_

t ,

r}:
1.

 

 

5 (a Q .5 c C" 3: LIBRARY
Michigan State
University

This is to certify that the
dissertation entitled

PROTON TRANSFER REACTIONS IN DIHYDROGEN
BONDED SYSYEMS

presented by

SIMONA MARINCEAN

has been accepted towards fulfillment
of the requirements for the

Ph.D. degree in ORGANIC CHEMISTRY

 

 

Q” { (gyro/D...

Major Professor’s Signature

08.11.2003

 

Date

MSU is an Affirmative Action/Equal Opportunity Institution

PLACE IN RETURN Box to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/CIRC/DateDue.p65-p. 15

 

 

PROTON TRANSFER REACTIONS IN DIHYDROGEN BONDED
SYSTEMS
By

Simona Marincean

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHYLOSOPHY
Department of Chemistry

2003

ABSTRACT
PROTON TRANSFER REACTIONS IN DIHYDROGEN BONDED
SYSTEMS
By
Simona Marincean
As possible substrates for one-photon infrared-pumped reaction

(IRPR), the structures and decomposition paths of complexes between All-14‘
and three proton donors, H20, HF, and HCl, have been studied by ab initio
methods at the MP2//6-311++G** level of theory. Dihydrogen bonded
complexes, [A1H4...HA]‘ (A=OH, F, Cl), were obtained as minima on the
potential energy surface (PES). In each case only one transition state was
found for the proton transfer and H2 loss process. For each cluster, the
geometry and energy characteristics of reactants, complex, transition state,
and products were analyzed with [AlH4. . .HC1]‘ emerging as the best IRPR
candidate. The calculated intrinsic reaction coordinate (IRC) confirmed the
one-step proton transfer and H2 loss with no intermediate. Classical
trajectories for the vibrationally activated process were calculated on the ab
initio potential energy surface, beginning with varying degrees of excitation

of the modes associated with proton transfer, 1411 and 1973 cm". Proton

transfer from HCl and loss of H2 were calculated to occur on the
femtosecond time scale for several of the initial conditions, including the
first vibrationally excited level.

The influence of the dihydrogen bonding on the activation parameters
of ketone reduction by tetrabutyl ammonium borohydride, NBu4BH4, has
been studied. 2-Hydroxycyclopentanone was chosen as substrate because it
was the most thoroughly studied in the previous works in this group. The
reaction kinetics were found to be first order in ketone and in borohydride
with a rate constant of k=2.20 :t 0.160 x 10'3 l-mol"-s'l at room temperature.
Formation of a dihydrogen bonding complex between the hydroxyl
substituent and the BH,’ lowered the activation enthalpy by 6.6 kcall.
Comparison of the activation entropy for the 2-hydroxycyclopentanone and
cyclopentanone reductions, -29.6 and —22.9 e.u. respectively, pointed to a
higher organization degree in the transition state when the substrate bears an
(Jr-hydroxyl group. When a hydrogen bonding solvent was used the
activation parameters were lowered even more as expected, due to increase
number of proton donor partners. Direct analysis of the products ratio
showed a smaller stereochemical control than reported previously. The

influences of different factors on the stereochemical outcome are discussed.

TABLE OF CONTENTS

LIST OF TABLES .......................................................................................... v

LIST OF FIGURES ........................................................................................ vi

1. Literature Review .................................................................................... l
1.1. Selective Bond Dissociation via Photon Excitation ........................ 1
1.2. IR Laser Pump Spectroscopy in Hydrogen Bonded Systems ........ 12
1.3. Dihydrogen Bonding ..................................................................... 26
1.4. Proton Transfer in Dihydrogen Bonded Systems .......................... 37

2. Dynamic Studies for Proton Transfer in Dihydrogen-bonded

Complexes involving All-14’ .................................................................. 56

3. Kinetic and Mechanistic Investigations of 2-Hydroxycyclopentanone

Reduction by Borohydride .................................................................. 74
4. Methods of Calculation ......................................................................... 9O
5. Experimental Section ............................................................................ 94
BIBLYOGRAPHY ...................................................................................... 98

iv

LIST OF TABLES
Table 5.1 Dihydrogen Bonded Complexes: Geometries ....................... 62
Table 5.2 Energetics of Proton Transfer Reaction for Dihydrogen Bonded
Complexes ............................................................................ 63
Table 6.1. Reported Reduction Results of ot-Hydroxycycloalkanones and of
Corresponding Unsubstituted Ketones ............................................ 76

Table 6.2. Rate Constants for the Reduction of

2-Hydroxycyclopentanone by NBu4BI-I4 ......................................... 81
Table 6.3. Activation Parameters for Ketones Reductions. . . . . . . . . ...........83
Table 6.4 Work-up Procedures Results ........................................... 84

Table 6.5 Trans: Cis Ratios in 2-Hydroxycyclopentanone Reduction ....... 86

LIST OF FIGURES

Figure 2.1 OH Frequencies for EtOH Oligomers in CCl4 ...................... 13
Figure 2.2 Laenen Model for EtOH Excitation ................................. 15
Figure 2.3 Bakker Model for the OH Excited Mode ............................ 20
Figure 2.4 10M NaOD Solution in D20 ........................................... 22
Figure 3.1 Dihydrogen Bonded Complex ........................................ 26
Figure 3.2 Alane Piperidine Adduct ............................................... 29
Figure 3.3 Cyclotrialumazane ...................................................... 30

Figure 3-5 [1r(PMes)4(H)(0H)]+ and [F6(H)2(H2)(PEt2Ph)4l

Complexes ............................................................................. 3 1
Figure 3.6 Bifurcated Hydrogen Bonds in Ir Hydride Complexes ............ 34
Figure 3.7 Dihydrogen Bonds in W Hydrides Complexes ..................... 35
Figure 4.1 NH4 Structures ........................................................... 38
Figure 4.2 Proton Transfer in the H30' Complex ............................... 39

Figure 5.1 [All-I4...HZO]’ Complex Structures: Reactant, Transition State,

and Products .......................................................................... 59
Figure 5.2 [All-I4...HF]' Complex Structures: Reactant, Transition State, and

Products ................................................................................. 60

Figure 5.3 [AlH4...HCl]' Complex Structures: Reactant, Transition State,

and Products .......................................................................... 61

vi

Figure 5.4 IRC Calculation for Proton Transfer and H2 Loss in

[AlH4. . .HCl]' ........................................................................ 66
Figure 5.5 Excited Modes .......................................................... 66
Figure 5.6. Distance Evolution in Vibrationally Excited Complex at 1411
cm”1 ........................................................................................ 68
Figure 5.7 H(1)-H(2) Distance Evolution in Vibrationally Excited Complex
at 1973 cm" .......................................................................... 70
Figure 5.8 Evolution of the 1411 and 1973 cm'1 Modes Upon Transition
State Relaxation ...................................................................... 71

Figure 5.9 Evolution of the Modes Other Than 1411 and 197 3 cm'1 Upon

Transition State Relaxation ......................................................... 72
Figure 6.1 Time Evolution of Ketone Concentration ........................... 80
Figure 6.2 Time Evolution of Borohydride Concentration ..................... 80

Figure 6.3 Integrated Rate Laws for Reduction of

2- Hydroxycylopentanone by NBu4BH4 .......................................... 81
Figure 6.4 Logarithmic Representation of the Eyring Equation .............. 82
Figure 6.5 2-Hydroxycyclopentanone ............................................ 87
Figure 6.6 2-Hydroxycyclopentanone-BH4' Complexes ....................... 87

vii

1. Literature Review

1.1 Selective Bond Dissociation via Photon Excitation

There are two main ways to achieve excitation of molecules in order
to activate them for reaction. The most familiar distributes kinetic energy via
thermal activation raising a fiaction of the molecules above the activation
barrier, as dictated by the Boltzman distribution. The second way is to
supply external energy by a mechanism that is “not subject to the
equilibrium distribution law of canonical ensembles”.l Chemical activation,
single/multi photon excitation, collisions in molecular beams, and
photosensitization reactions can accomplish the latter. This thesis probes the
behavior of such-nonequilibrium excitations in the recently popular
dihydrogen bonded systems.

The notion of selective bond cleavage via vibrational activation
("molecular scissors") has long been recognized as a natural corollary to the
identification of infi'ared absorption bands with molecular fimctional groups.
In principle, deposition of sufficient vibrational energy into a chemical
bond's stretching mode Should induce its rupture, resulting in an
infrared-pumped reaction (IRPR). But many vibrational quanta are needed to

break a typical bond, so this idea was not seriously considered until the

advent of pulsed lasers, light sources intense enough to accomplish rapid
multiphoton excitation.2

Unfortunately, though lasers have been crucial in the development of
today's understanding of molecular dynamics, the main result is the
knowledge that the timescales for multiphoton absorption and, most
importantly, for vibrational relaxation are typically much faster than for any
intended reaction. Early measurements showed that intramolecular
vibrational relaxation takes places on a picosecond time scale.3 The
challenge to those who would design IRPRS is thus clear: the reaction time
must be competitive with the relaxation time of the vibrational mode in
question; the energy deposited must of course be adequate to surmount the
activation barrier (including thermal and tunneling contributions) for the
process; and the overall reaction should be irreversible or at least exothermic
(i.e. the back reaction should have a high barrier).

Zewail pointed out that vibrational relaxation takes place via two
modes: energy dedistribution and phase loss, and it is important to know
their respective contributions to the apparent relaxation processes as criteria
for tuning of laser pulse shape and duration.2 Experiments done by exciting

C-H stretches in naphthalene found relaxation times under around 0.1 ps

pointing up the need for subpicosecond lasers.3 Zewail also made the point

of differentiating between the vibrational modes of interest, i.e. the ones to
be excited, and the other modes that are not excited directly but interact with
the first ones and that can function as energy Sinks.

Early attempts of laser control of reactions are characterized by
multiphoton excitation. King et al. induced reactions in tetrazine in neat and
mixed crystals at low temperature (Scheme 1.1). Isotopic composition in the
products was controlled by tuning the laser to absorption band of the

molecule with a specific isotope component.4

Scheme 1.1
N/\N h
I I .—.»
N\/N

The reaction takes place from the electronic excited states and exhibits
molecular selectivity but showed that lasers could drive reactions.
Multiphoton excitation of terminal and non-terminal olefinic C-H
stretches in CH2=CH-CH2-NC induced isomerization,5 which was explained
by a reaction time much faster than IVR in this particular case (Scheme 1.2).

Scheme 1.2

CH2=CH-CH2-NC ‘9 CH2=CH-CH2-CN

Isomerization and fragmentation were obtained via multiphoton excitations
of the CH2 stretch and wag vibrations in cyclopropane in the gas phase.6
Products were the same as resulted fi'om thermal excitation (Scheme 1.3) but
their distribution was different. Excitation of the CH2 wag favored the
isomerization path.

Scheme 1.3

A ——> CH2=CH-CH3

A ——> CH2=CH2 + ICH2

In a nice review about IR-laser chemistry,7 Quack et al., referring to
mode selective excitation, cautioned that different reactive populations will
be induced by excitation via laser as opposed to chemical or thermal
activation or by different excitation frequencies and therefore different
products distributions which are “no proof whatsoever for mode selective
reaction control”. In an attempt to use different chromophores to probe the
mode selective excitation they used F-CHz-CHz-CHz-CHD-F but no mode

selectivity was observed (Scheme 1.4).

Scheme 1.4

nhv1 DhV2
F< CH2 CHzD
CH2/ \CH2/ \F

1

HF + CH2=CH-CH2-CHDF HF + FCH=CH-CH2-CHD

Among theoretical approaches to actively control product formation in
a reaction, of major impact are work done by Rice et al., Brumer and
Shapiro, and Rabitz et al. Rice suggested that selection control of the
products in a branching reaction could be achieved by transfer of vibrational
amplitude from ground state to an excited state and, alter a specific time
delay, back to ground state. Gerber et al. applied this formalism for the
ionization and dissociative ionization of Na; and demonstrated control of
different product channels, (Scheme 1.5), by choosing the selective delay
time between pump and control laser pulses.8

Scheme 1.5
Na; -—> N32. —* Na;+ + e'
—> Na+ Na+ + e'

An improved version, the Tannor-Rice method, used the phase to

control the reaction outcome in addition to the original pump-dump delay. If

previously the transfer from the excited state to the ground state was not
affected by phase, now the phase was locked to that of the pump pulse and
therefore interference occurred. Scherer studied this method experimentally
in generating fluorescence—detected dynamics of 12.9

Brummer and Shapiro used an approach that relied on the fact that a
molecule at an energy where dissociation is possible is characterized by a
continuum of degenerate eigenstatesm'12 Control over the probability of a
desired product quantum state was obtained by producing a linear
combination of these degenerate states. They chose to take advantage of the
interference between amplitudes of different states that connect the reactants
and products. Gordon et al. applied this method in studying the ionization
and dissociation of HI and DI by one and three-photon excitation and they
showed that the phase difference between the ionization and dissociation
products is dependent on the excitation energy.” '4

A more sophisticated version of this approach, called “incoherent
interference”, considers that a single field transfers population fi'om ground
state to the degenerate target states.15 ’ '6 A second pulse will connect the final

states to another empty state. Provided the fields are strong enough they will

interfere and control can be obtained by tuning the interference.

Brumer and Shapiro found out that efficiency is at a maximum when the
pulse relative to the target states is applied before the excitation pulse but
overlapping with it. This method is called stimulated Raman adiabatic
passage (STIRAP) and Kobrak and Rice developed an extended version of it
that allows for product selectivity in a chemical reaction.“ ‘8 Experimental
work to confirm this approach has been done by Shnitman in proving laser
control over a branching photochemical dissociation of Naz (Scheme 1.6).
Control was demonstrated over the Na(3 d), Na(3p) branching ratio by using
a two-photon incoherent interference scenario based on pump and dump
19,20

laser pulses with variable frequencies.

Scheme 1.6
Na2—> Na (35) + Na (3p)

-—> Na (38) + Na (48)
'—> Na (38) + Na (3d)

Recently Abrashkevich et al. studied computationally the possibility
of control over dissociation of collinear CHzBrI to yield either CHzBr +1 or
CH21 + Br. The authors reported that the control level depends strongly on
the superposition created by the first step excitation that was achieved with

experimentally controllable laser amplitudes and phases.21

The Hamiltonian at the Born-Oppenheimer level remains unknown for
polyatomic molecules. A method of closed loop control introduced by
Rabitz offers a way to surmount this problem in finding the appropriate laser
pulse. This theoretical concept is denoted by optimal control theory,
(OCT).21'23 The goal of OCT is to design an electric field that allows for
desired manipulation of the system. An important feature is the competition
between the maximization of the expectation value of an observable and the
constraints imposed by the electric field. All the objectives are included in
the cost functional which is optimized with respect to the electric field in
such a way that it yields best possible control performance. Although the
OCT simulations had to be carried out on simple systems due to
computational cost, they showed that successfirl control fields are possible.

