The formation of inorganic particles at ultra-thin organic films


Material Information

The formation of inorganic particles at ultra-thin organic films
Physical Description:
viii, 127 leaves : ill. ; 29 cm.
Whipps, Scott, 1968-
Publication Date:


Subjects / Keywords:
Chemistry thesis, Ph. D   ( lcsh )
Dissertations, Academic -- Chemistry -- UF   ( lcsh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1996.
Includes bibliographical references (leaves 121-126).
Statement of Responsibility:
by Scott Whipps.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 026497196
oclc - 36717790
System ID:

This item is only available as the following downloads:

Full Text







For Mom


Within a short time after joining Dan Talham's research group, I knew that
I had made the right decision. Over the last few years, Dan has taught me how
to thoughtfully approach and thoroughly investigate research problems using
any and all resources at hand. In the process he also managed to instill in me
some of his philosophy of science including, "If it didn't happen twice, it didn't
happen." In addition to having a sharp sense of humor, Dan Talham has
always been a very reasonable man who stuck by me through some very
difficult times. For that I am grateful.
Deserving thanks also are my fellow group members: Liang-Kwei Chou,
Brian Ward, Candace Seip, Missy Petruska, and Gail Fanucci. These people
with their unique personalities made the lab a much better place to be.
Grad school at UF wasn't all lab work (which helps explain why it took so
long to finish). The lasting friendships I made here will give me nothing but
good memories. Just a few of these people are Don Cameron, Brian Flynt,
Scott "Gordo" Gordon, Rob Guenard, Brendan Boyd, Jason Portmess, Paul
Whitley and Karl Zachary. I hope the Market Street Pub can stay in business
without us.
A special thank you goes to Jeff O Palko. Without his large contribution
of help and advice, the COM research would not have been nearly as good nor
as enjoyable. I would also like to thank Dr. Saeed Khan for the sharing of his
lab and his insights.

Dr. Augusto Morrone and Eric Lambers at the Major Analytical
Instrumentation Center need to be thanked for allowing me to use their TEM

and XPS instruments as do Professor Randy Duran and Tim Herod for the use

of and help with the BAM experiment.
I'm thankful that my family has always been supportive of me whether or

not they understood exactly what I was doing down in Florida. My father,
Richard, and brothers Greg, Larry and Ted have shaped me to the person I now
Lastly and most importantly, I would like to thank my mother, Cathleen, to

whom this dissertation is dedicated. In addition to a million other things to thank
her for, she made me believe in myself and realize that any goal is attainable as

long as you keep trying.


ACKNOWLEDGMENTS ............................. .... ..............................................iii

ABSTRACT ........................................................... .........................................vii


1. INTRODUCTION.................................................................................................. 1

Scope of Dissertation.................................................................................... 1
Previous Work in the Talham Research Group........................................ 2
Biomineralization ................................................................................................. 5
Vehicles ............................................................................................................ 7
Langmuir-Blodgett Films ...................................... .................................... 10

POLYMERIZED LANGMUIR-BLODGETT TEMPLATE.................................... 16

Introduction ............................................................. .................................. 16
Experimental.............................................................................................. 22
Materials .................................................. ................. ........................ 22
Substrate Preparation................................ ... ................... 23
Instrumentation............................... ......... .......................................24
Procedure....................................................... ....25
Results/Discussion .............................................. ................................. 26
Characterization of Diynoic Acid Template................................. ....26
Brewster Angle Microscopy ..................................................................28
UV-Vis Spectroscopy.............................. ...... .................33
Attenuated Total Reflectance IR.................................................33
X-ray Photoelectron Spectroscopy................................................ 39
Transmission Electron Microscopy/Diffraction ...............................43
C conclusion ..................................... ...................................... ......................5 8


Introduction ................................................................ ................ 59
Experimental.............................................. .............................................62
M materials ................................................ .................... .......................62

Instrum entation................................................. .. ............................ 62
Procedure.............................. ................. ........63
Results/Discussion ............................................. .................................. 64
Fourier Transform Infrared Spectroscopy .............................................64
X-ray Photoelectron Spectroscopy............................ ............ ....68
Transmission Electron Microscopy/Diffraction ....................................72
C conclusion ............................. ........... ............................. 7


Introduction ....................................... .................. ...............79
Experimental.......... ........... .... ............ .... ..... ........... ..........84
M aterials....................... ............... ... ...... .. ................... 84
Instrumentation.............................. ... ............. .................84
Substrate Preparation............................. ........................ .....85
Methods............................ -.......... .. ....85
Results/Discussion .............. ... ........ ...... ... ................. ....87
Monolayer Studies ........................................ .................. ... 87
Blank Studies ........................................ ..... ........ ....89
ATR-IR Experiments................................ ..............................89
Brewster Angle Microscopy......... .............................91
Transmission Electron Diffraction...................................... 98
Scanning Electron Microscopy ................................. .....102
Conclusion........................ ........................... 117

5. CONCLUSIONS, PERSPECTIVE AND FUTURE WORK...........................118

LIST OF REFERENCES ................................. .. .......................... .........121

BIOGRAPHICAL SKETCH........................... .... ....................127

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy




December, 1996

Chairman: Dr. Daniel R. Talham
Major Department: Chemistry

Particles of cadmium iodide were formed at Langmuir-Blodgett films by
reacting the cadmium-containing film with HI gas. In one case a polymerized
diynoic acid film was used as the template for particle growth. The films were
characterized by attenuated total reflectance Fourier-transform infrared
spectroscopy (ATR-FTIR) and ultraviolet-visible spectroscopy (UV-Vis). ATR-
FTIR was also used to follow the reaction that formed cadmium iodide particles.
X-ray photoelectron spectroscopy (XPS) and transmission electron
microscopy/diffraction (TEM/TED) techniques, respectively, identified the
particles as being Cdl2 and determined crystal orientation with respect to the
template. It was found that the polymeric diynoic acid template was able to
guide the formation of many large crystals with the same orientation [(001) face
parallel to film plane] over areas 2.6 pm in diameter. This polymer template did
not, however, appear to limit diffusion of Cdl2 between inorganic layers any

better than a monomeric template. It was suggested that this was due to the
template molecule's weak van der Waals forces and the possibility of boundary
defects being present around its many small domains.
In a similar study, Langmuir-Blodgett films of octadecylxanthate were
also used as a template to grow Cdl2 particles. The xanthate head group was
chosen for its potential to interact favorably with iodide layers in the crystal. The
xanthate template did exert some control over particles formed, as evidenced by
TEM data, where overlapping crystals from different layers in a sample had the
same crystallographic orientation over small areas less than 1 Am in diameter.
Overall, however, the xanthate template was unsuccessful in controlling crystal
orientation. This was due to the template decomposing after the reaction with
HI, and giving off CS2 head groups, as seen by an XPS experiment.
The formation of kidney stones at biological membranes was effectively
modeled using monolayers of phospholipids held upon supersaturated
solutions of calcium oxalate monohydrate (COM the main component of
kidney stones). Various lipid head groups were studied and it was found that
the nucleation process appears to be governed largely by electrostatics. The
head groups with more anionic character formed more COM crystals per unit
area. In addition, the COM crystal face which nucleated predominantly at the
monolayer was the (10T), a calcium-rich face.



Scope of Dissertation

This dissertation is concerned with the formation of inorganic particles at
organic Langmuir-Blodgett films. The focal point of this work has been to study
the role of a thin film in directing crystal growth, i.e. controlling the crystal's size,
shape and orientation. In this way it may be possible to engineer materials at
the molecular level, having possible applications in semiconductor as well as
optical and magnetic devices, to name a few. Chapter 1 will summarize
previous work done in the Talham group on particle formation at Langmuir-
Blodgett (LB) film templates as well as provide a literature review on the subject
of biomineralization. Chapter 2 describes the formation and characterization of
oriented cadmium iodide particles within polymerized diynoic acid LB films. In
this project, a polymerized LB film was used in order to limit diffusion of
cadmium iodide between layers during particle formation, thereby inducing
thinner particles. Chapter 3 is also about Cdl2 particles, but this time they are
being formed at an octadecylxanthate LB template. The xanthate head group
(-OCS2-) was chosen for having a potentially favorable interaction with iodide,
where the aim was again to make a more 2-dimensional crystal. Chapter 4
describes the nucleation and growth of calcium oxalate monohydrate (COM)
crystals underneath floating phospholipid monolayers. Since calcium oxalate is
a major component of kidney stones, the lipid systems studied can serve as a

model for kidney stone growth at renal membranes. Various lipids were studied
and compared for their ability to form COM. The final chapter presents
conclusions and a perspective of the body of work as well as gives possibilities
for future work.

Previous Work in the Talham Research Group

Within the Talham group, previous work conducted by Pike investigated
the role of an organic Langmuir-Blodgett (LB) film in directing the growth of
inorganic particles."3 LB films of cadmium arachidate (Cd-arac) were
transferred as Y-type bilayers (where the amphiphiles are aligned hydrophilic
head to head, hydrophobic tail to tail, and Cd2+ is bound in between layers of
anionic head groups, as depicted in Figure 1-1) to a solid support. Upon
exposing the LB film to HX gas, where X = I, Br, or CI, particles of the
corresponding CdX2 are formed within the film. Investigation of the particles
was carried out using transmission electron microscopy/diffraction to see how
the particles were oriented in the sample, and to learn what effect the LB film
had on controlling the particle's size, shape and orientation.
Transmission electron diffraction patterns showed that the Cdl2 particles
formed within Cd-arac films were oriented exclusively with their (001) face
parallel to the LB film plane. In this orientation, the metal halide layers were
parallel to the LB layers. This preferred orientation can be explained by the
Cd-arac film's ability to organize Cd2+ ions into an arrangement similar to that
found in the (001) face of cadmium iodide, resulting in metal halide planes
being parallel to LB layers upon crystal growth.

long-chain tail

Cd Cd Cd Cd Cd Cd Cd CdCd head group

Solid Support

Figure 1-1 Cartoon depiction of a Y-type cadmium arachidate

When an area of diffraction was used that included many crystals, the
diffraction patterns obtained were similar to that of a single crystal, where
discrete spots are seen, as opposed to polycrystalline rings. This indicated
there were areas of up to 2.6 mm diameter where the in-plane crystal
orientations were the same. This uniform in-plane orientation of Cdl2 particles
suggested a lattice match between the cadmium iodide and cadmium
arachidate LB film. The lattice match may have been responsible for the
directed growth of particles. A nearly exact match exists between the Cd-arac
LB film and the (001) face of Cdl2 having only a 0.5% lattice mismatch.
In contrast to the Cdl2 system, particles of CdBr2 formed in cadmium
arachidate displayed two distinct particle shapes and orientations. One type of
particle had a round shape and was oriented with its (001) face parallel to the
film, while the other particle type was needle-like in shape and was oriented
with the (001) face perpendicular to the LB film plane. The (001) face of CdBr2
has a 7% lattice mismatch with the Cd-arac LB film, which is much larger than
that seen in the cadmium iodide system. This may have been the reason for the
larger number of allowed orientations in the CdBr2 system. The needle-like
CdBr2 particles were found to exist in three preferred orientations that were
rotated by 1200 with respect to one another. Pike speculated that these
orientations could have been a consequence of the in-plane hexagonal close-
packing of the LB film.1
When Cd-arac films were exposed to HCI to form CdCl2, the resulting
particles were also arranged in two distinct orientations, having their (001) faces
parallel or perpendicular to the LB film plane as given by their diffraction
pattems. This result was similar to that obtained in the CdBr2 system. However,
unlike CdBr2, the presence of two types of CdC12 particles was not detectable
from the electron micrographs.

