Rochester Institute of Technology Rochester Institute of Technology

RIT Scholar Works RIT Scholar Works

Theses

8-14-1993

Organometallics: (I) Synthesis of a macrocyclic complex Organometallics: (I) Synthesis of a macrocyclic complex

praeseodymium and (II ) effect of wilkinson's catalyst on the praeseodymium and (II ) effect of wilkinson's catalyst on the

cleavage of cyclohexene oxide by borane cleavage of cyclohexene oxide by borane

Carolyn Ruth

Follow this and additional works at: https://scholarworks.rit.edu/theses

Recommended Citation Recommended Citation Ruth, Carolyn, "Organometallics: (I) Synthesis of a macrocyclic complex praeseodymium and (II ) effect of wilkinson's catalyst on the cleavage of cyclohexene oxide by borane" (1993). Thesis. Rochester Institute of Technology. Accessed from

This Thesis is brought to you for free and open access by RIT Scholar Works. It has been accepted for inclusion in Theses by an authorized administrator of RIT Scholar Works. For more information, please contact ritscholarworks@rit.edu.

ORGANOMETALLlCS: (I) SYNTHESIS OF A MACROCYCLIC COMPLEXOF PRAESEODYMIUM and (II) EFFECT OFWILKINSON'S CATALYST ON THE CLEAVAGE OFCYCLOHEXENE OXIDE BY BORANE

CAROLYN RUTH

AUGUST, 1993

THESIS

SUBMITTED IN PARTIAL FULFILLMENT OF THEREQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE

APPROVED:

Terence C. MorrillProject Advisor

Department Head

Rochester Institute of TechnologyRochester, New York 14623

Department of Chemistry

PERMISSION GRANTED

Title of Thesis: ORGANOMETALLlCS: (I) SYNTHESIS OF A

MACROCYCLIC COMPLEX OF PRAESEODYMIUM and (II) EFFECT OF

WILKINSON'S CATALYST ON THE CLEAVAGE OF CYCLOHEXENE OXIDE

BY BORANE

I, C AtZo LyAi RLlT H_. hereby grant permission to

the Wallace Memorial Library of the Rochester Institute of Technology to

reproduce my thesis in whole or in part. Any reproduction will not be for

commercial use or profit.

Signature of Author

ACKNOWLEDGMENTS

The author would like to thank the Department of Chemistry at

Rochester Institute of Technology for the support of this work as well as Dr.

T.C. Morrill, research advisor, and Drs. R. Gilman, K. Turner, and W. Hallows

for serving as members of the research advisory committee.

ABSTRACT

(I) The template synthesis and characterization of the complex of

praeseodymium acetate from 2,6-diacetylpyridine and ortho-

phenylenediamine was attempted. The attempted synthesis produced a

green precipitate with a high melting point and poor solubility in organic

solvents, unlikely characteristics for compound 1 seen on the

page. This project was set aside due to the apparent failure to produce the

desired compound.

(II) The effect ofWilkinson's catalyst on the treatment of cyclohexene

oxide with borane-THF at72 for 5 1/2 hours was also investigated.

Oxidation of the organoborane formed by both the catalyzed and

uncatalyzed reactions produced cyclohexanol. Another significant product

was one which, according to GC-MS, had a molecular weight of 172. This

product was not completely identified. Studies of the progress of both the

catalyzed and uncatalyzed reactions were conducted using TLC. TLC

showed no evidence of the starting material, cyclohexene oxide, after 30

minutes of reflux. No change in the product mixture measured by TLC

occurred after 30 minutes for the catalyzed reaction and after 60 minutes for

the uncatalyzed reaction, indicating that the reactions were probably

complete at that point.

TABLE OF CONTENTS

SECTION PAGE

I. INTRODUCTION 1

II. PART ONE: HISTORICAL REVIEW 3

III. PART TWO: HISTORICAL REVIEW 9

IV. EXPERIMENTAL SECTION: Instrumentation and Chemicals 19

V. PART ONE: EXPERIMENTAL SECTION 21

VI. PART TWO: EXPERIMENTAL SECTION

A. Catalyzed Reaction 24

B. Uncatalyzed Reaction 30

C. Reaction Study Using TLC 35

VII. PART ONE: RESULTS AND DISCUSSION 41

VIII. PART TWO: RESULTS AND DISCUSSION 45

IX. REFERENCES 53

X. APPENDIX: IR, GC, GC-MS, and NMR Spectra 56

iii

INTRODUCTION

This thesis deals with two different research projects, both of which

are described below.

Part One describes the first research project: the attempted

synthesis and characterization of the praeseodymium (III) macrocyclic

complex (1) derived from 2,6-diacetylpyridine and orf/70-phenylene diamine.

2 OAc

CI-

n H20

(1)

Part Two describes the second research project: a study of the

hydroboration of cyclohexene oxide (2 ) and the effect of Wilkinsons catalyst

1

on this reaction. The products formed and the effect of the catalyst upon the

progress of the reaction were examined.

+ BHyTHFcatalyst

THF

Ethanol

NaOH

H202

(2)

no catalyst

+ BH3THF ==>3

THF

Ethanol

NaOH

H202

(2)

PART ONE: HISTORICAL REVIEW

Nuclear magnetic resonance is the most useful technique for organic

structure determination. 1H NMR, which is frequently used, reveals the

number of nonequivalent protons present in a compound, the number of

neighboring protons, and finally the total number of each type of proton. As

with any spectrometric technique, resolution is a major concern, where

resolution is the separation between 1H NMR signals. One method to

improve resolution is to increase the strength of the instrument magnetic

field in order to separate overlapping chemical shifts of nonequivalent

protons.

Solvent induced shifts have also proven to be useful in the

simplification of NMR spectra as demonstrated by Hinckley in 1969 (1). In

addition, paramagnetic ions were known to electrostatically interact with

electron rich substrates to cause NMR signal shifts as demonstrated by

Sanders and Williams in 1970 (2). As a result, the field of lanthanide shift

reagents expanded greatly during the next decade or so.

As the name suggests, LSRs (lanthanide shift reagents) employ

lanthanide metals to enhance the nonequivalence of nuclei to produce, in

many cases, first-order proton NMR spectra. LSRs require three basic

characteristics. First, the lanthanide metal cations must be paramagnetic.

Secondly, these LSRs must function as Lewis-acids: the Lewis-acidic site is

clearly the metal cation and these coordinate with the Lewis-base sites on

the organic substrate. Third, the LSRs should be soluble in nonpolar

organic solvents commonly used for NMR analysis. The NMR solvent

should be a nonpolar organic compound that is less Lewis-basic than the

substrate in order to avoid competition between the substrate and solvent for

the LSR (3).

In Figure 1 below, a proton NMR spectrum of 0.1 M toluene in

deuterated chloroform shows the effect of the binuclear shift reagents used

for spectrum (b) when compared with spectrum (a). The binuclear shift

reagents have peaks of their own, of course, but fortunately these peaks do

not appear in the same region as the toluene peaks. In Figure 2, the

structure of the shift reagents are shown (4).

"B Hc

Pr

7 6 5 4 3 2 1 Oppm

Proton NMR spectrum of 01 M toluene in CDCIj with (a) no shift reagent.

ndfb/0 2M PrfJFodi, and 0.2 M Ag?od)

o<CH>>

oQuu'CF,CF,CF,

oQuu.'CF,CF,CF,

Figure 1 Figure 2.

Paramagnetic ions exert a strong magnetic field of their own which

has the capability of inducing shifts and thus spreading out NMR spectra (5).

In order for this secondary magnetic field to be influential, the nucleus of the

substrate must be in the vicinity of the cation. This is achieved due to the

equilibria established in the following equation:

4

LSR + S = LSR-S where S = organic substrate

Crown ethers, macrocylic ethers of ethylene glycol, have been used

to encapsulate lanthanide metal cations. Employing crown ethers with the

appropriate cavity size made it possible for the metal cation to form

coordinate covalent bonds to the oxygen atoms in the ether. At the same

time, the hydrocarbon exterior around the metal cation made solubility in

nonpolar solvents possible. DeCann characterized the 15-crown-5

praeseodymium (III) nitrate complex and attempted to employ it as an NMR

shift reagent (5). The complex, however, was only soluble in polar solvents

and dissociated in solution. It was not any more effective as an NMR shift

reagent than praeseodymium (III) nitrate alone.

Nitrogen-containing macrocycles were shown to have some

advantages over crown ethers in 1983 by Hiroshi and Tsukube (6). The

hexaaza and tetraaza nitrogen containing macrocycles, shown in Figure 3,

were selective for ammonium cations over potassium cations as opposed to

a typical crown ether which was less selective among cations. The host

cation also bound more tightly to the nitrogen atoms of the polyamines than

to the oxygen atoms of the polyethers. This enabled the complex to be more

stable in solution.

c. . ; c 3ft.CM.fth

ft.CM2fth

Figure 3.

