

中 華 民 國 99 年 5 月 30 日


關鍵字:氫鍵、低能障氫鍵、分子振動、紅外光譜、分子力學、BO 分子<br />

力學、密度泛涵計算<br />

摘要<br />

利用分子力學與量子計算所得之位能面解薛丁格方程式分析氫鍵中之質<br />

子的動力學。研究系統包括八維多項式模型位能與(Dih)2H + (Dih: 4,5-dihydro-<br />

Himidazole) 陽離子。此研究目的在於了解氫鍵中質子之運動特性,並藉以釐清<br />

低能障氫鍵是位能面上的真實低能漲氫鍵或只是平均位置看似位於兩重原子<br />

之中而已,此動力學分析亦可預測氫鍵中質子之紅外光譜,輔助分析實驗光<br />

譜。利用 B3LYP/ D95+(d,p) 計算與簡諧近似法預測(Dih)2H + 陽離子中的連結<br />

氫鍵是低能障氫鍵。量子振動分析顯示此一質子其實是侷限於一端的低能量區<br />

而其能量並不足以穩定類過渡態結構的低能量氫鍵結構。進一步的分子動力學<br />

分析則顯示質子靠近一端的氫鍵結構所佔機率甚大。然而在溫度 300K 時,原<br />

子的運動足以支持振動輔助之質子穿隧機制,紅外光譜可觀測到一典型之侷限<br />

質子具有的寬闊的 NH-延伸振動譜線。此氫鍵之不對稱性亦可預測 NH 的倍頻<br />

譜線。<br />

Keyword: hydrogen bond, low barrier hydrogen bond, molecular vibration, IR spectrum, MD,<br />

BOMD, DFT<br />

Abstract<br />

The harmonic approximation applied to B3LYP/D95+(d,p) quantum chemical computations<br />

predicts the hydrogen bridge (HB) in the (Dih)2H + cation (Dih: 4,5-dihydro-Himidazole) to be a<br />

low barrier hydrogen bond (LBHB). Explicit quantum calculations for the linking proton show,<br />

that proton delocalization does not provide enough energy to stabilize a transition state like<br />

geometry to permanently support a LBHB. The analysis of the (Dih)2H + trajectories shows<br />

further that the majority of the (Dih)2H + conformers supports single well HBs with a localized<br />

proton. However, the increased mobility of the atoms at 300 K supports a vibrationally assisted<br />

proton tunneling mechanism. Furthermore, the increase in atomic mobility results in the wide<br />

broadening of the NH stretch modes typically observed for systems with delocalised protons. The inherent<br />

asymmetry of the N···H + ···N bridge suggests an increased probability to observe overtones of the NH<br />

stretch vibration directly in weakly bond cluster ions with long hydrogen bridges.<br />

2


1 Introduction<br />

The theoretical analysis of the proton mobility in hydrogen bridges (HB) has a long history<br />

[1–4] and biochemical research has joined the discussion last by the analysis of low barrier<br />

hydrogen bonds (LBHBs; HBs where the proton ground state vibrational energy ɛ0 is<br />

higher than the barrier E ‡ for the proton movement) for the energetics [5–9] and proton<br />

tunneling for the kinetics of enzymic reactions [10–14].<br />

The harmonic approximation commonly used to analyze the movement of atoms in<br />

molecules [15, 16] works well for the majority of chemically relevant molecules, but with<br />

proton ground state energies comparable to barrier heights in many HBs, the movement<br />

of the proton cannot be described well with classical models and the quantum nature of<br />

the proton needs to be considered explicitly [17–41].<br />

Two systems are frequently used to probe hydrogen bridges. The Zundel ion H 5 O + 2 and<br />

related species [42–52] and carboxylic acid dimers [53–60]. The focus on proton solvation<br />

stems from the importance of proton transfer processes for chemical reactivity and the<br />

richness of experimental data available, while the clear separation of the stretch mode of<br />

the linking proton from the remaining modes of the carboxylic acid dimer can be used to<br />

simplify the theoretical analysis.<br />

Proton transfer along a hydrogen bridge can be modeled by the movement of a proton<br />

in a double well potential. The potential energy function (PEF) can be described by three<br />

parameters: The distance between the minima, which is linked to the distance between the<br />

heavy atoms spanning the bridge (later called dNN); the height E ‡ of the barrier separating<br />

the minima and finally, the energy difference ∆E between the minima. The value of ∆E<br />

2


can be linked to the difference in basicity ∆PA of the linked bases [61–64] and/or to the<br />

environment of the hydrogen bonded cluster [32, 65–69].<br />

The influence of the environment on the properties of the hydrogen bridge can be<br />

readily observed in the infrared (IR) spectra of the cluster. Hayd and Zundel showed that<br />

thermal fluctuations in the environment manifest themselves in a varying electric field,<br />

which again broadens IR peaks linked to the movement of the linking proton [70]. These<br />

interactions between the HB and its environment become important as the HB length<br />

increases, since short HBs often have a large polarizability [71].<br />

Cluster with a N···H + ···N bridge (later abbreviated as NHN + ) tend to have long<br />

nitrogen-nitrogen distances and thereby tend to form weak hydrogen bridges [72, 73].<br />

The harmonic analysis of the proton movement in these cluster ions suggests, that some of<br />

them can be labeled as LBHBs and that they fall thereby into the group of highly polar-<br />

izable hydrogen bridges [74]. Moreover, the harmonic analysis showed a clear separation<br />

of the NH stretch mode of the HB from the remaining modes of the cluster, which can be<br />

used to simplify the theoretical analysis.<br />

The harmonic analysis [74] suggests a LBHB in (Dih)2H + (Dih: 4,5-Dihydro-1H -<br />

imdazole, C3H6N2) and a ∆E value very close to zero as expected for a homo-molecular<br />

hydrogen bridged system.<br />

HN NH N NH HN N HN NH (1)<br />

The Mulliken [75–78], NBO [79–82] and APT [83, 84] charges in the ground (+0.4, +0.5,<br />

+0.9) and in the transition state (+0.4, +0.5, +1.1) indicate that a proton is moving<br />

between the two Dih molecules as suggested by the cation’s Lewis structure (Eq. 1). The<br />

3


small changes in the proton charge during the transfer indicates that any reduction in<br />

electron density at the proton caused by the cleavage of a NH bond is compensated by<br />

the formation of the new one [85] and is typical property of highly polarizable hydrogen<br />

bridges [1].<br />

These properties makes the (Dih)2H + cation a suitable candidate for the analysis of the<br />

coupling [86, 87] between the proton movement in a HB and its environment. However, a<br />

complete quantum treatment of the 23 atom (Dih)2H + cation by VSCF [88, 89] or quantum<br />

Monte Carlo [90, 91] is beyond the scope of this work. Moreover, the recent discussion<br />

about proton storage in bacteriorhodopsin [92, 93] showed that a ”diffuse vibrational<br />

band is not unique to protonated water cluster” [93] and that the discussion of LBHB<br />

type hydrogen bridges may be reduced to that of the PEF which describes the proton<br />

movement [33, 34, 94, 95]. Hence, our study of NHN + bridges started with a numerical<br />

analysis of model PEFs (Section 3.1). A subsequent set of PEFs was obtained from the<br />

optimized geometry of (Dih)2H + at the B3LYP/D95+(d,p) level and related structures<br />

to investigate whether the delocalization of the proton between the two bases provides<br />

enough energy to stabilize a LBHB with a transition state like geometry (Section 3.2).<br />

Thermal effects were included into the model by Born-Oppenheimer molecular dynamics<br />

(BOMD) simulations (Section 3.3). The BOMD simulations were done at the same level<br />

as electronic structure calculations to maintain a smooth transition between the static<br />

and the dynamic model. Potential energy scans of the proton movement in (Dih)2H +<br />

conformers observed during the BOMD simulations were used to generate PEFs for the<br />

quantum analysis of the proton transfer in these clusters. Hence the data triple (dNN, E ‡ ,<br />

∆E) for the quantum model calculations shows the same thermal fluctuations as expected<br />

