Syracuse University Syracuse University

SURFACE SURFACE

Syracuse University Honors Program Capstone Projects

Syracuse University Honors Program Capstone Projects

Spring 5-2017

Synthesis of substituted tetrahydropyrans using metal-catalyzed Synthesis of substituted tetrahydropyrans using metal-catalyzed

coupling reaction coupling reaction

Amanda Lieu Syracuse University

Follow this and additional works at: https://surface.syr.edu/honors_capstone

Part of the Chemistry Commons

Recommended Citation Recommended Citation Lieu, Amanda, "Synthesis of substituted tetrahydropyrans using metal-catalyzed coupling reaction" (2017). Syracuse University Honors Program Capstone Projects. 1040. https://surface.syr.edu/honors_capstone/1040

This Honors Capstone Project is brought to you for free and open access by the Syracuse University Honors Program Capstone Projects at SURFACE. It has been accepted for inclusion in Syracuse University Honors Program Capstone Projects by an authorized administrator of SURFACE. For more information, please contact surface@syr.edu.

i

Synthesis of Substituted Tetrahydropyrans Using Metal-Catalyzed Coupling Reactions

A Capstone Project Submitted in Partial Fulfillment of the Requirements of the Renée Crown University Honors Program at

Syracuse University

Amanda Lieu

Candidate for Bachelor of Science and Renée Crown University Honors

Spring 2017

Honors Capstone Project in Chemistry

Capstone Project Advisor: _______________________ Nancy Totah, Professor Capstone Project Reader: _______________________ Michael Sponsler, Professor Honors Director: _______________________

Chris Johnson, Interim Director

ii

© (Amanda Lieu May 2017)

iii

Abstract

The purpose of this research is to explore the use of cobalt- and iron-catalyzed cross coupling reactions of alkyl iodides containing beta-tetrahydropyran groups to develop a general method to synthesize substituted tetrahydropyrans. Substituted tetrahydropyrans can be used towards the total syntheses of biologically active natural products. The starting iodide compounds underwent sp3-sp3 and sp3-sp2 cobalt- and iron-catalyzed cross-coupling reactions with different Grignard reagents. It was expected that the formation of the coupled product would be faster than the formation of the side products (elimination and homocoupling) and thus yield the desired substituted tetrahydropyrans. The results of this study showed that cobalt-catalyzed reactions are best used for sp3-sp3 coupling though in low yields while the iron-catalyzed reactions produced better yields for the sp3-sp2 couplings. However, the reactions still resulted in the formation of undesired elimination and homocoupling side products. In the future, optimization will be made to minimize the formation of the side products and to increase yields of the substituted tetrahydropyrans which can then provide a further basis for the use of cobalt- and iron-catalyzed cross-coupling reactions in the synthesis of natural products.

iv

Executive Summary

Many biologically active natural products have a significant impact in the drug discovery

field and are therefore a consistent source of new medicines. Often these natural products are

complex molecules that are difficult to isolate from the biological sources. To overcome this

challenge, these natural products are built from simple molecules by the means of total synthesis.

Total synthesis can not only help obtain natural products in better yields and in a short amount of

time, it can also expand the library of chemical structures and help discover non-natural derivatives

that may contain important biological activities.

The skeletons of tetrahydropyrans are often found in biologically active natural products,

making the synthesis of complex tetrahydropyrans crucial. To form new carbon-carbon bonds by

metal-catalyzed coupling reactions is a common method to synthesize complex tetrahydropyrans.

Metal-catalyzed cross-coupling reactions are processes frequently used to join an organic halide

and an organometallic compound and involve the coupling of the carbon centers of those

compounds with the aid of a metal catalyst. We are interested in the studies of cobalt- and iron-

catalyzed reactions because of the availability and low toxicity of cobalt and iron catalysts.

The focus of this research is to develop a general method to synthesize tetrahydropyran

derivatives from alkyl halides containing beta-tetrahydropyran groups and Grignard reagents by

the use of sp3-sp3 and sp3-sp2 metal-catalyzed cross coupling reactions. The problems that have

been encountered with this chemistry are the competition between the formation of the coupled

products and the formation of the side products such as elimination and homocoupling. The

development of a general method is hoped to help overcome these difficulties and allow for the

expansion of this chemistry.

v

First the starting tetrahydropyrans that were used in the metal-catalyzed coupling reactions

were synthesized (Figure 1). After the synthesis of the simple starting materials, the synthesis of

substituted tetrahydropyrans could be carried out with Grignard reagents. The Grignard reagents

that were used are also shown below alongside the starting tetrahydropyrans. (Figure 1).

Figure 1

The cobalt- and iron-catalyzed cross-coupling methods were conducted with the same

general reaction conditions. With the cobalt-catalyzed reactions, the starting tetrahydropyrans

were treated with 0.1 equivalents of catalytic CoCl2 and 1.5 equivalents of Grignard reagent

(Scheme 1). With the iron-catalyzed reactions, the tetrahydropyrans were treated with 0.05

equivalents of Fe(acac)3, 0.1 equivalents of TMEDA, and 0.05 equivalents of HMTA alongside

with 1.5 equivalents of the Grignard reagents (Scheme 2).

Scheme 1

Scheme 2

O I O IO1 2

TMS MgClMeO MgBr MgCl

Cl

MgBr MgCl3 4 5

6 7

O I + RMgXCoCl2 (5 mol%)

THF, 0°C - rt O R1

O I + RMgXFe(acac)3 (5 mol%)/TMEDA/HMTA (1:2:1)

THF, 0°C - rt O R1

vi

The resulting mixtures from these cobalt- and iron-catalyzed coupling reactions often

contained the leftover unreacted iodide 1. Difficulties were encountered separating the coupled

products from the starting iodide 1 by column chromatography due to the similar polarities led to

the treatment of the crude mixtures with activated zinc dust and ammonium chloride (Scheme 3).

This will lead to the conversion of the iodide compound 1 to the elimination product 8 which is

easier to separate from the coupled product, as compound 8 is a more polar alcohol.

Scheme 3

Overall, the project resulted in the formation of substituted tetrahydropyrans despite the

presence of homocoupling side products and the cobalt- and iron-catalyzed cross-coupling

methods were evaluated. The cobalt-catalyzed method was better for sp3-sp3 coupling while the

iron-catalyzed method gave better yields for sp3-sp2 coupling and fewer products. The results of

this project can provide a further basis for the use of cobalt- and iron-catalyzed cross-coupling

reactions towards the synthesis of complex tetrahydropyrans.

O I +Zn, NH4Cl

EtOH, reflux, 1-1.5 h O RO R+

HO1 8

vii

Table of Contents

Abstract……………………………………….……………….………….. iii Executive Summary………………………….……………….………….. iv List of Abbreviations………………………….……………….……..….. ix Acknowledgements …………..………………………………………...… xi Chapter 1: Introduction………………………………………………… 1

I. Importance of Natural Products…………….………………… 1

II. Biologically Active Natural Products Containing Tetrahydropyran Units…………….……………………………. 2 III. Metal Catalyzed Coupling Reactions…………………………… 5

a) The Kumada Coupling……………………………...……7

b) Cobalt- and Iron-Catalyzed Coupling Reactions……..….10

IV. Metal-Catalyzed Coupling Reactions in Systems Containing b-Alkoxy Groups ………………………………….…………….15

a) Metal-Catalyzed Coupling Reactions with Acyclic Compounds Containing Alkoxy Groups …………………16

b) Metal-Catalyzed Coupling Reactions with Cyclic Compounds Containing Alkoxy Groups………………….19

V. Project Objectives………………………………………………… 23

Chapter 2: Results and Discussions……………………………………….. 26

I. Synthesis of Tetrahydropyran Starting Materials…………………..26

II. Cobalt-Catalyzed Coupling Reactions……………….…………….30 a) Coupling Reactions of (Trimethylsilyl)methylmagensium Chloride……………….30 b) Coupling Reactions of Aryl Grignard Reagents …………...34 c) Attempted Coupling Reaction of Benzyl Grignard ………...36 d) Attempted Coupling of an Iodo Lactone …………………...37 III. Iron-Catalyzed Coupling Reactions……………….………….........38

a) Coupling Reactions of (Trimethylsilyl)methylmagensium Chloride………….……..38 b) Coupling Reactions of Aryl Grignard Reagents …….………39

viii

c) Attempted Coupling Reaction of Benzyl Grignard …….. 40 d) Attempted Coupling of an Iodo Lactone ……………..… 41 IV. Conclusions and Future Work…………………………...…..… 41

Chapter 3: Experimental………………………………………………… 44 References....…………………………………………….…………..…… 51 Appendix...…………………………………………………….…….….… 56

ix

List of Abbreviations Ac acetyl acac acetylacetonate t-Bu tert-butyl °C degrees Celsius

d doublet (spectral)

dr diastereomer ratio

ED50 dose effective in 50% of test subjects

equiv equivalent

g gram(s)

h hour(s)

HMTA hexamethylmethylenetetramine

Hz hertz

L liter(s)

LD50 dose that is lethal in 50% of test subjects

m multiplet (spectral); milli

Me methyl

MIC minimum inhibitory concentration

min minute(s)

mol mole(s)

NHC N-heterocyclic carbene

NMR nuclear magnetic resonance

Ph phenyl

x

Rf retention factor (in chromatography)

rt room temperature

s singlet (spectral); second(s)

t triplet (spectral)

TBDMSO tert-butyldimethylsilyloxy

TBS tert-butyldimethylsilyl

temp temperature

Tf trifluoromethanesulfonyl (triflyl)

THF tetrahydrofuran

TLC thin-layer chromatography

TMEDA N,N,N’,N’-tetramethyl-1,2-ethylenediamine

TMS trimetylsilyl

UV ultraviolet

vol volume

xi

Acknowledgments I would like to thank Professor Nancy Totah for her aid and allowing me the opportunity to conduct this project under her supervision. Thank you to Mike Robinson for providing guidance, and Brianna Shores, Clark Coffman, Sarah Medhi, Ivanna Pohorilets, and Wenhong Lin for their in-lab support; Dr. Michael Sponsler and Dr. John Chisholm for being my readers. I would lastly like to thank Karen Hall, and Hanna Richardson as well as all the people in the Honors Program for the Crown Wise funding and for all of their help.

1

Chapter 1

Introduction

I. Importance of Natural Products

Natural products are complex molecules that come from biological sources such as marine

creatures, fungi, bacterial, and plants. Natural products have a significant impact on the science

community, especially in the field of drug discovery which involves finding chemicals with

biological activities that can be used to target certain diseases such as cancer and bacterial

infections. Drug discovery relies on small molecule chemical scaffolds to produce medicines and

natural products contribute 83% of new therapeutics. Therefore, natural products have been a

successful source of new medicines.i Studies on natural products have provided many useful

molecules and analogs that can be used in the pharmaceutical industry. However, the isolation of

natural products is often difficult because it is a time consuming and complicated process. The

natural products are often isolated in small quantities which makes it challenging to characterize

them and to use them in biological testing which is needed confirm any biological activity. This is

a common source of problems for isolation of natural products from plants and marine organisms.ii

The total synthesis of complex molecules can help overcome the challenges of isolation of

natural products. Often, the starting point of the total synthesis is a simple and chemically stable

molecule and which is much easier to work with than attempting isolation from natural sources.

The role of total synthesis can improve on the isolation of natural products by scaling up and by

allowing small-scale productions of the targeted natural product in a short amount of time. These

benefits can help overcome difficulties in obtaining the natural product in better yields and it

speeds up the drug development process.iii The total synthesis of complex molecules can not only

2

expand the collection of natural products but it also gives chemists a chance to discover non-natural

derivatives that may contain improved biological activities. This would then further expand the

collection of chemical structures and benefit the drug discovery field.iv

Oxygen heterocycles like the tetrahydropyran skeleton 9 are found as substructures of

many natural products containing important biological activities (Figure 2).

Figure 2

Therefore, the development of creative synthetic approaches to tetrahydropyrans have been

the subject of significant attention in many research groups. By developing more synthetic

methods to access tetrahydropyrans, the chemistry of these tetrahydropyran-containing natural

products can be further understood. This is important because then the library of natural

products/chemical structures can be further expanded more. It is also important to continue

developing new methods to synthesize tetrahydropyrans in order to discover new efficient ways to

make tetrahydropyrans. The more synthetic methods of tetrahydropyrans available, the more range

of techniques scientists can choose from therefore increasing their chances to eventually synthesize

tetrahydropyran-containing natural products.

