Self-organization of discogenic molecules at the air-water interface

HAL Id: jpa-00247614

https://hal.archives-ouvertes.fr/jpa-00247614

Submitted on 1 Jan 1992

HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Self-organization of discogenic molecules at the air-water interface

Nicholas Maliszewskyj, Paul Heiney, J. Kent Blasie, John P. Mccauley Jr., Amos B. Smith Iii

To cite this version:

Nicholas Maliszewskyj, Paul Heiney, J. Kent Blasie, John P. Mccauley Jr., Amos B. Smith Iii. Self-

organization of discogenic molecules at the air-water interface. Journal de Physique II, EDP Sciences,

1992, 2 (1), pp.75-85. �10.1051/jp2:1992114�. �jpa-00247614�

Classification Physics Abstracts

68.10G 82.65D 61.30

Self-organization of discogenic molecules _at the air-water interface

Nicholas C.

Maliszewskyj II),

Paul A.

Heiney (~),

J. Kent Blasie

(~),

John P.

Mccauley,

Jr. (~) ^{and Amos B.}

Smith,

III

(2)

(1)

Department

of

Physics

and

Laboratory

for Research _on the Structure of Matter,

University

of Pennsylvania,

Philadelphia,

PA 19104, U-S-A-

(2)

Departrnent

of

Chemistry

and Laboratory for Research _on the Structure of Matter,

University

of Pennsylvania,

Philadelphia,

PA 19104, U-S-A-

(Received it

September

J99J,

accepted

27

September

J99J)

Abstract. The behavior of several

species

of

discogenic

molecules at the air-water interface has been determined

by measuring

surface pressure vs. molecular _area isotherms.

Alkylthiotripheny-

lene derivatives do not appear to form

monolayers

at the interfaces, due _to lack of

amphiphilic

character. Two _truxene derivatives _were found to

display

molecular _{areas on} the order of 65 A2,

most

likely

due to the molecules lying _« edge-on _» to the water surface.

Hexacyclens generally

lie with the _core group

parallel

to the surface of the _water with the tail groups

extending

_away from the interface _{; a} higher density

phase

_may

correspond

to conformational changes within the

molecule.

1. Introduction.

The

organization

of

amphiphilic

molecules such _as

lipids

and soaps into

monolayers

on water surfaces is _a well known

phenomenon [I]. Recently,

it has become clear

[2-7]

that

Langmuir monolayers

can also be formed from the

disc-shaped

molecules of which discotic

liquid crystals [8]

are

composed.

Such molecules

typically,

but not

universally,

have flat,

conjugated

cores and 4-8

aliphatic

tails. The bulk

liquid crystalline phases

_may be nematic («

ND »)

_or

hexagonal-columnar

(«

Dhd »),

^the latter structure

consisting

of _a

hexagonal

_array ^of columns

of molecules. The column axis is

approximately

normal _to the molecular _core and the

aliphatic

tails extend _{more or} less

radially

from the column axis.

By

contrast, ^surface pressure isotherm measurements have shown that several

quite

different conformations _are

possible

for

discogenic

molecules at the air-water interface.

For

phthalocyanines [3],

benzene hexa-n-alkanoates

[2, 5],

and

hexacyclens [6, 7],

the

molecular _areas

extrapolated

from measurements of

spreading

_pressure 17 _{versus area}

A are consistent with _a model for _a condensed

phase

in which the _core lies flat in the

surface,

with the

aliphatic

tails

projecting

_away from the surface.

By

contrast, the measured molecular

area of several

alkoxytriphenylene

derivatives is much smaller than the _area of such _a

X X la

X,R

m

SCSH,,

16

X,R

-

SC~H,~

lc

X,R

₌

SC7H,s

id X ~

COO(CH~)~CH~

R m

OCSH,,

1e

X,R

=

OCSH,,

R

2a R =

COO(CH~),~CH~

2b R

m

COOPhO(CH~),OCH~

R R

R

~~'~~

f~~'h

O

R

~N ~j

^{a R} =

C6H40C,~H~

R~O

b R

"