The field of lanthanide metal/nitrogen containing macrocyclic

complexes and their potential as proton NMR shift reagents has been

investigated by many researchers. In 1979, Alan Hart and co-workers

illustrated that the template condensation of 2,6 diacetylpyridine with

ethylenediamine gives the 18-member hexaaza macrocyclic complexes of

the type [M(N03 )3 L] where M=La or Ce and L= 2,7,1 3,1 8,-tetramethyl-

3,6,1 4,1 7,23,24-hexaazatricyclo [1 7.3.1 8-12] tetracosa-1(23),2,6,8,10, 12

(24), 1 3,1 7,1 9,21 -decaene (7). The complexes are kinetically stable in water

with respect to dissociation of the macrocycle. The Ce complex in particular

has proven to be a potential aqueous NMR shift reagent.

L.M. Vallarino and co-workers illustrated in 1984 that the 18-member

hexaaza-macrocyclic ligand having the formula C22H26 N6- can De frmed

via the templating effect, used by Hart, with any one of the lanthanide ions

(excluding radioactive Pm) along with the condensation of 2,6- diacetyl

pyridine and 1,2-diaminoethane (8):

rNHj H2N

^0.1 M N.C1 r

N... j y

Nn 2

OAc'

S Ln(OAc)j J"

. ! JG"

O O

^-v

nHjO

U Pr. Od. Eu. U

Research involving lanthanide metal/nitrogen containing macrocyclic

complexes and their potential as proton NMR shift reagents has continued at

Rochester Institute of Technology under the direction of T.C. Morrill and his

research students. Mary DiSano, during 1988-89, synthesized a hexaaza

macrocyclic ligand coordinated to Pr (3+) as a potential lanthanide shift

reagent (9). Nitrate, carbonate, picrate, and chloroacetate counteranions

were employed in order to put on anions that readily dissociate compared to

the original praseodymium acetate template complex. Unfortunately, a

successful lanthanide shift reagent was not synthesized. Insolubility of the

complexes in suitable NMR reagents, difficulties in purification of the

complexes, and failure of the complexes to induce meaningful NMR shifts in

the spectrum of 1- heptanol were some of the problems encountered.

During 1990, Eileen George, an undergraduate student at RIT,

attempted the reaction seen below (10). In this reaction, the acetate and

chloride anions on a hexaaza macromolecule are changed to

tetraphenylborate anions with the hope that the tetraphenylborate ions

would increase solubility in nonpolar NMR solvents and because of their

bulkiness would coordinate to the metal cation less strongly. The inhibited

coordination to the cation would, theoretically, allow the cation to coordinate

to basic sites in organic substrates and thus improve the resolution of the

NMR spectra of those substrates. This project was only partly successful

with respect to the attempt to change the anions on the complex to

tetraphenylborate anions. Analysis by IR and NMR showed that only one

tetraphenylborate anion had been substituted for the acetate and chloride

anions and not three as are indicated in the reaction below:

N.i N ^N-,

I3:- NN 3B(CH)

f '"i*:1+ 12nA'

+ xsNaB(CH) ? > J' ' *

^y.- ^y.n H O

In the summer of 1 991,Dr. Robert Bryant from the University of

Rochester suggested that a lanthanide macromolecule made with four

aromatic rings instead of two would make a complex that was more planar

and hence would make a better contrast agent for MRI (11). From that

suggestion came this thesis project: the synthesis of a hexaaza

macromolecule (1 ) from 2,6 - diacetylpyridine, ortfto-phenylenediamine, and

praeseodymium acetate using the template synthesis method successfully

employed by A. Hart (7) and L. Vallarino (8), and then, if the synthesis was

successful, to test it as a NMR shift reagent:

-A-a

NH2NaCl

+ Pr(OAc)3NH, CH3OH

2 OAc

a-

nH20

(1)

PART TWO: HISTORICAL REVIEW

The synthetic procedure which can lead to net anti-Markovnikov

hydration of an alkene is hydroboration-oxidation. It was developed by H.C.

Brown and his coworkers at Purdue University in the 1950's as part of a

broad program designed to demonstrate how boron-containing reagents

could be applied to organic chemical synthesis (12).

Diborane, B2H5, is a colorless gas which can dissolve in aprotic

solvents like tetrahydrofuran (THF). These solvents form"loose"

Lewis acid-

base complexes with borane, BH3 (13):

H

H* A /\

h'\\* 2

(_J' Mq

THF

Hydroboration is a reaction in which borane (BH3), or alternatively,

RBH2 or R2BH compounds add to a carbon-carbon bond. New carbon-

hydrogen and carbon-boron bonds are created. An alkylborane is formed

from an alkene in a hydroboration reaction. The B-H bond always adds syn

across a C=C bond (12):

R

0^

H

H BH3:THF

R h H rh

THF LI

The mechanism in Figure 4 below shows how the B-H unit adds to the

pi bond in an alkene (14). Because the pi bond is electron rich and borane

is electron poor, it is reasonable to formulate an initial Lewis acid-base

complex requiring the participation of the empty p orbital on BH3, as in the

borane-THF complex. Subsequently, one of the hydrogens is transferred by

means of a four-center transition state to one of the alkene carbons, while

the boron shifts to the other without any additional intermediates. The

stereochemistry is syn. All three B-H bonds are reactive in this way.

Hydroboration is not only stereospecific but also regioselective. The boron

binds to the less-hindered (less-substituted) carbon.

Q-0

Emprj p orfriul

Mechanism of Hydroboration:

-Q G

cc-

H-B H

H>UXH

H

Boraiw-ilkent compkv

HjBX

Figure 4.

10

Following hydroboration, the organoborane is oxidized by treatment

with hydrogen peroxide in aqueous base. This is the oxidation stage of the

sequence; hydrogen peroxide is the oxidizing agent, and the organoborane

is converted to an alcohol (12):

VJUbh, H202

Cr^ OH

EtOH

O

The mechanism for the oxidation stage is (15):

H202 + 'OH >HOO"

+ H20

\/R

B-R + "OOH > [ B V ] >

/ /0N

0(0H

"OH + B-OR

/

H20

B-O-R > B-OH +ROH

' Hydrolysis

B-OH contains two more B-H bonds, so the above hydroboration-

oxidation process can take plaGe two more times.

Additional advantages of a hydroboration-oxidation synthesis of

alcohols are simplicity of procedure, relatively mild reaction conditions (for

11

example, temperature), high overall yields, and absence of skeletal

rearrangements (15).

In 1985, a paper was published by Mannig and Noth (16) on catalytic

hydroboration using Wilkinson's catalyst and a relatively unreactive

hydroborating reagent: 1,3,2-benzodioxabarole (catecholborane).

Catecholborane reacts with alkynes and alkenes, but only at elevated

temperatures:

aII B-H ? R-CC-H

*

Z"0^IL / 70 C H B-0

2 hn.O

catecholborane

C^O RR" R h

R"

Bx \-Sf Vh ?

sc-c'

- y-4{o

L0/ R'

- lOOj R^\

catecholborane

In the presence of Wilkinson catalyst, [chlorotris(triphenyl phosphine)

rhodium (I) ] (3), seen in Figure 5, these reactions proceeded rapidly (25

minutes) at room temperature (20).

/Rh

PCPb3)

(Pb3)P<':P(Pbj)

(3)

Figure 5.

12

Moreover, when molecules such as 5-hexene-2-one, which contain

both a double bond and a carbonyl group, were reacted with catechol

borane in the presence of Wilkinson's catalyst, the carbon-carbon double

bond was completely reduced and the keto carbonyl was not reduced. See

reaction below. Therefore, Wilkinson's catalyst activated the double bond

and not the carbonyl (16).

CO- -*--co

(3)

0^

The complex mechanism for the result above was proposed by

Mannig and Noth in 1985 and can be represented by a catalytic loop (17).

See Figure 6. Wilkinson's catalyst is a 16 electron complex. In solution,

Wilkinson's catalyst dissociates with one of its triphenylphosphine ligands

and after oxidative addition of the B-H bond from catecholborane (and

possibly complexation with a solvent molecule), Wilkinson's catalyst

becomes a stable 18 electron complex. See (c) below. This complex then

binds with an olefin present in solution. The next step of the reaction

involves insertion of the double bond into the Rh-H bond. Once this is

complete, irreversible reductive elimination of the organoborane returns

Wilkinson's catalyst back to a 16 electron complex which can undergo

reaction again.

13

L-PPh,

reductive

elimination

RnljCI WlllunMO I CJtJiyS,

t

RhL,a

Ch<

c-ci

u C /

***** V. /\ J.-i _Xolefin

':i'

W Nl bntf,n9migration

Figure 6.

Wilkinson's catalyst was first used in 1966 for the homogeneous

hydrogenation of double and triple bonds (18). In fact, the hydroboration

mechanism above is merely a modification of the homogeneous

hydrogenation mechanism.

KR'"

(3)

H2,25

C

Ethanol

RR"

*R'"

R' HR

H

Before Wilkinson's catalyst, hydrogenations were performed

heterogeneously with metals such as platinum or palladium (19). These

reactions were often performed at high temperatures and pressures, and the

yields were often low. Wilkinson's catalyst hydrogenations can be

14

performed at room temperature and standard pressure, with the time of

reaction being short and the yields high.

One interesting point about the ability of Wilkinson's catalyst to

activate double and triple bonds is the possibility of activating other pi

systems, or in the case of cyclopropane rings,"pseudo"

pi systems.