Related to OCT are optimal control experiments, OCE, based on
adaptive control techniques and Rabitz et al. have suggested them.”30
In this approach a loop is closed around the system, observables are used to
refine the candidate laser field applied until the control objective is achieved
as best possible. The elements involved are: an input laser design, the
apparatus needed to generate the laser pulses, application of the laser field to
the quantum system, observation of the control outcome, and a learning

algorithm that analyzes the results of the prior experiments and suggests a

new control field. The advantage by comparison with OCT resides in the
ability of the system to “solve its own Schrodinger equation”.

Selective bond dissociation has been achieved in a series of ketones.31
Starting out with acetone it has been demonstrated that optimizations to the
point where one pathway can be favored over the other can be achieved in a
reasonable laboratory time, 10 min, and also that strong fields are able to
alter the relative ion yields and therefore the information content in the mass
spectrum which is used to measure the control observable. The control over
selective bond cleavage was studied with trifluoroacetone and acetophenone.
In the first case the dissociation pathways are not identical, (Scheme 1.8)
and indeed CF 3 signal was enhanced compared to the initial one but not at
the expense of suppressing other pathways, i.e. the CH3 branch.

Scheme 1.8

CH3-CO-CF 3 —" CH3-CO + CF3
—’ CF3-CO + CH3

For acetophenone one pathway was favored over another in contrast with the
dissociation energy (preferential cleavage of the phenyl ring was obtained
even though the energy required to cleave the Ph-CO bond is 15 kcal /mol
higher than that for CH3-CO bond). Also a rearrangement pathway was

observed and limited control over it was achieved (c, Scheme 1.9).

Scheme 1.9

O

0
0:0

CH3
—-> + CH3

CH3

Using the same method Gerber showed control over different
dissociation pathways in organometallic compounds.32 Gerber demonstrated
the possibility of adaptive control in the liquid phase when
4-dicyanomethylene-2-methyl-6-p-dimethylaminostyryl-4H-pyran was
excited selectively over [Ru(dpb)3](PF6)2 (dpb is 4,4’-diphenyl-2,2’-
bipyridine) when both compounds were dissolved in methanol.33

All the control experiment approaches are based on multiphoton laser
excitation in the UVE domain. As one can see from the examples given the
reactions that are subjected to investigation are unimolecular. The
limitations arise from that fact that UV spectroscopy is not functionality

specific and also the majority of reactions of interest for chemists are at least

10

bimolecular. In bimolecular reactions two features add to the complexity of
the problem: orientation of the reactants, and collisions.34

Althorpe et al. studied the mechanism of the D2 + H —> D + HD
reaction, using multiphoton excitation. They reported two mechanisms to
take place, the dominant one is characterized by scattering of the newly
formed HD molecule in the opposite direction than the initial approach of
the H atom. The second mechanism, which is observed with a 25 fs delay,

ejects the HD molecule in the same direction as the initial approach.35

11

1.2. IR Laser Pump Spectroscopy in Hydrogen Bonded Systems

Infrared laser Spectroscopy allows for time-resolved studies of
vibrational dynamics. General principle is to excite with a laser pulse the
bond of interest and with a second delayed pulse to probe the dynamics of
the system. Hydrogen bonded complexes have been studied due to their
importance in modern chemistry and biology and because little is known
about their structure and dynamics.

Heiweil et al. were the first ones to use IR pump-probe Spectroscopy
investigate monomeric EtOH in CCl4, (<0.007 mole fraction). They
bleached the OH frequency and monitored its recovery with a 20 ps laser
pulse tuned to 3625 cm'I and reported a dissociation time of 70 :1: 10 ps.36

Graner, Ye and Laubereau performed classical IR spectroscopy on
different concentration solutions of EtOH in CCl4, and they concluded that
internal hydrogen bonds were responsible for a peak at 3300 cm".37'39
Excitation of a 0.42 M solution with a 10 ps laser at this frequency followed
by a probe pulse at different delay times allowed for study of hydrogen
bonds dynamics. The excitation energy exceeded the strength of the
hydrogen bond in EtOH, which was determined to be 6 kcal/mol,4°’ 4‘ and

thus cleavage was achieved. The dissociation was faster than the duration of

the laser pulse and therefore its time constant could not be determined while

12

the recovery of the hydrogen bonds had a time constant of 20 :h 5 ps. Careful
examination of the signal curve and of the transmission increase led the
authors to conclude that the absorption onset is delayed by 5 i 3 ps, which
they labeled “predissociation lifetime”, i.e. lifetime of the excited molecules
before dissociation. Cleavage of hydrogen bonds was considered to be an
efficient relaxation channel.

Hole-burning spectroscopy is techniques that typically bleaches a
spectral band of a molecule and provides information about the vibrational
levels of the molecule. IR hole-burning spectroscopy started with the
availability of the IR lasers. Laenen and Labereau used this technique to
study EtOH solutions in CCl4 using a laser with an excitation pulse of 2 ps
and a probe pulse of 1 ps.42 At 1.2 M concentration of EtOH, Oligomers

present are characterized as expected by specific peaks, Figure 2.1.

Monomer Dimer Trimer
H Et Et\ Et\
Et—OT \OTH-u-O H O "-70 H----O H
Va Vb Va Vb Vc Va
v, = 3630 cm" v, = 3500 cm" v. = 3450 cm“

Figure 2.1 OH Frequencies for EtOH Oligomers in CCl,

l3

EtOH trimers, tetramers/pentamers, and larger polymers are characterized by
stretching vibrations of internal OH groups at 3450, 3330, and 3240 cm-1,
respectively.43

Upon excitation of the internal OH groups, Laenen et al. demonstrated
transient hole burning with a lifetime of 1 ps, estimated from relative
transmission change as a function of the delay time. The spectral holes were
attributed to transitions from ground state (0) to the first excited vibrational
level (1). Bleaching that was observed at early delay times, -2 ps, persisted
for long times, >80 ps, and was considered evidence for rapid spectral
relaxation and/or vibrational energy redistribution of the OH groups. Also
induced absorption was reported on the red excitation frequency at early
delay times as a result of excited state absorption (ESA) from first to second
excited levels of the OH-stretching mode. At longer delay time, 5 ps, an
induced absorption in the blue wing of the excitation frequency illustrated
the formation of smaller hydrogen bonded Oligomers as a result of hydrogen
bonds breaking. In order to deduce quantitative information from the

measured spectra a five states model was proposed, Figure 2.2.

14

En rgy
oligomeric broken
subensemble 2 Oligomers Oligomers

 

Figure 2.2 Laenen Model for EtOH Excitation
The short intense pulse at 3330 cm", indicated by the thick arrow, produced
transitions to the first excited vibrational level (state 1), which was observed
as a transient spectral hole. The excited subensemble of Oligomers relaxed
via spectral redistribution and/or energy transfer to a larger ensemble while
maintaining the population of the vibrational excited state (state 2). Because
the excitation energy exceeded the hydrogen bond energy, dissociation could
be considered an important decay channel for the vibrational excitation with
formation of smaller Oligomers observed in the spectrum, (state 3).
Reassociation involved local temperature changes leading to a
nonequilibrum distribution of hydrogen bonds at a higher temperature, (state
4), followed by therrnalization, (state 5). The probing transitions are
indicated by thin arrows and the relaxation pathways by dashed arrows. BSA
was considered a direct measure of the excited state population, generated

by the pump pulse. The time constant deduced from the corresponding

15

signal transient (relative transmission change as function of delay time for
pulse at 3330 cm'1 and probe at 3100 cm”) was consequently interpreted as
the population lifetime and determined to be 1.4 d: 0.3 ps. The difference
between the hole and the population lifetime was explained as a result of
structural changes around the individual excited OH groups and /or
vibrational energy transfer to neighboring EtOH molecules. Computations of
transient spectra for different delay times using the proposed model and
comparison with the experimental results led to a dissociation time of 2.0 ps.
The reassociation was much slower, 10 ps.

At a lower concentration, 0.17 M EtOH in CC14, new features about
the hydrogen bond dynamics were observedf”4 At this concentration,
monomer and dimer bands are observed at 3633 and 3500 cm", respectively.
Excitation in the same oligomer region, at 3350 cm", induced again transient
holes with a lifetime of 1.2 ps and BSA with 1.7 ps. Spectral relaxation via
energy migration along the hydrogen bonded chain was found to be fast.
Breaking of hydrogen bonds was a very effective relaxation channel with a
quantum yield of 0.9. Reassociation occurred on a much longer scale, 23 ps.
As expected at 0.05 M EtOH in CCl4, when only monomers are observed,
the lifetime of the excited state is 8 ps,45 five times longer than the one

corresponding to 0.17 M solution.

16

Bakker et al. also investigated EtOH solution in CC14 with a
concentration of 1.2 M, using a femtosecond laser (100 fs, 1 mJ).46 Upon
excitation of the oligomer band at 3330 cm”, they noted that the relaxation
had two components: a fast one, with a constant of 250 fs, and a Slower
decay at a transmission value that was slightly higher then the initial one,
with a time constant of 15 ps. Excitation at different frequencies showed the
same pattern with an increase of the predissociation lifetime up to 900 fs
when the excited fiequency was 3450 cm". This dependence on excitation
frequency proved the inhomogeneous character of hydrogen bonds and also
that the predissociation rate is influenced by the hydrogen-bond strength.
In the model that the authors proposed for the process, the excitation of the
O-H stretch vibration is followed by a fast delocalization over the hydrogen
bonded oligomer and breaking of hydrogen bonds.

Early studies of MeOH Oligomers in a N2 matrix showed that upon
irradiation in the mid-IR cyclic trimers could be transformed in open chains
and also that this conversion could be reversed by increasing the
temperature.47 Also work done by Bakker et al. on both MeOH dissolved in
CCl4 and MeOH in zeolites, showed that excitation of internal OH stretch
vibrations led to dissociation that was faster then the duration of the laser

pulse (18 ps).48 Calculations to fit the data led to time constants of 0.5 and

17

10 ps for solution and zeolites, respectively. The reassociation time for both
cases was so much larger that in the zeolites it was outside the
experimentally measurable range. The argument was that the solvent keeps
the dissociated partners in close proximity by a cage effect while in zeolites
that arrangement is not likely since in the zeolites experiments the methanol
concentration was kept well below complete saturation of the zeolite.
Applying a new technique, 3D IR-Raman spectroscopy, Dlott et al.49
investigated the vibrational energy relaxation in neat methanol at room
temperature. 3D IR-Raman spectroscopy uses resonant vibrational pumping
by a tunable mid-IR pulse and incoherent anti-Stokes Raman probing. The
first two dimensions were represented by a series of incoherent anti-Stokes
Raman spectra at a given mid-IR pump frequency, and the third dimension
involved changing the IR pump frequency within the range of CH and OH
stretching transitions. The authors reported no observation of nonassociated
MeOH transition created by vibrational predissociation as a result of OH
stretch pumping, but they stressed the point that the experiments were done
on neat MeOH in which the molecules are involved in large hydrogen-
bonded networks and essentially all molecules have multiple hydrogen

bonds. The instantaneous generation of OH bending (1380 cm") transients

when the OH stretch (3250 cm") was pumped was considered evidence of

18

vibrational predissociation. Interestingly they also noted that even though
the OH mode is not directly coupled to the C-H stretch, excitation of the
methyl-rocking mode together with a significant population of the C-0
stretch mode is present.

In contrast with Bakker’s and Laenen’s work, the Fayer group showed
that excitation of OD non-donating groups in MeOH dissolved in CC14 with
a laser pulse tuned to 2690 cm'1 led to dissociation with a time constant of 2
ps proving that relaxation does not have to be necessarily in the excited
O-H...O stretch.50 More work on deuterated alcohols, MeOD, EtOD, and
PrOD, from the same group led to dissociation times of 2-3 ps upon
excitation of non-donating OD groups. The mechanism was rationalized in
terms of sequential hydrogen bond dissociation via intramolecular energy
redistribution followed by energy transfer into the intermolecular modes.5 ‘
When they measured the OD relaxation of groups that both donate and
accept hydrogen bonds in a MeOD solution in CCl4 both mechanisms were
observed. Of interest is that in isotopically mixed solution the mechanism
proposed by Fayer et al. is not observed and the explanation was that energy
redistribution over the oligomer led to OH hydrogen bond rather than OD

hydrogen bond breaking.52

l9

If one wants to study hydrogen bonds in their most important context,
H20 would seem the system of choice. The Bakker group investigated the
vibrational dynamics of water using a femtosecond IR pump-probe laser
system.53 Extensive studies of OH stretch in HDO dissolved in D20 brought
evidence for a Stokes shift, i.e. upon excitation a peak was observed at a
frequency lower than the incident frequency, with a lifetime of 740 fs. In
describing the data quantitatively the hydrogen bond was assumed to be a

Brownian oscillator coupled to the excited OH mode, Figure 2.3.

A v=2

  
   

 
    
 

-------—---,

Free Energy

 
 

 

add

 

 

R(O-H...O)

Figure 2.3 Bakker Model for the OH Excited Mode

20

The Brownian oscillator model is characterized by harmonic potential
energy functions for the hydrogen bond mode, which are displaced by a
distance d with respect to each other in the ground and excited states of the
OH-stretching mode. Also because the OH stretch absorption band had an
approximately Gaussian shape and was broad, the authors assumed the
hydrogen bond mode to be strongly overdamped, which implies that the
wavefunctions associated with the hydrogen bond vibrations are strongly
coupled to each other and to other low-frequency liquid modes. The
limitations occurred from the not so good description of the spectral
relaxation at large delay times. Using the same model Kropman et al. studied
the OD stretch of HDO dissolved in H20.54 The unexpectedly long
vibrational lifetime for OD, 1.8 ps, twice as much as the correspondingone
for OH, prOmpted the authors to suggest that despite the fact that the energy
gap law would predict the opposite, the number of accessible modes is
probably larger for OH.

Using the same laser system, Bakker et al. studied the dynamic
behavior of the OH stretch modes in HDO molecules dissolved in a 10 M
NaOD solution in D2055 In the IR Spectrum of this solution three OH

groups could be distinguished, Figure 2.4.

21

\ / \ \
D ’D D Q Q
‘s I' ‘ /D e lo
,o—H-mo ’O-I-l----,O-D----O\ ,o-o----,o—H----o\
,o o o o o o
x o D
o-o / /
\ O O
D \ \
D D
a. neutral DOH/020 b. DOH solvating 00' c. HO' solvated by
cluster 020

Figure 2.4 10M NaOD Solution in D20

In the range 3200-3 500 cm'1 are the majority of HDO molecules,
which donate hydrogen bonds to other D20 or HDO molecules, Figure 2.4,
a. At lower frequencies are HDO molecules that form strong hydrogen
bonds with 0D, Figure 2.4, b. Finally, resonances at 3600 cm'1 were
attributed to HO' ions, Figure 2.4, c. The excitation was applied in all three
regions and it was found that persistent spectral holes could be observed in
the OH stretch in all three environments, evidence for a more rigid hydrogen
bonded network than that of pure liquid water, for which rapid broadening of
the hole was observed, with a lifetime of 600 fs.