C6H40C~H~

~~"~ c R

=

C~HS

R d R

«

C17H35

R

~s~/~ eR=C~H~OC,,H~~

~~

^~

_~/~~

~~~~~~~

~

~~

~

~'

~N ~j ^~

j

^R

~ 4a.d

Scheme 1.

molecule

lying flat, implying

that these molecules most

likely

_are oriented _«

edge-on

_» to the air-water

interface,

with at least two tails

penetrating

into the

subphase [4, 7].

^Ther-

modynamic

transitions between different conformations may be

possible;

in

particular,

benzene hexa-n-alkanoate isotherms have been

interpreted

both in _terms of _{a structure} for which the _core lies

parallel

to the interface with the tails close

packed [5],

and in _terms of _a transition in which the entire molecule tilts from _a conformation in which the _core lies

parallel

(« flat

»)

to the interface to one in which it sits

perpendicular

_(«

edge-on »)

to the interface at

high compression [4].

We have measured surface pressure vs. molecular _area isotherms _{at room}

temperature

^for three classes of

thermotropic

discotic mesogen as shown in Scheme I

hexacydens,

truxenes, and

triphenylenes.

Thermal data and literature references for these

compounds

_are

given

in table I. The

goal

was to further

explore

the role of molecular architecture in

determining

both

liquid-crystal-phase

structure and _{structure at} the air-water interface.

Small

quantities

of solute

(-10~~g)

_were

weighed

_{on a} Cahn electrobalance _{to an}

accuracy of 10 ^~ g. The

hexacyclens

3a-3f and 4a-4d _are

quite hygroscopic [9, 10],

and for

these

compounds

it _was necessary to remove water

by heating

the

sample gently

to 55 °C

under

dynamic

_vacuum for at least 36 hours before

weighing

to obtain

reproducible

results.

After

weighing,

the solutes _were dissolved in 10 ml of HPLC

grade methylene

chloride.

Methylene

chloride is known to be _a

good spreading

solvent, immiscible in _water and

quite

volatile

(boiling point -40°C) [6].

The

subphase

for H-A measurements was water

purified

in _a

Millipore

filtration system. ^A

commercially

available Lauda film balance _was used to determine the surface pressure

by

the movable

barrier/Langmuir

float method to an

accuracy of

0.lmN/m.

All measurements were

performed

at room temperature,

T

=

21 _± 3 °C. After

performing

a

cleaning

_sweep of the surface, 0.1ml of the solution _was

spread

_on the interface and at least 10 min allowed for

evaporation

of the solvent. The film

was then

compressed

at a barrier

speed

of 0.75

cm/min,

so that the duration of _an

experiment

was

approximately

30 min. All measurements were made 3 times from different solutions and the

reproducibility

_was within 5 fb.

In _a

typical

17- A

isotherm,

_steep rises in pressure are associated with _a pure

phase,

while

approximately

horizontal

regions

_are associated with coexistence between _two

phases.

There

is _more than one convention for

determining

the molecular _area from such isotherms. For

example,

^{the molecular} area of 3a is

variously reported

as 160

Ji2 _[7]

_and

22512 _[6],

even

though

the measured isotherms _are

quite

similar. The difference lies in whether the molecular

area is defined

by

the inflection

point

of the vertical rise in pressure

corresponding

to the condensed

phase,

_or

by

the onset of that vertical rise. It is the former convention _to which _we adhere in this paper.

As

recently

observed

[11],

17-A isotherms in

Langmuir

films _are in

practice

often

performed

_away from

thermodynamic equilibrium.

^We and others

[2, 4, 5, 7]

have in _some

cases observed _a local maximum, _{or «}

bump

_», ⁱⁿ

spreading

pressure vs. molecular _area. Such

a feature is _an indication that the isotherm is in fact

measuring

_a

nonequilibrium

_process such

as irreversible second

layer

^formation or a

hysteretic

first order transition. With _care,

however,

the isotherms _are still useful tools for

determining

molecular surface _areas close _to

monolayer completion.

2. Results.