4


for a real (Dih)2H + in vacuo.<br />

Section 4 focuses on comparing our data with published data on LBHB formation and<br />

stabilization, hydrogen tunneling and assisted proton transfer reactions and published IR<br />

spectra with an emphasis on the over tones of the NH stretch vibration.<br />

2 Computational Setup<br />

2.1 The wave function of the proton<br />

The proton is assumed to move on a straight line between the heavy atoms of the hydrogen<br />

bridge and the the Hamilton Operator of the proton movement is given by<br />

ˆH = ˆ T + ˆ <br />

V = − 0.047818 kcal ˚A 2 <br />

d2 +<br />

mol dx2 8<br />

i=0<br />

ai x i<br />

where the coefficients ai of the PEF originate from different sources as discussed below.<br />

The eigenfunction of the harmonic oscillator [1, 2, 34] or gaussian functions [36, 40]<br />

are commonly used to expand the proton’s wave function. Hermite polynomials as basis<br />

seem to be the natural choice, as the PEF of a double well hydrogen bridge (DWHB)<br />

can be regarded as the interaction of two harmonic oscillators [1] or as the perturbation<br />

of a harmonic potential [96]. Nevertheless, intergral evaluation is cumbersome and the<br />

quantum calculations tend to converge slowly [1]. On the other hand, integral evaluation<br />

with gaussians is straight forward and therefore, the wave function ψ of the proton was<br />

expanded into a set of evenly spaced (∆xi = 0.05 ˚A) gaussians χi with a common exponent<br />

α<br />

ψ =<br />

N<br />

i=0<br />

ci χi =<br />

1<br />

2 α 4<br />

π<br />

5<br />

N <br />

i=0<br />

−α(x−xi) 2<br />

ci e<br />

(2)<br />

(3)


while the eigenvalue problem was solved by combining the Cholesky decomposition scheme<br />

for the overlap matrix with the Jacobi algorithm [97, 98]. The orbital exponent α was<br />

optimized for every calculation (αopt) to minimize the value ɛ0. This optimization was<br />

done with a linear search algorithm with variable step size at the beginning of the search<br />

and a quadratic search algorithm at the end.<br />

Tests with harmonic potentials reproduced the energy eigenvalues with an error smaller<br />

than 10 −6 kcal/mol and tests with general PEFs showed that the maximum error for the<br />

energy eigenvalues is typically much better than 0.001 kcal/mol.<br />

2.2 Proton Position<br />

LBHB formation has been described as a resonance effect [67, 99–101] of the proton despite<br />

principle limitations of the model [102–104]. Hence, special effort was put into the quantum<br />

chemical analysis of the proton position in this work.<br />

The expectation value xi provides the average value for the position of the proton in<br />

the i-th state. Highly polarizable hydrogen bridges tend to have broad PEFs with a flat<br />

bottom region and are therefore expected to have energetically narrowly spaced states. To<br />

account for possible excitations, the expectation values xi for the individual states were<br />

weighted by their thermal population possibility.<br />

x = 1<br />

Z<br />

N<br />

i=0<br />

xi e −ɛ i/R T = 1<br />

Z<br />

N<br />

i=0<br />

The position xψ of the maximum in ψ 2 0<br />

〈ψi|x|ψi〉 e −ɛ i/R T<br />

and Z =<br />

N<br />

i=0<br />

e −ɛ i/R T<br />

(4)<br />

describes the place with greatest possibility to<br />

observe the proton in a single experiment and can be used to distinguish between a LBHB<br />

(xψ = 0) and a symmetric double well HB (xψ = 0) where x = 0 marks the geometric<br />

center of the HB.<br />

6


A direct measure for the proton delocalization is given by the fraction η, which is part<br />

of the CEDE classification scheme for HBs proposed by Karingithi et al. [37].<br />

η = ψ2 0 (x‡ )<br />

ψ 2 0 (xψ)<br />

The expression ψ 2 0 (x‡ ) is the value ψ 2 0 at the position x‡ of the maximum in the barrier<br />

separating the minima in the PEF. A delocalized proton has an η value of 1 while η equal<br />

to 0 indicates a localized proton, which does not percolate the barrier. Hence, a single-well<br />

HB (SWHB) has a default η value equal to zero.<br />

2.3 Relative Transition Probabilities<br />

The relative transition probability between the initial state i and the final state f scales<br />

with the square of the transition dipole moment µf←i [2, 15]. The proton is the only<br />

particle carrying a charge in the model calculations and hence the dipole moment µ of the<br />

system can be written as e x [16], which can be used to simplify the expression for µf←i.<br />

µf←i = 〈ψf |µ|ψi〉 = 〈ψf |e x|ψi〉 = e 〈ψf |x|ψi〉 = e xf←i<br />

The relative transition probabilities are defined here as the product of the energy difference<br />

between the two states, the square of xf←i and the Maxwell-Boltzman factor of the initial<br />

state.<br />

<br />

Af←i = (ɛf − ɛi) × 349.75<br />

cm−1 <br />

× 〈ψf |x|ψi〉<br />

kcal/mol<br />

2 × e −ɛ <br />

N<br />

i/RT<br />

e<br />

i=0<br />

−ɛ −1<br />

i/RT<br />

Equation 7 was used to estimate the IR spectrum of the linking proton in thermally<br />

fluctuating PEFs by summing the values of Af←i in a given frequency interval (∆˜νf←i =<br />

125 cm −1 ) and constructing a smooth envelope function between 62.5 and 5562.5 cm −1<br />

7<br />

(5)<br />

(6)<br />

(7)


with cubic splines [105]. The final curve was divided by the number of data points used<br />

for its construction to account for different sample sizes.<br />

2.4 The potential energy function<br />

The interaction of a proton with one base can be described with a Lennard-Jones type<br />

interaction potential<br />

VB(x) =<br />

−A<br />

+ ZA |x − xN|<br />

B<br />

|x − xN| ZB<br />

where the parameters A and B were calculated from the targeted well depth Vmin and<br />

position xmin of the minimum relative to the origin of the potential |xN − xmin| = 1 ˚A.<br />

A = Vmin |xN − xmin| ZA<br />

− 1<br />

ZA<br />

ZB<br />

and B = A ZA<br />

ZB<br />

(8)<br />

|xN − xmin| ZA−ZB (9)<br />

The final PEF for the proton movement along the hydrogen bridge is the sum of two<br />

monomer potentials VB(x) each designated by its own subsript.<br />

VHB(x) =<br />

−A1<br />

|x − xN1 |ZA 1<br />

+<br />

B1<br />

|x − xN1 |ZB 1<br />

−<br />

A2<br />

|x − xN2 |ZA 2<br />

+<br />

B2<br />

|x − xN2 |ZB 2<br />

Initial tests showed that choosing V (1)<br />

min = −242 kcal/mol, ZA1 = ZA2 = 1 and ZB1 = ZB2<br />

= 6 yields PEFs similar to those observed for NHN + bridges between amines [74] and<br />

hence these parameters were used as a starting point for the model calculations.<br />

The data points generated by equation 10 were fitted to an 8 th order polynomial<br />

(Equation 2) for the quantum calculations and shifted in energy to ensure V (x1) = 0 at<br />

the global minimum x1 of the PEF.<br />

8<br />

(10)


2.5 Electronic structure calculations<br />

Geometry optimizations and electronic structure calculations for the (Dih)2H + cation were<br />

done at the D95+(d,p)/B3LYP level [106–109] as defined in Gaussian 03 [110]. The nature<br />

of the localized stationary points was tested with subsequent frequency calculations. The<br />

potential energy scans for the generation of the PEF for the proton movement were done<br />

with frozen geometries and the proton moving on a straight line between bases with the<br />

origin of the coordinate system in the geometric centre of the hydrogen bridge and a step<br />

size of 0.01 ˚A.<br />

2.6 Born-Oppenheimer molecular dynamics<br />

Born-Oppenheimer molecular dynamics (BOMD) simulations of the (Dih)2H + cation were<br />

done with GAMESS US [111] at the B3LYP/D95+(d,p) level. The initial geometries and<br />

velocities for the final 4 ps NVE production runs were taken from the last 1.6 ps of a<br />

4.6 ps NVT equilibration run with 1 fs integration steps and a target temperature of 300<br />