The cyclopropane molecule has attracted a great deal of interest

because of its highly strained bond angles and its unusual chemistry, which

is in several respects similar to that of alkenes. Walsh (20) proposed an

orbital model for bonding in three-membered rings designed to account for

this similarity. Because the H-C-H bond angle in cyclopropane is close to

1 20 (21), Walsh chose sp2 hybridization for the carbon, with one hybrid

being used for each C-H bond and the third pointing directly into the center

of the ring. See (a) in Figure 7 below. The unhybridized p orbital then lies

in the ring plane. See (b) in Figure 7 below (22).

<

(bi

Figure 7.

15

Walsh proposed that these atomic orbitals interact in two sets. The

three sp2 hybrids pointing toward the ring center overlap and form three

molecular orbitals, one bonding and two antibonding (22). These orbitals

are shown in (a) of Figure 8 below. The bonding member of the set holds

two electrons which are delocalized in the region inside the ring. The three

p orbitals interact through overlap that is between the end-on type

characteristic of sigma bonds and the side-on type characteristic of pi bonds

to form the three delocalized orbitals shown below in (b) of Figure 8. This

time two are bonding and one is antibonding. Each of the bonding pi-type

orbitals has one node, but the net effect is bonding.

7 Vi

<.*)

Figure 8.

The Walsh model makes clear the pi nature of cyclopropane and also

indicates that significant electron density will be concentrated outside the

equilateral triangle defined by the internuclear lines (22). The possibility

thus exists that hydroborating reagents activated by Wilkinson's catalyst

could facilitate the ring opening reaction of a cyclopropane ring.

Research in the area of using Wilkinson's catalyst to activate

cyclopropane rings in a hydroboration reaction has been carried on at

16

Rochester Institute of Technology under the direction of Dr. T.C. Morrill. One

of his students, Kevin Gillman, worked with quadricyclene (tetracyclo

[3.2.O.2-704'6

] heptane) (23). Quadricyclene contains about 96 kcal/mole

of strain energy due to its restricted polycyclic nature. When Gillman

allowed quadricyclene to react with catecholborane and Wilkinson's

catalyst, the product was notricyclanol in 9.9% yield:

1. (PPhj^RiiCl

Catecbolborme

THF ,5C

2. H202 , OH

quadricyclene notricyclanol (9.9%)

Donna Chen, in the same laboratories, attempted the catalyzed

hydroboration of norcarane (bicyclo[4.1.0.]heptane) using BH3-THF. GC-MS

analysis showed that only the starting material, norcarane, appeared in the

ether extract (24):

o + BH3:THF

l.(PPh3)3Rha

Reflux

ZHjOj.OH

norcarane

Cyclohexene oxide, (7-oxabicyclo[4.1.0]heptane), (2) seen below is

very similar to norcarane except that it contains a 3-membered heterocyclic

ring. The oxygen atom has unshared pairs of electrons to which the rhodium

17

atom in Wilkinson's catalyst would be attracted. The research project

developed for this thesis was to study the hydroboration of cyclohexene

oxide (2) and the effect of Wilkinson's catalyst on this reaction.

+ BH3THF

catalyst

THF

no catalyst

+ BH3THF=7^r>

3THF

Ethanol

NaOH

H202

Ethanol

NaOH

H202

(2)

18

EXPERIMENTAL SECTION

Instrumentation and Chemicals:

GC Spectra were obtained with a Hewlett Packard 5890

Series II Chromatograph. The Helium flow rate was 1.3 mL/min. A

J&W Durabond-5 capillary column was used that is relatively

nonpolar and composed of 95% dimethyl-(5%)-diphenyl-

polysiloxane.

GC Parameters Used: Initial temperature =80 (2 minutes)

Ramp: 10/minute (7 minutes)

Final temperature =150 (6 minutes)

Injection Port =225

Detector =225

GC-Mass Spectroscopy was performed on a Hewlett Packard

5995 Spectrometer which was equipped with a J&W Durabond-1

column composed of 100% dimethyl-polysiloxane (nonpolar). The

parameters used were the same as those listed for the gas

chromatograph above.

Infrared Spectra were obtained with a Perkin Elmer 1760X

FT-IR Spectrometer and also with a Perkin Elmer 681

Spectrophotometer. The solid samples in Part One were analyzed

as KBr pellets and the liquids obtained in Part Two were

dissolved in deuterated chloroform and analyzed between salt

cells.

19

The reagents and solvents used were purchased from

Aldrich Chemical Company, Inc., Fisher Scientific, Inc., and J.T.

Baker, Inc. No further purification of these chemicals was done.

Dry nitrogen was prepared by passing reagent grade nitrogen

from Linde Air Products through a Fieser's solution (25) followed

by passage through concentrated H2S04 and finally through

anhydrous KOH.

Proton NMR spectra were obtained on a Bruker 200 MHz,

courtesy of Fisons Pharmaceutical Company. Deuterated

chloroform was used as solvent.

TLC was done using silica gel glass plates, Si-250F,

obtained from J.T. Baker, Inc. The plates were developed with a

3:2 pentane:ethyl ether mixture and then sprayed with a developer

and heated on a hot plate. The development spray (26) was

composed of 0.5 mL p-anisaldehyde in 9 mL 95% ethanol

containing 0.5 mL H2S04 and a few drops of glacial acetic acid.

Liquid chromatographic (LC) columns were made using

fluorescent silica gel of pore size 60 obtained from E. Merck,

Germany.

All mobile phase solutions for TLC and for liquid

chromatographic columns were prepared in volume/volume ratios.

Melting points were determined with a Mel-temp apparatus

from Lab Specialties Inc. and required no correction as per

calibration by benzoic acid.

The thermogravimetric analysis was obtained from a

Perkin-Elmer TGS-2 instrument heated at a rate of 20

degrees/min. under nitrogen gas at a flow rate of 55 mL/min.

20

PART ONE: EXPERIMENTAL SECTION

Attempted Template Synthesis of the Praeseodymium (III)Macrocyclic Complex derived from 2,6-Diacetylpyridine and

Orf/70-phenylenediamine

c -a

NH2NaCl

2 ^JL *L^ +2 B J + Pr(OAc)3

NH2CH3OH

O

; %KN^^^2 OAc

-

a-

nH20

(1)

Praeseodymium acetate hydrate, 0.3720 g (1 mmol), was mixed with

anhydrous methanol and swirled into a 250 mL round-bottom flask. Then

0.3264 g (2 mmol) of 2,6 - diacetylpyridine dissolved in 20 mL methanol and

0.2163 g (2 mmol) of o/t/70-phenylenediamine dissolved in 5 mL methanol

were both poured into the flask. To this mixture was added 0.0292 g NaCl in

5 mL methanol.

The mixture refluxed at65 for 5 hours. The color gradually changed over a

period of 3 hours from cloudy green to cloudy yellow. The reaction mixture

was allowed to cool to room temperature. A pale green precipitate

21

separated from the yellow reaction liquid and this was removed by filtration

through filter paper. The precipitate had a mass of 0.2536 g after being

dried in a 1 00 oven for one hour.

An IR spectrum of the green precipitate showed a strong broad band

between 3200 and 3400 cm"1

(OH) and two strong peaks between 1375 and

1575 cm-1 (COO"). There was also a weak narrow band at 1600 cm-1 (C=N).

The precipitate was slightly soluble in water and DMSO; it was insoluble in

benzene, chloroform, pyridine, ethyl acetate, carbon disulfide, and acetone.

While attempting to determine a melting point, the precipitate seemed to

shrink, looked slightly wet and became a darker green at 160. Between

220and 260, the green solid stuck to the sides of the melting point tube,

while a greenish liquid rested on the bottom of the tube. At 280, the

material turned brown (decomposed).

The green precipitate was placed in a crucible for a combustion test. It

burned slowly and completely charred.

The precipitate tested negative for the presence of chloride ion. The

alcoholic silver nitrate test produced no precipitate and the Beilstein test

showed an orange flame.

Thermogravimetric analysis suggested that 4.9% of the green precipitate

was water.

Tests to determine purity with TLC were unsatisfactory due to the difficulty of

dissolving the precipitate in organic solvents.

This reaction was also carried out in acetonitrile and in dioxane as solvents.

Although some of the results obtained from product analysis were

22

unsatisfactory (especially solubility), complete characterization of the

products was not done.

This research project was set aside in October, 1992. The reasons for this

decision are discussed in Part One of the Results and Discussion section.

23

PART TWO: EXPERIMENTAL SECTION

Treatment of Cyclohexene Oxide with BoraneTHF Complex in

the presence of Wilkinson's Catalyst

+ BH3-THF

Ethanol

catalyst NaOH

THF H202

(2)

A 3-neck round-bottom flask was flushed by a five-minute flow of

deoxygenated and dried nitrogen gas. Then, 100 mL (0.1 mole) of 1 M

BH3-THF was injected into the flask using a syringe followed by the addition

of 0.1 g CIRh[P(C6H5)3]3, Wilkinson's catalyst. After the catalyst was

dissolved, 10.0 g (0.1 mole) of cyclohexene oxide was added. The color of

the reaction mixture changed from clear orange-brown to clear dark brown

within a half hour. A reflux was maintained at72 for 5 1/2 hours along with

continuous stirring and a nitrogen atmosphere.