The behavior upon excitation was rationalized using the brownian
oscillator model again. Two components were distinguished. The first
component is due to the HDO molecules that are bonded to D20 as
DO-H...OD2, which have a shorter vibrational lifetime than the HDO

molecules in “pure” base free liquid, i.e. no D0' or HO' present, only HDO

22

and D20, 600 fs compared to 740 fs. The second component corresponds to
HDO molecules that are hydrogen bonded to D20 or 0D and that
participate in proton transfers between HDO and 0D or deuteron transfers
between OD' and D20. The lifetime of vibrational relaxation for this
component is ~ 160 fs and was attributed by the authors to the time scale of
deuteron hopping. The large difference between the time scales of these two
components led to the conclusion that the complexes that generate them are
localized in regions of hydrogen bonded network and that they are not
interchangeable.

All the work in the Bakker group was rationalized in the light of the
brownian oscillator model, which assumes that the motion of the OH is
strongly damped due to strong coupling among the hydrogen bonding and
liquid modes and thus the dynamics are based on a hydrogen bond length
value resulting from interactions with the surrounding liquid. Based on
thorough molecular dynamic simulation studies of the HOD/D20 system,
Hynes demonstrated that there is not a simple linear dependence between the
frequency of the OH and the length of the hydrogen bond in which that OH
is involved and thus interpretation of the experimental data should be based

on a range of distances.56

23

Examples of ionic hydrogen bonded systems studied by IR pump-
probe spectroscopy are few and the most representative is the work done is
in the Strauss group on Tutton salts, (NH4)2[Cu(HzO)6](SO4)2 and
(NH4)2[CoxNi1-x(HzO)6(SO4)2, in solid the state at low temperature.“ 58
Upon excitation the hydrogen bonds were broken due to reorientation. The
lifetime of burning holes was estimated to be on minute time scale. The
temperature independence of rate constants prompted the authors to suggest
the occurrence of tunneling.

Another studied ionic system is gas-phase clusters,Cl(HzO)n(CCl4)m, in
which excitation led to loss of a H20 molecule if m=0 and CCl4 if there is
one CCl4 molecule.59

Recently the Baker group investigated the vibrational dynamics of
hydrogen-bonded HCl-diethyl ether complexes.62 Upon excitation of the H-
Cl stretching mode, 2500 cm”, with a 500 fs laser the relaxation was found
to take place in two steps. The first step was assigned to the vibrational
energy transfer to the hydrogen bond mode with a time constant of 0.9 i 0.2
ps, causing a shift in the HCl stretching mode absorption. The second step
corresponded to the relaxation of the hydrogen bond with a time constant of
3.1 i 0.5 ps. The authors also reported no evidence for a proton transfer

from the HCl to the ether, i.e red shift due to formation of a vibration lower

24

in energy than the HCl stretch, due probably to the fact that the energy is
rapidly transferred to the hydrogen bonding mode.

Complexes in which the hydrogen bond acceptors are different
molecules were investigated by SchriVer et al. for HI with different partners:
water, acetylene, ethylene oxide, dimethyl ether, and acetone in matrix.6°’ 6'
Photoirradiation led to cleavage of the hydrogen bonds. Since the laser only
accessed frequencies in the range 1600-2500 cm", excitation of CH3

followed by energy redistribution and proton transfer were proposed but no

accurate evidence could be found.

25

1.3. Dihydrogen Bonding

Dihydrogen bonding has been in recent years the subject of extensive
experimental and theoretical research. It can be defined as an attractive

interaction between the hydrogen of a proton donor, A-H (A=O, N, X) and

the o M-H bond, where M is less electronegative than hydrogen,Figure 3.1.

Figure 3.1 Dihydrogen Bonded Complex

With strengths and directionality comparable with those found in
traditional hydrogen bonds, dihydrogen bonds play an important role in solid
and solution chemistry, controlling structure, reactivity and selectivity.
Possible applications include stereochemical control, catalysis, crystal
engineering and material chemistry and numerous reviews have summarized
the achievements63 with the most recent one from the Epstein group.64

1.3.1 Dihydrogen Bonding in Main Group Hydrides

The ability of boron hydrides to serve as proton acceptors in hydrogen
bonds was observed by M. P. Brown et al. in 1968. Variable temperature
infrared spectroscopy studies of LBH3 (L=Me3N, Et3N, Py, Et3P) and

Me3NBH2X (X=Cl, Br, I) in the presence of proton donors such as methanol,

26

propanol, and p-fluorophenol in CCl4 led to values for the association energy
in the range 1.7-3.5 kcal/mol. Based on the perturbation of N-H stretch with
temperature and concentration, they also suggested 3 NH. . .H3B interaction
in MezNI-IBH3 and (RNI-IBH2)3 (R=Pr, Butt“57

Epstein et al. studied NEt3BH3, P(OEt)3BH3, and Bu4NBI-I4 in the
presence of different proton donors by solution NMR and IR Spectroscopy
and they noted the occurrence of O-H...H-B interactions with properties
similar to those found in classical hydrogen bonds.“69 The association
energies increased with the proton donors’ acidities, falling in the range 1.1-
3.7 and 2.3-6.5 kcal/mo], for neutral and ionic hydrogen bonds,
respectively.70

In the Jackson group the X-ray and neutron crystal structures of
NaBI-I4-HzO and NaBD4-D20 were determined and O-H...H-B dihydrogen
bonding was observed with three H...H close contacts, 1.79, 1.86, and 1.94
A. Also the O-H vectors point toward the middle of the B-H bonds,
suggesting association with the o-bond electrons, rather than B or H atoms.71

Crabtree et al. searched the Cambridge Structural Database (CSD) for
close intermolecular B-H. . .H—N contacts of which 26 were found in the
range 1.7-2.6 A. These compounds have a bent geometry with NH. . .H-B

angles in the range of 95-120°.72 The authors suggested that this structure is

27

due to the presence of negative charges on both B and H atoms so that the
bending allows close proximity between protonic NH and B with maximum
attractive electrostatic interaction. Theoretical study on a NH3BH3 dimer, at
the B3LYP/PCI80 level of theory, confirmed two H...H interactions, at 1.82
A distance, and 98.8°NH...H-B angle, falling in the range found by the CSD
search. The association energy was predicted to be 6.1 kcal/mol per
dihydrogen bond, which could be responsible for the difference in melting
points between the isoelectronic borane (+104 °C) and ethane (-181 °C).”
In a subsequent paper on NH3BH3, the geometry found by Cramer and
Gladfelter, at the MPZ/cc-pVDZ level, has bifurcated dihydrogen bonds,
with H...H distance of 1.99 A, NH...H-B angle of 886°, and association
energy of 15.1 kcal/mol.73 The X-ray structure of (NH3BH3)2 dimer reported
by Crabtree et al. shows three intermolecular H...H interactions, the shortest
at 2.02 A, and NH...H-B angle of 106°.74 The N-H vectors point toward the
middle of the 8-H bonds , suggesting again that the interaction is between

the proton and the o-bond electrons. The N-H. . .H-B dihydrogen bonding

interaction was characterized was also characterized in compounds other

74, 75

than aminoborane by crystallographic methods and by ab initio

calculations.”79

28

The hydrides of Ga and Al, elements in the same group as B, are also
capable of forming dihydrogen bonded complexes. Intramolecular N-H. . .H-
Al close contact was observed in an alane piperidine adduct by Raston et al
with a H. . .H distance of 2.31 A, Figure 3.2, in an eclipsed conformation.80
The authors noted that the structure undergoes H2 loss to form an

amidoalane.

 

Figure 3.2 Alane Piperidine Adduct
Ab initio studies by Cramer and Gladfelter for (NH3A1H3)2 dimer, at
the MP2/cc-pVDZ level, Showed two intermolecular H. . .11 contacts at 1.78
A, and an interaction energy of 6 kcal/mol per dihydrogen bond.“
Cyclotrialumazane, [(NH2A1H2)3]2 dimer has been found to have a C3,,

symmetry with 6 H. . .H contacts and dimerization enthalpy of -9.9

kcal/mol, Figure 3.3.83

29

‘ \ ‘\ \I
H “.H. H
i H “H l4
: H7<HFNJ7AIH
- H

Figure 3.3 Cyclotrialumazane
The ability of Ga to participate in dihydrogen bonds was confirmed by
the neutron diffraction crystal structure of cyclotrigallazane.81 Theoretically
the strength of the dihydrogen bond was predicted to be 3 and 5

kcal/mol/bond for [(NHzGaI-Iz)3]2 and (NH3GaH3)2 respectively.

Extensive ab initio theoretical work investigated the occurrence of

dihydrogen bonds in complexes involving other main groups hydrides.”89
Based on thorough theoretical investigation of complexes between group I,

II, and IV hydrides with HF, Grabowski demonstrated correlation between

H. . .H distances and dihydrogen bonding energies.” 89

1.3.2 Dihydrogen Bonding involving Transition Metal Hydrides
In 1990, Koetzle et al. reported the neutron diffraction crystal

structure of [Ir(PMe3)4(H)(OH)]+ in which a H. . .H contact of 2.4 A and an

Ir-O-H angle of 104° suggested an attractive interaction, Figure 3.5.90

30

 

 

 

 

L = PMe; L = PEtzH

Figure 3.5 [Ir(PMe3.)..(H)(OH)]+ and [Fe(I-I)2(I-Iz)(PEt2Ph)4] Complexes
A shorter H. . .H distance, 1.862 A, was observed in [Fe(H)2(H2)(PEt2Ph)4]
and prompted the authors to suggest an intramolecular hydrogen bond,
Figure 3.5.9]

The Morris and Crabtree groups independently discovered the first
example of intramolecular dihydrogen bonding in Ir hydride complexes.
Crabtree92 showed evidence for Ir-H. . .H-O interaction facilitated by

tautomerization of the ligand into an iminol, 1.

 

X=PPh3

31

The calculated H. . .H distance was 1.8 A. The coupling between the
Ir-H and O-H hydrogens, 3 Hz, led to the conclusion that the interaction had
a covalent character. A higher coupling, 5.5-5.6 Hz, was found when the
ligand was pyridine93 (Y=H, Cl, Br, I), in which the H...H distance was

calculated to be 1.7 A for Y = Cl, 2.

PhaF},’

O

Y——lr-—H"'
,lfiaph3

\

/
N ,0

2

In order to estimate the dihydrogen bond strength, Crabtree measured
the rotation barrier for the NH; group in a series of 2-aminopyridine
complexes93 (3), which represents the sum of the intrinsic rotation barrier
and dihydrogen bond. The strongest dihydrogen bond was found to be 5.0
kcal/mol when Y=H. Also the Y ligand weakens the interaction in the order
F>Cl>Br>I>CN>CO>H which correlates with electronegativity decrease.
Thus, as the hydride becomes more negative, the NH...HIr interaction

becomes stronger.

32

\

/
N N
Phael .4
Y—lr—H”
,L‘Pph,
3

Ir-H. . .H—N interactions were demonstrated by the Morris group94 via
X-ray diffraction and solution NMR for an Ir hydride complex, The H. . .H
close contact distances, calculated from relaxation times, were 1.75A.
Theoretical calculations by Hoffmann et al. on a model compound, L=PH3,
confirmed the interaction and concluded that its nature is mostly

electrostatic.82

The interaction was turned off by the presence of a competing hydrogen

bonding solvent, tetrahydrofuran.94 Morris et al. also showed evidence for

33

occurrence of bifurcated dihydrogen bonds with estimated distances of 1.8

and 1.86 A (Figure 3.15).95

Tcya _-I+ PCy3 HN_ |+2
\\\\s |\\\\S

N'H )s‘ .
H-l-N/_ s‘/H-N
\
/NH

Figure 3.6 Bifurcated Dihydrogen Bonds in Ir Hydride Complexes
The first example of intermolecular dihydrogen bonding involving a
transition metal hydride was reported by Crabtree et al. for a Re

polyhydride, 5, based on compound’s X-Ray structure.96

45 \ '
L H \HH
"PPh3
5

The geometry and energetic characteristics corresponded to a dihydrogen

bond and were confirmed by theoretical calculations on model compound

34

[ReH5(PH3)3]-NH3.97’ 98 The intermolecular A-H. . .H-Re were reported in
complexes between imidazole, 2,4,6-Me3C6H50H, 2-tBu-6-MeC6H30H,
pyrrole, Ph-NH—Ph, Ph-NH-Bu, Ph-NH-Me and ReH5(PPh3)2 with
interaction energies in the range 3.0-5.8 kcal/mol.99 Substitution of
ReH5(PPh3)2 with WH4(PMePh2)4 allowed Crabtree to demonstrate the
ability of W to participate in this type of interaction. Re-H. . .H—N
interactions were studied in solution by Crabtree using UV-VIS
spectroscopy.loo

Shubina et al. explored the dihydrogen bonding in solution between

tungsten hydrides and alcohols (Figure 3.12).

L

I goo
ON—W—H-----H—OR
OC/ 1'.

L = PMea, PEta, P(i-Pr0)3. PPh3
R = Ph, CH(CF3)2, C(CF3)3

Figure 3.7 Dihydrogen Bonds in W Hydrides
Using IR and NMR spectroscopy they excluded the hydrogen bonding to CO
and NO groups and demonstrated occurrence of the O-H. . .W-H interactions.
The association energies fall in the range of expected dihydrogen bonds and

correlate to the acidity of the proton donors. They proposed a linear

35

OH. . .H-W orientation in contrast to the previously found bent geometry of

dihydrogen bonds.101

The same group reported the occurrence of intermolecular O-H.. .H-
Re in solution and theoretically.'°2’ '03 The ability of the solvent to interact
via .11.. .H-0 and the bulk of the ligand appear to influence the dihydrogen
bonding. The Scheiner group reported elegant theoretical studies on Mo and
W hydrides complexes, 6, and confirmed the correlation between the

dihydrogen bond strength and the acidity of the proton donor or the donating

ability of the cis ligand.104

L
gm

|
L'—M—H
oc’l “HA

\

M = Mo, W

L = PH3, NH3

LfiNchi

A = F, 0H, H20
6

When the proton donor is H3O”: proton transfer takes place with the

formation of an nz—Hz dihydrogen complex. All the optimized complexes

have strongly bent geometry.

%

1.4. Proton Transfer in Dihydrogen Bonded Systems

Dihydrogen bonds have been proven to have strength and
directionality comparable with conventional hydrogen bonds. What makes
them unique is their ability to undergo proton transfer and H2 loss with or
without participation of an nz-Hz intermediate (Scheme 4.1).

Scheme 4.1

+

H _ 'Hz
rim-1 A ——-M_A
H

 

M—H + H—A —->M—H""H—A

 

 

The simplest systems to be studied theoretically are the ones in which
the proton acceptor is H', the hydride itself. Ab initio calculations by Cremer
and Kraka, at the MPZ/6-31+G* level on NIH and H30' predicted
association energies of 9.1 and 18.7 kcal/mol, respectively, in complexes
with geometries consistent with dihydrogen bonded complexes. The proton
transfer barriers were 10.4 and 7.0 kcal/mol, respectively. In the case of
NH; a minimum with tetrahedral geometry was also optimized, higher in
energy than the dihydrogen bonded structure by 1.3 kcal/mol.105 This entity
should be envisioned as a Rydberg system.

The Gordon group investigated the NH; ion at the MP2/6-311++G**

level and reported geometries similar with the ones found by Cremer and

37

Kraka.106 However the energy of the tetrahedral structure was estimated to
be 10.0 kcal/mo] higher (Figure 4.1).

