UVA Chemical Filters:

A Systematic Study

Jacqueline F. Cawthray, B. Science (Hons)

A thesis submitted for the degree of

Doctor of Philosophy

in

The University of Adelaide

Department of Chemistry

February 2009

Chapter 5

151

5 Cyclodextrin Complexation Studies

5.1 Introduction

Cyclodextrins (CDs) are cyclic oligosaccharides consisting of a number of D-

glucopyranose units that are linked by α-1,4 glycosidic bonds to form a well-defined cavity

(FIGURE 5.1). The naturally occurring CDs are α-, β- and γCD, and are composed of 6, 7

and 8 D-glucopyranose units, respectively. The CD molecule can be depicted as a

truncated cone having a relatively rigid structure. The inner cavity is lined with hydrogens

and ether-like glycosidic oxygens, making the interior of the annulus relatively

hydrophobic. The secondary hydroxyl groups are located at the wider end of the cone with

the primary hydroxyl groups at the narrow end, making the exterior hydrophilic.

O

HO

OH

O

O

OHOH

O

OOH

OH

O

OH

HO

O O

O

OHOH

O

OOH

OH

OH

OH

HO

OH

OH

OH

n

4

65

32

1

O

FIGURE 5.1: Structures of naturally occurring cyclodextrins.

Cyclodextrins are of interest due to their ability to include either all or a substantial part of

a guest molecule inside their annuli, to form inclusion complexes, otherwise known as

Primary hydroxyl rim

Secondary hydroxyl rim

n = 1 α-CD n = 2 β-CD n = 3 γ-CD

Chapter 5

152

host-guest complexes [375,376]. The inclusion complex is held together by the spatial

entrapment of a guest molecule in the CD cavity without the formation of covalent bonds.

The formation of such inclusion complexes is a reversible and dynamic process in which

free guest molecules are in thermodynamic equilibrium with included guest molecules.

There are a number of energetically favourable interactions that describe the driving force

for inclusion of a guest molecule (for reviews see Refs. [377,378]). In aqueous solution,

the hydrophobic annulus of the CD molecule is occupied by water molecules, which can be

readily replaced by an appropriate non-polar guest molecule (FIGURE 5.2). The

hydrophobic effect, therefore, is one of the main driving forces involved in complex

formation [379,380]. The presence of a guest molecule gives rise to a net gain in enthalpy

due to exclusion of the cavity-bound water and an increase in hydrophobic interactions.

Once inside the cyclodextrin cavity, the guest molecule undergoes conformational

adjustments to maximise the weak van der Waals interactions with the CD. The stability

of the inclusion complex is also influenced by steric interactions with the size of the guest

molecule relative to that of CD cavity being a determining factor [381]. Although the

inclusion complexes formed are held together by secondary bonding forces only, their

stability can be as high as 105 mol dm-3 [382].

X

Y

X

Y

+

FIGURE 5.2: Schematic representation of inclusion of an aromatic guest molecule within

the CD cavity in an aqueous environment; water is represented by the small circles.

The physicochemical properties and complexing ability of CDs can be improved by

chemical modification of naturally occurring CDs (for review see Ref. [383]). Many CD

derivatives exist in which the primary and/or secondary hydroxyl groups of the naturally

occurring CD are substituted with various functional groups to give modified CDs that

have slightly altered properties from those of the parent CD. For example, modified βCDs

such as 2-hydroxypropyl-β-cyclodextrin (HPβCD) and randomly methylated βCD in which

one or several OH groups have been replaced by the relevant alkyl substituent, have a

much higher solubility than native βCD [384]. This is due primarily to disruption of the

Chapter 5

153

intramolecular hydrogen bonding between O2-H and O3-H that, in substituted CDs, is

replaced with intermolecular hydrogen bonding with the solvent in the modified CDs.

Depending on the method of preparation, the hydroxypropyl groups are often randomly

substituted onto the hydroxyl groups of βCD shown in FIGURE 5.3. Consequently,

HPβCD and other substituted CDs are characterised by the degree of substitution, which

refers to the average number of substituents per CD molecule. The degree of substitution

and substitution pattern influences both the solubility and inclusion complex forming

ability of HPβCD in addition to the properties of the inclusion complex formed

[381,385,386].

R Solubility (mg/ml)

βCD H 18.5

HPβCD CH2CH(OH)CH3 >600

FIGURE 5.3: Structures and solubilities of βCD and HPβCD.

The formation of CD inclusion complexes can result in the advantageous modification of

certain physicochemical properties of the guest molecule including improved water

solubility and chemical stability [387]. The low toxicity of CD and its ability to act as an

excipient has meant many natural and substituted CDs have been approved by regulatory

authorities worldwide. Consequently, CDs are used in a wide range of applications in the

pharmaceutical [387-389] food, cosmetic [390,391] and chemical industries.

O

H

H

RO

H

OOR

HH

OR

7

Chapter 5

154

5.1.1 Use of Cyclodextrins with Sunscreens

It is now clear that wavelengths in the UVA region (320 – 400 nm) of the solar spectrum

can cause a wide range of detrimental biological effects [22,84,122,161]. Chemical

sunscreen filters are a popular and effective method of photoprotection against UVA [257].

An important characteristic of any effective UV chemical filter is photostability.

Photodegradation of the sunscreen filter leads to a permanent loss of protection and, in

addition, the degradation products have the potential to cause toxic or allergic reactions.

The UVA chemical filter, 4-tert-butyl-4′-methoxy dibenzoylmethane (BMDBM), is used

worldwide in sunscreen formulations [203]. However, several studies have demonstrated

that BMDBM undergoes UV-induced photodegradation leading to loss of protection

[205,225,258,392]. The degradation products of BMDBM are potentially harmful, causing

damage to relevant biological molecules [333,393,394]. As with other β-diketones,

BMDBM exists in a keto-enol tautomeric equilibrium (FIGURE 5.4) with the enol form

strongly favoured. The enol tautomer, stabilised by an intramolecular hydrogen bond,

absorbs strongly in the UVA region (λmax ~359 nm) whilst the keto absorbs in the UVB

region (λmax ~270 nm). Following UV-irradiation, ketonisation and subsequent

photodegradation of the keto tautomer leads to a permanent loss of absorbance in the UVA

region, reducing the effectiveness of the sunscreen [205,225].

FIGURE 5.4: Keto-enol equilibrium in 4-tert-butyl-4′-methoxydibenzoylmethane

(BMDBM).

Various methods have been used to prevent or minimise the photodecomposition of

BMDBM including the complexation of BMDBM with natural or modified CDs has on the

photostability of BMDBM. Of the naturally occurring CDs, βCD forms the most stable

inclusion complexes [213]. The smaller cavity of αCD can result in steric interference

O O O O

O O

Enol Keto

hv

H

Chapter 5

155

preventing stable inclusion complexes forming whilst the larger annulus of γCD generally

does not allow for optimal interactions between the guest and host. The photostability of

BMDBM is enhanced by inclusion in HPβCD [371,395] and randomly methylated βCD

[213] and forms the claim of at least one patent [396]. However, the stabilising effect of

HPβCD is reduced in lotion vehicle (oil-in-water emulsions) due to competitive

displacement of BMDBM from the CD cavity by emulsion excipients. This can by

minimised by incorporation of the inclusion complex of BMDBM and HPβCD into lipid

microparticles (lipospheres) [397]. In contrast, inclusion complexes of BMDBM and βCD

were found to have very little effect on photostability of BMDBM [289]. However it has

been demonstrated that the photodegradation products are themselves included in the CD

annulus, therefore limiting any potential toxic or allergic reactions [289,395].