Manganese halide particles were also investigated by reacting the

appropriate hydrohalide gas with LB films of manganese arachidate. In
general, the MnX2 particles were smaller than their Cd counterparts. In contrast
to the cadmium halides, manganese halides did not produce single crystal
diffraction patterns when analyzing many crystals at once. Rather, diffuse rings
were obtained. This result indicates that the particles did not share a common
in-plane orientation. For all three manganese halide particle systems, it was
found that the degree of particle orientation was not as high as that seen in the
cadmium halides.
These studies showed that an organized LB film could be used to direct
the growth of inorganic particles. In the case of the cadmium halides, the
complementarity of the layered CdX2 structure and the layered LB film may
have played a role in orienting the particles. In Chapters 2 and 3, we will
explore the effects of changing the organic LB template. A polymerizable
diynoic acid template will be investigated in Chapter 2 in order to see the effect
a polymerized template has upon particle formation. Chapter 3 will investigate
the effects of using a different head group, a xanthate, has upon the formation of
inorganic particles.


Biomineralization refers to the processes by which living organisms form
minerals. This phenomenon is widespread, with organisms from all five
kingdoms and 55 different phyla forming over 60 known minerals through this
process.4 Approximately half of these minerals contain calcium and most
commonly serve as structural support in skeletons or in teeth. The remarkable

control nature exerts over the crystal's size, shape and crystallographic
orientation, as well as the properties of the resulting materials such as high
strength, resistance to fracture, and esthetic value serve as goals for synthetic
materials scientists to reproduce. Biomimetic chemistry5 is the field of science
that attempts to improve synthetic material design by adapting biological
mechanisms that control the formation of minerals at the molecular level.
Nature uses three key points in controlling how inorganic particles are
deposited within organic polymer matrices to give organized crystalline arrays.6
These are ion concentration regulation at the organic matrix/solution interface,
growth modification by soluble molecules within the organic matrix, and
chemical mediation of crystal nucleation and growth at the organic/inorganic
interface by specific interactions at the molecular level. A transfer of chemical
information from the bio-organic matrix to the forming inorganic crystal faces
can result in a controlled or regulated crystallization. This allows forming nuclei
to be stabilized and crystal growth to be directed.
One of the best understood examples of controlled biomineralization can
be seen in the well-studied case of shell formation in molluscs.7 The mollusc
(Nautilus repertus) forms its shell of aragonite, a form of CaCO3, upon a thin
organic sheet of protein containing acidic polypeptides in the b-pleated sheet
conformation. Each CaCO3 crystal is oriented with its ab face parallel to the
organic sheet. There is a structural correspondence, or epitaxy, between the b-
pleated organic layer and the Ca-Ca distances in the ab plane of the overlying
crystal lattice. Negatively charged aspartic acid residues attached to the b-
pleated sheet every 4.96 A are closely matched with the Ca-Ca distance of 4.7
A along the a axis of aragonite.8 This close geometric correspondence
between protein sheet and crystal lattice allows for an organized binding of
calcium which lowers the activation energy for nucleation and orients the nuclei

along a preferred direction of growth. Epitaxially controlled crystal growth in
organisms, though often suspected, is seldom proven due to the difficulty of
characterizing the complex biopolymer matrices where nucleation occurs.
Addadi et al. proposed a model for nucleation and crystal growth for the
mollusc shell9 where sulfate and carboxylate groups on the b-protein sheet act
cooperatively in oriented CaCO3 nucleation. In this model, sulfates concentrate
calcium ions without permanently binding them to create a localized area of
supersaturation. The high concentration of Ca2+ is then used by the structured
domains of carboxylates (from aspartic acid) to nucleate CaCO3 from its (001)
face. The authors created an artificial model substrate of sulfonated polystyrene
films absorbed with poly(aspartate). It was found that sulfate-bound substrates
did concentrate calcium as seen by 45Ca scintillation counting, and that the
combined presence of sulfate and carboxylate groups increased the number of
oriented CaCO3 crystals by a factor of three.9

In addition to the functionalized protein sheet discussed above, another
type of organic assembly that nature employs to control mineral formation is the
micellular vesicle.10 Micellular vesicles, usually composed of lipids, are used
by the organism to compartmentalize a small volume prior to biomineralization.
The vesicle can maintain the supersaturation, redox and pH gradients
necessary to control crystal nucleation and growth. In addition, vesicle size and
shape can affect the size and shape of the crystals forming within.
An example of biomineralization within vesicles is demonstrated by
magnetotactic bacteria. These aquatic bacteria (several species exist) react to
the earth's magnetic field in order to direct themselves toward suitable
environments. A row of discrete magnetic crystals made of either magnetite11

(Fe304) or greigite12 (Fe3S4) allows these bacteria to sense changes in the
earth's magnetic field. These magnetic particles are single domain, membrane-
bound crystals that have uniform shape and a narrow size distribution (50 90
nm). The organic membrane that surrounds the crystal plays a central role in
the process by providing an enclosed, microscopic regulated area for the
crystal formation (a solid state reaction) to occur. Some species of bacteria
formed crystals which displayed unique morphologies that could not be
reproduced synthetically. This indicated to the authors that some sort of direct
biological intervention took place such as the addition of a face specific crystal
growth inhibitor.12
Phospholipid vesicles lend themselves readily as model systems for
biomineralization. By forming the vesicles in a solution containing a species of
interest, the species can be encapsulated within the vesicle and is then held
captive for subsequent solid state reactions with membrane-permeable species
such as OH- and H2S. Restricting a chemical reaction to the vesicle interior
enables control to be exerted upon supersaturation levels, stereochemical
requirements for ion binding and nucleation, and the spatial organization of
crystal growth and morphology.
In one study, the effects of vesicular compartmentalization on the reaction
of Fe3+ and Fe2+ with OH- were investigated.13 The precipitation of Fe(ll) and
Fe(lll) oxides inside the vesicles was accomplished by raising the pH of the
extravesicular solution. This caused OH- to diffuse into the vessicle where the
desired reaction could occur. Iron oxide formed inside the 300 A diameter
phospholipid vesicles of phosphatidylcholine differed in structure, morphology,
and size compared to precipitates formed in bulk solution. For example,
intravesicular solutions of Fe(ll) produced Fe304 crystals 30-50 A in diameter
upon reaction with hydroxide. In contrast, bulk oxidations produced mixtures of

a- and g-FeOOH. The authors attributed the observed differences between
vesicle and bulk reactions to the vesicle membrane's ability to regulate the rate
of OH- diffusion into the intravesicular solution, thereby regulating the rate of
precipitation. Another difference was that all particles formed within vesicles
were either spherical or disk-shaped, whereas bulk precipitates were not
rounded. It was presumed by the authors that the initial stages of crystal growth
within vesicles were partially constrained by the curvature of the vesicle
membrane, resulting in rounded crystal morphologies.13 The influence of the
curved organic substrate in these crystallization reactions is readily apparent.
The membrane surface acts as a charged template for ion localization and
crystal formation. Also, the hydrophobic membrane acts as a barrier to the
external solution, limiting diffusion into the reaction site while the shape and
dimensions of the vesicles impart spatial restrictions on crystal development.10
These factors result in changes of crystal structure, size, and morphology
compared to the same reactions carried out in the absence of vesicles.
Surfactant vesicles have been used to produce semiconductor particles
of CdS and ZnS.14 The main point of using vesicles in this work was to limit the
size of the particles formed, which have potential use in the areas of molecular
electronics and catalysis. Semiconductor particles formed using
dihexadecylphosphate measured 2 8.5 nm in diameter and were found by
electron microscopy to nucleate at the phosphate head groups on the vesicle
surface. The authors stated that the discrete nuclei, spatially organized and
isolated on the vesicle surface, would not be as likely to aggregate as would be
nuclei in free solution, and that this may be why particle sizes were restricted
below 10 nm.14
One drawback of vesicles is that they are rather delicate and cannot
withstand extreme reaction conditions such as high temperature. This is not

surprising for these organic assemblies since weak hydrophobic van der Waals
forces are what hold the vesicle together. One study found a more robust
alternative to surfactant vesicles in the use of a biological molecule: the cage-
shaped iron-storage protein ferritin.15 Ferritin was used to make nanometer
sized FeS particles by reacting the native iron oxide in the ferritin
(5Fe2O3-9H20) with H2S gas. Oxides of manganese and uranium were also
formed in the ferritin cage once the native iron oxide was removed using
dialysis techniques. However, electron diffraction analysis showed these
oxides to be amorphous and their identities were not determined.

Langmuir-Blodgett Films
Another route to obtaining an organized assembly of organic molecules
which can be used as nucleation sites for studying biomineralization and
biomimetic chemistry is the Langmuir-Blodgett (LB) film. LB films serve as a
versatile model for biomineralization. In these experiments, a monolayer film of
an amphiphilic molecule is compressed and held upon a subphase solution
that is supersaturated with the inorganic species of interest. Crystal formation
may then occur at the solution/monolayer interface where the surface of
hydrophilic headgroups are located. LB films can be tailored in such a way as
to help regulate crystal growth in a desired manner. By changing the head
group of the LB film, for example, from a sulfate to a carboxylic acid, one can
control the chemical nature of the sites where crystal nucleation and growth
occur. In addition, by regulating the degree of film compression, one can
control the organization and packing of molecules. Most importantly, this allows
for control over the spacing of head groups, where ion concentration and
nucleation occur.

In one of the early studies of crystallization at LB monolayers, it was
determined that structural information could be transferred from the monolayer
to the crystals growing underneath1617 using a chiral LB system. In this work,
Landau et al. made monolayer films from chiral a-amino acid amphiphiles that
terminated with either R- or S-glycyl head groups. The subphase was a
supersaturated solution of racemic glycine. When an R-monolayer was
compressed, the head groups simulated an R-layer of a glycine crystal exposed
at its (010) face. This R surface gave a stereochemical match for S-glycine
molecules to come out of solution and nucleate. Conversely, when using an S-
monolayer, the head groups simulate the (010) glycine face and pull R-
enantiomers out of solution to form oriented crystals. In each case, structural
information from the monolayer was transferred to the crystal in determining
which crystal face of glycine formed at the interface. The authors illustrated the
importance of head group spacing by varying it through the incorporation of
perfluorinated or cholestanoyl moites which decreased and increased head
group spacing, respectively. When head group spacing was reduced, some
control over orientation was lost, as crystals nucleated at both the (010) and
(010) faces for a given enantiomeric monolayer. Increasing the distance
between head groups nearly eliminated crystallization at the
monolayer/solution interface altogether.16 In a similar but more thorough study
of glycine crystallization under a-amino acids that used grazing angle x-ray
diffraction to determine structure of the floating monolayers,'1 Landau et al.
concluded that the packing of the head groups determines nucleation rates and
the degree of orientation of the attached growing crystals.
The Landau group has also investigated the growth of sodium chloride
crystals beneath monolayers of long chain carboxylic acids such as stearic and
arachidic acid.17 In these studies the crystal nucleated from a crystal face that

