Quantum Tunneling in Hydride Transfer Reactions in Solution

By Mortezaali Razzaghi, Master of Science

A Thesis Submitted in Partial Fulfillment of the Requirements for the Degree of Master of Science in the field of Chemistry

Advisory Committee: Dr.Yun Lu, Chair Dr. Chin-Chuan Wei Dr. Sarah Luesse

Graduate School Southern Illinois University Edwardsville August, 2013

UMI Number: 1549836

All rights reserved INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion.

UMI 1549836 Published by ProQuest LLC (2013). Copyright in the Dissertation held by the Author. Microform Edition © ProQuest LLC. All rights reserved. This work is protected against unauthorized copying under Title 17, United States Code

ProQuest LLC. 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, MI 48106 - 1346

ABSTRACT

QUANTUM TUNNELING IN HYDRIDE TRANSFER REACTIONS IN SOLUTIONS by MORTEZAALI RAZZAGHI

Chairperson: Professor Yun Lu

The secondary kinetic isotope effects for the hydride transfer reactions from aliphatic alcohols to four carbocations (NAD+ models) in acetonitrile were determined. The results suggest that the hydride transfer takes place by tunneling and that the rehybridizations of both donor and acceptor carbons lag behind the H-tunneling. This is quite contrary to the observations in alcohol dehydrogenases where the importance of enzyme motions in catalysis is manifested.

ii

ACKNOWLEDGEMENTS

I would like to extend my sincere appreciation to the following individuals, for without them this thesis would not have been possible. To Dr. Lu for your guidance in research, patience in data collection, and friendship throughout my tenure in your research lab .I would like to thank Dr. Sadegh Khazaeli who helped me to join SIUE. I would like to thank Dr. Wei and Dr. Luesse who are my committee members. I would like to thank the department of chemistry faculty at Southern Illinois University Edwardsville (SIUE) for their thoughtful conversations and help in my two years here, especially, Dr. O’Brien, Dr. Shabestary, Dr. Shaw. To the following graduate students that have worked with me in Dr. Lu’s research lab over the past two years: Blake Albert Hammann, Qi Liu, Sadra Kashefolgheta, Cindy Shaw, Mortaza Derakhshani Molayousefi, Mahdi Raghibi Broujen, Mina Jafari, Binita Maharjan, Jonathan Lefton and Scott Alvarado. To my mother and father, thank you for pushing me to pursue my master’s degree and for supporting me in everything that I have done, and making me who I am today. To all of my friends and family, thank you for your support, encouragement, and the interest in my research. I would like to dedicate this thesis to my parents for their loving support, encouragement, and guidance in my life.

iii

TABLE OF CONTENTS ABSTRACT ........................................................................................................................ ii ACKNOWLEDGEMENTS ............................................................................................... iii LIST OF FIGURES ........................................................................................................... vi LIST OF TABLES ........................................................................................................... viii Chapter I. INTRODUCTION ...................................................................................................... 1 1.1 Hydrogen Transfer Reactions and our General Research Purpose .................. 1 1.2 Mechanistic Analysis Tools and Theories Used in the Study of the ADHs ..... 4 1.2.1 Kinetic isotope effects (KIEs) ........................................................ 4 1.2.2 Equilibrium isotope effects (EIEs)................................................. 6 1.2.3 Linear-free energy relationships (LFER): Hammett correlation analysis .................................................................................... 7 1.2.4 Hammond’s postulate .................................................................... 9 1.2.5 The Rule of geometric mean (RGM) ........................................... 10 1.3 Semi-Classical Limits on Kinetic Isotope Effects............................................ 11 1.4 Semi-Classical Limits on the Relationship between D-KIE and T-KIE Swain- Schaad Relationships ................................................................. 13 1.5 Semi-Classical Limits on the Temperature Dependence of KIEs ................. 13 1.6 Failure of the Semi-Classical Model: Quantum Tunneling Models .............. 14 1.7 Marcus-Like Hydrogen Tunneling Model: Full Tunneling Model of Hydrogen Transfer ................................................................................ 20 1.8 Alcohol Dehydrogenase................................................................................. 27 1.9 Significance of Research ............................................................................... 30 II. EXPERIMENTAL ...................................................................................................... 34 2.1 Instrumentation .............................................................................................. 34 2.2 Materials ........................................................................................................ 34 2.3 Preparation of Reaction Solvent .................................................................... 39 2.4 Kinetic Study Procedure ................................................................................ 39 III. RESULTS AND DISCUSSION ................................................................................ 44 3.1 The Imbalanced Transition States .................................................................. 45 iv

3.2 The H-Tunneling Mechanism......................................................................... 47 IV. CONCLUSIONS ....................................................................................................... 49 REFERENCES ................................................................................................................. 51

v

LIST OF FIGURES Figure

Page

1-1: Reaction Coordinate and Free Energy Difference (ΔG). .............................................2 1-2: Differing Zero-Point Energies of Protium- and Deuterium-Substituted Molecules as the Cause of Primary Kinetic Isotope Effects ..................................................6 1-3: Correlation of Acid Dissociation Constants of Benzoic Acids with Rates of Basic Hydrolysis of Ethyl Benzoates..............................................................................8 1-4: Hammond's Postulate and Reaction Energy Diagram ...............................................10 1-5: Semi-Classical Model of KIEs. ..................................................................................12 1-6: Transition-State Structure for the Glucose-6-Phosphate Reaction. ............................15 1-7: An Example of the Ground-State Tunneling along the Reaction Coordinate ...........16 1-8: One-Dimensional Tunneling Correction Model. ........................................................17 1-9: Transition State Theory Model with Tunneling Correction. ......................................19 1-10: An Illustration of a Full Tunneling Model for Hydrogen Transfer Reaction Based in Equation (1-20) ...................................................................................25 1-11: The Free Energy Profile and Arrhenius Plots for a C-L Bond Breaking, Where L is H, D or T ...................................................................................................26 1-12: Oxidation of an Alcohol to Relevant Ketone or Aldehyde Catalyzed by ADH with NAD+ as Cofactor. ...................................................................................27 1-13: The Reaction of Hydride Transfer from an Alcohol to Xanthylium Carbocation (NAD+ Model) to Form Relevant Ketone and Xanthene to Mimic the Rate Determining Step of the Hydride Transfer Reactions in ADH. ............32 1-14: Xanthylium Carbocation Structure as a Model of NAD+ Cofactor. .........................33

vi

Figure

Page

2-1: Synthesis of 9H (D)-Xanthen-9-ol. ............................................................................35 2-2: Synthesis of 9H (D)-Thioxanthen-9-ol. ......................................................................36 2-3: Synthesis of 9H(D)-Xanthylium Perchlorate. ............................................................37 2-4: Synthesis of 9H(D)-Thioxanthylium Perchlorate .......................................................38 2-5: Synthesis of αH and αD-Cyclohexanol ......................................................................39 2-6: Kinetic Scans Determined by Analysis of the Reaction of αD-IPA (0.012M) with Xn(H)+(3.75×10-5M) at 60oC. .....................................................................41 2-7: Plot of Absorbance vs. Time of the Reaction of αD-IPA (0.012M) with Xn(H)+(3.75×10-5M) at 60oC. .............................................................................42 2-8: Determination of the Rate Constant of the Pseudo First Order Reaction of αD-IPA (0.012M) with Xn(H)+(3.75×10-5M) at 60oC. ......................................43

vii

LIST OF TABLES Table

Page

1-1: Selected Properties of the Hydrogen Isotopes. .............................................................1 1-2: De Broglie Wavelengths...............................................................................................3 1-3: Semi-Classical Behaviour of Hydrogen Effects. ........................................................14 1-4: Violations from Semi-Classical and Tunnel Correction Models ...............................20 3-1: 1o and 2o KIEs for the Hydride Transfer Reactions from Aliphatic Alcohols to the Hydride Acceptor Cations. ..........................................................................45

viii

1

Chapter I

Introduction 1.1

Hydrogen Transfer Reactions and our General Research Purpose

Hydrogen is the lightest element in the periodic table. It possesses three isotopes which are protium (H), deuterium (D) and tritium (T) (Table 1-1). These isotopes show greater difference than the isotopes of other elements in terms of their physical and chemical properties. The primary difference is in their mass, which affects the fundamental bond vibrational frequencies and thus zero point energies (ZPE) to the largest extent1. Table 1-1: Selected properties of the hydrogen isotopes1. Isotope Protium Deuterium Tritium 1 2 3 H or H H or D H or T Symbols Natural abundance 99.985% 0.0156% < 1 in 1017 atoms 1.0078 2.0141 3.0160 Isotopic mass /u

The hydrogen transfer reactions, in a form of proton, hydrogen atom and hydride appear common in the biological systems. Examples include hydrogen transfers as a hydrogen atom in soybean lipoxygenase-1 (SLO)2 and proton and hydride transfers in reactions catalyzed by thymidylate synthase (TSase)3 and alcohol dehydrogenase (ADH)4. Enzymes enhance the rate of the chemical reactions by many orders of magnitude with extraordinary selectivity5. Most enzymes are proteins (with exception of a few catalytic RNA molecules) and catalyze almost every biochemical reaction. Enzymes are classified by the type of the reaction they catalyze. For example, the enzymes which catalyze

2 transfer of electrons (hydride ions or H atoms) are called oxidoreductases. SLO, TSase and ADH are examples of oxidoreductases. An enzymatic reaction can be described by energy profile throughout the reaction (Figure 1). A simple enzymatic reaction can be written as: 1-1 Where E, S and P represent enzyme, substrate and product, respectively; the ES and EP are the transient complexes of the enzyme with substrate and products. The energy hill in the reaction coordinate diagrams called the transition state (TS), which is an extremely unstable state that involves bond breakage and/or formation, and/or charge development in order to form the product6.