The aqueous solubility of BMDBM (1.5 µg/ml [213]) is enhanced through inclusion by

natural and modified CDs. Water-soluble sunscreens are useful in cosmetic formulations

and hair care preparations [203]. Solubility studies of BMDBM and parent CDs (α-,β-,

γCD) and their derivatives (HPα-, HPβ-, HPγCD, randomly methyl βCD) show that

HPβCD and randomly methyl βCD are the most effective at increasing the aqueous

solubility of BMDBM [213,371]. The increased solubility of the inclusion complexes

formed with HPβCD over those formed with βCD can be attributed to the greater solubility

of HPβCD itself.

There are a number of published reports focused on the complexation of BMDBM with

CDs, however the focus has been on determining the influence of CD complexation with

BMDBM has on photostability, solubility and transdermal penetration

[213,289,371,395,397,398]. The methods used in previous investigations have

demonstrated the ability of CDs to form inclusion complexes with BMDBM, mostly in the

solid state, but there is little available information regarding the nature of the inclusion

complexes formed, particularly in solution. The objective of the present study is to gain

further insight into the mode of inclusion in the complexes formed in solution between

BMDBM and the cyclodextrins βCD and HPβCD.

The 1D and 2D 1H NMR techniques are widely used for providing evidence for inclusion

complexes in solution and for studying the mode of inclusion between the guest and host

CD (for review see Ref. [399]). As both techniques are particularly sensitive to changes in

the electron environment of a proton that occurs upon close contact or short-range

Chapter 5

156

association, they are particularly suitable for investigating the non-covalent interactions

present in CD inclusion complexes. When either part or all of a guest molecule is enclosed

in the CD cavity, the resonances of the CD interior protons (H3 and H5) are shifted in the

spectrum while the exterior protons (H2 and H4) remain relatively unaffected (FIGURE

5.5). In a similar manner, the resonances of the guest molecule also experiences

complexation-induced chemical shifts. The formation of inclusion complexes is a dynamic

process involving the CD moving on and off the guest molecule. Consequently, the

observed chemical shift at a given temperature in the NMR experiment is dependent on the

rate of chemical exchange. If the exchange is rapid on the NMR timescale, a single

resonance is observed whose chemical shifts is the weight average of the chemical shifts of

the individual states. Conversely, if the exchange is slow on the NMR timescale, separate

resonances for each state is observed. Intermediate rates show broad or partially averaged

resonances. Since the rate of exchange depends on the change in Gibb’s free energy (∆G),

temperature-dependent 1H NMR studies can be used to investigate the exchange process.

FIGURE 5.5: β-cyclodextrin interior and exterior protons.

The 2D NMR technique, 1H-1H ROESY (Rotating frame nuclear Overhauser Effect

SpectroscopY), is a spin-lock technique that identifies through-space interactions (NOEs)

via spin-spin relaxation. In the 1H ROESY NMR spectra, cross-peaks arise from spatially

separated protons that are ≤ 4 Å apart even in the absence of covalent bonding. Therefore, 1H ROESY NMR provides structural information regarding the functional groups of the

guest that are included within the CD annulus and how deep in the annulus the guest sits.

The intensity of the observed NOE between spatially separated protons is related to both

separation distance and concentration.

O

H

H

HO

H

OOH

HH

(H6)(H2)

(H1)

(H5)

(H3)

(H4)

OH

7

H5

H6 H6

H4

H2

H1

H3

Chapter 5

157

β-Cyclodextrin and its hydroxyalkyl derivative, HPβCD were chosen as they are

particularly suitable as the host CDs in this study as their cavity sizes and volumes are well

suited for inclusion of the hydrophobic tert-butyl and phenyl groups of BMDBM (FIGURE

5.6). The cavity volume of βCD is somewhat extended in HPβCD due to substitution by

the hydroxyalkyl group. The degree of substitution of HPβCD used in this study is 3.0-8.0

which gives an average number of 5 hydroxypropyl groups per βCD molecule. Both βCD

and HPβCD have been the subject of extensive toxicological studies and are considered

safe [400]. Both are approved for use by the Therapeutics Goods Association in Australia

making them suitable for pharmaceutical applications. They are particularly suited for

incorporation into sunscreen formulations as their large molecular size and hydrophobic

nature prevents penetration into the skin [401].

FIGURE 5.6: Dimensions of β-cyclodextrin (βCD) and hydroxypropyl-β-cyclodextrin

(HPβCD) compared with the dimensions of the phenyl group in BMDBM.

5.2 1H NMR Studies of βCD and HPβCD Complexes

The 1H and 1H ROESY NMR spectra are reported for solutions of BMDBM and either

βCD or HPβCD in which the mole ratio of BMDBM to CD is varied. A variable-

temperature 1H NMR study of the solution of BMDBM and either excess βCD or HPβCD

are also presented.

15.4 Å

6.0-6.5 Å

Cavity volume 262 Å3

7.9 Å

4.1 Å

R

HH

H H

5.0 Å

R = OCH3 or C(CH3)9

Chapter 5

158

5.2.1 βCD Complexes/NMR data of BMDBM-.

The guest molecule, BMDBM, and its βCD inclusion complexes are insufficiently water

soluble for NMR studies. Previous studies of BMDBM inclusion complexes with CDs

have either relied upon more sensitive techniques for analysis of BMDBM concentrations

such as HPLC or examined complexation in solids or suspensions [213,371,397,398]. To

achieve the concentrations required to obtain an 1H NMR spectrum, it was necessary to

prepare all solutions in 0.1 mol dm-3 NaOD/D2O such that pD ≈ 12. As the pKa of βCD is

12.20 [402], under the basic conditions employed the βCD hydroxyl groups, OH(2) and

OH(3), are partially deprotonated. This results in an increase in the solubility of βCD. The

pKa of BMDBM is 11.34 (this work; Chapter 2.5) therefore under these conditions, it will

exist predominately as the β-diketonate of BMDBM (herein called BMDBM−) (FIGURE

5.7) having only slightly improved water solubility over the parent species.

FIGURE 5.7: Generation of the β-diketonate of BMDBM (BMDBM −). Only one of the

chelated enols are shown.

Solutions were prepared by adding BMDBM to a solution of βCD in 0.1 mol dm-3

NaOD/D2O. The resulting suspension was gently heated then filtered to remove the

suspended material, which was probably BMDBM and/or its βCD complex. As a

consequence of the poor solubility of BMDBM, even under the basic condition used, the

calculated concentrations of BMDBM− were not achieved. Instead, the mole ratios of βCD

to BMDBM were determined by integration of the appropriate 1H NMR signals. In the

following text the numbering scheme shown in FIGURE 5.8, whereby the CD protons are

denoted as H1-6 and the aromatic protons of BMDBM as Ha-Hd, have been used for the

sake of clarity.

O O

O

+ OHO O

O- OH

+ H2O

BMDBM BMDBM�

H

Chapter 5

159

To enable a comparison between free and included BMDBM− the 1H NMR spectrum of

BMDBM− in 0.1 mol dm-3 NaOD/D2O was obtained (FIGURE 5.10a). To achieve the

concentration required to obtain a 1H NMR spectrum, the solution was rapidly heated to

high temperatures and the suspended material was removed by filtration. The 1H NMR

spectrum indicates that the solution consisted of a mixture of BMDBM− and degradation

products. It was possible to assign all resonances to either BMDBM− or to degradation

product by integration of the 1H NMR signals. The tert-butyl resonance appears as two

singlets, assigned as t-bu and t-bu in FIGURE 5.10a. The BMDBM− aromatic resonances

as two sets of doublets assigned as Ha-Hd and Haʹ-Hd . The methoxy resonance is

observed as two singlets appearing at δ 3.87 (assigned to degradation products) and δ 3.84.