does not occur in nature. However, monolayers were found to stabilize crystal
faces that do not occur naturally. The nucleation of NaCI is less specific than
glycine since the structure of the monolayer cannot simulate the first layer of the
crystal face as seen with the chiral glycine study discussed above. The
nucleation of sodium chloride was primarily due to electrostatic interactions
between Na+ ions and monolayer head groups. In contrast to bulk solution
precipitation of NaCI, where cubic crystals showing their (100) faces result,
NaCI grown under monolayers of stearic and arachidic acid gives sodium
chloride attached to the monolayer at the (111) face. This face is comprised
entirely of sodium ions. By concentrating a layer of sodium ions underneath the
layer of negatively charged carboxylate groups, nucleation of the (111) face
may be induced.17
Mann et al. have studied the controlled nucleation and growth of CaCO3
beneath monolayers of stearic acid,19,20 octadecylamine,21,22 eicosyl sulfate
and eicosyl phosphonate.23 In these experiments, a LB monolayer was
compressed and held upon a supersaturated calcium carbonate subphase
In control experiments, where crystals precipitated in the absence of a
monolayer, CaCO3 was in the form of calcite crystals located around the sides
and bottom of the LB trough. However, when the supersaturated solution was
underneath an organized stearic acid monolayer, the vaterite form of CaCO3
formed exclusively at the monolayer/solution interface, oriented with the (001)
faces parallel to the monolayer plane.19 Compression isotherm data indicated
to the authors that Ca2+ was incorporating into the head groups of the film to
form a layer of calcium which could mimic a crystal face.20 This electrostatic
step alone could not explain the selective formation of vaterite over calcite,
since the (001) face of both forms is comprised entirely of Ca2+ ions. It was

necessary to investigate the stereochemistry of the forming face. In calcite, the
carbonate anions are oriented parallel to the (001) face, whereas in vaterite
they are perpendicular. This latter arrangement is equivalent to the orientation
of carboxylate head groups in the monolayer. The stereochemical arrangement
of carboxylate groups in the LB film mimics the (001) face of vaterite and
therefore favors vaterite nucleation over that of calcite.
The observed selectivity with which a preferred crystal face nucleates at
the monolayer is related to a stereochemical complementarity between the
oxygen atoms of the surfactant's head group and the oxygen atoms of
carbonate ions in the nucleating crystal face. Monolayers having sulfate and
phosphonate head groups, when compressed on supersaturated calcium
bicarbonate solutions, promoted the nucleation of calcite at the (001) face.23
Electrostatics and symmetry also play key roles in crystal nucleation with the
negatively charged sulfate and phosphonate films. The pseudohexagonal
close packing of both the sulfate and phosphonate monolayers is similar in
symmetry to the hexagonal array of calcium ions in the (001) face of calcite. In
this way, calcium bound to the negatively charged monolayers will create a
surface of calcium ions which mimics the calcite (001) face.23
In contrast to stearic acid monolayers, which produced vaterite
nucleated exclusively from the (001) face, nucleation of CaCO3 under a
positively charged monolayer of octadecylamine produced vaterite nucleated at
both the (001) and (110) faces in equal amounts.21 Due to the positively
charged head group, the binding of Ca2+ to the monolayer is not possible. It
then follows that a layer of Ca2+ that mimics a crystal face could not be formed,
as observed with the carboxylate monolayers. The authors speculated that a
different mechanism was responsible for the nucleation. Bicarbonate anions
from the subphase solution could act as bidentate ligands, binding to the -NH3+

headgroups. This may provide a stereochemical recognition between
carbonates organized within the boundary layer and carbonates in the
nucleating faces of vaterite.22
The Mann group has also studied the nucleation of BaSO4 underneath
monolayers of long chain sulfates,24 carboxylates 25 and phosphonates.26
Oriented barium sulfate nucleated underneath eicosylsulfate monolayers
exclusively with the (100) face parallel to the monolayer. However, a different
crystal face, the (001) face of BaSO4, had a good potential lattice match with a
hexagonal array of sulfate headgroups. The fact that the (001) face was not
nucleated showed that an epitaxial relationship was not the primary factor in
determining crystal orientation. The authors concluded that monolayer-induced
oriented nucleation depended on the organic film's ability to mimic both the
lattice geometry of cations and the stereochemistry of oxyanions in the
nucleating crystal face. At least two recognition processes are in effect at the
monolayer/inorganic interface. There is a geometric matching between close-
packed head groups and Ba-Ba distances in the crystal lattice as well as a
stereochemical complementarity between the sulfate head groups of the
monolayer and the sulfate anions of the nucleating crystal face.
Using a technique other than precipitation to grow inorganic particles at
an organic template, Langmuir-Blodgett films of arachidic acid have been
reported to induce the epitaxial growth of PbS crystals.27 In these experiments,
lead sulfide crystals were formed by forming a LB film of lead arachidate on the
trough and then exposing it to H2S gas. The uniformly sized, equilateral
triangle-shaped crystals had their [111] axes perpendicular to the monolayer
and their [112], [121], and [211] axes parallel to the monolayer, arranged in 3-
fold symmetry at 1200 angles. In control experiments where no monolayer was
present, only irregularly shaped, non-oriented PbS crystals formed. The growth


of PbS under arachidic acid monolayers was believed to be epitaxial in nature
by the authors due to a nearly perfect lattice match between the nucleating
(111) face of PbS and the (100) plane of the hexagonally close-packed
In Chapter 4, we will use LB films of lipids to model biological
membranes for the formation of calcium oxalate monohydrate in experiments
similar to those performed by Mann et al. described above. This project has
potential biological relevance, since calcium oxalate monohydrate is a major
component of kidney stones, and its nucleation and growth at lipid monolayers
is analogous to stone formation in the kidneys.




An objective of this work has been to prepare single layers of an
extended inorganic lattice. Once such a system was achieved, it would allow
for the study of properties such as magnetism and conductivity within the two-
dimensional limit of a monolayer. The strategy used in our group is to take an
organic Langmuir-Blodgett film containing metal ions, for example, cadmium
arachidate, and react it with a hydrogen halide gas to produce metal halides
within the film.1-3'28 The layered inorganic compound Cdl2 was chosen for
study since its layered nature may favor the formation of an extended
monolayer over bulk crystals within the LB films. In addition to acting as a
template for controlling inorganic crystal growth, it was hoped that the LB film of
arachidic acid would act as a barrier against the diffusion of Cdl2 between
layers which would lead to bulk particles. It was found that the arachidic acid
template was successful in forming Cdl2 particles which were exclusively
oriented with their 001 crystal faces parallel to the layer plane. However, this
template produced only particles; no evidence for the formation of an extended
inorganic lattice monolayer was observed.1
The focus of the current work is to use a LB template that may better limit
diffusion between inorganic layers. For this reason a polymerizeable,

diacetylenic acid, 10,12-tricosadiynoic acid, was chosen. This diacetylenic
acid has been shown to form a stable film on the Langmiur trough.29,30 Once
polymerized by UV radiation, the resulting close-knit polymer film may
effectively prevent or limit the diffusion of Cdl2 between layers when reacting
the film with HI gas.
In studies by other groups, polymerized LB films of tricosa-, pentacosa-,
and heptacosadiynoic acid have been investigated as photolithographic
resists.3133 The resist films were evaluated by the resolution obtained (e.g. the
minimum distance between two separate photolithographic features) after
exposure by excimer laser, x-ray, and electron beam. Ogawa reported that films
of pentacosadiynoic acid, which were transferred from Ca2+ containing
subphases, performed best as resists, having a resolution of 0.3 mm.32 The
authors believed that the resolution was limited by the lens and mask used in
the experiments, and, therefore, higher resolutions were possible with the
diynoic acid films.
Diynoic acids as well as a few of their derivatives have been investigated
for their nonlinear optical properties.3 In one study,35 a new optical device was
proposed based upon a reversible color change of the polymer film when it is
annealed for 10 min. at 600 C. The proposed device could operate as an
optical switch. A probe light controls the color phase of the film which in turn
determines whether or not the input light is allowed to pass through the film.
The polymerization of 10,12-tricosadiynoic acid from its cadmium salt in
monolayer films was first described by Tieke.36 Exposure to UV radiation
results in the formation of a linear, one-dimensional polymer backbone that has
conjugated double and triple bonds, as depicted in Figure 2-1. The polymer
films have been characterized by infrared spectroscopy and electron

C' c CC Uv

mo nomer


S C Cpolymer C


Figure 2-1. Polymerization scheme for a diynoic acid

Diynoic acid films undergo color changes from a colorless monomer to a
blue polymer after approximately 30 min. exposure to UV light. Further UV
exposure or heating results in a change of color from blue to a bright red color.
A phase change in the film from a monoclinic to an orthorhombic arrangement
of the alkyl chains is generally accepted as the reason for the thermochromic
behavior of the film. Rearrangement of alkyl chains in these films stresses the
conjugated polymer backbone and results in a shift of the p and p' electronic
energy levels.37
An improvement in the layered order of polydiacetylene films was
observed upon pre-annealing them before polymerization.38'41 After annealing
for 10 20 hr at 42 700 C, LB films of 10,12 tricosadiynoic acid (Cd salt) gave
more intense x-ray diffraction peaks as well as more orders of diffraction
compared to films that were not annealed.38'39 This indicated to the authors a
growth in domain size within the film. In another study on the effects of pre-
annealing,41 polymerized LB films without a pre-anneal step had absorption
maxima at 640 nm, while those with a pre-anneal step had their maxima at 704
nm. The authors concluded that the extremely ordered structure of the pre-
annealed films (as given by x-ray data) caused the red shift in absorption peaks.
Domain size within the polydiacetylene films is an important issue when
trying to form an extended inorganic lattice within them. Large, uniform
domains are desirable for this purpose. An electron diffraction study revealed
that a polydiacetylene film consists of crystalline domains.42 Individual domains
in adjacent layers were not in registry with one another and the cadmium ions
were not located in fixed sites, presumably due to the inclusion of water which
partially solvated the Cd2+ ions in the interlayer regions.42 The polymer
backbones grew within the layer plane, restricted in their length by the domain
size, which averaged from 3 100 mm in diameter.43 Tieke investigated some

factors which affect the domain size. Choice of spreading solvent, subphase
temperature, surface pressure of monolayers, the use of mixed monolayers, and
annealing all play a role in domain size.43 Studies were performed using
polarizing microscopy. It was found that aromatic spreading solvents increased
average domain size by a factor of 10. A subphase temperature of 15 200 C,
and a surface pressure above 15 mN/m during film transfer also increased
domain size. In contrast to results from x-ray studies,38,39 Tieke reported that
annealing the films disordered some molecules and rendered them
unpolymerizeable. This led to reduced domain sizes.
Another study44 used polymerized films of 10,12-nonacosadiynoic acid to
grow quantum-state (Q-state) particles of CdS, CdSe, CdTe, and CdSxSex-l.
The particles were formed by exposing the cadmium-containing film to H2X gas,
where X = S, Se, or Te. Films were analyzed by UV-vis spectroscopy. Neither
forming an inorganic monolayer nor controlling particle orientation was the goal
of the authors, but it was rather to form particles as small as possible which had
shifts in their optical bandgap compared to the bulk materials and posessed
non-linear optical properties. The resulting particles ranged in size from 40 50
A. Smaller particle size was obtained by using mixed monolayer films of
dihexadecylphosphate and the diynoic acid. The reduction in particle size was
attributed to the dihexadecylphosphate causing a reduction in the domain size
of the films, where particle nucleation and growth occurs.4
In contrast to the previous study, where producing smaller particles is
desirable, the work of the current project which will be presented in following
sections, aimed to produce large, thin particles approaching monolayer
thickness. A cartoon depiction of a cadmium containing diynoic acid bilayer is
given in Figure 2-2 as an overview of the steps used to form cadmium iodide
particles within the LB film template. Results from this new work show that the


1. Pre-anneal

2. Photopolymerize


Cd 2+
Polymer Template

3. HI(g)

Figure 2-2 Cartoon depicting the formation of cadmium iodide particles at a
polymeric template

Cd 2+



polymer diynoic acid template produced large Cdl2 crystals that were singly
oriented across areas up to 2.6 mm in size. These results are similar to those
obtained with the monomeric arachidic acid template.1,2 The polymeric diynoic
acid template was not able to produce a single layer, extended lattice of Cdl2 as
desired, at least in part due to the its inability to limit diffusion between inorganic
layers. This inability to limit diffusion was attributed to boundaries between
domains in the film serving as avenues for metal ions to diffuse between layers,
thus forming bulk Cdl2 particles rather than a single layer.