Figure 1-1: Reaction coordinate and free energy difference (ΔG).6 The TS stabilization is the key point of the rate enhancement in the enzyme-catalyzed reactions. The energy gain of the TS structure (ΔGuncat-ΔGcat) is very important in multiple disciplines, such as drug design and life sciences. Different models have developed for enzyme-substrate interactions; the Pauling’s theory about the TS stabilization has

3 been recognized to be the most important principle for the chemical rate enhancement by enzymes7. In order to understand the origin of the TS stabilization one needs to study the mechanism of enzyme-catalyzed reactions. ADHs containing coenzyme NAD+, which catalyze the oxidation of an alcohol to aldehyde, have become one of the paradigmatic systems for studying the H-transfer reactions. In the past, it has been found that the classical theories cannot explain the kinetic results of this enzymatic system 5. The electron tunneling in the biological molecules has been recognized and accepted for many years8 and a similar recognition for hydrogen in the enzymatic reactions may have a similar phenomenon9. The mass of hydrogen is 1750-fold heavier than the electron, but the distance between hydrogen acceptor and donor (e.g. 3.2-3.5 Å between the two carbon atoms in C-H-C or 2.5-2.7 Å between the two oxygen atoms in O-H-O) results in 0.5 to 1.0 Å distances in which hydrogen will be transferred8,10. The non-classical property of the hydrogen in the enzymatic reactions becomes apparent when the H-transfer distance is comparable to the de Broglie wavelength of 0.63 Å (Table 1-2)11. Therefore, the matching of a particle’s wavelength to the distance (energy barrier width) that it must cross is the key feature of its quantum (tunneling) behavior12. Table 1-2: De Broglie wavelengths; λ=h/(2mE)1/2 , where m is the mass of the particle and E=20kJ/mol.11 Particle property eH D T C Mass (amu) 1/1750 λ (Å)

27

1

2

3

12

0.63 0.45 0.36 0.18

4 The importance of the electrostatic interactions, hydrogen bonding and desolvation in TS stabilization and catalysis has long been recognized. But the effects of heavy atom motions (protein dynamics) on the reaction coordinate and rate enhancement of enzymatic reactions remain controversial13. The aim of this thesis is to study the mechanism of the hydride transfer reactions from secondary alcohols to a NAD+ model compound, which mimics the model of ADH. Comparison of the results of ADH catalyzed reactions to those of the uncatalyzed reactions derived using the same mechanistic tools gives an insight into the role of the physical features of the hydrogen tunneling and enzyme motions in ADH as well as in general enzyme catalysis.

1.2

Mechanistic Analysis Tools and Theories Used in the Study of the ADHs

1.2.1

Kinetic isotope effects (KIEs)

Substitution of an atom by its isotope is a very valuable tool to study the mechanism of the reactions. Isotopic substitution is applicable to any nuclei, but most studied are for hydrogen transfer, because its isotopes have the largest relative mass difference. The fact that the isotopically substituted molecules react at different rates is called the kinetic isotope effect (KIE). There are two types of isotope effects termed as primary and secondary. The KIE is expressed as the ratio of the reaction rates of the light (L) and heavy (H) atoms (KIE=kL/kH). The primary KIEs (1o KIEs) correspond with processes in which an isotopically labeled bond is made or broken in the rate determining step14.

5 The assumption of the Eyring’s theory for reaction rate15 is that the reactants are in equilibrium with transition state, and in the rate determining step of the reaction TS decomposes into the products. The difference of the TS with ordinary chemical molecules is in that one of its vibrations has been replaced by an internal translation. In a typical reaction which involves hydrogen transfer from a C-H bond in a molecule (donor) to another (acceptor), the C-H stretching vibration becomes translational motion. The 1o KIE originates from the zero point energy (ZPE) difference between carbon-hydrogen (C-H) and carbon-deuterium (C-D) bond which occur when the reactants are converted to the TS. The stretching vibration of the C-H and C-D bond is quantized with the frequency νH and νD, and with the associated zero point vibration energy of 1/2hνH and 1/2hνD, respectively. The stretching frequency for C-H vibration is near 2900 cm-1, the ZPE is about 4.15 kcal/mole. The corresponding stretching frequency for C-D vibration is about 2100 cm-1, the ZPE is about 3.0 kcal/mole. The difference between these two ZPEs is 1/2h (νH-νD) ≈ 1.15 kcal/mole. In TS this stretching vibration becomes translational degree of freedom and the vibration is lost (1/2h (ν *H-ν*D) = 0), where ν*H and ν*D are the (zero) frequencies for the stretching vibrations of the C-H and C-D bonds in the TS, respectively. Therefore, the net difference in the activation energy for the reaction of C-H and C-D is 1.15 kcal/mole, at 300 K according to the Eyring’s equation (Equation 1-2), the calculated 1o KIE is about 7.

⁄

1-2

where ΔG⧧ is the free energy difference of activation, κ is the transmission coefficient, kB is the Boltzmann’s constant, h is the Planck’s constant, R is the gas constant and T is the absolute temperature16.

6

Figure1-2: Differing zero-point energies of protium- and deuterium-substituted molecules as the cause of primary kinetic isotope effects.

The secondary (2o) KIEs is observed when the isotopically substituted atom bond is not broken in the reaction. The values of 2o KIEs are smaller than those of 1oKIEs and are usually in the range from 0.7 - 1.5. They can be normal (2o KIEs > 1) or inverse (2o KIEs < 1). Depend on the position of the substituted isotope; they can be as α- or β-2o KIEs. In the α-2o KIEs the isotopic atoms are attached to the reaction center of the reactant (the atom undergoing bond cleavage), and in the β-2o KIEs the isotopically labeled atom is attached to the adjacent position of the one undergoing bond cleavage. In general the 2oKIEs are due to the changes in hybridization and hyperconjugation14. 1.2.2

Equilibrium isotope effects (EIEs)

Isotopic substitution can also alter the equilibrium (thermodynamic isotope effect), which is termed as equilibrium isotope effects (EIEs). Measuring EIE is a powerful tool for the investigation of enzyme kinetic mechanisms. EIE originates from the change in

7 bond order of the reactants to the products, while KIE results only from the changes from the ground state to the TS17. Comparison of the 2o KIEs with EIEs give valuable information about the TS structure. Since the 2o KIE results from the changes in hybridization along the reaction coordinate, a 2o KIE close to unity indicates a very early transition state and a 2o KIE close to EIE predicts a late transition state18. 1.2.3

Linear-free energy relationships (LFER): Hammett correlation analysis

Substituent groups and chemical properties of compounds are correlated. For example, substituent groups affect the acidic strength and the stability of carbocation, carbanion and radical species. The effect of substituent groups on reactivity can be used to give insight into the reaction mechanisms. In 1930s, Hammett reported a numerical linear relationship between the acid strengths of substituted benzoic acid and the rates of many other chemical reactions, such as the hydrolysis of substituted ethyl benzoate (Figure 13).19 Figure 1-3 shows the linear relationship between log k/ko against log K/Ko, where k is the rate constant for hydrolysis of a substituted ethyl benzoate, ko is the rate constant of the hydrolysis of the unsubstituted ethyl benzoate, K is the acidic dissociation constant of the substituted benzoic acid and Ko is for the unsubstituted benzoic acid. In Figure 1-3 the slope of the line is defined as the following equation.

(

In Equation (1-3) substitution of

)

and

(

)

1-3

gives

which

shows the direct proportionality in free-energy changes that is called free-energy relationships (LFERs)20.

8

Figure 1-3: Correlation of acid dissociation constants of benzoic acids with rates of basic hydrolysis of ethyl benzoates19.