K. The points chosen as seeds had to be close to the average total harmonic energy of<br />

the aforementioned last 1.6 ps of the equilibration run and match three different target<br />

temperatures close to 300 K. Computational details are provided within the supplementary<br />

material.<br />

Time series decomposition of data from the BOMD trajectories was done with a moving<br />

average digital filter [112] with a symmetric averaging interval. Automatic peak detection<br />

in these data streams was done with a simple algorithm to locate local extrema in the<br />

filtered signal after additional bezier smoothing [105] followed by a visual verification of<br />

the located points.<br />

9


3 Results<br />

3.1 Calculations with the model potential<br />

Figure 1 summarises the results from calculations with the model PEF. The first set of<br />

calculations (Figures 1a and 1b) analyses the dissociation of a hydrogen bridge with a<br />

fixed ∆PA value of −0.1 kcal/mol. A ∆PA value not equal to zero was chosen to account<br />

for a small asymmetry typically observed in PEFs for the proton movement in cluster<br />

structures obtained from unconstrained geometry optimisations.<br />

The model calculations suggest that the IR spectrum for clusters with a short heavy<br />

atom distance dNN = |rN1 − rN2 | is dominated by the fundamental 1←0 transition as<br />

indicated by the high values for relative transition probability A1←0 (Figure 1a). As the<br />

value of dNN increases, the ground state energy ɛ0 and that of the first vibrationally excited<br />

state ɛ1 degenerate and consequently the first hot band (A2←1) appears in the predicted<br />

IR spectrum. At the very large heavy atom distances dNN, the predicted IR spectrum is<br />

dominated by the first overtone (A2←0) and the transition between the second and fourth<br />

vibrational level (A3←1). The last two transitions correspond to the fundamental vibration<br />

of the proton trapped in the potential energy well at either side of the hydrogen bridge<br />

and thereby mark the breaking of the bridge.<br />

Figure 1b shows the frequencies ˜νf←i of the aforementioned transitions as a function of<br />

dNN. A frequency ˜νf←i is shown only if the corresponding relative transition probability<br />

Af←i has a value greater or equal to 1 ˚A cm −1 . The value of ˜ν1←0 decreases continuously as<br />

the systems passes from a SWHB to a DWHB reflecting thereby the degeneration of ɛ0 and<br />

ɛ1 with increasing values of dNN. As the system passes from a LBHB to a DWHB, the first<br />

10


hot band appears in the predicted IR spectrum. Its intensity (Figure 1a) and frequency<br />

increase until the monomer transitions ˜ν2←0 and ˜ν3←1 appear in the same frequency region<br />

(Figure 1b) and the first hot band vanishes from the predicted IR spectrum.<br />

The fundamental 1←0 and the 2←3 transition are commonly called the tunneling<br />

transitions [2], as they are not observed in long HBs which do not support proton tunneling.<br />

Janoscheck et al. [2] identified the 2←1 and the 3←0 transition as stretch vibrations. At<br />

a HB length of approximate 3.1 ˚A, both transitions have the same frequency until they<br />

vanish at approximately dNN = 3.3 ˚A as the vibronic proton wave functions do not overlap<br />

anymore. From there on, the proton stretch peak in the predicted IR spectra originates<br />

from the proton vibrations of protons captured by the unbound monomers. In so far it<br />

is possible to take the 1←0 peak as an indicator for a shared proton, the 2←1 peak as<br />

indicator for a proton in a strong HB while the 2←0 transition indicates a weak or broken<br />

HB.<br />

Figures 1c to 1e show the relative transition probabilities of the aforementioned transi-<br />

tion as a function of the asymmetry of the potential energy profile for the proton movement<br />

(∆PA = |V (2)<br />

min | − 242 kcal/mol) for various fixed heavy atom distances dNN. As the value<br />

of dNN increases, the relative transition probability A1←0 for the tunneling peak decrease<br />

rapidly. Moreover, Figure 1c shows that symmetric potentials with ∆PA equal to zero<br />

support a tunneling transition stronger than asymmetric ones. At very long heavy atom<br />

distances and large ∆PA values, the value of A1←0 increases again as both the ground and<br />

the first excited state are again bound in a single well.<br />

The ∆PA dependence is more pronounced for the stretch peak of the linking proton<br />

(Figure 1d). Large values of A2←1 require similar energies for the ground and the first<br />

11


excited as observed in systems with an extended heavy atom distance. At very long dNN<br />

values, the vibronic proton wave functions do not overlap anymore and value of A2←1<br />

decreases. However, at this elongated dNN values, seizable value for A2←1 can be observed<br />

only for very symmetric potentials with ∆PA values close to zero.<br />

The opposite behaviour is observed for the 2←0 transition (Figure 1e). Large values<br />

for A2←0 can be observed only in asymmetric potentials with ∆PA values significantly<br />

differing from zero. Moreover, an asymmetry potential can result in large values of A2←0<br />

at distances much shorter than expected from Figure 1a. The intensities of the 2←1 and<br />

the 2←0 transition in the IR spectrum of a hydrogen bridge are therefore likely to provide<br />

essential information about the symmetry of the potential in which the proton moves.<br />

3.2 Static study of the (Dih)2H + cation<br />

The DFT geometry optimization of the (Dih)2H + cation resulted in an asymmetric struc-<br />

ture with the proton localized at one side of the HB (Figure 2a). The proton binding ring<br />

is nearly planar with an NCCN dihedral angle of 0.93 ◦ as expected for the C2v symmetry<br />

of the free DihH + cation, while the other ring is twisted with a NCCN dihedral angle of<br />

15.55 ◦ similar to that observed in the free Dih molecule. In the transition state with E ‡<br />

= 0.881 kcal/mol, the proton stays in the center of the HB and the NCCN dihedrals have<br />

nearly identical values of −10.6 ◦ and −10.3 ◦ . Furthermore, the NN-distance (dNN = 2.57<br />

˚A) is 0.11 ˚A shorter than in the optimized clusters (2.68 ˚A). The frequency analysis of the<br />

optimized structure predicted the NH stretch frequency of the localized proton to be 2172<br />

cm −1 , which corresponds to an harmonic zero point energy (hZPE) of 3.1 kcal/mol. With<br />

hZPE much larger than E ‡ , the (Dih)2H + ion can be classified as a LBHB.<br />

12


The PEF for the proton moving in the frozen framework of optimized (Dih)2H + cation<br />

is very asymmetric with the second minimum 2.523 kcal/mol above the global one. The<br />

barrier separating the minima is 3.071 kcal/mol high and slightly off center at −0.05 ˚A.<br />

The asymmetry in the PEF is the direct result of the cation’s different dihedral angles of<br />

the Dih rings and it cannot be removed by simple shortening the nitrogen-nitrogen distance<br />

dNN. This principle asymmetry is also visible in the proton’s stationary wave function for<br />

the optimised geometry: The proton is effectively localised in the deeper minimum and<br />

the wave function’s tail penetrates the barrier, but no second maximum is observed at the<br />

other side.<br />

The proton in a LBHB is often compared with the proton in the transition state<br />

[40, 61, 113, 114]. The resulting PEF for the proton movement in the transition state is<br />

symmetric and has a small barrier of 0.135 kcal/mol in the center of the HB. In order to<br />

observe a permanently delocalized proton, the reduction the proton’s ground state energy<br />

ɛ0 resulting the removal of the barrier has to compensate the energetic costs of a transition<br />

state like geometry. The summary of the corresponding energy analysis is presented in<br />

Figure 2b. The horizontal line labeled ZPE is the ground state energy ɛ0 of the proton in<br />

the PEF obtained from the fully optimized (Dih)2H + tautomer with the proton localized<br />

at either of the Dih molecules. The framework energy ∆EFr = E ‡ (dNN, xglob) − Eopt is the<br />

energy E ‡ (dNN, xglob) of the (Dih)2H + cation in a generalized transition state geometry at<br />

a given nitrogen-nitrogen distance dNN and the proton at the position xglob of the global<br />

minimum of the PEF relative to the energy Eopt of the fully optimized (Dih)2H + tautomer<br />

(dNN = 2.68 ˚A, xglob = 0.239 ˚A). ∆EFr provides therefore a measure for the energetic costs<br />

of the framework distortion and has been found to be minimal at dNN equal to 2.63 ˚A.<br />