The polymerizeable 10,12-tricosadiynoic acid (CH3(CH2)9C:::C-
C:::C(CH2)8COOH, 98%) was obtained from TCI America (Tokyo, Japan) and
stored at 00 C away from light until use. Octadecyltrichlorosilane and cadmium
chloride hemiheptahydrate (CdCl2 2.5 H20, 98%) were obtained from Aldrich
Chemical Company (Milwaukee, WI) and stored under N2 until use. Hydrogen
iodide gas (98%) was purchased from Matheson Gas Products (Atlanta, GA)
and used as received. Spectrograde chloroform and hexadecane from Aldrich
were used as received. Titanium 300 mesh transmission electron microscope
grids were purchased from the Ted Pella Company (Redding, CA) as was the
formvar resin used to coat TEM grids. Silicon monoxide (325 mesh powder)
was obtained from Aldrich. Single crystal (100) n- and p- type silicon wafers
were purchased from the Semiconductor Processing Company (Boston, MA)
and were cut into 20 x 15 x 0.5 mm pieces for use as XPS substrates. For all

FTIR measurements, attenuated total reflectance (ATR) silicon crystals, 50 x 10
x 3 mm parallelograms, were purchased from Wilmad Glass Company (Buena,
NJ). The water used in LB experiments was purified using a Sybron/Bamstead
Nanopure system (Boston, MA) and had an average resistivity of 18 megaohm-

Substrate Preparation

UV-Vis experiments used standard 75 x 25 x 1 mm glass microscope
slides that were cleaned using the RCA method.45 In the RCA procedure,
substrates were first immersed in a 5:1:1 by volume mixture of water, 30%
hydrogen peroxide, and concentrated ammonium hydroxide for 10 min. at 700
C. Substrates were then immersed in a 6:1:1 mixture of water, 30% hydrogen
peroxide, and concentrated HCI for 10 min. at 700 C followed by a thorough
rinsing with water. The substrates were then rendered hydrophobic by self-
assembling octadecyltrichlorosilane46 (OTS) as follows. Substrates were
immersed in a 2% by volume solution of OTS in hexadecane for 1-3 hr. They
were then rinsed for 30 min. with chloroform in a Soxhlett extractor.
The silicon substrates used in FTIR and XPS experiments were first
cleaned in a RF plasma etch by placing the substrate in an argon plasma for 15
min. on each side. The substrates were then cleaned following the RCA
procecure and made hydrophobic through the self-assembly of OTS as
described above.
The substrates used in TEM/TED experiments consisted of TEM grids on
top of a microscope slide, covered by about 600 A of Formvar resin. An electron
beam evaporator was used to deposit 300 A of SiO on top of the Formvar

coated grids, enabling the substrate to then be self-assembled with OTS. In the
OTS procedure, hexane was substituted for chloroform in the rinse step.


LB film deposition was carried out using a KSV Instruments (Stratford,
CT) KSV 5000 LB system equipped with a home built double barrier teflon
trough. The surface pressure was measured by a platinum Wilhelmy plate
suspended from a KSV microbalance.
UV-Vis spectra consisted of 20 scans and were recorded using a
Hewlett-Packard 8452A diode array spectrophotometer. Brewster angle
microscopy was carried out using a BAM1 Brewster angle microscope by
Nanofilm Technology (Gottingen, Germany). The instrument was equipped with
a video camera and recorder which allowed for the acquisition of data in real
Fourier transform infrared spectra were taken with a Mattson Instruments
(Madison, WI) Research Series-1 FTIR spectrometer using a narrow-band
mercury cadmium telluride detector. A Harrick (Ossining, NY) TMP stage was
used for attenuated total reflectance (ATR) experiments. Spectra consisted of
1000 scans at 2 cm-1 resolution and in each case were referenced to the clean
OTS-coated silicon ATR crystal.
Transmission electron microscopy/diffraction was carried out on a JEOL
(Peabody, MA) JEM 200CX electron microscope. The microscope was
operated at an accelerating voltage of 100 kV and beam exposure to the
sample was kept to a minimum to avoid sample degradation.
X-ray photoelectron spectra were obtained using a Perkin-Elmer (Eden
Prairie, MN) PHI 5000 Series spectrometer. All spectra were taken using the

Mg Ka line source at 1253.6 eV. The spectrometer has a resolution of
approximately 1.7 eV, an anode voltage of 15 kV, and a power setting of 300 W.
Survey scans were performed at a 700 takeoff angle with respect to the sample
surface using a pass energy of 89.45 eV. Multiplex spectra, consisting of 10
scans over each peak, were run over a 30 40 eV range with a pass energy of
17.90 eV.


Langmuir-Blodgett films were prepared by spreading 100 mL of a 1.0
mg/mL solution of the diynoic acid onto a 4.0 x 10-4 M Cd2+ subphase that was
cooled to 120 C while at pH 6.2. Approximately 15 mL of a 0.05 M KOH
solution was added to the subphase in order to adjust to the desired pH.
Subphase pH was again measured at the end of LB experiments to ensure that
pH had not changed. Diynoic acid solutions were prepared in the dark in order
to avoid polymerization. Chloroform was used as the spreading solvent and
films were allowed to spread for 10 min. before compressing. Films were
compressed to a target pressure of 25 mN/m at a rate of 5 mN/m/min with
maximum barrier speeds of 15 mm/min.
The transfer of LB films to solid supports was accomplished by using
dipping speeds of 10 mm/min on the downstroke and 20 mm/min on the
upstroke. Samples typically consisted of one bilayer for XPS work and ten
bilayers for TEM experiments. Transfer ratios of the film under these optimum
conditions were unity. Before polymerizing the films, the samples were pre-
annealed by placing them in a drying oven at 620 C for approximately two
hours. This step was done to better preserve the layered structure of the films
during polymerization.38 Photopolymerization of the transferred LB films was

accomplished by placing the sample 20 cm away from a 300 W medium
pressure Hg lamp and irradiating the sample for 2.5 hr.
Cadmium iodide particles were then made by exposing the polymerized
Cd-diynoate film to hydrogen iodide gas. To accomplish this, samples were
held in a 30 mL Schlenk tube fitted at the top with a rubber septum and were
evacuated thoroughly before introducing 10 mL of HI gas into the reaction tube
using an air-tight syringe. Samples were allowed to react for 5 min. before
opening the reaction tube and purging out the HI gas with house nitrogen to
stop the reaction. Much care was taken to aviod sample contamination during
this procedure.


Characterization of Divnoic Acid Temolate

The material used to form the organic polymer Langmuir-Blodgett (LB)
template, 10,12-tricosadiynoic acid, was first characterized in the monomer form
at the air-water interface on the LB trough. Monolayers of the diynoic acid
collapse at 8 mN/m on a pure water subphase and at 35 mN/m when the
subphase is 4.4 x 10-4 M CdCl2, as shown in Figure 2-3. The much higher
collapse point on the cadmium subphase reflects the strength of the interactions
between cadmium ions and carboxylate head groups.
All further studies were carried out on a cadmium subphase since
cadmium ions need to be incorporated into the films in order to form cadmium
iodide particles. The importance of pH control is illustrated in the


z 40.0-

Ca A

| 20.0-


0.0 20.0 40.0 60.0

Mean Molecular Area (A,)

Figure 2-3 Isotherms of 10,12-tricosadiynoic acid on A) 0.44mM Cd(ll) subphase and
B) pure water subphase. (Both isotherms at 16 C and pH 5.4)

isotherms given in Figure 2-4. At pH 5.0 the carboxyl groups of the diynoic acid
appear to be protonated and cannot interact with metal ions, resulting in the low
collapse point. At pH 5.4 the monolayer is much more rigid due to
carboxylate/cadmium ion interaction, collapsing at 32 mN/m and having a mean
molecular area (Mma) of 23.04 A2.
The isotherm at pH 5.4 in Figure 2-4 shows a pronounced transition
beginning at a Mma of 33 A2. It is suggested that this transition corresponds to
the diynoic acid molecule standing to a more upright position from one where
part of the molecule (the alkyl chain nearest the carboxyl group) is lying at the
air-water interface. Figure 2-5 shows a cartoon depiction of the proposed
A creep test, where barrier movement was monitored while a constant
surface pressure was maintained on the film for two hours, was performed on
10,12-tricosadiynoic acid. At the same conditions used for film transfer, the film
was very stable, as shown in Figure 2-6, with a creep of only 0.08 mm/min.

Brewster Anale Microscoov

Brewster angle microscopy (BAM) was used to obtain images of the
diynoic acid monolayer as it was compressed through its isotherm. The
micrograph shown in Figure 2-7 was taken at a surface pressure of 20 mN/m.
The BAM micrograph shows the existence of non-uniform, intricate,
feather like structures which appear to consist of many domains. These
domains form immediately upon spreading the diynoic acid solution on the
subphase and remain even after the film has collapsed. The monolayer
existing as many smaller domains rather than one homogeneous domain is an
important factor in controlling the formation of Cdl2 particles. This topic will be




S20.0- B

o. 0.0\

0.0-- -i'-I--i- '-r -I

0.0 20.0 40.0 60.0

Mean Molecular Area (A~

Figure 2-4 Isotherms of diynoic acid at A) pH 5.4 and B) pH 5.0
(both at room temperature)


Figure 2-5. Cartoon of Proposed Transition


E 200



0 50 100

Time (min.)

Figure 2-6 Creep test of 10,12-tricosadiynoic acid on a Cd(ll) subphase







C 20

O 10

0 310

4---- Image

20 30 40 50

Mean Molecular Area (A2)

Figure 2-7 Isotherm (top) and Brewster angle micrograph
(bottom) of cadmium diynoate monolayer taken at a
surface pressure of 20 mN/m.

discussed in greater detail when the transmission electron microscopy data are
Pre-annealing the samples before polymerization has been shown to
increase Bragg peaks in X-ray diffraction studies by as many as seven orders.40
After annealing, the sample is exposed to UV radiation for 2.5 hr until the
polymer is converted to its red form. The sample is then exposed to 5mL of HI
gas (at atmospheric pressure) for 5 min., forming cadmium iodide in the
inorganic layers and protonating the carboxylate groups of the template.