We will discuss in _{tum our} studies of

triphenylene derivatives,

_truxene

derivatives,

and

hexacydens.

Triphenylene

derivatives: The

liquid crystalline properties

of

alkoxy-

and

alkylthio- triphenylene

derivatives have been

extensively

studied

[12-15].

H-A isotherms of

JOURNAL DE PHYSIQUE II T 2, N'l,JANVJER lW2 5

Table I. Bulk transition temperatures and molecular areas at the water

su~fiace,

derived at the

inflection point ofthe first

state

ofthe

condensed

phase of

the

monolayer. References

_are

given for

the appearance

ofthese compounds

in the literature. Molecular _{areas are} measured

from

the

inflection points

in the

first

_« state »

of

the condensed

monolayer phase jkom

the isotherms in

figures 1,

2, 4 and 5.

Bulk

Liquid-Crystalline

Phases Reference

(I

K1 _~ I

95 .c

i

la Kl _- K2 _-

D~d

-1

[27]

17

54 .c xi .c 91 °c

16 K

- H _-

D~d

++

[13-15]

17

62c mc 91c

lc K

- D~~ _-

[27]

21

72 x ~c 941.c

id

D~d

-

[27]

53

167c

le K - D ++

[4,12]

68

69 °c 122 >c

~~ ~

6~ ^~~ 8~ ^~~~ _1/C ^~~~ _2~C

^{~~~ ~~~ ~~}

2b K ₊₊ D~d +-

ND

- D~d +-

N~

+-1

[17-23]

74

9u .c 117 c 179 °c 284 °c 297 °c

3a K _- T

[6], [7], [10]

159

97 °c ml °c

3b K ++ I

[IQ]

168

125 °c

3c

glass

_-

[10]

17

157 °c

3d K

+-

[10]

122

92 °c

3e K _+- T ++

[10]

176

58 °c 144 °c

3f K _+-

l10]

165

v6 °c

4a K

-

[10]

163

71 >c

4b K _- 155

25-61 °c

4c K ₊₊

l10]

157

118 °c

4d K _+- I

[10

12

42 °c

compounds

la-le _are shown in

figure

I. Our isotherm for le agrees within 6 fb with the

previous

determination

[4]

of _a 7-9

mN/m

_pressure and

70h2

area per molecule _at the inflection

point. Compound

ld

gives

_a similar molecular _area, with _a smaller

compressibility

in the condensed

phase.

This

supports

^the

hypothesis

of _{an «}

edge

on » arrangement ^{for both}

compounds.

The

alkylthiotriphenylenes (la-c)

lack

polarizable

atoms which would serve to

anchor the molecule _to the

surface, suggesting

that

they might

be less

likely

to form

N° DISCOGENIC MOLECULES AT AIR-WATER

~~

Q lall~R-SC,H~~

40

~

R

O ~

O

_R

10

~

0

70 60

50

lbll~R=SCJf~,

40 30

_

20

~

¹⁰

~ o

Z

E

~~

©$

lcll~R=SC~II~,

20

idX=COOICH~)sCBj

lo

R=OC,H~~

o

iell~RmOC~lI~~

loo 200

Area (A~/molecule)

Fig.

I. 1I A isotherms for several derivatives of

triphenylene.

Note the extremely small molecular

areas for the

alkylthiotriphenylenes

la-c- The low molecular _areas and

high

surface pressures indicate

that these

compounds

_are not

forming monolayers,

^but rather oily

phases

or

microcrystallites

at the surface.

Triphenylenes

id and le _are

proposed

to lie

edge-on

_to the water surface, with at least two tails

penetrating

into the

subphase.

monolayers

on water.

Indeed,

the very small measured molecular _areas for these

compounds

is smaller than the known _area for

packing

of

hydrocarbon

chains attached _to

hydrophilic

groups

(m18h~), _leading

us to conclude that these

alkylthiotriphenylenes

do not form

Langmuir

films

[16].

Truxene derivatives The

liquid crystalline properties

of the _truxene

derivatives,

hexa-n-

alkanoyloxytruxene (2a)

and

hexa-n-alkoxybenzoyloxytruxene (2b),

_are also well understood

[17-23].

The

spreading

behavior of these mesogens is shown in

figure

2. The H A

diagrams

all show the formation of

relatively incompressible monolayers

with surface pressures ^at the inflection

point

_on the order of 17.5

mN/m.

The _areas measured at the inflection

points

of the isotherms _are much smaller than would be

expected

if the molecules _were assumed to lie with

the _cores flat _on the surface

(about

154

h2

_{for 2a and 380}

h2

_{for 2b). The}

presence of

highly electronegative

_oxygen atoms to serve as

polar anchoring points

to the water surface and

relatively high

surface pressures

(m

²⁰

mN/m) coupled

with

non-negligible

molecular _areas indicate that the molecules _are not

being pulled

into the

subphase

and _are not so

hydrophobic