(A comparison between the methoxy resonance of free and included BMDBM− is not

possible as it is masked by the βCD resonances and, consequently, has not been included in

the 1H NMR data presented).

FIGURE 5.8: Labelling scheme used for BMDBM − and βCD protons.

The degradation products are likely to be the result of cleavage at or near the centre of

BMDBM resulting in two fragments. There are a number of possible ways this could

occur, one such possibility is illustrated in FIGURE 5.9. Integration of the peaks assigned

to the fragments indicates that the area ratios of each fragment are significantly different.

As Fragment A, containing the hydrophobic tert-butyl group, would be somewhat less

soluble than Fragment B it is likely to have been filtered from solution. As the aim was to

obtain a 1H NMR spectrum of BMDBM− for comparison purposes, it was not deemed

necessary to investigate the degradation products further. The rapid heating to high

temperatures was not necessary when preparing solutions containing βCD and BMDBM−,

consequently there were no indications that degradation had occurred in these solutions.

O O

OHa

Hb HcHd

H5

H6 H6

H4

H2

H1

H3

Chapter 5

160

FIGURE 5.9: Two possible positions where bond cleavage of BMDBM − can occur

leading to different fragmentation patterns.

5.2.1.1 NMR data of 1 : 2 mole ratio of BMDBM− : βCD.

In the 1H NMR spectrum of a solution in which the mole ratio (determined by integration

of the 1H NMR signals) of BMDBM− and βCD was 1 : 2 (FIGURE 5.10b) the tert-butyl

resonance appears as a singlet and the BMDBM− aromatic resonances Ha-Hd as doublets

arising from AABBʹ spin-spin splitting. Unambiguous assignment of the methoxy

resonance was not possible as the βCD H3 resonance appears in the same region of the

spectrum. The tert-butyl and Ha-Hd resonances of BMDBM− appear as a sharp singlet and

well-resolved doublets respectively. The differences between the chemical shifts of free

and included BMDBM− (TABLE 5.1) are greatest for BMDBM− tert-butyl, Ha and Hb

protons. These observations are consistent with the formation of βCD·BMDBM− as either

a single includomer or two includomers in fast exchange. In processes involving fast

exchange, the observed resonances consist of the time-averaged resonances of free βCD,

BMDBM− and the inclusion complexes formed. The inclusion complexes differ in the

inclusion orientation of βCD with respect to BMDBM−, all being in thermodynamic

equilibrium. The nature of the inclusion complex formed between βCD and BMDBM− are

explored further by 1H ROESY NMR spectroscopy as detailed below.

OHa′

Hb′ Hc′Hd′

tBu′

O O

Fragment A Fragment B

OHa′

Hb′ Hc′Hd′

tBu′

O O

Fragment A′ Fragment B′

Chapter 5

161

TABLE 5.1: 1H NMR chemical shifts (ppm) corresponding to BMDBM in the absence and

presence of βCD in a 1 : 2 mole ratio in 0.1 mol dm-3 NaOD/D2O at 298 K.

δfree δcomplex (∆λa)

t-Bu 1.28 1.42 (+0.14)

Ha 7.52 7.35 (-0.17)

Hb 7.79 7.61 (-0.18)

Hc 7.82 7.72 (-0.10)

Hd 6.99 7.01 (+0.02) a Upfield displacements are negative

FIGURE 5.10: Partial 1H 600 MHz NMR spectra of BMDBM tert-butyl and aromatic Ha-

Hd resonances of (i) solution of BMDBM (Solution A) and (ii) solution of BMDBM and

βCD in a 1 : 2 mole ratio (Solution B) in 0.1 mol dm-3 NaOD/D2O at 298 K. The

degradation productions of BMDBM in (i) are denoted by Haʹ, Hb , Hc , Hd and t-Bu.

The spectra are not plotted to a constant vertical scale.

O O

OHa

Hb HcHd

6.877.27.47.67.88 1.21.5(ppm)

Hd Ha Hb Hc

Hd Ha Hb Hc

Hdʹ Hcʹ

Hbʹ Ha t-Buʹ

t-Bu

t-Bu (i)

(ii)

Chapter 5

162

The 1H ROESY NMR spectrum of a solution in which the mole ratio of BMDBM− and

βCD was 1 : 2 is shown in FIGURE 5.11.

FIGURE 5.11: 1H 600 MHz ROESY NMR spectrum of a solution BMDBM with βCD in a

1 : 2 mole ratio in 0.10 mol dm-3 NaOD at 298 K. The cross-peaks enclosed in the boxes

correspond to intermolecular interactions between the protons indicated on the F1 and F2

axes.

Strong cross-peaks are observed between BMDBM− tert-butyl resonances and those of

βCD H3, H5 and H6 protons of the annular interior. Strong cross-peaks are also observed

βC

D p

roto

ns

H4

H2

H5

H6

H3

Hc Hb Ha Hd t-Bu

βCD H1

HOD βCD H2-6

H5

H6

H6

H4

H2

H1

H3

O O

OHa

Hb HcHd

Chapter 5

163

between BMDBM− Ha and βCD H3 and H5 as well as BMDBM− Hd and βCD H3. A

somewhat weaker cross-peak can be observed between BMDBM− Hb and βCD H6. The

interactions, if any, between the methoxy protons of BMDBM− and βCD protons are not

observed as they are masked by the βCD resonances. No interactions are observed

between BMDBM− protons and the exterior βCD H1, H2 or H4 protons.

The experimental observations taken from the 1H and 1H ROESY NMR support the

formation of a 1 : 1 host-guest inclusion complex, βCD·BMDBM−, and the subsequent

formation of a 2 : 1 complex, (βCD)2·BMDBM−. The single resonances observed for each

of the tert-butyl and Ha-Hd protons supports a fast exchange between free βCD, free

BMDBM−, βCD·BMDBM− and (βCD)2·BMDBM−. As the process is rapid on the NMR

timescale, the observed resonance is a time- and weight-averaged resonance for each of the

different chemical environments.

The different possible inclusion orientations of βCD relative to BMDBM− results in a

number of includomers existing in thermodynamic equilibrium as shown in FIGURE 5.12.

A visual comparison of the cross-peak intensities in the 1H NMR ROESY, taking into

consideration the number of BMDBM− protons giving rise to the cross-peaks with βCD

protons, supports the preferential inclusion of the tert-butyl phenyl group to the methoxy

phenyl group to form βCD·BMDBM− includomer A and βCD·BMDBM− includomer Aʹ in

FIGURE 5.12. This would be facilitated by the greater hydrophobic nature of the tert-

butyl phenyl group compared with the methoxy phenyl group. Support for the proposed

mode of inclusion is provided by the chemical shift changes of BMDBM− tert-butyl group

and Ha-Hb protons. The upfield shift for Ha and Hb protons indicates a more hydrophobic

environment, consistent with the positioning of βCD over the phenyl group. The

downfield displacement of the tert-butyl protons suggests a close proximity to an

electronegative oxygen atom, placing it either near the narrow (primary) hydroxyl end

(includomer A in FIGURE 5.12) of the CD cavity or the wider (secondary) hydroxyl rim

(includomer Aʹ in FIGURE 5.12). The larger size of the βCD annulus relative to the tert-

butyl group means the tert-butyl group can insert either end without experiencing

significant steric effects. This is evident in the AM1 optimised geometry of includomer A′

shown in FIGURE 5.13 [403]. Consequently, several inclusion orientations of BMDBM−

relative to the βCD are possible and the experimental NMR spectra represents a

population-weighted average of the spectrum of βCD·BMDBM− complexes involving

different threading orientations of βCD over the tert-butyl phenyl group.