II. Biologically Active Natural Products Containing Tetrahydropyran Units

Since many natural products with interesting biological activities contain tetrahydropyran

units, it is given that tetrahydropyrans are an area of interest in the total synthesis of natural

products. Some examples of biologically active natural products that contain tetrahydropyran units

are apicularen A (2), pyranicin (3) and ambruticin (4) (Figure 3).

O9

3

Figure 3

Apicularen A 2 was first isolated from a variety of strains of the myxobacterial genus

Chrondromyces by Kunzev and its structure was determined by X-ray crystallography by Jansen.vi

It is known to have high cytostatic activity against ovarian, prostate, lung, kidney, cervical,

leukemia, and histiocytic cancer cells with LD50 values between 0.23-6.79 nM.v It has been

prepared before by five total syntheses, with the first total synthesis done by de Brabander.vii The

most recent total synthesis of apicularen A was done by Uenishi; an efficient Cu(I)-catalyzed

coupling process was used to complete the synthesis.viii

Pyranicin 3 was isolated by McLaughlin from the stem bark of the Goniothalamus

giganteus tree in 1997.ix This natural product has demonstrated anticancer activity, specifically

selective in vitro cytotoxicity against human pancreatic cancer cells with ED50 value of 10-2

μg/mLix and in vivo cytotoxicity against the growth of cancerous leukemia cells with LD50

value of 9.4 μM .x A total of five total syntheses of pyranicin have been done with the first

O

O

H HHO

OHOHO

pyranicin (11)(anticancer activity)

OH

ambruticin (12)(antifungal activity)

apicularen A (10)(anticancer activity)

O

OHOH

CO2H O

OO

O

OH

OH

NH

O

4

total synthesis was accomplished by Nakata and Takahashixi while the most recent total

synthesis was done by Crimmins and Jacobs.xii In the thirteen-step total synthesis of pyranicin

performed by Phillips, an alkyl-alkyl Pd-catalyzed Suzuki coupling process was used in the

final steps of the synthesis.xiii

Ambruticin 4 was isolated from the myxobacterium Polyangium cellulosum by Ringel in

1977.xiv It is an antifungal agent that has pronounced activity against pathogens such as

Coccidioides immitis, Histoplasma capsulatum, and Blastomyces dermatitidisxiv and it has

exhibited potent inhibitory activity against the yeast strain Hansenula anomala with a MIC value

of 0.03 μg/mL.xv The first total synthesis of ambruticin was accomplished by Kendexvi, while

the total synthesis performed by Jacobsen involved the use of a palladium-catalyzed Kumada

coupling process to form a bond for the natural product.xvii

The tetrahydropyran is a core structure in many natural products.xviii It is important to

synthesize tetrahydropyrans because natural products containing tetrahydropyran show a diverse

range of biological activities that have potential to be used in medicine such as anticancer and

antifungal activities as shown in the examples above. The synthesis of tetrahydropyrans can also

provide organic chemists more challenges to develop more efficient methods which can also lead

to synthesis of new derivatives of natural products that also contain important biological activities.

While there are efficient ways to synthesize natural products containing tetrahydropyrans like the

ones in Figure 2, there needs to be more methods involving the formation of carbon-carbon bond

on tetrahydropyrans. In order to form new bonds, new chemistry must keep being developed which

must be tested first on simple molecules such as tetrahydropyrans. This is what keeps

tetrahydropyrans an area of interest in organic synthesis.

5

III. Metal-Catalyzed Coupling Reactions

Of fundamental task in organic synthesis is the ability to form carbon-carbon bonds. A

carbon-carbon bond can be used to connect two simple molecules together to form a new

compound that is more complex. There are many known methods of carbon-carbon bond

formation between different types of carbon centers such as the sp3-sp3, sp3-sp2, and sp2-sp2 carbon

atoms and there are still more methods of carbon-carbon bond formation being developed.xix

Metal-catalyzed coupling reactions are useful for the preparation of new carbon-carbon bonds

between sp3, sp2, and sp hybridized carbon atoms.

Metal-catalyzed coupling reactions are processes in which two carbon centers are brought

together with the aid of a metal catalyst. These reactions are frequently performed using an organic

halide and an organometallic compound, resulting in the formation of a new carbon-carbon bond.

The general mechanism of metal-catalyzed coupling reactions consists of three phases (Figure 4).

The first step involves the oxidative addition of the organic halide to the metal catalyst. Then the

organometallic compound will undergo transmetalation; once this occurs both participating

compounds will be on the same metal catalyst and result in the formation of an dialkylmetal

intermediate. The final phase is the reductive elimination of alkylmetal intermediate which will

regenerate the metal catalyst and give the coupled product.xx

6

Figure 4

The most used organic halides with metal-catalyzed coupling reactions are aryl and vinyl

halides. Aryl and vinyl halides are commonly used because of their general availability and

applicability, their selectivity during reactions (for example, they can undergo oxidative addition

rather than beta elimination), and their high tolerability of different functional groups.xxi These

benefits have made the use of aryl and vinyl halides more developed than other electrophiles in

metal-catalyzed coupling reactions.

The use of alkyl halides (R-X), especially sp3 R-X, in these processes has proven to be

more difficult. Unlike aryl and vinyl electrophiles, alkyl halides have a greater tendency to undergo

side reactions like beta-elimination rather than the desired coupling process for many reasons – for

example, the C(sp3)-H bond in alkyl halides are more electron rich than the C(sp2)-H bond in

aryl/vinyl halides.xxi Additionally this causes sp3 alkyl halides go through oxidative addition much

more slowly compared to aryl and vinyl halides and even compared to sp3 alkyl and vinyl halides.

L2M0

L2MIIR

X

R X

R' YX Y

MIIL2

R

R'

R R'

Oxidative Addition

Transmetalation

Reduction Elimination

7

This problem can lead to the formation of homocoupling or beta-elimination side products.xxii

Another problem with alkyl halides is that the alkylmetal intermediate tends to undergo

intramolecular beta-elimination.xxiii However, there has been progress in developing metal-

catalyzed coupling reaction conditions to overcome these problems with sp3 alkyl halides.

a) The Kumada Coupling

Throughout the studies of metal-catalyzed coupling reactions, many different coupling

processes with different organometallic compounds have been developed. One of the well-known

partners with organic halides in cross-coupling reactions are Grignard reagents. Grignard reagents

offer many advantages to their use in metal-catalyzed coupling reactions, one being that Grignard

reagents are good nucleophiles. Grignard reagents also can undergo cross-coupling without

suffering from beta-elimination; this is seen even with alkyl (sp3) Grignard reagents with beta-

hydrogens. The coupling process between organic halides and Grignard reagent came to be known

as the Kumada coupling (Scheme 4). This coupling process is typically uses palladium and nickel

catalysts and can be used to form new carbon-carbon bonds between alkyl, aryl, and vinyl groups.

Scheme 4

The Kumada coupling has been used to produce many interesting chemical structures. In

one specific example, Danishefsky studied the synthesis of 17- and 19-membered ring homologues

through a ring-closing metathesis strategy. The metathesis precursor was synthesized through the

Kumada coupling from vinyl iodide 13 (Scheme 5). The vinyl iodide 13 was converted to 1,5 diene

R + R' MgX

Metal Catalyst (typically nickel or palladium)

solvent/ligandRX R'

8

15 in the presence of catalytic PdCl2(dppf) and 3 equivalents of butenylmagnesium bromide 14.xxiv

Scheme 5

The 1,5 diene product 15 was obtained in a 75% yield. From this result, it was likely that

this Kumada coupling reaction could be used to synthesize more alternative unconjugated dienes

that can eventually lead to the synthesis of large ring analogues. The reaction was also able to

proceed even with the presence of other heteroatoms such as nitrogen and sulfur.

The selectivity of the Kumada coupling can be highlighted in this nickel-catalyzed

coupling reaction with a phenol 16 that contains both a bromo and a fluoro group. Manabe and

Wang reported that the coupling process between the phenol 16 and 3 equivalents of the Grignard

reagent 17 (Scheme 6) in the presence of the nickel catalyst, Ni(dppbz), occurred at the o-fluoro

position, not the p-bromo position to give the coupled product 18. This is unusual because it is

expected that the bromo group is more reactive than the fluoro group. During this process, the

transition state is proposed to have the magnesium coordinated to the phenol group which will then

facilitate the oxidative addition of the Ni metal center to the adjacent C-X bond.xxv

Scheme 6

N

SOTBS

I

+MgBr

1413

PdCl2(dppf) (23 mol%)Et2O, 12h, rt

75%

N

SOTBS

15

(3 eq)

OH

Br

F+ n-C12H25MgBr

(3 eq)

OH

Br

n-C12H25NiCl2(dppbz)5 mol%

Toluene, rt65%

16 17 18

9

Manabe and Wang noted that para- and double-cross-coupled products were obtained in

trace amounts. They also addressed that the directing effect of the hydroxyl group has a stronger

effect on the order of coupling than the reactivity order of the halogen groups. This research

demonstrated that the Kumada coupling can have a high selectivity during reactions.

Palladium and nickel are the most widely employed catalysts for metal-catalyzed coupling

reactions.xxvi While palladium-catalyzed coupling reactions have a wide substrate scope, the use

of Grignard reagents is limited due to the incompatibilities some functional groups that are highly

electrophilic such as esters, nitriles, benzyl, and nitro substituents. Since Grignard reagents are

commercially available and relatively reactive compared to other reagents such as organoboron or

organolithium compounds, it is ideal to use Grignard reagents in carbon-carbon bond formation.

The use of bulky ligands is also often required in palladium-catalyzed coupling reactions. The main

disadvantage that comes with using these ligands are that they are sterically hindering, which can

hamper the results of the coupling process.xxvii Nickel, despite being another popular metal catalyst

due to its high reactivity (it even has a higher reactivity than palladium), has less of a general scope

than palladium catalysts. Nickel catalysts also have low selectivity because of their tendency to

undergo side reactions, and they have low tolerability of functional groups. However, nickel

catalysts are much less expensive than palladium catalysts (the cost of NiCl2 is $0.10 per mmol

while the cost of PdCl2 is $5.80 per mmol).xxviii Another disadvantage that comes with palladium

and nickel catalysts is their high toxicity.xxix Because of these disadvantages with palladium and

nickel catalysts, interest in other metal catalysts began to develop.

10

b) Cobalt- and Iron-Catalyzed Kumada Coupling Reactions

The studies on the use of cobalt- and iron-catalyzed coupling reactions have been recently

developing due to the advantages they have over palladium and nickel. Compared to palladium

and nickel catalysts, iron catalysts are inexpensive and have a lower toxicity. Iron is an abundant

metal, making it easier and cheaper to obtain. Iron is also heavily integrated in many biological

processes including many safe disposal pathways, and from this it is observed that iron has a low

toxicity.xxx Iron catalytic systems have been proven to be superior to palladium and nickel catalyst

systems in terms of reaction efficiency because iron-catalyzed coupling reactions give better yields

of product and their reaction time is shorter compared to palladium- and nickel-catalyzed coupling

reactions.xxxi Cobalt catalysts, like iron catalysts, are also inexpensive. Cobalt catalysts provide

other advantages compared to palladium and nickel catalysts, one being that it can work under

mild reaction conditions and it is highly chemoselective. Furthermore, cobalt catalysts typically

only require the use of simple ligands rather than bulky ligands which are often used in palladium-

catalyzed coupling reactions.xxix It has also been reported that some cobalt catalytic systems do not

suffer from beta-hydride elimination side products with alkyl halides, unlike palladium-catalyzed

reactions.31,xxxii Due to these advantages, cobalt and iron catalysts can offer new means to develop

metal-catalyzed coupling reactions economically, ecologically, and efficiently.

The advantages of iron- and cobalt-catalyzed coupling reactions have attracted the attention

of researchers to expand iron- and cobalt-catalytic reaction scope. For example, Cahiez

investigated the iron-catalyzed coupling process between primary and secondary alkyl halides and

alkenylmagensium bromides (Scheme 7). In one specific example from the study below, Cahiez

successfully produced the coupled product 21 in a 69% yield from an alkenyl Grignard reagent 19

11

and a secondary alkyl halide 20 in the presence of 5 mol% Fe(acac)3/TMEDA/HMTA catalytic

system. The stereochemistry from the alkenylmagnesium bromide 19 was also retained as the

coupled product 21 was isolated as a Z/E mixture of 97:3, indicating that the reaction is highly

stereoselective.xxxiii

Scheme 7

Cahiez reported furthermore in the study that this is the first iron-catalyzed coupling

process between alkenyl Grignard reagents and secondary alkyl halides, which expands the scope

of the iron-catalyzed coupling reactions even further. The yields of these coupled product (Figure

5) range from 15-84%. The coupling process also performs well with primary alkyl halides as

shown with the coupled products 25 and 26. It has also shown during this study the iron-catalyzed

coupling reaction works with alkyl halide containing functional groups. This further shows the

tolerability of the functional groups in substrates that can be used alongside the iron-catalysts as

seen with the coupled products 27 and 28.