The reaction flask was then placed in an ice bath until the temperature of the

reaction mixture was approximately 5. Then 40 mL of 95% ethanol,

followed by 30 mL of 3M NaOH (aq) and 40 mL of 30% hydrogen peroxide

were added with continued stirring and a pause of several minutes after

each addition. The temperature rose to20

with some fizzing after the

addition of ethanol. The mixture was allowed to cool 5-10 degrees. When

the NaOH was added, the temperature rose to approximately30 and a

24

white solid formed. The reaction mixture gradually changed to a pale yellow

color after the addition of hydrogen peroxide. It was then allowed to sit

overnight with continued stirring.

The heterogeneous mixture of brown liquid and white solid in the reaction

flask was treated with 4 X 100 mL portions of ethyl ether. The separated

ether fraction was combined with anhydrous magnesium sulfate to remove

dissolved water, filtered, and then reduced in volume to 15 mL. At this point,

a cloudy, pale yellow oil was seen.

The white precipitate which formed when 3 M NaOH (aq) was added to the

reaction mixture became tinged with a pale yellow color after sitting in the

reaction mixture overnight. It was separated from the water fraction by

filtration through filter paper followed by flushing with distilled water. The

flushing removed some of the yellow color from the precipitate. The

precipitate was air dried; the mass was 2.4 grams.

The above reaction, workup, and extraction procedure was repeated three

times. The volume of the yellow oil obtained was 20 mL the second time and

18 mL the third time. Examination of the white precipitate was done just

once in the manner described later in this section.

Flame tests were done on BH3-THF and on the yellow oil to test for the

presence of boron. The flame of BH3-THF was yellow with green edges.

The flame of the yellow oil was predominantly blue with some yellow in it.

The yellow oil produced a burning sensation on the skin. It tested positive

for peroxide using acidified sodium chromate since the color of the chromate

ion changed from yellow to blue when mixed with a sample of the yellow oil.

25

GC analysis of the pale, yellow oil showed it to be a mixture of two primary

products which were in a 5:1 ratio for trial one, 6:1 ratio for trial two, and a 5:1

ratio for trial three. The GC retention times of these two products were,

respectively, 4.2 and 10.6 minutes.

The retention times of the two products by GC-MS analysis (appendix) were

4.2 and 10.5 minutes. The molecular weight of the 4.2 minute product is 100

and the molecular weight of the 10.5 minute product is 172.

TLC of the yellow oil revealed three primary spots. The Rf values were 0.9,

0.7 and 0.2. The colors of the sprayed spots were dark blue, rose, and

brown respectively. The rose-colored spot had a small blue"cap"

on it. TLC

of reagent grade cyclohexanol was done for comparative purposes. A rose-

colored spot with an Rf value of 0.7 was observed. See drawings below.

Rose-

Brown<a <r

fy-o.2.

TLC of Unpurified Ether Extract TLC of Reagent Grade Cyclohexanol

Attempts were made using liquid chromatography to separate the various

products in the yellow oil so that the TLC spots could be associated with

specific peaks seen in the GC analysis and the mass spectra. It was found

that separating the 0.9 spot (blue) from the 0.7 spot (rose) could be achieved

fairly well by eluting the column with pure hexane. The substance which

caused the fastest blue spot on the TLC plate travelled fastest through the

26

liquid silica gel column. When the column was flushed with mixtures of

hexane and pentane in various proportions (4:1 ,3:1

,1 :1 ,

1 :3), small

amounts of the substances which caused the blue and rose spots on the

TLC plates came out together. Significant quantities of the substance

causing the rose-colored spot came out when ethyl ether was added to

pentane in the mobile phase. Mixtures of pentane and ethyl ether in various

proportions (4:1,3:1

, 2:1 ,1 :1

,1 :2, 1 :3, 1 :4) were tried. The substance

causing the rose-colored spot came out easily with even small amounts of

ethyl ether in the mobile phase, but other substance(s) causing the brown

spot on TLC plates started to come off the column with it.

The fractions coming off the liquid column which were relatively free of the

0.7 rose-colored spot were found now to contain two principal substances,

both showing blue spots on the TLC plates: Rf values of 0.8 and 0.7 were

measured. A TLC plate with only one spot was not found. See drawing

below of a fraction free of the rose-colored spot.

Faint Blue.

Dark

OorKinc.

TLC of Hexane-eluted fraction

The fractions showing the two blue spotswere subjected to GC analysis.

Only one primary peak at 10.6minutes was seen in addition to the solvent

peak. No peak at 4.2 minutes appeared. The NMR spectrum of the fractions

showing two blue spots is shown in the appendix.

27

An IR spectrum was obtained using the same sample dissolved in

deuterated chloroform that was used for the NMR spectrum just mentioned.

Strong peaks seen were: 2850-2950 cm-1 (C-H aliphatic stretch); 1450 cm-1,

1375 cm-1

(CH2, CH3 bends); 1075 cm-1 (C-0 stretch). Weak peaks

(possible contamination) seen were: 3625 cm-1 (free OH); 3500-3100 cm-1

(OH).

Fractions from the column which were rich in the substance which produced

the rose-colored spot with the Rf value of 0.7 were also subjected to GC

analysis. A large peak at 4.2 minutes was seen with these fractions.

The fractions which were rich in the substance which produced the rose-

colored spot were dissolved in deuterated chloroform and analyzed by IR.

Peaks seen were: a sharp, narrow peak at 3625 cm-1 (free OH); a broad

peak from 3550 to 3150 cm-1 (OH); 2850-2950 cm-1(C-H aliphatic stretch);

1450 cm-1, 1375 cnr1(CH2, CH3 bends); 1075 cm-1,1140cm-1 (C-0 stretch).

Since the GC-MS library file suggested that the product with the retention

time of 4.2 minutes was cyclohexanol, reagent grade cyclohexanol was

subjected to TLC, GC, IR, and GC-MS analysis on the same instruments

used for analyzing the yellow oil and the chromatographed fractions.

Reagent grade cyclohexanol produced a rose-colored spot with an R{ value

of 0.7 on the TLC plates, a peak at 4.2 minutes on the GC, and and an IR

spectrum virtually identical to the one described in the preceeding

paragraph. The mass spectra of the 4.2 minute product and of reagent

grade cyclohexanol are seen in the appendix and are also virtually identical.

The GC-MS library file had no suggestions as to the identity of the product

with the molecular weight of 172, so no comparative studies with reagent

grade samples on TLC, GC and GC-MS were done.

28

All fractions collected from the liquid chromatography column with hexane,

hexane-pentane mixtures, and pentane-ethyl ether mixtures were clear and

colorless. None of these fractions tested positive for peroxides with acidified

sodium dichromate. The yellow dichromate solution remained yellow in

color when mixed with all of the fractions just mentioned.

The white precipitate produced by the addition of 3 M NaOH (aq) to the

reaction mixture was only very slightly soluble in chloroform and benzene

and insoluble in methanol and DMSO. The IR spectrum showed a broad

peak from 3600-3100 cm-1 (OH). No other identifiable functional peaks were

seen. It had a melting point of68-69

. Immediately after melting, it

appeared to bubble within the melting point tube. When this same

precipitate was heated in an open beaker, a liquid was driven off leaving a

white solid. The melting point of this solid from which the liquid had been

driven off was above 250.

29

Treatment of Cyclohexene Oxide with BoraneTHF Complex in

the Absence of Wilkinson's Catalyst

Ethanol

no catalyst NaOH+ BH3-THF

THF HoO2^2

(2)

A 3-neck round-bottom flask was flushed by a five-minute flow of

deoxygenated and dried nitrogen gas. Then 100 mL (0.1 mole) of 1 M

BH3-THF was injected into the flask using a syringe followed by the addition

of 10.0 g (0.1 mole) of cyclohexene oxide. The reaction mixture immediately

began boiling on its own. A reflux was maintained at72 for 5 1/2 hours

along with continuous stirring and a nitrogen atmosphere. The reaction

mixture remained clear and colorless throughout the reflux.

The reaction flask was then placed in an ice bath until the temperature of

the reaction mixture was approximately 5. Then 40 mL of 95% ethanol,

followed by 30 mL of 3 M NaOH (aq) and 40 mL of 30% hydrogen peroxide

were added with continued stirring and a pause of several minutes after

each addition. The temperature rose to 20 with some fizzing after the

addition of ethanol. The mixture was allowed to cool 5-10 degrees. When

the NaOH was added, the temperature rose to approximately30 and a

white solid formed. The reaction mixture did not change color significantly

after the addition of hydrogen peroxide. It was then allowed to sit overnight

with continued stirring.

30

The heterogeneous mixture of colorless liquid and white solid in the

reaction flask were treated with 4 X 100 mL portions of ethyl ether. The

separated ether fraction was combined with anhydrous magnesium sulfate

to remove dissolved water, filtered, and then reduced in volume to 25 mL. At

this point, a pale yellow oil was observed.

The white precipitate which formed when 3 M NaOH (aq) was added to the

reaction mixture became tinged with a pale yellow color after sitting in the

reaction mixture overnight. It was separated from the water fraction by

filtration through filter paper followed by flushing with distilled water. The

flushing removed some of the yellow color from the precipitate. It was then

air dried. This precipitate melted at68

; it bubbled in the melting point tube

immediately after melting. No further examination of this white precipitate

was done.