as to form

oily droplets

at the surface.

R R

/ '

/ R

b~~

R

2a R

= COO lClI~)uCBj

f

Z E

~

~~

30

20 2b R

= COO

@

O(CIIz)iocBj

io

°

ioo 200

Area (h~/molecule)

Fig.

2. 1I A isotherms for _two derivatives of _{truxene cores.} The low molecular _areas of 2a and 2b

are reminiscent of those of ld and le and _are consistent with _a model in which the molecules lie edge-on

to the _water surface. The larger _area of 2b _can be attributed to the presence of

bulky phenyl rings

in the

substituted tails.

The _most

plausible explanation

of the observed molecular _areas is that, like the benzene and

triphenylene

derivatives discussed

above,

the molecules lie

edge-on

to the interface.

Taking

the bulk _core-core distance of 4.5

h

_{for the}

core-core distance in the

Langmuir film,

the

corresponding long

axis of the molecular

projection

on the _water surface _must be _on the order of 15-20

h,

consistent with the transverse dimensions of the molecule. The _«

edge-on

_»

hypothesis requires

that two

hydrocarbon

tails penetrate ^the

subphase

to some extent.

Figure

3 shows _a

possible

conformation for 2a at the

surface,

and illustrates the _manner in which the

(hydrophobic)

tails that

penetrate

^the

subphase

most

likely

coil up so as to minimize their surface _area

[24].

The _cross sectional _area of the molecule shown in

figure

3 is consistent with the measured molecular _areas from

figure

2.

Fig. 3. Model of 2a at the air-water interface,

assuming

an edge-on arrangement ^{of the} truxene cores at the interface. We

postulate

that the tails

submerged

in the water adopt _a coiled _{structure to} reduce the

free energy of the

hydrophobic hydrocarbon

tails in direct _contact with _water.

Hexacyclen

derivatives _: We have studied _a

variety

of hexasubstituted

[18]-N~

_azacrowns, or

hexacyclens (Scheme1);

H-A isotherms _are shown in

figure

4 and 5.

Among

the

hexacyclen compounds

3a-3f and 4a-4d,

only

3a and 3e form

liquid crystalline phases.

The bulk

mesophase

structure of 3a is

unclear,

and has been

variously

described _as tubular-

columnar

[25]

or smectic

[10].

The interfacial behavior of this molecule,

however,

is much

more

straightforward,

and _we have

duplicated previous [6, 7]

H- A isotherms.

Compounds

3a, 3b, and 3e _are all similar _to each

other, sharing

the _same hexamide _core and

benzoyl linkage,

and

differing only

in the

length

of the attached

alkyl

^tail. Substitution of

a hexamine _core

(amounting only

to

substituting

the

carbonyl

functional group for _a

methylene linkage) produces

4a and 4b. All of these

compounds display

isotherms characteristic of a transition between _a

low-density, low-pressure

state and _a

high-density, high-pressure

state. The molecular _areas in the

low-density

state are on the order of 160

l~,

while those for the

high-density

state are either 50 _or 100

Ji~.

The transition between

the two states

generally

involves _a

non-equilibrium «bump»

_as discussed above. The

generally accepted [6, 7] interpretation

of the

low-density

state is that the

macrocycle

_core lies flat _on the water surface and the

alkyl

tails

project

_away from the _water surface _a space-

filling

model of such _an arrangement

yields

_a molecular _area in agreement ^{with that} observed.

R/)/~Q

~$ l~

N N

~~UN/~~

Q ~ ^R

RQ

3aRm©OICII~)iic~

3b R m

@ OICII~)sCB~

o

20

lo

3cRm@

11

*

^~

E

80

11

60

40 3d R

=

lCIIz)i~CB~

20

o

3eR-©OiC%)ioCl3i

3fR

"1C%)AC'&

100 200 300

Area (A~/molecule)

Fig. ^4. 1I A isotherms for derivatives of the hexamide _core 3. Note the distinct second phase at low molecular _areas for several of these

compounds,

corresponding ^either to

multilayer

formation _{or a}

more

compact ^molecular conformation.

~~

/)R

4*~*©°(~~~a)i1~3

~N _N

~§~)

R

40

20 4bRm@OICII~)~CH~

z

fi

E tQ 20

io

o

60

40

4d

20 100

Area

Fig.

5. 1I A isotherms for derivatives of the hexamine _core 4. The second phase is clearly manifest in 4a and 4b.