Figure 5

+

5 % Fe(acac)3/TMEDA/HMTA(1:2:1)

0°C, THF, 45 min69%19 20 21

MgBr

Br

O

O

CN

2384%

2773%

2867%

2255%

2475%

2680%

2567%

12

Oshima has performed series of cobalt-catalyzed coupling reactions of primary and

secondary alkyl halides with 1-trimethylsilylethenylmagnesium bromide 30 and 2-

trimethylsilylethynylmagnesium bromide 31 (Scheme 8). Under the cobalt-mediated cross-

coupling reaction conditions of catalytic Co(acac)3 and 3-4 equivalents of the Grignard reagents,

the alkenylation and alkynlation of the cyclic bromide 29 proceeded smoothly, giving the desired

coupled products 32 and 33 in high yields of 85% and 71% respectively for the alkenylation and

alkynylation process.xxxiv

Scheme 8

Overall, Oshima was successful in the formation of new carbon-carbon bond through

alkenylation and alkynylation of the thirteen different alkyl halides (Figure 6). The yields ranged

from 26 to 90% and the reactions proceeded under mild reaction conditions. The alkenylation and

alkynylation can still proceed with primary and secondary linear alkyl halides to give the coupled

products 34-36. The cobalt-catalyzed coupling reaction also gives coupled products 37-40

containing oxygen and nitrogen, and this shows that the cobalt-catalyzed coupling reaction

conditions can tolerate the presence of other heteroatoms. This study is also interesting because it

shows that the formation of carbon-carbon bond between primary/second alkyl halides and alkenyl

Grignard reagents can proceed smoothly with cobalt catalysts without experiencing the beta

elimination side product.

Co(acac)3

TMEDA, 25°C, 15 min+

BrMg

SiMe3

29

3032

85%Br

orBrMg SiMe3

31

SiMe3

SiMe3

3371%

or

13

Figure 6

Fürstner reported the use of iron-catalyzed coupling reaction towards the synthesis of

latrunculin B. In the presence of catalytic Fe(acac)3 at -30˚C, the coupling of the ester 41 and the

Grignard reagent 42 proceeded smoothly affording the alkyne 43 in a great yield of 97% (Scheme

9). The reaction proceeded stereoselectively and very rapidly under mild conditions.xxxv

Scheme 9

An interesting note on the reaction is that there was no addition of the Grignard reagent to

the ester group. This is important because it shows the iron-catalytic systems have the possibility

to not go through the elimination process. Fürstner also reported that this is the first case of an

iron-catalyzed coupling reaction with an enol triflate, thus further expanding the scope of the use

of iron catalysts.

Casar reported a study of cobalt-catalyzed cross-coupling reactions between aryl Grignard

reagents and allylic and vinylic bromides in the presence of the natural ligand sarcosine. This

C8H17

SiMe3

SiMe3

O

O

SiMe3 O

O

SiMe3

C8H17

SiMe3

TsN

SiMe3

TsN

SiMe3

3490%

3664%

3574%

4067%

3982%

3854%

3780%

O

OEt

TfO

+MgBr Fe(acac)3 (10 mol%)

THF, -30°C97%

O

OEt41 42 43

14

method was used to prepare a key intermediate that can be used towards the synthesis of a drug

product. The ester 44 was treated with catalytic CoBr2 and sarcosine and 1.4 equivalents of the

Grignard reagent 45 that was freshly prepared beforehand to give the coupled product 46 in a 97%

isolated yield (Scheme 10).

Scheme 10

The reaction demonstrated high alpha-regioselectivity and good functional group tolerance

despite the presence of the ester in the vinyl halide 44 and the fluorine groups in the aryl Grignard

reagent 45. The coupled product 46 was later used to synthesize the antidiabetic drug sitagliptin.

IV. Metal-Catalyzed Coupling Reactions in Systems Containing b-Alkoxy Groups

An area of study with metal-catalyzed coupling reactions that has yet to be investigated is

the identification of methods in cases where the organic halide contains an oxygen substituent that

is in the beta position to the metal center. It is important to continue researching efficient methods

for these cases so researchers can perform metal-catalyzed coupling reactions without suffering

from the elimination side reaction of the oxygen. The organic halide containing the beta oxygen

substituent can undergo elimination when the carbanion forms from the metal complex instead

which will then form an alkene in place of the coupled product (Figure 7).

Br O

O +F

F F

MgBr CoBr2/sarcosineTHF, -20°C

97%

O

OF

F F44 45 46

HN

O

OHsarcosine

15

Figure 7

The formation of the elimination product can be a major competing process alongside with

the cross-coupling process. This can prove to be an obstacle during the studies of carbon-carbon

bond formation via metal-catalyzed coupling reactions with compounds containing oxygen

substituent in the beta position. If more methods can be developed to overcome the elimination

process, the scope of reactions with b-alkoxy groups containing compounds can be expanded.

a) Metal-Catalyzed Coupling Reactions with Acyclic Compounds Containing b-

Alkoxy Groups

Cahiez reported a cobalt-catalyzed coupling process between aromatic Grignard reagents

and functionalized inactivated alkyl bromides. In the presence of catalytic Co(acac)3 and 1.1

equivalents of the aryl Grignard reagent 48, the bromo acetate 47 was smoothly converted to the

coupled product 49 in a great yield of 88% (Scheme 11). The elimination side product, ethylene,

was not observed during this reaction.xxxvi

X O R

Metal catalyst (M)R'-Y

M O R

R' O R

O R

Cross-coupling reaction Elimination reaction

+ HO R

16

Scheme 11

The reaction was shown to tolerate the presence of the acetyl group which indicates that

this cobalt-catalyzed coupling process is highly chemoselective. The reaction was also able to

proceed under mild conditions of 0˚C in only 40 minutes with the catalytic system being obtained

by mixing two inexpensive substances (cobalt acetylacetonate and TMEDA).

Cossy reported the first iron-catalyzed coupling process between sp3 alkyl halides and

alkenyl Grignard reagents. In one specific example within the study, the bromo acetal 50

successfully coupled with the alkenyl Grignard reagent (2 equivalents) 51 to give the alkene 52 in

a moderate yield of 50% (Scheme 12). This reaction occurred under the presence of the inexpensive

and nontoxic FeCl3 as a catalyst (5 mol%) and 1.9 equivalents of TMEDA in THF at 0˚C, and was

eventually warmed to room temperature. Despite the moderate yield of the coupled product, no

elimination product was observed.xxxvii

Scheme 12

Kang reported the synthesis of substituted tetrahydrofurans by the iron catalyzed coupling

of functionalized alkyl iodides with aryl Grignard reagents (Scheme 13). In the presence of

catalytic FeCl2 and 2 equivalents of phenylmagnesium bromide 54, iodide 53 gave the

Br OAc +MeO

MgBr Co(acac)3 (5 mol%), TMEDA (5 mol%)THF, 0°C, 40 min

88% MeO

OAc

47 48 49

Br O

O+

MgBrFeCl3 (5 mol%), TMEDA (5 mol%)

THF, 0°C-rt50% O

O

50 51 52

17

tetrahydrofuran 56 in 11% yield. In this sequence, initial formation of an aryl iron complex is

followed by intramolecular cyclization of iodide 53 to give an intermediate tetrahydrofuran

derivative 55. Reductive elimination then gives the final product 56. The low yield in this reaction

can be attributed to beta elimination of oxygen in the starting alkyl iodide 53 to give allyl alcohol

as the major decomposition product.xxxviii

Scheme 13

Better yields were obtained in substrates that contained acetal and dimethyl substituents.

In these modified systems (Figure 8) yields ranged from 46-64%. Little stereoselectivity was

observed.

OI

+ BrMgFeCl2 (5 mol%)

THF, 0°C, 2h11% O53 54 56

O

FeIIIPhBr

55

18

Figure 8

Oshima reported the cobalt-catalyzed coupling reactions between iodoacetals and allyl

Grignard reagents that resulted in the synthesis of substituted tetrahydrofurans. In one interesting

example in the study, the treatment of the iodoacetal 62 with the allyl Grignard reagent 63 in the

presence of CoCl2(dppp) afforded the gamma lactone 68 in a yield of 71% (Scheme 14). The

sequence of this reaction starts off with the intramolecular cyclization of the iodoacetal 62,

resulting in a radical intermediate 64. Then, the radical intermediate 64 goes through a ring-

opening process to give another radical intermediate 65 which will then go through oxidative

addition to the cobalt-allyl complex 66 via single electron transfer. After reductive elimination

from the cobalt complex 66 and Jones oxidation, the gamma lactone 68 will have been

synthesized.xxxix

OEtO57

46% (1.3:1)

OEtO58

58% (1.2:1)

OEtO59

58% (1.6:1)

OMe

OEtO OEtO

S

6064% (1.4:1)

6146% (1.9:1)

19

Scheme 14

The beta-oxygen elimination may have been prevented by the presence of the allyl ligand.

This allowed the formation of the carbon-carbon bond on the iodoacetal 62. This method can be

further extended to alkyl halides containing β-alkoxy groups.

b) Metal-Catalyzed Coupling Reactions with Cyclic Compounds Containing b-

Alkoxy Groups

Oshima studied the synthesis of alkenylated and alkynylated cyclic compounds through the

cobalt-mediated sequential tandem cyclization and alkenylation/alkynlation between cyclic

iodoacetals and 1-trimethylsilylethenylmagnesium bromide 30 and 2-

trimethylsilylethynlmagnesium bromide 31. In the presence of catalytic Co(acac)3 and 3

equivalents of the Grignard reagents, the alkenylated/alkynlated lactones 70 and 71 were

successfully synthesized from the cross-coupling of the iodoacetal 69 in yields of 65% and 30%

respectively (Scheme 15). In this sequence, an alkyl radical is produced from the iodoacetal 69

followed by radical cyclization. Next, the alkyl cyclic radical is then captured by the cobalt

complex. Reductive elimination and Jones oxidation will then give the alkenylated and alkynylated

lactones 70 and 71.xxxiv

I

OC4H9O

C4H9O O C4H9O O+ CoCH2CH=CH2

C4H9O O

O O

62

64 65 66 67

68

+ BrMg

63

1. CoCl2(dppp)THF, -40°C

2. CrO3/acetone71%

CoCH2CH=CH2

20

Scheme 15

The alkenylated and alkynlated lactones 70 and 71 are known precursors of prostaglandins

and other related compounds. Other substituted cyclic compounds were synthesized using the same

cobalt-catalyzed coupling reaction condition in yields of 44-78% (Figure 9). Some

diastereoselectivity was observed in these reactions.

Figure 9

Cárdenas developed a novel iron-NHC catalytic system for the cross coupling of primary

alkyl halides and alkyl Grignard reagents. NHC ligands have been used in the past to interfere with

Co(acac)3

TMEDA, 25°C, 15 min+

BrMg

SiMe3

TBDMSO

O I

OBu

TBDMSO

O

SiMe3

O

69

3070

65%or

BrMg SiMe3

31or

TBDMSO

O

O

7130%

SiMe3

7260%

OO

SiMe37344%

OO

SiMe3

OO

SiMe37478%

OO

7566%

SiMe3

21

the formation of beta elimination product in alkyl Pd(II) intermediates. Under the presence of

catalytic Fe(OAc)2 (2.5 mol%) – NHC ligand system and 3.7 equivalents of the alkyl magnesium

bromide 70, the carbon-carbon bond formation occurred on the tetrahydropyran 1 to give the final

product 72 in a 63% yield (Scheme 16). The reaction proceeded in either of two different

mechanistic pathways, one being that the tetrahydropyran 1 undergoes oxidative addition by the

single electron transfer and then undergoes transmetalation with the Grignard reagent 76 to form

the intermediate iron(III) complex 77. The second mechanistic pathway involves the reverse order,

instead the transmetalation of the Grignard reagent 76 and then the oxidative addition of the

tetrahydropyran 1 to give the intermediate Fe(III) complex 77. Both mechanistic pathways will

end with the reductive elimination of the intermediate Fe(III) complex 77 to give the final product

78.xl

Scheme 16

This method widened the scope of the iron-catalyzed coupling reactions and it provided a

new method to use on alkyl halides containing beta oxygen substituents.