UV-Vis Spectroscovp

The polymerization of cadmium diynoate multilayers on a glass
microscope slide can be followed by UV-vis. To the naked eye, a monolayer of
the monomer is colorless, but turns blue after five minutes of exposure to UV
light. Further irradiation turns the sample red until no further changes occur
after 2.5 hr of exposure. Figure 2-8 follows the photopolymerization of eight
bilayers of the monomer through both the blue and red forms of the polymer.
Arrows indicate the direction of band growth as polymerization occurs.

Attenuated Total Reflectance IR

Attenuated total reflectance Fourier transform infrared spectroscopy
(ATR-FTIR) was used to analyze effects that the pre-anneal and polymerization
procedures had upon the organic template as well as to determine the
completeness of reaction when forming cadmium iodide. The pre-anneal step
had no effect on the IR spectrum of the template. The main effect polymerization

Figure 2-8 UV-vis spectra of 8 Cd-diynoate bilayers after 0, 5, 10, 30,
60, 90, and 120 minutes of exposure to UV light. Arrows
indicate direction of band growth.

had on the template IR spectrum was to broaden the asymmetric methylene
stretching band at 2921 cm-1 (See Figure 2-9). This increase in the full width at
half maximum (FWHM) from 26 cm-1 in monomer form to 34 cm-1 in the
polymer indicates that the film becomes more disordered during polymerization.
More specifically, the methylene chains of the molecule become less close-
packed as the polymerization progresses. Polymerization also resulted in the
appearance of a small peak at 1711 cm-1 that corresponds to the protonation of
some carboxylate groups.
Figure 2-10 is the ATR-FTIR spectra of a polymerized Cd-diynoate
multilayer sample before and after exposure to HI gas. The carboxylate band at
1537 cm-1 before HI treatment (top spectrum) disappears completely after 5
min. of HI exposure. This information coupled with the resulting carboxylic acid
band at 1693 cm-1 (bottom spectrum) indicates that all of the template
carboxylate groups have been protonated in the reaction to form cadmium
iodide particles.
It has been reported that the mode of hydrogen bonding (either facial or
lateral, See Figure 2-11) between carboxylic acid groups causes IR shifts in the
C=0 stretching band.47 Facial hydrogen bonding between COOH groups gives
a band at 1710 cm-1 while lateral hydrogen bonding gives its C=0 stretch at
1724 cm-1. From the position of the COOH band at 1693 cm-1 in the polydiynoic
template, it is believed that facial H-bonding is occurring. This is nearly the
same peak position seen in the spectrum of neat diynoic acid at 1691 cm-1. By
comparison, the carbonyl band of the arachidic acid template 2 had a slightly
higher energy at 1701 cm-1, but appears to also form facial H-bonds as in the
diynoic template. In order for facial H-bonding in the diynoic template to occur,
there can be nothing between the opposing carboxylic acid head groups, as
this would interfere with the hydrogen bonding. This indicates that the diynoic


Figure 2-9 The effect of polymerization on the width of the asymmetric
methylene band.


8 2921.0 cm-1

After HI
1693.4 cm-1

3000 2500 2000


Figure 2-10 ATR-FTIR spectra of a polymerized Cd-diynoate multilayer sample
before and after exposure to HI gas.

0 0

I -
S0 0H O,"'H, 0 0

Facial H-bonding Lateral H-bonding

Figure 2-11 Hydrogen bonding modes between carboxylic acid groups

template produced discrete Cdl2 particles (rather than a single layer) and left
much of the template's interlayer space empty for facial H-bonding. Had a
single layer of Cdl2 been formed, it would have blocked facial H-bonding and
only allowed lateral H-bonding, which would have shifted carbonyl bands to
higher frequencies.

X-ray Photoelectron Spectroscooy

X-ray photoelectron spectroscopy (XPS), also known as electron
spectroscopy for chemical analysis (ESCA), was used to confirm the presence
of cadmium iodide formed in samples after exposure to HI gas. A typical
multiplex spectrum of a HI treated template is given in Figure 2-12.
The predicted XPS peak intensity ratio for Cdl2 formed in the polymer
template was 1.0 cadmium to 1.7 iodide. This ratio was calculated by taking
into consideration the effect of photoelectron attenuation by an overlayer and
the photoelectron's kinetic energy.48 An experimental ratio of 1 cadmium to 3
iodide was found in the XPS experiment, which lies outside experimental error
of the predicted ratio. The high iodine content cannot be explained by the
presence of some excess 12 formed by the reaction of HI with residual water,
because one would expect any 12 formed to sublime readily under the high
vacuum conditions imposed by XPS analysis. A more likely explanation for the
higher than expected iodine content is that some of the HI gas is adding onto
unsaturated carbon-carbon bonds in the polymer backbone in addition to the
expected reaction to form Cdl2 crystals. This line of reasoning will be supported
by XPS and TEM data to be presented later. It would not be unreasonable to
expect to see some changes in the ATR-FTIR spectra of the polymer template

10 -- -- ---

7 1(3d)
6 s BE = 629.5, 618.0 eV
5 s % Conc =74 Cd(3d)
4 BE =410.8, 404.1 eV
3 % Conc= 26

600 500 400

Binding Energy (eV)

Figure 2-12 XPS multiplex spectrum of cadmium iodide formed at the
polymer template.

after exposure to HI. For example, the intensity of C-C double and triple bond
stretches should decrease after HI adds onto these bonds. However, the
intensities of these bands are very weak and they do not lend themselves to
When the photopolymerization step was skipped and the template was
exposed to HI gas while still in the monomer form, TEM/TED showed no
diffraction from Cdl2 crystals (see Figure 2-14). However, when XPS analysis

was carried out on an identical sample, a 1 Cd to 4 I ratio was obtained. These
results can be explained as follows. When the monomer template is exposed to
HI, it reacts both at the unsaturated C-C bonds and at the inorganic layers.
However, under the high vacuum conditions used in XPS and TEM/TED, the
monomeric template molecules could be volatilized, since weak van der Waals
forces are all that hold the monomer template molecules together. Cadmium
iodide could have also been removed from the sample by the high vacuum so
that no diffraction would be seen in the TEM experiment. However, some
monomer template molecules remain on the sample that contain both
unreacted Cd2+ and iodine that has added onto triple bonds which were
detected by XPS in a 1 Cd to 4 I ratio.
When the XPS multiplex peaks for iodine were examined more closely, it
was found that the iodine peaks on the polymerized template were symmetrical,
while those on the unpolymerized template were asymmetrical, having a slight
shoulder at lower binding energy. The shoulder indicates that two types of
iodine were present, each with slightly different binding energies as a result of
being in different chemical environments. The peaks were analyzed using a
peak fitting program, as shown in Figure 2-13. The iodine peaks were fitted
using a Gaussian curve that had the same FWHM as in an arachidic acid
template sample where the iodine signal is known to be solely from Cdl2. The

633 632 631 630 62 628 627
Bihing Energy (eV)

Figure 2-13 Peak fitting of iodine 3d3 XPS peak from a monomer

shoulder iodide peak from the diyne template had the same binding energy as
the iodide in the arachidic acid template, indicating that the shoulder iodine
peak at 628.4 eV is from iodide in Cdl2 particles. A photoelectron ejected from
an iodide ion in the Cdl2 lattice would be expected to have a lower binding
energy than the photoelectron ejected from an iodine atom in a covalent
environment. This is due to more electron-electron repulsion involved with the
anionic iodide than with the covalently bound iodine.
There is a nearly 2 to 1 ratio of covalently bound iodine (major peak) to
iodide incorporated into the Cdl2 lattice (shoulder). A possible explanation for
this is that the gas appears to react preferentially with the triple bonds rather
than with Cd2+ when the unpolymerized template is exposed to HI. The
absence of diffraction from the monomer template after exposing it to HI is to be
expected if the HI is not producing Cdl2, but reacting first with triple bonds.

Transmission Electron Microscopy/Diffraction

Transmission electron microscopy and transmission electron diffraction
(TEM and TED) were used to gain structural and orientational information
concerning the polymer template and the cadmium iodide particles. When a
TEM sample of the polymer template (10 bilayers, no exposure to HI gas) was
analyzed it was found not to diffract. Figure 2-14 gives these data.
The domains present in the micrograph of Figure 2-14 are aligned such
that the template has a directional grain like appearance. The domains, which
form upon spreading the 10,12-tricosadiynoic acid on the water surface,
aggregate into superstructures that look like feathers at lower magnification.

Figure 2-14

a) Transmission electron micrograph of 10 bilayers of polymer
template before exposure to HI gas. (Magnification = 10k)

b) Transmission electron diffraction pattern of 10 bilayers of polymer
pictured in A) showing no difraction (selected area of diffraction =


Once the polymer template is exposed to HI gas, the formation of
cadmium iodide crystals can be confirmed through its TED pattern. Figure 2-15
shows Cdl2 diffraction patterns as well as the area of the micrograph to which it

corresponds. The diffraction pattern in Fig. 2-15 b), taken from the area marked
A on the micrograph, shows no diffraction. This region is believed to only
consist of the organic polymer since it does not diffract. Note that area A has a
different appearance from the polymer template in Figure 2-14 (no HI), with
smaller domains separated by very dark regions. The difference in appearance
can be explained by the fact that the template in Figure 2-15 could have been
chemically altered during exposure to HI, which adds onto unsaturated bonds
as discussed earlier.
The next two diffraction patterns, c) and d), in Figure 2-15 are both from
the Cdl2 crystal containing area labelled B in the micrograph. In diffraction
pattern c), several different orientations of Cdl2 are evident, as given by the

presence of diffraction rings rather than discrete spots. For this diffraction
pattern a selective aperture of diffraction (SAD) of 2.6 mm was used. However,
when the area of diffraction was reduced to 0.7 mm at the same location, only
one orientation of Cdl2 was evident. This result is displayed in diffraction
pattern d) of Figure 2-15.
Figure 2-16 shows a bright field micrograph of Cdl2 crystals, the
diffraction pattern resulting from the entire area pictured (SAD 5.5 mm), and a
dark field micrograph of the same area. In the dark field micrograph, the white
regions within the dark Cdl2 crystals are most responsible for the (100)
diffraction spots. These are the innermost hexagon of the diffraction pattern.
Within the entire diffraction area, which measures 5.5 mm across and contains
over a dozen cadmium iodide crystals, crystal orientations vary by less than two

Figure 2-15

a) Transmission electron micrograph of cadmium iodide particles
(area B) formed at polymer template (area A). (10 bilayers,
Magnification = 8k)

b) Transmission electron diffraction pattern taken from polymer area
A in micrograph above showing no difraction. (SAD =2.6pm )

1.o pm

Figure 2-15 (cont.)

c) Transmission electron diffraction pattern from Cdl2 particle in area
B of micrograph. Several crystal orientations are evident using a
SAD of 2.6 um.

d) Transmission electron diffraction pattern of same particle as in c),
but using a smaller SAD of 0.7 pm. Only one crystal orientation is

B SAD 2.6pm

Figure 2-16

a) Bright field micrograph of Cdl2 particles formed at a 10 bilayer
polymer template. (Magnification = 10k)

b) Transmission electron diffraction pattern taken from entire area of
micrograph above. (SAD =5.5im )

1. m, ",
. -" ." l

* I' I ., t
1,.0^ pm

CT -- 1
~1 ~ f

Figure 2-16

c) Dark field micrograph of same Cdl2 particles shown in a).
(10 bilayers, Magnification = 10k)

rI 1

degrees. This is given by the spread of diffraction spots for a given diffraction
Pre-annealing effects are evident by comparing Figures 2-15 (not pre-
annealed) and 2-16 (pre-annealed). The pre-anneal step consisted of placing
the sample (still in monomer form) in a 620C drying oven for 3 hr. This
procedure did not appear to affect samples as evidenced by a lack of change in
the XPS or ATR-FTIR spectra. However, TEM micrographs show that the
organic template's appearance was changed by the pre-anneal, taking on a
reptilian or alligator skin appearance. In addition to visual differences, the pre-
anneal step had the effect of growing a few large Cdl2 crystals (3-5 mm) rather
than many smaller ones (< 1mm). This can be interpreted in two ways. First, it
is likely that the pre-annealing process allows for the organic template
molecules (not yet polymerized) to rearrange into larger domains that allow for
larger crystals to form during exposure to HI gas. Or second, the pre-anneal
step could make the template contract or could cause defects which would
allow for the diffusion of Cd2+ between layers, resulting in the formation of large
bulk crystals.
Appearances can sometimes be deceiving when analyzing crystals in
multilayer samples by TEM. For example, the large Cdl2 structure shown in
Figure 2-17 appears to be a single crystal, but its diffraction patterns at several
different locations on the structure show that at least two different orientations of
Cdl2 are present. Therefore the structure cannot be a single crystal despite its

appearance and likely consists of several crystals that are vertically stacked in
different layers.