It is conceivable that the

high-density

_state

corresponds

to

multilayer formation;

ellipsometry

measurements could

help

resolve this

question. Altematively,

the second state may

quite likely correspond

to a conformational

change

^{in the}

monolayer.

For

example,

there may be more than _one conformation of the

macrocyde

core : an «

expanded

» conformation in which the

nitrogen

atoms are

relatively

far away from each

other,

and _{a «}

collapsed

»

conformation in which the _core

occupies

much less volume. If _{we assume a} conformation in which _a

collapsed macrocycle

^{lies flat} on the _water surface and the

alkyl

tails _are constrained to extend away from the water

surface, thereby incorporating

the known cross-sectional _area per

alkyl

chain of 18

h~,

_{then the minimum molecular}

area is

approximately

110

Ji~,

in

good

agreement ^{with the measured} areas for 3a, 3b and 4a. If _{we assume a} linear conformation of the molecule

[10],

in which three chains group

together

on either end of the molecule, we

calculate _a molecular _area of

m 55

h~,

in reasonable agreement ^with our results for 3e and 4b.

Completely removing

the

alkyl

tails leaves _us with

compounds

3c and 4c.

Compound

4c has

an isotherm with

only

_a first _state

observable,

whereas 3c has _an isotherm with _a measurable molecular _area of 17

Ji~,

smaller than that for _a

single alkyl chain,

and

certainly

too small to support

interpretation

_{as a}

monolayer.

As this

compound

is

weakly

soluble in water, the low

molecular _area for 3c _most

likely

arises from the molecules

dissolving

into the

subphase.

Finally, removing

the

phenyl ring

and associated oxygen and

leaving

the

alkyl

tail

gives

compounds 3d, 4d,

and 3f.

Compounds

3d and 4d share similar

spreading behavior,

with

comparable

areas per molecule and absence of _a second _state. The molecular _areas associated with these _two

compounds

_appear to be consistent with the

hexacyden

_core

laying

flat _on the

water surface with the

alkyl

tails

extending

_away from the

subphase.

As 3f is

structurally

similar to 3d and

4d,

one would expect to see similar

spreading

behavior.

However,

the isotherm measured

clearly displays

the two state

phenomenon

observed in

compounds

such _as 3a.

3. Conclusions.

A number of

quite

dissimilar conformations _are

possible

for

discogenic

_mesogens at the air-

water interface.

Depending

_on the details of the molecular structure,

discogenic

molecules

with flat

conjugated

_cores

(such

_as

benzene, triphenylene,

_or truxene

derivatives)

_may either

lie flat _on the ~vater surface with the tails

extending

_away from the

interface,

or

edge-on

with

two tails

penetrating

the

subphase

to some extent. In the

hexacyclen compounds,

which have

quite

flexible _cores, the molecules may ^make a transition between _a conformation with the

cores

lying

flat on the water surface and _a second state in which either the _{cores are}

collapsed

(I.e. compressed

such that the

projection

of the _{core on} the _water is _as small _as

possible)

and the

alkyl

chains _are close

packed,

_or in which the molecule is

essentially

linear

(quite

similar to the

edge-on

model for _more

rigid compounds).

De Gennes first

suggested

^{in 1971}

[26]

that

elongated

molecules at interfaces could form two-dimensional nematic

phases,

with

short-range positional

order and

quasi-long-range

orientational order.

Discogenic

molecules such _as

phthalocyanines

and benzene-hexa-

alkanoates lie with their _core groups flat _on the water

surface, projecting roughly

circular cross-sections on the water surface. These materials _are therefore _not

good

candidates for

two-dimensional nematics.

Fatty

acid molecules exhibit tilted

monolayer phases, projecting elliptical

cross-sections _on the water

surface,

but in this _case the three-dimensional aspect ^of the molecules becomes _more