Chapter 5

164

FIGURE 5.12: Formation of βCD·BMDBM − and (βCD)2·BMDBM

− inclusion complexes

involving the possible inclusion orientations of βCD.

O O

O

O O

O

O O

O

O O

O

O O

O

O O

O

O O

O

O O

O

Includomer B

Includomer B′

Includomer D′

Includomer A′

Includomer C′Includomer C

Includomer A

Includomer D

Chapter 5

165

FIGURE 5.13: Geometry of includomer A′ optimized at the AM1 level of theory. The

carbons in βCD are coloured grey and blue in BMDBM to distinguish between the two

molecules.

Chapter 5

166

The interconversion of includomer A and Aʹ occurs through decomplexation of the

(βCD)2·BMDBM− includomer A to BMDBM− and βCD followed by formation of

βCD·BMDBM− includomer Aʹ. In the 1H NMR ROESY, the comparatively weaker cross-

peaks between BMDBM− tert-butyl protons and βCD H6 coupled with the absence of

cross-peaks between BMDBM− Ha and βCD H6 suggests the preferential formation of

βCD·BMDBM− includomer A′ over βCD·BMDBM− includomer A.

Evidence of the complexation of the methoxy phenyl group is provided by the cross-peaks

in the 1H NMR ROESY observed between BMDBM− Hd and βCD H3 protons and

between Hc and βCD H6. There are two possible inclusion orientations of βCD relative to

BMDBM−, βCD·BMDBM− includomer B and includomer Bʹ It is not possible to say with

any certainty if one includomer is preferred over the other as the methoxy interactions with

βCD protons are masked and the observed cross-peaks for Hc are either weak or noise.

This does support, however, the previously proposed notion that the tert-butyl group is

complexed in preference to the methoxy phenyl group although formation of

βCD·BMDBM− includomer A and Aʹ does not preclude formation of βCD·BMDBM−

includomer B and B.

If βCD·BMDBM− includomer Aʹ is the dominant 1 : 1 inclusion complex as previously

proposed then the variation in the intensities of the cross-peaks for Hc and Hd with βCD

protons and the absence of cross-peaks for Hd with βCD H5 can be attributed to the

presence of (βCD)2·BMDBM− includomer Cʹ in addition to includomer Dʹ where the βCD

orientation over the methoxy phenyl group is reversed. If the ROESY interactions of Hc

and βCD H6 protons were due to the presence of (βCD)2·BMDBM− includomer Dʹ only

BMDBM− Hd protons would be positioned deep in the βCD cavity and cross-peaks of

similar intensity with both βCD H3 and H5 would be observed. The difference in chemical

shift between free and included Hc and Hd of BMDBM− is much less than that observed

for Ha and Hb (TABLE 5.1), however the chemical environment of the methoxy phenyl

protons are not expected to be as sensitive to the hydrophobic environment of the βCD

cavity because of the deshielding effect of the methoxy group.

The chemical environments of the tert-butyl and Ha-Hb protons in βCD·BMDBM− are not

expected to be affected significantly by the addition of the second βCD to the methoxy

phenyl end of the molecule to form (βCD)2·BMDBM− includomers Cʹ and Dʹ.

Accordingly, only one chemical environment would be observed for the tert-butyl and Ha-

Chapter 5

167

Hb protons in βCD·BMDBM− and (βCD)2·BMDBM−. The same principle applies to the

chemical environment of the methoxy phenyl protons if the first βCD were to be positioned

over the methoxy phenyl group in βCD·BMDBM− and the subsequent complexation of a

second βCD to the tert-butyl phenyl group to form (βCD)2·BMDBM−.

5.2.1.2 NMR data of BMDBM with excess βCD.

In the 1H NMR spectrum of a solution, solution B, in which the mole ratio of BMDBM−

and βCD was 1 : 6 (as determined by integration of 1H NMR signals), the tert-butyl

resonance appears as two singlets and the BMDBM− aromatic resonances assigned to Ha-

Hd appear as three sets of doublets (FIGURE 5.14b). Assignment of the methoxy

resonance is not possible due to it having a similar chemical shift as that of the βCD H3

resonance. Integration of all BMDBM− proton signals reveals that the larger tert-butyl

resonance having an observed chemical shift of δ 1.41 is due to two tert-butyl groups

possessing coincident chemical shifts, appearing with an area ratio of 2.7 : 1. This is

possible if either they possess coincident chemical shifts or the chemical environments of

the two are similar. The 1H NMR spectrum of free BMDBM− as shown in FIGURE 5.14a

is the same as presented earlier in FIGURE 5.10a.

The 1H NMR chemical shifts of BMDBM− resonances for solutions of free BMDBM−,

BMDBM− and βCD in a 1 : 2 molar ratio and BMDBM− and βCD in a 1 : 6 mole ratio are

compared in TABLE 5.2. The distinction between the different chemical environments of

the tert-butyl phenyl and methoxy phenyl groups has been made by using different

coloured text. The chemical shifts of the dominant includomer observed for solution B

(blue text in FIGURE 5.14 and TABLE 5.2), are very similar to those of solution B. In

solution A, these resonances were proposed to be due to predominately (βCD)·BMDBM−

with a lesser amount of (βCD)2·BMDBM. In solution B, βCD is in much greater excess

and therefore it seems reasonable to assume that the major resonances (t-Bu, Ha-Hd in

FIGURE 5.14) are predominately due to (βCD)2·BMDBM−. The change in populations of

(βCD)·BMDBM− and (βCD)2·BMDBM− in solution A and B is supported by the changes

in chemical shifts.

Using 1H NMR ROESY, it was possible to assign the resonances of the minor includomers

in solution to two different tert-butyl phenyl groups and two different methoxy phenyl

groups. Using this method it was not possible to connect either of the tert-butyl phenyl

groups to a particular methoxy phenyl group and vice-versa as the distance between the

Chapter 5

168

two phenyl groups is too large to generate cross-peaks in the 1H NMR ROESY or for either

phenyl group to significantly influence the chemical environment of the other. Therefore,

the assignment of these resonances to a particular includomer cannot be made with

certainty.

FIGURE 5.14: Partial 1H 600 MHz NMR spectra of BMDBM −- tert-butyl and aromatic

Ha-Hd resonances of (i) solution of BMDBM − and (ii) solution of BMDBM − and βCD in

a 1 : 6 mole ratio (Solution B) in 0.1 mol dm-3 NaOD / D2O at 298 K. The degradation

products of BMDBM in (i) are denoted by Haʹ - Hd and t-Bu. The coloured text makes

the distinction between the resonances assigned to the different includomers. The spectra

are not plotted to a constant vertical scale.

O O

OHa

Hb HcHd

(i)

(ii)

6.877.27.47.67.88 1.21.5

(ppm)

Hd Ha Hb Hc

Hd Ha Hb Hc Hdʹ Hcʹ

Hbʹ Ha t-Buʹ

t-Bu

t-Bu t-Bu

Hd Hd Hc Hc

Hb

Hb

Ha t-Bu Ha

Chapter 5

169

TABLE 5.2: 1H chemical shifts (ppm) corresponding to BMDBM − in the absence and

presence of βCD in 0.1 mol dm-3 NaOD/D2O at 298 K .