Figure 2-17

a) Transmission electron micrograph of what appears to be a single
large Cdl2 crystal at the polymer template. (10 bilayers,
Magnification = 10k)

b) Transmission electron diffraction pattern taken at area indicated by
arrow in micrograph above. Pattern shows more than one
orientation and therefore, did not come from a single crystal.
(SAD =2.6gm )



_______.ff^ 2^f^Jp


The polymeric diynoic acid template was able to effectively guide the
formation of large, singly oriented cadmium iodide crystals over areas 2.6 mm in
diameter. However, the polymer template did not appear to limit diffusion
between inorganic layers significantly better than did the monomer arachidic
acid template. This inability to limit diffusion is attributed to the weaker van der
Waals forces of the diynoic acid template molecules before polymerization
compared to the arachidic acid molecules. The two short C10 and C8 alkyl

chains of the diynoic acid molecule do not have as strong of intermolecular
forces as does the one long C19 chain of arachidic acid. Weaker van der

Waals forces allow for many smaller diynoic acid domains to exist compared to
the arachidic acid system. If polymerization leaves these smaller diyne
domains intact, it may result in the formation of crystals at slightly different in-
plane orientations as seen in this work. In addition, defect sites between
domains could provide avenues for metal ions to diffuse between layers,
forming bulk cadmium iodide particles rather than the desired cadmium iodide




This chapter, as the previous one, is concerned with forming inorganic
particles at a pre-formed organic Langmuir-Blodgett film. The LB film may act
as a template which directs the formation of these particles, controlling their
size, shape and the orientation. The central objective of this project was to form
an extended lattice of CdX2 (X = Cl, I) at the template that would have a
thickness approaching that of a single layer. This could be thought of as an
inorganic crystalline monolayer of CdX2. In order to achieve this, the interaction
between template and the forming crystal must be strong enough to suppress
the formation of 3-dimensional structures driven by the crystal's lattice energy.
A sufficiently strong interaction between the template and forming crystal would
prevent the formation of bulk cadmium iodide crystals in favor of an extended
ultrathin inorganic lattice.
Previous work in the Talham group1"3,28 concerning the growth of
oriented inorganic particles at LB templates used a long chain carboxylic acid
(arachidic acid, 20 carbons) to construct the organic template. This arachidate
template was successful in producing Cdl2 particles oriented exclusively with
their layer planes (the 001 faces) parallel to the LB layers. It was believed that

the complimentarity in lattices between the layered cadmium iodide structure
and the layered LB film was responsible for controlling particle orientation.
Though successful in orienting particles, the arachidate template, terminating
with its COO- head group, was not able to produce single layers of the inorganic
lattice. Rather, particles with an average size of 0.4 mm were formed.
The approach used in the current project was to change the head group
of the LB template to a xanthate, R-O-CS2-. The terminal atoms are sulfur atoms
instead of oxygen atoms in the carboxylate film. The potential advantage of the
xanthate head group is to enhance the interaction between it and the 001
surface of Cdl2, which is a surface of iodide ions. Within the cadmium iodide
structure, there are two types of bonding interactions present. One is the ionic-
covalent bonding between cadmium and iodide ions within the anion-metal-
anion layer. These bonding interactions are too strong to be altered by the
presence of the LB template. However, the other type of interaction is a much
weaker anion-anion interaction between layers in the Cdl2 structure. This weak
iodide-iodide interaction is van der Waals in nature. If interactions between the
template headgroups and the (001) iodide layer of the crystal can be induced
by the proper choice of headgroup, these interactions will compete with iodide-
iodide van der Waals interactions of the crystal. In this way, crystals
approaching single layer thickness may be favored over bulk crystals.
The sulfur-containing xanthate is a softer, more easily polarizable head
group than the oxygen-containing carboxylate. This should lead to a stronger
soft/soft interaction of the template head group with the iodide ions of Cdl2, and
induce the formation of a thinner oriented crystal.
Compounds having the -CS2 functionality such as xanthates and
dithiocarbamates are in general strong chelators and have been used to extract
metal ions from mining wastewater. This fact leads one to believe that these

functionalities should be effective in LB experiments, binding metal ions from
the subphase. In this work, it is necessary for the hydrophilic head group to
bind metal ions from the subphase and incorporate them into the film in order to
produce crystals in a later step. Complexes of xanthate and dithiocarbamate
compounds have also been used as antioxidants, antiwear lubricants, and
vulcanization accelerators.49
A limited amount of literature exists pertaining to the use of xanthates or
dithiocarbamates in Langmuir-Blodgett experiments. In one study, LB and self-
assembled (SA) films of the double chain compound
dioctadecyldithiocarbamate were prepared on gold and silver surfaces and
compared using IR techniques.50 It was found that both SA and LB methods
produced highly ordered films, but films made by the LB technique had better
surface coverage. Also studied was the single chain compound,
octadecyldithiocarbamate. This molecule did not form a stable monolayer on
the trough and could not be used to produce LB films.
The dioctadecyldithiocarbamate molecule has also been used to make
Cu2+ containing multilayer samples for an EPR study.51 It was found that the
orientation of the molecules was largely influenced by the dipping direction of
monolayers from the water. In another paper52 the strong affinity of
dioctadecyldithiocarbamate for metal ions was utilized to make Cu2+ and Ni2+
sensors based on the LB technique. The sensor used energy transfer from an
underlying fluorescent dye layer to a dioctadecyldithiocarbamate layer with its
-CS2 head groups in contact with a metal containing aqueous phase. Metal
complexes formed instantaneously at the LB film/liquid interface and were
detected by the presence of their characteristic absorption bands in
fluorescence measurements.

It will be shown in the following sections that the goal of forming an
inorganic monolayer of Cdl2 inside an LB film template was not achieved using
octadecyl xanthate. It was found that crystals from different layers (lying directly
above one another) were in registry with one another in some areas less than 1
mm in diameter. However, it was also found that most of the xanthate template
decomposed during the reaction with HI during Cdl2 formation. The headgroup,
which was hoped to control crystal growth, was given off as CS2 gas. For this
reason the xanthate template failed to control crystal orientation within the layer
plane of samples.


Octadecanol (98%) was obtained from Aldrich Chemical Co. (Milwaukee,
WI). Analytical reagent grade hexane, ethanol, tetrahydofuran and carbon
disulfide were obtained from the Fisher Chemical Company (Pittsburgh, PA).
All other materials as well as substrate preparation techniques used were the
same as described in Chapter 1.

All instrumentation used to study the xanthate system was the same as
that described in Chapter 1 except for a method used to take transmission IR
spectra of xanthate multilayer samples after reacting with HCI gas. A silicon
wafer was used as the substrate and was held with the layer planes
perpendicular to the incident IR beam. Spectra consisted of 5000 scans at 4

cm-1 resolution. A deuterium triglycine sulfate (DTGS) detector was used in
order to investigate the S-C band region at 1200 800 cm-1.

Sodium octadecylxanthate [CH3(CH2)170-CS2- Na+] was synthesized1
by adding a 10 20% excess of CS2 dropwise via addition funnel to
octadecanol and NaOH dissolved in THF while stirring with a magnetic stirrer.
The reaction mixture was refluxed for three hours before the yellowish crude
product was Soxhlett extracted in hexane for approximately one hour using a
cellulose thimble filter. Three recrystallizations gave the product in 80% yield.
Identity of the product was confirmed by CHN analysis (experimental : 61.29%
C, 10.85% H ; theoretical : 61.96% C, 10.05% H), nmr, and FAB mass
spectrometry, where the most intense peak at m/z = 346 corresponded to [M -
Langmuir-Blodgett films of cadmium octadecylxanthate (Cd-xanthate)
were prepared as follows. Chlorobenzene was used as the spreading solvent
as this solvent produced a clear solution with the xanthate after heating with a
heat gun. Other common spreading solvents did not form a clear solution. Due
to chlorobenzene's high rate of thermal expansion, it required great care to
ensure the desired amount of material was being spread on the LB trough. The
100 mL of 1 mg/mL octadecylxanthate spreading solution was withdrawn by
syringe just as the meniscus of the cooling solution reached the calibration
mark of the 10 mL volumetric flask.
The subphase used was 4.0 x 10-4 M CdCl2, chilled to 170 C, and
adjusted to pH 6.7 with 0.05 M KOH solution. A delay period of ten minutes
passed before compressing the films in order to allow the spreading solvent to
evaporate. The xanthate films were compressed at the rate of 5 mN/m/min

using maximum barrier speeds of 10 mm/min until the films reached a target
pressure of 30 mN/m. The monolayer was then allowed to sit at the target
pressure for 30 min. to allow for the film to stabilize and for the system to reach
equilibrium. Dip speeds of 10 mm/min were used on both the downstroke and
upstroke to transfer the film to a suitable substrate with transfer ratios of unity.
When preparing multilayer samples, care was taken not to completely withdraw
the sample from the subphase as not to disturb the monolayer. This technique
was found to improve film transfer. One bilayer samples were used for XPS
experiments and 8-10 bilayer samples were used for TEM work. All substrates
used were made hydrophobic with OTS.