Figure 4.1 NH,’ Structures

H
‘‘‘‘‘‘‘ H_N'0u””H Hllllllo-N
H \H H, \H
dihydrogen bonded complex tetrahedral structure

Intrinsic reaction coordinate calculations showed that this structure is not
connected to the transition state or the products of the proton transfer
reaction. The authors performed a dynamic reaction coordinate calculation
on the tetrahedral structure and they noted that proton transfer took place
upon excitation with 32.5 kcal/mol. It should be noted that the tetrahedral
structure is higher in energy than the transition state for proton transfer.
Since its first mass spectrometric observation in 1982 by Paulson and
Henchman,107 the H30' ion has received considerable attention.
In a mass spectrometric study by Miller et al., rate constants for formation of
the H30‘ anion and its reaction with a large number of reagents have been
measured. The association energy was found to be 14.4 i: 1.0 kcal/mol and
dissociation of H30' into HO' and H2 required 4.5 d: 1.0 kcal/mol.108
Investigation of the HO +H2 —> H20 +H reaction via photodetachment

Spectroscopy of H30, paralleled by ab initio calculations, led to values for

38

the dihydrogen bond and barrier for proton transfer in good agreement with

the previously calculated ones, 18.7 and 3.80 kcal/mol, respectively.109

H‘ + H,O

H,+ HO'
18.2

3.8

i , H,...HO‘
H...H,O

 

Figure 4.2 Proton Transfer in the 1130' Complex
Dihydrogen bonding could influence the reactivity of the bonds
involved in the interaction. Crabtree et al. reported that the hydrogens
involved in dihydrogen bonding, H, and Hb, in an Ir hydride complex, could
undergo scrambling, whereas the non-interacting hydrogen, Hc, was

exchanged much more Slowly with either H, or Hb, (Scheme 4.2).l '0

39

Scheme 4.2

 

 

 

 

 

L = PPh3

R = Me, n-Bu, p-TOLYL, Ph, p-FCQH4, 3,4-CeH3F2

The AG‘ was found to be 16 kcal/mol (R = Ph) and gets smaller as the R
group becomes more electron withdrawing which is consistent with a
mechanism involving proton transfer from OH to give an nz—Hz complex.
Rotation of H2 and proton back transfer completes the process. When the

solvent was benzonitrile, the H2 ligand was displaced by PhCN in a rate-

determining step. Heating the dihydrogen bonded complex at 80 °C in a
sealed tube led to H2 loss with formation of an Ir-O 0 bond. The process

could be reversed at 20 °C in CH2C12 with exposure to H2 (Scheme 4.3).

40

Scheme 4.3

 

An analogous hydrogen exchange process was observed for the

aminopyridine complexes (Scheme 4.4).“2

41

Scheme 4.4

 

/ N’R -—> /
13113946 ”fl. Prise/,6” NR
Hd—Ir—Hg' Hd—lr
\PPha l\PPh3
0 Ho
R= H, Ph

The reaction is first order with respect to the complex with a large negative
value for AS’t = -32 en, pointing in the direction of a highly organized
transition state.

Chaudret et al. documented the dynamic equilibrium between a
dihydrogen bonded complex and an nz-Hz complex, using NMR
spectroscopy.1 '1 In benzene or toluene, the RuH2(dppm)2-PhOH complex
exists as a mixture of dihydrogen bonded trans and cis isomers, but only the
trans isomer is involved in a dynamic equilibrium with the dihydrogen
bonded complex, which is 17 kcal/mol more stable. The driving force for the
reversibility of the process was suggested to be the strength of the
dihydrogen bond. As expected in the presence of a more acidic proton
donor, such as (CF3)2CH-OH, the nz-Hz complex undergoes H2 loss

(Scheme 4.5).

42

Scheme 4.5

+ROH

\\p/N\#{;’
R

”Ru.

\\\\\\\\O
\P/ \

/ \/p\ OR

 

Theoretical work112 done by Scheiner et al. on
HOH...H2Ru(PH2CH2PH2)2 model complex estimated a barrier for proton
transfer of 10 kcal/mol and a reaction enthalpy of 10.7 kcal/mol. However

when HF was used the only minimum localized on the potential energy

43

surface (PBS) was the nz-Hz complex, which was found to be 23.8 kcal/mo]
more stable than the fiee species.

The Epstein group investigated proton transfer by solution IR and
NMR spectroscopy in dihydrogen bonded complexes between
(triphos)Re(CO)2H and different proton donors: phenol, tetrafluoroboric
acid, chloroacetic acid, hexafluoro-Z-propanol, and perfluoro-Z-methyl-Z-
propanol at 200 — 260 K.113 The nz-Hz complexes were more stable than the
corresponding dihydrogen bonded ones and increases in temperature

induced H2 loss with the formation of the Re-O 0 bond (Scheme 4.6).

Scheme 4.6

ihph
PI

 
  

T, “ I I
fl“ ””1! ..\\\\\CO

 

For the HzRe(CO)(NO)(PMe3)2 0CF3COOH, and W(H)(CO)2(NO)(PMe3)20

(CF3)2CHOH systems the nz-Hz complex was unstable. In the case of

44

HRu(Cp)(CO)(PCy3) with (CF3)3COH the 112-112 complex is stabilized by
ion pair formation. The barrier for proton transfer in the latest complex was
found to be 15 kcal/mol.”4 The interaction between (Cp)Re(H)(NO)(CO) or
(Cp)Ru(H)(CO)(PH3) hydrides and proton donors of different strength, such
as H3O+, H20, and CF3OH, was studied by Scheiner et al in order to model
the reaction investigated by Epstein.”5 They found that the dihydrogen
bonded complex strength and the proton transfer barrier are strongly
influenced by the proton donor ability of the acid.

Chaudret and coworkers documented the proton transfer in a
dihydrogen bonded complex between a Ru hydride and (CF3)2CHOH when
the solvent is CDCle/CDF3 (2:1) at low temperature.”6 This Freon mixture
was selected to increase the solution’s dielectric constant at low temperature.
The proton transfer took place only at a temperature low enough to induce
the required change in the dielectric constant. The same behavior could be
observed when the solvent was CD2C12 (dielectric constant of CDzClz
increases from 9 at room temperature to 17 at 170 K).

N-H. . .H-Ru dihydrogen bonds are able to undergo proton transfer as
observed by Lau et al.”7 (Scheme 4.7). H/D exchange for both hydridic and

protonic hydrogens in D20, proved the equilibrium between the dihydrogen

45

bonded and nz-Hz complexes. The hydrogenolysis of the Ru-N bond could

be done at 60 °C under 60 atm of H2 pressure.

Scheme 4.7

@ (Cam —-- © (CH2)n

—

Lthp'llelH _____ H/N-uulllMe thPIIIII'-Ru H N...lllllMe
Ph2P \ Lph pl \
Me 2 H Me

t‘Hz +

@ (CH2)n

ll ----
ph2p| II RU\N...ulIIMe
LthP

 

+H2

Me

The covalent system was found to catalyze the reduction of CO; to HCOOH,
although in small yields. The proposed mechanism involves a heterolytic

ll7

splitting of the H2 ligand (Scheme 4.8).

46

Scheme 4.8

+

(0) (CH2). + H2 (0) (orb)n

LPhZPmmR”\N-umume LthpmuhR” H N---mIIMe

 

 

 

 

 

thp Me thp H Me
- HCOOH ll
0 (Cl-12)" O (CHzln
thPlllw-Ru Nf-umlMe + 002 thp"""'RU NHHIIIMe
Len P \0 H/ \ Ph P’ \H """ H/
2 Me 2 M6
0 H
n= 2, 3

The detection of the dihydrogen bonded complex by NMR suggested
that the C02 insertion is the rate determining step. Subsequent theoretical
calculations supported this conclusion.1 '8

Noyori et al. reported that theoretical studies on hydrogen transfer
between alcohols and carbonyl compounds catalyzed by Ru(II) complexes
containing NH or NH; ligands predicted the involvement of NH/NHZ group

119

in the reaction (Scheme 4.9). The implications of the intramolecular

N-H...H-Ru dihydrogen bond in the catalyst complex were not investigated.

47

Scheme 4.9

 

 

— - x
Q --> © —* ©
/RI|U~Y HZC'----H---RIU~Y H390“. ZN/
Hco- -H/ H/|l:\/ (QWHEJ H L]
1.. H _
Y=O, NH

The Morris group used a diamine, NH2C(CH3)2C(CH3)2NH2 (trnen),
and provided experimental evidence for the catalytic cycle predicted by
Noyori. Moreover they concluded that the N-H...H-Ru is responsible for
enantioselectivity observed in these reactions.‘20

Chaudret et al. reported hydrogen exchange in a related complex, 7,

using NMR spectroscopy, and the activation barrier for the process was

found to be 11 kcal/mol.

  

(C\Hzl2
+
Ph3pllllln-Ru\ /N.ulllllMe
Ph P H ----- H
3 Me
7

However theoretical calculations121 predicted that the exchange mechanism
does not involve proton transfer within the dihydrogen bond (Scheme 4.10).

Instead the evidence pointed to direct proton transfer to the metal itself.

48

Scheme 4.10

<F>—> ——-» WW

P\\\“" \R” HH_NM"2 P\\\““ "\R"

—]+ MezHN
NHMez

. Ru\
. P\“‘
13““R

A three-center intramolecular dihydrogen bonded complex and its

dynamics were reported by Jalon et al.122 The fast exchange between the
hydridic and protonic hydrogens was believed to occur via an nz-Hz

intermediate (Scheme 4.11) with a barrier of 13.6 kcal/mol as determined

via NMR Spectroscopy.

49

Scheme 4.1 1

 

I uu\|\\|\\\\\\“

Ru...-

—/\N/
\ / .\\/

When a solution of the dihydrogen bonded system in CDgOD was exposed
to H2 atmosphere at room temperature and 1 atm, more than 90% of the Hz
was exchanged for D2 in about half an hour.

The Morris group reported a similar exchange in a ruthenium
polyhydride complex, K[( 1 , 10-diaza-l 8-crown-6)] [RuH5(PiPr3)2] (Scheme
4.12).'23 Upon exposure to D; at room temperature and 1 atm for 5 min, the
intensities of the protonic and hydridic hydrogens were depleted by 100%

and 90%, respectively. When 18-crown-6 ether was used only 13 %

50

depletion was observed after 10 days. Also replacement of Ru hydride by the
less basic Os hydride induced a slower exchange. Control experiment
excluded RuH2(H2)(PiPr3)2 as an intermediate.

Scheme 4.12

Intramolecular N-H. . .H-Re were considered to be responsible for

hydride fluxionality in ReH5(PPh3)2L, (L = N—acetyl-2-aminopyridine), 8.

51

The activation energy for rotation, involving H1, H4, and H5, is 0.7 kcal/mol
smaller with the NHAc group in ortho than in para suggesting that the
transition state is stabilized by dihydrogen bonding.124 However theoretical
calculations showed that the two hydrogens involved in the dihydrogen
bonding are forced to be 1.49 A apart at the transition state.125
Stereoselectivity and reactivity can be influenced by dihydrogen
bonding. By stabilizing one particular transition state among many others,
they can control the product distribution and/or the stereochemical outcome.
This concept was demonstrated by Crabtree et al. in the selective imination

of an Ir aldehyde complex with ortho- versus para- hydroxyaniline(Scheme

4.13).126

52

Scheme 4.13

 

 

OH

When an equimolecular mixture of reactants, 2-aminophenol, 4-
aminophenol, and Ir aldehyde complex, was used a 4.2:1 products ratio was
obtained which corresponded to a kl/k2 ratio of 6. The authors concluded
that the driving force was the dihydrogen bonding which stabilized the
product and the corresponding transition state. The two products do not

interconvert indicating that the product distribution is due to kinetic control.

53

The dihydrogen bonding was estimated to be stabilizing by 4.5 kcal/mol
while AAG“ between the two reactions was found to be 1.1 kcal/mo]. The
difference was suggested to come from the unfavorable ring conformation
required for dihydrogen bonding at the transition state.

Aime et al. showed that dihydrogen bonding could control the
regiochemistry of ligand attachment in transition metal complexes.127 When
an unsaturated Os complex was reacted with EtNH2 or EtzNH only the syn
product was observed (Scheme 4.14). The intramolecular bond would not be

possible in the anti isomer. When Et3N was used no reaction occurred.

Scheme 4.14

R

...... H \

H H """""" \N°““R

08(C0)4 03(CO)4
L /

(co)3ol/ \Os(CO) ——> (0010! \LHCO)
\H/ 3 3 \H/ 3

L = EtzNH, EtNHz
R = Et, H
Treatment with NH; followed by acetaldehyde or acetone in CHC13 led to

the exclusive formation of the syn product with an intramolecular

dihydrogen bond (Scheme 4.15).

54

Scheme 4.15

.............. “Tye

1. NH3 H

I OS(C0)4\ 2 CH C(R)=O I 08(C0)4

/ 3 / \l

(OC);.,Os\Os(CO)3*(00130s 08(00):
\H / CHCI3 \H/

When CHC13 was replaced with a more polar solvent capable of hydrogen
bonding, such as methanol, the N-H. . .H-Os was disrupted. NMR
spectroscopy studies showed that the outcome in this system is dictated by
thermodynamic control.”8’ ‘29

In the Jackson group, it was proved that O-H. . .H-B dihydrogen bonds
could accelerate and direct the borohydride reduction of ketones to
alcohols.130 The reductions of 2-hydroxycyclobutanone or 2-
hydroxycyclopentanone with NBu4BH4 in non hydrogen bonding solvents,
such as CH2C12 and ortho-C6H4C12, were accelerated by 150 times relative to
the reductions of corresponding unsubstituted cycloalkanones. At the time
the findings appeared to indicate high stereoselectivity but as will be

discussed in chapter 4 more recent results suggest a less clear-cut picture.

55

2. Dynamic Studies for Proton Transfer in Dihydrogen-bonded

Complexes involving A1H4'

In principle, deposition of sufficient vibrational energy into a
chemical bonds stretching mode should induce its rupture, resulting in an
Infrared-pumped reaction (IRPR). At the same time the vibrational
relaxation is very fast, usually beating any process that actually effects
chemical change. The challenge to those who would design IRPRS is thus
clear: the reaction time must be competitive with the relaxation time of the
vibrational mode in question; the energy deposited must of course be
adequate to surmount the activation barrier (including thermal and tunneling
contributions) for the process; and the overall reaction should be irreversible
or at least exothermic (i.e. the back reaction should have a high barrier). We
have used ab initio methods to explore a class of candidate reaction systems
that appear to meet these requirements.

Dihydrogen bonded complexes A-H. . .H-M (A = O, N, Halogen; M =
metal) offer several attractive features which should make them good
substrates for reaction via single-photon infrared excitation. As in any
A-H. . .3 hydrogen bond, the A-H stretch vibrations are typically isolated at

the high energy end of the IR spectrum, in the 3-4000 cm'1 range (i.e. ca.

56

8-12 kcal/mol), and are preorganized to align the corresponding normal
mode motions with the reaction coordinate for proton transfer. Also, a range
of potential energy surface behaviors is possible; depending on the
components' relative acid/base strengths, the complex dissociation energies
become larger while the barrier to proton transfer decreases. Unique to the
hydridic-to-protonic class of hydrogen bonds, proton transfer fi'om A-H to
H—M triggers irreversible loss of H2 and new A-M bond formation. The most
active atoms in this process are both hydrogens (proton and hydride), light
particles capable of quantum mechanical tunneling, which may allow them
to move on a timescale competitive with vibrational relaxation.