OI

+O O

BrMg Fe(OAc)2 (2.5 mol%)IMes•HCl (6 mol%)

THF, rt63%

O OO

1 76 78

LnFe(III)

I

O

OO

77

22

Cossy looked at the synthesis of substituted piperidines by the cobalt-catalyzed coupling

reaction between iodopiperidines and Grignard reagents. During this study, two iodopiperidines

containing beta alkoxy groups 79 and 81 were coupled to the aryl Grignard reagent 54 under the

reaction conditions of catalytic CoCl2 (5 mol%) and the ligand (R,R)-tetramethylcyclo-

hexanediamine (TMCD) to produce two different coupled products 80 and 82. When the

iodopiperidine 79 was treated with the aryl Grignard reagent 54 and the catalytic system of CoCl2

and TMCD, it will go through a 5-exo-trig cyclization process before proceeding to the cross-

coupling process to give the bicyclic product 80 in a yield of 81% (Scheme 17). But when the

iodopiperidine 81, which only contains an extra carbon atom on the ether chain, underwent the

cobalt-catalyzed coupling reaction it only went through the cross-coupling process giving the

substituted piperidine 82 in a yield of 71% (Scheme 18). This occurred because the cross-coupling

process was faster than the 6-exo trig cyclization step.xli

Scheme 17

Scheme 18

No beta-elimination side product was observed in both reactions due to the presence of

the TMCD ligand. Diastereoselectivity was also observed.

Fürstner reported the study of iron-catalyzed coupling reactions with alkyl halides and

sterically hindered aryl Grignard reagents. Under the presence of catalytic iron complex 84, the

NBoc

I

O+

BrMg CoCl2 (5 mol%), TMCD (6 mol%)THF, 0°C to rt, 3 h

81% NBoc

O

HPh

79 54 80

NBoc

I

O+

BrMg CoCl2 (5 mol%), TMCD (6 mol%)THF, 0°C to rt, 3 h

71% NBoc

O

Ph

81 54 82

23

formation of the carbon-carbon bond between the tetrahydropyran 1 and the Grignard reagent 83

was successful giving the final coupled product 85 in the yield of 78% (Scheme 19). The use of

the iron complex 84 allows for the cross-coupling of the alkyl halide and sterically hindered

Grignard reagent due to the coordination environment around the Fe(II) metal center.xlii

Scheme 16

This reaction was also shown to be applicable to other alkyl halides containing beta oxygen

substituents (Figure 10). The coupled products from these alkyl halides containing beta oxygen

substituents were synthesized with yields ranging from 62-89%.

Figure 10

O I +BrMg

84 (5 mol%)THF, 70°C

78% O

1 83 85

FeEt2PPEt2

Cl

Cl

Et2P

PEt2

84

O

OOO

O

8680%

OMeO

TBSO OTBS

8789%

O

MeO

8877%

O

O

O

MeO

8962%

24

V. Project Objectives

Previous work from the Totah lab has shown the utility of iron-catalyzed coupling reactions

between a Grignard reagent and alkyl iodides for the preparation of carbon-carbon bonds in

complex molecules. The interest was to explore the use iron-catalyzed cross-coupling reaction to

incorporate the linear side chain of spirastrellolide A in a single step. In the presence of the catalytic

Fe(acac)2, TMEDA, and HMTA, the alkyl iodide 90 smoothly cross coupled with isopropenyl

magnesium bromide 91 to give the coupled product 92 (Scheme 20).xliii

Scheme 20

Despite the concerns about beta oxygen substituent, the coupled product was synthesized

in a moderate yield of 52%. This led to the utilization of the same iron-catalyzed cross-coupling

reaction condition in the final step of the spirastrellolide side chain. First, the vinyl iodide 93 was

converted to a Grignard reagent 94. Then, the newly made Grignard reagent 94 underwent an iron-

catalyzed coupling reaction with the alkyl iodide 90 in the presence of catalytic Fe(acac)2,

TMEDA, and HMTA (Scheme 21). Through this method, spiroketal 95/96 that contains the linear

side chain of spirastrellolide A was successfully synthesized in 57% yield in a diastereomeric

mixture of 1:1.xliii

O

TBSO

O Pr +MgBr

Fe(acac)2 (5 mol%)

THF, TMEDA, HMTA52%

O

TBSO

O PrI

90 91 92

25

Scheme 21

The use of the iron-catalyzed coupling reaction to contribute towards the synthesis of

natural product has therefore been demonstrated in the preparation of the linear side chain of

spirastrellolide A. The formation of the coupled product was achieved despite the presence of

sensitive functional groups and substituents such as the skipped diene and beta oxygen substituent.

The formation of the coupled product also occurred in a single transformation and yielding the

desired coupled products in moderate yields, making this iron-catalyzed coupling reaction efficient

to use. Thus, the results of this study have sparked an interest of developing a general method to

prepare tetrahydropyran derivatives through the metal-catalyzed cross coupling reactions between

alkyl iodides containing beta-tetrahydropyran groups and Grignard reagents.

OO

I1. tBuLi, -78°C; MgBr22. 90, Fe(acac)3:TMEDA:HMTA (1:2:1)

THF, 0°C - rt, 16h57% (dr = 1:1)

OO

O

TBSO

O Pr

+

OO

OO Pr

TBSO

OO

MgBr +O

TBSO

O PrI

90

93

94

95

96

26

Chapter 2

Results and Discussion

I. Synthesis of Tetrahydropyran Starting Materials

Interest in the synthesis of tetrahydropyran-containing natural products has led us to

consider the development of a general method to prepare tetrahydropyran derivatives by the metal-

catalyzed cross-coupling process between alkyl iodides with beta-tetrahydropyran groups and

Grignard reagents. The model starting substrates that we are interested in using for the

development of the general method are the alkyl iodides 1 and 2 (Figure 11).

Figure 11

The alkyl iodide 1 is a basic tetrahydropyran with the oxygen in the beta position from the

iodide. The second alkyl iodide 2 also contains a lactone group. The general method that is

developed will be a Kumada coupling between these alkyl iodides and a Grignard reagent (Scheme

22).

Scheme 22

We have decided to focus on the use of iron and cobalt catalysts during this study for a few

reasons, as these metal catalysts have a low toxicity, are inexpensive and easy to obtain compared

O I O IO1 2

O I1

or

O IO2

+ RMgXCo or Fe catalyst

THF, 0°C - rt O Ror

O RO

27

to other catalysts. The iron catalyst has been used successfully before in previous works in the

Totah lab.xliii The cobalt catalyst would operate under more simple reaction conditions which could

work with alkyl iodides with beta-oxygen substituents. Prior work in the Totah lab with cobalt

catalysts suggest that the cobalt-catalyzed reactions would allow for the formation of cross-

coupled products even in the presence of a lactone.xliv

The goal of this project is to expand the scope of metal-catalyzed coupling reactions

between alkyl iodides containing beta-tetrahydropyran groups and Grignard reagents. By further

exploring the capabilities of this method, it will eventually lead the way to discover an efficient

method to synthesize tetrahydropyran subunits into more complex structures. But in order to start

this project, the starting tetrahydropyrans 1 and 2 must be synthesized which are already known

compounds.

Two methods were evaluated for the synthesis of 2-iodomethyltetrahydropyran 1. First,

direct conversion of alcohol 97 to the corresponding iodide 1 was explored (Table 1). Treatment

of 0.30 g of hydroxymethyltetrahydropyran 97 with I2, pyridine, and triphenylphosphine in

refluxing benzene gave the desired iodide 1 in only 22% yield (entry 1). Though iodide 1 had not

been prepared in this way before, the synthesis of related primary iodides with this method was

known to give good yields.xlv For this reason, the reaction was tried again, this time on larger scale

and for a longer reaction time (entry 2). Under these conditions a much better yield of 87% was

obtained. The higher yield of this reaction may be the results of running the reaction on a larger

scale compared to the first experiment and due to letting the reaction run longer than the previous

reaction (4 hours compared to 2.5 hours). Seeing the high yield of the second experiment, the

reaction was tried again on an even larger scale to provide more of the iodide 1 to work with (entry

3). While the desired iodide 1 was successfully synthesized, there were problems with the

28

separation between the iodide 1 and the side product triphenylphosphine oxide. After attempting

another separation of the two products by flash column chromatography which did not result in

the isolation of the iodide 1, the two products were dissolved in hexanes and were left at -20˚C.

By leaving the reaction mixture in hexanes, it was hoped that the triphenylphosphine oxide would

precipitate due its poor solubility in hexanes. Afterwards, the mixture would be filtered through

Celite to separate the triphenylphosphine oxide from the iodide 1. However, even this separation

method did not work due to the large amount of triphenylphosphine oxide that was produced from

the reaction. Since the separation of the iodide 1 and the by-product triphenylphosphine oxide was

unsuccessful, we were unable to use the iodide 1 in further reactions. In order to be able to prepare

the iodide 1 in a large scale, another preparation was found and used.

Table 1

Entry Starting Material Time Yield

1 0.30 g 2.5 h 22%

2 1.00 g 4.0 h 87%

3 2.00 g 4.0 h N/A

Due to these problems, an alternative “green” protocol was investigated for the synthesis

of iodide 1.xlvi This reaction, developed by the Iranpoor lab, involves the iodoetherification of 5-

hexene-1-ol 8 with sodium iodide and oxone dissolved in deionized water (Table 2). According to

the Iranpoor procedure, an immediate reaction was expected to occur. However, when alcohol 8

was subjected to the conditions reported, little reaction was observed, even when the mixture was

stirred overnight (entry 1). Isolated material consisted mostly of the starting alcohol 8. We

O I

1

O OH

97

PPh3, pyridine, I2benzene; reflux

29

considered that the quality of the oxone could be the reason for this result. In the second experiment

(entry 2), the reaction was run on a larger scale, and the number of equivalents of oxone used was

doubled (1.0 equivalents). Even with these modified conditions, little product was observed. TLC

and crude 1H NMR still indicated that the reaction mixture contained mostly starting material, even

after 144 hours of stirring. After purification via column chromatography, the iodide 1 was

obtained in 27% yield. Due to the low yields obtained in these first two reactions, a new batch of

oxone was purchased and the reaction was attempted again (entry 3). In the third experiment, the

reaction was run on the same scale as in the second experiment and 1.0 equivalents of oxone was

used. This time however, the reaction was only stirred overnight. While the TLC and crude 1H

NMR indicated there was still starting material presence, in this experiment iodide 1 was produced

in greater yield, 58%.

Table 2

Entry Starting Material Equivalents Oxone Time Yield

1 0.10 g 0.5 24 h NR

2 0.50 g 1.0 144 h 27 %

3 0.50 g 1.0 24 h 58 %

After using both sets of reaction conditions, we were able use both preparation methods to

produce the iodide 1. While the first reaction condition gave the iodide 1 in a higher yield (87%)

than the second reaction condition (58%), there is problem with the purification when it comes to

the separation of the synthesized iodide 1 and the by-product triphenylphosphine oxide. This

makes the “green” reaction conditions more preferable to use to prepare the iodide 1 since there is

O INaI, oxone

Water8 1

HO

30

little problem with the separation between the iodide 1 and the starting material 8. Another

additional advantage that the “green” methods offers is reduced waste by the use of deionized

water as the solvent. The starting alcohol 8 used in the “green” method is also cheaper to obtain

than the starting alcohol 97 used in the first method.

The other iodide compound 2 was prepared by another researcher in the Totah lab. The

carboxylic acid 98 was treated with iodine and sodium bicarbonate in acetonitrile to give the iodo

lactone 2 in a yield of 42% (Scheme 23).

Scheme 23

II. Cobalt-Catalyzed Coupling Reactions

a) Coupling Reactions of (Trimethylsilyl)methylmagnesium Chloride

With preparation of the iodide 1, the feasibility of the cobalt catalyzed coupling reaction

with the Grignard reagent 3 was explored (Table 3).

Table 3

Entry Equivalents of CoCl2 Concentration Time Ratio of 99:1:8a

1 0.05 1.0 M 3 h 1:1:1

2 0.10 1.0 M 2 h 2:1:1

3 0.20 1.0 M 1 h 4:1:2

4 0.10 0.33 M 1 h 2:1:1 a ratio determined by 1H NMR of the crude reaction mixture.