δfree δcomplex (∆λa) δcomplex (∆λa)

1 : 2 mole ratio b 1 : 6 mole ratio t-Bu 1.28 1.42 (+0.14) 1.41 (+0.13) 1.41 (+0.13) 1.43 (+0.15)

Ha 7.52 7.35 (-0.17) 7.38 (-0.14) 7.34 (-0.18) 7.52 (+0.00)

Hb 7.79 7.61 (-0.18) 7.61 (-0.18) 7.74 (-0.05) 7.90 (+0.11)

Hc 7.82 7.72 (-0.10) 7.73 (-0.09) 7.83 (+0.01) 7.96 (+0.14)

Hd 6.99 7.01 (+0.02) 6.99 (+0.00) 6.95 (-0.04) 7.06 (+0.07) a Upfield displacements are negative b Mole ratio of BMDBM− to βCD in solution.

These observations are consistent with the formation of three distinct includomers in slow

exchange on the NMR timescale that appear in the area ratio 2.7 : 1 : 1. Evidence for the

slow exchange process is provided by the reduced resolution of the βCD resonances in

solution B as compared with those of the solution having lower βCD concentrations

(solution A) where the exchange process was rapid as shown in FIGURE 5.15. The

greatest change observed is for the resonances assigned to the interior βCD protons H3, H5

and H6. The downfield shift of the H5 resonance means it is no longer possible to assign

individual resonances to H5 and H6. There is no significant change in the resonances

assigned to the exterior βCD protons H2 and H4. Furthermore, as the concentration of

βCD in solution B is greater than for solution A, the ratios of 1 : 1 and 1 : 2 inclusion

complexes will change as will the annuli CD 1H chemical shifts.

Chapter 5

170

FIGURE 5.15: Partial 1H 600MHz NMR spectra of βCD H2-H6 resonances of (i)

solution of BMDBM − and βCD in a 1 : 2 mole ratio (solution A) and (ii) solution of

BMDBM − and βCD having a 1 : 6 mole ratio (solution B) in 0.1 mol dm-3 NaOD/D2O at

298 K. The spectra are not plotted to a constant vertical scale.

The 1H ROESY NMR spectrum of a solution of BMDBM− and an excess of βCD is shown

in FIGURE 5.16. This shows strong intermolecular interactions between the proton

resonances of BMDBM− previously assigned to the dominant includomer and those of the

βCD interior H3, H5 and H6 protons. Strong cross-peaks are observed between BMDBM−

tert-butyl protons and βCD protons H3, H5 and H6. The colour scheme used to distinguish

between includomers is the same as used previously. As the tert-butyl resonance is

coincident with the dominant tert-butyl resonance, it is not possible to ascertain if there are

interactions between the minor tert-butyl protons and βCD protons. The apparent cross-

peaks between BMDBM− tert-butyl protons and βCD exterior protons H2 and H4 are

possibly due to noise running horizontally along the spectra. Strong cross-peaks are also

observed between the BMDBM− Ha and βCD H3, H6 and H5. Cross-peaks can be seen

between BMDBM− Hb and Hc protons and βCD H5/H6 protons. Additionally, there are

interactions between BMDBM− Hd with βCD H3 protons. For the minor includomers, a

weak cross-peak is observed between BMDBM− Ha and βCD H5 and H6.

H5

H6 H6

H4

H2

H1

H3

3.43.53.63.73.83.9(ppm)

H5 H3 H6

H2 H4

H2 H4 H6 H5

H3

(i)

(ii)

Chapter 5

171

FIGURE 5.16: 1H 600 MHz ROESY NMR spectrum of a solution of BMDBM − and βCD

in a 1 : 6 mole ratio (solution B) in 0.10 mol dm-3 NaOD at 298 K. The cross-peaks

enclosed in the boxes correspond to intermolecular interactions between the protons

indicated on the F1 and F2 axes. For an expansion of BMDBM − aromatic Ha-Hd

resonances, refer to FIGURE 5.14.

The observations from the 1H NMR and the 1H ROESY NMR of solution B provides

additional evidence in support of the formation of (βCD)2·BMDBM− in which βCD is

O O

OHa

Hb HcHd

H5

H6 H6

H4

H2

H1

H3 βC

D p

roto

ns

H4

H2

H6,

H5

H3

βCD H1

HOD βCD

H2-H6

t-Bu

Chapter 5

172

positioned over the tert-butyl phenyl group whilst a second βCD envelopes the methoxy

phenyl group. In solution A, the resonances (tert-butyl, Ha-Hd) were due to a fast

exchange on the NMR timescale between (βCD)·BMDBM− and (βCD)2·BMDBM− and

having a greater population of (βCD)·BMDBM−. In solution B, the higher mole ratio of

βCD supports the observed increase in population of the 2 : 1 species, (βCD)2·BMDBM−.

The possible inclusion orientations of βCD with BMDBM− and the equilibrium between

these possible includomers is the same as discussed previously for FIGURE 5.12. In

solution A, the dominant 1 : 1 inclusion complex, (βCD)·BMDBM−, was previously

assigned to includomer A' . This was supported by chemical shift changes between free

and included BMDBM− and also by 1H ROESY NMR interactions. There is no change

observed for the chemical shifts of the resonances tert-butyl, Ha-Hd in the 1 : 1 inclusion

complex, (βCD)·BMDBM− with those in the 2 : 1 complex, (βCD)2·BMDBM−. This is

attributed to the distance between the phenyl groups, where changes in the chemical

environment of Ha and Hb are too far away to influence Hc and Hd. The relative

intensities confirm the preferential inclusion of BMDBM− tert-butyl phenyl group.

5.2.1.3 Temperature-Dependence Studies of βCD Inclusion Complexes with BMDBM

The nature of the inclusion complexes formed were investigated further by variable-

temperature 1H NMR studies (298-323 K) of a solution of BMDBM− and βCD in a 1 : 6

mole ratio. The 1H NMR spectrum obtained at 298 K and 323 K are shown in FIGURE

5.17. The entire spectrum, referenced to the HOD solvent resonance, is shifted downfield

with increasing temperature. The dielectric constant of the solvent changes as temperature

increases, influencing its shielding properties. No significant broadening of the resonances

assigned to BMDBM− tert-butyl and aromatic protons Ha-Hd can be observed with

increasing temperature. The three sets of doublets previously assigned to Ha-Hd of

BMDBM− exist in the area ratio 2.7 : 1 : 1 at 298 K indicating a major includomer and two

minor includomers where the two minor includomers exist in the same ratio. As

temperature increases to 323 K, this ratio decreases to 2 : 1 : 1 where the ratio of the two

minor includomers, relative to each other, is not influenced by temperature.

Chapter 5

173

FIGURE 5.17: Partial variable-temperature 1H 600 MHz NMR spectra of BMDBM −

tert-butyl and aromatic Ha-Hd resonances of a solution of BMDBM − and βCD in a 1 : 6

mole ratio (solution B) at (i) 298 K and (ii) 323 K in 0.1 mol dm-3 NaOD/D2O. The spectra

are not plotted to a constant vertical scale.