The above reaction, workup, and extraction procedure was repeated three

times. The volume of the yellow oil obtained was 20 mL the second time

and 15 mL the third time.

Flame tests were done on BH3-THF and on the yellow oil to test for the

presence of boron. The flame of BH3-THF was yellow with green edges.

The flame of the yellow oil was predominantly blue with some yellow in it.

The yellow oil produced a burning sensation on the skin. It tested positive

for peroxide using acidified sodium chromate since the color of the chromate

ion changed from yellow to blue when mixed with a sample of the yellow oil.

GC analysis of the pale, yellow oil showed it to be a mixture of two primary

products which were in a 1 2:1 ratio for trial one, a 9:1 ratio for trial two, and a

1 2:1 ratio for trial three. The GC retention times of these two products were,

respectively, 4.2 and 10.6 minutes.

31

The retention times of these two products by GC-MS analysis were 4.2 and

10.2 minutes respectively. The molecular weight of the 4.2 minute product is

100 and the molecular weight of the 10.2 minute product is 172. See mass

spectra in the appendix.

TLC of the yellow oil revealed three primary spots. The Rf values were 0.9,

0.7 and 0.2. The colors of the sprayed spots were dark blue, rose, and

brown respectively. The rose-colored spot had a small blue"cap"

on it. TLC

of reagent grade cyclohexanol was done for comparative purposes. A rose-

colored spot with an Rf value of 0.7 was observed. See drawings below.

TLC of Unpurified Ether Extract TLC of Reagent Grade Cyclohexanol

Attempts were made using liquid chromatography to separate the various

products in the yellow oil so that the TLC spots could be associated with

specific peaks seen in the GC analysis and the mass spectra. It was found

that separating the 0.9 spot (blue) from the 0.7 spot (rose) could be achieved

fairly well by eluting the column with pure hexane. The substance which

caused the fastest blue spot on the TLC plate travelled fastest through the

32

liquid silica gel column. When the column was flushed with mixtures of

hexane and pentane in various proportions (4:1 ,3:1,1:1,1 :3), small

amounts of the substances which caused the blue and rose spots on the

TLC plates came out together. Significant quantities of the substance

causing the rose-colored spot came out when ethyl ether was added to

pentane in the mobile phase. Mixtures of pentane and ethyl ether in

various proportions (4:1, 3:1 , 2:1 , 1 :1

,1 :2, 1 :3, 1 :4) were tried. The

substance causing the rose-colored spot came out easily with even small

amounts of ethyl ether in the mobile phase, but other substance(s) causing

the brown spot on TLC plates also came off the column with it.

The fractions coming off the liquid column which were relatively free of the

0.7 rose-colored spot were found now to contain two principal substances,

both showing blue spots on the TLC plates: Rf values of 0.8 and 0.7 were

measured. A TLC plate with only one spot was not found. See drawing

below of a fraction free of the rose-colored spot.

Faint Blux

Darkfeiu.1

OorKBUe.

TLC of Hexane-eluted fraction

The fractions showing the two blue spots were subjected to GC analysis.

Only one primary peak at 10.6 minutes was seen in addition to the solvent

peak. No peak at 4.2 minutes appeared.

33

No NMR or IR spectra of the samples showing a GC peak at 1 0.6 minutes

was obtained because of the difficulty of collecting sufficient quantities of it.

Fractions from the column which were rich in the substance which produced

the rose-colored spot were also subjected to GC analysis. A large peak at

4.2 minutes was seen with these fractions.

The fractions which were rich in the substance which produced the rose-

colored spot were dissolved in deuterated chloroform and then analyzed by

IR. Peaks seen were: a sharp, narrow peak at 3625 cm-1 (free OH); a broad

peak from 3550 to 3150 cm-1 (OH); 2850-2950 cm-1 (C-H aliphatic stretch);

1450 cm-1, 1375cm-1

(CH2> CH3 bends); 1075 cm-1, 1140cm"1 (C-0 stretch).

Since the GC-MS library file suggested that the product with the retention

time of 4.2 minutes was cyclohexanol, reagent grade cyclohexanol was

subjected to TLC, GC, IR, and GC-MS analysis on the same instruments

used for analyzing the yellow oil and the chromatographed fractions.

Reagent grade cyclohexanol produced a rose-colored spot with an Rf value

of 0.7 on the TLC plates, a peak at 4.2 minutes on the GC, and and an IR

spectrum virtually identical to the one described in the preceeding

paragraph. The mass spectra of the 4.2 minute product and of reagent

grade cyclohexanol are seen in the appendix and are also virtually identical.

The GC-MS library file had no suggestions as to the identity of the

substance with the molecular weight of 172, so no comparative studies with

reagent grade samples on TLC, GC, and GC-MS were done.

All fractions collected from the liquid chromatography column with hexane,

hexane-pentane mixtures, and pentane-ethyl ether mixtures were clear and

colorless. None of these fractions tested positive for peroxides with acidified

sodium dichromate since the yellow dichromate solution remained yellow

when mixed with all of the fractions just mentioned.

34

Reaction Study Using TLC of the Treatment of

Cyclohexene Oxide with Borane-THF in the Presence of

Wilkinson's Catalyst

+ BH3THFcatalyst

THF

Ethanol

NaOH

H202

(2)

Samples were removed from the reaction mixture with a capillary

tube at the following intervals: 1, 2, 3, 5, 10, 15, 30, 60, 120,

180, and 240 minutes. The samples were spotted on TLC plates,

developed, sprayed, heated, and examined. Samples of the solvent

and starting materials - THF, borane-THF, cyclohexene oxide

dissolved in THF, and Wilkinson's catalyst dissolved in THF - were

also spotted on the TLC plates, developed, sprayed, heated, and

compared with the collected reaction samples.

Drawings of the TLC plates on which only solvent and starting

materials were spotted are seen below.

35

THF Cyclohexene

Oxide/THF

BlueLint R^t-o

8luc C n<l

bh3/thf Wilkinson's Catalyst/THF

Line

-Blue- Li *e.^='0

"7ellU>A*4L

Spat-

36

Drawings of the TLC plates on which reaction mixture samples at

1, 5, 15, and 30 minutes were spotted are seen below. The TLC

plates of samples collected after 30 minutes were virtually

identical to the 30 minute plate. This TLC reaction study was

done only once.

1 minute 5 minutes

to

Brown.Specks

.Band

FlzjzjmSpot

-DarK Blue.

os-

U2V& Slue.

15 minutes 30 minutes

k

Blue

-ftjS*.

Likt

sr^e.

37

Reaction Study Using TLC of the Treatment of

Cyclohexene Oxide with Borane-THF in the Absence of

Wilkinson's Catalyst

Ethanol

no catalyst NaOH+ BH3THF

THF h,02^2

(2)

Samples were removed from the reaction mixture with a capillary

tube at the following intervals: 1, 2, 3, 6, 10, 15, 30, 60, 120,

180, and 240 minutes. The samples were spotted on TLC plates,

developed, sprayed, heated, and examined. Samples of the solvent

and starting materials - THF, borane-THF, and cyclohexene oxide

dissolved in THF - were also spotted on the TLC plates, developed,

sprayed, heated, and compared with the collected reaction

samples. Drawings of the solvent and starting materials are seen

in the previous section dealing with the TLC reaction study of the

Wilkinson's catalyst reaction.

Drawings of the TLC plates on which reaction mixture samples at

3, 6, 10, 30, and 60 minutes were spotted are seen below. The

TLC plates after 60 minutes were virtually identical to the 60

minute plate.

38

3 minutes 6 minutes

Fuz-zu.Spot?"

DjirlC

BlueR5C

^c+

10 minutes 30 minutes

U <^T

>Tue_

39

60 minutes

The reaction study of the treatment of cyclohexene oxide in the

absence of Wilkinson's catalyst was done twice. The appearance

of the TLC plates was virtually identical for both studies.

40

PART ONE: RESULTS AND DISCUSSION

As early as 1979, the template synthesis and properties of

macrocyclic complexes of La and Ce nitrates from 2,6-diacetylpyridine and

ethylene diamine were reported by Alan Hart and coworkers (27). The

template synthesis and properties of the complexes of several lanthanide

acetates and perchlorates using 2,6-diacetylpyridine and ethylene diamine

was demonstrated by L.M. Vallarino and coworkers in 1985 (8).

We were interested in the template synthesis and possible shift

reagent properties of a complex of praeseodymium acetate using2,6-

diacetylpyridine and orfno-phenylenediamine:

^aNH,

NH,

NaCl

+ Pr(OAc)3CH3OH

2 OAc

CI-

n H20

(D

41

The template synthesis and properties of the complexes of Ce, Pr, and

Nd nitrates from 2,6-diacetylpyridine and orffro-phenylenediamine were

reported in 1985 by Wanda Radecka-Paryzek (28 ). Professor Tom Bell of

SUNY at Stony Brook in 1989 suggested a simple lanthanide complex of the

macrocycle formed from 2,6-diacetylpyridine and orf/70-phenylenediamine

had been claimed, but he was doubtful that any bona fide complexes of this

ligand existed (29).