OH

O I2, NaHCO3

acetonitrile42% O I

2O

98

O I THF, 0°C - rt O TMS1 99

+ TMS MgCl 5% CoCl2

3

+

8HOO I

1

+

31

In the initial experiment (entry 1), the reaction conditions reported by the Oshima lab were

used to effect an sp3-sp3 cobalt catalyzed coupling.xlvii In this reaction, the iodide 1 was combined

with 0.05 equivalents of anhydrous CoCl2 catalyst, followed by the addition of

(trimethylsilyl)methylmagnesium chloride at 0°C. Evaluation of the crude reaction mixture by 1H

NMR showed the presence of the desired coupling product 99, as well as the alcohol 8, and

unreacted starting material 1. The ratio of 99, 1, and 8 was 1:1:1, although it is difficult to

determine exactly much of 8 was formed since it is volatile (boiling point range = 78-80°C) and

some could have been lost when the reaction mixture was concentrated. Formation of alcohol 8

likely occurred due to transmetallation between the Grignard reagent 3 and the intermediate

organocobalt species formed from iodide 1 (Scheme 24) From here, elimination of the beta alkoxy

oxygen results in formation of the carbon-carbon double bond and ring opening of the

tetrahydropyran to give compound 8.

Scheme 24

In an effort to increase the rate of the cobalt-catalyzed coupling reaction over the

elimination the amount of CoCl2 catalyst was doubled (0.10 equivalents) as shown in entry 2. This

modification showed some success. While the crude reaction mixture still consisted of iodide 1,

the coupling product 99, and the elimination product 8, the ratio of the ratio of 99, 1, and 8

increased to 2:1:1 with the coupling product 99 predominating. Like in the first experiment, there

O I O Co

CoCl2+ TMS MgCl

O Co

TMS

O O HO

O TMS1

399

8

32

was some uncertainty in the amount of 8 in the crude mixture since some could have been lost

during the concentration.

Due to the result above, in the third experiment the amount of the cobalt catalyst was

increased to 0.20 equivalents to see if the ratio of coupling and elimination products could be

increased further. The crude 1H NMR still indicated the presence of starting material 1 and

elimination products 8 in addition to the desired product 99. However, the ratio between 99, 1, and

8 was increased to 4:1:2. Since there was less starting material, this implied that the reaction rate

increased when the equivalents of CoCl2 increased. But, the ratio of the coupled product 99 and

the elimination product 8 remained the same. This result suggested that that increasing the

equivalents of the cobalt catalyst more would most likely only increase the reaction rate rather than

lead to further optimization of the product ratio.

In a fourth experiment, the amount of cobalt catalyst was reduced back to 0.10 equivalents,

but the concentration of the reaction mixture was changed from 1.0 M to 0.33 M. This change was

based on a successful procedure from the Oshima lab where a more dilute reaction concentration

was used.xlviii Unfortunately, this change had little effect on the outcome of the reaction. 1H NMR

of the crude reaction mixture indicated that iodide 1 was still present and the ratio of 99, 1, and 8

was still approximately 2:1:1.

In a final alteration of the reaction conditions tetramethylethylenediamine (TMEDA) was

added to the mixture (Scheme 25). Though the exact role of TMEDA in coupling reactions of this

type has not yet been determined,xlix it was anticipated that coordination of this basic ligand to the

cobalt catalyst could help to stabilize reactive intermediates and prevent the formation of the

elimination product 8. In this experiment, after combining the iodide 1 with 0.10 equivalents of

CoCl2, 0.10 equivalents of TMEDA was added to the reaction mixture. Afterwards, the

33

(trimethylsilyl)methylmagnesium chloride added at 0°C. However, in this experiment only the

starting iodide 1 was recovered. Neither the coupling product 99 or the elimination product 8 were

observed in the crude 1H NMR. It is unknown why this reaction did not proceed. One possibility

may be that the TMEDA used was not dry.

Scheme 25

Attempts to purify the product of these reactions by column chromatography was

complicated by the fact that the coupling product 99 could not be visualized by thin layer

chromatography (TLC). The product was not UV active, nor did it not appear with any of the stains

tried (p-anisaldehyde, cerium ammonium sulfate, KMNO4, iodine). Further, while the elimination

product 8 could be readily separated by chromatography, the silyl product 99 and starting iodide

1 did not separate well.

To aid in the isolation of the coupled product 99, the crude reaction mixture (obtained

from experiments in Table 1) was treated with activated zinc dust in the presence of solid

ammonium chloride and refluxed for 1 hour (Scheme 26). Under these conditions, the iodide 1

should undergo elimination, giving product 8 and leaving the coupling product 99 behind. This

process was expected to make the isolation of the coupled product 99 easier due to the difference

in polarity between the elimination product 8 and the coupled product 99.

Scheme 26

O I THF, 0°C - rt, 1 h+ TMS MgCl

5% CoCl2 (0.1eq), TMEDA

O TMS

199

Not observed3

O I EtOH, reflux, 1hO TMS

1 99

+ Zn, NH4Cl+

8

HOO TMS

99

34

In practice, the iodide 1 was successfully converted to the elimination product 8, The crude

1H NMR spectrum indicated the presence of only the elimination product 8 and the coupled

product 99. However, due to the lack of methods to visualize the coupled product 99 by TLC we

were unable to find the final product upon attempted purification so a pure sample of this product

was not obtained for characterization.

Though conditions for the cobalt catalyzed coupling had not been fully optimized, we

decided to test the capabilities of this reaction using different types of Grignard reagents, focusing

on those that would provide UV-active products to ensure that the coupling product could be

visualized by TLC. In these reactions, the conditions identified in Table 1, entry 2 were used.

b) Coupling Reactions of Aryl Grignard Reagents

The cobalt catalyzed coupling reaction of iodide 1 with several aryl Grignard reagents was

next explored. In the first experiment, 3-methoxyphenylmagnesium bromide 4 was used. In the

presence of 0.10 equivalents of CoCl2, the iodide 1 was treated with 1.5 equivalents of this

Grignard reagent 4 at 0 °C and the resulting mixture stirred for 1 hour (Scheme 27). As expected,

the course of the reaction could be followed by TLC.

Scheme 27

Here too, the crude 1H NMR showed the presence of the starting iodide 1 and the

elimination side product 8, in addition to the desired sp3-sp2 coupling product 100. Upon

purification by flash column chromatography, the coupled product 100 was successfully separated

from these side products, though was isolated in a low yield of 10%.

O I THF, 0°C - rt, 1 h10% O

+5% CoCl2 (0.1eq)MgBr

OCH3

OCH3

1 4 100

35

The next aryl Grignard reagent utilized was o-tolylmagnesium chloride 5. Attempted

coupling reaction with the iodide 1 under the conditions described above gave similar results. Both

the starting iodide 1 and the elimination product 8 were present in the crude reaction mixture

(Scheme 28).

Scheme 28

In order to facilitate purification, the crude reaction mixture was treated with activated zinc

dust and ammonium chloride to convert the leftover iodide 1 into the elimination product 8. As

seen in the reaction of (trimethylsilyl)methylmagnesium chloride (Scheme 26), this process

allowed for the complete removal of iodide 1 from the reaction mixture. Separation of the tosyl

derivative 101 from the elimination product 8 then proceeded smoothly, especially since the

coupled product could be visualized by UV and upon staining the TLC plate with KMnO4.

However, upon purification by flash column chromatography it was determined that the coupled

product 101 was contaminated by the homocoupling side product, 2,2’-dimethylbiphenyl, 102

(Figure 12). Metal catalyzed homocoupling reactions of Grignard reagents are known.l The 1H

NMR indicated that the ratio between the coupled product 101 and the homocoupling product 102

is 1:1, giving an estimated yield of the coupled product of 21%.

Figure 12

To test the tolerance of this cobalt-catalyzed coupling reaction with different functional

groups, the Grignard reagent 4-chlorophenylmagensium bromide 6 was used. In the presence of

O IO

+MgCl

1

1. CoCl2 (5 mol%)THF, 0°C - rt, 2.5 h

2. Zn, NH4ClEtOH, reflux, 1.5 h

~21%5 101

102

36

CoCl2, it was expected that the iodide 1 would undergo oxidative addition more rapidly than the

aryl chloride to give the coupled product 103 (Scheme 29).

Scheme 29

Unfortunately, this reaction did not yield the desired coupled product 103. Instead, the

crude mixture solidified indicating that polymerization had occurred. The polymerization most

likely resulted from the presence of the chloride functional group.

c) Attempted Coupling Reaction of Benzyl Grignard

The sp3-sp3 cobalt catalyzed coupling was attempted a second time using

benzylmagnesium chloride 7. This Grignard reagent was combined with iodide 1 in the presence

of catalytic CoCl2 under the reaction conditions described previously (Scheme 30).

Scheme 30

Despite previous problems with sp3-sp3 cross coupling, this Grignard reagent was expected

to be much easier to work with since desired coupled product 104 would be visible under UV light.

1H NMR of the crude reaction was difficult to interpret, but TLC suggested the presence of product.

In an effort to isolate some of the coupled product 104 the crude mixture was purified via flash

column chromatography. Unfortunately, only the homocoupled product biphenyl 105 was

obtained in yield of 47% (Figure 13). None of the desired cross-coupled product 104 was isolated.

O I+

MgBr

THF, 0°C - rt, 3 h O

1

Cl

ClCoCl2 (5 mol%)

6103

not observed

O I

1

+MgCl

7

THF, 0°C - rt, 1 hCoCl2 (5 mol%) O

104not observed

37

Figure 13

d) Attempted Coupling of an Iodo Lactone

Previous work in the Totah lab suggested that in the presence of a cobalt catalyst, the

coupling reactions between (trimethylsilyl)methylmagnesium chloride 3 and vinyl iodides

proceeded in high yield, even in the presence of other reactive functionality like esters (Scheme

31).xliv

Scheme 31

Based on this result, the cobalt-catalyzed coupling reaction was applied to the iodomethyl

lactone 2 (Scheme 32). In this case, 3-methoxyphenylmagnesium bromide 4 was chosen as the

coupling partner as this was the only aryl Grignard reagent that was used successfully without

formation of homocoupling product. The same reaction conditions as in the previous reactions

examples were used, e.g. 0.10 equivalents of catalytic CoCl2, 1.5 equivalents of Grignard 4, and a

concentration of 1.0 M.

Scheme 32

10547%

I+ TMS MgCl

5 mol% CoCl2THF

106 3 107

O

O

O

O

TMS

OOI

+MgBr

OCH3THF, 0°C - rt, 1 h O

5% CoCl2

OCH3

O

2 4 108not observed

38

Unfortunately, only formation of the delta-unsaturated carboxylic acid 98 (Figure 14) was

observed. As with the alcohol 8 this product is formed by elimination of an intermediate

organometallic species. None of the desired coupled product 108 was observed.

Figure 14

III. Iron-Catalyzed Coupling Reactions

a) Coupling Reactions of (Trimethylsilyl)methylmagnesium Chloride

Due to past successes with iron catalysis in the Totah lab (cf. Schemes 20 and 21),xliii the

feasibility of the iron-catalyzed coupling reactions of alkyl iodides containing beta-

tetrahydropyrans was also explored. As with the cobalt catalyst, the coupling reaction between the

iodide 1 and (trimethylsilyl)methylmagnesium chloride 3 was investigated first. For these

reactions, a procedure reported by the Cahiez group was utilized (Scheme 33).xxxiii Thus, iodide 1

was combined with Fe(acac)3 catalyst, TMEDA, and HMTA (hexamethylmethylenetetramine), then

1.5 equivalents of the (trimethylsilyl)methylmagnesium bromide 3 was added.

Scheme 33

After 3 hours, evaluation of the reaction mixture by 1H NMR showed that none of the

desired coupling product 99 had formed. Only the starting iodide 1 and the side product resulting

from elimination and ring opening of the tetrahydropyran (8) were present. This result is perhaps

not surprising because sp3-sp3 coupling has been proven to be difficult in metal catalyzed coupling

HO

O

98

O I + TMS MgClO TMS

199

not observed3

Fe(acac)3 (5 mol%)/TMEDA/HMTA (1:2:1)

THF, 0°C - rt, 3 h

39

reactions. The Csp3 bond to a halide is more electron rich than a Csp2 bond with a halide, so the

reductive elimination steps are much slower in the coupling of sp3 hybridized halidesxxi making

beta elimination of the oxygen to give 8 a highly competitive process.

b) Coupling Reactions of Aryl Grignard Reagents

Next, the use of the iron catalyst was tested in the cross-coupling reaction of aryl Grignard

reagents and the iodide 1. This compound was treated with 3-methoxyphenylmagnesium bromide

4 in the presence of Fe(acac)3/TMEDA/HMTA under the conditions described above, except that

this reaction ran overnight (Scheme 34).