13


A structure can be labeled as stabilized by proton delocalization, if the sum ΣE of ɛ0<br />

and ∆EFr falls below the value of ZPE. The ΣE curve has its minimum at 2.61 ˚A, which<br />

suggests a very small stabilization energy of 0.639 kcal/mol. The PEF at this point has a<br />

barrier of 0.508 kcal/mol, but the hydrogen bridge is still a LBHB with ɛ0 = 1.15 kcal/mol.<br />

3.3 Dynamic study of the (Dih)2H + cation<br />

The results of the final NVE BOMD simulations are summarized in Table 1 with a focus on<br />

the NH stretch vibration along the hydrogen bridge. The analysis of the proton movement<br />

in run number 2a suggests a periodicity of 15 fs for the NH stretch vibration, which agrees<br />

very well with an estimated value of 15.3 fs from the harmonic frequency analysis. The<br />

time series decomposition of dNN in run number 3 (supplementary material) suggests a<br />

periodicity of 225.4 fs for the NN stretch vibration, which perfectly reproduces a periodicity<br />

of 225.4 fs from the harmonic frequency analysis. A repetition of this analysis with the first<br />

and second run yielded values of 222.5 and 223.6 fs in good agreement with the harmonic<br />

frequency analysis. The proton in the hydrogen bridge vibrates therefore 15 times faster<br />

than its environment, which justifies the adiabatic separation of both movements in the<br />

model.<br />

The value of dNN varies between 2.47 and 2.98 ˚A in run number 2 (Figure 3a): 248<br />

geometries (6 %) have dNN values smaller than the transition state geometry (2.57 ˚A) while<br />

the majority (2227 structure, 55%) has a NN distance larger than 2.68 ˚A. The average<br />

value for dNN (2.70 ˚A) is therefore slightly larger than that observed in the optimised<br />

geometry (2.68 ˚A).<br />

Despite the periodic occurrence of short NN distances in which the proton automati-<br />

14


cally moves towards the centre of the hydrogen bridge, the event of proton crossing appears<br />

to be aperiodic (Figure 3b). One of the reasons for this aperiodicity is the uncorrelated<br />

twisting of the Dih rings as indicated by the large standard deviations for δ1 and δ2. The<br />

dihedral angles of the Dih rings sway between ± 40 ◦ and usually differ strongly in size<br />

resulting in asymmetric potential energy functions which localise the proton at one side<br />

of the bridge.<br />

A small section of points (run number 2a, grey area in Figure 3a) between 2530 and<br />

2977 fs from run number 2 was chosen for the quantum analysis of the proton movement.<br />

Figure 3c shows the projection of the proton position onto the NN line in comparison with<br />

the global maximum of ψ 2 0<br />

of the proton’s ground state wave function. The position (xψ)<br />

of the maximum of ψ2 0 follows the classical proton position (xH +) most times, though the<br />

ability of the quantum particle to tunnel into the barrier moves xψ closer to the centre of<br />

the hydrogen bridge than its classic counter part x H +.<br />

Thermal fluctuations can deepen the potential energy well of the unoccupied hydrogen<br />

position relative to that of the occupied one. Hence, the maximum of ψ 2 0<br />

can be observed<br />

on the opposite site of the hydrogen bridge five times during the BOMD simulation. It<br />

stays there about 10 ± 2 fs 1 , which can be compared with the average periodicity of 15 fs<br />

for the NH stretch vibration and 9.1 fs for the non-bonding NH stretch vibrations (3657<br />

and 3675 cm −1 ). Moreover, a time interval of approximately 10 fs is equal to the period<br />

length of a frequency of 3336 cm −1 which lies in the region NH stretch vibrations typically<br />

observed for amines (3300 to 3500 cm −1 ) [115, 116]. Hence, it is possible to say that the<br />

1 Peak ➀ is only half as wide as the remaining ones. Its exclusion from the average calculation yields a<br />

time of 11 ± 1 fs for the maximum of ψ 2 0 being on the other side of the hydrogen bridge.<br />

15


maximum of ψ2 0 stays on the opposite of the hydrogen bridge for approximately the length<br />

of the period of NH stretch vibration.<br />

Whether these times with the maximum of ψ 2 0<br />

on the opposite site of the hydrogen<br />

bridge provide an opportunity for proton tunneling or not depends on the localisation of<br />

the H + . Figure 3d shows the value of η as a function of time for run number 2a. LBHBs<br />

(marked grey) can have η values as low as 0.8 and a η value of 0.8 was therefore chosen as<br />

a criterion for the effective proton delocalisation in DWHBs. Using this criterion, peaks ➀,<br />

➁, ➃ and ➄ indicate a high possibility for the proton to move to the other side of the HB,<br />

while the low η value for peak ➂ suggests that the proton stays at the position predicted<br />

by the BOMD simulation and thereby represents an excited sate of the proton in the HB.<br />

Figure 3e shows the correlation between ˜ν1←0 and ˜ν2←1 in run number 2a. The left<br />

corner with low values for ˜ν1←0 and medium large ˜ν2←1 values marks cluster geometries,<br />

which support long double well hydrogen bridges (DWHBs). In these hydrogen bridges<br />

ɛ0 and ɛ1 are nearly degenerate and the first hot band has become the dominant transi-<br />

tion. The upper right corner of the triangle is marked by short SWHBs with their nearly<br />

harmonic PEFs. The connecting line between these two points is marked by hydrogen<br />

bridges, which preserve their symmetry while stretching the NN distance. LBHBs can be<br />

found in a small region at the centre of this line. The lower right corner is marked by<br />

long, strongly asymmetric DWHBs with transition state energies for the proton movement<br />

close to the energy of the first excited state. The lower arm of the triangle marks therefore<br />

hydrogen bridges with increasing asymmetry.<br />

The quantum calculations for the proton movement in run number 2a showed that only<br />

6.5% of all hydrogen bridges in Figure 3e can be labelled as LBHBs and that short SWHBs<br />

16


are also rare. Hence, the majority of clusters in run number 2a supports asymmetric<br />

hydrogen bridges. This principal asymmetry becomes visible by looking at the averaged<br />

results (Table 1) from the individual quantum calculation for the proton: The value of η<br />

plunges down to 0.18, the energies ɛ0 and ɛ1 rise while the value of ˜ν1←0 increases to 938<br />

cm −1 .<br />

Initial calculations on the proton movement in the frozen geometry of the optimized<br />

(Dih)2H + cation showed, the ion’s dipole moment can be reasonably well approximated<br />

with a linear fit (supplementary material). Hence, equation 7 can be used to estimate<br />

the IR intensities of the proton vibration in (Dih)2H + . Figure 3f shows the corresponding<br />

simulation of the IR spectrum for a proton in a fluctuating PEF at 300 K. The spectrum is<br />

clearly dominated by the fundamental tunneling peak at 1008 cm −1 . Small contributions<br />

from the first hot band (1155 cm −1 ) blue shift the maximum of the envelop to 1069<br />

cm −1 . The next big contributor is the first overtone with its maximum at 1828 cm −1 .<br />

The dominance of the 2←0 transition over the 2←1 is the direct result of the underlying<br />

asymmetry as a high A2←1 value requires a symmetric hydrogen bridge (Figure 1d), while<br />

asymmetric hydrogen bridges generally result in high values for A2←0 (Figure 1e). The<br />

smallest contribution to the spectrum arises from the 3←0 transition (2445 and 3051<br />

cm −1 ), which has been assigned to the proton stretch in the free monomers (Figure 1a),<br />

but also benefits from the general asymmetry of the PEF.<br />

The average dNN value of 2.680 ˚A at 80 K reproduces that of the optimised geometry<br />

2.681 ˚A very well and the gaussin envelope of the dNN distribution function (Figure 4a)<br />

reaches values close to zero for dNN values close to that observed in the transition state<br />

geometry. Within the time window of observation, no proton transfer was recorded for<br />

17


un number 4.<br />

Figure 4b shows the correlation between ˜ν1←0 and ˜ν2←1 in run number 4a. Areas of the<br />

plot assigned to long DWHBs and short SWHBs are hardly populated while the majority<br />

of points cumulates in the centre with a clear tendency towards asymmetric DWHBs. This<br />

property of the trajectory is reflected in the properties of the dihedral angles δ1 and δ2:<br />