Fourier Transform Infrared Spectroscopy

Using the optimum dipping parameters given in the experimental section,
the deposition of cadmium octadecylxanthate (Cd-xanthate) bilayers was
followed by transmission infrared spectroscopy, as shown in Figure 3-1. The
asymmetric methylene band at 2922 cm-1 was used to follow the build-up of
material. Multiplying the band's intensity by its full width at half maximum
(fwhm) gives an estimate of the peak area; this is then plotted as a function of
the number of bilayers depositted. The plot is linear through three bilayers,
indicating uniform depositions. After four bilayers, however, the peak area
deviates negatively from linearity, indicating that less cadmium octadecyl
xanthate is being transferred. The lower transfers are believed to be the result



S *

S4.00 -


0 4 8
# Bilayers Cd-Octadecylxanthate

Figure 3-1 Deposition of Cd-octadecylxanthate bilayers followed by

of removing the sample from the LB trough for IR measurements, which tends to
disrupt the system. If a sample was dipped continuously without removing, as
many bilayers as desired could be transferred with transfer ratios of unity.
The reaction of Cd-xanthate with HCI gas to form CdCl2 particles was
followed by IR in order to see the effect HCI gas had on the organic LB film
template. Figure 3-2 shows the spectra of 20 bilayers of Cd-xanthate both
before and after treating with HCI gas. Overall, the spectrum shows little
change after being exposed to HCI for one hour. Of great importance is the
preservation of order in the organic template during the formation of inorganic
particles. This is desired so that the organic template can play a role in
controlling the size and orientation of particles being formed. The position,
intensity, and fwhm of the asymmetric methylene stretch at 2922 cm-1 indicate
that the crystallinity of the alkyl chains is not disturbed by the reaction.
However, the intensities of the asymmetric S-C-S stretching band from
the xanthate head group at 1044 cm-1 and the asymmetric C-O-C stretching
vibration53 at 1202 cm-1 decrease after reacting with HCI. The asymmetric S-C-
S band did not shift to higher wavenumbers (as would be expected if
protonated), indicating that the reaction did not go to completion. No new peaks
appeared in the spectrum after the reaction that corresponded to protonating
the xanthate head group. These observations indicated that the template may
be unstable. Some of the xanthate head groups may have been lost through
decomposition of the LB film, giving off carbon disulfide gas in the process.
Further evidence for the decomposition of xanthate will be presented in the XPS

Before HCI

Sasvm CHo


After HCI

~"C *LLkil




Figure 3-2 ATR-IR spectra of 20 bilayers of Cd-xanthate before and
after exposure to HCI gas.




X-ray Photoelectron Spectroscopy

Figure 3-3 contains the XPS multiplex spectrum and relative abundance
table for one bilayer of Cd-xanthate that has not been reacted with HI gas.
Within this bilayer, an XPS peak intensity ratio of 4.5 sulfur atoms to 1 cadmium
ion was predicted. This calculated prediction takes into consideration the effect
of photoelectron attenuation by an overlayer and the photoelectron's kinetic
energy.2,48 Experimentally, however, only 2.4 sulfur atoms per cadmium ion
were found. The ultra high vacuum conditions (10-9 torr) inside the XPS sample
chamber caused approximately 47% of the xanthate head groups from the LB
template to be given off as CS2. The xanthate head groups susceptibility to
decomposition before reacting to form inorganic particles raised doubt as to its
effectiveness as a template.
An XPS multiplex spectrum and relative abundance table of the Cd-
xanthate bilayer after the reaction with HI to form Cdl2 particles is given in
Figure 3-4. The spectrum shows the complete formation of cadmium iodide as
expected, with a Cd:l ratio of 1:2. What was not expected was the almost
complete absence of sulfur in the spectrum. The sulfur content was less than
one tenth of the predicted value. On the average, the sulfur content dropped by
80% in XPS samples after gasing.
The proposed reason for the loss of sulfur in Cd-xanthate samples after
reacting with HI is straightforward and is illustrated in Figure 3-5. When HI is
introduced to the Cd-xanthate bilayer, the iodide reacts with cadmium to give
Cdl2 while the proton bonds to the sulfur of the xanthate headgroup to form a
xanthic acid. This xanthic acid is unstable, and readily decomposes54,ss to give
off CS2 gas, leaving octadecanol behind. These XPS results raised serious


7 *


1 .0 35.8 366.8 25.0 2 0.6 150.0

Binding Energy (eV)

Relative Abundance

Cd S

Expected 1.0 4.5
Found 1.0 2.4

Figure 3-3 XPS multiplex spectrum and relative abundance table for 1
bilayer of Cd-xanthate before exposure to HI gas.

BE = 410.8, 404.3 eV
% Conc = 29.4

BE = 161.8 eV
% Conc =70.6

C(ts) reference at 282.5 eV



LU 7




6ea.6 550s. 5M0. 0 .0e 4o.e 358.0 36.0 25se. 260.6

Binding Energy (eV)

Relative Abundance

Cd I S


Figure 3-4 XPS multiplex spectrum and relative abundance table for 1
bilayer of Cd-xanthate after exposure to HI gas.


C dS SC d l d+
Cd2+ H I(g)
Cd 2c --, /cZ7 + CS2(g)
C. C.* n -A ii


xanthic acid


Figure 3-5 The loss of sulfur as seen by XPS after reacting with hydrogen
iodide is attributed to the formation of an unstable xanthic acid,
which decomposes to octadecanol and releases carbon disulfide gas.

doubt as to the effectiveness of the xanthate template. If a considerable amount
of xanthate headgroups were leaving during the reaction to form Cdl2 particles,
the template would not be effective in controlling particle growth and orientation.
Recall that it was the headgroup which was intended to influence the inorganic
particle. Once the ramification of these results were realized, further study of the
xanthate system was not pursued. It was worthwhile, however, to investigate
the cadmium iodide particles that were formed with transmission electron

Transmission Electron Microscopy

Figure 3-6 shows a transmission electron micrograph of Cdl2 particles
(dark features) formed at the octadecylxanthate template and its corresponding
diffraction pattern. These data were representative of the Cdl2 particles formed
in the 10 bilayer samples of Cd-xanthate. The micrograph is at a magnification
of 20k and the diffraction pattern was taken using a selective aperature
diameter of 2.6 mm, which essentially encompasses the entire area seen in the
micrograph. The diffraction pattern shows rings rather than discrete diffraction
spots as seen in previous work with arachidic acid LB templates. The rings
were indexed and found to arise from hkO diffraction planes. This means that
these particles lie with their (001) faces parallel to the layer plane of the sample,
as seen with the arachidic acid template. However, the presence of rings rather
than spots indicates that the particles do not have the same orientation within
the layer plane. The particles are able to exist at many orientations within the
layer plane rather than being locked into a single orientation by the template.
This lack of control by the xanthate template in limiting the orientation of
particles is not surprising when one considers that the xanthic acid head groups

Figure 3-6


Transmission electron micrograph of Cdl2 particles formed at an 8
bilayer xanthate template. (Magnification = 20 k)

b) Transmission electron diffraction pattern taken from micrograph
above. Polycrystalline rings indicate particles do not have same
in-plane orientation. (SAD = 2.6 plm)



1.0.a m | H 20 k

are unstable and are given off as CS2 gas as evidenced by XPS data. With
fewer template headgroups present to interact with forming particles, the Cdl2
would be able to adopt many more in-plane orientations which would give rise
to the observed diffraction rings.
Although octadecylxanthate in general was not successful as a template,
due to the instability of xanthic acid and its inability to control inorganic particle
orientation, there was one piece of evidence that showed that the xanthate
template did exert some control over Cdl2 particles formed. The TEM
micrograph in Figure 3-7, taken at a magnification of 100 k, has checkered or
fringed areas (indicated by arrows) which are known as Moire patterns. A Moire
pattern arises in electron microscope images when the electron beam is
diffracted by two overlapping crystals that have equal lattice patterns.56 In this
case, the Moire pattern is due to two overlapping Cdl2 crystals from different
layers within the sample. Despite the fact that we are looking at two separate
crystals as given by the Moire pattern, the diffraction pattern from this area is
that of a single crystal, having cadmium iodide's characteristic hexagonal
pattern and showing only one orientation. From the diffraction pattern, it is
evident that the overlapping crystals from different layers have the same in-
plane orientation, possibly due to the effect of the xanthate template over the 0.7
mm diameter area.


In working with Langmuir-Blodgett films of cadmium octadecylxanthate, it
was found that the well ordered system could be used to form Cdl2 crystals with
their 001 faces oriented exclusively parallel to the LB layer planes. This aspect

Figure 3-7

a) Transmission electron micrograph of Cdl2 particles formed at an 8
bilayer xanthate template. Arrows indicate presence of Moire
patterns. (Magnification = 100 k)

b) Transmission electron diffraction pattern taken from micrograph
above. (SAD = 0.7 pm)


100 k

SAD 0.7 Rm|

0.1 m

was similar to results obtained using arachidic acid films. Within the layer
plane, however, crystals in the xanthate template do not have similar
orientations. This was in contrast to the control arachidic acid films have over
in-plane crystal orientation of cadmium halides. The reason for the xanthate
template's lack of control over in-plane orientation is that the xanthate gave off
its CS2 head group during the reaction with HI gas. Once this occurred, the
organic film lost its ability to control the in-plane orientation of the particle as
seen in electron diffracton data.
It was hoped that the soft/soft interaction between sulfur atoms of the
template and iodide ions of the Cdl2 would overcome the lattice energy that
leads to the formation of bulk Cdl2 crystals. However, the interactions between
template and inorganic lattice were not achieved due to the decomposition of
the template during the reaction to form Cdl2. Although overall the
octadecylxanthate template was unsuccessful in controlling the orientation of
inorganic particles, there was evidence that overlapping crystals from different
layers in a sample were in registry with one another for small areas of less than
1 mm diameter.




Kidney stone disease continues to be a significant health problem that,
although treatable through surgery, as yet has no cure. In recent years, one
percent of all hospital admissions in the United States were due to kidney stone
disease.57 Calcium oxalate monohydrate, Ca(C204) H20 is the major
component in 60 80% of kidney stones58,59 and from this point on will be
referred to by its abbreviation, COM.
An objective of the work to be presented here is to show how LB films of
lipids can be used as an effective model system for the formation of kidney
stones at biological membranes. Information gained on the nucleation and
growth mechanisms of COM from an in vitro model may give insight on how to
prevent kidney stone formation. Figure 4-1 shows cartoon depictions of both a
biological membrane and COM crystals forming under a lipid LB monolayer.
This figure uses "ball and stick" notation, where the hydrophilic head groups are
represented by balls and the lines are hydrophobic aliphatic chains. Some
important differences exist between the biological membrane and the LB model.
The biological membrane is much more complex, being composed of a mixture
of lipids with different head groups, as well as having various proteins

Figure 4-1. Membrane Modeling

incorporated into the bilayer structure. The LB model consists of a monolayer of
a single lipid, floating on a supersaturated COM solution, where the level of
supersaturation is defined as the ratio of the activity of the experimental COM
solution to the activity of a saturated COM solution.60
The Langmuir-Blodgett (LB) technique is used to form well-ordered
monolayer and multilayer assemblies of amphiphilic molecules and is
described elsewhere in greater detail.6'-65 An advantage of the LB technique is
the ability to have control over the organization of a film at the molecular level.
Langmuir-Blodgett films have been used by Landau et al. in one of the early
studies of crystallization at monolayers.16"17 Here it was determined that
structural information could be transferred from the monolayer to the crystals
growing underneath. This group has also investigated the growth of sodium
chloride crystals beneath monolayers of long chain carboxylic acids such as
stearic and arachidic acid.17 The monolayers were found to nucleate and
stabilize the (111) face of NaCI, a face that does not occur naturally.
Mann et al. have studied the controlled nucleation and growth of CaCO3
beneath LB monolayers of stearic acid,19.20 octadecylamine, 21,22 eicosyl sulfate
and eicosyl phosphonate.23 In these experiments, the monolayer was
compressed and held upon a supersaturated calcium carbonate subphase
solution. In control experiments, where crystals precipitated in the absence of a
monolayer, the calcite form of CaCO3 was formed around the sides and bottom
of the LB trough. However, in the presence of a stearic acid monolayer, a
different form of CaC03, vaterite, formed exclusively at the monolayer/solution
interface, oriented with the (001) faces parallel to the monolayer plane.19
Compression isotherm data indicated to the authors that Ca2+ was
incorporating into the head groups of the film to form a layer of calcium ions.