Scheme 34

Unlike in the iron catalyzed sp3-sp3 coupling (Scheme 33), here the crude 1H NMR

indicated the presence of coupled product 100. Some of the starting iodide 1 and the elimination

product 8 were also present. Upon purification by flash column chromatography, the coupled

product 100 was obtained in a yield of 31%, still modest, but an improvement over the much lower

yield obtained in the cobalt-catalyzed process with this same Grignard reagent 4.

The iron-catalyzed coupling reaction was also attempted between the iodide 1 and o-

tolylmagnesium chloride 5 (Scheme 35). As seen in the previous coupling reactions, 1H NMR

indicated that the crude mixture contained the desired coupled product 101 as well as recovered

starting material 1. However, unlike with the cobalt catalyzed coupling, there was no sign of the

elimination product 8. This result suggested that the iron-catalyzed coupling reaction could be used

to minimize the formation of the elimination product 8.

O IO

+

MgBr

OCH3

OCH3

1 4 100

Fe(acac)3 (5 mol%)/TMEDA/HMTA (1:2:1)THF, 0°C - rt, 24 h

31%

+HO

8

40

Scheme 35

Unfortunately, purification by flash column chromatography did not result in separation of

the coupled product 101 from the iodide 1. This is due to the fact that the two compounds have

very similar polarities and thus very similar Rf values making make it very difficult to isolate the

coupled product. Based on the 1H NMR of the purified material, the iron-catalyzed coupling

reaction with the Grignard reagent 5 provided a yield of the coupled product 101 with less

formation of the homocoupling product 102 and no presence of the elimination 8. Presumably

separation of the iodide 1 from the desired product 101 could have been achieved if the crude

mixture had been heated with Zn dust and ammonium chloride. However, the crude mixture was

not treated with Zn dust and ammonium chloride due to the presence of the homocoupling product

102. Since it is difficult to separate the coupled product 101 from the homocoupling product 102,

we did not attempt to purify the crude mixture by column chromatography.

Finally, the iron-catalyzed coupling reaction of the p-chloro Grignard reagent 6 was

explored in order to determine if the iron based coupling would proceed without polymerization

(Scheme 36). In this case, the coupled product 103 was not observed. Only the starting iodide 1

and the elimination product 8 were present by 1H NMR of the crude reaction mixture. Unlike the

crude mixture in the cobalt-catalyzed coupling reaction (Scheme 29), the resulting crude mixture

however did not solidify which means that there was no polymerization.

Scheme 36

O IO

+MgCl

1 5 101

Fe(acac)3 (5 mol%)/TMEDA/HMTA (1:2:1)THF, 0°C - rt, 2.5 h

HO

8observed

O I +MgBr

O

1

Cl

Cl

6103

not observed

Fe(acac)3 (5 mol%)/TMEDA/HMTA (1:2:1)THF, 0°C - rt, 3 h

41

c) Attempted Coupling Reaction of Benzyl Grignard

The sp3-sp3 iron-catalyzed coupling reaction of iodide 1 and benzylmagnesium chloride 7

was examined (Scheme 37). Under the standard reaction conditions, none of the coupled product

104 was observed. Only the starting iodide 1 and the elimination product 8 were present by 1H

NMR of the crude reaction mixture. In keeping with the outcome of the earlier iron-catalyzed sp3-

sp3 coupling shown in Scheme 29, this result was not unexpected since oxidative addition and

reductive elimination are slow. Unlike in the cobalt-catalyzed process with Grignard reagent 7,

there was no sign of the homocoupled product 105.

Scheme 37

d) Attempted Coupling of an Iodo Lactone

The iron-catalyzed coupling reaction was also conducted using the iodomethyl lactone 2

and 3-methoxyphenylmagnesium bromide 4 (Scheme 38). In the presence of 5%

Fe(acac)3/TMEDA/HMTA (1:2:1), the iodolactone 2 was treated with 1.5 equivalents of the aryl

Grignard reagent 4. Evaluation of the crude reaction mixture showed the carboxylic acid 98 as the

only product. As in the cobalt-catalyzed coupling reaction this result indicates that elimination of

the beta-alkoxy functionality is faster than formation of the cross-coupled product was slower than

the elimination process.

Scheme 38

O I + O

1

MgCl

7104

not observed

Fe(acac)3 (5 mol%)/TMEDA/HMTA (1:2:1)THF, 0°C - rt, 3 h

OO I+

MgBr

OCH3

O

OCH3

O HO

O

98 observed2 4

108not observed

Fe(acac)3 (5 mol%)/TMEDA/HMTA (1:2:1)THF, 0°C - rt, 3 h

42

IV. Conclusions and Future Work

The purpose of this research was to explore the metal-catalyzed coupling reactions of 6-

iodomethyltetrahydropyrans 1 and 2 in the interest of developing a general method for the

synthesis of substituted tetrahydropyrans. Both the cobalt- and iron-catalyzed coupling reactions

of these substrates with different Grignard reagents were evaluated in sp3-sp3 and sp3-sp2 carbon-

carbon bond forming processes.

In the cobalt catalyzed reactions, both sp3-sp3 and sp3-sp2 coupling reactions were

observed. The sp3-sp3 coupling reactions produced either the coupled products or the

homocoupling products. The sp3-sp2 coupling reactions in general produced the coupled products

except in the presence of other functional groups such as the chloro group that caused

polymerization.

In the iron catalyzed reactions, only sp3-sp2 coupling reactions was also observed. The

coupled products or even the homocoupling products were not observed in the sp3-sp3 coupling

reactions. The sp3-sp2 coupling reactions successfully produced the coupled products in better

yields than the cobalt-catalyzed reactions; however the coupling product was still not observed

when the Grignard reagent contained another functional group.

Comparing the results of the cobalt-catalyzed and iron-catalyzed coupling reactions, the

sp3-sp3 coupling reactions worked best with the cobalt catalysts. The sp3-sp2 coupling reactions

produced better yields of the coupled product with the iron catalysts than with cobalt catalysts.

However, both metal-catalyzed coupling reactions did not work in the presence of some functional

groups like the chloro group. The formation of the cross-coupled products still competed with the

formation of the elimination and homocoupling side products in the presence of both metal

catalysts.

43

In the future, the reaction conditions for both the cobalt- and iron-catalyzed coupling

reaction should be optimized in order to improve the yield of the coupled products and minimize

the formation of side products such as the elimination and homocoupling products. Different

changes that can be done during optimization are by trying different cobalt and iron catalysts or

different ligands than TMEDA. Once the reaction conditions are optimized, this study can provide

a further basis of the use of cobalt- and iron-catalyzed coupling reactions in synthesis towards

complex tetrahydropyrans.

44

Chapter 3

Experimental

General Methods:

All air sensitive reactions were performed in oven-dried glassware under an atmosphere of

argon. Tetrahydrofuran was dried over sodium/benzophenone ketyl and was distilled just prior to

use. All other reagents were reagent grade and were purified as necessary. Analytical thin layer

chromatography was performed on EM silica gel 60 F254 glass plates (0.25 mm). Flash column

chromatography was performed using SiliaFlash P60 silica gel (40-63 µm) from SiliCycle, Inc. 1H

NMR spectra were recorded on a Bruker Avance DPX-300 or ASCEND-400 spectrometers.

Chemical shifts are reported in ppm, downfield from tetramethylsilane using residual CHCl3

(d 7.27 ppm) as the internal standard.

Experimental Procedures:

Iodomethyltetrahydropyran 1: Alcohol 97 (1.00 g, 8.61 mmol) was dissolved in benzene (179

mL). Triphenylphosphine (3.39 g, 12.92 mmol), pyridine (2.16 mL, 26.71 mmol), and iodine (3.06

g, 12.06 mmol) were added sequentially, and the resulting mixture was refluxed for 4 h and then

cooled to room temperature. The reaction mixture was partitioned between ether (180 mL) and

deionized water (180 mL) and the layers separated. The organic layer was washed with Na2S2O3

(180 mL) and brine (180 mL) and dried over Na2SO4. The solution was concentrated to

approximately 45 mL in vacuo, rediluted with 1:1 hexane:ether, and stored at -20°C for 5 nights.

The reaction mixture was filtered through celite and the solvent was removed in vacuo. The

O I

1

O OHPPh3, pyridine, I2

benzene, reflux, 4 h87%

97

45

product was purified by flash chromatography (SiO2; hexane:ethyl acetate, 10:1) to provide the

iodide (1.70 g, 87%) as an orange liquid.

Iodomethyltetrahydropyran 1: Alcohol 8 (0.50 g, 4.99 mmol) and NaI (0.75 g, 5.00 mmol) were

dissolved in deionized water (15 mL). Oxone (1.53 g, 2.49 mmol) was added and resulting mixture

was stirred for 24 h. An additional 0.50 equivalent of oxone (1.53 g, 2.49 mmol) was added to the

reaction mixture and stirring continued for another hour. After that time, the reaction mixture was

washed with aqueous Na2S2O3 and the product was extracted with diethyl ether. The combined

organic layers were dried over Na2SO4, filtered, and the solvent removed in vacuo. The product

was purified by flash chromatography (SiO2; hexane:ethyl acetate, 20:1) to provide the iodide

(0.655 g, 58%) as an orange liquid.

TLC: Rf = 0.56 (hexanes: ethyl acetate, 5:1). 1H NMR (CDCl3, 400 MHz): d 4.05 (1H, d, J = 9.8

Hz), 3.50 (1H, t, J = 10.2 Hz), 3.30 (1H, m), 3.18 (2H, m), 1.86 (1H, dd, J = 6.3, 1.0 Hz), 1.56-

1.45 (4H, m), 1.32 (1H, m). Characterization data for this compound was consistent with that found

in the literature.xlvi

2-(2-trimethylsilylethyl)tetrahydro-2H-pyran 99: Anhydrous cobalt(II) chloride (0.0175 g,

0.135 mmol) was dried with a heat gun under high vacuum until the cobalt salt became dark blue

O INaI, oxoneWater, 24 h

58% 1HO

8

O I O TMS1 99

+ TMS MgCl

3

1. CoCl2 (5 mol%)THF, 0°C - rt, 1 h2. Zn dust, NH4Cl

EtOH, reflux1 h

46

and then placed under argon. Iodomethyltetrahydropyran 1 (0.300 g, 1.33 mmol) was dissolved in

THF (1.3 mL) and quantitatively transferred to the flask containing the dried cobalt(II) chloride.

The reaction mixture was cooled to 0°C and the Grignard reagent 3 (2.00 mL of a 1.0 M solution

in THF, 1.99 mmol) was added dropwise slowly over the course of 1 hour. The reaction mixture

was warmed to room temperature (25°C) and stirred for 1 hour. The reaction mixture was poured

into saturated ammonium chloride solution (20 mL) and extracted with ethyl acetate (20 mL x 2).

The combined organic layers were dried over Na2SO4, filtered, and the solvent removed in vacuo.

The crude mixture was dissolved in ethanol (8.50 mL) and solid NH4Cl (0.115 g, 2.15 mmol) and

zinc dust (0.275 g, 4.21 mmol) were added. The resulting mixture was refluxed for 1 hour then

cooled to room temperature. The reaction mixture was poured into ether (42.5 mL) and filtered

through celite. The layers were separated and the organics were washed with water, Na2S2O3, water

again, brine. The organic layer was dried with Na2SO4, filtered, and the solvent removed in vacuo.