Their average values are close to zero and the corresponding standard deviations similar<br />

to the values observed at 300 K. This uncorrelated movement of the dihedrals prevents<br />

the formation of symmetric cluster geometries which could support proton transfer. The<br />

position (xΨ) of the maximum of the squared ground state wave function (Ψ2 0 ) follows<br />

therefore that of the proton (x H +) in the BOMD trajectory with only one exception as<br />

shown in Figure 4c. This peak overlaps with a region of LBHBs with η values close to 1 and<br />

hence marks an opportunity for a successful tunneling process, while four opportunities<br />

were observed in an interval of similar length at 300 K.<br />

This result contrasts the observation that the average η value in run number 4a (η =<br />

0.28) is significantly larger than that in run number 2a (η = 0.18). The small amplitude<br />

atomic movement in run number 4a generates cluster geometries which resemble that of<br />

the optimised structure and thereby η values close to 0.26. At 300 K, the trajectory<br />

runs through areas of more distorted geometries with η values close to zero. The average<br />

value of η is therefore expected to be much lower than that of the optimised structure.<br />

Meanwhile, the large amplitude of the atomic movement in run number 2a generates more<br />

cluster geometries which support proton transfer either by shortening the NN distance or<br />

by equilibrating both sides of the hydrogen bridge. The thermal motion thereby supports<br />

the proton tunneling process in run number 2a while simultaneously reducing the LBHB<br />

18


character of the hydrogen bridge.<br />

The same behaviour can be observed in the thermally averaged expectation value x of<br />

the hydrogen position. The proton movement shows a large amplitude sampling both sides<br />

of the hydrogen bridge at 300 K (Figure 3c), while the proton stays at one side at 80 K<br />

(Figure 4c). The values of x stay closer to center of the HB as those of xψ and x H + in both<br />

cases. However, the average of the absolute value of x is to large to classify the system as<br />

a symmetric DWHB or as a LBHB. Moreover, the excited states of the proton movement<br />

are not populated at 80 K and the properties of the proton are solely determined by the<br />

ground state wave function.<br />

Figure 4d shows the IR spectrum estimated from the PEFs gathered from run number<br />

4a. The most dramatic changes can be observed in the low frequency part of the spectrum.<br />

The tunneling peak is red shifted by 193 cm −1 and significantly taller as the result of the<br />

tighter distribution of dNN values (Figure 4a), while the peak for the first hot band is<br />

completely missing. The peaks ˜ν2←0 and ˜ν3←0 are blue shifted by 173 and 343 cm −1 while<br />

the minor peak (2445 cm −1 ) of the bimodal ˜ν3←0 band vanishes. The small decrease in<br />

the intensities of the ˜ν2←0 and ˜ν3←0 peaks can be linked to the tighter distribution of dNN<br />

values at 80 K which does not cover the long distances necessary for large value of A2←0<br />

and A3←0. These intensity changes are small compared to changes in the ˜ν1←0 peak and<br />

the intensities of these peaks appear to be almost constant.<br />

19


4 Discussion<br />

It has been argued that the vanishing barrier for the proton movement in LBHBs and<br />

the associated energy gain play a significant role in enzymic catalysis [5–9, 117]. The<br />

calculations with the static model show that the barrier does not completely vanish and<br />

that proton delocalisation can indeed stabilise the (Dih)2H + ion with transition state<br />

like geometries. Nevertheless, the energy gain is much smaller than the estimated upper<br />

limit of 4 kcal/mol published by Perrin and Nielson [118]. The observed energy gain<br />

of 0.64 kcal/mol is equal to the kinetic kinetic energy of the 23 atoms in (Dih)2H + at<br />

9 K. As all atoms of the (Dih)2H + ion need to be aligned for the necessary symmetric<br />

PEF, the observation of a LBHB becomes increasingly unlikely at temperatures markedly<br />

higher than 9 K. The analysis of the BOMD trajectories underpins this argument as only<br />

14% of all (Dih)2H + conformers at 80 K (Ekin = 5.5 kcal/mol) can support a LBHB<br />

compared with 5.8% at 300 K (Ekin = 21 kcal/mol). In that respect, the definition of<br />

LBHB supporting environments [6, 9, 99, 117–120] has to be amended by the constraint<br />

that symmetry breaking modes are thermally unaccessible in the temperature region of<br />

interest. Moreover, the necessity of such an amendment increases with the number of<br />

atoms necessary to be lined.<br />

The quantum proton represented either by xψ or by x generally follows the movements<br />

of its classic counterpart x H + (Figure 4c), however with a lower frequency. On several<br />

occasions, xψ and x are on the opposite of the hydrogen bridge and high η values suggest<br />

a high tunneling probability for the ground state proton. The only execption from this<br />

general behaviour is peak ➂ in Figure 3c, where the proton would be trapped in the<br />

20


first excited state for a short time [3]. Moreover, the thermal analysis indicates that the<br />

excited states of the proton are sparsely populated and that the dynamics of the proton<br />

in (Dih)2H + are dominated by the properties of the ground state and that the adiabatic<br />

approximation holds [4].<br />

The comparison of Figures 3c and 4c suggests that the possibility of proton tunneling<br />

increases with the temperature. A similar effect has been reported for the protonated<br />

1,8-bis(dimethylamino)naphthalene proton sponge [121]. Its electron density map at 120<br />

K suggests two proton positions in the HB, while a single position in the centre is observed<br />

at 300 K. Two mechanisms are commonly used to explain the temperature dependence<br />

of the proton position in enzymes [10–14, 122, 123]. In the case of vibrationally enhanced<br />

tunneling the vibrations of the environment compress the reaction coordinate and thereby<br />

facilitate the tunneling process. In the second case, vibrationally assisted tunneling, the<br />

normal modes of the environment equilibrate the relative energies of the two proton po-<br />

sitions to support the tunneling process. The peaks ➀, ➁, ➃ and ➄ in Figures 3c and 3d<br />

indicate tunneling possibilities while the reaction coordinate for the proton transfer (dNN)<br />

is markedly stretched in agreement with the second mechanism.<br />

The smallest number of proton transfer reactions was observed in run number 1 and<br />

the analysis of the (Dih)2H + conformers with minimum dNN values in this run show, that<br />

short dNN distances create PESs with a single minimum; however, this minimum is not<br />

necessarily located in the centre of the hydrogen bridge. The asymmetric PESs show a<br />

narrow minimum which locates the proton firmly at one side of the HB [124] and do not<br />

support proton transfer of any kind. Hence, vibrationally assisted tunneling seems to be<br />

more important for the (Dih)2H + ion at 300 K than the vibrationally enhanced process.<br />

21


The quantum calculations on the proton movement also provide the means to interpret<br />

the IR spectra of hydrogen bonded cluster. A single IR peak below 1700 cm -1 for an<br />

NH or OH stretch mode is an indicator for a symmetric HB in according Johnson and<br />

Rumon [125] and proton tunneling peak is usually indicated by a continuum below 1600<br />

cm -1 with maximum values between 500 and 950 cm -1 [126]. Quantum calculation for the<br />

proton in proton sponges [38] predict frequencies for the tunneling peak between 509 and<br />

788 cm -1 and similar calculations for the proton in H5O + 2<br />

[1] yield frequencies between 4<br />

and 759 cm -1 depending on the oxygen-oxygen distance. A more recent study [37] reports<br />

frequencies for the tunneling peak between 17 and 245 cm -1 for malon-aldehydes and 1378<br />

cm -1 for the maleate anion with a symmetric SWHB. Test calculations with the model<br />

potential show that only if the hydrogen bridges have the same length, the ˜ν1←0 value can<br />

be used to gauge the delocalization of the proton as the effective width of the potential<br />

energy well has a much stronger influence on ˜ν1←0 than the barrier height E ‡ .<br />

Thermal and temporal averaging of cluster geometries tends to result in more sym-<br />

metric structures and elongated heavy atom distances and consequently low ˜ν1←0 values.<br />