Since Dr. Robert Bryant from the University of Rochester in the

summer of 1991 (11) suggested that a lanthanide macromolecule made

with four aromatic rings instead of two would make a complex that was more

planar and hence would make a better MRI reagent, the decision was made

to attempt the synthesis of compound 1 using the successful synthetic

procedures employed by Hart and by Vallarino.

The green precipitate obtained from the attempted synthesis was

disappointing with respect to solubility characteristics. Since it precipitated

from the reaction solvent (methanol), NMR analysis could not be done in

methanol-d^ Furthermore, it was only slightly soluble in the polar solvents

water and DMSO, and insoluble in benzene, chloroform, pyridine, ethyl

acetate, carbon disulfide, and acetone as well as methanol. The solubility

problems made its use as a MRI reagent very limited.

Poor solubility of the green precipitate also made determination of

purity using TLC plates practically impossible.

The IR spectrum showed strong OH andCOO"

stretches. There was

an extremely weak peak at 1600cm-1 (C=N stretch) compared to the IR peak

seen on a spectrum for the complex formed using 2,6-diacetylpyridine and

ethylenediamine as done by Eileen George (10). If the synthesis had

produced 1,the C=N peak was expected to be much stronger.

The high melting point and also the negative test for chloride ion were

further indications that the product formed in our attempted template

synthesis was not 1 . This project was set aside in October, 1 992.

42

In the fall of 1992, Dr. Morrill had another opportunity to discuss these

macrocycles withTom Bell (30). Dr. Bell suggested that the properties we

had observed could be associated with the structure in Figure 9, a complex

that had been produced and verified in his laboratories. (Pr3*) was not part

of the complex Bell and his coworkers synthesized, but was probably part of

the complex we had synthesized since praeseodymium imparts a green

color to its compounds.) The poor solubility problems and high melting point

might be explained by the cross bonding between the acetyl groups.

Figure 9.

Before this project had been set aside, we also attempted to repeat

the work of Radecka-Paryzek (28) since she had claimed success in the

synthesis of the praeseodymium nitrate complex formed from 2,6-

diacetylpyridine and orf/70-phenylenediamine:

43

X-CCNH2

+ Pr(N03)3

NH,CH3OH

3

2 H20

We followed the synthetic method briefly described byRadecka-

Paryzek in the literature, which seemed virtually identical to the procedure

followed by Hart and by Vallarino, but our results were disappointing.

According to the Radecka-Paryzek paper published in 1985 (28), a yellow-

brown precipitate was obtained from the reaction above. The IR spectrum of

this precipitate included the following bands: 3450-3350 cm-1 (OH stretch);

1608 cm-1 (C=N stretch); 1595, 1568, 1455 and 1000 cm"1 (pyridine

vibrations); 1465, 1295, 1040, 823 and 743 cnr1 (coordinated nitrate ion

stretch). Her results were also verified by NMR, Mass and UV spectra.

We obtained a yellow-brown precipitate from our three attempted

syntheses of the praeseodymium nitrate macrocyclic complex seen above.

The IR spectrum showed the following absorption bands: 3350 cm-1 (N-H

stretch); a small peak at 3100cm-1 (aromatic C-H stretch); two small peaks

at 2900 and 2950cm-1 (aliphatic C-H stretch); small peak at 1700

cm-1

(C=0 stretch); peaks at 1600 and 1650 cm"1 (C=N stretches); 1495cm-1

(CH2 aliphatic bend).There was no band for the OH stretch. We were not

able to determine whether our synthetic method was exactly the same as

that of Radecka-Paryzek. No current address for her was available.

44

PART TWO: RESULTS AND DISCUSSION

The work described in Part Two of the Experimental Section

developed out of earlier studies done at RIT in which attempts were made to

open cyclopropane rings using two B-H bond sources and in some cases,

Wilkinson's catalyst.

As mentioned in Part Two of the Historical Review, Kevin Gillman (23),

in our laboratories, hydroborated the cyclopropane ring of a strained

polycyclic compound (quadricyclene) with Wilkinson's catalyst:

l.(PPh3)3Rha

Catecholbonne

THF .5*C

2.H202,OH

quadricyclene notricyclanol (9.9%)

Donna Chen (24), also in RIT laboratories, attempted to hydroborate

the less strained norcarane using Wilkinson's catalyst without success. In

addition, a mixture of norcarane and Wilkinson's catalyst showed no NMR

changes after the attempted complexation.

onorcarane

-f BH3:THF

l.(PPh3)3RhCl

Reflux

2.H2Oj.OH

45

These results indicated that most three-membered rings were difficult

to cleave in spite of their pi bond character unless there was extreme ring

strain as in the case of quadricyclene. We thought that a three-membered

ring containing oxygen might be hydroborated more easily because of the

unshared pairs of electrons on the oxygen atom, thus giving the ring more

electron density. We also thought that using an elevated temperature would

facilitate this process.

H.C. Brown and coworkers found that the reduction of epoxides with

borane-THF was slow at room temperatures (31). Addition of trifluoroborane

(32) or a catalytic amount of sodium borohydride (33) to borane-THF,

however, was shown to effect ring opening faster at room temperature. We

simultaneously tested the effectiveness of an elevated temperature (72)

and of Wilkinson's catalyst in the opening of the epoxide ring of cyclohexene

oxide:

bh3-thfcatalyst

Ethanol

NaOH

THF H,0

(2)

+ BH3-THFno catalyst

THF

T->2

Ethanol

NaOH

H,02^2

(2)

If an epoxide ring is opened and an organoborane is formed, the

standard oxidation procedure by alkaline hydrolysis of the alkoxyboranes

which are formed yields alcohols (Part Two: Historical Review).

The treatment of cyclohexene oxide with borane-THF using

Wilkinson's catalyst was repeated three times with consistent results. A pale

yellow oil was obtained from the ether extract. Approximately 60% of the

46

yellow oil was one product, according to GC analysis. The rose-colored TLC

spot with an R, value of 0.7, the IR and mass spectra, and the GC retention

time of 4.2 minutes of this product matched the results found with reagent

grade cyclohexanol. Thus, approximately 60% of the ether-extracted

product mixture from the Wilkinson's catalyst reaction was cyclohexanol.

The treatment of cyclohexene oxide with borane-THF in the absence

of Wilkinson's catalyst was also repeated three times with consistent results.

A pale yellow oil was obtained from the ether extract. Approximately 70% of

the yellow oil was one product, according to GC analysis. The rose-colored

TLC spot with an Rf value of 0.7, the IR and mass spectra, and the GC

retention time of 4.2 minutes of this product also matched the results found

with reagent grade cyclohexanol. Thus, approximately 70% of the ether-

extracted product mixture from the uncatalyzed reaction was cyclohexanol.

A summary of the results of the catalyzed and uncatalyzed reactions

is shown in the Table below:

Catalyst Temp Reflux Time Primary Product % of the

of Ether Extract Ether Extract

YES 72 5.5 HOURS CYCLOHEXANOL 60%

NO 72 5.5 HOURS CYCLOHEXANOL 70%

Since the yield of cyclohexanol was higher when no catalyst was

used, it can be concluded that no catalyst is needed if the temperature is

held at 72 and that the presence of the catalyst may actually interfere

slightly with theproduction of the alcohol.

A suggested mechanism for the synthesis of cyclohexanol by means

of the hydroboration-oxidation of cyclohexene oxide is as follows:

47

2BH,

H20

OH0+BH2

The reaction studies done using TLC, although very qualitative,

indicated that there was an excess of unreacted borane in the reaction

mixture. Borane was the light blue spot (Rf value close to zero) which

appeared on the TLC plates throughout the 5.5 hour reflux. (Ap-

anisaldehyde spray and heat was applied to the TLC plates to impart color

to the spots.) Cyclohexene oxide, which appeared as a fuzzy spot (Rf value

of 0.95), disappeared from the TLC plates within the first half hour. From

this, it can be concluded that cleavage of the epoxide ring was achieved

fairly quickly at 72. In addition, there was no evidence of cyclohexene

oxide in the GC spectrum of the product mixture (cyclohexene oxide

produces a GC peak at 3.9 minutes). TLC plates derived from the

Wilkinson's catalyst reaction showed no change in appearance after 30

minutes. TLC plates derived from the uncatalyzed reaction showed no

change in appearance after 60 minutes. Refer to Part Two of the

Experimental Section for drawings of these changes. Sustaining the reflux

for the full 5.5 hours was probably unnecessary with both reactions.

The treatment of cyclohexene oxide with borane-THF (with and

without Wilkinson's catalyst) produced one other product of interest. We are

48

uncertain of the identity of this product, largely because of our inability to

obtain adequately pure samples of it. According to GC-MS, this product had

a M.W. = 172. It produced one of the fastest travelling blue spots on TLC

plates, had a GC retention time of 10.6 minutes, and showed strong peaks

representing an aliphatic C-H stretch, CH2 and CH3 bends, and a C-0

stretch in the IR spectrum. The IR spectrum suggests an ether. Ethers

which could possibly result from the reactions we tried are trans -1 ,2 -

diethoxycyclohexane, (4), or c/'s-1,2-diethoxycyclohexane, (5), (M.W. = 172).