In 193721 Hammett presented LFERs for equilibrium and rate data respectively as:

(

)

1-4

(

)

1-5

The σ and ρ are the substituent effect and reaction constant respectively. The σ indicates collective measure of the total electronic effect of the substituent and the magnitude of the ρ reflects how sensitive the reaction is to the electronic effect of substituent. Hammett equation shows that for electron-withdrawing groups (EWD) tron-donating groups (EDG)

and for elec-

In the Hammett equation ρ is positive for all reac-

tions that are favored by EDGs and negative for those favored by EWGs. The sign and

9 magnitude of the ρ provides information about the TS structure and intermediates for the reaction. 1.2.4

Hammond’s postulate

Transition state is not an intermediate and thus is not feasible to detect during the reaction. Therefore it is impossible to prove its structure experimentally. Hammond’s postulate relates the transition state structure to the structures of reactants, intermediates and products in individual steps in reaction mechanisms22. Hammond’s postulate states: “If two states, as for example, a transition state and an unstable intermediate, occur consecutively during a reaction process and have nearly the same energy content, their interconversion will involve only a small organization of the molecular structures.” This postulate is valuable for the elementary processes which are highly endothermic or exothermic. In the exothermic steps it predicts that there is early transition state (the transition state structure resembles reactants) and for endothermic single step process a late transition state (the transition state structure resembles products) will be expected.23 Hammond’s postulate can be depicted as in Figure 1-4.

10

Figure 1-4: Hammond's postulate and reaction energy diagram: (a) early TS resembles reactant, (b) midpoint TS resembles neither reactant nor product, (c) late TS resembles intermediate or product22.

1.2.5

The Rule of geometric mean (RGM)

The rule of the geometric mean (RGM) states that there is no isotope effect on isotope effect. In other words, the two isotopes should exert their effects independently. Thus, secondary KIE will be independent of the primary isotope effect. Equation (1-6) shows the RGM as:

(

)(

)

1-6

where the subscript refers to the primary isotope (transferred atom) and superscript to the secondary isotope (nontransferred atom). This equation is true only if:

11

(

)

(

)

1-7

The RGM is valid in the classical systems, in which transferred isotope should not affect the secondary KIE. However it has been found that this is not valid in enzymatic systems, specifically ( )is smaller than( )

1.3

24

Semi-Classical Limits on Kinetic Isotope Effects

Equation (1-2) can be written as:

1-8 Where the Q and Qǂ are total partition functions of the GS and TS respectively and ΔE is the energy difference between GS and TS. As mentioned above, a KIE originates mainly from ZPE difference, in the semi-classical model, According to the Figure 1-2, the heavier isotopes have the lower vibrational ZPEs at both GS and TS. To reach the TS the molecules containing the heavier isotopes require higher thermal activation energy than molecules containing the lighter isotopes. Based on the KIE definition and ZPE differences, Equation (1-8) can be written as follows:

1-9 where ΔΔGǂ is the difference in the free energy difference of the activation between heavy and light isotopes.12,25

12

Figure 1-5: Semi-classical model of KIEs. Different activation energies (ΔEa) for H, D and T resulting from their different ZPEs at the GS and TS 17.

This semi-classical model of KIE provides the upper limit for primary KIEs (at 25 ᵒC and C-H frequency of 3000 cm-1), for carbon-hydrogen bond cleavage it is 6.9 for kH/kD and 15.8 for kH/kT26. In the semi-classical theory, the magnitude and being normal or inverse 2o KIEs in many types of reactions are related to the EIE. The position of the TS is determined by comparison of the 2o KIEs and EIEs, when the 2o KIEs are very close to unity, an early TS is suggested and if it is very close to the EIEs, a late TS is predicted.25

13

1.4

Semi-Classical Limits on the Relationship between D-KIE and T-KIE - Swain-Schaad Relationships

The Swain-Schaad relationships have been used as mechanistic tools. The SwainSchaad equation relates the KIEs of three hydrogen isotopes. Semi-classical theory shows the relationships of hydrogen isotopes by a simple exponential which only depends on the reduced mass of the different isotopes and ZPE difference between isotopes27:

1.5

(

)

1-10

(

)

1-11

(

)

1-12

Semi-Classical Limits on the Temperature Dependence of KIEs

For enzymatic reactions explained by the TS theory, it assumes there is a dynamic equilibrium between the ground state and transition state, according to the Boltzmann distribution. The rate of the reactions and temperature dependence of KIEs is frequently expressed via an Arrhenius equation, which is an empirical equation and relates rate of the reactions to temperature as follows: 1-13 where k is the rate constant, Ai is pre-exponential factor (the limit of the rate at infinite temperature) of particle i, Ea is the activation energy that is needed to go from reactants

14 ground state to the transition state, R is the universal gas constant and T is absolute temperature. In infinite temperature, the Arrhenius pre-factor for protium transfer (AH) is not expected to be very different from the values for AD (Arrhenius pre-factor for deuterium) and AT (Arrhenius pre-factor for tritium).28 Since KIE is ratio of the rate constants of light isotope (L) and heavy isotope (H), the temperature dependence KIE is17: (

)

1-14

Table (1-3) summarizes the semi-classical limits of the hydrogen isotope effects for a C-H bond on Arrhenius pre-exponential factors, magnitude of KIEs and Swain-Schaad relationships. Table 1-3: Semi-classical behaviour of hydrogen effects. Arrhenius pre-exponential factors Reference ⁄ ⁄ ⁄ [17] 0.3-1.7 0.5-1.4 0.7-1.2 KIE 1 < KIE < EIE EIE < KIE < 1 o 1 1.

These models fail to explain temperature independent small KIEs, with a large activation energies for isotopically sensitive step. Table (1-4) summarizes some enzymatic reactions which violate the semi-classical and tunnel correction model’s predictions28. According to the Table (1-4) there are KIEs in some systems which greatly exceed the Semiclassical limit, and in all of the examples in this table, the AL/AH surpassed unity. In attempt to explain these observations, many scientists adapted the Marcus theory of electron tunneling to the situation of H-tunneling36.

20 Table 1-4: Violations from semi-classical and tunnel correction models28. Parameter c

Expected valuesa

EXP

≤ 4.5

1o KIE

6.9

AH/AD

0.7-1.2

Observed values

Enzymeb

Reference

10

YADH htADH above 30 o C SLO MMO AADH SLO MADH SADH ChOx PRO htADH above 30 o C htADH below 30 o C

11

16 80 100 20 18 13 5.8 11 3.6

AH/AT

0.6-1.7

4.1

AH

1013 s-1

1017 s-1

a

39 40 41 42 40 43 44 45 46 39 39

Predicted values are based on Bell 47 and Melander and Saunders 26. b YADH: yeast alcohol dehydrogenase; htADH: thermophilic alcohol dehydrogenase; SLO: soybean lipoxygenase; MMO: methane monooxygenase; AADH: aromatic amine dehydrogenase; MADH: methylamine dehydrogenase; SADH: sarcosine dehydrogenase; ChOx: Choline Oxidase; POR: protochlorophyllide oxidoreductase. c EXP refers to Swain-Schaad relationships (interrelationships between KH, KD, KT) 27.

1.7

Marcus-Like Hydrogen Tunneling Model: Full Tunneling Model of Hydrogen Transfer

Rudolph A. Marcus won Nobel Prize in chemistry in 1992 for his fundamental contributions to the theory of electron transfer reactions in chemical systems.48 He discovered that the fluctuations in reaction media play important role in the electron transfer reactions. The rate constant k of a reaction is given by: ⁄

1-16

ΔG*is given by:

(

)

1-17

21 where A is the pre-exponential factor, ΔG0 is the standard free energy difference of reaction and λ is the “reorganization term”, composed of solvational (λo) and vibrational (λi) components. 1-18 In order to rationalize the deviations from the above models, scientists used Marcus theory of electron tunneling49 as a model to explain the hydrogen transfer. They termed this new model as Marcus-like hydrogen tunneling model. This model is a full tunneling model in which the Marcus electron tunneling model modified by adding a term which accounts for the fluctuations of the donor-acceptor distance (DAD).36 This model is termed with a various name in literature such as “vibrationally enhanced tunneling” model50, “environmentally coupled tunneling” model 51, “protein promoting vibration” model52 and others17, 53. The full tunneling model derived from the work by Kuznetsov and Ulstrup54 and used to reproduce temperature dependent KIEs in soybean lipoxigenase2. The major feature of the Marcus-like hydrogen tunneling model is the separation of the light atom motion (for example hydrogen tunneling) from the heavy atom motions, which includes protein environmental motion and the surrounding solvent. In this model there are three major processes that govern the reaction rates. They are: 1) heavy atom motion that causes the system to be in the tunneling ready state (TRS). It means that the heavy atom motions make the reactants and products potential energy surfaces degenerate so the tunneling can take place. This process is the mass (isotope) independent and temperature dependent called the Marcus term (reorganization term, λ) which is indicated by the pre-integral term (first exponential) in Equation (1-19) and Panel A in Figure (1-10); 2) tunneling of the transferring atom (hydrogen or deuterium) takes place