In so far, the strong red shift observed for ˜ν1←0 peak for temporal averaged geometries<br />

has to be regarded as a false positive, specially since averaging the individual IR spectra<br />

suggests a significant blue shift. Nevertheless, the predicted ˜ν1←0 values for the optimized<br />

structures are at the upper limit of those expected for a LBHB and that the maxima of<br />

simulated IR still in range or slightly above.<br />

The observation of IR absorption in the first hot band requires a large heavy atom<br />

distance (Figure 1a, dNN > 2.6 ˚A) and a symmetric potential energy function (Figure 1d,<br />

∆PA = 0). As the distance between the heavy atoms increases with the excitation of the<br />

22


NN stretch vibration, thermal excitation is expected to increase the value of A2←1 and<br />

˜ν2←1 to be part of combination adsorption [2]. The harmonic frequency calculation for<br />

the (Dih)2H + ion suggests a NN stretch frequency of 148 cm −1 and the combination peak<br />

for ˜ν2←1 is therefore expected to be blue shifted by a similar value. However, the general<br />

asymmetry of the thermally excited (Dih)2H + ion drastically reduces the value of A2←1<br />

and hot band absorption is expected to play only a minor role at 300 K (Figure 3f) and<br />

to be negligible at 80 K (Figure 4d). The NHN + hydrogen bridges in proton sponges tend<br />

to be highly symmetric, but their short NN distance still prevents the observation of hot<br />

band absorption [38, 121].<br />

A different behavior is observed for the first overtone (˜ν2←0). Though forbidden in a<br />

harmonic potential energy well and to be expected for even longer heavy atom distances<br />

than the 1 st hot band (Figure 1a), the first overtone contributes significantly to the sim-<br />

ulated spectra (Figures 3f and 4d). The asymmetry of the thermally excited (Dih)2H +<br />

ion makes its observation possible. A similar phenomenon has been reported by Roberts<br />

et al. for the proton transfer in hydroxide solutions [46]. During each proton transfer<br />

attempt, the frequency of the first the overtone of the OH stretch vibration drops into the<br />

experimentally probed frequency region. But, calculations by McCoy et al. [90] for the free<br />

H3O − 2<br />

ion suggest the intensity of the the first overtone signal to be very small as expected<br />

for the short OO distance and the high symmetry of the ion. Moreover, the calculations<br />

on the Zundel ions suggest [50, 90], that combinations of the heavy atom stretch and out<br />

of plane bending vibrations can cover the overtone vibration. Weakly bound NHN + ions<br />

therefore seem to be ideal candidates for the direct observation of these overtones.<br />

23


5 Conclsions<br />

The standard quantum chemical analysis predicts the (Dih)2H + cation to form a LBHB.<br />

The quantum analysis of the proton movement in (Dih)2H + shows that the proton wave<br />

function penetrates the barrier, which causes a a low NH stretch frequency of 802 cm −1<br />

and the proton is essentially localized on one side of the hydrogen bridge. This asymmetry<br />

can be traced back to the conformation of the Dih rings. The proton binding Dih ring is<br />

planar as observed in the free DihH + ion while the other ring is twisted as observed in the<br />

free Dih molecule. The twisted ring binds the moving proton weaker than the planar ones<br />

which results in the observed asymmetry of the hydrogen bridge.<br />

Thermal excitation of the cluster ion increases the chance of a proton transfer reac-<br />

tion via a vibrationally assisted tunneling mechanism. However, the ion’s trajectory runs<br />

through conformational regions, which strongly localize the proton. The overall quantum<br />

delocalization of the proton decreases therefore despite an increase in the proton transfer<br />

reactions. At high temperatures, the (Dih)2H + ion resembles a system with the proton<br />

hopping between two basins. At low temperatures, the proton is localized in one basin,<br />

but its wave function penetrates the barrier which causes a low frequency tunneling peak<br />

(˜ν1←0) in the IR spectrum.<br />

The BOMD simulations reproduce correctly the periodicity results from the harmonic<br />

frequency analysis and thereby link seamlessly to standard quantum calculations. The<br />

most dramatic change can be observed in the position of the proton tunneling peak. It is<br />

red shifted from 2170 cm −1 in the harmonic approximation to 1069 cm −1 in the explicit<br />

proton quantum calculation and extremely broadened. The quantum calculation for the<br />

24


proton movement suggests the IR absorption form the first overtone should be observable<br />

at 1828 cm −1 overlapping with Dih ring vibrations and out of plane bending vibrations of<br />

the linking proton (harmonic approximation 1602 cm −1 ). Moreover, the long NN distance<br />

in (Dih)2H + and the thermally induced asymmetry predict a sizeable intensity for the<br />

first overtone, which makes weakly bound NHN + ions suitable candidates for the direct<br />

spectroscopic analysis of overtones from proton stretch vibrations in hydrogen bridges.<br />

25


References<br />

[1] Janoschek, R.; Weidemann, E. G.; Pfeiffer, H.; Zundel, G. J. Am. Chem. Soc. 1972,<br />

94 , 2387–2396.<br />

[2] Janoschek, R.; Weidemann, E. G.; Zundel, G. J. Chem. Soc., Faraday Trans. 2<br />

1973, 69 , 505–520.<br />

[3] Hammes-Schiffer, S.; Tully, J. C. J. Phys. Chem. 1995, 99 , 5793–5797.<br />

[4] Staib, A.; Borgis, D.; Hynes, J. T. J. Chem. Phys. 1995, 102 , 2487–2505.<br />

[5] Gerlt, J. A.; Gassman, P. G. J. Am. Chem. Soc. 1993, 115 , 11552–11568.<br />

[6] Warshel, A.; Papazyan, A.; Kollmann, P. A.; Cleland, W. W.; Kreevoy, M. M.; Frey,<br />

P. A. Science 1995, 269 , 102–106.<br />

[7] Cleland, W. W.; Frey, P. A.; Gerlt, J. A. J. Biol. Chem. 1998, 273 , 25529–25532.<br />

[8] Gerlt, J. A.; Kreevoy, M. M.; Cleland, W. W.; Frey, P. A. Chem. & Biol. 1997, 4 ,<br />

259–267.<br />

[9] Guthrie, J. P. Chem. & Biol. 1996, 3 , 163–170.<br />

[10] A.Kohen; J.Klinman Chemistry & Biology 1999, 6 , R191–R198.<br />

[11] Scrutton, N. S.; 1, J. B.; Sutcliffe, M. J. Eur. J. Biochem. 2001, 264 , 666–671.<br />

[12] Sutcliffe, M. J.; Scrutton, N. S. Phil. Trans. R. Soc. Lond. A 2000, 358 , 367–386.<br />

[13] Villà, J.; Warshel, A. J. Phys. Chem. B 2001, 105 , 7887–7907.<br />

[14] Bruno, W.; Bialek, W. Biophys. J. 1992, 63 , 689–699.<br />

26


[15] Zülicke, L.; Quantenchemie - Band 1 ; 1 st edition; Dr. Alfred Hüthig Verlag, Heidel-<br />

berg 1978.<br />

[16] Levine, I. N.; Quantum Chemistry; 5 th edition; Pearson Education Taiwan Ltd.,<br />

Taipei 2006.<br />

[17] Duong, M. P. T.; Kim, Y. J. Phys. Chem. A 2010, ASAP, 1–8.<br />

[18] del Amo, J. M. L.; Langer, U.; Torres, V.; Buntkowsky, G.; Vieth, H.-M.; Pérez-<br />

Torralba, M.; Sanz, D.; Claramunt, R. M.; Elguero, J. J. Am. Chem. Soc. 2008,<br />

130 , 8620–8632.<br />

[19] Siwick, B. J.; Cox, M. J.; Bakker, H. J. J. Phys. Chem. B 2008, 112 , 378–389.<br />

[20] Latanowicz, L.; Filipek, P. J. Phys. Chem. A 2007, 111 , 7695–7702.<br />

[21] Barroso, M.; Arnaut, L. G.; Formosinho, S. J. ChemPhysChem 2005, 6 , 363–371.<br />

[22] Zhang, C.; Dyer, M.; Alavi, A. Journal of Physical Chemistry B 2005, 109 (47),<br />

22089–22091.<br />

[23] Manca, C.; Tanner, C.; Coussan, S.; Bach, A.; Leutwyler, S. J. Chem. Phys. 2004,<br />