See Figure 10 below.

cc ccOC2H5

OC2H5

Trans^ ,2-diethoxycyclohexane C/s-1 ,2-diethoxycyclohexane

(4) (5)

Figure 1 0.

A literature reference was found for the trans form, compound 4, of

this particular ether (34 ). No references were found for cis -1,2-diethoxy

cyclohexane in the index of Chemical Abstracts going back to1952. In the

reference for the trans form of the ether, the NMR peaks were listed as

follows: 1.15 (t,6H), 1.5-2.3 (m,8H), 3.15 (m, 2H), 3.55 (q, 4H). The NMR

spectrum was recorded using Varian EM-390 and Varian FT-80A

spectrometers.

The NMR spectrum for our product with a GC peak of 10.6 minutes

shows the following peaks: multiplets at 0.8, 1 .2, 1.6, 2.0, 2.9, 3.15, and 3.5.

This spectrum was recorded using a Bruker 200 MHz spectrometer.

Although there is agreement of this spectrumwith some of the NMR peaks

49

associated with pure trans -1,2-diethoxycyclohexane,

there are additional

peaks, not related to the trans diether, and these indicate a mixture.

This same sample, which had been eluted with pure hexane from the

LC column, had also shown two significant blue spots on the TLC plate, so

the inconclusive results of the NMR spectrum were not surprising.

Trans -1 ,2-diethoxycyclohexane is not commercially available. We

have not synthesized it from the frans-1,2-cyclohexanediol, so comparative

studies with our reaction product were not done.

We suggest the following mechanism for the synthesis of 4 from the

hydroboration-oxidation of cyclohexene oxide:

O BH,

O BH,

The white precipitate which formed when 3 M NaOH (aq) was added

to the reaction mixture had a melting point of68-69 and showed a broad

peak representing the OH functional group on the IR spectrum. The

50

recorded melting point of Na2B4O7-10H2O, sodium tetraborate

decahydrate, is 60. The recorded melting point of Na2B407-8H20, sodium

tetraborate octahydrate, is 75. The white precipitate produced by the

reactions in this thesis could be a impure mixture of these two borate

compounds and/or other borates or organoborates. Both the white

precipitate and the borate compounds just mentioned have an Rf value of

0.0 using the same stationary and mobile phases that were used for all other

TLC tests performed. The white solid and the borate compounds mentioned

above would not burn when subjected to a Fisher burner flame. The

production of a borate compound by a hydroboration-oxidation reaction is

common.

Approximately 25% of the GC peak areas from the ether extracted

products resulting from these reactions had GC retention times of three

minutes or less. No special effort was made to identify these short retention

time compounds.

The most interesting part of this research was trying to identify the

product with the molecular weight of 172, which we referred to as our

"mysterycompound."

It was noted earlier that a higher percentage of

cyclohexanol occurred in the ether extract when no Wilkinson's catalyst was

used. It should also be noted that a higher percentage of the "mystery

compound"

occurred in the ether extract when Wilkinson's catalyst was

used. Since the quantities of the yellow oil extracted by ether were

approximately the same for the catalyzed and the uncatalyzed reactions, this

means that more of the "mysterycompound"

was obtained in terms of

quantity, not only in terms of percentage of the ether extract, when

Wilkinsons catalyst was present.

We suggest that if research continues on the hydroboration of

cyclohexene oxide, larger quantities ofWilkinson's catalyst should be tried

to see if this would further improve the yield of the "mysterycompound."

Another suggestion stems from the unavailability of a commercial

51

source of compound 4. Attempts should be made to synthesize 4 in our

laboratories by using the commercially available trans-1 ,2-cyclohexanediol.

Once 4 is synthesized, GC, TLC, and GC-MS comparison studies can be

made with the hexane-eluted fractions from the LC column.

More effort also needs to be expended in the area of finding optimal

temperature and time parameters for this reaction.

In summary, the following significant information was learned from the

experimental work described in Part Two: (1) the epoxide ring of

cyclohexene oxide was completely cleaved by borane-THF at 72 in less

than 0.5 hours; (2) at 72, Wilkinson's catalyst was not essential to this

cleavage; (3) TLC plates showed that there was no cyclohexene oxide

remaining in the reaction mixture after 0.5 hours of reflux, indicating that the

extended 5.5 hour reflux time was unnecessary; (4) we observed

cyclohexanol as the primary product found in the ether-extracted product

mixture; (5) approximately 70% of the ether-extracted product mixture was

cyclohexanol in the case of the uncatalyzed reaction, while approximately

60% of the mixture was cyclohexanol in the case of the catalyzed reaction,

suggesting that the catalyst may have interfered slightly with alcohol

formation; and (6) there was some evidence that another minor product

formed was frans-1,2-diethoxycyclohexane.

52

(1

(2

(3

(4

(5

(6

(7

(8

(9

REFERENCES

Hinckley, C.C. J. Amer. Chem. Soc. 1 969, 91, 5160-5162.

Sanders, Jeremy, K.M.; Williams, D.H. J. Amer. Chem. Soc. 1 971, 93,641-645.

Sievers, R.E., Ed. "Nuclear Magnetic Resonance Shift Reagents";Academic: New York, 1973; pp. 1-2, 21-22.

Morrill, T. C. "Lanthanide Shift Reagents in Stereochemical Analysis";VCH: 1986; pp.1 59-1 60.

DeCann, Dale. MS Thesis, Rochester Institute of Technology, June,1979.

Hiroshi, Tsukube. J. Chem. Soc. Chem. Commun. 1984, 315-316.

Arif, A.M.; Gray, C.J.; Hart, F.A.; Hursthouse, M.B. J. Chem. Soc. Dalton

Trans. 19 87, 1665-1673.

DeCola, L; Smailes, D.L.; Vallarino, L.M. Inorg. Chem. 1986, 25, 1729-

1732.

DiSano, Mary. MS Thesis; Rochester Institute of Technology, Rochester,

New York, June, 1989.

(10) George, E.S. Spring Research 1989 Report; Rochester Institute of

Technology, Rochester, New York, 1989.

(11) Bryant, Robert. University of Rochester, private communication, 1991.

(12) Carey, F.A. "Organic Chemistry"; McGraw-Hill, Inc: New York, 1992; pp.

225-226.

(13) Brown, H.C. "Hydroboration"; W.A. Benjamin, Inc: New York, 1962.

53

(14) Vollhardt, K.P.C. "Organic Chemistry"; W.H. Freeman: New York, 1987;p. 478.

(15) Noland, W.E., Editor-in-Chief. "Organic Syntheses, Collective Volume6"; John Wiley & Sons: New York; pp. 923-924.

(16) Mannig, D.; Noth, H. Angew. Chem. Int. Ed. Engl. 1985, 24, 878-879.

(17) Evans, D.A.; Fu, G.C. J. Org. Chem. 1990, 55, 2280.

(18) Harmon, R.E.; Parsons, J.L.; Cooke, D.W.; Gupta, S.K.; Schoolenberg,J. J. Org. Chem. 1969, 34, 3684.

(19) Rylander, P.N. "Catalytic Hydrogenation Over Platinum Metals";Academic Press: New York, 1967.

(20) Walsh, A.D. Trans. Faraday Soc. 1949, 45, 179.

(21) Gunthard, Hs.H.; Lord, R.C.; McCubbin, F.K. J. Chem. Phys., 1956, 25,768.

(22) Lowry, T.H. "Mechanism and Theory in Organic Chemistry, Third

Edition"; Harper & Row Publishers, Inc.: New York, 1987, pp. 31-33.

(23) Gillman, Kevin. MS Thesis, Rochester Institute of Technology,

Rochester, New York, August, 1991.

(24) Morrill, T.C.; Gillman, K.; Feng, P.; Chen, D. Paper presented at

National ACS Meeting, San Francisco, California, April, 1992.

Abstract No. 289: "Hydroboration of Highly Strained Cyclopropane

Rings Promoted by WilkinsonsCatalyst."

(25) Fieser, L.F.; Fieser, M. "Reagents for Organic Reactions"; John Wiley &

Sons: New York, 1967; p. 393.

(26) Gordon, A.J. ; Ford, R.A. 'The Chemist's Companion"; John Wiley &

Sons; New York, 1972; p. 379.

(27) Backer-Dirks, J.D.J. ; Gray, C.J.; Hart, F.A.; Hursthouse, M.B.; Schoop,

B.C. Chem Commun. 1979, 774.

54

(28) Radecka-Paryzek, W. Inorg. Chim. Acta A 985, 109, L21.

(29) Bell, Tom. SUNY at Stony Brook, Letter, July, 1989.

(30) Bell, Tom. SUNY at Stony Brook, private communication, Fall, 1992.

(31) Brown, H.C.; Heim, P.; Yoon, N.M. J. Am. Chem. Soc. 1 970, 92, 1637.

(32) Brown, H.C.; Yoon, N.M. Chem. Commun. 1968, 1549.

(33) Brown, H.C.; Yoon, N.M. J. Am Chem. Soc. 1968, 90, 2686.

(34) Barluenga, J.; Alfonso-Cires, L.; Campos, P.J.; Asensio, G.

Tetrahedron 1984, 40, 2563-2568.