The observations from the variable-temperature 1H NMR studies supports the presence of a

slow exchange between the three distinct includomers of BMDBM− and βCD where the

coalescence temperature has not been reached within the temperature range studied. If

initially the exchange between the major and two minor includomers is slow on the NMR

timescale at 298 K, coalescence of the proton resonances assigned to the includomers is

expected as temperature increases. However, no coalescence is observed indicating the

rate of exchange is not increasing and the coalescence temperature, where the peaks merge,

has not been reached. The includomers are still in slow exchange even at the higher

temperatures although the relative populations of includomers are changing. The change

6.97.27.57.88.1 1.31.51.7(ppm)

Hb Hc

Hb Hb

Hc

Ha Ha

Hd Ha

Hc

Hd Hd

Hb Hc

Hc

Hb Hc Hd Ha

Ha Ha Hd Hd

t-Bu

t-Bu, t-Bu

t-Bu

t-Bu, t-Bu

(ii)

(i)

O O

OHa

Hb HcHd

Chapter 5

174

in the populations, indicated by the area ratios being 2.7 : 1 : 1 at 298 K and 2 : 1 : 1 at 323

K indicates a change in the equilibrium position. The equilibrium constant for the

exchange process, involving βCD moving on and off BMDBM−, is a function of the

temperature as indicated by:

RT

GK

∆In

−=

where ∆G is Gibbs free energy, R is the gas constant and T is temperature. That the

relative populations of the minor includomers does not change with temperature implies

these species have similar energies.

5.2.2 HPβCD Complexes

5.2.2.1 NMR data of BMDBM with HPβCD in D2O

In contrast to βCD and its inclusion complexes, the inclusion complexes of HPβCD with

BMDBM are sufficiently soluble to allow acquisition of a 1H NMR spectrum in D2O. In

the absence of HPβCD, BMDBM cannot be detected by 1H NMR spectroscopic methods

due to the limited aqueous solubility of BMDBM (1.5 µg/ml [213]). An excess of

BMDBM was added to a solution of HPβCD (0.01 mol dm-3) in D2O and stirred for 25 hrs.

The resulting suspension was filtered to remove undissolved BMDBM and, therefore the

concentration of BMDBM is not known with any accuracy. The mole ratio of BMDBM

and HPβCD, determined by integration of the 1H NMR signals, is approximately 1 : 9.

Phase solubility studies show the solubilising effect of HPβCD on BMDBM is greater than

for the parent βCD [213,371]. The increase in solubility of BMDBM with HPβCD is

attributed to the greater solubility of HPβCD itself over that of βCD.

In the 1H NMR spectrum of a solution of BMDBM and excess HPβCD in D2O (FIGURE

5.18) the tert-butyl resonance appears as two singlets. The BMDBM aromatic resonances

Ha-Hd appear as two sets of poorly resolved doublets arising from the AAʹBBʹ spin-spin

splitting pattern of the para-substituted phenyl rings. The BMDBM vinylic proton Hv of

the enol tautomer of BMDBM can be observed as a broad singlet having an area ratio less

than one due to exchange with the solvent. The enolic hydrogen involved in the

intramolecular hydrogen bond in the enol form of BMDBM is not observed for similar

Chapter 5

175

reasons. The methoxy resonance cannot be distinguished from those of the HPβCD

resonances.

FIGURE 5.18: Partial 1H 600MHz NMR spectrum of BMDBM tert-butyl and aromatic

Ha-Hd resonances and HPβCD (0.01 mol dm-3) in D2O at 298 K. The vertical scaling of

the spectra is not constant.

As with other β-diketones, BMDBM exists in a keto-enol tautomeric equilibrium (FIGURE

5.4) with the equilibrium position heavily influenced by the nature of the solvent [293].

Consequently, there are two different complexing environments of BMDBM, the keto (K)

form and the enol (E) form as shown in FIGURE 5.19. The detection of BMDBM in the 1H NMR solution may be the consequence of the formation of an inclusion compound

between the keto form and HPβCD (K·HPβCD) and/or the enol form and HPβCD

(E·HPβCD). There is also the possibility for formation of 1 : 2 (BMDBM : HPβCD)

inclusion complexes or either the keto and enol forms. It is possible to distinguish between

the two tautomers by 1H NMR based on the separate resonance signals for the vinylic and

methylene protons of the enol and keto tautomers respectively. The vinylic proton of the

1.41.56.877.27.47.67.888.28.4

(ppm)

Hd Ha Hb Hc t-Bu Hv

O O

OHa

Hb HcHd

H

Hv

Chapter 5

176

enol tautomer of BMDBM can be observed as a broad singlet whereas the methylene

resonance of the keto tautomer is not observed.

FIGURE 5.19: Keto-enol tautomerisation of BMDBM in the absence and presence of

HPβCD.

O O

O

O O

O

H

O O

O

HO O

O

E

E•HPβCD

O O

O

HO O

O

K

K•HPβCD

K•(HPβCD)2 E•(HPβCD)2

Chapter 5

177

There are two possibilities for the absence of the methylene resonance; either the keto is

not present in solution or the methylene resonance is masked by the HPβCD resonances.

In polar, protic solvents, the keto-enol equilibrium of β-diketones shifts towards the keto

tautomer as the intramolecular hydrogen bond of the enol is replaced by hydrogen bonding

with the solvent, stabilising the keto form. However, the enol form is still the dominant

tautomer. This is confirmed by investigation of the keto-enol equilibrium of BMDBM in

different solvents. The poor aqueous solubility of free BMDBM prevents acquisition of a 1H NMR spectrum of BMDBM in D2O. However, a quantitative evaluation of the

tautomeric equilibrium of BMDBM in different solvents indicates the enol form

predominates. Integration of the relative intensities of the vinylic and methylene protons

shows 100% enol in CDCl3 and 90% in d6-DMSO, which is in agreement with similar

studies of BMDBM [205]. Consequently, the absence of the keto tautomer cannot be

stated with any certainty but the enol is assumed to be the predominate form. The

geometries of both tautomers of BMDBM have been optimised at the B3LYP/6-31+G(d,p)

level [403], and while the enol is a planar molecule, the keto tautomer is not as shown in

FIGURE 5.20. Moreover, tautomerisation from the enol to the keto form requires

significant structural changes. Therefore, preferential inclusion of the enol tautomer is

expected.

FIGURE 5.20: Structures of chelated enol and keto forms of BMDBM optimised at the

B3LYP/6-31+G(d,p) level of theory [403]. (Only one of the chelated enol forms is shown).

The 1H ROESY NMR spectrum of a solution of BMDBM and an excess of HPβCD in D2O

shows intermolecular interactions between the tert-butyl resonance of BMDBM and

HPβCD H2-H6 resonances (FIGURE 5.21) indicating that the tert-butyl protons are within

~4 Å of the HPβCD interior H2-H6 protons. A weak cross-peak is observed for BMDBM

aromatic Hd resonance, however this is possibly due to interaction between BMDBM Hd

Enol Keto

Chapter 5

178

and methoxy protons rather than an interaction between BMDBM Hd protons and HPβCD

H2-6 protons. No cross-peaks are observed between BMDBM aromatic Ha-Hc and those

of HPβCD. This may be due to either no interaction or the interaction is sufficiently weak

at the present concentrations that it is not detected.

FIGURE 5.21: 1H 600 MHz ROESY NMR spectrum of a solution of BMDBM and HPβCD

in a 1 : 9 mole ratio in D2O at 298 K. The cross-peaks enclosed in the boxes correspond to

intermolecular interactions between the protons indicated on the F1 and F2 axes.