In the Jackson group, R. Custelcean has explored this mode of
reaction in a number of solid state complex materials using ordinary thermal
excitation. Of particular relevance is the solid state kinetic study of
LiBH4-TEA (TEA = triethanolamine; IUPAC name tris(2—
hydroxyethyl)amine) in which a value of 2'20 kcal/mol was deduced for the
rate determining initial proton transfer.131 This value is quite Similar to the
activation energy value found by Mesmer and Jolly for the solution
hydrolysis of BH4' in H20.132

As noted above, a dihydrogen bonded complex may offer an

opportunity for an IRPR. The H. . .H association preorganizes the system,

57

and therefore this reaction may be able to proceed rapidly enough to outpace
vibrational energy redistribution, a process that normally occurs much faster
than chemical reactions.

Because Al belongs to the same group as B and Ga and aluminum
hydrides are widely used reducing reagents, we have considered that it
would be of interest to study A1H4’ as a candidate for M-H and to model the
proton transfer reaction for H. . .H complexes involving All-14'. Three A-H
partners have been considered: H-OH, H-F, and H-Cl.

2.1 Reactants, Transition States and Products

The common features of a dihydrogen bonded complex, A-H...H-M,
as noted in chapter 1.2, are close H...H contacts distances, smaller than the
sum of the Van der Waals radii of the atoms involved in the interaction, bent
geometry, elongation of the A-H bond, and association energies comparable
with those corresponding to traditional hydrogen bonded complexes.

In the calculated simulations, at the MPZ/6-3l 1++G** level, all the proton
donors considered, H20, HF and HCl, formed dihydrogen bonded

complexes with A1H4’, as defined by a H...H contact distance <2.4 A, the

sum of the Van der Waals radii.

58

Figure 5.1 [AIH4.. .H20] Complex Structures: Reactant, Transition
State, and Products

 

 

Reactant

Transition State

 

Products

59

Figure 5.2 [AIH4...HF]' Complex Structures: Reactant, Transition State,
and Products

 

Transition State

 

Products

60

Figure 5.3 [AIH4...HCI]' Complex Structures: Reactant, Transition
State, and Products

 

Reactant

 

Transition State

 

Products

61

The complexes geometry is slightly bent, with the H(A)-H(Al)-Al
(A=F, Cl) angles falling in the range 140-150°. In the [A1H4...H20]'
complex, Figure 5.1, both protons from H20 form dihydrogen bonds and
thus the small values for the H(O)-H(Al)-Al angle, 109.5 and 110.1°. As
expected the elongation of A-H (A=OH, F, Cl) increases with the acidity of
the proton donor and the H...H close contact distances vary in the same
direction, the smallest value corresponding to the [A1H4...HCl]'complex.

A weaker elongation was observed for the H-Al bond involved in
dihydrogen bonding together with shrinkage of the other H-Al bonds. The
elongation and the bending in geometry are consistent with known
experimental structures, and are interpreted as an indication that the
interaction is between the proton and the o—bonding electrons of the Al-H
bond. Table 5.1 summarizes the relevant geometrical findings.

Table 5.1 Dihydrogen Bonded Complexes: Geometries

 

 

 

 

 

A H—A‘I H(A)-H(AI)' H(A)-H(Al)—Al"
Free Complex Distance Angle
O 0.960 0.967 2.009; 2.009 109.5; 110.1
P 0.917 0.956 1.371 143.0
CI 1.273 1.381 1.226 149.4

 

 

 

 

 

a. in A, b. in degrees

62

The association energy — calculated as difference between the
complex’s energy and the sum of the free partners energies, All-14' and HA,
and corrected by the zero point vibrational energy (ZPVE) — has been
predicted to be 10.14 kcal/mo] for the [Ali-I4...H20]’ complex and
corresponds to 5.07 kcal/mol/dihydrogen bond. Interestingly, the strength of
the dihydrogen bond does not increase with the acidity of the proton donor,
15.45 and 13.87 kcal/mo] for [All-I4...HF]' and [All-I4...HC1]' respectively.
The calculations have also reproduced the experimentally observed shift of
the A-H stretch to lower frequency upon association, a feature common to
most hydrogen bonded complexes (see Table 5.2).

Table 5.2 Energetics of Proton Transfer Reaction for Dihydrogen

 

 

 

 

 

Bonded Complexes
A AEMP.,,' AE.‘ AEm.‘ H-A Vibration”
Free Complex
OH 10.2 16.0 27.5 3884; 4003 3797; 3847
F 15.5 19.0 20.9 3325 2200
C1 13.9 6.7 31.4 3087 1978

 

 

 

 

 

a. in kcal/mo], b. in cm"

The potential energy surface (PBS) for each proton transfer/hydrogen
loss reaction was initially explored via a relaxed scan in which the H. . .H

distance was stepped from the complex value to the distance in molecular H2

63

 

with reoptimization of all other geometrical parameters at each selected
H...H distance. In each case, only one maximum was found, suggesting that

the reaction occurs via a concerted process, Scheme 5.1.

Scheme 5.1
..—-H *
HaAl—H+H—A —> H3Al-—,:H —> HaNk ~ 5 ——-> HaAl—A+H2
\“ :zH
H\A Ax
complex transition state

The structure with the highest energy was optimized to a transition
state and a vibrational analysis was performed in order to check the nature of
the saddle point. All the transition states have both the proton and the
hydride in close proximity to the A1 at distances in the range of 2.0 A (see
Figures 5.1, 5.2, 5.3).

Previous calculations133 at the CSD(T)/TZ2P(f,d) level done by
Schleyer et al. concluded that AlH5 is a weak complex between alane (A1H3)
and a slightly elongated H2 molecule with an association energy of 1.7
kcal/mol at 0 K. At room temperature Ale was found to be unstable toward
dissociation by 3.1 kcal/mo]; evidently formation of the weak tricentric bond
barely perturbs the 104 kcal/mo] H-H bond strength, consistent with the high
(22 kcal/mol) barrier calculated for scrambling between H2 and NH,

hydrogens.

64

As expected, the barrier for proton transfer drops dramatically with
decreasing H. . .H distance in the energy minimized complex. For the
[All-I4”'HC1]' complex (see Figure 5.3) this barrier has been calculated to be
only 6.69 kcal/mol with an H(Al)-H(Cl) distance of 0.78 A, just 0.03 A
longer than in free H2 at the same level of theory.

For all the systems studied the net reaction is strongly exothermic,
Table 5.2, yielding H2 and A1H3A' as products. The best candidate for
subsequent dynamic studies was considered to be [AlH4. . .HCl]' complex
because it has a small barrier for proton transfer, which at the same time is
much smaller than the association energy. Also the reverse reaction has a
huge barrier by comparison and competitive processes have far smaller
chances to happen. In order to check that the transition state connects the
reactant and the products, an intrinsic reaction coordinate (IRC) calculation,
steepest descent path in mass-weighed cartesian coordinates from the
transition state to both reactant and products, has been performed and the

result has been interpreted as a single step process, Figure 5.4.

65

20

 

 

 

 

 

Free Partners

10 TS
_ 0 _
2 Complex
a
3'; -1o -
>2
5’ -2o _
c
m

-30 _

Produ
to “3
0 10 20 30 40

Reaction Coordinate, hohrlamu"2

Figure 5.4 IRC Calculation for Proton Transfer and H2 Loss in
[All-I,...HCI}'
2.2 Dynamics
Once the reactants, products, and transition state geometries were located on
the PBS, the reaction dynamics could be explored via ab initio classical
trajectories. Among the modes associated with the dihydrogen bonded

complex the ones relevant to proton transfer are 1411 and 1973 cm", Figure

5.5.

   

O
1411 cm" (Al-H stretch) 1973 cm" (Cl-H stretch)

Figure 5.5 Excited Modes

66

These modes represent mixtures of symmetrical and antisymmetrical Al-H
and Cl-H stretches involved in the H...H interaction.

In order to model the IRPR on [AlH4...HCl]', we have performed
dynamic reaction path (DRP) calculations. A method implemented in the
GAMESS package, DRP allows for deposition of a selected energy quanta
in a specific mode and classical treatment of the reaction dynamics, i.e.
calculation of ab initio trajectories, as a result of this excitation.

The Al-H stretch, 1411 cm", has been excited with different energies
in the range 1.5-4.5 quanta, corresponding to 6.4-19.2 kcal/mol, and all the
other modes with the ZPVE energy. It is important recognizing that with the
first excitation energy, 6.4 kcal/mol, the system is brought close to the
activation barrier, 6.69 kcal/moi, while in all the other situations the energy
deposited is well above the activation barrier. The time evolution of the
H(1)-H(2) distance, H(l) and H(2) are the proton and the hydride involved
in dihydrogen bonding, is depicted in Figure 5.6. For 1.5 and 4.5 hv the
proton transfer is followed by loss of the H2 molecule within 400- 800 fs via
assistance from Cl'. The differences reside in the time scale for proton

transfer and the hydrogen atom that is involved in the process.

67

a as:
82 com 8m 2:. So com 8*. 8m 8m 2: o

fi fi q q q _ q ‘ d o

 

31.8:

I
V

51.3:

 

O
(B
co
y ‘aoueisga

31.3:

n

m o

A : gag/...;

 

 

 

 

 

 

>923 :ouflfixo End 6 “>9. md no "Emfi "a ”Em... no

5:

g2? :

...—3 :3 an 53:50 eczema £73335; 5 =o€£o>fl 3:5»:— .w.m PEER

68

While for 4.5 quanta the particles that react are the ones that are
dihydrogen bonded initially, for 1.5 quanta, the All-14' fragment rotates,
moving a different hydride into position to interact with the proton and react
to form H2. For the intermediate energies, 2.5 and 3.5 quanta, the proton is
transferred but undergoes reversion without liberation of H2.

The H—Cl stretch, 1973 cm", was excited with energies in the range
1.5-3.5 quanta, 9.0-22 kcal/mo], Figure 5.6. In all calculations proton
transfer has been observed with the formation of complexed H2 for time
intervals that increased in length with the excitation energy. Only for 3.5
quanta, i.e. 22 kcal/mol, has H2 loss accomplished within 200 fs. It is worth
noting that in all simulations, for 1411 and 1973 cm'1 modes, proton transfer,
reversible or not, is observed.

It is of importance to know the extent of the reactant vibrational
modes involvement in the proton transfer/Hz loss reaction. In order to
investigate this aspect we have performed a DRP calculation in which the
starting point was the transition state with 0.05 kcal/mol energy deposited

into the imaginary mode corresponding to the reaction path.

69

coo F

e 65:.
8a 8w 2: 8m 8m 84 com 8m 2: o

 

1 q _ _ q 4 4 q i o

 

v ‘aouasm
70

O

 

 

 

 

E 2 a ”.53 b 25 a
E:

P I...~NVI

was at: a 5355 Beam 2.23.2.3» a 8:33. 8:333:13: 5. 2:5

The amplitude and momentum of the modes were mapped onto the normal
modes of the dihydrogen bonded complex, i.e. the ground state. In other
words the normal modes of the reactant have been considered basis vectors
to describe the reaction dynamics. The modes that we considered to be
associated with the proton transfer, 1411 and 1973 cm", were the only ones
that at the ground state had amplitude changes characterized by periodicity,
Figure 5.8, whereas the other were characterized by monotonic decay of
amplitude, Figure 5.9.

Figure 5.8 Evolution of the 1411 and 1973 cm'1 Modes Upon Transition

State Relaxation

 

    
   

1411 mode
1973 mode

41.4 -

Amplitude, bohrlamu"2
,5

 

 

 

O 20 40 60 80 100
Time, f8

71

Figure 5.9 Evolution of the Modes Other Than 1411 and 1973 cm“1

Upon Transition State Relaxation

.3
U1

 

A

 

O

Amplitude, bohrlamum
9': o
01 0|

 

I
A

 

0 10 20 30 4O 50 60 70 80 90 100
Time,fs

Additional trajectories need to be computed in this 15-dimensional
phase space but the rapid geometry and momentum changes require a small
step size and consequentially long computational times, which do not readily
allow for direct collection of statistically significant number of runs.
However, the surprisingly general occurrence of proton transfer (albeit not
always leading to dissociation of H2) has been very encouraging for the
potential of processes like theses for IRPR. It must be recalled that theses
simulations are all classical. Presumably the contribution of tunneling would

only serve to enable the reaction.

72

The particles involved in reaction are light and able to tunnel and we
attempted to calculate a reaction rate based on the IRC, which includes
tunneling correction based on a formalism presented by Cukier et al.‘34
The IRC curve in cartesian coordinates has been fitted to a quartic potential,
V(r) = ar2 + br3 + cr4 (r is the reaction coordinate). The reactant well and
barrier curvatures have been expressed as frequencies. The rate constant has
been evaluated with and without an excitation energy found to be 0.836 ps'l.
Presence of an excitation energy of 6 kcal led as expected to an increase in
the rate constant by 70 times. For values above the activation barrier the rate
constant has reached a plateau around 0.086 fs". For a more complete
understanding we hope soon to be able to compute ab initio trajectories in
which the hydride and the proton benefit from a quatum mechanical
treatment, based on a new formalismI35 that will be introduced in the

GAMESS package.136

73

3. Kinetic and Mechanistic Investigations of

2-Hydroxycyclopentanone Reduction by Borohydride

Dihydrogen bonds have been at the core of extensive research since
1994. In the intervening years, their structures, and association energies have
been explored, and their importance in reactivity and selectivity in solution
and solid state have been demonstrated. Dihydrogen bonded complexes
involving borohydrides are of particular importance — borohydrides are
widely used as reducing agents.

Sharma, Huang, and Jackson have demonstrated the existence of
H...H close contacts in the NaBH4-HZO crystal structure. Experimental and
theoretical work by Epstein has also suggested the ability of borohydrides to
dihydrogen bond with alcohols. Hydrolysis of B11; in acidic or neutral
conditions has been proven to involve proton transfer as rate determining
step (RDS), followed by H2 loss and B-0 bond formation. The activation
parameters, AH9e and AS’, were determined to be 20.6 d: 1 kcal/moi and —
22.3 :i: 3 e.u., for neutral conditions, and 8.0 :I: 1 kcal/mol and —3 :t 3 e.u., for
acidic hydrolysis, respectivelym'I39

Work done by S. Gatling in the Jackson group has aimed to explore

the stereoselective control and reactivity enhancement facilitated by

dihydrogen bonds in a typical borohydride reduction reaction. The

74

substrates, 2-hydroxycyclobutanone and 2-hydroxycyclopentanone, have
been selected based on semiempirical calculations. Because of their ring
fiameworks, intramolecular hydrogen bond was thought to be unlikely in
these structures and therefore rate enhancements or directing effects would
not be expected fiom carbonyl activation by OH group. Also the chosen
solvents, chlorinated hydrocarbons (1,2-dichlorobenzene and
dichloromethane) and the borohydride’s counterion, tetrabutyl ammonium,
were unable to participate in hydrogen bonds with either the substrate or the
borohydride. The results, summarized in Table 6.1, showed that the
hydroxycycloalkanones were reduced at least 100 times faster than the
corresponding simple cycloalkanones.