The crude mixture showed the presence of the desired coupling product 99 by NMR. This mixture

was subjected to flash chromatography (SiO2; hexane:ethyl acetate, 20:1), but the coupled product

was not isolated due to lack of methods to visualize the compound during purification. 1H NMR:

(CDCl3, 400 MHz) d 3.99-3.95 (1H, m), 3.63 (2H, t, J = 6.4 Hz), 3.16-3.10 (1H, m), 1.83-1.80

(2H, m), 1.60-1.18 (4H, m), 0.89-0.78 (1H, m), 0.59 (1H, td, J = 14.1, 4.6 Hz), 0.44 (1H, td, J

=13.2, 4.8 Hz), -0.03 (9H, s)

47

2-(3-methoxybenzyl)tetrahydro-2H-pyran 100:

a) Preparation by cobalt-catalyzed coupling method:

Anhydrous cobalt(II) chloride (0.0176 g, 0.136 mmol) was dried with a heat gun in vacuo

until the cobalt salt became dark blue and then placed under argon. Iodomethyltetrahydropyran 1

(0.302 g, 1.33 mmol) was dissolved in THF (1.3 mL) and quantatively transferred to the flask

containing the dried cobalt(II) chloride. The reaction mixture was placed in a water bath at 0°C

and the Grignard reagent 4 (2.00 mL of a 1.0 M solution in THF, 1.50 mmol) was added dropwise

slowly over the course of 1 hour. The reaction mixture was warmed to room temperature (25°C)

and stirred for 2 hours. The reaction mixture was poured into saturated ammonium chloride

solution (20 mL) and extracted with ethyl acetate (20 mL x 2). The combined organic layers were

dried over Na2SO4, filtered, and solvent removed in vacuo. The crude mixture was purified by

flash chromatography (hexanes:ethyl acetate, 40:1 from fractions 1-14; hexanes:ethyl acetate, 20:1

from fractions 15-22) to give the coupled product 100 (0.027 g, 10%) as a colorless oil.

b) Preparation by iron-catalyzed coupling method:

Iodomethyltetrahydropyran 1 (0.302 g, 1.33 mmol), iron(III) acetylacetonate (0.024 g, 0.0691

mmol), hexamethylenetetramine (HMTA) (0.0096 g, 0.068 mmol), and TMEDA (0.02 mL, 0.133

O I THF, 0°C - rt, 1 h10% O

+5% CoCl2 (0.1eq)MgBr

OCH3

OCH3

1 4 100

O IO

+

MgBr

OCH3

OCH3

1 4 100

Fe(acac)3 (5 mol%)/TMEDA/HMTA (1:2:1)

THF, 0°C- rt, 24 h31%

48

mmol) were dissolved in THF (0.80 mL). The reaction mixture was then placed in a water bath at

0°C and the Grignard reagent 4 (2.00 mL of a 1.0 M solution in THF, 1.99 mmol) was added

dropwise slowly over the course of 1 hour and the reaction mixture allowed to stir overnight. The

reaction mixture was washed with 1.0 M aqueous hydrochloric acid (4 mL) and extracted with

hexanes (4 mL x 3). The combined organic layers were dried over Na2SO4, filtered, and the solvent

removed in vacuo. The crude mixture was purified by flash chromatography (hexanes:ethyl

acetate, 40:1 from fractions 1-22; hexanes:ethyl acetate, 20:1 from fractions 23-40) to give the

coupled product 100 (0.085 g, 31%) as a colorless oil.

TLC: Rf = 0.435 (hexanes:ethyl acetate, 10:1). 1H NMR: (CDCl3, 400 MHz) d 7.21 (1H, t, J = 8.0

Hz), 6.83-6.76 (3H, m), 4.01-3.97 (1H, m), 3.80 (3H, s), 3.53-3.39 (2H, m), 2.87 (1H, dd, J = 13.6,

6.6 Hz), 2.63 (1H, dd, J = 13.6, 6.5 Hz), 1.83-1.80 (1H, m), 1.65-1.39 (4H, m), 1.35-1.25 (1H, m).

Characterization data for this compound was consistent with that found in the literature.li

2-(2-Methylbenzyl)tetrahydro-2H-pyran 101: Anhydrous cobalt(II) chloride (0.0178 g, 0.137

mmol) was dried with a heat gun in vacuo until the cobalt salt became dark blue and then placed

under argon. Iodomethyltetrahydropyran 1 (0.305 g, 1.35 mmol) was dissolved in THF (1.3 mL)

and quantitatively transferred to the flask containing the dried cobalt(II) chloride. The reaction

mixture was placed in a water bath at 0°C and the Grignard reagent 5 (2.00 mL of a 1.0 M in THF,

1.50 mmol) was added dropwise slowly over the course of 1 hour. The reaction mixture was

warmed up to room temperature (25°C) and stirred for 2 hours. The reaction mixture was poured

O I +MgCl 1. CoCl2 (5 mol%)

THF, 0°C - rtO

+2. Zn dust, NH4Cl

EtOH, reflux1 5 101 102

49

into saturated ammonium chloride solution (20 mL) and extracted with ethyl acetate (20 mL x 2).

The combined organic layers were dried over Na2SO4, filtered, and the solvent removed in vacuo.

The crude mixture was dissolved in ethanol (13 mL) and solid NH4Cl (0.163 g, 3.05 mmol) and

zinc dust (0.389 g, 5.95 mmol) were added. The resulting mixture was refluxed for 1.5 hours, then

cooled to room temperature. The reaction mixture was poured into ether (60 mL) and filtered

through celite. The filtrate was washed with water, Na2S2O3, water again, and brine. The organic

layer was dried over Na2SO4, filtered, and the solvent removed in vacuo. The crude mixture was

purified by flash chromatography (SiO2; hexane:ethyl acetate; 20:1) to provide a mixture of the

coupled product 101 (ca. 21% yield) and the homocoupled product 102 (0.05345 g) as a colorless

oil. These compounds were isolated as a 1:1 mixture of 101:102 that could not be separated.

2-(2-Methylbenzyl)tetrahydro-2H-pyran 101: TLC Rf = 0.55 (hexanes:ethyl acetate, 10:1). 1H

NMR: (CDCl3, 400 MHz) d 7.17-7.01 (4H, m), 3.91-3.87 (1H, m), 3.42-3.29 (2H, m), 2.82 (1H,

dd, J =13.8, 6.2 Hz), 2.57 (1H, dd, J = 13.8, 7.0 Hz), 2.24 (3H, s), 1.73-1.70 (1H, m), 1.53-1.48

(2H, m), 1.46-1.19 (4H, m). Characterization data for this compound was consistent with that

found in the literature.51

2,2'-dimethyl-1,1'-biphenyl 102: TLC Rf = 0.81 (hexanes:ethyl acetate, 10:1). 1H NMR: (CDCl3,

400 MHz) d 7.17-7.01 (8H, m), 1.97 (6H, s). Characterization data for this compound was

consistent with that found in the literature.lii

50

1,1'-(1,2-ethanediyl)bisbenzene 105: Anhydrous cobalt(II) chloride (0.0173 g, 0.133 mmol) was

dried with a heat gun in vacuo until the cobalt salt became dark blue and then placed under argon.

Iodomethyltetrahydropyran 1 (0.300 g, 1.33 mmol) was dissolved in THF (1.33 mL) and

quantitatively transferred to the flask containing the dried cobalt(II) chloride. The reaction mixture

was placed in a water bath at 0°C and the Grignard reagent 7 (2.00 mL of a 1.0 M solution in THF,

1.99 mmol) was added dropwise slowly over the course of 1 hour. The reaction mixture was

warmed to room temperature (25°C) and stirred for 1 hour. The reaction mixture was poured into

saturated ammonium chloride solution (20 mL) and extracted with ethyl acetate (20 mL x 2). The

combined organic layers were dried over Na2SO4, filtered, and the solvent removed in vacuo. The

crude mixture was purified by flash chromatography (SiO2; hexanes for fractions 1-19;

hexanes:ethyl acetate, 40:1 for fractions 20-45) to provide the homocoupled product 105 (0.114 g,

47%) as a colorless liquid. TLC: Rf = 0.72 (hexane:ethyl acetate, 10:1). 1H NMR: (CDCl3, 400

MHz) 7.37-7.33 (4H, m), 7.28-7.25 (6H, m), 3.00 (4H, s). Characterization data for this compound

was consistent with that found in the literature.liii

O I + CoCl2 (5 mol%)

1 7

MgCl

105THF, 0°C - rt

47%

51

References

1 Harvey, A. L.; Clark, R. L.; Mackay, S. P.; Johnston, B. F. Current strategies for drug discovery through natural products. Expert Opin. Drug Discov. 2010, 5, 559-568, DOI: 10.1517/17460441.2010.488263 2 Bucar, F.; Wube, A.; Schmid, M. Natural Product Isolation – How to Get From Biological Material to Pure Compounds. Nat. Prod. Rep. 2013, 30, 525-545, DOI: 10.1039/C3NP20106F 3 Karpf, M. From Milligrams to Tons: The Importance of Synthesis and Process Research in the Development of New Drugs. in Pharmaceutical Process Chemistry (eds Shiori, T.; Izawa, K.; Konoike, T.), Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, Germany. doi: 10.1002/9783527633678.ch1 4 Ball, P. Chemistry: Why Synthesis? Nature. 2015, 528, 327-329. doi:10.1038/528327a 5 Kunze, B.; Jansen, R.; Sasse, F.; Hofle, G.; Reichenbach, H. J. Antibiot. 1998, 51, 1075-1080

6 Jansen, R.; Kunze, B.; Reichenbach, H.; Hofle, G. Apicularen A and B, Cytotoxic 10-Membered Lactones with a Novel Mechanism of Action from Chondromyces Species (Myxobacteria): Isolation, Structure Elucidation, and Biosynthesis. Eur. J. Org. Chem. 2000, 2000, 913-919. DOI: 10.1002/(SICI)1099-0690(200003)2000:6<913::AID-EJOC913>3.0.CO;2-L 7 Bhattacharjee, A.; Seguil, O. R.; De Brabander, J. K. Total Synthesis and Biological Evaluation of Apicularen A and Synthetic Analogs. Tetrahedron Lett. 2001, 42, 1217-1220 DOI: 10.1016/S0040-4039(00)02245-0 8 Palimkar, S. S.; Uenishi, J. Total Synthesis of (−)-Apicularen A. Org. Lett. 2010, 12, 4160-4163. DOI: 10.1021/ol101753y 9 Alali, F. Q.; Rogers, L.; Zhang, Y.; McLaughlin, J. L. Unusual Bioactive Annonaceous Acetogenins from Goniothalamus giganteus. Tetrahedron. 1998, 54, 5833– 5844. DOI: 10.1016/S0040-4020(98)00286-5 10 Takahashi, S.; Yonezawa, Y.; Kubota, A.; Ogawa, N.; Maeda, K.; Koshino, H; Nakata, T.; Yoshida, H.; Mizushina, Y. Int. J. Oncol. 2008, 32, 451– 458 11 Takahashi, S.; Kubota, A.; Nakata, T. Total Synthesis of a Cytotoci Acetogenin, Pyranicin. Org. Lett. 2003, 5, 1353– 1356. DOI: 10.1021/ol034323m 12 Crimmins, M. T.; Jacobs, D. L. Asymmetric total synthesis of pyranicin. Org. Lett. 2009, 11, 2695. DOI: 10.1021/ol900814w 13 Griggs, N. D.; Phillips, A. J. A Concise and Modular Synthesis of Pyranicin. Org. Lett. 2008, 10, 4955– 4957. DOI: 10.1021/ol802041c

52

14 Ringel, S. M.; Greenough, R. C.; Roemer, S.; Connor, D.; Gutt, A. L.; Blair, B.; Kanter, G.; von Strandtmann, M. J. Antibiot. 1977, 30, 371−375 15 Gerth, K.; Washausen, P.; Höfle, G.; Irschik, H.; Reichenbach, H. J. Antibiot. 1996, 49, 71−75. 16 Kende, A. S.; Mendoza, J. S.; Fujii, Y. Total Synthesis of (+)-Ambruticin. Tetrahedron. 1993, 49, 8015. DOI: 10.1016/S0040-4020(01)88025-X 17 Liu, P., Jacobsen, E. N. Total Synthesis of (+)-Ambruticin. J. Am. Chem. Soc. 2001, 123, 10772-10773. DOI: 10.1021/ja016893s 18 Tang, Y.; Oppenheimer, J.; Song, Z.; You, L.; Zhang, X.; Hsung, R. P. Strategies and approaches for constructing 1-oxadecalins. Tetrahedron. 2006, 62, 10785-10813, DOI: 10.1016/j.tet.2006.08.054. 19 Brahmachari, G. Design for carbon-carbon bond forming reactions under ambient conditions. RSC Adv. 2016, 6, 64676-64725. DOI: 10.1039/C6RA14399G 20 Hartwig, J. F. Organotransition Metal Chemistry, from Bonding to Catalysis. University Science Books: New York, 2010. ISBN 1-891389-53-X 21 Frisch, A.C.; Beller, M. Catalysts for Cross-Coupling Reactions with Non-activated Alkyl Halides. Angew. Chem. Int. Ed. 2005, 44, 674–688. DOI:10.1002/anie.200461432 22 Jana, R.; Pathak, T. P.; Sigman, M. S. Advances in Transition Metal (Pd,Ni,Fe)-Catalyzed Cross-Coupling Reactions Using Alkyl-organometallics as Reaction Partners. Chem. Rev. 2011 111, 1417-1492. DOI: 10.1021/cr100327p 23 Terao, J.; Kambe, N. Cross-coupling reaction of alkyl halides with grignard reagents catalyzed by ni, pd, or cu complexes with π-carbon ligand(s). Acc. Chem. Res. 2008, 41, 1545-1554. doi:10.1021/ar800138a 24 Rivkin, A.; Njardarson, J. T.; Biswas, K.; Chou, T.; Danishefsky, S. J. (2002). Total syntheses of [17]- and [18]dehydrodesoxyepothilones B via a concise ring-closing metathesis-based strategy: Correlation of ring size with biological activity in the epothilone series. J. Org. Chem. 2002, 67, 7737-7740. doi:10.1021/jo0204294 25 Wang, J.; Manabe, K. High ortho preference in ni-catalyzed cross-coupling of halophenols with alkyl grignard reagents. Org. Lett. 2009, 11, 741-744. doi:10.1021/ol802824b 26 Buchwald, S. L.; Miyaura, N. Cross-coupling reactions: a practical guide; Springer: Berlin, 2002.