22 in accord with the Frank-Condon principle which this process is indicated by second exponential term in Equation (1-19) and Panel B in Figure (1-10). In hydrogen (and not electron) transfer reactions the probability of the tunneling is determined by the hydrogen wave function overlap. According to the second exponential in the equation (1-19) which describes the wavefunction overlap, the overlap decreases exponentially with increasing the mass. In this process the heavy atom motion modulates the fluctuations of the tunneling barrier; and 3) the third process in Marcus-like hydrogen tunneling model is that the hydrogen tunneling actually takes place through the barrier. This process is termed as a gating term in the equation (1-19) and when the hydrogen donor and acceptor reached enough close together the hydrogen wavefunction overlaps increases and then the probability of tunneling increases and tunneling takes place.36, 28

(

)

(

) ⁄

⁄

∫

⁄

1-19

In Equation (1-19) Const. is a constant that depends on the electronic coupling, ΔGo is the reaction driving force, λ is the reorganization energy (Marcus term), R is the gas constant, T is the absolute temperature, mH is the mass of the transferring particle, wH is the frequency of the transferring particle, rH is the transferring distance, Planck’s constant divided by 2π, Ex is the potential energy ( tion of the donor-acceptor coordinate rx (

√

is the ) as a func-

), and kb is the Boltzmann con-

stant.55 In Equation (1-19) the factors outside of the integral determines the rate of reaching a tunneling ready state (TRS) which is governed by heavy atom motions and it is isotopically insensitive (mass independent). Therefore the temperature dependence of reaction

23 rates is related to this factor. The terms inside the integral (the Boltzmann and tunneling terms) are mass sensitive (isotopically sensitive), which their temperature dependence depends on the free energy as a function of DAD. Thus the Marcus-like hydrogen tunneling model can explain temperature-dependent or temperature-independent KIEs with temperature-dependent or temperature-independent rates.25, 36 Figure (1-10) presents the effects of the heavy atom motions on the ZPE of the reactant (blue) and product (red). Three different panels shows the potential energy surfaces of the reactants and products along the reaction coordinate. The heavy atom motions coordinate is shown in Panel A which is known as the Marcus term or reorganization term, and Panel B is related to the position of the hydrogen atom. The hydrogen atom position coordinate is orthogonal to the heavy atom coordinate. The top panels show that the hydrogen atom is placed in the reactant energy well and due to the ZPE differences between the reactant and product they are not degenerate. The system reaches to the tunneling ready state (TRS) when the reactants and products are degenerate, and such degeneracy is necessary for tunneling of hydrogen atom. The middle panels of A and B shows a TRS states which is the result of the heavy atom motions;, it means that the heavy atom motions bring the system to the TRS with high tunneling probability. After tunneling the degeneracy of potential wells of the reactant and products breaks down and the hydrogen goes to the product side which is shown in the bottom panels. The reorganization energy (λ) of the heavy atom motions and the reaction driving force (ΔGo) determine the rate of reaching the TRS. The effect of the donor-acceptor distance (DAD) on the hydrogen atom wave function overlap at TRS showed in middle of the panel C in Figure (1-10). The magnitude of the overlap integral of the hydrogen atom

24 wave functions on the reactant (blue) and product (red) potential energy wells (bottom panel C) determines the transmission probability (P). The panel C shows the fluctuations in the DAD and hydrogen transfer at each DAD as a function of the transmission probability and the population at each DAD are presented. The middle part of Panel C shows the consequences for hydrogen wave function overlap at three different DADs. In Panel C the vertical dashed line shows that the ZPE is greater than the energy barrier height in the relevant DAD and in practice the reaction goes over the barrier.25 The most important part of the Marcus-like hydrogen tunneling model is that it separates the coordinates of the heavy atom motions from wave function overlap. The wave function overlap depends on the mass of the transferred particle which is independent of temperature and occurs only after heavy atom environment prepare the condition for tunneling.56

25

Figure 1-10: An illustration of a full tunneling model for hydrogen transfer reaction based in equation (1-20). The effects of the heavy atom motions on the zero point energy of the reactant (blue) and product (red) potential energy well.25

So far we discussed three different models to explain the experimental data of KIEs. Figure (1-11) show the comparison between three models. Figure 1A and 1B shows models which explain KIEs based on the ground-state differences in ZPEs or tunneling corrections respectively. The aberrant experimental observations of KIE is soybean lipoxygenase changed the theoreticians view regard to the degree of tunneling from the ground-state tunneling (tunneling from the deep potential wells) which is showed in Figure 1C.54, 56

26

Figure 1-11: The free energy profile and Arrhenius plots for a C-L bond breaking, where L is H, D or T. (A) Semi-classical KIEs which predicts that at higher temperature the Arrhenius prefactors are close together and being independent of isotopically different labels. (B) Semi-classical KIEs with a tunnel correction which predicts a large value for both Ea and AL in the case of heavier isotope(s). (C) Full tunneling model which predicts no or little difference in the magnitude of Ea among the isotopes and AL that are greatly elevated for H in relation to D or T56.

Interpreting KIEs using Marcus–like tunneling models suggest that the temperature dependence of the DADs sampling determines the temperature dependence of KIEs. It explains that when there is a very narrow distribution of DADs (closely packed reaction active site) at TRS, KIEs are temperature independent because the DADs distribution is temperature independent. In the case of temperature dependent KIEs, this model explain that there is a loose active site and a wide range of DADs are available for TRS, so the distribution of DADs is sensitive to temperature 25. In the next part we are going to use Marcus-like hydrogen tunneling model to explain the experimental results in alcohol dehydrogenase which is one of the most investigated enzymatic H-transfer systems.

27

1.8

Alcohol Dehydrogenase

Alcohol dehydrogenases (ADHs) are the enzymes for which scientists studied the Htunneling in the hydride transfer reactions. The reaction of oxidation of an alcohol to an aldehyde or ketone is catalyzed by ADH enzymes. Nicotinamide is a cofactor of ADH as shown in Figure (1-12). The ADH enzyme (for example, the yeast ADH) serves as an excellent model system for studying the mechanism of an enzyme catalyzed hydrogen transfer reaction. Kinetic isotope studies by Klinman and coworkers on the yeast ADH (yADH) showed that the oxidation of aromatic alcohol to aldehyde by NAD+ cofactor is the rate determining step.57 Furthermore, it is thermodynamically possible to study this reaction in both forward (alcohol to aldehyde) and reverse (aldehyde to alcohol) directions under the same conditions.57, 58

Figure 1-12: Oxidation of an alcohol to relevant ketone or aldehyde catalyzed by ADH with NAD+ as cofactor.

Klinman et al. studied the mechanisms of the hydride transfer in yADH using benzyl alcohols and aldehydes. They used two mechanistic study tools, kinetic isotope effects and linear free energy relationships (Hammett correlations), in order to elucidate the transition state structure.57, 58, 59 The 2o KIEs and the substituent effects results were contradictory. In the oxidation of benzyl alcohol to the aldehyde the 2 o KIEs is close to

28 EIE which is 1.35 (late TS). On the other hand, in the reverse reaction the 2o KIEs is close to unity (early TS). Therefore, the 2o KIE results suggest that the TS structure resembles aldehyde rather than alcohol59. The LFER studies contradict the KIE results. Using the para-substituted benzyl alcohols and benzaldehydes, the rate constant measurements give insight into the electronic effects of different substituents. The results from LFERs suggest more alcohol-like TS than aldehyde.57, 58 The comparison of the KIEs and LFERs results indicates deviations from the classical models in this reaction, so the tunneling effect is used to interpret these contradictories.25 Other measurements of the α-2o KIEs on the NAD+ cofactor carried out by Cleland and coworkers. They measured the α-2o KIEs of the reduction of NAD+ to NADH (sp2 to sp3) for which the relevant EIE should be inverse (0.89 < EIE < 1). They obtained a significantly inflated α-2o KIEs which was greater than unity. These surprising results were interpreted as a 1o-2o coupled motions .29, 60 The computational studies of this system also supported their observation along with the tunneling effect contribution to these non-classical behavior. It has been documented that the part of the catalytic power of the hydrogen transfer enzymes not only comes from the alteration of the height of the reaction energy barrier, but also from altering the contribution of tunneling which is related to the shape of the energy barrier.61 The rule of the geometric means (RGM) (which states that the secondary isotope effect is independent of the transferring isotope in the primary position) is one of the strongest evidence for tunneling and coupled motions. The deviation from RGM is an indicator for the tunneling and coupled motion contribution in the hydrogen transfer reactions.24 The RGM studies by Cleland and coworkers showed the deviations from this rule and strongly supported the tunneling

29 effects and 1o-2o coupled motions in the formate dehydrogenase (FDH) enzyme system.31 Other indication for contribution of tunneling and coupled motions in hydrogen transfer reactions is Swain-Schaad Exponents (SSEs); any deviations from these exponents is explained as that the reaction is involving the reaction in tunneling and coupling motions. Researchers used a mixed labeling experiment to study the connection among rate constants obtained from yADH in the experiments.11 They measured 2o KIEs of the reaction with H/T in secondary position and H at primary position, while the secondary D/T KIE is measured with D at primary position. They defined the mixed-labeling SSE (mSSE) as: ( ( where the