At the same time the analysis of the products has led to almost
exclusively the trans diol product suggesting that the hydride is delivered
from the OH substituted face of the substrate. The influence of competitive
hydrogen bonding additives, and of replacement of the OH with a
substituent incapable of hydrogen bonding were also studied.

We thought that it would be of interest to explore the influence of the
dihydrogen bonding on the activation parameters of ketone reduction by

tetrabutyl ammonium borohydride, NBu4BH4. We have chosen

75

2-hydroxycyclopentanone as substrate because it was the most thoroughly

studied in the previous works.

Table 6.1. Reported Reduction Results of a-Hydroxycycloalkanones and
of Corresponding Unsubstituted Ketones

 

Substrate Solvent ty, Cis diol Trans diol

 

0:0 DCBa 2 hrs

6” DCB“ <1 min 0.1 99.9

E)” DCB 17 hrs

DCB 7 min 1.4 98.6

CHzClz 8 min 0.1 99.9

 

a: DCB = 1,2-dichlorobenzene

The overall ketone reduction reaction is given by the equation

4R2C=O + NBuCPBmG —-> NBuPl<chH0).B]9 5.1

with the following steps:
R2C=O + man?) B1449 —> NBuflRZCHOBHgG 5.2

R2C=O +NBu§9[R2CHOBH3]e —>NBu§9[(R2CHO)zBH2]GS.3

76

R2C=O +NBuS9[(R2HCO)ZBH2]e—’NBU§)[(R2CHO)3BH]G 5.4
ch=o +NBuS9[(chH0),BH]9 —>N13u§+>[(RZCHO),B]e 5.5

Previous studies on ketone reductions by borohydrides have shown
that the rate law is second order and that the intermediate species are more
reactive that NBu4BI-I4.14o Wigfield and Gowland have examined the
reduction of different ketones with a mixture of NaBH, and NaBD4. The
absence of isotopically mixed borohydride species at the end of the reaction,
NaBHnDM, has been considered a proof that disproportionation has not
taken place."“

The OH from 2-hydroxycylopentanone and the borohydride, BH4', are
able to interact and thus a dihydrogen bonded complex is formed prior to the
reduction. We have considered that the reaction has second order kinetics,
first order in both ketone and borohydride, and that the rate determining

step, RDS, is the transfer of the first hydride. The overall process is be

 

described by Scheme 6.1.
Scheme 6.1
HO—ch=o + BH,’ === BHi----HO—R2C=o—>HO—RzHC—OBHa'
dihydrogen bonded complex 1
(HO—RZHC 0).,B'
cis and trans

77

The rate law is given by

6x/6t = k-[ketone]'[NBu43H4] 5.6
The ketone and borohydride concentrations were calculated as

[ketone], = [ketone]o - 4x 5.7
[NBmBHdt = [NBu43H410 — x 5.8
where x is the conversion. Application of the above expressions in equation
5.6 led to

ax/at = k-{[ ketone]o - 4x}-{[NBu4BH4]o — x} 5.9
Integration of equation 5.9 gave

1n{([NBu4BH4]o - X)/([ket0ne]o - 4X)} = (4'INBmBHdo - [ketoneloflot +
1n( [NBu4BH4]o/[ketone]o) 5. 10
From the plot of ln{([NBuBH4]o — x)/([ketone]o - 4x)} versus time the rate
constant for a specific temperature was determined as

k = slope/(4-[NBu4BI-I4]o - [ketone]o) 5.1 1
The Eyring equation gives the relationship between the activation
parameters and the rate constant for a specific reaction:

k/T = kB/h-exp(AS¢/R) -exp(-AH¢/R) 5.12
From the logarithmic representation of the rate constant (k) versus (l/T) we
obtained the activation enthalpy as

AH¢ = slope-R 5.13

78

and the activation entropy as
AS;é = {intercept - ln(kB/h)} 'R 5.14
The reduction of 2-hydroxycyclopentanone was monitored at different
temperatures over a range of 40 K by solution IR spectroscopy. The
concentrations of ketone and borohydride were 0.1 and 0.4 M, respectively,
to ensure an excess of the reducing agent at all times. The progress of the
reaction was followed via the carbonyl’s absorption. We determined the
concentration of ketone during the reaction using a calibration line based on
the carbonyl peak (1748.6 cm“'). The borohydride concentration was
determined with the help of equation 5.7 and 5.8. The OH peak, which was
observed at 3450 cm'l in a solution of ketone in 1,2-dichlorobenzene
(1,2 DCB), shifted to 3250 cm’l, indicative of a dihydrogen bonding
interaction, in the presence of borohydride and its appearance remained
unchanged during the reaction. Through all the runs we did not observed H2
evolution. As expected the reaction kinetics were found to be first order in

ketone (see Figure 6.1) and in borohydride (see Figure 6.2).

79

ln([borohydride])

In([ketone])

-0.91

-0.92 -

-0.93
-O.94

-0.95 »

-0.96

-o.97 ~
-0.98 —

-0.99

 

 

 

 

 

 

 

 

9 =
o fi‘fifiz’fiwou«m»..«««m»,T,.EZ‘1,K
_ O. A““‘: ...-....l...: =283 K
'0 n T = 298 K '
T=313 K “AA
.0
0 500 1000 1 500 2000
Time, s
Figure 6.1 Time Evolution of Ketone Concentration
1" 0
CI. ...00...
ill- ”.°"00« u o 274 K
O Q =
0 A‘ 'I I...- T =283 K
.. ‘“ Ill-..-
.. T=313 K “n "I .
0o “M l
T = 298 K i
0 500 1000 1500 2000
Time, s

Figure 6.2 Time Evolution of Borohydride Concentration

Application of the ketone and borohydride concentrations in equation 5.10

allowed us to obtain the integrated rate laws that have been depicted in

Figure 6.3.

80

9‘
.5

 

 

 

 

8 4.9 _ -
3 o
g“ T=313K “‘,T=298K
3.9 .
E 34 °' 1“ _.. T=283K
m . ' AA‘ "In _
g 2.9 .. “A I ....-ll. T-274
I
2- 2.4 .:='..-. ....e000000009.000.000.00009°’09000
2' 1.9 “......OOQ...
" 1.4
0 500 1000 1500 2000
Time, [a]

Figure 6.3 Integrated Rate Laws for Reduction of
2-Hydroxycylopentanone by NBu4BH4

Using equation 5.11, we have determined the rate constants, and their values

have been summarized in Table 6.2.

Table 6.2. Rate Constants for the Reduction of
2-Hydroxycyclopentanone by NBu4BH4

 

 

Temperature, K k, [l-mol"°s"] - 103
274 0.33 :1: 0.027
283 1.00 :t 0.033
298 2.20 :1: 0.160
315 8.102t0.890

 

81

The representation of equation 5.12 has been depicted in Figure 6.4 and the
activation parameters have been determined using the equations 5.13 and

5.14 (Table 6.3)

-10

 

-11 a

-12 _

Inlkl'l']

-13 _

-14 _

 

 

-15
0.0031 0.0032 0.0033 0.0034 0.0035 0.0036 0.0037

1”

 

Figure 6.4 Logarithmic Representation of the Eyring Equation
For comparison the reduction of cyclopentanone by NBu4BH4 in 1,2-
dichlorobenzene and in 2-propanol have also been performed and the
activation parameters have been determined (Table 6.3).

Dihydrogen bonding between the OH substituent and BH,’ lowers the
enthalpy of activation by 6.6 kcal. A larger effect on the enthalpy is
observed when the solvent is 2-propanol. While in the
2-hydroxycyclopentanone/BH4' system there is one dihydrogen bond, in the
cyclopentanone/BHUZ-propanol there one hydrogen bond between the

carbonyl and the OH from the alcohol and dihydrogen bonds between the

82

BR; and 2-propanol. Thus activation of both carbonyl and borohydride will
lower the activation enthalpy of the reaction. At the same time the presence
of interactions requires a higher degree of organization in the transition state,
more so when there is an excess of proton donors, and hence the larger
negative value for the activation entropy.

Table 6.3. Activation Parameters for Ketones Reductions

 

 

Substrate Solvent AH“, [kcal/moi] AS‘, [e.u.]
OH
6“) DCB 12.2 i 1.89 -29.6
0:0 DCB 18.8 i 1.08 -22.9
2-iPrOH 9.8 i 1.07 -44.3

 

We were interested in further investigating the influence of different
factors on the stereochemical outcome. When we used the same
work-up procedure employed by dr. S. Gatling and analysis by gas
chromatography we were able to reproduce the results previously obtained
for the 2-hydroxycyclopentanone/I\IBu4BI-I4 system in dichloromethane. It is
of importance to note that the actual work-up procedure is different from the
one reported. Instead of the HzOleaOH quenching followed by

dichloromethane extraction of the aqueous layer and concentration (which

83

was the reported work-up) we treated the reaction mixture with HCl,
followed by neutralization of the aqueous layer and removal of water
(procedure used by dr. S. Gatling).I42 In both procedures the alcohols
mixtures were derivatized to the corresponding trimethylsilyl ethers and
were analyzed by gas chromatography.

Equipment upgrade in our laboratories allowed us to examine the
products of reaction directly via I-IPLC. We analyzed the different layers of
the reaction mixture at each step in the work-up procedure and we noted that
there were losses at each step. Moreover the trans and cis diols were not
hydrolyzed to the same extent and also were not extracted with the same
efficiency, (see Table 6.4).

Table 6.4 Work-up Procedures Results

 

 

Work-up Layer Trans'I Cis' Diols"

HzOz/NaOH H20 52 48 5 1
CH2C12 8 1 19 49

6N HCl H20 97 3 43
CH2C12 45 55 57

 

a.% of cis or trans diol in the respective layer
b. % of diols (cis and trans) in the respective layer

We used instead a single step procedure without separation or

derivatization, treatment with concentrated H2804. The results are

84

summarized in Table 6.5 together with the ones previously obtained by S.
Gatling. We observed a smaller directing effect, trans:cis ratio 63:37 as
opposed to 99.7:0.3 reported by dr. S. Gatling in similar conditions. One
reason for the discrepancy in the results is the stability of the cyclic borates
corresponding to the 1,2 cis diol. The equilibrium constant was determined
to be 1.9 in favor of the borates formed by the 1,2 cis hexanediol.143 Also in
the acidic work-up the aqueous layer has a much smaller volume than the
organic layer( 3 ml H20 and 15 ml CHzClz) and thus hydrolysis of the
borates is not favored.

In order to clarify if the lower trans:cis ratio observed could be
attributed solely to the reduced number of separation steps , we have run the
reaction at higher dilution and lower temperature, and we observed that the
trans diol is slightly favored. This result was interpreted as evidence of the
increased ability of the dihydrogen bonded complex to react at low
temperature and high dilution before the intermediate alkoxyborohydride
species participate in reaction.

Subsequent ab initio calculations at the B3LYP/6-311G level found
that 2-hydroxycylopentanone is in fact capable of intramolecular hydrogen
bonding, despite Jackson and Gatling earlier assumption t0 the contrary,

Figure 6.5. We calculated a O-H...O(=C) close contact of 2.35 A and an OH

85

frequency shift of 65 cm'1 compared to the non hydrogen bonded conformer.
The hydrogen bonded structure was found to be more stable by 4.8 kcal/mol.

Table 6.5 Trans:Cis Ratios in 2-Hydroxycyclopentanone Reduction

 

 

[Ketone] [BIL'] [Trans] [Cis] Additive
0.321“ (0.25) 0.356 62.8 (99.7) 372 (0.3)

0.163a 0.500 62.4 37.6

0.078a 0.016 72.6 27.4

0.140” 0.132 71.7 28.3

0.008“ 0.531 62.0 38.0

0.0113 0.012 65.6 34.4

0.0138 0.013 72.0 27.0

0.005a 0.005 69.2 30.8

0.004b 0.002 79.6 20.4

0.349“(0.25) 0.331 (0.25) 63.5 (93.5) 36.5 (6.5) TBABr
0.357‘1 (0.25) 0.324 (0.25) 64.2 (79.7) 35.8 (20.3) TBACl
0.376‘(0.25) 0.310(0.25) 65.8 (54.5) 34.2 (45.5) TBAF
0.242° 0.250 70.0 30.0

0.242d 0.250 46.5 53.5

 

a: NBuBH4, CH2C12
b: temperature 0 °C
c: NaBH4, THF

(1: LiBI-I4, THF

The values in parenthesis are previously reported results.

86

 

Figure 6.5 2-Hydroxycyclopentanone
When BH4' was introduced into the calculations we have been able to find

two minima for the optimized dihydrogen bonded complexes.

 

Figure 6.6 2-Hydroxycyclopentanone-BH4' Complexes

In the structure with the lowest energy, the BH4‘ ion lies almost in the plane

of the ring suggesting a small face selectivity for the reduction. The structure

87

l—

that has the B11; coming from the OH substituted face of the ring, B in
Figure 6.6, is higher in energy by 0.7 kcal/mol.

Presence of additives capable of hydrogen bonding to the ketone’s OH
should be able to slow the reaction rate and decrease selectivity as a result of
competing with the BH.{ for interaction with the OH. Dr. Gatling’s reported
that addition of tetrabutylammonium bromide (TBABr),
tetrabutylammonium chloride (TBACl), and tetrabutylammonium fluoride
(TBAF) decreased the reaction rate and the trans:cis ratio of the diols in the
order TBABr<TBACl<TBAF in accordance with Br'<Cl'<F' trend as
hydrogen bond partners. We have observed in similar conditions barely any
influence. Based on the structure that we have obtained for the dihydrogen
bonded complex, A in Figure 6.6, one explanation may be that the hydrogen
bonded complex adopts a conformation with the X' (X = Br, Cl, F) in the
plane of the ring also. Also dr. Gatling used all TBAX without purification
and in their commercial form they contain water, which is a very good
dihydrogen bonding partner for the borohydride, in other words the change
in the reaction rate cannot be attributed only to the TBAX. In our set of

experiments we used recrystallized TBAX (water removal was checked via

IR).

88

We chose NBu4BH4 to study the reduction of ketones by borohydrides
because the counterion, NBuf, is incapable of hydrogen bonding to either
substrate or borohydride. However in practical synthetic organic chemistry
the widely used borohydrides are NaBH4 or LiBH4 in solvents such as
tetrahydrofuran (THF) and we were interested to explore the streochemistry
of the reaction in those conditions. The trans:cis ratio was dictated by the
solubility of the borohydrides; for NaBH4 which has a lower solubility, the
ratio was essentially the same as in the system NBu4BH4/CH2C12, but for

more soluble LiBH4, an almost equimolar trans:cis ratio was observed.