Hc Ha Hd Hb t-Bu H2-6 H1

HPβCD Methyl

H2-

6

HPβCD

HPβC

D p

roto

ns

O O

OHa

Hb HcHd

H

H5

H6 H6

H4

H2

H1

H3

Chapter 5

179

The low concentration of BMDBM in solution and corresponding absence of cross-peaks

makes it difficult to identify the mode of inclusion with any certainty. However, it seems

that there is preferential inclusion of the enol tautomer with HPβCD positioned over the

tert-butyl group. The equilibrium shown in FIGURE 5.12 for inclusion of BMDBM with

βCD is, in principle, also possible for inclusion complexes with HPβCD. From the 1H

ROESY NMR data it is likely that HPβCD·BMDBM includomer A and/or includomer Aʹ

are the dominant inclusion complexes present. It is possible that the keto tautomer can

form inclusion complexes with HPβCD. Keto-enol equilibrium studies of the β-diketone,

benzoylacetone (1-phenyl-1,3-butadione), show both the keto and enol tautomers form

inclusion complexes with βCD [404]. The planar geometry of the enol tautomer of

benzoylacetone allows deeper protrusion inside the CD cavity than the keto tautomer, it is

stabilised in the hydrophobic cavity interior of CD and the keto-enol equilibrium is shifted

to the enol tautomer.

5.2.2.2 NMR data of 1 : 1 mole ratio of BMDBM with HPβCD

For further characterisation of HPβCD inclusion complexes, it was necessary to prepare

solutions of BMDBM and HPβCD in 0.10 mol dm-3 NaOD/D2O to achieve higher

concentrations than was possible in D2O. As discussed previously for βCD solutions, the

basic conditions increase both the solubility of HPβCD and, to a lesser extent, BMDBM

with the pKa of HPβCD is expected to be comparable to the pKa of βCD (pKa 12.20 [402]).

Solutions were prepared by adding BMDBM to a solution of HPβCD in 0.1 mol dm-3

NaOD/D2O. The resulting suspension was gently heated then filtered to remove any

suspended material, which was probably BMDBM and/or its HPβCD complex. As a

consequence of the poor solubility of BMDBM, even under the basic condition used, the

calculated concentrations of BMDBM− were not achieved. Instead, the mole ratios of

HPβCD to BMDBM− were determined by integration of the appropriate 1H NMR signals.

The 1H NMR spectrum of a solution (solution A) in which the mole ratio of BMDBM− and

HPβCD was ~1 : 1 (determined by integration of NMR signals) is shown in FIGURE 5.22.

The tert-butyl resonance appears as a broad singlet and BMDBM− aromatic Ha-Hd

resonances appear as doublets with replication. There is some broadening of the aromatic

Ha-Hd resonances, which is more noticeable for Ha and Hb than for Hc and Hd. The

differences in broadening are consistent with different rates of exchange at the methoxy

and tert-butyl phenyl groups between free and included BMDBM−. The methoxy

resonances are not easily distinguished from the HPβCD H2-H6 proton resonances. Three

Chapter 5

180

separate chemical environments can be observed for BMDBM− Ha-Hd appearing in the

area ratio of ~1 : 1 : 6. Integration of the broad singlet assigned to the tert-butyl resonance

is consistent with this ratio, indicating either three chemically distinct tert-butyl groups that

possess coincident chemical shifts or the tert-butyl protons experience no significant

difference in chemical environments in the different includomers. The chemical shifts and

chemical shift differences of all BMDBM− protons in the absence and presence of HPβCD

are presented in TABLE 5.3.

FIGURE 5.22: Partial 1H 600MHz NMR spectra of BMDBM − tert-butyl and aromatic

Ha-Hd resonances of (i) solution of BMDBM − and (ii) solution of BMDBM −and HPβCD

having ~1 : 1 mole ratio (Solution A) in 0.1 mol dm-3 NaOD/D2O at 298 K. The

degradation productions of BMDBM in (i) are denoted by Haʹ - Hd and t-Bu. The

coloured text makes the distinction between the resonances assigned to the different

includomers. The spectra are not plotted to a constant vertical scale.

t-Bu′

1.21.5(ppm)

6.877.27.47.67.88

t-Bu′ t-Bu

O O

OHa

Hb HcHd

(ii)

(i)

t-Bu

Hc Hb Hc Hb

Hb Hc

Ha

Ha Ha Hd

Hd Hd

Hcʹ

Hbʹ Ha

Hc Hb Ha Hdʹ Hd

Chapter 5

181

TABLE 5.3: 1H chemical shifts (ppm) corresponding to BMDBM − in the absence and

presence of HPβCD in 0.1 mol dm-3 NaOD / D2O at 298 K.

δfree δcomplex (∆λa)

~1 : 1 mole ratiob

t-Bu 1.28 1.41(+0.13) 1.41(+0.13) 1.41(+0.13)

Ha 7.52 7.35(-0.17) ~7.34(-0.18) 7.52(+0.00)

Hb 7.79 7.63(-0.16) 7.76(-0.03) 7.91(+0.12)

Hc 7.82 7.71(-0.11) 7.81(-0.01) 7.98(+0.16)

Hd 6.99 7.01(+0.02) 6.98(-0.01) 7.06(+0.07) a Upfield displacements are negative b Mole ratio of BMDBM− : HPβCD in solution

This spectrum is similar to that of a solution of BMDBM− and βCD (1 : 6 mole ratio)

(FIGURE 5.14). As the cavity of HPβCD has a similar structure to that of the parent βCD

therefore a similar mode of inclusion is expected. A difference between the two is the

concentration of observed includomers with similar CD concentrations. At comparable

βCD concentrations (1 : 2 mole ratio BMDBM− to βCD) BMDBM− existed primarily as

βCD·BMDBM−. This can be attributed to the higher solubility of HPβCD itself relative to

βCD. The solubility of HPβCD is influenced by the nature of the substitution groups, the

degree of substitution and the pattern of substitution. The degree of substitution of HPβCD

used in this study is 3.0-8.0 which gives an average number of 5 hydroxypropyl groups per

βCD molecule.

Aided by 1H NMR ROESY, it is possible to distinguish between the three different tert-

butyl and methoxy phenyl groups, indicated by coloured text in FIGURE 5.22 and TABLE

5.3. For the minor species present in solution, it is not possible to relate the different

chemical environments of any one tert-butyl phenyl groups to a particular methoxy phenyl

group and vice-versa with any certainty. This is due to the distance between the phenyl

groups, where changes in the chemical environment of Ha and Hb are too far away to

influence Hc and Hd. Therefore, the assignment of these resonances to a particular

includomer can not be made with certainty.

The 1H ROESY NMR spectrum of a solution in which the mole ratio of BMDBM− and

HPβCD was ~1 : 1 (determined by integration of NMR signals) is shown in FIGURE 5.23.

Chapter 5

182

FIGURE 5.23: 1H 600 MHz ROESY NMR spectrum of a solution of BMDBM − and

HPβCD in a 1 : 1.3 mole ratio in 0.10 mol dm-3 NaOD at 298 K. The cross-peaks enclosed

in the boxes correspond to intermolecular interactions between the protons indicated on

the F1 and F2 axes. Not all BMDBM − resonances are labelled.

Strong intermolecular interactions are observed between the dominant BMDBM− tert-butyl

resonance and aromatic Ha resonance with HPβCD H2-H6 resonances. A weaker

interaction is observed between the dominant BMDBM− Hb and βCD H2-H6 resonances.

Cross-peaks are observed for the dominant BMDBM− aromatic Hd resonance but this may

CD

pro

tons

OC

H3

H2-

H6

Hc Hb Ha Hd

t-Bu CD H1

HOD

CD H2-6 CD

CH3 OCH3

O O

OHa

Hb HcHd

H5

H6 H6

H4

H2

H1

H3

Chapter 5

183

be due to correlation with the signal assigned to the methoxy resonance of the BMDBM−,

which overlaps with the HPβCD protons. The absence of cross-peaks between the minor

BMDBM− Ha-Hd resonances assigned to the minor includomers may be due to the low

concentration of these species. The interactions between the tert-butyl and Ha-Hb protons

with HPβCD H2-H6 protons is similar to that observed for βCD·BMDBM− (FIGURE

5.11) although no interactions are observed between BMDBM− Hc-Hd and HPβCD

protons. This supports the view that the βCD spectra represents a fast exchange process

occurring between βCD·BMDBM− and (βCD)2·BMDBM−.