⁄ ⁄

) )

1-20

is the rate constant of the reaction, in which the i is the isotope at the pri-

mary position and j is the isotope at the secondary position. Semi-classical models, which predict

, cannot explain any deviations from this limit.62 After

finding deviations from the SSE limit and accepting the contribution of the tunneling and coupled motions, scientists tried to relate the dynamics and enzyme structure role in the tunneling. They did a series of mutations in the active site of the horse liver ADH (HLADH) and measured the mSSE for each mutant. By mutation they changed the size of the alcohol binding pocket (decreasing or increasing the DAD) their results showed that the hydrogen transfer distance (DAD) affects the degree of the tunneling. These results suggest that the enzyme fluctuations in active site may provide a proper distance for hydrogen tunneling.63 The temperature dependence of KIEs is also important tool to study the hydrogen tunneling. Thermophilic ADH (htADH) shows nearly temperature

30 independent 1o KIEs in the range of 30-65 oC, which is its physiological temperature range, and also inflated mSSE. On the other hand, it shows temperature dependent 1 o KIE below 30 oC and a mSSE value in the range of the semi-classical predictions. The htADH enzyme optimal function takes place at elevated temperature (30-65 oC) and the experimental results show significant tunneling contribution in this temperature range and also temperature independent 1o KIEs, which suggest that the enzyme provides a proper conformation for tunneling. On the other hand the temperature dependent 1 o KIE at the lower temperatures (below its physiological temperature) suggests that there is no proper conformation of the enzyme structure for tunneling at these temperatures.39 These findings have changed with the growing acceptance of the Marcus-like hydrogen tunneling model to explain all of these experimental results in a different way as explained above. This model links the enzyme motions to the catalytic power of the enzymes.25

1.9

Significance of Research

Enzymes possess the ability to accelerate the rate of the chemical reactions by factors as large as 1020 with extraordinary selectivity. Understanding how enzymes catalyze the reaction is very important in many areas such as biomedical and industrial application. Recent literature has demonstrated the importance of the quantum mechanical tunneling and heavy atom motions (proteins dynamics and environmental motions) contribution in the catalytic power of the enzymes. The most recent model which the scientists used to explain the experimental results from enzymatic catalyzed hydrogen transfer reactions is the Marcus-like model, which links the protein motion to the catalytic ability of the enzymes. It has been suggested that enzymes provide proper geometry change, such as

31 reorganization of the reaction active site, for efficient hydrogen tunneling and also shorten the DAD through physical motions and vibration. This can be tested by comparing the results of the enzymatic reaction with those in their model reactions in solution. This comparison can help us to give more insight into the physical features of the proteins, because in the model reactions in the solution the motions governed in proteins are eliminated and replace by the solvent molecules. Lu’s research group applies the same mechanistic tools which are used for enzymatic catalyzed reaction in order to understand the reaction mechanism occurring in the solution and give insight to the enzyme active site. As described before, Marcus-like hydrogen tunneling model explains the experimental results of the ADH reactions. First, the 2o KIE is affected by the crowdedness of the enzyme. 2o KIE for D-transfer is less inflated than that for H-transfer. 20 D-tunneling possesses shorter DAD than H-tunneling, leading to steric hindrance effect decreasing 2o KIEs. Second, the tunneling ready state is imbalanced with rehybridization preceding the hydride tunneling. Third, the temperature independence of 1o KIEs was observed. All these demonstrate that enzymes have evolved to reorganize the reaction coordinate for efficient hydrogen tunneling. The hypothesis of this thesis is that we study ADH enzymatic model reaction in the solvent (less restricted environment active site) to compare our results with those from enzymatic reaction (more restricted environment active site). We studied the hydride transfer reaction from an alcohol to the carbocation in acetonitrile (Figure (1-13)).

32

Figure 1-13: The reaction of hydride transfer from an alcohol to xanthylium carbocation (NAD+ model) to form relevant ketone and xanthene to mimic the rate determining step of the hydride transfer reactions in ADH.

We modeled the rate determining step of an ADH enzymatic system in solution with various alcohols and NAD+ models. The carbocation in the figure (1-13) can act as the model of the cofactor. In this reaction, the hydride transfer takes place between two carbons and is best described as a model for ADHs5. The xanthylium carbocation is an effective model to NAD+, because the two phenyl groups on both sides of this carbocation directs the hydride transfer to a position which is equivalent to that of the NAD+ cofactor as shown in Figure (1-14). Reactivity of the xanthylium carbocation as NAD+ model increases as the nitrogen atom in NAD+ is replaced with the more electronegative oxygen atom. The 1o and 2o KIEs can be measured by replacing the alcohol’s α and β hydrogen with deuterium. In the case of Xn+ by labeling the hydrogen at position 9 by deuterium, we can study the β-secondary KIE on donor anf α-2o KIE on acceptor side. The 1o KIEs and their temperature dependencies can also be studied.

33

Figure 1-14: Xanthylium carbocation structure as a model of NAD+ cofactor.

34

Chapter II

Experimental 2.1

Instrumentation

NMR spectra were obtained on an Oxford instrument with the Varian software operating at 300MHz. 1H NMR and 13C NMR signals are reported in part per million (δ) from TMS as an internal standard. UV-Vis absorption spectra were measured on an Agilent 8453 UV-Vis Spectrophotometer or 8452A UV-Vis Spectrophotometer. All reactants and sample volumes for UV-Vis analysis were delivered using Hamilton SoftGrip Adjustable Pipettes. Melting points measurements were made on a Thomas Hoover Capillary Melting Point Apparatus. Denver Instrument APX200 used to weight measurements. The rotary evaporator was a Büchi Rotavapor R-200 one.

2.2

Materials

All raw materials were purchased from Sigma-Aldrich, Acrōs Organics and Fisher Scientific and were used as received with exception of acetonitrile (CH3CN). CH3CN was distilled twice with first time over KMnO4/K2CO3 and second time over P2O5. Synthesis of 9H(D)-xanthen-9-ol: According to the literature general procedure64 with some changes, the Xanthen-9-one (2.6 g, 13.3 mmol) was dissolved in 30 mL methanol/THF (1:1) and stirred for 20 minutes. NaBH4 (2.055 g, 54.3mmol) was slowly added to the well-stirred solution of the xanthen-9-one at room temperature. After 2 hours, the mixture was poured onto 40 mL ice cold water and stirred for another 1 hour. The

35 extraction of the product was carried out using diethyl ether for three times and the organic layer dried by brine water and over magnesium sulfate anhydrous. Recrystallization was carried out from ethanol. The product was characterized by melting point and 1

HNMR. m.p. 123-125 ᵒC (literature m.p. 127-128 ᵒC), 1HNMR (300 MHz, CDCl3) δ

(ppm) 7.61-7.68 (d, 2H), 7.34-7.44 (td, 2H), 7.17-7.25 (m, 4H), 5.83-5.90 (d, 1H), 2.012.10 (d, 1H). The 9D-xanthen-9-ol was synthesized using the same procedure with NaBH4, replaced by NaBD4 and CH3CH2OH bye CH3CH2OD. The product characterized by 1HNMR spectrum of the 9D-xanthen-9-ol doesn’t show δ (ppm) 5.83-5.90 (d, 1H).

Figure 2-1: Synthesis of 9H(D)-xanthen-9-ol. Synthesis of 9H(D)-thioxanthen-9-ol: According to the literature general procedure 64 with some changing in the that Thioxanthen-9-one (2.2 g, 10.4 mmol) was dissolved in 30 mL methanol/THF (1:1) and stirred for 20 minutes. NaBH4 (1.377 g, 33 mmol) was slowly added to the well-stirred solution of the thioxanthen-9-one at room temperature. After 2 hours, the mixture was poured onto 40 mL ice cold water and stirred for another 1 hour. The extraction of the product was carried out using diethyl ether for three times and the organic layer dried by brine water and over magnesium sulfate anhydrous. Recrystallization was carried out from ethanol. The product was characterized by melting point and 1HNMR. m.p. 123-125 ᵒC (literature m.p. 127-128 ᵒC), 1HNMR (300 MHz,

36 CDCl3) δ (ppm) 7.58-7.68 (d, 2H), 7.40-7.49 (td, 2H), 7.17-7.35 (m, 4H), 5.55-5.62 (d, 1H), 2.01-2.10 (d, 1H). The 9D-thioxanthen-9-ol was synthesized using the same procedure with NaBH4, replaced by NaBD4 and CH3CH2OH bye CH3CH2OD. The product characterized by 1HNMR spectrum of the 9D-thioxanthen-9-ol doesn’t show δ (ppm) 5.55-5.62 (d, 1H).