89

4. Methods of Calculation

The methods presented in this chapter refer to the calculations from
chapter 2. Calculations used the MPZ level of theory and the 6-311'H-G”
basis set within GAUSSIAN 94,144 GAUSSIAN 98“” and GAMESS‘“5
packages. Complexes were obtained by optimizing a structure with the
hydride, All-14', and the proton donor, HA, initially at large distance. H...H
contact distances smaller than 2.4 A and association energy were used as
1 criteria 0 validate or refute a dihydrogen bonded complex. The interaction
energy was calculated as the difference between the complex’s energy and
the sum of the free partners' energies including the zero point vibrational
energy corrections. Because previous calculations in the Jackson group had
shown this level of theory to be quantitatively successful in estimating halide
and borohydride ion-molecule association energies, we did not correct for
basis set superposition error (BSSE). Vibrational analyses were performed to
confirm the nature of all stationary points on the PES, i. e. the dihydrogen
bonded complexes were minima with no imaginary fi'equencies.

The potential energy surface (PBS) for the proton transfer/Hz loss
reaction was investigated via a relaxed scan with respect to the proton-
hydride distance. The H...H distance was stepped from the value in the

dihydrogen bonded complex to the value in the H2 molecule and at each

90

selected distance all the other parameters were reoptimized. A vibrational
analysis was performed on the highest energy structure to confirm/refute its
transition state nature, i.e. one imaginary fiequency corresponding to proton
transfer. Subsequently the structure was reoptimized as a transition state.
The last structure from the PES relaxed scan was fully optimized and
considered the product geometry. The intrinsic reaction coordinate (IRC),
steepest descent path in mass-weighted coordinates calculation, was then
traced in order to verify the concerted or step-wise nature of the process and
also to check that the transition state connects the reactant and products.

Once the most suitable candidate ([A1H4. . .HCl]‘) for IRPR was

found, the modes associated with the proton transfer were identified and a
series of dynamics studies was performed using the dynamic reaction path
(DRP)147 method implemented in the GAMESS package. DRP method
allows for deposition of excitation energy in a specific vibrational and
calculation of classical ab initio trajectories resulted from that excitation.
Two types of DRP calculations were done. First we deposited in each of the
modes associated with proton transfer in the dihydrogen bonded complex, i.
e. ground state, variable amounts of energy while all the other modes had
ZPVE energy and analyzed the reaction dynamics. We also excited the

transition state with 0.05 kcal/mol and investigated the energy transfer as the

91

system relaxed to the dihydrogen bonded complex. The excitation energy
was deposited in the imaginary mode corresponding to the reaction path and
the amplitude and momenta of the modes were mapped onto the normal
modes of the dihydrogen bonded complex.

For the tunneling corrections a new IRC calculation, in cartesian
coordinates, was performed and the curve was fitted to a quartic potential:
V(r) = ar2 + br3 + cr4
The microscopic rate constant was defined as
14E] =(riE1)"e'Sm’“
where, E is the excitation energy. The well vibrational period, r[E], and the
semiclassical action, S[E], were given as
1031 = 2(m/2)ml[E-V(r)]'”2dr
SlEl/h = 2(2m)m/hl{[V(r)-E]dr} “2
For the quartic potential, the reactant well curvature was defined by
m. = 1V"(r>.= .../m1”
where r0 is the position of the reactant well minimum. The barrier curvature,
00), was expressed as
6.. = [Wm-elm)”

The barrier height was defined by

Vb = V(r = 0) —V( r = r0)

92

Application of all the above expressions led to the rate constant

k[E] = wo/(Zn)eXp{-2fl[Vb-El/(hwb)}

and when the excitation energy became close to the barrier

k[E] = (00/{2n{ 1+ exp{21r[Vb-E]/(h0)b)} }}

The above equations were used to calculate the rate constants when different

excitation energy values were considered.

93

5. Experimental Section

General

All reagents were purchased fiom commercial sources and purified when
necessary. Reductions were monitored by FTIR using a REACT IR MP
Mobile spectrometer. Calibration curves were obtained for all reactants. The
cis/trans diol ratios were analyzed by HPLC.

HPLC Column Characteristics:

Name: BIORAD Aminex HPX-87H

Conditions - mobile phase: 5 mM H2SO4 solution

low rate: 0.6 ml/min

temperature: 65 °C

detector: UV 210 nm

RI

injection size: 10 pl

Kinetic Studies of Reductions with Tetrabutylammonium Borohydride.
Standard solutions of 2-hydroxycyclopentanone (0.2M) and
tetrabutylammonium borohydride (0.8M) in 1,2-dichlorobenzene were
prepared. The solvent was purchased dry from Aldrich.
Tetrabutylammonium borohydride was recrystallized from ethyl acetate and

dried under vacuum for 24 h at room temperature. In a 50 ml three-neck

94

round bottom flask fitted with a magnetic stirring bar, a reflux condenser
connected to a nitrogen line, and a thermometer, 10 ml of the
tetrabutylammonium borohydride solution was placed and allowed to
thermally equilibrate. After the desired temperature was reached, 10 ml of
the ketone solution (kept at the same temperature) was added and the IR
spectra were recorded beginning 20 seconds after the mixing and at intervals
determined by the reaction length at a specific temperature.

When the carbonyl compound was cyclopentanone, to 20 ml solution of
tertabutylammonium borohydride (0.6M), 1.4 ml (0.016 moles)
cyclopentanone were added. Cyclopentanone was purchased from Aldrich
and redistilled at (bp = 130°C) at atmospheric pressure under nitrogen.

1H NMR (CDC13, 300 MHz): 1.66, multiplet (4H); 1.95, multiplet (4H).

IR: 1746 cm" (CO).

Preparation of 2-hydroxycyclopentanone148
A. l,2-Bis(trimethylsilyloxy)cyclopentene.
A 1 liter, three-neck flask was fitted with a mechanical stirrer, a dropping
funnel, a reflux condenser and maintained under an oxygen-flee, nitrogen
atmosphere. The flask was charged with 300 g reagent grade toluene (dried

over MgSO4) and 15 g freshly cut sodium. The temperature was slowly

increased to 110°C. The stirrer was operated at full speed to disperse the

95

sodium. A mixture of 75 g (0.69 mole) chlorotrimethylsilane, and 24 g (0.15
mole) dimethyl glutarate in 50 ml toluene was added over 3 hours. After
reacting for 12 hours, the mixture was cooled and filtered. The precipitate
was washed several times with anhydrous diethyl ether. Toluene was
removed under reduced pressure and the crude 1,2-
bis(trimethylsilyloxy)cyclopentene was vacuum distilled, b.p. 53°C (0.3 mm
Hg), (22.5 g, 61 . 1%).1H NMR(CD2C12, 300 MHz): 0.2, multiplet (18H);
1.0-2.0, multiplet (6H).

B. 2-Hydroxycyclopentanone.

In a 250 ml, three neck flask fitted with a magnetic stirrer, a dropping
funnel, a nitrogen bubbler and a reflux condenser, 115 ml reagent grade
methanol was placed. Dry oxygen-free nitrogen was vigorously bubbled
through the methanol for 1 hour. The 22.5 g of
l,2-bis(trimethylsilyloxy)cyclopentene were added dropwise for 2 hours at
room temperature. The methanol and methoxytrimethylsilane were removed
under reduced pressure, the crude 2-hydroxycyclopentanone was dissolved
in 10 ml water and a catalytic amount of concentrated H2SO4 was added.
After reacting for 6 hours the 2-hydroxycyclopentanone was extracted with
3 x 10 ml CH2C12. The solution was dried (over MgSO4) and the solvent

was removed under reduced pressure and the

96

2-hydroxycyclopentanone was vacuum distilled, b.p. 43°C, (0.9 mm Hg),
(5.6 g, 56%). IR: 3415 cm" (OH), 1746 cm" (CO); 1H NMR(CD2C12, 300
MHz): 4.03, triplet (ll-I); 3.02, singlet, broad (ll-I); 1.5-2.5, multiplet (6H)
Cis/trans diols Ratio Analysis

Standard solutions of carbonyl (0.5M) and tetrabutylammonium borohydride
(0.5M) in CH2C12 (dried over P205) were prepared. 0.5 ml of each solution
were added to a vial, shaken well, and then allowed to react for up to 6
hours. Out of each reaction mixture 0.1 ml were taken and treated with
concentrated H2SO4 and diluted up to 1 ml with the mobile phase that had as
internal standard ethylene glycol (0.25M). The additives
tetrabutylammonium fluoride (dried under reduced pressure at 70°C for 24
hrs), tetrabutylammonium chloride (recrystallized from CHzClzlEtzO) and
tetrabutylammonium bromide (recrystallized from CH2C12/Et20) were added
to the ketone solution before the addition of the tetrabutyl ammonium
borohydride solution. NaBI-I, was recrystallized from i-PrOH/CHzClz,

tetrahydrofuran was dried over P205 and LiBH4 used as purchased.

97

BYBLIOGRAPHY
1. Oref, L.; Rabinovitch, B. S. Acc. Chem. Res. 1979, 12, 166.
2. Zewail, A. H. Phys. Today 1980, 33, 27.
3. Zewail, A. H. Acc. Chem. Res. 1980, 13, 360.
4. King, D. S.; Hochstrasser, R. M. J. Am. Chem. Soc. 1975, 97, 4760.
5. Reddy, K. V.; Berry, M. J. Chem. Phys. Lett. 1979, 66, 223.
6. Hall, R.; Kaldor,‘ A. J. Chem. Phys. 1979, 70, 4027.
7. Lupo, D. W.; Quack, M. Chem. Rev. 1987, 87, 181.
8. Baumert, T.; Helbing, J .; Gerber, G. Adv. Chem. Phys. 1997, 101, 47.

9. Scherer, N. F.; Carlson, R. J .; Matro, A.; Ruggiero, M. Du; Romero-

Rochin, V.; Cina, J. A.; Fleming, G. R.; Rice, S. A. J. Chem. Phys.
1991, 95, 1487.

10.Shapiro, M.; Brumer, P. Chem. Phys. 1989, 139, 221.
11.Shapiro, M.; Brumer, P. J. Chem. Phys. 1986, 84, 4103.
12.Shapiro, M. Brumer, P. Acc. Chem. Res. 1989, 22, 407.

13.Fiss, J. A.; Zhu, L. C.; Suto, K.; He, G. Z.; Gordon, R. J. Chem. Phys.
1998, 233, 335.

14.Zhu, 1.; Suto, K.; Fiss, J. A.; Wada, R.; Seideman, T.; Gordon, R. J.
Phys. Rev. Lett. 1997, 79, 4108.

15.Shapiro, M.; Chen, Z.; Brumer, P. Chem. Phys. 1997, 217, 325..

16.Chen, Z.; Brumer, P.; Shapiro, M. Chem. Phys. Lett. 1992, 198, 498.

17.Kobrak, M. N.; Rice, S. A. Phys. Rev. A 1998, 57, 2885.

98

18.Kobrak, M. N.; Rice, S. A. J. Chem. Phys. 1998, 109, 1.

l9.Shnitman, A.; Sofer, I.; Golub, 1.; Yogev, A.; Shapiro, M.; Chen, Z.;
Brumer, P. Adv. Chem. Phys. 1997, 101, 285.

20.Shnitman, A.; Sofer, I.; Golub, 1.; Yogev, A.; Shapiro, M.; Chen, Z.;
Brumer, P. Phys. Rev. Lett. 1996, 76, 2886.

21.Abrashkevich, D. G.; Shapiro, M.; Brumer, P. J. Chem. Phys. 2002,
1 16, 5584.

22.Pierce, A. P.; Dahleh, M. A.; Rabitz, H. Phys. Rev. A 1988, 37, 4950.

23.Shi, S. H.; Woody, A.; Rabitz, H. J. Chem. Phys. 1988, 88, 6870.

24.Shi, S. H.; Rabitz, H. J. Chem. Phys. 1990, 92, 364.

25.Judson, R. S.; Rabitz, H. Phys. Rev. Lett. 1992, 68, 1500.

26.Gross, P.; Neuhauser, D.; Rabitz, H. J. Chem. Phys. 1993, 98, 4557.

27.Toth, G. J .; Lorincz, A.; Rabitz, H. J. Chem. Phys. 1994, 101, 3715.

28.Phan, M. Q.; Rabitz, H. Chem. Phys. 1997, 21 7, 389.

29.Phan, M. Q.; Rabitz, H. J. Chem. Phys. 1999, 110, 34.

30.Geremia, J. M.; Zhu, W. S.; Rabitz, H. J. Chem. Phys. 2000, 113,
10841.

31.Lewis, R. J .; Menkir, G. M.; Rabitz, H. Science 2001, 292, 709.

32.Assion, A.; Baumer, T.; Bergt, M.; Brixner, T.; Kiefer, B.; Seyfried,
V.; Strehle, M.; Gerber, G. Science 1998, 282, 919.

33.Brixner, T.; Damarauer, N. H. ; Niklaus, P.; Gerber, G. Nature 2001,
414, 57.

99

34.Henriksen, N. E. Chem. Soc. Rev. 2002, 31, 37.

35.Althorpe, S. C.; Femandez-Alonso, F.; Bean, B. D.; Ayers, J. D.;
Pomerantz, A. E.; Zare, R. N.; Wrede, E. Nature 2002, 416, 67.

36.Heiweil, E. J .; Casassa, M. P.; Cavanagh, R. R.; Stephenson, J. C. J.
Chem. Phys. 1986, 85, 5004.

37.Graener, H.; Ye, Q.; Laubereau, A. J. Chem. Phys. 1989, 90, 3413.

38.Graener, H.; Ye, Q.; Laubereau, A. J. Chem. Phys. 1989, 91, 1043.

39.Ye, Q.; Graener, H. J. Phys. Chem. 1989, 93, 5963.

40.Liddel, U.; Becker, D. E.; Spectrochimica Acta 1957, 10, 70.

41.Biondi, A.; Simkin, D. J. J. Chem. Phys. 1956, 25, 1073.

42.Laenen, R.; Rauscher, C.; Laubereau, A. J. Phys. Chem. A 1997, 101 ,
3201.

43.Barnes, A. J .; Hallam, H. E. Trans. Faraday Soc. 1970, 66, 1932.
44.Laenen, R.; Rauscher, C. J. Chem. Phys. 1997, 106, 8974.
45.Laenen, R.; Rauscher, C. J. Mol. Struct. 1998, 448, 115.

46.Woutersen, S.; Emmerichs, U.; Bakker, H. J. J. Chem. Phys. 1997,
107, 1483.

47.Coussan, S.; Bakkas, N.; Loutellier, A.; Perchard, J. P.; Racine, S.
Chem. Phys. Lett. 1994, 217, 123.

48.Bonn, M.; Bakker, H. J .; Kleyn, A. W.; van Santen, R. J. Phys. Chem.

1996, 100, 15301.

49.Iwaki, L. W.; Dlott, D. D. J. Phys. Chem. A 2000, 104, 9101.

100

50.Levinger, N.; Davis, P. H.; Fayer, M. D. J. Chem. Phys. 2001, 115,
9352.

51.Gaffney, K. J .; Piletic, I. R.; Fayer,M. D. J. Phys. Chem. A 2002, 106,
9428.

52.Gaffney, K. J .; Davis, P. H.; Piletic, I. R.; Levinger, N. E; Fayer, M.
D. J. Phys. Chem. A 2002, 106, 12012.

53.Bakker, H. J .; Woutersen. Phys. Rev. Lett. 1999, 83, 2077.