53

27 Knappke, C. E. I.; von Wangelin, A. J. 35 years of palladium-catalyzed cross-coupling with grignard reagents: How far have we come? Chem. Soc. Rev. 2011, 40, 4948. doi:10.1039/c1cs15137a

28 Anaikov, V. P. Nickel: The “Spirited Horse” of Transition Metal Catalysis. ACS Cat. 2015, 5, 1964-1971. DOI: 10.1021/acscatal.5b00072 29 Gosmini, C., Bégouin, J., Moncomble, A. Cobalt-catalyzed cross-coupling reactions. Chem. Commun. 2008, 28, 3221-3233. DOI: 10.1039/B805142A 30 Bauer, I.; Knölker, H. Iron Catalysis in Organic Synthesis. Chem. Rev. 2015, 115, 3170-3387. DOI: 10.1021/cr500425u 31 Ohmiya, H.; Tsuiji, T.; Yorimitsu, H.; Oshima, K. Cobalt-Catalyzed Cross-Coupling Reactions of Alkyl halides with Allylic and Benzylic Grignard Reagents and Their Application to tandem radical Cyclicsation/Cross-Coupling Reactions. Chem. Eur. J., 2004, 10, 5640-5648. DOI: 10.1002/chem.200400545 32 Cahiez, G.; Moyeux, A. Cobalt-Catalyzed Cross- Coupling Reactions. Chem. Rev. 2010, 110 (3), pp 1435–1462. DOI: 10.1021/cr9000786 33 Cahiez, G.; Duplais, C.; and Moyeux, A. Iron-Catalyzed Alkylation of Alkenyl Grignard Reagents. Org. Lett. 2007, 9, 3253-3254. DOI: 10.1021/ol7016092 34 Ohmiya, H.; Yorimitsu, H.; Oshima, K. Cobalt-Mediated Cross-Coupling Reactions of Primary and Secondary Alkyl Halides with 1-(Trimethylsilyl)ethenyl- and 2-Trimethylsilylethynylmagnesium Reagents. Org. Lett. 2006, 8, 3093-3096. DOI: 10.1021/ol0611144 35 Fürstner, A.; de Souza, D.; Parra-Rapado, L.; Jensen, J. T. Catalysis-based total synthesis of latrunculin B. Angew. Chem. Int. Ed. 2003, 42, 5358-5360. doi:10.1002/anie.200352413 36 Cahiez, G.; Chaboche, C.; Duplai, C.; Moyeux, A. A New Efficient Catalytic System for the Chemoselective Cobalt-Catalyzed Cross-Coupling of Aryl Grignard Reagents with Primary and Secondary Alkyl Bromides. Org. Lett. 2009, 11, 277-280. DOI: 10.1021/ol802362e 37 Guérinot, A.; Reymond, S.; Cossy, J. (2007). Iron-catalyzed cross-coupling of alkyl halides with alkenyl grignard reagents. Angew. Chem. Int. Ed. 2007, 46, 6521-6524. doi:10.1002/anie.200702206 38 Kim, J. G.; Son, Y. H.; Seo, J. W.; Kang, E. J. Iron-catalyzed tandem cyclization and cross-coupling reactions of iodoalkanes and aryl grignard reagents: Iron-catalyzed tandem cyclization and cross-coupling reactions. Eur. J. Org. Chem. 2015, 2015, 1781-1789. doi:10.1002/ejoc.201403511

54

39 Tsuji, T.; Yorimitsu, H.; Oshima, K. Cobalt-catalyzed coupling reaction of alkyl halides with allylic grignard reagents. Angew. Chem. Int. Ed. 2002, 41, 4137-4139. doi:10.1002/1521-3773(20021104)41:21<4137::AID-ANIE4137>3.0.CO;2-0 40 Guisán-Ceinos, M.; Tato, F.; Buñuel, E.; Calle, P.; Cárdenas, D.J. Fe-catalysed Kumada-type alkyl-alkyl cross-coupling. Evidence for the intermediacy of Fe(I) complexes. Chem. Sci. 2013, 4, 1098-1104. DOI: 10.1039/C2SC21754F 41 Gonnard, L., Guérinot, A. and Cossy, J. Cobalt-Catalyzed Cross-Coupling of 3- and 4-Iodopiperidines with Grignard Reagents. Chem. Eur. J. 2015, 21, 12797–12803. doi:10.1002/chem.201501543 42 Sun, C., Krause, H., & Fürstner, A. A practical procedure for Iron-Catalyzed Cross-Coupling reactions of sterically hindered Aryl-Grignard reagents with primary alkyl halides. Adv. Synth. Catal. 2014, 356, 1281-1291. doi:10.1002/adsc.201301089 43 Lam, T., & Totah, N. I. (2015). Studies toward the synthesis of spirastrellolide A. preparation of a C31–C47 model. Tetrahedron Letters, 56(23), 3349-3352. doi:10.1016/j.tetlet.2015.01.026 44 Riaz, M. T.; Pohorilets, I.; Hernandez, J. J.; Rios, J.; Totah, N. I. Syracuse University, Syracuse, NY, unpublished work, 2016. 45 Totah, N. I.; Schreiber, S. L. Thermodynamic Spiroketalization as an Efficient Method of Stereochemical Communication. J. Org. Chem. 1991, 56, 6255-6256. DOI: 10.1021/jo00022a003 46 Firouzabadi, H.; Iranpoor, N.; Kazemi, S. Direct halogenation of organic compounds with halides using oxone in water – A green protocol. Can. J. Chem. 2009, 87, 1675-1681 47 Affo, W.; Ohmiya, H.; Fujioka, T.; Ikeda, Y.; Nakamura, T.; Yorimitsu, H.; Oshima, K.; Imamura, Y.; Mizuta, T.; Miyoshi, K. Cobalt-Catalyzed Trimethylsilylmethylmagnesium-Promoted Radical Alkenylation of Alkyl Halides:  A Complement to the Heck Reaction. J. Am. Chem. Soc. 2006, 128, 8068-8077 48 Ohmiya, H.; Yorimitsu, H.; Oshima, K. Cobalt(diamine)-Catalyzed Cross-coupling Reaction of Alkyl Halides with Arylmagnesium Reagents: Stereoselective Constructions to Arylated Asymmetric Carbons and Application to Total Synthesis of AH13205. J. Am. Chem. Soc. 2006, 128, 1886-1889 49 Hammann, J. M., Haas, D., & Knochel, P. (2015). Cobalt-Catalyzed negishi Cross-Coupling reactions of (hetero)arylzinc reagents with primary and secondary alkyl bromides and iodides. Angew. Chem. Int. Ed. 2015, 54, 4478-4481. doi:10.1002/anie.201411960 50 Dongol, K. G.; Koh, H.; Sau, M; Chai, C. L. L. Iron-Catalysed sp3–sp3 Cross-Coupling Reactions of Unactivated Alkyl Halides with Alkyl Grignard Reagents. Adv. Synth. Catal., 2007, 349, 1015–1018. DOI:10.1002/adsc.200600383

55

51 Molander, G. A.; Argintaru, O. A.; Aron, I.; Dreher, S. D. Nickel-catalyzed cross-coupling of potassium aryl- and heteroaryltrifluoroborates with unactivated alkyl halides. Org. Lett. 2010, 12, 5783-5785. doi:10.1021/ol102717x 52 Lee, D.; Qian, Y.; Park, J.; Lee, J.; Shim, S.; Jin, M. (2013). A highly active and general catalyst for the stille coupling reaction of unreactive aryl, heteroaryl, and vinyl chlorides under mild conditions. Adv. Synth. Catal. 2013, 355, 1729-1735. doi:10.1002/adsc.201300075 53 Chen, T.; Yang, L.; Li, L.; Huang, K. Homocoupling of benzyl halides catalyzed by POCOP-nickel pincer complexes. Tetrahedron, 2012, 68, 6152-6157. doi:10.1016/j.tet.2012.05.075

56

Appendix

9.59.0

8.58.0

7.57.0

6.56.0

5.55.0

4.54.0

3.53.0

2.52.0

1.51.0

0.5ppm

1.2601.2821.2871.3091.3131.3181.4481.4731.4751.4931.5011.5311.5471.5591.5611.7781.8121.8151.8411.8461.864

3.1783.2963.4443.4733.4983.502

4.0194.0234.048

1.55

3.23

2.09

2.030.95

0.95

1.00

O I1

57

-0.58.5

8.07.5

7.06.5

6.05.5

5.04.5

4.03.5

3.02.5

2.01.5

1.00.5

0.0ppm 0.835

0.8460.8550.8650.8730.8851.1791.1971.2141.2461.3671.3721.3861.3991.4211.4261.4361.4451.4581.4661.4771.4821.4951.5061.5161.5281.5431.5591.5691.5801.5981.6141.6491.8001.8101.8261.8312.0502.0682.0862.1033.0993.1033.1183.1303.1423.1573.1613.4433.4603.4783.4953.6183.6343.6503.9493.9543.9593.9773.9823.986

7.269

9.002.22

0.93

0.95

1.18

17.57

10.90

2.10

1.80

0.90

0.809.89

1.93

0.95

1.63

0.79

O TMS99

58

9.59.0

8.58.0

7.57.0

6.56.0

5.55.0

4.54.0

3.53.0

2.52.0

1.51.0

0.5ppm 1.483

1.4931.5041.5081.5141.5351.5451.5561.5681.5781.5871.5971.6091.6181.6401.6501.8011.8061.8141.8241.8321.8372.6072.6232.6412.6572.8502.8662.8842.9003.3933.3993.4233.4293.4523.4583.4703.4753.4873.4913.4963.5033.5083.5133.5183.5303.5353.8043.9723.9783.9834.0014.0064.012

6.7566.7776.7896.8076.8267.1887.2077.2277.264

1.24

2.122.11

1.06

1.01

1.01

2.02

3.00

1.02

2.93

0.98

O

OCH3

100

59

9.59.0

8.58.0

7.57.0

6.56.0

5.55.0

4.54.0

3.53.0

2.52.0

1.51.0

0.5ppm 1.331

1.3401.3491.3741.4021.4061.4561.4771.4861.4991.5081.5181.5291.6971.7031.7191.7281.7341.9682.2402.5462.5632.5802.5982.7962.8112.8302.8463.2913.2983.3213.3273.3503.3563.3653.3813.3923.3983.4033.4073.4193.4243.8723.8773.8823.9003.9063.911

7.0097.0237.0317.0347.0397.0497.0597.0727.0817.1077.1167.1217.1297.1337.1427.1477.174

3.55

2.03

0.98

3.25

2.89

1.00

1.00

2.02

1.00

4.983.32

O+

101 102

60

9.59.0

8.58.0

7.57.0

6.56.0

5.55.0

4.54.0

3.53.0

2.52.0

1.51.0

0.5ppm

3.004

7.2537.2737.2907.3417.3597.378

4.00

5.894.07

105

61

62

63