Figure 2-2: Synthesis of 9H(D)-thioxanthen-9-ol Synthesis of 9H(D)-xanthylium perchlorate: According to the general procedure in the literature65 9H(D)-xanthen-9-ol (0.396g, 2 mmol) was dissolved in 10 mL acetic anhydride in 50 mL round bottom flask at room temperature with N2 gas bubbling for 10 minutes, then the reaction vessel cooled to 0 ᵒC in the ice bath. Perchloric acid (70%, 0.518 mL, 6 mmol) was added dropwise to the solution and stirred for additional 30 minutes. The resulting crude solid product was filtered under vacuum filter and rinsed with cold, dry diethyl ether until the filtrate became transparent. The recrystallization was carried out from acetonitrile. The product was characterized by melting point and 1

H NMR. m.p. 207-208 ᵒC (Literature m.p. 208-209 ᵒC) 1H NMR (300 MHz, CD3CN) δ

(ppm) 10.35 (s, 1H), 8.57-8.73 (m, 4H), 8.32-8.44 (dd, 2H), 8.05-8.14 (td, 2H). The melting point and 1HNMR spectrum of the 9-D-xanthelynium perchlorate are: m.p. 207208 ᵒC, 1H NMR (300 MHz, CD3CN) δ (ppm) 8.58-8.77 (m, 4H), 8.34-8.46 (dd, 2H), 8.06-8.18 (td, 2H).

37

Figure 2-3: Synthesis of 9H (D) -xanthylium perchlorate. Synthesis of 9H (D) -thioxanthylium perchlorate: according to the general procedure65 9H (D) -thioxanthen-9ol (0.395g, 1.83 mmol) was dissolved in 10 mL acetic anhydride in 50 mL round bottom flask at room temperature with N2 gas bubbling for 10 minutes, then the reaction vessel cooled to 0 ᵒC in the ice bath. Perchloric acid (70%, 2.0 ml, 23 mmol) was added dropwise to the solution and stirred for additional 30 minutes. Then about 25 mL of cold diethyl ether was added to the reaction mixture and the crystals formed. The resulting crude solid product was filtered under vacuum filter and rinsed with cold, dry diethyl ether until the filtrate became transparent. The recrystallization was carried out in acetonitrile. The product was characterized by melting point and 1H NMR. m.p. 207-208 ᵒC (Literature m.p. 208-209 ᵒC) 1H NMR (300 MHz, CD3CN) δ (ppm) 10.28 (s, 1H), 8.75-8.92 (m, 4H), 8.41-8.50 (dd, 2H), 8.20-8.28 (td, 2H). The melting point and 1HNMR spectrum of the 9D-thioxanthenylium perchlorate are: m.p. 207-208 ᵒC, 1H NMR (300 MHz, CD3CN) δ (ppm) 8.74-8.98 (m, 4H), 8.418.52 (dd, 2H), 8.20-8.30 (td, 2H).

38

Figure 2-4: Synthesis of 9H (D)-thioxanthylium perchlorate. Syntheses of αH and αD-cyclohexanol: In 100 mL round bottom flask 3 ml of methanol 2.8 mL (26.4 mmol) of cyclohexanone were dissolved and put in an ice bath. 1.0 g (26.4 mmol) of the NaBH4 was added in three batched while the solution was stirred with a spin rod. The reaction mixture stirred for 30 minutes and then allowed to warm up to the room temperature. The reaction was quenched using 15 mL of iced cold deionized water while stirred for a while. The extraction was carried out using methylene chloride three times and then the organic layer was washed with deionized water and dried on MgSO4 anhydrous. Then the drying agent filtered and methylene chloride evaporated using rotary evaporator and product was characterized by 1HNMR (300 MHz, CDCl3) δ (ppm) 3.61 (s, 1H), 1.83-1.98 (d, 2H), 1.63-1.81 (d, 2H), 1.44-1.61 (m, 3H), 1.08-1.38 (m, 3H). The αD-cyclohexanol was synthesized using the same procedure except NaBH4 which is replaced by NaBD4. The product was confirmed by 1

HNMR which is the same as α-H-cyclohexanol except the singlet peak at 3.61 ppm

which is eliminated.

39

Figure 2-5: Synthesis of αH and αD-cyclohexanol

2.3

Preparation of Reaction Solvent

Acetonitrile (HPLC grade, fisher scientific) was first refluxed over K2CO3 and KMnO4 for several hours and then refluxed over P2O5 to remove any trace of reducing agent and water in order to do kinetic experiments. Tetrahydrofuran (THF) and diethylether were purchased from Scientific Fisher and distilled over potassium and sodium, with benzophenone as an indicator, respectively to remove any trace of water.

2.4

Kinetic Study Procedure

Reaction setup: Kinetics reactions of the (T)Xn+ were done in time based measurement with a thermostated cell holder attached to the UV-Vis instrument. In order to maintain the temperature of the cell holder to that of the water bath, water lines were insulated with foam pipe and aluminum foil. The reported temperatures in the experiments are the cell holder temperature. The water bath temperature is higher than cell holder, because of the heat loss in the water lines. In 1 cm quartz cuvette cell 2 mL of the target alcohol solution in AN placed and the solution was allowed to equilibrate to the desired temperature for approximately 6 minutes. The UV-Vis set up or time based measurement and the blank spectrum of the alcohol solution was run. Then a 32 µL of the 0.0025 M (T)Xn+ solution in AN was injected into the blank siolution via 100 μL adjustable pi-

40 pette and stirred quickly in 10 seconds to make a homogeneous mixture, the cuvette was capped then the time based measurement was initiated. The kinetics for (T)Xn+ decay were monitored over two half-life times at the maximum absorption wavelength of 373 nm of the cation (λmax of Xn+ is 373 nm)(Figure 2-6). The general procedure was repeated 4 times for each H-transfer and D-transfer in different days. Figure (2-7) shows an example of the absorbance (Abs) vs. time plot for the reaction IPA with Xn +.

Absorbance (A.U.)

41 1.20 1.15 1.10 1.05 1.00 0.95 0.90 0.85 0.80 0.75 0.70 0.65 0.60 0.55 0.50 0.45 0.40 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00 280 290 300 310 320 330 340 350 360 370 380 390 400

Wavelength (nm)

Figure 2-6: Kinetic scans determined by analysis of the reaction of αD-IPA (0.012M) with Xn(H)+(3.75×10-5M) at 60oC.

42 1.1 1

Absorbance (A.U.)

0.9 0.8 0.7 0.6 0.5 0.4

0.3 0.2 0.1 0 0

100 200 300 400 500 600 700 800 900 1000 1100 1200 1300

Time (second)

Figure 2-7: Plot of absorbance vs. time of the reaction of αD-IPA (0.012M) with Xn(H)+(3.75×10-5M) at 60oC.

43 0.4 Ln (abs.)

0.2 Time (sec)

0

-0.2

0

200

400 600 800 y = -1.9559E-03x + 3.3036E-01 R² = 9.9999E-01

1000

-0.4 -0.6 -0.8 -1 -1.2 -1.4 Figure 2-8: Determination of the rate constant of the pseudo first order reaction of αD-IPA (0.012M) with Xn(H)+(3.75×10-5M) at 60oC.

44

Chapter III

Results and Discussion The 1o and 2o KIEs are the ratio of the second-order rate constants (k) for the reactions of normal alcohols or carbocations and versus those with heavy isotope (deuterated alcohols or carbocations). The pseudo first-order rate constants of the reaction were determined by following the decay of carbocation in reactions with large excess amount of alcohols. The second-order rate constants were calculated by dividing the pseudo firstorder rate constant by [alcohol]. 1o KIEs for the reaction of α-H/α-D alcohols with the carbocations are shown in Table 3-1, which were measured by Blake Albert Hammann66. The results of the 1o KIEs show that the hydride transfers are rate-determining step in the relevant reactions. Also in this table are included the 2o KIEs at the β-H6/D6 position of the 2-propanol and the 9-H/D position of the Xn+ and the TXn+. As we discussed in Chapter 1, the Hammett correlations and 2o KIEs give information about the TS structure, the former of which gives information about the electronic structure of the TS and the latter determines the geometry of the reaction center of the TS. 2o KIEs close to EIE imply late TS, while those close to unity are an indication of the early TS. For Hammett correlation, observed larger/smaller reaction constant (│ρ│) accompanied a later/earlier TS on the reaction coordinate in terms of the change in charge.