121 , 2578–2590.<br />

[24] Golo, V. L.; Volkov, Y. S. Int. J. Mod. Phys. C 2003, 14 , 133–156.<br />

[25] Makri, N.; Miller, W. H. J. Chem. Phys. 1987, 86 , 1451–1457.<br />

[26] Smedarchina, Z.; Siebrand, W.; Fernańdez-Ramos, A.; Cui, Q. J. Am. Chem. Soc.<br />

2003, 125 , 243–251.<br />

27


[27] Smedarchina, Z.; Siebrand, W.; Fernández-Ramos, A.; Gorb, L.; Leszczynski, J. J.<br />

Chem. Phys. 2000, 112 , 566–573.<br />

[28] Mei, H. S.; Tuckerman, M. E.; Sagnella, D. E.; Klein, M. L. J. Phys. Chem. B 1998,<br />

102 , 10446–10458.<br />

[29] Brougham, D. F.; Horsewill, A. J.; Ikram, A. J. Chem. Phys. 1996, 105 , 979–982.<br />

[30] Althorpe, S. C.; Clary, D. C. J. Chem. Phys. 1994, 101 , 3603–3607.<br />

[31] Wales, D. J. J. Am. Chem. Soc. 1993, 115 , 11180–11190.<br />

[32] Surez, A.; Silbey, R. J. Chem. Phys. 1991, 94 , 4809–4816.<br />

[33] Benoit, M.; Marx, D. ChemPhysChem 2005, 6 , 1738–1741.<br />

[34] Somorjai, R. L.; Hornig, D. F. J. Chem. Phys. 1962, 36 , 1980–1987.<br />

[35] Totsuji, C. J. Phys. Soc. Japan 2004, 73 , 1054–1058.<br />

[36] Totsuji, C.; Matsubara, T. J. Phys. Soc. Japan 1994, 63 , 2760–2768.<br />

[37] Karingithi, R. N.; Shaw, C. L.; Roberts, E. W.; Molina, P. A. J. Mol. Struc.<br />

(Theochem) 2008, 851 , 92–99.<br />

[38] Bieko, A. J.; Latajka, Z.; Sawka-Dobrowolska, W.; Sobczyk, L.; Ozeryanskii, V. A.;<br />

Pozharskii, A. F.; Grech, E.; Nowicka-Scheibe, J. J. Chem. Phys. 2003, 119 , 4313–<br />

4319.<br />

[39] Garcia-Viloca, M.; Gelabert, R.; González-Lafont, À.; Moreno, M.; Lluch, J. M. J.<br />

Am. Chem. Soc. 1998, 120 , 10203–10209.<br />

28


[40] Garcia-Viloca, M.; Gelabert, R.; Gonzalez-Lafont, A.; Moreno, M.; Lluch, J. M. J.<br />

Phys. Chem. A 1997, 101 , 8727–8733.<br />

[41] Fillaux, F.; Nicolaï, B. Chem. Phys. Lett. 2005, 415 , 357–361.<br />

[42] Ortega, I. K.; Escribano, R.; Herrero, V. J.; Maté, B.; Moreno, M. A. J. Mol. Struc.<br />

2005, 742 , 147–152.<br />

[43] Headrick, J. M.; Diken, E. G.; Walters, R. S.; Hammer, N. I.; Christie, R. A.; Cui,<br />

J.; Myshakin, E. M.; Duncan, M. A.; Johnson, M. A.; Jordan, K. D. Science 2005,<br />

308 , 1765–1769.<br />

[44] Gutberlet, A.; Schwaab, G.; Birer,<br />

Havenith, M.; Marx, D. Science 2009, 324 , 1545–1548.<br />

Ö.; Masia, M.; Kaczmarek, A.; Forbert, H.;<br />

[45] Iftimie, R.; Tuckerman, M. E. Angew. Chem. Int. Ed. 2006, 45 , 1144–1147.<br />

[46] Roberts, S. T.; Petersen, P. B.; Ramasesha, K.; Tokmakoff, A.; Ufimtsev, I. S.;<br />

Martinez, T. J. Proc. Nat. Acad. Sci. 2009, 106 , 15154–15159.<br />

[47] Loparo, J. J.; Roberts, S. T.; Tokmakoff, A. J. Chem. Phys. 2006, 125 , 194522–<br />

194533.<br />

[48] Eaves, J. D.; Loparo, J. J.; Fecko, C. J.; Roberts, S. T.; Tokmakoff, A.; Geissler,<br />

P. L. Proc. Natl. Sci. Acad. USA 2005, 102 , 13019–13022.<br />

[49] Asthagiri, D.; Pratt, L. R.; Kress, J. D. Proc. Natl. Sci. Acad. USA 2005, 102 ,<br />

6704–6708.<br />

29


[50] Kaledin, M.; Kaledin, A. L.; Bowman, J. M. J. Phys. Chem. A 2006, 110 , 2933–<br />

2939.<br />

[51] Stoyanov, E. S.; Reed, C. A. J. Phys. Chem. A 2006, 110 , 12992–13002.<br />

[52] Stoyanov, E. S.; Stoyanova, I. V.; Reed, C. A. J. Am. Chem. Soc. 2010, 132 , 1484–<br />

1485.<br />

[53] Ortlieb, M.; Havenith, M. J. Phys. Chem. A 2007, 111 , 7355–7363.<br />

[54] Brogan, A. P.; Dickerson, T. J.; Janda, K. D. Angew. Chem. Int. Ed. 2006, 45 ,<br />

8100–8102.<br />

[55] Cassidy, C. S.; Lin, J.; Tobin, J. B.; Frey, P. A. Bioorganic Chemistry 1998, 26 ,<br />

213–219.<br />

[56] Marushkevich, K.; Khriachtchev, L.; Räsänen, M. J. Phys. Chem. A 2007, 111 ,<br />

2040–2042.<br />

[57] Madeja, F.; Havenith, M. J. Chem. Phys. 2002, 117 , 7162–7168.<br />

[58] Marechal, Y.; Witkowski, A. J. Chem. Phys. 1968, 48 , 3697–3705.<br />

[59] Najeh, R.; Houcine, G.; Flakus, H. T.; Jablonska, M.; Brahim, O. J. Comp. Chem.<br />

2009, 31 , 463–475.<br />

[60] Witkowski, A. J. Chem. Phys. 1970, 52 , 4403–4406.<br />

[61] Gilli, G.; Gilli, P. J. Mol. Struc. 2000, 552 , 1–15.<br />

[62] Smets, J.; McCarthy, W.; Maes, G.; Adamowicz, L. J. Mol. Struc. 1999, 476 , 27–43.<br />

30


[63] Chandra, A. K.; Zeegers-Huyskens, T. J. Org. Chem. 2003, 68 , 3618–3625.<br />

[64] Lankau, T.; Yu, C.-H. Phys. Chem. Chem. Phys 2007, 9 , 299–310.<br />

[65] Schiøtt, B. Chem. Commun. 2004, 5 , 498–499.<br />

[66] Pan, Y.; McAllister, M. A. J. Am. Chem. Soc. 1997, 119 , 7561–7566.<br />

[67] Perrin, C. L.; Ohta, B. K. J. Mol. Struc. 2003, 664 , 1–12.<br />

[68] Scheiner, S. J. Phys. Chem. B 2005, 109 , 16132–16141.<br />

[69]<br />

Özen, A. S.; Doruker, P.; Aviyente, V. J. Phys. Chem. A 2007, 111 , 13506–13514.<br />

[70] Hayd, A.; Weidemann, E. G.; Zundel, G. J. Chem. Phys. 1979, 70 , 86–91.<br />

[71] Hayd, A.; Zundel, G. Journal of Molecular Structure 2000, 500 , 421–427.<br />

[72] Raub, S.; Marian, C. M. J. Comp. Chem. 2007, 28 , 1503–1515.<br />

[73]<br />

Özen, A.; DeProft, F.; Aviyente, V.; Geerlings, P. J. Phys. Chem. A 2006, 110 (17),<br />