54. Kropman, M. F. ,Nienhuys, H. —K.; Woutersen, S.; Bakker, H. J. J.
Phys. Chem. A 2001, 105, 4622.

55.Nienhuys, H. —K.; Lock, A. J .; van Santen, R.; Bakker, H. J. J. Chem.
Phys. 2002, 117, 8021.

56.Rey, R.; Moller, K. B.; Hynes, J. T. J. Phys. Chem. A 2002, 106,
l 1994.

57.Strauss, H. L. Acc. Chem. Res. 1997, 30, 37.

58.Cha, Y. -H.; Strauss, H. L. J. Phys. Chem. A 2002, 106, 3531.

59.Choi, C. —H.; Kuwata, K. T.; Cao, Y. —B.; Okumura, M. J. Phys.
Chem. A 1998, 102, 503.

60.Schriver, A.; Racine, S.; Schrems, O.; Schriver-Mazzuoli, L.
Spectrochimica Acta 1994, 50A, 287.

61 .Schriver, L.; Schriver, A.; Perchard, J. —P. J. Mol. Struct. 1990, 222,
141.

62. Giebels, I. A. M. E.; van der Broek, M. A. F. H.; Kropman, M. F .;
Bakker, H. J. J. Chem. Phys. 2000, 112, 5127.

101

63.Custelcean, R.; Jackson, J. E. Chem. Rev. 2001, 101, 1963.
64.Epstein, L. M.; Shubina, E. S. Coordin. Chem. Rev. 2002, 231, 165.
65.Brown, M.P.; Walker P. J. Spectrochimica Acta 1974, 304, 1125.
66.Brown, M. P.; Heseltine, R. W. Chem. Commun. 1968, 1551.

67.Brown, M. P.; Heseltine, R. W.; Smith, P. A.; Walker, P. J. J. Chem.
Soc. A 1970, 410.

68.Shubina, E. S.; Bachmutova, E. V.; Saitkulova, L. N.; Epstein, L. M.
Mendeleev Commun. 1997, 83.

69.Epstein, L. M.; Shubina, E. E.; Bakhumotva, E. V.; Saitkilova, L. N.;
Bakhumutov, V. L.; Chistyakov, A. L.; Stankevich, I. V. Inorg. Chem.
1998, 37, 3013.

70.Shubina, E. S.; Belkova, N. V.; Bakhmutova, E. V.; Saitkulova, L. N.;
Ionidis, A. V.; Epstein, L. M. Russ. Chem. Bull. 1998, 47, 817.

71 .Jackson, J. E.; Huang, R.; Eckert, J .; Sharma, M.; Argyriou, D.;
Sheldon, R. Manuscript in preparation.

72.Richardson, T. B.; de Gala, S.; Crabtree, R. H. Siegbahn, P. E. M. J.
Am. Chem. Soc. 1995, 117, 12875.

73.Cramer, C. J .; Gladfelter, W. L. Inorg. Chem. 1997, 36, 5358.

74.Klooster, W.T.; Koetzle, T. F .; Siegbahn, P. E. M.; Richardson, T. B.;
Crabtree, R. H. J. Am. Chem. Soc. 1999, 121, 6337.

75.Singh, P.; Zottola, M.; Huang, S.; Show, B. R.; Pedersen, L. G. Acta
Cryst. 1996, C52, 693.

76.Popelier, P. L. A. J. Phys. Chem. A 1998, 102, 1873.

102

 

 

77.Rozas, I.; Alkorta, 1.; Elguero, J. Chem. Phys. Lett. 1997, 275, 423.

78.Alkorta, I.; Elguero, J .; Foces-Foces, C. J. Chem. Soc. Commun. 1996,
1633.

79.Crabtree, R. H.; Siegbahn, P. E. M.; Eisenstein, O. Rheingold, A. L.;
Koetzle, T. F. Acc. Chem. Res. 1996, 29, 348.

80.Atwood, J. L.; Koutsantonis, G. A.; Lee, F.-C.; Raston, C. L. Chem.
Comm. 1994, 91.

81 .Campbell, J. P.; Hwang, J. W.; Young, V. G.; Von Dreele, R. B.;
Cramer, C. J .; Gladfelter, W. L. J. Am. Chem. Soc. 1998, 120, 521.

82.Liu, Q.; Hoffmann, R. J. Am. Chem. Soc. 1995, 117, 10108.

83.Alkorta, 1.; Elguero, J.; Foces-Foces, C. Chem. Commun. 1996, 1633.

84.Remko, M. Molecular Physics 1998, 94, 839.

85.Kulkarni, S. A. J. Phys. Chem. A. 1998, 102, 7704.

86.Kulkami, S. A.; Srivastava, A. K. J. Phys. Chem. A 1999, 103, 2836.

87.Lundell, J.; Petersson, M. Phys. Chem. Chem. Phys. 1999, I, 1601.

88.Grabowski, S. J. Chem. Phys. Lett. 1999, 312(5-6), 542.

89.Grabowski, S. J. J. Phys. Chem. A 2000, 104, 5551.

90.Stevens, R.; Bau, R.; Milstein, D.; Blum, O.; Koetzle, T. F. J. Chem.
Soc. Dalton Trans 1990, 1429.

91 .Park, S.; Ramachandran, R.; Lough, A. J.; Morris, R. H. J. Chem. Soc.
Chem. Commun. 1994, 2201.

103

92.Lee, J. C.; Rheingold, A. L.; Miller, B.; Pregosin, P. S.; Crabtree, R.
H. Chem. Commun. 1994, 2001.

93.Peris, E.; Lee, J. C.; Rambo, J. R.; Eisenstein, O.; Crabtree, R. H. J.
Am. Chem. Soc. 1995, 117, 3485.

94.Lough, A. J .; Park, S.; Ramachandran, R.; Morris, R. H. J. Am. Chem.
Soc. 1994, 116, 8356.

95.Park, S.; Lough, Morris, R. H. Inorg. Chem. 1996, 35, 3001.

96.Wesse1, J .; Lee, J. C.; Peris, E.; Yap, G. P. A.; Fortin, J. B.; Ricci, J.
S.; Sini, G.; Albinati, A.; Koetzle, T. F.; Eisenstein, O.; Rheingold, A.
L.; Crabtree, R. H. Angew. Chem. Int. Ed. 1995, 34, 2507.

97.Pate1, B. P.; Wessel, J.; Yao, W.; Lee Jr., J. C.; Peris, E.; Koetzle, T.
F.; Yap, G. P. A.; Fortin, J. B.; Ricci, J. S.; Sini, G.; Albinati, A.;
Eisenstein, O.; Rheingold, A. L.; Crabtree, R. H. New J. Chem. 1997,
21, 413.

98.Sini, G.; Eisenstein, O.; Yao, W.; Crabtree, R. H. Inorg. Chim. Acta
1998, 280, 26.

99.Peris, E.; Wessel, J .; Patel, B. P.; Crabtree, R. H. Chem. Commun.
1995, 2175.

100.Desmurs, P.; Kavallieratos, K.; Yao, W.; Crabtree, R. H. New J.
Chem. 1999, 23, 1111.

101.Shubina, E. S.; Belkova, N. V.; Krylov, A. N.; Vorontsov, E. V.;
Epstein, L. M.; Gusev, D. G.; Niedermann, M.; Berke, H. J. Am.
Chem. Soc. 1996, 118, 1105.

102.Belkova, N. V.; Shubina, E. S.; Ionidis, A. V.; Epstein, L. M.;
Jacobsen, H.; Messmer, A.; Berke, H. Inorg. Chem. 1997, 36, 1522.

103.Messmer, A.; Jacobsen, H.; Berke, H. Chem. Eur. J. 1999, 5, 3341.

104

 

 

 

104.Scheiner, S.; Orlova, G. J. Phys. Chem. A 1998, 102, 260.

105.Cremer, D.; Kraka, E. J. Phys. Chem. 1986, 90, 33.

106.Matsunaga, N.; Gordon, M. S. J. Phys. Chem. 1995, 99, 12773.

107.Paulson, J. F.; Henchman, M. J. Bull. Am. Phys. Soc. 1982, 27, 108.

108.Miller, T. M.; Viggiano, A. A.; Stevens Miller, A. E.; Morris, R. A.;
Henchman, M.; Paulson, J. F.; Van Doren, J. M. J. Chem. Phys.
1994, 100, 5706.

109.de Beer, E.; Kim, B. H.; Neumark, D. M.; Gunion, R. F.; Lineberger,
W. C. J. Phys. Chem. 1995, 99, 13627.

110.Lee, J. C.; Peris, E.; Rheingold, A. L.; Crabtree, R. H. J. Am. Chem.
Soc. 1994, 116, 11014.

111.Ayllon, J. A.; Gervaux, C.; Sabo-Etienne, S.; Chaudret, B.
Organometallics 1997, 16, 2000.

112.0rlova, G.; Scheiner, S. J. Phys. Chem. Soc. 1994, 116, 11014.

113.Shubina, E. S.; Belkova, N. V.; Bakhmutova, E. V.; Vorontsov, E.
V.; Bakhmutov, V. I.; Ionidis, A. V.; Bianchini, C.; Marvelli, L.;
Peruzzini, M.; Epstein, L. M. Inorg. Chim. Acta 1998, 280, 302.

114.Epstein, L. M.; Shubina, E. S. Ber. Bunsenges. Phys. Chem. 1998,
102, 359.

115.0rlova, G.; Scheiner, S. J. Phys. Chem. A 1998, 102, 4813.

116.Grundemann, S.; Ulrich, S.; Limbach, H. —H.; Golubev, N. S.;
Denisov, G. S.; Epstein, L. M.; Sabo-Etienne, S.; Chaudret, B.;
Inorg. Chem. 1999, 38, 2550.

117.Chu, H. S.; Lau, C. P.; Wong, K. Y.; Wong, W. T. Organometallics
1998, 17, 2768.

118.Musashi, Y.; Sakaki, S. J. Am. Chem. Soc. 2000, 122, 3867.

105

119.Yamakawa, M.; Ito, H.; Noyori, R. J. Am. Chem. Soc. 2000, 122,
1466.

120.Abdur-Rashid, K.; F aatz, M.; Lough, A.; Morris, R. H. J. Am. Chem.
Soc. 2001, 123, 7473.

121 .Ayllon, J. A.; Sayers, S. F.; Sabo-Etienne, S.; Donnadieu, B.;
Chaudret, B. Organometallics 1999, I8, 3981.

122.Caballero, A.; Jalon, F. A.; Manzano, B. R. Chem. Commun. 1998,
1879.

123.Abdur-Rashid, K.; Gusev, D. G.; Lough, A. J .; Morris, R. H.
Organometallics, 2000, 19, 834.

124.Patel, B. P.; Kavallieratos, K.; Crabtree, R. H. J. Organomet. Chem.
1997, 528, 205.

125.Bosque, R.; Maseras, F.; Eisenstein, O.; Patel, B. P.; Yao, W.;
Crabtree, R. H. Inorg. Chem. 1997, 36, 5505.

126.Yao, W.; Crabtree, R. H. Inorg. Chem. 1996, 35, 3007.
127.Aime, S.; Gobetto, R.; Valls, E. Organometallics 1997, I6, 5140.

128.Aime, S.; Ferriz, M.; Gobetto, R.; Valls, E. Organometallics 1999,
18, 2030.

129.Aime, S.; Ferriz, M.; Gobetto, R.; Valls, E. Organometallics 2000,
19, 707.

130.Gatling, S. C.; Jackson, J. E. J. Am. Chem. Soc. 1999, 121, 8655.
131.Custelcean, R.; Jackson, J. E. J. Am. Chem. Soc. 2000, 122, 5251.
132.Mesmer, R. E.; Jolly, W. L. Inorg. Chem. 1962, 1, 608.

133.Schreiner, P.R.; Schaefer, H. F.; Schleyer, P. von Rague J. Chem.
Phys. 1995, 103, 5565.

106

134.Cuk_ier, R. I.; Zhu, J. J. Chem. Phys. 1999, 110, 9587.

135.Webb, S. P.; Iordanov, T.; Hammes-Schiffer, S. J. Chem. Phys.
2002, 117, 4106.

136. Private communication with Professor S. Hammes-Schiffer.
137.Davis, R. E. J.Am. Chem. Soc. 1962, 84, 892.

138.Kreevoy, M. M.; Hutchins, J. E. C. J. Am. Chem. Soc. 1969, 91,
4329.

139.Prepperberg, I. M.; Halgren, T. T.; Lipscomb, W. N. J. Am. Chem.
Soc. 1976, 98, 3442.

140.Garrett, E. R.; Lyttle, D. A. J. Am. Chem. Soc. 1953, 75, 6051.
141.Wigfie1d, D. G.; Gowland, F. W. Can. J. Chem. 1976, 56, 786.

142.Personal communication with dr. S. Gatling

143.Henderson, W. G.; How, M. J .; Kennedy, G. R.; Mooney, E. F.
Carbohydrate Res. 1973, 28, 1.

144.Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.;
Johnson, B. G.; Robb, M. A.; Cheeseman, J. R.; Keith, T. A.;
Peterson, G. A.; Montgomery, J. A.; Raghavachari, K.; Al-Laham,
M. A.; Zakrzewski, V. G.; Ortiz, J. V.; Foresman, J. B.; Cioslowski,
J .; Stefanov, B. B.; Nanayakkara, A.; Challacombe, M.; Peng, C. Y.;
Ayala, P. Y.; Chen, W.; Wong, M. W.; Andres, J. L.; Replogle, E.
S.; Gomperts, R.; Martin, R. L.; Fox, D. J .; Binkley, J. S.; Defi'ees,
D. J .; Baker, J .; Steward, J. P.; Head-Gordon, M.; Gonzales, C.;

Pople, J. A. Gaussian 94 (Revision D.3); Gaussian, Inc.: Pittsburgh,
PA, 1995.

145.Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, R.; Zakrzewski, J. R. V. G.; Montgomery Jr., J.
A.; Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.;
Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J .;
Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B. ; Pomelli, C.;

107

Adamo, C; Clifford, S.; Ochterski, J.; Petersson, G. A.; Ayala, P.
Y.; Cui, Q.; Morokuma, K.; Salvador, P.; Dannenberg, J. J .; Malick,
D. K; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.;
Cioslowski, J .; Ortiz, J. V.; Baboul, A. G.; Stefanov, B. B.; Lin, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R.
L.; Fox, D. J .; Keith, T.; Al-Laham, M. A.; Peng, C. Y.;
Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.;
Chen, W.; Wong, M. W.; Andres, J. L.; Gonzalez, C.; Head-
Gordon, M.; Replogle, E. S.; Pople, J. A. Gaussian 98 (Revision
A.l 1.1); Gaussian, Inc.: Pittsburgh, PA, 2001.

146.Schmidt, M. W.; Baldridge, K. K.; Boatz, J. A.; Elbert, S. T.;
Gordon, M. S.; Jensen, J. J .; Koseki, S.; Matsunaga, N.; Nguyen, K.
A.; Su, S.; Windus, T. L.; Dupuis, M.; Montgomery, J. A. J.
Comput. Chem. 1993, 14, 1347.

147.Taketsugu, T.; Gordon, M. S. J. Chem. Phys. 1995, 103, 10042.

148.Dr. S. Gatling thesis

108

 

 

)llllllllnl