The calcium ions in this layer could then mimic the calcium positions found in a
specific crystal face of vaterite and proceed to grow from this nucleated face."
The Mann group has also studied the nucleation of BaS04 underneath
monolayers of long chain sulfates,24 carboxylates25 and phosphonates.26 The
authors concluded that monolayer-induced oriented nucleation depended on
the organic film's ability to mimic both the lattice geometry of cations and the
stereochemistry of oxyanions in the nucleating crystal face. At least two
recognition processes are in effect at the monolayer/inorganic interface. There
is a geometric matching between close-packed head groups and Ba-Ba
distances in the crystal lattice as well as a stereochemical complementarity
between the sulfate head groups of the monolayer and the sulfate anions of the
nucleating crystal face.
We are interested in studying the role that the lipid plays in crystal
formation. More specifically, we want to determine the effect that varying the
chemical nature of the lipid has on crystal formation in terms of the amount, size,
shape, and orientation of COM crystals.
Figure 4-2 shows the structures of lipids used in this study. The
phospholipids dipalmitoylphosphatidylglycerol, dipalmitoylphosphatidylserine,
and dipalmitoyl-phosphatidylcholine vary only in their head group structure.
The lipid head groups are believed to be an important factor since they form the
surface where nucleation occurs. The COM formation experiments were
performed at pH 7.0, at which the lipids in Figure 4-2 are known to exist in the
fully ionized state.66

0 0

0 =





0 0

0=P- 0


Figure 4-2. Structures of the lipids used to study COM formation:
dipalmitoylphosphatidylglycerol, dipalmitoylphosphatidylserine,
and dipalmitoylphosphatidylcholine.



Dipalmitoyl-L-a-phosphatidyl-DL-glycerol (99%), dipalmitoyl-DL-a-
phosphatidyl-L-serine (98%), and dipalmitoyl-L-a-phosphatidylcholine (99%)
were purchased from Sigma (St. Louis, MO) and used as received without
further purification. Calcium chloride (97%), sodium oxalate (99.99+%),
tris(hydroxymethyl)aminomethane hydrochloride (99+%), from Aldrich
(Milwaukee, WI) and sodium chloride (99+%) from Fisher (Pittsburgh, PA) were
used as received. The water used in LB experiments was purified via a
Sybron/Barnstead Nanopure system (Boston, MA) and had a resistivity of 18


Langmuir-Blodgett experiments were carried out on a custom made
teflon double barrier trough measuring 12 x 60 cm and controlled with a KSV
Instruments Model 5000 LB system (Stratford, CT). Surface pressure was
measured with a platinum Wilhelmy plate suspended from a KSV microbalance.
IR spectra were recorded with a Mattson Instruments Research Series-1
FTIR spectrometer (Madison, WI) using a narrow-band mercury cadmium
telluride detector and a Harrick TMP stage (Ossining, NY) for ATR experiments.
Spectra consisted of 1000 scans at 2 cm-1 resolution and were ratioed to bare
ATR crystal background spectra.

TEM/TED analyses were conducted on a JEOL JEM 200CX electron
microscope (Peabody, MA), operated at an accelerating voltage of 100 kV.
Scanning electron microscopy experiments were done on a JEOL 35C
microscope (Peabody, MA) at an accelerating voltage of 15 kV.
A Nanofilm Technology Brewster angle microscope 1 (Gottingen,
Germany) equipped with a CCD camera and video cassette recorder was used
to observe the surface during compression isotherms and the formation of COM
crystals at the monolayer/subphase interface.

Substrate Preparation

Silicon ATR crystals from Wilmad Glass Co. (Buena, NJ) were cleaned
and made hydrophilic using a piranha etch (one hour immersion in freshly
prepared 3:1 H2SO4 : H202) and the RCA cleaning method (ten minute
immersion in 70C solution of 5:1:1 H20: H202: NH40H followed by a ten
minute immersion in 70C solution of 6:1:1 H20: H202: HCI and a thorough
water rinse.) Glass microscope coverslips used as substrates for SEM analyses
were cleaned using a piranha etch. Titanium and copper grids (300 mesh)
from the Ted Pella Co. (Redding, CA) used for TEM/TED analysis were mounted
onto microscope slides for support, covered with an approximately 600 pm thick
film of Formvar resin (Ted Pella Co.) and made hydrophilic by immersion in a
1% (w/v) aqueous poly-L-lysine solution from Fisher.


Typical parameters used in LB experiments such as isotherms and SEM
sample preparation include linear compression rates of 5 mN/m/min, barrier

speeds of 10 mm/min, and target pressures of 20 mN/m. All lipids were
dissolved in a 5:1 chloroform: methanol spreading solvent, at concentrations
near 1.0 mg/mL.
Supersaturated calcium oxalate monohydrate subphases were prepared
as follows. An aqueous solution that was 150 mM NaCI and 5mM Tris HCI
buffer was adjusted to pH 7.0 by adding the appropriate volume of 0.05 M KOH
(aq). This solution was split into two equal volumes before adding the
appropriate amounts of CaCI2 to one and Na2C202 to the other. This was
done in order to avoid precipitation of COM. After these solutions completely
dissolved and just prior to their use, they were combined and filtered (<1 .m
pore size) to give a final solution that was either 0.35 or 0.5 mM in COM. These
subphase solutions had relative supersaturation levels (RS) of 5 and 10,
respectively. The computer program EQUIL v.1.3, was used for the calculation
of relative supersaturation levels.
The transfer of lipid monolayers to a solid support for analyses by SEM
and TEM/TED was accomplished by carefully draining the subphase from the
trough, which effectively lowered the monolayer onto a substrate that had been
placed in the subphase before the monolayer was applied. Substrates were
positioned at an angle of approximately 15" with respect to the bottom of the
trough to allow for better drainage. The Langmuir-Blodgett technique was used
for monolayer transfer to ATR crystals, using dipping speeds of 10 mm/min.


Monolayer Studies

Pressure-area isotherms were obtained for dipalmitoylphosphatidyl-
glycerol, dipalmitoylphosphatidylserine, and dipalmitoylphosphatidylcholine
repeating the same experimental conditions as previous authors.67,68 The
isotherms obtained were consistent with those previously published. Hereafter,
the lipids will be referred to by their head groups: glycerol, serine, and choline.
The lipids were also studied by performing creep tests on the supersaturated
COM subphase used to grow crystals. All lipid monolayers were found to be
very stable on COM subphases, with no loss of film after several hours of
holding a constant surface pressure of 20 mN/m.
Pressure-area isotherms of glycerol in Figure 4-3 show a dependence
upon subphase calcium concentration. In the range of 6 x 10-6 M to 6 x 10-3 M
Ca2+ studied, the smallest mean molecular areas (Mma) were obtained for
middle concentrations, while larger Mma's were observed for both the higher
and lower calcium concentrations. This result is attributed to an optimum
concentration of calcium in the range of 6 x 10-5 to 6 x 10-4 M which binds the
monolayer together most effectively. At lower concentrations the monolayer is
expanded due to less calcium binding, while at higher [Ca2+] an excess of
calcium incorporated into the film caused the expansion.
In the complete absence of Ca2+, lipid monolayers showed higher
surface pressures at large (>60 A2) Mma's. The addition of a very small
amount of Ca2+ lowered the surface pressure, due to the Ca2+ ion organizing



C 40.00




35.00 40.00 45.00 50.00

Mean Molecular Area (A)

Figure 4-3. Isotherms of dipalmitoylphosphatidylglycerol over varying
subphase calcium concentrations. A) 6 mM; B) 0.006 mM;
C) 0.06 mM; D) 0.6 mM.

the lipid film into more compact domains. The interactions of phospholipids
with calcium ions is well documented in the literature.6-75

Blank Studies

Blank experiments were conducted where a supersaturated COM
subphase at a relative supersaturation level of 10 was allowed to sit in the LB
trough without the presence of a lipid monolayer. After periods of up to 21 hr,
prepositioned substrates did not have any COM crystals on them as seen by
SEM. The remaining subphase was then passed through filter paper with pore
size s 0.2 mm. The filter paper was inspected by SEM, and in each case, no
COM crystals were observed.

ATR-IR Experiments

ATR-IR experiments of lipid monolayers76,77 showed the lipid film to be
well-ordered and gave evidence of COM formation. Figure 4-4 is the ATR-IR
spectrum of a glycerol monolayer after it had been on a supersaturated COM
subphase for 21 hr. The broad band at 3300 cm-1 is from the hydroxy groups of
the glycerol head group. Another band assigned to the lipid is the asymmetric
methylene stretch, whose position (2918 cm-1) and shape (FWHM = 19 cm-1)
are consistent with a well-ordered, closely packed film, with its alkyl chains in an
all trans conformation.62
Evidence of COM formation is seen by the presence of the a(HOH) water
deformation band of COM at 1655 cm-1 and a non-fundamental COM vibration

0.015 asym. CH2
2918 cm-1
1381 cm-1

- 68(H-0-H)
M 1655 cm-1

0.005 -


3500 3000 2500 2000 1500

Wavenumbers (cm-1)

Figure 4-4. ATR-FTIR spectrum of a glycerol monolayer after floating on a
supersaturated COM subphase for 21 hours. Bands at 1381 and
1656 cm-1 are attributed to COM and give evidence for COM
crystal formation.

at 1381 cm-1. The band at 1381 cm-1 is a combination of O-C-O and C-C
vibration modes of the oxalate anion. 78.79 The COM bands did not lend
themselves for quantitation for two reasons. First, the bands were of very low
intensity. Second, the bands were not seen to increase regularly with time as
would be expected if the COM crystals were increasing in size or number. This
was due to crystals being unevenly distributed beneath the monolayer. In this
way, a sample of monolayer taken after four hours of crystal growth may have
more COM present than one taken after ten hours.
The IR spectra did show that COM was being formed, but it did not tell if
the COM was being formed at the lipid/solution interface or simply precipitating
out of bulk solution. To investigate this it was necessary to use Brewster angle

Brewster Angle Microscopy (BAM)

BAM experiments showed conclusively that COM was forming at the
lipid/solution interface rather than simply precipitating out of bulk solution. In
the BAM experiment, polarized light from a He-Ne laser is directed toward a
floating monolayer at the Brewster angle (530 for water), the angle at which the
intensity of reflected light is at a minimum. This reflected light is then recorded
on a CCD camera through a microscope and stored on videotape by a VCR.
The Brewster angle microscope sees only what is taking place at the interface,
by measuring differences in refractive indices at the monolayer/water interface.
Figure 4-5 is a BAM image of a glycerol monolayer early in a
compression isotherm, before surface pressure begins to increase. Individual
flower-shaped domains, similar to those observed by Losche et a/.,80 using
fluorescence microscopy, are apparent before they are compressed together to

Figure 4-5 Brewster angle micrograph of glycerol lipid domains at low surface
pressure. Image dimensions are approximately 2mm x 2 mm.

Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EVX4JBS0B_WES3UK INGEST_TIME 2014-04-14T19:05:14Z PACKAGE AA00014234_00001