5860–5868.<br />

[74] Lankau, T.; Yu, C.-H. Int. J. Quant. Chem. 2009, 109 , 2286–2296.<br />

[75] Mulliken, R. S. J. Chem. Phys. 1955, 23 , 1833–1840.<br />

[76] Mulliken, R. S. J. Chem. Phys. 1955, 23 , 1844–1846.<br />

[77] Mulliken, R. S. J. Chem. Phys. 1955, 23 , 2338–2342.<br />

[78] Mulliken, R. S. J. Chem. Phys. 1955, 23 , 2343–2346.<br />

[79] Foster, J. P.; Weinhold, F. J. Am. Chem. Soc. 1980, 102 , 7211–7218.<br />

31


[80] Reed, A. E.; Weinhold, F. J. Chem. Phys. 1983, 78 , 4066–4073.<br />

[81] Reed, A. E.; Weinstock, R. B.; Weinhold, F. J. Chem. Phys. 1985, 83 , 735–746.<br />

[82] Reed, A. E.; Weinhold, F. J. Chem. Phys. 1985, 83 , 1736–1740.<br />

[83] Cioslowski, J. J. Am. Chem. Soc. 1989, 111 , 8333–8336.<br />

[84] Proft, F. D.; Alsenoy, C. V.; Peeters, A.; Langenaeker, W.; Geerlings, P. J. Comp.<br />

Chem. 2002, 23 , 1198–1209.<br />

[85] Lankau, T.; Yu, C.-H. accepted by J. Comp. Chem.<br />

[86] Giese, K.; Ushiyama, H.; Takatsuka, K.; Kühn, O. J. Chem. Phys. 2005, 122 ,<br />

124307–124307.<br />

[87] Johannissen, L. O.; Hay, S.; Scrutton, N. S.; Sutcliffe, M. J. J. Phys. Chem. B 2007,<br />

111 , 2631–2638.<br />

[88] Chaban, G. M.; Jung, J. O.; Gerber, R. B. J. Phys. Chem. A 2000, 104 , 2772–2779.<br />

[89] Swalina, C.; Hammes-Schiffer, S. J. Phys. Chem. A 2005, 109 , 10410–10417.<br />

[90] McCoy, A. B.; Diken, E. G.; Johnson, M. A. J. Phys. Chem. A 2009, 113 , 7346–<br />

7352.<br />

[91] Swalina, C.; Wang, Q.; Chakraborty, A.; Hammes-Schiffer, S. J. Phys. Chem. A<br />

2007, 111 , 2206–2212.<br />

[92] Mathias, G.; Marx, D. Proc. Nat. Acad. Sci. 2007, 104 , 6980–6985.<br />

32


[93] Phataka, P.; Ghoshb, N.; Yub, H.; Cuib, Q.; Elstner, M. Proc. Natl. Sci. Acad. USA<br />

2008, 105 , 19672–19677.<br />

[94] Kreevoy, M. M.; Liang, T. M. J. Am.Chem. Soc. 1980, 102 , 3315–3322.<br />

[95] Anderson, G. R.; Lippincott, E. R. J. Chem. Phys. 1971, 55 , 4077–4089.<br />

[96] Lin, C.-K.; Theoretical Studies on Spectroscopic Properties of Diamond Nitrogen–<br />

Vacancy Defect Systems; Ph.D. thesis; National Tsing Hua University, Hsinchu,<br />

Taiwan 2008.<br />

[97] Press, W. H.; Flannery, B. P.; Teukolsky, S. A.; Vetterling, W. T.; Numerical Recipes<br />

in C ; 2 nd edition; Cambridge University Press, Cambridge 1994.<br />

[98] Schnakenberg, J.; Algorithmen in der Quantentheorie und der statistischen Physik;<br />

1 st edition; Verlag Zimmermann-Neufang, Ulmen 1995.<br />

[99] Perrin, C. L.; Ohta, B. K. Bioorganic Chemistry 2002, 30 , 3–15.<br />

[100] Gilli, P.; Bertolasi, V.; Pretto, L.; Lycka, A.; Gilli, G. J. Am. Chem. Soc. 2002,<br />

124 , 13554–13567.<br />

[101] Bertolasi, V.; Gilli, P.; Ferretti, V.; Gilli, G. J. Am. Chem. Soc. 1991, 113 , 4917–<br />

4925.<br />

[102] Truhlar, D. G. J. Chem. Educ. 2007, 84 , 781–782.<br />

[103] Jensen, W. B.; Kerber, R. C. J. Chem. Educ. 2006, 83 , 1290–1291.<br />

[104] Kerber, R. C. J. Chem. Educ. 2006, 83 , 223–227.<br />

33


[105] Williams, T.; Kelley, C.; et al.; Gnuplot 4.0 ; Patchlevel 0; 2004.<br />

URL http://www.gnuplot.info/faq/<br />

[106] Becke, A. D. J. Chem. Phys. 1993, 98 , 5648–5652.<br />

[107] Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37 , 785–789.<br />

[108] Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J. J. Phys. Chem.<br />

1994, 98 , 11623–11627.<br />

[109] Hertwig, R. H.; Koch, W. Chem. Phys. Lett. 1997, 268 , 345–351.<br />

[110] Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheese-<br />

man, J. R.; Montgomery, Jr., J. A.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam,<br />

J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani,<br />

G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.;<br />

Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai,<br />

H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Adamo, C.;<br />

Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi,<br />

R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Sal-<br />

vador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain,<br />

M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.;<br />

Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu,<br />

G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.;<br />

Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.;<br />

34


Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A.; Gaussian 03 ;<br />

Revision C.02; Gaussian, Inc., Wallingford CT 2004.<br />

[111] Schmidt, M. W.; Baldridge, K. K.; Boatz, J. A.; Elbert, S. T.; Gordon, M. S.;<br />

Jensen, J. H.; Koseki, S.; Matsunaga, N.; Nguyen, K. A.; Su, S. J.; Windus, T. L.;<br />

Dupuis, M.; Montgomery, J. A.; GAMESS US; Version: 12 Jan. 2009 (R3); Ames<br />

Laboratory, Iowa State University 2009 (J. Comput. Chem. 14 (1993) 1347 - 1363).<br />

[112] Smith, S. W. The scientist and engineer’s guide to digital signal processing.<br />

URL http://www.dspguide.com/<br />

[113] Garcia-Viloca, M.; Gonzalez-Lafont, A.; Lluch, J. M. J. Phys. Chem. A. 1997, 101 ,<br />

3880–3886.<br />

[114] Gilli, P.; Bertolasi, V.; Pretto, L.; Antonov, L.; Gilli, G. J. Am. Chem. Soc. 2005,<br />

127 , 4943–4953.<br />

[115] Reusch, W. Virtual Textbook of Organic Chemistry.<br />

URL http://www.cem.msu.edu/∼reusch/VirtualText/intro1.htm<br />

[116] Falbe, J.; Regitz, M. (editors); Römpp Chemie Lexikon; 9 th edition; Georg Thieme<br />

Verlag, Stuttgart 1995.<br />

[117] Cleland, W. W.; Kreevoy, M. M. Science 1994, 264 , 1887–1890.<br />

[118] Perrin, C. L.; Nielson, J. B. Annual Review of Physical Chemistry 1997, 48 , 511–<br />

544.<br />

35


[119] Ramos, M.; Alkorta, I.; Elguero, J.; Golubev, N. S.; Denisov, G. S.; Benedict, H.;<br />

Limbach, H.-H. J. Phys. Chem. A 1997, 101 , 9791–9800.<br />

[120] Nishihira, J.; Tachikawa, H. Journal of Theoretical Biology 1999, 196 , 513–519.<br />

[121] Grech, E.; Malarski, Z.; Sawka-Dobrowolska, W.; Sobczyk, L. J. Phys. Org. Chem.<br />

1999, 12 , 313–318.<br />

[122] Agrawal, N.; Hong, B.; Mihai, C.; Kohen, A. Biochemistry 2004, 43 , 1998–2006.<br />

[123] Sutcliffe, M. J.; Scrutton, N. S. Trends Biochem Sci. 2000, 25 , 405–408.<br />

[124] Fuhrmann, C.; Daugherty, M.; Agard, D. J. Am. Chem. Soc. 2006, 128 , 9086–9102.<br />

[125] Johnson, S. L.; Rumon, K. A. J. Phys. Chem. 1965, 76 , 74–86.<br />

[126] Malarski, Z.; Rospenk, M.; Sobczyk, L.; Grech, E. J. Phys. Chem. 1982, 86 , 401–<br />

406.<br />

36

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!