These observations are consistent with BMDBM− existing primarily as HPβCD·BMDBM−

where the cavity of HPβCD is positioned over the tert-butyl and Ha protons. Again, a

similar mode of inclusion to βCD is expected for HPβCD and, therefore, the equilibrium

shown in FIGURE 5.12 and reproduced in FIGURE 5.24 is, in principle, applicable here

also. As there are two possible threading orientations of HPβCD such that either the tert-

butyl group is closest to the primary or secondary hydroxyl lined rim, two includomers

corresponding to HPβCD·BMDBM− includomer A and includomer Aʹ are possible

(FIGURE 5.24). As inclusion by HPβCD is a dynamic process, it is likely that the

observed resonances assigned to HPβCD·BMDBM− are a time- and population-average of

the individual resonances of these two includomers. This is supported by the observed

broadening of the dominant BMDBM− Ha and Hb resonances. As the HPβCD resonances

are less resolved due to substitution than those observed for the unsubstituted βCD, it is not

possible to determine if there is a preferred threading orientation of the CD. The proposed

mode of inclusion is similar to that proposed for the analogous βCD inclusion complex

(βCD·BMDBM−). This is consistent with HPβCD having a similar structure and cavity

size to that of the parent βCD.

Chapter 5

184

FIGURE 5.24: Formation of HPβCD·BMDBM − and (HPβCD)2·BMDBM

− inclusion

complexes involving the possible inclusion orientations of HPβCD.

O O

O

O O

O

O O

O

O O

O

O O

O

O O

O

O O

O

O O

O

Includomer B

Includomer B′

Includomer D′

Includomer A′

Includomer C′Includomer C

Includomer A

Includomer D

Chapter 5

185

5.2.2.3 NMR data of BMDBM with excess HPβCD

The 1H NMR spectrum of a solution of BMDBM− and an excess of HPβCD is shown in

FIGURE 5.25. The tert-butyl resonance appears as two singlets with very similar chemical

shift and the aromatics, Ha-Hd as doublets with replication. The aromatic resonances Ha-

Hd appear in the area ratio is approximately 1 : 1 : 1.5. The chemical shifts of the

replicated aromatic Ha-Hd (TABLE 5.3) are similar to those observed in the 1H NMR

spectrum of BMDBM− with equimolar HPβCD. This suggests that the βCD and HPβCD

includomers formed with BMDBM are similar and therefore, the increase in solubility of

HPβCD includomers can be attributed to the increase in solubility of HPβCD itself.

FIGURE 5.25: Partial 1H 600MHz NMR spectra of BMDBM − tert-butyl aromatic Ha-Hd

resonances of a) solution of BMDBM − and b) solution of BMDBM − with an excess of

HPβCD in 0.1 mol dm-3 NaOD/D2O at 298 K. The spectra are not plotted to a constant

vertical scale.

O O

OHa

Hb HcHd

6.877.27.47.67.88 1.31.4(ppm)

Hcʹ

Hbʹ Ha

Hc Hb Ha Hdʹ Hd

t-Bu, t-Bu, t-Bu

t-Bu

Hc

Hb

Hc Ha Ha Ha

Hd

Hb Hc Hd Hd

t-Buʹ

Hb

(ii)

(i)

Chapter 5

186

The 1H ROESY NMR spectrum of a solution of BMDBM− and an excess of HPβCD is

shown in FIGURE 5.26. Strong intermolecular interactions are observed between the tert-

butyl resonances with HPβCD H2-6 resonances. There are also strong cross-peaks

between two of the BMDBM− Ha aromatic resonances, Ha and Ha and those of HPβCD

H2-H6 protons. Additionally, a cross-peak is observed between BMDBM− Hb aromatic

resonance with the HPβCD H2-H6 resonances.

FIGURE 5.26: 1H 600 MHz ROESY NMR spectrum of BMDBM − and an excess of

HPβCD in 0.10 mol dm-3 NaOD at 298 K. The cross-peaks enclosed in the boxes

correspond to intermolecular interactions between the protons indicated on the F1

(BMDBM −) and F2 (HPβCD) axes.

H5

H6

H6

H4

H2

H1

H3

O O

OHa

Hb HcHd

CD

pro

tons

OC

H3

H2-

H6

Hc Hb Ha Hd t-Bu

CD H1

HOD

CD H2-6

CD CH3

OCH3

Chapter 5

187

5.2.2.4 Temperature-Dependence Studies of HPβCD Inclusion Complexes with

BMDBM

To study the nature of the HPβCD inclusion complexes formed further, variable-

temperature 1H NMR studies (298-323 K) were carried out on a solution of BMDBM− and

an excess of HPβCD. The 1H NMR spectrum obtained at 298 K and 323 K are shown in

FIGURE 5.27.

FIGURE 5.27: Partial variable-temperature 1H 600MHz NMR spectra of BMDBM − tert-

butyl and aromatic Ha-Hd resonances of a solution of BMDBM − with an excess of

HPβCD at (i) 298 K and (ii) 323 K in 0.1 mol dm-3 NaOD/D2O. The HPβCD methyl

resonance (δ 1.49 ppm) has been omitted for clarity. The spectra are not plotted to a

constant vertical scale.

1.31.51.7(ppm)

t-Bu, t-Bu, t-Bu

t-Bu, t-Bu, t-Bu

Ha

6.877.27.47.67.888.28.4

Hc Hb

Hc Ha Ha Hd

Hb Hc Hd Hd Hb

Ha

Hd

Hd

Hd

Ha Ha

Hc Hb

Hb Hc

Hb

Hc

(ii)

(i)

O O

OHa

Hb HcHd

Chapter 5

188

The entire spectrum, referenced to HOD solvent resonance, is shifted downfield with

increasing temperature as observed and discussed previously for the analogous studies of

BMDBM with βCD in Chapter 5.2.1.3. The area ratio of the three sets of doublets

previously assigned to BMDBM− aromatic Ha-Hd resonances changes from 1 : 1 : 1.25 at

298 K to 3 : 1 : 3 at 323 K. No significant broadening or coalescence of the resonances

assigned to BMDBM− tert-butyl or aromatic Ha-Hd protons can be observed with

increasing temperature. This indicates the coalescence temperature has not been reached

and the includomers are in slow exchange at the higher temperatures although the relative

populations of includomers are changing. The change in populations indicates a change in

the equilibrium positions where the equilibrium constants for the exchange process

involving HPβCD moving on and off BMDBM− is a function of temperature. Similar

observations were made in the analogous variable-temperature studies of BMDBM with

βCD. One difference in the two studies is that in the case of βCD the relative populations

of the minor includomers did not change with temperature whereas a change is observed

with increasing temperature for the relative populations of the HPβCD includomers.

The 1H NMR data of the inclusion complexes of βCD and HPβCD with BMDBM are in

agreement with phase solubility studies reported for these complexes [405] whereby

HPβCD forms more stable includomers. This is indicated by the association constants K of

1.44 × 103 mol dm-3 [213] and K of 2.23 × 103 mol dm-3 for formation of 1 : 1 inclusion

complexes of BMDBM with βCD and HPβCD, respectively [213,371]. Additionally, the

association constant K of 13 mol dm-3 for formation of a 1 : 2 inclusion complex of

BMDBM with HPβCD is reported whereas the analogous βCD includomer were not

detected.