45 3.1

The Imbalanced Transition States

The β-H6/D6 2o KIE values of 2-propanol are all normal (>1) and close to unity in the reactions with all carbocations as shown in Table 3-1. The corresponding EIE for alcohol sp3 to ketone sp2 conversion is 1.53, the 2o KIE results strongly suggest an early transition state with respect to the rehybridization of the alcohol moiety. The Hammett correlation studies from the reactions of substituted benzyl alcohols with Xn+, TXn+ and PhXn+ gave rise to large negative values, suggesting late TS’s65,67. The results of the 2o KIEs (geometric structure of TS, early TS) and Hammett correlations (electronic structure of TS, late TS) are not consistent with each other, suggesting an imbalanced alcohol moiety in the transition state structure. Table 3-1: 1o and 2o KIEs for the hydride transfer reactions from aliphatic alcohols to the hydride acceptor cationsa. Alcohol

a

CL3CY(OH)CL3

CH3CY(OH)CH(CH3)2

C(CH2)5CYOH

Cations

Xn(L)+

PhXn+

(TXn)(L)+

PhTXn+

Xn(L)+

Xn(L)+

1ᵒ KIE b

2.75 (0.10)

3.26 (0.14)

2.73 (0.28)

3.47 (0.11)

2.45 (0.18)

2.36 (0.11)

2ᵒ KIE c

β-H6/D6

α-H/D

βH6/D6

βH6/D6

α-H/D

β-H6/D6

α-H/D

α-H/D

Y=H

1.05 (0.02)

0.99 (0.01)

1.05 (0.04)

1.05 (0.03)

0.97 (0.02)

1.04 (0.01)

1.03 (0.02)

1.00 (0.02)

Y=D

1.04 (0.02)

0.98 (0.03)

1.00 ((0.03)

1.04 (0.00)

0.96 (0.03)

0.98 (0.01)

0.99 (0.02)

0.98 (0.01)

2ᵒ EIE

1.52

0.89

1.52

1.52

0.89

1.52

0.89

0.89

in MeCN at 60 oC, numbers in parenthesis are standard deviations; b L = H, Y = H or D; c β-H6/D6 (on 2-propanol), α-H/D (on Xn+ or TXn+).

46 The same study of 2o KIEs at the 9-αH/D position of the carbocation in reaction with alcohol give the inverse 2o KIEs, which are consistent with sp2 to sp3 conversion of the reaction center. These values are also close to unity. The corresponding EIEs for this process (Xn+ and TXn+ to the XnH and TXnH products) are 0.89 according to the literature value of C4-H/D position in NAD+ conversion to NADH. These 2o KIEs suggest early TS on the carbocation moiety for the hydride transfer reactions. The Hammett correlation gave rise to small positive ρ values for reactions with substituted Xn+ and TXn+, suggesting that the rehybridization and the bond change are concerted on the cation side. Above analysis shows an imbalanced TS in both charge development and change in rehybridization in alcohol and in charge development in between two reactants. The C-H bond cleavage on the alcohol precedes the rehybridization on C, but the C-H bond formation and the rehybridization on the carbocation appear concerted. Moreover, the rehybridizations on alcohol and on cation seem concerted based on the observations that 2o KIEs on both are closer to unity and far from their EIEs, but the positive charge gain on alcohol precedes the positive charge loss on the cation, according to the difference in Hammett correlations (large negative Hammett constant on alcohols vs. small positive Hammett constants on cations). The imbalanced charge change on both reactants suggests a negatively charged transferring nucleus that characterizes its “hydridic” nature. Our results strongly suggest a non-classical hydride transfer mechanism.

47 3.2

The H-Tunneling Mechanism The hydride transfer reaction is an endothermic reaction because of the for-

mation of very unstable α-hydroxy carbocation. Thus according to the Hammond’s postulate, a late TS is predicted. According to this classical view, our 2o KIE of the β-H6/D6 of 2-propanol should be close to the corresponding EIE = 1.52, and α-2o KIEs on Xn+ should be closer to EIE = 0.89, as mentioned in table (3-1). But the observed 2o KIEs are just closer to unity. So the α-2o KIEs on carbocation is inflated and β-H6/D6 of 2propanol is deflated. The inflated 2o KIEs can be interpreted in terms of hydrogen tunneling effect and 1o/2o coupled motions of the hydrogen61 but the β-2o KIEs on alcohols cannot. Since the Marcus-like hydrogen tunneling model explains well on both KIEs and Hammett correlation results in ADH reactions5, 11, 25, we will use this model to explain our experimental results of the ADH model reaction. The Marcus-like hydrogen tunneling model links the protein motion to the catalytic power of the enzyme in enzymatic reactions. This model states that the tunneling takes place at tunneling ready state (TRS) where the potential energy well of the donor and acceptor are degenerate, and the enzyme provides the shorter enough donor acceptor distance (DAD) for sufficient tunneling. In this model both the energy degeneracy and short DADs sampling require energy and are mediated by heavy atom motions.5, 12, 25 Therefore, the observed 2o KIEs which are close to unity can be explained in terms of the small degree of rehybridization of H/D toward the formation of the TRS (i.e., the small degrees of rehybridization of the donor and acceptor carbons).

48 The Marcus-like model can also explain the observed indistinguishable 2o KIEs (within experimental error) on hydride transfer and deuteride transfer reactions of Xn + and TXn+ (Table (3-1), 1.05 vs. 1.04 for both systems). Within this model, the results suggest that the 2o C-H/D bonds reorganize to the same extent for both processes although H can tunnel through longer distances than D so that the latter transfer takes place in a more crowded environment. This may be explained in terms of the loose reactive complex in a not very restrictive solution environment so that the steric difference between the two reactants are so small that does not affect the reorganization of 2 o H/D. In sharp contrast to the reactions of Xn+ and TXn+, the 1o isotope effect on 2o KIEs becomes pronounced in the reactions with PhXn+ (1.05 (hydride transfer) vs. 1.00 (deuteride transfer) and PhTXn+ (1.04 vs. 0.98 (closer to 1.00 within experiment error)). This may be explained in terms of the steric effect caused by the 9-Ph group in the latter systems so that the D-transfer with shorter DAD experiences higher steric effect than Htransfer resulting in a resistance on 2o H/D reorganization thus a lower 2o KIE than that in the H-transfer reaction. Table 3-1 shows the different 1o KIEs for (T)Xn+ and Ph(T)Xn+ carbocations which can also be explained by the Marcus-like hydrogen tunneling model. The 1o KIEs for Ph(T)Xn+ is larger than that of (T)Xn+. The model defines the 1o KIE to be the ratio of the wavefunction overlap between donor and acceptor of H transfer over that of D transfer at the TRS. It predicts that the longer the DAD the larger the 1 o KIE17, 36. Since the 9-Ph group in PhXn+ and PhTXn+ creates steric hindrance effect at the reaction center as compared to H in Xn+ and TXn+, the former reactions would have longer DADs at the TRS, thus larger 1o KIEs.

49

Chapter IV

Conclusions Kinetics of the hydride transfer reactions from 2-propanol and substituted benzyl alcohols to the carbocations of Xn+, ArXn+, TXn+ and ArTXn+ were determined in MeCN. These reactions are ADH model reactions where the cations model the NAD+ and the MeCN the enzyme reaction environment. The Hammett correlations as well as the 2o KIEs on both the alcohols and the cations were determined. The data showed deviations from the values predicted from the classical TS theory, suggesting nonclassical H-tunneling effect. First, the Hammett correlations show the imbalanced development of charge in both reactants on going to the TS, with the gain of the positive charge on alcohol preceding the loss of positive charge on the cation. This indicates a negative charge borne by the “in-flight” nucleus. Second, the 2o KIEs on both the alcohols and the cations were observed to be closer to unity and far from the EIEs, suggesting early TRS’s, but the Hammett study as well as the endothermic nature of the reactions predict otherwise results showing deflated 2o KIE on alcohol and inflated 2o KIE on cations. While the inflated 2o KIEs can be explained by the traditional H-tunneling model and 1o/2o H coupled motions, the deflated 2o KIEs cannot. The Marcus-like Htunneling model that emphasizes the effect of dynamic on H-tunneling was then used to explain the 2o KIEs. Within that model, the closer to unity 2 o H/D KIEs can be explained in terms of small reorganization on both reactants toward the formation the TRS. The Marcus-like model also explains the observed insignificant difference in between the 2o KIEs on hydride and deuteride transfers to Xn+ and TXn+, and the pro-

50 nounced change on hydride and deuteride transfers to the more sterically hindered PhXn+ and PhTXn+. That is, the reactive complexes are loose in such less restrictive solution environment so that the DAD difference for H- and D-transfers to (T)Xn+ would not create much difference in steric effect that affects the degree of the reorganization of the 2o H/D, but that for H- and D-transfers to the more sterically hindered Ph(T)Xn+ would do otherwise resulting in the 1o isotope effect on 2o KIEs. In the imbalanced TRS’s of these reactions, rehybridization lags behind H-tunneling. The results are opposite from the observations in ADHs where same mechanistic tools were used for the study. Therefore, our results indicate H-tunneling effect and imbalanced TRS. More importantly, they strengthen the proposal that protein motions promote the Htunneling in ADHs by advancing the geometric change and providing favorable orbital conditions for H-tunneling.