55

APPENDIX

56

<SBwple> <Co(wient>

Sntft Reaoent

6.302 mg( 6.302 dq)

<Reference>

92/07/22 12-47 Platinum

0.000 no <Sampllno>

25. , .1 0 sec

TG/DTA<Name>

Ruth

<Date>

c

o1-

o

-2.5

<Temp prooranlC) lC/<tn ) lnml>

1* 25.0-300 0 5 00 0.00

<6as>

nltrooen 100 0 nl/nln

q 0 ni /min

100.5

-7.5L

25 95

Rochester Institute of Technoloov

165 235

TEMP C (Heating)

TGA ANALYSIS OF PRODUCT FROM ATTEMPTED SYNTHESIS OF

PRAESEODYMIUM ACETATE MACROCYCLIC COMPLEX

IT to TYPE KIOTM *R(*I|

t.144 tu rv .It! .Ml?*

t.l4 ? vv .Ill .11743

t.tll 11*1 VP It* .11*11

2.2*7 SI229M p 41 41.42711

t.47* 1*3*2 IP ,122 .11492

t.891 *sr4i rv ,114.14*14

t.*IS 7*71 VI ,121.ItiSt

i.i4i 1*91 PV ,111 .11171

1.111 11*1 VV .111 .11919

1.1*9 sis vv .129 .11299

1.11* 78i ri .119 .l**27

.*? 11*7 ip .172 .129*1

!. *47*2I PI ,177 l.ll71

TL **E*1.I74I*I7

facto*.1.IIIIE*!!

,^

MT>2

VJaI

GC Spectrum of Cyclohexene Oxide(Dissolved In Deuterated Chloroform)

Ota*

*T ARE* TYPE yiDTH AMAX

2.124 12274272 PV .131S7.21**

2. If* 94.81* VI .21 4. 1**43

2.1*4 2*741 IV .44 .**

l.*S3 729 VI .27 .337

l.*3 17**2 PP .72 .343

4.272 114239 PI .1723S.*2*74

TOTAL AKEA-2. 132SE*7

MUL FACT0*1.?*

yS

Sov^

*yr

LtfcloUunoL-}

GC Spectrum of Cyclohexanol

(Dissolved in Deuterated Chloroform)

41141

T

1.44*

.4

l.tlj

I.I

1.114

.74

1.111

1.474

t.lSt

l.tll

>.44l

a. u*

1.111

J*

14444

*

1411

74

1*14

4411411

4ttJ

11154

urn

14444

14*1

1444

45

tan

4714111

lllllii

TTM IBTM

.1*4

v.l

v .It

7 .a*

Po.it

vv.41

.11

.14

.>

.4t

.til

vy .44

?> .tit

4.tat

rt .441

*T4, I.ttS4ttt7l '4CI01.|.t444t.tt

J>*

J>

1141

.11744

111

a*

.14447

.41141

I.iaaat

.77l|

.11711

.17111

.14741

tllll

tl7lt

.14

.till!

GC Spectrum of Mixture: Cyclohexene Oxide and Cyclohexanol

(Dissolved In Deuterated Chloroform)

.iLu

AREA*

ST AREA TYPE UIOTH AREA):

2.042 393600 PB .024 14 .22772

2. 13* 143 *P .006 .00343

2.24* 315* PP .140 . 12338

2.348 72964 PV .024 1 .74884

2.424 383680 ve .029 9 . 19624

3.268 1003 pp .044 .82404

3. 867

^"TT25B

72

2544i'02~

pp

pe

.046 .82898

...87*

-*P588293>

3.712 2393 IV .040 .06220

3.773 3023 VB .067 12044

6.23* 1707 *p. 101 04891

6.941 12109 PB .043 29823

7.818 1404 PO .296 83365

9.336 2124 vv . 112 03891

10.220 336 PV .017 eeeos

^8.536 343230

'4172138

VI . 133 13 06835_}

TOTAL AREA-

IUL FACTOR'l ,0**E4 0

*

GC SPECTRUM OF ETHER EXTRACT FROM CATALYZED REACTION

flREflM

IT n*Eft TYPE HtOTH area*

2.*27 1711077 PI .123 11.72131

.22* (221 IP .31 .1111

1.131 1131*11 PV .2( 3.32239

2.123 1*23231 VI .22 3.1*7*8

2.731 *33 IP .11 .89320

2.1*1 lit PI .111 .08*1

2.3*3 7*33 IV .11( .133*9

3.(71 19*7 VP .111 .1232*

1.1*8 (311 pp .171 .9321*

CTTZ i37i3*ee re .229 *.9i.iiT~;3.73* 12*1 V .33 .88932

3.113 11721 VI .171 .8392I

.33 379*7 PI .13 .29317

*.*** 1171 PV .1(2 ees9i

10.210 3331 pp .It .92(9*

CIS. 63* 11257*8 PI .11* s.t**i>

TOTftL ft*Ert-)l.*7(2E07

MUL FftCTO-l.l**OE**0

GC SPECTRUM OF ETHER EXTRACT FROM UNCATALYZED REACTION

DC

O

O

CX

OJ

X

o

Q

Wt-

<C

a

E-

w

Q

&u

O

z:

L>

EC

a,co

J

o

z<

X

oJ Zu cej ob Cb

o

2 Ow a

O X

< uw

a aw

Q HW <

CM DC>4 W

J HC 3

Z tij

<: a

i

ce z

< Q

CQ W

>

O OCO

Z CO

:o -

ce at-

uu

Cu

CO

a

Q

Oce

CuZ

ce>4 ooe fc.

< o

z ce?1 O

ce J

Cu x

uQ

u a

O H<c <

ce ceE- W

1 u

ce a

u

x z

H -

w

a

o >i-i

z o

3 CO

ce co

H -

u au

Cl,CO

XE-

Socogce wS

o z

&=

o

Cd

Cd

I Cu

Cd"

X U2 _.

CdCd~

z^

^ COoet-

CJCd

a,

CO

ce

co

B.BE5u

4. BE3-

2-BCSd

B

fcckn 2B Ci.779 ntn) of OftTR: MIXTURE . 0

%7 39 B3

I'l 4'

SI

X

SB 4B SB

a? =769

'/ /I \ I

"

1 1 1 i1 1 1| 1|

79 97

/

JU-

BB 7B

Hitt/Ch*PBBB 9B

TIC of ORTRsmJ fXTURL. 0

(J

c

D

S.BCB -i

4.BC6-;3.BEB-

|C

38. BEE:

a. X.BEB:

B- Jl 1 JTim* (nln.)

GC-MS of Cyclohexene Oxide

u

c

II

B

C

3

cr

Scan 313 C4.B32 111

7

4-r

of ORTR: MIXTURE. 0

B.BC3-

4.BES"

E. BE3-

Bji

9

/44

\

J.lll

34

I ,, \

B2

IBB

\1

3B 4B SB BB 7B BB 9B IBB

TIC of ORTR: MIXTURE. O

u

c

m

c

a

a

a.

S. BEB

4.BCB

3. BEB

C.BE6

I. BEB

B JLTi *> tun. )

GC-MS of Cyclohexanol

10000-j90Q&1

eooo]7000J 2?

6000-j>

S32ef 41

4S00|/

3000J20001

leaolj |0jl 1 . .Ill

Scan 338 (4.302 min) of DATFhNOWCK. D

5?

.u.

67

\ 7.

/

62/

:103j

9000

'8000

;7000

6000

"5000

[4000

[3000

100 ;2000

N^-1000

^0

MASS SPECTRUM OF PRIMARY PRODUCT IN ETHER EXTRACT

Scr 356 (4.456 -in) of DfiTfi:CYCHEXOL. DIO000 /

57

: I0CD[

9000 90on

8000 8000

7000 27 7000

8000 / BOOQ

5000

4000? 67

V /

5DDD

4000

3000 \ 71 30DO

zcoo / 100 :20QO

IOQQIi'

.11 1 , .iil! 1 ...II. . . 1 ,.\ [IBQO

MASS SPECTRUM OF BAKER-ANALYZED REAGENT CYCLOHEXANOL

in 537 (4.S&S nir.) of DR7R: NOWCK.O

B.Brs:

c B.BCSit

4. BE3:c

I t.BES:

B

B744

\34

l ,.,li.ill. ^a

te

B7

/

-..,, SI. ^

3B 4B SB 6B 7B SB BB IBB

Matt /Chargt

vS.BEB'

4. BEB

J C-BEB

C 1.BE6

Ti vie Coin.)

GC-MS OF PRIMARY PRODUCT IN ETHER EXTRACT

J 3.0C5

1.0E5

I.OC5"

JJ

C

B.OE6

r 5. 0E6

c 4.1E6

* 3.0E6

I a.ncs

c I. DEB

SO 100 1&J3

TIC of OflrRiNOpJCK. 0

k.

4 & a

Ti < (PMPl.j

4

10 12- t

GC-MS OF PRODUCT (M.W.=172) IN ETHER EXTRACT

O a

\ o o^" * *

^*

so Si ?">>ru ommtnv

8*010 >

f

*p eeeeeesaao<m<o <%

o*b

*

r in *em rm n *

xuff in

S3

^3SS feSSSSSS ssssss 3& ssssrssls