important, making

them

again inappropriate

models for two-

dimensional nematics. On the other

hand,

molecules such _as the

triphenylene

and truxene alkanoates

project elongated

_cross sections _on the _water surface due

solely

to the molecular

shape.

This raises the

interesting possibility

that similar

compounds

_may exhibit two-

dimensional

liquid crystal phases.

Future progress in

understanding

the _structure of

mesogenic compounds

at the air-water interface will

clearly rely

on structural

probes

such _as

high

resolution x-ray diffraction.

Acknowledgements.

We thank R.

Fischetti,

V. Skita, S. Xu and T.

Lubensky

for useful discussions. We thank A. R. McGhie and S. H. J. Idziak for assistance with

calorimetry

measurements. This work

was

supported by

the National Science Foundation Grants DMR MRL 88-19885 and DMR 89-01219.

References

Ii GAINES G. L. Jr., Insoluble Monolayers at Liquid ^{Gas Interfaces} (Interscience, New York 1966).

[2] RONDELEz F., KOPPEL D. and SADASHIVA B. K., J.

Phys.

Fiance 43 (1982) I37I.

[3] KALINA D. W. and CRANE S. W., Thin Solid Films 134 (1985) 109 ORTHMANN E. and WEGNER G., Angew. Chem. 98 (1986) 1114.

[4] ALBRECHT O., CUMMING W., KREUDER W., LASCHEWSKY A. and RINGSDORF H., Colloid Polym.

Sci 264 (1986) 659.

[5] RONDELEz F., BARET J. F. and BoIs A. G., J. Phys. France 48 (1987) 1225.

[6] MALTHETE J., POUPINET D., VILANOVE R. and LEHN J. M., J. Chem. Soc., Chem. Commun.

(1989) 1016.

[7] MERTESDORF C. and RINGSDORF H., Liq. Cryst. 5 (1989) 1757.

[8] CHANDRASEKHAR S., SADASHIVA B. K. and SURESH K., Pramana 9 (1977) 471.

[9] TATARSKY D., BANNERJEE K. and FORD W. T., Chem. Mater. 2 (1990) 138.

j10] IDzIAK S. H. J., MALISzEWSKYJ N. C., HEINEY P. A., MCCAULEY J. P. Jr., SPRENGELER P. A.

and SMITH A. B. III, J. Am. Chem. Soc. l13 (1991) 7666.

[I ii SCHLOSSMAN M. L., SCHWARTZ D. K., PERSHAN P. S., KAWAMOTO E. K., KELLOGG G. J. and LEE S.,

Phys.

Rev. Lett. 66 (1991) 1599.

[12] LEVELUT A. M., J.

Phys.

France 40 (1979) L81 J. Chim. Phys. 80 (1983) 149.

[13] KOHNE V. B., POULES W, and PRAEFCKE K., Chem. Zeit. 108 (1983) l13.

[14] GRAMSBERGEN E. F., HOVING J. H., _DE JEU W. H., PRAEFCKE K. and KOHNE B., Liq.

Cryst.

1 (1986) 397.

[15] HEINEY P. A., FONTES E., DE JEU W. H., RIERA A., CARROLL P. and SMITH A. B. III, J.

Phys.

France 50 (1989) 461.

[16] This

interpretation

_was

strengthened by

the observation of _a

patchy

structure when crossed

polarizers

_were used _to examine _a

deposition

of

density

similar _to that _on the surface of the

trough

_{or on} the surface of _water in _a Petri dish,

suggesting

that three-dimensional

microcrystallites

_or

oily droplets

_were

floating

_on the _water surface.

[17] DESTRADE C., MALTHETE J., NGUYEN H. T. and GASPAROUX H.,

Phys.

Lett. A 78

(1980)

82.

[18] DESTRADE C., GASPAROUX H., BABEAU A., NGUYEN H. T. and MALTHETE J., Molec. Cryst. Liq.

Cryst.

67 (1981) 693.

[19] NGUYEN H. T., MALTHETE J. and DESTRADE C., J.

Phys.

France 42 (1981) L417.

[20] DESTRADE C., FOUCHER P., MALTHETE J. and NGUYEN H. T., Phys. Lett. 88 (1982) 187.

[21] FONTES E., HEINEY P. A., OHBA M., HASELTINE J. N. and SMITH A. B. III, Phys. Ret,. A37 (1988) 1329.

[22] LEE W. K., WINTNER B. A., FONTES E., HEINEY P. A., OHBA M., HASELTINE J. N. and SMITH A.

B. III, Liq. Cryst. 3 (1989) 87.

[23] WARMERDAM T., FRENKEL D. and ZIJLSTRA R. J. J., Liq. Crysi. 3 (1988) 149.

[24] Figure 3

actually

shows _{a structure} obtained via

potential

_energy minimization of _an isolated (gas

phase)

molecule, not

including

the effects of the air-water interface _or the

surrounding

molecules. The energy was minimized using the conjugate gradient method in the Dreiding force field available _on

Polygraf,

_a

product

of

Biodesign,

Inc.

[25] LEHN J. M., MALTHETE J. and LEVELUT A. M., J. Chem. Soc.. Chem. Commun. (1985) 1794.

[26] DE GENNES P. G., Farad. Symp. 5 (1971) 16.

[27] Transition temperatures _were measured in _our laboratory using Differential

Scanning

Calorimetry.