51

REFERENCES (1) Housecroft, C. E.; Sharpe, A. G.; Inorganic Chemistry; 2nd Edition; Pearson Prentice Hall; Harlow; England; 2005; p 237. (2) Knap, M. J.; Rickert, K.; Klinman, J. P. J. Am. Chem. Soc. 2002, 124, 3865-3874. (3) Hong, B.; Maley, F.; Kohen, A. Biochemistry. 2007, 46, 14188-14197. (4) Nagel, Z. D.; Meadows, C. W.; Dong, M.; Bahnson, B. J.; Klinman, J. P. Biochemistry. 2012, 51, 4147-4156. (5) Nagel, Z. D.; Klinman, J. P. Chem. Rev. 2010, 110, PR41-PR67. (6) Nelson, D. L.; Cox, M. M. Principles of Biochemistry; 5th Edition; W. H. Freeman and Company; New York; 2008; p 189-192. (7) Wang, Z.; Abeysinghe, T.; Finer-Moore, J. S.; Striund, R. M.; Kohen, A. J. AM. Chem. Soc. 2012, 134, 17722-17730. (8) Gray, H. B.; Winkler, J. R. Q. Rev. Biophys. 2003, 36, 341-372. (9) Klinman, J. P. Procedia Chemistry. 2011, 3, 291-305. (10) Frey, P.; Hegeman, A. D. Enzymatic Reaction Mechanisms. Oxford University Press. 2007, p 30-34. (11) Cha, Y.; Murray, C. J.; Klinman, J. P. Science. 1989, 243, 1325-1330. (12) Kohen, A.; Klinman, J. P. Acc. Chem. Res. 1998, 31, 397-404. (13) Hay, S.; Scrutton, N. S. Nature Chemistry. 2012, 4, 161-168. (14) Carey, F. A.; Sundberg, R. J. Advanced Organic Chemistry, Part A: Structure and Mechanisms. 5th Edition; Spirnger; 2007, p 332-335. (15) Johnston, H. S., Gas Phase Reaction Rate Theory, Ronald Press Co., New York, 1966. p 128. (16) Westheimer, F. H. Chem. Rev. 1961, 61, 265-273. (17) Kohen, A.; Limbach, H-H. Isotope Effects in Chemistry and Biology. CRC Press, Taylor & Francis. 2006. (18) Streitwieser, A.; Jagow, R. H.; Fahey, R. C.; Suzuki, S. J. Am. Chem. Soc. 1958, 80 (9), 2326-2332. (19) Carey, F. A.; Sundberg, R. J. Advanced Organic Chemistry, Part A: Structure and Mechanisms. 5th Edition; Spirnger; 2007, p 335-344. (20) Williams, A. Free-Energy Relationships in Organic and Bio-Organic Chemistry, Royal Society of Chemistry, Cambridge, UK, 2003. (21) Hammett. L. P. J. Am. Chem. Soc. 1937, 59, 96-103. (22) Carey, F. A.; Sundberg, R. J. Advanced Organic Chemistry, Part A: Structure and Mechanisms. 5th Edition; Spirnger; 2007, p 289-290. (23) Hammond, G. S. J. Am. Chem. Soc. 1955, 77, 334-338. (24) Amin, A.; Price, R. C.; Saunders, W. H. J. Am. Chem. Soc. 1988, 110(12), p 40854086. (25) Roston, D.; Islam, Z.; Kohen, A. Molecules. 2013, 18, 5543-5567. (26) Melander, L.; Saunders, W.H. Reaction Rates of Isotopic Molecules, 4th ed.; Wiley: New York, NY, USA, 1987. (27) Swain, C. G.; Stivers, E. C.; Reuwer, J. F.; Schaad, L. J. J. Am. Chem. Soc. 1958, 80 (21), 5885-5893. (28) Nagel, Z. D.; Klinman, J. P. Nature Chemical biology. 2009, 5(8), 543-550.

52

(29) Cook, P. F.; Oppenheimer, N. J.; Cleland, W. W. Biochemistry. 1981. 20, 18171825. (30) Hermes, J. D.; Cleland, W. W. J. Am. Chem. Soc. 1984, 106, 7263-7264. (31) Hermes, J. D.; Morrical, S. W.; O’Leary, M. H. Cleland, W. W. Biochemistry, 1984, 23, 5479-5488. (32) Huskey, W. P.; Schowen, R. L. J. Am. Chem. Soc. 1983, 105, 5704. (33) Glickman, M. H.; Wiseman, J. S.; Klinman, J. P. J. Am. Chem. Soc. 1994, 116, 793-794. (34) Hwang, C-C.; Grissom, C. B. J. Am. Chem. Soc. 1994, 116, 795-796. (35) Sikorski, R. S,; Wang, L.; Markham, K. A.; Rajagopalan, P. T. R.; Benkovic, S. J.; Kohen, A. J. Am. Chem. Soc. 2004, 126, 4778-4779. (36) Stojkovic, V; Kohen, A. Israel Journal of Chemistry. 2009, 49, 163-173. (37) Bell, R. P. Trans. Faraday Soc. 1959, 55, 1-4. (38) Stern, M. J.; Weston, R. E. J. Chem. Phys. 1974, 60, 2808-2814. (39) Kohen, A.; Cannio, R.; Bartolucci, S.; Klinman, J. P. Nature. 1999, 399, 496-499. (40) Glickman, M. H.; Klinman, J. P. Biochemistry. 1995, 34, 14077-14092. (41) Nesheim, J. C.; Lipscomb, J. D. Biochemistry. 1996, 35, 10240-10247. (42) Hothi, P.; Hay, S.; Roujeinikova, A.; Sutcliffe, M. J.; Lee, M.; Leys, D.; Cullis, P. M.; Scrutton, N. S. ChemBioChem. 2008, 9, 2839-2845. (43) Basran, J.; Sutcliffe, M. J.; Scrutton, N. S. Biochemistry. 1999, 38, 3218-3222. (44) Harris, R. J.; Meskys, R.; Sutcliffe, M. J.; Scrutton, N. S. Biochemistry. 2000, 39, 1189-1198. (45) Fan, F.; Gadda, G. Biochemistry. 2007, 46, 6402-6408. (46) Heyes, D. J.; Sakuma, M.; de Visser, S. P.; Scrutton, N. S. J. Biol. Chem. 2009, 284, 3762-3767. (47) Bell, R. P. The Tunneling effect in Chemistry. Chapman and Hall, New York, 1980. (48) http://www.nobelprize.org/nobel_prizes/chemistry/laureates/index.html. July 15, 2013. (49) Marcus, R. A.; Sutin, N. Biochimica et Biophysica Acta. 1985, 811, 265-322. (50) Sutcliffe, M. J.; Scrutton, N. S. Eur. J. Biochem. 2002, 269, 3096-3102. (51) Francisco, W. A.; Knapp, M. J.; Blackburn, N. J.; Klinman, J. P. J. Am. Chem. Soc. 2002, 124, 8194-8195. (52) Caratzoulas, S.; Schwartz, S.C. J. Chem. Phys. 2001, 114, 2910–2918. (53) Marcus, R. A. J. Phys. Chem. B. 2007, 111, 6643-6654 (54) Kuznetsov, A. M.; Ulstrupt, J. Can. J. Chem. 1999, 77, 1085–1096. (55) Meyer, M. P.; Tomchick, D. R.; Klinman, J. P. Proc. Natl. Acad. Sci. USA. 2008, 105, 1146-1151. (56) Klinman, J. P. Biochemistry. 2013, 52, 2068-2077. (57) Klinman, J. P. J. Biol. Chem. 1972, 247, 7977-7987. (58) Klinman, J. P. Biochemistry. 1976, 12, 2018-2026. (59) Welsh, K. M.; Creighton, D.J.; Klinman, J.P. Biochemistry. 1980, 19, 2005-2016. (60) Cook, P. F.; Blanchard; J. S.; Cleland, W. W. Biochemistry. 1980, 19, 4853-4858. (61) Huskey, W.P.; Schowen, R.L. J. Am.Chem. Soc. 1983, 105, 5704–5706. (62) Kohen, A.; Jensen, J. H. J. Am.Chem. Soc. 2002, 124, 3858-3864.

53

(63) Bahnson, B. J.; Park, D-H.; Kim, K.; Plapp, P. B.; Klinman, J. P. Biochemistry. 1993, 32, 5503-5507. (64) Zhu, X-Q.; Dai, Z.; Yu, A.; Wu, S.; Cheng, J-P. J. Phys. Chem. B. 2008, 112, 11694–11707. (65) Dauben, H. J.; Honnen, L. R.; Harmon, K. M. J. Org. Chem. 1960, 25, 1442. (66) Hammann, B. A. A Comparison of Imbalanced H-Tunneling Ready States in Alcohol Dehydrogenases and Model Reactions. SIUE, Edwardsville, August 2012. (67) White, O. R. A Study of the Mechanisms of Hydride Transfer Reactions from alcohols and 1,4-dihydropyridines to NAD+ and NADH Models. MS Thesis. SIUE. Edwardsville. May 2013.